Professional Documents
Culture Documents
CARL BOZZUTO
Electronic Book
Appendices
Emissions Calculator
PUBLISHED BY
www.alstom.com
Disclaimer
The information in this book has been obtained by Alstom Power, Inc. from sources that are believed to be reliable. Neither Alstom Power, Inc., nor any of the authors employed thereby, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process disclosed in this book, or represents that the use of any such information, apparatus, product, or process would not infringe privately owned rights. Any such information, apparatus, product, or process disclosed in
this book is for illustrative purposes only. Any reference to any plant or unit is for historical and illustrative purposes only. Trademarks used in this book including, Alspa™, CFS™, Controlled Circulation®, Combined Circulation®, Combustalloy™, Dynamic™ Classifier, FICIRC™, Flowpac®, KNX™,
Ljungström®, LNCFS™, Mer-Cure™, Raymond®, RSFC™, TFS 2000™, and TFS 2000™ R are trademarks of the Alstom Group. Teflon® is a registered trademark of the Dupont company. This book is published with the understanding that Alstom Power, Inc. and the authors employed thereby are
supplying information but are not attempting to provide or render engineering, operational, or any other professional services. If any such advice or services are required, the assistance of an appropriate professional should be sought.
Clean Combustion Technologies
Clean Combustion Technologies
A Reference Book on Steam Generation
and Emissions Control
FIFTH EDITION
EDITED BY
CARL BOZZUTO
PUBLISHED BY
ISBN 978-0-615-26919-1
Disclaimer
The information in this book has been obtained by Alstom Power, Inc. from sources that are
believed to be reliable. Neither Alstom Power, Inc., nor any of the authors employed thereby,
makes any warranty, express or implied, or assumes any legal liability or responsibility
for the accuracy, completeness, or usefulness of any information, apparatus, product, or
process disclosed in this book, or represents that the use of any such information,
apparatus, product, or process would not infringe privately owned rights. Any such
information, apparatus, product, or process disclosed in this book is for illustrative
purposes only. Any reference to any plant or unit is for historical and illustrative
purposes only. Trademarks used in this book including, Alspa™, CFS™, Controlled
Circulation®, Combined Circulation®, Dynamic™ Classifier, Combustalloy™, FiCIRC™,
Flowpac®, KNX™, Ljungström®, LNCFS™, Mer-Cure™, Raymond®, RSFC™, TFS 2000™,
and TFS 2000™ R are trademarks of the Alstom Group. INCONEL® and Teflon®
are registered trademarks of Specialty Metals Corporation and the DuPont company, respec-
tively. This book is published with the understanding that Alstom Power, Inc. and the authors
employed thereby are supplying information but are not attempting to either provide or render
engineering, operational, or any other professional services. If any such advice or services are
required, the assistance of an appropriate professional should be sought.
Transcontinental Printing
Printed in Canada
Contents
Foreword xi
Acknowledgements xiii
Introduction xv
Chapter One Fossil-Fueled Power Plant Design Anthony Munisteri and Carl Bozzuto 1-1 to 1-33
System Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
Laws of Thermodynamics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
First Law of Thermodynamics • Second Law of Thermodynamics
CYCLES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-4
Carnot Cycle • Rankine Cycle • Regenerative Rankine Cycle • Reheat Cycle • Supercritical
Pressure Cycle • Combined Cycle • Cogeneration
Designing a New Fossil-Fueled Power Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-17
Retrofitting an Existing Fossil-Fueled Power Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-20
Documentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21
Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-22
Fixed Charges • Fuel Costs • Economies of Scale • Present Worth and Levelized Costs •
Economic Analyses • Other Considerations
Summarizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-31
Chapter Two Combustion Technology Doug Hart and Bob Kunkel 2-1 to 2-25
HISTORICAL INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
COMBUSTION FUNDAMENTALS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
Combustion Equations • Concept of the Mole • Methods of Combustion Calculations
Combustion Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
The Combustion Reaction • Suspension Firing • Design of a Practical Fuel-Firing System •
Horizontally Fired Systems • Tangentially Fired Systems • Vertically Fired Systems • Formation and
Control of NOx in Steam Generating Equipment • Firing Systems that Minimize NOx Formation •
Stokers • Ignition Systems
Chapter Three Boilers Kevin Taugher, George Stamatelopoulos, Scott Darling, Glenn Selby, Donald 3-1 to 3-105
airley, Jesse Bolinger, Denis Bruno, Mark Malo, Thomas Mastronarde, Edward Ortman,
B
Harpreet Singh, Charles Tsirovasiles, and Anisa Brown
Pulverized Coal-Fired Boilers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
Converting Coal’s Energy to Electric Power • Impact of Fuel on Boiler Design • Furnace Sizing •
Dry-Bottom Versus Wet-Bottom Furnaces • Arrangement of Upper Furnace Heating Surface •
Design of Large, HP Furnace Wall Systems • Tangential Firing and Circulation System Design •
Supercritical-Pressure Steam Generators • Boilers for Process Use and Power Production •
Advanced Technology for Lignite-Fired Boilers
Fluidized Bed Steam Generators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-44
Historical Perspective • General Description of FBC • Characteristics of FBC • Types of FBC
Systems • Chemical Processes • Heat Transfer in Fluidized Bed Boilers • FBC Boiler Efficiency •
BFB Versus CFB • BFB Steam Generators • CFB Steam Generators • Hybrid CFB • Pressurized
FBC Systems
Heat Recovery Steam Generators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-76
Overview of Combined Cycle • Related Parameters • HRSG Designs • Fundamentals of How HRSG
Equipment Works • Design Considerations • Essential Components • Emissions Reduction • Future Trends
vi | Clean Combustion Technologies
Chapter Four Steam Turbines for Electric Power Generation Eric Pickering, 4-1 to 4-33
hee Loong Chong, and Beat Zimmerli
C
The Basics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1
Steam Turbine Thermodynamics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2
Development of the Modern Steam Cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Steam Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4
Steam Cycles for Coal/Fossil Fuel Applications • Combined Cycle Applications • Non-Reheat
Applications
Component impacts on efficiency and design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5
Effect of Exhaust Pressure • Boiler Feed Pump Drive (turbine or electric motor) • Reaction/Impulse
Blading • Three-Dimensional Blading • Partial/Full Arc Admission • Sliding- and Constant-
Pressure Operation
Mechanical Design of the Turbine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-9
Welded rotors • The HP Turbine • The IP Turbine • The LP Turbine
Major Turbine Auxiliary Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-15
Lube Oil System (Bearings) • Gland Steam System • Control Fluid/Hydraulic System • Turbine
Electronic Controls • Turning Gear and Jacking Oil Systems
The Future of Steam Turbines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-18
Generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-18
Principles of Control • Generator Options • Main Excitation System • Main Components •
Generator Externals • Stator • Rotor • Exciter • Cooling Gas Circuits • Water-Cooled Stator Bars •
Hydrogen Management and Seals • AVR and Electrical Generator Protection
Chapter Five Control of Power Plant Stack Emissions Alan Ferguson, 5-1 to 5-137
L eif Lindau, Steve Francis, Kjell Porle, Dennis Laslo, Stefan Ahman, Lars-Erik Johansson,
Mike Rini, Larry Czarnecki, John Wheeldon, and Carl Bozzuto
ENVIRONMENTAL FACTORS IN POWER PRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
POWER PLANT EMISSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Particulate Emissions • Sulfur Oxide Emissions • Nitrogen Oxide Emissions • Mercury Emissions •
Integrated Emission Controls
CONTROL OF PARTICULATE EMISSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4
Influence of Coal And Combustion Characteristics • Influence of Volumetric Gas Flow •
Measurement Techniques • Dust Collection by Mechanical Means • Electrostatic Precipitation
of Dust • Fabric Filtration • Operating Issues
CONTROL OF NOx EMISSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-32
SCR Chemistry • SNCR Chemistry • SCR Systems • SNCR Systems • Economic Considerations •
Post-NCR and Post-SCR Effects
CONTROL OF SO2 EMISSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-61
FGD Systems • Economic Considerations
Control OF MERCURY EMISSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-102
Background Chemistry • Mercury Chemistry in the Flue Gases Upstream of the Air Pre-Heater •
Mercury Collection Using Activated Carbon • Mercury Control and FGD • Alstom Mercury Control
Technologies
CONTROL OF CO2 EMISSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-110
Oxygen Firing • Post-Combustion Capture • Electro-Catalytic Oxidation • Benefit of Using a PC
Plant with Advanced Ultra-Supercritical Steam Conditions • Chemical Looping
Chapter Six Auxiliary Equipment Rory Eastman, Gus Shearer, Gary Mooney, Dennis Piontek, 6-1 to 6-83
hil Lafave, Doug Martin, Roger Martinsen, Peter Kelly, and Carl Bozzuto
P
Pulverizers and Pulverized Coal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1
Historical Perspective • Pulverizing Properties of Coal • Classification and Size Consist • Sampling
Pulverized Coal • Grinding • Methods of Pulverizing and Conveying Coal • Direct Firing
Arrangements • Principal Types of Pulverizers • Alstom Pulverizer Design
AIR PRE-HEATERS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-32
Regenerative Air Pre-heaters • Gas–Gas Heater for FGD Reheat Applications • Gas–Gas Heater for
SCR Reheat Applications • Tubular Air Pre-Heaters
Ash-Handling Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-43
Factors in System Selection • Types of Systems • Operational Controls • Material- and
Energy-Balance Considerations
Auxiliary Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-75
Pumps • Fans • Heat Exchangers
Contents | vii
Chapter Seven Controls Jack O’Rourke, Marc Thomas, Jean-Paul Drouin, Mark Wagner, and Rich Donais 7-1 to 7-45
A TOTAL CONTROL SOLUTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1
RELIABLE AND VERSATILE OPEN COMMUNICATION NETWORKS . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2
A RANGE OF CONTROLLERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3
OPERATOR CONTROL FACILITIES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4
Operator Mimic Viewer • Functional Diagram Viewer • y = f(t) Curves and y = f(x) Curves Viewer •
Alarms and Events Management • Data Properties Display • Bar-Graph Viewer • Variables Trend Viewer�
STEAM GENERATOR OPERATIONAL CONTROL SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
Furnace Explosions • Boiler Protective Systems • BMS/BPS for Tangentially Fired Utility Boilers�
CONTROL OF UTILITY STEAM GENERATORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23
Boiler Follow Systems • Coordinated Control Systems • Combustion Control System
CONTROL OF TANGENTIALLY FIRED STEAM GENERATORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-28
Unit Airflow Control • Secondary Air Damper Control—In-furnace NOx Reduction •
Feedwater/Separator Storage Tank Level Control (Once-Through Steam Generators) • Wet to Dry
Transfer • Once-Through Operation • Steam Temperature Control (General) • SH Steam
Temperature Control—(Drum Units) • Steam Temperature Control—(Once-Through Units)
boiler protection and BURNER MANAGEMENT SYSTEMS For Fluid Bed Units. . . . . . . . 7-34
Fluidized Bed Steam Generators—Degrees of Automation • Basic Safety Functions • Combustor
Temperature Monitoring�
CONTROL OF CFB BOILERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-36
Fluid Bed Steam Generator CCS • Airflow Control • Primary Air Control • Secondary Air Control •
Furnace Temperature Control • Furnace Differential Pressure Control • SO2 Emissions Control •
Steam Temperature Control—General • Superheat Temperature Control • Reheat Steam
Temperature Control�
FURNACE IMPLOSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40
The Implosion Phenomenon • The ID Fan • Control Implications�
Chapter Eight Materials for Fossil-Fueled Steam Generators Terry Totemeier and Ian Perrin 8-1 to 8-19
Functional Requirements for Pressure Part Materials. . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
Basic Metallurgy of Steel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Iron–Carbon Phase Diagram • Transformation Diagrams • Grain Structure
Fabrication Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-5
Bending, Forging, and Rolling • Welding • Heat Treatment
Steels for Fossil-Fueled Steam Generator Pressure Parts. . . . . . . . . . . . . . . . . . . . . . . 8-6
Specifications and Codes • Carbon Steels • Low-Alloy Ferritic Steels • Advanced Ferritic Steels •
Austenitic Stainless Steels
Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-11
Tensile Properties • Hardness • Toughness • Creep–Rupture and Allowable Stresses
Oxidation and Corrosion of Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-14
Steam Side Oxidation • Fireside Corrosion
MECHANISMS OF Pressure Part Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-16
Materials for Advanced Supercritical Plants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-16
Chapter Ten Operation of Steam Generators Gary Mattice and Frank Gabrielli 10-1 to 10-39
BASIC OPERATING PRINCIPLES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1
Steam Temperatures • Boiler Efficiency • Start-Up from Cold Condition: General • Start-Up of
Supercritical Pressure Boilers
PULVERIZER OPERATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-10
Lubrication • Mill Operation • Slagging and Fouling • Excess Air • Fuel Fineness • Secondary Air
Distribution
FEEDWATER/BOILER WATER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-14
Oxygen Control • pH Control • Importance of Water Analysis
COMMISSIONING FUNCTIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-16
Pre-Boiler Cycle • Chemical Cleaning of Boilers • Detergent Flush/Alkaline Boilout • Acid Cleaning •
Superheater/Reheater Cleaning and Flushing • Post-Acid Cleaning Activities • Steam-Line Blowing
OPERATION OF FLUIDIZED BED BOILERS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-30
Start-up • Normal Operation • Shutdown/Restart • Plant Main Fuel Trip (MFT������������� )������������
• Loss of Power •
Tube Rupture
PROBLEMS IN OPERATioN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-32
Furnace Explosions • Unintentional Fires External to the Furnace • Coal Handling in Power Plants
OPERATOR TRAINING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-36
The Systems Approach • The Needs Analysis • The Role of the Power Plant Simulator • Other
Support Materials • Completeness • Program Validation
Chapter Eleven Construction and Maintenance Paul Lafferty, Phil Lafave, Jim Sutton, 11-1 to 11-69
ruce Carney, Jim Geyer, Charlie Hart, Richard Stone, and Bill Herman
B
Field Construction of Steam Generating Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1
Planning and Organizing for Field Construction • Field Construction of a Boiler • CFB Boilers •
Maintenance and Repair of Steam Generating Equipment • Maintenance Foresight • Equipment
Manuals: An Asset to Effective Maintenance • Determining the Need for Spare Parts •
In-service Maintenance • Pulverizer Maintenance • Planned or Scheduled Outages • Forced
Outages • Component Maintenance • Maintenance Staffing • Beyond Maintenance and Repair
Maintenance and repair of Steam Turbines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-49
Inspections • Timing of Inspections • Spare Parts • Water Quality
Maintenance and repair of Generators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-51
Repair Actions
Maintenance and Repair of Emission Control Equipment. . . . . . . . . . . . . . . . . . . . . . 11-53
Electrostatic Precipitators • Fabric Filters • Spray Tower Absorbers for SO2 Removal • Dry
Scrubbing Systems for SO2 Removal
Chapter Twelve Retrofits Pat Jennings, Keith Berger, Jim Sutton, Phil Lafave, and Walt Lacey 12-1 to 12-57
BOILER RETROFITS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1
Unit Uprating • Emissions Control • Fuel Switching • Maintenance Upgrades
Steam Turbine Retrofits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-29
Retrofit Drivers • Retrofit Design Features • Turbine Retrofit Technology • Retrofit Installation
Emissions Control Equipment Retrofits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-43
Upgrades for Electrostatic Precipitators • ESP to Fabric Filter Conversion • Upgrades to Existing
Fabric Filters • Upgrade of SO2 Removal by Open Tower Spray Absorbers
Chapter Thirteen Performance Testing of Power Plants Matt Dooley 13-1 to 13-35
OBJECTIVE OF TESTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-1
Performance Evaluation/Routine Monitoring • Component Evaluation • Design Modification •
Contractual Guarantees Verification�
ASME PERFORMANCE TEST CODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1
PTC 46 Overall Plant Performance • PTC 4 Fired Steam Generators • PTC 6 Steam Turbines •
PTC 6.2 Steam Turbines in Combined Cycle Applications�
OTHER TEST METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-3
U.S. Environmental Protection Agency (EPA) Reference Methods • High Pressure-Intermediate
Pressure (HP-IP) Packing Leakage • Air Heater Leakage • Furnace Outlet Temperature
TESTING RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-3
Heat Rate • Boiler Efficiency
MEASUREMENTS AND INSTRUMENTATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-6
Temperature • Pressure • Flow • Power • Fuel Analysis�
Validation And Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-17
Combustion Calculations • Excess Air • Calculating Flue Gas Analysis • Enthalpy Equation Constants
from Draft of PTC 4.4 • Boiler Efficiency • Boiler Heat Output—Turbine Cycle Heat Input�
Contents | ix
Chapter Fourteen Maintaining Availability: Condition Assessment and Remaining 14-1 to 14-19
Life Analysis Phil Lafave, Charlie Hart, Jack Matton, Brian Morris, Charles McDaris,
Dan Gelbar, Jim Geyer, Michael Rogers, and Jerry Chase
AVAILABILITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1
Productivity Indices • Measuring Availability • Availability Statistics • The Power Producer’s
View of Availability • Reserve Margins • Problems and Their Root Causes • Availability/Reliability
and Life Continuation�
MAINTAINING AVAILABILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-7
Why a Need for Life Continuation? • Assessment of Remaining Life of Steam Generator Components
SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-17
Chapter Fifteen Alternative Fuels Doug Hart, William Bailey, and Carl Bozzuto 15-1 to 15-11
SOLID FUELS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1
Biomass • Fossil Fuel–Based Products • Municipal Waste Fuels
LIQUID FUELS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-9
Orimulsion • Alcohol Fuels
GASEOUS FUELS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-11
Index 1 to 22
Foreword
The world today displays a wide spectrum of economic development. This translates into
disparate needs, and challenges society to create and implement strategies that promote
improved standards of living, sustain and empower growth, and protect the environment.
Society requires dependable production of electric power. Electricity provides the ability
to increase productivity, comfort, and technological devel- the increased emphasis on cleaner generation of electricity, it
opment throughout the world. There are many ways that becomes more important to understand all of the processes
power is generated including the use of nuclear, hydro, wind, that must work together in order to produce electricity to
and solar power technologies. However, fossil fuels (coal, meet increased demand. To that end, there are 15 chapters
natural gas, and oil) are the mainstay of power production, that cover the primary components of a power plant as well
providing more than 70% of power generation. Although all as controls, materials, operations, maintenance, and retrofits.
sources of energy will be needed in the future, fossil fuels are This material represents a compendium of works that provide
expected to dominate energy production for a long time to an overview of the technology intended for students and those
come, and to provide the diversity of energy supplies needed that are relatively new to the industry. These chapters are in
to provide reliable power delivery. the hardcover portion of the book. With this edition, a new
Fossil power generation has evolved considerably over feature has been added in the form of a CD. The CD includes
the years with increased efficiency and reduced emissions searchable .pdf files of all 15 chapters and the 7 appendices
through the use of advanced designs and new technolo- that go with the book. In this manner, the “look up” type of
gies. Average heat rates (the amount of energy required information that is included in the appendices, as well as in
to generate a kW-h of electricity) improved from 25,000 the preceding chapters, will be easily accessible without the
BTU/kW-h in the early 1900s to 6,600 BTU/kW-h for the need to have the physical book present. In addition, a new
best combined cycle plants in the early 21st century. At the feature has been included in the form of an executable file: an
heart of these technologies is the combustion process and emissions calculator. This feature allows the user to perform
the associated environmental control processes that allow for combustion calculations, make English to metric (or other
clean, efficient power production. This book will provide the conversions), compare emissions levels and rates in differing
reader with an overview of the various design considerations, formats, and perform boiler efficiency and draft loss calcula-
equipment, and technologies that will maintain fossil power’s tions. These features are intended to provide a more valuable
prominence in supporting society. resource in a more convenient and easily usable format.
In this fifth edition, the subject matter has been broadened
to include more of the overall power plant equipment. With Carl Bozzuto
xi
Acknowledgements
We dedicate this edition of Clean Combustion Technologies, with thanks, to the many people
who contributed their talent and time, in great or small measure, to revising and updating its
technical base, and encouraging and supporting its publication.
To Joseph Vasile for recognizing the value of the combustion technology textbook
and for helping to continue this legacy for the new engineering challenges ahead.
To Carl Bozzuto for his vast knowledge and technical capabilities in the editing,
rewriting, and proofreading of the entire book.
To Anne Guerin-Moens for her leadership, support, and vision which made
Clean Combustion Technologies possible.
Authors
Stefan Ahman Scott Darling Bill Herman Doug Martin
William Bailey Rich Donais Pat Jennings Roger Martinsen
Donald Bairley Matt Dooley Lars-Erik Johansson Thomas Mastronarde
Keith Berger Jean-Paul Drouin Kurt Johnson Gary Mattice
Jesse Bolinger Rory Eastman Peter Kelly Jack Matton
Carl Bozzuto Alan Ferguson Bob Kunkel Charles McDaris
Anisa Brown Woody Fiveland Walt Lacey Gary Mooney
Denis Bruno Steve Francis Phil Lafave Brian Morris
Bruce Carney Frank Gabrielli Paul Lafferty Anthony Munisteri
Paul Chapman Dan Gelbar Dennis Laslo Jack O’Rourke
Jerry Chase Jim Geyer Leif Lindau Edward Ortman
Chee Loong Chong Charlie Hart Mark Malo Ian Perrin
Larry Czarnecki Doug Hart John Marion Eric Pickering
(continued)
xiii
xiv | Clean Combustion Technologies
Authors (continued)
Dennis Piontek Glenn Selby Kevin Taugher Mark Wagner
Kjell Porle Harpreet Singh Marc Thomas John Wheeldon
Mike Rini George Stamatelopoulos Terry Totemeier Beat Zimmerli
Michael Rogers Richard Stone Dave Towle
Gus Shearer Jim Sutton Charles Tsirovasiles
Contributors
Iqbal Abdulally Tony Criswell Vince Llinares Peter Rufli
Arthur Adams Veronica Delsol Gail Marken Benjamin Sargent
Bevan August Lothar Eckert Carmen Marrone Jim Scholze
John Banas David Elkins Sharon Marshall Hermann Seelinger
Barbara Barberie Allan Ferry Pascal Michelin Jean-Claude Semedarde
Kenneth Barna Cathy Fischer Tapan Mukherjee Ted Sharp
Dave Bintz John Flynn David Muraskin Cheryl Smith
William Birmingham Art Fournier Don Mylchreest Wendy Smith
Ron Blodgett Tom Gibbons Peter Nilsson Charles Soothill
John Bona Luis Gonzales Michelle Nowak Paul Traccarella
Mike Borden Hans Haldermann Nsakala Nsakala Bill Urko
Daniel Bradfield Glenn Hill Kevin O’Boyle Rod Vanstone
Christian Breton Kelley Huebner Richard Pangrazzi Anke Vietgen-Zeidler
Oliver Briggs Thomas Hurd Eric Peckins Bruce Wilhelm
Lynn Broders Bill Keegan Joe Pereira Randy Wilson
Graham Butt Heinrich Klotz Allen Pfeffer Manfred Wirsum
Ray Chamberland Edwin Kramer Steve Renals Andrea Woods
Michelle Clayton Heinrich Lageder Dan Roberts John Zeppa
Tricia Cole Chris Lech Troy Robinson
John Crean Frank-Udo Leidich Christoph Ruchti
Introduction
The world today faces a critical challenge as all nations strive to satisfy basic human
requirements—food, shelter, clothing, and work—which are so dependent on adequate
supplies of energy. The great increase in the use of energy has been met mostly by fossil
fuels—primarily coal, oil, and gas. Environmental concerns, security of supply, and economic
impacts all must be balanced as the demand for energy con- with Alsthom to form GEC Alsthom. In 1998, GEC and
tinues to increase. Real economic growth and energy use are Alcatel sold off part of their shares to allow the company
still inextricably linked. to be publicly listed on the Paris stock exchange as Alstom.
While the search for ultimate solutions to provide In 1999, ABB and Alstom formed a joint venture company
adequate energy supplies continues, near term, interim for their power businesses (ABB Alstom Power Ltd.). In
approaches must be considered for meeting the immediate 2000, Alstom purchased ABB’s interest in the joint venture
growth in demand for energy. Technological improvements company from ABB, leading to the creation of what Alstom
in the mining, drilling, moving, processing, and using of fossil is today in the power industry. The pursuit and realization
fuels can, of course, stretch energy resource reserves, as can of technical innovation has led to Alstom becoming a global
a determined effort at conservation of energy. Similarly, the leader in power generation.
utilization of advanced clean coal technologies such as coal In 1918, Combustion Engineering, Inc. (C-E), initiated
gasification, fluidized bed combustion, or ultra supercritical the development of pulverized coal firing. Overcoming the
pulverized coal technologies is capable of widening the use limitations imposed upon boiler capacity by virtue of the
of the world’s vast coal resources. Developing new ways to size restrictions associated with the use of stokers, pulverized
realize the best possible use of the world’s energy supplies is coal firing opened the door to making it possible to realize
what Alstom traditionally does best. a large increase in the capacity of coal-fired boilers. Capital-
Alstom represents a significant number of companies with izing on the pulverized coal firing system that was being
roots going back to the 1880s. Such names as Grieve Grate employed for steam locomotives, C-E successfully adapted
Company and the American Stoker Company brought this pulverized coal-firing technology for use with the
together two well-known manufacturers of fuel-burning stationary utility boilers at the Oneida Street and Lakeside
equipment in 1912 to form what became Combustion Stations of the Milwaukee Electric Railway and Light Co.
Engineering, Inc. With the merger of Société Alsacienne This pioneering work, done by C-E jointly with Milwaukee
de Constructions Mécaniques (SACM) and Thomson- Electric’s engineers and the Bureau of Mines, marks one of
Houston in 1928, Alsthom was born. In 1969, Compagnie the most important developments in the history of the art of
General d’Electricite (CGE) became the majority sharehold- solid fuel combustion. As a result, in 1980, the Oneida Street
er of Alsthom. During the 1980s, ASEA and Brown Boveri Station, where pulverized coal-firing was first employed by
merged to form ABB. In 1990, ABB purchased Combus- a utility, was named an Historic Engineering Landmark by
tion Engineering, Inc. During that time, UK GEC merged the American Society of Mechanical Engineers. Through the
xv
xvi | Clean Combustion Technologies
acquisition of two boiler companies, the former Heine Boiler PULVERIZED COAL FIRING DEVELOPS
Company of St. Louis and the former Hedges-Walsh and The pioneering work that was done with regard to the
Weidner Boiler Co. of Chattanooga, C-E was able to aug- pulverized coal firing system that was employed at the utility
ment its manufacturing capabilities. Alstom’s present major boiler plants of the Milwaukee Electric Railway and Light
manufacturing facility in Chattanooga is on the site of the Co. was followed in 1927 by the introduction of tangential
former Hedges-Walsh and Weidner Boiler Co. firing technology, in which fuel and air are introduced into
Up until this time, steam generating units originally were a furnace from the furnace’s four corners instead of through
an assembly of separate components that were bought from the furnace’s front or rear walls. First installed commercially
different manufacturers. However, with the addition of these at the U.S. Rubber Co. plant in Detroit, Michigan, tangen-
two boiler companies, it was now possible to combine all of tial firing technology provides the optimum in the rapid
the major components (i.e., the boiler, the superheater, the and intimate mixing of fuel and oxygen to promote efficient
water-cooled furnace, the fuel-burning equipment, and the combustion. Later, in 1940, a major modification to the
heat-recovery equipment) of a steam generating unit into basic concept of tangential firing technology was applied to a
one coordinated design. For many years, C-E licensed its boiler at Duke Power Co.’s Buck Station. With this modifi-
technology relating to steam generating units to other com- cation, the fuel and air nozzles in the corners of the furnace
panies such as Stein Industry, EVT, Sulzer Brothers, and were allowed to tilt vertically. This new design enabled the
International Combustion, Ltd. All of these companies that flame envelope that is created by the action of the tangential
are named here are today now a part of Alstom. firing system to be moved up and down within the furnace,
In electric power generation systems, the steam that is thereby changing the heat absorption pattern that is applied
produced by steam generating units is employed in steam to the furnace’s walls. Because the temperature of the hot
turbines. Such steam, which is at a high temperature and gases that enter the superheater surfaces is capable of being
pressure, is expanded in the steam turbine thereby causing varied in this manner, a means is thus provided through
the steam turbine to rotate, which in turn causes a generator which control can be exercised over steam temperature. Such
that is connected to the steam turbine to rotate as well. When a tangential firing system continues to provide the basis
the generator is rotated a conductor is made to move through throughout the world for the design of utility boilers that are
a magnetic field, thereby causing an electric current to be supplied by Alstom.
generated. This fundamental mode of operation is the basis Another piece of equipment that is employed in steam
upon which power generation is today predicated. Companies generating units, which has become a standard in the
such as Stal Laval, ASEA, Brown Boveri (BBC), Alsthom industry, is the Alstom Bowl Mill, which was introduced
Atlantique, and General Electric Corp. were the major early commercially by a forerunner company of Alstom in 1933.
manufacturers of steam turbine equipment in Europe. The first Originally based on the technology developed by Raymond®,
all welded rotor for a steam turbine was installed in 1929 at the a forerunner company of Alstom, the design of the Alstom
Esdori plant in Japan by a forerunner company of Alstom. Bowl Mill has over the years been modified and increased
In the search for means for realizing higher efficiency for in size starting with a coal capacity of 3,500 lb/hr coal to
utility plants, the pressures and temperatures at which utility more than 100 tons/hr today, in order to keep pace with the
boilers are operated have been increased. To this end, in requirement for supplying pulverized fuel to larger and larger
the mid-1920s, most utility boilers supplied steam to steam steam generating units.
turbines at 400 psig, with a few such utility boilers being
capable of supplying steam at a pressure of as high as 600 WELDED DRUMS PERMIT THE USE OF
psig. By the late 20s, some utility boilers capable of operat- HIGHER PRESSURES, TEMPERATURES
ing with throttle pressures of 1,200 pounds per square inch A major contribution in the development of modern steam
were sold. During this same period, the temperature of the generating units has been the creation of the welded boiler
steam increased gradually from between 500°F and 600°F to drum. Until the late 1920s, heavy pressure vessels were
a range of 700°F to 750°F, which represented a substantial constructed from formed plates that were riveted together.
advancement considering the types of materials that were The joints of these formed plates were caulked to prevent
available at that time. leakage. This form of construction served to physically limit
Introduction | xvii
the operating steam pressure to which the non-welded boiler that the surge toward the use of the reheat cycle in the utility
drum could be subjected. C-E, at its Chattanooga, Tennes- industry began in earnest. The first post-war steam generat-
see plant, developed and has perfected processes, techniques, ing unit having a reheat cycle was sold in 1947 to Boston
and machines for the fusion welding of heavy metal plate in Edison Co. for its Edgar Station. In the following years,
order form cylindrical shells from which boiler drums could designs for steam generating units that had a reheat cycle
then be fabricated. The first such welded boiler drum was became the standard.
dedicated as a National Historic Landmark by ASME on C-E designed and installed the first Controlled Circula-
May 2, 1980—50 years after this welded boiler drum was tion® steam generating unit at the Montaup Electric Co. in
first successfully tested. Welded construction enabled steam Somerset, Massachusetts, in 1942. In accordance with this
pressures to be increased. In 1931, the first boiler to operate design, the circulation of water is ensured and controlled
at a pressure of 1,800 psig was installed at the Phillip Carey through the utilization of pumps that are designed to
Co. in Lockland, Ohio. By 1953, the first steam generating provide a positive flow through the heat-absorbing tubes in
unit capable of operating at a pressure of 2,650 psig was the walls of the steam generating unit. The available head
designed for the Kearny Station of Public Service Electric produced by such pumps enables distribution orifices to be
and Gas of New Jersey (PSE&G). employed in the flow circuits through the heat-absorbing
With improvements in the strength of materials that tubes in the walls of the steam generating unit. With the
permitted such materials to be operated at higher tempera- Controlled Circulation® steam generating unit design, it is
tures and pressures, the outlet steam temperatures of steam possible to overcome the problem of decreasing thermal
generating units began to climb from the 750°F tempera- circulation as the operating pressure of the steam generating
ture level. By 1939, the outlet steam temperatures reached unit approaches the critical pressure point of 3,208 psia. The
925°F at the River Rouge plant of the Ford Motor Co. Controlled Circulation® steam generating unit design has
The first steam generating unit to exceed a 1,000°F total been well accepted around the world by the utility industry.
steam temperature was sold to PSE&G for its Sewaren A license agreement was executed in 1953 between C-E
Station in 1949. This steam generating unit was designed and Sulzer Brothers of Switzerland giving C-E the rights
to be operated at a temperature of 1,050°F at the super- to build and sell in the United States the Sulzer Brothers’
heater outlet. In 1953, the Kearny steam generating unit of Monotube steam generating unit design. (The Monotube
the same utility was designed to operate at a temperature steam generating unit design comprises a once-through type
of 1,100°F and a pressure of 2,376 psig at the superheater steam generating unit in which the water that is introduced
outlet and at a temperature of 1,050°F at the reheater outlet. at the inlet of the steam generating unit passes continuously
In 1954, C-E supplied the supercritical pressure unit for through the tubing in the walls where the water is heated
Philadelphia Electric Co.’s Eddystone Station, which was to the desired outlet temperature without being subjected
designed for outlet conditions of a temperature of 1,210°F to any other type of internal recirculation.) In 1954, the
at a pressure of 5,300 psig. This supercritical pressure unit Philadelphia Electric Company purchased from C-E the
is still in operation and continues to represent the highest first steam generating unit embodying this Monotube
temperature and pressure steam conditions to date at which technology for the Eddystone Plant. This steam generating
a steam generating unit has been operated. unit was designed to operate at a pressure of 5,300 psig and
In its search for higher efficiencies for steam generating a temperature of 1,210°F at the superheater outlet, which
units, the utility industry adopted the reheat cycle. This represented the highest pressure and the highest tempera-
concept involved the reheating of steam in a section of the ture operating conditions for a steam generating unit of
steam generating unit after some of the steam’s energy had commercial size ever. This unit went into operation in 1959,
been extracted through the expansion of the steam in the and was followed by several other supercritical and subcriti-
initial section of the turbine. This reheated steam, which is cal pressure installations of the Monotube design as well.
now at or near the steam’s initial high temperature, is then The next advancement in high-pressure steam generat-
returned to the final section of the turbine. Although a few ing unit technology was the Combined Circulation® steam
steam generating units that embodied a reheat cycle had generating unit design, which made use of the principle
been built in the 1920s, it was not until after World War II of pumped circulation of water in the walls of the steam
xviii | Clean Combustion Technologies
generating unit (recirculation of the fluid by pumps) coupled which had purchased both C-E and the Flakt Company,
with the once-through flow employed in the Monotube combined together the operations of these two companies
technology. This technology was followed by sliding pres- in order to form what is now known as the Alstom Envi-
sure, supercritical designs. Current developments are aimed ronmental Control Systems group. This group includes
at increased steam temperatures and pressures for higher electrostatic precipitators, fabric filters, selective catalytic
efficiency and reduced emissions. reduction (SCR) systems, mercury capture technology,
and presently is now in the process of developing new CO2
BROADENED CAPABILITIES capture technology.
Air preheaters are capable of providing heat recovery from In the latter part of the 1980s, the Brown Boveri Com-
the hot flue gases that leave steam generating units, thereby pany of Switzerland and Germany and ASEA from Sweden
improving the efficiency of these units. During the 1920s, merged to form ABB Ltd. In 1939, BBC introduced the first
the Ljungström® continuous, regenerative-type, air preheater electrical generating gas turbine in Neuchatel, Switzerland.
was developed. This development of the Ljungström® air pre- In 1989, General Electric Corp (UK) and Alcatel Alsthom
heater occurred in connection with the development of the first merged to form GEC Alsthom. ABB subsequently pur-
turbine locomotive. In the United States, the first Ljungström® chased C-E in 1990. In 1998, GEC Alsthom was renamed
air preheater installations for industrial power plant steam Alstom. In 1999, ABB and Alstom merged their respective
generating units occurred in 1923. In 1925, the Air Preheater power businesses into a joint venture company known as
Corporation, which was jointly owned by the Ljungström ABB ALSTOM Power Ltd. In 2000, Alstom purchased
Turbine Manufacturing Co. of Sweden and the James Howden from ABB, ABB’s interest in the joint venture company, lead-
Co., Ltd., was founded. The Air Preheater Corporation was ing to the formation of the present day Alstom S.A., which
acquired by C-E after the depression, along with an exclusive is headquartered in Paris.
license to manufacture the Ljungström® air preheater design The power generation industry has gone through a lot
in the United States and Canada. In accordance with the of consolidation over the years. The advances that have
Ljungström® air preheater design, a rotating heating element, been made over more than a century in the development
made up of closely spaced metal plates, is packed in baskets of energy conversion processes from fossil fuels have been
through which hot flue gases leaving the steam generating substantial. To this end, energy conversion efficiencies have
unit and cold combustion air flowing into the steam generat- increased from near 10% to over 40% (HHV basis) during
ing unit are alternately made to pass. This Ljungström® air that time frame. Moreover, the amount of emissions from
preheater design, which is manufactured in Alstom’s Wells- steam generating units has been reduced substantially
ville, New York facility, has been universally accepted within compared to the amount of emissions produced by earlier
the power generation industry and has been employed with designs. Those individuals who have worked in the power
steam generating units that are supplied by all steam generat- generation industry during some part of that time frame
ing unit manufacturers. have been able to witness not only tremendous growth in
C-E was one of the pioneers, along with the Flakt Compa- the art and science of fuel burning and steam generation,
ny in Sweden, in the development of flue gas desulfurization but also remarkable changes in the lifestyles of people
equipment for the removal of sulfur dioxide (SO2). Research throughout the world that have been brought about at least
and development work on this technology was begun by C- in part because of such growth. Today, new developments
E in 1963. A pilot facility was constructed in C-E’s Windsor, are underway that are expected to open up new vistas for
Connecticut development laboratory. This was followed in the use of fossil fuels for power generation purposes. In
1966 and 1967 by experimental field work that was done this book an attempt will be made to cover many of these
on a steam generating unit at Detroit Edison Co.’s St. Clair new developments.
Station. The first commercial scrubber installations in the The Power Systems Sector of Alstom is capable of design-
United States for the removal of SO2 went into service in ing, manufacturing, and supplying a broad range of products,
1968 at Union Electric Co’s Meramec Station and at Kansas systems and services to the power generation and industrial
Power and Light Co’s. Lawrence Station. In the 1990s, ABB, markets and has at last count a presence in 70 countries
Introduction | xix
around the globe. This range of products, systems, and ser- Throughout this book, these products, systems and
vices, by way of exemplification, includes: services, and their associated technologies will be referred
to as Alstom products or Alstom technologies, regardless
Plant Engineering of their origin.
Steam Turbines Alstom continues to embrace the same values that
Gas Turbines have kept it at the cutting edge of power generation
Combined Cycle Plants technology. A dedication to service and the courage to
Boilers lead—attributes that have been major influences through-
Environmental Control Equipment out the history of the power generation industry—still
Energy Recovery Systems guide its operations today. As it faces the technical
Clean Coal Combustion Technology challenges of the future requirements for energy and the
Controls environment, Alstom will continue to lead in the develop-
Maintenance and Retrofit Services ment of Clean Combustion Technologies.
Repowering and Rehabilitation Capabilities
Fossil-Fueled Power
Plant Design
Chapter One
1-
1- | Clean Combustion Technologies
through the walls of the boiler to water chapters. An isometric view of the
that flows through those walls, thus major components of a fossil-fueled
generating steam. Next, the steam is power plant is shown in Figure 1-2.
separated from the water and is sent to Fossil-fueled power plants can be de-
a steam turbine. The rotor of the steam signed so they are suitable for myriad
turbine rotates as the steam passes purposes, including electric power
over the turbine blades. This rotation production, thermal energy applica-
in turn causes a generator set, which tions, and standby/emergency power
is connected to the rotor, to rotate as systems. From the early concepts of
well, thus producing electric power. industrial power to the later, central
When the motive force provided by the station designs and back again to de-
steam is fully consumed, this steam is signs for smaller, cogeneration systems,
sent to a condenser to return the steam power plant technology and its applica-
to a liquid state so it can be pumped tions have continued to evolve.
back to the boiler. Thus, the water that The design of a fossil-fueled power
flows through the walls of the boiler is plant begins with the application of
continuously reused. certain basic, fundamental principles
After most of the useful energy that govern the relationship between
has been transferred to the water the properties of matter, which define
that flows through the walls of the the conversion of energy from one
boiler, the products of the combus- form to another. These properties
tion process must be cleaned. In of matter are known as the first and
this regard, various emission control second laws of thermodynamics, and
devices are employed. Such devices they provide a quantitative method of
typically include particulate control describing sequentially the processes of
systems, NOx removal systems, and a working fluid (predominantly water
SO2 removal systems. Other types in liquid or gaseous form [i.e., steam])
of emission control devices could be as functions of temperature, pressure,
used for controlling mercury and, enthalpy, and entropy.
eventually, CO2 as well as other sub- The first law of thermodynamics
stances. Each of these components is basically a “conservation” law. It
and systems will be discussed in later states that matter and energy can be
neither created nor destroyed. Thus,
the heat that is released during the
combustion process must end up
somewhere (preferably, mostly in the
steam). Moreover, the products that
result from the combustion must
be accounted for, because they do
not just disappear. Simply stated, at
steady state, the inputs must equal the
outputs. This fact enables heat and
mass balances to be created, which
provide an accounting as to where all
the matter and energy go during the
combustion process.
Figure 1-2 | Isometric view of a fossil-fueled power plant
Equation 1-1
1- | Clean Combustion Technologies
where in and out refer to input and The steady-flow, adiabatic expansion
output, respectively, or of a fluid through a turbine is governed
by the following equation:
V 2 V2
W = m h + + z − h + + z I = mi si = − m( sin − sout )
2g 2g
in out
Equation 1-6
Equation 1-3
two reservoirs that are each at a fixed calculated. Consequently, when realistic
temperature. The Carnot cycle con- irreversibilities are introduced into the (A) Flow System
sists of several reversible isothermal calculations for the efficiency of the Car- High-Temperature b
Heat Source
and isentropic processes. Figure 1-3A not cycle, the actual efficiency is reduced,
illustrates a flow system for executing and equipment sizes must be increased Q1
WT
the Carnot cycle. The temperature– to attain the desired work output. c
entropy diagram shown in Figure 1-3B a Q2
depicts the state changes experienced Rankine Cycle
WC
by the working fluid. The cornerstone of the cycle for a mod- d
The classic Carnot cycle is such ern steam power plant is a modification
Compressive
that no other cycle can have a better of the Carnot cycle proposed by W. J. Low-Temperature
Work
efficiency between the same specified M. Rankine,8,9 a distinguished Scottish Heat Sink
WT b P1 c
Q1 d e T1
T
dT
a d
Q2
T2 e
b f
f
Boiler Condenser
a Pump S
Figure 1-8 | Reheat regenerative cycle for a 600-MW, subcritical-pressure, fossil-fueled power plant (U.S. units)
28.9 kg, 3394 h, 2.52 MPa abs 16.0 kg, 3,212 h Boiler
1.21 MPa abs Feed Pump
30.1 kg, 3,212 h, 1.27 MPa abs
Figure 1-9 | Reheat regenerative cycle for a 600-MW, subcritical-pressure, fossil-fueled power plant (SI-metric units)
Figure 1-10 | Reheat regenerative cycle for a 800-MW, supercritical pressure, fossil-fueled power plant (U.S. units)
Figure 1-11 | Reheat regenerative cycle for a 800-MW, supercritical pressure, fossil-fueled power plant (SI-metric units)
the regenerative, reheat, supercritical reduce the temperature of the heat that
pressure heating processes. With is rejected. A penalty is associated with
such supercritical pressure plants, employment of the open Brayton cycle.
cycle efficiencies exceeding 45% This penalty comprises the compression
(higher heating value basis) have work that is required by the air compres-
been reached. sor of the combustion turbine. New
combined cycle plants generally fall into
Combined Cycle two categories: a combustion turbine
In combined cycle plants, a Brayton plus an unfired heat recovery steam
cycle is integrated with a conventional generator (HRSG), and a combus-
Rankine cycle to recover waste heat. tion turbine plus a supplementary fired
To this end, combined cycle plants use HRSG. In the latter, the supplementary
one of a number of possible con- firing is designed to occur at atmospheric
figurations of combustion turbines, pressure, and it makes use of the thermal
waste heat recovery boilers, and steam energy and the remaining oxygen that
turbines. Such combustion turbines are is contained in the exhaust products of
predominantly fired with natural gas combustion to burn additional fuel.
but can also be fired with jet fuel, light For a combustion turbine plus
distillate, or with proper pre-treatment, unfired HRSG (Fig. 1-12), all the fuel is
a wide range of residual and crude oils. subjected to combustion in the turbine,
In an open Brayton cycle, a combus- and the HRSG depends entirely on
tion turbine engine is employed in the combustion turbine as the HRSG’s
power generation applications. More heat source. Because the power from the
specifically, the open Brayton cycle con- steam turbine is produced without any
sists of the following four processes: additional fuel input, and because only a
small decrease occurs in the efficiency of
1. A reversible isentropic compres-
the combustion turbine as a result of the
sion in the inlet air compressor.
increased backpressure of the HRSG,
2. A constant-pressure fuel
the overall efficiency of the combined
combustion.
cycle plant that employs a combustion
3. A reversible isentropic expansion
turbine plus unfired HRSG is improved
in the turbine and exhaust nozzle,
over that of the simple (or open) cycle
which is designed to drive the
combustion turbine.
Fuel compressor and the generator.
The combustion turbine plus sup-
4. Discharge of the products
plementary fired HRSG makes use of
of combustion.
the oxygen that remains in the exhaust
Compressor Turbine The discharge of the exhaust of the combustion turbine through
Generator
products of combustion makes this an incorporation of a supplemental firing
Air
“open cycle.” The open Brayton cycle is system, which is designed to be located
intended to function such that the work- in the connecting duct between the
Steam Heater Turbine
ing fluid temperature is increased as a combustion turbine and the HRSG.
Generator Train
Generator consequence of the high-temperature Depending on the size of the combus-
Exhaust capabilities of the combustion turbine, tion turbine and the temperature limit
while the steam turbine is employed to of the HRSG, it is possible to double
Figure 1-12 | Gas turbine plus unfired heat
recovery steam generator
(HRSG) Technologies
Clean Combustion
Figure: 01.12/ kb
PMS5415 K
Fossil-Fueled Power Plant Design | 1-15
Cogeneration
Cogeneration is the simultaneous pro- Steam Heater Turbine
duction of power (usually electricity) Generator Train
Fuel Generator
and thermal energy (usually steam or
hot water) from a single fuel-consuming Exhaust
process. Cogeneration also is often re-
ferred to as “combined heat and power.” Figure 1-13 | Gas turbine plus furnace fired steam generator
Cogeneration plants are most suitable
for applications that have coincident
Clean Combustion Technologies
electric and thermal loads and that range a given facility, the designFigure:
and configu-
01.13/ kb
in size from small industrial and insti- ration of cogeneration plants can vary
PMS5415 K
tutional plants to large central stations greatly from site to site. For example,
serving the electric and thermal needs of cogeneration plants can use either a
a municipality or petrochemical instal- topping cycle (electricity is produced
lation. Similar electric and thermal load first, and the exhausted thermal energy
profiles enable cogeneration facilities to is used thereafter) or a bottoming cycle
operate at a higher annual load factor (electricity is produced through the
and higher thermal efficiency than could recovery of waste heat from high-
be obtained by separate plants to handle temperature processes). In either case,
the power need and the thermal need. the prime movers in each such cycle
The design of cogeneration facili- can be conventional boilers, combus-
ties differs from that of conventional tion turbines, or reciprocating engines.
Rankine cycle plants. The main goal of Moreover, the thermal needs can be
a Rankine cycle plant is to economically met by the turbine exhaust, extraction,
optimize thermal efficiency during the or the waste heat that is recovered
production of electric power, whereas from the combustion of the fuel. Fur-
the goal of the cogeneration facility may thermore, if the amount of energy that
be to maximize the use of waste heat for is produced by such cycles exceeds the
heating purposes. The overall efficient facility’s own needs, the excess energy
use of the heat that is liberated from (thermal and/or electric) can be used
combustion of the fuel, however, is still for other needs.
greater than what could be achieved Figure 1-14 shows a schematic dia-
with a separate plant for each purpose. gram of a basic closed power/process
Because of the varying nature of the cycle wherein a backpressure turbine
power needs and the thermal needs of is employed. In such a cycle, after
1-16 | Clean Combustion Technologies
Spill b
b Valve
WT
d c
Q1
WT d
Q1 Q2
h
Pressure Turbine Q2
Reducing c a
Valve Heater e
Desuperheater S
Boiler
a
e
Pump
WP
Figure 1-16 | Heat balances of a steam topping cogeneration cycle (SI-metric units)
life-cycle cost analysis for each poten- that have, for instance, been designed
tial site. Such a life-cycle cost analysis for use in accordance with one operat-
focuses on a number of contributing ing scenario may, in practice, later be
factors, including (but not limited to) required to operate in accordance with
available real estate, proximity and another operating scenario that had not
robustness of the closest electrical been contemplated when the plant was
transmission lines, sources of fuel and originally designed.
cooling water, environmental consider- The major factors that determine
ations, and any relevant public opinion. what cycle should be used—that is,
Additional considerations include such the conventional Rankine cycle, open
matters as availability of municipal Brayton cycle, or combined cycle—are
services, availability of labor (construc- generally the operating scenario, fuel
tion and operation), any relevant zoning availability, and cost. Typically, the open
regulations, present use of any adjacent Brayton cycle (i.e., the simple cycle) is
Life cycle cost analysis considers properties (i.e., airport), access to the used for peaking service, whereas the
all of the important factors potential site, climate at the potential Rankine cycle or combined cycle are
in selecting a power plant, site, and tax treatment applicable to the normally selected for intermediate or
potential site. Room for possible growth baseload service. Fuel availability and
leading to the lowest cost and at the potential site as well as freedom cost are generally the drivers insofar
environmentally sound choice. from events such as flood and earth- as selection of the prime mover is
quake are also deemed to be essential. concerned. For example, fluidized bed
When considered in conjunction combustors are most commonly con-
with an operating cost analysis, which sidered for use when the fuel being
compares the proposed fossil-fueled employed is waste coal. Combined
power plant with existing power cycle plants are most commonly
plants, load profiles will indicate in considered for use when natural
which portion of the dispatch curve gas is the lowest-cost fuel. Conven-
and under which scenario the new tional boilers are most commonly
plant will operate—namely, whether considered for use if coal is the most
the new facility will be employed as a cost-effective fuel. When compar-
baseload plant, an intermediate plant, ing fuel costs, consider the following:
or a peaking plant. Operating scenarios, base price of the fuel; cost of the fuel
as the term is used here, are generally delivered; cost of fuel handling; cost of
defined in terms of how many hours labor; cost of plant services, includ-
annually the fossil-fueled power plant ing (but not limited to) auxiliary
is expected to operate. To this end, power loads; and fixed charges associ-
baseload fossil-fueled power plants ated with operation and maintenance
commonly operate 6,000 or more (O&M) of the fuel handling and firing
hours annually, intermediate power equipment. The type of cycle will drive
plants between 2,000 and 6,000 hours the selection of the main cycle compo-
annually, and peaking power plants less nents, such as the boiler, steam turbine,
than 2,000 hours annually. Note that combustion turbine, and/or HRSG.
no clear-cut line of demarcation exists Each one of these main cycle compo-
among these three operating scenarios. nents needs to be designed to match
In this regard, fossil-fueled power plants the operating scenario of the proposed
Fossil-Fueled Power Plant Design | 1-19
Tying together the intricate com- industrial and commercial plants are
ponents and systems associated with typically designed for electrical and
operation of a fossil-fueled power thermal loads that may differ between
plant is the control system. The control daytime and nighttime operation,
system is normally interconnected with between weekday and weekend/
the power plant operations through holiday operation, and between
field monitoring instrumentation and seasons. To address marked swings
final control devices. The control system in the demand to which industrial
is designed to provide indications, and commercial fossil-fueled power
alarms, and an automatic response plants can be subjected, first consider
to certain critical operating condi- making alterations in production to
tions. In addition, the control systems smooth the peaks in generation before
are designed to provide trending and moving on to implement changes in
reporting capabilities, to assess plant sizing and/or selection of equipment.
and equipment performance, and to To this end, by referencing electric
enable the operator to take manual and thermal load curves, it is possible
control of the operation of the plant to determine what changes, if any,
should this become necessary. Larger should be implemented in the size of
fossil-fueled power plants have typi- the boiler, turbines, and auxiliaries
cally used distributed control systems that are normally employed in in-
that are integrally connected with a dustrial and commercial fossil-fueled
combination of smaller (programmable power plants. Other issues that come
logic controller) control systems for into play commonly involve consid-
discrete systems within the power plant, eration of reliability of service (in
such as water/wastewater treatment terms of life cycle, safety, or high-cost
and emission control systems. Smaller production environments) as well as
fossil-fueled power plants that are regulatory requirements.
designed as peaking stations may even
be designed to allow remote operation Retrofitting an Existing
with no dedicated operating staff at Fossil-Fueled Power Plant
the plant site. As with the electrical An existing, operational fossil-fueled
distribution system, the control system power plant can be a most valued asset
of the fossil-fueled power plant em- for the owner. The owner’s most prized
bodies sufficient levels of redundancy possessions, generally speaking, are the
to allow continuous operation of the power plant’s valid operating permits.
power plant under the most stringent Any improvements to such an exist-
of failure scenarios. ing power plant that do not cause the
The design for industrial and com- operation to exceed the limits of the
mercial fossil-fueled power plants is existing, valid operating permits will be
very similar to what has been described a benefit to the present infrastructure
above for such power plants that are of the plant. The most important aspect
operated as central stations. The major to consider in evaluating the economic
difference between such plants that are benefit of a retrofit is the expected
operated as central stations and indus- operating life of the existing power plant
trial and commercial plants is that the after the retrofit has been implemented.
Fossil-Fueled Power Plant Design | 1-21
concern of both the design engineer power plant is a life-cycle cost analysis.
and the power plant owner. To this Initial capital costs as well as long-term
end, the initial cost of such equipment operating, maintenance, and replace-
must be balanced against the cost ment costs need to be balanced against
associated with operating such equip- the predicted revenue streams that will
ment over the lifetime of the power be generated by the power plant. By
plant. Before discussing how such the application of economic discount
comparisons are made, it is necessary factors, the present worth of the asset
to consider the two general categories (i.e., the fossil-fueled power plant)
of power plant costs. can be established. This methodology
Much of the money spent by the is most often termed discounted cash
owner of a fossil-fueled power plant flow analysis. To account for the time
is for goods and services that are value of money, the value of the future
intended to be consumed within a revenue streams from and the costs for
relatively short time after their acquisi- the power plant need to be discounted Proper economic analysis
tion. This category of costs includes insofar as time is concerned. Initial provides information needed
outlays for wages and salaries, operat- capital costs include such items as to make decisions to provide
ing and maintenance supplies, and licensing, permitting, engineering,
fuel. Such costs are most commonly design, procurement, construction,
electric power or steam at the
termed expenses, and the payments for legal, owner’s oversight, and interest lowest cost to the consumer.
these costs are normally made from the during construction on the fund-
revenue generated by the operation of ing raised to build the plant. On the
the power plant. other hand, the long-term operating,
The other category of costs for the maintenance, and replacement costs
owner of the power plant are items include such items as fuel, plant labor,
with a useful life that continues for third-party services, and maintenance
an extended period of time and that activities. Note here that cash flow
normally produce revenue in the analysis can account for any number
future. The money spent to construct of variables, including inflation, taxes,
a power plant is a typical example of interest rates, construction times, risk,
such an expenditure, and such costs maintenance outages, and so on. A
are generally termed capital investments. positive net present value indicates
Expenditures of this nature are not or- the construction and operation of the
dinarily paid directly from the revenue asset will provide a net benefit to the
that is generated from operation of the owner based on the assumptions that
power plant. The basic reason is that have been made. Variations on the as-
the current revenue being generated sumptions that have a major effect on
from operation of the power plant the present worth that is determined
would most often be insufficient to for an asset include the financing
cover large capital expenditures. In method, discount factor, timing and
addition, the equipment for the power the magnitude of the projected rev-
plant is expected to provide service well enue/cost streams, and the projected
into the future. operating life for the power plant. Un-
Critical to the decision making fortunately, it is not always possible to
and performance monitoring that are know every detail of the factors used
associated with owning a fossil-fueled in these assumptions.
1-24 | Clean Combustion Technologies
IC × FCR Table 1-1 | Typical Fixed Charge Rate Estimates, Utility Financing*
C=
100
Component Percentage
Equation 1-8
Interest 8–10
power plant to the fuel costs per equipment than it does to fabricate a
kilowatt-hour of power produced by smaller piece.
the plant is generally sufficient. The When the total power plant costs
equation that is normally employed is increase at a slower rate, percent-
the following: age-wise, than the rate at which
the designed output increases, then
economies of scale are possible.
FC × NPHR
F= Experience in the construction of fos-
10,000
sil-fueled power plants indicates that
Equation 1-10 economies of scale can be realized in
terms of the steam generating plant,
where the steam turbine generator set, the
condenser, the switchgear, and conse-
F = the operating fuel cost in
quently, in total costs.
cents/kWh
This type of cost behavior is often
FC = the purchased fuel cost in
represented by some form of power
$/million BTU
relationship, in which the power
NPHR = the net plant heat rate in
relationship is less than unity and,
BTU/kWh
more specifically, usually between 0.7
Thus, if the purchased fuel costs for and 0.9. If this power relationship is
a coal-fired power plant are $1.80/ denoted by k, then this type of cost
million BTU and the net plant heat behavior can be expressed as follows:
rate is 9,300 BTU/kWh, then the
operating fuel cost, F, for such a power k
IC1 S1
plant would be 1.7 cents/kWh. =
IC2 S2
Economies of Scale Equation 1-11
Many of the components that make where
up the total costs for a fossil-fueled
IC1 = the capital cost of a fossil-
power plant are relatively indepen-
fueled power plant that is of
dent of plant size or have been found
size 1
to increase slowly as the size of the
IC2 = the capital cost of a fossil-
power plant increases. Examples of
fueled power plant that is of
this might include a guard shack at
size 2
the entrance to the power plant or
S1 = the size of fossil-fueled power
a control room. These items might
plant 1
have the same cost for a 300-MW
S2 = the size of fossil-fueled power
power plant as they would for a
plant 2
600-MW plant. Likewise, as the size
of the equipment used in the power For example, if the cost of a 300-
plant increases, the surface to volume MW fossil-fueled power generating
ratio of such equipment normally unit is expected to be $2,000/kWh
decreases. As such, this means it ($600 million), then the total cost for a
generally takes less material per unit 500-MW fossil-fueled power generat-
of volume to fabricate a large piece of ing unit will, when k = 0.8, be
Fossil-Fueled Power Plant Design | 1-27
0.8
500 MW
300 MW × $2 , 000 / kWh × 300,000 kWh = $903,000,000
Based on the above calculation, the cost per kilowatt-hour of the larger fossil-fu-
eled power plant is now $1,806/kWh. As such, economies of scale can be an
important factor in selecting the size of a power plant for a given application.
Of course, this is not the only factor that needs to be considered. Other factors
include financing requirements, environmental impacts, site considerations (e.g.,
amount of space available for construction), electrical grid requirements, proposed
mode of operation, and any applicable regulations.
C1
C0 =
(1 + r )t
Equation 1-12
where
r = t he interest rate or the discount rate
1
The expression is most often termed the discount factor, or the present
( 1 + r )t
worth factor. As such, the present worth for a payment of $100,000 that is to be
1-28 | Clean Combustion Technologies
made 3 years later, assuming a discount costs. The concept of levelized costs
rate of 9% per year, would be calculated is an averaging process that uses level-
as follows: izing, wherein more weight is given to
costs that occur early in the life of the
1 power plant. To obtain levelized costs,
$100 , 000 × = $77 , 200
(1.09)3 it is only necessary to divide the total
present worth of the payments that
Equation 1-13 will be required by the sum of the total
present worth costs. For instance, level-
In other words, a payment of $77,200 ized total costs (LC) can be expressed
that is made now is equivalent to a pay- as follows:
ment of $100,000 that is made 3 years
Ci
later, assuming that the $77,200 would
be capable of being invested to generate (1 + r)t
LC =
a rate of return of 9% per year. 1
The total present worth costs is (1 + r)t
equivalent to the sum of all the costs Equation 1-14
associated with a fossil-fueled power
plant project that have been properly The levelized cost that results from
discounted to a common point in time. Equation 1-14 is a constant, annual
As such, the total present worth costs value designed to be equivalent to (i.e.,
can be used to compare two different have the same total present worth
possible implementations of the power costs as) that of the non-uniform
plant project, each of which involve series of actual annual costs for the
different cost expenditures at different fossil-fueled power plant project.
points in time. Thus, if the non-uniform annual
A concept closely related to total expenses for the power plant were
present worth costs is that of levelized assumed to be $157 million, $165
million, $167 million, $163 million,
and $170 million, respectively, over the
Table 1-2 | LevElized Costs and Present-Worth Costs of a New
first 5 years of operation, then the total
Generating Unit
present worth costs and the levelized
Present-Worth Expenses Present-Worth costs, using a 9% per year discount rate,
Year
Factor 106$ Expenses, 106$
would be as shown in Table 1-2.
1 0.917 157 143.9
Note here that the discount rate
2 0.842 165 138.9 designed to be used in such equations
3 0.772 167 128.9 is usually equivalent to the after-tax
4 0.708 163 115.4 return that is desired from the invest-
5 0.650 170 110.5 ment in the fossil-fueled power plant
Totals 3.889 637.6 project. This discount rate gener-
ally is the minimum return that any
$637.6 × 106 plant project must provide for it to
Levelized Expenses =
3.889 be deemed acceptable from a finan-
= $164 × 106 / yr cial standpoint. Also note, however,
that different discount rates can be
Fossil-Fueled Power Plant Design | 1-29
be removed from the system. The water performance during periods of high load
in the system must remain in balance. demand. Operating at peak performance
The appropriate amount of air must be maximizes efficiency, which reduces the
introduced and the product gases must level of emissions per unit of output.
be exhausted. These steps all require Chapter 11 describes the construc-
auxiliary equipment. As these pieces of tion and maintenance practices needed
equipment all require energy to operate, to build an efficient plant and maintain
design for performance and efficiency is such a plant over decades of operation.
a key to clean combustion technology. Proper maintenance allows the plant to
Chapter 7 covers the controls for a operate at an effective performance level,
modern plant. The operation and con- which reduces overall emissions, leading
trol of such large pieces of equipment to cleaner power generation.
involves special considerations in order Chapter 12 discusses retrofit ap-
to achieve the desired levels of efficiency, plications. Technology does not stand
emissions, reliability, load following, still. As advances are made, retrofit
and safety for the clean production of possibilities emerge. These retrofits may
electric power. be added on to existing equipment to
Chapter 8 describes some of the reduce emissions or substituted for older
materials that are used in large power equipment to improve efficiency, which
plants operating at high temperatures lowers overall emissions as well.
and pressures. Suffice it to say that these Chapter 13 provides some of the test
plants have to be built out of something procedures used on such large equip-
that is available at reasonable cost and ment. It would be difficult to know if a
still capable of withstanding years of unit was performing at peak efficiency if
operation under demanding conditions. it could not be tested and measured. As
As steam conditions are increased to noted in chapter 10, peak efficiency mini-
improve efficiency and reduce emis- mizes fuel use and associated emissions.
sions, improved materials will need to Chapter 14 describes some of the
be demonstrated. considerations for improving the main-
Chapter 9 provides some insights tainability and reliability of the plant
into the needs of the water chemistry equipment. These practices are impor-
for these plants. Water chemistry is tant because power plants must last a
important for both the boiler and the long time. Maintaining these plants in
steam turbine, as well as for some of the a safe and efficient operating condi-
auxiliary equipment. Poor maintenance tion provides for more production of
of the water chemistry is still one of the electricity with fewer additional plants.
leading causes of outages for boilers. Efficient operating conditions also
Good maintenance of water chemistry reduce overall emissions.
reduces forced outages, which reduces Chapter 15 provides some infor-
the number of startups and shut downs. mation on alternatives to the more
This, in turn, reduces the amount of fuel traditional fossil fuels. These alternative
that is wasted and the emissions associ- fuels are of interest both for cost con-
ated with that fuel use. siderations and environmental reasons.
Chapter 10 describes some of the op- Biomass-type fuels can be produced
erating considerations involved in large sustainably. To the extent that they can
fossil-fired power plants. These plants be utilized to produce energy, less CO2
have to be able to start up and shut is generated overall, leading to a cleaner
down safely as well as operate at peak environment overall.
Fossil-Fueled Power Plant Design | 1-33
References
1. E. A. Bruges, Available Energy and the Second Law Analysis. New York: 10. C. D. Weir, “Optimization of Heater Enthalpy Rises in Feed-Heating
Academic Press, 1959. Trains,” Institution of Mechanical Engineers. Proceedings, 174: 769–796,
1960. [Discussion by R. W. Haywood, pp. 784–787.]
2. J. H. Keenan, “A Steam Chart for Second Law Analysis,” Mechanical
Engineering, 54: 195–204, 1932. 11. G. Chiantore et al., “Optimizing A Regenerative Steam-Turbine Cycle,”
Transactions of the ASME. Journal of Engineering for Power, 83, Series A:
3. M. W. Thring, “The Virtue of Energy, Its Meaning and Practical Signifi-
433–443, 1961.
cance,” Institute of Fuel Journal, 17: 116–123, 1944.
12. J. K. Salisbury, Steam Turbines and their Cycles. Huntington, NY: Robert
4. C. Birnie and E. F. Obert, “Evaluation and Location of the Losses in a
E. Krieger, 1974, Part 3: Cycle Analysis.
60,000 KW Power Station,” Proceedings of the Midwest Power Conference.
Chicago: Illinois Institute of Technology, 1949, pp. 187–193. 13. “The Reheat Cycle—A Re-Evaluation.” Combustion, 21 (12): 38–40,
1950.
5. A. Keller, “Evaluation of Steam-Power-Plant Losses by Means of the
Entropy-Balance Diagram,” Transactions of the ASME, 72: 949–953, 14. Papers given at the Symposium on the Reheat Cycle sponsored by
1950. the ASME and held in New York, November 29–December 3, 1948.
Transactions of the ASME, 71: 673–749, 1949.
6. C. A. Meyer et al., “Availability Balance of Steam Power Plants,” Transac-
tions of the ASME. Journal of Engineering for Power, 81, Series A: 35–42, 15. J. K. Salisbury, “Analysis of the Steam-Turbine Reheat Cycle,” Transac-
1959. tions of the ASME 80: 1629–1642, 1958.
7. E. Clapeyron and R. Clausius, “Memoir on the Motive Power of Heat,” 16. J. K. Salisbury, “Power-Plant Performance Monitoring,” Transactions
in S. Nicolas, L. Carnot, E. Clapeyron, and R. Clausius, Reflections on the of the ASME. Journal of Engineering for Power, 83, Series A: 409–422,
Motive Power of Fire; and Other Papers on the Second Law of Thermody- 1961.
namics. Gloucester, MA: Peter Smith, 1962.
17. M. S. Reddy, F. Afshar, and R. J. Hollmeier, “Evaluation of Alternative
8. E. F. Obert and R. A. Gaggioli, Thermodynamics, latest edition. New York: System Designs in a Cogeneration Plant,” Proceedings of the American
McGraw-Hill. Power Conference. Chicago: Illinois Institute of Technology, 1982.
9. N. A. Hall and W. E. Ibele, Engineering Thermodynamics. Englewood
Cliffs, NJ: Prentice-Hall, 1960, pp. 447–520.
Combustion Technology
Chapter Two
The basic equations used to describe how fuel and air are transformed into gaseous products
of combustion are discussed in the first part of this chapter. Fuel-firing systems for burning
coal and other solid fossil fuels are described in the second part.
HISTORICAL INTRODUCTION
1781, Antoine L. Lavoisier substituted
From our earliest existence on earth, the theory of oxygenation for the theory
The control and utilization
humans have been fascinated by fire, of phlogiston and provided experimen- of heat energy has been a
but we did not achieve a quantitative tal evidence that combustion was the fundamental underpinning of
understanding of the combustion pro- union of the substance burned with the civilization. The combustion
cess until about the year 1880. Before oxygen of the atmosphere.
that date, one can trace the development In 1755, Joseph Black discovered
of fossil fuels has provided
of many hypotheses concerning the carbon dioxide, and, in 1781, Henry the principal means for the
nature and properties of fire, including Cavendish demonstrated the compound liberation of heat energy from
some that were expressed in super- nature of water. At about the same time, this natural resource.
natural terms of fear and awe. Even the Lavoisier made the precise measure-
now-discredited phlogiston theory of ments and formulated the volume and
combustion, however, did not prevent weight relationships that underlie the
enterprising engineers from designing modern theory of combustion.
and constructing boilers to generate Beyond this, Amedeo Avogadro
steam for the earliest steam engines. established in 1811 that the number
Phlogiston was a hypothetical, of molecules in a unit volume under
mysterious substance. It sometimes was standard conditions is the same for
presumed to have the property of nega- all gases. During this same period,
tive weight and combined with a body John Dalton enunciated the law of
to render it combustible. First proposed partial pressures, and, in 1803, his
by G. E. Stahl in 1697, the phlogis- study of the physical properties of
ton theory dominated the chemical gases led to formulation of the atomic
thought of the eighteenth century. Even theory, including the law of combin-
such a perceptive observer as Joseph ing weights. A related observation
Priestly, who in 1774 discovered the was made by Joseph Louis Gay-
unique power of oxygen for supporting Lussac in 1808—that gases always
combustion, accepted the phlogiston combine in volumes that bear simple
theory. In the years between 1775 and ratios to each other.
2-
2- | Clean Combustion Technologies
are the whole-number values for the Table 2-2 | Combustion Equations
main isotopes of each substance.
Combustible Molecular Heat Release,*
A molecular weight of any sub- Reaction
Constituent Weight BTU/lb
stance in the gaseous state and under
Carbon 12 C + O2 → CO2 14,100
the same conditions of temperature
Hydrogen 2 H2 + 0.5O2 → H2O 61,000
and pressure will occupy the same
volume. This relationship is very Sulfur 32 S + O2 → SO2 4,000
significant. The volume will, of course, Hydrogen
34 H2S + 1.5O2 → SO2 + H2O 7,100
sulfide
vary numerically for different units of
weight and for different conditions of Methane 16 CH4 + 2O2 → CO2 + 2H2O 23,900
temperature and pressure. For combus- Ethane 30 C2H6 + 3.5O2 → 2CO2 + 3H2O 22,300
tion calculations, the pound (lb) and the Propane 44 C3H8 + 5O2 → 3CO2 + 4H2O 21,500
cubic foot (ft3) are the units commonly Butane 58 C4H10 + 6.5O2 → 4CO2 + 5H2O 21,300
used in the United States, and unless Pentane 72 C5H12 + 8O2 → 5CO2 + 6H2O 22,000
otherwise stated, the temperature and
*Higher heating value/lb combustible
pressure are understood to be 32°F and
14.7 psia (atmospheric pressure). Thus,
a molecular weight of 32 lb of oxygen at 1 mol C + 1 mol O2 → 1 mol CO2
32°F and atmospheric pressure will have
the same volume as a molecular weight Equation 2-2
directly proportional to the absolute Because there are 4.32 lb dry air/lb
temperature and inversely propor- oxygen, the stoichiometric combustion
tional to the absolute pressure. Because of 1 lb of carbon requires 11.52 lb of
combustion processes in steam boiler dry air, or 11.68 lb of wet air (with
furnaces usually take place at practically 1.3% water vapor).
constant atmospheric pressure, pressure Each equation balances. There are
corrections are seldom necessary. the same number of atoms of each
Returning to the combustion element and the same weight of
equation for carbon and oxygen and reacting substances on each side of
applying these concepts, it is possible the arrow, but not necessarily the
to write this reaction in several ways. same number of molecules, moles, or
For purposes of molar analysis, carbon volumes. Thus, one atom of carbon
may be treated as a gas: combined with one molecule of oxy-
gen gives only one molecule of carbon
dioxide, and two moles of hydrogen
C + O2 → CO2
plus one mole of oxygen
Equation 2-1
2- | Clean Combustion Technologies
yield two moles of water vapor. It known as the mole method and is based
will be evident from a consideration on the chemical relationships previously
of the mole–volume relationship that explained. The second uses the firing of a
percentage by volume is numerically million BTU as a basis for calculation.
the same as mole percentage.
Because a mole represents a definite The Mole Method
weight as well as a definite volume, Table 2-2 gives the basic combustion
it is a means of converting analyses reactions for the carbon, hydrogen, and
by weight into analyses by volume, sulfur in coal. In the following analysis,
700 and vice versa. Volumetric fractions assume a high-volatile bituminous coal
of the several constituents of a gas burned at 23% excess air, and perform
650
can be multiplied by their respective calculations on the basis of 100 lb of
Molar Volume, ft3/mol
600
molecular weights, with the sum of the as-fired fuel. The fuel analysis as fired
550 products then being equal to the ap- is given in Table 2-3.
500 parent molecular weight of one mole of The calculation of air weight for
450 V = 0.7297 t + 335.65 gas. The percentage by weight of each combustion must be made on the basis
V = ft3/mol component can then be determined. of an oxygen balance, because oxygen
400 t = °F
Finally, the density of any gas at any is the only element common to all
350
0 100 200 300 400 500 temperature is found by dividing the oxidizing reactions. Oxygen contained
Temperature, °F molecular weight of the gas by the molal in the fuel must be deducted from the
volume at that temperature (Fig. 2-1). calculated quantity needed, because
Figure 2-1 | Gas density determination it is already combined with carbon,
Methods of Combustion hydrogen, or other combustible con-
Clean Combustion Technologies Calculations stituents of the coal.
Figure: 02.01/ SPS
PMS5415 K
Two methods of combustion calculations The molar relations are as given in
are presented in this text. The first is Table 2-4.
Fuel in Products, F . As defined earlier, F is that portion of the fired fuel that ap-
pears in the gaseous products of combustion. Because all quantities are to be those
required for, or resulting from, the firing of a million BTU, F must be calculated
on that basis. If a fuel contains no ash, then F is obtained through dividing
1,000,000 by the as-fired heating value of the fuel. For solid fuels, where ash and/
or solid combustible loss must be considered,
Atmospheric Air for Combustion, A . In accordance with the mole method, the
theoretical weight of dry air (zero excess) may be calculated from the fuel analysis
and the formula:
Equation 2-7
Equation 2-8
Values of Awet range from 570 lb/106 BTU for pure hydrogen to greater than
800 lb/106 BTU for certain cokes and meta-anthracite coals, as shown in Figure 2-2
Bituminous Coal
for various fuels burned in steam generators. Any calculated values of wet
Anthracite
Coke
Natural Gas
air for combustion that differ substantially from these values should lead to
Coke-Oven Gas
Refinery Gas
Lignite
800 cross-verification of the ultimate analysis and the observed HHV. The analysis
Blast-Furnace Gas
A, lb Air/106 BTU Fired
Oil
Bagasse
750 and HHV of the fuel have to be from the same sample to avoid errors in determi-
nations of air and gas weight.
700
Wood
Hydrogen
C, Correction Factor
100 1.50
0.98 2.00
Equation 2-10
0.97 3.00
If, however, all heat in the fuel does
0.96 4.00
not come from carbon alone (so that
the air is not strictly proportional to the 0.95 5.00
carbon burned), C will not be exact. For 4,000 6,000 8,000 10,000 12,000 14,000
HHV of Fuel, as Fired
high-carbon, low-volatile fuels, it will
be nearly exact and will result in only a
Figure 2-3 | Graphical solution of Equation 2-11
small error—even for fuels that are low
in fixed carbon and high in hydrogen. Clean Combustion Technologies
In all cases, the error involved by using Figure: 02.03/ SPS
Wa
Moisture in Combustion Air,PMS5415 K
Equation 2-10 is quite within the limits
of accuracy for all other combustion For heat balance calculations, the Relative Humidity
0.12
calculations. Finally, C can be expressed moisture in air will be 1.3% of air Saturation
0.10 80%
lb/106 BTU fired in the case of some An ideal fuel-burning system fulfill-
refinery gases. Thus, ing these functions would have the
following characteristics:
W f = Wc + Wh
• No excess oxygen or unburned
Equation 2-13
combustibles in the end products
where of combustion
• A low rate of auxiliary ignition-
Wc = H2O × 104/HHV
energy input to initiate the com-
Wh = 9 × H × 104/HHV
bustion reaction
where, in turn, both Wc and Wh are in • An economical reaction rate
lb/106 BTU fired, H2O is the percentage between fuel and oxygen compat-
water by weight in the as-fired fuel, H is ible with acceptable nitrogen and
the hydrogen in the fuel (% by weight as sulfur oxide formation
fired), and HHV is the higher heating • An effective method of handling
value of the as-fired fuel in BTU/lb. and disposing of the solid impuri-
ties introduced with the fuel
Dry Gas, Pd • Uniform distribution of the
The dry gas content of the combustion product weight and temperature in
products is used in the calculation of relation to the parallel circuits of
the dry gas loss item of a boiler heat the heat-absorbing surface
balance. The dry gas may be deter- • A wide and stable firing range
mined by subtracting the water vapor • Fast response to changes in
from the total products; thus, firing rate
• Equipment with high availability
Pd = P − (Wa + W f ) and low maintenance requirements
Equation 2-14
In actual practice, some of these
where characteristics must be compromised
to achieve a reasonable balance between
Pd = dry gas, lb/106 BTU fired
combustion efficiency and cost. For
P = total products of combustion
example, firing a fuel with the stoichio-
Wa = moisture in air
metric air quantity (no excess above the
Wf = moisture from fuel
theoretical amount) would require an
infinite residence time at temperatures
Combustion Systems above the ignition point at which com-
In the process of steam generation, plete burnout of the combustibles takes
fuel-burning systems provide con- place. Thus, every firing system requires a
trolled, efficient conversion of the quantity of air in excess of stoichiometry
chemical energy of fuel into heat to attain an acceptable level of unburned
energy, which in turn is transferred carbon in the products of combustion
to the heat-absorbing surfaces of the leaving the furnace. This amount of
steam generator. To do this, fuel-burn- excess air is an indicator of the burning
ing systems introduce the fuel and air efficiency of the firing system.
for combustion, mix these reactants,
ignite the combustible mixture, and The Combustion Reaction
distribute the flame envelope and the The rate and degree of completion
products of combustion. of a chemical reaction, such as in
Combustion Technology | 2-
Figure 2-9 | Selective furnace utilization and steam temperature control are accomplished by tilting nozzles in a tangentially fired system
1,400 2-15 is rate determining as a result of its mixed flame fronts (Fig. 2-12).
1,200 high activation energy of 317 kJ/mol. A
1,000
better understanding of thermal NOx Fuel NOx
800
600
can be derived from bench-scale tests Although the kinetics involved in
400 that measure NOx in a heated mixture the conversion of organically bound
200 of N2, O2, and argon.3 These tests show nitrogen compounds found in fos-
that thermal NOx can be predicted by sil fuels are not yet well understood,
the following equation: numerous investigators have shown fuel
80 NOx to be an important mechanism
% Conversion to NOx
1/2
d NO = K1e ( − K2 /T ) N 2 O2 dt in NOx formation from fuel oil as well
70
as the dominant mechanism in NOx
Equation 2-18
60 generated from the combustion of coal.
where Bench-scale tests (Fig. 2-13) burning
50 fuel oils in a mixture of oxygen and
[…] = mole fraction
carbon dioxide (to exclude thermal
T = temperature
0 0.4 0.8 1.2 1.6 2.0 NOx) have shown a remarkable corre-
% Fuel Nitrogen
t = time
lation between the percentage N in the
K1, K2 = constants
fuel oil versus NOx.6 Figure 2-13 also
Figure 2-13 | Fuel NO and percentage From this equation, it can be seen illustrates that the percentage of fuel-
conversion of fuel nitrogen to that thermal NO can be decreased nitrogen conversion is not constant
fuel NO: liquid fuels
Clean Combustion Technologies by reducing time, temperature, and but, rather, decreases with increasing
Figure: 02.13/ SPS concentrations of N2 and O2. The fact fuel nitrogen.
PMS5415 K
that temperature in this equation is an Similar bench-scale tests with various
exponential clearly demonstrates its coals have not produced similar results.
Combustion Technology | 2-15
unstable and will reduce to harmless outlet NOx levels for liquid fuels 400
N2 under fuel-rich conditions or to with various nitrogen levels. The tests 300
NO under air-rich conditions. Low- shown in the figure used air for an
200
NOx, pulverized-coal firing systems oxidant, and the total NOx levels
are primarily designed to minimize represent both thermal and fuel NOx. 100
0.4 0.6 0.8 1.0
volatile nitrogen conversion to NOx However, the increased effectiveness First-Stage Stoichiometric Ratio
by establishing early ignition and of staging with higher fuel-nitrogen
controlling O2 availability by using content is clear. Figure 2-16 shows Figure 2-16 | Effect of initial fuel–air mixing:
air- and fuel-staging techniques. a similar but less dramatic effect on staged, Utah bituminous coal
In air staging, a portion of the total a Utah bituminous coal. This and Clean Combustion Technologies
air required to complete combustion other tests have shown that for coal, Figure: 02.16/ SPS
PMS5415 K
is withheld initially, and the balance of unlike oil, decreasing the first-stage
the air is mixed with the incomplete stoichiometric air below an optimum
products of combustion only after the level will increase NOx levels, which
oxygen content of the first-stage air suggests the possible requirement of
is consumed. By varying the amount an oxidant to reduce the intermediate
of first-stage air, the suspension-fired HCN, NH3, and other radicals.
combustion of a coal particle or oil The most important design criteria
droplet can be interrupted at different relevant to the control of both thermal
2-16 | Clean Combustion Technologies
and fuel NOx for coal firing can be nitrogen conversion by driving the
summarized as follows: major fraction of the fuel-nitrogen com-
pounds into the gas phase under overall
• Coals with the lowest fuel-nitrogen
fuel-rich conditions. In this atmosphere
content and the lowest fuel oxy-
of oxygen deficiency, there occurs a
gen/nitrogen ratios generally will
maximum rate of decay of the evolved
produce the lowest NOx.
intermediate nitrogen compounds to
• The fuel NOx can be minimized
N2. Following the admission of the
by controlling the quantity of air
remaining air, the slow burning rate
permitted to mix with the fuel in
reduces the peak flame temperature to
the early stages of combustion.
curtail the thermal NOx production in
• The contribution of thermal NOx
the latter stages of combustion.
to total NOx can be reduced by
Early studies of NOx emissions from
operating at the lowest practical
all types of steam generators indicated
excess-air percentages as well as
that those from tangentially fired units
by minimizing gas temperatures
were about half those from horizontally
throughout the furnace by using
fired systems. Tangentially fired steam
low-turbulence diffusion flames
generators inject the fuel and air streams
and large, water-cooled furnaces.
from the furnace corners, tangential to
an imaginary circle in the center of the
Firing Systems that Minimize furnace. The impingement of laterally
NOx Formation adjacent streams creates a single, rotat-
Firing systems designed to minimize ing flame envelope that promotes bulk
NOx formation operate to control mixing for complete combustion. The
long diffusion flames emanating from
each corner, plus the large amount of in-
ternal gas recirculation generated by the
cyclonic fireball, moderate fuel and air
Overfire Air mixing and form the basis of an inher-
Nozzles ently low NOx combustion system.
Windbox
Low-NOx, Tangential Firing
Side Ignitor
with Overfire Air
As shown in Figure 2-17, the first
Secondary
modification added air compartments
Secondary Air Nozzles within the windbox above the up-
Air
permost coal nozzle. Termed overfire
Dampers
air (OFA) ports, these compartments
divert approximately 20% of the total
Coal
Nozzles combustion air to a burning zone
above the windboxes. As a result, the
fireball at windbox level is at or near
Oil Gun stoichiometric air conditions.
Damper
Drive Units Figure 2-18 shows the effect on NOx
production of varying quantities of OFA.
Note the effect of total excess air on
NOx levels. This results primarily from
the reduction in available oxygen within
Figure 2-17 | Tangentially firing windbox with overfire air ports
Clean Combustion Technologies
Figure: 02.17/ (mh)
Combustion Technology | 2-17
0 % Excess Air
27
24
10 20
16
% Reduction NOx
Separated
20 Overfire
Air Nozzles
30
Close-Coupled
40 Windbox Overfire Air Nozzles
0 20 40 60 80 100
Side Ignitor
Overfire Air Damper, % Open
CFS™
Figure 2-18 | Effect of overfire air on NOx Secondary
production Air
Dampers
Clean Combustion Technologies Low NOx
Figure: 02.18/ SPS Coal Nozzles
the fireball adjacent
PMS5415 K to the windboxes.
With a constant windbox-to-furnace
differential, reducing total excess air
Oil Gun
increases the proportion of OFA at Damper
any given OFA damper position. This Drive Units
decreases the available oxygen below the
OFA ports in even greater proportion to
the change in total excess air.
Global and local staging techniques
have progressively developed and are Figure 2-19 | Windbox for low-NOx, tangentially fired system (CFS™, Concentric Firing System)
Without LNCFS™
400 air entrainment into the expanding the use of post-combustion treatment.
fuel/primary air streams, and it provides The combination of in-furnace control
300
more favorable conditions for low NOx and post-combustion treatment allows
With LNCFS™
200
in the pre-fireball flames (Fig. 2-23). the design engineer the opportunity
Furnace aerodynamics are such that the to meet emissions requirements at a
2.0 3.0 4.0 5.0 6.0
stream separation is short-lived. The lower cost.
Excess Air (% O2)
concept remains advantageous, however,
because it increases oxygen concentra- Stokers
Figure 2-23 | NOx vs. excess air with
and without the Low-NOx tion along the furnace waterwalls in Stokers are firing systems that feed
Concentric Firing System the firing zone. In 1983, offset air was and burn solid fuels in a bed at the
Clean Combustion
(LNCFS™) Technologies
Figure: 02.23/ SPS commercialized as the Concentric Firing bottom of a furnace. In all cases, the
PMS5415 K System (CFS™). Units retrofitted with fuel is burned on some form of grate,
Combustion Technology | 2-19
through which some or all of the grate rotates around to its return run.
air for combustion passes. The grate Traveling-grate stokers can burn every
surface can be stationary or moving. type of coal that is mined, with the
Stokers are classified according to exception of caking bituminous coal.
the way fuel is fed to the grate; the In addition, by-product and waste
three general classes in use today are fuels, such as coke breeze or anthra-
underfeed stokers, overfeed stokers, cite dredged from river bottoms, can
and spreader stokers. be burned effectively. This machine
has also been used as a part of chemi-
Underfeed Stokers cal process operations that produce
In an underfeed stoker, the incoming coke and CO2. Steaming rates ranging
fuel (usually coal) is pushed through from 10,000 to 300,000 lb/h are
one or more troughs, called retorts, achievable.
that are located below the burning
fuel bed and air-emission grates. Here,
Active Combustion Zone
moisture is driven off, and volatile
constituents are distilled, passing up Coking Zone
through the incandescent bed. As the
retorts are recharged, the char from Ash Zone Ash Zone
the previous fill is pushed into and
becomes part of the burning bed. On
both sides of the retort, the “overflow”
burns to completion on air-admitting Auxiliary Auxiliary
Air Main Air Air
grates, or tuyeres. Ash is removed by Chamber
Chamber Chamber
dumping grate sections (Fig. 2-24). Dumping Dumping
Boilers with underfeed stokers of the Grate Grate
single-retort design have been built
in capacities ranging from 5,000 to
50,000 lb/h steam flow. Multiple-
retort stoker units have been built
for capacities ranging from 40,000 to
300,000 lb/h steam flow. Figure 2-24 | Cross-section through underfeed stoker showing stages of combustion
Overfeed Stokers
In an overfeed or traveling-grate Idler
CleanShaft Drive Shaft
Combustion Technologies
stoker, gravity feeds the fuel (again, Figure: 02.24/ (ks)
primarily coal) at one end of the air- PMS5415 K
admitting grate surface (Fig. 2-25).
The incoming bed depth is adjusted
by a gate under which the coal passes
before entering the furnace. The rotat-
ing grate surface moves away from
the feed end through the furnace;
combustion is completed as air passes
up through the grate. The ash residue
is continuously discharged as the
Figure 2-25 | Arrangement of traveling-grate stoker
2-20 | Clean Combustion Technologies
For more detailed discussions about loadings for spreader stokers are higher
the designs and performance of these than those for previously described
first two classes of stokers, the reader is designs, in which the fuel quiescently
referred to the first edition of Combus- enters the burning zone. Practically all
tion Engineering: A Reference Book on types of coal (except anthracite) and a
Fuel Burning and Steam Generation.13 wide variety of cellulosic fuels, including
wood wastes, bagasse, furfural residue
Spreader Stokers sludge, rice hulls, and coffee grounds,
The third major class of stokers—and have been successfully burned on
the machine of choice in modern spreader stokers.
steam generation applications—is the The most popular type of spreader
spreader stoker (Fig. 2-26). This class stoker is that incorporating a continu-
combines the principles of suspension ous ash discharge grate (Fig. 2-26).
burning and thin-bed combustion. This machine, unlike the traveling-
Feeder/distributor devices continuous- grate stoker, moves toward the fuel
ly project fuel into the furnace above feeders, which are throwing new fuel
an ignited fuel bed on the grate. Fines toward the back of the unit. All the
are burned in suspension, while larger fuel is burned before reaching the
particles fall and burn on the grate. front end, from which ash is continu-
The spreader stoker method of fuel ously dumped. The return side of
firing provides quick response to chang- the grate carries siftings, which fall
es in boiler demand, and it generally through the top side of the grate to
never has more than a few minutes of the rear and discharges them into
fuel inventory on the grate. Flash drying a hopper. Grate speed is regulated
and rapid release of tarry hydrocarbons to maintain an ash-bed depth of 2
enable this system to fire caking-type to 4 inches at the discharge. Typical
coals without concern for matting or operating speed ranges from 2 to 20
clinkering. The air-cooled, non-agitated ft/h and is varied via gear or hydraulic
ash bed forms few clinkers despite low- drive units.
fusion-temperature fuels. Particulate
Coal Size Specifications for
Spreader-Stoker Firing
Ignition-System Design
IGNITOR FUNCTIONAL DEFINITIONS
An electrical spark is the typical
To fully understand the hardware, it may be controlled conditions to limit the potential for initiating source of fuel ignition. High-
necessary to review the definitions contained abnormal operation, as well as to limit the voltage, free-air sparks can reliably
in NFPA 85E, “Standard for Prevention of charge of fuel to the furnace in the event that
Furnace Explosions in Pulverized Coal-Fired ignition does not occur during light-off. Class 2
ignite high-calorific-value gases and
Multiple Burner Boiler-Furnaces, 1985 ignitors may be operated as Class 3 ignitors.16 light (distillate) oils. High-energy,
Edition.”14 surface-shunted arcs are used to
Class 3 (Interrupted Ignitor) reliably ignite distillate and heavy
Class 1 (Continuous Ignitor) A small ignitor applied particularly to gas
An ignitor applied to ignite the fuel input and oil burners to ignite the fuel input to the
(residual) oils. Safety considerations
through the burner and to support ignition burner under prescribed light-off conditions. dictate that because of its relatively low
under any burner light-off or operating The capacity of such ignitors generally does ignition energy, only a limited quantity
conditions. Its location and capacity are such not exceed 4% of full load burner fuel input. of fuel be exposed to the electrical spark.
that it will provide sufficient ignition energy As a part of the burner light-off procedure,
The heat released from this initial fuel
(generally in excess of 10% of full load burner the ignitor is turned off when the timed trial
input) at its associated burner to raise any for ignition of the main burner has expired. input then becomes the ignition energy
credible combination of burner inputs of both This is to ensure that the main flame is for a different and/or larger fuel input.
fuel and air above the minimum ignition self-supporting, is stable, and is not depen- Essentially, the ignition system provides
temperature.15 dent upon ignition support from the ignitor.
a controlled transition from spark to
The use of such ignitors to support ignition
Class 2 (Intermittent Ignitor) or to extend the burner control range shall main-fuel firing through an incremental
An ignitor applied to ignite the fuel input be prohibited.17 increase in the ignition energy.
through the burner under prescribed light-off Not making too large of a step-
conditions. It is also used to support ignition Class 3 Special
increase in fuel input is also important
under low load or certain adverse operating (Direct Electric Ignitor)
conditions. The range of capacity of such ignit- A special Class 3 high-energy electrical to avoid undesirable furnace pressur-
ors is generally 4 to 10% of the full load burner ignitor capable of directly igniting the main ization during light-off. Figure 2-28
fuel input. It shall not be used to ignite main burner fuel. This type ignitor shall not be shows two methods of achieving a
fuel under uncontrolled or abnormal condi- used unless supervision of the individual
step-increase in fuel input for tangen-
tions. The burner shall be operated under main burner flame is provided.18
tial firing. In the example shown on the
left of Figure 2-28, the lowest elevation
of ignitors is fired first, and after a suit-
ignition energy at one instant may able period of time, the elevation of oil
Auxiliary-Air be insufficient the next. From the warm-up guns is ignited. As the fur-
Compartment rate-igniting standpoint, there is nace temperature stabilizes, the main
never excess ignition energy: The coal nozzles adjacent to the oil guns are
Coal
more stable a fire, the more likely the brought into operation. Thereafter, ad-
Compartment ignition energy substantially exceeds ditional ignitors can come into service
the minimum necessary to maintain to light off elevations of main coal
Warm-Up Oil
ignition. nozzles directly.
Compartment Years of experience with various In the alternative example shown on
fuels and firing systems have led the the right of Figure 2-28, ignitors light
Ignitor National Fire Protection Association off successive elevations of warm-up oil
(NFPA) to establish practical guide- guns. Adjacent elevations of main coal
Figure 2-28 | Alternative step ignition
lines for defining ignition-energy nozzles are then ignited by the oil guns.
processes for tangential firing
requirements based on fuel type and Experience indicates that firing
Clean Combustion Technologies ignition-system function. For NFPA by these steps/input methods will
Figure: 02.28/ (KA)
PMS5415 K definitions, see the accompanying box minimize furnace pressurization and
Ignitor Functional Definitions. unnecessary overpressure trips.
Combustion Technology | 2-23
logic and discriminating flame scanners The HEA ignitor can ignite
Rigid (or Flexible) Ignition would be required. (Refer to Chapter warm-up fuel oils ranging from
Spark Rod Exciter
7 for more information on flame scan- No. 2 to No. 6 and also crude oils.
Firing Spark
Tip Rod
ners.) A Class 3 application would be The ignitor is a self-contained electri-
Retractor made only when a limited amount of cal discharge device that produces a
ignition fuel is available, such as with a high-intensity spark. Use of a high-
bottled-gas supply system. resistance transformer (to produce a
full wave charging circuit and control
Guide Pipe Flexible Cable HEA Ignitors. The HEA ignitor was spark rate) enables the sealed power
developed specifically for residual oils. supply to store maximum energy
It eliminates dependence on premium and then deliver a greater percent-
Figure 2-30 | High Energy Arc (HEA) ignitor
fuels, such as natural gas and No. 2 fuel age of this energy through insulated
oil for main-fuel ignition and boiler cables to the ignitor tip. A high spark
Clean Combustion Technologies warm-up. The HEA ignitor directly energy also eliminates coking of the
Figure:
(A) Flame 02.30/JT
Pocket is Created on Ignitor Tip
PMS5415 K
spark-ignites a heavy-oil warm-up gun, ignitor tip. The HEA ignitor consists
which in turn ignites the main coal of four basic components: an exciter,
nozzles. The HEA ignitor is used with a flexible cable, a spark tube and
a discriminating scanner that verifies guide pipe, and a retractor assembly
the operation of the warm-up gun. The (Fig. 2-30).
warm-up gun is sized for an oil flow of A key to the successful application
about 10% of the energy capacity of an of spark-ignition is the presence of
(B) Flame Pocket Becomes Entrained in
Recirculation Zone and Expands Towards
adjacent coal nozzle, with the discrimi- a strong recirculation pattern in the
Oil-Gun Tip nating scanner proving the presence of primary combustion zone (Fig. 2-31).
a flame. The combination of an HEA The recirculation provides the energy
ignitor and a No. 6 fuel oil-fired warm- required to vaporize and heat the oil
up gun satisfies the NFPA definition of to its ignition point, thus maintain-
a Class 1 or Class 2 ignitor. The HEA ing stable ignition after the spark has
ignitor alone would be a Class 3 ignitor. been deactivated.
The complete HEA ignition system The discriminating flame scanners
(C) Flame Pocket Reaches Oil-Gun Tip consists of: operate in the visible-light range,
and Stabilizes
which provides excellent sensitiv-
• An HEA ignitor
ity to the presence of a flame. These
• A warm-up oil compartment
scanners therefore can discriminate
capable of producing a stable flame
between different flames based on
at all loads
both frequency and intensity. This
• A flame-detecting system sensitive
insures safe operation of the overall
only to its associated oil gun
HEA system.
Figure 2-31 | High Energy Arc (HEA) ignition • A control system coordinating all
sequence the components and providing Doug Hart
Clean Combustion Technologies unit safety Bob KunkEL
Figure: 02.31abc/ (KA)
PMS5415 K
REFERENCES
1. “Standard Atmosphere Tables and Data for Altitudes to 65,800 feet,” 9. S. L. Chen, M. P. Heap, D. W. Pershing, and G. B. Martin, “Influence of
National Advisory Committee for Aeronautics Report 1235 Internation- Coal Composition on the Fate of Volatile and Char Nitrogen During
al Civil Aviation Organization (ICAO), Montreal, Canada, and Langley Combustion,” 19th Symposium (International) on Combustion, The
Aeronautical Laboratory, Langley Field, VA. 1952. Combustion Institute, 1982, p. 1271.
2. Ya. B. Zel’Dovich, “The Oxidation of Nitrogen in Combustion and 10. Ibid, p. 1271.
Explosions,” Acta Physicochimica (USSR), 21: 577–628, 1946.
11. M. S. McCartney and J. Grusha, “Development and Evolution of the
3. D. J. MacKinnon, “Nitric Oxide Formation at High Temperature,” Air ABB Combustion Engineering Low NOx Concentric Firing System,”
Pollution Control Association Journal, 24: 237–239, 1974. 1991 Joint Symposium on Stationary Combustion NOx Control,
Washington, D.C., 1991, Vol. 1.
4. C. P. Fenimore, “Formation of Nitric Oxide from Fuel Nitrogen in Ethyl-
ene Flames,” Combustion and Flame, 19: 289–296, 1972. 12. R. A. Buchs, L. K. Bailey, J. V. Dallen,
������������������������������������������
T. D. Hellewell, and C. W. Smith,
“Results from a Commercial Installation of Low NOx Concentric
5. G. De Soete, “Mechanism of Nitric Oxide Formation from Ammonia
Firing System (LNCFS™),” International Joint Power Generation Con-
and Amines in Hydrocarbon Flames,” Revue de L-Institut Français du
ference, Boston, MA, 1990.
Petrole et Annuales Des Combustible Liquides, 28: 95–108, 1973.
13. Otto De Lorenzi, ed., “Combustion Engineering: A Reference Book on
6. W. W. Habelt, “The Influence of Coal Oxygen to Coal Nitrogen Ratio on
Fuel Burning and Steam Generation,” First Edition, New York: Com-
NOx Formation,” Presented at the 70th Annual American Institute of
bustion Engineering-Superheater, Inc., 1947.
Chemical Engineers Meeting, New York, November 13–17, 1977.
14. NFPA 85E, “Standard for Prevention of Furnace Explosions in Pulverized
7. D. W. Pershing and J. O. L. Wendt, “Pulverized Coal Combustion: The
Coal-Fired Multiple Burner Boiler-Furnaces, 1985 Edition,” p. 85E-9.
Influence of Flame Temperature and Coal Composition on Thermal
and Fuel NOx,” 16th Symposium (International) on Combustion, The 15. Ibid, p. 85E-9.
Combustion Institute, 1976, p. 389
16. Ibid, p. 85E-9.
8. J. D. Freihaut, W. M. Proscia,
�����������������������������������������������
and J. D. Seery, “Fuel-Bound Nitrogen
17. Ibid, p. 85E-9.
Evolution During Devolitilization and Pyrolysis of Coals of Varying
Rank,” Joint Symposium on Stationary Combustion NOx Sources, New 18. Ibid, p. 85E-9.
Orleans, 1987.
2-25
Boilers
Chapter Three
Among the ash properties described • The ash fusibility temperatures (both
in Appendix B, those that Alstom in terms of their absolute values and
considers to be particularly important in terms of the spread or difference
when designing and establishing the between initial deformation tempera-
size of coal-fired furnaces include: ture and fluid temperature)
3- | Clean Combustion Technologies
• The ratio of basic to acidic ash general, high-sodium lignite also tends
constituents to have high slagging tendencies.
• The iron/calcium ratio Among the most important design
• The fuel ash content in terms of lb criteria in large, pulverized-fuel,
ash/million BTU (kg ash/million J) tangentially fired furnaces are the net
• The ash friability heat input in BTU/h-ft² (W/m²)
over furnace plan area (NHI/PA) and
These characteristics translate into the vertical distance from the top fuel
the typical furnace sizes shown in nozzle to the furnace arch in ft (m).
Figure 3-2, which are based on the six Furnace dimensions must be adequate
coal ranks shown in Table 3-1. This to establish the necessary furnace
size comparison illustrates the philoso- retention time to properly burn the
phy of increasing the furnace plan area, fuel as well as to cool the gaseous
volume, and the fuel burnout zone (the combustion products. This cooling
distance from the top fuel nozzle to the must be adequate to ensure that the
furnace arch) as lower-grade coals with gas temperature at the furnace outlet
poorer ash characteristics are fired. plane is well below the ash-softening
Figure 3-2 is a simplified character- temperature of the lowest-quality
ization of actual furnaces built to burn coal that is burned. Specifically, the
the fuels listed in Table 3-1. Wide vari- furnace outlet plane is the entrance to
ations exist in fuel properties within the close-spaced convection surface.
coal ranks as well as within several sub- The latter is defined as non-platen
classifications (e.g., sub-bituminous A, surface on less than 12 inch (300 mm)
B, and C), each of which may require horizontal centers. In addition to the
a different size of furnace. Note that in tilting of fuel and air nozzles, the heat
absorption characteristics of the walls
are maintained using properly placed
Plan Plan Plan Plan Plan wall blowers to control the furnace
Area W Area 1.08 W Area 1.16 W Area 1.26 W Area 1.29 W
= 1.25 = 1.56 = 1.63 outlet gas temperature by removing ash
= 1.00 = 1.15
D 1.06 D 1.08 D 1.24 D 1.26 D that has been deposited on the furnace
walls below the furnace outlet plane.
To arrive at an appropriate value
of NHI/PA and furnace plan area,
a thorough evaluation of all fuels
expected to be burned in the unit is
2.1 h
required. Depending on the analyses
1.52 h
h 1.17 h
and range of coals to be burned, large,
1.15 h
pulverized fuel-burning units have an
1.30 H 1.45 H NHI/PA that generally varies from
H 1.05 H 1.07 H
1.4 to 2.0 × 106 BTU/h-ft2 (4.4 to 6.3
× 106 W/m²). The distance from the
Medium- High- Low- Medium- High- upper fuel nozzle to the furnace arch is
Volatile Volatile Slagging Slagging Slagging a function of furnace width and depth,
Bituminous Bituminous Lignite Lignite Lignite
or
the upper furnace superheater and
Sub-Bituminous reheater surface arrangement, and of
course, the fuel and ash characteristics.
Figure 3-2 | Effect of coal rank on sizing of a pulverized-fuel furnace (constant heat output)
A widely spaced, steam-cooled, platen- 100 MWe, the furnace volumetric heat
type heating surface is usually required release rate in BTU/h-ft³ (W/m³)
in the upper furnace to provide further generally is not a controlling design pa-
cooling of furnace gases before those rameter. For such units, the release rate
gases enter the convection surface. range is from 15,000 BTU/h-ft³ for
These sections potentially are subject very good bituminous coals to approxi-
to the most severe fouling conditions, mately 10,000 BTU/h-ft³ for lignite
and wide spacing helps to prevent (from 150,000 to ~100,000 W/m³).
ash plugging caused by bridging. To Similarly, the burner zone, volumetric
perform satisfactorily in this environ- heat release rate in BTU/h-ft³ (W/m³)
ment, the platens are constructed with is not a meaningful design parameter
the tubes at a tangent to each other in for tangentially fired boilers.
the direction of gas flow to minimize
ash deposition. DRY-BOTTOM VERSUS
A furnace design parameter known WET-BOTTOM FURNACES
as “burner zone heat release” is some- All current pulverized-coal reheat
times specified. This term does not units are of the dry-bottom type. The
have the same significance in tangen- ash dislodged from the furnace walls is
tial firing as it does in other firing below the ash melting point and leaves
systems. In tangential firing, not all of the furnace bottom in a substantially
the fuel is consumed in the windbox dry condition. The wet-bottom, or
zone because of the manner in which slagging-bottom, furnaces are offered
the fuel and air are introduced into today only for very special applications,
the furnace. In a unit designed for wall such as in furnaces for gasifying coal.
firing, the fuel and air are intimately In the wet-bottom design, the lower
mixed at the burner throat and are part of the flame has to sweep the fur-
consumed directly in front of the nace floor at all loads to maintain the
burner. A wall-fired boiler may have its fluidity of the ash. This requirement
burners moved farther apart, or more imposes a definite limitation in the use
burners can be added, to give any value of slag-tap furnaces. Slag-tap freez-
of burner zone release rate, because ing can occur during operation at low
little interaction occurs between rows load or when fuel is fired only in the
of burners. The vertical distance from upper furnace for steam temperature
the lowest coal nozzle to the point control. In addition, wet-bottom units
where the furnace walls are bent to have higher nitrogen oxide production.
form the hopper section of the furnace The dry-bottom furnace, on the other
is a function of the furnace width and hand, and particularly those designs
depth and of the slagging potential using tilting fuel nozzles, can provide a
of the fuel. This dimension generally wider control range for steam tempera-
is between 12 and 20 ft (between ture and can handle coals with widely
3.7 and 6 m) in a tangentially fired varying ash characteristics. This latter
furnace. The hopper section typically characteristic, more than any other,
slopes at 50° to the horizontal. originally was responsible for the great-
With the exception of dry-bottom, er application of dry-bottom units.
pulverized coal-fired units with sizes In the typical dry-bottom furnace,
of between approximately 50 and bending the front- and rear-wall tubes
3- | Clean Combustion Technologies
surfaces can fall into the hopper. The of coal in the same unit. However, such
next heating surface has little risk a unit would not usually be economically Coal ash properties have
of blockage. feasible. For example, consider the wide a major influence on the
variation in pulverizer size and primary
Convection Pass Design for Two-Pass design of any coal-fired boiler.
air temperature requirements for a high-
and Tower Boilers calorific-value, low-moisture, low-ash Thorough consideration of these
A number of coal ash properties have coal versus a sub-bituminous coal or properties is the key to successful
a significant influence on superheater, lignite with high moisture and ash con- performance and operation.
reheater, and economizer convection tents. Selecting an optimum preparation
surface design and on the associated and firing system for one rank of fuel can
considerations of fouling and erosion. result in a much less favorable situation
These include: for some other rank of fuel.
Operating experience on units firing
• Ash-softening temperature
certain coals shows the need for wide
• Base/acid ratio
transverse tube spacing throughout the
• Iron/calcium ratio
unit to reduce the fouling rate and pos-
• Sodium and potassium content
sible bridging of ash deposits between
• Silica and alumina content
adjacent assemblies. This arrangement
• Chlorine content
minimizes serious fouling problems,
• Ash friability
which have an adverse effect on fan
In sizing and locating these sections power requirements and unit avail-
within the unit, a proper balance is ability. Because the gas temperature
required to maintain a thermal head is reduced along its flow path to well
with which to transfer heat from gas below the ash-softening temperature
to steam while optimizing heating of the worst coal that is expected to be
surface use and avoiding undesirably burned, the transverse spacing of the
high metal temperatures. Also, to banks of convection pass tubes also is
limit erosion of pressure parts from reduced, depending on the fuel type
fly ash, the flue gas velocity must not and the ash characteristics. The tube
exceed reasonable limits. Depending spacing selected at the entrance to the
on the ash quantity and abrasiveness, convection pass depends mainly on
the design velocity generally is 40 to the maximum gas temperature that
60 ft/s (12 to 18 m/s). A boiler that is possible in this zone under upset
burns coal yielding a heavy loading conditions and the propensity of the
of erosive ash (usually indicated by a ash to adhere.
high silica/alumina content) would In the superheater and reheater
use the lower velocities. Such velocities sections beyond the ash adhesion
are based on the predicted average gas temperature zone, the tube spacing
temperature entering the tube section is further reduced to the minimum
at the maximum continuous rating of required for effective penetration of
the steam generator fired at normal soot blower jets. In addition, gas veloci-
percentage of excess air. ties are maintained sufficiently low to
Considering these factors in furnace prevent erosion. To avoid serious plug-
sizing and convection pass design, it ging, the transverse spacing in this area
is not impractical to propose a steam must be wide enough to allow pieces of
generator capable of burning any kind accumulated ash that are loosened by
3-10 | Clean Combustion Technologies
soot blowers to pass through the tube plant will use a significant amount
bank and not bridge the span between of the “worst” coal. Finally, if coals
adjacent rows of tubes. Table 3-2 is a are to be blended, information must
guide for establishing transverse clear- be provided on both the separate
ances for platen and spaced convection analyses and the possible blends,
surfaces. This guide is based on fouling including values for eutectic fusibility
and non-fouling coals and on gas tem- temperatures.
peratures at maximum boiler load.
Design of Large, HP Furnace
Wall Systems
table 3-2 | TRANSVERSE CLEARANCE GUIDE
The design of various heat-absorbing
Temperature Range Inches (cm) of Clearance equipment involves a finite pressure
°F °C Non-fouling Coal Fouling Coal drop or resistance to fluid flow. Air
2,000–2,400 1,093–1,316 Platen–22 (56) Platen–22 (56) heaters, economizers, superheaters,
1,750–2,000 954–1,093 Spaced–7 (18) Spaced–12 (30) and reheaters all have a multiplic-
1,450–1,750 788–954 Spaced–3 (8) Spaced–6 (15) ity of parallel flow paths or circuits
through which the fluid being heated
Up to 1,450 Up to 788 Spaced–2 (5) Spaced–3 (8)
must pass. The uniformity of the flow
distribution largely depends on the
Coal Properties Used for Design magnitude of the pressure drop across
Properties of the coal to be used in a the apparatus. The water-cooled fur-
boiler design are established by the nace generally handles the fluid it is
purchaser’s specifications. It is prefera- heating in many more parallel circuits
ble to show an analysis that represents than are found in other heat-transfer
an actual coal as the performance devices. Therefore, the problem of
coal and to include low, average, and distributing water to each tube, and
high analyses to confirm limits for the providing adequate circulation for
boiler design. Important items of the cooling all the tubes, is correspond-
analysis are calorific value, moisture ingly more difficult.
and ash contents, grindability index, The tubes that form the furnace
and the ash analysis. If the fuel walls are of different lengths and are
specifications are stated as ranges, it is subject to varying rates of heat absorp-
customary to consider the average as tion. Great care must be exercised
the performance fuel but to design the to ensure that each tube receives a
unit to handle fuels with the worst sufficient flow of water to prevent the
analysis. If coals from two or more overheating of individual waterwall
sources may be fired, individual analy- tubes as well as an excessive differ-
ses rather than a calculated average are ence in temperature between adjacent
desirable, with “low” and “high” repre- tubes. Flow instability, which can
senting the composite. Justification for occur in furnace circuits, is consider-
using individual analyses comes from ably more difficult to control when
the necessity to evaluate pulverizer a large number of tubes operate in
capabilities as well as ash influences parallel. The designer must calculate
on unit performance. The “worst” coal the heat absorption per furnace tube
can become the design coal when a from a forecast of the normal absorp-
good probability exists that the power tion pattern for the entire furnace.
Boilers | 3-11
from dependence on the difference in economizer. Figure 3-7 shows how the
steam and water densities to provide furnace wall operating pressure and the
circulating head. Using a recirculat- percentage steam in the mixture in the
ing pump means that the designer is wall tube affect the available thermal
ensured of positive circulation even be- head for circulation. The available head
fore heat is applied. Because the pump is the difference between the weight of
helps to overcome friction loss in the water in the unheated downcomer and
waterwall tubes, it is possible to sub- the static head of the steam–water
stitute smaller-diameter tubes for the mixture in the heated waterwall tubes.
larger-diameter tubes that are required Figure 3-8 shows a typical Alstom
with a limited thermal circulation thermal circulation, reheat steam
head. The net result is a lower-tube- generator for burning pulverized sub-
weight, lower-metal-temperature wall bituminous coal for low to medium
with lower thermal stresses. subcritical pressures.
Another advantage to incorporat- Tangential fuel nozzles are located
ing a recirculating pump is the greater at five elevations in the four corners of
flexibility of boiler layout. More freedom the boiler furnace. Vertical downcom-
exists in the arrangement of boiler ers, connecting the steam drum to the
heating surface, because with ensured bottom waterwall header system, sup-
positive circulation, a horizontal ply the boiler water. The lower ends
evaporation surface may be used to any of the front, rear, and side wall tubes
extent desired. The Alstom system for connect to the bottom waterwall
controlled, forced recirculation at the headers, and the upper ends connect
high-subcritical-pressure level is called a to the upper headers. Relief (riser)
Controlled Circulation® system. A 2,850 tubes transport the mixture of steam
psig (19.7 MPa gage) drum pressure and water from the waterwalls to the
limit has been established for Controlled drum, where the steam is separated
Circulation® steam generators. This from the boiler water.
limit permits operation of the 2,400 psig To ensure uniform distribution of
(16.6 MPa gage) throttle pressure cycle steam and water, the connecting tubes
at 5% overpressure, or 2,520 psig (17.4 are spaced evenly along the drum and
MPa gage) at the turbine, and as high as
2,680 psig (18.5 MPa gage) at the boiler
superheater outlet.
Operating Pressure
2,000 psig
30
Thermal Circulation Furnace Design. 2,200 psig
Net Available Static Head, psi
2,400 psig
Thermal circulation results from a 2,600 psig
density difference between the mixture 2,800 psig
20
3,000 psig
of saturated water and feedwater in
the downcomers (downtakes) and the
lighter steam–water mixture in the 10
furnace wall tubes. The temperature
and corresponding density in the
downtakes can be changed somewhat 0
0 20 40 60 80 100
in the initial design of a unit by vary- % Steam (mass basis) Leaving Top of Furnace
ing the water temperature leaving the
Figure 3-7 | Typical available static head for circulation based on a 200-ft high furnace; this
available head is independent of the type of circulation
Clean Combustion Technologies
Figure: 03.07/ SPS
PMS5415 K
3-14 | Clean Combustion Technologies
length, arrangement, and heat absorp- and flow resistances can be made at this
tion. Strainers or screens between the time to increase or decrease the average
circulating pump discharge and the circulation and so achieve satisfactory
orifices prevent large particles of foreign cooling of the furnace-wall.
material from plugging the orifices at Tubes in the furnace enclosure
the entrance to the generating circuits. have different lengths and dissimilar
The strainers are particularly valuable configurations, and they receive varying
during the initial start-up of a new unit amounts of heat. Any tube absorbing
because, during this period of opera- more heat than the average tube has a
tion, foreign material is brought into the higher circulating water flow but prob-
boiler from preboiler-cycle, feedwater ably also a high percentage of steam
piping and auxiliary equipment. leaving. Typically, a 50% increase in
absorption above the average, equally
Design Of A Controlled Circulation® received throughout the length of a
Furnace Wall System. The orificing tube, produces a 10 to 20% increase in
and selection of the pump head for water flow entering the tube. For every
a Controlled Circulation® boiler are tube receiving more than the average
best understood by first reviewing a tube, another tube is receiving less
non-pump-assisted, thermal circulation water. Thus, each tube must be ana-
unit design. The designer of a thermal lyzed on the basis of its individual heat
circulation furnace of a given plan area absorption to arrive at its own mass
determines the total number of parallel percentage of steam (quality) leaving.
vertical tubes making up the outer walls Realize that in designing a waterwall
of the furnace. A common selection is to section (i.e., a group of parallel tubes
use 2½-inch (6.4 cm) (outer diameter) having a common [bottom] inlet header
tubes on 3-inch (7.6 cm) centers. The and a common upper [outlet] header),
next step, based on experience with such as that shown in Figure 3-11, the
units of similar physical size and heat pressure drop or head loss between
duty, is to establish the size and num- the headers must be identical for every
ber of unheated downcomers from the tube. This situation exists irrespective
steam drum to the lower waterwall of any pump assistance. The flow in
header system. Along with this, it is each tube and the percentage steam in
necessary to establish the arrangement, the mixture leaving differ, depending
diameter, and number of relief tubes on the following factors:
connecting the upper waterwall headers
Upper to the drum. An average is then calcu- • Heat absorption (overall and local)
Header lated for the overall percentage of steam • Effective inside diameter of tubing
“Average”
leaving the furnace, based on the heat ab- (as affected by manufacturing toler-
Tube sorbed by the “average” furnace tube and ances)
A B C
on the flow resistances in that tube, the • Effective tube length (accounting
downcomers, the relief tubes, the drum for bends)
internals, and all associated entrance and • Amount of tubing “shaded” from
Lower
Header exit losses. Within the limits of available heat
thermal head and physical space, adjust- • Location relative to feeder tubes
ments in downcomer and relief tube size • Orifices (if used)
Figure 3-11 | Schematic of furnace-wall
section; head loss between
headers is identical for each tube
Boilers | 3-19
Tube A, for instance, may be longer that the circulating pumps move the
but have a lower overall heat absorp- maximum mass of water before firing
tion than the “average” tube B. Without commences, because the water is most
orificing, tube A will have a lower flow dense at that time. The lower curve
than tube B. Tube C, if shorter than shows that the designer of a Controlled
tube B but fully exposed for its entire Circulation® unit can preferentially
length to high heat, will have a higher increase the flow in tubes with less fa-
flow than the average tube (i.e., tube B). vorable configurations to provide more
All tubes, however, must have the same cooling at all loads. The orifice pressure
differential pressure from header to drop designed into Controlled Circula-
header. An equilibrium condition will tion® circuits is additive to the heated
always result in the flow-plus-orifice circuit resistance loss and static head.
losses being additive to the weight of Orificing individual tubes, or groups of
the saturated mixture in each tube at tubes, provides either adequate mass
any given moment. This is because the flow or a desired exit quality under
header-to-header differential consists a wide range of postulated operating
of the following components: conditions. The orifice pressure drop
has a dampening effect on the almost-
• The weight of the steam–water
random behavior of a purely thermal
mixture in each tube
circulation unit. Because the orifice
• The pressure drop caused by the
drop occurs at the beginning of a cir-
resistance to flow of the steam–
cuit, before heat has been added in the
water mixture in each tube
furnace, the principal variable is mass
• The orifice pressure drop (if any)
flow through the opening, with no ef-
The flow may or may not be enough fect of differential specific volume. This
to cool the tube properly, and it may is a very significant difference between
even reverse (from upward to down- pure forced circulation, in which most
ward, and vice versa) from time to of the pressure drop is in the heated
time, depending on the variables acting tube circuits themselves, and Alstom’s Total Furnace-Wall Flow
on it. In tubes where mass flow is lower selectively orificed, pump-assisted, Controlled
Average Mass Velocity, %
with higher mass flow. For any given Controlled Circulation® reheat boiler 0
0 100
overall steam content, water between designed to provide steam to a large, Heat Absorption, %
the lower and upper headers redistrib- high-subcritical-pressure, turbine genera- Flow in Least Favorable Tube
utes from tube to tube according to the tor. This is a single, open-furnace unit. It Controlled
above relationships. has a superheater with both pendant and Circulation®
Tube Mass Velocity, %
Figure 3-12 gives a comparison horizontal surfaces and a reheater with 100
between thermal circulation and Con- both radiant wall and pendant surfaces.
trolled Circulation® performance. In This boiler has superheater outlet condi- Thermal Circulation
the upper curve, average mass flow in tions of 2,620 psig and 1,005°F (18.1 0
0 100
furnace wall tubes is shown as a func- MPa gage and 541°C), with single reheat Heat Absorption, %
tion of design heat absorption. Note to 1,005°F (541°C). It is designed to
Figure 3-12 | Comparison of furnace-
Cleanwall performance:
Combustion thermal
Technologies
circulation
Figure: 03.12/ SPS versus Controlled
®
PMS5415Circulation
K systems
3-20 | Clean Combustion Technologies
Riser Tubes
Furnace Side
Economizer
Wall
Furnace Front
Wall Furnace Rear Wall
Downcomers
Economizer Inlet
Pulverizers
Figure 3-13 | Large Controlled Circulation®, tangentially firing, pulverized coal-fired reheat steam generator
80
There are many recorded instances in tube in its upper half because of
which an operator has kept a Con- soot blower cleaning or shedding,
60 trolled Circulation® unit in service which can result in a calculable
Area of through a peak period, or until a week- decrease in the circulation rate
40 Permissible
Operation end, when it became more convenient within an individual tube or group
to shut down. An additional capability, of tubes. Flow reversal can occur,
20
that of forced furnace wall cooling with with local overheating becoming a
0 600 1,500 2,850 the pumps, allows boiler cooldown possibility.
Pressure in Steam Drum, psig
with uniform stresses without using • Controlled Circulation® and
the feedwater pump or dumping feed- thermal circulation boilers have
Figure 3-14 | Maximum permissible capacity
as a function of operating water to the condenser. the same thermal circulating head
pressure in the drum for available if they have the same fur-
Clean Combustion ®
Technologies
Controlled Circulation boilers Operating Pressure Flexibility. The nace height, the same economizer
Figure: 03.14/ SPS
PMS5415 K curve shown in Figure 3-14 shows water leaving temperature, and the
that a Controlled Circulation® steam same circulation ratio (Fig. 3-7).
generator can be operated through a The pumps and orifices incorporat-
wide range of pressures at full load. ed in Controlled Circulation® units
Operation within the permissible area do not destroy the available ther-
allows great latitude in establishing a mal head or lessen the difference in
desirable sliding-pressure pattern. density between furnace wall tubes
and unheated downcomers.
Thermal Circulation Versus Controlled • To ensure adequate available head
Circulation® Systems. In choosing be- under all operations of firing and
tween thermally induced circulation and maneuvering the boiler, Alstom
a Controlled Circulation® system, the adds orifices so that all tubes will
following must be taken into account: receive water under all conditions,
and Alstom shrouds the thick-
• Although thermal circulation walled steam drum to cool or heat
does increase in response to heat it in a controlled manner under all
applied, the increase in flow is rates of load change. The circu-
not directly proportional to the lating pumps complement these
increase in heat flux. That is, engineered modifications to the
although a larger amount of water basic thermal circulation design,
usually circulates in an individual and they provide the pressure head
tube when more heat is applied, to overcome the incrementally
the increase may not be completely increased resistance. A most impor-
adequate to provide proper tube tant point is that increases in flow
protection. are created with pump assistance
• Slag can form heavily—and does that are much greater than those
form unpredictably—on different available through the thermosy-
areas of the furnace at different phonic phenomenon. This point
times. It can be removed by soot is illustrated in Figure 3-15 by
blowers or shed, equally unpre- comparing the pressure drops in
dictably, because of its mass. It is the various components of typical
Boilers | 3-23
Figure 3-15 | Comparison of component pressure drops and available static head: Thermal
Supercritical-Pressure circulation boiler versus Controlled Circulation® boiler
Steam Generators Clean Combustion Technologies
Figure: 03.15/ SPS
Alstom supercritical-pressure steam high-subcritical pressures and PMS5415
in the K
generators can be classified in three supercritical-pressure range. A later
basic design categories according to section of this chapter addresses the
their operating pressure regimes. In the subject of sliding (or “variable”) pres-
first, which includes units designed for sure operation in detail and covers the
constant-pressure operation, supercriti- significant advantages to the steam
cal pressures are maintained in both generator and turbine while operating
furnace walls and superheater over the in such a mode. The three designs to
normal operating range. The furnace meet user requirements for throttle
wall arrangement and the use of boiler pressures above the 3,800 psig (26
throttling valves are as described later MPa gage) level are described in the
under Typical Combined Circulation® following sections.
Unit. For units that are to have partial
sliding-pressure capability, supercritical Constant-Pressure Supercritical Units
pressures are maintained in the furnace The Combined Circulation® system is
walls. Only the superheater follows a a once-through steam generator with
sliding-pressure program. The boiler a superimposed, controlled, forced-
throttling-valve arrangement is modified recirculation system for the furnace
from the basic Combined Circulation® walls. When once-through flow is
design to allow increased throttling. For insufficient at low loads, the furnace
units designed to have full sliding pres- wall system is automatically protected
sure, the furnace wall and superheater by recirculation of fluid with a boiler
pressures may vary with load, including circulating pump. The recirculation
operation at subcritical pressure. pump eliminates the need for a high-
Sliding pressure is a highly capacity bypass system for furnace
desirable way to operate central wall protection while still allowing
station steam generators at both once-through supercritical flow at
3-24 | Clean Combustion Technologies
higher output. This type of design cal pressure. Boiler throttling valves are
permits furnace wall tubes of sufficient used at the furnace wall outlet to keep
inside diameter to maintain an adequate the wall system pressurized (Fig. 3-16).
mass flow at all operating conditions. The superheater operates in the slid-
The recirculating pump handles both ing-pressure mode below approximately
recirculated flow and feedwater flow 30% load. Above that range, turbine
and, thereby, assists the feed pump, re- load control is accomplished by using
ducing some of its power requirement. the turbine throttle valves.
In the Combined Circulation® design,
all furnace walls are single-pass upflow, System Integration. The simplified
with no mixing headers. Proper design flow diagram (Fig. 3-17) illustrates
tolerances and use of tube orifices com- several features that have significantly
pensate for flow unbalance or uneven advanced system integration. First, the
heat absorption. The mixing vessel or supercritical, once-through steam gen-
sphere is substituted for the convention- erator closes the steam and feedwater
al drum to properly mix the recirculated connections of the steam power cycle
flow from the furnace walls with the in a continuous, heat-addition loop
unit throughput. The constant-pressure, without the division established by the
Combined Circulation® supercritical water level in the drum of subcritical
units, then, are capable of operating recirculating units. This fact is recog-
with once-through flow, but they cannot nized by an automatic, non-interacting,
be operated with full sliding pressure. feed-forward control system geared
The furnace walls must be kept above to the dynamic characteristics of the
the supercritical pressure (3,208 psia entire plant. Second, system integra-
[22.1 MPa abs]) to avoid the film tion permits starting the turbine and
boiling and tube overheating that can steam generator simultaneously under
occur during the transition to subcriti- conditions that provide an optimum of
turbine protection for accelerated cold
and hot starts. This facility is combined
with unique features to remove solids
BT Valves Turbine Throttle Valves and oxides from both pre-boiler and
BE Valve BTB boiler systems. Third, the Combined
4,000 Valves Furnace-Wall Outlet Pressure Circulation® principle eliminates the ba-
sic requirement of other once-through
3,000
systems for minimum feedwater flow
Pressure, psig
Figure 3-19 | A Combined Circulation® tangentially firing, coal-fired reheat steam generator
• All components of the recirculating by raising the actual flow to a safe level
system, except for the recirculating regardless of the feedwater flow. At low
line, are in series with the once- loads, the recirculated flow is high, but
through flow. this flow decreases as the load increases.
• Drainable, welded tube panels form At approximately 60% load, the pressure
all the walls. drop through the furnace wall system
equals the head produced by the circu-
These characteristics provide a num- lating pump. The stop-check valve in the
ber of important advantages, among recirculating line automatically closes.
which is the design flexibility available The circulating pump then ceases to
from the presence of a circulating pump, add to the quantity of furnace flow but
with which the flow quantity through continues to contribute its positive head
the parallel tubes around the furnace on the once-through flow and, therefore,
periphery can be distributed. Other acts as a booster to the feed pump. At
benefits are improved temperature this time, the pump may be shut down.
performance over the entire load range
100
and elimination of the bypass system as Combined Circulation® Performance.
Recirculation Flow a requirement of the furnace wall. In Figure 3-21, fluid temperatures,
80 These advantages will be understood both primary and reheat, throughout a
most clearly after considering the flow typical Combined Circulation® super-
Volumetric Flow, %
Total Flow
60 performance of a typical furnace wall critical steam generator at maximum
system, with the volumetric flow at continuous load are plotted against
40
the furnace wall inlet being a func- the percentage of heat absorption.
Once-Through tion of load (Fig. 3-20). The through The primary fluid takes approximately
20
Flow
flow, as maintained by the feed pump, 83% of total output, showing a gradual
0 increases in direct proportion to load. and continuous rise of fluid tempera-
10 20 30 40 50 60 70 80 90 100
Load, % The recirculated flow, as maintained by ture from the feedwater inlet to the
the circulating pumps, supplements the superheater outlet. The curve of these
Figure 3-20 | Flow in typical Combined through flow over the low-load range in temperatures follows the characteristic
Circulation® furnace-wall system a manner that protects the furnace walls line of supercritical fluid, in which
Clean Combustion Technologies
any heat absorption or change of heat
Figure: 03.20/ SPS content is accompanied by a change
1000PMS5415 K Platen SH in fluid temperature. Desuperheating,
Final
Panel SH
SH occurring at the outlet of the panel
900
Horizontal SH
superheater, provides precise control of
Initial the average steam temperature. It also
Economizer RH
800
adjusts for unbalance of pickup across
Temperature, °F
Pressure, psig
steam turbine throttle valves. This
causes the usual temperature changes 2,000
in the first stage of the turbine and is a SH Inlet Pressure
factor in establishing permissible rates (BT Valve Outlet)
1,000 Throttle Pressure
of load change. This design is extreme-
ly well suited for base load operation.
Cycling and two-shifting operation, 0
0 25 50 75 100
however, may require greater flexibility Synchronization Steam Flow, % MCR
at reduced unit load to protect the
steam generator and turbine during the Figure 3-22 | Sliding pressure operation of Combined Circulation® steam generators (BE, boiler
increased number of load swings. extraction; BT, boiler throttling; BTB, boiler throttling bypass; MCR, maximum
Figure 3-22 shows the rearrangement continuous rating; SH, superheater)
of a constant-pressure, boiler-throttle- Clean Combustion Technologies
Figure: 03.22/ SPS
valve set to obtain sliding-pressure and adopts a full sliding-pressure ap- K
PMS5415
turbine operation to 80% rating proach (Fig. 3-23). The furnace walls Supercritical-
Pressure
while maintaining the furnace walls are allowed to enter the subcritical 4,000 Regime
at supercritical pressure. This design pressure range, along with the super-
has the usual advantages of operating heater circuits, over the entire load 3,500 Subcritical-
Pressure
the turbine in a sliding-pressure mode, range as shown. Benefits to plant op- Regime
3,000
including increased range of reheat eration, in addition to those described
Waterwall
temperature control, reduced thermal above, include:
Pressure, psig
2,500 Pressure
stresses, and improved turbine heat rates Turbine
1. Reduced pressure levels at lower
as a result of variable-pressure operation 2,000 Throttle
loads that unload all cycle compo- Pressure
with full arc admission below 80% load.
nents in terms of pressure, from 1,500
Note that the boiler feed-pump power
the feedwater pump to the HP
is the same for this modified design as 1,000
turbine, thereby prolonging life
for constant-pressure operation. With
span.
use of boiler throttle valves at all loads
2. Improved overall power plant 0 20 40 60 80 100
below 80%, however, some increased Boiler load, % MCR
heat rates when considering
wear and maintenance can be expected.
power consumption of the boiler
This design, then, often is well suited as Figure 3-23 | Pressure versus load for full
feed pump and other auxiliaries. sliding-pressure supercritical
a field modification to existing Com-
3. Simplified and faster start-up design (MCR, maximum
bined Circulation® units that are now
procedures that can be employed continuous rating)
Clean Combustion Technologies
required to cycle or for new units that Figure: 03.23/ (KA)
in the design.
will cycle a minimal number of times PMS5415 K
during their operating lives. The superheater circuitry for the
full sliding-pressure design, as well as
Full Sliding-Pressure Design the start-up system, the operational
The third once-through design controls, and the auxiliary equipment,
eliminates the boiler throttling valves are essentially the same as those for
3-30 | Clean Combustion Technologies
Figure 3-24 | A supercritical-pressure, coal-fired steam generator for full sliding-pressure operation
Live Steam
4,045 psi 279 bar (design pressure)
1,010 °F 542 °C
2,788 t/h 774.4 kg/s
Reheater Steam
1,000 psi 69 bar (design pressure)
1,055 °F 568 °C
2,475 t/h 687.6 kg/s
Feedwater
523 °F 273 °C
Fuel
Bituminous Coal
flow via the flash tank because of in all operational situations where the
the sudden volume increase during turbine is not able to take the whole
start-up (i.e., swelling phenomenon). steam flow (Fig. 3-29). This usually is
Another solution is to return the necessary during start-up or shutdown
condensate to the feedwater tank of the unit or during operational mal-
(Fig. 3-30, right). In this case, the con- functions (i.e., turbine trip). In these
densate, or a part of it, is recirculated operating cases, the HP/LP bypass
to the boiler via the feedwater pump. systems take over the pressure control
This type of start-up system should of the boiler and also ensure sufficient
be used for boilers that are operated flow through the superheater and
mainly at base load. Especially during reheater heating surfaces. Furthermore,
a hot start-up, the feedwater tank the HP bypass valves can serve as a
is not able to recover all the energy; pressure relief (safety) device. Protec-
tion of the boiler reheater part must be
accomplished by separate safety valves.
Separator Separator
Eco Eco Vertical Versus Spiral or Helical
Leveling Vessel Leveling Vessel
Furnace Walls
The principal concern with a sliding-
pressure, supercritical-pressure design
Flash is the requirement for once-through
Tank operation. The mass flow in the fur-
Circulation nace wall tubes must be sufficiently
Pump high to avoid overheating or DNB
Feedwater
Flash while generating steam at subcritical
Tank pressures and to avoid excessive metal
temperatures as well as uneven steam
Deaerator Feedwater outlet temperatures when operating
Pump
at supercritical pressure at higher
HP Feedwater
Heater
boiler loads.
To accomplish these objectives, the
spiral wall design can be used. The
Figure 3-30 | Start-up and low-load system (Eco, economizer; HP, high pressure)
principle of the spiral or helical wall
furnace is to increase the mass flow per
consequently, water and energy losses tube by reducing the number of tubes
Clean Combustion Technologies
6P cannot be totally avoided. In addition,
Figure: 03.30/ (KA) that are needed to envelop the furnace
P P P P P P PMS5415 K the system increases the start-up times without increasing the spacing between
P
3P P
compared to those with a start-up the tubes. This is done by arranging the
P
system via the recirculation pump. tubes at an angle and spiraling them
around the furnace. For instance, the
30°
Bypass Design number of tubes required to cover the
The main function of the HP and LP furnace wall can be reduced in half by
bypass systems is to lead the super- putting the tubes at a 30° angle (Fig.
heated steam to the reheater inlet and 3-31). Note that the centerline spacing,
the reheater steam to the condenser or pitch, is made the same as that on a
Figure 3-31 | Basic principle of spiral-wall
furnace (P, pitch)
vertical wall to prevent fin overheat- of the spiral arrangement. With this
ing. In addition, by spiraling around design, it is customary to revert to
the furnace, every tube is part of all vertical tube construction in the up-
the walls, which means that each tube per portions of the furnace, where the
acts as a heat integrator around the heat absorption rates are lower. This
four walls of the combustion chamber requires use of bifurcated/trifurcated
(Fig. 3-27). The spiral wall concept sections of tubing or an intermediate
thus addresses two major challenges header. Spiral wall configurations also
of the full sliding pressure, supercriti- entail difficult tube routing around all
cal pressure boiler: openings for the firing equipment in
the lower furnace.
1. Achieving the required mass flows
In larger sizes, the mass flow per
to avoid overheating and excessive
tube increases such that cooling flow
metal temperatures by reducing
can be adequate. Vertical walls can be
the number of tube circuits
used under these conditions. In order
2. Minimizing differences in
to ensure low temperature differences
tube-to-tube heat absorption by
at the evaporator outlet and sufficient
exposing each tube to all four
cooling throughout the furnace walls,
furnace walls
rifled tubing is used, as in the Con-
Spiral wall furnaces have been in trolled Circulation® design. The vertical
operation worldwide for many years wall design has individual tube orifices
and have given satisfactory performance. to control the heat absorption in each
Spiral wound furnace designs, however, circuit (in the same manner as in Com-
involve one performance penalty. bined Circulation® units). Analyses
Because of the high mass flow, the based on operating experience and
pressure drop in the lower furnace design practices for Combined Circula-
walls generally is higher than that for tion® units have demonstrated that
vertical wall supercritical or subcritical satisfactory temperature differentials
units, which increases the boiler feed throughout the entire operating load
pump power requirements. range can be achieved. Thus, the two
While on the process side the problems of DNB (tube overheating)
spiral wall design achieves all of the and waterwall steam-temperature un-
required performance parameters, balance can be addressed in a vertical
the furnace wall tubes are at an angle. wall design by employing a tangentially
Thus, the furnace wall support system fired system and rifled tubing.
is more difficult to build. The load
must be transmitted through the Heat Rate: The Why of
fins between the tubes by means of Supercritical Pressure
weld attachments and tension strips. In Chapter 1, it was noted that im-
Consequently, the spiral wound fur- proved thermodynamic cycle efficiency
nace is somewhat more expensive to can be gained by using high steam
manufacture and erect compared with pressures and temperatures. The data
a vertical tube unit. There typically are in Table 3-3 give some insight regard-
fourfold as many tube-to-tube butt ing the relative improvement of net
welds in the furnace walls as a result plant heat rate for several steam cycles.
3-36 | Clean Combustion Technologies
Table 3-3 | Net Plant Heat Rate Large incremental decreases in plant fuel, condenser temperature, and steam
Improvement heat rate (improved efficiency) are generator exit gas temperature. These
very difficult to achieve. Increasing the and other items noted in Chapters 1
2,400 3,500
PSI PSI superheat temperature of a subcriti- through 6 must all be considered when
cal unit to 1,050°F (565°C) in lieu of selecting a pressure and temperature. A
Temperature,
—Btu/kWhr— 1,000°F (540°C) provides only 0.8% detailed history regarding development
°F
1,000/1,000 9,000 8,860
improvement in the heat rate. Increas- of the once-through steam generator
ing pressure to 3,500 psi (24 MPa) and large supercritical units is given
1,050/1,000 8,930 8,800
at 1,000°F/1,000°F (540°C/540°C) in Chapter 25 of the 1966 edition of
1,000/1,025/
1,050
——— 8,630 yields an improvement of 1.5%, and Combustion.
the double-reheat ramp cycle improves Advances in overall power station
For 4,500 psi/1,110°F/1,150°F the heat rate would
4.1% above 2,400 psi (17 MPa) at efficiency have received consider-
approach 7,600 BTU/KWhr
1,000°F/1,000°F (540°C/540°C). able attention in Europe as well (Fig.
Many factors affect plant heat rate be- 3-32). The
�������������������������
European power plant
sides the cycle itself, such as load regime, industry has been steadily increas-
ing the cycle conditions as well as
further optimizing various power
55
3,770 psi/1,075 °F/1,110 °F
plant components. The various con-
Introduction of 1st Supercritical Plants
50 Once-Through Technology 3,480 psi–4,060 psi 260 bar/580 °C/600 °C tributions to achieve this significant
2,540 psi/1,005 °F/1,005 °F (240–280 bar increase in net plant efficiency from,
175 bar/540 °C/540 °C evaporator pressure)
45 as a percentage, the mid-thirties to
g ( %)
Boilers for Process Use and hottest tubes where the flue gas enters
Power Production the bank. The effectiveness of this
A typical pulverized coal-fired indus- surface is a function of the boiler op-
trial power plant boiler is a thermal erating pressure and the superheater
circulation, water tube steam gen- outlet temperature. As Figure 3-34
erator, without steam reheat, with a shows, the percentage of total heat
steam drum operating pressure below absorbed in a boiler bank is reduced
2,000 psig (13.8 MPa gage) and an significantly at higher pressures by
electrical generating capability of less two factors:
than 100 MWe. The lower design
1. In the superheater, the greater
pressure, lower steam flow, and lack
heat absorption required to heat
of a reheater alter design consider-
the steam to a higher temperature
ations to some degree from those of
reduces the temperature of gas
the utility boilers already discussed.
entering the boiler bank.
Operating pressures are selected to
2. The water temperature in the
supply the process steam require-
boiler bank is essentially at
ments of the industrial facility. Typical
the saturation temperature
applications are to provide an easily
corresponding to the boiler
controlled source of heat for driving
operating pressure. As the pres-
chemical reactors or drying products.
sure increases, the temperature of
Many applications generate steam at
the water inside the tubes rises,
higher pressures and temperatures
thereby reducing the logarithmic
to operate a “topping” steam turbine
mean temperature available for
generator, in which the exhaust from
heat transfer.
the turbine has sufficient energy
for the industrial process. These
cogeneration applications provide a Evaluating Pulverized Coal-Fired
substantial increase in overall energy Systems for Industrial Boilers
efficiency compared with providing Reduced efficiency caused by carbon
these quantities of steam and electric- loss is a major factor in the selec-
ity separately. tion of a solid fuel firing system. A
properly designed pulverized coal-
Boiler Design as a Function fired boiler can maintain an efficiency
of Operating Pressure loss resulting from unburned carbon
A common feature of most industrial of less than 0.4%. On the other hand,
boilers is a convective section of heat- a continuous-ash-discharge spreader-
ing surface known as a “boiler bank.” stoker fired unit will typically have a
A boiler bank consists of several carbon loss of 4 to 8%, depending on
hundred vertically oriented tubes con- the amount of reinjection. The pulver-
nected to the steam drum at their top ized coal-fired unit offers a lower
and to a lower drum at their bottom. carbon loss because of the increased
These tubes are a very effective means combustion efficiency obtained with
of absorbing heat from the flue gas the finer coal particles that enter the
to heat the boiler water, generating furnace (normally, 70–80% will pass
a significant amount of steam in the through a 200-mesh screen.)
3-38 | Clean Combustion Technologies
200 psig 600 psig, 750 °F 1,800 psig, 1,005 °F at 1,800 psig, 1,005 °F at
Saturated Steam Superheater Outlet Superheater Outlet Superheater Outlet
240 °F Feedwater 240 °F Feedwater 350 °F Feedwater 350 °F Feedwater
Figure 3-34 | Heat absorption distribution for pulverized coal suspension-fired boilers (SH, superheater)
Clean Combustion
Advanced Technologiesfor
Technology
Figure: 03.34/ km
being ignited to ensure a good ignition
Lignite-Fired
PMS5415 K Boilers and burnout behavior in pulverized
Brown coal and lignites typically coal-fired units. The drying is com-
are medium- and high-volatile coals bined with the grinding process in
(50–60% of volatile matter on daf [dry special milling equipment, the beater
ash free] basis) that are used either in wheel mill. Unlike with bituminous
circulating fluidized bed or in modern, coal, hot flue gas is used as a drying
highly efficient, pulverized coal-fired medium instead of hot primary air.
boilers with tangentially fired systems. The origin of the beater wheel mill is
They are characterized by a relatively the KSG Lopulco firing system, shown
low heating value of 1,750 to 4,750 in Figure 3-35, developed in the early
BTU/lb (4–11 MJ/kg, as received), 1920s by KSG, an Alstom predecessor
mainly because of the high moisture company.
content, which can vary between 20 Depending on the moisture content
and 70%, whereas the ash content can of the brown coal and the require-
be very low (2%) or very high (up to ments for stable coal firing at part load,
35%). Other typical characteristics of three different firing system arrange-
the fuel are the low ash-softening point ments, which are shown schematically
(typically between 1,830 and 2,190°F in Figure 3-36, are applied:
[1,000 and 1,200°C]) and, very often,
the xylite content, resulting in a fibrous 1. Direct injection without vapor
structure of the fuel. separation: The raw coal is
Because of its high moisture content, transferred via belt conveyors
brown coal needs to be dried before into the mills, where it is mixed
Boilers | 3-39
6 6 8
1 7 7 18
13 1
17
2 10
2
4
4 1 5
4
11 4
5 2 3
5
12
3 3 16
9 14
Clean Combustion
into a classifier, Technologies
where coarse particles and maintenance cycles for these
Figure: 03.36/ (KA)
are separated
PMS5415 Kand led back to the beater types of mills.
wheel and fine particles exit the mill In brown coal-fired units, six or
and are carried by the transporting gas, eight beater wheel mills are arranged
consisting of a mixture of recirculated circumferentially around the combus-
flue gas, air, and volatile matter to the tion chamber, as shown in Figure 3-38.
burner openings. In brown coal-fired Full load is achieved with (n − 1) or
units, no primary air fan is installed. (n − 2) mills in operation, with the
The beater wheel mill takes over mills that are out of service undergoing
the role of the primary air fan. The inspection or maintenance. Tangential
pre-drying and grinding capacity are firing has proven to be the preferred
the main design parameters to be option for the burner arrangement
matched with the layout of the firing because of the homogeneous tempera-
system design. In modern designs, ture distribution of the gases inside
the beater wheel speed can be ad- the furnace. Each mill serves one
justed according to the demand for burner compartment at each corner
drying, either by hydraulic coupling of the furnace. Unlike bituminous
or by frequency converter. Such mea- coal, tangentially fired system arrange-
sures significantly improve the service ments, because of the position of the
Boilers | 3-41
PMS5415 K
Zone of Heavy
MW
1,200
Neurath F/G
1,050 Neideraußem K
psi bar °F °C
900 300 Schwarze Pumpe 600
1,100
4,000
Neurath D 1,050
3,000
600 200 560
Capacity
SH Outlet Temperature
SH Outlet Pressure
Frimmersdorf Q
300 100 520
1,000 950
Frimmersdorf G
150 50 500
0 0 900 480
1950 1960 1970 1980 1990 2000 2010
Year Primary Measures
for NOx Reduction
Clean Combustion
Fluidized Bed Steam Technologies
Figure: 03.42/ SPS been referenced in the literature since
Generators
PMS5415 K the late 1800s. The person most often
For decades, fluidized bed reactors have cited as initiating fluidized bed technol-
been used for non-combustion applica- ogy is Fritz Winkler, who in the 1920s
tions in which the thorough mixing developed the Winkler coal gasifier,
and intimate contact of the reactants in which employed fluid bed concepts. He
a fluidized bed result in high product received a patent for his work in 1928.
yield with improved economies of time With a knowledge of Winkler’s
and energy. Although other methods of work and a wartime demand for
burning fuels also can provide very high petroleum products, a Fluidized
combustion efficiency, fluidized bed Catalytic Cracking (FCC) process
combustion (FBC) can burn nearly was developed during the early 1940s.
any fuel efficiently at a temperature low This process greatly increased both
enough to avoid many of the prob- the quantity and quality of the refined
lems encountered when using other product. Over the years, many im-
technologies. provements were made to the original
FCC process, and today, various types
HISTORICAL PERSPECTIVE of fluid bed processes are used in the
Fluidization and observations of refining of petroleum and the produc-
phenomena related to fluidization have tion of other chemical feedstocks.
Boilers | 3-45
One was the 160 MWe demonstration the solids allows the particles to move
boiler at the Tennessee Valley Author- freely, giving the bed a liquid-like, or
Increasing
ity’s Shawnee Steam Plant in Paducah, fluid, characteristic. Recycle Rate
Kentucky.1 This project utilized BFB The transition from fixed bed to fluid
Velocity
technology for coal firing and began bed is illustrated in Figure 3-43, which Minimum
operation in 1988. The other was plots the drop in gas pressure through Fluidization Velocity
the 110 MWe demonstration boiler the bed versus the gas velocity. For a
Figure 3-43 | Plot of gas pressure drop
through a fluidized bed versus gas velocity
3-46 | Clean Combustion Technologies
fixed bed, the pressure drop is propor- Figure 3-44 shows a generalized
tional to the square of the velocity. As fluidized bed combustor, in which fuel is
velocity is increased, the bed becomes burned in a bed of hot, non-combustible
fluidized, and the velocity at which this particles suspended by an upward flow
transition occurs is called the “mini- of fluidizing gas. Typically, the fuel is a
mum fluidization velocity” (Vmf ). The solid such as coal, although liquid and
Vmf depends on many factors, includ- gaseous fuels can be readily used. The
ing particle diameter, gas and particle fluidizing gas typically is the combus-
density, particle shape, gas viscosity, and tion air and the gaseous products of
bed void fraction.2,3 At velocities above combustion. Where sulfur capture is
Vmf , the pressure drop through the bed not required, the ash may be supple-
remains nearly constant and is equal mented by an inert material, such as
to the weight of solids per unit area, sand. For applications that require
because the drag forces on the particles sulfur capture, limestone is used as the
just overcome the gravitational forces. sulfur sorbent and forms a portion of
Further increases in velocity bring about the bed. Bed temperature is main-
changes in the state of fluidization tained in the range of 1,550 to 1,650°F
(discussed later in this chapter). (850–900°C) by the use of a heat-ab-
sorbing surface within or enclosing the
CHARACTERISTICS OF FBC bed. This temperature range is optimal
FBC offers the power engineer the abil- for the chemical processes needed to
ity to efficiently utilize a wide variety of capture sulfur and minimize NOx
fuels, including fuels with high ash, high emissions, and also avoids ash softening
moisture, high sulfur, low volatiles, and in nearly all fuels. Efficient combustion
low heating value that are unsuitable for can be achieved at this relatively low
other firing systems. Furthermore, FBC temperature range because of the long
can utilize such fuels while meeting strin- fuel residence time and high gas–solid
gent stack emission limits. This unique mixing rates in the bed.
combination of fuel flexibility and low The characteristics of fluidized bed
emissions is responsible for the develop- firing also can be illustrated by com-
ment and growth of FBC technology. parison to the other key combustion
technologies—stoker and pulverized
firing—as shown in Figure 3-45. With
stoker firing, fuel feed size is relatively
Flue Gas large and gas velocity low, with fuel
burned in the resulting fixed bed. Fur-
nace temperatures and fuel residence
Bed times are high to provide efficient
combustion with the given fuel feed
Tube Bundle size and low bed mixing rates. With
Fuel
pulverized firing, fuel feed size is very
Sorbent
small and gas velocity high, with the
fuel burned in the resulting entrained
bed (also called “suspension firing”).
Air
Furnace temperatures are very high
to provide efficient combustion with
Figure 3-44 | Generalized fluidized bed combustor
Clean Combustion Technologies
Figure: 03.44/ kn
PMS5415 K
Boilers | 3-47
Stoker Firing (fixed bed) Fluidized-Bed Firing Pulverized Firing (entrained bed)
BFB CFB
Gas
Gas
Gas
Gas
Air Air
Fuel Fuel &
Fuel & Fuel
Sorbent
Sorbent
Figure 3-45 | Relationships between stoker, fluidized bed, and pulverized firing of solid fuels (BFB, bubbling fluidized bed; CFB, circulating fluidized bed)
the low solids residence time, which is in ash and/or moisture. Fuels with over
Clean Combustion Technologies
equal to the gas residence time (a few 70% ash and03.45/ENB
Figure: over 60% moisture have
seconds). With fluidized bed firing, fuel PMS5415 K burned in a fluid bed.
been successfully
sizing and gas velocity are nominally The high thermal inertia of the bed also
between those for stoker and pulver- allows good performance when firing
ized firing, but furnace temperatures are low-volatile fuels, such as anthracite,
much lower. In addition, fluidized bed anthracite culm, and petroleum coke.
firing systems typically use solids recycle
to maintain bed inventory and increase Fuel Flexibility
solids residence time (calculated as bed Because furnace temperatures are below
solids inventory/throughput) to values the ash-softening temperature of most
on the order of hours. The high solids fuels, FBC boilers are not as sensitive to
residence time and high mixing rates in fuel ash characteristics. A wide range of
the bed allow efficient combustion at fuels with varying ash content and ash
the relatively low furnace temperatures properties can be fired in a given boiler.
with FBC. These characteristics of flu-
idized bed firing lead to the advantages Low NOx Production
described below. Emissions of NOx are considered to
come from two sources: oxidation of
Ability to Burn Low-Grade Fuels nitrogen in the air (thermal NOx) and
The high thermal inertia of the bed mass oxidation of nitrogen and/or nitrogen
allows stable ignition and combustion of components in the fuel (fuel NOx). At
very-low-grade fuels, such as those high the low furnace temperatures in FBC,
3-48 | Clean Combustion Technologies
thermal NOx is essentially zero. Design decreases with increasing height in the
features such as staged combustion can furnace. This type of fluidized bed is
significantly reduce fuel NOx. Further- called a “circulating fluid bed” (CFB)
more, selective non-catalytic reduction because of the high rate of material
(SNCR) systems can be added, leading circulating from the furnace to the par-
to very low NOx emissions. ticle recycle system and then back to the
furnace. The CFB is further character-
Low SO2 Emissions ized by very high solids mixing rates.
Emissions of SO2 can be controlled The difference in fluidization state will
within the combustor by addition of a be reflected in different applications and
sorbent material, typically limestone. performance of BFB and CFB.
The sulfur sorbent also can react
with other fuel constituents, such as CHEMICAL PROCESSES
vanadium, reducing the downstream Within the bed, several interrelated
corrosion potential. chemical processes occur, includ-
ing combustion, sulfur capture, and
TYPES OF FBC SYSTEMS nitrogen–oxygen conversion.
As shown in Figure 3-45, there are two
basic types of fluid bed firing systems, Fuel Combustion
each operating in a different state of Even at the relatively low temperatures
fluidization. The state of fluidization associated with FBC, the combustion
depends mainly on the bed particle of fuel in a fluid bed is a rapid process.
diameter and the fluidizing velocity. The combustion rate is a function of
At relatively low velocities and coarse the reactivity of the fuel and the avail-
bed particle size, the fluid bed is dense, able fuel surface area. The combustible
with a uniform solids concentration portion of solid fuel can be considered
and a well-defined surface. This type of to consist of volatile matter (the gas-
fluidized bed is called a “bubbling fluid eous components driven off as the fuel
bed” (BFB), because the air in excess of is heated) and fixed carbon (the solid
that required to fluidize the bed passes remaining, called “char”). Volatile com-
through the bed in the form of bubbles. bustible matter generally burns more
The BFB is further characterized by rapidly than the residual char and can
modest bed solids mixing rates and rela- be viewed as a separate process from
tively low solids entrainment in the flue char combustion. Char combustion
gas. Little recycle of the entrained mate- is a much slower process and requires
rial to the bed is needed to maintain bed substantial solids residence time and
inventory, but substantial recycle rates mixing. With sufficient residence time
may be used to enhance performance. and mixing, the concentration of char
At higher velocities and with finer within the fluidized bed at any given
bed particle sizes, the fluid bed surface time is a small percentage of the total
becomes diffuse as the solids entrain- bed material. The combustible loss is
ment increases, such that a defined bed determined by the amount of char that
surface no longer is present. Recycle escapes the system without burning,
of entrained material to the bed at because the loss from unburned volatiles
high rates is required to maintain bed is insignificant. Both the fly ash and bot-
inventory. The local density of the bed tom ash streams will contain some char,
Boilers | 3-49
but the majority of combustible loss is Equations 3-1 and 3-2 indicate that
typically in the fly ash stream. a mole of calcium is required to capture The ability to capture sulfur
In a BFB, the carbon loss by entrain- one mole of sulfur. Then, defining the during the combustion process
ment from the bed (elutriation) can Ca/S molar ratio as moles of calcium
be on the order of 10% when firing a in the limestone feed to moles of sulfur
is a key feature of fluid bed
fuel such as coal. Recycle of elutriated in the fuel feed, Ca/S = 1/1 could combustion.
material back to the bed is an effective theoretically provide 100% sulfur
means for retaining the char within the capture. In practice, however, this is
system long enough for efficient com- not possible for two reasons. First,
bustion. In a CFB, the high fluidizing the sulfation process takes place on
velocity produces vigorous mixing, and the surface of the lime particles, and
the recycle rates that are attainable with the resulting sulfate shell prevents
a high-efficiency particle separator allow access to the lime in the particle core.
a long solids residence time. For CFB, Second, the time available for the sul-
combustion efficiencies of 98% for less fation reaction is limited (equal to the
reactive fuels such as coal, and of more furnace gas residence time of ~3–6
than 99% for more reactive fuels such as seconds), inevitably allowing some
biomass, can be achieved. For BFB with SO2 to escape capture.
sufficient recycle, combustion efficien- The porosity of the particle surface
cies approaching that of a CFB can be formed during calcination is a strong
achieved for most solid fuels. factor in sulfur capture and results
from the CO2 vacating the CaCO3
Sulfur Capture structure. Ideally, calcination results
The use of limestone as a sorbent allows in a highly porous particle with an
sulfur emissions to be controlled within exposed surface area that is much
the fluidized bed during the combus- higher than that of a smooth particle
tion process. Limestone consists of having a similar diameter. As sulfur
calcium carbonate (CaCO3) and various capture proceeds, calcium sulfate tends
impurities, such as magnesium carbon- to block the pores. Optimum porosity
ate (MgCO3). Lime (CaO) is formed provides the maximum surface that
by calcining the limestone to drive off can be fully sulfated. The presence of
carbon dioxide (CO2): MgCO3 tends to enhance porosity
and, therefore, limestone utilization
CaCO3 → CaO + CO2 (endothermic) (that fraction of calcium reacting with
Equation 3-1 sulfur), even though it does not par-
ticipate in the sulfur capture process.
Sulfur in the fuel is converted to This is because, in calcining to MgO,
SO2 during the combustion process. the MgCO3 increases porosity without
Although nearly all the sulfur is oxi- contributing to pore blockage.
dized, some of the inorganically bound The calcination and sulfation
sulfur may be retained in the ash. The processes begin at around 1,300°F
SO2 combines with the calcined lime (700°C), and the reaction rates increase
in the reaction: with increasing temperature. The most
favorable temperature for simultaneous
calcination and sulfation, however, is
SO2 + CaO + ½O2 → CaSO4 (exothermic)
approximately 1,550°F (840°C). Above
Equation 3-2
3-50 | Clean Combustion Technologies
80
burned”), limiting sulfation capacity. however, can closely approach the
70 Also, at high temperatures, the sulfation sorbent utilization of a CFB.
reaction can reverse. Figure 3-46 indi-
60 cates the dependence of sulfur capture NOx Production
50
on temperature. Emissions of NOx are generated
1,500 1,550 1,600 1,650 The amount of limestone needed from conversion of atmospheric ni-
Temperature, °F
for sulfur capture depends on several trogen in the combustion air (thermal
factors, including the required sulfur NOx) and/or from conversion of
Figure 3-46 | S
ulfur removal versus fluidized
bed temperature
capture, fuel sulfur content, limestone nitrogen in the fuel (fuel NOx). In
reactivity, limestone particle size, and FBC, the furnace temperatures are
Clean Combustion Technologies limestone calcium content. Sulfur low enough to avoid thermal NOx,
Figure: 03.46/ SPS capture in an FBC boiler typically is in and the proper staging of combustion
PMS5415 K the range 85 to 95%, although capture in the furnace can significantly reduce
rates of more than 98% have been fuel NOx.
demonstrated. As the capture rate in- In a CFB, staged combustion is
SO2 Capture, %
fully contacting the sorbent materials secondary benefit, the smaller particle
in the bed. size results in greater sorbent surface
For proper performance, the designer area, improving sulfur capture.
must provide for good mixing without In the freeboard, large solids that
promoting gas bypassing. This is accom- are ejected (elutriated) from the bed
plished by ensuring that the gases in surface separate from the gas flow
the bubbles are rapidly exchanged with and then fall back into the bed. The
gases from the non-bubbling regions freeboard is considered to terminate
of the bed. The non-bubbling part of at the level where the gas velocity
the bed is essentially at the minimum significantly increases, usually on enter-
fluidization condition and is referred ing a convective heat-transfer section.
to as being in the “emulsion phase.” The The freeboard can be considered as a
bubbles rise through the emulsion and second reactor in series with the fluid-
produce agitation. The effectiveness of ized bed. The freeboard serves as an
gas interchange between the bubbles important region in which additional
and emulsion is a function of the bubble combustion and sulfur capture occur.
size and the apparent viscosity of the For coal-fired applications, the fly
emulsion phase. ash can contain significant amounts of
One parameter that can be adjusted unburned carbon and underutilized
to provide for efficient gas mixing is limestone. Typically, fly ash is collected
the mean bed particle size. At a given and recycled back to the bed to enhance
gas velocity, the volume of gas in the combustion efficiency and sorbent utili-
bubbles depends on the Vmf , which is zation. The amount of recycle generally
affected by particle characteristics. The is described as a recycle ratio, equal to
particle size and density both influ- the recycle mass flow divided by the fuel
ence the apparent viscosity of the bed, mass flow. For BFB systems burning
which in turn affects the average bub- bituminous coals, the recycle ratio usu-
ble size and the rate of bubble growth. ally is selected in the range of from 1:1
With a mean bed particle size of 1,000 to 3:1. When firing a bituminous coal,
microns, near-optimum conditions can with the flue gas flow approximately 10-
be maintained over a range of operat- fold the coal flow, a recycle ratio of 1:1
ing conditions with fluidizing velocities gives a solids concentration in the free-
from approximately 2 to approximately board of approximately 10% by weight.
10 ft/s (~0.6 to 3.0 m/s). This solids concentration provides
For a specific application, it may high sulfur capture and combustion
be desirable to choose a finer or a efficiency. For biomass-fired applica-
coarser bed particle size. For example, tions, fuel reactivity is very high and
a low-reactivity, high-sulfur fuel, such fuel sulfur low (generally not requiring
as petroleum coke, may require a sulfur capture or limestone feed), so fly
deep bed with a low design superficial ash recyle is not required.
velocity. This would increase the solids Bed material usually consists of fuel
residence time for both the fuel burn- ash and the products of limestone calci-
ing and the sulfur capture processes. nation and sulfation. For fuels with little
In such instances, the bed particle size or no ash or sulfur, such as biomass, an
could be reduced while still maintain- inert material must be added to form the
ing desirable bubble conditions. As a bed. Properly sized sand typically is used
Boilers | 3-55
of water gage (1½ to 2 psi [10 to 12.5 Overbed feed is most appropri-
kPa]). Considering the pressure that is ate for highly reactive fuels with low
lost in the conveying lines and splitters, sulfur content and for fuels very high
the feed system must move the fuel in ash. With reactive, low-sulfur fuels
from atmospheric pressure to between (e.g., certain sub-bituminous coals,
5 and 10 psig (35 and 70 kPa gage). lignite, and biomass), the combustion
The advantage of overbed feed is efficiency can be quite high, even with
that it is much simpler than under- minimal or no recycle. Also, because
bed feed. The fuel is sized as in a the required sulfur capture is low,
typical spreader stoker application to SO2 bypassing the bed may not be
1 to 2 inches (25 to 50 mm) topsize. a concern. With high-ash coals, the
With the type of equipment in- fuel is relatively dense and contains
volved, this size of material is much sufficient inerts so that a small particle,
easier to feed when wet, so fuel dry- such as ¾ inch (20 mm) or smaller,
ing is not required. can be uniformly spread. Efficiency of
The principal limitations in using sulfur capture can still be a problem,
overbed feed arise where high perfor- but the fuel ash can often contain a
mance is required, especially with fuels large amount of calcium. The amount
such as coal and coke, which require of limestone additive required for
sulfur capture. From a process point sulfur capture may then be negligible
of view, overbed feed injects the coal or no greater than that in the system
downstream of the primary reactor. designed for high-sulfur fuel. Over-
For combustion, this may not be a bed feed is more is appropriate when
major problem, because high recycle simplicity and reliability, rather than
rates and overfire air can enhance performance, are the main concerns.
fuel burnout in the freeboard. The
bed, however, is ineffectively used. Distribution of Heat Release
Larger particles that settle on the bed Within the BFB combustor, fuel
will liberate volatiles, a large source heat is released both in the bed and
of SO2, at the top of the bed. Fines in the freeboard. Bed temperature is
will burn before reaching the bed and maintained at the desired level by heat
liberate SO2 in the freeboard. All this transfer to furnace wall tubes and, if
SO2 bypasses the sorbent in the bed. needed, to tubes immersed in the bed.
Screening of the coal to eliminate fines Because the heat-transfer rate to these
as well as water sprays to avoid fines tubes is very high, it usually is cost-
burning in the freeboard have been effective to maximize the amount of
tried to offset this limitation. Overbed heat released in the bed. For fuels such
feed also requires closer attention to as coal and an underbed feed system,
the removal of rock and other inert 85 to 90% of fuel heat is released in
material being fed to the bed. This the bed. With an overbed feed system,
material can be quite large and, along 70 to 80% of fuel heat is released in
with any ash particles formed during the bed, shifting heat release to the
combustion, will sink to the bottom freeboard, where heat-transfer rates are
of the bed. A layer of large, dense much lower. For fuels such as biomass,
particles can then form at the bottom with relatively high moisture and
of the bed and disrupt fluidization. volatile content, and an overbed feed
Boilers | 3-57
system, only 30 to 40% of fuel heat is tive simplicity and reliability of overbed
released in the bed, with the remainder feed has led to its much more exten-
being released in the freeboard. For sive use. For coal firing, overbed feed
this reason, BFBs that fire biomass do requires fuel sizing in the range of 1
not require in-bed tubes. Also, such to 2 inches (25 to 50 mm) topsize and
boilers typically have an extended free- can accommodate 10% or greater fuel
board region, with several levels of air surface moisture.
feed, to provide for adequate combus- With underbed feed, the fuel must
tion of the fuel volatiles and a relatively be properly sized for pneumatic
uniform gas temperature profile. conveyance. High topsize leads to a
requirement for excessive transport line
Heat Duty Distribution velocity and, as a result, a potential for
The locations of evaporator, super- erosion. With coal, topsize typically is
heater, and reheater heating surfaces limited to ¼ inch (6 mm). Furthermore,
within the combustor and convec- the fuel must have a surface moisture
tion pass are designed to minimize of less than 6% to avoid pluggage in the
total cost while providing the proper transport lines. Sorbent size is selected
performance characteristics over the to maximize sorbent utilization, striking
complete operating range. Usually, the
combustor (furnace) enclosure is com- TABLE 3-5 | Typical Bubbling Fluid Bed Process Parameters
posed of welded, water-cooled tubing Fuel Overbed Feed Underbed Feed
and performs evaporative duty. In-bed Topsize 1–2˝ (25–50 mm) 3⁄8˝ (10 mm)
tube bundles, if required, are used for Moisture — < 6 percent (surface)
evaporation and for superheater and
Fines < 25 percent minus 16 mesh < 20 percent minus 30 mesh
reheater service. The convective pass
Sorbent
typically contains superheater, reheater,
economizer, and air heater surfaces. Topsize ⁄16˝ (5 mm)
3
successfully addressed those issues. The superheater and economizer, plus the
tube bundles can be used for evaporative, air heater. The FBHE contains the su-
superheat, or reheat duty. If multiple perheater surface and may contain an
FBHEs are provided, solids flow to one additional evaporator surface. When
or more can be used to control furnace fluid cooled, the walls of the FBHE
temperature. This feature allows furnace perform evaporative duty.
temperature to be maintained at a For reheat cycles in a unit without an
desired level over a wide range of load FBHE, a portion of the reheater is in
as well as fuels. Furthermore, if one the convection pass, and a portion is in
FBHE contains a finishing reheat the furnace. Reheat steam temperature
surface, solids flow to that FBHE can control is via a split backpass with
be used to control reheat steam tem- gas biasing dampers or steam bypass
perature, more or less independently around the convective pass reheat
of the other temperatures. surface. In a unit with an FBHE, there
The advantage of internal surface usually are at least two FBHEs. A por-
is lower cost, in that the complex- tion of the reheater is in the convective
ity of the FBHE (e.g., plug valve, pass, and a portion is in one of the
solids transfer ducts, and fluidizing FBHEs. Reheat steam temperature TABLE 3-6 | typical circulating
blower/piping/controls) is avoided. control is via solids flow to the FBHE fluid bed process
The advantage of the FBHE is in the containing the reheat surface. The split parameters
potential control of furnace tempera- reheater (with part in convective pass Fuel
ture, enhancing emission control at and part in the FBHE) is designed Top size 3–20 millimeters
part loads and with a range of fuels. to put heat into the reheat circuit at 500–1,000
d50
For large CFBs (e.g., >350 MWe), low loads, when hot solids may not be microns
however, there may not be room in the available to the FBHE, for matching Sorbent
furnace for the internal surface that is the turbine steam temperature without Topsize 1,000 microns
needed, so an FBHE may be required requiring a turbine bypass.
d50 100–300 microns
in any case. The heat duty distribution described
furnace
The distribution of evaporator, above minimizes the heating surface
superheater, and reheater duty within and cost in a subcritical, drum-type 1,560–1,650°F
Temperature
(850–900°C)
the primary solids recirculation loop cycle without a turbine bypass (the
16–20 ft/s
(primary loop) and convective pass is most common cycle in the United Velocity
(5–6 m/s)
set to minimize cost while providing States). Other heat duty distributions
100–1,000
the required performance over the may be best for other cycles and for Particle Size
microns
load range. The major considerations meeting certain special performance
30–50˝ WG
in selecting the heat duty distribution requirements. Pressure Drop
(10–20 kPa)
have been described by Gottung et al.5 Recycle Rate 10 to 100/1
In a unit without an FBHE, the fur- Process Parameters
Performance
nace walls (and wingwalls, if needed) Table 3-6 lists typical CFB process
Carbon Loss 1–2 percent or less
perform evaporative duty, the internal design parameters. Proper fuel feed
surface (if needed) handles superheat size is extremely important to both 1.5–2.5 (for
Ca/S 90 percent SO2
duty, and the convective pass contains operation and performance. If the capture)
the superheater, economizer, and air feed is too coarse, there will be insuf-
SO2 100 ppm or less
heater. In a unit with an FBHE, the ficient material entrained in the gas
CO 100 ppm or less
furnace does evaporative duty, and the leaving the furnace, leading to insuffi-
convective pass typically contains the cient solids recirculation. This results NOx 100 ppm or less
3-62 | Clean Combustion Technologies
capture above 95% has been achieved slowly, at a rate dictated either by the
in several commercial plants. Levels of allowable rate of drum metal tempera-
CO generally are in the range of 100 ture increase by or the allowable rate
to 200 ppm, and levels of NOx below of refractory temperature increase.
100 ppm are typical with SNCR. Both limitations typically are in
the range of 100 to 200°F/h (55 to
Part Load Operation 110°C/h). When solid fuel permis-
Turndown over the load range from sive temperature is reached (typically
100 to 60% is accomplished by reduc- 1,000–1,300°F [540–700°C] for coal),
ing both fuel and air flows. In the solid fuel is added. Furnace tem-
process, grate and furnace velocity are perature is further increased by adding
sufficient to produce adequate mixing solid fuel and backing out start-up fuel.
and solids recirculation and to avoid Above a load of 25 to 35%, the boiler
temperature maldistribution and can run on solid fuel alone.
backsifting of bed material into the air
plenum. Furnace temperatures drop CFB System and Components
moderately. At loads below approxi- The following sections describe the
mately 60%, fuel flow is decreased, but major subsystems within the CFB
airflows are held constant to maintain boiler and discuss typical equipment
velocity. Furnace temperatures drop and major performance criteria.
more rapidly. At some load, usually
in the range of 25 to 35%, furnace FUEL PREPARATION. As with a BFB, the
temperatures drop low enough that CFB fuel preparation system is depen-
support fuel, via a start-up burner, dent on fuel type, fuel characteristics,
must be added to maintain stable range of fuels to be fired, incoming fuel
combustion. size, and required fuel sizing to the
With an FBHE, furnace tempera- boiler. With the variety of fuels fired
tures can be maintained through a given in CFBs and the subsequent variety
load range. As load is reduced, solids in fuel preparation systems that is
flow to the FBHE can be lowered, required, a complete treatment here is
reducing primary loop heat absorption not possible. Only general remarks can
and maintaining furnace temperature. be provided.
At some load, either all solids flow is For coals and waste coals, the prepa-
stopped or a further reduction in solids ration system usually consists of one or
flow would produce an undesirable drop more stages of crushing. Many types
in steam temperatures. Below this load, of crushers, including impact mills,
furnace temperatures will drop with hammer mills, rod mills, and cage
decreasing load. mills, have been applied. For coals, the
system usually is straightforward. For
Start-Up waste coals, the need to limit oversize
Start-up is accomplished by means of material can add complexity via the
start-up burners located in the lower addition of screening and recycle steps.
furnace walls and/or in the primary Drying has been utilized occasionally,
air duct. Minimum primary air flow but justifying the added costs through
is established. The start-up burn- savings in operating costs and im-
ers are used to heat the bed material proved reliability typically is difficult.
3-64 | Clean Combustion Technologies
For biomass, the preparation system system as described above. This location
also is very dependent on the fuel type is advantageous when firing high-
and source. Again, the system may moisture fuels, such as lignite and
need to consider separation/removal of biomass, because this mixes the fuel with
non-combustibles. hot solids, partially pre-drying the fuel.
This feed location also has the advantage
SORBENT PREPARATION. Sorbent can of eliminating a separate fuel feed open-
be purchased at the proper size and ing in the furnace. The seal-pot feed may
delivered via truck to the site, where it be used for any solid fuel when deemed
is loaded into a day bin. Alternatively, to be convenient and economical.
coarse limestone can be stored on site, Because of the high internal mix-
then crushed to the proper size, dried, ing rates, only a few feed points are
and loaded into the day bin. Various required for adequate mixing and
types of crushers, including rod mills, dispersion of the solid fuel within the
roll crushers, and air-swept roller mills, furnace. Typically, one feed point is
have been used. Because the sized ma- provided for every 250 to 500 ft2 (23
terial is relatively fine and typically will to 46 m2) of furnace plan area. For ex-
be conveyed pneumatically from the ample, a 150-MWe unit with a furnace
day bin to the furnace, the sized lime- plan area of 985 ft2 (92 m2) has four
stone must be dried to no more than feed points. It is important to note that
1% moisture. This is accomplished via CFB combustion technology evolved
oil- or gas-fired burners feeding hot from other industrial applications, in-
gas to the crushers. To reduce cost and cluding calcination of alumina hydrate
complexity, Alstom has developed an to alumina. The CFB calciners are fired
alternate system that combines the with oil or gas, because any fuel ash
limestone preparation and feeding would contaminate the alumina prod-
systems into a single system, called the uct. This experience indicates that CFB
“Just-In-Time ( JIT) limestone system,” boilers, although generally designed to
which is described under Sorbent Feed fire solid fuel, can very easily handle
on the next page. liquid or gaseous fuels.
Liquid and gaseous fuels for load car-
FUEL FEED. The solid fuel feed system rying are fired in what are called “lances.”
usually consists of a belt feeder (typical- Lances are fuel feed pipes that carry only
ly gravimetric) followed by a fuel chute fuel plus an atomizing medium—these
or pipe leading to the lower furnace. pipes carry no combustion air—and are
Fuel from the feeder falls by gravity into located in the lower furnace. The lance is
the furnace. The feeder is pressurized intended to disperse the fuel within the
with cold primary or secondary air, bed, where it is combusted in the fluid-
and the head of fuel in the standpipe at izing air stream. Full load can thus be
the feeder inlet forms the pressure seal obtained on liquid or gaseous fuels with
between the day bin and the feeder. An adequate lance capacity. Because these
isolation valve at the feeder discharge fuels have much shorter bed residence
protects the feeder from furnace gases if times compared with solid fuels, they
the pressure seal is lost. require more feed points for proper
Fuel also can be fed to the solids fuel distribution and performance.
return from the seal pot using the same The lances can be either retractable or
Boilers | 3-65
300˝ WG (62 to 75 kPa). Note that 2 inches (25 to 50 mm) thick over the
if these streams are significant and not tube crown and is held in place with
pre-heated in the air heater, the air short metal studs.
heater performance is affected, because The upper furnace contains the
less air is available to cool the flue in-furnace surface (if used) and the
gas. A positive displacement blower gas outlet(s) to the cyclone(s). The
supplies conveying air at the higher density of the bed in this region is
pressures required for pneumatic relatively low, and the density drops
transport of material. off very slowly with increasing furnace
Fluidizing air nozzles are located in height. Because all the air has been fed
the furnace floor for proper air distribu- in the lower furnace, the upper furnace
tion. The nozzles are designed to avoid operates under excess air (oxidizing)
backsifting of bed material into the air conditions. Physically, this section
plenum and pluggage during both op- usually is rectangular, straight walled,
eration and shutdowns. Nozzle pressure formed from finned or fusion-welded
drop is selected to minimize auxiliary waterwall tubing, and unlined to maxi-
power requirements while ensuring mize heat absorption.
adequate air distribution during low- The furnace floor, containing the
load operation. Secondary air ports are fluidizing nozzles (also called the
located along the furnace walls at one or “grate”), typically is water cooled. The
more elevations above the floor. air plenum below the floor can be
either water cooled or formed from
FURNACE. The furnace consists of two plate. If necessary for large units, two
zones: the lower furnace and the tapered, lower furnace sections can
upper furnace. The lower furnace is be used with a single upper furnace,
that portion containing the fluidizing forming the dual grate (or “pantleg”5)
nozzles, secondary air ports, fuel feed configuration (see Fig. 3-54). This
ports, and solids recycle ports. The configuration allows secondary air
density of the bed in this region is rel- feed to the center of the furnace, a
atively high, with density decreasing problem that is encountered when
rather rapidly with increasing furnace scaling CFBs to large sizes. The dual-
height (see Fig. 3-51). Because of the grate configuration is used for Alstom
staged air feed, this region is sub- CFBs in sizes of greater than 350
stoichiometric. Physically, this section MWe. In small sizes, the furnace can
usually is rectangular, tapered, formed be bottom supported, but in larger
from finned or fusion-welded water- sizes (>50 MWe), top support nearly
wall tubing, and lined with refractory always is used.
material to protect the tubing from ero-
sion by the dense bed and corrosion in Circulation System. The walls of the
the sub-stoichiometric atmosphere. The furnace are cooled by natural circu-
optimum refractory lining is hard (to lation for subcritical units. When
minimize erosion), thin (to minimize provided, an FBHE evaporator bundle
weight), and reasonably conductive (to can be cooled using natural circulation
maximize furnace heat absorption). or pumped circulation. Use of natural
The refractory lining typically is 1 to circulation generally requires inclined
Boilers | 3-67
viability of the lining. The exterior Again, proper design of the tube
must be left uncovered to avoid bundle supports is essential. Fluidiz-
Upper problems with refractory anchors. ing velocity is low (1–2 ft/s [0.3–0.6
Header m/s]). The fluidizing medium is air.
The advantages of the steam-cooled The particle size is small, and the
cyclone are reduced weight, reduced re- carbon content of the material is low.
fractory maintenance, and reduced heat All these conditions lead to essen-
losses. The advantage of the plate/re- tially no erosion or corrosion of the
Typical
fractory cyclone is reduced capital cost. in-bed tube bundles. Also, because
Steam of the high bed density, rates of heat
Tube SEAL POT. The seal pot is a non- transfer are very high. The enclosure
Inside
mechanical valve that moves the sol- can have either refractory lined steel
ids collected by the cyclone back into plate or water-cooled tubing.
the furnace against the furnace back-
pressure. Solids flow down the inlet SNCR. Even though NOx emissions
Lower
Header
side, up the outlet side, then back to from the furnace are low because of
the furnace. The bottom portion of low furnace temperatures and staged
the seal is fluidized so that material combustion, emission regulations
Figure 3-55 | Steam-cooled cyclone
in the seal can seek different levels on may require additional NOx control.
each side of the seal. The difference In a CFB, this is accomplished using
Clean Combustion Technologies in level corresponds to the pressure an SNCR system. This system uses
Figure: 03.55/ (STM) Insulating
Insulating difference across the seal. Then, solids an ammonia-based reagent (anhy-
SteelPMS5415 Fire Brick
BlockK
Casing entering the seal inlet displace solids drous ammonia, aqueous ammonia,
out of the seal on the outlet side. or urea) to reduce NOx emissions.
The seal pot is constructed of steel The reagent is injected into the gas
plate or pipe with a multiple-layer, stream at the cyclone inlet duct
refractory lining. Fluidizing nozzles and/or outlet duct using nozzles
along the bottom of the seal provide located on the side of the duct.
the required fluidizing air. On units NOx reductions of 50% and higher
with an FBHE, a plug valve is located can be achieved. The required gas
in the lower portion of the seal pot to temperature range for effective NOx
Anchor Bricks Brick Lining
regulate the flow of solids from the reduction is 1,400 to 1,700°F (760 to
seal pot to the FBHE. 925°C). This becomes an issue at part
Figure 3-56 | P
late-refractory cyclone
construction
loads, when gas temperatures at the
FBHE. The FBHE is a bubbling bed cyclone can become too low. At these
Clean Combustion Technologies heat exchanger consisting of one loads, ammonia flow is stopped to
Figure: 03.56/ (STM) or more compartments separated avoid excessive ammonia slip.
PMS5415 K
by weirs and containing immersed
tube bundles (see Fig. 3-52). Hot CONVECTIVE PASS. The convective pass
solids from the seal pot enter the is of the same basic design as that
FBHE, where they are fluidized, used in pulverized fuel- or stoker-
transfer heat to the tube bundles, fired boilers. The enclosure walls are
and flow back to the furnace. The usually formed from finned or fusion-
tube bundles can be an evapora- welded tubing and are steam or water
tor, superheater, or reheater surface. cooled. Where gas temperatures are
Boilers | 3-69
sufficiently low, duct plate can be used Use of steam or hot-water pre-
to form the enclosure. The convection heaters, or of a cold-air bypass,
pass can contain superheater, reheater, usually is recommended to increase
evaportive, and economizer surface. the air heater average cold-end tem-
Gas velocities are typical of solid fuel- perature to minimize acid (H2SO4)
fired boilers. Retractable or rotary condensation/corrosion in colder
soot blowers can be used to keep the climates and/or at low loads. The air-
heat-transfer surfaces clean. bypass option offers the advantage of
cooling the unit faster on a shutdown.
AIR HEATER. A prime consideration in Note that because of the low SO2/
selecting an air heater type for CFB SO3 in the flue gas (from limestone
applications is the high air-to-gas injection in the furnace), the acid
pressure differential resulting from the dew-point temperature (and the
high primary air pressures required associated average cold-end tempera-
plus the draft loss across the cyclone ture) is relatively low. The CFB boiler
separator and backpass. Low leakage can operate at relatively low exit gas
designs, such as tubular and heat pipe temperatures, offering the potential
air heaters, have been used. The devel- for higher boiler efficiency than with
opment of advanced leakage control other boiler types.
systems for Ljungström® regenerative
air heaters, as described in Chapter ASH REMOVAL/COOLING. The ash removal
6, has led to their use on most large system includes both the bottom ash
(>200 MWe) CFBs. and fly ash systems.
Because of the different air pressures,
the primary and secondary air streams Bottom Ash. The main functions of
are heated separately, in series or in the bottom ash removal system are
parallel, with the gas stream. With large to control the flow of bottom ash to
amounts of fluidizing air for FBHEs maintain the desired bed inventory
and FBACs, it sometimes is economical and to cool the ash to a temperature
to provide a separate air heater for this suitable for the bottom ash convey-
fluidizing air. This is because the pri- ing system. Bed pressure, measured
mary and secondary air heaters become just above the furnace floor, is the
large to achieve a given stack tempera- indicator of inventory, and bottom ash
ture if a significant amount of the total flow is adjusted to maintain desired
air bypasses the air heaters. bed pressure. The bottom ash system
With tubular air heaters, the most also can help control accumulation
common design for CFB applications of oversize material. In a CFB, such
is a gas-over-tubes/air-through-tubes accumulation can produce an unfavor-
design. The dust-laden gas passes able pressure profile with excessive
over the tubes and, because the tubes material in the lower furnace and
are arranged in-line, they can be little in the upper furnace, resulting
easily cleaned with soot blowers. Gas- in poor performance. The best and
through/air-over designs, although most direct way to control oversize
potentially more difficult to clean, also accumulation, however, is with proper
have been used successfully. design of the fuel-sizing equipment to
3-70 | Clean Combustion Technologies
the bed density will be too low mixing, creating zones of low O2
and unpredictable, leading to low near the feed nozzles that reduce
rates of heat transfer, high furnace combustion efficiency and SO2
temperatures, insufficient loading capture, both of which need O2.
of the FBHEs (especially at part • Cyclone diameter: This impacts
load), and excessive NOx/SOx collection efficiency. Excessive
emissions. diameter reduces collection
• Furnace depth: This impacts efficiency. Inadequate collection
secondary air penetration into efficiency reduces solids recirculat-
the center of the furnace. Solids ed to the furnace and bed density,
(including fuel) residence times in increasing furnace temperature
a CFB are very high, meaning that and NOx/SOx emissions.
fuel particles will find their way to
all parts of the furnace, including Alstom’s approach to scale-up has
the center. With excessive furnace been to design larger boilers using
depth, secondary air (fed from the “modules” that have been proven
walls of the furnace) cannot pen- on smaller boilers, as illustrated in
etrate the bed and reach the fuel Figure 3-57. Thus, a 300 MWe unit
in the furnace center, decreasing essentially consists of three 100 MWe
combustion efficiency. furnace/cyclone “modules” without
• Fuel feed points per furnace increasing furnace height, furnace
plan area: This impacts combus- depth, feed points/plan, or cyclone
tion efficiency and emissions. diameter beyond the values proven
An inadequate number of feed at the 100 MWe scale. For more
points limits fuel distribution and compact layout of larger units, the
dual grate (pantleg) configuration is
used while avoiding excessive furnace
depth. This configuration allows
< 50 MWe 100–200 MWe 200–300 MWe
secondary air feed to the center of the
furnace on either side of each grate,
ensuring high combustion efficiency.
This configuration can be scaled to
600 MWe and larger.
Single Grate
Supercritical CFB
Supercritical steam conditions allow
an increase in power plant efficiency
> 300 MWe 400–600 MWe
in accordance with the selected steam
conditions. Combining this increased
efficiency with the inherent fuel flex-
ibility and emission control of CFB
technology offers obvious advantages.
Dual Grate
(Pant-Leg)
Design of a supercritical CFB involves
bringing together two technologies:
supercritical boiler design and CFB
Figure 3-57 | Scale-up (modular approach)
40
35 PC Furnace
Height Above Lower Hopper, m
30
25
Cyclone
20 CFB Furnace Overflow
Duct
15 Cyclones
10
Backpass
5
0 Furnace
0 100 200 300 400
Heat Flux (average), kW/m2
Coal Silos
Figure 3-58 | Heat flux profile: Pulverized coal
versus circulating fluidized bed
on the larger side, typically above 300 8.7 MPa gage, and 513°C) designed
MWe, to match the size of available to fire wood (Fig. 3-60). The unit
steam turbines. has a single-grate furnace, with one
3-74 | Clean Combustion Technologies
Utility Application
Two examples of CFB designs for
utility applications also are described.
The first boiler is a 1,922,000 lb/h,
Figure 3-61 | A 524,000 lb/hr circulating fluidized bed 2,524 psig, and 1,005°F (244 kg/s,
17.5 MPa gage, and 541°C), with
reheat, designed to fire bituminous coal
Clean Combustion Technologies
Figure: 03.61/ (STM)
(Fig. 3-62). This design has a single-
PMS5415 K grate furnace with wingwall surface,
three cyclones, three seal pots, and two
FBHEs (one with a superheat surface
and one with a reheat surface). The
convective pass contains a superheater,
reheater, and economizer surface. A
regenerative air heater and FDA also
are included.
The second boiler is a 2,260,000 lb/h,
2,524 psig, and 1,004°F (287 kg/s,
17.5 MPa gage, and 540°C), with
reheat, designed to fire a high-ash
anthracite (Fig. 3-63). This design has
a dual grate furnace, four cyclones,
four seal pots, and four FBHEs (two
containing a superheat surface and
two containing a superheat and reheat
surface). The convective pass contains a
superheater, reheater, and economizer
surface. Also included are four rotary
ash coolers and a regenerative air heater.
Figure 3-62 | A 1,922,000 lb/h circulating fluidized bed with reheat
HYBRID CFB
The hybrid CFB, called the FiCIRC™
CFB, combines a BFB with a recir-
culation loop of fines to increase
performance over that attainable with
a BFB alone. The basic FiCIRC™
module is shown in Figure 3-64 and
includes a bubbling bed furnace, an
air plenum, an air distributor with
fluidizing nozzles (tuyeres), an in-bed
tube bundle, and a cyclone/standpipe.
The use of relatively small-diameter
cyclones with direct return of col-
lected material to the bed produces a
large amount of fine material circulat-
ing through the freeboard. The result
is high combustion efficiency, high
limestone utilization, and low NOx/
CO emissions. Ammonia injection
into the freeboard is very effective in
further reducing NOx.
The boiler design is relatively sim-
ple, with a plate/refractory furnace
enclosure and cyclones plus remov- Figure 3-63 | A 2,260,000 lb/hr circulating fluidized bed with FBHEs and reheat
able in-bed evaporator tube bundles
bolted to the furnace enclosure. Fuel
Clean Combustion Technologies Steam Drum
feed is over the bed via spreader Figure: 03.63/ (GM)
or in-bed via pneumatic injectors PMS5415 K
from the sidewalls. Hot gases from
the cyclones enter a convective pass Waterwalled
Cyclone Convection Unit
containing a superheat and econo- Separators
mizer surface, tubular air heater, and Primary
baghouse/electrostatic precipita- Superheater
tors (see Fig. 3-64). Boiler height is
relatively low, and the boiler typically
Fines Economizer
is bottom supported. Return
Larger boilers are designed using
multiple furnace/cyclones modules. Air Heater
Because of this modular approach Fluid Bed
Furnace
and the relatively small size of the
Air Heater
basic module, maximum FiCIRC™ Inlet
boiler size is in the range of 50 to
70 MWe. Flue
Air Inlet
Gas
PRESSURIZED FBC SYSTEMS Air Heater Outlet
In-Bed Evaporator
This chapter has introduced the vari- Tube Bundle Outlet
fluidized bed steam generators, and the gas turbine. The turbine exhaust
it has described the implementation is further reduced in temperature
of designs for boilers operating at or by passing through heat recovery
near atmospheric pressure. Combus- equipment, such as evaporative and
tion in that regime is referred to as feedwater-heating surface.
“atmospheric FBC” (AFBC). Fluidized The commercialization of this tech-
bed operation at pressures 10 to 20 nology has been difficult, mainly for
fold atmospheric pressure, taking place two reasons:
in large cylindrical or spherical pres-
sure vessels, in combination with axial 1. The optimal furnace tempera-
compressors and gas turbines is termed ture of the FBC process is lower
“pressurized FBC” (PFBC). than today’s state-of-the-art
In addition to the reduced emis- combustion temperature of gas
sions of SOx and NOx that are turbines. Thus, the efficiency
possible with fluid bed combustion, advantage achievable by com-
PFBC offers the potential for a gain bined cycle operation cannot be
in overall thermal efficiency because fully exploited.
of the incorporation of a gas tur- 2. Difficulties in adequate cleaning
bine in the cycle. Another potential of the flue gas entering the gas
advantage of the PFBC system is that turbine blades from particles and
all the equipment operating above alkalies, as well as the extended
atmospheric pressure is smaller in outage time required to repair
size than it would be at atmospheric equipment inside the pressurized
pressure, making shop assembly and vessel, led to lower availability
barge delivery of components an at- during commercial operation of
tractive option. PFBC units.
Two of the many approaches to
the pressurized cycle are the PFBC
Heat Recovery Steam
turbocharged cycle and the PFBC
Generators
combined cycle. Each has many
possible variations that are being Overview of Combined Cycle
investigated throughout the world in Gas turbine (GT) combined cycles
both design studies and operational account for almost one-fifth of
units. In the turbocharged cycle, hot global electrical capacity and electrical
flue gas from a PFBC boiler (~800°F production. Much of this capacity is
[~425°C]) is expanded through a relatively new, having been installed in
gas turbine that produces enough the 1990s and 2000s. North America,
power to drive the turbocharging Asia, and Europe account for most of
compressor. One version of the higher the capacity. Because they can be built
efficiency power-producing PFBC relatively quickly, combined cycles are
concept has the gas leaving the pres- responsive to the power industry’s
surized fluidized bed combustor at business cycles. Typically, 30 to 40
approximately 1,600°F (~870°C). GW are added each year, but annual
The gas is cleaned in tandem, high- orders have run as high as 100 GW
temperature cyclones and then sent to during market cycle peaks.
Boilers | 3-77
High-Pressure RH DSH
DSH Spraywater Spraywater
High-Pressure Intermediate-Pressure Low-Pressure
High-Pressure Steam Drum Steam Drum Steam Drum
Steam Outlet High-Pressure
DSH Spraywater IP/LP Optional
Feedwater Stack
Silencer
High-Pressure
DSH
Optional
Duct Burners
Gas Flow
IP/LP ECON
HPECON3B
HPECON5B
HPECON3A
HPECON5A
HPECON6
HPECON1
HPECON2
HPECON4
Optional
HPEVAP
IPECON
LPEVAP
IPEVAP
HPSH1
HPSH2
HPSH3
RHTR1
RHTR2
LPSH
IPSH
Stack
Damper
RH DSH
Spraywater High-Pressure
RH Feedwater
DSH
Low-Pressure
RH Steam
Cold
Steam Outlet
Reheat
Outlet
Inlet
Inlet Duct Module 1 Module 2 Module 3 Module 4 Module 5 Stack
Optional
CO/SCR Cavity Transition Duct
Notes:
1. Duct burner would be supplied in either the inlet duct or downstream of RHTR1.
2. If duct burner is located downstream of RHTR1 then Module 1 would consist
of HPSH1 and RHTR1. Module 2 would be HPSH2-HP Evap.
3. HP and DSH is shown as exit stage. It could also be interstage (located between
HPSH/RHTR1 and HPSH/RHTR2). If necessary, both might be supplied.
Figure 3-72 | Triple pressure with reheat HRSG (HP, high pressure; IP, intermediate pressure; LP, low pressure)
for a triple-pressure-with-reheat
1,000 Gas-Turbine Exhaust Temperature High-Pressure
HRSG illustrates the way in which
Intermediate-Temperature
heat-absorbing sections operating at 900 High-Pressure Economizer
High-Temperature High-Pressure
certain temperature levels are located 800 Economizer Low-Temperature
High-Pressure
in the gas stream to minimize the Superheater Intermediate-Pressure Economizer
700
Evaporative Bank
Temperature, °F
amount of heat-transfer surface. 600 Feedwater
Steam/Water Temperature IP
Often twenty discrete heat exchange Preheater
500 Economizer
sections are distributed in descending LP
400 Evaporative
order based on the gas temperature Intermediate-Pressure
available and the fluid temperature 300 Superheater
requirements. Note that the LP 200 High-Pressure Evaporative Bank
and IP economizer, evaporator, and 100
0 10 20 30 40 50 60 70 80 90 100
superheater sections are arranged in Percent Heat Transferred from Exhaust Gas
an alternating sequence with sections
of the HP economizer to optimize Figure 3-73 | Temperature profile of unfired HRSG with three operating pressure levels
boiler performance. Furthermore,
note that the HP superheater and Clean Combustion Technologies
IP reheater sections are arranged in Figure: 03.72/ SPS
an economizer means the economizer
an alternating sequence to meet the can be isolated from the waterPMS5415
flow pathK
temperatures required at the HP and and become a fired pressure vessel. The
IP steam turbines, respectively, as isolatable economizer must be protected
temperature requirements at these by a pressure relief device.
steam turbines typically are similar.
Pressure Relief Valves for Isolatable
Isolated Economizers Economizers. Traditional sizing criteria
Hot Water Extractions. The HRSGs for pressure relief devices on isolat-
may have one or more hot water able economizers in Section I of the
extractions where hot water is extracted ASME Boiler and Pressure Vessel
from an economizer for use in another Code (B&PVC) is based on relieving
process in the plant. These extractions steam. Although the calculation is based
can be used for processes such as fuel on steam relief, the valve typically will
pre-heating or combustion turbine relieve water, which will then flash to
rotor air cooling. To provide water at steam as the pressure is quickly reduced
the desired temperature, the extraction across the valve. Water is not com-
is located downstream of one or more pressible. If the economizer is isolated,
economizer sections. The extraction changes in temperature cause quick
flow typically is required at all operating increases in pressure, lifting the valve.
conditions. To ensure the availability of Unless the economizer is steaming, the
the extraction flow when the feedwater pressure is quickly relieved, and the
control valve is closed, the feedwater valve closes. If the pressure relief valve is
control valve is located downstream relieving water, it will not reseat quickly.
of the extraction point. Locating the Operation of all valves in the feedwater
feedwater control valve downstream of system should be reviewed to make sure
3-88 | Clean Combustion Technologies
the valve is not being held in a relieving oil, the condensate pre-heater may
condition by the discharge flow and have to be bypassed.
pressure of the feedwater pump.
Design of the pressure relief valve Operation and Control
must accommodate relieving either The typical HRSG comprises a
steam or water. If a traditional, spring- unit in the functional hierarchy of
loaded safety valve is used, it will the power plant that generally is
be quickly damaged when relieving controlled by a Distributed Control
water. Spring-loaded safety valves are System (DCS). Depending on the
designed for relieving steam. Steam plant configuration and operating phi-
is compressible and acts as a cushion losophy, the HRSG operation can be
when the valve closes, preventing the a combined cycle, with or without a
disk from slamming into the seat. steam turbine functional unit. In gen-
Water is not compressible, and so eral, HRSG operation is coordinated
does not provide this cushion. When with a specific gas turbine functional
a spring-loaded safety valve relieves unit and may be coordinated with
water, the disk and seat are damaged a specific steam turbine unit, steam
after few cycles. A pilot-operated valve header system, or a combined cogen-
limits differential pressure across the eration steam turbine and process
seat to minimize the impact between steam header system.
the disk and seat and, therefore, is The hierarchy of control is broken
appropriate for either steam or water down into the following function group
service. controls: drum level, steam temperature,
blowdown tank level, steam vent, and
Recirculation. If the overall plant steam drain control. Optionally, ad-
design requires the HRSG to include ditional function group control can be
an LP drum acting as a storage tank provided for feedwater pre-heater tem-
for the HP and IP feedwater pumps, perature control, duct burner firing rate
care must be taken to avoid corro- control, HRSG inlet damper bypass
sion in the “cold end” of the HRSG. control, and so on. The direct control
In this arrangement, cold condensate interface to the process is the DRIVE
flows through a condensate pre-heater control level, made up of the actuators
heat exchanger located in the gas path for control valves (feedwater and fuel
before entering the LP drum. Typi- flow), motor-operated valves (high-en-
cally, a recirculation pump is supplied ergy steam vents and drains, dampers),
to recirculate the hot water from the and solenoid air-operated power block
condensate pre-heater outlet back to valves (desuperheater spray water).
the condensate pre-heater inlet. The
recirculation flow is varied such that Start-up. Typically, the gas turbine and
the mixed temperature of the conden- the HRSG are started together. If a
sate flow and the recirculation flow is diverter damper is between the gas
greater than the water dew point of turbine and the HRSG, the HRSG
the exhaust gas. With a higher-sulfur can be started at a later time. The
fuel, such as conventional No. 2 fuel following permissives must be met in
Boilers | 3-89
order to initiate HRSG start-up: (i.e., a rise in drum water level as-
stack damper open, ready lists com- sociated with increased water volume
plete (gas turbine ready list, steam in the evaporator section because of
turbine ready list, electrical ready list, increased temperature and steam con-
balance-of-plant ready list, HRSG tent). During start-up, the drum level
ready list), and HRSG cold-start set points are reset to normal operating
valve alignment table complete. Once condition. Typically, the set points are
all pressure sections of the HRSG are automatically reset at specific steam
filled with water and levels are verified flow rates. The start-up set point flow
to be at start-up set points, the gas rate will be monitored by single-element
turbine start can be initiated, followed measurement via HP steam flow. Three-
by the HRSG start-up. For combined element measurement drum control
cycle start-up, the gas turbine starts will automatically take over when the
using its normal start-up procedure respective steam flow is greater than
and loading curve. approximately 30% of normal steam Faster start-up times allow
For a start-up to proceed, the follow- flow (typically the lowest repeatable combined cycle plants to
ing systems related to the HRSG must and accurate steam flow measurement
respond to load demand in a
be in operation: available). As the HRSG is brought
up to the normal operating pressures timely fashion.
1. Feedwater storage and feedwater
and temperatures, steam vents and
pump systems
drains are cycled. The intent is to
2. Drain and vent system
warm up the HRSG unit, remove
3. Blowdown system
any condensate, and minimize pres-
In addition, the following function sure losses and waste heat releases
groups in the HRSG control system through the vents and drains. The
must be enabled: duct burner is permitted to operate
above a determined gas turbine load.
1. LP drum level control
2. IP drum level control
Drum Level Control for all Drum level
3. HP drum level control
Control Systems
4. Blowdown tank level control
Normal Level Control. The drum level
5. HRSG drains
will be corrected by the drum pres-
6. HRSG vents
sure. Level control is maintained by
The gas turbine and HRSG gas monitoring drum level and adjusting
path also must be purged of all the feedwater control valves to
gaseous, or suspended, combustibles provide more or less feedwater to
before gas turbine flame is initiated. the system.
The purge will be in accordance with Measurement devices used in level
local codes. control include:
The control system drum water
level set points are set to start-up set • Steam outlet flow (pressure and
points. The drum level set points may temperature compensated)
be lower than the normal operating • Feedwater to drum flow (pressure
set points to allow drum level “swell” and temperature compensated)
3-90 | Clean Combustion Technologies
pressure and temperature in the fossil fuels, such as coal and heavy oil.
boiler steam drum. Thus, HRSGs do not have to contend
• Hot side: The steam–water with erosion and fouling, and unless a
mixture density in the evaporator corrodent-laden oil is being fired, corro-
will be approximately 25 lb/ft3 sion problems are minimal.
(400 kg/m3). Variations in boiler Materials are almost always iron-
pressure have a lesser effect on the base alloys, ranging from plain carbon
mixture density. steels through low-alloy steels to the
creep strength-enhanced ferritics and
As boiler pressure increases, the dif- stainless steels. Other high-alloy
ference between the densities of water iron-base alloys, erosion-resistant
and steam, which is the motive force materials, and the nickel-base materi-
for natural circulation boilers, becomes als have not frequently been used in
smaller. Thermal head differential is HRSGs because of the relatively low
the resulting differential and ranges gas temperatures and ease of fuel
between approximately 25 psi and 10 psi handling. The product forms used are
(1.7 and 0.7 bar), with the greater also common to other boilers with:
differential being possible in lower
• Tubing for heat-absorbing surfaces
pressure boilers.
• Piping for headers, manifolds,
Circulation ratio is defined as
links, and lines
the mass flow of water entering the
• Plate and sheet for drums, casings,
downcomer divided by the mass flow
and ductwork
of steam in the water–steam mixture
• Structural members for support
leaving the evaporator tube circuits.
• Castings and forgings for valves
Because circulation is dependent on
and pipe fittings
the thermal head, which is dependent
on boiler pressure, the flow of water In contrast to other utility and
into the downcomer increases as load industrial boilers, HRSGs do not
decreases. At low loads of 50% of the use heat-absorbing waterwalls for
maximum continuous rating or less, gas containment and flow direction.
the circulation ratio will be much Typically either a cold casing or a hot
higher, because less steam is being casing design is used. In the more
generated. Natural circulation boilers common cold casing design the duct
generally are designed for circulation is formed from carbon steel outer
ratios ranging from 3:1 up to 10:1. plates, a layer of insulation, and a
system of overlapped inner liner
HRSG Materials sheets. Both the insulation thickness
The HRSGs have water temperatures and the liner sheet material are varied
and pressures similar to those encoun- as the gas temperature changes from
tered in subcritical utility and industrial 1,100 to 1,200°F (600 to 650°C) at
boilers, but the gas temperatures are the inlet to between 212 and 300°F
considerably lower, precluding most (100 and 150°C) at the stack. All
common overheating incidents unless of the heat-absorbing tubes have
supplemental firing with duct burners is extended external surfaces in the
employed. Turbine exhaust gas is much form of spiral wound fins. Because
cleaner than that associated with other the turbine exhaust gas has very little
3-94 | Clean Combustion Technologies
particulate, these fins are not prone When shop fabricated pressure
to pluggage and can be closely spaced, part modules are supplied, these
which results in the HRSG being a are the largest and heaviest HRSG
very compact steam generator. components for the project.A pres-
Module 2
Material selections follow standard sure part module is assembled by
industry practices for structural and connecting from 3 to 30 harps. The
non-pressure containing components number of pressure part modules per
and boiler code limitations for pressure HRSG can vary from two modules
parts. The structurals can nearly always for small HRSGs (single wide, two
utilize plain-carbon steels in readily modules deep) (Fig. 3-75) to up to
Module 1
available product forms. Pressure parts 18 modules for the largest HRSGs
(basically plates for large drums, tubes (triple wide, six modules deep) (Fig.
and pipes for the remaining cylindrical 3-76). Depending on the degree of
components, and forgings and castings shop fabrication, module weights can
for fittings) employ carbon steels for range from 20 to more than 300 tons,
low-temperature sections, chromium– with approximate dimensions of 18
molybdenum low alloys (1¼Cr-½Mo × 13 × 98 ft (5.5 × 4 × 30 m). If the
and 2¼Cr-1Mo) for the next level, and infrastructure, including roads, bridges,
Gas Flow
creep strength-enhanced ferritics (9Cr- docks. and so on, cannot support the
Figure 3-75 | Single wide × two modules 1Mo-V) or type 304 stainless steel for weight or accommodate the size of a
deep HRSG, top view the hottest sections. module, then the module will have to
Construction is effected mainly be split into two or more parts during
Clean Combustion Technologies through cutting, forming, and welding. the initial design phase.
Figure: 03.75/MH
PMS5415 K Machining is mostly limited to that Shipping limitations also must
required for generating appropriate weld be considered when designing other
preparations. All welding utilizes fusion HRSG components. The HRSG
processes with filler metal additions structure, inlet duct, stack, platforms,
except for the tube finning process, dur- stair towers, steam drums, and piping
ing which an electric resistance process all have to be designed with these
is used. The fusion processes include gas limitations in mind. For the HRSG
metal arc, gas tungsten arc, submerged structure, large shop assemblies are
arc, and shielded metal arc. required to reduce construction costs.
These assemblies can be as large as
Shipping Considerations 4.9 × 13 × 98 ft (1.5 × 4 × 30 m)
In addition to all the performance and and weigh over 50 tons. Special
structural inputs, the HRSG designer trucks and rigging equipment are
must be aware of shipping limitations, required to move steel assemblies of
loads, and requirements. These practical this size. Inlet duct panels follow the
issues often create hard design con- same rules as the main HRSG struc-
straints that change the configuration of ture but typically are smaller. The
HRSG components. Available modes HRSG stack can be split into 180,
of transport, infrastructure, and eco- 120, 90, or 72° sections, depending
nomics are all factors influencing design. on the diameter and over-the-road
Transportation limitations or require- width clearance. If the stack shell
ment oversights lead to costly delays section is too wide, it will have to be
and, potentially, damage to the product split to the next smaller size until it
before it is delivered to the customer. can be transported to the site. For
Boilers | 3-95
Module 1C
Module 2C
Module 3C
Module 4C
Module 5C
Module 6C
Module 1B
Module 2B
Module 3B
Module 4B
Module 5B
Module 6B
Gas CL Unit
Flow
Module 1A
Module 2A
Module 3A
Module 4A
Module 5A
Module 6A
Figure 3-76 | Triple wide × six deep heat recovery steam generator
Essential Components
Duct Burner System
Temperature limits at the gas turbine
inlet force the turbine to use excess air,
above the optimal stoichiometric ratio
to burn the fuel. The excess oxygen may
be used for supplementary firing of fuel
gas after the gas turbine to increase flow
or temperature of the steam generated.
Supplementary burners also are called
“duct burners,” since they are located
either in the inlet duct or in the inter-
Figure 3-78 | Heat recovery steam generator pressure part modules bank duct after the superheater section.
The duct burners in HRSGs typically
use natural gas as fuel, but they also
Ź}Ź~
ČńŇčŋŌĞĤŶĥ may fire oil.
ʼnňŅʼn
In contrast, locations with high labor The duct burner consists of burner
costs usually can provide equipment elements designed to withstand flue gas
for lifting and prefer the largest, heavi- temperatures and typically are made of
est, most shop-assembled modules stainless steel. The number of burner
that can be economically transported. elements is dependent on firing capacity
A range of options is available for and the duct dimension. Each burner
pressure part modules. The same element comes equipped with flame
HRSG can have pressure parts stabilizers or baffles to ensure that
shipped in 20 ton bundles (Fig. 3-77), proper flame length is maintained. The
Boilers | 3-97
These baffles must address leakage at locations in the HRSG steel struc-
sidewalls, between side-by-side mod- ture. In so doing, seismic loads from
ules, above upper headers and below the finned tubes are transferred to the
lower headers, and in gaps between main steel at the tube restraint eleva-
headers. tions. Figure 3-81 illustrates a typical
HRSG tube restraint.
Tube Supports
Finned tubes are held in place by HRSG Structure
passing them through tube restraints The steel structure consists of two
located at several elevations vertically main sections: the inlet duct and the
in the boiler. Tube restraints are sized HRSG. Each section is made up of a
to the specific diameter of the finned series of parallel steel moment frames
tube. The restraints serve to restrict connected together by a continu-
lateral movement of the tubes and to ous, ¼ inch (6 mm) thick steel duct
limit vibrations induced by exhaust casing. A substantial portion of the
gas flow. Vortices are produced at pressure part mass is supported at the
regular frequencies, depending on the roof elevation. The steel frames serve
gas properties and velocity at each as the vertical load carrying system.
tube row, when the gas turbine ex- Occasional loads, wind and seismic,
haust flows over the finned tubes. The are resisted by moment frames and
resulting differential pressure loads casing plate in transverse and gas
excite the finned tubes. If the excita- flow directions, respectively. Internal
tion frequency is close to a natural and external HRSG components are
frequency of the tubes, destructive attached to wide flange roof beams,
vibrations will occur. Tube restraints which are then supported by wide
are located such that the first natural flange columns that bear on a founda-
frequency of the tubes is significantly tion system.
higher than the vortex shedding fre-
quency. Tube restraints are designed Cold Casing Design
according to the mechanical loading The HRSG incorporates “cold casing”
placed on them, and materials are construction. Internal insulation is
selected based on the local exhaust selected to ensure the external casing
gas temperature. The tube restraining has a typical surface temperature of
system also includes restraint stops, 140°F (60°C), allowing the use of
which coincide with main column carbon steel casing plate and stiffeners
and minimizing the need for expan-
sion provisions for the overall boiler
structure. Insulation materials are se-
lected for low casing temperature, safe
and efficient operation, and minimum
maintenance. The floor of the HRSG
is sloped, has drains, and insulation
suitable for water washing.
An internal liner is utilized at the
internal surface of the insulation
Figure 3-81 | Heat recovery steam generator tube restraint
Section #2
Section #3
Section #4
Section #5
means, including, but not limited to,
sound barriers and enclosures.
Cavity #3
Cavity #4
0.6
Harps “B”
Drums
Inter-Bank Burner
1.0
HRSG evaporative circuits incorporate
0.4
Cavity #1
Cavity #5
Cavity #2
steam drums, which ensure steam purity
0.6
Harps “A”
1.0 and reduce the potential for water surges
0.4 encountered during cold starts. Figure
3-84 shows a typical drum arrangement.
On the water side, incoming feedwater
Baffles, with Extensions, in Low Baffles Between
Temperature Region of Tube Bank, to
is distributed along the length of the
Modules 100% Solid
Minimize Risk of Resonance from HP drum by a feedwater distribution header.
Evaporation Tubes Nozzles in the distribution header direct
the feedwater in the downward direction
Figure 3-83 | Acoustic baffle positioning to minimize turbulence and aid in cir-
culation. The water in the drum is then
Clean Combustion Technologies directed to downcomers.
Figure: 03.83/ km stack. These silencers are effective at On the steam side, the saturated
PMS5415
Saturated Steam K attenuating noise in the exhaust gas steam–water mixture from the evapo-
Steam
Separator Feedwater of the HRSG system. Noise emanat- rative circuit enters the drum through
ing from HRSG walls is addressed risers and is directed through separa-
via duct design or shrouding. Adding tion devices and dryers to remove the
mass to duct walls can attenuate water from the steam. Dry saturated
noise and usually is accomplished steam leaves the drum exiting through
by increasing the thickness of the saturated steam outlets located on the
Water
external casing plate. To a greater top of the drum.
extent, however, shrouds are employed Other aspects of drums include:
to attenuate wall noise from HRSGs.
Typically, shrouds can be designed as • Construction: The drum is fusion
Steam Steam side-wall shrouds that use existing welded, fabricated from carbon
HRSG structural steel for support or steel plate, and equipped with two
Downcomer (not shown)
as full, free-standing enclosures that manway openings, one at each end
Figure 3-84 | Typical heat recovery steam
effectively cover the HRSG sidewalls of the drum.
generator drum arrangement and roof. • Downcomers: Dedicated external
Clean Combustion Technologies Aside from gas turbine exhaust, the downcomers are used for proper
Figure: 03.84/ (STM) HRSG designer also must consider a circulation in the evaporative steam
PMS5415 K number of intermittent noise sources, generating sections.
including start-up vents, safety valve • Safety Valves: The drum is fitted
exhaust systems, pumps, and other with safety valves, each rated at a
operating equipment. Vents typically percentage of the drum’s design
are addressed with commercially avail- evaporation rate as required by
able silencing equipment, sized for the boiler design code.
Boilers | 3-101
may not remove undesirable gases. It for release of exhaust gas to the atmo-
also is required that the flow entering sphere. For large HRSG installations,
the HRSG be as uniform as possible stack diameters range from 20 to 23 ft
to ensure good thermal performance. (6 to 7 m). Diameter, too, is depen-
The inlet ductwork is fabricated as dent on air permit requirements but
shop-assembled panels consisting of also includes consideration for veloc-
a carbon steel outer casing, insula- ity and draft loss as critical design
tion, and internal liner. Design and inputs. Typically, HRSG stacks are
fabrication is similar to that of the fabricated from carbon steel, because
casing panels. operating temperatures range from
The outlet duct serves as the transi- 212 to 300°F (100 to 150°C), with
tion from the HRSG to the exhaust operating pressures in the range of
stack. The typical geometry is a 2˝ WG to 4.8˝ WG (5 to 12 mbar).
Figure 3-86 | Compact heat recovery steam reducing cross section as the exhaust The stack is bottom supported from
generator inlet duct gas makes the turn from horizontal its base using a series of anchor
Clean Combustion Technologies to vertical, coupled with a geometry bolts around a stiffened base ring.
Figure: 03.86/ (mh) change from rectangular cross section The stack is not normally insulated,
PMS5415 K
(HRSG) to a circular cross section but insulation can be used in cold
(exhaust stack). Figure 3-87 shows environments or where corrosion may
a representative outlet duct. The be a concern.
ductwork from the HRSG to the Stacks may include a variety of ad-
stack is carbon steel casing, reinforced ditional equipment:
with stiffeners, provided with ex-
panded metal standoffs for personnel • Silencer: Reduces gas turbine
protection from the hot surfaces. The exhaust noise to acceptable levels
ductwork is complete with necessary • Damper: Provides protection for
doors and expansion joints. Internal internal parts during unit standstill
insulation and liners are not typi- periods
cally used in outlet ductwork, because • Platforms: Provides access to emis-
exhaust gas temperature typically is sion testing locations
212 to 300°F (100 to 150°C) when it • Access doors: Provides access to the
enters the outlet duct. HRSG or damper
• Trolley: Can be used for painting
Steel Stack trolleys or other accessibility devices
The final component in the exhaust gas
system of the HRSG, the stack directs
turbine exhaust gas to the atmosphere.
Emissions Reduction
The stack is the primary location for
emissions compliance testing and Selective Catalytic Reduction
validation. Figure 3-88 shows a typical (SCR) Process
HRSG stack. The process of SCR involves the cat-
Stacks are configured as straight alytic reaction of ammonia (NH3),
cylinders, with typical heights from which is injected into the flue gas
130 to 200 ft (40 to 60 m) for HRSG containing NOx to produce molecu-
applications. Stack height is largely lar nitrogen (N2) and water vapor
dependent on air permit requirements (H2O). These reactions take place
Boilers | 3-103
broken, allowing the free atoms to turbines will drive increasingly efficient
recombine, forming CO2: combined cycles. For the HRSG,
this means that exhaust gas flows will
2CO+O2 → 2CO2 increase, requiring larger boilers. Steam
Equation 3-7 cycle parameters also will increase,
requiring higher steam temperatures
The catalyst is composed of and pressures from the HP section of
stainless-steel foil substrate, which is the HRSG.
corrugated and coated with alumina The HRSGs operating at higher
wash coat. The wash coat is impreg- pressures (2,300–3,200 psi [160–220
nated with platinum group metals. bar]), requiring that the HP evapo-
The catalyzed foil is folded and en- rator be based on once-through
cased in welded steel frames to form circulation principles, are an emerging
individual modules weighing approxi- trend. Higher steam temperatures
mately 44 lb (20 kg). The individual (1,100–1,160°F [600–625°C]) will
CO catalyst modules are installed require use of the same high-strength
in the frame structure within the alloys that are currently being de-
HRSG casing. The CO catalyst frame veloped for advanced, supercritical,
is installed in the HRSG inter-bank coal-fired boilers. The combined cycle
upstream of the SCR catalyst, where plant will continue to support a range
the flue gas temperature is predicted of load applications and often will
to be between 500 and 1,000°F (260 be required to cycle on and off daily
and 540°C) for maximum efficiency as well as to support grid frequency
of CO destruction. with rapid-response loading charac-
teristics. The mechanical design of
Future Trends HRSGs must continue to evolve to
Gas-fired combined cycle technol- improve fatigue tolerance, even as the
ogy is available now for base load, for size, temperature, and pressure of the
peaking, and as a bridge between the boiler continues to increase.
older, coal-fired steam plants of the
past and the high-efficiency, coal-fired
plants of the future designed with Kevin Taugher
George Stamatelopoulos
CO2 capture and sequestration as Scott Darling
an essential consideration. For the Glenn Selby
near term, the dominant technolo- Donald Bairley
Jesse Bolinger
gies for production of electricity that Denis Bruno
will support reduced CO2 emissions Mark Malo
will be nuclear, gas-fired combined Thomas Mastronarde
Edward Ortman
cycle, large-scale wind farms, and
Harpreet Singh
solar applications. Continued techni- Charles Tsirovasiles
cal advancement of utility-scale gas Anisa Brown
1. Combustion Engineering, Inc., “TVA Utility AFBC Project, Phase 3. D. Kuni and O. Levenspiel, Fluidization Engineering. Malabar, FL: Robert
II, Final Report: Volume 1—200-MW Demonstration Plant Final E. Krieger Publishing Co., Inc., 1969.
Design; Volume 2—800-MW Commercial Plant, Proposed Design
4. B. Lecker and L.-E. Amand, “Emissions from a Circulating and a Stationary
and AFBC Research and Development Requirements”; prepared under
Fluidized Bed Boiler: A Comparison,” Proceedings of the Ninth International
TVA Contract 51863A, January 18, 1981. Windsor, CT: Combustion
Conference on Fluidized Bed Combustion, Boston, MA, May 4–7, 1987.
Engineering, 1981.
5. E. J. Gottung et al., “Design Considerations for Circulating Fluidized Bed
2. S. L. Goodstine et al., “Industrial Application of Fluidized Bed Combus-
Steam Generators,” Proceedings of the Tenth International Conference on
tion, Phase 1, Task 4—Sub-Scale Unit Testing and Data Analysis, Final
Fluidized Bed Combustion, San Francisco, CA, May 1–4, 1989.
Report, Volume 1,” prepared for the U.S. Department of Energy under
Contract EX-76-C-01-2473, December 1979; published by the U.S.
Department of Energy (NTIS) DOE/ET10389-T3.
3-105
Steam Turbines for Electric
Power Generation
Chapter Four
The first demonstration of a steam turbine was the Hero engine, which was developed as a
novelty. Although first invented two millenia ago, useful application of the steam turbine did
not occur until the nineteenth century, with the work by pioneers such as Gustav Delaval and
Charles Parsons.
Steam turbines evolved at the end tively compact generating device—the generates power. A bladed rotor is
of the nineteenth century as the most steam turbine generator. Therefore, shown in Figure 4-2.
common prime mover in the gen- the steam turbine generator is the From humble beginnings as a
eration of electricity. Steam turbines preferred generating device for central desktop novelty, modern turbines
also commonly are used for marine station applications. have grown to sizes of up to 1,600
propulsion, industrial drives, and large In addition to electric power MW or more for nuclear applications.
pump drives. Because Alstom provides generation, steam turbines play an Fossil fuel applications of up to 1,300
steam turbines primarily for generation important role in the supply of steam MW are operating in North America.
of electric power, this chapter focuses and heat for industrial and district These machines may be more than
principally on this application. heating applications. By passing steam 100 ft in length and weigh thousands
Steam turbines convert the heat through the steam turbine and then of tons. Large machines typically
energy from steam into mechanical using the exhaust steam for other consist of multiple casings (modules)
energy. The source of heat to generate applications, both steam and electric- arranged on a single shaft (tandem
the steam is irrelevant to the steam ity are cogenerated, resulting in very compound) connected directly to a
turbine, making these turbines appli- efficient use of energy. generator. Less common turbo sets
cable to almost any fuel source. Heat consist of multiple shafts driving one
sources range from coal, biomass, The Basics or more generators (cross compound).
solar, nuclear, natural gas, and com- The steam turbine converts mov- Figure 4-3 shows an example of a
bustion turbine exhaust (combined ing steam to mechanical energy tandem compound steam turbine. A
cycle) to waste products. aerodynamically, through a circu- typical utility reheat turbine would
The steam turbine has many lar arrangement of stationary and consist of a high-pressure (HP) sec-
advantages for electric power genera- rotating airfoils (blades) as shown tion, an intermediate-pressure (IP)
tion. Because turbines favor relatively in Figure 4-1. Each stationary and section, and one or more low-pressure
high speeds, they are suitable for rotating row set is called a “stage.” The (LP) sections (Fig. 4-4).
direct drive of alternating current stationary blades direct the steam Because utility turbines typically
(AC) generators. High speeds also into the rotating blades. Lift on the are connected directly to the genera-
allow large flows and result in a rela- rotating blades turns the rotor and tor, their speed is governed by the
4-
4- | Clean Combustion Technologies
Steam Cycles
5,400 psia
1,292 °F
Steam Cycles for Coal/Fossil
4,060 psia 1,328 °F FUEL Applications
1,166 °F Live steam temperatures continue
–15
1,202 °F
4,060 psia R&D Ongoing to increase with the development of
1,112 °F (Thermie) in
Europe, U.S.
advanced casing, blade, and rotor ma-
1,148 °F
Reduction in Heat Rate, %
–10
4,060 psia Ni-base terials. Contemporary live and reheat
–8 1,076 °F steam temperatures for supercritical
–6 Supercritical 1,112 °F
3,500 psia R&D Ongoing plants often exceed 1,100°F (595°C)
–4 Subcritical 1,000 °F/1,050 °F (COST) in
Current
Europe, U.S.
for large (>500 MW) utility units.
–2 2,400 psia Market
1,000 °F/1,000 °F Mature Technology Introd.
Development programs in Europe are
0 Europe, Japan, U.S., expected to achieve inlet temperatures
Mature Technology Korean Republic. Ready Japan, Japan
World Market for Market Introduction Denmark, of 1,290°F (700°C). Materials studies
in Asian countries Germany in the United States are looking at
1950s 1960s 1970s 1980s 1990s 2000s 2010s
1,400°F (760°C). These advanced cycle
Time parameters result in a steam cycle effi-
ciency of greater than 50% and a plant
Figure 4-6 | Development of main steam temperatures and pressures for steam turbines efficiency of 48% (based on Lower
(HR, Heat Rate) Heating Value [LHV]) (Fig. 4-7).
Clean Combustion Technologies For smaller units (<400 to 500
Figure: 04.06/ (Laserwords)
PMS5415 K the expansion line by reheating steam MW), the gains in efficiency for
55 from the HP turbine exhaust before supercritical cycles are less than those
Combined
Cycle readmitting into the IP turbine. for larger units. Traditionally, cycles
50
Hard
Reheating HP exhaust steam typically for medium-sized units have remained
45 results in a 5% increase in cycle effi- subcritical. Increased temperatures of
�th
Coal Fired
ciency and has the benefit of reducing 1,050°F (565°C), however, are common.
40
Lignite moisture in the LP turbine exhaust. Subcritical cycles typically have six
Fired
35 By the end of the 1960s, typical or seven feedwater heaters, and super-
1970 1980 1990 2000 2010 subcritical cycles had inlet steam critical cycles typically have seven or,
Year
parameters of 2,415 psia and 1,000°F more commonly, eight. If an extraction
(16.65 MPa and 540°C) and reheat is taken from the HP turbine, this is
Figure 4-7 | Plant efficiency has improved
temperatures of 1,000°F (540°C). In known as a HARP (Heater Above
significantly
Clean Combustion Technologies the United States, these cycle param- Reheat Pressure) cycle. Typical final
Figure: 04.07/ (Laserwords)
PMS5415 K eters were typical of utility coal plants feedwater temperatures range from
through the 1990s. approximately 500°F (260°C) for
During the 1960s, supercritical small subcritical cycles to nearly 700°F
and double reheat cycles also were (370°C) for some high-temperature,
introduced. Supercritical cycles had supercritical applications.
some acceptance in the United States
but generally were used in Europe and Combined Cycle Applications
Japan because of the higher efficiency. During the 1980s and 1990s, de-
Double reheat units have been limited velopment of larger, more efficient
in application since that time, how- combustion turbines resulted in the
ever, because of their higher capital common use of heat recovery steam
costs and more complex operation. generators and steam turbines to
Steam Turbines for Electric Power Generation | 4-
extract energy from the combustion cycle efficiency. The variation of the
turbine exhaust. In this way, plant cycle design condenser pressure may be
efficiencies approaching 60% (LHV) more significant than cycle changes
can be achieved. from subcritical to supercritical
Because ample low-grade energy is steam parameters. By lowering the
available in the exhaust stream of the exhaust pressure, the expansion line
combustion turbine, a combined cycle of the steam extends further into the
typically does not use regenerative moisture zone, increasing the exhaust
heating (feedwater heating extraction). moisture. At lower pressure, the steam
Conversely, the available low-grade en- also has a much higher specific volume,
ergy can be used to generate additional so the LP turbine exhaust area must be
IP and LP steam that can be admitted much larger. Therefore, the LP turbine
to the IP and LP turbine. As a result, a must be designed and sized to handle
combined cycle steam turbine applica- the volume flows for the design cycle.
tion may have 35% more mass flow Aspects of the LP turbine component
at the LP exhaust compared with the design will be discussed later.
HP inlet. A conventional boiler steam
cycle may have 35% less steam flow at Boiler Feed Pump Drive
the LP exhaust because of feedwater (turbine or electric motor)
heating extraction. Other associated cycle equipment also
affects the cycle efficiency. The boiler
Non-reheat Applications feed pump drive can be a major plant
For smaller industrial plants or other load—in some cases, more than 40
cycles with generally lower available MW (mechanical) for a large super-
steam temperatures, turbine inlet critical unit. Therefore, this drive must
pressures are limited to approximately be carefully considered in the overall
2,000 psig (14 MPa(s)) because of the plant concept.
high moisture content (i.e., wetness) Boiler feed pumps are driven by
in the exhaust steam. Typically, these either a turbine or an electric motor.
units are smaller and use lower steam Motor drives may be variable-speed,
parameters for ease of operation and synchronous drives or conventional
reduced capital cost. AC motors driving the pump through
a variable hydraulic drive. Synchronous
Component impacts on motors are more efficient than con-
efficiency and design ventional AC but have higher capital
cost. Electrical motors have advantages
Effect of Exhaust Pressure
during facility start-up, provided that
The turbine exhaust pressure (con- electric power is available. For smaller
denser pressure) profoundly affects facilities, motor-driven drives are a less
both the cycle efficiency and the expensive solution than turbine drives.
turbine component design. By lower- Because the motor requires electric
ing the exhaust pressure, more energy power, however, the main generator
can be extracted from the steam. This must be larger.
results in less energy being rejected Boiler feed pump turbine (BFPT)
to the condenser and in a higher drives more commonly are used for
4- | Clean Combustion Technologies
large units. The BFPT is driven with rotating blade row. Energy is transferred
steam extracted from the cold reheat by momentum transfer and by accelera-
(HP turbine exhaust) or from the IP tion of the steam out of the rotating
turbine. With increasing load, more blade (also called a “nozzle effect”).
steam is available from the main tur- Impulse blading generally is much
bine, matching the steam available with heavier than reaction blading, and
the steam required by the BFPT. Use more energy can be extracted per stage
of a BFPT reduces the LP steam flow, from the steam. In practical terms, this
in turn reducing the exhaust losses in means that fewer stages are required.
the LP turbine (exhaust losses will be Because of the higher pressure drop
discussed later), making the cycle more across the nozzle assembly (called a
efficient. Because the BFPT drive does “diaphragm”), however, impulse stages
not require electric power, the genera- require a reduced diameter at the
tor can be smaller. For very large units, stationary stage to reduce leakage.
BFPT drives can be less expensive than Additionally, a heavier root attachment
motor drives. is required to hold the heavier blade.
Figure 4-8 shows the relative difference
Reaction/Impulse Blading between the types of blade technology.
The more detailed theory of blade type Reaction blading generally is lighter
will not be discussed in this text, but but requires more stages. With reac-
some explanation regarding the impact tion blading, a drum-type rotor can
of reaction and impulse blading is nec- be used (Fig. 4-9), and both rotating
essary for understanding steam turbine and stationary blades can be directly
technology. By definition, a purely mounted in T-slots in an inner casing
impulse stage has a pressure drop only or blade carrier. Because of the pres-
across the stationary blade (sometimes sure drop across the rotating blades,
called a “nozzle”). Pressure drop across substantial axial thrust is created on
the rotating blade (sometimes called the rotor. Therefore, reaction turbines
a “bucket”) is, by definition, zero. Im- typically require a balance piston for
pulse stages, as the name implies, rely single-flow sections.
only upon the transfer of momentum Comparing the two technolo-
to the rotating element. gies (impulse blading and reaction
A purely reaction stage has an equal blading), both of which have been
pressure drop across the stationary and successfully employed for more than
100 years, usually is more theoretical
Typical Low Reaction (Impulse) Typical Reaction Type than practical. To summarize, howev-
Fixed & Rotating Blades Fixed & Rotating Blades er, impulse technology requires fewer
stages, which can result in a shorter
section length. Reaction stages tend to
have less leakage loss across each stage,
but some leakage results from the bal-
ance piston.
Control stages (discussed below)
utilize impulse technology. Down-
Fixed blade diaphragm Rotating blade Fixed blade Rotating blade stream blading for Alstom designs
typically use predominantly reaction
Figure 4-8 | A typical impulse stage has a heavier stationary element (often a diaphragm
assembly), whereas Clean
the reaction stage Technologies
Combustion has similar profiles for both stages
Figure: 04.08/ (TDB)
PMS5415 K
Steam Turbines for Electric Power Generation | 4-
Three-Dimensional Blading
Early blade profiles were uniform
throughout the length of the blade.
With the evolution of three-dimensional
modeling technology and multi-axis
machining capability, the ability to
manufacture blades with different
curvature and profile at the root versus
the tip became possible. Because the
velocity at the tip is different than the
velocity at the root, optimum perfor-
mance can only be obtained by varying
the profile along its length (Fig. 4-10).
This is particularly important with Low Reaction Reaction
lower-pressure blades, for which the (impulse)
thus, the output of the turbine. Sliding- achieved compared to a typical control
and constant-pressure control modes stage design. Full arc admission designs
will be discussed in more detail later. can be used with constant pressure.
They are less efficient in this mode at
Partial Arc Admission part load, however, because all of the
By sequentially opening the control steam is throttled.
valves, steam can be admitted to only
one arc segment of the turbine inlet an- Sliding- and Constant-
nulus. At the inlet, only the blades being Pressure Operation
struck in areas where the valve is open Sliding- and constant-pressure operat-
will generate power. In this mode, it is ing modes refer to the pressure within
possible to open some valves entirely, the boiler. In the sliding-pressure
negating the pressure drop across that mode, the pressure within the boiler
valve (i.e., no throttling), while other is allowed to increase or decrease with
valves are completely closed. This main- load changes. Constant-pressure opera-
tains the steam velocity—and, thus, tion requires that the pressure within
efficiency—across this stage. Partial arc the boiler be kept constant. It also is
admission requires the use of a control possible to operate using both of these
stage. The control stage uses impulse modes (hybrid operation).
blades, so there is minimal pressure In sliding-pressure operation, live
drop across the rotating blades. If a steam pressure typically is maintained
pressure drop were to occur across the by a variable-speed, boiler feed pump
control stage, the steam would tend to drive (turbine or motor). The heat rate
flow around the inlet annulus rather advantage here is that at lower loads,
than where it is directed by the inlet the boiler feed pump power require-
nozzles when operating with only one ments are substantially reduced. Most
inlet section open. combined cycle and modern supercriti-
cal applications use sliding pressure.
Comparing Partial Versus Because there is no need for partial arc
Full Arc Admission control with sliding-pressure operation,
Each mode has advantages shown in these steam turbines do not normally
Table 4-1. Full arc admission designs have control stages.
not using a control stage result in a Subcritical, drum-type boilers
higher full load (valves wide open) typically have constraints on the
efficiency. An improvement of approxi- operating pressure of the boiler, so
mately 0.2 to 0.5% in heat rate can be constant-pressure or hybrid operation
is desired. In the United States, most
TABLE 4-1 | Partial Versus Full Arc Admission
older, subcritical fossil fuel units use
constant-pressure operation in com-
Advantages Disadvantages
bination with partial arc admission.
No throttling across fully open valves Often, however, larger, base load units
Lower efficiency of first
when at part load
Partial Arc stage have no control stage to achieve better
Higher part load efficiency for
(Nozzle Control) Higher cost (more
subcritical cycles full load performance.
control valves)
Faster response to step load changes In some cases, the turbine valves are
Lower efficiency in operated partially closed to respond
Full Arc Higher full load efficiency
throttle control mode quickly to load increases and frequency
Steam Turbines for Electric Power Generation | 4-
control. The required step change pressure down or up to obtain the flow
capability will vary with electrical grid necessary to achieve the desired output
requirements. The difference between at the valve point (valves fully open or
the valves wide open load and the closed). This combined use of sequen-
operating load is called the “throttling tial valve control and sliding pressure is
margin.” Without a control stage, all called “hybrid control.” For units operat-
control valves open in parallel to admit ing at a variety of load conditions, this
more steam; with a control stage, con- may be the most efficient mode across
trol valves open sequentially. Sequential the load range (see Fig. 4-11).
operation with partial arc admission
gives a finer degree of control and, Mechanical Design
hence, better response. Residual energy of the Turbine
in the boiler and piping provide the
Welded rotors
instantaneous energy for a step load
change. At the new load point or for The rotors of Alstom steam turbines
ramp load changes, steam pressure (and gas turbines) typically are built
must be maintained by the boiler feed from forged disks that are welded
pump during this response. together using a combination of the gas
Sequential operation also yields a tungsten arc and submerged arc welding
series of valve points where all valves procedures (Fig. 4-12). After welding,
are either fully open or closed. Opera- the shafts are stress relieved under con-
tion at the valve point is more efficient trolled temperature conditions.
than at load points adjacent to it (Fig. The resulting rotor is a robust,
4-11). Therefore, operation at these drum-type rotor. Stresses at the welds
points is desirable. If the valve points at and throughout the rotor are low for
a nominal pressure do not exactly match this design compared with disk or
the load point, it is possible to slide the mono-block designs. This allows the
Change of STG Net Heat Rate vs. Load Subcritical Unit Change of STG Net Heat Rate vs. Load Supercritical Unit
5.0% 5.0%
Constant Pressure Mode Constant Pressure Mode
4.5% (valves open sequentially) 4.5% (valves open sequentially)
Hybrid Mode Hybrid Mode
4.0% 4.0%
(3 valves fully opened and (3 valves fully opened and
1 valve closed) 1 valve closed)
3.5% 3.5%
Hybrid Mode Hybrid Mode
Net Heat Rate Change
3.0% (2 valves fully opened and 3.0% (2 valves fully opened and
2 valves closed) 2 valves closed)
2.5% 2.5%
Sliding Pressure Mode Sliding Pressure Mode
2.0% (without control stage) 2.0% (without control stage)
1.5% 1.5%
1.0% 1.0%
0.5% 0.5%
0.0% 0.0%
50% 60% 70% 80% 90% 100% 50% 60% 70% 80% 90% 100%
–0.5% Load –0.5% Load
Figure 4-11 | Efficiency varies with turbine inlet configuration and valve operation. Maximum efficiency occurs at the valve points (STG; steam
turbine generator)
The HP Turbine
Austenitic Blade Material
The HP section of the turbine
expands the live steam from the
boiler superheater outlet down to the
Intermediate-Pressure Turbine
reheat pressure (cold reheat), where
High Temperature Resistant Material
it is piped back to the boiler reheater.
Conventional The high temperatures and pressures
Rotor Materials combine to create high stresses in the
blading, inlet valves, and inner casings.
Because of the high pressures, thicker
materials are required. Therefore,
care must be taken to accommodate
thermal stresses induced by transient
loading conditions.
Figure 4-17 | Steam is guided smoothly into the blading via inlet scrolls (HP, high-pressure; IP, intermediate-pressure; LP, low-pressure)
Intermediate-Pressure Turbine
Steam In
Direction of
Steam Flow
Through the
Turbine
High-Pressure Turbine
condenser. Issues with this large LP energy back to the steam—that is,
exhaust volume flow are two-fold: energy lost to the condenser. These
efficiency, and choking. High-volume losses, combined with the kinetic en-
flow through a fixed area translates ergy losses discussed earlier, represent
into high velocity. The kinetic energy exhaust losses.
leaving the last-stage blades is lost to
the condenser, reducing cycle efficiency. LP Selection
(Note that the kinetic energy in the Exhaust losses can be plotted versus
steam leaving any upstream stage can volume flow, as shown in Figure 4-21.
be recovered in the downstream stages. The right side of the curve (high vol-
This lost kinetic energy is the major ume per flow) parabolically increases
component of the combined exhaust because of increasing kinetic energy
losses that will be discussed later.) losses. Typically, however, the left side
Steam, like any gas, has a sonic (low volume per flow) is more vertical
velocity (speed of sound). As the steam and less defined. Choking will occur
velocity reaches the sonic velocity, a at the extreme right end of the curve.
shock wave forms in the blading. Once Maximum efficiency occurs where
the shock wave is formed, steam flow the exhaust losses are minimized. For
cannot be increased by downstream a given steam flow, exhaust volume
conditions. This point is appropri- flow will increase with decreasing
ately called the “choke point,” and it exhaust pressure.
represents a maximum volume flow Because the LP exhaust pressure
condition for a given LP turbine. will vary seasonally with the condens-
If the volume flow and exhaust er cooling medium (water or air), the
steam velocity are too low, turbulence size of the total LP annulus area must
and uneven flow result along the length be selected to match that experienced
of the last-stage blades. At extremely throughout the year. Ideally, the LP
low flows, partial recirculation of steam exhaust area should be selected to op-
may occur, typically along the rotor. erate at or near the minimum exhaust
These phenomena mechanically add loss point. Because the left side of the
curve is more vertical, the LP should
be selected to avoid operation too far
LP Turbine Exhaust Losses below the minimum exhaust loss.
The LP annulus area can be
Increasing Losses due
to Increasing Velocity increased by increasing the last-stage
blade length or by increasing the
number of flows. Increasing the LP
Increasing Losses due to
annulus area effectively reduces the
Exhaust Losses
Electronic Electronic
Safety Controller
System
Hydraulic Supply Oil
2-of-3
SSA
SSA
STA
STA
Accumulator Tripping Unit E E
H H
Live
Steam
2 � 100% Filters
Hydraulic
Drain Oil
HP
Pressure Control Valve M M IP
HP Exhaust
2 � 100% Lube Oil and
Steam
Hydraulic Pumps Jacking Oil Supply
Reheater
Figure 4-24 | The control oil system (hydraulic system) supplies the valve actuators with pressurized fluid to actuate the valves and control the
turbine. If control oil pressure is lost, the stop valves will automatically close (E, open; H, closed; HP, high pressure; IP, intermediate
pressure; SSA, control valve; STA, stop valve) Clean Combustion Technologies
Figure: 04.24/ (Laserwords)
PMS5415 K
Some of these systems are included turbine generator system, the control
in the base control system. Others, system also is the main interface with
such as vibration monitoring and the plant distributed control systems
synchronization, may be in specialized (see Chapter 7). Modern systems
subsystems. Critical functions and also can monitor turbine stress and
parameters will have triple redun- turbine life. Communications systems
dancy, with two out of three voting also allow remote monitoring of sys-
among the channels to ensure safety tems and condition assessment. All of
while preventing inadvertent trips. these can generate increased reliability
Less critical parameters may have only and equipment lifetime.
double redundancy, and non-critical
parameters may be monitored only on Turning Gear and
a single channel. The central processor Jacking Oil Systems
itself is redundant on the same bus to Large steam turbines and generators
ensure system reliability. have extremely heavy rotors with long
The speed of the system is impor- bearing spans. Over long periods of
tant in regulating the turbine. The storage, these rotors can develop a set
control system must respond quickly sag. Also, at elevated temperatures as
to load and system requirements the turbine cools, thermal gradients
as well as to emergency situations can cause a bow in the rotor. There-
requiring systematic, controlled fore, turbines have a turning gear to
shutdown. As the “nerve center” of the maintain a low rpm that keeps the
4-18 | Clean Combustion Technologies
temperature of the rotor constant while als that can survive the environment
the unit cools. for an extended lifetime. Currently,
Turning gears have various me- materials are available to accommodate
chanical means to engage with the inlet temperatures of approximately
rotor. Some units utilize a ratchet 1,150°F (620°C). Efforts are underway,
mechanism; others use a pinion gear to however, to develop steam cycles using
turn the rotor. These can be motor or inlet temperatures of 1,290°F (700°C),
hydraulically driven. Smaller units also and materials programs are looking at
have manual backup mechanisms. materials for 1,400°F (760°C). These
As long as a film of oil is main- elevated temperatures may require
tained in the bearings, the force steam or water cooling. Other energy
required to maintain rotation is sources may introduce new require-
relatively small. At operating speeds, ments as well. In addition to traditional
the rotation of the shaft generates energy sources, such as nuclear,
enough hydrodynamic force such that fossil, and gas, new technologies may
the required lube oil pressure is mini- introduce further requirements to this
mal. In case of an extended outage or “mature” technology.
from a standstill, more oil pressure is
required than can be supplied from Generators
the auxiliary lube oil pumps. The jack- The generator (or turbogenerator)
ing oil system provides HP oil to the converts mechanical energy supplied
bearings during start-up or turning by the turbine(s) into electrical energy.
gear operation. The generator consists of a stationary
In case of a black trip and extended active part, the stator; a rotating part,
outage, special consideration must the rotor; and the generator externals.
be given to maintaining power to the The electrical energy is generated in
turning gear and jacking oil systems the key component of the generator,
to protect the turbine. This can be the stator winding (armature wind-
accomplished with battery-powered, ing) (Fig. 4-25, 1), which connects to
DC motor-driven pumps or via an the generator terminals (bushings)
emergency diesel generator. (Fig. 4-25, 2).
The stator winding outputs a
The Future of Steam Turbines three-phase AC current at the power
Because steam turbines can use virtu- frequency (50–60 Hz) and a voltage of
ally any high-quality energy source 10 to 27 kV, depending on the power
to develop power, they will remain an rating. Bus ducts (Fig. 4-25, 3) connect
important part of our electric power the generator terminals to the step-up
generating portfolio for the foreseeable transformer, which adapts the power
future. Increasing demands for better to the grid voltage (Fig. 4-25, 4). By
efficiency are driving steam parameters the synchronous generator principle,
to higher levels. For the conventional the speed of the turbo set is locked to
steam cycle, this means higher inlet the power frequency. The speed for
temperatures and pressures. full-speed generators is 3,000/3,600
For the steam turbine, increased rpm; that for half-speed generators is
steam temperatures require materi- 1,500/1,800 rpm.
Steam Turbines for Electric Power Generation | 4-19
Electrical Energy
3 4
Field Current
2 to Rotor Via
1 Slip Rings
5
Mechanical Energy
of concern are voltage and frequency. power goes up and down around some
The plant parameter to influence the average power. The average is called
frequency is the active power deliv- the “active power,” and the variation
ered into the grid. In the grids, which around the average is the reactive power
usually are large in size, the turbine (commonly referred to as volts-amperes
valve is governed by continuously reactive or var). An analogy would
comparing the actual grid frequency be climbing up and down a hill. In
with a reference value (50/60 Hz) climbing up and down a hill without
(Fig. 4-27). carrying anything, no net work is
If the demand for power from the done, and no active power is required.
grid increases, then the speed of the If, however, some cargo is carried up
turbines supplying the grid slows the hill and left there, then net work
down. Because the synchronous is done, and active power is required.
generator locks the speed with the grid Reactive power is required to get up
frequency, this frequency decreases. the hill; this power is recovered in the
The response of the turbine governor absence of friction going back down the
is to open the steam or fuel valves to hill. Active power cannot be delivered
increase the power output from the without reactive power. Thus, reactive
turbines and so restore grid frequency power is needed to start motors or acti-
to the reference value. vate devices, and is a key consideration
The plant parameter to control the in the design requirements of a steam
grid voltage is the reactive power deliv- turbine and generator.
ered into the grid. The reactive power The generator excitation is respon-
results from the current component, sible for the control of reactive power.
which is out of phase with the voltage. The excitation current can be controlled
Reactive power normally is associated such that the grid current is in phase
with AC systems. In an AC system with the voltage (power factor = 1).
(single phase), the voltage and the cur- Increasing the excitation current results
rent alternate up and down (60 times in a phase shift of grid current to volt-
per second in the United States and age (power factor < 1; overexcited or
50 times per second in Europe). This lagging current mode). Decreasing the
excitation current results in a reverse
phase (power factor < 1; underexcited
Turbine Valve Generator Excitation or leading current mode). Overexcited
mode tends to increase the grid voltage,
and underexcited mode tends to lower
it. Stability of the turbo set on the grid
is heavily influenced by the excitation;
usually, an overexcited mode makes the
system more stable. This is the reason
why, in case of a disturbance on the grid,
the generator excitation is driven to a
ceiling value for a short duration. In case
Turbine Generator of severe disturbances on the grid, the
generator may fall out of step; in that
Figure 4-27 | Turbine valve and generator excitation
case, the generator is switched off the exhausts it back to ambient. The ad-
grid and resynchronized. vantage is that the generator efficiently
The control for generator excitation is matches the gas turbine output at low
the AVR. It is a closed loop that senses temperatures (Fig. 4-28).
the generator terminal voltage, com- More common is the totally enclosed
pares it with a set value, and controls water–air cooling system (TEWAC), in
the excitation current to counteract any which air circulates in the generator and
deviation in voltage. An important ad- is cooled by heat exchangers (coolers)
ditional component is the power system integrated in the generator casing. The
stabilizer, which dynamically modulates advantage is the clean air inside, an im-
the excitation current to counteract low- portant feature for reliability (Fig. 4-29).
frequency oscillations between the grid The rotor winding is directly cooled,
and the turbine generator. Operation of which means the air passes through the
a single generator supplying a local grid rotor winding and removes the heat
is called “island-operating mode.” In that by direct contact with the conductors.
mode, the turbine valve governor acts
directly to maintain constant speed
of the turbo set and, thus, constant
frequency. The AVR maintains the
grid voltage.
Generator Options
Basic generator features, such as speed,
cooling system, and excitation, are de-
termined by the various turbine ratings
and plant requirements. The generator
speed is determined by the number
of magnetic poles (a basic generator
design feature) and the grid frequency
of 50/60 Hz. Two-pole generators run Figure 4-28 | Open ventilation system
at 3,000/3,600 rpm and are the general Clean Combustion Technologies
choice for gas and steam turbines in Figure: 04.29/ (MF)
PMS5415 K
the power range of 40 to 1,000 MW.
Four-pole generators (half-speed
generators) run at 1,500/1,800 rpm
and often are used for small genera-
tors (less than ~40 MW) with geared
drives. They also are found on the large
half-speed steam turbines, which are
common in nuclear power plants.
The cooling of the generator
depends mainly on rated power. The
simplest coolant is air. The open ven-
tilation system uses ambient air, feeds
it straight through the machine, and
Figure 4-29 | Totally enclosed water–air cooling system
field that rotates with the speed of Rotor Winding Conductors Brushgear
Compartment
the rotor. The magnetic field lines
bridge the air gap between the rotor
surface and the stator bore and close Figure 4-32 | Alstom TOPAIR generator
in the stator core. The stator core is
a cylinder built up from as many as rises, the current in the rotorClean
winding
Combustion Technologies
250,000 insulated steel sheets. Slots must be increased to maintain Figure: 04.32/ (MF)
the volt- Stator Current
PMS5415 K (alternating current)
are formed in the inner surface to age in the stator winding.
L3
support the stator winding (armature The stability of the generator is L2
winding), which is built up by series controlled by the AVR, which senses
L1 ~
~ ~
connection of the bars in the stator the generator voltage and acts on the
slot. The stator winding consists of excitation (rotor current) to maintain a
three individual windings, one for pre-set voltage under all load condi-
each phase (Fig. 4-33, L1–L3) of the tions (for constant-voltage regime). For
generator output. It carries the main static excitation, the AVR controls a Rotating
Magnetic Fields =
current in the generator and connects stationary, thyristor-controlled rectifier,
directly to the generator terminals. which feeds the excitation current via Rotor Current
The stator terminals are connected by brush gear and slip rings (collectors) (direct current)
bus ducts to the step-up transformer, to the rotor winding. For brushless
which matches the grid voltage. excitation (rotating excitation), the Figure 4-33 | Generator working principle
The outer cylindrical part of the AVR controls the input of an exciter
core is called the “yoke.” It carries the machine, which is mounted on the
cylindrical component of the mag- non-turbine end of the main generator
Clean Combustion Technologies
netic fields. and in which a rotating diode wheel is Figure: 04.33/ (A.R.)
The rotating magnetic field in- integrated. PMS5415 K
duces an alternating voltage in the
stator winding (see Fig. 4-33). When Generator Externals
the speed of the generator rotor is For air-cooled turbogenerators, the
synchronized with the grid frequency, stator and rotor are supported on a
connection is made via the transformer common frame. This design allows
to the grid. As the load on the genera- transport with the rotor installed.
tor increases and the stator current Small units have a rectangular base
4-24 | Clean Combustion Technologies
frame on which the generator stator two-part housing: the housing frame
pedestal bearings and exciter stator are and the housing cover. The housing
mounted. The base frame transfers all frame carries the heat exchangers, the
mechanical forces of the bearings and inner stator, and on integral housing
stators to the foundation (Fig. 4-34). extensions, the bearing pedestals and
The base frame supports a box-shaped the brush gear or rotating exciter. To
enclosure (acoustic cover), which serves keep the railway profile, the housing
for internal air flow, noise protection, cover and bushings are installed on
and when requested, weather pro- site. The normal cooling is TEWAC,
tection. The top of the enclosure is with the heat exchangers located at
prepared both for open ventilation (air the bottom of the housing frame.
filters) and for a totally enclosed air cir- An additional outer cubicle can be
cuit with air–water heat exchangers (i.e., provided for noise protection. The
TEWAC). All essential components are main interfaces to the plant surround-
inside the enclosure. This applies to the ings are the stator winding terminals
bearings, the exciter, and the wiring and mounted on top of the housing, the
piping. Further components may be in- individual bearing oil supplies, and
tegrated inside the enclosure, including the excitation current leads of the
a gearbox to the turbine and medium- brush gear. The external components
voltage equipment, such as voltage and are the AVR and the static excitation
current transformers, neutral point cubicles, where applicable. Current
circuitry, and overvoltage protections. and potential transformers also are
The only external component is the mounted outside the generator.
generator excitation control module (i.e., Hydrogen and hydrogen/water-
the AVR). cooled turbogenerators (except
Larger air-cooled units consist of a four-pole generators) are all equipped
with bearing pedestals supported
on the foundation. Transport is by
Air-Water Heat Exchanger components, with the main compo-
nents being the stator central housing
Voltage Transformer
and rotor.
Current Transformer Hydrogen-cooled units consist of
a central housing part that contains
Neutral Point Circuit
the inner stator and rotor, two end
covers that support the gas–water
heat exchangers, and pedestal bear-
ings. Because the housing contains
the hydrogen, it is of heavy design
and contributes significantly to the
Bearing
overall weight of the generator. The
openings for the shaft in the end cov-
Base Frame ers are fitted with hydrogen-tight oil
Exciter
seals located inboard of the generator
Enclosure
bearings. The generator is designed
for table mounting and needs a recess
Figure 4-34 | Totally enclosed water–air cooled generator
Stator
The inner stator consists of a cylindrical
magnetic core with the stator winding Stator Core
back and, in some designs, by tie bolts twisted together to minimize eddy
Magnetic Force Acting on Core through the core. This forms a rigid cy- current and circulating current heat-
lindrical structure to resist the rotating ing in the winding. The bars are series
Stator 2f ovalization forces produced by the connected, upper bar to lower bar, to
magnetic fields in the air gap (Fig. 4-37). form coils. For two-pole windings, the
The stator core connects to the connection is made between nearly
Rotor
bedplate or housing via a spring sup- opposite slots. Thus, the array of all
port system to attenuate vibrations connections forms a cone-shaped end
transmitted to the foundation. Press winding. The line and neutral ends of
plates are fitted at the ends of the core the three phases of the stator winding
to transmit axial pressure from the are connected to the generator termi-
keybars and tie rods and to protect the nals by the phase ring connections.
Deformed Undeformed
Stator Core
end laminations from axial magnetic Usually, rated generator voltage is
Stator Core
fields that induce eddy currents in the defined phase to phase and, depend-
core ends. Various designs of the press ing on size, is in the range of 10 to
Figure 4-37 | 2f Ovalization (2f, double
frequency) plate and core end screen are available, 27 kV. Terminal currents are in the
Clean Combustion Technologies including steel plates with a copper range of 2 to 30 kA.
Figure: 04.37/ SPS screen, aluminum plates, or conical To support the bars against
PMS5415 K
rings of bonded lamination material. magnetic forces, the slot is closed by
The stator winding system con- fiberglass wedges. The tightness of
sists of a two-layer arrangement of these wedges is maintained either
Roebel bars located in the slots. Each by fiber-reinforced ripple springs
slot carries a lower and an upper between the top bar and wedge or by
bar (Fig. 4-38). The Roebel bars are a double-tapered wedge, which inher-
built up from two (or four) parallel ently acts as a spring.
stacks of insulated strands, which are The main insulation of the bars
consists of a mica-tape wound around
the bar in several overlapping layers
to produce a ground wall insulation
1-High Voltage Ground Installation 10 thickness according to the rated gener-
2-Stator Core Lamination ator voltage. The bars are impregnated
3-Stator (Roebel) Bar with resin, either by vacuum pressure
4-Roebel Transposition 2 impregnation or by using resin-rich
5-Conductor Strand
6-Conducting Paint tape. After curing, the bars are inserted
9
7-Slot Bottom Strip in the stator slots and connected to-
8-Graphitic Felt gether in the end form stator windings
1
9-Nomex Strip
10-Slot Wedge
8 to coils. Some air-cooled generators are
11-Separator manufactured using a global vacuum
pressure impregnation process. To
11 4 achieve this, the stator is fully wound,
6 including end winding support. The
insulation is dry (without resin). Resin
3
5 impregnation is made by submerging
the wound stator in a large, resin-filled
7 pressure vessel.
The stator bars are provided with a
Figure 4-38 | Stator winding system
Clean Combustion Technologies
Figure: 04.38 / (HR)
PMS5415 K
Steam Turbines for Electric Power Generation | 4-27
conductive coating around the main For larger generators with water-
insulation, which prevents electric cooled (directly cooled) windings,
surface discharges that might erode the the bracket support structure may be
main insulation. In the slot region, this replaced by a massive, cone-shaped,
is done by conductive paint or taping fiberglass ring. Usually, the phase ring
(slot corona protection); in the end connections also are supported by the
winding region, a coating of non-linear, end winding support (Fig. 4-39).
voltage-dependent resistance is applied The stator winding insulation ma-
(grading paint or taping). terials are specified to allow conductor
At each end of the stator, the bars temperatures up to at least 311°F
form the typical conical end winding. (155°C), according to Class F (311°F
Like the other sections of the bar, the [155°C]) of the electric machine
stator end winding is subject to double- standards. According to long-standing
frequency magnetic forces. To keep the practice, rated thermal utilization is
vibrations within acceptable levels and to well below 266°F (130°C; Class B).
prevent deformation under short-circuit
condition, the bars are fixed to an end Rotor
winding support structure. For indirectly The main components of the rotor of
cooled windings, the support must allow a turbogenerator are a cylindrical steel
enough space for coolant flow around forging (rotor body), the rotor wind-
the bars. In general, this is done by ing contained in slots in the rotor
taping the bars in the cone to triangular body, the retaining rings (end bells)
brackets, which in turn are suspended by to support the winding heads, and
a system of concentric rings. A well- one or two fans for coolant drive. The
designed support structure allows axial rotor body consists of a steel forging
expansion of the bars while keeping the with shaft extensions on both sides.
end winding stiff in the radial and tan- The drive end is fitted with a coupling
gential directions. In larger air-cooled to the turbine; the non-drive end
machines and in hydrogen-cooled carries the slip rings or the rotating
machines, the cone inner surface is ad- exciter (Fig. 4-40).
ditionally supported by an inner ring. On two-pole units, the center part
6
2
6
7
4 1-Stator Bars 2
5 4
2-Outer Support Ring 5
3-Inner Support Ring
4-Re-Tightening
1 5-Press-Plate 1
6-Phase Ring Connectors
3 7-Water Cooling Circuit
3
Figure 4-39 | Indirectly and directly cooled generators Clean Combustion Technologies
Figure: 04.39/ (HR)
PMS5415 K
4-28 | Clean Combustion Technologies
Ň 4
ņ AVR 5
Ņ
7 6
Ź } Ŷ
Figure 4-44 | Conventional and overhung bell exciters Figure 4-45 | Rotating excitation (AVR,
Ź}Ź~
automated voltage regulator)
ČńňčňňĞĤŸŶĥ Clean Combustion Technologies
ʼnňŅʼn
Figure: 04.45/ SPS
PMS5415 K
Steam Turbines for Electric Power Generation | 4-31
The ends of the stator bar connect The generator housing is pressure tested
to water boxes. On bars with con- and equipped with shaft seals. The shaft
ventional copper tubes, the strands seals consist of two rings, which are
and tubes of the stator bar are brazed assembled from segments. The rings are
together to form a combined electri- floating and close around the shaft with
cal connection and water box. This a small clearance. The seal oil enters the
brazed connection is a weak point of gap between the rings and fills the clear-
design on many generators, because ance between ring and shaft. Because oil
microscopic failures in the brazed pressure is kept slightly above hydrogen
Ring Manifold joint can allow water into the stator pressure, the seal oil flows through the
winding insulation. This problem is clearance in the hydrogen side ring
Teflon® Hoses
avoided by making the electrical con- and prevents hydrogen from escaping.
nection of the copper strands between The same—with more pressure drop
Water Boxes
copper plates and hydraulic connec- and, thus, more flow—applies to the
Electrical Connections
Stator Bars with tion of the stainless-steel tubes in a air side ring. An array of baffling rings
Stainless Steel Tubes separate, welded water box (Fig. 4-48). collects the oil flow. This single-circuit
The water passes through all bars seal is very common, however, because
Figure 4-48 | Water-cooled stator bars in parallel. This is achieved by a ring it combines the highest pressure with
manifold at one end of the generator, the lowest hydrogen leakage rate. More
which supplies water via electrically complex twin-or triple-circuit seals are
Clean Combustion Technologies
Figure: 04.48/ (MF) insulated, Teflon® hoses to each bar. available as well.
PMS5415 K The water is collected on the other end The drain oil flow from the hydro-
in a similar manner. Usually, sensors gen seals passes to the seal oil auxiliary
are fitted to measure the water outlet package unit, which essentially consists
temperatures to help detect corrosion of a pump and degassing tanks. The
or blockage of the cooling tubes. degassed seal oil (vacuum oil) is pumped
A separate auxiliary unit provides back into the shaft seal. Because the seal
circulation and cooling of the wa- oil system is a vital safety issue, its com-
ter and maintains the water quality. ponents are arranged with various levels
Regeneration of the water normally is of redundancy. The seal oil unit operates
done in an ion exchanger. The water as long as hydrogen is in the generator
quality is permanently monitored by and when the generator is at a standstill.
sensing the electrical conductivity. The The seal oil unit generally is a few yards
water pressure is kept below the hydro- (a few meters) below the generator level,
gen pressure so that, in case of leaks, and the drain seal oil returns in slightly
water does not leak into the machine. inclined ducting to prevent production
of foam.
Hydrogen Management A gas unit is connected to the
and Seals generator. The gas unit serves for safe
Hydrogen-cooled generators need rigid filling with hydrogen, for maintaining
hydrogen management to maintain hydrogen supply, and for evacuation of
operational performance and safety. the hydrogen. The gas unit is con-
This is achieved by the design of the nected to a hydrogen supply, commonly
generator itself and by the auxiliary consisting of a battery of gas baffles in
equipment connected to the generator. a free area, and to a purging medium
Steam Turbines for Electric Power Generation | 4-33
This chapter presents the various types of commercially available equipment for the control
of particulate and gaseous emissions from power plants. No reference is made to national or
local legislation or regulations, because these change from year to year. Such regulations are
best found in the Federal Register in the United States or obtained from local environmental
regulatory agencies in other countries.
ENVIRONMENTAL FACTORS IN
emissions from power plants are re- • More sophisticated treatment
POWER PRODUCTION
stricted by environmental regulations. of raw water, makeup water, and
The installation of highly efficient Changes in fuel availability as well as wastewater
emission control systems including air-quality standards have significantly • Condenser cooling water systems
electrostatic precipitators (ESPs), impacted the electric utility industry. that do not add heat to inland or
fabric filters (baghouses), selective Discounting inflation, the cost of in- coastal waters.
catalytic reduction (SCR), and flue stallation for a major facility has more
gas desulfurization (FGD) systems than doubled since the imposition of The auxiliary power requirements of
has become increasingly necessary. environmental laws. Because of the equipment to satisfy environmental
Power plant owners must include such permits needed to satisfy environmen- regulations will demand a significant
systems in their planning in order to tal statutes and the delays occasioned percentage of the total installed capac-
obtain permission to start construction by debating these in public, the time ity of a power plant.
of a new facility. Additional controls required to plan and construct a fossil
that are applied to other gaseous fuel power plant has increased by as POWER PLANT EMISSIONS
emissions, such as SO3, mercury, and much as 2 to 3 years. Besides the emis- Four types of emissions from fuel-
condensable particulates, and to liquid sion control equipment described in burning processes are judged to be
and solid wastes that result from the this chapter, a coal-fired power plant significant from an air-quality stand-
processing, handling, and disposal must have: point: particulate matter, sulfur oxides,
of fossil fuels and their products of nitrogen oxides, and mercury. Carbon
combustion add considerably to the • Very high static pressure, induced dioxide has recently received a great
complexity and cost of installation, draft fans, or additional booster fans deal of attention and, while presently
operation, and maintenance of power to accommodate draft requirements unregulated in the United States,
plants. The amount and complexity of • FGD solid waste and ash benefi likely will be regulated in the future
emission control systems will become ciation, handling, and disposal and is already subject to a cap-and-
even more of an issue as carbon dioxide systems trade program in the European Union.
5-
5- | Clean Combustion Technologies
Historically, particulate matter has mined from the ground, the inorganic
0.03 Lb Ash in received the greatest attention, because fraction (or mineral matter content)
Effluent / 106 BTU Fired
100.0 it is easily seen and often labeled as a can be considered to be a local sam-
99.8
public nuisance. pling of the earth’s crust. As such, trace
quantities of numerous elements can
99.6
PARTICULATE EMISSIONS be found in coal ash. Particle size varies
Efficiency, %
99.4
Emissions from coal-fired boilers vary from less than 0.01 micron in diameter
99.2 considerably, depending on the ash to more than 100 microns. Figure 5-2,
99.0 content of the coal and on the type of based on test data,1 shows the particle
0.1 Lb Ash in
98.8 Effluent / 106 BTU Fired firing. A pulverized coal (PC)-fired size distribution of fly ash at equivalent
98.6 unit can be expected to have 80 to particle diameters to be between 2 and
0 10 20 30 40 50 60 70 80 90% of the coal ash leaving the furnace 50 microns for typical coals used in PC
Lb Ash Entering Collector / 106 BTU Fired
with the flue gas. The balance of the boilers.
ash leaves through the boiler bottom
Figure 5-1 | Required particulate collection
ash removal system, hoppers under SULFUR OXIDE EMISSIONS
efficiency
economizers and air heaters, and pulver- The quantity of sulfur varies widely
Clean Combustion Technologies izer reject systems. Figure 5-1 illustrates for different coals and may range from
Figure: 05.01/ SPS
90%Kof Data the removal efficiency required of less than 0.5 to more than 5%. Sulfur
PMS5415 80%
100 Most particulate-collection equipment for generally is present in two forms:
70 Probable 50%
high-ash-content coals and very low organic and inorganic. The inorganic
Equivalent Particle Diameter, �m
50 Distribution
40 levels of effluent ash. The require- compound pyrite (FeS2) is present as
30
ment often is expressed as an emission discrete particles within the fuel and
Bahco @ 2.50
20
requirement rather than in terms of typically accounts for 20 to 50% of
Actual
10 Specific collection efficiency. Emission levels as the total sulfur. Pyrite as a percent-
7 Gravity low as approximately 0.004 gr/scfm age of total sulfur is from less than 10
5 Avg: 2.34
4 Max: 2.92 (10 mg/Nm3) at 6% O2 are not un- to more than 55% for the low-sulfur
3 Min: 1.90 usual in Europe for green field plants, Appalachian Basin coal and from
2
1 5 10 20 40 60 80 95 99 whereas older plants may have to approximately 25 to 70% for the
% Less Than Size Indicated meet emissions in the range of 0.015 higher-sulfur Illinois Basin fuel. The
to 0.040 gr/scfm (35–100 mg/Nm3). coal milling and classifying operations
Figure 5-2 | Fly ash distribution Because fly ash is not homogeneous, associated with PC steam generators
its properties (e.g., specific gravity, can separate a substantial portion of
Clean Combustion Technologies particle shape, and particle size) are the pyritic sulfur from the coal. An
Figure: 05.02/ SPS very difficult to describe in absolute additional amount of inorganic sulfur
PMS5415 K terms. Methods for determining these is retained in the bottom ash, and some
properties include photomicrographs, sulfur is either retained or adsorbed on
sedimentation, elutriation, and inertial the alkaline fly ash, leaving the boiler in
impaction devices. Fly ash is the com- solid form and eventually captured in
bination of inert or inorganic residue in the particulate control equipment.
PC particles with varying amounts of As fuels burn, most of the sulfur
carbon or coke particles resulting from is converted to SO2 and SO3. In
incomplete combustion. In general, the general, 90% or more of the sulfur in
inorganic ash particles consist primarily the fuel will oxidize to gaseous sulfur
of silicates, oxides, and sulfates, together oxides. Typically, concentrations will
with small quantities of phosphates and be from 0.05 to 0.30% by volume
other trace compounds. Because coal is in the products of combustion. The
Control of Power Plant Stack Emissions | 5-
Conversion to SO3
0.1
SO2 by molecular oxygen, the oxida- and the boiler surface area. The net
tion of SO2 in the flame by atomic effect of the two mechanisms (the gas
oxygen, and the catalytic oxidation of phase and the catalytic ash effects) is 0.01
SO2. With decreasing temperature, estimated to correspond to a conver-
SO3 progressively becomes the more sion of 0.7 to 1.4%. 0.001
thermodynamically stable form. The 500 1,000 1,500 2,000
NITROGEN OXIDE EMISSIONS Temperature, °K
equilibrium conversion in the reaction
During the combustion process
SO2 + ½O2 → SO3 in a conventional fossil fuel-fired O2 = 1% O2 = 6%
boiler, oxides of nitrogen (collectively
Figure 5-3 | SO3/SO2 equilibrium ratio
increases with decreasing temperature referred to as NOx) form in the high-
(Fig. 5-3). There is some influence on temperature region both in and around
Clean Combustion Technologies
the conversion of the flue gas oxygen the flame zone. The oxidation of both Figure: 05.03/ SPS
level within the range of practical atmospheric nitrogen (thermal NOx) PMS5415 K
interest in boilers, with higher oxygen and nitrogen contained in the fuel (fuel
increasing the thermodynamic equilib- NOx) is the cause. The rate of forma-
rium conversion ratio. tion is influenced by the temperature
Because SO3/SO2 conversion rates level and the local oxygen concentration.
of only some few percent are observed Reducing both the flame temperature
in practice at the boiler exit, it is safe and the excess air level can help to
to assume that the production of control thermal NOx formation. Fuel
SO3 is kinetically limited during the NOx is related to the available nitro-
residence in the boiler system. In a gen in the fuel and is influenced most
chemical kinetics simulation using the significantly by the oxygen concentra-
rate constants for nine homogenous, tion in the combustion region. (In
gas-phase reaction paths, modeling addition to the information on post-
both high- and low-sulfur coals,2 it furnace NOx control in this chapter,
was calculated that approximately 0.7% refer to Chapter 2 for a discussion of
of the SO2 was converted to SO3. The NOx formation during combustion
conversion mechanism is active only and to Chapter 12 for techniques to
during the first 2 seconds of residence reduce NOx in furnaces.)
time in the boiler, which corresponds
to a cooling of the flue gases down to Mercury Emissions
approximately 1,900°F (1,300°K). At Mercury often is a trace element in
lower temperatures, the reaction rates coal. When the coal is burned, some
become negligible. of the mercury is oxidized, and some
Evidence also indicates catalytic remains as the element in the vapor
conversion processes based on the state. Mercury has been considered to
catalytic activity of fly ash. Ash deposit be a surrogate for other heavy metals
iron compounds also can contribute to that also might be present in the coal.
the production of SO3 in the tempera- Mercury emissions are being restricted
ture range of 800 to 1,200°F (700 to in many areas.
5- | Clean Combustion Technologies
INFLUENCE OF VOLUMETRIC
GAS FLOW Smog Clouds and Fog Mist Drizzle Rain
No matter what type of fly ash collector Tobacco Smoke H2SO4 Mist
is to be installed on a steam generator, Lung-Damaging Dust Pollens Beach Sand
it is essential that it be designed for Insecticide Dusts
the correct gas volume. Manufacturers
Oil Smoke Pulverized Coal
design and size their equipment based
Carbon Black Stoker Fly Ash
on their experience with this same type
Pulverized-Coal Fly Ash
of equipment. This experience includes
tests on both laboratory or pilot-plant Coal Dust
6.6% H2O rings eliminate fringe effects. A voltage Cyclone-type, mechanical collectors are
1010
is applied between the two electrodes, now used only occasionally as pre-
109
and the current is noted. The resistiv- collectors to ESPs and fabric filters.
108 ity is then calculated and reported in Mechanical collectors can be either dry
13.5% H2O ohm-centimeters (Ω-cm). The conduc- cyclone collectors or wet scrubbers.
107
20% H2O tivity cell is placed in an oven, and the
106
resistivity can be measured at different Mechanical Cyclone Collectors
310 366 422 478 534 590 646
Temperature, °K temperatures. The humidity in the These devices achieve particulate
oven also should be noted; normally, removal by centrifugal, inertial, and
Figure 5-6 | Laboratory measurement of it is possible to vary the humidity. The gravitational forces developed in a vortex
fly ash resistivity for varying effect of temperature and humidity separator. The dust-laden gas is admitted
Clean Combustion
humidity Technologies
(% by volume) in the
Figure:
flue 05.06/
gas SPS on resistivity thus can be investigated. either tangentially or axially over whirl
PMS5415 K The influence of other gas components, vanes (Fig. 5-7) to create a high velocity
however, cannot normally be taken in the cylindrical portion of the device.
Clean Gas Outlet into account. Figure 5-6 illustrates Particles are subjected to a centrifugal
typical bulk resistivity as a function of force and an oppositely directed viscous
Dust Laden Gas Inlet temperature and humidity as measured drag. The balance between these two
Whirl Vanes
in a laboratory. forces determines whether a particle will
move to the wall or be carried into the
In Situ Resistivity vortex sink and then passed on to the
Central Vortex Sink A combination high-voltage and clean-gas outlet tube.
ground-potential probe, inserted in The high-velocity, downwardly
the gas stream, has a negative high- directed vortex is reversed at the bot-
Downward Directed
voltage probe consisting of a flat disk tom of the cylindrical section. In this
Peripheral Flow that is concentric about a needle-point reversal, inertia and gravity as well as
electrode. The needle-point electrode centrifugal force inject the dust into
Dust to Removal System precipitates particulates electrically the hopper. The action of the reversed-
from the dispersoid. The flat, charged gas circulation takes place over a small
disk is separated from the needle point diameter but at relatively great velocity.
Figure 5-7 | Principle of cyclone-separator
operation
Clean Combustion Technologies
Figure: 05.07/ (ks)
PMS5415 K
Control of Power Plant Stack Emissions | 5-
Efficiency, %
60
unit. Because these collectors depend
3.0 Typical for Dust
primarily on differential inertia, collec- 40 Specific Gravity of 2.50
2.0
tion efficiencies vary with particle size, 1.0
particle density, gas temperature, and 20
of parallel units are required when high pressure-drop venturi scrubbers are ef- 6�� WG Pressure Drop
60
gas flows prevail. fective in removing particles in the range
of 0.1 to 1.0 microns. Typically, a scrub- 40
Wet Scrubbers ber operating at a 6˝ WG (1.1 cm Hg)
20
In the analysis of wet particulate scrub- pressure drop should capture practically
bing, a number of important facts are all particles greater than 5 microns and 0
0.5 1.0 1.5 2.0
known. First, the dust particle must approximately 90% of the particles in Effective Particle Size, �m
impact on the water droplet. The the range of 1 to 2 microns. Figure 5-9
impaction efficiency is a function of shows a typical efficiency curve for a Figure 5-9 | Wet-scrubber fractional efficiency
VV wet scrubber operating at a 6˝ WG curve
the dimensionless group r s , where Clean Combustion Technologies
Dg (1.1 cm Hg) pressure drop. Figure: 05.09/ SPS
Vr is the relative velocity between the PMS5415 K
water droplet and the dust particle, Venturi Scrubbers
Vs is the settling velocity for the dust In a venturi scrubber, dust-laden gases
particle, D is the diameter of the are wetted continuously at the venturi
water droplet in microns, and g is the throat. Flowing at 12,000 to 18,000
acceleration of gravity. Because g and fpm (61 to 91 m/sec), the gases produce
Vs are constant for a dust particle of a shearing force on the scrubbing liquor
5- | Clean Combustion Technologies
because of the initial high velocity through a diffuser into the lower cham-
differential between the two streams. ber of a separator vessel. Impingement
This shearing force causes the liquor to of the stream into the liquid reservoir
atomize into very fine droplets. removes most of the particulates. A
Impaction takes place between the venturi-type scrubber operating in the
dust entrained in the gas stream and the pressure range of 30 to 40˝ WG
liquid droplets. As the gas decelerates, (5.6 to 7.5 cm Hg) can collect nearly
collision continues, and agglomerated, 100% of particles in the range of 0.2 to
dust-laden liquor droplets discharge 10 microns.
ELECTROSTATIC PRECIPITATION
Collected Dust OF DUST
Discharge Electrodes
Clean Gas In electrostatic precipitation,
+ suspended particles in the gas are elec-
–
– – ––
– trically charged, then driven to collecting
– – – electrodes by an electrical field. The elec-
–
Dust-laden – –
Gas – – trodes are rapped to cause the particles
– –
– – to drop into collecting hoppers. The
– – –
– –– – charges and the electrical field are gener-
– – –
– ated by high-voltage discharge electrodes
Corona – –
– – that are placed in between the grounded
–
–
Dust Particles collecting electrodes (Fig. 5-10).
– –
– This process differs from mechanical
– Collecting Electrodes
+ or filtering processes, in which forces
are exerted directly on the particulates
rather than on the gas as a whole. Ef-
fective separation of particles can be
Figure 5-10 | Operating principle of electrostatic precipitation
achieved with a low power expenditure
Clean Combustion Technologies
and with a low draft loss compared to
Figure: 05.10/ (MH) fabric filters and scrubbers and with
Supporting Insulator
PMS5415 K Rectifier Circuit little or no effect on the composition
of the gas.
The most commonly used design
Charging is the single-stage ESP—known
previously as the Cottrell type—that
combines the ionizing and collect-
ing process (Fig. 5-11). The discharge
electrodes are suspended on insulators
connected to a high-voltage source—
the power supply.
The dust-laden gas flows horizon-
tally between grounded collecting
Ground
Discharge Electrodes plates. This system is powered from a
(spiral wires) transformer rectifier (T/R), and the
Collecting Electrodes
(plate curtains) assembly is called a “bus section.” There
usually are several independent bus
Figure 5-11 | Principal design
Clean of a single-stage
Combustion electrostatic precipitator
Technologies
Figure: 05.11/ (HR)
PMS5415 K
Control of Power Plant Stack Emissions | 5-
sections in length, and for large plants To meet a demand for ultra-
with high gas flows, there are parallel high efficiency collectors of rugged
bus sections. One or several (parallel) construction and high reliability, Eu-
casings enclose the sections to form ropean manufacturers in possession
the ESP plant. The T/R sets often of the basic patents disseminated by
are placed on the roof of the ESP. The Frederick Gardner Cottrell developed
high voltage is unidirectional—usually the rigid-frame precipitator. Actually,
negative—and can exceed 100 kV for this design more closely approximat-
certain applications. ed Cottrell’s original design than the
The unipolar discharge of gas ions U.S.-style, weighted-wire designs did.
starts at a critical voltage—the onset The term rigid-frame refers to the rug-
voltage. The ionization is visible as a ged pipe-frame or mast-construction
corona at the discharge electrode. The discharge electrode, which largely
negative ions move toward the positive precludes wire breakages and electri-
collecting electrode; the positive ions cal short circuits. The basic design
migrate toward the discharge electrode. incorporates segmented collecting
The movement of the ions constitutes plate configurations and profiles
the corona current. When a higher volt- (for close fabrication tolerances over
age is applied, the current increases. The heights above 50 feet [15 m]), rigid
ions become attached to particles car- discharge frames and electrodes, and
ried by the gas stream. The charges on much greater division of rapping
the particles and the electrical field move (often with individual rapping for
the fly ash to the appropriate electrode. each discharge frame). Modern ESPs
Higher voltages and currents enhance are robust and have high reliability.
the precipitation process for most Availabilities for new ESPs often
fly ash. The particles attached to the exceed 99%.
positive electrode dissipate their charge
and become electrically neutral. The fly Precipitator Design Factors
ash layer on the electrodes is removed Precipitator designs consider:
mechanically by rapping. In a two-stage
• Specific collection area (SCA)
precipitator, the ionizing and collect-
• Treatment time
ing stages are separated. This design is
• Gas velocity
used for low particulate loadings and
• Spacing
minimum ozone generation, and it is
• Electrode configuration and rapping
common for indoor ventilation systems.
• Number of bus sections and fields
Single-stage ESPs can be either hang-
• Power supply
ing, weighted-wire, or rigid-frame units.
• Automatic voltage control
As initially applied in the United States,
• Hopper size and slope
precipitators were designed to provide
a minimum plate area at low cost. De-
signs used interlocking collecting plates Specific Collection Area. The SCA
and weighted-wire discharge electrodes relates the size of one precipitator to
with a maximum height of 36 feet another in terms of the effective
(11 m). Roof-mounted, gang-rapping collecting-electrode surface area in
vibrators removed particulate matter square feet per thousand cfm (square
from the collecting plates. meters per cubic meter per second) of
5-10 | Clean Combustion Technologies
gas treated. The area considered is the 5 ft/s (1.5 m/s) can be used. Excessive
flat, projected surface of both sides gas velocities will cause re-entrainment
of the collecting electrodes; the gas of collected ash and, thereby, reduce
volume is the actual volume at the de- the collecting efficiency.
sign operating temperature and at the
prevailing absolute pressure in the ESP. Spacing. Spacing is the distance between
Depending on the physical, chemical, adjacent collecting plates. Through the
and electrical characteristics of the coal years, the spacing has been increased,
ash, precipitator SCAs range from 300 because experience has shown that cost
to more than 1,000 ft2/cfm (60 to 200 savings can be achieved. Standard spac-
m2/cubic meter per sec) for collection ing is approximately 16 inches (400 mm)
efficiencies above 99.5%.19 for fly ash applications. Many years ago,
the spacing normally varied between 9
Treatment Time. This refers to the and 10 inches (230 and 250 mm),
length of time the flue gas will spend in and that can still be seen in old ESPs.
the presence of the electrical field at de- During the 1970s, it became common to
sign gas velocity. The treatment length use 12 inches (300 mm). Since the mid-
is not the total front-to-rear dimension 1980s, the common consensus is that the
of the “box” but, rather, is the horizon- wider spacing (e.g., 16 inches [400 mm])
tal dimension from the plane of the does not need longer treatment time in
front face of the first collector to the order to achieve a certain efficiency com-
rear plane of the last surface, less any pared to the narrower spacing.20,21 Thus,
space for walkways, beams, and so on. the actual collecting area is reduced in
proportion to the spacing for a given
Gas Velocity. Superficial gas velocity is a task. For some power plants, a space of
critical design parameter. It is calculated 12 inches (300 mm) remains the main
by dividing the actual gas volume (at alternative; on others, spacings of more
design temperature and absolute pres- than 16 inches (400 mm) are applied.
sure) by the face area of the precipitator It should be noted that the SCA
immediately in front of the first col- value for an ESP depends on the spac-
lecting electrode. Effective plate height ing. For example, the SCA value for 16
is multiplied by the inside, face-to-face inches (400 mm) is three-fourths of
dimension between the outer two col- the SCA for 12 inches (300 mm) for
lecting electrodes—one at each side of the same treatment time. The wider
the sidewall casings—to give the face spacing implies that the ESP weight
area. The calculated gas velocity is the is reduced for the same “box.” Fur-
average velocity. Real velocity at any spot thermore, fewer components are used
may deviate somewhat from the average inside, and ESP reliability is increased.
velocity because of imperfections from
the distribution of the gas. Electrode Configuration and Rapping.
Designs for medium- and high- The shape of the electrodes and the
resistivity fly ashes and for low emis- rapping effectiveness can drastically
sions often use average velocities affect the collecting efficiency of a
between 3 and 4 ft/s (0.9 and 1.2 given ESP. Properties such as cur-
m/s). For fly ash particles that easily rent distribution along the collecting
agglomerate, velocities on the order of electrode, the amount of voltage and
Control of Power Plant Stack Emissions | 5-11
current achieved in the precipitator the voltage and current that are used in
before sparking occurs, and the rapping the ESP. From time to time, sparking
sequences are important parameters in or arcing can occur between the high
achieving efficient ESPs.22,23 voltage and the grounded parts in the
ESP. During a spark, the ESP is short
Number of Bus Sections and Fields. circuited and not efficient. It therefore
For reliability and effectiveness, ESPs is important to control the spark rate
are divided into bus sections, each of and power input.
which normally is energized by one Since the mid-1990s, a new type of
independent, electrical, high-voltage T/R set has been available. This T/R
power supply. There may be several converts the main voltage with a fre-
bus sections in parallel and in series. quency of up to 50 kHz to an almost
Each set of parallel bus sections is pure DC voltage. The high-frequency
called a “field.” It is common to have power supply is small and lightweight
three or more fields in the direction compared to conventional units. The Improved power supplies and
of gas flow. The amount of plate area high voltage and the control units are T/R sets provide a nearly pure
per power supply—a T/R set—affects placed together in one box. For many DC voltage, allowing for a
the current and voltage achieved in an applications, a higher collecting effi-
electrical section. Small bus sections ciency can be achieved for a given ESP
higher collection efficiency and
are more efficient per square foot than with this type of power supply.24 reduced emissions.
are large ones because of better control
of tolerances and less variations in gas Automatic Voltage Control. The auto-
and particulate data. This implies that matic voltage control system keeps the
a lower SCA value can be used for precipitator operating at its optimum
a given task. The ESPs of European voltage and current. This digital con-
design often use large bus section areas trol system allows a much faster and
(e.g., 60,000–80,000 ft2 [5,600–7,400 more reliable control of sparking and
m2]); the U.S. design has a higher other parameters. Advanced micro-
sectionalization (more T/Rs), implying processor controls (one for each power
smaller bus sections. Cost estimations, supply) can automatically sense the
available space, redundancy, or spare operating conditions and change the
capacity might decide the final number set values of the current and voltages
of bus sections. in order to achieve the highest possible
collecting efficiency. The optimum set-
Power Supply. The conventional T/R tings vary with the type of coal and the
set produces a high direct current volt- boiler load, and the controls are able to
age based on the mains 50/60-Hz handle such changes.
frequency. The T/R set charges particles Intermittent energization, or semi-
flowing through the ESP. This occurs pulse, is a standard feature for modern
once during a portion of every half- automatic voltage control systems.
wave. The result is that the high direct The intermittent energization com-
current (DC) voltage has peak and prises blocking of a selected number
valley voltages—that is, ripple—be- of half-waves of the mains. The ripple
cause there is a continuous corona is then enlarged, which results in
discharge but an intermittent charging. improved collecting efficiencies for
The collecting efficiency depends on high-resistivity ashes.25
5-12 | Clean Combustion Technologies
Equation 5-1
Hopper Size and Slope. The proper de-
sign of the collecting hoppers forming where
the floor of a precipitator is important,
= efficiency of collection (%)
because a breakdown in the hopper
e = base of natural logarithms
outflow that results in overfilling can
(2.718...)
severely damage the electrical internals.
A = collecting-electrode surface
Alstom has made comprehensive rec-
area (ft2 [m2])
ommendations concerning the design
w = migration velocity (ft/s
and heating of precipitator and fabric
[m/s])
filter hoppers and their fly ash removal
V = actual volumetric gas flow
systems.26 These recommendations,
rate (1,000 ft3/s [m3/s] at
some of which are detailed later in this
temperature and absolute
chapter, help to avoid costly outages
pressure)
attributable to these portions of the
A/V = SCA
collection/transport complex.
It should be noted that w cannot be
Precipitator Design Analysis measured but, instead, must be calcu-
MIGRATION VELOCITY, w . It was tradi- lated from the equation. The calculated
tional to consider the semi-empirical w is furthermore the average particle
Deutsch–Anderson equation, devel- velocity through the ESP regardless of
oped in 1922, in the analytical design particle size. A significant discrepancy
of a precipitator for low efficiencies exists between theoretical and practical
(up to 98%). In the 1970s, a modi- values of w, with the theoretical value
fied form of the Deutsch–Anderson being much higher than the actual
formula, the so-called Matts–Ohnfeldt value.28,29 This “loss” of performance
equation,27 was used to better predict is caused by such factors as uneven
the performance of an ESP for most gas flow, particle diffusion, electrical
applications. wind, particle charging time, and loss
The ESP collection efficiency is of particles from collecting surfaces by
related to the total surface area of the re-entrainment. Because the value of
collecting electrodes per unit volume w can hardly be estimated from theory,
of gas, or the specific collecting area and because of these uncertainties, it is
(SCA). The SCA value is inversely necessary to rely upon field experience
proportional to the particle migra- with a variety of precipitator instal-
tion velocity. The expression “particle lations.19,26,30–34 When experience is
migration velocity” is the velocity the limited, designs have to be established
Control of Power Plant Stack Emissions | 5-13
more charges in the electrical field and between the high-tension discharge
gets a higher migration velocity than a electrodes and the collecting electrodes.
small particle does. This implies that This current must pass through the
larger particles are efficiently collected layers of collected ash that normally coat
in the front of an ESP but that the fine the plates. The ash therefore must be
particles take a longer time to migrate to able to conduct the ionic current to the
the collecting electrode. Thus, the mi- grounded metal surfaces of the plates.
gration velocity is not constant for low The electrical conductivity required is
and high collecting efficiencies, because very small—approximately 10–8 mho/m
for a high efficiency, more of the fine or S/m (Siemens/m), corresponding to
particles must be precipitated and the a resistivity of 1010 Ω-cm.
average migration velocity is reduced. In The transport of charges through
order to use Equation 5-1, the particle the ash layer produces a voltage or
size distribution must be known. potential drop across the layer in
The difficulties encountered with accordance with Ohm’s law. When
Deutsch–Andersen equation led to the the resistivity of the ash is in the
development of the modified equation: critical zone (in the range from 1011
to 1012 Ω-cm, depending on condi-
tions and design), the voltage drop
− Awk
k
across the ash layer can amount to
V
= 1 − e × 100 several kilovolts. The existence of this
voltage drop may be higher than the
breakdown strength across the layer
Equation 5-2 and can be sufficient to cause localized
sparking because of ionization of the
where wk is now an empirically gas. Positive ions that are generated
observed migration velocity and k is a migrate toward the high-voltage elec-
constant (~0.5, but varying between trode and will cause particle charging
0.4 and 0.6 depending on the specific of the opposite polarity compared
ash and application). The other terms to the original charging. The whole
are as defined as for Equation 5-1. process is called “back corona,” and
5-14 | Clean Combustion Technologies
pre-heater. The flue gas temperature ash resistivity from between 1011 and
Gas conditioning can reduce was approximately 650°F (340°C). 1012 to 5 × 109 Ω-cm. Experience has
ash resistivity, leading to higher Many of these plants could not meet revealed that predicting the effect of
the required emission levels, and SO3 is difficult for very-high-resistivity
collection efficiency and reduced
investigations revealed that the ash coals—for example, when the silicon
emissions. layer on the electrodes developed back and aluminum oxides exceed 85% of
corona after some time of operation. the fly ash. An alternative concept is,
The reason was found to be so-called then, to use dual conditioning—a com-
“sodium depletion.” In other words, bination of SO3 and NH3 injection in
the ash layer on the electrodes became the flue gas. These two gas components
inhomogeneous after some time of combine and, when mixed into the fly
operation. Sodium, which reduces ash, reduce resistivity significantly.
resistivity, migrated to the outer ash
layers, while the inner ash layer was COHESIVE AND ADHESIVE PROPERTIES OF
lacking sodium. Thus, a high-resistivity, FLY ASH. These properties of fly ash
thin inner layer was developed, and the are important for collection by ESPs,
corona current had to pass through it. where the ash forms in compacted lay-
As a result, back corona was generated ers on the electrode surfaces. Typically,
in this layer. Manual cleaning of the these layers are less than 0.5 inch (13
electrodes had to be done frequently. mm) thick but, in some cases, may
Most of these “hot” ESPs have been build up to 1 inch (25 mm) or more.
rebuilt and are today placed after the Although some ash falls into the hop-
air pre-heaters.44,45 pers because of either the mechanical
Another way of overcoming the vibrations frequently present in plant
high-resistivity problem is to use a sub- structures or its own weight, rapping
stantially lower flue gas temperature in or jarring of the electrodes usually is
the ESP. Since the late 1990s, several necessary to effectively remove the
large boilers have been equipped with collected ash layers. The ash generally
ESPs operating successfully at ap- breaks away in isolated patches, so
proximately 200°F (93°C). The use of the collecting plate surfaces present a
low-sulfur coal, together with a more somewhat spotty appearance.29
alkaline fly ash, prevents clogging and Fine solid particles tend to cohere
corrosion—no acid dew point can be and form agglomerated masses, which
measured in this flue gas. Such ESPs have a small but definite stability. The
are characterized by very small collect- coherence of the particles is caused by
ing areas. molecular attractive forces at the points
of contact of the particles. The degree
Gas Conditioning. Migration veloc- of coherence is appreciable only for
ity declines as resistivity increases, as relatively fine particles of less than ap-
is characteristic of many low-sulfur proximately 20 to 40 microns. Coarse
bituminous and sub-bituminous coals. or granular particles (e.g., sand) do not
Adding SO3 to flue gas can reduce ash cohere in the dry state. On the other
resistivity.46 For some western-U.S., hand, fume particles, which typically
low-sulfur coals, an increase of 30 ppm are less than 1 micron in size, usually
of SO3 in the gas has reduced the fly show a high degree of coherence.
Control of Power Plant Stack Emissions | 5-17
The adhesion of particles to the col- stable and fall with as little loss of
lecting electrodes also is the result of particles as possible. If the falling ag-
molecular attractive forces, and similar glomerates have little coherence, some
to cohesion, the adhesive forces are may break up and be redispersed into
larger for fine particles. Thus, adhesion the gas stream, partially nullifying the
and retention of the collected layers of collecting action of the precipitator.
ash on the electrodes are determined The presence of coarse particles
basically by particle size, and both tends to reduce the cohesive and
increase with decreasing particle size. adhesive forces and, therefore, the ag-
Other factors, however, such as particle glomerating effect, which holds masses
shape and composition, influence the of fine particles together. Excessive
adhesion and cohesion of particles amounts of coarse, gritty particles in
as well. The presence of condensed an ash make the ash difficult to hold
moisture, with possible cementing ef- on the collecting plates and also lead to
fects, also can profoundly influence the loose, unstable agglomerates when the
buildup of particles on electrodes and plates are rapped. On the other hand,
other surfaces. the elimination of all coarse particles
In ESPs, the accumulation of ions above 10 to 20 microns may, in some
on the collected dust layers produces cases, make removal of the ash difficult.
an additional retention force of major There will be a much greater tendency
importance. These ions tend to charge for the particles to build up on the
the collected particles so that they discharge electrodes. A very fine ash
are attracted to the collecting plates containing no coarse particles presents
by relatively strong electrical forces. a more difficult rapping problem.27
Furthermore, any particles that do
escape from the collected layers of ash Construction of ESPs
(assuming that no back corona condi- The weighted-wire type of pre-
tions prevail) tend to be immediately cipitator has been well described
recharged electrically by the corona elsewhere.3–16,30,47–49 Features of
ions and forced back to the collecting the rigid-frame design can be found
plates. The particle-retaining force of elsewhere as well.3–16,38,44,49
the corona current increases with the In the rigid-frame design, the carbon
electrical resistivity of the particles and steel collecting plates are constructed
with the strength of the current. Thus, of profiled segments up to 26 inches
it often is advantageous to operate (660 mm) wide and sometimes more
precipitators, particularly the outlet than 50 feet (15 m) long. They attach
sections, at relatively high corona cur- at the top to back-to-back channels
rent densities.29 spanning the insulator housing. At the
Agglomerated masses of particles, bottom, anvil or spacer bars maintain
removed from the electrodes by rap- plate alignment and act as anvils for plate
ping, fall into the hoppers by the force rapping. The collecting electrode plates
of gravity. These agglomerated masses are rapped individually by mechani-
are bound together by molecular cohe- cal hammers indexed on a shaft driven
sive forces. For best performance, these from outside the precipitator shell.
falling agglomerates must be relatively For applications when the ash is easily
5-18 | Clean Combustion Technologies
the rigid-frame and a maximum of 36 FABRIC FILTRATION Table 5-1 | Percent of Fly ash
feet (11 m) for the weighted-wire unit. Fabric filters (or baghouses, as they Collected In Each Row
In many instances, a rigid-frame unit commonly are called) have a long his- of Hoppers
will fit inline, whereas the weighted- tory of application in the capture of
Total number of rows, front to rear
wire design must be arranged in one of particulate matter from process gases.
Row
the other, less desirable ways. Applications range from nuisance- 3 4 5 6
No.
particulate collection (for the control
1 84.1 74.9 66.9 60.2
FLY ASH RECEIVING HOPPERS. During of silo dust or conveyor transfer sta-
operation, the fly ash accumulated on tion dust) to process applications in 2 13.4 18.8 22.2 24.0
the collecting surfaces of the precipita- which the fabric filter equipment is an 3 2.1 4.7 7.3 9.3
tor is periodically shaken loose and integral part of the process equipment 4 — 1.2 2.4 3.9
dropped into the hoppers. The level of used to capture the product (as with 5 — — 0.8 1.6
dust in each hopper will rise until that carbon-black manufacturing) and to 6 — — — 0.6
particular hopper is emptied. If for any strictly particulate capture for environ- TOTAL 99.6 99.6 99.6 99.6
reason emptying the hopper is delayed mental compliance (as with coal-fired
until the dust level approaches the el- boiler applications). In general, fabric
evation of the bottom of the discharge filters have found increasing acceptance
electrodes, they will be electrically in applications where gas-borne par-
short circuited to ground through the ticulate matter must be efficiently and
mass of collected dust. The affected dependably removed from gas streams.
precipitator section will cease func- Fabric filtration presently is con-
tioning in a normal manner but will sidered by many as the best available
continue to collect some dust by acting technology to control particulate
as a settling chamber. emissions from a gas stream. The
If the dust and flue gas entering the development of new and improved
precipitator are reasonably well distrib- fabrics and finishes since the early
uted, all precipitator hoppers in any 1970s has had a dramatic effect on the
row perpendicular to the gas flow will application potential of fabric filters,
collect substantially the same quantity which in early years were limited
of dust per unit of time. Much more primarily to the natural material cloths.
fly ash, however, will be collected These new fabrics and finishes have
in the rows of hoppers closer to the greatly extended the allowable operat-
precipitator inlet than in the rows ing temperature window, the expected
toward the rear of the precipitator. service life, the resistance to chemical
The quantities can be approximated attack, and the cake release properties.
with reasonable accuracy by use of The preference of fabric filtration
the Deutsch–Anderson equation. An over other types of particulate removal
example is shown in Table 5-1 for four equipment generally is attributable
precipitators having overall efficiencies to the superior performance of fabric
of 99.6% and from three to six rows of filters in the following areas:
hoppers in the direction of gas flow. In
modern fly ash precipitators, then, the • Outlet emissions are nearly inde-
inlet row of hoppers can collect from pendent of the magnitude of the
40 to 100 fold as much fly ash as does inlet dust loading as well as the
the rearmost row.50 resistivity of the dust.
5-20 | Clean Combustion Technologies
Filter bag attachment hardware con- service and one of these in the cleaning
sists of a cap installed in the upper-top mode, is limited to 2.5 for these types
portion of the bag. This arrangement of units.
allows attachment of the bag to the The cleaning frequency of the
tensioning device that is hung from individual compartments will depend,
the compartment bag-hanging frame. in part, on the particle size of the dust,
An adjustable clamp or sewn-in band the cohesive nature of the filter cake
fastens the bottom cuff of the bag that is formed, and the inlet dust load-
to the tubesheet located in the lower ing. All other parameters remaining
section of each compartment. Both equal, the frequency of cleaning will
arrangements allow direct attachment also increase as the inlet dust loading
to the tubesheet thimble. Filter bags increases, which results in a greater
for reverse air service contain multiple time of operation by the filter unit in
rings sewn in the bag length to prevent the net (less than full complement of
total collapse during the cleaning cycle compartments operating) condition.
(see Fig. 5-17). Ideally, a properly designed fabric
Although the bag hardware rep- filter unit should operate in the gross
resents only a small fraction of the condition the majority of the time,
total cost of the fabric filter system, because this represents the most ef-
this hardware can cause problems if ficient filtering mode (lowest air/cloth
not properly selected with respect ratio) and lowest differential system
to both design and material. Any pressure drop.
component that contacts the cloth This fabric filter design using ap-
needs to be carefully chosen to ensure propriate cloth materials has been
that premature cloth damage will not applied to a wide range of tempera-
occur during normal operation of tures, from near ambient to more than
the unit. For example, edges need to 500°F (260°C). An advantage of this
be smooth, and excessive corrosion particular system is the ability to use
should be prevented. fiberglass cloth that can withstand
Depending on the inlet dust loading wide ranges in temperature without
per unit volume of gas treated and the physical damage. The reverse air
particle size distribution, the designer system, being a gentle method of filter
of the fabric filter unit will adjust the bag cleaning, is compatible with most
design air/cloth ratio and/or the total available bag materials.
number of system compartments re- One of the most attractive features
quired to ensure a reasonable air/cloth of this type of fabric filter system is
ratio under all operating conditions. the extended bag life for the fiberglass
Of concern in these calculations is bags. The typical bag is approximately
the effective air/cloth ratio during 12 inches (0.3 m) in diameter, with bag
the cleaning mode and the maximum lengths of up to 36 feet (11 m). A large
ratio in the event of full load operation number of operating units have expe-
with one compartment isolated for any rienced bag life in excess of 10 years
maintenance activity plus one addition- between changes. The negative aspects
al compartment in the cleaning mode. of this design are the increased capital
Generally, the maximum air/cloth cost associated with the low air/cloth
ratio, with two compartments out of ratio (particularly considering the
Control of Power Plant Stack Emissions | 5-23
to fall between the bags and into the bags and, therefore, can fall through
discharge hopper. Within a single the rising gas stream into the hopper.
pulse jet unit or single compartment, With the advent of large collectors
rows of filter bags are pulsed sequen- to control emissions from hot pro-
tially under the control of a solid-state cesses (e.g., boilers), the online pulse
programmer, which usually is mounted jet filter with gas entering the compart-
adjacent to the collector casing. ment through the hopper has been
The original operating concept for shown to have several disadvantages:
all pulse jet types of collector was
on-line pulse cleaning, with the gas en- • Online maintenance and inspec-
tering each compartment in the bottom tions are not practical.
area (or hopper) and flowing upward • Compressed air requirements are
between and through the bags. The gas very large, with increased filter area
flow through the collector is continu- needed to maintain the required
ously maintained, and the rows of filter pressure drop.
bags are pulsed sequentially, either on a • Physical properties of the fiber-
timed sequence or under control of the glass fabrics required for hot boiler
collector differential-pressure monitor. processes are inconsistent with the
Although still an established arrange- harsher cleaning mechanism of on-
ment for many nuisance dust collection line pulsing at high pulse pressures.
applications, this gas flow arrangement
and cleaning method can be inefficient Reverse air systems, as previously
compared to off-line pulse cleaning. described, are essentially multi-
Consider the following points: compartment units with off-line cleaning.
The carryover of the reverse air operat-
• The pressure pulse is from 50
ing scheme to smaller industrial boilers
to 100 ms in duration. Only
produced the modular reverse air
during this period is the gas flow
systems (i.e., those that are made up of
stopped through a row of bags
totally shop-assembled compartment
being cleaned.
modules). These designs substantially
• Dust pulsed from one bag row is
reduced field construction labor costs.
immediately re-entrained on an
This same concept was then ap-
adjacent bag row because of the
plied to the pulse jet units using
upward gas flow direction and,
shop-assembled modular units and
therefore, does not reach the hopper.
off-line cleaning. This allowed pulse
• Only after repeated pulsing will
jet systems composed of totally shop-
the captured cake gravitate to
assembled modular systems, including
the hopper.
the extra module for off-line clean-
The main reason that dust can be ing. With this method of operation,
collected using online pulse jet cleaning features include:
with gas entering the compartments
through the hoppers is dust ag- • Off-line cleaning, which is more
glomeration. The dust particles form effective, because the dust has little
agglomerates that are heavy and dense tendency to migrate to adjacent
enough to resist attraction to adjacent filter bags. Instead, the cake has an
Control of Power Plant Stack Emissions | 5-25
size, the pulse jet style of fabric filter greater effect than that in a unit with a
often is preferred by many end users. greater number of compartments.
The typical bag diameter is approxi- In the design of the reverse air unit,
mately 6 inches (0.15 m) or less, with bag aspect ratios typically are 32:1,
bag lengths typically of up to 26 feet with a maximum of 36:1. Bag aspect
(7.9 m). Progress is being made on ratio is defined as the ratio of the bag
extending the bag length even more length divided by the bag diameter.
based on the quality of the fabrics Here, the concern is the ability to
and finishes available. This increase effectively clean the upper section of
in filter area per bag will help to the bag on this inside type of collector
further reduce the total installed cost unit if the length becomes too long.
of the filter. The negative aspect of Most designs have used the more
this design is the decreased bag life. common bag length of 32 feet (9.8 m)
Because of the more aggressive clean- in the typical boiler flue gas applica-
ing system, a bag life of 3 to 4 years is tion. In recent years, however, many
common in most installations. successful units with a ratio of 36:1
have been used.
Design of Equipment Similar design criteria also have been
Overall system designs involve selec- developed for the pulse jet units in
tion of the cleaning mechanism and defining acceptable ratios or compo-
type of fabric filter to be used for a nent designs. As an example, typical
particular service along with selec- filter bag lengths for most industrial
tion of a suitable bag material and applications range from 12 to 16 feet
bag finish. To make these choices, the 3.6 to 4.9 m), although some designs
designer needs a complete knowledge have been up to 20 feet (6.1 m) in
of the process on which equipment is length. For larger gas volume applica-
to be installed and knowledge regard- tions, such as coal-fired boilers, a bag
ing the limiting factors of the options length of 26 feet (7.9 m) has been used
available for application.51 on a large number of installations. The
Fabric filter designs have some application of felted cloths on pulse jet
established criteria, such as typically units is quite typical. The use of woven
acceptable air/cloth ratios and bag fiberglass cloth is acceptable, especially
aspect ratios. As noted previously in for the higher boiler flue gas operating
the discussion of reverse air systems, temperatures, provided the method of
the gross air/cloth ratio that gener- cleaning the bags is appropriate for this
ally is acceptable for the reverse air material. As indicated in the section
system is at or less than 2.0, with a below on filter materials, higher-
maximum of 2.5 in the net condition temperature felts also are available
of two compartments off-line and one with some limitations. Filter bag cage
of those compartments in the clean- materials commonly are of carbon
ing mode. In the design of a reverse air steel with a baked-on epoxy coating.
unit, or any off-line cleaning unit, the Other coatings are available, as is alloy
number of compartments will affect construction. High side inlet collectors
the value of the net condition, because have advantages with respect to hopper
with fewer compartments, the removal capacity and hopper re-entrainment.
of one compartment from service has a Designs that have the gas entering
Control of Power Plant Stack Emissions | 5-27
the hoppers are acceptable in some Fiber manufacturers and felt sup-
instances and allow lower equipment pliers have done a large amount of
setting heights in the overall system developmental work to improve the
design. performance and features of felts used
The selection of the cleaning in fabric filters. Different fiber shapes
mechanism design, the number of and sizes, along with varying methods
compartments needed, the air/cloth of assembling the fibers into felt mate-
ratio necessary, the bag material and rial, have been investigated and tested
finish, the maximum bag length, and a in laboratories and some pilot demon-
variety of other design considerations stration units. As more and more fabric
are specific to the service intended and filters are requested by end users, the
to the user’s preference. Regardless of technology will continue to advance.
the design selected, the majority of In addition to the developments in
emissions from a fabric filter occur felt material construction, bag suppliers
either during or immediately after bag have improved quality control of manu-
cleaning. The cleaning system and the facturing and have increased production
bag material must be properly matched capability. This has impacted the cost
to achieve low emissions from the of bags. Improvements in the quality of
installed fabric filter system. coatings and membranes for filter bags
also have resulted in improved filter
FABRIC FILTER MATERIALS. Specification performance and bag life, particularly in
of the weave, thread count, weight, very harsh gas environments.
and finish of the cloth for a required
filter bag lists only a few of the pos-
sible combinations of manufacturing Table 5-2 | Summary of filter media*
FABRIC FILTER PERFORMANCE. Unlike represent 10% or more of the total cost
fabric filters, most other particulate of the equipment, a significant operat-
emission control devices depend on ei- ing expense can occur if the bag life is
ther pressure drop or collection surface shorter than expected. Typical bag life
area to reach the required collection is 3 to 5 years (depending on the type
efficiencies. Upsets to the predicted of cleaning mechanism selected and the
gas volumes or particulate loadings design parameters used in that selec-
typically will cause these other types tion), so this operating expense item
of equipment to have higher emission can represent a significant cost over
rates despite their maintaining the the life of the equipment. In some
same fractional efficiency. The fabric cases, the shortened bag life also could
filter does not behave in this fashion, require an outage to change the bags
however, because the material col- and, thereby, result in loss of revenues
lected on the filter bags is the filtering from production. Shortened bag life
medium that captures the incoming can be caused by various operating
particulate matter. Although upset and design problems, among which
conditions may change the flange- are the following:
to-flange pressure drop, resulting in
an excessively fast buildup of cake, • Bag blinding, usually caused by
these conditions do not materially operating a fabric filter for frequent
change the final emission rate. It can or long periods at or below the
be said that a fabric filter represents a dew point. Such operation tends
constant-emission device. to plug or blind the cloth. Bag
The actual performance of a fabric blinding results in unacceptable
filter depends on specific items, such unit pressure drop that ultimately
as the air/cloth ratio, the permeability will become fan or process limiting
of the fabric/cake (particulate bleed- and require filter bag replacement.
through), the loading and nature of The effect on the filter cloth is that
the particulate (irregularly shaped or material builds up in the passages
spherical), the particle size distribution of the cloth body and restricts the
(fine vs. coarse), and to some extent, gas passage such that even after
the frequency of the cleaning cycle. By repeated cleanings, the lodged
careful selection of design components, material cannot be removed.
fabric filters can achieve removal levels • Bag erosion, resulting from high-
of higher than 99.9% of incoming velocity streams of fly ash-laden
particulates, or an outlet loading of less gases where the gas enters the filter
than 0.005 grain per dry standard ft3 bag or directly impacts the cloth
(0.11 mg/Nm3). in any way. Proper selection of the
air/cloth ratio and filter bag length
FABRIC FILTER BAG LIFE. A primary controls the high entry velocity
concern in the operation of a fabric at the bag-inlet inside collectors.
filter is the bag life, which affects bag Higher air/cloth ratios with longer
replacement frequency and resultant filter bags tend to increase this veloc-
cost. Because the initial cost of the ity in the entrance, or “neck” area,
filter bags, excluding installation, can of the bag. Velocities generally are
Control of Power Plant Stack Emissions | 5-29
will have to handle the volume, re- such conditions, that solenoids will fail
Removing ash from the receiving sulting in increased gas side pressure and that valve seats will become worn,
hoppers as continuously as drop. resulting in an inflow of air into the
With both precipitators and fabric hopper. With a pressure-pneumatic
possible is key to high availability
filters, if a plugged hopper cannot be ash removal system, air can be forced
of ash collection equipment. cleared and discharged through the into the hopper at pressures as high as
ash system in a normal manner, the 100 psig (6.9 bar) in a dense-phase sys-
ash must be emptied onto the hopper tem. With vacuum systems, the motive
room floor or the ground (creating a force for inducing air into precipitator
fire and safety hazard) and be manually hoppers or baghouse hoppers is the
removed.50 suction maintained in the precipita-
tor or baghouse by the boiler induced
SOLUTIONS TO FLY ASH HOPPER PLUGGING. draft fans, which can create a suction
The problem can be solved by: of as much as –20 to –30˝ WG (–3.7
to –5.6 cm Hg) in the precipitator or
• Removing the ash as continu-
baghouse. With either a pressure or
ously as possible so that it can
vacuum system, then, a pressure dif-
remain hot and uncompacted, thus
ferential can result in the introduction
preventing the hoppers from being
of large quantities of cool air into the
used for storage of the hygroscopic
hoppers, which can lead to the conden-
fly ash.
sation of moisture from the flue gas.
• Preventing air-in leakage that cools
Actual field temperature mea-
the fly ash below the acid or water
surements on both “hot” and “cold”
dew point.
precipitators have shown that even
• Regularly inspecting and maintain-
very active hoppers can have low (es-
ing, or replacing, fly ash intakes.
sentially ambient) temperatures. One
• Modifying hopper design for
plausible explanation for such a low
improved fly ash outflow.
internal temperature is leakage of air
• Enclosing hopper areas and using
into the hoppers through the fly ash
adequate insulation thickness to
intake valves located at the bottom of
minimize heat losses and cooling of
the hoppers. Very seldom have such
ash system hardware.
valves been considered “maintenance
• Heating hoppers such that the
items” that should be inspected and
walls do not drop in temperature
considered for replacement on a set
and have condensation form.
schedule. It has become apparent in
These suggestions, and other recom- plant after plant, however, that such
mendations on the design and sizing of valves cannot be expected to last for-
fly ash hoppers, are discussed further ever and that they are a source of the
below and in the following chapter on low-temperature conditions occurring
ash-handling systems. in many hoppers.26
vapor, aqueous ammonia, or urea solution is injected into the hot flue gas
at a location in the boiler where the temperature is sufficient to drive the fol-
lowing reactions:
The remaining flue gas then passes through the rest of the boiler and the ex-
haust gas train to the stack.
SCR SYSTEMS
In the SCR process, the exhaust gas exits the boiler economizer or similar heat
recovery section and proceeds to the SCR system. Near the entrance of the SCR
system, a small amount of ammonia is distributed throughout the exhaust gas
stream to achieve the proper molar ratio blend of ammonia and NOx. The am-
monia and NOx-containing exhaust gas is introduced into the SCR reactor that
contains the catalyst. The chemical reactions between the NOx and the ammonia
occur as the exhaust gas passes through the catalyst layer(s). The exiting exhaust
gas stream, now with a much lower content of NOx and ammonia, continues
through the balance of the heat recovery equipment and other air emissions con-
trol equipment. The exhaust gas stream with the SCR end products finally exits
through the stack.
In practice, most SCR systems can achieve NOx removal efficiencies of between
80 to 90%. Because of the relatively high cost of an SCR system, lower NOx
emission levels usually can be obtained by less expensive, alternate technologies. In
theory and laboratory experiments, SCR simulations can obtain very high levels
of NOx removal. Practical SCR system design and operating limitations tend to
influence the amount of NOx removal that may be obtained in an operating unit.
The factors that determine the NOx removal capability include the operating
environment, installed equipment costs (capital costs), catalyst life, and long-term
system operating and maintenance costs.
Tail-End Dust Systems. The tail-end Smaller reactor (than with high-dust Exhaust gas temperature may limit
systems) operation
SCR system is located downstream
of several other gas cleaning devices,
which are used to remove particulates, lower than typical SCR systems do,
SOx, and heavy metal compounds exhaust gas reheating is required (see
from the exhaust gas stream. The Fig. 5-21). The reheating can be done
exhaust gas coming to the SCR system by directly or indirectly fired equip-
is essentially free of catalyst-damaging ment or by steam heat exchangers.
compounds, so no special equipment is Gas-gas heat exchangers between the
required to address these related issues. SCR incoming and exiting gas often
Because the other cleaning devices are included to minimize the reheating
often operate at temperatures much demand of these systems. Advantages
5-36 | Clean Combustion Technologies
and ammonia, is directed to the remaining heat recovery and emission control
equipment.
Catalytic Reaction Concepts. The catalytic reactor can be simply described as a plug
flow reactor. The following equation relates the amount of NOx reduced in the
reactor with the ammonia reaction rate at the catalyst:
Equation 5-5
CNOx,i VA 1
VA = =
FNOx,i Q Av
Equation 5-6
where the area velocity (Av) is related to the space velocity (Sv) as shown:
Q S
= v = Av
AV A
Equation 5-7
The fraction of NOx reduced across the reactor also can be defined as
(CNOx,i − CNOx,o ) = 1 − (C
x NOx =
CNOx,i
NOx,o CNOx,i )
Equation 5-8
5-38 | Clean Combustion Technologies
where
CNOx,i = concentration of NOx entering the reactor (inlet)
CNOx,o = concentration of NOx exiting the reactor (outlet)
The SCR reaction follows a first-order relationship with respect to NOx and
can be stated as
where k is the reaction rate constant. Substituting the reaction rate expression
above and integrating the right-hand side of Equation 5-5 yields
1
K = AV ln
1 − x NOx
Equation 5-10
where K is the global reaction rate constant. The equation above is useful for
conditions in which ammonia is not a limiting factor, but there are conditions
when the amount of ammonia injected into the SCR system is restricted to mini-
mize the amount of unreacted ammonia exiting the reactor. Using the definition for
stoichiometric ratio (SR) between the inlet NOx and NH3 concentrations entering
the catalyst:
CNH3 ,i
SR =
CNOx,i
Equation 5-11
Cho52 related the reaction rate of NOx to the reaction rate of ammonia across
the catalyst by
x
− rNOx = − rNH3 = kCNH3 = kCNH3 ,i 1 − NOx
SR
Equation 5-12
and using this reaction rate equation, the reactor equation can be restated as
SR
K = AV ln
SR − x NOx
Equation 5-13
This form is useful when the ratio between inlet NH3 and NOx concentrations,
SR, is less than 1.
The global reaction rate constant consists of rate constants for both the mass
transfer rate of NOx and NH3 reactants from the bulk exhaust gas to the catalyst
Control of Power Plant Stack Emissions | 5-39
surface (km) and the rate of reaction and pitch is defined as the unit dimen-
between NOx and NH3 on the cata- sion of the catalyst cell or channel:
lyst surface (kr):
pitch = a + b
K–1 = k–1 –1
m + kr Equation 5-17 b
Equation 5-14
Pressure drop across the catalyst
The term km mostly depends bed can be estimated using the Darcy
on the gas velocity in the catalyst equation:
cells, the catalyst cell geometry and a
dimensions, and the diffusivity of lv 2
p = 4 f
NOx in the exhaust gas. The term kr 2b
Figure Clean
5-22 Combustion
| Square-channel configuration
Technologies
primarily depends on the catalyst’s
Equation 5-18 Figure: 05.22/ SPS
chemical composition and internal PMS5415 K
pore structure as well as the exhaust where
gas temperature.
∆p = draft loss across the catalyst
Catalyst activity tends to degrade
body
over time because of a variety of fac-
f = friction factor
tors, such as catalyst poisoning, fouling,
l = catalyst body length
and erosion. Deactivation of SCR cata-
v = gas velocity through the cata-
lysts can be commonly described as an
lyst body
exponential decay function over time:
= gas density
K = K0(1 – D)t b = catalyst channel hydraulic
diameter
Equation 5-15
where and the friction factor for the catalyst
can be determined using the Haw-
K = catalyst activity at time t
thorne correlation53:
K0 = initial catalyst activity at time
zero
D = deactivation rate A 0.0445bNRe
4f = 1+
t = time NRe l
Equation 5-19
For a given catalyst configuration,
the area density is the amount of cata- where
lyst surface per unit catalyst volume.
NRe = Reynolds number
For a square-channel configuration
A = 53 for triangular channels, 57
(see Fig. 5-22), the area density can be
for square channel, and 96 for
calculated using the expression
parallel plates (dimensionless)
4b 4b
A = = SCR Process Chemistry
(a + b ) 2
pitch 2
Before the catalytic reactions can occur
Equation 5-16 between NOx and NH3, these reac-
where tants must first move from the bulk
a = catalyst wall thickness dimension fluid exhaust gas phase to the catalyst
b = catalyst channel opening body surface, as shown in Figure 5-23.
dimension This involves film diffusion, or external
5-40 | Clean Combustion Technologies
O2, NH3, NO
Effect of NOx and NH 3 .As stated earlier,
the NOx-NH3 reaction is considered
Step 3: Chemisorption of Reactants NH3 to be first order with respect to NOx.
Step 4: Reaction between Molecules
This means that the fraction of NOx
reduced across a given catalyst body is
independent of the inlet concentration
of NOx when ammonia is not a limit-
NO
ing factor. When ammonia is a limiting
3H H factor, the fraction of NOx removed
3H 2O
Step 5: Reverse Diffusion of Reaction Products N O
N (xNO) is directly proportional to inlet
Step 6: Regeneration of Active Site, Diffusion
SR between NH3 and NOx. As the
SR exceeds unity, the relationship devi-
ates because of greater inefficiencies,
Figure 5-23 | Diffusion and chemisorption on catalyst and xNO never exceeds 100%. As the
Clean Combustion Technologies
value for SR increases, the ammonia
Figure: 05.23/JT mass transfer, as denoted by km. At slip, or unreacted ammonia, exiting the
PMS5415 K the catalyst surface, the reactant gas catalyst beds also increases. For many
species must continue their journey practical operating environments, the
through the small pores that exist on amount of ammonia slip often is a lim-
the catalyst body surface. The catalyst iting factor with regard to the amount
active sites, where the chemical reac- of ammonia delivered to the SCR
tion occurs, are found within these system. For this operating scenario, the
small pores. The internal reaction rate, amount of NOx removal is strongly
kr , accounts for both the diffusion of dependent on the inlet NOx concen-
the reactants within the pores and the tration. These relationships are shown
reaction on the catalyst active sites. in Figure 5-25.
At the start of the catalytic reac-
tion cycle, ammonia is adsorbed on Effect of Temperature. The catalysts
a catalytic active reaction site (Fig. used in SCR systems have a strong
5-24, top portion of reaction schematic). dependence on operating temperature.
Then, NO (NOx) combines with the Upper and lower temperature limits
adsorbed ammonia to form an NOx- define the operating temperature
NH3 complex (see Fig. 5-24, right-hand window for a given catalyst and ap-
side of the reaction schematic). The plication design. At low temperatures,
NOx-NH3 complex decomposes to the SCR reaction is strongly kineti-
form nitrogen and water vapor, which cally limited by the reaction rate on
Control of Power Plant Stack Emissions | 5-41
90
volume, the activity of the catalyst can Equation 5-22 25
be suppressed. In these cases, cata- 80
lyst formulation and design must be The formation of SO3 is undesirable. 70
specially considered for the application, Increased SO3 concentrations raise the
or additional air may be delivered to acid dew point of the exhaust gases, 60
20 100
than 1% by volume, the influence on that can influence the amount of fine 15
10
catalyst activity is not significant. particulate matter in the gas stream 5
(PM2.5). For combustion systems 0.6 0.8 1.0 1.2
Effect of SOx.The presence of SOx, firing very low-sulfur natural gas, a NH3 / NOx Mole Ratio
SO2, and SO3 can have an effect on higher degree of SO2 oxidation and
limiting the operating temperature resulting SO3 levels usually can be
Figure 5-25 | NOx reduction and ammonia
range of the catalyst. The catalyst in tolerated. For higher-sulfur fuels, such slip versus NH3/NOx mole ratio
the SCR system also will oxidize SO2 as many types of coal, a lower degree Clean Combustion Technologies
Figure: 05.25/ SPS
PMS5415 K
5-42 | Clean Combustion Technologies
quantity of these two potential poisons that is present in the fuel sources. Because
of the volatile nature of these compounds, such analysis should involve the fuel itself
rather than the post-combustion ash. The poisoning effect of fuels with high levels
of arsenic or with high arsenic and low calcium levels may be mitigated by adding
additional calcium, often as limestone, to the fuel supply that is fed to the burners.
The alkali compounds also can bind with the catalyst active sites via another
route. These compounds, often present as salts in the ash, can leach out and com-
bine with the catalyst when excessive amounts of moisture form on the catalyst
body. High-calcium fuels pose a pore-masking risk as well, because calcium may
combine with SO3 to form calcium sulfate when both of these are present in the
catalyst pores. Poisoning by alkalies can best be avoided by careful ash deposit
maintenance and by keeping the catalyst dry during off-line storage. The effect of
catalyst poisoning may be permanent, but it can be addressed using careful wash-
ing treatments performed off-line.
REAGENT UTILIZATION. As noted in the reaction equations between NOx and NH3,
the reaction stoichiometry requires one molar amount of ammonia to react with
one molar amount of NOx. Careful SCR system design can result in operation at
a very high ammonia usage efficiency. Thus, the amount of ammonia required for
a given system can be determined by the amount of exhaust gas to be treated, the
amount of NOx to be removed, and the amount of ammonia slip (or unreacted
ammonia) that can be tolerated:
( )
CNH3 ,i = CNOx,i − CNOx,o + CNH3 ,o = CNOx,i x NOx + CNH3 ,o
Equation 5-25
typically occurs on two levels: overall gas flow rate. Feed-forward signals
ammonia flow delivery control and can be taken from boiler load or inlet
localized ammonia distribution. Both NOx concentration measurements.
features are crucial for effective ammo- This simple approach works well for
nia use and successful NOx removal relatively low NOx removal efficiencies
within the SCR system. Each of these or for applications in which maintain-
features involves unique and indepen- ing ammonia slip levels is not difficult
dent equipment in order to meet their to accomplish.
respective objectives. Supplemental measurements of O2
The overall control of ammonia fed or CO2 may be used to account for
to the SCR system is designed to meet changes in NOx concentration caused
the overall NOx removal efficiency by air in-leakage. In general, reliable
and ammonia slip targets. Several NH3 slip measurements are difficult
strategies can be used to control the to make at ppm-level concentrations.
total ammonia fed to the SCR system. Some in-duct analytical methods
One feed-forward method relies on have shown promise as supplemental
determining the inlet NOx load to operating information. For applica-
the SCR system based on inlet NOx tions in which ABS formation is a risk,
concentration and exhaust gas flow the symptoms of salt deposits, such as
rate measurements. An appropriate heat exchanger draft loss or thermal ef-
amount of ammonia is fed to the SCR ficiency, may be monitored for changes
system based on the SR required to and an indication of excessive ammo-
achieve an NOx removal efficiency nia slip.
target. The outlet NOx concentration Very little mixing occurs within the
measurement signal is used to trim the SCR reactor. The treatment of NOx
ammonia flow and avoid excessive am- emissions with ammonia depends on
monia injection. creating an ammonia flow profile that
Another control method relies on matches the exhaust gas NOx profile.
measuring NOx removal load on the To accomplish this objective, distri-
SCR system based on both inlet and bution of ammonia throughout the
outlet NOx concentrations and the exhaust gas duct is a critical objective.
exhaust gas flow rate. The amount of Carefully designed ammonia distribu-
ammonia is adjusted to achieve the tion equipment is located upstream
desired NOx removal load or efficiency of the SCR reactor to accomplish this
level. The outlet NOx concentration task. An SCR system also may rely on
signal may be used to trim excessive ammonia slip exiting an SNCR system
ammonia flow. Inlet factors, such as for some of the ammonia feed. For
boiler load and inlet NOx load, can these systems, the ammonia control
be used as feed-forward signals to and distribution equipment must
anticipate changes in the ammonia account for the ammonia slip profile
load demand. exiting the SNCR system.
Still another control method is For a typical ammonia distribution
based on maintaining a NOx emission system, the ammonia is introduced
level either as a concentration or as a in relatively high concentrations at
load based on the outlet NOx con- various locations across or around the
centration measurement and exhaust duct. At this point in the SCR system,
5-46 | Clean Combustion Technologies
the exhaust duct contains a few small equipment also must be provided. Al-
regions with feed gas that has a very though it is not typically considered as a
high ammonia concentration and that is component, the ductwork between the
surrounded by exhaust gas that has zero SCR system components will influence
ammonia content. This high-ammonia the exhaust gas behavior, consequently
feed gas must diffuse and blend with the affecting the performance of the respec-
NOx-containing exhaust gas. Gas mix- tive downstream components.
ers often are located either upstream or
downstream of the ammonia injection Catalyst. Most SCR catalysts are com-
location in order to blend the ammonia posed of a chemical blend of vanadium
throughout the exhaust gas stream. oxide (V2O5) and titanium dioxide
A simple design assumption for (TiO2). The vanadium oxide compo-
the ammonia distribution equipment nent provides the catalytically active
is that all of the individual ammonia sites for the SCR reaction, whereas the
injection points will require equal titanium dioxide component provides
amounts of ammonia feed gas. For the support and the micropores for
many real-world applications, however, distributing the active sites. Other
this assumption is not true. A need catalyst formulations may be based on
exists to bias the amount of ammonia zeolite materials, precious metals (e.g.,
at each injection location to meet that platinum), or other base metals (e.g.,
location’s corresponding NOx load. iron), each with their own respective
Most injection equipment also features ranges of operating temperature.
some form of flow adjustment to The physical catalyst structure can
modify the relative ammonia delivered exist as a homogeneous structure, in
to each injection location. Another which the catalyst body is entirely
strategy to obtain proper ammonia catalyst material, or as a heterogeneous
and NOx distribution is to rely on catalytic coating, supported by a me-
intensive mixing downstream of the chanically strong but catalytically inert
ammonia injection location to smooth structure. The inert structure may be
both NOx and ammonia distribution made into a mesh fabricated from metal
profiles. For cases with severe and or a ceramic material, such as fiberglass.
variable NOx concentration profiles, Catalysts used in ash-free exhaust gas
it may be necessary to create multiple applications may consist of a thin coat-
ammonia injection zones within a ing of catalyst on a metal foil or rigid
single SCR system. Each ammonia ceramic body, such as cordierite.
injection zone would be equipped with The catalyst body structure is
separate and complete control systems designed to maximize both the surface
in order to control the amount of am- area in a given catalyst volume and the
monia in each zone. frontal open area as well as to produce
a low resistance to gas flow. Most cata-
SCR Components lyst bodies are monoliths composed
Any typical SCR system is composed of many small cells or channels with
of the following basic components: thin walls that are oriented parallel to
catalyst, reactor, and ammonia the direction of gas flow. Some catalyst
injection-distribution system. For products are composed of packed beds
high-dust applications, ash-handling of small, shaped pellets or beads.
Control of Power Plant Stack Emissions | 5-47
the large particles on screens or other ammonia is injected directly into the
similar devices that are periodically exhaust gas duct, dual-fluid nozzles are
cleaned, driving the large ash down- used, where high-pressure air is sup-
ward. The second means of addressing plied to atomize the ammonia solution
the large-particle ash is removal of into a fine spray for evaporation and
the ash from the SCR system. This distribution.
generally is done using conventional Although ammonia injection nozzle
ash hoppers, typically located upstream location and arrangement may take
of the SCR reactor and adjacent to an into account local NOx load varia-
upward flow-oriented duct. Hoppers tions, additional ammonia mixing may
should be designed to ensure that the be required to achieve the proper
ash introduced into the hopper does ammonia to NOx ratio distribution
not re-entrain back into the exhaust across the catalyst inlet face. Some am-
gas flow. monia injection system designs utilize
the mixing effects that the ammonia
Ammonia Injection Grid. One of the supply piping provides. Another ap-
primary components of any ammo- proach to achieving good ammonia to
nia injection system is the injection NOx ratio distributions is to use static
nozzle, which is used to introduce the gas-mixing devices. Static mixers may
ammonia into the exhaust gas and be perforated plates or some variation
distribute the ammonia across the of baffles. Baffle shapes may be ellipti-
exhaust gas duct. Nozzle designs can cal, triangular, or rectangular, and they
range from very simple (i.e., a hole in may be interconnected or freestanding.
a pipe) to very complex, depending These devices may be placed upstream
on the requirements of the applica- of the ammonia injection nozzle grid
tion and the other components of the to address NOx load imbalances.
ammonia subsystem. Regardless of the Mixers also may be placed downstream
design, the ammonia injection nozzle of the nozzle grid location to achieve
also must be able to perform effectively better diffusion of ammonia. To obtain
across a range of ammonia injection a suitable ammonia to NOx ratio dis-
rates and, in high-sulfur or high-dust tribution in any given application, there
applications, to avoid plugging. The usually is a design balance between
total number of nozzles, or the spacing injection nozzle spacing and the use
between the nozzles for a given ap- of static mixers. A high nozzle density
plication, is influenced by the mixing offers better tuning opportunities,
distance between the injection location but with more operating complexity.
and the catalyst beds as well as by the Aggressive gas mixing offers simplicity
design philosophy of the ammonia and robustness, but with a draft loss
subsystem. The performance of a given penalty. Additionally, the duct arrange-
ammonia injection nozzle arrangement ment design often influences which
often is validated using models, such as mixing strategies are feasible.
computational fluid dynamics or physi- The supply of ammonia to the injec-
cal scale models, at the design stage. tion nozzles is accomplished through
For ammonia-dilution air systems, a piping distribution system from a
the nozzles are based on single-fluid central ammonia vapor supply. This
designs. For systems in which aqueous piping system is designed to maintain a
Control of Power Plant Stack Emissions | 5-51
Mixing
Ammonia Injection Grid
Chamber
Dilution Air
FIC
Flue Gas Duct
PIC
QIC
PIC
Unloading
Compressor
LI TI PI
PIC
Circ. Heat
Evaporator
Pump Exchanger
Railway
Tank Car Storage Tank
as reducing agents into the flue gas key design considerations. All of these
through a series of injectors. The factors must take into account the
amount of reagent injected is suitable exhaust gas operating environment,
to reduce the NOx concentration and composition, and flow. Certain boiler
to avoid creating excessive ammonia types have inherent mixing dynamics,
slip at the reactor exit. The injectors such as circulating fluidized bed (CFB)
are designed to distribute the reagent boilers, which can assist the degree of
across the reaction zone and blend mixing possible. Powerful modeling
with the exhaust gas. Furnace model- techniques, such as computational fluid
ing usually is required to optimize the dynamics, often are used to develop the
injector location and mixing parame- optimal injector designs for the best
ters. At this high-temperature location, mixing possible.
the injected reagent vaporizes and Upon entering the hot exhaust gas
decomposes into ammonia radicals. stream, the reagent droplet begins to
These radicals then react with and evaporate and release NOx-reducing
reduce the NOx into nitrogen, nitrous vapors as it travels across the SNCR
oxide, and water. The resulting exhaust reactor section in the boiler. The injec-
gas, with lowered NOx and ammo- tor nozzle design and spray droplet
nia contents, is directed through the characteristics must achieve suitable
remaining steam generation sections of distribution and mixing while ensuring
the boiler. adequate droplet evaporation. Modeling
methods often are used to help deter-
Thermal SNCR Reaction Concepts. The mine the best suitable reagent injection
SNCR thermal reactor has to accom- criteria. After evaporation is complete,
plish four basic tasks to successfully the reducing reagent must continue to
operate: heat, up to the point where the reagent
begins to decompose into ammonia
1. Mixing the injected reagent
radicals. These radicals then can react
throughout the exhaust gas
with the NOx in the hot exhaust gas.
stream.
2. Evaporating the liquid reagent
SNCR Process Chemistry. As stated
supply, if supplied as a solution.
above, the NOx-reducing reagent
3. Decomposing the ammonia or
must decompose to form radicals.
urea reagent into ammonia-free
For ammonia, the decomposition
radicals.
reactions are:
4. Reacting the free radicals with the
NOx compounds.
NH3 + OH ↔ NH2 + H2O
The design of the reagent injectors is Equation 5-27, Step 1
one of the most critical aspects of the
SNCR system. The placement of the
injectors and their relation to the boiler NH3 + O ↔ NH2 + OH
arrangement is a primary consider- Equation 5-27, Step 2
ation. The choice of the injector nozzle
spray direction, the spray pattern, and
H + NH3 ↔ H2 + NH2
the droplet size and velocity also are
Equation 5-27, Step 3
5-56 | Clean Combustion Technologies
For urea, the reactions are: The NOx reduction reactions are
considered to be the following:
CO(NH2)2 → NH3 + HNCO
Equation 5-27, Step 4 NH2 + NO → N2 + H2O
Equation 5-27, Step 10
HNCO + H ↔ NH2 + CO
Equation 5-27, Step 6 HN2 → N2 + H
Equation 5-27, Step 12
The contribution of hydrogen, oxy-
gen, and hydroxyl radicals are proposed The termination reaction is consid-
by reactions such as the following: ered to be the following:
H + O2 ↔ OH + O NH2 + O ↔ NO + 2H
Equation 5-27, Step 7 Equation 5-27, Step 13
0
Effect of Temperature. The reactions 1,200 1,400 1,600 1,800 2,000
Temperature, °F
that make up the SNCR process are
very dependent on temperature. Typi- Ammonia Urea
cally, the SNCR reactions best occur Figure 5-33 | Effect of temperature on NOx reduction
in the temperature range of 1,600 to Clean Combustion Technologies
Figure: 05.33/ SPS
2,100°F (870 to 1,150°C), although PMS5415 K
this window is somewhat reagent de- of heat and mass transfer, and the
pendent. Additives, often proprietary, longer the time for the NH2-NOx
may be used to modify the operating reactions to occur. All of these condi-
temperature range, and the degree of tions favor a higher degree of NOx
NOx reduction by SNCR at various reduction. In SNCR applications, the
temperatures is shown in Figure 5-33. residence time may be as brief as 0.01
When the exhaust gas temperature is second or as long as several seconds. In
too low (<1,600°F [<870°C]), then general, SNCR applications may have
the reaction kinetic rate is very slow, residence times of approximately 0.3
and high levels of ammonia slip can second. Residence times much longer
occur. If the exhaust gas temperature than 0.5 second do not normally pro-
is too high (>2,100°F [>1,150°C]), vide much additional NOx reduction.
then the ammonia reagent can become The effect of residence time on the
oxidized and begins to form NOx, as NOx reduction operating temperature
indicated in Equation 5-27, Step 13. window is shown in Figure 5-34.
Effect of Residence Time. The residence Effect of Inlet NOx Concentration. The
time of the SNCR reactor is the time reaction rate for NOx reduction is di-
during which the boiler exhaust gas rectly dependent on the concentration
and reducing reagent are at the effective of both NOx and ammonia (reduc-
operating temperature conditions to- ing) reagents. As the concentration
gether. The residence time classically is of NOx increases, the rate of NOx
determined by the volumetric flow rate reduction also increases, as shown
of the exhaust gas relative to the boiler in Figure 5-35. As the level of NOx
gas path dimensions: The greater the decreases, the amount of NOx reduc-
residence time, the greater the degree tion as well as the optimal operating
5-58 | Clean Combustion Technologies
50
between NOx removal and ammonia
40 slip is depicted in Figure 5-37, which
30 shows that the increasing ammonia
slip rate does not equally match the
20
gains in NOx reduction.
10 Some practical designs tend to
0 favor operation at a relatively high
1,500 1,600 1,700 1,800 1,900 2,000 2,100
Temperature, °F
operating temperature, even though
this can result in lower NOx removal
30 ppm 70 ppm 200 ppm
efficiencies. The design relies on this
Figure 5-35 | Effect of inlet NOx concentration on NOx reduction
Clean Combustion Technologies
Figure: 05.35/ (JG)
PMS5415 K
Control of Power Plant Stack Emissions | 5-59
condition as a means to gain some the boiler wall and direct the reagent
50
reagent heating benefits and avoid stream from the boiler wall toward the
45
creating excessive ammonia slip while exhaust gas. A lance injector is a pipe
40
metering pump and flow control for on compressed air or steam to provide 0
0 5 10 15 20
reagent dosing. The reagent usually is the required energy for atomization. Ammonia Slip, ppm
supplied from a single, common source
for all injection locations. The reagent ECONOMIC CONSIDERATIONS
Figure 5-37 | Effect of ammonia slip on NOx
flow is diluted and blended with The decision of whether to use an reduction Technologies
Clean Combustion
pressurized, filtered water in order to SCR or an SNCR system for NOx Figure: 05.37/ (JG)
PMS5415 K
maintain consistent pressure and flow emission control is based primarily on
qualities, regardless of varying reagent the NOx removal efficiencies that a
flow rates. given application requires. In almost
all cases, an SNCR system is less
Reagent Injection and SNCR Reactor. costly to install than an SCR system.
The diluted reagent flow to each injec- The SNCR system, however, does not
tion location is distributed to each offer the same high capability for NOx
individual injection point based on removal that an SCR system does.
flow metering equipment and control The advantages of an SNCR system
valves. Each injection point is equipped are relatively low capital equipment
with an injector. Two types of injectors costs, simple operation, and minimal
are used in SNCR systems: wall and impact on boiler plant operation. The
lance. Wall injectors are mounted to reagent consumption cost makes up
5-60 | Clean Combustion Technologies
the majority of the SNCR system op- form condensable salts with sulfate and
erating costs. The disadvantages of the nitrate anions in the exhaust gas when
SNCR system are the aforementioned they cool to ambient temperatures.
lower capability for NOx removal, the Some emission permits and particulate
somewhat limited retrofit ability, and emission testing methods specifically
the operational problems with respect address these types of emissions. The
to changes with boiler load and other potential presence of ammonia should
operating conditions. be considered for these applications.
The major advantage of the SCR For high-dust applications, the
system is the relatively high capability ammonia slip exiting the SNCR or
for NOx removal. Other advantages SCR system can deposit on the ash
are that it is easier to retrofit (no pres- as the exhaust gas temperature cools.
sure part disruption), tolerant of boiler This ash is collected by the particulate
load changes, and relatively simple to collection equipment. In general, most
The application of SCR or operate. The disadvantages are the particulate collection equipment will
SNCR depends upon the type relatively high capital equipment cost, tolerate small amounts of ammonia
sensitivity to fuel qualities, and greater on the ash. The concern is when the
of fuel, the type of boiler, and
impact on boiler plant operations. The collected ash is disposed of or sold for
the required level of NOx SCR system operating costs mostly use in construction products. If the
reductions. result from reagent consumption, ash is wetted or used in a water-based
catalyst replacement, and operating mixture with an alkaline pH, am-
draft losses. monia vapor can evolve from the wet
ash. This can be a nuisance, or even a
potential safety risk, if personnel are in
POST-SNCR and POST-SCR EFFECTS
close contact with the ammonia vapors.
Ammonia Slip For most applications, the ammonia
Ammonia that is not consumed by content in the ash does not affect the
the NOx-NH3 reaction will exit the quality of the end product containing
SNCR or SCR system as ammonia the ash.
slip, and this ammonia can pose several
risks. For low-dust exhaust gas applica- AMMONIA SALTS. When high levels of
tions, one potential risk is the emission ammonia are present with high levels
of ammonia as ammonia species. For of SO3, a potential exists for the
applications with emission scrubbers formation of ABS and ammonium
located downstream of the NOx treat- sulfate salts:
ment equipment, the ammonia can
transfer from the exhaust gas into the NH3 + H2O + SO3 → NH4HSO4 (ABS)
scrubbing liquor. The choice of scrub- Equation 5-28
ber and associated liquor treatment
equipment should consider the impacts
of ammonia. Spent liquor discharge 2NH3 + H2O + SO3 → (NH4)2 SO4
also will be a source of liquid-based Equation 5-29
ammonia emission.
Another potential concern regarding These solid salts can form when
ammonia emission is its potential as the temperature of the exhaust gas is
a particulate emission. Ammonia can cool enough and the concentrations
Control of Power Plant Stack Emissions | 5-61
of NH3 and SO3 are high enough. the SCR system also should consider
These salts are somewhat corrosive and the impact of additional, fresh catalyst
can cause plugging problems because layers that may be added in the later
of their melting point behavior. The years of operation. One special risk
ABS salt has a very sticky nature and is when the fuel contains appreciable
can bind with particulate emissions amounts of vanadium metal (i.e., heavy
to create plugging concerns. This po- oils). In these cases, the vanadium
tential effect should be considered for emissions from the oil combustion will
equipment located downstream of the deposit on the catalyst and increase the
SCR system, such as the GGH or the amount of SO3 formation during the
particulate collection equipment. With lifetime of the catalyst.
respect to heat exchangers, special As stated above, SO3 formation can
equipment designs and materials may contribute to ammonium salt formation.
be used in conjunction with SNCR The SO3 in the exhaust gas also can
and SCR systems to minimize the be considered as a potential source
formation of the ash-ABS deposits and for condensable particulate emissions
better facilitate equipment cleaning if or may be regulated directly as an
they do form. SO3 emission concern. Equipment
Because of the detrimental effects located downstream of the SCR
of ABS in high-ash systems, care must system can face impacts from the
be taken to ensure that no localized elevated SO3 level when the exhaust
high ammonia concentrations exist gas cools. The increase in SO3 across
downstream of the SNCR or SCR the SCR system will result in a higher
system. Localized ABS deposits can acid dew point temperature for the
cause problems even when the overall exhaust gas, which in turn may result
conditions do not favor ABS forma- in greater corrosion impacts. This
tion. A small region of high ammonia often is observed at the GGH and the
concentration can lead to very large exhaust gas ductwork immediately
areas of ABS formation if rotating heat following this device. Wet scrubbing
exchangers are used or exhaust gas flow systems may collect some of the SO3
patterns shift because of plugging. emissions; the design and operation
of such equipment should take this
SO3 Emissions into account. Lastly, any SO3-specific
One of the detrimental side reactions emission control equipment will be
of the SCR catalyst is the oxidation of impacted by the higher SO3 levels.
SO2 to SO3:
CONTROL OF SO2 EMISSIONS
2SO2 + O2 → 2SO3 Reduction of SO2 emissions has been
Equation 5-30 achieved through one or a combination
of the following:
In general, the concentration of SO3
in the exhaust gas doubles as it passes • Fuel switching (to lower-sulfur
across the SCR catalyst beds. As the fuels)
catalyst ages, the amount of SO3 formed • Coal cleaning (use of fuel desulfur-
tends to decrease. Any estimate of the ization methods)
lifetime potential for SO3 formation in • FGD systems
5-62 | Clean Combustion Technologies
The first two—low-sulfur fuel and such as spray dryer absorbers (SDAs)
coal cleaning—can be utilized in and flash dryer absorbers (FDAs), have
retrofit applications where appropri- particulate collection as an integral part
ate emission regulations can be met. of the environmental equipment and
In many locations (particularly the are less susceptible to particulate-based
European Union and the United problems.
States), air permits for new power
plants require a percentage reduction FGD SYSTEMS
in SO2 emissions regardless of the The FGD technology was first tried
sulfur content of the fuel. in London, England, in 1935, when a
Fuel desulfurization processes local power station used the alkaline
range from conventional coal washing Thames River water in a wooden
to coal liquefaction and gasification. absorber for scrubbing flue gas. The
Coal washing is effective in reduc- visibility of the vapor plume caused
ing the pyritic sulfur in the coal. It abandonment of the process early in
generally does not remove the organic World War II, however, because it was
sulfur from the coal matrix, however, an easily seen target. After the war,
so it is limited to sulfur reduction of FGD technology remained dormant
less than 50%. Coal liquefaction was until the mid-1960s, when the United
developed in Germany and used in States and Japan began to institute
South Africa. The economics were environmental legislation. Since then,
not favorable at low oil prices, but more than 50 FGD processes, differing
there has been renewed interest with in their chemical reagents and resultant
higher oil prices. Coal gasification has end products, have been developed.
been developed as an alternative to Initially, FGD processes were dif-
PC or CFB technology. At present, ferentiated as either regenerable or
only a small number of demonstra- non-regenerable and as either wet or
tion plants are in operation. The CFB dry. A non-regenerable, or “throwaway,”
technology captures the sulfur in the system produced waste products of
fluidized bed during the combustion various sulfur compounds that were
process (see Chapter 3). Therefore, stored in ponds or treated for use as
this chapter will focus on post-com- landfill. A regenerable system recov-
bustion FGD processes. ered the SO2 in some commercially
The FGD systems capture the useful form, such as elemental sulfur
primary acidic gas SO2 and typically or sulfuric acid. The original alkali was
provide the additional benefit of also “regenerated” for repeated use in the
capturing the secondary gases, such absorption/regeneration cycles. A fur-
as SO3, HCl, and hydrogen fluoride ther differentiation was made between
(HF). Additionally, FGD systems wet processes and dry processes: Wet
provide capture of particulate matter, processes saturate the flue gas with
although particulate matter capture water vapor, and dry processes do not.
can be detrimental in wet FGD The most dominant FGD process
systems, either by causing chemistry today is the limestone forced oxidation
problems or by fouling the gypsum system, which is not a throwaway pro-
by-product. Semi-dry FGD systems, cess, because the gypsum by-product
Control of Power Plant Stack Emissions | 5-63
Wet FGD–
Air Fabric Filter
SCR Limestone/ Wet Stack
Pre-heater or ESP
Gypsum
Arrangement A: Common Arrangement for Control of NOx, SO2 and Particulate Matter (PM) With a Wet Stack
Wet FGD–
Air Fabric Filter
SCR Cooler Limestone/ Heater Dry Stack
Pre-heater or ESP
Gypsum
Arrangement B: Adding Gas-to-Gas Heat Exchangers (Cooler and Heater) Allows a Dry Stack
Wet FGD–
Air Fabric Filter
SCR Limestone/ WESP Wet Stack
Pre-heater or ESP
Gypsum
Arrangement C: Adding a Wet Electrostatic Precipitator (WESP) to Arrangement A adds SO3 Removal
Dry FGD
Air Fabric Filter
SCR with Dry Stack
Pre-heater or ESP
Landfill
A heat exchanger (see the section SCR/fabric filter or ESP with MERCURY
immediately following), typically a gas- Capture/Wet FGD/Wet ESP/CO 2 Cap-
gas, rotating, Ljungström®-type heat ture/Wet Stack. The state of the art
exchanger, extracts heat from the wet in emission control includes potential
FGD inlet flue gas and then heats the mercury capture as well as CO2 cap-
saturated flue gas exiting the wet FGD ture and sequestration. Subsequent
absorber. Seals are used to minimize sections will address these technolo-
leakage, but a small amount of untreat- gies in more detail.
ed flue gas can bypass the absorber.
Another heat exchange system (called Reheaters and Stacks
the “non-leak heat exchanger,” because The wet stacks in some of the arrange-
there is zero leakage) is available where ments described above are common
the heat exchange is accomplished in North America. Japan and Europe
through gas-liquid coupling to finned typically eliminate the opaque water
tubes. This equipment is larger and plume by heating the gas to tempera-
higher in capital cost because of the tures well above the dew point. The
costs involved with corrosion preven- decision to reheat the flue gas is almost
tion and heat transfer (size); thus, always a local issue and is driven by
these units are infrequently used. the need to satisfy one or more of the
following objectives:
SCR/fabric filter or ESP/Wet Scrubbing/
• Eliminate or reduce the visible
Wet ESP/Wet Stack. In a wet scrubber,
plume of condensed water vapor.
SO3 is difficult to control, because
• Reduce corrosion in downstream
it condenses to very small droplet
ducts, dampers, fans, and stack liners.
diameters and passes through the
• Enhance plume rise and dispersion
gas absorber as a particle. Adding a
of residual emissions.
wet ESP after the wet FGD is very
• Meet government-mandated regu-
effective in capturing this submicron
lations for the exit temperature of
particle and in preventing the high
stack gas.
opacity that SO3 often causes in con-
centrations higher than approximately The following reheat options are the
4 to 5 ppm. most common:
• No reheat (wet stack operation)
SCR/fabric filter or ESP/Semi-Dry
• Bypass reheat
FGD/Dry Stack. The use of a spray dry
• Direct reheat with an in-duct heat
absorber or FDA with an ESP or
exchanger or direct combustion
fabric filter cools the flue gas, but not
burner
to saturation as a wet FGD absorber
• Indirect reheat with hot air mixing
does. This allows use of a dry stack
• Regenerative reheat with rotary
at the expense of slightly lower SO2
regenerative or shell/tube-type heat
removal. Use of a fabric filter for par-
exchangers
ticulate control also will capture the
• Reheat with liquid-coupled heat
fine SO3 particles. The mixture
exchangers
of calcium/sulfur salts and fly ash
may be used as roadbed material or No Reheat/Wet Stack. The “no reheat”
simply landfilled. option yields significant savings in
Control of Power Plant Stack Emissions | 5-67
energy and has been dominant in Gas or oil is burned, and the combus-
the United States since 1980. High- tion product gas (at 1,200–3,000°F Stack gas reheat can reduce
efficiency mist elimination, resulting [650–1,650°C]) is mixed with the flue or eliminate, at the expense
in minimal amounts of water entrained gas to raise its temperature.
of additional energy costs, the
into the outlet duct and stack, have Direct firing requires some care in
produced good wet stack operating mixing the hot combustion gas with white plume of water vapor that
results. Stack “rain-out” is minimized the cool, scrubbed gas. If mixing is forms when the flue gas cools.
by using smooth stack liners and by not effective, hot spots can develop
operation at moderate velocities of 50 downstream from the heater, causing
to 60 ft/s (15 to 18 m/s). damage to the duct lining. Combustion
chambers must be lined with refrac-
Bypass Reheat. Bypass reheat is a viable tory materials capable of withstanding
option for overall SO2 removal require- the high temperatures. Also, mainte-
ments of less than 70 to 80%. In such nance of flame and flame stability are
cases, a fraction of the flue gas, usually required for effective operation. The
15 to 20%, is allowed to bypass the disadvantages of direct firing are the
absorber and is mixed with flue gas that availability and cost of clean-burning
has been treated through the absorber. gas and oil.
The bypassed gas warms the scrubbed
gas by 20 to 40°F (10 to 20°C). Bypass Indirect Reheat. Indirect reheat con-
reheat offers the advantages of low sists of heating ambient air through an
capital and operating investment as well external heat exchanger and then mix-
as ease of operation. The maximum ing the hot air with the saturated flue
amount of reheat that can be obtained gas to raise its temperature. The major
and the applicability of this mode of components are a forced draft fan, heat
reheat, however, are limited by the con- exchanger, ducting, isolation damper,
straints of emission standards. and air-gas mixing chamber.
The advantage of indirect hot air re-
Direct Reheat. This type of reheat uses heat over in-line reheat is that the heat
a heat exchanger directly in the wet flue exchanger is removed from the direct
gas ductwork downstream of the mist gas path. The potential for corrosion
eliminators. The heat exchanger con- and plugging of the heat exchanger
sists of multiple rows of tubes through thereby is significantly reduced. The
which the heating fluid is circulated. disadvantages include the need for
The tubes typically are staggered. an additional fan to convey hot air;
Bare-surface and finned-surface tubing the relatively large amount of space
have been used. The gas passes over required for the reheat system (com-
the tubes and picks up thermal energy pared with other reheat methods); the
from the surfaces of the heating tubes. increase in stack gas volume, which
The initial rows of tubes must provide may be undesirable because of the
sufficient heat transfer to evaporate any limited capacities of induced draft (ID)
residual mist leaving the mist elimina- fans and stacks; and the higher energy
tor. The subsequent rows of tubes serve consumption, which is needed to heat
to boost the gas temperature. air from the ambient temperature level.
A direct combustion reheater elimi- Another variation of the indirect
nates the need for heat exchangers. hot air reheat method involves use of
5-68 | Clean Combustion Technologies
the air pre-heater associated with the In the shell-and-tube GGH, the
boiler to provide hot air for stack gas untreated flue gas (absorber inlet)
reheat. In this case, the temperature of flows vertically downward through the
the combustion air entering the boiler tubes. The gas tubes, typically with a
would be lower than usual, because diameter of 3.1 inches (80 mm) and a
part of the heat content of the flue wall thickness of .04 inch (1 mm), are
gas is used to provide the hot air for constructed of polytetrafluoroethylene
stack gas reheat. Also, the pre-heater (PTFE; Teflon® material). The treated
is larger, because it is designed to heat flue gas (absorber outlet) flows across
a greater-than-normal amount of the outside of the tubes. The leakage
air (but to a lower-than-normal exit of gas between the two streams in the
temperature). shell-and-tube GGH is essentially
zero (typically <0.01%). The leakage
Regenerative Reheat/Ljungström® is attributable to the small diffusion of
GGH. In the GGH system, the sensible gases through the PTFE. The useful
heat from the hot, unscrubbed, inlet life of PTFE in this type of service
flue gas is extracted and transferred to is claimed to be from 10 to 14 years.
the desulfurized, moisture-saturated, The permeability of the PTFE will
outlet flue gas. The principal device likely increase over its life. Water wash
used for this purpose is a close relative lances are used to wash the inside of
of the air pre-heater: the GGH of the the tubes.
rotary, regenerative, Ljungström® type. The lower tubesheet of the heat
An alternative to the Ljungström® type exchanger is made of perfluoroalkoxy
is the shell-and-tube GGH. The shell- Teflon® material (PFA). This material
and-tube heat exchanger is sometimes is twice the cost of PTFE, but unlike
referred to as a “no-leak” GGH. This PTFE, it can be welded. The tubesheet
designation is not strictly correct, how- must allow for expansion of the PTFE
ever, because it is a liquid-gas exchange, tubes. Thermal expansion of PTFE is
not a gas-gas exchange, as implied with approximately tenfold greater than that
the term GGH. of carbon steel. The upper tubesheet
The Ljungström® GGH transfers is an alloy construction. The casing is
heat by rotating heating elements carbon steel, with a 0.6 inch (1.5 mm)
(typically called “baskets”) at approxi- PFA “foil” lining. The tubes must have
mately 1 rpm between the hot and cold an internal pressure that is greater than
gas. Each element is heated by the hot the external pressure; otherwise, the
(untreated) gas and then loses its heat tubes would collapse. For this reason, a
to the cool (treated) gas. The flow of C-position fan is not acceptable.
gas in the Ljungström® GGH is coun-
ter-current. The untreated and treated Regenerative Reheat/Liquid-coupled
gases flow in opposing directions. heat exchange. The basic equipment
Separate inlet and outlet ports are in this type of reheat system consists
provided for the treated and untreated of two separate shell-and-tube heat
gas streams. A small amount of leakage exchangers mounted in the absorber
(typically 0.5–1.5%) occurs between inlet duct (untreated gas) and in the
the two streams. absorber outlet duct (treated gas).
Control of Power Plant Stack Emissions | 5-69
Heat is transferred by water (or other in the scrubber. The solution was to
heat-transfer fluid) that circulates add a crystallization tank and increase
between the two heat exchangers. the L/G ratio. Adding a crystalliza-
The heat-transfer fluid flows on the tion tank operating at a high solids
inside of the tubes, and flue gas flows concentration allows calcium/sulfur
across the outside of the tubes. Teflon® compounds to precipitate in the tank
(PTFE), Teflon®-coated, and enam- instead of fouling the interior of the
eled steel have been proposed for tube absorber. Because of the precipitation
materials. Diffusion of acid gases, such reactions, the absorbent recycled to
as HCl or H2SO4, through PTFE the scrubber has a lower concentra-
tubes, resulting in acidification of tion of scale-forming gypsum. A high
the heat-transfer fluid, is a concern. L/G ratio minimizes the increase in
Chemical dosing is required to prevent supersaturation across the scrubber
this problem. and, thereby, further diminishes the
Liquid-coupled reheat is an expen- potential for scale formation. Finally, Limestone deposits are plentiful
sive alternative and is used only when the recirculation of a slurry provides and widespread throughout the
leakage of SO2 and fly ash cannot be preferential sites for crystallization, world, making calcium-based
tolerated because of requirements for reducing the likelihood of precipitation
low emissions. Both capital and operat- on the absorber surfaces.
systems the most common for
ing costs for liquid-coupled exchangers Many subsequent attempts at coun- sulfur capture technologies.
are high. tering scale and the other problems
resulted in failure. Absorbers were
Absorber Island chosen or designed to operate in an
The widespread availability of lime or erosive, corrosive, scaling environment,
limestone alkali helped make the wet and early adopters tried packed towers,
FGD scrubbing technology the oldest marble beds, turbulent contact absorb-
and most common method of FGD. ers (TCAs), tray towers, and venturi
Because calcium salts have very low spray towers. The U.S. EPA built a
solubility, however, these technologies large test facility at the Tennessee Val-
require absorbers with high liquid/gas ley Authority’s Shawnee station with a
(L/G) ratios, crystallization holding marble bed absorber, TCA, and venturi
tanks, and tolerance to corrosive and spray tower in parallel, 10-MW gas
erosive slurries. paths.55 Support also was available
The first successful, closed-loop, from a small EPA pilot TCA glass ab-
lime/limestone process is credited to sorber at its headquarters in Research
the English firm of J. Howden and Triangle Park, North Carolina.56
Company.54 From 1931 to 1933, The TCA first used “ping pong”-like
pilot-plant tests were conducted balls, which through constant motion
in which the scrubber effluent was prevented scale buildup, making it an
recycled to the absorber. Initially, scale early favorite. This absorber was only
formed in the absorber, but the critical a short-term success, however, because
breakthrough in system chemistry the plastic balls failed quickly, became
came during 1933, when the problem lodged in the support grids, and caused
was identified as one of excessive gyp- high pressure drop and, eventually, a
sum (CaSO4·2H2O) supersaturation shutdown for cleaning. Foam rubber
5-70 | Clean Combustion Technologies
was later substituted, at a higher cost, In fact, the U.S. EPA initially man-
for plastic. With time, spray tow- dated spare modules in the United
ers became the dominant solution. States. The scale-free operations of
In the spray tower, scale was still a forced oxidation scrubbing, however,
problem, but because the buildup was demonstrated that the spare modules
not catastrophic, the absorber could were not required. A Dutch company
operate longer. built a single-module, forced oxida-
Borgwardt56 and others57,58 identi- tion, limestone, gypsum spray tower in
fied the mechanism for scale formation. the Netherlands and set the stage for
The solid calcium sulfite crystals were similar designs globally.
actually a mixture of calcium sulfite Another breakthrough in wet FGD
hemihydrate and calcium sulfate—a came from Rochelle and King,61 who
co-precipitate. The upper limit of suggested using small concentrations of
calcium sulfate in this co-precipitate is organic acids in the circulating fluid to
approximately 18%. Sulfite oxidation enhance performance. Later, field tests
rates above this amount resulted in proved the organic acids did, indeed,
supersaturated slurries and, eventu- improve performance substantially, but
ally, scale. Oxidation rates with these the acids were lost at unanticipated
“natural oxidation” systems were not quantities through not only material
a controllable parameter; thus, scaling balance losses in the process water but
could not be controlled. Operation, also oxidative degradation losses,62
however, was possible with either especially in forced oxidation systems.
complete oxidation or no oxidation, Additionally, co-precipitation of the ac-
and in both cases, scale-free operation ids in calcium sulfite crystals provided
was possible. another mechanism for loss.63
No oxidation operation was Even with these losses, organic acids
achieved by inhibiting oxidation with proved to be cost-effective, especially
additives such as sodium thiosulfate.59 adipic acid and a waste stream mix of
Complete oxidation of the sulfite to organic acids (glutaric acid, succinic
sulfate could be obtained by simply in- acid, and adipic acid),64 commonly
jecting air into the recirculating slurry referred to by the DuPont trade name
tank and converting the calcium sulfite DBA (dibasic acid). Successful pilot
to gypsum.60 This latter process, also testing,65 full-scale testing,66 plus
called “forced oxidation,” had the added strong U.S. EPA approval67 led to
advantage of providing a potentially commercial use in many installations.
salable by-product, wallboard gypsum, In the late 1990s, Alstom supple-
instead of a calcium sulfite “sludge.” mented the open spray tower
Of the two scale-free adaptations of technology with a high-efficiency,
limestone scrubbing, the forced oxida- limestone, forced oxidation, pump-
tion, gypsum process quickly became free, crossflow tray absorber called the
prevalent. The cleaner operation also Flowpac® system. The first installa-
allowed certain cost savings. Scaling in tion of the Flowpac® system was at
the absorbers was so problematic that a power plant in Sweden.68 Second-
most utilities included spare modules generation Flowpac®69,70 absorbers are
that could be isolated and cleaned installed in several locations on oil,
while the boiler continued operations. coal, and industrial smelter gases. The
Control of Power Plant Stack Emissions | 5-71
Flowpac® system is unique in that it uses the energy of the oxidation air in an
airlift to also move and agitate the slurry, eliminating the need for agitators and Demister
pumps (Fig. 5-41).
N = kl(C* – C)
N = kg(y – y*)
Equation 5-31
where
Equilibrium, pH
6
CaCO3 + 2H+ → Ca2+ + CO2 + H2O Gypsum Precipitation. Precipitation 5.5
Equation 5-38 is the process by which a substance 5
separates from solution by a chemical 4.5
and change. Precipitation of gypsum will 10 100 1,000 10,000
[Ca2+] + [Mg2+] Concentration, mM
occur if the gypsum relative saturation
KA2 + (RS) of the solution is greater than
H
ACa 2 + = 1.0. Gypsum RS is defined as the ratio
PCO2 Forced Oxidation Natural Oxidation
of the ion activity product of calcium
Equation 5-39 Figure 5-42 | Recycle tank equilibrium pH
and sulfate over the solubility product versus dissolved calcium and
Clean Combustion Technologies
constant of calcium sulfate dihydrate. Figure: magnesium
05.42/ (JG) concentrations
where Specifically, PMS5415 K
Ksp
(dimensionless)
PCO2 = partial pressure of CO2 Equation 5-41
concentration. The sprayed slurry falls eliminator section sets a practical limit
countercurrent to the gas flow and col- of 12 to 13 ft/s (3.7 to 3.9 m/s) through
lects in an air-sparged, agitated tank at the absorbing section. A short transition
the bottom of the tower. The scrubbed piece with a gradually expanding area
gas travels upward, through a chevron connects the mist eliminator section to
mist eliminator, and out through a the absorber section. The number of
cone-shaped top. absorbing stages required to obtain the
desired level of SO2 removal governs the
Spray Tower Reaction Tank. The reac- overall tower height.
tion tank usually is an integral part of The corrosive and erosive envi-
the absorber. (This tank is also called ronment within the tower requires
the “hold tank,” or the “recycle tank”). special materials. Type 318L and 317L
The reaction tank performs several stainless steels have been successfully
process duties: applied in numerous instances. Under
certain conditions of scrubber-slurry
• Providing adequate slurry and
pH, temperature, and chloride concen-
solids residence times to allow
tration, the use of nickel-based alloys
limestone dissolution and gypsum
with higher molybdenum and chro-
precipitation reactions
mium contents can provide superior
• Mixing fresh limestone slurry with
corrosion protection.74 The use of such
spent absorber slurry
high-grade alloys requires very careful
• Dispersing compressed air for
fabrication. In general, failures of high-
forced oxidation of sulfite to sulfate
grade alloys incorporated in the tower
• Providing residence time for the
shell and internal support members oc-
gypsum crystals to grow
cur primarily because of faulty welding
• Providing adequate NPSH for the
rather than corrosion.
absorber recycle pumps to operate
In contrast to the spray tower shell,
properly
the gas ducts upstream and down-
The reaction tank is sized to satisfy stream of spray towers represent much
the kinetics of limestone dissolution more severely corrosive environments.
and gypsum precipitation reactions. Absorber inlet and outlet ducts as well
The former usually is the slowest, as reheat mixing zones require special
although at higher rates of sulfur consideration with respect to mate-
removal, the latter may dominate. The rial selection. Protective linings, such
agitators provide solids suspension, as inorganic cements or borosilicate
mixing, and air dispersion in the case glass blocks over carbon or stainless-
of forced oxidized designs. The materi- steel plate, typically are used for the
als of construction must be resistant to absorber inlet ducting.
both corrosion and erosion.
Spray Section: Pumps, Piping, and
Spray Tower Shell. Velocity consider- Nozzles. There can be independent
ations dictate the tower cross-sectional absorption stages, each consisting
area. Although higher gas velocities of a spray pump, piping, manifold,
through the absorbing section improve headers, and spray nozzles. If a spare
SO2 removal, the maximum allowable pump is included, a complete spare
velocity through the close-coupled mist absorption stage also is provided. This
5-76 | Clean Combustion Technologies
scale from forming on the vanes. droplet carryover, which would result
(These are process design issues, from deposition on the vanes. The reaction tank performs
which are covered in detail in the several key functions including
chemistry of calcium-based wet Reaction Tank. Usually located at the
FGD systems.)
the dissolution of limestone, the
bottom of the spray tower, the reaction
2. The design should allow no back tank, or recycle tank, serves many crystallization of gypsum, the
side or downstream washing of purposes. It provides residence time oxidation of sulfites, and the
the last mist eliminator vane. This primarily for the dissolution of solid provision of liquid head for the
results in too much washwater to limestone and for the various chemical
recycle pumps.
carry over when there is gas flow. reactions. It also provides residence
3. There is very little penetration time for gypsum crystallization and
of the washwater in to the vanes. crystal growth. The tank is used to
Only the visible surface of the inject air for oxidation of sulfite to
vanes (relative to the washwater) sulfate. Finally, the tank provides liquid
will be cleaned. head for the many recycle pumps.
A secondary absorption reaction,
If the wash system is to be supplied which occurs in the FGD process,
separately, the wash system design involves the conversion of sulfite and
(nozzle pressure, angle, and layout) bisulfite to sulfate:
should be approved by the system
supplier in order to maintain valid mist HSO3– + ½O2 (aqueous) → SO42– + H+
eliminator performance guarantees by Equation 5-42
that supplier.
The cleaned flue gas passes through a SO32– + ½O2 (aqueous) → SO42–
separation section with two stages: the
Equation 5-43
bulk entrainment separator, and the mist
eliminator. The bulk entrainment sepa- This oxidation reaction will, to some
rator consists of 6-inch (15 cm) FRP extent, occur naturally because of
vanes mounted at a 45° angle on 2-inch oxygen in the flue gas, and it can
(5 cm) parallel spacing. It extends be promoted, or forced, through
across the entire face area of the spray the contact of sparged air with the
tower absorber. The mist eliminator absorber slurry. Several FGD system
located above the bulk entrainment operating factors may influence the rate
separator is made from V-shaped FRP of natural oxidation, including the oxy-
vanes arranged in a series of chevrons gen/sulfite SR, ionic strength of the
across the gas flow path. Two rows of absorbent slurry, absorbent slurry pH,
chevrons ensure droplet impingement gas/liquid temperature, gas pressure,
and minimize mist carryover. Wash and L/G ratio in the absorber.
nozzles are installed above and below As with the absorption of SO2, oxida-
the first stage mist eliminator and tion of sulfite involves the mass transfer
below the upper stage mist eliminator. of a gas to the liquid phase. The rate of
The stages are washed a portion at a absorption can be expressed by two-film
time such that the wash rate doesn’t ex- mass transfer theory as follows:
ceed the total makeup water required.
The first stage is washed more often.
N = Kg A(y – y*)
Efficient cleaning prevents excessive
Equation 5-44
5-78 | Clean Combustion Technologies
where
N = rate of oxygen mass transfer
Kg = overall mass transfer coefficient (resistance to mass transfer)
A = interfacial area for mass transfer
y – y* = driving force for mass transfer (concentration gradient)
In general, the rate of sulfite oxidation is nearly always faster than the rate of oxygen
transfer from the gas phase. Thus, oxidation in an FGD system is usually mass trans-
fer limited, and the rate of chemical reaction can be assumed to be instantaneous.
Oxygen Transfer to the Slurry from Air Bubbles. Borgwardt60 used oxygen
75
transfer data from Jackson et al. to estimate FGD oxidation transfer rates.
Jackson measured oxygen transfer using single bubbles rising through a tall tank
with dissolved sulfite. As such, the mass transfer of oxygen was controlling. The
dynamics of an FGD oxidizer will be more complex, however, because the slurry
is agitated and has a flow field toward the tank exit. The bubble dynamics also are
different (agglomeration and velocity vectors other than straight up)—all of which
affect mass transfer. In the case of an agitated lance system, the bubbles can take a
cyclonic route to the surface. Borgwardt’s experiments, however, indicated excellent
agreement with the data of Jackson et al. and, thus, became the basis for FGD
oxidizer design. A brief summary of their findings follows:
Urza and Jackson76 empirically determined the oxygen transfer efficiency of a
bubble column to be
h
70 , 000 × Kl a × h × P + s
e 1.0 + 0.264 e 2
+ ln =
3.0 + 0.791e 1−e mo H
Equation 5-45
where
e = oxygen transfer efficiency (fraction)
Kla = overall liquid side mass transfer coefficient
H = Henry’s law constant
P = barometric pressure (atm)
h = static sparger depth without aeration
hs = sparger depth with aeration
mo = air (oxygen) rate
Borgwardt then developed an empirical equation for Kla based on temperature
and the data from Jackson et al.:
−0.666
293 m1o.09
Kl a = 2.08 0.214
T P + hs / 2.0
Equation 5-46
Next, Borgwardt combined the two equations for 122°F (50°C) as follows:
e 1 + 0.264 e
+ ln = 0.0474 mo0.09h
3.0 + 0.791e 1−e
Equation 5-47
Gas Velocity. The design velocity usually is limited by the maximum velocity
acceptable to the in-line mist eliminator and/or the gas side pressure drop, and
it typically is in the range of 10 to 13 ft/s (3 to 4 m/s) for modern absorbers. In
general, increasing gas velocity increases pressure drop, but it also increases mass
transfer and, therefore, SO2 removal efficiency. Higher gas velocity in the spray
tower translates to smaller cross-sectional areas and, thereby, lower initial costs of
the absorber tower.
L/G Ratio. Increasing the L/G ratio at a given gas velocity increases perfor-
mance. As a power plant reduces load, at a constant liquid flow rate, the L/G
ratio increases and the gas velocity decreases. These opposing effects are almost
equivalent; thus, performance is maintained as load decreases.
Limestone Delivery and Storage. For feeder that discharges the rock into the
on-site grinding, limestone commonly feed end of the grinding apparatus.
is received as ¾ × 0 inch (20 × 0 mm)
crushed rock. The handling proper- Limestone Grinding. The limestone
ties of this material are similar to particle size required for proper SO2
those of coal. Although rain shelters absorption/gypsum production is 95%
are common, limestone rock may be by weight less than 44 microns (325
stockpiled outside in most climates. U.S. Standard Mesh). Grinding of the
Conventional earth-moving equip- limestone rock can be performed at
ment, such as front-end loaders and the site of the FGD system or at the
bulldozers, are used to push or move quarry where the mineral is mined. In
the rock into subgrade drop door bins. North America, on-site grinding us-
Foreign material, such as roots, wood, ing wet mills is the most common. In
and iron, is separated from the rock at Europe and Asia, dry grinding at the
this stage. From the drop door bins, quarry is common.
belt conveyors move the limestone to Three types of limestone grinding
day bins located above the mills. The apparatus have been commercially used
limestone is then gravity fed, using a in FGD applications: horizontal ball
control slide gate, onto a gravimetric mills, vertical tower mills, and roller
mills. The first two types are used in
Cyclone-Manifold wet grinding systems. Raw limestone
Distribution System is mixed with water or process liquor
Limestone during the grinding operation, produc-
Feed Bin
ing a slurry product of finely ground
limestone particles. The third type, the
roller mill, pulverizes the limestone
without the addition of water, produc-
Feed Feeder
Controller Motor ing a dry powder.
Gear Box
Water Supply Wet Ball Mills. Horizontal wet
Air Clutch
ball mills have been—and continue to
Ball Mill
be—the most common apparatus used
for grinding limestone. As depicted in
Cooling High Pressure Figure 5-43, grinding is accomplished by
Water Lift Pump tumbling the limestone in a cylindrical
vessel that is partially filled with forged
Oil
Low Pressure Pump steel balls and water. The ball charge in-
Reservoir cludes balls ranging from ¾ to 4 inches
Water Supply (20 to 100 mm) in diameter. Fresh or
Density
reclaimed process water is added in
Meter proportion to the rock to maintain con-
stant slurry density. The resultant slurry
Classifier
Mill Tank Feed Pump Limestone Storage output from the mill overflows into a
Slurry Feed Pump mill product tank and is then pumped to
a bank of hydrocyclones and classified.
Figure 5-43 | Wet ball mill
Oversize particles are captured in the by the mill pump. As the finely ground
hydrocyclone underflow and returned material overflows the main body into
to the mill. Fine particles exit in the the coarse classifier, the largest particles
hydrocyclone overflow and are sent are removed and recycled to the bot-
to the limestone slurry storage tank. tom of the grinding chamber via a mill
From the limestone slurry storage tank, sump and the mill pump. The coarse
pumps deliver the required flow (based classifier overflow reports to a collec-
on the amount of absorbed SO2) to tion sump. From the sump, it is further
each operating absorber. classified into final product by the
The grinding system operates external hydrocyclone classifiers.
as a closed circuit. That is, rock is
recirculated until it has been ground Limestone Powder. For cases in
small enough to be captured in the which the limestone is ground at the
hydrocyclone overflow. In practice, quarry, it is delivered to the FGD site
however, it is not uncommon for the in bulk tank trucks or rail wagons as
discharge end of the mill to produce a ready-to-use, fine powder and then
a small stream of reject rocks. These pneumatically conveyed to storage
usually are rocks containing higher- silos. Such silos may be sized for a
than-average amounts of silica and 15- to 30-day supply. Volumetric screw
alumina compounds. These rejected
hard rocks typically are less than 1% of
the rock feed to the mill.
conveyers transfer the powder into the size and the size distribution. The
limestone day bins. The powder falls greater the top size, the greater the
by gravity onto a covered conveyor required energy input.
weigh belt, which then dumps the • Size of the ground product. The
powder into a wetting-and-mixing smaller the product size, the greater
tank. Mixers to pre-wet the powder the required energy input.
also are used at times. In the mixing • Mill capacity. The capacity in ton/h
tank, water is metered in to make the (kg/h) of the limestone grinding
needed slurry. Typical slurries are pre- system is governed by the following
pared at 25 to 35% solids by weight. factors:
• Steady-state limestone consump-
Mill Type Selection. Ball mills are tion rate. This is determined
preferred when the limestone feedstock by knowing the maximum SO2
size is medium to large (0.8–1.0 inch removal rate, the design stoichi-
[20–25 mm] top size), the rock has ometry, and the minimum CaCO3
a high work index, and the required content (purity) of the limestone.
product grind is 80 to 95% less than 44 • Hours per day the mill will
microns. Tower mills become competi- operate. In most plants, grind-
tive when ultrafine grinds are required, ing is performed during one or
but the feed size to a tower mill gener- two shifts. Thus, if the mill is to
ally is limited to ⅜ inch. A moderate operate for 8 hours in any 24-
work index favors a tower mill. Dry hour period, the capacity of the
roller mill pulverizers are preferred grinding system must be equal
when a source of heat and large rock to three-fold the steady-state
sizes are available. No quantitative data limestone consumption rate.
are available regarding the compara-
tive reliability of the various mill types. Dewatering. Dewatering of the waste
Anecdotal evidence, however, suggests solids typically is performed in a
that wet ball mills are the most reliable primary and a secondary stage. Usually,
and that dry roller mill pulverizers are in the case of gypsum, the primary
the least reliable. dewatering concentrates the slurry
from 15% solids to 50% solids, and the
Mill Sizing. Mill sizing is performed secondary dewatering further concen-
by the mill vendor. For approximations, trates the gypsum to a moist cake of
detailed crushing and grind calcula- 10% moisture. Primary dewatering can
tions, including horsepower, can be be accomplished either with a thickener
found in Bond.79,80 The major factors (highest flexibility but rarely used) or,
influencing the mill size are: more commonly, with a hydrocyclone.
Secondary dewatering is needed
• Bond work index. This is a mea- to get the product moisture that is
sure of the limestone hardness and desired for almost all of the final by-
its resistance to grinding. Typical products discussed in this section. The
work index values range from 6 to choices are:
13. Soft stones have values of 6 to 8.
Hard stones have values of 11 to 13. • Rotary drum vacuum filter (up to
• Size of the feed rock. This 88% or more for forced oxidation
includes both the top (largest) solids)
Control of Power Plant Stack Emissions | 5-85
solids, than the feed. The degree of tank via gravity feed. For a two-
solids separation depends on the feed stage hydrocyclone arrangement, the
composition and characteristics as overflow of the first stage is routed
well as on the configuration, size, and to a storage tank before entering the
operational parameters of the hydrocy- second stage.
clone equipment.
Quite a few equipment parameters Secondary Dewatering. The second-
can be selected or changed on a hydro- ary dewatering system makes a final
cyclone setup. Hydrocyclones can be reduction in the moisture content of
modified for different capacities and the waste slurry. This system receives
separation performances: type of inlet the underflow stream from the primary
(involute vs. tangential), size of inlet, dewatering system and separates that
size of vortex finder, length of cylinder stream into a filtrate stream and filter
section, and cone angle. In general, long cake stream. The solids, concentrated
hydrocyclones help separation, whereas as filter cake (65– 92% solids by
short hydrocyclones inhibit maximum weight), are transferred to the solids
recovery of solids. The inlet and the treatment subsystem. The liquor
vortex finder fittings affect the capacity recovered as filtrate is returned for the
and the cut point. The effects of these plant’s use.
changes are used to fine tune a particu- Secondary dewatering is ac-
lar cyclone for an application and do complished by vacuum filtration or
not have to be fixed for initial sizing. centrifugation, using one of the follow-
Hydrocyclones are designed only ing devices:
for on/off operation. Variable flow
• Horizontal vacuum belt filter
operation is not recommended. The
• Rotary drum vacuum filter
hydrocyclone underflow stream is
• Centrifuge
routed either directly to the secondary
dewatering device or to a filter feed The choice of device will depend
on the degree to which the solids are
oxidized and on the concentration of
solids required in the filter cake.
discharged into the slurry vat. The vat the rotary drum vacuum filter the
agitator maintains suspension of the preferred secondary dewatering device
slurry solids. The pick-up zone is the for all systems except forced oxidation,
portion of the drum that is submerged commercial-grade gypsum production.
in the slurry vat. The rotary drum has not provided
The drum’s surface (the cylinder the degree of dewatering necessary for
walls) is the filter medium. It is a producing commercial-grade gypsum.
porous, canvas-like material through A rotary drum vacuum filter is sized
which the vacuum pump takes suction based on the quantity and particle size
via the internal piping. In the pick-up of the solids to be processed. Because
zone, the suction draws the slurry into the particle size depends on the degree
contact with the filter medium. The of solids oxidation, filter sizing is dif-
filter medium is selected based on the ferent for each of the three types of
particle size of the slurry solids that the oxidation systems. Design guidelines
filter allows the maximum amount of for rotary drums used in natural,
liquor to pass through while holding inhibited, and forced oxidation systems
virtually all of the solids on the surface. are presented in the following para-
The suction is continued in the drying graphs. A more complete presentation
zone as the filter surface rotates out of is given for a natural oxidation system.
the vat. Entrained moisture is drawn Exceptions to the natural oxidation
through the filter cake and filter media. design are given for inhibited and
The blow discharge zone is the forced oxidation systems rather than
last zone before the rotation cycle repeating the procedure in its entirety
repeats. Here, the suction is replaced for each system.
with a positive pressure. Air is forced Some systems require a surge tank,
through the piping to the filter media or a filter feed tank, upstream of the
by the discharge blower. This causes secondary dewatering devices. This is
the filter cake to bellow out from true of systems with batch operation of
the filter media. A horizontal knife, the secondary dewatering devices and
known as a “doctor blade,” catches for systems in which it is anticipated
the dislodged filter cake and directs that the operation will either exceed
it onto the discharge conveyor. The or be significantly less than the design
filtrate removed from the filter cake conditions. For surge protection, a
is a mixture of slurry liquor, air, and retention time or storage capacity of
suspended solids. The liquor and sus- two hours generally is acceptable. For
pended solids are separated from the batch operation, the retention time
air by the vacuum receiver. The vac- is equal to the number of hours that
uum pump discharges the air stream the dewatering devices will be inactive
to an outdoor location. The filtrate between periods of operation. Because
pump takes suction at the bottom of of the suspended solids in the slurry, a
the vacuum receiver and delivers the filter feed tank requires an agitator. In
filtrate slurry to a holding tank or the general, the agitator blade centerline is
primary dewatering device. located at a distance off the tank floor
The lower initial capital cost and equal to one-third of the tank diam-
a simpler method of operation make eter. The agitator blade must remain
5-88 | Clean Combustion Technologies
submerged and operating at all times with the following exceptions. For
during plant operation. Continuous a filter feed slurry with a suspended
operation of an agitator requires a solids concentration of 30 to 40%, a
minimum level of submergence of the filtration rate of no greater than 250
agitator’s blade, which typically is two lb/h/ft2 (1,220 kg/h/m2) should be
feet above the blade’s centerline. This used to produce a filter cake with a
minimum level or volume of liquid concentration of 75 to 80% solids.
(not active volume) should not be in- Design guidelines for a filter feed
cluded when calculating the retention tank with a forced oxidation system
time of the tank. are identical to those given for a natu-
Forced oxidation controls the degree ral oxidation system.
of solids oxidation by introducing
compressed air into the reaction tank Horizontal Vacuum Belt Filter. A
slurry. The additional oxygen can horizontal vacuum belt filter uses a
produce a solids oxidation of 99% belt as the filtering medium. The filter
CaSO4 . 2H2O or greater. The particle feedbox assembly spreads the slurry
size of forced oxidized solids is much uniformly across the entire width of
larger than that in naturally oxidized the moving filter cloth. The slurry is
or inhibited oxidized solids. Therefore, dewatered by both gravity and vacuum
the slurry of a forced oxidation system as it travels on the filter media above
is more easily dewatered. the drainage/transport belt. The
A standard rotary drum for a filtrate is collected in the vacuum pan
forced oxidation system can be de- assembly after being drawn from the
signed using the same guidelines given slurry through the filter cloth and
for a natural oxidation system, but drainage belt. The washbox sprays the
moving cake with washwater to
rinse dissolved solids from the cake,
and this liquid also is drained from
Slurry Feed
the cake. A series of rollers separates
the filter media from the drainage
Cake Wash
belt and breaks up the filter cake.
The cake is discharged as the filter
media travels around a small roller.
The abrupt change in direction
breaks the cake loose so that it falls
away from the filter media. The fil-
ter media and drainage belt are then
washed, tensioned, and guided in
preparation for the next cycle. The
filtrate removed from the filter cake
is a mixture of slurry liquor, air, and
Product Discharge
suspended solids. The liquor and
suspended solids are separated from
the air by the vacuum receiver. The
Figure 5-46 | Horizontal vacuum belt filter
vacuum pump blows the air outdoors. The filtrate pump takes suction at the bot-
tom of the vacuum receiver and delivers the filtrate slurry to a holding tank or the
primary dewatering device (Fig. 5-46).
Horizontal vacuum belt filters can produce filter cakes with higher solids con-
tent than rotary drum filters—namely, 85 to 92% solids compared to 55 to 85%
for drum filters. Typically, belt filters are favored over vertical bowl centrifuges,
because they can operate continuously and require less capital cost. Because of
these qualities, belt filters are the preferred secondary dewatering device for forced
oxidation systems, especially if commercial-grade gypsum is required (typically,
90% solids minimum).
Dry Scrubbing
The idea of using a spray dryer as a gas cleaning device dates back to at least 1958,
when a Czech patent was issued covering use of a spray dryer and a dust collector
in series. Dry scrubbing, or dry FGD, was pioneered in the United States as an en-
hancement to dry injection of nahcolite or trona ahead of a fabric filter. The intent
was to develop a simple process with a dry product using one single-unit opera-
tion. The initial work was focusing on using soda ash solutions as the absorbent,
but it was soon discovered that lime slurry produced very good results as well.
Today, lime is exclusively used for power applications.
The most important chemical reaction in lime based dry FGD is the formation
of calcium sulfite hemi hydrate [CaSO3 . ½H2O(s)]. Its solubility is very low, thus
being a chemical driver for the dry FGD process. As seen from Table 5-14 below,
calcium sulfite is much less soluble than gypsum.
In the dry FGD processes, lime [CaO] is first converted to hydrated lime Table 5-14 | Solubility
[Ca(OH)2] before the reagent is contacted with the flue gas. The hydrated lime products of calcium sulfite
may be injected into the flue gas either as slurry or as dry or moistened powder. and calcium sulfate
A prerequisite for the lime to react with SO2 is that the humidity of the flue gas
Solubility
is high enough so that the particle surface picks up about one monolayer of water product
molecules. This has been shown to happen at a water concentration of the flue gas Compound Formula
(Ksp. 25°C
corresponding to a relative humidity of 40 to 50%. [77°F])
The reaction is then believed to proceed predominantly according to Calcium
CaSO3 · ½H2O 6.8 × 10−8
sulfite
Calcium CaSO4 · 2H2O
SO2 + Ca(OH)2(s) → CaSO3 . ½H2O(s) + ½H2O 3.14 × 10−5
sulfate (gypsum)
Equation 5-49
5-90 | Clean Combustion Technologies
Other acid species react with lime in a dry FGD system. Hydrogen chloride (HCl)
may form either calcium hydroxy chloride (CaOHCl) or CaCl2 . 2H2O according to
Recycling of the dry product from than that of an SDA, is termed the
the dust collector can be employed. FDA. This is enabled by adding water
This increases the utilization of the to the recycle in a dust mixer imme-
lime, because more active alkali is fed diately before the injection into the
to the SDA. Moreover, the drying rate reactor vessel. Table 5-15 summarizes
is increased, because the water required the dry FGD process designs.
to cool the flue gas is spread out over a
larger surface area than in the case of a Spray Dryer Absorber.When using an
single-pass system. The slurry feed to SDA the lime is hydrated according to
the atomizing device, however, must
still be pumpable, which limits the CaO + H2O → Ca(OH)2 + heat
amount of recycle that can be handled. Equation 5-55
This process version is thus defined as
a low-recycle system. The hydration reaction is highly
A third version is defined as the exothermic and takes place in the
high-recycle system. In this case, the liquid phase. This process is commonly
feed of lime and recycle are dry. Water referred to as “slaking the lime.” The
is added separately, either in a mixer temperature in the slaking reactor is
immediately before the injection into controlled by the feed ratio between
the reactor vessel or directly into the CaO and water.
reactor by pressure nozzles. Because The reactor usually is one of three
the recycle rate is very high, the surface types: a paste mixer or pug mill, a
area available for the cooling water is detention unit, or a ball mill. The
much increased, yielding a very short mill may be arranged in a vertical or a
gas residence time in the reactor. This horizontal position. The ball mill pro-
means that the reactor vessel can be vides for slaking while simultaneously
built significantly smaller than the SDA. reducing the size of the impurities
The dry solids content of slurry for that always are present in the pebble
a single-pass system typically varies be- lime. Both pug mill and detention
tween 2 and 15%. For a recycle system, slakers use a screen or settling unit
the dry solids content is kept between to remove grit or rock-like impurities
35 and 55%. The exact value is deter- from the milk of lime. Pug mill slak-
mined by the fly ash and SO2 contents ers, which allow the highest ratios of
of the flue gas, its temperature, and the pebble lime to water, can most reliably
required removal rate of SO2. achieve the high slaking temperatures
Two types of reactor vessel designs so important for the spray dryer
exist. The addition of water can take scrubbing process.
place inside or outside the reactor
vessel. If injected inside the vessel,
Table 5-15 | Dry flue gas desulfurization Process Designs
pressure nozzles spraying into a bed
of recycle solids introduce water. This Process Reagent Recycle Reactor retention
Reactor/Dryer
Version Added As Ratio time (s)
often is termed a CFB scrubber. The
Single pass Lime slurry — Spray dryer absorber 10–15
reactor residence time is reduced to
about one-third of that for an SDA. Low recycle Lime slurry 2–10 Spray dryer absorber 10–15
The final version, in which the dry- Dry hydrated Flash dryer absorber
High recycle 50–100 0.5–5
lime Circulating fluidized bed
ing time is an order of magnitude less
5-92 | Clean Combustion Technologies
thereby making OH–, or hydroxyl ions, available in the liquid phase for neutrali-
zation of hydrogen ions (H+). Hydrogen ions are produced as the sulfur dioxide
Control of Power Plant Stack Emissions | 5-95
molecules diffuse across the droplet surface, dissolve in the liquid, and dissociate
according to the reaction
Equation 5-57
The dissolution of sulfur dioxide tends to lower the pH (or raise the hydro-
gen ion concentration) in the liquid phase, whereas the dissolution of lime tends
to raise the pH. Slaked lime has an equilibrium pH of approximately 12.0;
sulfur dioxide at 1,000 ppm by volume in the gas phase in equilibrium with
pure water will lower the pH to approximately 2.5. Equations 5-56 and 5-57
characterize the acid-base neutralization that occurs in each droplet as it passes
through the SDA.
Although it is clear how the OH– ions in Equation 5-56 are removed through
neutralization, it is not immediately apparent what happens to the Ca2+. The
calcium ions also must be purged from the solution in order to allow more lime
to dissolve. The most important reaction is the precipitation of calcium and sulfite
ions to form calcium sulfite crystals:
Ca2+ + SO2–
3 + ½H2O → CaSO3 . ½H2O(s)
Equation 5-58
Mass and Heat Transfer. The gas-phase heat transfer to the surface of the drop-
let and the mass transfer of water molecules away from the surface govern the rate
of water evaporation. The rate of heat transfer is modeled, in the classical sense, by
where
T = the gas-phase bulk temperature
Ts = the adiabatic saturation temperature of the gas/liquid system, generally
assumed to be the temperature at the surface of the droplet
h = the heat-transfer coefficient
Q = quantity of heat transferred
The rate of water mass transfer is modeled by
dM/dt = Dwatera(Hs – H)
Equation 5-60
5-96 | Clean Combustion Technologies
SO2 Removal, %
70
cates that freshly slaked lime exhibits evaporated. The maximum humidity
65 SR = 1.0
superior reactivity compared with that the gas could attain is HS and the
60
powdered hydrate mixed in water. minimum temperature TS (i.e., the
Ideally, the lime should be a light- saturation values). In practice, however, 55
between the SDA inlet and outlet tem- aid in the precipitation of fresh
90
peratures. It is directly related to the calcium sulfite (Equation 5-58).
85 amount of water supplied to the atom-
izing system. It also is a key indicator All of these mechanisms probably play
80
SO2 Removal, %
of how much evaporative heat is avail- some part in the observed performance
75 able in the flue gas. Normally, the spray improvement seen with the use of
SR = 1.4
70
down temperature is not a controlled recycle. The performance effect from
SR = 1.2 variable unless an air heater flue gas by- using a ratio of up to 2 pounds (kg) of
65 pass is available for this purpose. Most recycle solids for each 1 pound of fresh
SR = 1.0
60 often, spray down temperature is set by Ca(OH)2 solids in the atomizer feed is
0 100 200 300 the requirements for maximizing boiler shown in Figure 5-54. For a 2.5% sulfur
SDA In–Out, °F
efficiency. A lower SDA inlet tempera- coal, the SR decreases from 1.2 to 1.0%
ture typically is the result of a higher at 74% SO2 removal in the SDA.
Figure 5-53 | Effect of spray down
Clean Combustion Technologies boiler efficiency. This also results in a In the high-recycle systems (CFB
temperature (SDA, spray dryer
Figure: 05.53/ (JG)
PMS5415 absorber;
K SR, stoichiometric lower spray down temperature. scrubber, FDA), the stochiometric value
ratio) Spray down temperature has a that the reactor sees is significantly in-
significant and direct effect on the creased. A corresponding linear increase
90
efficiency of SO2 removal. As shown in performance is not realized. For the
in Figure 5-53, there are conditions in SDA, a limitation is set by the maxi-
85
SR = 1.2 which an increase of 50°F (28°C) in mum solids content of the recycle slurry
SO2 Removal, %
80 spray down temperature can mean a that is still pumpable. For high-recycle
SR = 1.1
75 decrease in SR from 1.4 to 1.2 at 75% systems, however, this limit does not
SR = 1.0
SO2 removal in the SDA if all other exist, and high recycle rates also can be
70 SR = 0.9
significant variables are held constant. realized for high-sulfur fuels.
65
Typical western U.S. coals have a chlo- waste applications. Often, both SO2 90
ride content of 100 to 200 ppm, whereas and HCl are present together, with
SO2 Removal, %
some eastern U.S. coals are much richer removal requirements established for 80 SDA
in chlorides (700–800 ppm). both species. In other instances, chlo- and FF
70
The effect of the chlorides on the rides are introduced through the use of SDA
performance of a dry FGD system high-chloride make-up water and are 60
is a function of chloride and sulfur concentrated in the feed slurry when
50
contents as well as the amount of recycle of solids is employed. 0 1,000 2,000
fly ash that enters the system. Thus, Chlorides in the feed slurry impact Inlet SO2 Concentration, ppm (vol.),
Wet
knowledge about the chloride con- operation in two important ways:
tent of the fuel is quite important for
1. By changing the apparent Figure 5-56
predicting performance and operating Clean| Combustion
Effect of inlet SO2
Technologies
saturation temperature of the gas Figure: 05.56/ (JG) (FF, fabric filter;
concentrations
variables for dry FGD. Very high levels PMS5415 SDA,K spray dryer absorber)
leaving the system
of chlorides, as seen at various waste
2. By increasing the relative solubil-
incineration plants, require major ad-
ity of calcium ions in the slurry
justments to the approach to saturation
liquid or moistened feed
temperature. For municipal solid waste,
a common HCl concentration in the The dry FGD systems operating
flue gas is 300 to 800 ppm by volume, with chlorides present respond to a
5-100 | Clean Combustion Technologies
reaction. The net effect of this is that coals is of interest. Some data from
1
more molecular bromine will be pres- the U.S. Geological Survey database
ent compared to molecular chlorine: COALQUAL are displayed in Figures 0.8
5-58 and 5-59. The database lists
Distribution Function
Cl approximately 3,500 data points for 0.6
<< 1
Cl total bromine in coals from all over the
United States. At least 50% of the 0.4
Br
≈1 coals have bromine contents of less
Brtotal 0.2
than 10 ppm. There also is some
correlation between the bromine and 0
For practically all coals, the chloride chlorine contents of the coal: With 1 10 100
Bromine, ppm
content in the coal is far higher than the coal chlorine content of greater than
bromine content (Cl/Br = 50–500), 500 ppm, there is a high likelihood
Figure 5-58 | Distribution of bromine content
but in the flue gas, the amount of that the coal also contains bromine at Clean Combustion Technologies
data in the U.S. Geological
molecular bromine may well be larger greater than 10 ppm. Figure: 05.58/
SurveySPS
COALQUAL database
than the amount of molecular chlorine Full-scale experiments performed PMS5415 K
Hg + Br2 → HgBr2
Hg + Cl2 → HgCl2 100
chloride and bromide were added to will provide additional residence time
40% of the burners, and the change but also new reaction paths for the
in oxidation rate was determined. halogenation reactions. The effect of
Adding up to the equivalent of 900 an SCR was investigated85 by measur-
ppm of chlorine to the entire coal ing the mercury species upstream of
stream produced a marginal increase of the SCR and downstream of the air
oxidation from 35 to 40%. Addition of heater at 10 different plants burning
the equivalent of 100 ppm of bromine coals with chlorine levels ranging from
to the coal produced an increase in 130 to 1,400 ppm (Fig. 5-60). The re-
oxidation rate from 35 to 72%. This sults clearly show that the halogenation
demonstrates in full scale the impor- of the mercury is augmented by the
tant role of bromine in the coal to SCR. Other results show that when
control the oxidation of mercury in the catalyst ages, the effect becomes
the flue gas in a boiler. The process to weaker. For low-halogen coals, the
add bromine compounds to the coal effect can be much less. A catalyst can
for the purpose of controlling mer- only provide increased kinetics, not the
cury chemistry was invented by Prof. halogen reactants.
Bernard Vosteen and is the subject of
U.S. Patent 6,878,358. This process is Mercury Collection Using
licensed to and marketed by Alstom Activated Carbon
under the trade name KNX™. Some of the mercury can be caught
The halogenation reaction is kineti- with the fly ash in a conventional flue
cally limited. Other factors, which also gas cleaning system. An early inves-
contribute to an increased ratio of Hg2+ tigation of this86 showed that this
to Hgtot at the air pre-heater outlet, are native collection in a conventional
the residence time in the system and the fly ash collection system is enhanced
presence of UBC, which acts as a cata- by a high UBC in the ash, a low ash
lyst for the reaction. A high-dust SCR filtration temperature, and use of a
fabric filter rather than an ESP. Best
results were obtained with high UBC,
corresponding to a loss-on-ignition
100
of 10 to 20% and a filter temperature
80 below 210°F (99°C). Such values of
operational parameters often conflict
60 with other requirements for a modern
Hg(0), %
PC boiler system.
40
The UBC fraction in the ash was
20 surmised to have some similarity in
structure to activated carbon, because
0
1 2 3 4 5 6 7 8 9 10 both are of a carbonaceous nature and
Plant Number have a disproportionately high contri-
bution to the fly ash specific surface
Upstream SCR Downstream AH
area. Experiments with commercially
Figure 5-60 | Result of mercury testing at 10 different plants burning coals with chlorine contents
of 130 to 1,400 ppm (AH, air heater; Hg(0), elemental mercury; SCR, selective
catalytic reduction)Clean Combustion Technologies
Figure: 05.60/ (JG)
PMS5415 K
Control of Power Plant Stack Emissions | 5-107
bag filter cake provides excellent mass emitted Hg0 can be somewhat higher
transfer between PAC and the flue than the amount going into the FGD
gas. For bituminous, high-halogen, system. This phenomenon is called “re-
eastern U.S. coals, it was observed early emission of mercury,” and it limits the
on that most of the mercury exists as total performance of the wet FGD as
Hg2+ and that this system often gives a a collection system for Hg2+. Various
variable, but significant, native mercury ways to modify the chemistry of the
collection efficiency. By adding con- wet FGD system to stabilize the dis-
trolled amounts of PAC, the mercury solved mercury without impairing the
collection efficiency can be stabilized. main function of the wet FGD system
For western U.S., sub-bituminous, to collect SO2 and to produce gypsum
low-halogen, coals (Cl, <100 ppm) are being investigated.
with a very high fraction of Hg0, the
native collection efficiency of the dry Alstom Mercury Control
A multi-faceted approach to FGD system is close to zero. Adding Technologies
mercury capture involving conventional PAC to this system does
KNX™ Technology
little to improve mercury collection
higher temperature oxidation
efficiency. This problem can be solved The KNX™ Technology comprises
and longer residence times by using the Alstom KNX™ additive adding a bromine additive compound
with a modified activated along with injection of ordinary PAC, to the coal in trace quantities to control
carbon provides for a which produces good mercury collec- mercury oxidation for all types of coal,
retrofit technology that uses tion results.87 Another solution is to irrespective of its composition. The
use injection of the more expensive, process is under exclusive license from
considerably fewer additives. bromine-impregnated PAC. Vosteen Consulting of Germany.
In the wet FGD system pre- The first application was in con-
dominantly used today, with nection with low-halogen, western
limestone-based forced oxidation U.S. coal and dry FGD, in which the
producing gypsum, the metallic additive changes the mercury sorp-
mercury entering the system is not tion process so that a high collection
collected at all because of the low efficiency can be reached with the
aqueous solubility of the metallic addition of regular PAC. The addi-
vapor. The Hg2+ fraction is collected, tive provides mercury bromination in
because it has a high aqueous solubility the boiler backpass in order that inert
and because the mercury halides also Hg0 is converted to the more reactive
form complexes of the type (HgCl4)2− Hg2+ species. The additive also has
and (HgBr4)2−, giving a low backpres- shown good mercury control effects in
sure of Hg2+ from the solution. The other systems firing low-halogen coal,
dissolved mercury complexes, however, in which a typical amount of UBC can
can be reduced by other dissolved provide adequate absorption when the
ions in the solution. The amount of interaction between mercury and car-
mercury in the solution is very small bonaceous material is augmented by the
in comparison to several other species. effect of the additive. This process also
The reduction occurs first to Hg22+, integrates well with the wet FGD sys-
which easily decomposes to Hg2+ and tem, in which the upstream conversion
Hg0. In such cases, the amount of of mercury to Hg2+ occurs regard-
Control of Power Plant Stack Emissions | 5-109
less of coal quality, and this mercury ing ESP. Because the amount of PAC
species can then be absorbed in the is approximately 0.2% of the typical
aqueous phase of the scrubber liquid. flue gas fly ash content of PC boiler
It is a simple system, which is unique flue gas, it is within the normal toler-
in that it does not employ PAC. ances of ash variability. The impact on
ESP performance normally is negligi-
Mer-Cure™ Technology ble, and the existing ESP often can be
The Mer-Cure™ technology is an used without modifications. To achieve
advanced, PAC-based mercury control a uniform distribution of the injected
technology in which the mass transfer material throughout the flue gas in the
limitations of simple PAC injection short ducts that usually are available
to an ESP have been overcome by the upstream of the air heater, a detailed
application of several new and unique fluid mechanics design is necessary.
features:
Filsorption Technology
• The PAC base sorbent is impreg-
The Filsorption technology uses a
nated with an additive that
combination of PAC addition and a
accelerates the sorption process
bag filter (other than dry FGD). The
kinetics.
PAC is added upstream of the bag
• The sorbent is mechanically pro-
filter by admixing to the flue gas. In
cessed before injection to provide a
suspension, some interaction between
sorbent with maximal surface area
sorbent particles and flue gas occurs.
and uniform dispersion in order to
The PAC material is then deposited on
enhance mass transfer from the gas
the bags. Here, the diffusion distance
to the sorbent.
between PAC particles and the bulk
• The sorbent is injected upstream
gas is very short, leading to much
of the air pre-heater in order to
improved, gas side mass transfer. The
prolong the flue gas contact time
PAC particles reside in the gas stream
and take advantage of the cata-
for the duration of the cleaning interval
lytic effect of the sorbent on the
of the bags, which is on the order of
oxidation reactions at the higher
10 minutes. This gives ample time for
temperature.
reaction between the gas and the PAC
By this multifaceted approach using particles. The mass transfer limitations
the increased oxidation that results for the sorption of mercury are thus
from the injection upstream of the eliminated, and good mercury collec-
air heater, the improved mass trans- tion efficiency can be obtained using
fer from the processing of the PAC standard PAC. The fabric filter can be
material just before injection, and the installed to perform fly ash collection
improved kinetics from sorbent chemi- as well. As in the ESP case, the injected
cal impregnation, the performance of PAC is then diluted by the ash. The
this process is generally better than dilution is so high that conventional fly
that of conventional PAC injection ash discharge and handling technology
in an ESP. The PAC consumption is can be used.
modest, and the system can be installed If the fly ash is used in cement pro-
without any modifications to the exist- duction, a dedicated fly ash collection
5-110 | Clean Combustion Technologies
device can be located ahead of the fabric 1% per year (BTU/unit of GDP).
The most cost effective means filter. The product from a fabric filter Examples include more efficient ap-
of reducing CO2 emissions is downstream of a dedicated ash collector pliances, fluorescent light bulbs, and
will be high in PAC, which can be prone improved auto mileage.
efficiency improvement.
to auto-ignition if allowed to remain For power plant equipment manufac-
in large volumes for extended periods turers, efficiency improvements lead to
of time in non-ventilated areas. When the development of higher-temperature
the PAC resides on the bags, the heat and -pressure steam conditions for
produced by the slow oxidation of the steam turbines and boilers as well as
PAC is readily dissipated by the passing lower power requirements for auxiliary
flue gas. Only when the PAC is kept in equipment. The reduction of required
a hopper for a long time can auto-igni- auxiliary power leads to higher net
tion occur. The remedy is to design the output from power plants, improving
filter with steep ash hopper angles and overall efficiency. The development of
to have continuous ash discharge. “ultra-supercritical” steam technology
is manifested in the world’s first power
CONTROL OF CO2 EMISSIONS plant with steam conditions at 1,300°F
Concerns about CO2 buildup in the (700°C) at a plant in Wilhelmshaven,
atmosphere have led to the potential Germany. Such plants can lead to an
for CO2 emission regulations cover- improvement in efficiency from subcriti-
ing steam generating and power plant cal steam conditions of over 33%. (See
equipment. Currently, there are a wide Chapters 3 and 4 on supercritical boilers
array of considerations, ranging from and supercritical steam turbines.)
carbon taxes to cap-and-trade propos- Should longer-term stabilization of
als to no control regimes. It is not the CO2 in the atmosphere be required,
purpose of this text to debate the mer- larger reductions in CO2 emissions
its of such concerns; that is a topic for will be required. Aside from abstinence
political, social, and scientific debate. regarding the use of fossil fuel altogether
Even the threat of controls, however, (requiring development of an economi-
is of considerable consequence to a cal substitute), emissions avoidance
potential owner of plant equipment. implies the capture and sequestration
Because Alstom supplies equipment (or storage) of CO2 from emitting
throughout the world, development sources. Capturing and effectively stor-
of CO2 control technology is a critical ing carbon dioxide (CO2) produced by
research and development undertaking. fossil-fueled power plants could, if done
The most cost-effective means of on a large enough scale and in enough
reducing CO2 emissions is efficiency countries, play a significant role in miti-
improvement. As the cost of emitting gating growth in global greenhouse gas
CO2 increases (by regulations, taxes, emissions. It also could fundamentally
or allowance costs), the value of effi- alter the carbon footprint of conven-
ciency improvements increases. These tional fossil fuels—particularly cheap,
improvements can occur throughout carbon-rich coal—and allow them to
the range of economic activity. Indeed, remain significant sources of energy for
the historical improvement in efficien- decades to come, without the potential
cy of the U.S. economy has averaged CO2 impact on the global climate.
Control of Power Plant Stack Emissions | 5-111
The potential role that carbon cap- every living person in the United States
ture and storage technology could have would have to consume 20 tons/yr
on the future of the global economy (tonnes) of sodium bicarbonate, or
and energy mix is substantial. Oil, approximately 100 pounds (45 kg) of
natural gas, and coal currently account tablets per day. Of course, CO2 also
for 70% of the world’s primary energy is produced through the use of other
supply and, collectively, produce 80% fuels as well—not to mention people
of man-made global CO2 emissions. and animals. It is the scale of the
Although there may be some uses for problem that makes the solution truly
the captured CO2, the scale of the challenging.
problem is enormous. The MIT Coal One of the difficulties involved with
Report88 points out that the under- the capture of CO2 has to do with the
ground storage for CO2 would require fact that most combustion processes
a pipeline system three-fold the size of utilize air as the source of oxygen.
the existing natural gas pipeline system Along with the air comes substan-
in the United States. tial amounts of nitrogen. Thus, the
A number of technologies are be- resulting product gases then contain
ing developed to address the issue of substantial amounts of nitrogen,
capture and sequestration of CO2. diluting the CO2 and making it more
In addition to the technology, both difficult (and expensive) to separate
regulatory and liability issues need to (i.e., capture) from the rest of the gas.
be resolved. Ownership of the CO2 One means of avoiding this dilution
after the storage site is filled also is a problem is to use relatively pure oxygen
key issue. A number of demonstra- as the reagent. There are two approach-
tion projects are either planned or es to utilizing the oxygen. The first is
underway to explore these issues and direct combustion with oxygen (often
demonstrate the feasibility of safe cap- referred to as oxy firing). The second
ture and storage of CO2. It is not that is to use the oxygen to first gasify the
CO2 has never been captured before. fuel. This gasification process pro-
The commercial process for producing duces a reducing gas (mostly CO and
sodium bicarbonate (a typical antacid) hydrogen). The CO in this gas can be
is to capture CO2 from flue gas after converted to CO2 by use of the water
natural gas combustion using naturally gas shift reaction. Here, water vapor is
occurring sodium carbonate. The scale reacted with the CO to produce CO2
of the amount of capture now required, and hydrogen:
however, makes utilization for all but a
fraction of the CO2 nearly impossible. CO + H2O = CO2 + H2
The United States consumes more
than 1 billion tons (tonnes) of coal The CO2 can be separated from
annually, which in turn produces more this gas using processes developed in
than 3 billion tons/yr (tonnes/yr) of the refinery and petrochemical indus-
CO2. If all of this CO2 were converted tries. The hydrogen subsequently is
to sodium bicarbonate, the produc- combusted with air for the purpose
tion level would be more than 6 billion of generating heat and/or electricity.
tons (tonnes). That would mean that Because this fuel gas can be burned in
5-112 | Clean Combustion Technologies
in Fig. 5-62). Again, if a forced oxida- mixture must be increased from 21%
tion system is used in the FGD unit, by volume to between 25 and 30% by
then in order to avoid introducing volume in wet RFG. For dry RFG, the
nitrogen into the flue gas, the oxidant oxygen content would have to be 37
must be oxygen (possibly mixed with to 40% by volume. The oxygen flow is
CO2 to aid dispersion). fixed by stoichiometry, but the RFG
The flue gas that is not recycled is flow decreases and the CO2 content
compressed before storage. In some increases.
instances, this storage might be at As with air, the RFG is split into
the site of the power plant, but more primary and secondary flow streams
commonly, it will be located elsewhere. into which the oxygen is introduced.
The compressed gas may be trans- Sufficient oxygen must be present in
ported in a pipeline to a storage site or the primary RFG stream to achieve
utilized for enhanced oil recovery. The ignition and keep the flame close to
raw flue gas leaving the FGD contains the burner tip and so help to reduce
moisture, nitrogen, oxygen, and trace NOx formation. The exact amount
amounts of other species, such as SO2 of oxygen required depends on the
and NOx. These must be removed or burner design. An important design
reduced either before or during the safety consideration limits the oxygen
compression process in order to satisfy concentration of any one stream to
certain pipeline criteria. For example, less than 40% by volume. Systems
the oxygen must be reduced to below with oxygen purities of greater than
100 ppmv, and the moisture to below 50% by volume flowing above a
110 ppmv, to limit pipeline corrosion critical velocity can cause certain pipe
potential. If the gas is to be injected materials to ignite. Carbon steel is
directly, however, less cleanup may one such material. The presence of
be required, and essentially all of the oils and hydrocarbon-based greases
gas can be sequestered. This situation exacerbate this tendency. Below an
represents the only true “zero emission” oxygen concentration of 40% by vol-
power plant. ume, the pipe work can be designed as
Flames produced from coal burning for air, but it must be cleansed of oils
in air are brighter than similar flames and grease. Another approach is to
produced with an O2/RFG mixture, limit the oxygen concentration to 23%
even when the amount of oxygen by volume in the RFG flows to the
present is the same. This results from burner and add the additional oxygen
the higher specific heat of CO2,, which at the burner tip through a separate
lowers the adiabatic flame tempera- feed line. Such an approach may be
ture. The flame also propagates less required at low-load conditions to
rapidly, resulting in increased release maintain flame stability and prevent
of volatiles and formation of soot. The ignition failure.
additional soot that is formed can be The greater density of CO2 means
seen as bright specks in the O2/RFG that the volumetric flow of the RFG
flame. To obtain an adiabatic flame passing through the burner is less than
temperature similar to that of air, the that of the air in an air-fired unit. Be-
percentage of oxygen in the combustion cause the coal carrying capacity of a gas
Control of Power Plant Stack Emissions | 5-115
the primary RFG and the primary air cases of air, wet RFG, and dry RFG:
1.2
should be the same. This requirement Wet RFG Dry RFG
• The air case has the highest furnace
results in the volumetric flow of the 1.0
exit gas temperature and the lowest
secondary RFG stream being only 70%
total emissivity. The wet RFG case
that of the secondary air flow in an 0.8
has the lowest furnace exit gas
air-fired unit. In an unmodified burner
temperature and the highest total 0.6
designed for use with air, this lower
emissivity. In combination, these
gas flow will result in reduced swirl
two factors result in the furnace 0.4
and, possibly, an unstable, elongated 0.15 0.2 0.25 0.3 0.35 0.4 0.45
heat absorption being similar for all Oxygen Fraction at Burner Inlet
flame. To avoid this condition, addi-
three cases.
tional swirl must be introduced to the
• The heat transferred to the convec- Figure 5-63 | Effect of oxygen fraction and
secondary flow. For tangentially fired Clean Combustion
recycled flueTechnologies
gas moisture on
tion pass is similar for each case.
units, a higher velocity may be required Figure:furnace
05.63/ heat
SPS transfer
Because of the higher CO2 content, PMS5415 K
to generate the required swirl in the
the specific heat and gas density are
center of the furnace.
higher for the two RFG cases, and
The RFG used to moderate the
the gas velocity and gas mass flow
combustion temperature is high in CO2
and moisture. The resulting higher
specific heat of the gas will lower the Table 5-16 | Furnace Heat Transfer Characteristics
mean furnace gas temperature for Recycled Flue Gas
oxygen combustion with 21% oxygen by Air Wet Dry
volume. This lower bulk temperature in Recycle ratio (recycled flue
turn lowers the radiation component, re- — 2.55 1.36
gas/O2 mass ratio)
quiring an increase in heat-transfer area Fraction of oxygen
to achieve the same duty. Fortuitously, At burner inlet 0.21 0.25 0.38
the CO2 and water vapor also are strong In flue gas 0.033 0.033 0.033
radiators, having higher gas emissivities. Adiabatic flame temperature
3,500 (1,930) 3,220 (1,770) 3,790 (2,090)
This factor will help to compensate for (°F [°C])
the reduced gas temperature driving Gas temperatures (°F [°C])
Entering convection pass 2,120 (1,160) 1,980 (1,080) 2,000 (1,090)
force. Testing has established that for Exiting convection pass 800 (430) 720 (380) 530 (270)
oxygen-fired combustion to achieve sim- Convection pass temperature 1,320 (730) 1,260 (700) 1,470 (820)
ilar furnace heat-transfer performance drop
to that with air, the oxygen content Gas emissivity 0.36 0.58 0.47
must be increased. Figure 5-63 presents Total emissivity 0.50 0.70 0.65
data for a retrofit application using wet Heat absorbed
and dry RFG. For the wet RFG, the 8.36 (94.9) 8.42 (95.6) 8.43 (95.7)
(BTU/ft2-s [kWh/m2])
furnace heat transfer is similar to that Overall heat-transfer coefficient
of air when the oxygen content is 25% for convection pass 8.2 (46.4) 8.6 (48.6) 6.9 (39.4)
by volume. For the dry RFG, an oxygen BTU/ft2-h [W/m2])
(������
content of 38% by volume is needed to Gas velocity (ft/s [m/s])* 32 (9.8) 25 (7.5) 16 (5.0)
provide the same heat-transfer perfor- In-furnace gas residence time (s) 2.1 2.5 3.8
mance. Additional data are presented in
*Entering convection pass.
Table 5-16.
5-116 | Clean Combustion Technologies
In all the designs, the vent stream is slightly longer (by perhaps 5 hours)
requires treatment to remove the SO2 than that for an air-fired boiler. This
and NOx. An SCR can be used, but longer time probably arises from the
this will be much smaller than that transition to oxygen-fired operation
used on an air-fired PC unit. The SO2 that is made after 20 hours with the
can be removed by passing the exit gas plant at 30%. This requires bringing
to an existing FGD unit or by build- the air separation unit into service,
ing a compact unit especially for the starting the RFG flow, and adjusting
vent stream. The remaining CO2 and the burners for the new combustion
nitrogen can be discharged into the conditions. The CO2 plant enters
flue gas stream at a convenient location service after 35 hours with the output
or passed to a dedicated stack. still at 60%. Up to that time, the flue
Fireside corrosion may be increased gas has been passing to the stack. Full
in the immediate vicinity of the oxy- load is reached in a little over 40 hours,
gen-fired burner because of the higher with the exact time depending on the
oxygen concentration. This threat can operation of the CO2 plant.
be offset by using coatings. If the heat
fluxes are kept the same as those for Post-Combustion Capture
air-fired units, standard materials of An alternative to oxygen firing is
construction most likely can be used. post-combustion capture of CO2
If both the oxygen content and the using a solvent to chemically absorb
heat fluxes are increased, some material the CO2 from the flue gas. The higher
upgrades probably will be required. the generating efficiency, the less CO2
Startup considerations of the boiler released per megawatt-hour. This
and CO2 purification plant are sum- lower release helps lower the cost
marized in Figure 5-65. From cold-start of capture and storage. Thus, super-
conditions, the 60% minimum load critical units are being considered as
condition is reached in 30 hours, which a starting point for post-combustion
capture plants. One solvent currently
in use for small-scale applications
CO2 Recovery Commences is monoethanolamine (MEA). Two
1,000 MW early plants using Alstom CFB tech-
nology were installed with partial flue
Oxy-Fuel Operation on Oil
gas scrubbing of CO2 for food-grade
600 MW
Full Load applications. Research is underway to
identify solvents that are less costly
300 MW for power industry applications.
Turbine Roll Coal-Firing Commences For most scrubbing processes, the
200 MW
Turbine Master in Automatic design consists of two stages. In the
first, the flue gas is passed through an
Synchronization absorption column, where the solvent
Light Off on Oil Using Air removes the majority of the CO2 by
chemical absorption. In the second, the
solvent is sent to a stripping column,
0 +4 +6 +8 +20 +30 +35 +40
Time, hours where heat is applied to release the
CO2 and regenerate the solvent. The
Figure 5-65 | Startup of an oxygen-fueled supercritical PC plant
Clean Combustion Technologies
Figure: 05.65/ SPS
PMS5415 K
Control of Power Plant Stack Emissions | 5-121
flow sheet is similar to a standard ge- raises the solvent temperature to ap-
neric gas treatment process (Fig. 5-66). proximately 250°F (120°C). This steam
The flue gas and aqueous solvent represents a parasitic loss to the power
solution are contacted countercurrently plant operation. To enhance desorption
in an absorption column, with the flue of the CO2 and reduce the amount of
gas entering at the bottom of the col- heat required, the stripper operates close
umn and the CO2-lean solvent at the to atmospheric pressure. The hot, lean
top. To minimize solvent degradation, solvent is withdrawn from the reboiler
the inlet flue gas temperature should inlet and then passed to the rich-lean
not exceed 150°F (65°C). solvent heat exchanger, where it is cooled
For an MEA system, the lean solvent ahead of re-entering the absorber.
typically enters the absorber at 110°F Any MEA solvent carried over from
(45°C), and because the CO2 absorption the absorber by the CO2-depleted
reaction is exothermic, the CO2-rich flue gas is recovered by a water wash
solvent leaves the bottom of the absorber system and returned to the center of
at 140°F (60°C). The rich solvent passes the absorber. The water added also
to a rich-lean solvent heat exchanger, helps maintain the solvent dilution at
where it is heated by the hot, lean solvent the required level. Solvent carried over
leaving the reboiler. The hot, rich solvent from the stripper by CO2 is recovered
then enters at the top of the stripper in a condenser and returned to the
column, with additional heat being top of the stripper. A small portion
provided by a steam-heated reboiler that of the lean solvent is extracted from
To CO2 Compressor
Vent Gas to
Stack
Condenser Liquid
Separator
Water Wash
Solvent Make-Up
Absorber Stripper
Reboiler
Flue Gas Rich/Lean Steam
Exchanger
Condensate
Rich Solvent
To Reclaim Unit
Lean Solvent
the reboiler and then fed to a reclaim • Low heat demand (i.e., low heats of
unit, where any degradation products reaction [absorption and desorp-
are precipitated out following reaction tion] at a low temperature) lowers
with caustic soda. This small amount steam usage and the size of the
of material can be kept in suspension reboiler. The lower the temperature
and injected into the boiler for disposal of the heat supplied for a given
by incineration. duty, the lower the grade of heat
The MEA also reacts with the other that is required. The reduced steam
acid gases in the flue gas, such as SO2 demand improves the overall effi-
and NO2, forming heat-stable salts that ciency, which reduces the specific
must be removed from the process and cost of CO2 capture.
replaced with make-up amine. To limit • High reaction rates (absorption
the increase in operating cost, it typically and desorption) lower the size of
is recommended that the flue gas leaving the contactor required for a given
the boiler be pre-treated to reduce both capture rate.
the SO2 and the NO2 to low levels. An • High CO2 loading lowers the
economic trade-off study regarding the amount of solvent that has to be cir-
cost of acid gas scrubbing versus that of culated. This lowers pumping duty
make-up MEA will determine the level and, as the absorber and regenerator
of cleanup needed. Using low-NOx operate at different temperatures,
burners and SCR, NOx emissions can lowers the sensible heat duty.
be reduced to below 20 ppmv. Because • Low volatility lowers the amount
NO2 usually represents only 5% of of solvent carried over from the
the total NOx, the NO2 content is absorber and so lowers the heat
approximately 1 ppmv. The NO present consumed by evaporation.
in the NOx does not react with amines. • High stability lowers solvent decom-
A benefit of these reactions is that the position and reduces make-up costs.
flue gas discharged to the atmosphere • Low corrosivity allows use of low-
contains virtually no SO2. cost construction materials.
Solvent improvements include a • Low toxicity reduces process haz-
new formulation that has an increased ards and environmental impact.
CO2 absorption rate. This decreases • Tolerance to impurities in the flue
the required packing volume in the gas reduces solvent degradation.
absorber, allowing a smaller, less
expensive absorber to be used. The Current research is focusing on
improved solvent also has a higher achieving these properties by modifying
CO2 carrying capacity, resulting in the structure of amines. MEA is a pri-
less solvent circulation and a corre- mary amine, and recent studies suggest:
sponding reduction in reboiler steam
demand. The reclaim stage has been • Moving to secondary and tertiary
improved as well, allowing degrada- hindered amines
tion products to be removed with • Increasing the size of the amino
reduced solvent loss. group
The desirable characteristics of a • Increasing the number and posi-
CO2 solvent are as follows: tion of hydroxyl groups
Control of Power Plant Stack Emissions | 5-123
In sterically hindered amines, the nitrogen atoms of the ammonia are substituted
by a bulky molecule. By slowing down the rate of carbamate formation, the CO2
capacity of the amine is increased. A primary hindered amine reacts as follows:
Because only one molecule of the hindered amine is used, compared to two
molecules of MEA, the carrying capacity theoretically can be doubled.
Absorber inter-cooling can be used to limit the maximum temperature reached
in the absorber. For a two-stage absorber, semi-rich solvent is extracted from the
bottom of the upper section, cooled, and then returned to the top of the lower
section. This process increases the CO2 absorption rate, allowing a smaller and less
expensive absorber to be used. The CO2 loading in the rich solvent is increased,
resulting in less solvent circulation and a corresponding reduction in reboiler
steam demand.
A split flow configuration can be used to reduce steam demand. In this opera-
tion, CO2 regeneration is achieved by combining the steam stripping with a flash
stage. One portion of the rich solvent passes to the stripper, and as the flow is
reduced, steam demand also is reduced. The other portion of the rich solvent is
heated indirectly by the hot, lean solvent leaving the reboiler and is regenerated
in a flash drum. This regeneration is achieved without any additional steam, the
energy being provided by the heat contained within the hot, lean solvent. The
semi-lean solvent leaving the flash drum contains more CO2 than the lean solvent.
When the two streams are combined, the solvent passing back to the absorber has
a higher CO2 loading. To achieve the same recovery, the solvent circulation rate
must be increased. This increase is only marginal, however, being offset by the two
previous improvements.
In order to decrease the partial pressure of CO2 in the gas stream leaving the
flash drum and, thereby, increase regeneration, live steam can be added. To avoid
an increase in steam consumption, this steam is extracted by flashing the reboiler
condensate down to the flash drum pressure. Increasing flash regeneration reduces
the CO2 loading of the solvent returning to the absorber, which decreases the
solvent circulation rate and, in turn, the reboiler steam requirement.
Membrane gas absorption contactors are being investigated to replace the con-
ventional packed-column designs. The membrane forms a gas-permeable barrier
between the solvent and the gas. The gas diffuses through the pores to be absorbed
in the liquid. The use of fine membrane material results in a high specific surface
and very compact equipment. Operation is independent of gravity, and the process
is not influenced by the L/G ratio. Thus, these flows can be adjusted independently.
Because the gas and liquid flows are kept separate, solvent entrainment and foam-
ing are eliminated. The design has no contact packing, so flooding and channeling
are eliminated as well. The membrane minimizes contamination of the solvent with
oxygen, SO2, NOx, and dust and, therefore, reduces solvent degradation reactions.
This eliminates the need for inhibitors and allows the solvent formulation to be
5-124 | Clean Combustion Technologies
optimized for CO2 capture. By building the membrane gas absorption from cross-
The chilled ammonia process flow modules, a range of CO2 removal rates can be satisfied. Units handling as much
attempts to capture CO2 as 11,000 tons/day (10,000 tonnes/day) of CO2 are thought to be feasible. These
units will be much more compact and lighter than conventional absorbers.
with less energy consumption
As an alternative to amine-based scrubbing, Alstom is developing the chilled
by using a more favorable ammonia process. The chilled ammonia process uses ammonium carbonate as
solvent (ammonia) and by the solvent, capturing CO2 as ammonium bicarbonate, which is then regenerated
regenerating the solvent at a to release the CO2 and reform the carbonate. The process equipment includes an
absorber and stripping column similar to the amine process shown in Figure 5-66,
higher pressure to reduce CO2
but the operating temperature range is different. The chilled ammonia plant is
compression costs. shown in Figure 5-67.
The reversible chemical reaction is as follows:
The heat of reaction is lower than that of MEA, but ammonium carbon-
ate can dissociate in solution, allowing ammonia to evaporate readily at typical
absorber temperatures. This would increase the energy consumed by evapora-
tion and produce a dilute ammonia solution in the water wash used to remove
the ammonia from the flue gas. The energy required to concentrate this stream
is high, thus impacting the economics. To overcome this limitation, the rate of
CT
Air
Stack
50 °C
DCC1
Make-Up Water CO2
FGD
Figure 5-67 | Chilled ammonia system (ABC, ammoniumFlue Gasbicarbonate; AC, ammonium carbonate;
Water CT, cooling tower; DCC,
Rich Slurry Leandirect
Slurrycontact cooler; CO
FGD,
2
flue gas desulfurization; G, gas; HP, high pressure; HX, heat exchanger)
The flue gas then enters an ammonia scrubber that removes unconverted SO2,
acid gases, and HgCl2. The NO2 and SO3 combine with the ammonia to form
aqueous ammonium sulfate and nitrate. Mercury is removed from the liquor
before it is crystallized. The cleaned flue gas passes through a wet ESP to remove
aerosols and any residual fine particulate matter and is directed to the bottom of
the ammonia scrubber. The CO2 in the flue gas reacts with ammonia according to
the following reaction:
Because ammonia is highly volatile, the flue gas leaving the absorber must
pass through a water wash, creating a stream of dilute aqueous ammonia that is
sent back to the barrier scrubber. This interdependence between the two control
systems imposes a limitation on these systems, because they can only be used
in conjunction with the electro-catalytic barrier systems. Without the barrier
process, the dilute ammonia stream would have to be concentrated for reuse,
and this requires significant energy, negating the thermal advantage gained.
Chemical Looping
Alstom is developing a longer-term concept that represents a new high-temperature
process whereby calcium-based compounds are “looped” in a regenerative
manner to extract oxygen from air for purposes of combustion (Fig. 5-69) or
gasification and then to extract CO2 from reformed syngas for purposes of
Control of Power Plant Stack Emissions | 5-127
Oxidizer CaSO4
1 ata
2,000 °F
Hot Solids
Cold Solids
Product Gas
CaS
CaSO4 Calciner
0.5 ata
Hydrogen 1,600 °F
CaCO3
Reducer
WGS/deCO2 To
CO2
6 ata Sequester
1,600° – 1,800 °F CaO
Coal CaCO3
Steam
CaSO4 + 4C CaS + 4CO
CaO + CO2 CaCO3
chemical looping can be applied to Table 5-20 | Chemical Looping product options
advanced steam cycles, advanced gas
CO2
turbine combined cycles, and in the Option Suboption Products Applications
Capture
future, fuel cell power cycles.
1 — CO2/Steam Yes Coal-fired advanced steam cycles
Table 5-20 also shows that product
Feedstock for transportation/liquid fuels
gas from chemical looping can be used in 2 A SynGas No Feedstock for chemical applications
petrochemical/oil refinery applications Industrial applications
to provide, for example, hydrogen for Coal-fired advanced steam cycles
increasing refinery yields. The product Coal-fired advanced gas turbine
gas also can be specifically tailored combined cycles
2 B* H2/CO2 Yes
Coal-fired advanced cycles
to provide a feedstock for producing Industrial/oil refinery/petrochemical
Fischer-Tropsch or other liquid fuels applications
from coal (or any other carbonaceous Coal-fired advanced steam cycles
fuel [e.g., biomass, waste, or opportu- Coal-fired advanced gas turbine combined
nity fuels]). cycles
3 — H2/CO2 Yes
Coal-fired advanced cycles
Table 5-21 shows potential perfor- Industrial/oil refinery/petrochemical
mance and range of usage for some applications
of the applications in Table 5-20. * Additional Water Gas Shift System and CO2 Scrubber is required.
Referring to Table 5-21, Indices 1
and 4 represent the performance for
conventional steam plants and IGCC,
* CC, combined cycle; CFB, circulating fluidized bed; IGCC, integrated gasification combined cycle; PC, pulverized coal.90
† This column equals the plant efficiency with 95% CO2 capture divided by the plant efficiency without CO2 capture.
‡ Also applicable to other steam cycles.
5-132 | Clean Combustion Technologies
1. “Criteria for the Application of Dust Collectors to Coal-Fired Boilers: 17. Code of Federal Regulations, Title CFR 40, Parts 53 to 80, “Protection
The Results of an ICCI/ABMA Joint Technical Committee Survey,” of the Environment,” revised as of August 26, 2005. Washington, D.C.:
Proceedings of the American Power Conference, 29: 572–580, 1967. Office of the Federal Register, National Archives and Records Adminis-
tration, U.S. Government Printing Office, 2005.
2. R. Srivastava et al., “Emission of SJ3,” JAWMA, 54: 750, 2004.
18. R. E. Bickelhaupt, “A Technique for Predicting Fly Ash Resistivity,” U.S.
3. Institute of Clean Air Companies, Inc., “Terminology for Electrostatic
Environmental Protection Agency Report EPA-600/7-79-204, Wash-
Precipitators,” Publication EP-1. Washington, D.C.: Institute of Clean
ington, D.C., 1979.
Air Companies, Inc., 2000.
19. G. W. Driggers, A. A. Arstikaitis, and L.A. Hawkins, “Computer Model
4. Institute of Clean Air Companies, Inc., “Factors to Consider when Select-
Use for Precipitators Sizing,” Proceedings of Fourth Symposium on
ing Electrostatic Precipitators,” Publication EP-5. Washington, D.C.:
Transfer and Utilization of Particulate Control Technology, Houston,
Institute of Clean Air Companies, Inc., 1982.
Texas, October 11–14, 1982.
5. Institute of Clean Air Companies, Inc., “Gas Flow Model Studies,” Pub-
20. D. O. Heinrich, moderator, “Panel Discussion on Wide Spacing
lication EP-7. Washington, D.C.: Institute of Clean Air Companies, Inc.,
Precipitators,” Proceedings of the Second International Conference on
2004.
Electrostatic Precipitation, November 1984, Kyoto, Japan.
6. Institute of Clean Air Companies, Inc., “Structural Design Criteria for
21. D. O. Heinrich, “Are Non-Deutschian Phenomena Required to Explain
Electrostatic Precipitator Casings,” Publication EP-8. Washington, D.C.:
Wide Spacing?” Proceedings of the Second International Conference on
Institute of Clean Air Companies, Inc., 1993.
Electrostatic Precipitation, November 1984, Kyoto, Japan.
7. Institute of Clean Air Companies, Inc., “Power Consumption Measure-
22. K. Porle and S. Maartmann, “On the Choice of Electrode Geometry
ment for Electrostatic Precipitators,” Publication EP-9. Washington,
and Voltage Waveforms for ESPs,” Fourth International Conference on
D.C.: Institute of Clean Air Companies, Inc., 1988.
Electrostatic Precipitation, September 1990, Beijing, China.
8. Institute of Clean Air Companies, Inc., “Fundamentals of Fabric Collec-
23. L. Lillieblad, M. Thimanson, K. Porle, and H. Jacobsson, “On Dust Cake
tors and Glossary of Terms,” Publication F-2. Washington, D.C.: Institute
Removal in Electrostatic Precipitators,” Eighth International Conference
of Clean Air Companies, Inc., 1972.
on Electrostatic Precipitation, May 2001, Birmingham, Alabama.
9. Institute of Clean Air Companies, Inc., “Operation and Maintenance of
24. M. Kirsten, C. Mauritzson, M. Thimanson, and A. Karlsson, “Advance
Fabric Collectors,” Publication F-3. Washington, D.C.: Institute of Clean
Switched Integrated Rectifiers for ESP Energization,” Eighth
Air Companies, Inc., 2003.
International Conference on Electrostatic Precipitation, May 2001,
10. Institute of Clean Air Companies, Inc., “Flue Gas Desulfurization Birmingham, Alabama.
Terminology,” Publication FGD-1. Washington, D.C.: Institute of Clean
25. H. Jacobsson, M. Thimanson, K. Porle, and M. Kirsten, “Back-Corona
Air Companies, Inc., 1982.
Control with Help of Advanced Microprocessor Enhances Performanc-
11. Institute of Clean Air Companies, Inc., “Gaseous Emissions Equipment: es,” Sixth International Conference on Electrostatic Precipitation, June
Product Definitions and Illustrations,” Publication G-1. Washington, 1996, Budapest, Hungary.
D.C.: Institute of Clean Air Companies, Inc., 1972.
26. J. G. Singer, “Design for Better ESP/Fabric-Filter Hopper Operation and
12. Institute of Clean Air Companies, Inc., “Cyclonic Mechanical Dust Col- Maintenance,” Presented at Air Pollution Control Association ‘83, 76th
lector Criteria,” Publication M-2. Washington, D.C.: Institute of Clean Annual Meeting and Exhibition, Atlanta, Georgia, June 19–24,1983.
Air Companies, Inc., 1969.
27. S. Matts and P.-O. Ohnfeldt, “Efficient Gas Cleaning with the SF Electro-
13. Institute of Clean Air Companies, Inc., “Standardized Method of static Precipitator,” Fläkt Review, 6(7): 105–122, 1963.
Particle Size Determination and Collection Efficiency,” Publication M-5.
28. C. W. Penney, “Weaknesses in Conventional Theory of Electrostatic
Washington, D.C.: Institute of Clean Air Companies, Inc., 1975.
Precipitation,” ASME Paper 67-WA/APC-1. New York: American
14. Institute of Clean Air Companies, Inc., “Wet Scrubber Terminology,” Society of Mechanical Engineers, 1967.
Publication WS-1. Washington, D.C.: Institute of Clean Air Compa-
29. H. J. White, “Effect of Fly Ash Characteristics on Collector Perfor-
nies, Inc., 1975.
mance?” ASME Paper 54-A-259. New York: American Society of
15. Institute of Clean Air Companies, Inc., “Basic Types of Wet Scrubbers,” Mechanical Engineers, 1954.
Publication WS-3. Washington, D.C.: Institute of Clean Air Compa-
30. H. J. White, Industrial Electrostatic Precipitation. Reading, MA: Addison-
nies, Inc., 1980.
Wesley, 1963.
16. Institute of Clean Air Companies, Inc., “Wet Scrubber System: Major
31. S. Oglesby, Jr., and G. B. Nichols, “A Manual of Electrostatic Pre-
Auxiliaries,” Publication WS-4. Washington, D.C.: Institute of Clean
cipitator Technology. Part 1—Fundamentals,” Report PB-196
�����������
380
Air Companies, Inc., 1975.
(AVT’D-O610). Springfield,
������������������������������������������������
VA: National Technical Information
Service, 1970.
5-134
32. S. Oglesby, Jr., and G. B. Nichols. ”A Manual of Electrostatic Precipitator 46. H. J. White, “Resistivity Problems in Electrostatic Precipitation,” Journal
Technology. Part II—Application Areas,” Report PB-196 381 (APTD- of the Air Pollution Control Association, 24(4): 313–338, 1974.
0611). Springfield, VA: National Technical Information Service, 1970.
47. G. C. Schneider et al., “Selecting and Specifying Electrostatic Precipita-
33. J. P. Cooch and G. H. Marchant, Jr., “Electrostatic Precipitator Rapping tors,” Chemical Engineering, 82(11): 94–108, 1975.
Re-entrainment and Computer Model Studies,” Report FP-792, Vol. 3.
48. Y. Goland, “Mechanical and Structural Design Consideration for Inter-
Palo Alto, CA: Electric Power Research Institute, 1978.
nals in Electrostatic Precipitators,” ASME Paper 76-WA/Pwr-7. New
34. G. W. Driggers, L.A. Hawkins, and R. W. Gray, “Applying Performance York: American Society of Mechanical Engineers, 1976.
Results of C-E Rigid Frame Precipitators over a Wide Range of Coal
49. J. Makansi, “Particulate Control: Optimizing Precipitators and Fabric
Characteristics,” Presented at the ASME-IEEE Joint Power Generation
Filters for Today’s Power Plants;” Power Special Report, December
Conference, Indianapolis, Indiana, September 25–29, 1983.
1986. New York: McGraw-Hill, 1986.
35. Kermeth J. McLean, ”Survey of Australian Experience in Collecting
50. Joint Technical Committee of the American Boiler Manufacturers
High-Resistivity Fly Ash with Electrostatic Precipitators, Final Report,”
Association and Industrial Gas Cleaning Institute, Inc., “Design and
Report PB-221 139 (EPAR2-73-258). Springfield, VA: National Tech-
Operation of Reliable Central Station Fly Ash Hopper Evacuation Sys-
nical Information Service, 1972.
tems,” Proceedings of the American Power Conference, 42: 74–85, 1980.
36. Kenneth J. McLean, “Some Effects of High-Resistivity Fly Ash on Elec-
51. J. A. Hudson et al., “Design Considerations and Initial Startup of Shaw-
trostatic Precipitator Operation,” Presented at Electrostatic Precipitator
nee Baghouses,” Proceedings of the American Power Conference, 42, 1990.
Symposium, Birmingham, Alabama, 1971.
52. S. M. Cho, “Properly Apply SCR for NOx Removal,” Chemical Engineer-
37. Z. Herceg and K. J. McLean, “Efficiency of Electrostatic Precipitators
ing Progress, 90: 35–45, 1994.
and Relationship to Corona Voltage-Current Characteristics,” Presented
at the 64th Annual Meeting of the Air Pollution Control Association, 53. R. D. Hawthorne, “Afterburner Catalysis—Effects of Heat and Mass
Atlantic City, New Jersey, June 27–July 1, 1971. Transfer Between Gas and Catalyst Surface,” AIChE Symposium Series,
70: 428–438, 1974.
38. L. A. Hawkins and H. L. Wheeler, “Characterization of Discharge Elec-
trode Performance: Results of Laboratory and Pilot Plant Experiments,” 54. J. L. Pearson, G. Nonhebel, and P. H. N. Ulander, “The Removal of
Presented at the Sixth Joint EPA/EPRI Symposium on Transfer and Smoke and Acid Constituents from Flue Gases by a Non-Effluent
Utilization of Particulate Control Technology, New Orleans, Louisiana, Water Process,” Journal of the Institute of Fuel, 8: 119–156, 1935.
February 25–28, 1986.
55. M. Epstein, EPA Alkali Scrubbing Test Facility: Advanced Program,”
39. K. Wark and C. F. Warner, Air Pollution: Its Origin and Control. New PB 245 279. Springfield, VA: National Technical Information Service,
York: IEP, 1976. 1975.
40. H. J. White, “Electrostatic Precipitation of Fly Ash: Part II, Section 3—Fly 56. R. H. Borgwardt, “EPA/RTP Pilot Studies related to Unsaturated
Ash and Furnace Gas Characteristics,” Journal of the Air Pollution Con- Operation of Lime and Limestone Scrubbers,” Combustion, 47: 37–42,
trol Association, 27(2): 114–120, 1977. 1975.
41. C. C. Shale, J. H. Holden, and G. E. Fasching, “Electrical Resistivity 57. B. F. Jones, P. S. Lowell, and F. S. Meserole, “Experimental and Theo-
of Fly Ash at Temperatures to 1,500°F.,” U.S. Bureau of Mines Report retical Studies of Solid Solution Formation in Lime and Limestone
of Investigations 7041. Washington, D.C.: U.S. Government Printing Scrubbers,” Vols. 1 and 11. Final Report. PB-264 953 (EPAJ600/2-
Office, 1968. 76/273a) and PB-264 954 (EPA/600/2-76/273b). Springfield, VA:
National Technical Information Service, 1976.
42. A. B. Walker, “Operating Experience with Hot Precipitators on Western
Low-Sulfur Coals,” Proceedings of the American Power Conference, 39: 58. P. C. Rader, M. R. Gogineni, and K. Poglitsch, “Co-Precipitation: A
582–594, 1977. Method for Gypsum Scale Prevention in Lime/Limestone Flue Gas
Desulfurization Systems,” Presented at the 72nd Annual Meeting of the
43. W. F. Frazier and W. Borowy, “Evaluation of Utility Cold-Side Elec-
American Institute of Chemical Engineers, San Francisco, California,
trostatic Precipitators,” Presented at the ASME-IEEE Joint Power
November 25–29, 1979.
Generator Conference, Milwaukee, Wisconsin, October 20–24, 1985.
59. G. T. Rochelle, D. R. Owens, J. C. S. Chang, and T. G. Brna, “Thiosul-
���������
44. R. M. Kotan, E. G. Drdla, P. Yosick, and K. Hognefelt, “Omaha Public
fate as an Oxidation Inhibitor in Flue Gas Desulfurization Processes: A
Power District Nebraska City Unit No. 1 Electrostatic Precipitator
Review of R&D Results,” Presented at the Ninth Symposium on Flue
Upgrade,” Presented at the Ninth Particulate Control Symposium, Wil-
Gas Desulfurization, Cincinnati, Ohio, June 4–7, 1985.
liamsburg, Virginia, October 1991.
60. R. H. Borgwardt, “Sludge Oxidation in Limestone FGD Scrubbers,”
45. R. E. Bickelhaupt, “An Interpretation of the Deteriorative Performance
EPA-600/7-77-061. Washington, D.C.: U.S. Environmental Protection
of Hot Side Precipitators,” Journal of the Air Pollution Control Associa-
Agency, 1977.
tion, 30: 812, 1980.
5-135
61. G. T. Rochelle and C. J. King, “The Effect of Additives on Mass Transfer 76. I. J. Urza and M. L. Jackson, “Pressure Aeration in a 55-ft Bubble
in CaCO3 or CaO Slurry Scrubbing of SO2 from Waste Gases,” Indus- columns,” Industrial & Engineering Chemistry Process Design and Devel-
trial & Engineering Chemistry Fundamentals, 16: 67–75, 1977. opment, 15: 106–113, 1975.
62. Y. J. Lee and G. T. Rochelle, “Oxidative Degradation of Organic Addi- 77. A. J. Toprac and G. T. Rochelle, “Limestone Dissolution in Stack Gas
tives for FGD: Products, Kinetics, and Mechanism,” Environmental Desulfurization Processes—Effect of Type and Grind Modeled by
Science and Technology, 21: 266–272, 1987. Mass Transfer,” Presented at the Annual AIChE Meeting, New Orleans,
Louisiana, November 8–12, 1981.
63. R. Ruiz-Alsop and G. T. Rochelle, “Co-precipitation of Organic Acids
with CaSO3 Solids,” Industrial & Engineering Chemistry Response, 27: 78. D. Laslo and E. Bakke, “The Effect of Dissolved Solids on Limestone
2123–2126, 1988. FGD Scrubbing Chemistry,” Presented at the ASME 1983 Joint Power
Generation Conference, Indianapolis, Indiana, 1983.
64. J. H. Lester and D. Danley, “Buffered Flue Gas Scrubbing System Using
Adipic Acid By-Product Stream,” U.S. Patent 4423018, December 1983. 79. F. C. Bond, “Crushing and Grinding Calculations: Part I,” British Chemi-
cal Engineering, 6: 378–385, 1961.
65. S. Wang and D. Burbank, “Adipic Acid Enhanced Lime and Limestone
Testing at the EPA Alkali Scrubbing Test Facility,” PB 82 230 624. 80. F. C. Bond, “Crushing and Grinding Calculations: Part II,” British
Springfield, VA: National Technical Information Service, 1982. Chemical Engineering, 6: 543–548, 1961.
66. O. W. Hargrove, J. D. Colley, R. L. Glover, and M. L. Owen, “Full- 81. R. Meij, “The Fate of Trace Elements at Coal-Fired Power Plants,”
Scale Utility FGD (Flue Gas Desulfurization) System Adipic Acid Proceedings of the Second International Conference on Managing
Demonstration Program. Volume 1. Process Results,” PB 83 238 683. Hazardous Air Pollutants, Washington, D.C., 1993.
Springfield VA: National Technical Information Service, 1983.
82. F. Frandsen, “Trace Elements from Coal Combustion,” Ph.D. Thesis,
67. J. D. Mobley and J. C. S. Chang, “The Adipic Acid Enhanced Limestone Department of Chemical Engineering, Technical University of Den-
Flue Gas Desulfurization Process: An Assessment,” PB83131219. mark, 1995.
Springfield, VA: National Technical Information Service, 1981.
83. U.S. Geological Survey, Fact Sheet 095-01.
68. K. Nolin, “Newly Developed FGD Plant for the Oil-Fired Condensing
84. K. Dombrowski et al., “Mercury Control Evaluation of Furnace Halogen
Power Station in Karlshamn, Sweden,” VGB PowerTech, p. 37, 2000.
Injection at TUX’s Monticello Unit 3,” Presented at the International
69. K. Nolin and D. Schreyer. “Flowpac®—Major WFGD Advance in Flue Conference on Air Quality, Arlington, Virginia, September 19–21, 2005.
Gas Contact,” Presented at the EPA/EPRI Mega Symposium, Wash-
85. J. A. Withum, “Mercury Emissions from Coal Fired Facilities with
ington, D.C., September 2004.
SCR-FGD Systems,” DOE/NETL Mercury Control Technology
70. M. Maripuu et al., “Design of the Flowpac® WFGD System for the Conference, Pittsburg, PA, December 12, 2006.
Amager Power Plant,” Presented at Power-Gen, Orlando, Florida,
86. L. Lindau, “Mercury Sorption to Coal Fly Ash,” Staub-Reinhaltung der
November 2006.
Luft, 43, pp. 166–167, 1983.
71. J. Chang and D. Laslo, “Chloride Effects on Limestone FGD System
87. S. Sjostrom, “Evaluation of Sorbent Injection for Mercury Control,”
Performance,” Presented at the EPA/EPRI FGD Symposium, Hol-
U.S. Department of Energy Cooperative Agreement DE-FC26-
lywood, Florida, 1982.
03NT41986, Topical Report 41986R07.
72. D. Laslo, J. Chang, and J. Mobley, “Pilot Plant Tests on the Effect of
88. “MIT Study on the Future of Coal.” Cambridge, MA: MIT, 2007.
Dissolved Salts on Lime/Limestone FGD Systems,” Presented at the
EPA/EPRI FGD Symposium, New Orleans, Louisiana, 1983. 89. N. Y. Nsakala et al., “Greenhouse Gas Emissions Control by Oxygen
Firing in Circulating Fluid Bed Boilers,” Alstom Power–U.S. Depart-
73. J. C. S. Chang and N. Kaplan, “SO2 Removal by Limestone Dual Alkali,”
ment of Energy (DOE) Report, PPL Report PPL-03-CT-09, U.S.
Environmental Progress (3), 267–274, New York, November, 1984.
DOE Project Manager: Suresh C. Jain, May 15, 2003.
74. W. H. D. Plant and W. L. Mathay, “Nickel-Containing Materials in
90. H. E. Andrus, Jr., et al., “Hybrid Combustion-Gasification Chemi-
Flue Gas Desulfurization Equipment,” Technical Paper 10072. Toronto:
cal Looping Coal Power Technology Development, Phase 1 Report,”
Nickel Institute, 1999.
Alstom Power–U.S. Department of Energy (DOE) Contract DE-
75. M. L. Jackson et al., “Oxygen Transfer in a 23-Meter Bubble Column,” FC26-03NT41866, U.S. DOE Project Manager: Suresh C. Jain,
AIChE Symposium Series, 71(151): , 1976. December 2004.
5-136
SUGGESTED READINGS
F. A. Ayer, ed., Proceedings: Symposium on Flue Gas Desulfurization, Hol- Krebs Engineers, “The Use of Hydrocyclones in Classification and Separa-
lywood, Florida, 1977. Report EPA-600/7-7 8/058. Springfield, VA: tion,” Presented at the Sixth Annual Meeting of the Association for
National Technical Information Service, 1978. Crystallization Technology, Charlottesville, VA, 1995
G. P. Behrens et al., “The Evaluation and Status of Flue Gas Desulfurization C. C. Leivo, “Flue Gas Desulfurization Systems; Design and Operating
Systems.” EPRI Report CS-3322, Project 982-28. Prepared by Radian Considerations. Vol. II,” Technical Report PB-280 254 (EPk600/7-78/
Corporation, January, 1984. O3oB). Springfield, VA: National Technical Information Service, 1978.
D. C. Borio et al., “Design of Spray Tower Absorbers for Lime/Limestone E. C. Lewis, G. W. Driggers, and K. W. Malki, “Laboratory Evaluation of
Wet Scrubbers,” ASME Paper 79-WA/Fu-9. New York: American Several Alloys in High Chloride FGD Environment—Progress Report,”
Society of Mechanical Engineers, 1979. Presented at Solving Corrosion Problems in Air Pollution Control
Equipment, Denver, Colorado, August 11–13, 1981.
D. Burbank and S. C. Wang, “EPA Alkali Scrubbing Test Facility:
Advanced Program — Final Report,” PB 8O 204 241. Springfield VA: F. W. Linkand and W. H. Ponder, “Status Report on the Weilman Lord/
National Technical Information Service, 1980. Allied Chemical Flue Gas Desulfurization Plant at Northern Indiana
Public Service Company’s Dean H. Mitchell Plant,” PB-282 091, Spring-
“Evaluation of Advanced Coal Technologies with CO2 Capture: Canadian
field, VA: National Technical Information Service, 1978.
CPC Phase 1 Studies of Coal Technologies with CO2 Capture,” Report
1004880. Palo Alto, CA: Electric Power Research Institute, 2004. T. Ohishi, “Mitsubishi CO2 Recovery Technology from Flue Gas: Experi-
ence and R&D Facilities,” Presented at the Ninth International CO2
D. Frahotta and P. C. Rader. “Lime/Limestone Air Quality Control
Capture Network Meeting, Copenhagen, June 16, 2006.
Systems: Effect of Magnesium on System Performance,” ASME Paper
76-WA/APC-10. New York: American Society of Mechanical Engineers, C. Rader et al., “The Role of Crystallization in the Design of Lime/Lime-
1976. stone Wet Scrubbing Systems for Flue Gas Desulfurization,” Presented
at the Symposium on Crystallization and Energy Systems, 83rd National
H. N. Head, “EPA Alkali Scrubbing Test Facility: Advanced Program.
Meeting of the American Institute of Chemical Engineers, Houston,
Third Progress Report,” PB 274 544 Springfield, VA: National Technical
Texas, March 20–24, 1977.
Information Service, 1977.
S. Reddy et al, “Fluor’s Econamine FG PlusSM Technology: An Enhanced
K. A. Hoff et al., “The Kvaerner–Gore Membrane Process for CO2
Amine-Based CO2 Capture Process,” Presented at the U.S. Department
Removal: Mathematic Model and Experimental Verification,” Presented
of Energy’s Second National Conference on Carbon Sequestration,
at the AIChE Annual Meeting, Los Angeles, California, November 2000.
Alexandria, Virginia, May 5–8, 2003.
“Just Catch™ Development Project for CO2 Capture from Gas Power
Y. Yagi et al, “Improvement of CO2-Capture Technology from Flue Gas,”
Launched by Aker Kvaerner,” Available at http://www.akerkvaerner.
Presented at the Seventh International Conference on Greenhouse Gas
com/Internet/MediaCentre/PressReleases/Group/2006/AKPressRe-
Control Technologies, Vancouver, Canada, September 5–9, 2004.
lease_1025215.htm
5-137
Auxiliary Equipment
Chapter Six
Although most of the attention is given to the main plant components, no plant can operate
properly without adequately designed auxiliary equipment. This equipment includes
pulverizers, air preheaters, ash-handling equipment, pumps, fans, feedwater heaters,
and condensers. This chapter provides an overview of this type of equipment.
Pulverizers and Pulverized respect are capable of burning almost operational flexibility, and allowed
Coal Systems any fuel which is used for making the practical use of an extremely wide
Suspension firing of coal as a fine steam.” Henry Kreisinger, former range of fuels.
powder has been the predominant Director of Research at Combustion Since the first information at-
method of coal combustion in utility Engineering, made this statement tributed to the Niepce brothers
scale steam generators since the 1950s. at a 1937 meeting of the American was published in France in the
Suspension firing has dominated over Society of Mechanical Engineers in early 1800s, there have been many
lump-coal combustion techniques for Windsor, Canada.l examples of engineers whose visions
a number of reasons. Pulverized coal is One of the most significant engineer- of future developments in pulverized
easily ignited and controlled with char- ing achievements in the 20th century coal technology have far outreached
acteristics similar to those of natural was the commercial development of the materials and technical under-
gas. Pulverized coal is easily trans- methods for firing coal in pulverized standing of their time. In his 1824
ported with air, allowing multiple fuel form. This development made pos- engineering classic, Reflections on the
injection points, which makes furnace sible extremely large, technologically Motive Power of Fire,2 Sadi Carnot
design and size essentially unlimited. advanced, steam-generating units oper- and his colleagues provided a critical
Virtually all coal types can be reduced ating with minimized emissions, high thermodynamic analysis of the pyreo-
to a fine powder, making suspension thermal efficiency, high reliability, and lophore, an engine fired by powdered
firing technology universally applicable increased safety. coal. During the 1890s, Rudolf Die-
in all geographic regions. Worldwide, practically every type of sel conducted his first experiments
coal mined is now being burned suc- on the internal combustion engine
HISTORICAL PERSPECTIVE cessfully in pulverized form. Similarly, bearing his name using pulverized
“A change of coal upsets the operation many other types of low-grade, waste, coal as the primary fuel. At this time
of a pulverized-coal plant to a much and by-product solid fuels also may pulverized coal firing was achieving
smaller degree than it does a stoker be fired economically and efficiently in its first real commercial success in the
fired plant. Pulverized-coal furnaces this manner. Pulverized fuel firing has cement industry.3,4 In the early 1900s,
can be readily adapted to burn other contributed to the reduction of labor Thomas A. Edison made improve-
fuels that burn like gas, and in that costs in steam power plants, increased ments in the firing of pulverized coal
6-
6- | Clean Combustion Technologies
in cement kilns, greatly increasing power industry was needed. John An-
their efficiency and output.5,6 derson, then chief engineer of power
Largely developed as an empirical plants of what is now WE Energy,
art, pulverized coal firing progress provided the necessary leadership. He
has been marked by the efforts of effectively enlisted the support and ac-
devoted engineers whose success may tive participation of exceptionally able
be attributed to persistence in the face engineers from his own organization,
of many discouraging obstacles. In the public utility industry, equipment
general, theoretical understanding has suppliers, and the U.S. Bureau of
followed rather than preceded practical Mines. Anderson’s pioneering efforts
accomplishment in the field of pulver- resulted in the construction of pulver-
ized coal firing. ized coal installations in the Oneida
In the U.S., the rapid increase in Street Station and at Lakeside.12–19
oil prices in the 1890s provided the Two outstanding test reports stand as
principal incentive to develop the use engineering classics. Based on ex-
of pulverized coal for firing cement perimental work at the Oneida Street
kilns, the first industrial application and Lakeside Stations, these reports
to achieve commercial success. E. H. contain important basic information
Hurry and H. J. Seaman of the Atlas on pulverized coal firing and establish
Portland Cement Company began a a pattern for subsequent research,
series of experiments relating to the use including many of the activities of the
of pulverized coal in 1894, and in the ASME Furnace Performance Factors
following year, pulverized coal was suc- Committee.20–25
cessfully used in a rotary cement kiln. Studies on velocities and charac-
Since that time, pulverized coal has teristics of pulverized coal particles
been the dominant fuel in the cement were reported by E. Audibert26 and
industry. By the time of World War I, John Blizard27 who a few years earlier
powdered coal, the term then gener- had published a comprehensive study
ally used for what is now designated of the state of the art of pulverized
“pulverized coal,” had gained sufficient coal.28 Research linking studies of
acceptance for the American Society inflammability of coal mine dust to
of Mechanical Engineers (ASME) desired combustion properties appears
to sponsor a symposium bringing in an article by Henri Verdinne.29
together the accumulated experience in W. Nusselt published results of
several fields of application.7–11 research on the ignition times of
By the end of World War II, pulver- coal particles in 1924,30 and P. Rosin
ized coal still had not achieved its full reported on studies of thermody-
potential, despite an increasing number namic data based on heat liberation in
of applications, ranging from the 1925.31 The first of a series of papers
cement industry to the metallurgical on the study of heat transfer in boilers
industry, to the steam locomotive, and at Yale University by W. J. Wohlenberg
to several stationary boilers. Although and his colleagues was also published
all of the elements for success appeared in 1925.32 Despite the extensive
to be present, integration of the many theoretical studies that were made in
ideas to adequately demonstrate pul- the 1920s, much of the progress was
verized coal firing in the central station achieved on an empirical basis with
Auxiliary Equipment | 6-
Hardgrove Index
size or design coal pulverizers. The the moisture content changes. 60
energy required for pulverization is Pulverizing a small, air-dried sample Davenport
50
dissipated in a number of ways and of properly sized coal in a miniature Mine
affected by several process variables. It mill determines its grindability. Results 40
Kincaid
cannot be accounted for in the specific may then be converted into a grindabil- Mine
30
manner that is applicable to a boiler ity index factor that, with appropriate
or power plant heat balance. For this correction curves, can be used to 0 10 20 30 40 50
Moisture Content, %
reason, both the design and the ap- predict, or interpret, mill capacity.
plication of pulverizers have retained The Hardgrove method was
Figure 6-1 | Variation of grindability index
many of the elements of an engineering developed to measure the quantity of Clean Combustion
with Technologies
moisture content, North
art, as the demands on pulverizer per- new material that will pass through Figure: 06.01/ SPS
Dakota lignites (average of
PMS5415
standardK and corrected values)
formance have continually increased. a 200-mesh sieve after a standard
grinding duration. The apparatus for
PULVERIZING PROPERTIES OF COAL this method, shown in Figure 6-2, is Integral Motor
To predict pulverizer performance extremely simple. and Reducing Gears
on a specific coal with some degree of A 50-g (1.76 oz) sample of air-dried
Revolution Counter
accuracy, the ease with which the coal coal, sized to less than 16 and greater
can be pulverized must be estimated. than 30 mesh, is placed in the mortar Grease Seal
passing through the 200-mesh sieve is capacity because of the inability to re-
used in the preparation of a calibra- move the fines efficiently and as quickly
tion chart, from which the grindability as they are produced. Agglomeration of
of the coal sample is determined in fines has the same effect as coarse coal
accordance with ASTM Standards during the combustion process, because
D 409, Grindability of Coal by the the surface available for the chemical
Hardgrove-Machine Method. Four reaction is reduced. Since in-mill drying
coal samples, obtained from ASTM, is the accepted method of preparing coal
standardized especially for this for pulverized fuel burning, air with a
purpose and representing grindability sufficient mass flow and temperature is
indices of 40, 60, 80, and 100, are used necessary in the milling system transport
180 °F Leaving Mixture Temperature to calibrate each grindability machine air. The process of flash drying coal
and associated apparatus, before the contributes to the recirculating load in
% H2O
700 Entering - Leaving equipment is used to test coals. an operating pulverizer. If high levels
16-2.0 Often, too much emphasis is placed of moisture are present in bituminous
Air Inlet Temperature to Mill, °F
600 14-2.0 on grindability while other factors af- coals, a reduction in the grinding capac-
12-2.0
10-2.0 fecting mill capacity, such as moisture, ity of the pulverizer will occur. Curves
8-2.0
500 6-1.5 are almost entirely overlooked. Pulver- such as those shown in Figures 6-3 and
4-1.5
2-1.0 izer capacity is proportional to the 6-4 indicate the air temperature required
400 grindability index of the coal within to dry coal of various total moistures and
a range of 40 to 60 HGI (Hardgrove coal-air mixtures.
300 Grindability Index). Outside this Achieving rated pulverizer capac-
range, varying correction factors apply, ity depends on having sufficient heat
200 depending on machine design and input to dry the coal. If there is a
1 2 3 4 5 manufacturer. Corrections must also deficiency of hot air, the mill output
Lbs of Air Leaving Mill/Lb of Coal
be made for fineness of product and will be limited to the “drying capac-
moisture of the raw feed. ity” and not the “grinding capacity” of
Figure 6-3 | Temperature of air to mill,
Clean Combustion Technologies the mill. Thus, it may be possible to
Figure:Eastern
06.03/U.S.
(JG)coals Moisture
PMS5415 K obtain more capacity with a relatively
Usually, a reference to moisture in coal dry coal with a lower grindability
170 °F Leaving Mixture Temperature pertains to the total moisture content, index than with a high-moisture coal
% H2O Entering - Leaving which comprises what is commonly with a higher grindability index. Pre-
26-7.5 termed “equilibrium moisture” and dicting pulverizer capacity for a given
700 24-7.0
22-6.5 surface or free moisture. Equilibrium fuel-rank grindability and moisture
20-6.0
Air Inlet Temperature to Mill, °F
18-6.0 moisture varies with coal type or rank content involves independent analysis
600 16-5.5
14-5.0 and mine location, and would be of the machine’s grinding and thermal
12-5.0
10-4.5 more accurately called “bed” or “seam” capacities, as well as the impact of the
500 8-4.0
6-4.0 moisture. In reality, surface moisture is moisture content on grinding capacity.
400 the difference between total moisture
and bed moisture. Surface moisture Relation of Pulverizer Capacity
300 adversely affects both pulverizer per- to Grindability
formance and the combustion process. As stated previously, mill capacity is
200 The surface moisture produces agglom- not directly proportional to grindabil-
1 2 3 4 5
Lbs of Air Leaving Mill/Lb of Coal eration of the fines in the pulverizing ity for any given fineness distribution.
zone, and reduces pulverizer grinding This is in part due to the differences
Figure 6-4 | Temperature of air to mill,
Clean Combustion Technologies
Figure:Midwest-U.S.
06.04/ (JG) coals
PMS5415 K
Auxiliary Equipment | 6-
Hardgrove Grindability
required to produce more surface area 100 85
90 80
increases as a finer product is made, the 80 75
70
product fineness grindability index and 70 65
60 Fineness,
coal moisture content must be consid- 50 % Through 200 Mesh
ered together to develop reasonable 40
60 70 80 90 100 110 120 130
expectations for pulverizer throughput. Base Mill Capacity, %
The effects of a finer product and a
wider range for the coal grindabil- Figure 6-5 | Mill correction factors for grindability, fineness, coal type, and moisture—bituminous
ity index have become predominant coals with preheatedClean
air and static classifier
Combustion Technologies
considerations since the early 1990s, Figure: 06.05/ (JG)
Variation of Lignite Grindability PMS5415 K
as commercial pulverizing equipment
with Moisture
is required to support reduced emis-
sion combustion with a wide range of The U.S. Bureau of Mines at Grand
globally sourced coals. Although the test Forks, North Dakota, and others have
equipment does not indicate a direct reported the grindability of lignites at
proportion of capability between hard various moistures. The results show a
and soft materials, the value of these wide variation, as in Figure 6-1. Some
tests remains a key input used for per- think that such curves are of little
formance assessment. Correction factors value because it seems impossible to
developed by pulverizer manufacturers select the proper index from them.
on commercial equipment provide for Others think that these curves do
overcoming these discrepancies. Figure have significance. With the increased
6-5 gives correction curves for variations use of lignites, the solution to this
in fineness, grindability, and moisture. problem is important.
As a rule of thumb, for every point the In setting up the present ASTM
grindability index changes within the code for grindability, the test specifies
range of 40 to 60 HGI, there will be a the use of an air-dried sample. In the
corresponding change of about 1.5% in Alstom bowl mill, all of the surface
pulverizer capacity. Changes in fineness moisture and some of the equilib-
have a similar impact on pulverizer rium moisture are evaporated during
throughput, although depending on the pulverization with a hot air sweep. The
fineness range, the corresponding capac- moisture content of the pulverized
ity change will be between 1 and 5% per product leaving the mill is thought
%-200 mesh fineness. to be the moisture level that exists in
6- | Clean Combustion Technologies
the grinding zone. Thus, grindability After the predrying step is accom-
versus moisture indices above the plished in the pulverizing operation,
equilibrium level is of little interest. the low-rank sub-bituminous coals
Hardgrove indices, therefore, have and lignites manifest a higher degree
meaning to the pulverizer designer of reactivity than do the higher-rank
only below the equilibrium moisture bituminous coals. Recent investigations
level and in the general range of mois- confirm that this increased reactivity
ture contents between 10 and 25%. results primarily from the lack of ag-
The actual choice of the grindability glomerating properties and increased O2
index for pulverizer design capacity re- content of the fuel.33 A comparison of
quires consideration of total moisture, flammability indices with volatile matter
equilibrium (bed) moisture, heat input and heating value suggests that factors
to the pulverizer, and experience with other than these have a great influence
the coal’s drying and grinding behavior. on ignition temperature and particle
burnout. Because lower flammability
Relationship of Coal Rank temperatures are experienced primarily
to Required Fineness with lignitic and sub-bituminous coals,
Successful pulverized coal firing it is reasonable to ask what property is
depends on recognizing differences in common to lower-rank coals. Exclusive
coal constituents and the impact they of heating value, the most obvious dif-
have on efficient combustion. Experi- ference between the two groups is the
ence over the years has established that property known as “agglomerating char-
a relationship exists between the rank acter.” Sub-bituminous coals and lignites
of bituminous coals and the degree do not agglomerate. As applied to coals,
of fineness required for successful agglomeration is the property of coal
operation. To ensure complete com- particles to fuse into a coke-like mass or
bustion within the furnace confines for bond together into a firm cake when the
minimal carbon loss, high-rank coals particles are heated to temperatures of
must be pulverized to a finer size than 1,000°F (540°C) or above.
coals of lower rank. Since the 1990s, in While the determinations of both
furnace air staging technologies have volatile matter and heating value are
been widely applied as an effective well defined, tests for establishing the
means to reduce thermal NOx produc- agglomerating character of coals are less
tion. These systems create a lower commonly known. ASTM Standards
stoichiometric region at the point of D 388, Specifications for Classification
initial combustion, which calls for a of Coals by Rank, describes agglomerat-
finer coal product to support complete ing character as follows:
burnout. When firing certain coals in
the low-volatile content group in small The test is carried out by the ex-
pulverized coal furnaces, the fineness amination of the residue in the platinum
may be increased to as high as 80% to crucible incidental to the volatile matter
ensure adequate burnout of the carbon determination. Coals which in the volatile-
content. Table 6-1 lists the specifications matter determination produce either an
set by ASTM for classifying coal by agglomerate button that will support a
rank. Rank classifications are based on 500-g weight without pulverizing, or a
varying combinations of volatile-matter button showing swelling of cell structure,
content, heating value, and agglomerat- shall be considered agglomerating from
ing properties. the standpoint of classification.
Auxiliary Equipment | 6-
4. High-volatile B bituminous §
13,000 BTU/lb 14,000
coal
— — — —
(30,000 KJ/kg) (32,600 KJ/kg)
5. High-volatile C bituminous — — — — 11,500 BTU/lb 13,000
coal (26,700 KJ/kg) (30,000 KJ/kg)
10,500 BTU/lb 11,500 Agglomerating
(24,400 KJ/kg) (26,700 KJ/kg)
III. Sub-bituminous
10,500 BTU/lb 11,500
1. Sub-bituminous A coal — — — —
(24,400 KJ/kg) (26,700 KJ/kg)
9,500 BTU/lb 10,500
2. Sub-bituminous B coal — — — —
(22,100 KJ/kg) (24,400 KJ/kg)
3. Sub-bituminous C coal — — — —
8,300 BTU/lb 9,500 Non-
(19,300 KJ/kg) (22,100 KJ/kg) agglomerating
IV. Lignitic
6,300 BTU/lb 8,300
1. Lignite A — — — —
(14,700 KJ/kg) (19,300 KJ/kg)
6,300
2. Lignite B — — — — —
(14,700 KJ/kg)
* This classification does not include a few coals, principally nonbanded varieties, which have unusual physical and chemical properties and which come within the limits of
fixed carbon or calorific value of the high-volatile bituminous and sub-bituminous ranks. All of these coals either contain less than 48% dry, mineral-matter-free fixed carbon
or have more than 15,500 moist, mineral-matter-free British thermal units per pound.
†
Moist refers to coal containing its natural inherent moisture but not including visible water on the surface of the coal.
‡
If agglomerating, classify in low-volatile group of the bituminous class.
§
Coals having 69% or more fixed carbon on the dry, mineral-matter-free basis shall be classified according to fixed carbon, regardless of calorific value.
||
It is recognized that there may be nonagglomerating varieties in these groups of the bituminous class, and there are notable exceptions in high-volatile C bituminous group.
Reprinted from ASTM Standards D 388, Classification of Coals by Rank.
* Agg., agglomerating; HHV, higher heating value; MAF, moisture ash free; Non-Agg., non-agglomerating
† Agglomerating but non-caking.
Auxiliary Equipment | 6-
per linear inch designates the mesh of mesh (–200 mesh), as little as 3 to 5% Table 6-3 | Comparison of
a screen. Thus, a 200-mesh screen has greater than 50 mesh (+50 mesh) may Sieve Openings
200 openings to the inch, or 40,000 produce furnace slagging and increased
U.S. Standard Sieve
per square inch. The diameter of the combustible loss, even though combus-
Mesh Inches Millimeters
wire used in making the screen governs tion conditions are excellent for the
20 0.0331 0.840
the size of the openings. The U.S. finer coal. The small amount of oversize
Standard and W. S. Tyler are the most represents very little additional surface 30 0.0234 0.595
common screen sieves. The mesh and if it is pulverized to all –50 mesh and all 40 0.0165 0.420
opening of these and other interna- +200 mesh. As an illustration, assume a 50* 0.0117 0.297
tional screens are shown in Table 6-3 typical screen analysis of a high-volatile 60 0.0098 0.250
and Figure 6-6. bituminous coal sample, pulverized to 100* 0.0059 0.149
80% –200 mesh:
CLASSIFICATION AND 140 0.0041 0.105
SIZE CONSIST 99.5% –50 mesh 200* 0.0029 0.074
96.5% –100 mesh 325 0.0017 0.044
In the combustion of pulverized coal, it
80.0% –200 mesh
is important to have a proper percent- 400 0.0015 0.037
age of fine particles yielding a large This represents a surface area of W. S. Tyler Sieve
surface area. It is equally necessary approximately 1,500 cm2/g, with over Mesh Inches Millimeters
to reduce or eliminate the oversize 97% of the surface in the –200 mesh
20 0.0328 0.833
on the coarser screen. Despite the portion. By overgrinding and poor
28 0.0232 0.589
small percentage of less than 200 classification it would be possible, on
35 0.0164 0.417
48* 0.0116 0.295
USA England Germany France 60 0.0097 0.246
Tyler ATSM *IMM Std †Din ‡AFN Std
USBS *BESA 1171 Usuelle 100* 0.0058 0.147
0.03
No. No. No. No. No. No. No. 30 150 0.0041 0.104
400 400 17
0.04 325 325 200 300 0.06 300 40 200* 0.0029 0.074
270 270 18
0.05 50
0.06 230 230 240 19 250
220 60 325 0.0017 0.043
0.07 200 200 200 0.075 20 200 70
150 400 0.0015 0.037
0.08 170 170 170 0.09 190 80
0.09 150 140 21 170
0.1 120 150 0.1 150
100
115 120 100 120 0.12 22 120 * Commonly used screens in pulverized-coal practice
100 100 90 100 0.15 23 100 for combustion purposes.
0.15 80 80 90 150
80 70 85 0.2 80
65 24
0.2 70 60 72 60 200
60 60 50 60 0.25 25
Screen Opening, mm
Screen Opening, �m
0.3
48 50 40 52 0.3 26 50 300
42 45 35 44
35 0.4 27 40
0.4 40 30 36 400
32 35 25 30 0.5 28 30
0.5 28 500
30 25 0.6 29
0.6 20 600
24 25 22 0.75 20
0.7 16 30 700
0.8 20 20 18 800
0.9 16 18 1.0 31
1.0 12 16 1,000
14 16 1.2 16
12 10 14 32
14 12 1.5 14
1.5 10 12 8 33 12
10 2.0
9 10 8 34 10
2.0 5
8 8 7 2.5 35
7 7 6 3.0
3.0 6 36
6 5
5 5 4.0 37
4.0 4 4 38
5.0
at one stage of pulverization contains rings, races, and liners gradually erode
50% material that will pass through a and wear out as a result of abrasion
50-mesh sieve, and, if this –50-mesh and metal displacement in the grind-
material is removed from the grinding ing process. The power for grinding,
zone, it will contain a smaller percentage and the maintenance of the grinding
of –200-mesh material than if it had elements make up the major costs of
been permitted to remain in the grind- the pulverizing operation. In itself,
ing zone until the total quantity had “pure coal” is relatively nonabrasive.
been reduced to pass a 50-mesh sieve. However, silica and quartz commonly
As already noted, an abundance of found in the coal ash are quite abrasive.
fine particles is necessary to ensure Foreign materials such as slate, sand,
prompt ignition of coal in suspension and pyrites, commonly found in coal
burning. Substantial energy consump- as mined, are also quite abrasive. These
tion is required in the production of are the undesirable constituents that
this fine material. However, when produce rapid and sometimes excessive
grinding finer than necessary, power wear in pulverizing apparatus. The eco-
is wasted. The pulverizing equipment nomics of coal cleaning to remove such Table 6-4 | Mohs Scale of
must then be larger than actually abrasive foreign materials depends on Hardness
required. Removal of the fines from the many variables and must be deter- Hardness Mineral
pulverizing zone as rapidly as they are mined for each individual application. 1 Talc
produced and return of the oversize The resistance of a smooth planar 2 Gypsum
for regrinding eliminates unnecessary surface to abrasion is called its hard-
3 Calcspar
production of fines and reduces energy ness. It is commonly recorded in terms
4 Fluorspar
requirements. Better product sizing of 10 minerals according to the Mohs
and increased capacity result from the scale of hardness (Table 6-4). There 5 Apatite
continuous removal of the fines. This is no quantitative relation between 6 Feldspar
process is called “closed circuit grinding.” these materials. The diamond is much 7 Quartz
The pulverizing system component that greater in hardness above sapphire 8 Topaz
accomplishes this size control is known than sapphire is above talc. Hardness 9 Sapphire
as the “classifier.” The fineness require- of selected common materials is shown
10 Diamond
ments for utility boilers were satisfied in Table 6-5. The relatively low hard-
by static double-cone classifiers through ness of pure coal is compared to the
the late 1980s. By the early 1990s, many abrasive impurities usually found in the Table 6-5 | Common Materials
countries had implemented stationary commercial product. and Their Mohs Hardness
boiler emission regulations, which Material Hardness
drove the need for in-furnace emis- Coal Preparation
Coal 0.5–2.5
sions reduction technologies. These Coal should be prepared and deliv-
Slate 0.50–6.0
technologies require finer coal to sup- ered properly for safe, economical, and
port complete combustion and avoid efficient use in a pulverizing system. Mica 2.0–6.0
excessive carbon loss. The need for a Controllable continuity of flow to Pyrite 6.0–6.5
higher level of product fineness has the pulverizer must be maintained. Granite 6.5
been met through the implementation Organic foreign materials such as Marble 3.0
of integral rotating classifiers. wood, cloth, and straw should be Soapstone 1.0–4.0
removed. Such materials may collect
Kaolin clay 2.0–2.5
Abrasion in the milling system and become a
Iron ore 0.50–6.5
Pulverizing results in an eventual loss of fire hazard. They may also impair ma-
grinding-element material. Balls, rolls, terial or airflow patterns in the mill. Carborundum 9.5
6-12 | Clean Combustion Technologies
any pulverizer and especially important coal on the feeder belt is a solid-state
in a direct-fired system for smooth load cell across a weigh span on the belt. Raw-Coal Inlet
boiler operation. There are several This same belt-feeder design can also
types, including the belt feeder and the be used for volumetric measurement.
overshot roll feeder. Another acceptable method of continu- Hinged
Levelling
ously weighing the coal feed rate uses a
Gate
THE BELT FEEDER. The belt feeder series of levers and balance weights.
Stationary
uses an endless belt running on two Core
separated rollers receiving coal from THE OVERSHOT FEEDER. The overshot roll Revolving Blade
above at one end and discharging it feeder (Figure 6-10) has a multi-bladed
Hot-Air Slot
at the other. Varying the speed of the rotor that turns about a fixed, hol-
driving roll controls the feed rate. A low, cylindrical core. This core has an
leveling plate fixes the depth of the opening to the feeder discharge and is
coal bed on the belt. The belt feeder provided with heated air to minimize
can be used in either a volumetric or wet-coal accumulation on surfaces and Figure 6-10 | Overshot roll feeder
gravimetric type of application (see to aid in coal drying. A hinged, spring- Clean Combustion Technologies
Figure: 06.10/ (A.R.)
Figure 6-9). loaded leveling gate mounted over the PMS5415 K
The gravimetric type has gained wide rotor limits the discharge from the
popularity in the industry for accurately rotor pockets. This gate permits the pas-
measuring the quantity of coal delivered sage of oversize foreign material. Feeders
to an individual pulverizer. Generally, of this type may be separately mounted,
this type is applied to steam genera- or they may be integrally attached to the
tors with combustion-control systems side of a pulverizer. The roll feeder and
requiring individual coal metering to the belt feeder, by virtue of their designs
the fuel burners. The most common may be considered highly efficient
method of continuously weighing the volumetric feeding devices.
6-14 | Clean Combustion Technologies
METHODS OF PULVERIZING
AND CONVEYING COAL
Raw-Coal Bunker
Early coal-pulverizing installations To Boiler
Furnace
received undried coal and used ambi-
ent air in the mill system. Because no Raw-Coal
Feeder
heat was added to the system, the coal
feed was limited to that of very low Hot Air
moisture content. Therefore, maximum Motor
Pulverizer Exhauster
pulverizer capabilities were not real-
ized. Subsequently, external coal dryers
were added to the system. Because Figure 6-12 | Direct-fired pulverizing system
the height of the coal column above the ric damper through which a flow of
feeder should be about 1 foot (0.3 m) ambient air is induced by the suction
for every 3 inches of primary air, mill in the pulverizer. With this control, the
inlet pressure. If the primary air fan fan is designed for a constant, low-tem-
outlet pressure is 30 inches, then the perature mixture and has low power
coal column height will have to be at consumption, even though such mate-
least 10 feet (3.0 m). rial handling fans have a relatively low
Three basic air conveyance systems efficiency of 55 to 60%.
have been used for direct firing: The main disadvantage of the
suction system is the maintenance re-
• the suction system, in which an
quired on the exhauster. On the other
exhauster induces airflow through
hand, by using proper design tech-
the pulverizer and discharges the
niques and wear-resistant materials,
coal–air mixture under pressure to
the maintenance on an exhauster can
the furnace.
be minimized. Exhauster maintenance
• the pressurized-exhauster system,
costs are more than offset by the power
in which the pulverizer is pressur-
and capital cost savings of the system.
ized by the forced draft fan with
This justifies the continued use of the
both hot and ambient air and
suction system on smaller units.
discharges the coal–air mixture
through an exhauster, acting as a Cold Primary Air System
booster fan, to the furnace (this
In this system, the primary air fan han-
system is used only in conjunction
dles only ambient air. The fan is located
with pressurized furnaces).
ahead of the air preheater, and supplies
• the cold primary-air system, in
tempering air and heated air via a sepa-
which a primary-air fan forces air
rate primary air section of the air heater.
through the air heater, ductwork,
Although not as simple as the other
and pulverizer and then forces the
two systems, its chief advantage is in fan
coal–air mixture into the furnace.
power and maintenance. Because the fan
Systems for use on pressurized handles cold air, it can be smaller, run at
furnaces were common through the higher speeds, and use highly efficient
early 1970s. For practical operations airfoil blade shapes without concern for
reasons, many of these boilers have wear. Inlet vanes can control airflow and
been upgraded to balance draft fur- further add to fan efficiency.
nace operation. Some other advantages of this
system are that, with high efficiency
Suction System fans handling ambient air, design for
The suction system has a number of ad- higher pressure differentials is possible
vantages. It is quite easy to keep the area and larger mills with longer fuel pipe
around the pulverizer clean. To control runs are practical. Thus, mills may be
the airflow through the pulverizer, a located farther from the boiler. Because
damper is placed in the constant-tem- individual fans for each pulverizer are
perature coal–air mixture just in front not necessary, the space requirements
of the exhauster entrance. Control of for the pulverizer bays can be reduced.
the coal–air mixture temperature is by Moreover, experience has demonstrat-
a single, hot air damper and a baromet- ed that metering airflow on the inlet air
6-18 | Clean Combustion Technologies
the years. All grinding mill types use Table 6-7 | TYPES OF Pulverizers
either one, two, or all three of the basic
Speed: Low Medium High
principles of particle size reduction,
namely, impact, attrition, and crushing. Ring-roll or Impact or hammer mill
Type: Ball-tube mill
ball-race mill Attrition mill
The speed of these machines may be
classified as low, medium, and high. The
four most commonly used pulverizers
Exhauster Feeder
are the ball tube, the ring-roll or ball-
race, the impact or hammer mill, and
the attrition mill. Their respective speed
characteristics are shown in Table 6-7. Cold Air
Physically large per unit of capac- sound. Most installations require that
ity, these mills require considerable insulated enclosures be erected over
floor space. Because of their size and each mill for noise attenuation.
weight, the initial capital cost is also Maintenance in the grinding zone
quite high. The presence of a high of ball-tube mills is relatively easy to
circulating load within the mill results perform. Periodically, a ball charge is
in an overproduction of fines within added to the mill to make up for metal
the mill charge. Using an adequate lost in the grinding process. It may
external classifier permits the removal take years to wear out the cast liners,
of the fine product from the grinding but considerable downtime is neces-
zone and reduces the production of sary for their replacement. Over the
extreme fines. The comparatively poor long run, maintenance costs per unit
mixing of heated air with the partially of coal ground are about the same as
pulverized material reduces the dry- for ring-roll type pulverizers. Because
ing efficiency of this type of mill. of limited operational flexibility and
High-moisture coals produce a large comparatively high capital and operat-
reduction in mill capacity. ing costs, ball-tube mills are not usually
Ball-tube mills are not well suited supplied on modern thermal plants.
to intermittent operation, as the large Extremely high-ash or low-grindability
amount of heat stored in the coal and fuels are applications that may warrant
ball charge may produce overheat- consideration of tube mills.
ing and fires when the mill is idle.
The mass of this mill type makes it Impact Mills
necessary to use high-power, high- An impact mill consists primarily of
Figure 6-17 | Impact mill with external
classifier and exhauster starting-torque motors. In addition, a series of hinged or fixed hammers
integrally mounted these mills emit excessive levels of revolving in an enclosed chamber lined
with heavy-cast wear-resistant plates.
Grinding results from a combina-
Clean Combustion Technologies tion of hammer impact on the larger
Hot Air
Figure: 06.17/ (A.R.) Pulverized Coal and Air particles and attrition of the smaller
PMS5415 K particles on each other and across the
Hammers
grinding surfaces. An air system with
the fan mounted either internally or
Exhauster externally on the main shaft induces a
Raw-Coal Whizzer or Fineness flow through the mill. An internal or
Feeder Regulator
external type of classifier may be used.
(See Figures 6-17 and 6-18.)
This class of mill is simple and
compact, low in cost, and may be built
Mill Drive in very small sizes. Its ability to handle
Shaft
high inlet-air temperatures, plus the
return of dried classified rejects to
the incoming raw feed, makes it an
excellent dryer. The high-speed design
Feeder
Drive Unit results in high maintenance and high
power consumption when suitable
Figure 6-18 | Impact mill with hammers, classifier, and fan on common shaft
Clean Combustion Technologies
Figure: 06.18/(HR)
PMS5415 K
Auxiliary Equipment | 6-21
fineness for combustion is produced. speed and use primarily crushing and
Progressive wear on the grinding ele- attrition of particles, plus a very small
ments produces a rapid drop-off in amount of impact, to obtain size reduc-
product fineness. It is difficult, if not tion of the coal. The grinding action
impossible, to maintain fineness over takes place between two surfaces, one
the life of the wearing parts. Using an rolling over the other. The rolling ele-
external classifier permits maintenance ment may be either a ball or a roll, while
of fineness, but only at the expense of the member over which it rolls may be
a considerable reduction in capacity either a race or a ring. The ball diameter
as parts wear. Because the maximum is generally from 20 to 35% of the race
capacity attainable from impact mills diameter, which can be as large as 100
is lower than that of most other types inches (254 cm). If the element is a roll,
Stationary Ring
and product fineness is generally low, its diameter may be from 50 to 60% of
these machines are not supplied for the ring diameter, which can be as much
large or modern steam plants. as 130 inches (330 cm). Its face width, Grinding Ball
depending on mill size, will vary from
Attrition Mills 15 to 20% of ring diameter.
Driving Ring
No true attrition mill is used for coal When the rolling elements are balls
pulverizing because of the high rate of (Figure 6-19), they are confined between
wear on parts. However, a high-speed races. In the majority of designs, the Figure 6-19 | Diagram of ball-race mill
mill, which uses considerable attrition lower race is the driven rotary member,
grinding along with impact grinding, is while the upper race is stationary.
used for direct firing of pulverized coal. Some older designs also use a rotat- Clean Combustion Technologies
Figure: 06.19/ (A.R.)
In this mill, the grinding elements con- ing upper race. The required grinding
PMS5415 K
sist of pegs and lugs mounted on a disk pressure is obtained by forcing the Main
rotating in a chamber. The periphery of races together with either heavy springs Vertical
Grinding Shaft
the chamber is lined with wear-resistant or pneumatic or hydraulic cylinders. Ring
plates, and its walls contain fixed rows of Some additional grinding pressure is
lugs within which the rotating lugs mesh. obtained from the centrifugal force of
Grinding Roll Assembly
The fan rotor is mounted on the pulver- the rotating balls.
izer shaft. Instead of an external classifier, There are two general classes of mill
a simple, shaft mounted rejector type that use rollers as the rolling elements. In Figure 6-20 | Diagram of roller-type ring-roll
mill journal assembly
is used. This design uses wear-resistant one (Figure 6-20), the roller assemblies
alloy lug and peg facings and casing lin- are driven and the ring is stationary, Clean Combustion Technologies
ings to reduce the wear effect on fineness while in another (Figure 6-21), the roller Figure: 06.20/ A.R.
Journal Assembly
PMS5415 K
and to extend the periods between part assembly is fixed and the ring rotates.
Grinding
replacements. This mill type exhibits the Perhaps the most frequently used Roll
unfavorable wear life and performance application of the first class is the
characteristics of the impact mills and is, Raymond® roller mill, used in storage
therefore, found primarily on older, low- systems. Grinding pressure is obtained
capacity installations. from centrifugal force resulting from
the rotation of these rolling elements.
Ring-Roll and Ball-Race Mills The other classification of ring-roll
Ring-roll and ball-race mills comprise mills, commonly known as bowl mills,
the largest number of pulverizers used are those in which the ring rotates. Grinding Ring
for coal grinding. They are of medium These constitute the largest number
Figure 6-21 | Diagram of bowl-type ring-roll
Clean Combustion
pulverizerTechnologies
journal assembly
Figure: 06.21/ A.R.
PMS5415 K
6-22 | Clean Combustion Technologies
of mills used for grinding coal. Bowl pulverizing equipment to grind different
mills are manufactured by most major coals economically and efficiently. This
boiler companies and have become the capability is a reflection of the grind-
standard for modern power stations. ing pressures available, the method of
The rotating bowls of these mills oper- application of this force, the speed of the
ate between 20 and 70 rpm, with larger moving elements, abrasion, and power
mills running at the slower speeds. and size limitations of the particular
Generally, bowl mills are equipped units. A list of materials capable of
with self-contained or integral clas- being commercially pulverized by the
sifiers to regulate the fineness of the mill types described could be extended
finished product. The classifiers can be indefinitely. Therefore, the scope of
of the fixed-vane or rotating type. In Table 6-8 covers combustible materials
some cases, this device may be external of primary concern in power generation.
to the mill itself and is then termed an
“in-stream classifier.” Primary air fans ALSTOM PULVERIZER DESIGN
or exhausters create a flow of heated air The positive-pressure-type, Alstom
through the mills. This heated air dries HP ring-roll, bowl mill is shown with
the coal, removes it from the grinding Dynamic™ classifier in Figure 6-22.
zone, carries it through the classifying When fitted with an exhauster (such
zone, and conveys it to the burner in a that the pulverizer operates below at-
direct-fired system. mospheric pressure), it is designated an
When this type of mill is provided HPS or HPPS mill, a cross-sectional
with sufficient air at a temperature view of which is given in Figure 6-23
to produce a satisfactory mill outlet with static classifier. Table 6-9 shows
temperature, it can handle very wet the type of mill used with different
coals with only a small reduction in furnaces and air systems, for the two
capacity. The high ratio of circulating pressure regimes in which these mills
load (classifier rejects returning to the can operate.
grinding zone for further size reduc-
tion) to output, with the resulting Operation of the HP Pulverizer
rapid reduction of average moisture Raw coal enters the center of the
content, facilitates the grinding pro- pulverizer through a center feed pipe. It
cess. These mills require less power falls onto a rotating bowl. Centrifugal
than other types on a specific energy- force causes the coal to move outward
consumption basis, kilowatt hours per from the center and under the three
ton, giving a significant advantage in journal assemblies, where it is crushed
operating cost. Physically, these mills by large rolls. To prevent physical con-
are compact and occupy a relatively tact of the rolls and bowl liners when
small amount of floor space per unit the mill is run without coal, a stop
of capacity. Some designs of ring-roll limits the downward movement of the
mills are extremely quiet in operation. journal assembly. The force to pulver-
Fineness of the product is relatively ize the coal is applied to the journal
uniform throughout the life of the assembly by an external spring. As the
grinding elements and easily controlled journal rotates about its trunnion in
by classifier adjustment. response to increasing coal feed, the
Several references have been made spring is compressed and the force for
to the ability of the various types of grinding is increased. The partially
Auxiliary Equipment | 6-23
pulverized coal passes over the rim of Table 6-8 | Types of Pulverizers for Various Materials
the bowl and is entrained by the rising
Impact and Ball Ring
hot air stream and is flash dried. The Type of Material Ball-tube
Attrition race roll
pyrites and tramp iron that enter the Low-volatile anthracite x — — —
mill with the coal follow the same path
High-volatile anthracite x — x x
as the coal until they pass over the rim
Coke breeze x — — —
of the bowl. Being denser than coal,
they cannot be carried upward by the Petroleum coke (fluid) x — x x
air stream and fall into the under-bowl Petroleum coke (delayed) x x x x
plenum. Once there, rejected materials Graphite x — x x
are swept around by a set of pivoted Low-volatile bituminous coal x x x x
scrapers until they reach the tramp Med-volatile bituminous coal x x x x
iron opening. Reject materials fall into
High-volatile A bituminous coal x x x x
a hopper external to the mill, which
High-volatile B bituminous coal x x x x
can be emptied with the mill in service.
Initial classification of the partially High-volatile C bituminous coal x — x x
Incoming Coal
Discharge Valve Feed Pipe
Seal-Air Header
Discharge Knife-
Gate Valves
DYNAMIC TM
Classifier Assembly
Separator Top
DYNAMIC TM
Classifier Rotor
Spring
Assembly
Separator Body
Grinding Roll
Journal
Assembly
Bowl
Vane Wheel
Assembly
Millside
Assembly
Pyrite Scraper
Bowl Hub
Pyrite Chute
Seal Air
Header
Planetary
Gearbox
Figure 6-22 | Alstom HP ring-roll bowl mill for positive-pressure operation (shown with
Clean Combustion Dyamic™ classifier)
Technologies
Figure: 06.22 (MF)
PMS5415 K
Auxiliary Equipment | 6-25
Incoming Coal
Feed Pipe
Converter Head
Exhauster
Crossover Pipe
Converter Head Vane
Classifier
Assembly
Deflector Regulator
Assembly
Separator Top
Cone Assembly
Grinding Roll
Journal
Assembly
Bowl
Vane Wheel
Assembly Millside
Assembly
Pyrite Scraper
Bowl Hub
Pyrite Chute
Planetary
Gearbox
rolls to increase the total volume of the mill top; each lever operates half
wear material available, and by using of the vanes. Particle separation is ac- Dynamic™ classifiers provide
roll material with high wear resis- complished in a stationary-cone, cyclone improved fineness that allow
tance. Combustalloy™ wear material, classifier. As described previously,
an Alstom proprietary weld overlay, oversized coal particles are returned to
for combustion modifications to
provides effective wear life five times the grinding zone through the return reduce NOx emissions.
greater than standard NiHard and can spout at the bottom of the cone, where
be easily rewelded when worn. they mix with the incoming raw coal
The externally mounted spring as- feed to increase drying efficiency. The
sembly (Figures 6-22 and 6-23) has a classifier and outlet configuration has
major advantage in that maintenance been designed to minimize areas of
personnel do not have to enter the pul- localized wear and reduce the number
verizer to inspect or adjust the springs. of components exposed to erosive wear.
Since the assembly is located away For very abrasive coal applications the
from the coal flow, erosion is elimi- static classifier and outlet assembly
nated. Spring travel can be maximized can be lined with 90%-alumina tile for
with an external mechanism, to allow extended wear life. Knife gate valves, at
large ungrindable material to pass the outlet of a positive-pressure HP
under the rolls until it is rejected from pulverizer, isolate it from the fuel
the mill. A positive spring-preload piping leading to the furnace, as rec-
locking device prevents any change in ommended by the NFPA. Each valve
spring setting during operation. The has a set of replaceable valve seats,
entire spring assembly is a cartridge coated to minimize wear. Pneumatic
type to reduce change-out time, to cylinders operating in unison open
allow for stocking of spare assemblies, and close the valves. They are actuated
and to allow convenient inspection and by a single solenoid valve, which also
refurbishing. The spring assembly, as controls the supply of purge-seal air
well as the journal assembly and the to the fuel piping. Figure 6-22 shows
main bowl–hub air seal, have laby- the seal-air header for this purpose,
rinth-type seals that use air for sealing located downstream of the discharge
between stationary and moving parts. valves. HP pulverizers are avail-
The pressure of the filtered seal air is 8 able with manually adjustable coal
to 16 inches WG above the mill-inlet deflectors at the mill outlet to adjust
pressure, with the flow rate a function pipe-to-pipe coal distribution.
of mill size.
Dynamic™ Classifier
Static Classifier and Alstom has been supplying rotating
Discharge Valve Assemblies classifiers since the 1960s. By the late
The top of the mill is made larger 1980s, emission regulations and the
than the main body to reduce coal–air competitive economic pressure of the
stream velocity and mill aerodynamic power industry required staged firing
pressure drop and to optimize classifier conditions and the use of lower-cost
efficiency. Plant personnel can adjust fuels. These factors increased the
the position of the classifier vanes need for increased mill capacity and
when the pulverizer is operating, by finer pulverized coal for new boil-
using two manually operated levers on ers, as well as retrofits to existing
6-30 | Clean Combustion Technologies
units. The Dynamic™ classifier has pulverizers for effective fineness con-
been optimized to meet these needs. trol. Because of the uniform flow field
The Dynamic™ classifier must be a set-up within the HP mill internals,
reliable piece of processing equipment. there is no need for stationary pre-
Use of simple, reliable, off-the-shelf conditioning vanes. Because the rotor
components has proven an effec- speeds are low, pressure loss and wear
tive approach to meet this goal. The rates are minimized. Classifier speed is
Dynamic™ classifier uses two oversized, typically ramped as a function of feed
spherical, roller bearings in a simple rate and is easily adjusted if needed
grease-lubricated arrangement, with for significant changes in fuel or firing
lip-type grease seals and clearance- conditions. Because the fineness con-
type air seals to prevent leakage and trol is automated, classifier speed can
contamination. The drive system has easily be biased to trade between high
proven to be very reliable and compact. fineness and high coal throughput with
A single, small, horizontally mounted, no other adjustments required.
variable speed motor driving an integral,
self-contained, bath-lubricated gearbox Coal–Air Exhauster Fans
provides the initial reduction. Final For suction-type HP pulverizers, Al-
speed reduction is handled by a single- stom supplies material handling fans to
set, multi-strand V-belt. Speed control match the mill and fuel-piping system
is managed by a variable-frequency drive requirements. These exhausters are
typically mounted in the plant’s motor horizontal-shaft, straight-bladed, cen-
control center away from the mill bay. trifugal fans, (Figure 6-26) operating at
The heart of the Dynamic™ classifier constant speed and at temperatures be-
is the rotor assembly located inside tween 150 and 180°F (65 and 80°C).
the mill upper housing. The rotational They are designed in accordance with
speed of the classifier rotor controls the NFPA standards. There are replaceable
product fineness leaving the mill. For liners on the housing scroll and sides.
a utility-sized mill, rotor speeds are in The fan spider is removable to make
the range of 50 to 80 rpm to produce replacing fan blades easy. The fan wheel
fineness levels in the range of 70 to is supported by anti-friction bearings
85% passing a 200-mesh sieve. Alstom mounted in an external bearing hous-
uses a single rotating stage for all HP ing located between the exhauster and
its driving motor. Exhausters used with
HP pulverizers are driven by separate
Fan Casing
motors. The pulverizer and exhauster
Fan Fan Bearing motors are interlocked logically to
Intake Assembly ensure coal transport out of the mill.
Side Fan bearing lubrication is independent
of the mill lubrication system.
Whizzer
Wheel
Safety and Controls
The production and handling of pulver-
ized fuels can be hazardous. Because
fine particles in suspension or deposi-
Figure 6-26 | Centrifugal material-handling exhauster for Alstom suction-type pulverizer
Auxiliary Equipment | 6-31
tion are readily volatilized and become inerting takes place. Alstom’s pulver-
combustible, under certain conditions izer inerting and firefighting system is
explosions may take place. Notwith- designed to provide an early warning
standing these dangers, the industry of a potentially hazardous situation.
has achieved a remarkable safety record It uses readily available plant steam as
since the inception of pulverized coal the inerting medium because steam is
firing. To protect property and life, all less damaging to equipment than other
pulverizing equipment and related aux- inerting media and makes for easier
iliaries, including strength of equipment, restart of the pulverizer. Plant water
valving, and inerting, are designed in is used as the fire-extinguishing agent.
accordance with NFPA 85 as detailed It is hard-piped to fixed water spray
in Chapter 9, “Pulverized Fuel Systems.” nozzles installed in the pulverizer. The
Various controls and safety devices purposes of such a system are:
are used for the correct and proper
• to dilute the oxygen content of the
operation of the equipment. Pul-
mill when there is risk of explosion;
verizer output is controlled by the
• to clear the pulverizer of fuel by
regulation of feed rate in response to
means of an inert medium when
a load signal. Airflow and air tempera-
transport by air may be hazard-
ture are proportioned to feed rate by
ous; and
automatic control. Also included are
• to extinguish fires in the pulverized-
permissive interlocks for the proper
fuel system.
sequential operation of equipment,
flow alarms to indicate cessation of
coal flow to and from feeders, and System Design Features
load-limiting devices to prevent The Alstom pulverizer inerting and
overfeeding the mills. Anticipating firefighting system includes the follow-
actions and more responsive feedback ing features:
may often be included with the above
in pulverizer control systems. Control • The pulverizer is automatically
systems and instrumentation are inerted when conditions exist for a
covered in detail in Chapter 7. fire or potential explosion.
• Steam is used as the primary inert-
Pulverizer Inerting and ing medium, with CO2, water mist
Fire Extinguishing or other cold inert gas used to cool
To supplement the above operational the pulverizer.
controls, current practice is to install • The inerting system is capable of
systems to detect and extinguish mill supplying steam in sufficient quan-
fires and to reduce the possibility of tities to transport the combustible
destructive positive pressures. An ideal contents of the pulverizer to the
inerting system continuously purges furnace, while maintaining an inert
any combustible volatiles from the atmosphere within the mill. As an
pulverizer to the furnace and avoids alternative, a water-flushing system
“bottling up” of the pulverizer. The can be used to clear the mill of coal
system allows a mill to be returned via the pyrite system while the mill
to normal service quickly after the remains inerted.
6-32 | Clean Combustion Technologies
• Multiple water-spray nozzles are also make pulverized coal firing practical
strategically installed in the pulver- by providing the drying and transport-
izer to provide complete internal ing medium.
fire-extinguishing coverage, Two principal types of air preheat-
• The system monitors the entire ers are in use: the rotary regenerative
pulverized fuel system from the (Ljungström® air preheater) and the
feeders through the fuel piping. tubular recuperative air preheater.
• Audible and visual alarms are The choice of the size and type of air
activated in all critical areas and in preheater depends on economic and
the control room upon detection of engineering factors. The economic
a hazardous condition. factors include the original equip-
• Provision is made for interfacing ment cost, the fuel cost, and the fan
with existing plant control systems. power cost resulting from air-heater
draft losses. The engineering factors
In Chapter 10 on power plant op- include the air temperature required
eration, further information is given on for combustion and/or pulverized
the efficient and proper operation of coal drying, as well as unit reliability
pulverizer inerting and fire-extinguish- and installation space requirements.
ing systems.
Regenerative Air Preheaters
AIR PREHEATERS The Ljungström® air preheater trans-
The functions of combustion air pre- fers to the combustion air sensible
heaters have been described in Chapters heat in the flue gas leaving the boiler;
2 and 3. Although justified by the it does this through a regenerative heat
increased efficiency resulting from lower transfer surface in a rotor that turns
exit-gas temperatures, air preheaters continuously through the gas and air
streams at from 1 to 3 rpm (depending
on diameter). The principle is illus-
trated in Figure 6-27.
Hot Air
Outlet Figure 6-27 also illustrates the major
Hot
Flue-Gas components of a large, vertical-shaft,
Inlet Ljungström® air preheater designed
for gas flow downward and airflow
upward. The rotor, packed with an effi-
cient heat-transfer surface, is supported
Heat-Transfer
Surfaces through a lower bearing at the cold
Rotation
Rotor Housing end of the air preheater and guided
through a guide-bearing assembly
located at the top or hot end.
Axial Seals
Depending on its size, the rotor has
Rotor Cool Air between 12 and 48 radial members,
Hot-End Element Inlet which are attached to a center post. The
Cooled
Flue-Gas Intermediate Element rotor compartments are closed with
Outlet seal plates, as shown in Figure 6-28. The
Cold-End Element
rotor sealing system contains simple
leaf-type labyrinth seals bolted to the
Figure 6-27 | Ljungström® air preheater, bisector vertical design
total air for combustion is called the the air preheater rotor, which consist
“primary air.” of one gas and one primary air flow
Historically, separate air preheat- as in the trisector, but two separate
Hot ers were used for the combustion air secondary air flows (Figure 6-30).
Secondary
Air (secondary air) and the primary air. The The design has the primary air sector
Hot trisector air preheater is now widely “flanked” on either side by second-
Gas
used to provide the higher-pressure ary air; the aim is to reduce the total
Cooled primary air and the lower-pressure air-to-gas leakage of the unit. This
Gas secondary air, but heated by a single air method of leakage reduction is ideal
Hot
Primary preheater (Fig. 6-29). The trisector is for applications in which the pressure
Air
equipped with an additional radial seal- differential between the air and gas
Secondary ing plate at both the hot end and cold sides is high, such as in CFB applica-
Air
end and an axial sealing plate separating tions. (See Chapter 3.)
the primary and secondary air streams.
Pulverizers Primary Air
The heater is readily adaptable to coal Heating Surface Design
types and moisture contents, by varying The heating surface of the Ljungström®
Figure 6-29
Clean| Trisector Ljungström®
Combustion Technologiesair the size of the primary air opening, with air preheater uses combinations of flat,
Figure: preheater
06.29/ (MH)with primary and
a 35-, 50-, or 72-degree primary air or formed, pressed-steel sheets with
PMS5415 secondary
K air system on a
pulverized coal boiler opening normally furnished. corrugated, notched, or undulated rib-
Some advantages of the trisector bing. When used in combination, these
over a separate primary and secondary form longitudinal passages of the most
air preheater system include: desirable contour for the predeter-
mined spacing. Although the gas and
• No gas-biasing dampers and controls.
airflow are turbulent, the smooth path
• Because only one gas duct is
for their travel through the rotor offers
required, the need for ductwork,
low resistance. As an approximate rule,
expansion joints, and insulation is
1 inch (2.54 cm) of height of this highly
greatly reduced when compared
efficient heating surface recovers about
with a separate air heating system.
as much heat as 2 feet (0.6 m) of surface
• Since the average cold-end tem-
in a tubular heater, with equivalent resis-
peratures can be maintained using
tance to gas and air flow. The compact
only secondary air, the requirement
Figure 6-30 | The quadsector Ljungström® air arrangement of light metal sheets per-
for air steam is eliminated.
preheater mits a large amount of effective heating
• Equipment layout is simplified.
surface to be placed in a relatively small
Less structural steel is needed to
rotor. The metal sheets, commonly called
install the system and less cleaning
“heating elements,” are typically packed
equipment is required.
in containers at the factory.
• Fewer soot blowers and water-
Clean Combustion Technologies Normally, the heating surface is
Figure: 06.30/ (GM)
washing devices are required with
divided into two or more layers.
PMS5415 K one heater in lieu of two.
Advantages resulting from layering
• Even with a large variation in
include the ability to vary the configu-
primary airflow, there is relatively
ration of surface profiles throughout
little effect on overall heat recovery,
the preheater depth (to satisfy differ-
because heat not recovered in the
ent flow conditions) and to vary the
primary section will be picked up
material selection of sheets in each
in the secondary section.
layer (to accommodate the tempera-
The quadsector air preheater is de- ture and other conditions particular
signed with four flow streams through to each zone).
Auxiliary Equipment | 6-35
There are a number of materials ash, forms a sticky deposit that is dif-
commonly used for the heating surface, ficult to remove with conventional soot
with mild steel surfaces being widely blowers. Frequent outages for water
used for hotter parts of the rotor and washing may be required. In addition,
low-alloy corrosion-resistant steel used during water washing, ABS becomes
for colder sections that are exposed very corrosive, which can shorten the
to corrosion. In addition, porcelain operational life of the heating ele-
enamel-coated heating surfaces are the ment and the cold-end rotor structure.
industry standard for cold-end sections Experience has shown that if the ABS
in more severe applications or when collects in a region that bridges two
prolonged element life is required. layers of heating elements, the rate of
accumulation and subsequent conse-
Rotors quences are accelerated. Significant
Large rotors have between 24 and cleaning energy is lost when the clean-
48 compartments extending axially ing medium for the soot blower has Hot End
along the full radius of the rotor. to cross an opening between element
Each compartment consists of two layers. Any remaining cleaning energy
side plates connected by a series of will have little effect on deposits in the
circumferential stay plates and/or adjacent layer.
grating. Horizontal rotor designs For maximum cleaning effectiveness,
have the rotor mounted between two the ABS deposit should be trapped in Fouling Zone
Cold End
spherical roller bearings, while the the single deep layer of heating element
vertical rotors have the rotor shaft profile that is designed to aid the clean-
supported by a spherical roller, or ing of the layer (preferably a “closed
tilting pad, thrust-bearing. Various ro- channel” design) (Figure 6-31). In ad- Figure 6-31 | Preferred rotor arrangement for
tor designs are available, ranging from dition, the use of a corrosion-resistant, operation after an SCR
fully welded arrangements to modular smooth, enamel-coated surface in the Clean Combustion Technologies
Figure: 06.31/ (BB)
designs that use a pinned connection ABS deposition zone inhibits deposit PMS5415 K
to the shaft to reduce or eliminate adhesion and improves the cleaning
field welding. effectiveness. The use of specially
designed cleaning devices is also typi-
Designing for Operation after an SCR cally investigated to provide the most
Environmental regulations designed to cleanable solution.
reduce NOx emissions have resulted in
fuel changes, burner upgrades, and new Air Preheater Leakage
equipment and chemical process instal- The air-to-gas leakage in an air pre-
lations, including selective catalytic heater must be taken into account in
reduction (SCR) and selective non- the design and selection of associated
catalytic reduction (SNCR) systems fans. The leakage is composed of two
(Chapter 5). The operation of these main components:
systems generally results in some am-
monium bisulfate (ABS) deposits, as • “Direct” leakage is the air that
excess ammonia combines with sulfur passes into the gas stream
trioxide and water. The ABS deposits between the radial and axial seals
collect within the air preheater on and sealing surfaces as a result
metal surfaces within a temperature of the static pressure differential
range generally between 300 and between the air and gas streams.
375°F. ABS, in combination with fly The leakage across the sealing
6-36 | Clean Combustion Technologies
are most often located at the gas-outlet sometimes becomes necessary to water
side to prevent fly ash from being carried wash the heating surface to maintain
into the wind boxes. They are installed acceptable draft losses through the air
either as an integral part of the air preheater. In some instances if the air
preheater duct or in the gas-outlet duct preheater is subjected to factors that
work immediately adjacent to the unit. create serious fouling problems, this
Additional combined “multimedia” may be required more frequently than
cleaning devices are also available. during the scheduled boiler outages.
These devices operate as normal Most deposits forming on the air pre-
retractable soot blowers during boiler heater heat-transfer surface are highly
operation. They have the capabil- soluble in water, and can be easily
ity to quickly switch to operate with removed by washing with a sufficient
medium- or high-pressure water during quantity of it. A high-penetration,
boiler outages to provide water wash- low-pressure, stationary, multi-nozzle
ing capability. device is the standard washing appara-
Superheated steam (approximately tus and is available for all air preheater
300°F [150°C] superheat) or dry com- types and sizes. Adequate drainage is
pressed air is the recommended cleaning necessary before planning to wash an
medium. Although compressed air is air preheater. Washing can be on either
considered to be the premium cleaning the air or the gas side, depending on
medium, its merits do not stem from which has the best drainage.
any inherent cleaning ability, but from Washing is generally performed
its dryness as compared with steam. A during out-of-service periods, when
steam source must be selected to have the boiler has been shut down. During
pressure and temperature conditions shutdown is the best time to control
that, by proper control measures, may the washing operation and to make
be used to provide dry steam to the air a thorough inspection of the heating
preheater soot blower. surface, both during and after washing.
Soot-blowing pressures are set by The speed of the air-preheater rotor
the equipment suppliers for each type is typically reduced by means of an
of heat-transfer surface (typically in the auxiliary drive before admitting water
range of 100 to 250 psig [6.8 bar to 17 to the washing devices. The slower
bar]). A steam soot blower piping sys- speed permits the wash water to drain
tem should include an automatic drain from the rotor and heating surface
valve, thermocouples, and an automatic before the wash water enters the other
admission valve to the blowers. The stream. To ensure coverage of all heat-
automatic drain valve is open to free ing surfaces when using stationary,
drain discharge until the temperature- multi-nozzle devices, the washing cycle
sensing thermocouple indicates steam of must consist of at least one revolution
adequate quality. When using air, care of the preheater rotor.
should be taken to install a proper line An intermediate- or high-pressure,
of traps and separators to remove mois- low-volume wash is sometimes neces-
ture from the blowing medium. sary to remove particles of insoluble
material that are carried from the boiler
Water Washing by the gas stream. The larger particles
In cases in which soot blowing cannot can become wedged in the flow passages
readily remove residual deposits, it of the hot-end, heating surface. These
Auxiliary Equipment | 6-39
obstructed passages then fill up with fly minimize but not eliminate this risk.
ash, which restricts flow. This type of de- Early warning is key to minimizing
posit accumulation is difficult to remove the damage. Reliable indication of a
with low-pressure, multi-nozzle devices. fire is complicated by several process
Special mobile equipment, using one variables. First, localized hot spots in
high-energy jet or more, has been devel- the heat-transfer surface are moving
oped for this application. Some newer targets as the air preheater rotates.
designs of soot-blowing devices have a Second, the detection system must be
high-pressure washing function built in capable of indicating trouble during
to provide water washing to the plant at various boiler load conditions, from
all times, and shorten the required setup transient start-up to full-load, steady-
time. The procedure with these devices state conditions.
is similar to that followed in a regular Various designs and configurations
out-of-service washing using multi- of detection systems are available.
nozzle washing devices, except that the The most common systems use either The air heater not only heats
water nozzles are moved radially across thermocouple- or infrared-based moni- the air, but cools the flue gas,
the surface, allowing sufficient time toring. The systems are typically not
which improves efficiency and
for the rotor to complete at least one intended to automatically initiate fire-
revolution between each movement of fighting. Rather, they are designed to reduces emissions.
the nozzle of the cleaning device. Heat- signal an alarm to the plant operators,
transfer surfaces should be examined which allows further investigation and
frequently during any washing process. necessary corrective action in advance
After the deposits are removed, the unit of a major problem.
should be allowed to dry completely
before being returned to service. Air Preheater Temperature Terminology
Table 6-10 lists the most common
Air Preheater Fires and terms used when discussing the ther-
Detection Systems mal performance of Ljungström® air
While not a common event, fires do preheaters.
pose a significant threat to air pre-
heaters installed on fossil fuel fired Factors Affecting Air Preheater
boilers and furnaces. The origin of Performance
an air preheater fire can generally be There are many factors that combine to
traced to combustible deposits on the influence the operating performance of
heat-transfer surface. In its initial stage, an air preheater as shown in Table 6-11.
the fire is limited to the combustion of
these deposits and often confined to a Designing for Low Exit
relatively small area. With time, how- Gas Temperature
ever, this fire will progress to a metal Combustion of the sulfur in the coal
fire, with an associated rise in tempera- results in the formation of sulfur
ture into the 4,000 to 5,000°F (2,200 dioxide. About 1 to 4% of this sulfur
to 2,800°C) range. At this point, the is further oxidized to sulfur trioxide,
fire becomes far more damaging to the depending on the flue-gas constitu-
equipment and becomes significantly ents, temperature and catalytic activity
challenging to extinguish. within the boiler, and the presence of
Appropriate unit operation, an SCR. The SO2 and SO3 may then
coupled with routine maintenance, can combine with moisture in the flue gas
6-40 | Clean Combustion Technologies
Table 6-10 | Common Terms Used for Air Preheater Performance to form sulfurous and sulfuric acids.
Sulfurous acid will not form above
Exit-gas temperature with
This is the observed or measured average exit-gas the water dew-point temperature and
leakage (also termed
temperature and includes the dilution effect of leakage.
“corrected”) is seldom a problem. However, sulfur
This is the average temperature at which the gas would trioxide is hygroscopic, and will absorb
Exit-gas temperature with no leave the preheater if there were no leakage. This moisture at temperatures well above
leakage (also termed temperature cannot be measured directly, but it can the water dew point, resulting in the
“uncorrected”) be arrived at by a simple calculation accounting for the
cooling effect of the leakage. formation of sulfuric acid vapor. The
temperature at which this acid vapor
Difference between the gas entering temperature and
Gas drop condenses to form liquid sulfuric acid
the uncorrected exit gas temperature.
Difference between the average air leaving temperature is called the “acid dew point.”
Air rise While boiler efficiency can be
and the air entering temperature.
Difference between the gas entering temperature and improved by adding surface to reduce
Temperature head
the air entering temperature. the air preheater exit-gas tempera-
Gas Drop ture, this practice lowers the cold-end
Gas side effectiveness, % × 100 metal temperatures, possibly below
Temperature Head
the acid dew point. Consequently, steel
construction materials are subject to
corrosion from the sulfuric acid in the
Table 6-11 | Factors Affecting Air Heater Performance flue gas.
Changes in entering gas or air temperatures result in a change in the The acid dew point varies with the
Entering Gas
temperature head, which directly affects the drop in gas temperature, concentration of sulfur trioxide in the
Temperature
resulting in the exit-gas temperature changing in the same direction.
flue gas. High-sulfur coals result in the
An increase in gas mass flow rate to the heater will result in a higher existence of a dew point at a higher
Gas Flow exit-gas temperature, while conversely a lower exit-gas temperature
results from a lower gas mass flow entering the unit.
temperature, thus exposing more of
the air heater surface to corrosion and
Also called the “X” ratio, the heat-capacity ratio (HCR) =
fouling. Typical acid dew points for
mass flow of air × avg. specificheat of air
coal firing are between 280 and 320°F
mass flow of gas × avg. specificheat of gas
Heat-Capacity (135 and 160°C).
A decrease in the HCR results in an increase in the exit-gas temperature.
Ratio (HCR) For coal firing, the fouling potential
A 10 to 12% change in this ratio may alter exit-gas temperature by as
much as 30 to 35°F (15 to 20°C). Factors that affect the HCR include increases as the exit temperature of
tempering air, overall boiler-system infiltration, and air preheater the gas decreases. Water condensation
bypass for cold-end protection.
causes a marked increase in dew-point
The air and gas side pressure drops will change approximately in meter response at 140°F (60°C) or be-
proportion to the square of the air and gas weights through the heater.
If excess air is greater than anticipated, the air preheater pressure drop low. While operating at temperatures
will be greater than expected. A buildup of heating-element deposits below the water dew point would likely
Pressure Drop
will result in higher air preheater resistances, with the consequence lead to corrosion issues for unprotected
of an increase in pressure drop. Pressure drop will also vary directly
steels, the major concern is the rapid
with the mean absolute temperatures of the fluids passing through the
heater as a result of changes in flow density. fouling that would likely take place.
A change in the temperature of the fluid leaking past the seals will, by With low-sulfur coal, the corrosion
reason of density change, have a slight effect on the amount of leakage. and fouling potential is reduced and
Variations in pressure levels between the high- and low-pressure sides restricted to the extreme cold end of
Air Preheater of the heater will, likewise, alter the air preheater leakage. An increase
Leakage in the pressure differential will increase the leakage, while a decrease
the air preheater.
in pressure differential will reduce the leakage. Improper settings of the A number of means have been devel-
heater radial and circumferential seals will also result in an increase in oped to minimize the rate of corrosion,
leakage. as well as to provide for replacement of
Auxiliary Equipment | 6-41
corroded surface. Because corrosion oc- Gas–Gas Heater for FGD Reheat
curs on the lowest-temperature surface, Applications
air-heater designs have been developed The Ljungström® gas–gas heater
that incorporate replaceable cold-end (GGH) is used in applications in
sections. Other means to minimize cor- which heat can be recovered from the
rosion are aimed at increasing the metal flue gas in front of the flue-gas desul-
temperature. One such arrangement furization equipment and then used
directs a portion of the preheated air to reheat the flue gas. The GGH’s
to the inlet of the forced draft fan and main purpose is to lower the flue-gas
recirculates it through the air preheater. temperature to a level at which the
Thus, the temperature of air leaving flue-gas scrubbing equipment can
the fan and entering the air preheater is function and then provide buoyancy
increased, correspondingly increasing to the flue gas for its path up the
the cold-end metal temperature. stack. The GGH is a regenerative
Air bypass around the air preheater is heater with many of the same design
used to a limited extent. With reduced features and principles as the Ljung-
airflow, metal temperatures within the ström® air preheater.
air preheater are higher because of the One major difference of the FGD
influence of the higher gas-to-air ratio. GGH is a sealing system that is ca-
Also, because the overall heat recovery pable of lowering the leakage to as little
is less as a result of the reduced airflow, as 0.5%. The position of the induced
gas-outlet temperature rises, causing a draft, forced draft, and booster fans
rise in the cold-end metal temperature. determines the type of sealing system
The most common means of required and the achievable level of
increasing the cold-end metal tempera- undesirable untreated-to-treated gas
ture is the use of steam air preheated leakage, as a result of the pressure
coils located in the cold-air duct be- differential. Leakage of untreated gas
tween the FD fan and the air preheater. reduces the overall collection efficiency
These coils increase the temperature of the FGD system.
of the air entering the heater, cor- Various fan-assisted sealing sys-
respondingly causing an increase in tems are available to minimize the
the metal temperature. Steam bled entrained-leakage and the direct-
from the turbine is used as the heating leakage components. Entrained leakage
medium in the steam-coil air preheater. is minimized by blowing treated,
In supplying heat to the cold air, steam clean gas into the rotor as it passes
is condensed and the condensate from the untreated to the treated side.
returned to the appropriate stage of the Direct, radial leakage is minimized by
feedwater heating system. pressurizing the plenum between the
To assist designers and operators sector plates and the rotor with similar
in arriving at reasonable cold-end treated clean gas.
temperatures, average cold-end tem- The FGD GGH design raises signifi-
perature (ACET) recommendations cant concerns as a result of the very low
are prepared by the equipment suppli- exit gas temperatures. Corrosion is a ma-
ers, considering variables such as fuel jor consideration. The material selection
type, sulfur content, excess air level, throughout the heater is tailored to suit
and material selection. the operating conditions. For example,
6-42 | Clean Combustion Technologies
special design attention is given to the than a regenerative air preheater when
use of a suitable corrosion-resistant considering higher system pressures.
lining on the interior of the housing. The heater would typically be arranged
Special seal materials such as stainless for vertical gas flow through the tubes
steel and lined sealing plates are used. and horizontal air flow across the out-
Substantial corrosion allowances on side of the tubes usually 2 to 3 inches
the rotor plate work and enamel-coated (5 cm to 7.6 cm) in diameter. The tubes
heating elements are all evaluated. are in a staggered relationship for opti-
mal heat transfer. The air passes over all
Gas–Gas Heater for SCR the sections of the air preheater in se-
Reheat Applications quence, the effect of which is to provide
The SCR GGH is used in “tail-end,” counterflow heat transfer. Tube sheets at
or “low-dust,” flue gas sulfurization the top and bottom support and guide
(SCR) systems to assist in raising the the tubes. Most frequently, the bottom
flue-gas temperature up to the SCR tube sheet or sheets form the struc-
reaction temperature. The position tural support. The upper tube sheet is
of these SCR systems minimizes the welded to the outside casing, and the
effect that flue-gas contaminants have tubes pass through slightly oversized
on the catalyst design and operation, holes in the upper sheet, which allows
with the SCR and heater operating in for expansion when the equipment
a low-dust environment. is brought up to temperature. Many
The SCR GGH is designed with designs for use with sulfur-bearing
special features to reduce the negative fuels have separated cold-end sections,
impacts from the SCR related to SO2 to reduce the cost of tube replacement
oxidation, acid condensation, ammonia in the event of excessive corrosion.
slip, and formation of ammonia salts. In contrast to regenerative designs,
The potential for ABS fouling, together tubular, or recuperative, air preheaters
with the lower flue-gas ash concentra- have more severe cold-end corrosion
tion, means that there is a reduced problems. With variations in cleanliness
surface area of ash to adsorb the ABS of the tube wall, entering air tempera-
and transport it through the GGH. To ture, and flow intensities on the gas and
minimize the effects of ABS-enhanced air sides, very low metal temperatures
fouling, the GGH typically incorporates and correspondingly severe corro-
many of the design features of an air sion can occur. Although the tube
heater operating after an SCR. Similarly metal temperature may be considered
to the FGD GGH, the sealing arrange- to be the arithmetic average of the air
ment of the SCR GGH can incorporate temperature entering the air preheater
fan-assisted sealing systems to minimize and the gas temperature leaving, actual
the leakage of untreated to treated gas. field measurements have shown metal
temperatures as low as 120°F (49°C)
Tubular Air Preheaters below the mean of the air and gas tem-
A typical application of a tubular air peratures. Such conditions can result
heater might be an industrial boiler for in deposits that reduce heat transfer
which perhaps the tubular heater would and increase draft loss until they are
generally provide lower leakage rates removed, usually by water washing.
Auxiliary Equipment | 6-43
vertical velocity of the gases in a furnace Peak ash production rates (based on
Diameter of Particle, Microns (4 �m)
0 50 100 150 200 250 300
is 10 ft/sec (3 m/sec) or less. 100% capacity factor) are used to size
In this regard, a situation occurs ash-handling and transport systems.
10
in the design and operation of large, On the other hand, the average rate of
8
pulverized fuel furnaces that is of in- ash production, which considers the
Velocity, ft/sec
6 terest. As furnace plan areas increase projected plant capacity factor over
4 for a given quantity of net heat input the operating life of the plant, is used
2
to accommodate a very poor fuel for to size disposal sites.
example, the upward gas velocities
0
0 2 4 6 8 10 12 correspondingly decrease. At the same Fuel Source and Ash Content
Diameter of Particle, Thousandths of an Inch
time, the furnace wall area becomes A major factor affecting the choice of
325 200 140 100 60 50 greater, leading to the possibility of an ash-handling system is the type of
Equivelant in U.S. Mesh
larger accumulations of slag or ash fuel to be fired. Figure 6-36 illustrates
on the walls for a given heat input. the wide variation in ash “fired” into
Figure 6-35 | Velocity of airborne particles
(specific gravity 1.2) falling The greater ash deposits (and lesser a furnace as a function of fuel ash
Clean through
Combustion Technologies
still air at 70°F (21°C) upward velocities that result) mean content and moisture. For example, a
Figure: 06.35/ SPS
PMS5415 K that more ash will fall to the furnace boiler firing 15% ash sub-bituminous
bottom. Typically, the ratio of bottom coal generates almost three times
Lignite (40% moisture)
ash to fly ash increases with these the total ash as the same-size unit
15 Sub-Bituminous poorer fuels, which impacts the cost burning 10% ash, high-calorific value,
(low-volatile bituminous with 10% ash = 1.00)
the gas stream with directional changes The economizer, SCR, and air
in the gas flow. The recent addition of heater ash hoppers each are sized to
SCR systems for the removal of NOx collect about 5% of the ash, while a
has added another set of hoppers in precipitator, fabric filter, or other col-
this area, generally just downstream lection device would be sized to extract
of the economizer hoppers. The 80 to nearly 100% of the fly ash from
catalyst itself must be kept clear of the flue gas.
ash. The design of the ash-handling Note that the sum of all the
systems ahead of the SCR have assumed collection-point values can ex-
become more critical and the use of ceed the total ash in the coal by 50% or
continuous-removal systems more more because of the lack of precision
common. Fly ash hoppers included of the individual factors. Variations in
under the precipitators or baghouses fuel quality must also be considered.
collect the material extracted from the
flue gas stream by these particulate Ash Density
removal devices. Fly ash density depends primarily on
particle size, particle structure, and
Quantities of Ash carbon content. In general, the rela-
Most pulverized coal-fired steam tively large, coarse particles containing
generators are dry-bottom units in a high percentage of carbon have a
which 20 to 40% of the ash from the low density. These particles commonly
coal falls to the furnace bottom. Sixty have a porous, or lace-like, structure
to 80% of the ash leaves the furnace and result from the partial combus-
as fly ash. Some ash is collected in the tion of the pulverized coal. It appears
economizer/SCR/air heater hoppers that the volatile portion burns out,
and the rest in one of several available leaving black coke-like particles. These
dust-collection systems. particles have densities of about 35 to
However, with some fuels less than 65 lb/ft3 (560 to 1,040 kg/m3) (true
20% bottom ash has been measured, density). The finer ash particles, which
while with certain lignites, more than tend to be low in carbon content, have
45% bottom ash has been reported. As much higher densities, usually ranging
a guide to establishing the size of the from 90 to 180 lb/ft3(1,440 to 2,880
equipment in the total ash-handling kg/m3) (true density).
system, typical collection quantities In the hot state, freshly collected ash is
(mass basis) are used in the early, normally very fluid, with a lower density
budgetary phases of the design. A true than cold ash—generally between 30
ash analysis of the proposed fuel is and 50 lb/ft3 (480 to 800 kg/m3) bulk
preferred before final system selection. density. The fresh ash is aerated as the
For example, the 20 to 40% figure can individual particles are exposed to the
vary according to experience with the gas. With standing and cooling, de-
type of fuel burned. With low- and aeration occurs and the ash compacts. It
medium-volatility bituminous coals, becomes less fluid and can approach bulk
15 to 20% bottom ash is appropriate. densities as high as 90 lb/ft3 (1,440 kg/
With a high-volatility bituminous coal, m3). When cold, sub-bituminous coal
20 to 30% bottom ash is used. Systems ash, which is very fine, will usually have
for sub-bituminous coal and lignite are loosely packed bulk densities between
sized using 30 to 50% bottom ash. 65 and 85 lb/ft3 (960 and 1,360 kg/m3).
Auxiliary Equipment | 6-47
Bituminous coal ash, on the same basis, Ash-handling systems are further
will have a density between 60 and 70 differentiated by their conveying
lb/ft3 (960 and 1,120 kg/m3). frequency and conveying medium. The
The bulk density of dry bottom ash, major systems are:
as initially dropped into an ash receiver,
can be as low as 40 to 45 lb/ft3 (640 1. Bottom ash
to 720 kg/m3), but it can compact to a. Water-impounded hopper
above 75 lb/ft3 (1,200 kg/m3). A value with wet pipeline or sluiceway,
of 45 to 55 lb/ft3 (720 and 880 kg/m3) to a pond or closed recirculation
bulk density can be used for establish- system—intermittent, hydraulic
ing bottom-ash hopper or conveyor b. Dry hopper, dry pipeline—
storage capacity. Pulverizer rejects, when intermittent, pneumatic
principally iron pyrites (iron disulfide, c. Submerged scraper conveyor—
FeS2), will have a bulk density of at least continuous, mechanical
135 lb/ft3 (2,160 kg/m3). 2. Pyrites (pulverizer rejects)
a. Hoppers with wet sluicing
TYPES OF SYSTEMS (pipeline or open trough)—
The three principal methods of moving intermittent, hydraulic
ash from the collection points to the b. Hoppers with manual removal—
on-site disposal or storage location are intermittent, mechanical
hydraulic, pneumatic, and mechani- c. Hoppers with dry vacuum
cal. In a hydraulic system, a stream removal (infrequently used)—
of water conveys the ash in a closed intermittent, pneumatic
pipeline or open sluiceway. Commonly, d. Hoppers with drag conveyor
a jet or centrifugal pump provides the removal—continuous, mechanical
motive force. A pneumatic system (rare but possible)
transports the ash in a stream of air or 3. Economizer ash
flue gas induced by either an upstream a. Dry pipeline—intermittent,
pressure or a downstream vacuum. pneumatic
The oldest technique for ash removal b. Dry drag or screw conveyor—
is the mechanical method, the most continuous, mechanical
rudimentary form being a shovel or c. Wet pipeline using receiv-
hoe to rake ash out of the boiler, with ing tanks below economizer
the more sophisticated type being sub- hoppers—continuous/batch,
merged or dry scraper conveyors using hydraulic
chain and flight bars. d. Dry pipeline using receiv-
Two basic philosophies have evolved ing tanks below economizer
throughout the world for removing hoppers—continuous/batch,
ash from pulverized coal-fired steam pneumatic
generators: 4. Fly ash
a. Dry pipeline, vacuum—
• Collection, storage, and periodic intermittent, pneumatic
removal, usually called “intermit- b. Dry pipeline, pressure—
tent” or “batch removal systems” intermittent, pneumatic
• Collection and continuous removal, c. Dry pipeline, vacuum-pressure
commonly termed “continuous- combination systems—intermit-
removal systems” tent, pneumatic
6-48 | Clean Combustion Technologies
units above 400,000 lb (180,000 kg) Depending on the fuel ash content and TABLE 6-13 | freeboard
of steam/hr, intermittent-removal the boiler load, the usual time required water levels*
bottom-ash hoppers use the V-, W-, to evacuate a bottom-ash hopper is 1 to
Effective Ash
or triple-V-shaped “pant leg” design. 3 hours. The hopper volumetric capac- Hopper Width
Level (EAL) Below
Inside Refractory
This design has hopper floors sloping ity is measured below a mean ash level, NWL, ft
at angles of 35 to 55 degrees, allow- which is usually 12 to 24 inches (30 to 7´ 0˝ 0.8
ing gravity to help remove ash from 60 cm) below the normal water level, 8´ 0˝ 1.0
the hopper. During ash removal the depending on the bottom-ash hopper
9´ 0˝ 1.2
high-pressure water to the jet pump width (Table 6-13 and Fig. 6-40).
10´ 0˝ 1.4
is turned on and the clinker grinder is It is important to realize that the
started. The control system opens the ash falling from the furnace walls does 11´ 0˝ 1.6
hydraulic-cylinder-controlled sluice not pass through the furnace bottom 12´ 0˝ 1.8
gate, and the ash flows from the hopper “coutant” opening in equal increments 13´ 0˝ 2.0
through the sluice gate and grinder along the length of that opening. Figure 14´ 0˝ 2.2
into the jet pump, then to a transfer 6-41 shows that, for a large “open” fur-
* An interpolation can be made for other hopper widths.
tank, dewatering system, or settling nace (one without a center wall), all of
pond. Alternatively, a centrifugal the ash from each sidewall falls through
materials-handling pump evacuates the the first 3 or 4 feet (0.9 or 1.2 m) of the
hopper and transports the ash to the bottom opening. The ash falling from Water-Cooled
Furnace Bottom
disposal system. the front and rear walls can be as-
Operating experience with water- sumed to distribute fairly equally along
filled hoppers has shown that most of the furnace width. Thus, as shown by A Normal Water
the ash will flow toward and through the the figure, about 25% of the total ash Effective Ash
Level (NWL)
Level (EAL)
sluice gates by gravity. However, large to the furnace bottom will go through
slabs of fused ash do occur that will not the 3- to 4-foot portion near the side-
move by gravity alone. Their removal wall, and less than 2% of the bottom
requires the assistance of high-pressure ash will go through each remaining Ash Hopper Width
(125 psig [8.6 bar]) water-jetting nozzles foot of the bottom “slot.” It is of utmost (inside refractory)
located on the sloping floors of the hop- importance to take this distribution into
per. In addition, jetting nozzles in the account when designing equipment that
Figure 6-40 | Effective ash level below normal
wall opposite each sluice gate break up receives bottom ash. The normal water Clean Combustion Technologies
water level varies as a function of
Figure: 06.40/ SPS
the arching and packing of the ash that level should be at least 30 inches (76 ash hopper width (see Table 6-13)
PMS5415 K
may occur at the gate. cm) below the horizontal plane of the
Whenever practical, the hopper furnace tubes forming the furnace throat
should have a volume corresponding opening (“A” in Figs. 6-40 and 6-42). This
to the ash produced during 12 hours dimension is with the boiler pressure Side Wall
of operation at full boiler load while parts in the hot (expanded) condition. Rear Wall
firing the fuel with the lowest heat- This distance minimizes the amount of Front Wall
ing value and highest ash content for water that may splash on the furnace
which the boiler is designed. This tubing. The hopper is designed with 5 to
guideline is based on the normal prac- 10 foot (1.5 to 3 m) end-wall submer-
tice in the United States of emptying gence to provide a deep pool of water at 1-2% of
ash hoppers once each 8-hour shift. this critical portion of the hopper. Total Ash/Ft
of Furnace
The extra 4 hours allows time to deal Depending on the boiler width and Hopper
Bottom 20-25% of Total Ash
with any problem encountered when headroom available, the hopper con- at Each End
first starting up the system each shift. figuration should adequately provide
Figure 6-41 | Approximate distribution of
pulverized coal bottom ash along
the furnace hopper opening
Clean Combustion Technologies
Figure: 06-41/ RC
PMS5415 K
6-50 | Clean Combustion Technologies
the needed storage capacity. Center- ash hopper, together with the seal
Sidewall Furnace Normal wall boilers must have an odd number plates hung from the boiler tubes,
Submergence Waterwall Water Level
of sections. This places the center of a provides a seal against furnace pressure
A hopper section underneath the center fluctuations and accommodates the
wall so that there is maximum submer- downward expansion of the furnace.
Hopper gence to collect the ash properly from Nozzles located throughout the seal-
Slope both sides of the wall. trough periodically flush accumulated
(inside)
Hopper Slope CL For ease of maintenance, the ash sludge and sediment from the trough
(outside) Furnace and Ash Hopper hopper should be designed so that the bottom.
entire hopper and its discharge equip-
Figure 6-42 | Schematic arrangement of half
ment are located above grade. Pits REFRACTORY COOLING. At the top of the
of a triple-VTechnologies
Clean Combustion water-impounded
Figure:bottom-ash
06.42/ SPShopper. Dimension should be avoided. If they are used to hopper is a water distribution assembly
PMS5415 A is defined
K with the boiler in put the sluice piping below the level of for cooling the hopper refractory that
hot/expanded condition the basement floor, their depth should extends above the normal water level
be limited to a maximum of 3 feet (Fig. 6-44). The curtain of water flow-
High Water Level (0.9 m). These pits should have ad- ing over the refractory face also retards
Normal equate drains and sump pumps to avoid refractory deterioration during extend-
Clean Water Level
Out
flooding and submerging the grinders. ed periods when the hopper is drained
Plug of water during boiler operation.
OVERFLOW WEIRS. To maintain the nor- The water-distribution pipe must
mal water level (NWL) in the hopper, be equipped with a deflection shield
an NWL overflow should be provided. to prevent water from being sprayed
The bottom opening of the weir is at on the furnace tubes. Periodic flushing
Skimmer the normal water level (Fig. 6-43). To of this pipe with a high-volume water
Baffle protect from overfilling the hopper and flow through a drain connection in
possibly wetting the boiler tubes, there each section ensures that pieces of ash
Overflow is a second, independent “high-water- do not clog the holes.
Box
level” overflow. The opening at the Figure 6-45, a flow diagram for a
Hopper Wall
bottom of this weir is a minimum of 6 single-V bottom-ash hopper, shows
Weir Box inches (15 cm) above the NWL weir. all of the water flows entering and
A conductivity device in the outlet box leaving this hopper. Optimal operation
from the high-level overflow provides of a water-impounded hopper can be
Figure 6-43 | Arrangement of bottom ash an alarm when high overflow occurs. achieved by maintaining the water level
hopper overflow weirs and at all times, except for short periods
piping
Clean Combustion Technologies ASH HOPPER CONSTRUCTION. Ash hop- while discharging the ash from each sec-
Figure: 06-43/ RC
PMS5415 K pers are constructed of carbon steel tion. In this way, the design makes use
plate with structural framing similar of all the advantages of water impound-
to that of the furnace walls. A high- ing. The only other time the water level
temperature, water-resistant, gunnite has to be lowered with this arrangement
or cast refractory, supported by is when it is necessary to visually inspect
stainless steel anchors, lines the sides the hopper interior or to remove over-
and bottom slopes. Alternatively, on sized pieces manually. Each end of the
the sloped bottom portions of the hopper should be equipped with access
hopper, acid-resistant bricks backed by doors for maintenance.
concrete can be used.
A separate water seal-trough HOPPER SLUICE GATES. Bottom-ash hop-
around the upper periphery of the per sluice gates (Fig. 6-46) (usually
Auxiliary Equipment | 6-51
CLINKER GRINDERS. Single- or double- Figure 6-45 | Flow diagram for a single-V bottom-ash hopper
roll clinker grinders are provided at Clean Combustion Technologies
Figure: 06.45/ BAM
the outlet of each sluice gate. The the wear-prone internal parts of these
PMS5415 K
grinding elements are made from crushing machines. It is important to
heavy-duty, abrasion-resistant materi- provide easy access for their removal
Operating
als, such as Ni Hard, work-hardened while the steam generator is operat- Cylinder
manganese steel, or high-chrome ing. The designs shown in Figures 6-47
Ash and
alloy steel. Ash is ground by a shear- and 6-48 provide an easily removed Water
ing, splitting, or crushing action to front panel for such access, allowing from
Hopper
facilitate hydraulic or pneumatic maintenance personnel to replace worn Wear
transport (Figs. 6-47 and 6-48). Typi- grinding elements without entering Liners
Gate
cal particle sizes for crushed ash are the enclosure. Shaft bearings are of the Adjustment
approximately 2 to 3 inches (5.1 to spherical roller type, with provisions for
Gate
7.6 cm) in diameter. forced grease lubrication. They should Gate
Because of the severe duty, relatively be positioned away from the grinder Frame Renewable Gate Seat
frequent replacement is required for housing and packing gland areas.
Figure 6-46 | Bottom ash hopper sluice gates
Shaft seals have continuous grease or duplicate, backup, sluice gate, grinder,
Hammer Removable seal-water flow requirements during and pump for each hopper section
Segment Access Panel
periods of both grinder operation and should be considered (or the alternate
End Wear Plate Plain Segment shutdown. The selection of the proper use of a submerged scraper conveyor).
wear rings, lantern rings, and pack- The purpose of this redundant equip-
ing materials for optimal shaft sealing ment is twofold: first, to increase the
is an ever-changing science based on overall system reliability and, second,
hydraulic head and water quality. to increase the rate of ash removal
when quicker-than-design evacuation
Anvil JET PUMPS. A jet pump is a Venturi- is desired.
type solids/fluids transport device
Housing Rake with no moving parts (Figs. 6-49 and PIPING, FITTING, AND VALVES.Ash slurry
6-50). The pump has an inherently piping is supplied in several materials,
Figure 6-47 | Single-roll clinker grinder low efficiency, but its non-overloading including:
for reducing size of bottom- characteristic, ease of maintenance, and
ash; hammers
Clean Combustion and anvils are
Technologies • Centrifugally cast chrome–iron
Figure:replaceable
06-47/JT through removable lack of rotating parts offset this draw-
alloy with a hardness range of 280
PMS5415 K panel
access back. For these reasons, the majority
to 340 BHN in a standard alloy
of hydraulic bottom-ash and espe-
and a range of 475 to 550 BHN in
cially pulverizer-rejects systems use jet
a harder available alloy.
pumps instead of centrifugal pumps
• Internal surface-hardened carbon-
when water is readily available.
steel piping, referred to as induc-
To minimize the effects of abrasion,
tion-hardened pipe (especially
the jet pump is made of abrasion-
good for short pyrites lines or fly
resistant alloys. The combining tube
ash)
(diffuser) can be provided in high-
• Basalt-lined steel pipe
hardness, cast, chrome–nickel alloy
• Ceramic lined steel pipe
material with a minimum Brinell
Figure 6-48 | Advanced design of single-roll hardness number (BHN) of 550, or Ash in a water slurry usually travels
Clean clinker grinder
Combustion showing extended
Technologies an even higher chrome alloy with a along the bottom portion of pipes
Figure:bearing
06.48/supports
TB for better
PMS5415accessK to packing gland area minimum BHN of 700. in saltation flow. That is, it alter-
A jet pump is supplied at the outlet nately ‘‘rests’’ (settles out) and “jumps”
of each clinker grinder. Alternatively, (becomes re-entrained in suspended
a mechanical, centrifugal pump can be flow). Thus, most wear on slurry pip-
used. The advantages of each are listed ing is on the bottom third (the lower
in the box on the next page. 120-degree arc) of the pipe. Directly
As more high-ash coal is used for downstream of elbows, piping wear is
steam-generator fuel, producing greater higher than in straight runs because of
Ash and ash loading on bottom-ash systems, a the turbulence and impact created at
Water from the elbow. This calls for more frequent
Hopper
High Ash and replacement of downstream runs.
Pressure Nozzle Water Whenever there is a change in direc-
Water Diffuser Discharge
Input tion in the slurry piping, the fitting
located at the change of direction will
undergo accelerated wear as a result of
Nozzle Holder Pump Body turbulence and solids-impact erosion.
Thus, fittings, such as 45- and 90-degree
Figure 6-49 | Schematic arrangement of a jet Figure 6-50 | Jet pump for bottom-ash
pump for handling abrasive solids transport
Clean Combustion Technologies Clean Combustion Technologies
Figure: 06-49/JT Figure: 06.50/ TB
PMS5415 K PMS5415 K
Auxiliary Equipment | 6-53
elbows and laterals, are normally con- COMPARISON OF JET PUMPS VS. CENTRIFUGAL PUMPS
structed from alloy castings that range
from 400 to 550 BHN (Fig. 6-51). ADVANTAGES OF JET PUMPS ADVANTAGES OF A CENTRIFUGAL PUMP
White irons (high chrome) and Ni- • A jet pump is a simple device. There are • A centrifugal pump can obtain much
Hard casting materials are used with no moving parts; it is made up of only higher volumetric capacities than a jet
three principal components: nozzle, body, pump. A typical jet pump has a practical
the cast pipe above, while matching and diffuser. This makes maintenance and limit of about 3,500 gpm discharge flow.
basalt and ceramic fittings are used replacement easier and less expensive Centrifugal pumps can produce flows
with their respective pipeline mate- than for a centrifugal pump. of 35,000 gpm and higher. Larger MW
rial. Material hardness alone cannot • A jet pump can handle air without any plants with higher ash generation and
provide long life at points of change deleterious effects; lack of suction head is conveying rates and with more distant
never a problem. disposal areas will tend to use centrifugal
in direction. To add additional life
• A jet pump is self-regulating. The amount pumps.
to the fitting, the impact area has a of suction material depends on the • A centrifugal pump has a higher efficien-
thicker cross section than the normal discharge head. The higher the discharge cy than a jet pump. Stated in another
conveying pipeline. This thicker sec- head, the less suction material will be way, the power required for a centrifugal
tion is commonly termed an “integral taken—when there is no suction flow, pump to convey a given flow at a given
the only flow through the pump is the head is considerably less than the power
wearback.” motive (clear water) flow. This makes for required to give the jet-pump motive
Fittings can have connections that a self-cleaning action. If the discharge fluid the required head and flow to con-
have either a plain or a beveled end. line becomes clogged, the flow of more vey the given mixed water slurry head
Plain-end fittings are connected to the material into the pump is limited so that and flow. Often, a jet pump will require
the motive flow can flush the line. an increase in line size over a centrifugal
pipelines using sleeve couplings with
• There are no rotating parts subject to pump (8-inch (20 cm) instead of 6-inch
two gaskets held in place by friction. erosive ash. (15 cm) diameter, for example) thus
When plain-end sections are used, increasing the power differential.
piping harnesses and supports prevent • No separate motive-water pump is
the fitting from dislodging because required. This is a major horsepower sav-
ings in most cases.
of the reactive loads of the convey-
• In areas where water is limited and in
ing medium. Beveled-end fittings do
short supply, a centrifugal pump can
not require friction-type couplings to handle a given tonnage rate of bottom
secure the fitting to the pipe, but sup- ash with recirculated water and in many
ports should still be used at all fittings cases a smaller line size than a jet pump,
thus minimizing the water requirements.
for the reactive loads.
In conveying water-ash slurries,
it may be necessary to divert flow or
isolate branch lines when multiple and the gate before closing. Manual
sluice lines exist. Industry practice is to hand wheels or pneumatic operators are
provide heavy-duty knife-gate valves commonly used to open and close knife
for this service because of their profile gates. Automatic knife-gate systems can
characteristics. When the knife-gate have limit switches to provide feedback
valve is open, its orifice matches the to the control system.
diameter of the sluice pipe in which it Any slurry piping system must be
is installed. Minimizing the turbulence designed with minimum and maxi-
through the valve reduces erosion. In ad- mum velocities in mind. The minimum
dition, the insertion of the narrow valve velocity is that required to convey the Figure 6-51 | Chrome–nickel alloy 90-degree
in the line requires no special fittings. particle in the pipeline without set- Cleanelbow
Combustion Technologies
with integral wearback
Figure: 06.51/
helps TBwear from solids-
reduce
Also, knife-gate valves can have bonnets tling. This velocity is a function of the PMS5415 K
impact erosion and turbulence
to prevent leakage from the packing specific gravity of the material being experienced with changes in
and flushing ports for cleaning the seat conveyed, the pipeline size, and the flow direction
6-54 | Clean Combustion Technologies
slurry concentration. Typical ash-slurry from the associated water. Bottom ash
pipeline velocities are: is either removed from the site for other
use or remains on site in a permanent
• Bottom ash—7.5 to 8 ft/sec (2.3 to
impoundment (landfill).
2.4 m/sec) in horizontal lines with
9 to 10 ft/sec (2.7 to 3.0 m/sec) in
DEWATERING SYSTEMS. Ash from a con-
vertical risers (higher for slag)
veying sluice line can be deposited into
• Economizer ash—7.5 to 8 ft/sec
one of two dewatering bins. As the ash
(2.3 to 2.4 m/sec) in horizontal lines
enters the bin and settles, it displaces
with 9 to 10 ft/sec (2.7 to 3.0 m/sec)
water, which overflows into a trough ex-
in vertical risers
tending around the circumference of the
• Pyrites—10 ft/sec (3.0 m/sec)
bin top. The overflow water then drains
in horizontal lines with 12 ft/sec
by gravity to the settling tank, recircula-
(3.7 m/sec) in vertical risers
tion basin, or waste drain (Fig. 6-52).
• Fly ash—4.5 to 5 ft/sec(1.4 to
A series of baffle plates, concen-
1.5 m/sec)
tric with the outer shell of the bin,
minimizes ash carryover and undesir-
DISPOSAL OF BOTTOM ASH FROM HYDRAULIC able turbulence in the overflow trough.
SYSTEMS. Hydraulic ash systems dispose Submerged beneath the water level at
of the ash and water slurry in various the ash inlet location in the center of
ways, depending on the overall plant the bin, the inner baffle inhibits the
layout, ecological considerations, and finer material from reaching the over-
available disposal areas.40 As previously flow by targeting the larger particles
mentioned, the ecological considerations to the outer walls. The second baffle,
are presently a major influence on deci- extending above the water level of
sions made about the means of disposal the overflow trough, creates a barrier
of ash and slurry. The following methods to retain the floating material before it
explain how ash products are removed reaches the overflow trough. Eventually,
floating material settles and is released
through the discharge gate after the bin
is dewatered. Any particulate that can
flow under the inner baffles and over
to the overflow trough is still met with
either a serrated weir edge on the trough
or mesh screens to limit carryover of
fines to the water-recirculation system.
Dewatering bins are installed in pairs.
When one bin is filled, the input flow is
diverted to the empty bin. After a short
period of natural settling, a level of sur-
face water will exist. This surface water
is quickly drained off through vertical
decanting pipes or floating decanters.
After the surface water is drained, the
lower drain valves in the decanting pipes
or screens are automatically partially
Figure 6-52 | A complete water-recirculation system with jet pumps, two dewatering bins, a settling
tank, a surge tank, and recirculation pumps; pyrite hoppers and transfer tank shown also
Clean Combustion Technologies
Figure: 06.52/ BAM
PMS5415 K
Auxiliary Equipment | 6-55
Figure 6-61 | Dry drag conveyor with catenary shaft in drive section
Fly Ash Removal Systems ash from the hoppers. A fly ash intake
Fly ash from the hoppers serving air valve located at each hopper regulates
heaters, electrostatic precipitators the flow of the fly ash.
(ESPs), and baghouses is removed Fly ash intakes have carbon-steel
intermittently by either vacuum or cast-iron bodies and a swing
(negative-pressure) or pressure (positive- disk, which seals against a hardened
pressure) pneumatic type systems,
Clean Combustion or
Technologies seat. For maintenance, the outlet of
Figure:
combinations of 06.61/ (blb)
the two. each hopper has a manual isolation
PMS5415 K
gate. Each fly ash intake is actuated
VACUUM SYSTEMS. A vacuum system automatically by the system logic
(Fig. 6-62) uses a mechanical blower, that controls both the flow rate and
water exhauster, or a steam exhauster quantity of fly ash leaving the hop-
to create a vacuum that removes the fly per to avoid plugging the discharge
line. Because the system operates on
a vacuum, only one fly ash intake and
one conveyor branch line operate at
any given time. As each hopper is
emptied of fly ash, the system will step
to the next hopper in the same branch
line. When all hoppers in a branch line
have been emptied, the system will step
to the next branch line. The system
logic ensures the proper sequence of
events and positioning of valves. The
intake valve, shown in Figure 6-63, has
the moving parts and multiple access
areas needed to work with today’s
varied fuels, including high-calcium
ash from coals such as Powder River
Basin (PRB).
Figure 6-62 | Typical pneumatic vacuum system serving an assortment of collection hoppers using
a swing-gate-type material intake valve at each hopper
Clean Combustion Technologies
Figure: 06.62/ (JG)
PMS5415 K
Auxiliary Equipment | 6-63
ensures ash flow, especially for denser more ash than air), or the mixture may
and high-calcium ashes. begin as a dense phase and gradually
Vacuum transport systems usually change to a dilute phase at the discharge
move fly ash in a quite “airy” mixture end. Many low-velocity, dilute-phase
(i.e., very dilute phase), but the consis- systems operate at some time in a dense
tency of the mixture can vary with a phase because dense phase occurs when-
positive-pressure system. For example, ever saltation takes place in a pipeline.
the mixture might be in a dilute phase,
a dense phase (in which there is much AIRLOCK FEEDER OPERATION. It is custom-
ary in dilute-phase fly ash transport to
use the collecting hoppers for storage.
The airlock tank is used only during the
actual ash-moving process. Thus, the
inlet valve to the pressure tank is kept
closed until the transport cycle begins.
This exposes the ash to the unheated
tank metal for the shortest possible
time, but keeps the ash in the collecting
hoppers for the longest possible time.
In dense-phase transport, the inlet
valves of the pressure tanks are kept
open whenever the tanks are not full.
The ash, then, leaves the area that is heat-
ed by flue gas and drops into the tanks,
normally unheated, until the tank level
indicator causes the inlet valve to close.
Figure 6-66 | Typical water jet exhauster–based
Clean vacuum system with drain to a pond
Combustion Technologies The filled tank remains at essentially
Figure: 06.66/ BAM atmospheric pressure until its turn in the
PMS5415 K
transport “queue” is signaled. The time
that the ash remains in the tank before
fluidization and transport will depend
on the filling/emptying status of any
other tanks on the same conveying line.
pressure between any two zones before A vacuum conveying system features
the appropriate airlock gate opens and less equipment and less headroom
allows gravity or gravity-assisted flow required under each hopper: one fly
down through the airlock. The equal- ash intake valve versus two sets of
izing valves are exposed to ash-laden air airlock deck plate valves, an airlock
traveling at very high velocities. Their upper chamber, and an equalizer valve
design is as critical as the airlock. required for a pressure system. Also,
Either two 2-way (open–closed) a vacuum system uses the more posi-
equalizing valves can be used or one tive means of suction to evacuate the
3-way equalizing valve with each airlock. fly ash from the hoppers, whereas a
In a 3-way equalizing valve installation pressure system uses gravity alone
for a pressure system (Fig. 6-68), the to move the ash from the hopper to
3-way equalizing valve is first opened to the feeders. Of course, fluidization
vent the upper chamber of the airlock and rappers are available, but to be
Figure 6-68 | Equaling flow paths using a
and allow it to be at the same pressure effective, they must be used properly. Clean Combustion Technologies
3-way equalizing valve with an
Figure: 06.68/ (JG)
level as the hopper above. Vacuum systems are more dependent PMS5415 K
airlock
The top gate of the airlock is then on ambient air, which can be moist,
opened to allow ash to fall by gravity for conveying, while pressure systems
down into the upper chamber. On have the warmed air coming from
a timed cycle, the upper gate of the the blowers to aid in dealing with
airlock closes and the 3-way equalizing any sluggish ash. A vacuum system
valve switches position to allow the also has the advantage of being able
same blower pressure from below the to indicate, from the conveying-line
airlock to fill the upper chamber. An vacuum switches, whether a given
orifice before the airlock intake tee is hopper is full, empty, plugged, or rat
used to ensure slightly higher pressure holed. This is not always possible with
through the equalizing valve. The lower a pressure system, as multiple hoppers
gate of the airlock is then opened to al- are being evacuated simultaneously on
low gravity-assisted flow down into the quicker time cycles.
conveying line. After the volume of ash However, the practicality of a vacu-
in the airlock is discharged, the 3-way um system is limited by the amount of
equalizing valve returns to the first po- pressure differential that is available for
sition to vent the upper chamber and conveying. The actual pressure drop in-
start the continuous cycle over again. creases as the conveying rate increases.
Thus, if all else is constant, higher
COMPARISON OF VACUUM AND PRESSURE conveying rates may require a pressure
SYSTEM. Selecting a vacuum or pressure system. When plants are at high eleva-
fly ash removal system for a given unit tions, the available pressure decreases
depends on an evaluation of: (at 7,000 feet [2,100 m], approximately
5 inches Hg [12.7 cm/Hg] are lost over
• Conveying rate that of sea level), further limiting the
• Altitude application of vacuum systems. In addi-
• Number of hoppers to be evacuated tion, the pressure will drop as conveying
• Distance to be conveyed distance increases, making a pressure
• Number of disposal options system more feasible for longer convey-
desired ing distances.
6-66 | Clean Combustion Technologies
With the increase in options for the 20 to 40, or more) and the conveying
sale of fly ash depending on the texture distance dictates a pressure system,
and carbon content of the ash, more and it is usually economical to provide a
more plants need and want separate fly vacuum-pressure system (Fig. 6-69).
ash silos to receive and store “saleable” In this system, a transfer station is
and “non-saleable” ash. Re-injection of located close to the precipitator, or
high loss on ignition (LOI) ash also baghouse, and the fly ash is transported
comes into play as a possible option. by vacuum from the hoppers to the
These multiple disposal points are more transfer station. At the discharge of the
easily handled by pressure systems, station, a set of airlock feeders starts the
where the blowers are located back pressure system to convey the fly ash by
where the ash is first collected and need pressure to the main ash-storage silos.
only additional pipeline and valves to The capital and maintenance cost is
send the ash in any direction desired. reduced because there is less equip-
To understand a pneumatic trans- ment and more headroom under the
port system, it is important to realize hoppers. This reduction in equipment
that conveying distance is the total under the hoppers can more than
developed length of the pipeline. In offset additional costs associated with
pneumatics, the pressure drop through the transfer station and the positive-
an elbow can easily be the equivalent of pressure blowers. On the other hand,
50 feet (150 m) of pipe. Therefore, the a vacuum-pressure system uses more
optimal system would have a mini- power than a straight pressure system.
mum number of elbows.
FLY ASH HOPPER DESIGN. Problems have
VACUUM/PRESSURE SYSTEMS. When the been experienced in the removal of
number of precipitator or baghouse powdery fly ash from hoppers, particu-
collection hoppers is relatively high (say, larly with large precipitators on utility
steam generators. Because of the very
high ash content found in many coals,
and increasingly stringent particulate-
collection regulations, precipitators
on a large pulverized coal/fired unit
commonly have more than 40 hoppers
(sometimes as many as 128). Research
has shown that the incidence of hopper
operational and maintenance problems
is directly related to the number of
hoppers on a precipitator.48,49
Pneumatic ash removal systems are
not designed to handle wet mate-
rial. Thus, it is critical in the design
and operation of hopper systems
to maintain collected material at a
temperature sufficiently above the
water or acid dew point to keep it
absolutely dry, so that it will be as
free-flowing as possible.
Clean Combustion
Figure 6-69 | Typical vacuum-pneumatic Technologies
pressure system with transfer station
Figure: 06.69/ BAM
PMS5415 K
Auxiliary Equipment | 6-67
Dry dust in hoppers ordinarily will 12 inches (30 cm) (diameter or square).
flow freely by gravity and be transported In any case, 6- or 8-inch (15 or 20 cm)
pneumatically without difficulty. In outlets should be avoided wherever
order to do so, it: possible, unless such sizes have proven
successful with identical fuels.
• Must be kept essentially at the tem-
perature at which it was collected
HOPPER FLUIDIZING DEVICES. A fly ash
• Must not be exposed to any moisture
fluidizer, as it is used in a conical or
• Must not compact from its own
pyramidal hopper, is a porous mem-
weight to cause bridging above the
brane that allows pressurized airflow
hopper outlet
through it to be uniformly distributed
• Must have no clinkers formed as
to the material above, filling the voids
the result of oxidation of any com-
between the particles at a slight pres-
bustibles in the ash
sure and changing the effective angle
In power plant operation, it is frequent- of repose of the material to promote
ly difficult to keep the ash flowing freely. gravity flow. Properly located fluidiz-
With the steam generator at full load, ing devices can help evacuate hoppers
temperatures as low as 90°F (32°C) (particularly precipitators), if these
have been reported in hoppers of cold devices are well maintained and reliably
precipitators (those that follow the boil- supplied with dry air preheated above
er air heaters). Similarly, temperatures dew-point temperatures. If not, fluid-
20 to 40°F (–7 to 4°C) above ambient izers will only aggravate evacuation by
have been recorded in the hoppers of caking the ash, and provide unwanted
“hot” precipitators (those located ahead surface areas for ash accumulation and
of the boiler air heater). bridging. Where significant percent-
The problem in the industry is even ages of combustibles are present in the
more severe with the high-calcium ash collected fly ash, the fluidizing-medium
associated with certain sub-bituminous supply must be non-oxidizing to pre-
coals and various lignites. In addition vent destructive hopper fires.
to being highly hygroscopic, such ash
shows pozzolanic and cementitious HOPPER VIBRATORS. When their opera-
properties when exposed to moisture. tion is properly controlled and timed,
With the increased use of low-sulfur, hopper vibrators can help prevent
PRB coal to lower or avoid the high bridging and rat holing. In the auto-
cost of scrubbing equipment, this issue matic operation of vacuum pneumatic
is more common than ever before. systems, vibrators should be regulated
by ash evacuation controls to ensure
HOPPER SIZING. Hoppers under pre- operation only when an “empty hopper”
cipitators and baghouses should be signal is generated and the evacuation
considered only as funnels. Because of cycle is completed. Routine use of the
difficulties with removing fly ash that vibrators during evacuation of damp
has been stored for a long time while ash will further compact the ash and
exposed to flue gas containing moisture make evacuation difficult.50 It should
and sulfur, recommendations state that be possible to manually operate the
there should be no specified storage vibrators from each hopper to assist
time in collecting-equipment hoppers. maintenance personnel during emer-
Minimum hopper-outlet size should be gency evacuation.
6-68 | Clean Combustion Technologies
it can be discharged into the silo. Pres- the precipitator inlet duct. The vent fan
sure systems require only a target box places the silo under a slight negative
or an entry elbow on the silo roof. pressure, eliminating the possibility of
blowing dust out of the silo. Vent-fan
AIR/FLY ASH SEPARATION EQUIPMENT. As operation can be synchronized with that
stated in the discussion of vacuum of the conveying blowers.
systems, filter–receivers are the
primary mechanism to separate the SILO FLUIDIZATION. Fly ash, when not
fly ash from the conveying air. The allowed to agglomerate and when
stone box entry in the receiver tank fluidized with heated air, takes on the
(Fig. 6-64) allows capture of most of flow properties of a fluid. To expedite
the heavier ash particles. By using an fly ash removal, troughs with fluid-
airlock and equalizing valves at the izing air are provided at the bottom of
base of the receiver tank, the storage the silos. The fluidizing elements are
silo is never under vacuum and the normally constructed of canvas or po-
system can function continuously. The rous stone. Some plants have been using
bag filter on top of the receiver tank porous laminates of sintered, stainless
acts to precipitate the remaining fly ash steel wire. This provides a microporous
from the conveying air stream. These filter that allows controlled amounts of
multiple levels of separation protect heated air to pass and fluidize the fly ash.
the vacuum pump from unnecessary The possibility of stress cracking that is
particulate carryover. The bag filter common with fluidizing stones is elimi-
usually features pulse jet air cleaning nated by using this metallic material.
with a good air-to-cloth ratio. The Because of the microporous construc-
vacuum-producing device (blower, tion, the metallic cloth is highly resistant
vacuum pump, exhauster), which has to blinding. In addition, water will not
pulled the fly ash–air mixture down penetrate into the porous openings.
the pipeline and through the discharge This will allow washing of equipment
equipment, is located at the side or with water without having to remove
underneath the silo. Sound enclosures the fluidizing element. Fluidizing-air
can ensure proper noise abatement. A blowers and associated heaters provide
small, bin vent, bag filter is also needed the required air and temperature level
on a silo with a vacuum system to to the fluidizing-air troughs. Care must
vent the displaced air, aeration air, and be taken to properly size the blowers to
unloading system air. account for a full head of fly ash in the
On a pressure system, somewhat silo (Fig. 6-70).
simpler filtration equipment is used. Silo discharge equipment is designed
The conveying air enters the silo directly so that fly ash can be removed dry into
to join the displaced air, aeration air, enclosed vehicles, or so that it can be
and unloading-system air. All of this moistened (to 12–20% moisture by
air needs to be vented to keep the silo weight) for transport in open vehicles.
near ambient pressure. Large, bin vent, It is increasingly common for fly ash to
bag filters are the least expensive passive be sold for use in landfills, roads, and
form of venting. Alternatively, fans and the cement industry.
motor drives can be mounted on top Telescoping ash discharge chutes
of the storage silo. The vent fan should convey fly ash dry from the storage silo
be designed to handle the static head to an enclosed removal vehicle stationed
required to transport all the air back to below. Normally, each chute is capable
6-70 | Clean Combustion Technologies
of extending a full 18 feet (5.4 m) using process, which involves displacing the
a motor-driven, winch retract system. air volume in the receiving vehicle. An
A vent fan and piping in each chute, air-operated gate controls the inlet flow
vent back into the silo any fugitive to the chute and is placed just below a
dust released during the unloading manual maintenance gate. A clean-
out compartment should be included
between the maintenance gate and the
control gate.
be considered when selecting the type plant-wide DCS can easily handle
of control system and the degree of these few control signals and alarms.
automation that should be used. First,
the more automated the plant, the less MATERIAL- AND ENERGY-BALANCE
attention paid by the plant operators. CONSIDERATIONS
If little attention is paid to ash systems, Water-impounded, ash-receiving equip-
breakdowns are inevitable. When ment beneath a furnace, as previously
breakdowns occur, the operators may stated, helps to quench the ash as it falls
lack sufficient knowledge of the system from the furnace and to transport it to a
to take corrective action. For intermit- disposal point. The thermal shock to the
tent-removal bottom ash and pyrites hot ash as it enters the low-temperature
systems, many utilities select a push- water helps to break up the large pieces,
button automatic system, in which the while submerging the ash prevents sinter-
operator, at a local panel, performs the ing during the time that it may be stored
ash-removal sequence. The operator (longer time with intermittent-removal
can judge by the crushing sound of the systems, shorter time with an SSC).
grinder whether each hopper section Part of the heat released in the fur-
has been emptied. If a problem arises, nace is transmitted to the ash hopper by
the operator is on hand to analyze it radiation from the burning fuel. How-
and take action. ever, most of the heat input is from
Similarly, on a vacuum pneumatic the sensible heat given up to the water
system, the operator can make a check by the hot bottom ash falling from
by examining the strip chart that the furnace. The ash systems engineer
records the vacuum in the pneumatic must perform a heat-balance analysis
pipeline. Because the vacuum level can to determine the flow of cooling water
be related to the ash flow, the opera- required to absorb the incoming heat
tor can easily determine the removal and maintain the water temperature at
rate and time required to empty each a predetermined level.
hopper. Inconsistencies in removal time Field experience indicates that the
and/or removal rate per hopper are optimal temperature of the impounded
signs of a problem. water is about 140°F (60°C), and it
On multi-field precipitators, PLCs should not be higher than 160°F (71°C).
can be used to advantage. Because the Temperatures at this level are effective in
front fields will collect ash at a much rapidly cooling and fracturing the falling
faster rate than the rear fields, the op- ash. Lower temperatures require more
erator may choose to alter the sequence cooling water. Higher temperature may
in which hoppers are “visited” by the prove uncomfortable or unsafe in the
ash-removal system. Such adjustments area around the bottom-ash receiver.
accommodate unequal distribution
across the fields, or compensate for Bottom-Ash Cooling System Design
fields being removed from service. Heat is removed from bottom-ash hop-
Continuous-removal systems obvi- pers or submerged scraper conveyors by:
ously require the least-sophisticated
controls, but on/off switches, indicating • Evaporation of water from the
lights, and alarms should be provided. surface of the water pool
Programmable controllers are rarely • The heat removed in the water going
justified with such systems.41,42 The out the overflow or leakage points
6-72 | Clean Combustion Technologies
radiation transmitted)
0.6
Q fr = Afa × Rf × Ftr Curve is for l/w ≥ 5
the deaerator. The condenser is gener- the fans, their design is smaller than
ally of the shell-and-tube design. The natural draft cooling towers. Dry cool-
exhaust steam enters the condenser ing towers are mechanical draft cooling
shell at a sub-atmospheric pressure. towers, effectively being a steam-to-air
The steam is condensed by the circu- or a water-to-air heat exchanger.
lating water.
Air-Cooled Condensers. Air-cooled
Cooling Towers. Cooling towers are condensers condense exhaust steam
frequently used to exhaust process heat from the steam turbine and return
into the atmosphere. Cooling tow- the condensate to the boiler. They
ers are deployed where water usage is are used in electrical power plants
restricted. They are used to avoid using and waste-to-energy plants when
rivers, lakes, or other bodies of water cooling water is in short supply (i.e.,
for cooling. Cooling towers can also be dry regions) or when environmental
installed as a supplemental heat sink concerns over water usage, visibility, The cooling system provides
to limit the release of process heat into etc., are paramount. An air-cooled the heat sink that is required
rivers, lakes, or other bodies of water. condenser is made of modules by the second law of
Cooling towers can be classified as arranged in parallel rows. Each
wet or dry. Wet cooling towers are module contains a number of fin-tube
thermodynamics. The lower
designed to have the cooling water in bundles. An axial flow, forced draft the cooling temperature, the
direct contact with the atmosphere. fan located in each module forces the higher the plant efficiency.
Dry cooling towers do not have the cooling air across the heat exchange
cooling water in contact with the area of the fin tubes. Fin tubes are
atmosphere. Wet cooling towers lose critical to the technology; they are
some cooling water to evaporation as a necessary because of the low thermal
normal part of the process. Dry cooling conductivity, low density, and low heat
towers do not lose water to evapora- capacity of air. This larger surface area
tion. Because dry cooling towers do required to obtain a given heat-removal
not lose water to evaporation, they are rate, the area increasing with the
sometimes used where water resources design ambient air temperature. This
are restricted. fact also means that the footprint of
Cooling towers are also classified air-cooled condensers is larger than
as either natural draft or mechanical that of their water-cooled equivalents.
draft. Natural draft cooling towers Noise created by the large number of
circulate air by the natural draft created fans may introduce its own environ-
from heating the air. The design of the mental problem. Finally, an air-cooled
natural draft cooling tower shape is condenser is more expensive than its
intended to take advantage of this draft water-cooled equivalent.
to improve the performance. Large
hyperbolic cooling towers are examples
of an efficient design to optimize the Rory Eastman
Gus Shearer
performance of a natural draft cooling Gary Mooney
tower. Mechanical draft cooling towers Dennis Piontek
use circulating fans to move air through Phil Lafave
Doug Martin
the tower to improve heat transfer to
Roger Martinsen
the air. While the mechanical draft Peter Kelly
cooling towers need energy to operate Carl Bozzuto
REFERENCES
1. H. Kreisinger, “Combustion of Pulverized Coal,” Transactions of the 21. W. T. Ray and H. Kreisinger, “The Flow of Heat Through Furnace
ASME, 60 (Paper No. FSP-60-8): 289–296, 1938. Walls,” U.S. Bureau of Mines Bulletin 8. Washington: U.S. Bureau of
Mines, 1911.
2. N. L. Sadi Carnot, E. Clayperon, and R. Clausius, Reflections on the
Motive Power of Fire; and other Papers on the Second Law of Thermodynam- 22. H. Kreisinger and W. T. Ray, “The Transmission of Heat into Steam
ics; edited with an introduction by E. Mendoza. New York: Dover, 1962. Boilers,” U.S. Bureau of Mines Bulletin 18. Washington: U.S. Bureau of
Mines, 1912.
3. R. Diesel, Theory and Construction of a Rational Heat Motor, trans. by
Bryan Donkin. London: Spon, 1894. 23. H. Kreisinger and J. F. Borkley, “Heat Transmission Through Boiler
Tubes,” U.S. Bureau of Mines Technical Paper 114. Washington: U.S.
4. F. Klemm, A History of Western Technology. Cambridge, MA: MIT Press,
Bureau of Mines, 1915.
1964, pp. 342–347.
24. H. Kreisinger, J. Blizard, C. E. Augustine, and B. J. Cross, “An Investiga-
5. C. F. Herington, Powdered Coal as a Fuel. New York: Van Nostrand,
tion of Powdered Coal as Fuel for Power-Plant Boilers-Tests at Oneida
1918, pp. 68–72.
Street Power Station, Milwaukee, Wisconsin,” U.S. Bureau of Mines
6. F. L. Dyer and T. C. Martin, Edison—His Life and Inventions, with col- Bulletin 223. Washington: U.S. Bureau of Mines, 1923.
laboration of W. H. Meadowcraft. Vol. II. New York: Harper and Bros.,
25. H. Kreisinger, J. Blizard, C. E. Augustine, and B. J. Cross, “Tests of a
1929.
Large Boiler Fired with Powdered Coal at the Lakeside Station, Mil-
7. “Symposium on Powdered Fuel,” Spring Meeting, American Society of waukee,” U.S. Bureau of Mines Bulletin 237. Washington: U.S. Bureau
Mechanical Engineers, St. Paul-Minneapolis. Transactions of the ASME, of Mines, 1925.
36: 85–169, 1914.
26. E. Audibert, “Etude de l’Entrainement du Poussier de Houille par I’Air,”
8. R. C. Carpenter, “Pulverized Coal Burning in the Cement Industry,” Annales de Mines, 1(3): 153–191, 1922.
Transactions of the ASME, 36: 85–107, 1914.
27. J. Blizard, “The Terminal Velocity of Particles of Powdered Coal Falling
9. W. Dalton and W. S. Quigley, “An Installation for Powdered Coal Fuel in in Air or Other Viscous Fluid,” Journal of the Franklin Institute, 197(2):
Industrial Furnaces,” Transactions of the ASME, 36: 109–121, 1914. 199–208, 1924.
10. F. R. Low, “Pulverized Coal for Steam Making,” Transactions of the 28. J. Blizard, “Transportation and Combustion of Powdered Coal,” U.S.
ASME, 36: 123–136, 1914. Bureau of Mines Bulletin 217. Washington: Government Printing
Office, 1923.
11. “Topical Discussion on Powdered Fuel,” Transactions of the ASME, 36:
137–169, 1914. 29. H. Verdinne, “The Technique of Powdered Fuel Firing,” Fuel in Science
and Practice, 2: 146–151, 1923.
12. J. Anderson, “Pulverized Coal Under Central-Station Boilers,” Power,
51(9): 336–339, 1920. 30. W. Nusselt, “Der Verbrennungsvorgang in der Kohlenstaubfeuerung
(The Combustion Process in Pul�������������������������
verized Coal Furnaces),“ VDI
13. P. W. Thompson, “Pulverized Fuel at Oneida Street Plant,” Power, 51(9): Zeitschrift, 68(6): 124–128, 1924.
339–340, 1920.
31. P. Rosin, “Die Thermodynamischen und Wirtschafthichen Grundlagen
14. “Four-Day Test on Five Oneida Street Boilers Burning Pulverized Coal,” der Kohlenstaubfeuerung (Thermodynamic and Economic Bases of
Power, 51(9): 354–357, 1920. Pulverized Coal Firing),” Braunkohle, 24(11): 241–259, 1925.
15. “Pulverized Coal at Milwaukee,” Power, 51(9): 341–342, 1920. 32. W. J. Wohlenberg and D. G. Morrow, “Radiation in the Pulverized-Fuel
16. “The New Lakeside Pulverized-Coal Plant, Milwaukee,” Power, 52(10): Furnace,“ Transactions of the ASME, 47: 127–176, 1925.
358–360, 1920. 33. R. P. Hensel, “Coal Combustion,” presented at the Engineering Founda-
17. “Largest Station Using Pulverized Coal,” Power, 55(16): 604–610, 1922. tion Conference on Coal Preparation for Coal conversion, 1st, Franklin
Pierce College, Rindge, N.H., 1975. Paper updated in 1978. Combus-
18. H. Kreisinger and J. Blizard, “Milwaukee’s Contribution to Pulverized tion Engineering publication TIS-4599.
Coal Development,” Mechanical Engineering, 61: 723–726, 737.
34. R. P. Hensel, “The Effects of Agglomerating Characteristics of Coals
19. F. L. Dornbrook, “Developments in Burning Pulverized Coal–Thirty on Combustion in Pulverized Fuel Boilers,” Symposium on Coal
Year Review of Experience in Milwaukee Plants,” Mechanical Engineer- Agglomeration and Conversion, Morgantown, W.V., 1975. Sponsored
ing, 70: 967–974, 1948. by the West Virginia Geological and Economic Survey in Cooperation
20. H. Kreisinger, F. K. Ovitz, and C. E. Augustine, “Combustion in the With the Coal Research Bureau. Combustion Engineering publication
Fuel Bed of Hand-Fired Furnaces,” U.S. Bureau of Mines Technical TIS-4353.
Paper 137. Washington: U.S. Bureau of Mines, 1916.
6-82
35. V. F. Parry, “Production, Classification and Utilization of Western 46. R. G. Walsh and D. S. Piontek, “Financial/Cost Benefit Consider-
United States Coals,” Economic Geology, 45(6): 515–532, September- ations and Analysis for Replacement of Conventional Bottom Ash
October, 1950. Systems with a Submerged Scraper Conveyor,” Presented at Power-Gen
International Conference, Orlando, FL, Nov 28 – 30, 2006. PennWell
36. D. Magnum and P. L. Stanwicks, “HP Series Pulverizer Design, Testing,
Corporation, Tulsa, OK, 1979; also, Alstom Power publication TIS-8708.
Maintenance Cost,” Proceedings of POWERGEN ‘89 Conference,
Book 2, POWER-GEN ‘89, New Orleans, LA, December 5–7, 1989. 47. G. D. Mooney and J. J. Murphy, “Ash Handling Systems for Coal Fired
Boilers,” Coal Technology, volume 2: Storage and Handling. Houston, TX.
37. E. H. Tenney, “Pulverization and Boiler Performance,” FSP-54-7, Pro-
Dec., 1982.
ceedings of the Fourth National Fuels Meeting, Feb 11 to 13, 1931, pp.
55–65. Chicago: The American Society of Mechanical Engineers. 1931. 48. Joint Technical Committee of the American Boiler Manufacturers
Association and Industrial Gas Cleaning Institute, Inc., “Design and
38. E.G. Bailey, “Present Status of Furnace and Burner Design for Pulver-
Operation of Reliable Central Station Fly Ash Hopper Evacuation Sys-
ized Fuel,” Transactions of the American Society of Mechanical Engineers,
tems.” Proceedings of the American Power Conference, 42:74–85. Chicago:
September to December 1928, p. 177. New York: The American Society
Illinois Institute of Technology, 1980.
of Mechanical Engineers.
49. J. C. Singer, “Design for Better ESP/Fabric Filter Hopper Operation
39. W. E. Loftus, “Ash Handling, Storage and Utilization.” Proceedings of the
and Maintenance,” Presented at Air Pollution Control Association 76th
American Power Conference, 38: 707–717. Chicago: Illinois Institute of
Annual Meeting and Exhibition, Atlanta, June 19–24, 1983. AWMA,
Technology, 1976.
Pittsburgh, PA.; also, Alstom Power publication TIS-7402, Windsor, CT.
40. J. C. Singer and A. J. Cozza, “Material and Energy Balances of
50. J. C. Singer and A. J. Cozza, “Ash-Handling Options on Retrofitted and
Ash-Handling Systems,” Presented at the Joint Power Generation
Converted Steam Generators,” Presented at ASMEIEEE-ASCE Joint
Conference. Dallas, Sept. 10–13, 1978.ASME, N.Y., N.Y.; also. Alstom
Power Generation Conference, St. Louis, October 4–8, 1981. ASME,
Power publication TIS-5822, Windsor, CT.
New York, NY; also, Alstom Power publication TIS-6869.Windsor CT.
41. J. C. Singer and A. J. Cozza, “Design for Continuous Ash Removal:
51. J. C. Fleming and D. M. Rode, “Ash Removal from Industrial Boilers—
Alternative Concepts in Ash Handling,” Proceedings of the American
The Changing Scene,” Power, 126(9), 1982.
Power Conference, 41: 544–553, 1979. Chicago: Illinois Institute of
Technology. 1979; also, Alstom publication TlS-6211.5. 52. M. B. Caron, A. J. Cozza, J. C. Singer, and J. R. Young Jr., “Steam-Gen-
erator Availability as Affected by Ash-Handling Equipment,” Proceedings
42. J. E. Horne and A. Bosso, “Southwestern Public Service Company
of the American Power Conference, 44:214–225. Chicago: Illinois Insti-
Pioneering in Continuous Bottom Ash Removal,” Presented at Frontiers
tute of Technology, 1982; also, Alstom Power publication TIS-7118,
of Power Conference, October 11 and 12, 1982, Stillwater, OK; also,
Windsor, CT.
Alstom publication TIS-7261.
53. G. D. Mooney, R. J. Kowalik, and C. D. Hicks, “Dry Fly Ash System
43. C. R. Gretz and G. D. Mooney, “Bottom Ash Methods: A Look at Drag
Conversion for an Existing 480 MW Power Plant—Justification,
Conveyors for U.S. Applications,” Presented at Coal Technology Con-
Design and Construction.” Power-Gen International Conference, Dallas,
ference , Houston, Tx, Nov 13-15, 1979. PenWell Corporation, Tulsa,
TX, Dec 9-11, 1997. PennWell Corporation, Tulsa, OK.
OK 1979.
54. G. D. Mooney and R. G. Walsh, “Upgrades and Rebuilds in the
44. J. M. Chapman and D. S. Piontek, “A Modern Ash Handling System for
Aftermarket for Ash Handling,” Presented at Coal-Gen Conference,
a Large Multi-fuel Boiler Installation,” Presented at TAPPI Conference,
Las Vegas, NV, Dec. 6-8, 2005,PennWell Corporation, Tulsa, OK.; also
1985. TAPPI Jour., Vol 69, No. 7, pp55-60, 1979.
Alstom Power publication TIS-8703, Windsor, CT.
45. J. F. Bender, D. S. Piontek, and R. G. Walsh, “Case Study—Retrofit
55. G. D. Mooney and R. G. Walsh, “Matching Ash Systems to Fluidized
Submerged Scraper Conveyor for Bottom Ash—5 Years Later-Ray
Bed Boilers and Fuels,” Presented at Council of Industrial Boiler Own-
Nixon Unit 1,” Presented at Power-Gen International Conference,
ers (CIBO) Conference, Lexington, KY, May 8-10, , 2007.CIBO, Burke,
Orlando, FL, Nov 30- Dec 2, 2004. PennWell Corporation, Tulsa, OK,
VA.; also Alstom Power publication TIS-8710, Windsor, CT.
2004.
6-83
Controls
Chapter Seven
In today’s markets, customers are moving from the business of simply operating the plant to
the business of selling electricity. Their primary aim is to achieve faster payback and higher
returns on investment. Control systems are expected to meet the need for different solutions
and capabilities. In addition to state-of-the-art equipment and systems, the control system
technology is a key feature of any clean and efficient power plant. Historically, meters, gages,
and lights displayed equipment status to the operator, while recorders made a permanent record
of plant performance. Remotely oper- and flexible. This feature means that both for the design office and the
ated air cylinders and electric motors the system can be optimized for on-site high quality documentation.
gave plant operators the capability of all types of projects: new plants or
responding quickly and efficiently to retrofits, overall plant automation, all The advanced control system pro-
changing plant requirements. More power plants (thermal, hydro, nuclear, vides many features and benefits for
recently, liquid crystal displays (LCDs) diesel, and associated substations), and the plant owner and operator including
have replaced the panel board instru- machine control (turbine governing
mentation, to link the operator with and generator control). • Advanced technologies (object-ori-
past and present process information The range of control stations and ented engineering, Internet, fieldbus)
through sophisticated microprocessor- controllers provided allows for the • Intuitive, user friendly operator
based distributed control hardware. choice of the best technical and eco- interface (based on modular design
nomical solution. Three major parts concept, adaptable multifunction
A TOTAL CONTROL SOLUTION constitute the system: controllers, distributed control
The plant automation system in a flex- architecture, and open system
ible market covers both the traditional • The operator control interface and architecture based on international
distributed control system (DCS) control room environment with high standards)
functions and advanced plant manage- availability and high performance • Deterministic/high-performance
ment functions. This contributes to • The automation cells that execute communication networks
reduce operation and maintenance the control and protection func- • High reliability and availability,
(O&M) costs. Installation costs can be tions of the machines and connect with tolerance to data spikes dur-
optimized by close integration within to the process (the use of fieldbuses ing transient operations
the overall electrical engineering. A allows geographical distribution of • Galvanic isolation on input/output
fieldbus-based distributed architecture, the equipment) (I/O) and communication networks
integrating widely available hardware • The engineering station that provides • Integrated comprehensive plant
elements keeps the architecture open a set of advanced engineering tools management system (Fig. 7-1).
7-
7- | Clean Combustion Technologies
Controlog Controlog
Video Wall Web
Thin Clients
Optiplant Modiplant
EMS, PPS, Condition
IMS Reports Monitoring WAN
Firewall
Unit Network
Multi-Function
Controllers
OLC
Field Networks
CLC
Control Room
Engineering Tools
Intranet
An Integrated
Comprehensive
Plant
Management
System Integrated Training Simulator
To
Alarms
Fully Distributed
Automation/Machine
Control An Integrated Testing
Platform
Hot Standby
Redundancy:
Immediate
Normal Standby Normal Standby Normal Standby
Update Settings Switch-Over
to Standby
Controller
Process Commands Process Commands
SYSTEM_SET
644
644
is accepted or refused. In addition, a
1 1LCB31 1 1PAB31
variable known as an execution check
Explorer
+ Pan Zoom Selection Print Screen Mimic Total: 0
Detail Popup
0.0 x LP TURBINE UNIT OVERVIEW CNDS FILLING
11 LCA60AA002
0.0 x 0.0 x
EXHAUST
11OVE11 SURGE TANK
Window
0.0 x 11LCC51 11MAN31 FROM
0.0 x
0.0 x
10.1 % 11 x 60AA005
0.0LCA 11MAJ31 COMMON
1 x DEMIN WATER
1 x 10.1 % 11LAA31 HP HEATERS LP HEATERS 0.0 °C 11LCP31
START-UP VENT
test, or unlock faceplates. A locked
1 x CEP MIN FLOW CV START-UP 0.0 °C
1 x VENT /VENT 11LCP12
1 x 0.0 % 1.65 m 11LAD31 11LCC51
AA041
1 x 0.000 bar–A
1 x 1 x 0.0 % 382 11LAA10BB010
100
0 0
50s 10 20 30 40
Setpoint
10 mn 0.0 % LP HEATERS
AUX STEAM
Duration: 1 mn 11:10:41 EMERG DRAINS
11LB000 0.0 % 11LCC51 2210 mm
Devices
COND 11LCB20 11LCA20
11MAN31 11LDK31 11LCA24 AP01D 11LCA20
WATER BOX AA002 AA001
COND INJECTION AA002 AT001
50.0 A
11MAW31
382 kg/s 11LCM31
11MAJ20AC010 0.0 %
382 kg/s 11LCA34
11LCB30
11LCA30 11LCA30
SEQ MAIN CNDS SYSTEM
the display of the dynamic animated
Command
100.0 % AA002
AP01D
AA001
50.0 A AT001
11MAW30AC010
11MAJ31
CONDENSATE
POLISHING
40.0 bar
! ! applications diagrams generated for
Faceplate
11LDK41 DOSING
NH3 & N2H4
11MAJ30AC010 11LFN31
FWT_LEVEL CV SEL
11LCA60
AA002
11LCA60
AA005
EXTRACTION PUMPS SEL
11LCB10
AP010
11LCB20
AP010
11LCB30
AP010
EXTRACTION PUMPS GP SEQUENCE
AUTO
BOILER SAFETIES
TURBINE SAFETIES
every controller from the control log.
STEP: 0
11LCA00
EC002
11LCA00
EC101
11LCA00
EA100 MODE SEL
UNIT LOAD 0.0 MW
These views are displayed from the ex-
Unack: 18 Total: 18 1 2 3 4 +
Associated Time
2/18/2008 13:25:59.232 11PAB10CL010_XG02
ID
CWP1 HYDR V OIL LVL
Label Value Unit State
< MIN1
Status Int Doubt Valid plorer. The functional structure of the
Alarms 2/18/2008 13:25:59.232 11PAB10CL010_XG02
2/18/2008 13:25:59.232 11PAB10CL011_XG04
CWP2 HYDR V OIL LVL
CWP1 HYDR V OIL LVL
< MIN1
< MIN2 plant application studies is accessible
directly by the operator in the control
room. The operator can follow and
Figure 7-7 | Control mimic and faceplates (The “screen data” in Figures 7-7 to 7-14 are for
illustrative purposes only.)
check the status of the open and closed
Clean Combustion Technologies loops. Depending on the operator
Figure: 07.07 (MH) & (SPS) rights, the application parameters and
PMS5415 K the forced variables can be tuned and
ALSPA Controlog
modified (Fig. 7-8).
0 0
BOILER 0
CONDENSATE 0
Functional Structure
File Edit View
AUTO
Diagram Observation Observation Preferences Curves Observation Window
–
11HHF10CT003_XQ50
11HHA00EA001_XH02
B61g
XX
XXXXXXXX
XX XX XX
XXXXX
The curves viewer displays the curves
from a single variable or from a public or
XX 0
– CHARGE
XXXXXXXX
– COND08 XX XX XX
XXXXX 01HHA10EA001_SG3_C
– Cmd
XXXXXX_XX_XX
XX XX XX XX XX X
[XX][XXXXX]
XXXXX
0
FB_O ORDER
0
0
XXXXXX_XX_XX
XX XX XX XX XX X
[XX][XXXXX]
XXXXX
process variable relative to another ana-
ACTVT
AC
C_JMP STEP_JMP log variable. The y = f(x) curves viewer
displays pairs of variables grouped in
c30_1 CSD8SMV2J Logged in CIS_5 (localhost) User: USER5 FR EN ALSTOM 11:25:34
public or private y = f(x) groups. The
Figure 7-8 | Functional diagram viewer
curves viewer plots the curves with the (11LCA55CF902_XJ60) 11LCB31 11OVE11 11PAB31 FEEDWATER PUMP 3 (11LCA4OFFO09_XJ60) (11LCA60AA008_XE60)
% 40
a group (Fig. 7-10).
20
0
ALARMS AND EVENTS –20
-139.048
0.000
0.000
MANAGEMENT 0.000
32 33 34 35 36 37 38 39
mn
An alarm is an event representing an
02/20/2008 11:31:28 Duratio: 7 mn 44 s 02/20/2008 11:39:12
operational anomaly in the process or Vertical Rules Horizontal Rules Zoom/Pan Exports Imports Slope
in the system. It is triggered by a state Curve: Y: 109.83 kg/s [+– 0.00] Delta Y:
PMS5415 K
+
1 1LCB31 FEEDWATER PUMP 3 (L LLCA60AA008_XE60) 11BFADDEA111_XD11CG (11LCA55CF902_XJ60) (11LMAG10CL901_XJ60) FEED WATER...5 DIAGRAM
bar
%
40 120
30 80
alarm. Sixteen degrees of urgency are 20
10 20
available and each degree of urgency 0
10 20 10 40 50 60 70 80 32
0
%
is allocated to a different alarm color Vertical Rules Horizontal Rules Zoom/Pan Exports Imports Slope
D
11LAB20CP902_XJ60 BFP3 FWT FLOW A 0.000 280.000 0.000 kg/s
disappearance of an alarm.
MimicId: 11LCB31
Unack: 18 Total: 18 1 2 3 4 +
Time ID Label Value Unit State Status Int Doubt Valid
110QTHIS_ARCH
110QTHIS_ARF
110QTHIS_HHD
Alarm Sheet
110QTHIS_HHR
110QTHIS_LINK
110QTHIS_OVER
110QTHIS_SSVP FLUE GAS CLEANING
of dynamic vertical/horizontal bars.
The bar-graph viewer displays bar
110QTHIS_PC36035
SCR REAGENT PREPARATION INJECTION
110QTHIS_PC36035
110QTHIS_PC36035
110QTHIS_PC36035
110QTHIS_Source
Alarm tag and label
graphs for a single variable or a group
ALARMS
A1
ALARMSAS01 Alarm Part
ALARMSAS02
ALARMSAS03
Unack: 1515 Total: 1515 1 2 3 4 +
CONTROL SYSTEMS
10/27/2006 5:42:17 PM
HOME UNIT 1 UNIT 2 PUMP 1 BOILER 1 TURBINE 1 TURBINE 2 TURBINE 3
Search:
Explorer
room was a big step forward in power
Display Add to favorites View
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0.00 –100.00
10/27/2006 5:32:36 PM
unburned fuel accumulated in only HOME UNIT 1 UNIT 2 PUMP 1 BOILER 1 TURBINE 1 TURBINE 2 TURBINE 3
Search
60
20
NB_CYC_MES
17:32:12 55 56 100 6 9 83
17:32:22
55
55
56
56
100
100
6
6
9
9
83
83
17:32:32
55
55
56
56
100
100
6
6
9
9
83
83
must be supervised to check the results. Combustion must be kept efficient or the
unconverted chemical energy may accumulate and subsequently become explosive. Satisfactory boiler operation
The following factors influence the effective composition change of an explosive requires that the flow and
charge:
processing of fuel, air, ignition
• The facility for mixing energy, and products of
• The inert material in the fuel
combustion be properly
• The fuel/air ratio (a near-stoichiometric ratio develops the highest explosion
pressure) prepared, ratioed, directed,
• The kind of fuel and sequenced to prevent
A furnace explosion requires both sufficient explosive accumulation within the possibility of an explosive
the furnace and sufficient energy for ignition. The ignition requirements for an mixture in the furnace.
explosive charge are very small, making it impossible to protect against all pos-
sible sources of ignition, such as static electricity discharges, hot slag, and hot
furnace surfaces. Therefore, the practical means of avoiding a furnace explosion
is the prevention of an explosive accumulation. The factors determining the
magnitude of a furnace explosion (mass, change in composition, and reaction
time) are related in the explosion factor.
Each furnace has a limiting explosion factor. If the conditions create an explo-
sion factor exceeding this limit, a catastrophic explosion can result. Any lesser
reaction will produce a furnace “puff ” (a nondestructive explosion) or a temporary
upset.
To protect a furnace from an explosion, a safety system must ensure a minimum
reactive mass accumulation with a minimum available composition change and with
a maximum reaction time required. Only control of the composition of the furnace
atmosphere offers complete coverage in minimizing the explosion factor. After
firing has begun, furnaces always contain sufficient mass to have an explosion and
control of the time factor is impossible. Therefore, the composition change must be
controlled to prevent furnace explosions.
The mechanics of furnace explosions, although defining the actual process, do
not describe the furnace operations that provide the explosive accumulations. Ideal
furnace operation continuously converts reactive furnace inputs into unreactive
products as fast as the inputs enter the furnace. This precludes furnace explosions.
However, in practical furnace firing, unfavorable operations that create explosive
situations are difficult to avoid completely.
Several correctly timed events precede a damaging furnace explosion. The
furnace explosion event itself is the rapid change in composition of the furnace
atmosphere (not the furnace inputs). The change in furnace composition is not
spontaneous, and suitable ignition energy, which can be substantially less than
that required for continuous furnace input ignition, must be supplied after the
explosive composition is attained.
7-12 | Clean Combustion Technologies
1,500
produced by gas and oil fossil fuels
without being fooled by other heat
1,000
sources. Electrical shorts were com-
mon in UV tube technology, which
500 helped institute fault detection in
flame monitors. A mechanical shutter
0 would be placed in front of the detec-
tor to block the light from the flame.
200 300 400 500 600 700 800 900 A self-check circuit was designed to
Wavelength, nm
verify that the flame detector did not
Port 1
sense flame when the shutter is closed
Figure 7-20 | Coal fired curve (light is blocked from the detector).
Clean Combustion Technologies
Figure: 07.20/ SPS Poor response to coal flames and
PMS5415 K frequencies that are invisible to the overall lack of reliability led to the use
human eye (<400 nm or >750 nm), as of visible light technology for flame
well as the bright blue, orange, yellow, monitoring.
or red flames that are produced in the The use of visible light frequen-
visible light range (400–750 nm). In cies occurred almost simultaneously
addition to the light frequency, the with the use of solid state electronics to
intensity of the flame, measured in replace tube type technology. Silicon car-
direct current (DC), and the flicker bide photodiodes convert light from the
frequency of the flame is important. flame to an electrical signal that can be
Controls | 7-15
Secure Network
CHANNEL
Windbox 1 2 3 4 5 6 7 8
LINK
Side Ignitor
LINK
To BMS
AUX
ALARM
Secondary
Air
To BMS
Dampers
Coal Nozzles
To DCS
Oil Gun
To PCs
ALSTOM 0000 6813 6347 4459 13032 32047 42123 36215 100 %DC 100 %DC
protective system (BPS) or a burner XXX All
DC
XXXXX
0000
0000
5805
927
5361
6813
2330
0
9071
215
19604
1525
26518
3028
23894
4601 100 %AC 100 %AC
XXX 1
Spectrum
100 100
management system (BMS). The main %Freq %Freq
0000 7126 7248 5112 11858 40848 39520 30356
XXX
AC 0000 2081 2054 1760 5332 9825 11981 8629
XXX XXX XXX XXX
XXXXX 0000 338 312 0 833 1044 1332 1290
100 %Color 100 %Color
D1 D2 D3 D4 D5 D6 D7
Relative % Color
25
5
XX
Intensity
Rate
Color
X 0
Freq 84 0
Rate
XX
To implement this strategy, the and the plant equipment, using one
normal starting and stopping func- or more operator display stations. The
tions provided for the fuel firing human-machine interface is customized
equipment are provided in a separate according to the nature of the process
control system commonly identified as to which the control and supervision
a burner control system (BCS). This is is applied. This is particularly true
implemented in this manner due to the for the BPS due to the critical nature
extensive control system redundancy of the parameters to transmit to the
requirements needed to meet interna- operator’s level. Regarding the BPS
tional codes for the BPS. monitoring and control, several facili-
The aim of the BPS is to prevent the ties are provided to allow the operator
fuel firing equipment from starting if to analyze the status of the boiler
secure conditions are not met and to safeties: a boiler safeties set of dedi-
trip the boiler if emergency conditions cated mimics, an alarm monitoring, a
arise. A boiler trip is carried out by sequence monitoring, a sequence of
stopping all fuel inputs to the furnace. events log, and a historical data storage
and retrieval function.
BPS Logic System Architecture All critical status and commands are
The BPS logic system is implemented concentrated in a single mimic allowing
in a fail-safe and triplicate modular the operator to launch and manage the
redundant system. In case of a failure, furnace purge operation, the setting
the system continues to operate in of boiler safeties, the flame detection
a safe operational state because the monitoring, and all critical parameters
critical circuits are redundant and that require a boiler trip. Notably, from
perform identical functions simulta- this mimic, the operator can follow the
neously and independently. The data furnace purge sequence progress, the
output from each of the three circuits leak test sequence progress, and the
is voted in a majority-voting circuit, correct setting of boiler safeties. Along
before affecting the system’s output. If with the physical process data, the
one of the redundant circuits fails, its interface also displays the status of the
data output is ignored and the system logic protection system itself. All warn-
continues to output to the process the ings for the operator about required
value that agrees with the majority of maintenance action in order to avoid
the functional circuits. If for any rea- any inconvenient boiler tripping are
son the failed module is not replaced displayed. In the BPS control mimic,
before a second failure in a parallel the detailed list of the data involved in
circuit occurs, the second failure will the boiler safety monitoring (permit
cause the BPS to shut down to a fail- conditions as well as trip conditions)
safe state. is displayed. This allows the operator
to identify the missing conditions that
BPS Operator Interface prevent the furnace purge from being
The human-machine interface is carried out and the boiler safeties to be
a part of the overall digital control set. It is also the means to identify, on
system that provides a direct means of a boiler trip occurrence, which is the
communication between the opera- faulty condition (first out criteria) and
tor, at the central control room level, which are the induced protection faults
Controls | 7-19
that may have happened after the digital actions than to the process
boiler trip, if any. controls. The most important design
criterion of a fuel-firing protective
Burner Management system is that it be tailored specifically
System Architecture to the requirements and operating
To meet applicable U.S.-based NFPA characteristics of the firing system.
codes, the functions required to protect
the boiler and operate the fuel firing BMS/BCS Logic Implementation
equipment are provided in a stand- The logic portion of the BMS may be
alone system that partitions boiler implemented with computer-based pro-
protection and firing system control grammable logic controllers (PLCs) or
functions from other boiler control sys- (DCS), electromagnetic relays, or solid
tem functions, (i.e., feedwater control). state electronics. All of these installa-
This protective boiler control system tions must be able to tolerate the dust,
is designated as the burner manage- temperature variations, and electrical
ment system (BMS) and is physically noise encountered in a power plant. The
and functionally separated from normal method of powering the BMS
the remainder of the boiler control logic system is via an un-interruptible
systems. The system performs critical power source (UPS).
functions like flue gas path protection, The BMS may be designed in two
furnace purge, ignition time monitor- basic ways: “energize-to-trip” and “de-
ing, ignition trials release, evaporator energize-to-trip.” A “de-energize-to-trip”
protection, oil burners trip, and coal system closes or trips the fuel-firing
mills and associated feeders trip as well equipment on a loss of power, i.e., a
as supervision of placing fuel firing fail-safe design. An “energize-to-trip”
equipment into service. system requires the presence of power
Unlike the BPS system architecture to close or trip fuel-firing equipment.
described above, the normal starting Supply of an “energize-to-trip”
and stopping functions provided for system generally increases the steam
the boiler firing equipment are an in- generator’s availability, is less prone
tegral part of the burner management to nuisance trips, and has been most
system. Like the BPS system, the BMS commonly used in the utility boiler
prevents the fuel firing equipment from industry. The type of system to be
starting if secure conditions are not provided is determined by the needs/
met and trips the boiler if emergency requirements of the end user. Where
conditions arise. A boiler trip is carried an “energize-to-trip” system is provided,
out by stopping all fuel inputs to the DC power from the plant battery
furnace. system is also provided to the BMS to
Experience has shown that the ensure the capability of shutting off
burner management system cannot the fuel input in the event of a loss of
be treated as an auxiliary function the AC power source. Systems based
of the steam generator process (e.g., on electromagnetic relays and solid
combustion, feedwater, and steam tem- state electronics, which had been the
perature) control systems. Rather, the traditional building blocks of burner
BMS is a separate and distinct system, management system logic in the past,
more closely allied to the firing-system are now rarely built.
7-20 | Clean Combustion Technologies
ily those by which fuel and air are Fuel Air/ Side
• Inadequate waterwall circulation Auxiliary Air Ignitor
admitted to the steam generator. • High furnace-gas-side pressure (warm-up oil)
Typical examples are valve opera- Fuel Air Flame
• Low furnace-gas-side pressure Auxiliary Air Scanner
tors (oil and gas firing), feeders and • Low drum level (drum units)
pulverizer motor starters (pulverized- • High waterwall outlet temperature
coal firing), air damper drives, and fan Figure 7-24 | Typical air and fuel nozzle
(once-through units) Clean Combustion
arrangements Technologies
within a windbox
motor starters (Fig. 7-24). • High separator level (once-through Figure: 07.24/ SPS
• Sensing devices include not only PMS5415 K
units)
position information on the driven • High superheater-outlet pressure
devices, but also such items as fuel (once-through units)
pressures, temperatures and flow, • Flame failure
boiler drum level, and flame pres- • Loss of logic power to BPS/BMS
ence or temperature. • Loss of primary and redundant
• The logic system constitutes the BPS/BMS processors (if applicable)
heart of the BPS/BMS. All opera- • Operator’s emergency-trip push
tor initiated commands are routed buttons depressed
to the logic system. The status of
the sensing elements is monitored
Burner Management/Burner
continuously by the logic system.
Control–Degrees of Automation
Operator commands are passed
on to the driven devices only if its The BMS/BCS arrangement dis-
logic verifies that the proper safety cussed so far is commonly called a
permits have been satisfied. On the remote-manual system. With this type
other hand, the logic system will of system, the operator initiates the
shut down equipment automati- start-up and shutdown of each individ-
cally whenever continued operation ual piece of equipment from a remote
would endanger the furnace. operating panel. The system logic
ensures that the operator commands
are performed in the correct sequence
BMS/BPS FOR TANGENTIALLY and intervenes only when required
FIRED UTILITY BOILERS to prevent a hazardous condition.
This section describes the primary func- Higher levels of automation commonly
tions and interlocks associated with a specified are “automatic” and “automatic
tangentially fired steam generator. with load programming.” An automatic
7-22 | Clean Combustion Technologies
may have accumulated in the furnace is fuel guns may be used as an intermediate
completely removed prior to initiation step. On coal-fired units, for example,
of firing. To complete this function, a ignition fuel is normally kept in service
flow of air is passed through the steam until two adjacent coal elevations are be-
generator at a minimum rating (usually ing fired at 50% of their rated capability,
30% of that required for rated steam- thereby assuring sufficient ignition en-
generator capacity) for a minimum of ergy to maintain stable firing conditions
5 minutes. At the same time, windbox during the steam-generator start-up
dampers are maintained in a particu- phase of operation.
lar configuration, the fuel-admission
devices are proven closed (or off ), and Flame Monitoring
the flame-monitoring devices indicate The Alstom tangential firing configura-
“no-flame.” This combination of condi- tion (Fig. 7-25) operates during start-up
tions will provide the proper velocities with a separate, independent flame
and number of air changes through the emanating from each operating fuel
furnace and convection pass to ensure admission point (nozzle). Each flame
the removal of any fuel accumulations. is monitored independently by the
These purge-permit requirements also use of continuous, self-flame-proving
provide a check on the proper operation ignitors or with oil guns provided with
of the air damper, fuel-admission, and discriminating flame detectors.
flame monitoring sensing devices just As soon as any main fuel elevation
prior to firing. is fired at a rate exceeding 30% of its
On completion of the purge, the design rating, the characteristics of this
steam generator is ready to be fired. firing configuration change dramati-
Side Ignitor
The 30% minimum airflow require- cally (Fig. 7-26). Fuel Gun
ment is maintained throughout the Front-fired (horizontal-burner)
operating range of the steam genera- systems exhibit, at all firing rates, the Dampers
tor including low-load operation in individual burner characteristics that
order to ensure an air-rich furnace the tangential configuration exhibits
mixture. Initial firing is accomplished at low firing rates: that is, the flame Air
with a group of ignitors that have the produced at one firing location does Plenum
capability of lighting the ignition fuel not provide ignition energy for fuel
with an electric spark. A flame detector admitted at the other locations. The
Figure 7-25 | Tangential firing pattern at low
must be provided with each ignitor to flame monitoring system for such
firing rates
determine the presence or absence of a burners must be designed accord-
stable flame. In the case of the Alstom ingly. In general, it is not appropriate
ignitor, the flame detector is an integral to use front-fired flame monitoring Clean Combustion
SideTechnologies
Ignitor
FuelFigure:
Gun 07.25/ SPS
part of the ignitor, and a fuel-flow techniques on tangential firing configu- PMS5415 K
measurement is included to ensure the rations and vice versa.
Dampers
proper quantity of ignition energy.
Depending on the type of load- CONTROL OF UTILITY
carrying fuel and the firing-system STEAM GENERATORS
Air
arrangement, the load-carrying fuel may This section describes the major con- Plenum
obtain its ignition energy directly from trol loops required to operate a utility
the ignitors or an elevation of ignition steam generator.
Figure 7-26 | Tangential firing pattern: fireball
condition
+ Total: 0
UNIT COORDINATION
UNIT LOAD 350.0 MW AGC in order to maintain the process con-
MVAR HZ
trol variables (unit generation, steam
CONTROL MODE CONTROL MODE 150 59.98 DISPATCHING AGC CONTROL
CTRL REQUEST
BLR FOLLOW BLR FOLLOW ON OFF
FIXED SLIDING REQUEST
MODE MODE
TURBINE SAFTIES
FIREBALL DETECTED
TOTAL FUEL FLOW
TURBINE SPEED
78.1
3599
m3/h
RPM
BOILER SAFETIES 11HHX31
control) and the turbine governor valve
HFO/CO SEL COORDINATION MODE BMS MODE SEL TARGET LOAD SETPOINT LOAD GRADIENT HP PRESS SP BOILER LOAD SP
control. Figure 7-27 shows a typical
350.0 8.0 166.7 89.8
HFO CO AUTO MANU
350.0
MW MW/mn
8.0 166.7
bar
89.8
%
control room operator interface for
11EGC01 11CJD00
14.9 bar
EA001 EA114 11CJD00DU100 11CJD00DU102 11CJD00DP100 11CJD00DU101
the unit load control in a coordinated-
c10 PC9152 Logged in PC9152 (127.0.0.1) User: Alstom EN DE ALSTOM 8/6/2008 13:44:01
control system.
Figure 7-27 | Typical operator interface for the unit load control of a CCS
The preceding discussion sum- This device prevents the unit from
Limit circuits prevent the marizes the basic differences between accepting a load demand change at a
system from trying to operate boiler-follow and coordinated control rate that would create instability and
type systems. The Alstom CCS con- a resulting unit trip. Turbine gover-
beyond the current capability of
tains a number of additional features nors automatically change unit steam
the plant, thus maintaining safe designed to provide several layers of flow in response to an electrical-grid
operation. defense between the direct regulation frequency disturbance. To prevent
of the variables and the automatic the CCS from fighting the governor,
shutdown actions of the burner man- and to match the boiler demand to
agement system. the turbine action, the frequency bias
Each major subsystem, such as conditioner is required. This circuit
fuel, air, or feedwater, contains a flow simply adjusts required output to
tie-back loop to ensure the process match the governor action.
variable will quickly match required Limit actions occur when the CCS
output. Typically, triple or dual re- detects that one or more process
dundant transmitters are provided for variables cannot follow the output
critical process measurements. Control adjustment demand signal. If, for
algorithms contained within the con- example, a feedwater pump were out
trol system monitor the transmitters of service, the limit circuit would not
for failure or excessive deviation and allow required output to exceed the
transfer the control loop automatically capability of the pumps in service.
to manual if a transmitter malfunc- Without this limit, firing rate would
tion occurs. When triple redundant be allowed to exceed feedwater flow,
transmitters are provided, the control resulting in loss of drum level (drum
algorithms are designed to allow failure units) or high waterwall outlet tem-
of a single transmitter within the con- perature (once-through units). The
trol loop while maintaining automatic limit circuits also act whenever such
operation. When all process loops process variables as generation and
are well tuned and calibrated, and are throttle pressure deviate from their
functioning properly on automatic, the required values by a preset amount.
controlled variables will be maintained These deviations are indications that
well within acceptable limits. Various the steam generator or turbine, or both,
limit circuits are provided to take care are unable to follow the desired output.
of the times when this desirable condi- An extension of the limit circuits,
tion does not exist. which is not shown on Figure 7-28,
Referring again to Figure 7-28, note is the run-up and run-down circuits.
that the “output adjustment signal” (the These circuits act to return the re-
desired generation value established quired output within the capability of
by the operator or the load-dispatch the unit when a change in boiler unit
system) must pass through several conditions occurs. For example, if all
conditioning devices before emerging coal feeders are operating near maxi-
as unit load demand. mum output and one feeder trips,
The first such device is the the unit cannot maintain its existing
operator-set, rate-of-change limiter. load. The run-down circuit would
Controls | 7-27
cross-limiting features being imple- Furnace pressure is controlled by the in- 25 Limit Action
mented to ensure that a safe fuel/air duced draft (ID) fans, while unit airflow
ratio is maintained under all circum- is controlled by the FD fans. ID/FD fan 0
0 25 50 75 100
stances. This system is designed to control can be accomplished with vari- Fuel Flow, %
provide the following: able inlet vanes, control dampers, blade
Figure 7-29 | The deviation limit system’s
protection limits
Clean Combustion Technologies
Figure: 07.29/ SPS
PMS5415 K
7-28 | Clean Combustion Technologies
pitch, or control of fan speed. Furnace loops that are unique to tangentially
pressure control is accomplished by ma- fired or fluidized bed steam generators.
nipulation of the ID fan(s). The furnace
pressure control loop maintains furnace Feedwater/Steam Drum Water Level
pressure to approximately a –0.5˝ WG Control (Drum Boilers)
(–0.9 mm Hg). A feed forward from In a drum type steam generator, it
FD fan demand is provided to position is critical to maintain proper drum
the ID fan control device. level at the desired level set point. To
Three pressure transmitters are used achieve this objective, typically, three
to monitor furnace pressure and are independent drum level and pressure
arranged in a median select function measurements processed through me-
block to ensure the pressure signal dian select circuits provide the process
integrity to the furnace pressure con- inputs for the calculation of drum
troller. The furnace pressure controller level. Steam drum pressure is used to
compares measured furnace pressure compensate the raw drum level mea-
to the furnace pressure set point and surement. The measured differential
generates a demand that is summed pressure of the transmitters is adjusted
with the secondary air feed-forward for the differences in specific gravity of
signal. This ID fan demand is subject the water and steam in the steam drum
to overriding action as described in the and the reference leg subcooled water.
next section. From start-up to approximately 25%
unit steam flow, steam drum water
Furnace Pressure Directional Blocking level is controlled with a single ele-
To protect the furnace from potential ment control scheme. The demand for
damage due to excessive positive or feedwater is developed in response to
negative pressures, directional blocking the deviation in drum level from drum
of the FD and ID fans is provided. level set point.
If furnace pressure goes excessively Above approximately 25% unit
low, the control system responds as steam flow, drum level control is
follows: transferred to three-element control.
In three-element control, the feed-
• Further increases in demand to the
forward signal from unit steam flow
ID fan control devices are blocked.
provides the primary index of required
• Further decreases in demand to the
feedwater flow. This feed-forward
FD fan control devices are blocked.
signal, summed with drum level error,
If furnace pressure goes excessively provides a feedwater set point. A sec-
high, the control system responds as ond controller compares this set point
follows: to the total measured feedwater flow to
develop a demand for feedwater.
• Further decreases in demand to the
ID fan control devices are blocked. CONTROL OF TANGENTIALLY FIRED
• Further increases in demand to the STEAM GENERATORS
FD fan control devices are blocked.
This section describes the major con-
For unit airflow control, refer to the trol loops that are required for control
specific section associated with control of a tangentially fired boiler.
Controls | 7-29
of the combustion air, via use of the The control scheme relies on
CC-OFA, L-SOFA and H-SOFA stoichiometry-based function
(L, low; H, high) air compartments— curves—defining the L-SOFA and
over the boiler operating envelope H-SOFA flow requirements as a
(Fig. 7-30). The TFS 2000™ firing function of total measured unit airflow.
system relies on controlled stoichio- These stoichiometry-based curves in
metric conditions to reduce in-furnace combination with SOFA flow mea-
NOx formation levels. surement are used for modulating
The OFA control strategy uses the L-SOFA and H-SOFA damper de-
cascade arranged, elevation based, mands and resultant flows to maintain
CC-OFA, L-SOFA and H-SOFA the required stoichiometry profile.
damper position ramps. The eleva- These stoichiometry-based func-
tions of CC-OFA dampers are stagger tion curves control the main burner
ramped with total measured unit zone stoichiometry demand, L-SOFA
airflow. The L-SOFA elevations and stoichiometry demand, and H-SOFA
H-SOFA elevations are stagger ramped demand as a function of total mea-
with total L-SOFA demand and total sured unit airflow. For a given total
H-SOFA demand, respectively. measured unit airflow, the L-SOFA
stoichiometry control function curve
defines the L-SOFA flow contribu-
tion to the total measured unit airflow.
The cascaded L-SOFA damper ramp
curves establish the individual L-SOFA
damper positions or demand, for a
given total measured unit airflow. The
L-SOFA flow measurement feedback
Separated
Overfire Air trims and adjusts the overall L-SOFA
demand, maintaining the desired
L-SOFA contribution to the furnace
stoichiometry.
The H-SOFA flow demand for a
Close-Coupled given operating mode and airflow is de-
Overfire Air
termined by subtracting the sum of the
L-SOFA stoichiometry contribution
plus the main burner zone stoichiom-
Main Burner etry contribution from the furnace exit
Zone
stoichiometry or equivalence ratio (i.e.,
1.20 at 20% excess air). The resultant
required airflow flows through the
H-SOFA compartments to maintain
the correct final excess air. As described
for the L-SOFA above, the cascaded
H-SOFA damper ramp curves establish
the individual H-SOFA damper
Figure 7-30 | TFS 2000™ Low NOx firing system
(indexed from boiler master demand) Therefore, feedwater flow is the selected
average primary SH spray differential feedback. Three measurements are pro-
temperature. Its output represents the vided for each of these flows, with each
required adjustment to waterwall heat measurement individually temperature
transfer/steam generation to maintain compensated. The actual value used for
both the steam conditions and flue gas control is the median of three tempera-
temperatures entering the superheat ture compensated signals.
section so as to ensure adequate spray
water control range. A second control- STEAM TEMPERATURE
ler acts on a load dependent (indexed CONTROL (GENERAL)
from boiler master demand) separator On an Alstom tangentially fired steam
outlet temperature set point corrected generator, superheat and reheat steam
by the SH spray differential tempera- temperature are controlled by a com-
ture controllers output. This controller bination of fuel nozzle tilt positioning,
acts to adjust feedwater in response to gas recirculation, gas proportioning
firing system disturbances and the rela- dampers, desuperheating spray water,
tively fast effect they have on separator adjustment of feedwater to firing rate
outlet steam temperatures. The overall ratio, and control of excess air. The
combined feedwater feedback control method of control used is dependant
action is such that feedwater demand on the type of unit provided and the
is responsive to changes in the overall fuel being fired. Superheat spray water
unit heat transfer profile. control can be accomplished with
The combined feed-forward/feedback either one or two stages of spray water
demand signal is subject to a minimum injection (Figs. 7-31A and 7-31B).
economizer inlet flow set point (wet Reheat spray water control is accom-
mode) activated if the boiler circulation plished with a single stage of spray
pump is not in service and the unit is water with the desuperheater generally
being fired. This ensures the minimum located at the cold reheat (RH) inlet
economizer inlet cooling flow is main- (Figs. 7-31C).
tained by the feedwater supply system
in the event the start-up system is not SH/RH Spray Water Control
available. The feed-forward/feedback Where a single stage desuperheater is
demand signal is subject to a second provided for either SH or RH control,
“wet mode” feedwater demand devel- the control objective of the desuper-
oped to support separator storage tank heater is to control the final steam
level control. The resulting demand outlet temperature. Desuperheating
provides the set point to a feedwater spray control is a cascade type control.
master controller. The feedback to A master controller associated with
this controller is the lower of total final steam outlet temperature develops
feedwater flow (as measured before the set point for a slave controller. The
the start-up/recirculation system) slave controller positions the final spray
or economizer inlet flow. With the valve in response to the master output
start-up system in service (wet mode), as compared to the associated spray sta-
economizer inlet flow is greater than tion outlet temperature. Figures 7-31B
feedwater flow due to recirculation. and 7-31C show typical SH/RH steam
Controls | 7-33
and does not, in most cases, initiate the Therefore, the purpose of the pre-firing
start-up or normal shutdown of firing- purge is to be sure that any unburned
system components, it does exercise gaseous fuel that may have accumu-
authority over both the operator and lated in the furnace is completely
the process controls through its safety removed prior to initial firing. To do
interlocking features. this, a flow of air is passed through
If, for example, the combustion the steam generator at a minimum
control should drop the airflow below rating (usually 30% of that required
the minimum value (typically 25%), for rated steam-generator capacity)
the BPS/BMS would trip the fuel au- for a minimum of 5 minutes. At the
tomatically. Similarly, it will not permit same time, secondary air dampers are
the operator to start equipment out maintained in a suitable position and
of prescribed safe sequences and will the fuel-admission devices are proven
shut down equipment if prescribed closed (or off ). This combination of
operating practices are not followed. conditions will provide the proper Start-up of fluid bed combustors
Removing start-up burners before the velocities and number of air changes requires raising the bed
combustor temperature reaches the through the furnace and convection
temperature to the level where
minimum set point for firing main fuel, pass to ensure the removal of any
for example, would result in the shut- gaseous fuel accumulations. These the main fuel combustion is
down of the associated main fuel. purge-permit requirements also pro- self sustaining using start-up
The specific safety interlocking vide a check on the proper operation burners while continuously
included in the BPS/BMS depends of the air-damper and fuel-admission maintaining safe operating
on the physical characteristics of the devices just prior to firing.
firing system and type of fuel or fuels On completion of the purge, the
conditions.
being fired. All safety systems of this steam generator is ready to be fired.
type, however, are concerned with the The 30% minimum airflow require-
following key safety functions: ment is maintained throughout the
operating range of the steam generator
• A pre-firing purge of the furnace
including low load operation in order
• Establishment of the appropri-
to ensure an air-rich furnace mixture.
ate permits for firing the start-up
Initial firing is accomplished with a
burners
group of start-up burners that have
• Establishment of the appropriate
the capability of lighting the warm-up
permits for firing the main fuel
fuel with an electric spark. A flame
• Continuous monitoring of firing
detector must be provided with each
conditions and other key operating
start-up burner to determine the pres-
parameters
ence or absence of a stable flame.
• Emergency shutdown of portions
Depending on the type of load-
or all of the firing equipment when
carrying fuel and the firing system
required
arrangement, the load-carrying fuel
• A post-firing purge of the steam
may be placed into service when the
generator
combustor temperature reaches the
In a fluid bed steam generator, un- auto ignition temperature of the fuel
like a PC steam generator, a portion being fired. Start-up burners remain
of the bed material is made up of sol- in service until the combustor tem-
id fuel that cannot be removed from perature reaches the point where the
the furnace during a furnace purge. main fuel is self-sustaining.
7-36 | Clean Combustion Technologies
or other subsystem at any instant in the ID fan is now being replaced only
time can be described by the Perfect by preheated air rather than by the
Gas Law as follows: products of combustion in the firing
zone. As a result, the average tempera-
PV = MRT ture of the gases resident in the furnace
(or other subsystem) at any given time
Equation 7-2
following the fuel trip will decrease
where: rapidly. Because of the temperature
drop, the pressure in the furnace starts
P = absolute pressure
to decrease (Fig. 7-33).
V = v olume of system under con-
If the resident mass were to remain
sideration
constant in quantity, the absolute
M = resident mass (not mass flow)
pressure in the furnace would drop
R = the universal gas constant
in direct proportion to the absolute
T = absolute temperature, all in
temperature drop. A gage pressure of
consistent units
–1.0˝ WG (–0.2 cm Hg) is equivalent
Because V is fixed and R is ap- to 406˝ WG absolute (75.9 cm Hg
proximately constant, P is directly absolute). Thus, with a constant resi-
proportional to the product MT. dent mass, a 10% change in absolute
Thus, for two different conditions of temperature would cause a 10% change
pressure and temperature in a given in furnace pressure; that is, 40.6˝ WG
boiler system, this approximate rela- (7.6 cm Hg absolute). Fortunately, the
tionship holds: natural characteristics of FD and ID
fans and their control devices are such
P2 M2T2
= that a constant resident mass situation
P1 M1T1
is virtually impossible to achieve.
Equation 7-3 Assume, for the moment, that all
FD and ID fan dampers are fixed
During steady state operation, P
is held constant at approximately
atmospheric pressure by balancing
Convection Stack
the resident mass M and the existing Air Pass Air
temperature T. Furnace temperature in a FD Fan Heater Heater
steam generator is not directly controlled Furnace
and depends on the thermal balance Air Side Gas Side ID Fan
between the heat in (in the burning FD Fan Curve ID Fan Curve
fuel and heated air) and the heat out
Flow 1
(in the flue gas and heat transferred to P P1 P
P2 Flow 1
P2 Flow 2 P1 Flow 2
the pressure parts). The resident mass
is automatically balanced by control- Flow P2 Pressure Profile Flow
ling the flue gas flow out of the boiler to P1 Before the Trip
2 Atmospheric
maintain a given furnace pressure. P1 Pressure
P2
When the fuel input is terminated,
this balance no longer exists. The flue Pressure Profile at the Peak of the Excursion
gas being pulled out of the furnace by
Figure 7-33 | Pressure profile in a balanced-draft unit
Clean Combustion Technologies
Figure: 07.33/ SPS
PMS5415 K
7-42 | Clean Combustion Technologies
Time
crease in furnace pressure will cause a
PF
0 X System reverse flow down the stack, back into
0 Resistance the furnace.
VG
Pressure
PF
X 0 balanced-draft unit, flue gas (with its
Time
Volume Forced Draft Fan
mass at furnace temperature) will be
pulled out of the furnace until the
VA
Volume Flow
0
Gas
Change
VG
System PF that is directly related to the cutoff
Resistance
pressure of the ID fan correspond-
VA ing to the temperature entering the
fan. The higher the head capability
0 X
Time of the ID fan, the larger the pressure
excursions that may be experienced.
Figure 7-34 | Destructive negative pressure (DNP) with constant speed, centrifugal fans and
dampers locked
Clean Combustion Technologies
Figure: 07.34/ SPS
PMS5415 K
Controls | 7-43
Thus, although some authors have of gas through them. These devices
attempted to treat the ID fan char- have similar control characteristics
acteristics and the fuel cutoff as two (as compared to outlet dampers or
independent effects, they are actually variable-speed drives), but can have
inseparable. widely varying time constants, de-
With respect to PC and oil firing, pending on the type of actuator used
the difference is one of the rate of to move them.
temperature change. This rate de- Figure 7-35 shows the effect on the
pends on how fast a given percentage pressure/volume curve of such a fan
of the furnace volume undergoes a when either vanes or dampers are
replacement of combustion products closed to reduce the gas through-flow.
with preheated air. Coal-fired units On this curve, fan static-pressure capa-
have the following characteristics that bility is plotted against rated volume,
produce a slower temperature decay: and a system-pressure curve is shown,
a simple square relationship of static
1. The fuel cutoff is generally more
pressure to volume (static pressure is
gradual than with oil firing,
proportional to volume squared).
because of the pulverized fuel
The control device is regulated so
residual in the pulverizer and
that the required static pressure to
fuel piping.
satisfy the resistance of the physical
2. The ratio of resident mass to mass
system is developed at any given volu-
flow rate is higher than for an
metric flow rate. But, at the same time,
oil-fired boiler, because coal-fired
for the particular vane position, a fan
furnaces are significantly larger in
characteristic curve is established such
size than oil-fired only furnaces.
that the fan is capable of an infinite
number of discrete static-pressure
THE ID FAN values versus any given volumetric flow.
In connection with the implosion Consider an operating mode in
phenomenon, it is most important which the ID fan control vanes or
that engineers responsible for draft- dampers are capable of infinitely fast
system and furnace-framing-system response. Any change in the furnace
design understand how an ID fan pressure of a balanced-draft unit that
operates, how it is controlled, and
how it responds to changes in the 120 Static Pressure –
inlet pressure that it “sees.’’ Vanes or Dampers
ID fans are used singly, or in mul- 100 Wide Open
Rated Static Pressure, %
is sensed by the system will cause the given position, either accidentally or
vanes to change position to maintain purposely, furnace suction will vary,
the required system static pressure with the maximum static-pressure
exactly, and along the system-pressure capability of the fan being the limit.
curve. For instance, if there should be a Under these conditions, of course, FD
reduction in furnace pressure (caused, airflow continues, but is automatically
say, by a decrease in the fuel fired varied with boiler firing rate.
and the corresponding combustion Assume that a master fuel trip
airflow), the vanes will immediately occurs under this condition of vanes-
close to reduce the gas flow handled maintained-in-position. Assume
by the fan before the furnace pressure further that the FD fan dampers close
drops any further. With such infinitely quickly, so that there is a rapid decrease
fast vane control assumed, even with in the gas available to the ID fan. The
an instantaneous very large drop in ID fan will immediately ride up along
furnace pressure level (such as there the characteristic curve to the cutoff
could be if there were a master fuel point; after reaching the cutoff point,
trip or if the FD fan dampers were to reverse flow of gas down the stack will
accidentally close), the system curve occur (at some indeterminable flow
will always be followed and the ID fan rate) until the furnace suction that has
will make instantaneous corrections to been established by the decrease in the
maintain furnace pressure at the level average gas temperature is satisfied.
preset (below atmospheric). If the rate of reverse flow is not high
Another condition can be postu- enough, the furnace will fail structur-
lated, in which the control vanes or ally if the framing system is not capable
dampers are maintained in a fixed of withstanding the low absolute pres-
position as boiler load varies. This is sure that results from the fuel trip.
equivalent to a zero rate of damper Of course, what happens during a
response. When boiler load decreases master fuel trip, in which equipment
(and fuel flow, airflow, and gas flow de- is on automatic control, is somewhere
crease correspondingly), the developed between the above two postulated sit-
fan static pressure increases following uations. Fan dampers or vanes close in
a static-pressure versus volume curve an effort to reduce the gas flow leaving
sloping upward to the left; furnace the furnace, restricting the magnitude
suction (pressure below atmospheric) of the furnace pressure level reduc-
increases rapidly. The obvious limit tion. With good mechanical response
to the furnace suction that can be to the control signal (particularly if
developed is the highest static-pressure an “early” signal has been received
value on the curve, which for damper through the use of a feed-forward
settings below 50%, appears to be at or arrangement), fan static-pressure
about the fan no-load cutoff value (the capability will match the furnace
static head developed by a fan when its suction being developed as a result of
volumetric delivery is zero). the reduction in furnace gas tempera-
Under normal load-changing condi- ture, all while the volumetric gas flow
tions, if fan vanes are maintained at a through the fan is being reduced to
Controls | 7-45
A wide variety of materials are used in the construction of fossil-fueled steam generators,
ranging from structural steel support members to refractory ceramic insulation to wear-resistant
tungsten carbide composites in coal pulverizers to low- and high-alloy steels for high-temperature
pressure parts. The latter materials are the focus of this chapter, both because the capabilities
of alloys for pressure parts in the furnace and backpass set limits on the maximum
steam pressure and temperature (and, hence, efficiency) of the unit and because these
components tend to be the most combustion process, whether directly tubes exposed to the combustion
prone to failure in service as a result by radiation from the fireball or from environment.
of the demanding environment in hot, flowing combustion gases. These 3. Ease of fabrication, including
which they operate. This chapter fundamental requirements translate bending, forging, and welding.
presents an introductory descrip- into the following set of specific mate- Good fabricability usually is a
tion of the metallurgy, processing, rial requirements: trade-off with increased strength:
properties, and applications of alloys High-strength alloys tend to be
used for pressure parts in fossil-fueled 1. Strength at low and high more difficult to fabricate, requir-
boilers, and common failure modes temperatures to resist stresses ing greater care and precision.
observed. The chapter concludes with induced by high-pressure water 4. Appropriate thermal and
an overview of recent developments in and steam. For steam piping, physical properties for efficient
alloys for advanced supercritical steam headers, manifolds, and super- heat transfer and reduced
plants with high efficiency. heater and reheater tubing, the stresses associated with thermal
resistance to long-term creep transients. Efficient heat transfer
Functional Requirements for deformation and rupture is a criti- requires high thermal conductiv-
Pressure Part Materials cal property. ity, whereas transient thermal
Pressure parts in steam generators are 2. Compatibility with the expected stresses are reduced by both high
subject to highly demanding operating service environment. This thermal conductivity and low
conditions. The key functional require- includes resistance to oxidation and coefficient of thermal expansion.
ments for these components are to corrosion by high-temperature 5. Overall cost, including raw
contain water or steam at high pressure water or steam on the inside material costs, costs of fabri-
and temperature and, for waterwall, of tubes and pipes as well as cation, and costs of any heat
superheater, and reheater tubes, resistance to high-temperature, treatment required. A more
to absorb heat released during the fireside corrosion on the outside of expensive (per unit of weight),
8-
8- | Clean Combustion Technologies
A3 Line
An alloy may contain one or more
phases, depending on the composition
1,200 Ferrite (�) A1 Temperature Spheroidizing Annealing and
and temperature. The phase diagram in
Hardening Figure 8-1 depicts the phases that are
thermodynamically stable at a defined
1,000 temperature and carbon content in
Ferrite (�) and Carbide (C) a simple iron–carbon steel. At low
temperatures and carbon content above
800 approximately 0.025% (the solubility
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Carbon, % limit of carbon in ferrite), steel consists
of a mixture of two solid phases, ferrite
Figure 8-1 | Iron-rich section of the iron–carbon phase diagram
Temperature, °F
the bainite forms. Bainite is not as hard alloy may be controlled by combina-
and forms at slower cooling rates than tions of mechanical processing and heat
martensite does. If formed as part treatment. The ideal grain size for an
of a heat treatment operation, both alloy will depend on the specific proper-
of these phases are subsequently ties required; for example, a fine grain
tempered (i.e., heat treated below the size may give good low-temperature
A1 temperature) to reduce hardness tensile strength but reduced elevated-
and increase ductility. Much slower temperature creep strength.
cooling from the austenite phase field
will produce the equilibrium structure Fabrication Processes
of ferrite and cementite, which has a A number of fabrication processes
much lower hardness. are used to create boiler components,
With this knowledge, it some- such as waterwalls, steam drums,
times is possible to use the structure headers, and superheater assemblies,
observed in post-mortem metallurgi- from stock tubing, piping, and plate
cal examination of a pressure part received from steel manufacturers.
to estimate the temperature that the These processes include bending,
component has experienced in service, rolling, forging, and welding, and
which can be useful in failure mecha- depending on the alloy, they may
nism diagnosis. enhance or degrade the properties.
Heat treatment may be performed
Grain Structure either before or after fabrication to
A grain is a single crystal within a solid enhance or restore the structure
composed of many similar crystals. and properties.
Within any particular grain, all of the
atoms are arranged onto a single crystal Bending, Forging, and Rolling
lattice, but, at the boundary between Fabrication processes such as bend-
two adjacent grains, a transition zone ing and forging of tube and pipe
exists that is not directly aligned with and rolling of plate are collectively
Figure 8-6 | Comparison of fine (top) and
either grain. The grain boundary tran- referred to as either “cold working” or
coarse (bottom) grain structures
sition zone represents a crystal lattice “hot working,” depending on whether CleaninCombustion
an austeniticTechnologies
stainless steel
defect, which frequently is a point of the operation is carried out near room Figure: 08.06/
100× A.R.
magnification
PMS5415 K
8- | Clean Combustion Technologies
temperature (cold working) or at high base metals) and the high heat input
temperatures (hot working). Working of welding, which can greatly alter the
(deforming) metals at low temperatures structure of the base metal adjacent to
usually increases the hardness (resistance the weld (the so-called “heat-affected
to further deformation) and reduces zone”). High stresses also exist in the
ductility. Because of the work-hardening vicinity of weld joints in the as-welded
phenomenon, only limited deformation condition because of the constrained
can be performed at low temperature shrinkage of the weld metal and the
without reannealing the alloy to restore surrounding heat-affected zone. Even
its ductility. Typical cold-working in the absence of defects, the weld
operations include tube bending and thus represents a metallurgical (and,
roll-expanding tubes into header sockets. frequently, geometric) discontinuity in
Much greater deformation can be the structure of the component and is
imparted, using lower forces, if the alloy a common location for cracking. At-
is worked at high temperature. For tention to quality-control procedures
steels, this typically is in the austen- during welding processes is critical.
ite phase field, above approximately
1,600°F (870°C). The ductility is much Heat Treatment
greater, and continual annealing and Heat treatment is used to alter the
recrystallization occur during the structure and, hence, properties of
forming operation. Typical hot-working metals and can be used to strengthen,
operations include rolling of plate into soften, or improve ductility. Table
cylindrical sections and induction bend- 8-1 presents a list of common heat
ing of pipe. treatment procedures used on steels
for steam generator pressure parts.
Welding The specific times and temperatures
Pressure parts usually are joined by for each treatment type will depend
fusion welding, a process in which two on the alloy being treated. As with
metal pieces are joined by locally melt- welding, attention to quality-control
ing the contacting surfaces, which then procedures during heat treatment are
cool to a continuous solid. A filler metal critical, because some alloys are very
usually is added to the joint to fill in sensitive to small variations in heat
the gap between the pieces. Variations treatment conditions.
of fusion welding processes include the
heat source used for melting, the nature Steels for Fossil-Fueled
of the filler metal used (if any), and Steam Generator
the shielding gases used to protect the Pressure Parts
molten weld metal from oxidation. A wide range of steels is utilized in
A full discussion of welding the construction of fossil-fueled steam
processes and metallurgy is outside generator pressure parts, ranging from
the scope of this book. The key idea low-carbon steels used in the economiz-
to remember in terms of pressure ers and waterwalls to austenitic stainless
parts in steam generating equipment, steels used in high-temperature super-
however, is that welds have unique heat and reheat sections. These steels
microstructures and properties because may be roughly grouped into different
of the cast nature of the weld metal (as classes based on alloy content. Table 8-2
opposed to heavily worked, “wrought” lists the common steels used for steam
Table 8-2 | Steels Used in Fossil-Fueled Steam Generator Pressure Parts*
Product ASME Min UTS Min YS Chemical Composition Limits, % by weight
Nominal Composition EN Designation
Form Specification Grade ksi (MPa) C Mn P S Si Ni Cr Mo V Other
Low-Carbon Steel
Seamless
SA-192 47 (325) 26 (180) 0.06–0.18 0.27–0.63 0.035 0.035 0.25
Low Strength tube
Welded tube SA-178 A 47 (325) 26 (180) 0.06–0.18 0.27–0.63 0.035 0.035
Seamless
SA-210 A-1 37 (255) 60 (415) 0.27 0.93 0.035 0.035 0.10 min
tube
Medium Strength
Welded tube SA-178 C 37 (255) 60 (415) 0.35 0.80 0.035 0.035
Pipe SA-106 B 35 (240) 60 (415) 0.30 0.29–1.06 0.035 0.035 0.10 min 0.40 0.40 0.15 0.08 Cu: 0.40 max
Pipe SA-106 C 40 (275) 70 (485) 0.35 0.29–1.06 0.035 0.035 0.10 min 0.40 0.40 0.15 0.08 Cu: 0.40 max
High Strength Plate SA-299 A 40 (275) 75 (515) 0.28 0.84–1.62 0.035 0.035 0.13–0.45
Plate SA-515 70 38 (260) 70 (485) 0.35 1.30 0.035 0.035 0.13–0.45
Low-Alloy Ferritic Steel
Seamless
C-0.5Mo SA-209 T1 30 (205) 55 (380) 0.10–0.20 0.30–0.80 0.025 0.025 0.10–0.50 0.44–0.65
tube
Seamless SA-213 T12
1Cr-0.5Mo 13CrMo4-5 30 (205) 60 (415) 0.05–0.15 0.30–0.61 0.025 0.025 0.50 0.80–1.25 0.44–0.65
tube, pipe SA-335 P12
Seamless SA-213 T11
1.25Cr-0.5Mo 10CrMo5-5 30 (205) 60 (415) 0.05–0.15 0.30–0.60 0.025 0.025 0.50–1.00 1.00–1.50 0.44–0.65
tube, pipe SA-335 P11
Seamless SA-213 T22
2.25Cr-1Mo 10CrMo9-10 30 (205) 60 (415) 0.05–0.15 0.30–0.60 0.025 0.025 0.50 1.90–2.60 0.87–1.13
tube, pipe SA-335 P22
Advanced Ferritic Steel
W: 1.45–1.75
Nb: 0.02-0.08
Seamless SA-213 T23
2.25Cr-1.5W-Mo-V-Nb 7CrWVMoNb9-6 58 (400) 74 (510) 0.04–0.10 0.10–0.60 0.030 0.010 0.50 1.90–2.60 0.05–0.30 0.20–0.30 N: 0.030 max
tube, pipe SA-335 P23
Al: 0.030 max
B: 0.0005–0.006
Nb: 0.06-0.1
Seamless SA-213 T91
9Cr-1Mo-V-Nb X10CrMoVNb9-1 60 (415) 85 (585) 0.07–0.14 0.30–0.60 0.020 0.010 0.20–0.50 0.40 8.0–9.5 0.85–1.05 0.18–0.25 N: 0.030–0.070
tube, pipe SA-335 P91
Al: 0.040 max
W: 1.5–2.00
Nb: 0.04–0.09
Seamless SA-213 T92
9Cr-1.75W-Mo-V-Nb X10CrMoWVNb9-2 64 (440) 90 (620) 0.07–0.13 0.30–0.60 0.020 0.010 0.50 0.40 8.5–9.5 0.30–0.60 0.15–0.25 N: 0.030–0.070
tube, pipe SA-335 P92
Al: 0.040 max
B: 0.001–0.006
Austenitic Stainless Steel
Seamless
18Cr-8Ni SA-213 TP304H X6CrNi18-10 30 (205) 75 (515) 0.04–0.10 2.00 0.045 0.030 1.00 8.0–11.0 18.0–20.0
tube
Cu: 2.5–3.5
Nb: 0.30–0.60
Seamless SA-213 Super304H
18Cr-8Ni-3Cu-Nb 34 (235) 86 (590) 0.07–0.13 1.00 0.040 0.010 0.30 7.5–10.5 17.0–19.0 N: 0.05–0.12
tube S30432
Al: 0.003–0.030
B: 0.001–0.010
Seamless
18Cr-10Ni-Nb SA-213 TP347H X7CrNiNb18-10 30 (205) 75 (515) 0.04–0.10 2.00 0.045 0.030 1.00 9.0–12.0 17.0–19.0 Nb: 8xC–1.10
tube
Seamless
25Cr-20Ni SA-213 TP310H 30 (205) 75 (515) 0.04–0.10 2.00 0.045 0.030 1.00 19.0–22.0 24.0–26.0
tube
Seamless Nb: 0.20–0.60
25Cr-20Ni-Nb-N SA-213 TP310HCbN 43 (295) 95 (655) 0.04–0.10 2.00 0.045 0.030 1.00 19.0–22.0 24.0–26.0
Materials for Fossil-Fueled Steam Generators | 8-
tube N: 0.15–0.35
*ASME, American Society of Mechanical Engineers; EN, European Norm; UTS, ultimate tensile strength; YS, yield strength. EN equivalents are approximate and indicative only.
8- | Clean Combustion Technologies
generator pressure parts, their nominal levels of quality. The ASME boiler
The various codes and compositions, their applicable material and pressure vessel committees use
standards used for equipment specifications, and their minimum the ASTM materials specifications to
tensile properties. establish recognized grades of materi-
designs in power plants als to which their code rules for design,
throughout the world were Specifications and Codes allowable stresses, fabrication methods,
originally introduced as The grades of steels used in the and inspection apply. Both ASME
voluntary safety standards. construction of pressure parts are and ASTM material specifications are
specified by governing codes—for therefore equivalent.
example, the American Society of To supplement the rules of the
Mechanical Engineers (ASME) Boiler boiler and pressure vessel codes, a
and Pressure Vessel code in the United number of other standards control
States and the European Norm (EN) the design and fabrication of associ-
codes in Europe. The ASME code ated components and systems. These
originated as a safety code to pre- include ASME B31.1 Power Piping,
vent boiler explosions, and today, it and ASME B16.34 Valves Flanged,
promotes the acceptance of properly Threaded and Welding End. The Euro-
designed equipment on a national and pean equivalent of ASME B31.1 is EN
international basis by assisting in the 13480, Metallic Industrial Piping.
standardization of safety rules. The A basic limitation of all codes is that
rules of the code are kept up to date their rules must be expressed in suf-
by continuing review. Section I, Power ficiently general terms in order to cover
Boilers, of the ASME code provides the a wide range of applications. Because
basic design rules and requirements, they define only reasonable minimum
whereas Section II, Materials, pro- standards, cases in which a power plant
vides specifications, properties, and designer must make further investiga-
design-allowable stresses for materi- tion and exercise additional effort are
als used in boiler and pressure vessel sure to arise. A great variety of supple-
construction. Similar design rules and mentary information is available from
material specifications in the EN code publications of the various engineering
are found in EN 12952, Water Tube societies and from technical bulletins
Boilers, and EN 10216, Seamless Tubes of equipment suppliers.
for Pressure Purposes.
Another technical group, the Ameri- Carbon Steels
can Society for Testing and Materials Carbon steels are the predominant
(ASTM), prepares and publishes materials used in boiler pressure part
specifications for the purchase, test- applications because of their low cost,
ing, and examination of a variety of excellent fabricability and weldability,
materials. Similarly, the committees of and good combination of properties at
the American Welding Society write low and somewhat elevated tem-
specifications for welding practices and peratures. In addition to carbon, these
consumables. Such specifications are steels typically contain manganese
widely used to procure and acceptance- and silicon in small amounts (≤1%).
test materials in order to meet defined Manganese combines with harmful
Materials for Fossil-Fueled Steam Generators | 8-
residual sulfur in the steel to form be- in strength and ductility if the graph-
nign manganese sulfide inclusions, and ite nodules concentrate and align, as
it also contributes somewhat to solid can happen along heat affected zones
solution strengthening. Silicon is used associated with welds.
in the steelmaking process as a deoxi-
dizer and degasifier and can improve Low-Alloy Ferritic Steels
the oxidation resistance of the steel. In terms of temperature capability, the
Carbon steels for boiler appli class above carbon steels is low-alloy
cations are further classed into steels, which contain chromium and
low-strength, medium-strength, and molybdenum. This class, the so-called
high-strength grades, with commen- “Cr-Mo steels,” encompasses several
surate increases in carbon content. alloys containing chromium in levels
Carbon contents range from 0.1 to ranging from 0.5 to 2.25% and molyb-
0.2% in low-strength carbon steels, denum in levels ranging from 0.5 to
from 0.2 to 0.3% in medium-strength 1%. Molybdenum is a strong solid so- Carbon steels and low-alloy
carbon steels, and from 0.3 to 0.35% lution strengthening element, and 0.5% ferritic steels are the
in high-strength carbon steels. In- molybdenum originally was added principal materials for power
creased carbon content gives a higher to carbon steels in order to improve
fraction of cementite or pearlite in the the high-temperature creep–rupture
plant equipment in order
steel, producing higher strength at the strength. This C-½Mo steel was to maintain reasonable
cost of reduced ductility and weld- prone to graphitization, however, and costs as multi-tonnage
ability. Still higher strengths can be chromium subsequently was added quantities are required.
obtained by further carbon additions, to improve the high-temperature
but such steels are seldom found in oxidation resistance and to prevent
pressure parts because of weldability graphitization.
and ductility considerations. Further increases in high-temperature
The use of carbon steels as pressure strength and oxidation resistance were
parts in steam generating equipment realized by further chromium and
ultimately is limited by their high- molybdenum additions. The 2¼Cr-
temperature strength and resistance to 1Mo composition (Grade 22) gives
oxidation and corrosion. The maxi- an optimal combination of strength,
mum use temperature of carbon steels weldability, and oxidation resistance for
is approximately 850°F (454°C), water- and steam-touched tubing and
which typically restricts their use to piping in the temperature range of 900
water-touched applications, such as to 1,050°F (482–565°C). Above this
feedwater piping, economizers, and range, however, oxidation limits the use
furnace wall tubing. Carbon steels of these steels, and alloys with greater
also are prone to degradation of their chromium content must be used, such
microstructure at high temperatures as 9Cr-1Mo-V-Nb steel or austenitic
through a process termed graphitiza- stainless steels.
tion, wherein iron carbides slowly
convert to graphite (pure carbon) and Advanced Ferritic Steels
iron. This process can be accompanied The desire for a ferritic steel with signifi-
by a commensurate, marked reduction cantly greater strength and oxidation
8-10 | Clean Combustion Technologies
Engineering Stress
bon content and a coarser grain size. ing this period of elastic behavior, no 0.2% Yield Strength X
0.2%
Stainless steels typically are used as permanent deformation occurs. Strain is Plastic
Elastic Region
superheater and reheater tubing subject directly proportional to the stress. Region
to the highest steam temperatures or Continuing to increase the loading
severe fireside corrosion conditions. results in plastic (permanent) strains Engineering Strain
Despite their attractive combination and eventual breakage of the test X = Specimen Failure
of creep strength and environmental specimen. The yield strength, which has
Figure 8-7 | Stress–strain curves for two
resistance, these alloys have drawbacks several different definitions, describes
that prevent their more widespread the engineering stress (load divided by Cleansteels with different
Combustion strengths
Technologies
and ductilities.
Figure: 08.07/ TB Steel A is weaker
use in boilers. They are consider- the original cross-sectional area of a test PMS5415but more
K ductile than steel B
ably more expensive than low-alloy specimen) in the vicinity of this elastic-
steels because of the high concentra- to-plastic behavior transition. The
tions of chromium and nickel. Their tensile strength is the engineering stress
thermal expansion coefficients also when the maximum load during the
are considerably higher than those of test is reached. Ductility is a measure of
ferritic steels, which limits their ap- the amount of plastic deformation that
plication in thick-walled components a material will sustain before breaking.
such as headers and piping. Austenitic It usually is expressed as a percentage
stainless steels also are subject to stress elongation or reduction in cross-sec-
corrosion cracking in certain aqueous tional area of the test specimen. How
environments, making them unsuitable much plastic strain a steel will exhibit is
for application as water-touched tubing quite variable and generally is inversely
and piping. related to the tensile strength of the ma-
terial. High-strength steels typically will
Mechanical Properties exhibit much less ductility than softer,
Certain mechanical properties of steels lower-strength steels.
affect their fabrication and service. Most of the specifications covering
Among these are tensile properties at the steels used in the construction of
low and elevated temperature, hard- a boiler require room-temperature
ness, toughness, and creep–rupture tensile testing with minimum speci-
strength. The allowable stresses used fied values of tensile strength, yield
for the design of pressure parts are strength, and ductility. For design
derived from these properties. purposes, however, the tensile and yield
strengths of each grade of steel used
Tensile Properties need to be determined at temperatures
The standard tensile test provides up to the creep range of the grade.
data concerning yield strength, tensile Various organizations, including the
strength, and ductility. When metals producers and fabricators of steel,
are pulled (stressed) with a uni-axial, perform such tests, and the data from
increasing load, the material stretches these tests are used by codes to set
(strains). Diagrams such as the one maximum allowable stresses.
8-12 | Clean Combustion Technologies
steel are established from test data. establish rupture strengths of 13,000,
In this figure, creep-rate data for 7,000, and 3,300 psi (90, 48.3, and
1,000, 1,100, and 1,200°F (538, 593, 22.8 MPa), respectively.
and 649°C) are plotted versus stress. Table 8-3 lists the criteria used by
The intersection of these data lines the ASME and EN codes to determine
with that for 0.01% over 1,000 hours the allowable stresses as a function of
sets the respective creep strengths temperature. Figure 8-9 applies these
of 7,800, 5,000, and 2,400 psi (53.8, criteria to establish the allowable stress
34.5, and 16.6 MPa). The other for 2¼Cr-1Mo steel as an example.
curves in the figure show rupture life The relative influence of the different
versus stress for the same three tem- criteria varies with materials as well as
peratures. Lines through these data with temperature. Figure 8-10 shows the
are extrapolated to 100,000 hours to effect of temperature on ASME Boiler
Code allowable stresses for a number of
Table 8-3 | simplified criteria for Determination of Allowable alloys used for high-temperature service.
Stresses by American Society of Mechanical Engineers (ASME) and The carbon steels begin to lose strength
European Norm (EN) Methods above 700°F (371°C), and by 850°F
(454°C), these steels are down to
ASME Methodology EN Methodology
about one-half their room-temperature
Non-creep regime, minimum of: Non-creep regime, minimum of:
• 67% Yield strength • 67% Yield strength
values. The low-chromium ferritic al-
• 29% Tensile strength • 42% Tensile strength loys start to lose strength above 800°F
Creep regime, minimum of: Creep regime, (427°C) and are down to half-strength
• 67% of average stress to rupture in • 80% of average stress to rupture in at approximately 1,000°F (538°C).
100,000 hours 200,000 hours The austenitic stainless steels decline
• 80% of minimum stress to rupture in -or-
100,000 hours • 67% of average stress to rupture in
somewhat from room temperature to
• 100% of average stress to produce a 100,000 hours (if 200,000-hour data 1,000°F (538°C) because of reduction
creep strain of 0.01% over 1,000 hours are not available) in yield strength; above 1,000°F, creep
and rupture strength cause a rapid
decrease to half-strength or less by
20,000
1,200°F (649°C).
18,000
16,000 Oxidation and Corrosion
14,000 of Steels
12,000 As described above, the allowable
Stress, psi
Penetration, inches
0.004
have been measured for each alloy in is exfoliation (spalling) of oxides
0.003
controlled conditions, and as shown from the tube walls, which may col-
0.002
in Figure 8-11, the rate of growth lect in bends and block the flow of
0.001
decreases parabolically with time. For steam through a circuit or may pass
0
a given alloy, oxidation rates increase into the steam turbine and cause 0 4,000 8,000 12,000
with increasing temperature. Oxidation erosion of the turbine blades and Time, hours
rates decrease with increasing chro- vanes. Figure 8-12 shows an oxide
mium in an alloy; stainless steels show layer attached to the base metal of a Figure 8-11 | Oxide penetration (metal loss)
Clean Combustion
versus timeTechnologies
for 11⁄4Cr-1Mo steel
very low rates of steam oxide growth. superheater tube. Figure 8-13 shows Figure: in
08.11/
steamSPS
at 1,100°F (593°C)
Because of the predicable nature of a magnification of the tube with a PMS5415 K
steam oxidation rates, the thickness of double oxide layer. Factors that cause
steam scales found on the inner tube exfoliation are complex; thermal
walls can be used to estimate the local quenching of the oxide during
metal temperature, assuming that the start-up and shutdown operations
number of operating hours is known. is a potential cause. Limiting the
This can be useful in determining the thickness of steam side oxide scale by
cause of boiler tube failures. selecting an appropriate alloy for the
Other than an increase in tube specified steam temperature also will
wall temperature, the primary reduce the tendency for exfoliation.
8-16 | Clean Combustion Technologies
combustion products also may suffer ferent mechanisms; Table 8-4 presents
greater rates of fireside corrosion. In a list of mechanisms and the compo-
these areas, local chemically reducing nents in which they are observed. A
conditions allow enhanced attack by failure analysis investigation frequently
reduced sulfur species as well as carbu- is performed to verify the specific
200 microns rization, which can disrupt protective mechanism, which allows for a better
Figure 8-13 | Magnified tube cross section oxide scales. understanding of the root cause of
showing two layers of oxide scale Alloys containing greater concen- the failure and the actions that can
trations of chromium have a greater be taken to prevent future failures. A
resistance to most forms of fireside cor- full description of the various failure
rosion. For this reason, stainless-steel mechanisms and their features is out-
tubes with 18% chromium or greater side the scope of this book; readers are
Clean Combustion Technologies
Figure: 08.13/MH
are used for superheater and reheater referred to the Electric Power Research
PMS5415 K sections in units where aggressive fire- Institute’s handbook on boiler and heat
side corrosion conditions exist. In some recovery steam generator tube failure
particularly severe environments, alloys for additional information.
with 25% chromium may be necessary.
Waterwall panels also may suffer Materials for Advanced
from fireside corrosion, but stain- Supercritical Plants
less steels cannot be used for tubing During the last decade or so, renewed
because of the potential for stress emphasis on increasing the efficiency
corrosion cracking. For this applica- of fossil-fueled power plants has ap-
tion, protection can be provided by peared. A significant feature of this
the application of corrosion-resistant higher efficiency is that the level of
coatings over the low-chromium emissions per unit of power output
ferritic tubing. The most common is reduced; for example, an increase
Materials for Fossil-Fueled Steam Generators | 8-17
in efficiency of 2 percentage points Table 8-4 | Typical Boiler Tube Failure Mechanisms
corresponds to a 5% reduction in CO2
Applies to
emissions. Moreover, because CO2
Failure Mechanism Water- Steam-
capture and storage consumes energy Special
touched touched
that otherwise would contribute to case
tubes tubes
the overall power output, increased Short-term overheating X X
efficiency of generation will reduce the
Long-term overheating/creep X X
future impact of capture-and-storage
Corrosion fatigue X
methods on the cost of electricity to
the consumer. Hydrogen damage X
Consequently, a serious worldwide Fatigue cracking X X
effort is now underway to develop Fireside corrosion X X
fossil-fueled power plants operating Corrosion-enhanced thermal fatigue X
at higher steam temperatures to give Low-temperature creep crack growth X
greater generation efficiencies, and
Acid dew point corrosion X
significant progress has already been
Acid phosphate corrosion X
made toward development of the
advanced materials that are necessary Caustic gouging X
for operation at these temperatures. Stress corrosion cracking X
Figure 8-14 shows the substantial Pitting X
improvements in efficiency that can Dissimilar metal weld cracking X
be gained by operation at higher Fly ash erosion X
temperatures and pressures. The cur-
Soot blower erosion X
rent state-of-the art conditions are
Shockwave cracking X
main steam temperatures of 1,110°F
(600°C) and pressures in excess of Strain-induced precipitation hardening X
Piping
Superheater
Loop
Waterwall
Panel
Figure 8-16 | Location of COMTES 700 components within the Schloven F boiler
Clean Combustion Technologies
Figure: 08.16/ (RC)
PMS5415 K
Water Technology
Chapter Nine
Water is a basic engineering material used in the production of steam for power
generation and process use. Effective treatment or conditioning of water has made
possible more efficient steam cycles in the subcritical and supercritical pressure ranges.
Many industrial processes have special requirements for high-quality water that is free of
objectionable contaminants. These statements are indicative of the significant role of water
technology as an interdisciplinary professional field, which comprises elements of basic
chemistry as well as elements of required before raw water is suitable This chapter will discuss spe-
chemical and mechanical engineering. for introduction into the cycle (make- cific boiler applications of water
Water supplied to an industrial site up water) depends largely on the end technology, but many of the same
can originate from a river or stream, use of the steam, the steam purity considerations hold for various types
mountain runoff, wells, or a lake. The requirements, and the boiler design. of process operations. In practically
composition of impurities in water The pressure of a boiler generally every instance, a careful review of
varies over a wide range. Water may still is used to classify boiler designs. sources of water and of its condition-
be contaminated with chemical or Therefore, the water quality neces- ing is required to avoid economic
organic impurities, bacteria, dissolved sary for a particular application has, losses because of a slowdown in pro-
gases, suspended mineral matter, and by convention, been related to boiler duction, damage to equipment, and
dissolved solids of both scaling and pressure (and steam purity). This re- increased costs of operation.
non-scaling composition. lationship between boiler pressure and
When using water to generate steam water/steam purity has been reflected TECHNIQUES OF
for process or electric power, consider- in the guidelines of the American Boiler WATER TREATMENT
ation must be given to water treatment Manufacturers Association (ABMA). It The extent of water conditioning or
in order to prevent corrosion of the is recognized, however, that present-day treatment depends on both the original
construction materials, scaling of heat- water chemistry guidelines regarding supply source and the ultimate end use
transfer surfaces and other flow-related plants providing steam for electric gen- of the product. Table 9-1 lists common
components, and contamination of eration must reflect the steam purity impurities found in water, sets forth
steam. This involves treatment of the requirements as specified by the steam some of the resulting problems, and
raw water that is available on that turbine and gas turbine (if steam- or indicates common treatment methods.
particular site and introduced into the water-injected) supplier, because these The techniques of water treatment
cycle as well as conditioning of the require lower dissolved solids concen- cover a wide variety of raw-water
water and steam while being utilized in trations in the steam than are called for preparation schemes. This is covered in
the cycle. The extent of the treatment in the ABMA guidelines. detail elsewhere.1–5
9-
9- | Clean Combustion Technologies
Adapted from Betz Laboratories, Inc., Betz Handbook of Industrial Water Conditioning, 8th ed. Philadelphia, PA.: Betz Laboratories, 1980.
Table 9-2 | Analyses of Typical Surface and Ground Waters in the United States
Analysis
1 2 3 4 5 6 7 8 9 10 11 12 13
Number*
Silica (SiO2) 2.5 0.4 2.3 8.2 13 8.0 8.4 9.6 16 23 34 12 39
3+
Iron (Fe) 0.03 0.05 .09 .12 0.04 0.15 0.04 0.0 2.1 .09
Calcium (Ca)2+ 5 .3 27 32 1.7 72 79 40 3.4 7.2 70 26 72 7.2
Magnesium
1.7 7 10 0.4 6.4 28 16 1.5 2.5 24 10 33 4.2
(Mg)2+
Sodium (Na)1+ 1.4 3† 3.5 1.9 41 99 94 5.6 147 12† 138 358† 7.5
1+
Potassium (K) 0.6 1.0 0.7 4.7 4 0.4 1.6
Carbonate (CO3)2– 0 1 0 0 0 4 0 0 0 0
Bicarbonate
10 99 138 3 174 137 46 21 328 179 170 293 50
(HCO3)1–
Sulfate (SO4)2– 11 13 17 4.4 138 290 298 3.3 2.6 135 70 560 0.8
1–
Chloride (CI) 2.6 7 6.5 2.6 9.5 79 14 3.6 51 8 139 195 6.8
Fluoride (F)1– 0.1 0.1 0.6 0.4 0.4 0.1 0.2 0.8 2.5 0.0
1–
Nitrate (NO3) 0.3 0.2 0.2 4.0 0.2 3.6 0.7 0.0 0.0 1.1 0.0
Dissolved solids 34 130 171 23 386 661 554 42 392 392 503 1380 90‡
Total hardness as
20 98 121 6 206 315 166 15 28 276 106 316 35
CaCO3
Noncarbonate
6 16 8 3 64 197 128 0 0 126 0 76 0
hardness
Specific conduc-
tance (µS/cm 53.4 263 29.5 575 1,040 822 55.5 651 867
at 25°C)
Color 1 3 15 5 6 4 10 0 2.5
pH 6.9 8.1 8.2 5.8 7.7 8.4 7.6 7.0 8.0 7.6 7.9 7.5
* Analyses numbers are identified as follows. Values are in parts per million where this unit is appropriate:
1 = New York City, NY, Catskill supply (reservoir—finished). 10 = Jacksonville, FL, Well 1,064 ft deep.
2 = Detroit, Ml, Detroit River (raw). 11 = El Paso, TX, Well 703 ft deep.
3 = Chicago, lL, Lake Michigan (raw). 12 = Galesburg, IL, Well 2,450 ft deep.
4 = Fitchburg, MA, Pond (finished). 13 = Bremerton, WA, Anderson Creek.
5 = Omaha, NE, Missouri River (raw). † Computed by difference in epm and reported as sodium.
6 = Los Angeles, CA, Colorado River (raw). ‡ Sum of determined constituents.
7 = Pittsburgh, PA, Monongahela River (finished). µS/cm = micro Siemens/cm
8 = Macon, GA, Ocmulgee River (raw). epm = equivalents per million
9 = Houston, TX, Well 1,932 ft deep.
sulfide. Increasing the temperature, the chemicals form a floc, which assists
aeration time, and the surface area of in agglomerating impurities. Settle-
water improves the removal of gases. ment of the particles permits a clear
effluent from the coagulating chamber.
COAGULATION Figure 9-2 illustrates a typical clarifier.
Adding coagulating chemicals reduces Removal of colloids requires careful
surface-water contamination in the analysis of the impurities to establish
form of coarse suspended solids, silt, the nature of their electrical charge,
turbidity, color, and colloids. The one of the principal factors that
Water Technology | 9-
FILTRATION Distributors
Filters separate coarse suspended mat-
ter from raw water or remove floc or
sludge components from coagulation
or process softening systems. Gener-
Staggered
ally, gravity and pressure-type filters are Slat Trays
used for this purpose. Graded gravel
or coarse anthracite are the materials
commonly used in the filter bed. Dia-
tomaceous earth and special pre-coat
filters generally are used to remove oil
and reduce color in feedwater make-up.
Damper Air Inlet
CHEMICAL SOFTENING PROCESSES Blower
Reverse Osmosis
Figure 9-1 | Forced Draft Deaerator
Reverse osmosis is based on the
principle of osmosis. In osmosis, a
membrane separates two solutions
containing different amounts of
dissolved chemicals. The membrane Secondary Mixing and Reaction Zone Draft Tubes
Chemical Chemical
allows some compounds (like water)
Clarified Water
to pass through, but it does not allow Effluent Clean CombustionClear Water
Technologies
Escape Surface
Figure: 09.01/ (MM)
larger compounds to do so (i.e., a
Rotor-Impeller PMS5415 K
semi-permeable membrane). Pressure Raw Water
differences cause pure water to pass Concentrator
through the membrane from the dilute Return
Flow Zone
to the more concentrated solution, as Discharge Hood
shown in Figure 9-3. The pressure is Primary Mixing
called osmotic pressure. The natural and Reaction Zone
tendency is for water to move through Blow-Off and Drain Slurry Pool Indicated by Shaded Areas
the membrane from the dilute to the
Figure 9-2 | Typical clarifier
9- | Clean Combustion Technologies
concentrated solution until the chemi- regenerated with acid; the anion
cals reach equal concentrations on both exchange material is regenerated with
Dilute Concentrated sides of the membrane. caustic. If the cation resin is desig-
In reverse osmosis, pressure is ap- nated as Z and the anion material as
plied to the concentrated side of the R, the simple reactions in a two-stage
membrane (i.e., the contaminated side). demineralizer may be expressed as
Semi-Permeable This forces the osmotic process into
Membrane Water Flow reverse so that with adequate applied Cation: H2Z + CaSO4 → CaZ + H2SO4
pressure, pure water is forced from Anion: H2SO4 + R(OH)2 → R(SO4) + 2H2O
Figure 9-3 | In osmosis, water moves across the concentrated (contaminated) side
the membrane
Clean Combustion from the dilute to
Technologies to the dilute (treated) side. Treated Demineralization can yield a pure
Figure:the09.03/ (JG)
concentrated solution (Data
PMS5415 fromK“Water Treatment Notes:
water is collected in a storage container. water that is equal or superior to the
Reverse Osmosis Treatment The rejected contaminants on the best evaporated vapor. The anion
of Drinking Water,” Cornell concentrated side of the membrane and cation resins can be arranged in
Cooperative Extension, New York
are drained as wastewater. Figure 9-4 various combinations to produce the
State College of Human Ecology)
shows the process of reverse osmosis. best water most economically. Two-,
three-, or four-bed units, or a single
DEMINERALIZATION mixed-bed demineralizer, can be used
In demineralization, ion exchange to accomplish the required result.
removes ionized mineral salts. Cations Figure 9-5 illustrates possible combi-
such as calcium, magnesium, and nations of these systems to produce a
sodium are removed in the hydrogen specific water quality. Because strong
cation exchanger, and anions such base exchangers are temperature sen-
as bicarbonate, sulfate, chloride, and sitive, they should not be used above
soluble silica are removed in the anion 120°F (49°C). Cation-exchange
exchanger. Synthetic cation- and resins can tolerate temperatures of
anion-exchange resins are used in 250°F (120°C).
demineralization of water. Sulfonic, Use of finely sized resins in demin-
carboxylic, and phenolic resins are eralizers requires a water source that
used for cation exchange and amino is low in turbidity and organic matter.
or quaternary nitrogen for anion Clarification, filtration, and chlorina-
exchange. The cation exchanger is tion generally are required to reduce
organic matter to a low level. Failure
to reduce the organic level in the water
Feed Water produces a marked reduction in the
Storage Tank Under Pressure capacity of the exchangers because
of fouling of the resins. The advance
of demineralization to provide a
Treated Water Rejected Contaminants high-purity make-up water may be
(diluted) (concentrate)
attributed in part to the develop-
ment of highly basic anion-exchange
Water Flow Semi-Permeable Waste Stream
resins, which allow the removal of
Membrane to Drain soluble silica from raw water. Mixed-
bed demineralizers, in which the
Figure 9-4 | In reverse osmosis, pressure is applied to the concentrated solution, reversing
the natural directionClean Combustion
of flow and forcingTechnologies
water across the membrane from the
concentrated solutionFigure: 09.04/
into the more(JG)
dilute solution (Data from “Water
PMS5415
Treatment Notes: Reverse K Treatment of Drinking Water,” Cornell
Osmosis
Cooperative Extension, New York State College of Human Ecology)
Water Technology | 9-
Calcium carbonate has a very limited solubility; thus, the precipitated particles
will agglomerate at the heated surfaces to form a scale. Other substances, such as
calcium sulfate, also produce scales. In this case, the scaling mechanism is a func-
tion of a retrograde solubility, or a decrease of solubility in water with an increase
in temperature. These substances have a low thermal conductivity, and if left
untreated, even a thin scale will overheat the boiler tube.
pH Value at 77 °F
ppm because of its corrosivity at higher proceeds in water as follows: 11.0
10.6
concentrations. With demineralized- 10.2
quality boiler water, caustic (present HPO2– – –
4 + H2O → OH + H2PO 4 9.8
9.4
either as a treatment chemical or a 0 40 80 120 160 200
contaminant) can reach high concen- The latter reaction is complete at PO4 Concentration, ppm
trations because of the absence of large low pH (i.e., 5–7) but is less than 0.1%
quantities of other competing ions. at pH 10. Thus, dibasic phosphate Figure 9-9 | Coordinated phosphate pH
control
Whirl and Purcell7 recognized neither hydrolyzes nor dissociates in Clean Combustion Technologies
the ability of sodium phosphates to the normal boiler water pH range. Figure: 09.09/(DE)
suitably alkalize and control pH in Additions of disodium phosphate have PMS5415 K
boiler water without the negative little effect on pH in the range of 10 to
reaction associated with the use of 11. Monobasic phosphate dissociates
caustic. They developed the pH control to the dibasic form, and the addition of
method known as “the coordinated one mole of monosodium phosphate
phosphate pH control.” This method (NaH2PO4) is capable of neutralizing
is illustrated by the relationship of pH one mole of sodium hydroxide.
and the phosphate concentration in As discussed in later sections, the
boiler water, as shown in Figure 9-9. use of monosodium phosphate and
Values noted on the curve represent even disodium phosphate is limited,
pH values obtained by dissolving because under certain conditions, a
stoichiometrically pure trisodium potential exists to form acid phosphate
phosphate (Na3PO4). Conditions compounds, which can be corrosive.
below the curve represent solutions of Trisodium phosphate therefore is the
trisodium phosphate and disodium treatment of choice to ensure boiler
hydrogen phosphate (Na2HPO4). water pH control in drum-type units.
The area above the curve represents
solutions of trisodium phosphate and SIGNIFICANCE OF
caustic. In the coordinated phos- IRON–WATER REACTION
phate pH control, specifications can A number of influences characterize
be maintained by adding trisodium the reaction of boiler steel and water:
phosphate (Na3PO4), disodium phos-
• The chemical constituents present
phate (Na2HPO4), or monosodium
in solution
phosphate (NaH2PO4). Combinations
• The quantity of dissolved gases
of trisodium phosphate and disodium
• The structural characteristics of the
phosphate are preferred.
steel
Alkalinity control is attained by the
• The rate of heat input to the boiler
addition of phosphate ions to water to
metal
produce a captive quantity of hydrox-
ide (OH)– by the reversed hydrolysis Although the reaction is under-
reaction as follows: stood, there generally is insufficient
information to explain the mechanism
PO3– – 2– in exact detail. Many theories have
4 + H2O → OH + HPO 4
been forwarded to clarify this problem,
9-10 | Clean Combustion Technologies
and despite a lack of fundamental data, product and inhibits reactivity. The
experience and experimental evidence control of this reaction has been well
have permitted the development of established in the protection of metal
controls that satisfactorily contain the surfaces existing ahead of the boiler.10
corrosive activity. This section will deal The reaction of water and steel is
with existing techniques developed to spontaneous and rapid at high boiler
understand and control reactions in temperatures. The only reason that
boilers. boiler steel can survive normal operat-
The fundamental reaction of iron ing conditions is that the corrosion end
and water is to produce iron hydroxide product, magnetite (Fe3O4), forms a
and hydrogen as follows: protective barrier on the metal surface
to stifle further corrosion.
Fe + 2H2O → Fe(OH)2 +H2 In the simplest analysis, the function
of alkalinity control is to maintain an
It is an established fact that the environment in which this barrier is
end product of the reaction in boil- stable and protective. The objective of
ers is magnetite (Fe3O4). The control water treatment in boilers is to protect
of corrosion therefore is based on the the barrier against the aggressive action
knowledge of the rate-controlling step of impurities introduced into the boiler
in the overall reaction. with the feedwater. The work of Berl
Schikorr8 established a mechanism and Van Taack11 (Fig. 9-10) has been
that accounted for the production of used to relate corrosion of steel over
magnetite. In the simplest form, the a range of pH values. In interpreting
reactions would be: the results of Figure 9-10, it may be
concluded that the protective oxide
Fe + 2H2O → Fe(OH)2 +H2 is solubilized at pH values below 5.0
3Fe(OH)2 → Fe3O4 + 2H2O +H2 and above 13.0. Minimum corrosion is
indicated at pH values of 9.0 to 11.0.
From a consideration of physi- It should be noted that in actual plant
cal and chemical relationships, it can feedwater, other parameters, such as
be shown that the formation of iron oxygen and dissolved solids or salts in
hydroxide is the rate-controlling step the water, can influence the solubility
in the Schikorr hypothesis. Therefore, of the protective oxide or corrosion of
the rate of the overall reaction is based the metal.
Approximate pH Values at 25 °C
on the solubility and stability of this
1 2 3 4 5 7 9 10 11 12 13 14 product. The initial reaction above is ROLE OF OXYGEN
150 addition of alkalinity would reverse the water. Oxygen can react with iron hy-
100 reaction to the left. droxide to form either a hydrated ferric
50 Corey and Finnegan9 found that oxide (hematite) or magnetite. This
0 iron placed in contact with deaerated reaction is a function of the oxygen
3,650 400 10,000 100,000
365 40 200,000 and chemically pure water will produce concentration, pH, and salt content (or
36.5 4 4,000 40,000
3.65 0.4 20,000 an equilibrium pH of approximately cation conductivity) of the water.
HCl, 0.36 0
ppm NaOH, ppm 8.3. Increasing the alkalinity reduces Dissolved oxygen in water can cause
the solubility of the iron corrosion general or localized corrosion. Oxygen
Figure 9-10 | Relative general corrosion rate
of carbon steel versus pH
Clean Combustion Technologies
Figure: 09.10/ SPS
PMS5415 K
Water Technology | 9-11
reacts with hydrogen ions at the metal less than 0.010 ppm. To achieve a low
surface and depolarizes the surface residual, it is necessary to12: Control of oxygen in the boiler
locally, causing iron to dissolve. If the water is critical to avoiding
• Exclude air leakage into the con-
area becomes progressively anodic, it
permits more iron to dissolve, gradu-
denser corrosion problems. Efficient
• Judiciously control the addition of boiler operation requires control
ally creating a pit. The most severe
undeaerated water to the conden-
corrosion action occurs when a deposit of oxygen in the feedwater.
sate or feedwater
covers a small area. Creation of a dif-
• Prevent the addition of aerated
ferential aeration cell about the deposit
heater drips into the condensate
can lead to a severe local action. The
• Ensure the exclusion of air into the
metal beneath the deposit is lower in
feedwater cycle during short out-
oxygen compared with areas surround-
ages of the boiler
ing it, becomes anodic, and is attacked.
Pitting is most prevalent in stressed Boiler start-up is a likely source for
sections of boiler tubing, such as at a certain amount of oxygen infiltra-
welds and cold-worked sections and at tion into the feedwater. Normally,
surface discontinuities in the metal. pressure in a deaerator is not attained
Generally, oxygen is excluded or until steam is admitted to the turbine
removed from the feedwater of any and bleed steam is available for heat-
power plant. Conditions exist, however, ing. It is possible to introduce more
in which the cation conductivity (or oxygen into the boiler at this time
salt content) is very low (not to exceed than it will be in a few months of
0.3 µS/cm), so a small amount of normal operation. Admitting auxiliary
oxygen (<1 ppm) can actually reduce steam to a deaerator to pressurize the
the degradation of the protective oxide. unit to between 3 and 5 psig (0.2 to
Under these conditions, oxide layers 0.35 bar) can prevent much of this
consisting of hematite and magnetite problem, but the inclusion of auxiliary
are formed. These layers form an inter- boilers or steam in a plant cycle is
locking structure, which have a much rarely done today. Plants that are
lower corrosion product release rate. designed to cycle and predict frequent
This mechanism forms the basis for a starts and stops have been equipped
specialized treatment called “oxygen- with vacuum-type deaerators to better
ated treatment” (OT), which will be control oxygen ingress.
discussed in a later section.
Power plants employing tight cycles Removal of Residual Oxygen
to prevent oxygen infiltration and Chemical agents added to the feed-
condenser leakage generally are free water or, in some cases, to the boiler
of corrosion problems. On the other water are available to react with small
hand, many cycles are vulnerable to residual quantities of oxygen. These
uncontrolled oxygen leakage into are known as “oxygen scavengers.” Two
the feedwater as a result of design or types of scavengers exist; volatile, such
operation. Efficient boiler operation as hydrazine (N2H4) or its substitutes,
requires the control of oxygen in the and non-volatile, such as sodium sulfite
feedwater. The normal guaranteed (Na2SO3). The use of hydrozine has
value of oxygen leaving the deaerating been problematical and, therefore, has
heater or a deaerating condenser is been limited. However, it still provides
9-12 | Clean Combustion Technologies
the best example when discussing the its ability to reduce the oxidized form
general behavior of oxygen scavengers. of copper. In this state, the general
Sodium sulfite usually is added to the corrosion of the copper alloy surfaces
boiler water in low-pressure boilers is reduced in the pre-boiler cycle.
that typically use softened water as Copper oxide is reduced with hydra-
make-up. Sodium sulfite, when added zine at temperatures as low as 150°F
to the boiler water, reacts with oxygen (66°C) (Iron oxide, or Fe2O3, also can
to form sodium sulfate as follows: be reduced at a temperature of 250°F
[120°C]).
2Na2SO3 + O2 → 2Na2SO4 Normally, hydrazine or its substi-
tutes, when needed, are added to the
Temperature affects the reactiv- cycle at the outlet of the condensate
ity and stability of sodium sulfite (as pump at a rate not to exceed a residual
well as hydrazine). Generally, a cobalt of approximately 10 ppb at the inlet
catalyst is added with the sodium of the economizer. Reducing agents
sulfite solution to speed up the reaction aid in curbing corrosion, especially
at lower temperatures. Sulfite decom- copper alloys (oxygen > 10 ppb and
position increases with temperature cation conductivity > 0.3 μS/cm),
and local concentration in boiler water. but they will not prevent metal attack
Sodium sulfite decomposes in boiler everywhere when oxygen is present
water as follows: in the boiler feedwater. At higher
temperatures, the reaction rate of
Na2SO3 + 2H2O → 2NaOH + H2SO3 oxygen with steel exceeds that of the
H2SO3 → H2O + SO2 reducing agents.
an optimum fit between their chemi- Table 9-4 | Stability of Volatile Water-Treating Materials
cal and physical characteristics (e.g.,
Chemical Formula Pressure (psig) Temperature (°F) % Decomposed
volatility, solubility, and stability) and
Ammonia NH3 4,270 (29 MPa) 1,202 (650°C)
the particular application.
Table 9-4 lists the stability char- Cyclohexylamine C6H11NH2 4,270 (29 MPa) 1,202 (650°C) 88
acteristics of volatile amines. Both Morpholine O(CH2)2NH 4,270 (29 MPa) 1,202 (650°C) 100
morpholine and cyclohexylamine Hydrazine N2H4 4,270 (29 MPa) 1,202 (650°C) 7,100
have temperature stability limits.
Ammonia, hydrogen, and carbon
decomposition products are formed or binding many cations (calcium
in the dissociation of these amines at and magnesium hardness as well as
high temperatures. Ammonia is stable heavy metals) into a soluble organic
and has been used at steam tempera- ring structure. Chelants are in the
tures as high as 1,200°F (650°C). As a neutralized, sodium salt form. They
result, ammonia is the recommended hydrolyze in boiler water to an or-
compound for use in controlling the ganic anion. Full hydrolysis depends
pH of condensate in high-pressure, on a relatively high pH level. The
high-temperature boiler systems. anionic chelant has reactive sites that
It is important to maintain ammonia attract coordination sites on cations.
levels generally below 1 ppm in systems These coordination sites are areas on
containing copper alloys to avoid the ion that are receptive to chemi-
copper alloy attack, because a copper cal bonding. In this manner, cations
ammonium complex is formed in the (hardness salts) entering the boiler as
presence of high ammonia concentra- contamination from the condensate
tions. The reaction can be further system combine with the chelant to
accelerated by the presence of abnor- form a stable chelate. Deposition of
mal concentrations of oxygen. Typical hardness on boiler internal surfaces
reactions are as follows: may therefore be prevented.
Although many substances have
Cu+ + 2NH3 → Cu(NH3)+2 chelating properties, ethylenediam
Cu2+ + 4NH3 → Cu(NH3)2+
4 inetetra acetic acid is the most suitable
for boiler feedwater treatment. As with
It has been established that an am- any internal chemical treatment, how-
monia concentration of less than 0.5 ever, the application and limitations
ppm will ensure a protective pH in the must be clearly understood. Because
cycle without attacking copper sur- of their thermal decomposition at
faces. Morpholine and cyclohexylamine saturation temperatures correspond-
are preferred to ammonia for pH ing to higher pressures, chelants are
control in situations where excessive limited to boiler pressures of less than
decomposition of these chemicals does 1,000 psig (0.7 MPa) and to those
not occur and an extensive condensing boilers typically using softened make-
system or process piping network exist. up water. In addition, overfeeding of
chelants can lead to corrosion of boiler
USE OF CHELANTS FOR tubes, because chelation of the protec-
INTERNAL CONTROL tive magnetite (Fe3O4) layer can occur
Chelants are weak organic acids (For a more detailed discussion, see
with the capability of complexing references 1–5.).
9-14 | Clean Combustion Technologies
Boiler Water–Related 1 2 3 4 5 6 7 8 9
Corrosion Damage 1 in
occurs, the tube normally fails by rates. As heat is transferred through the
thick-edge fracture before the wall boiler tube wall to a water–steam mix-
thickness is reduced to the point ture, however, a temperature gradient is
at which stress rupture would established, and the internal tube metal
occur. Hydrogen gas, produced temperature must reach a value slightly
in the corrosion reaction, diffuses higher than the bulk fluid temperature.
through the underlying metal, As boiler water is converted to steam,
causing fissuring of the structure. the dissolved solids concentrate in a
Brittle fracture occurs along the residual film, as shown in Figure 9-20.
partially separated boundaries, and The solids concentration in this film
in many cases, an entire section increases until the boiling point of the
is blown out of an affected tube. solution is elevated to the tempera-
Examples of these failures are ture at the tube wall. It is improbable
shown in Figures 9-18 and 9-19. that such high concentrations can be
Figure 9-16 | Crack-like corrosion
Ź}Ź~
penetrations halfway through reached in a clean tube under normal
ČńōčŅŊĤŷĥ
the tube wall, strongly Internal metal-oxide deposits, which boiling conditions. Theoretically, the
ʼnňŅʼn
suggesting that propagation permit boiler water solids to concentrate bulk of the film-temperature drop is
results from separate time
periods of pitting
during the process of steam genera- experienced within a laminar film of
tion, accelerate both types of attack. approximately 0.001 inch (0.025 mm)
When the boiler water contains highly in thickness. Free diffusion of ions
soluble alkaline treatment chemicals, into the turbulent fluid would prevent
such as sodium hydroxide or potassium the establishment of a concentration
salts, ductile attack is more probable. gradient. Only when porous internal
Hydrogen damage, on the other hand, deposits are formed in areas of heat
is more likely to occur when a low-pH absorption is it possible to produce very
boiler water environment results from high concentrations. The deposit itself
condenser leakage or some other type of acts as a diffusion barrier and, in so
system contamination.13 doing, increases the probability of high
Figure 9-17 | Ductile gouging According to an American Society local concentrations being reached.
of Mechanical Engineers and Edison Because porous deposits may exist,
Ź}Ź~
ČńōčŅŋĤŷĥ Electric Institute (ASME/EEI) spon- boiler water chemicals should be of a
ʼnňŅʼn
sored laboratory research program,14,15 type that will not become highly cor-
clean tubes are not susceptible to cor- rosive when concentrated. Experience
rosion even under high heat-transfer indicates that sodium hydroxide and
conditions unless unusually high con- potassium salts are particularly objec-
centrations of acid or alkali are present. tionable. Alkalinity can be produced
In addition, fouled tubes do not appear with sodium phosphate (coordinated
to be susceptible to corrosion when phosphate treatment) without the risk
contaminants are present if proper of extremely high pH values being
boiler water treatment is maintained. present on concentration. Control of
These observations can be ex- boiler water pH also may be achieved
plained by the “concentrating film” by volatile amines, such as ammonia,
theory. The Berl and Van Taack curve entering with the feedwater (required for
FigureŹ}Ź~
9-18 | Thick edge waterwall tube
failure caused by hydrogen (see Fig. 9-20) illustrates the relative once-through units).
ČńōčŅŌĤŷĥ
ʼnňŅʼn
damage corrosion of carbon steel under vari- Condenser cooling waters represent a
ous acid and alkaline conditions. Bulk potential source for contamination of the
boiler water concentrations are always boiler water. Depending on the chemical
well within the range of low corrosion constituents dissolved in the cooling
Water Technology | 9-17
Flow-Assisted Corrosion
Flow-assisted corrosion (FAC) of carbon
steel is defined as the acceleration or
Figure 9-19 | Another typical waterwall failure caused by hydrogen damage
increase in the rate of corrosion caused
Clean Combustion Technologies
by the relative movement between a cor- Figure: 09.19 (AR)
rosive fluid (either single-phase water or two-phase water and steam mixtures) and K
PMS5415
a metal surface. The dissolution of the protective magnetite layer is enhanced by the 750 NaOH, 106
�Tm = 85°
Temperature, °F
652° - 50,000
will increase the loss of metal as iron oxidizes to replace the dissolved film, repeatedly.
700 104
Flow-assisted corrosion imparts a distinct surface pattern, often called a “scalloped
Film
surface,” as indicated in Figure 9-21. This physical feature helps to distinguish FAC, or Temperature 103
675
metal “wastage,” from other wastage corrosion mechanisms. Flow-assisted corrosion Gradient
is not a mechanical abrasion- or erosion-related mechanism. It is a chemical/corro-
650 650° - 100 102
sion process mitigated by specific water-chemistry conditions within a particular fluid Saturation - 2,200 psig
�Tf = 15°
temperature range. Fluid temperature and fluid flow mechanics influence the location
in boiler, or heat recovery steam generator (HRSG) components, where FAC could
preferentially occur. A brief review of these parameters follows.16,17 Figure 9- 20 | Illustration of “concentrating
film” theory.Technologies
Clean Combustion Film temperature
pH and Oxidation–Reduction Potential Figure:gradient of 15°F (8.3°C) is
09.20/ (JG)
PMS5415developed
K because of heat
The pH and the oxidation–reduction potential (ORP) of the fluid affect the solu- transfer. Soluble solids in boiler
water are concentrated in
bility of iron oxide/hydroxide, as typically represented by the Sweeton and Baes
the surface film during steam
equilibrium reaction: generation to produce “non-
boiling equilibrium.” Although
⅓Fe3O4 + (2 – b)H+ + ⅓H2 ↔ Fe(OH)(2 – b)+
+ (4∕3 – b)H2O the caustic concentration in the
b
boiler water is only 100 ppm,
the concentration in the surface
Ferrous hydroxide, on the right-hand side of the equation, is the soluble prod- film must equal 220,000 ppm
uct. As indicated by this equation, the dissolution of magnetite is influenced by to elevate the boiling point by
15°F (8.3°C)
the ORP of the solution, as represented by the hydrogen molecule (H2). It also is
influenced by the pH, as represented by the hydrogen ion (H+). Thus, lowering
the pH and /or the ORP (toward negative potential or reducing conditions) will
increase the dissolution rate of the protective oxide layer and accelerate corrosion.
Normally, oxygen or oxidizing conditions would be thought to cause addi-
tional corrosion, and in the presence of dissolved solids, corrosion is increased.
9-18 | Clean Combustion Technologies
This factor was used to determine the sessment of the potential for FAC,
concentration of ammonia remaining because the absolute corrosion rates
in the liquid film of water. The pH of have a high degree of uncertainty.
the liquid was based on this calculated The process for normalizing the data
value. Within the two-phase region, (to derive an index) utilized field
the most conservative approach was experience (HRSGs and conventional
taken by assuming that the oxygen boiler plants) as well as data from
concentration resides entirely within erosion–corrosion studies already
the steam. The calculated values for the referenced. The index values of greater
two-phase regions are based on 0 ppb than 1 indicate a potential for tube life
of oxygen in the film of liquid water on degradation, whereas an index value
the tube wall. of less than 1 indicates no significant
As stated earlier, the intent of the reduction in tube life. Some of the
model is to provide an index or as- model predictions or FAC potentials
for specific areas are illustrated in
Figures 9-25 and 9-26.
Single Phase - Economizers
Geometry Factor = 0.23
Flow-assisted corrosion has oc-
curred in combined cycle plants where
10.00
organic amines and oxygen scavengers
(or hydrazine) are used for pH control
in feedwater and low-pressure boiler
FAC Index
hydrazine or equivalent) will increase pH = 8.8 Index > 1 indicates a potential for FAC and
substantially the potential for FAC. corrosion rates can increase substantially as the
ORP is driven toward the negative potential.
FAC Index
It also is indicated, as commonly is
1.00
agreed, that at a pH of greater than
9.3, the potential for FAC is low, even
Transition Zone
in a reducing environment. pH = 9.4
0.00
TUBE FAILURES CAUSED BY Negative Potential -100 mV Positive Potential
Oxidation Reduction Potential (ORP)
OVERHEATING
The preceding discussion on ductile Feedwater Oxygen = 0 ppb Feedwater Oxygen = 10 ppb
gouging and hydrogen damage (under Figure 9-27 | Flow-assisted corrosion (FAC) index versus oxidation–reduction potential (ORP)
deposit corrosion mechanisms) focused Clean Combustion Technologies
on the role played by internal deposits in Figure: 09.27/ (JG)
PMS5415 K
the corrosion of boiler tubes, but deposits Figure 9-30 illustrates the temperature
also promote the majority of tube fail- profile under a film-boiling condition.
ures caused by overheating. Figure 9-28 The reduced film coefficient increases Radiant Heat
illustrates the temperature conditions the ∆T across the film to 310°F Q = 100,000
existing in a typical furnace wall tube. (170°C), and the metal temperatures rise
BTU/Hr-Ft2
The heat flux of 100,000 BTU/h-ft2 correspondingly. It should be noted that
(315,000 w/m2) is established by the the flux is constant at 100,000 BTU/h-ft2 TF > 2,800 °F
750 °F
furnace flame temperature spectrum, (315,000 W/m2) for all cases.
and the bulk fluid temperature of 660 °F
640°F (338°C) is fixed by the boiler STRESS RUPTURE 640 °F
pressure. The temperature profile is The strength of carbon steel tubing Nucleate Boiling
then established by the ∆T required to is nearly constant to approximately
drive the 100,000 BTU/h-ft2 (315,000 840°F (444°C), at which point it be- Figure 9-28 | Furnace wall tube—nucleate
W/m2) flux through each segment of gins to lose strength rapidly. If the tube boiling
Clean Combustion Tçhnologies
resistance. With nucleate boiling on a metal temperature is increased gradually Figure: 09.28/(DE)
PMS5415 K
clean tube, the ∆T across the fluid film beyond this temperature, it will plasti-
Internal Deposit
is quite small (10°F [5.5°C]), and the cally deform (creep) and then rupture.
overall temperature differential (90°F The approximate time to rupture is a Radiant Heat
[50°C]) results from the resistance of function of the hoop stress and of the
Q = 100,000
the tube metal. temperature (Fig. 9-31). The hoop stress
Figure 9-29 shows what happens to (S) equals the product of the internal BTU/Hr-Ft2
the tube temperature profile if an inter- pressure (P) times the tube radius (r) 810 °F
TF > 2,800 °F
nal deposit exists. Approximately 60°F divided by the wall thickness (x): 720 °F
(33°C) is required to drive the 100,000 660 °F
640 °F
BTU/h-ft2 (315,000 W/m2) heat flux S = P( r⁄x)
through the deposit, so both the inside Nucleate Boiling
and outside metal temperatures must Simple visual examination of a With Internal Deposit
rise correspondingly. failed section reveals a great deal
Figure 9-29 | Furnace wall tube—nucleate
Clean Combustion Technologies
boiling with deposit
Figure: 09.29/(DE)
PMS5415 K
9-22 | Clean Combustion Technologies
about the tube temperature history. In history. The normal pearlitic micro-
deposit-caused overheating, local tube structure of carbon steel boiler tubing
1,040 °F temperatures rise gradually, and spe- can be seen in Figure 9-33. The light
cific points begin to creep before others areas are iron, and the dark areas are
950 °F
(blistering). Deposits tend to form pearlite, consisting of alternate layers
non-uniformly, and even a 20°F (11°C) of iron and iron carbide. The normal
Radiant Heat
difference in temperature can have a operating temperature of a boiler tube
Q = 100,000 pronounced effect on localized creep. will not alter this structure. If the
BTU/Hr-Ft2 Blistering is almost always caused by temperature of steel is maintained in
640 °F internal deposition, and conversely, the range of 930 to 1,340°F (500 to
TF > 2,800 °F
Film Boiling overheating because of deposits will 730°C) for an extended period of time,
almost always be evidenced by blisters. carbon migrates from the pearlite to
When an upset of boiling condi- form spheres of carbon. This process,
Figure 9-30 | Furnace wall tube: film boiling
tions, such as departure from nucleate known as “spheroidization,” progresses
Clean Combustion Tçhnologies
Figure: 09.30/(DE) boiling, elevates tube temperature, slowly at 930°F (500°C) and rapidly
Killed CarbonKSteel (0.12–0.17C)
PMS5415 a temperature excursion will occur. at temperatures near the lower critical
140,000 840 °F 1,000 °F 1,200 °F Metal temperature will jump quickly range (1,340°F [730°C]).
from the normal condition to a higher Figure 9-35 shows carbon steel
Stress, lb/in2
14,000
value, consistent with the upset. If the with various degrees of spheroidi-
1,400
temperature excursion is great, the zation. If the tube is heated to a
1400 °F tube may fail very quickly. For instance, temperature of greater than 1,340°F
140 if a tube is completely blocked dur- (730°C) and is then allowed to cool
100 1,000 10,000 100,000
Rupture Time, Hours ing operation, the tube temperature slowly to below 930°F (500°C), the
will rise immediately to the failure- normal pearlitic structure will be
Figure 9-31 | Stress and temperature versus point temperature. The rapid plastic re-established. If the tube is heated
time to rupture deformation of these very short-term to above 1,340°F (730°C) and then
Clean Combustion Technologies
Figure: 09.31/(DE) ruptures produces very thin-edged cooled very rapidly, however, a mar-
PMS5415 K fractures. In the case of major flow tensitic structure (Fig. 9-35) will be
interruptions, the tube bulges over an produced. This type of microstructure
extended length and fails like a balloon is likely to be found in very short-term
bursting. If a departure from nucle- overheating failures, in which the water
ate boiling excursion places the tube escaping through the rupture suddenly
metal temperature in the range of 840 quenches the overheated tube metal.
to 1,020°F (450 to 550°C), the tube will
take a longer time to fail. During this BOILER WATER AND FEEDWATER
period, inter-granular oxidation of the TREATMENT IN THE PREVENTION
metal will occur, and this will weaken OF CYCLE CORROSION
the structure. Long-term failures, there-
fore, tend to be thick edged. The higher DRUM-TYPE UNITS: Utility
the temperature and stress conditions BOILERS and HRSGs
are above the creep limit, the shorter the The water chemistry guidelines for
time to failure and the thinner the metal drum-type boiler are provided in
edge at the fracture. Figure 9-32 presents Tables 9-5 through 9-11. Phosphate
examples of a short-, an intermediate-, treatment is the prime choice for
and a long-term failure. HRSG drum-type boiler water. All-
Metallographic analysis is an impor- volatile treatment is an alternative for
tant tool for analyzing tube temperature special cases.
Figure 9-32 | Physical appearance of
overheating failures
Clean Combustion Technologies
Figure: 09.32 (TB)
Water Technology | 9-23
Water Chemistry Guidelines for Drum-Type Boilers volatility of ammonia in the low-pressure
boiler is so high that it does not influence
Table 9-5 | Demineralized Water boiler water pH significantly within the
limits of normal phosphate treatment.
Parameter Unit N A1 A2 S An
The water chemistry principles for
Specific conductivity* µS/cm <0.20 — 0.2–0.5 >0.5 C
drum-type utility boilers are similar to
Specific conductivity† µS/cm <1 — 1–1.5 >1.5 C
those for HRSG drum-type boilers.
Silica as SiO2 ppb <20 20–50 50–200 >200 M
These chemistry guidelines reflect good
Sodium as Na ppb <10 — — — M
industry practice and generally are in
Iron as Fe ppb <20 — — — M
accordance with published guidelines
Copper as Cu ppb <3 — — — M
from the EPRI, European Power Plant
TOC‡ ppb <300 — — — M
Suppliers Association, VGB, Interna-
*At demineralizer plant outlet (on-site demineralization plant or water from vendor).
tional Electrotechnical Commission
†At demineralized water storage tank outlet.
‡TOC, total organic carbon. (IEC), ASME, and ABMA. One
notable exception is that the feedwater
and boiler water qualities also reflect the
Table 9-6 | Condensate (at Condensate Pump Discharge) steam purity requirements as specified
N* by the steam turbine and gas turbine
Parameter Unit A1 A2 S An
Cu SS/Ti (if steam or water injected) supplier,
Specific conductivity µS/cm 2–6 3–11 — — — M as opposed to those provided in the
Conductivity after ABMA guidelines, which are limited
µS/cm <0.20 <0.20 0.2–0.5 0.5–1 >1 C
cation exchange to the perspective of boiler chemistry.
pH 8.8–9.3 9.0–9.6 — — — M This guideline refers to normal plant
Silica as SiO2 ppb <20 <20 — — — M operation. Higher impurity levels may
Iron as Fe ppb <20 <20 — — — M be encountered during commissioning
Copper as Cu ppb <3 <3 — — — M and the early operating period until
Cu, copper-alloy tubed condenser; SS/Ti, stainless-steel or titanium tubed condenser. the final cleanliness of the system has
been achieved. Such condition must be
analyzed on a project-specific basis in
order to verify acceptance for operation
Table 9-7 | Feedwater (at Boiler Inlet)
during this period.
N*
Parameter Unit A1 A2 S An Feedwater chemistry control is based
Cu SS/Ti
on the following general considerations:
Specific
µS/cm 2–6 3–11 — — — C
conductivity • Ammonia is used to control pH.
Conductivity • An oxygen scavenger is not needed
after cation µS/cm <0.20 <0.20 0.2–0.5 0.5–1 >1 C if the feedwater oxygen concentra-
exchange
tion is less than 10 ppb. If the use
pH 8.8–9.3 9.0–9.6 ≠ N† — ≤7, ≥11 M
of an oxygen scavenger becomes
Silica as SiO2 ppb <20 <20 ≥20‡ — — M
necessary, hydrazine is preferable if
Iron as Fe ppb <20 <20 — — — M
allowed by local safety regulations.
Copper as Cu ppb <3 <3 — — — M
• To minimize the risk of FAC in
Oxygen ppb <10 — — — M
the low-pressure systems, the pH
*Cu, copper-alloy tubed condenser; SS/Ti, stainless-steel or titanium tubed condenser.
range has a relatively high limit,
†If outside N.
‡Time permitted above 20 ppb: [hours] × [ppb − 20] < 105. If this time has expired, silica deposits may and the use of organic treatment
be expected. Checking turbine pressure and efficiency as well as turbine cleaning should be considered. chemicals is not recommended.
Normal feedwater quality should be achieved at 30% and higher unit load.
Boiler water chemistry control is
based on the following considerations:
Water Technology | 9-25
• General and specific corrosion pro- Water Chemistry Guidelines for Drum-Type Boilers— continued
tection of the pressure part surfaces
in the event of contamination ingress Table 9-8 | Phosphate Treatment of boiler water
• Achievement of the required steam
N
purity. Parameter Unit An
≤120 bar 121–197 bar
Specific conductivity C
There are several accepted methods
N µS/cm <40* <20*
for controlling boiler water chemistry.
A1 µS/cm 40–80 20–40
The majority of drum-type boil-
A2 µS/cm >80 >40
ers, however, still use some form of
S µS/cm — —
phosphate treatment. The particular
pH C
phosphate treatment (phosphate
N 9.0–10.0† 9.0–9.8†
concentration and corresponding pH)
A1 <9.0, >10.0 <9.0, >9.8
should be selected based on the follow-
A2 — —
ing main cycle considerations:
S <8 <8
• Condenser cooling water chemistry Phosphate as PO4 M
• Presence of condensate polishing N ppm 2–6 1–3
system Silica as SiO2 M
• Phosphate hideout history or N ppm ‡ ‡
potential
Estimated value; must be compatible with steam purity.
• Sodium limit in the steam †Within limits of Na/PO4 :
≥ 3.0 (>2.6)
The parameters of these guidelines ≤ Na3PO4 + 1 ppm NaOH
– min 0.1 pH above feedwater pH
cover the chemical species that are
‡See Figure 9-36.
common in steam–water cycles. If, in a
specific case, the cycle also may contain
species that are not covered by these
Table 9-9 | All-Volatile Treatment of boiler water
parameters, then the specification must
be expanded accordingly. Such species Parameter Unit Value An
Cation conductivity C
may, for example, come from water
N µS/cm <5
treatment, return condensate, and so on.
A1 µS/cm 5–10
Feedwater Quality and Steam Purity A2 µS/cm 10–20
S µS/cm >20
Dissolved Solids and Oxides. The water
pH C
chemistry guidelines presented for
N *
drum-type boilers address the influ-
<9.0,
ence of both dissolved and suspended A1
>10.0
(i.e., oxides) solids. Because the A2 —
feedwater is used for desuperheat- S <8, >10.5
ing (control of steam temperature), Na M
its level of dissolved solids must be N ppm <1
limited in order to prevent fouling of A1 ppm ≥1
steam-touched surfaces as well as to Silica as SiO2 M
achieve steam purity requirements. It N ppm † M
also should be recognized that metal-
*pH is controlled by feedwater pH.
lic oxides transported to the boiler †See Figure 9-36.
(or superheater/reheater/turbine) by
feedwater can foul these surfaces such
9-26 | Clean Combustion Technologies
Water Chemistry Guidelines for Drum-Type Boilers— continued more adherent/less soluble oxide film.
This further minimizes the potential for
Table 9-10 | Live Steam and Reheat Steam (at Boiler Outlet)
FAC. Oxygen scavengers such as hydra-
zine are not required (if copper alloys
N*
Parameter Unit A1 A2 S An are present in the condensate system,
Cu SS/Ti the use of hydrazine can be evaluated).
Specific Organic-based treatment chemicals
µS/cm 2–6 3–11 — — — M
conductivity
are not recommended and should not
Conductivity be required in a combined cycle power
µS/cm <0.20 <0.20 0.2–0.5 0.5–1 >1 C†
after cation exch.
plant generating electricity in which no
pH 8.8–9.3 9.0–9.6 — — — M
process steam is exported from the cycle.
Sodium as Na ppb <10 <10 10–20 20–40 >40 C
Sodium as Na‡ ppb <5 <5 5–10 10–20 >20 C Cation Conductivity. The cation con-
Silica as SiO2 ppb <20 <20 ≥20§ — — M ductivity specification values given in
Iron as Fe ppb <20 <20 — — — M the water chemistry guidelines do not
Copper as Cu ppb <3 <3 — — — M
include the influence of carbon dioxide.
During normal operating conditions,
*Cu, copper-alloy tubed condenser; SS/Ti = stainless steel or titanium tubed condenser.
†Switchable common monitor for high-pressure- and reheat steam.
levels of 0.06 to 0.2 µS/cm can be
‡Requirement in case boiler phosphate level is <2 ppm of PO4. achieved. Higher values are observed
§Time permitted above 20 ppb: [hours] × [ppb − 20] < 105. If this time has expired, silica deposits may be during start-up. Carbon dioxide, hav-
expected. Checking turbine pressure and efficiency as well as turbine cleaning should be considered.
ing entered the system with air during
shutdown, is dissolved in the water, and
Table 9-11 | Maximum permissible tolerance for carbon it requires some time to purge from the
dioxide (on-line instrument cation conductivity readings) system. The removal rate depends on
Type of Type of condenser the efficiency of the deaeration devices
Restrictions
operation Water cooling Air cooling (deaerator, condenser) as well as on
Normal <0.4µS/cm <0.8 µS/cm
cycle water pH. At cold start, normal
specification values usually are obtained
Steam production, <40%
Start-up >2 µS/cm >2 µS/cm nominal. Do not pass steam within a few hours for base loaded
to turbine plants. Cyclic units with frequent cold
1–2 µS/cm and 1–2 µS/cm and Maximum of 30 min per starts require a significant proportion of
decreasing decreasing start-up the operating time to reach normal val-
0.4–1 µS/cm 0.8–1 µS/cm Maximum of 6 h per start-up ues. Therefore, the question is whether
it is worthwhile to extend heat-up and
bypass operation to achieve a low cation
that damage and/or efficiency losses conductivity when solely influenced by
may occur. carbon dioxide. In several combined
cycle plants, start-up cation conductivity
Oxygen. The feedwater guidelines (see has been investigated using ion chroma-
Table 9-7) indicate that the oxygen tography. One such example is given in
concentration should be limited to less Figure 9-37. In that figure, it is seen that
than 10 ppb. At the same time, the use cation conductivity started well above
of oxygen scavengers, such as hydrazine 1 µS/cm, but with exclusion of carbon
or hydrazine substitutes, has been omit- dioxide, it was never larger than 0.3
ted. This is based on the experience from µS/cm when the turbine was on-line.
plants using OT, in which less reducing Such analytical techniques normally
conditions in low-cation-conductivity are not available in a power plant. Even
waters (<0.2 µS/cm) help to form a if the plant would have such equipment
Water Technology | 9-27
MW
1.0 Calculated All Anions 100
steam condenses, the chemicals dis- Calculated Anions without CO2
0.8 80
solve into the water. They do not form MW
0.6 60
a solid phase; thus, they do not form
a scale or deposit on heat-transfer 0.4 40
The use of hydrazine for wet lay-up is ingress. In those plants where phosphate
High pressure boiler practices not relevant for operating conditions and treatment can be used in the low-pres-
and guidelines are applicable to therefore it is possible. Environmental sure boilers, it will prevent the risk of
aspects regarding the use of hydrazine, FAC as well. The pH value (at 77°F
both utility boilers and the high however, need to be considered. [25°C]) is corrected from the contribu-
pressure boiler of an HRSG. AVT(O) is defined as AVT that tion of ammonia and depends solely
does not employ any reducing agent. on the phosphate concentration. The
Even without use of an oxidizing agent, heavy black line in Figure 9-38 is the
however, the electro-chemical potential upper boundary for the saturated steam
will be substantially higher than that sodium concentration in high-pressure
with AVT(R). boilers with a mechanical carryover ap-
AVT(O) favors the formation of proaching 0.2%.
hematite layers, which are less soluble The listed phosphate range is 1 to 6
and, hence, more stable than the mag- ppm (see also Table 9-8). This is limited
netite layers of AVT(R). As a result, by the sodium steam purity require-
it gives more protective margin against ments and the steam drum mechanical
FAC, and it minimizes orifice fouling. or moisture carryover performance. The
ABMA guidelines allow a higher level
Boiler Water Quality of dissolved solids (or sodium) in the
The boiler water quality and respective boiler water; consequently, for these high
treatment is designed to prevent corro- operating pressure ranges, the result-
sion and deposits in the boiler as well as ing sodium level in the steam would
to produce steam with a sodium content be higher than that allowed by turbine
that does not exceed specification suppliers. Therefore, the turbine steam
limits. Therefore, a low-level sodium purity requirements as specified by the
phosphate treatment has been selected. turbine supplier controls the level of
Figure 9-38 is a plot of phosphate phosphates and other dissolved solids in
versus pH. This type of strong alkaline the boiler water. The lower phosphate
treatment provides good buffering limit of 1 ppm for the high-pressure
capabilities in the event of impurity boiler water was chosen to provide
sufficient general pH protection and
buffering in case of impurity ingress. Be-
10.0
Na/PO4 Ratio cause there generally is no problem with
TSP+1 NaOH carryover at lower boiler pressures (these
4.0
Option Option 3.5 boilers are more in the critical tempera-
9.5 3.0 Standard ture range for FAC), a lower phosphate
2.6
2.3 limit of 2 ppm was chosen for low- and
2.0 intermediate-pressure boilers.
9.0 0.2% Carry-Over /
pH
HP Boiler
10.0
pH Feedwater
7.0
8.8
9.8 9.0
9.2
9.3
9.4
pH Boiler Water, without Ammonia
9.6 9.6
9.8
10.0
9.4
9.2
9.0
8.8
8.8 9.0 9.2 9.4 9.6 9.8 10.0
pH Boiler Water, with Ammonia
IP Boiler
10.0
pH Feedwater
7.0
8.8
9.8 9.0
9.2
9.3
9.4
pH Boiler Water, without Ammonia
9.6 9.6
9.8
10.0
9.4
9.2
9.0
8.8
8.8 9.0 9.2 9.4 9.6 9.8 10.0
pH Boiler Water, with Ammonia
Table 9-12 | Selection of boiler water treatment regimes (>1 ppm) is considered to be risky
for low-staffed power plants, because
Treatment Preference Use
it requires much more attention and
Na/PO4 = 3 Standard
control in order to be safely applied.
Where a buffer for caustic contaminants is In a typical combined cycle, the con-
Na/PO4 = 3.0–2.6 Option
required
densate, or feedwater, is relatively free
Na/PO4 = 3.0 − Where a buffer for acidic contaminants is of iron oxide during stable operation.
Option
(3.0 + 1 ppm of NaOH) required
Therefore, use of organic dispersants in
All-volatile treatment Option Where phosphate hideout would occur
the boiler water is not needed. These
Where solid alkalizers are not permitted organic substances can decompose to
Where boiler water is used for feed of other carbon dioxide and produce organic
boilers
acids. These acids can then circulate
Where boiler water is used for attemperation throughout the cycle. As a consequence,
spray
they could destabilize (if treated only
Where impurity ingress can be excluded with ammonia) the magnetite layer on
the low-pressure boiler surfaces, which
Phosphate treatments should be ap- can increase the potential for FAC.
Intermediate and low pressure plied within the range defined in Table Table 9-13 provides some guidelines in
guidelines are applicable to 9-8 and in Figure 9-38. It should be the event of contamination ingress as
observed that in this figure, phosphate measured by cation conductivity.
lower pressure boilers and is related to pH after the contribution
the intermediate and from ammonia has been subtracted. (A Once-through Units: Utility
low pressure boilers of a method for this calculation is given in Boiler and HRSG
multipressure HRSG. Verib.19) Figure 9-39 indicates the cor- The water chemistry guidelines for
rection factors to be applied. For routine once-through units are provided in
use, plant-specific empirical correction Tables 9-14 to 9-18.
factors may be established and used. In
specific cases, when phosphate hideout General Remarks
occurs, phosphate treatment with low All conductivities referred to in water
������
phosphate concentration or AVT can chemistry guidelines for once-through
be applied. Each boiler normally is units are for 77°F (����������������
25°C). Possible
controlled and treated independently, contributions from carbon dioxide may
although the water chemistry guidelines be excluded.
presented here for drum-type boilers list Operation is desirable at the lowest
the same level of treatment chemicals achievable impurity levels, with the
and conductivity. With demineralized shortest and least frequent excursions.
feedwater, there really is no need to The specification is related to the
allow higher dissolved solids in low- following conditions:
pressure boilers just because the boiler
• The system has no copper alloys.
can handle it without adversely affecting
• Conditioning is done with ammo-
steam purity.
nia and oxygen injection.
For phosphate treatment, trisodium
phosphate is used as the conditioning The guidelines for once-through
agent. For the optional treatment with units presented here are a general
a sodium/phosphate ratio between 3 specification, valid only for the plant
and 2.6, an adequate mix of trisodium type mentioned above. The criteria will
and disodium phosphates can be be reviewed for each specific applica-
used. The use of sodium hydroxide tion and at commissioning.
Water Technology | 9-31
*Boiler water pH under phosphate control should be higher (by a minimum of 0.1 unit) than the feedwater pH.
Except where defined otherwise in Water Chemistry Guidelines for Once-Through Units
the footnote of a table, the following
definitions apply to Tables 9-14
Table 9-14 | Demineralized Water (at Demineralizer Water
through 9-18:
Plant Outlet)
Parameter Unit N A1* A2 S† An
• N = Normal Value. Values are
consistent with long-term system Specific conductivity µS/cm <0.10 0.1–0.2 — ≥0.2 C
reliability. A safety margin has At storage tank outlet‡ µS/cm <1 ≥1 — — C
been provided to avoid concentra- Silica as SiO2 ppb <10 10–20 — ≥20 C
tion of contaminants at surfaces. Sodium + potassium as
ppb <5 — — — M
• A1 = Action Level 1. The poten- Na+K
tial exists for the accumulation Iron as Fe ppb <20 — — — M
of contaminants and corrosion. TOC§ ppb <300 — — — M
Return to normal values within 1 Oxygen — Saturation
week. Maximum exposure is 336
*Indicates there is a problem. Action shall be taken to bring values to normal. Operation may be contin-
cumulative hours per year, exclud- ued as long as feedwater/steam specifications are not jeopardized.
ing start-up conditions. †Do not put this water into the demineralized water storage tank.
‡Conductivity as measured at demineralized water storage tank outlet, including carbon dioxide.
• A2 = Action Level 2. Accumula- §TOC, total organic carbon.
tion of impurities and corrosion
will occur. Return to normal levels
within 24 hours. Maximum expo- Table 9-15 | Condensate (at Hotwell Outlet)
Water Chemistry Guidelines for Once-Through Units— continued and boiler water are less than 3
ppm (permissible value depends on
demineralizer performance/con-
Table 9-16 | Feedwater (All-Volatile Treatment:
straints). The feedwater (pH >9)
Measurements at Economizer Inlet)
can be pumped through the boiler
Parameter Unit N A1 A2 S An into the start-up separator, flash
Specific con- tank (if present), and then to waste.
µS/cm 4.0–11 — — — C
ductivity
2. When this limit is achieved, cir-
Conductivity culate water through the boiler,
after cation µS/cm <0.20 0.2–0.5 >0.5 >1 C
exchange start-up separator, flash tank,
and then back to the condenser.
pH 9.2–9.6 <9.2, >9.6 >10 ≤7* C
Place condensate demineralizers
Sodium +
potassium as ppb <5 5–10 10–20 >20 M or C
in service.
Na+K 3. When the total iron concentra-
Silica as SiO2 ppb <20 ≥20† — — M tion drops below 1 ppm, silica
is less than 100 ppb, and cat-
Iron as Fe ppb <20 — — — M
ion conductivity is less than 1
Hydrazine ppb ‡ M
μS/cm (parameters measured at
Oxygen ppb <10 ≥50 — — C
economizer inlet), unit firing can
*If dosing is limited during system refilling, ammonia will be below normal. In such a case, pH ≥5.5 is commence. Boiler and start-up
tolerated.
†Time permitted above A1 (see “HP-steam”): (hours) × (ppb – A1) <105
system operation should be as per
‡Hydrazine or equivalent should not be used if O2 is 10 ppb or less. the instruction manual.
4. Proceed to normal feedwater
control limits (parameters mea-
Table 9-17 | Feedwater (Oxygenated Treatment instead of
sured at economizer inlet). These
all-volatile treatment)
should be obtained before exceed-
Parameter Unit N A1 A2 S ing one-third of the unit load.
Specific conduc-
µS/cm 2.5–7.0 — — —
tivity Chemistry control in once-through
Conductivity after steam generating units is based on the
µS/cm <0.20 0.2 - 0.5 * >0.5 * >1
cation exchange
following features:
pH 8.5–9.0 — — —
Sodium + potas-
• The system has no phase-separating
ppb <5 5 - 10 10-20 >20 devices (i.e., steam drum).
sium as Na+K
Oxygen‡ ppb 50–150 — — — • Feedwater, boiler water, and steam
are the same fluid stream.
*Discontinue oxygen injection.
‡Target for oxygen injection.
• Sliding pressure operation
includes fluid conditions typical of
both supercritical and subcritical
operation.
Once-Through Boiler Cycle: Guidelines The above design considerations
for Initial Start-Up and Restarts After require that the concentrations of
Long Outages feedwater contaminants be kept to a
The guidelines for initial start-up are minimum and be within allowable tur-
as follows: bine steam purity limits (as specified by
the steam turbine supplier). Corrosion
1. Flush to waste until suspended products transported to the boiler (or
solids in the condensate, feedwater, superheater/reheater/turbine) from the
Water Technology | 9-33
condensate and feedwater system must Water Chemistry Guidelines for Once-Through Units— continued
be kept at concentrations low enough to
minimize fouling of heat-transfer sur-
Table 9-18 | High-Pressure Steam and Reheat Steam
faces and, thus, the potential for damage
(at OTSG* Outlet)
and/or efficiency losses.
For once-through systems, feedwa- Parameter Unit N A1 A2 S An
ter conditioning to minimize general Specific conduc-
µS/cm 2.5–7.0 — — — M
tivity
corrosion and the production of
iron oxide can be accomplished with Conductivity after
µS/cm <0.20 0.2–0.5 >0.5 >1 C
cation exchange
either AVT or OT. Because of the
greater concern for copper transport at pH 9.0–9.6† — — — M
pH
creases, as illustrated in Figure 9-40. generally is maintained below 10 ppb 7
For example, strong alkaline buf- by mechanical deaeration. Hydrazine 6
fers, such as sodium phosphate and or hydrazine substitutes are reducing 25 45 65 85 105 125 145 165 185
Temperature, °C
sodium hydroxide, do not exhibit this agents that will lower the ORP and
relationship with temperature. The increase the FAC potential in carbon
Figure 9-40 | Effect of fluid temperature on
pH remains relatively constant over steel components where fluid tempera- Clean Combustion
pH with 1.5Technologies
ppm of ammonium
the temperature range. Therefore, with tures are in the range of approximately Figure: hydroxide
09.40/ (JG)
PMS5415 K
AVT, relatively small amounts of salts
or other acid-producing substances 10
(organic contaminants or organic
additives) can have a significant effect Pure 1 ppm
9 10 ppm
on pH in higher temperature regions. MgCl2
100 ppm
NH4OH 1,000 ppm
This relationship is taken into account 8
when formulating the cycle water
Solutions pH
Potential vs Ag/AgCl, mV
material is volatile, there is no objec-
Metal Loss, Microns
600
tion to its use even in non-drainable 16
superheater and reheater sections of the 800
Potential
boiler. Where freezing is a problem, it 12
sometimes is necessary to drain portions 1,000
Loss Rate
of the steam generator. This should be 8
1.12 mm y-1
accomplished under a positive nitrogen 1,200
Loss Rate
pressure. Nitrogen blanketing is recom- 4 1.10 mm y-1
mended even when the boiler can be 1,400
maintained full of inhibited condensate. 0
20 60 100 140 180
The success or failure of standby protec- Time, hours
tion also depends on the proper design
and selection of chemical feed equip- Figure 9-44 | Influence of oxygen on flow-assisted corrosion
ment and on the ability to provide an
inert gas blanket. Base loaded units may
get along with simple, manual systems Because of their extreme solubility,
Clean Combustion Technologies
for the introduction of lay-up additives sodium salts are of concern only to the
Figure: 09.44/ SPS
and nitrogen. Systems with load char- extent of their contribution toPMS5415
the totalK
acteristics that require frequent boiler dissolved solids concentration.
shutdowns require more elaborate, fool- Boiler tube scale consisting of
proof, wet lay-up protection systems. calcium and magnesium components,
however, is not uncommon in units
INDUSTRIAL BOILER WATER operating below 600 psig (4 MPa)
TREATMENT and using softened make-up water.
Internal chemical treatment systems These scales often are the direct result
are designed primarily to prevent scale of loss of chemical control, permitting
formations resulting from residual dense calcium precipitates, such as
amounts of calcium and magnesium calcium sulfate or calcium silicate, to
hardness in the feedwater. External form. They also may be a more subtle
treatment and control must provide a result of a porous, hardness sludge
feedwater that is low in hardness, alka- or metal oxides, which permit boiler
linity, silica, metal oxides, and oxygen. water to penetrate to the tube surface,
9-38 | Clean Combustion Technologies
limits, and analytical procedures must Silica fouling of turbines was common
be established with the cooperation of until it was recognized that success-
the chelant supplier or water treat- ful control of the amount of silica in
ment consultant. the steam could be accomplished only
by controlling the amount of silica in
CARRYOVER AND the boiler water. A similar solution
STEAM SAMPLING will be required for other solids when
This section deals with the phenom- operating at high pressures. Vaporous
enon of carryover, methods of steam carryover contributes a significant pro-
sampling, and techniques of steam portion of total solids in the steam as
purity determination. Once again, these the drum operating pressures increase
subjects cannot be understood solely by above 2,600 psig (18 MPa).
rigorous theoretical analysis. Knowledge Table 9-20 shows findings of a labora-
of laboratory testing procedures and tory study on vaporous carryover.22 For
ability to interpret field operating expe- various salts, the table lists the percent-
rience are required. Despite intensive age vaporous carryover, which is the
research, our understanding in these ratio of the salt concentration in the
areas is incomplete, and much empiri- steam and boiler water. In each case, so-
cism remains in the techniques used for dium was measured and then converted
correcting problems that result from to the appropriate salt concentration.
incomplete separation of steam and When calculating total solids in the
water in boiler equipment. steam, vaporous carryover is assigned a
value of 0.1% at drum operating pres-
CLASSIFICATION OF CARRYOVER sures above 2,600 psig (18 MPa).
Carryover of boiler water in steam Mechanical carryover may be classi-
leaving the drum provides a path for fied under four headings:
introducing solid materials into the
1. Priming
steam. Modern separator designs can
2. Spray
mechanically reduce the moisture
3. Leakage
content of the steam to 0.2% or less.
4. Foam carryover
In addition to mechanical carryover
of boiler water, however, another Each type of mechanical carryover
mechanism exists that results in the results in troublesome deposits in the su-
contamination of steam with solid perheater or on the steam turbine blades.
materials. As the operating pressures Priming is the development of exces-
increase, the steam phase exhibits sive moisture in the steam because of
greater solvent capabilities for the salts
that may be present in the water phase.
Table 9-20 | Summary of Laboratory Results on vaporous
These salts will be partitioned in an
carryover
equilibrium between the steam and the
water, a phenomenon known as “vapor- Pressure (psig) 2,600 2,800 3,000
ous carryover.” Vaporous carryover Concentration (ppm) 15 500 15 500 15 500
will contribute an additional quantity Sodium sulfate 0.02% 0.03% 0.04% 0.07% 0.28% 0.48%
of boiler water solids directly to the Disodium phosphate 0.01% 0.07% 0.03% 0.18% 0.41% 0.74%
steam, independent of the efficiency of
Trisodium phosphate 0.02% 0.11% 0.04% o.30% 0.35% 1.3%
steam–water separation components.
Sodium chloride 0.04% 0.18% 0.09% 0.36% 0.39% 1.2%
Silica was the first material found to
exhibit significant vaporous carryover. Sodium hydroxide 0.02% 0.31% 0.08% 0.69% 0.55% 2.2%
9-40 | Clean Combustion Technologies
Table 9-21 | Recommended Boiler Water Limits and Associated Process Steam Purity (at Steady-State,
Full-Load Operation)
Drum-Type Boilers
Drum Pressure Range of Total Dissolved Range of Total Range of Total Dissolved
(psig, actual) Suspended Solids in Boiler
Solids* in Boiler Water, Alkalinity† in Boiler Solids†§ in Steam (ppm,
Water, ppm (Maximum)
(ppm, maximum) Water, ppm maximum expected value)
0–300 700–3,500 140—700 15 0.2–1.0
301–450 600–3,000 120—600 10 0.2–1.0
451–600 500–2,500 100—500 8 0.2–1.0
601–750 200–1,000 40—200 3 0.1–0.5
751–900 150–750 30—150 2 0.1–0.5
901–1,000 125–625 25—125 1 0.1–0.5
1,001–1,800 100 ‡ 1 0.10
1,801–2,350 50 ‡ N/A 0.10
2,351–2,600 25 ‡ N/A 0.05
2,601–2,900 15 ‡ N/A 0.05
Once-Through Boilers
1,400 and above 0.05 N/A N/A 0.05
*Actual values within the range reflect the ions in the feedwater. Higher values are for high solids; lower values are for low solids in the feedwater.
†Actual values within the range are directly proportional to the actual value of ions of boiler water. Higher values are for high solids and lower values for low solids
in the boiler water.
‡Dictated by boiler water treatment.
§These values are exclusive of silica.
Source: American Boiler Manufacturers Association
9-44 | Clean Combustion Technologies
1. E. Nordell, Water Treatment for Industrial And Other Uses, 2nd ed. New 15. J. A. Armantano and V. P. Murphy, “Standby Protection of
York: Reinhold Publishing Corp., 1961. High-Pressure Boilers,” Proceedings of the 25th International Water
Conference, Pittsburgh, PA September 28–30, 1964, pp. 111–124.
2. S. T. Powell, Water Conditioning for Industry. New York: McGraw-Hill,
1954. 16. N. Henzel, W. Kastner, and B. Stellwag, “Erosion Corrosion in Power
Plants Under Single- and Two-Phase Flow Conditions—Updated
3. Betz Laboratories, Inc., Betz Handbook of Industrial Water Conditioning,
Experience and Proven Counteractions,” American Power Conference, 50:
7th ed. Trevose, PA.: Betz Laboratories, 1976.
992–1000, 1988.
4. P. Hamer et al., eds., Industrial Water Treatment Practice. London: But-
17. C. Cragnolino, “A Review of Erosion–Corrosion of Steels in High-
terworths, 1961.
Temperature Water,” in Environmental Degradation of Materials in
5. Nalco Chemical Company, The Nalco Water Handbook, 2nd ed. New Nuclear Power Systems—Water Reactors, edited by G. J. Theus and J. R.
York McGraw-Hill, 1988. Weeks. Warrendale, PA: The Metallurgical Society, 1987, pp. 397–406.
6. R. E. Hall et al., “Phosphate in Boiler Water Conditioning,” Journal of the 18. W. Kastner and K. Riedle, “An Empirical Model for Calculating Metal
American Water Works Association, 21: 79–88, 1929. Loss due to Erosion Corrosion,” VGB Kraftwerkstechnik, 66: 1171–
1178, 1986.
7. S. F. Whirl and T. E. Purcell, “Protection Against Caustic Embrittle-
ment by Coordinated Phosphate-pH Control,” Proceedings of the 19. G. Verib. “Sodium to Phosphate Ratios.” PPChem 2002, 4(11), 687–690.
Third Annual Water Conference, Pittsburgh, PA, November 9–10,
20. D. Platt and D. A. Vinnicombe. Operating of a Drum Boiler Without
1942, pp. 45–60.
Hydrazine. Johannesburg, South Africa: ESKOM, 1994.
8. Schikorr, G, “Uber die Reakiionen zwischen Eisen, seiner Hydroxyden
21. G. J. Bignold, C. H. de Whalley, K. Garbett, and I. S. Woolsey, Water
und Wasser,” Z Elektrochemie. 35, 65–70 (1929).
Chemistry of Nuclear Reactor Systems, 3. Vol. 1. London: British Nuclear
9. R. C. Corey and T. J. Finnegan, “The pH, Dissolved Iron Concentration, Energy Society, 1983, p. 219–226.
and Solid Product Resulting from the Reaction Between Iron and Pure
22. S. L. Goodstine, “Vaporous Carryover of Sodium Salts in High-Pressure
Water at Room Temperature,” Proceedings of the American Society for
Steam,” Proceedings of the American Power Conference, 36: 784–789, 1974.
Testing and Materials, Preprint 101,1939.
23. P. B. Place, “Carryover Problems and Identification of Carryover Types:”
10. R. C. Corey and T. J. Finnegan, “The pH, Dissolved Iron Concentration,
Combustion, 18(9): 29–34, 1947.
and Solid Product Resulting from the Reaction Between Iron and Pure
Water at Room Temperature,” Proceedings of the American Society for 24. P. B. Place, “Steam Purity Determination. Part I. Evaluation of Test
Testing and Materials, Preprint 101,1939. Results,” Combustion, 25(10): 62–65, 1954.
11. E. Berl and F. van Taack, “Uber die Einwirkung von Laugen und Salzen 25. P. B. Place, “Steam Purity Determination. Part II. Methods of Sampling
auf Flusseisen unter Hochdruckbedingung und uber die Schutzwirki- and Testing,” Combustion, 25(11): 41–44, 1954.
ung von Natriumsulfat gegen den Angriff von Aetznatron und von
26. P. B. Place, “Steam Purity Determination. Part III. Interpretation of Test
Chlormagnesium (The Action of Caustic and Salts on Steel under Con-
Results,” Combustion, 25(12): 43–46, 1954.
ditions of High Pressure and the Protective Effect of Sodium Sulfate
Against the Attack of Sodium Hydroxide and Magnesium Chloride),” 27. B. J. Cross, “The Sampling of Steam for the Determination of Purity,”
Forschung sorbeiten 330. Berlin: VDl, 1930. Proceedings of the 11th Annual Water Conference, Pittsburgh, PA,
October 16-18, 1950, pp. 71–82.
12. H. A. Grabowski, et al., “Problems in Deaeration of Boiler Feedwater,”
Combustion 25(9): 43–48, March,1955. 28. “Standard Practice for Sampling Steam,” D 1006-06. In 2008 Annual
Book of ASTM Standards, Part 31: Water. Philadelphia: American
13. H. A. Klein and H. A. Grabowski, “Corrosion and Hydrogen Damage
Society for Testing and Materials, vol. 11.01, 71–75.
in High-Pressure Boilers,” Presented at the Second Annual Education-
al Forum on Corrosion, National Association of Corrosion Engineers, 29. “Standard Test Methods for Electrical Conductivity and Resistivity of
Drexel Institute of Technology, Philadelphia, PA, September 15–17, Water,” ASTM Standards D 1125–95(2005). In 2008 Annual Book of
1964. ASTM Standards, Part 31: Water. Philadelphia: American Society for
Testing and Materials, vol. 11.01, 96–102.
14. P. Goldstein and C. L. Burton, “Research Study on Internal Corrosion
of High-Pressure Boilers,” Transactions of the ASME Journal of Engineer- 30. “Standard Test Methods for Deposit-Forming Impurities in Steam,”
ing for Power, 91, Series A: 75–101, April 1969. ASTM Standards D 2186-05. In 2008 Annual Book of ASTM Stan-
dards, Part 31: Water. Philadelphia: American Society for Testing and
Materials, vol. 11.02, 103–109.
9-45
SUGGESTED READINGS
K. L. Atwood, “Solvent Selection for Preoperational and Operational C. M. Kennedy et al, “Experience with High Rate Ammoniated Mixed Beds
Cleaning of Utility Boilers,” Combustion, 42(3): 16–21, 1970. for Condensate Polishing of CIPSCO Coffeen Station.” Combustion,
39(9): 19–30, 1968.
K. L. Atwood and G. L. Hale, “A Method for Determining Need for
Chemical Cleaning of High-Pressure Boilers,” Proceedings of the American H. A. Klein, “A Field Survey of Internation Corrosion in High-Pres-
Power Conference, 33: 710–720, 1971. sure Utility Boilers,” Proceedings of the American Power Conference, 33:
702–709, 1971.
K. L. Atwood and J. A. Martucci, “The Application of Hydrochloric
Acid with Ammonium Bromate for Scale Removal in Utility Boilers,” H. A. Klein, “Corrosion of Fossil Fueled Steam Generators,” Combustion,
Proceedings of the 28th International Water Conference, Pittsburgh, PA, 44(7): 5–20, 1973.
December 11–13, 1967, pp. 167–179.
H. A. Klein, “Use of Coordinated Phosphate Treatment to Prevent Caustic
R. J. Barto, D. M. Farrell, F. A. Noto, and S. L. Goodstine, “Intelligent Corrosion in High-Pressure Boilers,” Combustion, 34(4): 45–52, 1962.
Chemistry Management System (ICMS)—A New Approach to Steam
H. A. Klein and K. L. Atwood, “Chemical Cleaning of Utility Boilers,”
Generator Chemistry Control,” Proceedings of the American Power Confer-
Proceedings of the American Power Conference, 26: 762–778, 1964.
ence, 48: 1025–1031, 1986.
H. A. Klein and P. Goldstein, “The Effects of Water Quality on the
R. D. Brown and D. A. Harris, “Large Coal-Fired Cycling Unit,” Presented
Performance of Modern Power Plants,” Presented at the 24th Annual
at the ASME-IEEE-ASCE Joint Power Generation Conference, Port-
Conference of the National Association of Corrosion Engineers, Cleve-
land, OR, September 28–October 2, 1975.
land, OH, March 18–22, 1968, Paper 12.
B. Chojnowski and R. D. B. Whitcutt, “Corrosion Failure: One Cause and a
J. J. Kurpen, “Externally Regenerated Condensate Demineralization Systems
Cure in an Operational Boiler,” Combustion, 48(5): 28–33, 1976.
for Once-Through Steam Generators,” Presented at the Liberty Bell Cor-
W. H. Clayton et al., “Design for Peaking Cycling,” Presented at the Joint rosion Course, Philadelphia, PA, September 16–18, 1969.
Power Generation Conference, Pittsburgh, PA, September 27–30, 1970.
J. J. Kurpen and D. L. Dixson, “Operating Experience in Cycle Cleanup for
F. Gabrielli and W. R. Sylvester, “Water Treatment Practices for Cyclic Supercritical Pressure Units,” Proceedings of the American Power Confer-
Operation of Utility Boilers,” Proceedings of the International Water ence, 30:883–896, 1968.
Conference, Pittsburgh, PA, October 31–November 2, 1978, pp.
J. A. Levendusky and L. Olejar, “Condensate Purification Applications of
193–208.
the Power Process in High-Pressure Utility Plant Cycles,” Proceedings of
F. Gabrielli et al., “Contamination Prevention of Superheaters and Reheat- the American Power Conference, 29: 840–856, 1967.
ers During Initial Start-Up and Operation,” Proceedings of the American
E. W. Lolir and S. K. Love, :The Industrial Utility of Public Water
Power Conference, 38: 296–310, 1976.
Supplies in the United States, 1952: Part l,” Geological Survey Water
F. Gabrielli et al., “Prevent Corrosion and Deposition Problems in High- Supply Paper 1299. Washington, D.C.: U.S. Government Printing
Pressure Boilers,” Power, 122(7): 85–92, 1978. Office, 1954.
F. Gabrielli, N. C. Mohn, and W. R. Sylvester, “Water Chemistry Aspects of E. W. Lolir and S. K. Love, :The Industrial Utility of Public Water Sup-
Cyclic Operation for Older High-Pressure Drum-Type Boilers,” Proceed- plies in the United States, 1952: Part 11,” Geological Survey Water
ings of the American Power Conference, 45: 989–999, 1983. Supply Paper 1300. Washington, D.C.: U.S. Government Printing
Office, 1954.
F. Gabrielli, N. C. Mohn, and B. C. Teigen, “Deposit and Water Chemistry
Studies with Rifled Tubing,” Proceedings of the American Power Conference, H. M. Rivers, “Concentrating Films: Their Role in Boiler Scale and Cor-
46: 973-984, 1984. rosion Problems,” Proceedings of the 12th Annual Water Conference,
Pittsburgh, PA, October 22–24, 1951, pp. 131–145.
S. Glasstone, “Overvoltage and Its Significance in Corrosion,” Corrosion and
Material Protection, 3(6): 15–18, 1946. F. N. Speller, “Control of Corrosion by Deactivation of Water,” Franklin
Institute Journal, 193: 515–542, 1922.
S. L. Goodstine and J. J. Kurpen, “Corrosion and Corrosion Product Con-
trol in the Utility Boiler-Turbine Cycle,” Combustion, 44(11): 6–18, 1973. D. G. Stocky and K. L. Atwood, “Non-Destructive Testing for Location of
Corrosion Damage in High-Pressure Boilers,” Presented at the Meeting
N. D. Greene and M. C. Fontana, “A Critical Analysis of Pitting Corrosion,”
of the Southeastern Electric Exchange, New Orleans, LA, April 29–30,
Corrosion, 15(13): 41–47, 1959.
1971.
T. Hagewood et al., “The Control of Internal Corrosion in High-Pres-
R. Svoboda, F. Gabrielli, H. Hehs, and H. Sandmann, “Combined Cycle
sure Peaking Units,” Proceedings of the American Power Conference, 30:
Chemistry: Present and Next Generation,” Presented at the Seventh
939–948, 1968.
International EPRI Conference on Cycle Chemistry in Fossil Plants,
A. F. Kelly et al., “Modify Base-Load Turbines for Peaking Service,” Power, Houston, TX, June 3–5, 2003.
115(4): 62–63, 1971.
9-46
Ulmer, R. C. and H. A. Klein, “Impurities in Steam from High Pressure L. H. Vaughn and C. V. Runyan, “Corrosion Protection of Boilers and
Boilers.” Proceedings of the ASTM, 61:1396–1411,1961. Associated Equipment During Idle Periods,” Proceedings of the American
Power Conference, 33: 721–729, 1971.
R. Svoboda, E. Liebig, H. Sandmann, and F. Gabrielli, “Cycle Chemistry
Features in Advanced Combined Cycle Power Plants,” Presented at the C. W. Wages and C. W, Smith, “Operating Experience of a Deep-Bed Con-
Sixth International EPRI Conference on Cycle Chemistry in Fossil densate Polishing System,” Proceedings of the 38th International Water
Plants, Columbus, OH, June 27–29, 2000. Conference, Pittsburgh, PA, November 1–3, 1977, pp. 111–121.
9-47
Operation
of Steam Generators
Chapter Ten
The operation of a multi-million dollar steam generator and its associated power plant
equipment requires the constant exercise of intuitive reasoning and sound engineering
judgment. It is in operation that all the factors that went into the design and construction of
the facility are put to the test. A principal objective of proper operation is sustained service
between outages while, at the same time, obtaining the highest possible efficiency and
availability from all the plant components.
Operation of a steam generator is must know not only what he or she is safeguard systems, and an effective
a balance of inputs and outputs: the doing but also why it is done and what preventive-maintenance program.
better the balance, the smoother the results from the operator’s action.
operation. Producing steam from a The prime consideration for all BASIC OPERATING PRINCIPLES
boiler requires that the weight of water operation is the safety of people and In this section, certain basic operat-
entering the boiler equal the weight equipment. Whenever there is any ing guidelines for overall effective
of steam leaving the boiler. Firing the doubt about an unsafe condition, the operation of a large coal-fired unit are
furnace requires a balance of fuel and operator must take immediate action discussed.
air. To equalize these inputs and out- to return the unit to a known safe
puts, one must understand the system, condition, even if it means tripping the STEAM TEMPERATURES
not just the network of hardware that unit. Today, as throughout the history Maintaining the desired primary and
comprises it. This understanding is of steam generation, the two most dan- reheat steam temperatures requires
the principal ingredient of successful gerous conditions are the loss of water considerable operator attention. Even
operation. Too often operators have and the explosive mixture of fuel and the best control systems do not antici-
confused operation with control- air. Both result from an imbalance: the pate all of the factors affecting steam
system management. The operator first is caused by less water coming in temperature. Despite the equipment
must realize that a control system is than steam produced, even if some of installed for controlling superheater
hardware assembled to make operation that steam is escaping through a tube and reheater steam temperatures, cer-
easier, faster, and safer. All large steam leak; the second is caused by too little tain conditions may produce abnormal
generators require operators. The air for the fuel present. steam temperatures. For instance, with
control of all major functions can be Safe operation, then, is a result of a new coal-fired unit, it may be neces-
switched from automatic to operator comprehensive training programs sary to operate for a considerable time
control. To be effective an operator for operators, well-designed furnace before normal furnace seasoning allows
10-
10- | Clean Combustion Technologies
the unit to produce predicted steam can lead to better control of the steam
temperatures. “Normal furnace season- temperature and reduce the possibility
ing” is often defined as the condition of troublesome accumulations in the
of furnace wall slag or ash deposits furnace and convection passes.
that remain after soot blower opera-
tion. Low steam temperatures may also BOILER EFFICIENCY
result from: An effective operator is constantly
striving to obtain maximum efficiency
• insufficient excess air
from a unit. To do this, the opera-
• higher-than-design feedwater
tor must be aware of the effect of all
temperature
operating variables, and the adjust-
• reheater inlet temperature lower
ments required to maximize efficiency.
than specified
Two items within operator control that
• externally fouled superheater or
affect boiler efficiency are dry gas loss
reheater
There are many potential causes and unburned fuel loss.
• leaking desuperheater spray water
for temperature variations • poorly adjusted controls Dry Gas Loss
in furnace operations. Often
On the other hand, high steam Usually the largest factor affecting
these are interrelated. A proper temperature may result from: boiler efficiency, dry gas loss increases
understanding of the system is with higher exit gas temperatures or
• an “over-seasoned” furnace
needed to provide for proper • too high an excess air level
excess air values. Every 35 to 40°F
and timely operator action. (20 to 22°C) increment in exit gas
• feedwater temperature lower than
temperature will lower boiler efficiency
specified
by 1%. A 1% increase in excess air by
• reheater inlet temperature higher
itself decreases boiler efficiency by
than specified
only 0.05%. On most boilers, however,
• irregular ignition or delayed com-
increased excess air leads to a higher
bustion
exit gas temperature. Consequently,
• poorly adjusted control equipment
increases in excess air can have a two-
An operating variable that has a very fold effect on unit efficiency.
great effect on steam temperature is the Usually, coal-fired units are designed
cleanliness of the radiant and convective to operate with 20 to 25% excess air.
heating surfaces. Although all modern To operate a boiler most efficiently, an
coal-fired steam generators are equipped operator must have a reliable means
with automatic soot blower systems, of assessing the quantity of excess
the judicious supplemental manual air leaving the boiler. In situ oxygen
operation of certain blowers can im- analyzers that measure the oxygen at
prove overall unit operation. It can save the boiler or economizer outlet are the
valuable blowing medium and reduce best information source. These must be
required maintenance by minimizing checked regularly for proper calibra-
the number of blowing cycles. To be tion and maintained as necessary. The
most effective, a soot blower program operator should maintain the required
requires periodic furnace observations. excess air level by making sure the con-
Based on such observations and per- trols are in the correct mode and tuned
formance results, selective soot blowing with respect to the fuel being fired.
Operation of Steam Generators | 10-
the tests communicate the results not Because of the great differences in
to the rest of the plant, but rather to steam generating units, detailed start-
the company files. It is important, then, up procedures cannot be given. The
to review test results with the unit operator must be familiar with all of
operators and to establish operating the manufacturer’s instructions and the
procedures that will take advantage of plant operating procedures.
what is learned from the tests. At this point, safety tags can be
Plant restarts are expensive. They re- removed from breakers. As the boiler
quire considerable fuel to get a unit up is filling with treated water, all vents
to operating temperature and pressure should be open, as noted on the manu-
and bring it online. Extra maintenance facturer’s valve operating diagram. On
and online attention that keeps plant a thermal circulation unit, the water level
downtime to a minimum can pay off should be brought to where it just shows
in fuel savings. Retaining heat in the near the bottom of the water glass. On a
boiler during a weekend shutdown Controlled Circulation® unit, the water Good operating and maintenance
when there is no demand for steam can level should be brought near the top practices help to minimize
save fuel. A boiler can retain a good of the gage glass. This will prevent the
downtime and, thus, the number
deal of heat if its isolation dampers water from dropping from sight when
are in good condition and are closed the first boiler-water circulating pump of restarts. Restarts are expensive,
tightly. Precautions should be taken is started. The drainable portions of all consuming considerable fuel
to follow the purging requirements steam circuit headers, connecting links, without contributing to increased
both before bottling and prior to light and piping should be drained through efficiency. Minimizing the
off. The loss of the unit because of an lines free from back pressure. Reheater
explosion will be infinitely more costly drains and vents should be opened so
number of restarts improves
than any heat saving from failing to that residual moisture will be boiled overall efficiency leading to
perform a proper purge. off. These reheater drains and vents reduced annual emissions.
will have to be closed prior to raising a
START-UP FROM COLD vacuum in the condenser.
CONDITION: GENERAL During warm-up and until the unit
Prior to light off of any boiler, a super- is carrying load, there will be little or
visor of operations should inspect the no steam flow through the superheater
exterior of the unit. All doors should and reheater. To protect the superheat-
be checked, cleared of tags, and then er and reheater metals, the temperature
shut. Valves should be correctly posi- of the gas leaving the furnace should be
tioned for start-up in accordance with limited to the manufacturer’s recom-
the steam generator manufacturer’s mendation, usually 1,000°F (540°C).
valve operating diagram. All areas must The firing rate must be limited to
be free of debris that will hinder unit satisfy this requirement. Thermal
expansion. If repair work was done circulation boilers can be warmed up at
during the outage, special care must be a rate that does not exceed a satu-
taken to ensure that no permanent ties rated steam temperature rise of 200°F
were made to the furnace structure that (110°C) per hour. Controlled Circula-
will impair expansion. All personnel tion® boilers have no limitation on the
must be accounted for. Before light off, saturation steam temperature rise, only
the operator should check all instru- on the furnace exit gas, as previously
mentation and furnace safety systems. stated. Most modern boilers have
10- | Clean Combustion Technologies
traversing thermocouples or optical oil is the warm-up fuel, the air heater
pyrometers to monitor furnace exit gas soot blowers should be operated
temperature during warm-up. Older frequently to keep the heating surface
units may use temporary thermocouple clear of any flammable deposits. Once
probes for each start-up or have con- the unit is online, the load may be
servatively established firing limitations picked up as swiftly, as pulverizers
based on prior testing. can be brought into service. The usual
Usually the equipment will start in restrictions in most power plants
this sequence. Air heaters and boiler- are the warm-up and load increase
water circulating pumps (on positive limitations of the steam turbine. (See
circulation units) will be started first. Chapters 4 and 9.)
Next, the induced draft fans, followed
immediately by the forced draft fans Start-Up from Cold Condition:
will be started. This will establish Drum-Type Units
Safety first in any light off airflow through the plant equipment. On drum-type units, maintaining the
situation. On most units, the furnace draft will drum water level is of prime impor-
be established and then transferred tance. Normally, the operator carries
to automatic draft control. Airflow is out this function in the manual mode
raised to at least 30% of the full-load until pressure is raised. Before light
airflow, and the unit should be purged off, the drum water level should be
for at least 5 minutes to remove any brought in sight. It is best to start firing
unburned fuel or combustible gases with a low water level because, as the
that may remain in the furnace. During water starts heating, it will expand.
start-up, airflow should be maintained Drum vent valves should be wide open
at 30% of full-load airflow to ensure an so that the air will vent from the drum.
air-rich furnace mixture and to prevent The superheater and reheater should
any settling out of explosive mixtures. be drained of any condensate whenever
Once the unit is purged, oil pumps possible and then the valves opened or
may be started and gas or oil trip closed according to the valve operating
valves opened. At set intervals, the instructions.
operator should check for the proper For a Controlled Circulation®
functioning of any furnace safeguard boiler, two of the circulating pumps
system. Regardless of the level of ur- are started to initiate water circula-
gency for getting the unit returned to tion. To minimize flashing of steam
service, no interlock should be jumped in the downcomers, additional pumps
or bypassed. are not started until the boiler water
Light off of any fuel nozzle should reaches 250°F (120°C). The unit is
never be attempted without the now fired in accordance with the es-
required ignition energy source for that tablished furnace safeguards. During
nozzle. Any time an operator has any the warming period, the economizer
doubt about safe combustion in the recirculating line valves are open.
furnace, the fuel should be tripped and The water will swell as it is heated
the furnace purged before relighting. and the operator will manually blow
As pressure is raised, periodic in- down the boiler as required to maintain
spections are necessary to ensure that sight of the water in the gage glass. The
the unit is expanding as it should. If firing rate will be controlled to keep the
Operation of Steam Generators | 10-
furnace exit gas temperature below the theory, for increasing load, the operator
recommended value, usually between will open the turbine control valve. If
900 and 1,000°F (480 and 540°C). no other action were taken, the flow to
When drum pressure reaches 25 psig the steam turbine would increase and
(1.7 bar), it can be assumed that all the superheater outlet pressure would
air has been purged from the drum decrease. Upon seeing the outlet pres-
and water circuits. The operator can sure drop, the operator would increase
close the drum vent valves. As pressure the fuel and air inputs to return the
increases, it is necessary to progres- pressure to the desired level. Because
sively throttle the superheater drains of the increased steam flow, the water
and vents and to increase the firing level in the drum would change. Ini-
rate as required. If there is evidence tially, surging would cause the level to
of steaming in the economizer by er- rise, but shortly thereafter, it would fall
ratic drum control during the feeding rapidly because more steam is removed
of the water, the operator must be than water enters. A trained operator
certain that the recirculation line is will anticipate these interactions. As
open. On thermal circulation units it the turbine control valve is opened, the
may be necessary to feed more water operator will simultaneously increase
than required to maintain the level in fuel, air, and feedwater flow. A well-
the sight glass and to blow down to designed and well-executed operational
control the water level. control system will perform the same
The turbine should start rolling procedure as will a trained opera-
as soon as the minimum permissible tor. The more experienced and better
start-up pressure and temperature, trained the operator, and the more
specified by the turbine manufac- sophisticated the control system, the
turer, are reached. When bringing the higher the expectation that smooth
steam turbine up to speed, the firing operation will result.
rate must continue to be controlled
to prevent the furnace exit gas from Shutting Down Drum-Type Units
exceeding its temperature limit. Once Time requirements and procedures
the turbine is synchronized and the for shutting down a boiler depend on
minimum load established, this limit the nature of the shutdown (normal
can be removed and the firing rate shutdown to cold, normal shutdown
increased as required. Usually, by this to hot standby, emergency shutdown)
time all superheater and reheater vents and whether the unit is to be entered.
and drains are closed. If the steam For a scheduled shutdown, steam
drum water level is still under manual pressure can be reduced to the limit
control, it should now be placed on of the turbine before the unit is taken
automatic control. off-line. Thermal circulation units
should not exceed a cooldown rate of
Load and Pressure Control: 150°F (80°C) per hour of saturated
Drum-Type Units steam temperature decrease. Controlled
Opening or closing the turbine control Circulation® units can be rapidly cooled
valves increases or decreases the load if the circulating pumps are left in
on a drum-type unit. Changing the service. Normally, the drum vents are
firing rate regulates the pressure. In opened at 25 psig (1.7 bar) and the
10- | Clean Combustion Technologies
boiler is not drained until the water hotwell for recirculation through the
Very pure water is required for temperature is below 200°F (93°C). condensate and feedwater systems.
supercritical pressure operation Typically, at loads below 30%, the
START-UP of SUPERCRITICAL boiler operates in wet mode (and at
as there is no separation of PRESSURE BOILERS subcritical pressure), with recircu-
steam and water at these While there are many similarities lated waterwall flow superimposed
conditions. Any impurities in between the operation of drum- on feedwater flow. During wet mode
the make-up water will end up type and supercritical-type boilers, operation, the feedwater flow must
there are also some specific and very remain balanced with the superheater
somewhere in the steam system.
significant differences. Supercritical steam flow so that the separator
boilers demand a very precise balance storage tank water level is maintained
between inputs and outputs because within preset limits. If the feedwater
they do not have the flywheel effect flow is lower than the steam flow, the
inherent in the boiler water circuits of separator storage tank level will fall
drum type boilers. and eventually be lost. If the feedwa-
Although the firing system start-up ter flow exceeds the steam flow, the
and operation are the same as for separator storage tank level will rise
a drum-type unit, the fluid system and eventually flood the separators
start-up is completely different. The (in this case, corrective action is
supercritical boiler is furnished with taken to prevent such flooding). In
an integral start-up system. The unit wet mode operation, the waterwall
is initially fired, warmed up, and outlet is water and steam at
brought to partial load on the bypass saturated conditions.
system. Current vintage supercriti- Above 30% load, the boiler operates
cal boilers are of the sliding pressure in dry mode (once-through mode).
design, which uses start-up separators In dry mode operation, the waterwall
and a separator storage-tank. The outlet is steam only. All waterwall inlet
function of these components is to flow is evaporated, and a small degree
collect, separate, and then forward the of superheat is added before reach-
steam to the superheater. At the same ing the waterwall outlet. The slightly
time, the separated water is returned superheated steam discharging from
to the boiler water circuit. The water the waterwall is forwarded to the
level in the separator storage tank superheater with no water return from
is monitored and tightly controlled the separators to the separator storage
within preset limits during all phases tank. The boiler is now operating in
of operation. This has particular once-through mode. At this point,
importance for boiler start-up dur- the boiler pressure is still subcritical.
ing wet mode operation. Depending As the firing rate and load increase,
on the design, the start-up systems the pressure increases. Inherent in the
may or may not use a boiler water boiler design, the pressure slides up-
circulating pump (BCP). All designs ward as the feedwater and firing rates
use a flash tank whose function is to are increased. The pressure ultimately
receive any excess/drain water that is increases through the critical pressure
removed from the boiler water circuit. region and into the supercritical pres-
That excess/drain water is subse- sure region. Operation at supercritical
quently returned to the condenser pressure continues as long as the firing
Operation of Steam Generators | 10-
rate and the feedwater rate remain the separator storage tank to prevent
sufficiently high. As these reduce dur- the tank level from increasing and
Superheater Flash
ing normal operation load changes, flooding the separators, thus allowing
Tank
the pressure slides back down to the water to enter into the superheater.
subcritical region. There is no bump or With this design, the BFP provides Water Wall C
upset in boiler operation during these the minimum waterwall (evapora-
Economizer Low
pressure changes. The pressure simply tor) circulation and excess water is Load
increases or decreases through these continuously drained so as to maintain To Condenser
Underflow
pressures as the firing and feedwater the separator storage tank level within Governor
C
rates are changed. preset limits. Prior to boiler firing, the Deaerator
waterwall (evaporator) outlet is water
Circulation Systems only, and 100% of this is drained,
Boiler
C
The waterwall (evaporator) requires holding the separator storage tank High Pressure Feed
that a design preset, minimum inlet level at the set point. As the firing rate Heater Pump
water flow be maintained (or exceeded) increases, the waterwall (evaporator)
at all times. This is nominally 30% outlet becomes water and steam, and Figure 10-1 | Schematic diagram of a modern
maximum continuous rating (MCR) the quantity of drainage from the sepa- Clean Combustion
supercriticalTechnologies
boiler with a BCP
feed flow. This is required to ensure rator storage tank decreases. As firing Figure: 10.01/ (JG)
PMS5415 K
proper cooling water flow through the progresses, the waterwall outlet water/
furnace waterwall (evaporator) tubes steam ratio progressively reduces and Superheater High
Flash
at all times to prevent tube overheat- eventually becomes 100% steam with Water
Tank
Level
ing. There are different designs used on slight superheat. At this stage, there
the start-up systems and their method is no water drainage, and the boiler is C
Water Wall
of waterwall (evaporator) circulation. operating in once-through mode.
One design uses a boiler water circula- To Condenser
Economizer
tion pump and the other does not. (See Recommended Supercritical Unit
Figs. 10-1 and 10-2.) Start-Up Procedure1 Deaerator
When the design uses a boiler water 1. The BFP establishes a flow (5 to
C
circulation pump (BCP), that pump 10% of maximum continuous
remains in service until approximately rating) for boiler filling through High Boiler
Pressure Feed
35% (or higher) load, after which it the economizer to achieve the Pump
Heater
can be turned off. Up to 30% load, it normal separator storage tank
recirculates water through the waterwall level required to enable start of
(evaporator). When the load increases the BCP. After the BCP has been Figure 10-2 | Schematic diagram of a
Clean Combustion Technologies
modern supercritical boiler
through 30% and stable boiler dry mode started, the BFP flow is reduced Figure: 10.02/
without(JG)
a BCP
operation is attained, feedwater flow will to 0%. As the firing rate increases PMS5415 K
already be greater than the waterwall and the waterwall (evaporator)
minimum inlet flow requirement, allow- generates steam, the feed flow
ing the shutdown of the BCP. increases to maintain separator
When the design does not use a storage tank level at the set point.
BCP, the boiler feed pump (BFP) 2. The BCP is running and
provides the necessary minimum maintaining the waterwall
waterwall (evaporator) inlet water flow. (evaporator) inlet minimum flow
With this design, when the boiler is in requirement and the separator
wet mode operation, there is a continu- storage tank level is steady and at
ing requirement to drain water from the set point.
10-10 | Clean Combustion Technologies
fires is arbitrary, slow, and, worst of all, brought under control effectively and
dangerous, as it requires the presence expeditiously, an explosive condition
of a person in the immediate hazard can occur. Explosions also can occur
area. Temperature detection devices without the presence of an obvious fire,
should be used as the primary indica- if the necessary conditions to support
tor of a fire in progress. an explosion are present. All explo-
There are five types of pulverizer sions are initiated by fires, but all fires
fires. These are categorized by where do not result in explosions. Because of
in the fuel preparation system they this difference, independently designed
occur: feeder fires, above-bowl fires, and operated systems are available
under-bowl fires, exhauster fires, and to handle each condition safely. The
fuel-piping fires. Upon detection of explosion prevention system uses an
a pulverized fuel system fire, the fire automatic, steam-inerting sequence
extinguishing system must be acti- to reduce the potential for pulverizer
Pulverizer fires are a safety vated. A spray of water is introduced explosions when hazardous operating
as well as equipment concern. into the system at multiple locations. conditions exist. The system can also
Prompt measures should The timing is usually at the discretion safely transport the pulverized coal re-
of the operator. However, the pulver- maining in the mill to the furnace, while
be taken. izer must be kept inert until the water maintaining an inert atmosphere inside
injection sequence is completed. When the pulverizer. The fire control system
the pulverizer is empty of its contents, detects fires in operating pulverizers
it can be shut off and isolated. Water and alerts the control room operators.
injection should continue until all The operators can then initiate the fire
evidence of fire has disappeared. Entry extinguishing procedures, which include
for clean-up is allowable only after the water spray injection, steam-inerting, and
mill and its contents have cooled to transporting. A combination mill inert-
ambient temperature. Caution should ing and firefighting system integrates
still be exercised, as smoldering pockets the two subsystems into one complete
of fuel may persist. package.2–4
Before restart, the entire milling sys-
tem should be inspected and cleaned Slagging and Fouling
of any accumulations. The lubricants Chapter 2 and Appendix B describe
in the mill base and rolls should be ash characteristics and their effect on
checked for any evidence of carbon- slagging and fouling. This section will
ization. The compression of journal cover what the operator can do to
springs should be checked again by control slagging and fouling. Suc-
means of a hydraulic jack. If everything cessful boiler operation depends to
is satisfactory, the mill may be returned a significant extent on the ability of
to service. the operating staff to understand how
certain operating variables relate to
Explosions in Pulverizers the fuel properties and furnace sizing
Pulverizer fires occur more often than criteria. Operating variables that influ-
pulverizer explosions. If properly ence slagging and fouling are unit load,
handled, these fires are not overly excess air, fuel fineness, and secondary
dangerous. However, if a fire is not air distribution.
Operation of Steam Generators | 10-13
as sulfuric acid and caustic agents, flush. Operators may follow this flush
may inadvertently enter the feedwater with an acid cleaning. In some cases,
system. Chemicals incorrectly applied a lone detergent flush may be used
during boiler water treatment also can followed in a few years by an in-service
be corrosive, as, for example, sodium acid cleaning. Also, some utilities have
hydroxide used in conjunction with chosen to use a one-stage chemical
sodium phosphate compounds to cleaning. Finally, the superheater, steam
treat boiler water. Corrosion can piping, and reheater will be cleaned by
occur if the sodium hydroxide and steam or air purging. It is important
sodium phosphate are not added to that each of these processes be consci-
the water in the proper proportion. entiously undertaken and result in as
clean a system as possible.
IMPORTANCE OF WATER ANALYSIS
A comprehensive water analysis pro- PRE-BOILER CYCLE
gram should be maintained to ensure All pre-boiler systems of high-pressure
that the feedwater and boiler water boilers must be thoroughly flushed
chemistries are held within prescribed with alkaline or detergent solution to
limits. Continuous, automatic, ana- remove oils, siliceous materials, and
lytical instrumentation is preferred. If particulate matter that are present
automatic analyzers are unavailable or following fabrication, storage, and
are not operational, water tests should erection. It is important that these
be conducted daily or twice a shift; this materials be removed prior to initial
information is provided in the water operation. Otherwise, they will be
chemistry guidelines. Cation conduc- carried into the boiler. Optimal plant
tivity instrumentation or a sodium operating conditions will be realized
analyzer is of particular importance rapidly after start-up if the pre-boiler
in detecting condenser leakages or equipment is satisfactorily cleaned.
contamination of the feedwater. The The condensate system, the feedwater
operator must remember that many system, and the shell side of all heaters
potential tube failures can be avoided should be included in this cleaning
by continual attention to the control operation. This cleaning involves the
of the water and steam environment following basic operations:
throughout the station.
1. Manual cleaning of the condenser,
COMMISSIONING FUNCTIONS all feedwater heaters, and the
deaerator storage tanks.
Before a new unit can be put in service,
2. Gross flushing to waste to remove
the entire system must be cleaned to
the bulk of any loose material.
remove oil, grease, siliceous material,
3. Circulating cleaning solution
mill scale, rust, and any other debris.
throughout the system.
The condensate and feedwater systems
4. Rinsing to remove the cleaning
are cleaned before cleaning the boiler
solution.
so that none of the debris or dirt is car-
5. Laying up to protect metal sur-
ried into the boiler. These two systems
faces until initial operation.
are mechanically cleaned then flushed.
The economizer and boiler will be Cleaning the pre-boiler system
given an alkaline boilout or detergent normally requires the installation of
Operation of Steam Generators | 10-17
hydrostatic test are removed by blow- until the oils are emulsified and the
down during the cleaning process. turbidity has reached equilibrium.
The chemicals used for a detergent Following circulation, the solution
flush are nothing more than a concen- is drained and the boiler rinsed.
trated detergent readily available from The boiler water surfaces should be
major chemical vendors. A wetting checked to ensure cleanliness. The
agent is usually added to enhance unit is now ready for the next stage
cleaning effectiveness. Detergent is of commissioning.
added to the boiler water during or
after filling of the unit. Once filled, the Boilout Procedure
solution must be circulated to be effec- If a fired boilout is to be conducted,
tive and until analytical results show then the chemicals to be added to the
that the oils are neutralized and tur- boiler should be dissolved completely
bidity has stabilized, which will usually before introduction into the unit. In
occur within a matter of hours. thermal (natural) circulation units,
A fired boilout of the boiler is no these chemicals are most suitably
longer common. However, when such introduced into the boiler by blending
a boilout is used, the chemicals are them with water as the unit is being
typically a combination of sodium filled. This ensures a homogeneous
phosphate compounds. Organic deter- concentration throughout the boiler. In
gents are added to boilout chemicals positive (pumped) circulation boilers,
to improve their effectiveness. These the boilout chemicals may be pumped
materials must be used with care and into the boiler without concern about
according to the supplier’s recommenda- proper mixing with the fill water; the
tion. Their indiscriminate use may lead boiler water circulation pumps will
to foaming and carryover of chemical to ensure proper mixing. Gradual heating
the superheater. The temperature stabil- of the boilout solution in the boiler
ity of the organic detergent should be is accomplished by the use of ignitors
ascertained before use in a boiler. and warm-up burners. This ensures
a more even heating of the boiler
Detergent Flush Procedure surfaces. Steam pressure is raised to in-
Detergent flush of a boiler can be crease the saturation temperature and,
done at ambient temperature using a thereby, the thermal circulation of the
detergent solution of 0.6% by volume. boiler water. This promotes good mix-
If the fluid temperature is >150°F ing of chemicals in the boiler circuits.
(65°C), then 0.3% by volume may be The boiler pressure is raised to about
used. In all cases, circulation must be one-fifth of the normal operating
provided by natural circulation, boiler pressure or 300 psig (20 bar), which-
water circulating pumps, or external ever is lower. A pressure of not less
pumps, if no pump is provided on the than 100 psig (7 bar) is recommended
boiler. The detergent can be blended for lower-pressure industrial boilers.
with the boiler water during or after Excellent results have been obtained
filling. The detergent solution is in boiling out at a pressure range of a
usually circulated for 6 to 8 hours, or few pounds to 100 psig with positive
Operation of Steam Generators | 10-19
deposit on a tube with the cleanliness addition of acid to prevent the spillage
1,000
Maximum Pounds of Iron
of the boiler. Tube samples should of the solvent to this section. Thermal
Delivered to Boiler
750 Note:
be taken at yearly intervals from the circulation boilers are generally cleaned
Unless initial high-heat-flux areas of the boiler by this soaking process. Samples are
condensate is
500 discarded, assume (for example, several feet above the taken periodically to check the degree
each outage is equivalent
to 3 months of operation. windbox) or other areas that have, in of reaction in the boiler. Boilers are
250
100 200 300 400 500 600 700 the past, been prone to accumulating soaked for 4 to 6 hours. The acid is
Rated Steam Capacity, 1,000 Lb/Hr deposits. The information obtained drained by pressurizing with nitrogen.
1. Determine Feedwater Iron and Copper
from inspecting the tube samples, in This step is taken to prevent the oxida-
Concentration Weekly conjunction with operational factors, tion of the cleaned surfaces during
2. Assume All Iron and Copper Deposit is used to decide the need for clean- this time. Water is used to displace
in Boiler
3. Multiply No. Million Pounds of ing. These factors include the number nitrogen in rinsing the metal surfaces
Feedwater by Iron Concentration (ppm) of start-ups, the number of periods of the acid. The rinse water is subse-
4. Determine Cleaning Interval from
Maximum Permissible Iron (see note) with condenser leakage, any chemistry quently displaced with nitrogen. Two
5. Total Copper Level Dictates Chemical deviations, and the length of outages. rinses are usually sufficient to attain a
Solvent System
6. Maximum Cleaning Interval Not to
Chemical cleaning of industrial pH between 5 and 6. The boiler is then
Exceed 6 Years boilers should also be performed on a filled with water containing a strong
periodic basis. The primary purpose alkali (trisodium phosphate, soda ash,
Figure 10-4 | Operational chemical cleaning for cleaning these units is to prevent etc.) at a 1.0% solution. The water
guide for industrial-type
Clean Combustion Technologies boilers buildup of deposits to the point at quantity is adjusted to the operating
(based (JG)
Figure: 10.04/ on demineralized or
evaporated make-up water)
which overheating (or corrosion) may level, and the temperature is raised
PMS5415 K
occur. In cases in which deposits con- to correspond to a pressure of about
sist of hardness salts, it is imperative 100 psig (6.9 bar). About 2 hours are
to periodically examine tube samples required to effect neutralization of
from the unit to establish the need the acid and passivation of the metal.
to clean. In units where iron oxide The boiler is drained and inspected
and copper are the main impurities at the conclusion of the wash period.
in the feedwater, the information in Controlled Circulation® and once-
Figure 10-4 can be used as a guide in through boilers can be cleaned efficiently
determining the need to clean. How- because the circulation pumps can be
ever, good practice would still dictate used to equalize acid temperatures and
periodic examination of tube samples concentrations throughout the boiler.
to confirm the analytical evaluation. The acid is circulated intermittently
with one pump during the cleaning,
Acid Cleaning Procedure. In thermal which normally requires 4 to 6 hours.
(natural) circulation boilers, no ef- Organic-type acids are frequently
fective circulation can be obtained at used to clean once-through boilers. At
the low solvent temperature of 150 the conclusion of such a cleaning, the
to 170°F (65 to 77°C). Therefore, boiler is rinsed with water. When the
the distribution of acid strength and system has then been purged of acid,
temperature is obtained by blending condensate containing ammonia is
concentrated inhibited hydrochloric circulated to effect neutralization.
acid and hot water as the solution is in- Because of the close proximity of the
jected into the boiler. Superheaters are superheater to the drum and the po-
flooded with condensate prior to the tential of corrosion from the cleaning
Operation of Steam Generators | 10-23
procedures, with all control systems remain lodged in the reheater inlet
and protective interlocks functioning. header and cause blockage of reheater
The design, fabrication, and installa- tubes, with resulting overheating and
tion of any temporary piping system failure of reheater tubing when the
for the purpose of steam-line blowing, unit goes into operation. For this
as well as protection against overpres- reason, it is recommended that the
sure or overtemperature, require careful temporary piping be arranged such
attention by qualified engineers. that the main steam lines can be blown
to atmosphere first. Similarly, the
General cold-reheat lines should be blown to
Ideally, to obtain optimal cleaning, the atmosphere before they are connected
flow conditions in the system during to the reheater inlet header. Only after
steam-line blowing should equal those these precautionary measures have
during normal operation at maxi- been taken should steam be admitted
mum load. Because it is impossible to the reheater for blowing through the
to exactly duplicate these conditions hot reheat lines.
when blowing through the piping to The temporary piping is often
atmosphere, it is desirable to produce equipped with shutoff valves, which
equivalent conditions by using lower- are used as blowing valves. With this
pressure steam with a flow rate such arrangement, the superheater and the
that the product of steam flow times upstream portion of the temporary
velocity will equal that under normal, piping are maintained at drum pres-
full-load conditions. The determi- sure at all times when not actually
nation of the total obtainable flow blowing. If the unit is equipped with
quantity must be based on flow resis- main-steam stop valves, these may be
tances in the entire system, including used as blowoff valves if the valve de-
the temporary piping. Customarily, sign permits this type of service. When
the designer of the blowing system more than one steam line is provided
makes this determination. Most in any section, consideration should
high-pressure units obtain satisfactory be given to arranging the temporary
results with a blowing pressure in the piping so that each line may be blown
600 to 800 psig (41 to 55 bar) range. separately. In such situations, the
The actual steam blow is executed in piping system should be inspected to
three phases: main steam, cold reheat, ensure that flow unbalances and expan-
and hot reheat. sion stresses at the connections do not
To prevent foreign material from be- become a problem. Impact specimens,
ing transported from one section and installed in the blowoff piping during
deposited in the next, it is important the final blows, give an indication of
that the system be blown by blow- the cleanliness of the section. Polished,
ing each section separately. Particular square bar stock suitably mounted and
care must be taken when blowing supported has been used effectively for
the reheater. If the reheater were to this purpose. The degree of pitting of
be blown immediately in series with the specimen surface following a blow
the superheater, material too large to determines the end point of the blow-
pass through the reheater tubes would ing cycle for the particular section.
10-26 | Clean Combustion Technologies
turbine to the inlet of the cold reheat start-up, as well as the steam line blow-
piping. Temporary blowoff piping is ing, must be conducted with great
installed at the reheater inlet. Blowing care. The unit must be brought up
is executed in the same manner as for much more slowly than during sub-
the main steam line, but now the blow sequent normal start-ups so that all
is through the superheater, main steam equipment can be checked and expan-
line, cold reheat line, and temporary sion movements monitored closely.
blowoff piping to waste. Again, the When the unit is fired, the furnace
first blow is at reduced pressure, with exit gas temperature limitation must
subsequent blows from 600 to 800 psig not be exceeded.
(41 to 55 bar) down to 200 psig (14 During the blows, the drum-
bar) drum pressure. This cycle is also water level will be subject to extreme
repeated until the impact specimens fluctuations. As the temporary blowoff
are satisfactory. valve is opened, the drum-water level
will rise rapidly and may disappear
Blowing the Hot Reheat from sight in the gage glass. As the
Lines—Phase III blow progresses, the drum-water level
The temporary blowoff line at the will reappear and may drop out of
outlet of the cold reheat piping is sight. Consequently, it is important to
removed, and the cold reheat piping is establish the drum-water level at or
permanently connected to the reheat- slightly above normal operating level
er. A temporary blowoff connection before the start of each blow. Feed
is installed at the hot reheat piping as water flow must be established as soon
close to the turbine as possible and as the water level drops back in sight to
piped to waste. The same blowing prevent an excessively low water level
procedure is used as in phases I and with resulting suction loss of the boiler
II. When the specimen results are circulating pumps. After the procedure
satisfactory, the steam blows can be is completed, the temporary piping is
considered successful. removed and final piping connections
are made. The unit is now ready for the
Operating Procedures next commissioning function, setting
The unit is started in the normal the safety valves.
manner, following the cold start-up
procedures. All normal recommenda- Safety-Valve Setting. All safety valves
tions and limitations with respect installed on boilers or other pressure
to pumps, fuel-firing equipment, vessels should be test-operated before
ash-handling equipment, drains, and the boiler or vessel is placed in service.
vents should be followed, as if the unit This test should involve a check of
were being started for synchronizing the proper functioning of the valves
the steam turbine. Before starting the as to correct opening pressure, correct
first blow in each section, the econo- blowdown, proper mechanical opera-
mizer recirculating line valves should tion without chatter, and clean closure
be closed. Because the steam blowing without leakage.
operation is the first occasion that the In checking opening pressure and
unit is fired at any significant rate, the blowdown, a calibrated pressure gage
10-28 | Clean Combustion Technologies
must be used. The location of its con- company should retain the report as a
nection on the vessel, header, or other permanent part of their records.
component must be such that the Before testing safety valves for the
pressure it shows is a true indication of first time on new units, the steam lines
the pressure existing at the safety-valve will have been blown to eliminate as
inlet, both when it opens and when much of the foreign material as possible
it closes. When valves are located on from the superheater, reheater, and the
headers or steam lines through which connecting piping to prevent damage to
steam is flowing, the test pressure gage the safety valves. Damage to the valve
should be connected near the valve seating surfaces from grit blasting by
to eliminate any effect of pressure mill scale, weld beads, and other such
drop resulting from the flow. In some dirt is common during initial test-
cases the flow of steam being relieved ing. Adequate steam blowing prior to
through the valve itself could produce a popping valves will minimize damage.
In order to avoid damage to significant pressure drop. Valves mounted on the dead end of
the valves, maximum attention For high-pressure boilers (above headers are particularly prone to this
must be given to debris removal approximately 1200 psig (83 bar) damage. Inspection and manual cleaning
operating pressure) a manufacturer’s of the header by vacuum hose and/or
before testing and setting safety representative should initially test and magnets through the valve nozzle, with
valves. adjust safety valves. Certain local codes top works removed, is sometimes neces-
require that only licensed people may sary. With maximum attention given to
make adjustments on safety valves, removing debris before initial popping
so check with the insurance carrier of the valve, better success will be ob-
before making any adjustments. The tained with initial valve setting.
setting, or adjustment, of safety valves Leakage of safety valves in opera-
should be done by a competent person, tion usually results from one or more
familiar with their construction, of the following:
operation, and maintenance. A written
1. Seating-surface damage
report should be made of all testing
2. Externally imposed stresses on
and adjustment, including the open-
the valve body, which distort the
ing pressure, closing pressure, and/or
seating surfaces
blowdown and an indication of correct
3. Operation at pressure too close to
functioning of the valves as they are
the set pressure of the valve; the
left. In addition, the report should
valve manufacturers recommend
include the name and title of the re-
a minimum difference between
sponsible plant observer and the name
popping pressure and operating
and title of any insurance inspector,
pressure of 5% of the popping
or state or local boiler inspector, who
pressure. With less than a 5%
witnesses the test and adjustment. The
difference on a continuous basis,
name of the valve vendor’s representa-
problems with leakage and result-
tive making adjustments, if any, should
ing frequent valve maintenance
also be included, as well as the name-
may be expected.
plate data (serial number, set pressure,
blowdown, and valve location) of Safety of personnel and prevention
each valve on the unit. The operating of damage to the equipment should be
Operation of Steam Generators | 10-29
the prime concerns when testing and 4. The exhaust elbow attached to
adjusting safety valves. Tremendous the valve body is not of exces-
forces are involved when these valves sive length in the horizontal run
operate. All precautions necessary to (maximum recommended, 24
contain these forces and prevent ac- inches [61 cm]) that may result
cidents must be taken. For the initial in abnormal stresses being put on
popping of each valve, all personnel the valve body due to the reaction
should be kept at a safe distance. A force when the valve is blowing.
rope should be attached to the manual 5. The drain piping from the valve
lifting gear so that any uncontrollable body and drip pans, elbow, etc.,
valve chatter can be prevented from do- is installed and discharges into
ing serious damage. Before starting the a location that will prevent
actual setting, refer to the latest section injury to personnel when the
of the ASME Boiler and Pressure Ves- valve is blowing.
sel Code applicable to safety valves. 6. If flexible hoses are used between
the exhaust elbow and exhaust
Procedures Prior to stack, they should be of sufficient
Safety-Valve Setting length and installed so that they
Before firing the unit for setting the do not become “solid” in any posi-
valves, check that: tion of the valve.
7. All components of the safety
1. Hydrostatic test plugs have been valves, such as manual lifting gear
removed from the safety valves. and adjusting ring pins, are in
If a problem of scheduling a place and secure.
valve vendor’s representative
for removal of hydrostatic test
plugs is involved, the plug from Hydraulic Jack Method for
one valve must be removed prior Safety-Valve Setting
to the first fire. All plugs must The use of a hydraulic jack has made
be removed before the unit is setting safety valves easier. These jacks
brought up to full pressure for can be bought or rented from the
the first time. valve manufacturers. This method
2. There is no physical interference was developed originally for setting
that would prevent functioning supercritical pressure valves so that the
of the safety valves. Expansion of lift could be limited to reduce damage
the unit should be considered in to the seating surfaces. The hydraulic
this check. jack system contains a pump, hydraulic
3. The exhaust stacks from the piston, yoke, turnbuckle, and pressure
valves are firmly supported and gage. The hydraulic piston overcomes
restrained and that no physical some of the safety-valve spring force
interference that will put external so that the valve may be set at lower
stresses on the valves exists steam pressures. This method has also
between exhaust stacks and the been used widely on reheater safety
valves in the cold position, hot valves as well as with high-pressure
position, or in between. valves (drum and superheater) on
10-30 | Clean Combustion Technologies
air are stopped. Bed temperature can pump, which allows water to be fed to
remain above the firing permissive level the boiler during a power interruption.
for several hours, such that no purge
is required on restarting. On a hot TUBE RUPTURE
restart, the unit can be brought back to When limestone is used as the sulfur
full load within a few hours, depend- sorbent, the heated bed material
ing on the bed temperature level. If no will contain significant amounts of
hot restart is planned, the fuel and air calcined, dehydrated calcium oxide. If
flows are gradually reduced, so that the this lime comes in contact with suf-
rate of refractory temperature change ficient water, such as from a tube leak,
must be limited to no more than 250°F it will harden when allowed to dry, and
(121°C) per hour. will make cleanup and repair difficult.
The leaking water or steam may also
PLANT MAIN FUEL TRIP (MFT) damage relatively sensitive refractory
Stored heat in the bed material will linings. Quick action by the operators
continue to generate steam long can minimize the duration of the out-
after fuel flow stops. In a CFB, stored age and reduce the extent of cleanup.
heat in the cyclone refractory system When a tube leak is detected, it is
contributes to steam generation. important to maintain fluidization and
Precautions must be taken to prevent to drain the bed material as soon as
damage to backpass, steam-cooled sur- possible. Acoustic steam leak detectors
faces from this heat source. The drum are an asset for the timely detection of
level should be maintained by a slow tube leaks.
feed of water to the drum following an
MFT. If cyclone cooling is not possible PROBLEMS IN OPERATioN
in any other way, a small steam flow In spite of adherence to the general
can be induced through the backpass procedures and observance of the
by opening a steam vent to cool the cautions presented in this chapter,
backpass tubing. Steam-cooled surfaces operational problems can still occur.
in the combustor and/or FBHE are Tubes may rupture for a variety of rea-
protected from the loss of steam flow sons. The size and location of the break
in accordance with the manufacturer’s will determine what action is necessary.
instructions, which may call for On a drum-type unit, the best method
depressurization or a small, cooling for shutting down will be dictated by
steam flow by means of a steam vent or the ability to maintain normal water
turbine bypass. level in the boiler and the need for the
boiler in service. If the water level can
LOSS OF POWER be maintained, the unit can be kept
During a power loss, feedwater flow to in service until after a peak, or service
the boiler stops. As a result, stored heat may be stretched to a weekend outage.
can evaporate significant amounts of There is, however, the danger of high-
the water inventory. Whether this will pressure water from a break cutting
damage the boiler depends on the spe- other tubes. If the rupture can be
cific design. The conservative approach observed visually and it is ascertained
is to provide an emergency feedwater that it is blowing out into the furnace
Operation of Steam Generators | 10-33
and not damaging other tubes, the unit • at leaky fuel-inlet valves on idle
may be able to operate for a long time wind box compartments
provided that sufficient treated make- • when the fire is extinguished and
up water is available. the fuel is not shut off promptly
Unlike a waterwall tubing leak, • when the fuel does not burn as
a leak in a superheater, reheater, or rapidly as it enters the furnace
economizer element requires greater • if difficulty occurs in establishing
attention. Because of the physical ignition
arrangement of such surfaces, steam
cutting of adjacent tubes can result, Explosion Prevention
making a major repair job out of what Because most explosions occur during
might have been a simple and short periods of low fuel input, maintaining
one. If left unattended, economizer a minimum of 30% of full-load airflow
ruptures can lead to plugging of the is important to ensure an air-rich mix-
economizer and air heater because the ture in the furnace and to sweep out
water mixing with the fly ash can set, any accumulation of unburned fuel.
similar to concrete. Other preventive measures include:
A leak in a waterwall of a super-
critical unit can result in rapid and • Ensuring that all liquid or gaseous
extensive damage, not only to the fuel valves are tightly shut on idle
leaking tube, but also to nearby tubes. fuel compartments
Thermocouples are installed on indi- • Watching the fires closely at low
vidual outlet tubes and representative loads and shutting off all fuel
inlet tubes to alert the operator to such immediately if proper combustion is
leaks. Any decision to operate a once- not maintained. Most modern units
through unit with a known waterwall have flame scanners to trip the unit
tube leak must be made with the full automatically when poor ignition
knowledge of the serious damage that occurs. The scanners must be prop-
may be incurred. erly maintained. They should never
be removed from the safety system
FURNACE EXPLOSIONS or their outputs defeated
Furnace explosions usually occur • Always using the required ignition
during start-up, shutdown, or low- energy source when placing any
load operation. Generally, they result pulverizer in service
from the accumulation of unburned • Keeping adjacent pulverizers in
fuel in the furnace because of incom- service during low-load operation
plete combustion, loss of ignition, or • Regularly checking the proper
fuel-valve leakage. An explosion occurs functioning of any furnace safe-
when the proportion of unburned guard system
fuel and air is in the explosive range • Never defeating any portion of a
and some heat source increases the safeguard or interlock system
temperature of the mixture to the • Purging the furnace before shutting
ignition point. Unburned fuel, which off the fans on a unit trip
causes such fires, can accumulate in the • Emptying the pulverizers of all coal
furnace in several ways: before bottling up the unit
10-34 | Clean Combustion Technologies
Other operators keep the problem out • During reclaiming, avoid pushing
of the boiler room by limiting moisture the coal onto any water deposits or
pickup in the coal yard and removing muddy coal in the area of the coal
all large debris before the coal arrives at conveyor belt
the bunkers. • Minimize fine coal, as it absorbs
Coal usually arrives from the supplier water more readily than coarse coal.
reasonably dry and, if properly stored, Wet, fine coal does not flow easily.
will remain that way. If a small por- It will stick to bunkers and feed
tion gets wet, it can be moved aside pipes. Coarse particles will help
for drying and later use. On the other to eliminate bunker plugging and
hand, to minimize fugitive dust, coal keep the coal moving
is sometimes wetted before shipment
from the mine. Additional moisture
pickup during transport can result from Coal Yard Design and Management
leaky barges or rain. Stations receiving Good coal yards should be carefully
coal with a substantial moisture increase planned. Proper drainage is a must.
from mine to plant should consult with The yard should be properly graded,
the producer and the shipper to see if and all rocks, wood, and metal re-
improvements are possible. moved. A base layer of coal should be
spread, to a minimum depth of 2 feet.
Guidelines on Coal Piling This base should never be reclaimed
and Reclaiming and should not be included in the
• Compact the coal. Loose coal picks stockpile records, even for emergency
up moisture and encourages fires in use. The coal yard should be fenced,
the coal pile or otherwise isolated, from material
• Pile the coal for maximum runoff storage, scrap, or trash. Only coal, and
of rain. Piles should be rounded or the equipment needed to handle it,
pyramid-shaped, with the steepest should be allowed in the area. Work-
slopes possible. Consider cover- ers should be encouraged to look for
ing the coal pile if rain is a very foreign material in the yard and to
frequent problem remove it. The screens should not be
• Keep the coal pile free of valleys depended on to remove all debris.
and pockets. Water will collect in Avoid crushing coal until it is ready to
them and sink down into the coal enter the bunkers. Wait until immedi-
• Reclaim deep into the pile, not ately before the boiler is ready to burn
along wide areas of the top. When coal before filling the bunkers. There
properly piled and compacted, is a tendency in new plants to fill the
only the top layers will be high in bunkers as soon as the coal-handling
moisture. This top layer can be system is ready. This is understand-
reclaimed during drier spells able, since start-up personnel want as
• As coal is removed, rework the much as possible of the equipment to
pile to fill in the reclaimed area be operational and ready. However,
and eliminate gulleys, pockets, coal will gradually pack in the bunkers
and rivulets and will not flow easily when needed.
10-36 | Clean Combustion Technologies
1. Edward S. Sadlon and Guenter Scheffknecht, “The Supercritical Steam 3. Explosion Prevention Systems (NFPA 69). Quincy, MA: National Fire
Power Plant: Operational Success and Technological Advancement.” Protection Association, 2008.
Alstom, Windsor, CT, 2000.
4. Boiler and Combustion Systems (NFPA 85) Chapter 9, Pulverized Fuel
2. S. E. Kmiotek and R. E Hickey, Coal Pulverizer Inerting and Fire Fighting Systems. Quincy, MA: National Fire Protection Association, 2007.
System. Windsor, CT: Combustion Engineering, Inc., publication TIS-
5. George Thimot, “Want Better Coal Firing? Improve Coal Handling,”
8256.
Electrical World, August 1, 1972.
10-39
Construction and Maintenance
Chapter Eleven
11-
11- | Clean Combustion Technologies
Structural
Drum U-Bolts
Steel Framing
Steam Drum
Pressure-Part
Superheater Panels Support Steel
Finishing
Superheater
Superheater or or Reheater
Reheater Panels Convection
Superheater
Furnace or Reheater
Front Wall Economizers
Buckstays
Downcomers
Coal Silo
Windbox
Air Preheater
Coal Feeders
Furnace
Pulverized- Side Wall
Coal Piping to
Windboxes Forced
Draft Fans
Pulverizer
Boiler-Water
Circulating Pump
Primary Air Ducts
Tilting Tangential to Pulverizers
Fuel Nozzles
Pre-Construction
• Interface with the engineering team
At the award of a contract to furnish to determine the sizes and configu-
and install
CleanaCombustion
steam generating unit,
Technologies ration of the shippable pieces.
the initial construction
Figure: 11.01/MH team, con- • Interface with the procurement team
sisting PMS5415
of the SiteK Manager, Project to ensure that shipments are synchro-
Controls Manager, Scheduling Man- nized with the construction sequence.
ager, and Construction Engineer, will • Determine the site staff and select
be assigned at the start of engineer- candidates.
ing. The responsibilities of this • Interface with the applicable craft
team include: unions involved to coordinate labor
demand.
• Interface with the client’s and • Determine work to be subcon-
architects/engineers (A/E’s) staff tracted, prepare bid packages, and
to establish site services. negotiate contracts.
Construction and Maintenance | 11-
Canada, and in some cases, other coun- for Code Compliance meets the ASME
tries, the field erection must meet the B&PV Code for the erection of steam
requirements of the American Society generators and more. Hartford Steam
of Mechanical Engineers Boiler and Boiler Inspection and Insurance
Pressure Vessel (ASME B&PV) Code. Company serves as the third-party,
Specifically, the company must comply authorized inspection agency for
with the requirements of Section I the ASME B&PV Code and also
(Power Boiler), Section V (Nonde- as the registrar for the ISO 9001.
structive Examination), and Section IX The quality program is implemented
(Welding). Other sections may be through a Quality Systems Manual,
required for other components of the which meets the requirements of ISO
power plant, such as Section VIII 9001, and a Quality Systems Man-
(Pressure Vessels). Companies must ual for Code Compliance. The latter
be certified to perform certain activi- manual is reviewed by the Hartford
ties associated with the engineering, Steam Boiler Inspection and Insur- The International Standards
manufacture, and erection of the steam ance Company (Alstom’s authorized Organization (ISO) addresses
generator. They obtain their certifica- inspection agency) and is certified
quality assurance programs on
tion by providing evidence to a survey to meet the requirements of ASME.
team, which is made up of a represen- The quality program manuals are a worldwide basis.
tative of the ASME B&PV Code and supported by a subset of documented
a representative of the National Board procedures, which include quality
of Boiler and Pressure Vessel Inspec- common procedures, work instruc-
tors (NBB&PVI), that their quality tions, quality-assurance instructions,
control program addresses the require- and quality plans.
ments of the Code and assures that The quality control process begins
the Code will be met. The ASME then in the negotiation stage with a review
issues the applicable Code Stamps and of the owner’s and the engineer’s
Certifications for the specific work the specifications. After a contract has
company wants to perform. Similar been awarded, a specific quality as-
authorities and code bodies are found surance program is established that
throughout the world. meets the code requirements and
The ASME B&PV Code addresses any specific owner requirements. The
the specifics of the steam generator, specific quality-assurance program will
but another quality-assurance program address document control, purchasing
addresses even more, such as the tender- control, process control, product ID and
ing, planning, and business procedures traceability, inspection and testing, non-
of a company. This is the International conformance reporting, corrective and
Standards Organization (ISO) 9001 preventative action, material handling
Standard, which Alstom follows. As and storage, quality records, audits and
with the ASME B&PV Code, the ISO training, and other areas as required by
9001 also requires a quality-assurance the contract and scope of work.
program and certification. Alstom’s A significant percentage of the jobsite
quality programs for the ASME B&PV Quality-Control (QC) Superintendent’s
Code dovetail into the ISO 9001 time is dedicated to assuring that field
quality program. The Alstom Quality welding, inspection, and documentation
Systems Manual meets the requirements meet the requirements of the ASME
of the ISO 9001, and the construction B&PV Code. Inspectors are trained
division’s Quality Assurance Manual and certified to stringent criteria to meet
11- | Clean Combustion Technologies
a
weight of the individual girders and girder raised off the ground. Because
the capacity of the crane, the crane the girder rests on top of the structural
may need to be relocated for each of columns, its total length is longer than
the girder lifts. For the very heavi- the open area between the columns.
est girders and highest structures, it This requires the girder to go up at an
may be necessary to lift the girders angle in order to fit between the steel,
using two cranes. Coordinating two as can be seen in Figure 11-4B. Once
cranes to share the load as planned the girder has been raised above the
and maintain safe working distances column tops, the girder is rotated to
between the girder being lifted and its correct orientation. Then, the girder
the crane booms greatly increases the can be set down on the column tops,
difficulty of this process. as seen in Figure 11-4C. At this time,
Main girder erection is one of many the top flange of the girder is very close
rigging problems of the structural to the tubular framing on the boom. If
steel erection process. Lifting lugs are the girders begin to swing and touch
11- | Clean Combustion Technologies
the boom, the boom could collapse. by means of huge U-bolts that encircle
Once the girder is in place (Fig. 11-4D), the drum. Currently, utility boiler
intermediate steel, perpendicular to steam drums may be 7 feet (2.1 m) in
the girder, can be installed to tie all outside diameter and more than 100
of the upper steel together. Coordi- feet (30 m) in length, and may weigh
nated radio communication, highly as much as 330 tons (300 tonnes).
experienced operating engineers and
rigging supervisors, and first-class RAISING A DRUM. As shown in Figure
equipment are imperative for successful 11-5, the drum is often raised at an angle
Figure 11-5 | Lift of a steam drum in the girder erection. to the ground, because the bay may
boiler cavity not accommodate the full drum length.
Erection of Pressure Parts Based on the lifting lugs welded on the
Support Steel drum in the shop, the location of the
With all of the girders and intermedi- support steel for the load blocks can be
ate steel in place, aligned, and bolted, determined before construction begins.
the pressure parts support steel (see The drum lifting equipment is selected
Fig. 11-1, upper region) can be installed. based on the weight of the drum, the
This steel, located at the top of the unit, angle of lift, the movement to the front
supports all of the weight of the boiler. of the boiler cavity that will be required
Unlike conventional structures that are after the drum has been lifted, and the
built from the ground up, most modern configuration of the upper steel. Strand
boilers are built from the top down, with jacks, or block-and-tackle rigging, are
Ź}Ź~
the whole boiler suspended by hanger used based on the conditions encoun-
ČŅŅčńʼnĞŷ
ʼnňŅʼn
rods from the main girder elevation. tered.
This permits the boiler to expand down- The steam drum usually is brought
ward by as much as 15 inches (380 mm) inside the furnace cavity on a multi-
as it heats up during the start-up of the axle carrier and off-loaded with the
plant. Correctly setting the pressure lifting equipment. In the absence of
parts support steel is very important, access for a carrier directly under
because every part of the boiler is the rigging, the crane used for steel
located with respect to this steel. It erection can off-load the drum and
controls all elevations and is continually “walk it” to the lift area. In some
used as a reference to align the different cases, additional rigging is required to
components as they are erected. maintain adequate clearances between
the drum and the structure during
Installation of Steam Drums drum erection. This is accomplished
Supercritical boilers do not have steam by installing block-and-tackle rigging
drums. Therefore, this section is not to pull the drum horizontally during
applicable to supercritical boilers. the lift. Proper location of all rigging is
Subcritical boilers, including CFBs, do important because of the height of the
have steam drums. boiler and the lifting distance, which
The steam drum can be raised into can be well over 200 feet (61 m).
position at the top of the boiler once
the main girder steel has been erected. “DRIFTING” A DRUM. On some jobs, it is
As shown in Figure 11-1, this drum is advantageous to erect the support and
hung just below the main girder steel platform steel that is under the steam
Construction and Maintenance | 11-
positioning for welding. As soon as the some of the most difficult rigging
downcomers are in place, boilermaker on the boiler, contain sections of
welders can begin the fit-up and weld- tubes bent in complex configurations
ing procedures that will ultimately join and welded together on a common
the downcomers to the drum. A joint section of header. Each module can
can be finished in about 1½ days if a measure as much as 10 ft × 10 ft
two-person crew works on each weld. (3 m × 3 m) and 55 feet (17 m) in
On supercritical units, the steam length and weigh up to 70 tons (63
separators, storage tank, downcom- tonnes). They arrive on the jobsite
ers, and circulating pumps replace the partially encased in shipping rigs; these
drum and downcomers. The storage shipping rigs also are used to upend
tank is a single vessel oriented verti- the module to a vertical position.
cally and weighing up to 50 tons (45 A large crane and a cherry picker
tonnes). The separators and storage are needed to off-load the modules
tank are hung in the steel during from the railroad car to a flatbed truck,
erection of the structural steel. Before which brings them into the furnace
the front waterwalls are installed, the area. A set of 75 ton (68 tonnes)
downcomer manifold and downcom- capacity blocks are rigged above the
ers are installed, starting at the storage final position of the module. The lower
tank and working down. Later in the block is attached to a special spreader
construction sequence, the circulating beam, as shown in Figure 11-6.
pumps and links to the lower headers The lower end of the module is
are installed. picked up with a cherry picker, and the
main load blocks raise the header end.
INSTALLATION OF SUPERHEATER AND After the flatbed trailer is removed, the
REHEATER MODULES. The next boiler main blocks continue to raise the up-
components to be installed are the per end of the module while the cherry
superheater and the reheater picker lowers the other end, thus ori-
assemblies and modules. Figure 11-1 enting the module in a vertical position
shows their locations within the for final raising. Rotation to the vertical
boiler. These modules, which require position is accomplished quite easily
with specially designed rods between
the high-crown end bar of the module
}
and the spreader beam. The cherry
Ÿ
} Ÿ
picker and shipping frame are then
Ĥ~ĥ Ŷ} Ĥ
ĥ
removed, and the module is lifted with
}ď the 75 ton (68 tonnes) blocks.
}
Because of the thick-wall tubes that
}
Ŷ}
are required to retain high-pressure
Ź steam, the superheater modules are
Ź}~
} Ŷ} much heavier than the reheater modules.
When a complete section of superheater
}
module is in place, the header sections
Ŷ}}Ĥ
ĥ
}
are fitted together and prepared for
welding. Girth welds, made on the
Figure 11-6 | (A) Rigging for superheater; (B) raising of superheater module
Construction and Maintenance | 11-11
the grating, the waterwalls, which form slots be cut in the welded web between
the sides of the furnace, can be erected. the tubes. A bar welded to a small
The waterwalls are sections of vertical T-beam is inserted and pinned on the
tubes that are fusion welded together back; this device assists in keeping the
into a gas-tight panel. They conduct a panel straight along its width during
mixture of steam and water upward as the upending process. When the panel
heat is applied from the furnace side. is vertical, the tailing crane is released,
The walls extend from the base of the and the waterwall is lifted up to its
boiler (just above the ash hopper) up final position.
to the steam, or roof, level.
Waterwall panels may come from ERECTION OF ROOF TUBES. After the
the shop with the upper headers waterwalls are in place, the roof tubes
already attached. When two panels are can be erected, usually in one of two
joined in the field, only one girth weld ways. One approach uses single tubes
is then needed on the header. This that fit between the vertical elements.
eliminates the need for numerous field In other areas with larger openings,
welds between header nipples and the full roof tube panels can be raised
panel tubes. Such shop prefabrication from the ground with a set of blocks
can reduce the construction schedule and a tugger. While the roof tubes are
and costs by eliminating field welds, being fitted into place, the intermedi-
which often are performed under ad- ate waterwall panels are lifted and
verse weather or logistical conditions. fitted below the upper panels. Gener-
Other examples of shop prefabrication ally, the intermediate waterwall panels
of waterwall panels include welding of are attached to the upper panels by
buckstay stirrups, installing insulation a system of threaded rods and clips
pins, and fitting sealboxes. Another that are welded to the webs between
helpful shop procedure is the applica- the panel tubes. When this temporary
tion of white paint to each panel in hanging system is used, final panel
order to brighten the interior of the spacing for welding can be done with
boiler during erection; this technique the threaded rods.
improves working conditions.
Waterwall panels typically are ap- SUPPLEMENTARY FRAMING. Figure 11-7
proximately 12 feet (3.7 m) wide and shows the structural framing between
up to 80 feet (24 m) long. They are the furnace “nose” and the vertical front
brought into the furnace on a flatbed wall of the convection pass of a large,
truck. Because they are so long and high-pressure steam generator. All of
flexible, the biggest problem with the steelwork shown is supported by
these panels is upending them. This is the pressure parts. Both the pressure
performed by attaching either the fixed parts and these structural members
block-and-tackle rigging or the in- must be designed to withstand the
furnace crane to the upper end of the loads imposed as well as both positive
panel and then using a smaller hydrau- and negative gas pressures as dictated
Figure 11-7 | The structural framing between lic crane on the lower end to maintain by codes and owner specifications. The
the furnace “nose” and the
clearance. Often, “spreader beams” are horizontal tubing shown at the top of
vertical front wall of the
convection pass of a large, high- required to distribute the load during Figure 11-7 forms the floor of the gas
pressure steam generator upending. This requires that several pass directly behind the furnace nose.
Construction and Maintenance | 11-13
ERECTION OF AIR PreheatER. Depend- clearances between the rotor and seals,
ing on the location of the air heaters aligning the bearing and support struc- Extremely accurate
with respect to the backpass and the ture, and setting the rotor motor and measurements are required to
specific plans for erecting the back- roller pins. Boilermakers erect the main
pass, the air heater can be built early housing and raise the air preheater
align the air heater due to its
in the construction span or be left out into place. They also level and set the large size, its rotation during
and started after completion of the preheater in its final position within operation, and its expansion
backpass. The air preheaters are heat the boiler. and contraction due to
exchangers made of heat-absorbing
Erection of Backpasses temperature changes.
metal baskets assembled on a rotor
that is turned by a drive. Hot flue gases The backpass, like the furnace section,
leaving the boiler pass through one is made up of welded tube walls and
side of the heater. Incoming cold air internal-heat transfer surface. Unlike
from the forced draft fans passes in the furnace waterwalls, the backpass
counter-flow through the other side. walls are normally 1¾- or 2-inch (4.4
The “baskets” are heated by the hot or 5.1 cm) tubes on 4 to 5 inch (10 to
flue gases and rotate into the cold air 13 cm) centers. The surface is always
stream to raise the temperature of the horizontal elements. Superheater,
incoming air. Various seals prevent the reheater, and economizer surface is
gas and air flows from mixing. included in the backpass.
The air preheater most often is
oriented with a vertical shaft. Its ERECTION OF BACKPASS WALLS. The
main parts are the bearing, bear- backpass walls are erected in the same
ing supports, baskets, oil circulating manner as the furnace walls. Because
system, drive motor, seals, casing, and they do not extend as far as the furnace
insulation. On some projects, it is eco- walls, they generally are shorter, lighter,
nomical to fabricate the main sections and easier to handle. They usually are
of the preheater on the ground and erected with block-and-tackle rigging,
raise them into place in the struc- because a crane is not practical here.
tural framework with a crane from The sequence of erection is the side-
behind the unit. This method is very walls, front and rear walls, and then the
satisfactory if the same crane used for roof. The backpass walls carry steam
erecting the structural steel is avail- from the steam drum. Thus, they are
able and access exists. The air heater the first stage of superheat and feed the
baskets usually come assembled into horizontal surface in the backpass.
the rotor sections, forming large, pie-
shaped sections. These are raised with ERECTION OF BACKPASS SUPERHEATER,
a crane and lock into place on the REHEATER, AND ECONOMIZER. The internal
rotor shaft. After all of the pie-shaped surface in the backpass is supported
sections are installed and welded by economizer hanger tubes. These
together, the pin rack drive and the tubes extend from the outlet of the
radial and axial seals are installed. economizer elements through the
The extremely accurate measure- superheater and reheater elements to
ments required to align this large, the economizer outlet header above
rotating component are performed by the roof. Hangers are attached directly
millwrights. These include setting the to the roof tubes to support all of
11-14 | Clean Combustion Technologies
the weight hanging in the backpass. made up of two to six elements. These
The superheater and reheater surface are upended and raised in the same
generally is in the upper section of the manner as described for the super-
backpass. heater and reheater assemblies, except
Figure 11-1 identifies the location that the supports are mechanical, so
of the backpass components. These pressure part welds are not required
elements normally ship in assemblies for support.
of two elements attached to a hanger Two types of economizer heating
tube. If multiple elevations of surface surface are in common use on coal-fired
are in the backpass, they can be welded units. One consists of a pair of longitu-
together on the ground before instal- dinal fins located 180° apart along the
lation. An upending jig is required to top and bottom of the horizontal tube.
take the assemblies from horizontal The second is bare tubing, similar to a
to vertical in preparation for hanging. horizontal, low-temperature, superheat-
The assemblies are brought in under er surface. Spiral finned economizers
the backpass on a truck or transporter. are used on oil- and gas-fired units.
The rigging will be attached to the Finned economizer tubes require careful
assembly, an upending frame will be at- handling because of the heat-absorb-
tached, and a tailing crane will be used ing fins. They are grouped together in
to upend. Once the assembly is vertical, modules of three to eight elements,
the upending frame is removed, and each with saddle supports installed to
the assembly is raised to its final loca- hold them together. They are raised
tion. The assembly is then transferred into place from the ground with a
to a threaded rod-and-bracket arrange- tugger, rigged in the area of the outlet
ment on the hanger tubes, where it is headers. Pins and rods usually support
prepared for welding. The threaded rod the assemblies directly to lugs welded
allows minor adjustment for fit up of to the bottom of the intermediate
the weld joint. Normally, a two-bank headers. Because the economizer tubes
assembly needs to have the hanger often are fabricated in a staggered pitch
tubes welded before lower banks can to increase heat transfer and minimize
be erected. space requirements, fitting the last few
The economizer is a heat recovery elements sometimes presents a rigging
device that transfers heat from the exit- problem. It is difficult to pass hoisting
ing flue gases to the incoming boiler falls through the staggered tubes. This
feedwater. The economizer consists of often means that the already-installed
an inter-connected, horizontal tube elements must be spread apart and the
array. The upper ends of each econo- side walls moved out.
mizer assembly connect to the outlet
header, which is already in place above Erection of Windbox Assemblies
the roof tubes. The economizer banks After the intermediate waterwall pan-
are located in the backpass below els are in place, the windbox assemblies
the superheaters and reheaters. The (Fig. 11-8) can be brought into the
economizer is supported on straps that furnace and positioned for lifting. The
are attached to the lower end of the rigging of a furnace windbox assembly
economizer terminal tubes. The econo- is one of the most difficult tasks on the
mizer normally ships in assemblies project because of the weight, shape,
Construction and Maintenance | 11-15
Erection of Pulverizers
Coal feeders, coal pulverizers, and
coal piping to the windbox usually are Figure 11-8 | Windbox assemblies being raised into place
erected as part of the steam generator
contract. If shipping schedules permit,
coal pulverizers, such as those shown set on top of the foundation. The Cleanmill
Combustion Technologies
in Figure 11-1, can be set in place, in side assembly, where the grinding Figure: 11.08/AR
bowl
sections, as the structural steel is set. is located, is installed on top ofPMS5415
this. K
A large unit generally has 6 to 10 coal The separator body, which contains the
pulverizers, which are located on both three roller journals and covers, goes
sides or in front of the boiler at ground on top of this. The separator, where
level. Generally, one pulverizer on the coal is classified cyclonically to the
each side is left out until most of the desired fineness, is next. The very top
required components have been moved of the mill is the multi-port discharge
into the furnace cavity. assembly. Here, coal is fed into the mill
A typical bowl mill contains about from a central feeder pipe, and pulver-
six sections, the first of which is the ized coal is discharged to the coal
poured concrete foundation with four piping through the peripheral ports.
large anchor bolts. The mill base, which Mills weigh up to 150 tons
contains the gear case on many mills, is (136 tonnes) and are up to 30 feet
11-16 | Clean Combustion Technologies
(9 m) high. The crane that erects the plings join sections of the pipe. Figure
structural steel can easily set the heavy 11-9 shows the final tightening of a
sections of each mill. The lighter in- Victaulic coupling on an elbow that
ternals can be installed later. If the use connects to a windbox assembly.
of a crane is impractical, the different
mill parts can be rolled into place on Erection of Suction Manifolds
a track and then jacked into position. The suction manifold, also shown in
Ironworkers and millwrights do most Figure 11-1, is raised into place next.
of the construction work on the mills, It is the collecting header that distrib-
with the latter doing all of the align- utes the water from the downcomers
ment and settings because of the close to the boiler circulating pumps. The
tolerances that are required. During suction manifold is hung from the
the work on the bowl mills, hanging downcomers, which are welded to the
and setting of the coal piping to the steam drum. The downcomer leads
furnace can be progressing as well. join the top of the manifold. Lugs are
Welded joints and Victaulic cou- provided on the downcomers (and also
on the nipples of the manifold) so that
threaded rods can be inserted to hold
the manifold after it is lifted. These
also can be used to adjust the weld gap
for proper fit.
The manifold usually is lifted into
place using the crane in the furnace
cavity. If the crane is not used, blocks
must be rigged under the steam
drum, with the rigging reaching down
the full height of the boiler to hoist
the manifold into place. The boiler
circulating pump assemblies gener-
ally are raised and temporarily rigged
before the manifold is raised. Because
the lift is only about 8 tons (7.3
tonnes) for each pump assembly, the
crane in the furnace can be used for
this operation.
It often is advantageous to fit a
number of pump components together
on the ground first. The discharge valves
can be welded onto the pump casings
before erection. The pump casing is a
volute-shaped chamber into which the
pump impeller fits. The pumps force
the water from the manifold through
discharge links into the lower waterwall
headers. Check valves at the pump
Figure 11-9 | Fuel pipe connection being made to a furnace windbox
Ź}Ź~
ČŅŅčńōĞŷ
ʼnňŅʼn
atmosphere from the weld puddle. They usually employ the GTAW
Argon usually is the shielding gas, process, although the gas metal arc
although other gases, or combinations welding process can be used in the field
of gases, may be used. Historically, this as well. The main advantage of orbital
relatively slow process is used to put welding is the higher quality of the
in the root (first) pass in pressure part deposited weld. This improved quality
butt welds. results from the repeatability of the
The most extensively used process is machine and the decreased fatigue of
shielded metal arc welding (SMAW), the operators compared with manual
more commonly referred to as the “stick welders.
electrode” process. In this process, the Field welds are radiographically
flux-covered electrode is consumed in examined as required by code, contract,
the weld. The flux coating produces and self-imposed quality-control
the shielding gas, which is supplied ex- requirements. For this analysis, ra-
Almost all metal parts on a ternally in the GTAW process. It also diographic film is placed on one side
boiler are welded. Machine promotes electrical conductivity across of a weld, and a radioactive source is
the arc column, adds slag-forming temporarily placed on the other side.
welding improves the quality
materials that help prevent rapid oxi- (Sometimes, the source is placed inside
of the weld resulting from the dation of the weld metal, and in many a pipe weld for a panoramic shot, and
repeatability of the machine. cases, adds alloying materials to the other times, the source is placed on
weld. This process follows the GTAW the outside of a tube or pipe weld for
root pass and completes the pressure a double-wall shot.) When the film
part butt weld. It also is used to weld is developed, discontinuities may be
ductwork, waterwall seams, and other revealed, requiring interpretation by a
plate and pipe components. qualified radiographer. Any disconti-
The third process frequently used nuity, which is interpreted as a defect,
during the field construction of boiler must be ground out and rewelded.
components is gas metal arc welding. Depending on the material speci-
In this process, a machine feeds a base fication of the component and on
of flux-cored (flux-filled tubular) wire the diameter and thickness of the
from a spool into the weld. The wire is weld, post-weld heat treating may be
the consumed electrode, as in SMAW. required to reduce the residual stresses
The shielding gas may be supplied ex- induced by the welding. This stress-
ternally (as with GTAW), or it may be relieving process involves the placement
supplied by the flux (as with SMAW) of electric elements (or coils) or gas
if flux-cored wire is used. This process burners around the weld area of the
also is used for ductwork, waterwall component. A band around the weld
seams, and other plate welding. joint is brought up to a predetermined
With increasing frequency, machine temperature, typically 1,100 to 1,300°F
welding is being performed in the field. (590 to 700°C), at a controlled rate
In field machine welding, the equip- and then held there for a time period
ment performs the welding operation that is a function of the weld thickness.
under the constant observation and Thermocouples attached at the outside
control of a welding operator. Ma- of the heat band, or temperature-indi-
chines that make pressure part welds cating crayons, are used to monitor the
are called “orbital welding machines.” temperature. The component is then
Construction and Maintenance | 11-19
then repeated until no more leaks are a few. The turbine generator is a large
apparent. section of the mechanical work and
A steam generating unit is “built-in requires almost the same time span
accordance with the ASME Boiler for erection as the boiler does. Figure
Code” only when it has successfully 11-10 shows a turbine under construc-
passed such a hydrostatic test of the tion but with most of the components
entire unit, either in the shop or in the already in place.
field. The erector engages the services The overhead gantry crane, one of
of a steam boiler inspection agent who the first pieces of equipment erected,
is responsible for verifying the inspec- is used to raise and set most of the
tion and assuring the integrity of the turbine except for the stator. Because
boiler to the owner. The owner usually the stator is so heavy (200–400 tons
has its own insurance representative on [180–360 tonnes]), a special lifting
Figure 11-10 | Turbine pedestal with most
components in place
the job at the same time to look over rig is usually brought to the site just
the records and to ensure that proper for this one lift. This rig uses a system
inspections and tests are performed. of grip-type jacks with steel cables to
Once the hydrostatic test is lift the stator from the railroad car
complete, the insulation and lagging and up to the turbine-pedestal floor.
installation can proceed. Insulation Once the stator is up to elevation,
blankets, or blocks, are installed on special rolling beams are put under
all piping, waterwalls, drums, and the stator. The stator can be lowered a
ductwork. Metal lagging is installed few inches (centimeters) onto special,
over the insulation for protection of heavy-duty rollers and then rolled
Clean Combustion Technologies equipment and personnel. A coating of from the lift frame onto the concrete
Figure: 11.10/JG refractory is troweled on the top side of turbine floor. The stator is transferred
PMS5415 K the roof tubes, and a steel “skin” casing to the generator hole and is lowered
is welded over the top. If the unit is again using the upper section of the
an outdoor unit, the lagging serves to same lift rig. Figure 11-11 shows a
protect the boiler from the weather stator entering through an access
as well as to provide a satisfactory ap- opening from the ground.
pearance. If the unit is an indoor unit,
the boiler is completely enclosed by an ERECTION–COMMISSIONING INTERFACE.
outer building. This is desirable in cold With most of the major work com-
climates, making initial erection and pleted on the boiler, boilout and acid
subsequent maintenance easier during cleaning of the inside of the tubes
the winter months. can be done. Chemicals are circulated
through the boiler for prescribed
ERECTION OF MECHANICAL EQUIPMENT. lengths of time and then drained
The installation of mechanical equip- from the boiler. These procedures
ment proceeds while the boiler is remove internal scale from water-
being erected. Mechanical equipment bearing systems and ensure free water
includes the turbine generator, miscel- passage in all tubes, headers, and
laneous pumps and valves, the water drums. (See Chapter 10 for a descrip-
treatment systems, the boiler feed tion of boilout and acid cleaning of
pumps, and the condenser to name but high-pressure steam generators.)
Construction and Maintenance | 11-21
CFB BOILERS
The erection of a CFB boiler follows
much of the same logic as the erection
of a pulverized coal-fired unit. This
section will only describe the major dif-
ferences in the erection sequence and
in erection of the additional compo-
nents. Because the CFB has suspended
solids in the combustor, which both
enhance the capture and conversion
of sulfur oxides and enhance the heat Figure 11-11 | Stator lift rig
transfer, additional equipment (primar-
ily the cyclones) is included. The other
effect of the circulating solids is the seams between waterwall panels and
potential erosion that can take place in increased size and strength of the
the furnace, which in a CFB is referred buckstay system.
to as the “combustor.” This affects both
the placement of radiant surface in the Erection of Cyclones
combustor and the preparation of the The addition of the cyclones isClean
the Combustion Technologies
Figure: 11.11scanned/AR
walls during construction. biggest increase to the erectionPMS5415
scope. K
Cyclones can either be manufactured
Erection of Combustors plate or manufactured tubing and,
The erection of the combustor follows thus, water or steam cooled. This sec-
the erection of other units, but with tion will address the erection of plate
the following exceptions: First, the cyclones, which are more common.
potential for erosion requires that all Plate cyclones are ground assembled
surfaces inside the combustor, includ- in the same manner as the coal silos
ing all field welds and seams between and are erected in components as soon
components, be ground smooth. as the steel is erected to their support
Second, the internal operating pres- level. The cone section includes the
sure in a CFB boiler is up to fivefold support ring and weighs approximately
greater than that in a pulverized 30 tons (27 tonnes).This section is set
coal-fired unit. The higher pressure into the steel first, leveled, and bolted
requires additional welding on the to the support steel. The barrel section
11-22 | Clean Combustion Technologies
is then set on the cone section and After the cyclone, the inlet duct and
welded to it. Next, the inlet and top outlet are erected and welded. A spe-
section are installed and welded to cialty subcontractor is used to install
the barrel. Finally, the outlet section the refractory and brick linings. Refrac-
is supported independently from steel tory is installed wet, using air-driven
at a higher elevation and is erected guns, starting at the bottom and work-
after completion of the support steel ing toward the outlet. The refractory
to this point. will set up as gunned, but it needs to
be “dried” in order to achieve the de-
Solids Ducts and Fluid Bed Heat sired hardness. Dryout is accomplished
Exchangers when erection of the boiler is complete
Depending on the size of the boiler and at a point where heat can be intro-
and on the fuel burned, a fluid bed duced in the cyclones and surrounding
heat exchanger may be included. This components. This is done by installing
is a refractory-lined box, set at grade, temporary burners capable of bringing
that contains evaporator, superheater, the refractory temperature to 700°F
and/or reheater surface. A portion of (370°C) and then holding this tem-
the solids from the cyclone are routed perature for up to 24 hours to achieve
through this device to transfer the heat the required dryout and hardness. The
from the ash to the appropriate surface. total time to perform dryout, including
Because the fluid bed heat exchanger heat-up and cooldown, is 7 days.
is at grade, is relatively heavy, and in-
cludes substantial heat-transfer surface, Backpass
it needs to be erected in place at the The CFB backpass can be similar to
beginning of support steel erection and the pulverized coal-fired backpass. In
before the cyclones and solids return some cases, a two-pass backpass is
system is in place. used, with reheat surface in the front
Solids ducts return the solids from pass and superheat surface in the rear
the cyclones to the lower combus- pass. If the shipping dimensions are
tor (for recirculating) or to the fluid acceptable, the backpass can be shop
bed heat exchanger. These ducts are fabricated into large, vertical modules.
manufactured from plate and are This moves much of the difficult field
installed after the combustor is in welding to the shop, and it shortens
place. Because the ducts are fixed and the amount of time needed to erect
the combustor expands downward this section. This type of backpass
approximately 8 inches (20 cm), a module will weigh up to 50 tons (45
slip-type expansion joint and/or a tonnes), be 60 feet (18 m) tall, 15 feet
toggle section is used. (4.6 m) deep, and 12 feet (3.7 m)
wide. A rigging plan for rotating from
Refractory Linings horizontal to vertical (upending) and
The cyclone and solids ducts have then raising to its final position is
a refractory and brick lining that is included in the erection plan. This can
designed to maintain the outer steel be accomplished with a crane, block-
surface at acceptable temperatures. and-tackle rigging, or strand jacks,
Construction and Maintenance | 11-23
depending on the site conditions and modeling using virtual space, can help
equipment available. the designer plan for accessibility (Fig.
11-12). Thus, time can be saved not
Maintenance and Repair of only during the installation of new
Steam Generating Equipment equipment but also with component
A good maintenance program is one replacement during future maintenance.
of the keys to reliability of any steam Designers responsible for piping and
generator. To be successful, such a electrical work must be aware of areas
program requires managerial abil- requiring accessibility for maintenance.
ity, expertise, imagination, planning, Often, piping and wiring are field-run
Figure 11-12 | A scale model used to lay out
training, and the commitment of top from drawings having notes advis- a pulverizer bay for optimum
management. In addition, maintenance ing construction personnel of special maintenance access
must be closely integrated with both maintenance requirements.
the operation and power station engi-
neering functions. EQUIPMENT MANUALS: AN ASSET
TO EFFECTIVE MAINTENANCE
MAINTENANCE FORESIGHT When equipment is purchased, it is
Because plant layout can either facili- important to also obtain operation and
tate or impede a maintenance program, maintenance manuals that are com-
maintenance personnel should be a plete, detailed, and specific to the piece
part of the initial design team to ensure of equipment supplied. Material that
the plant layout facilitates maintenance. does not contribute to an understand-
Clean Combustion Technologies
Ideally, the person who will ultimately ing of the design, care, operation, and
Figure: 11.12/JG
be responsible for plant maintenance maintenance of the equipment should PMS5415 K
should be a member of the design be excluded. Manuals should include:
team, because only during the design
• Step-by-step disassembly
stage can truly adequate space be al-
procedures
located for maintenance.
• Step-by-step reassembly
As equipment is located, maintenance
procedures
personnel must have space to easily
• Preventive maintenance and lubri-
remove the largest replaceable part. The
cation instructions
steam generator should have enough
• Lists of all special tools required,
access doors so that all areas may be en-
and instructions for their use
tered easily for inspection. Access doors
• Tabulated dimensional data on set-
and aisles also must be large enough to
tings, clearances, and adjustments
accommodate components that must be
moved to a maintenance area. Sufficient The applicable plant department
headroom is a must. Overhead trolleys should contact the original equipment
and cranes for maintenance should manufacturers on a periodic basis to
be designed into the plant rather than verify if these procedures have been
retrofitted. All high-maintenance items improved on or if service or technical
should have adequate lay down and work bulletins have been issued on equip-
areas adjacent to the equipment. Scale ment still in operation. If so, copies of
models, as well as three-dimensional this documentation should be obtained
11-24 | Clean Combustion Technologies
for inclusion in their maintenance a given type does not wear at the same
records and plant documentation. rate, because different local conditions
can alter the wearing rate drastically.
DETERMINING THE NEED The use of manufacturer-supplied
FOR SPARE PARTS spare part sub-assemblies can provide
Often, a complement of spare parts significant time savings for repair
is ordered with the initial equipment work by eliminating time-consuming
purchase. If this is not done, the spare assembly of the individual parts by the
parts should be purchased before the plant personnel.
equipment is placed into service. Many When parts orders arrive, respon-
times, the supplier of the new equip- sible personnel should inspect the
ment will provide a Recommended material carefully. A defective part
Spare Parts list, with part numbers. may sit in a storeroom for years before
Items critical to plant production the defect is discovered—often just
should receive special consideration, at the time of need. Also, to ensure
and for those items, it may be prudent correctness and completeness, all
to maintain spare sub-assemblies. At incoming parts should be checked
the time of purchase, the applicable against purchase orders before those
plant department should ask for parts are stored. The inspections must
complete instructions from the manu- be performed carefully to prevent
facturer for error-proof ordering of damaging the preservative coating ap-
replacement parts. Catalogs specific to plied to many parts, such as bearings.
the equipment and with exploded-view This is critical to prevent the parts
drawings are helpful in selecting the from deteriorating during storage.
appropriate spare parts. Installation instructions and assembly
There should be a system for track- drawings supplied with the new parts
ing spare parts as well. Most plants and conversions must be forwarded to
maintain a computer-based, replace- the applicable plant engineering and
ment parts inventory program. If maintenance departments to enable the
establishing an in-plant program is not necessary tooling and work schedule to
feasible, most equipment manufactur- be prepared and the new part numbers
ers have created inventory management to be added to the plant inventory sys-
programs that can effectively manage tem. Incorrect parts must be identified
spare parts for planned and forced out- and action taken to delete them from
ages. Based on equipment history over plant inventory and prevent accidental
the entire fleet, some manufacturers reordering in the future. Maintenance
can create specialized sub-assemblies records must be updated to identify
so that the correct parts will be avail- these changes.
able at the time of the outage, along
with instructions for easy installation, IN-SERVICE MAINTENANCE
thus reducing the inventory necessary To ensure that a unit is available on
at the plant site. Some manufacturers demand requires constant effort by
have Web-based systems with original the maintenance staff. They must
drawings in order to simplify part implement a day-to-day, preventive
identification and ordering. Even so, maintenance program while the unit is
one must realize that all equipment of in operation, establish a well-planned
Construction and Maintenance | 11-25
service when the unit’s seasonal load sion ring and vane wheel segments,
demand increases. must be available in the event these
An effective pulverizer maintenance are needed to replace parts found
program is adjusted on the basis of the to be worn or damaged once the
actual plant operating experience and overhaul has started and the final
the service life of the pulverizer parts. scope of work has been established.
An essential factor in this is to main- If the pulverizer is equipped with an
tain an accurate log of each pulverizer’s exhauster, a spare exhauster fan wheel
running time, including stoppages for assembly and spare casing liners must
removal of foreign material, mainte- be available.
nance work performed, and results of Once the pulverizer has cooled, it is
inspections, in order to predict when locked out of service and isolated by
each overhaul is required. The life of closing the shutoff valves on the air in-
pulverizer parts depends on many fac- let, the burner isolation valves, and the
tors that can vary considerably from one pulverizer/exhauster discharge valve.
plant to another. The main factors are The access doors are removed, and to
the abrasive characteristics of the coal, aid cooling of the pulverizer parts, an
operating hours of the individual pulver- air mover may be used to ventilate the
izers, coal fineness required for correct interior. Once the journal assemblies
combustion, and pulverizer loading as a have been removed, there is free access
percentage of maximum capacity. to the inside of the pulverizer, and the
The grinding rolls and the bull ring worn bull ring or worn liners can be
(grinding ring) are the major replace- removed rapidly. When all obviously
able parts. (Further discussion of worn parts have been stripped from
grinding roll and bull ring wear is pro- the pulverizer, the remaining parts and
vided later in this chapter.) With good wear liners are carefully inspected for
planning and preparation, it is possible repair or replacement and any needed
to overhaul a pulverizer, replacing the replacements are added to the final
grinding rolls and the bull ring as well maintenance list.
as repairing or replacing other wear When the repair work inside the
parts, in a few days. Exhausters (if pulverizer has been completed, the
used) can be overhauled in the same overhauled journal assemblies are
time period using the same process. installed. All access doors are closed.
To aid a rapid overhaul, the plant The pulverizer is cleared of all “hold”
should have ready for installation a tags and then started, without coal, to
minimum of three complete journal facilitate adjustment of the grinding-
assemblies with the grinding rolls roll-to-ring clearance. The pulverizer
installed (Fig. 11-13). Often, such as- then is released for restricted service
semblies can be stored ready for use in and tested by operating it with coal,
the pulverizer overhaul area in order to during which the coal fineness test is
reduce handling time. Frequently used performed to confirm that the pulver-
bolts, nuts, and washers should be kept izer is operating correctly. Any required
in bins in the overhaul area. adjustments are made at this time, after
The remaining wear liners and which the pulverizer is released for
other parts, such as the bowl exten- normal operation.
Figure 11-13 | Three complete journal
assemblies stored on transport
dollies
Construction and Maintenance | 11-27
inspection is performed to detect any with readings taken during the outage,
obvious items requiring maintenance when the unit is cold, and after its re-
or more detailed investigation during turn to service. These readings are used
the outage. All items are noted, and a to confirm free and proper expansion
list is compiled. Operational checks of the unit.
of certain equipment and systems also
are conducted to determine whether SCAFFOLDING. For outages in which
each is functioning properly or requires extensive inspection and repairs to the Figure 11-15 | Interior view of unit with
adjustment, repair, or replacement. furnace are planned, fixed scaffolding, modular scaffolding system.
This includes: temporarily installed both outside and Note the easy entrance and
access to all areas
inside the furnace, can save time and
• Fuel-burning equipment, such as
money as well as improve safety and
burner tilt mechanisms, windbox
efficiency. When extensive repairs are
dampers, coal gates, and oil or gas
planned for furnace and penthouse
valves
areas, scaffolding rental fees are more
• Burner management and ignition
than offset by savings in working
systems Clean Combustion Technologies
hours. Adapted for quick installation,
• Combustion and feedwater control Figure: 11.15(scanned)/AR
such scaffolding is modular and made PMS5415 K
systems
of tubular carbon steel with steel loops
• Soot blowing equipment
at the end of each section (Fig. 11-15).
In addition, the settings on safety Fixed-length sections are safety-locked
valves are checked. Control and isola- to one another by inserting steel
tion valve performance is verified, and wedges into the loops.
any packing leaks are recorded. The Basically, the scaffolding for inte-
condition of actuators is checked to rior access provides a solid platform
confirm that oil seals are secure and near the top of the furnace. Typically,
lubricating oil is not contaminated 10 or more scaffold towers, each
with water or fly ash. Complete sets with three working platforms, are
of unit performance data, near to or positioned at the front of the furnace,
at full and control point loads, should adjacent to each side of the super-
be printed and analyzed. This review heater division panels. Scaffolding
method can indicate items that may also is provided across the unit at the
be out of calibration or problems with front and rear of the superheater or
equipment or subsystems, and it can reheater pendant assemblies.
provide an important historical refer- Furnace waterwalls are scaffolded
ence during future unit operation. Any from the lower slopes to a point ap-
problems should be noted and added proximately 6 feet (1.9 m) below the
to the outage activity list. All perma- lowest-element tubing of the super-
nent thermocouples, such as those on heater division panels. Walkway levels
superheater and reheater tubing as well are provided at approximately 7 foot
as waterwall panel outlets, are checked (2.1 m) intervals, beginning at the
to isolate any defects requiring repair bottom elevation of wall blowers and
during the outage. Readings from continuing to the top elevation of wall
expansion trams in important areas blowers. An extension of the water-
should be recorded for comparison wall platform at convenient working
11-30 | Clean Combustion Technologies
elevations provides interior access to plant locations and access points, and
the windbox. Exterior scaffolding also a plan should be created to properly
can be used at the penthouse front protect these areas. Typically, a permit
wall near the roof header and super- for a confined space will require regular
heater links, at the soot blowers for ambient air testing and access points,
lance removal, and at the economizer monitored by trained hole watch
outlet ductwork. personnel. The personnel and equip-
As typical views of scaffolding ment needed to support this protection
Figure 11-16 | View of a vertical modular
scaffolding system
show (see Figs. 11-15 and 11-16), should be included as part of any
maintenance and repair personnel outage plan.
have easy access to the unit during an
outage. This provides three significant PRESSURE PARTS: EXAMINATION OF
improvements over previous methods. EXTERNAL SURFACES. Within the furnace,
First, repair personnel can enter the the external surfaces of tubing are
interior more safely than by using examined for erosion, metal wastage,
cable-hung equipment. Second, and swelling, or developing problems in
equally important, continuous access to critical areas, such as the firing zone
all areas is possible, and various crafts and near the soot blowers. Suspicious
do not have to compete for access. This areas are investigated further by clean-
significantly reduces the time required ing with sandblasting, then checking
to transport personnel and materials the wall thicknesses by micrometer or
FigureClean
11-17 | Tubing isTechnologies
Combustion measured
ultrasonically for wall
to the areas needing repair. Finally, measuring those thicknesses ultra-
Figure: 11.16(scanned)/AR
PMS5415 thickness
K scaffolding allows a more complete and sonically (Fig. 11-17). This type of
Ź}Ź~
thorough inspection. inspection also covers:
ČŅŅčŅŋĞŷčč
The Outage Stage • Critical superheater and reheater
ʼnňŅʼn
sections
During the outage, every portion of
• Economizer tubing
the unit requires inspection. Sched-
• Steam-cooled wall surfaces and
uled repairs are then accomplished by
waterwall hanger tubing in the gas
a critical path analysis to ensure that
passes
all repairs are completed. Key areas
to be inspected are the pressure parts,
the fuel firing and transport equip- PRESSURE PARTS: EXAMINATION OF
ment, the boiler casing and structural INTERNAL SURFACES. The steam drum
supports, and such subsystems as ash- must be entered and thoroughly
handling and soot blowing equipment. inspected for deposits, loose parts,
Many locations in a power plant erosion, and areas that may allow
must be addressed as confined spaces, internal bypassing of water or steam.
either permit or non-permit. National Conditions found in the boiler drums
safety laws and local requirements and may dictate inspection of waterwall
codes define these spaces and how headers and other internal surfaces
personnel should be protected when for evidence of corrosion or deposits.
entering these areas. The most current Any lower waterwall headers large
editions of these regulating docu- enough to have accessways should be
ments should be reviewed closely to inspected. Waterwall distribution ori-
determine how they apply to specific fices and screens are examined closely
Construction and Maintenance | 11-31
for deposits, and orifices are checked correct position in reference to the
for proper size with go/no-go gages. configuration of the windbox nozzle.
Samples of furnace wall tubing may be Ignitor equipment should be checked
removed from various locations a short to determine the condition of the horn
distance above the uppermost fuel and all internal components for dete-
nozzles; such samples are examined in- rioration and alignment. Any defective
ternally to determine surface condition or questionable components should
and to detect any deposits. be replaced. Flame scanner alignment
In the confined areas, a video and orientation also must be checked
recording borescope can be inserted to verify not only the remote viewing Figure 11-18 | Fuel and air nozzles are
inspected from inside the
into high-temperature headers, either capabilities but also the position and
furnace
through a terminal tube that is cut condition of the guide pipe.
through and then sprung back for Any tilting fuel and air nozzles
access or through a removed hand- should be operated through their
hole plate or radiographic access plug. range of motion while being observed
This permits inspection of the internal from inside the furnace. Binding,
surfaces of the ligament and borehole non-parallelism, or other malfunctions
surfaces for defective conditions, such should be corrected. The tilt mecha-
as crazed cracking, borehole cracking, nism also is inspected from inside the
debris, girth weld cracking, and heavy windbox to determine the condition
oxide scale cracking (likely a thermal of the linkage. Windbox dampers are
fatigue mechanism). This type of inspected from within to determine
Clean Combustion Technologies
examination also is valuable in water- whether they operate through their Figure: 11.18/JG
touched headers. For example, the normal stroke and if damper-blade PMS5415 K
economizer inlet header can be sub- movements correspond correctly to ex-
jected to the effects of oxygen carried ternal linkage movement. No binding
in with non-deaerated boiler feedwater. should be evident.
This can manifest into deep pits, which If there are retractable oil guns, the
can link together to form line crevices retracting cylinders must be operated
or corrosion fatigue from the stop and to verify proper movement. A general
start of feedwater flow during start- inspection should be made of the
ups and other transient conditions. ignitor control cabinet and associated
In general, all of these areas require piping. Any damaged or defective com-
periodic inspection to prevent failures ponents should be noted and replaced.
and costly unplanned outages. A general inspection of the pulver-
izers and exhausters (if used) must be
FIRING EQUIPMENT. During the internal performed, and any repairs, including
inspection of the furnace, the condition replacement of worn parts, should be
of the fuel and air nozzles is deter- preformed as required. The inspection
mined, and any serious deterioration should be made even if plant practice
(Fig. 11-18) is noted for replacement is to overhaul the pulverizers online be-
or repair. This includes all low-NOx cause of the convenience of having them
equipment and compartments. all out of service simultaneously along
During this stage of the inspec- with the entire coal transport system.
tion, oil guns should be inserted to Each pulverizer, exhauster (if used), and
their normal firing position to verify coal feeder must be cleaned of all debris
11-32 | Clean Combustion Technologies
and spilt coal found inside it. This Fire Protection Association (NFPA)
opportunity can be taken to inspect Code requirements. The couplings
the gearing, change the gearbox oil, used to assemble the coal transport
and inspect the condition of all major system components must be checked at
internal components sections. These assembly to confirm that the instal-
include all shutoff dampers, control lation is correct, with the required
dampers, and inlet air ductwork. An engagement to each component.
NDE examination of the pulverizer Riffle distributors are inspected for
gearbox vertical shaft can be scheduled wear and debris. The riffle elements are
at this time if the gearbox condition is replaced or rotated as necessary. Coal
suspect. Instrumentation lines should pipe orifices are checked for wear and
be checked and confirmed to be open damage, especially if past clean airflow
and clean. Fire-extinguishing water tests have indicated a flow imbalance
spray nozzles and steam inerting lines greater than 5% of the average velocity
should be cleaned and inspected. in the subject coal transport line. If
Thickness testing, using NDE, should past clean airflow tests have indi-
be performed on the steam inerting cated such a flow imbalance, the coal
lines and components to identify any transport line in question should be
items requiring repair or replacement. opened and visually inspected for coal
Exhausters must be inspected for wear buildup, debris, and worn coal pipe
to the fan wheel assembly, including orifices. Components out of tolerance
the fan blades and fan spider. The ex- are replaced as necessary. A new clean
hauster casing liners must be inspected airflow test must be performed at the
for wear, cracking, and impact damage conclusion of the outage whenever
that could result in pieces breaking off components of the coal transport
and falling into the rotating fan wheel. system or burners are replaced in order
The feeders must be inspected as well, to confirm that the system is free of de-
as recommended by the manufacturer, bris and that the flow balance is within
and any necessary adjustments and the 5% limit.
maintenance performed.
Coal pipe elbows are examined CASING AND ROOF ENCLOSURE. The roof
internally for wear. In particular, the enclosure also is inspected, and the
final elbow at the coal nozzle must be condition of pressure parts, hangers,
inspected and the wear pattern noted. and other pressure part supports is re-
Coal pipe elbows fitted with deflec- corded. Any damage or leakage of the
tor blocks (“kicker blocks”) should furnace roof casing and seals should be
be inspected for wear and damage in corrected. The condition of insulation
the vicinity of the deflector/kicker and pipe seals should be recorded. The
blocks and attachment bolts, and for roof support steel and hanger rods
wear on the elbow internal surfaces. within the air space between the roof
Worn deflector/kicker blocks must be deck plates and insulation are checked
replaced. If the coal transport pip- for corrosion. Any leaks or damage to
ing contains components made from the external casing of the entire unit
brittle cast material, such as NiHard or noted during the pre-outage stage are
cast iron, each one should be marked repaired as necessary. The skin casing
and scheduled for replacement at the and seals within the dead air space be-
next possible opportunity by an identi- hind the deflection arch are inspected,
cal component that meets the National as are the casing, seals, and support
Construction and Maintenance | 11-33
steel within the dead air space beneath associated dampers and/or inlet vanes
the sloping furnace bottom. All are must be inspected as well. In addition,
inspected to determine general condi- the induced draft fans are examined for
tions and to note any maintenance erosion, corrosion, and/or deposits.
that should be performed during the Finally, all observation ports and ac-
current outage or in the future. cess doors are checked for tightness.
comprehensive report that covers all system or area before or during future
pre-outage, outage, and post-outage outages. Emphasis should be placed on
inspection and repair. items that were not completed and on
identifying the root cause for the fail-
The Pre-Outage Report. The pre- ure. These uncompleted items should
outage inspection report contains a be recommended as a high priority on
review of the unit history, the results the next outage work list.
of the visual inspections, the walk- The post-outage documents should
down of the unit while in operation, describe conditions immediately
and the pre-outage checkout of all after restart and any changes in the
thermocouples, ignitors, fuel-burn- operating parameters resulting from
ing equipment, safety valves, and soot the maintenance performed during
blowers. Any items requiring further the outage. Operating data, including
inspection or repair during the outage tube temperature measurements and
are noted separately. expansion data, can help illustrate this
item in the report. The final—and
The Outage Report. The outage most important—documentation lists
report contains the history of all work the recommendations of the inspec-
performed during the outage and forms tion team as well as an engineering
the basis for the maintenance activity department review of those findings.
and spare parts required for subse- Such recommendations should include
quent outages. Outage documentation which spare parts to order in prepara-
includes detailed inspection and work tion for the next outage.
reports that describe the conditions of After all the planning, inspection, and
both pressure and non-pressure parts of documentation have been completed,
the unit as well as the other plant equip- a critique provides all participants
ment. The engineer in charge of the involved in the outage an opportunity to
outage should compile the outage re- review their findings and determine im-
port. An effective report should include provements for future outages. A frank,
sketches, marked prints, or photo- open exchange sets the stage for the next
graphs, which can be useful when ideas outage and also the plan for continuing
or concepts rather than physical objects maintenance activities throughout the
are being described for the record. coming operating period.
Review and analysis of the damage
The Post-Outage Report. This post- found in many previous outages (as
outage report should include the status well as during forced outages) can help
of all the maintenance activities that to determine typical problem areas.
were scheduled and all additional items This analysis also can aid in identifying
that were added during the outage. root causes to eliminate these prob-
Each work item should include the lems in the future. Often, however, it
final status of the system or compo- is difficult to collect and analyze all
nent (i.e., a component replaced with these data when stored in individual
one from the storage facility or required records and reports. At a minimum,
an emergency order). This method it is beneficial to keep track of these
provides a gauge of the time spent in the repairs in an electronic database or
pre-outage stage and identifies if great- spreadsheet. Specialized equipment
er time should be spent in a particular repair and failure tracking software
Construction and Maintenance | 11-35
also is available, allowing easier track- should list alternatives, such as the use
ing and more detailed cause coding for of manpower or parts from another
future analysis. Future maintenance division or utility, the use of alternate
plans can be recorded and tracked in production methods, or the purchase
such software. Most of these software of power. Because downtime is critical,
programs also link the data entries to the repair plan details and schedule
component arrangement drawings in must be worked out immediately. Plant
order to provide a visual presentation personnel and the forced outage plan
of the problem locations. This allows should have lists of other critical work
better recognition of common problem that can be done during an outage. No
areas to determine causes. Some of items should be added to this list that
these programs are network or Internet will adversely affect the time during
based, allowing many users to view which the unit is out of service.
the results at any one time. This also Pressure part failure is a major
allows a team to work together for out- cause of forced outages. Because the
age planning and problem resolution. steam generator must be cooled before
Figure 11-20 illustrates examples of it can be entered, the cooling time
software entry and arrangement views. can be used for planning. The repair
of the failure will be the critical path
FORCED OUTAGES in returning the unit to service, so
A forced outage requires as much other maintenance may be performed
or more planning than a scheduled only if it does not interfere with the
outage, but by its nature, it allows the repair. To save time during furnace
least time for planning. Consequently, cooldown, crews can move parts,
a plant maintenance group must be tools, and scaffolds close to the steam
prepared with an emergency plan, generator access doors. Usually, it is
one that has been formulated before possible to determine the area of the
a failure occurs. This plan should be failure and its general magnitude from
in writing and periodically updated. It the outside, and if a leak is small, sky
mf-3270 DataObject
Connect Disconnect Field Window Generate
+
PT0Q OPMIS 09/28/08 14:47:22 PAGE: 1
OPEN QUOTE
N P R R
CLIENT PO NUMBER 25 25 SEPT AFTER OPEN VISIT CHANGES 09-28-08 01003764
* NOT RELEASED GAS PATH
CLIENT PO NO QOUTE 25 SEPT DIRECT 09-28-08 01003763
* NOT RELEASED +
TN 092 23/006
PF1 PF2 PF3 PF4 PF5 PF6 PF7 PF8 PF9 PF10 20–
PF11 PF12 PF13 PF14 PF15 PF16 PF17 PF18 PF19 PF20 25–
PF21 PF22 PF23 PF24 PA1 PA2 PA3 Clear Reset Enter
Er EOF Er Inp Dup Fld Mark Sys Req Attn Insert Home
Ready 2/20/2008 12:07:38
Figure 11-20 | Examples of software entry and arrangement views of maintenance activities on pressure part–related items.
code of construction for rules govern- For instances in which the welds fall
Defective
ing how to make a repair or alteration. in areas where access to one side of the Good Good
When it is deemed impractical to weld groove is difficult or impossible,
3 4
follow the original construction code it may be necessary to perform what is
1 2
requirements, however, the NBIC called a “window weld” (Fig. 11–22). Flame Cuts to
has provided some special rules and This weld is completed by cutting a Remove Defective Tube
guidelines for qualifying the repair hole (window) in the front of the tube
procedures and welder performance to allow access for welding the back
A
training/testing when making the side of the tube from the front. The
repair or alteration. Before the actual first pass, or root, of the weld must be
A
work begins, the Authorized Inspector welded using the GTAW process. The
and the owner must approve the repair remainder of the weld can be completed
plan and procedures. using either the GTAW or the SMAW
Section A-A
process. Fit-up of the tube ends and
GUIDELINES FOR WELDING REPAIR OF the replacement plug for the window
LOW–CARBON STEEL TUBES. When replac- is critical to obtaining a satisfactory
ing a length of damaged tubing, the weld. Much care must be taken when Window Width
Must Allow for Welding
cut and weld prep should be located at cutting the holes and the plug. The side Access at Sides
least 2 inches (5 cm) beyond the end of locations in which the two weld grooves
the damaged area. If the new cut line form a 90° angle require close attention. Figure 11-22 | Window welds may be used
places the new weld in the vicinity of Fit-up of the weld joints is crucial to repair damaged tubes when
Clean Combustion Technologies
access is
Figure: 11.22/ difficult
(DE)
an existing weld, the cut line should be to obtaining an acceptable final weld.
PMS5415 K
moved so that the old weld falls within Although it is difficult to make accurate
Replace
the removed section of tube. The mini- cuts on furnace tubes, it is important to Good Good
mum replacement tube length should get the existing tube ends squared and
be no less than 6 inches (15 cm). correctly chamfered, cut the replacement
The use of backing rings for a weld tube to the correct length, and allow for Alignment bar,
is permitted by code, but in some weld shrinkage. A tube-end scarfing tool one each side.
Tack weld to
instances, they either cannot be used should be used when possible. Remem- adjacent tubes.
or must be removed after the weld- ber that the weld and parent metal are Do not weld to
replacement tube.
ing is completed. For example, it may melted during the welding process and Similar pair of
be an engineering requirement that a that molten metal shrinks as it solidifies. bars may be used
at lower end of
backing ring not be left in place, because A butt weld in a tube will shorten the replacement.
it would create an orifice effect, an unac- total tube length by approximately 1⁄16 Wedges between
tube and fin will
ceptable stress riser, or a turbulent flow inch (1.6 mm). The required preheat maintain
in the component. If the component should be applied to the weld joint alignment.
is in a heat-absorbing area and is in before welding. A clamp or guide lug
contact with water or a mixture of water should be used to hold one end of the Flame cut just outside weld bead attaching
fin to defective tube. Reweld fins after
and steam, backing rings cannot be replacement tube in alignment while the completing tube welds.
used. Without a backing ring, a welding first weld is made (Fig. 11-23).
process capable of producing a defect- Both ends of the replacement tube
free root weld (usually GTAW) must should not be tack welded, particularly Figure 11-23 | Welding detail for replacing
a tube length in a fin-welded
be used. After the root weld has been if the existing tubes are rigidly support- Clean Combustion Technologies
furnace wall
completed, the remainder of the weld ed. The preheat requirements for the Figure: 11.23/ (DE)
PMS5415 K
groove can be made with any acceptable butt weld also apply to the tack welds.
welding process (usually SMAW). As a general rule, the weld at the lower
11-38 | Clean Combustion Technologies
end of the replacement tube should be the operating temperature and pres-
completed first. The upper end of the sure. The most common occurrence
replacement tube should not be welded of this situation is when dissimilar
until both the replacement tube and metal weld (e.g., ferritic to austenitic)
Replacement
Tube the existing tube have cooled to ambi- failures begin to appear. The ID of
3
ent temperature. The required preheat the ferritic tube should never be bored
1 can then be applied and the welding out to match the ID of the austenitic
performed in the same manner. tube. This will result in the ferritic tube
having less than the minimum wall
75°–90°
GENERAL GUIDELINES FOR ALLOY TUBE thickness required for the design tem-
0��– 1/4�� REPAIRS. If a tube is being replaced in perature and pressure at the butt weld
an area in which the alloy being used location. There are two ways to handle
changes, it is always preferable to make this situation: Have a short spool piece
the weld cut lines in the same alloy ma- machined out of the austenitic mate-
terial. The creation of dissimilar metal rial, with the ends machined to the
Existing Tube
welds (e.g., ferritic to austenitic) should matching weld joint geometry, or have
be avoided. In addition, the weld cut a short spool piece fabricated with two
line should be made in an area where short pieces, one ferritic and one aus-
the wall thicknesses and diameters are tenitic, that match the existing tubes,
Figure 11-24 | Fit-up detail to install a the same. The manufacturer’s mate- with the dissimilar weld being made
replacement tube section with rial diagram will provide approximate in the center by the fabricator. The
a wall thickness greater than
that of the existing tube
locations of all material, wall thick- latter option is the preferred method.
Clean Combustion Technologies nesses, tube diameters, and shop welds It allows the dissimilar metal weld (the
Figure: 11.24/(DE)
PMS5415 K
for specific areas in a panel. If possible, most difficult to perform in the field)
the new weld cut line should be made to be made in an environment where
at least 6 inches (15 cm) back from the welder access and weld position can
existing tube shop weld. This will allow be controlled. If the tube IDs match
a “safe end” with which to work. up but the outside diameters (ODs)
are different, additional weld material
CUTTING OUT A WELD. If it becomes must be deposited on the weld joint
necessary to cut out a weld, either to taper the larger-OD tube into the
shop or field, in which the two tubes smaller-OD tube.
being joined are of different alloys, wall
thicknesses, or outside diameters, spe- FIT-UP AND SHRINKAGE ALLOWANCE.
cial attention must be paid to this type Shrinkage resulting from welding in
of operation. The tube inside diameters alloy tubes is similar to that in carbon
(IDs) must match up to ensure that steel tubes. Allowance must be made
an acceptable weld is made. If the IDs for expansion from preheating, which
are not the same, it will be necessary to could close the root gap slightly. For
scarf out the ID of the smaller-ID tube SMAW with a backing ring, the root
in order to match the ID of the larger- gap opening must be sufficient to
ID tube (Fig. 11-24). allow full penetration and complete
This operation reduces the wall fusion with the backing ring during
thickness of the bored out tube, which the first pass. For GTAW, the root
could reduce its wall thickness to below gap needs to be sufficient to ensure
the minimum required thickness for complete fusion of the root without
Construction and Maintenance | 11-39
switches; control valves; speed control- matrix. Consequently, many abrasion- Elbow Without
lers; damper drives; and thermocouple resistant materials are not resistant to Deflector Blocks
inputs. If there are known problems in erosion (Fig. 11-27).
the control system, additional work is Figure 11-27 | Erosion and abrasion in coal
scheduled. For example, a dry run of Pulverizer Wear piping
the burner management system should In coal pulverizers, the major wear Clean Combustion Technologies
be conducted if some control logic is from a combination of abrasion Figure: 11.27/SPS
PMS5415 K
is not performing correctly, and the and erosion. The grinding parts of
calibration, updating, enhancing, and the pulverizer wear principally from
retuning of a control loop may be in abrasion; the pulverizer classifier sec-
order if poor control is observed dur- tion and the components of the coal
ing previous system operation. transport system wear principally from
Because the control system is the erosion. As discussed in Chapter 6, the
power plant component that can trip different types of Alstom pulverizers
11-44 | Clean Combustion Technologies
on plants that have switched coals and/ ceramic tile is made to the interior of
or been retrofitted for low-NOx firing the exhauster casing and exhauster
and now operate at fineness levels cross-over pipe, the installation must
exceeding the original design specifica- be performed carefully to ensure that
tion. Because worn parts reduce the each tile is fastened strongly into
pulverizer capacity and coal fineness place. Failure of even a single tile that
and can create unsafe operating condi- subsequently falls into the rotating
tions, such as coal spillage, it may be exhauster fan wheel can cause major
beneficial to replace the grinding parts damage to the fan blades and the
before their useable life has completely bearing assembly and necessitate an
expired in order to maintain the pul- emergency pulverizer outage to repair
verizer at its required load. the equipment.
Considerable laboratory research For whizzer wheel-type fan wheel
and field development of materi- assemblies, the service life of the fan
als have reduced the wear rate and blades can be increased by using ce-
increased the service life of the ramic tile-covered fan blades on which
pulverizer grinding parts and wear the edges are lined with tungsten
liners. The abrasion-resistant materi- carbide for increased impact protec-
als currently used for the bull ring are tion and abrasion resistance. This
heat-treated, nickel-chromium white design meets the fan blade weight
iron (Hi-Chrome), heat-treated white requirement to prevent overloading
iron (Crown 700), and ceramic/metal the fan spider.
Construction and Maintenance | 11-45
abrasive coals, wear to the pulverized ent location. System pressure drop
coal transport piping can become a considerations limit the number of Figure 11-29 | Coal-piping elbow with
deflector blocks
significant problem. The rate of wear deflector/kicker blocks that may be
Clean Combustion Technologies
generally is a function of: installed in each coal pipe. Sometimes, Figure: 11.29/SPS
the coal transport system flow charac- PMS5415 K
• Particle characteristics (hardness,
teristics and erosion rates from use of
sharpness, and size)
highly abrasive coals are so complex
• Particle velocity
that deflector/kicker blocks cannot be
• Angle of attack of the particles
completely effective. In such areas, coal
• Properties of the piping material
pipe components lined with special
The standard materials used for wear-resistant material are installed
the coal transport piping system are in place of the standard components
ductile iron elbows (Fig. 11-29) at the in order to control the wear rate and
Figure 11-30 | Coal pipe elbow failure
pulverizer outlet or exhauster outlet extend the service life.
and at the burner inlet, whereas all
straight piping between these points Experience with Wear-Resistant Coal
usually is carbon steel pipe. The ductile Transport Piping Materials
iron elbows usually are designed to a To reduce the rate of erosion and
two-pipe-diameter radius except for maintenance to the coal transport pip-
the burner inlet elbow, which has a ing, the practice of installing coal pipe
one-pipe diameter radius. elbows and other components made
The components of the coal from brittle cast material, such NiHard
transport system are assembled using and cast iron, often was used. These
mechanical couplings of either the materials were adopted because they
Rockwell or the Victaulic/Graylok provided wear resistance superior to
type. Both types provide strong assem- that of carbon steel. In 1988, however, Clean Combustion Technologies
Figure: 11.30/JG
bly and can accommodate the thermal the NFPA prohibited the use of brittle PMS5415 K
growth of the furnace/steam generator low-ductile materials in coal transport
and piping components. An extensive piping, because operating experi-
arrangement of pipe hangers supports ence found the NiHard and cast iron
the components of the coal transport components were subject to failure by
piping system. fracture. Each failure had the potential
Because the potential for erosion is to create a hazardous condition in
greater at the elbow and in the first 2 the plant through the release of large
to 3 feet (0.6 to 0.9 m) of straight pipe quantities of pulverized coal. Figure
directly downstream of it, deflector 11-30 is an example of an inlet elbow
blocks (kicker blocks) are installed in failure as the result of a coal transport
selected elbows both to absorb coal piping fire.
11-46 | Clean Combustion Technologies
reactive, but it is vital to keeping equip- moving parts subject to wear, such as a
ment in service for extended periods steam turbine. It is even more critical
of time. when the parts in question are ob-
The second type, anticipatory or scured from view because of the high
preventive maintenance, is the kind temperatures and pressures of normal
that prevents the strange noises and operation.
other undesirable indications from Maintenance issues for steam
occurring. Based on manufacturers’ turbines include solid particle impacts
recommendations, in-plant experi- (from foreign materials introduced
ence, improved data acquisition and to the turbine), erosion, deposits,
monitoring, and seat-of-the-pants corrosion, water impacts (from valve
judgment, preventive maintenance leaks), electro-erosion, vibration,
is done at planned intervals while seal damage, and bearing wear (from
the equipment is operating satis- contaminated or insufficient lubricat-
factorily. Its purpose is to eliminate ing oil). Wear, erosion, and corrosion
unscheduled outages resulting from can be detected through proper
component distress. With this type of inspection programs. These programs
maintenance, the timing is arbitrary. can be roughly divided into minor
Empirical data and the accumulation inspections (1–2 weeks in duration),
of favorable experience provide the intermediate inspections (2–4 weeks
basis for those decisions. in duration), and major inspections
The third type, predictive mainte- (6–8 weeks in duration).
nance, has the same aim as preventive
maintenance, but it employs more Inspections
sophisticated, computer-assisted meth- Minor inspections preferably are con-
ods of timing the actions. It is intended ducted when the plant is shut down for
to achieve the same or better results at reasons other than the steam turbine.
lower cost. Spot checks of control elements and
Maintenance and repair activities systems are performed, along with
beyond these three categories lead into inspections of the various safety
the realm of life continuation, which devices. Any operational problems are
is discussed in other chapters. Major reviewed, and corrective actions can be
efforts to obtain higher availability and taken to minimize further impacts.
indefinite continuation of equipment Intermediate inspections would
life are in the area of senior manage- be performed during a planned
ment decision making. outage. Bearings are checked, and
safety devices are tested. Steam valves
Maintenance and repair are inspected, and condenser and
of Steam Turbines feedwater heater systems are checked.
Reliable operation of the complete The low-pressure, last-stage blades
plant is critical to the overall economics are visually inspected. Boroscopic
of a power station. Regular, planned examination of all accessible turbine
inspections of plant equipment are components is carried out as well. All
carried out with the goal of avoid- control systems, lubrication systems,
ing unplanned shutdowns. This is and cooling systems are checked. If
particularly true of equipment that has deemed to be necessary, the turbine
11-50 | Clean Combustion Technologies
* Definition of the inspection types in accordance with the spare part category designations:
B-inspection: IB = OS + DM
C-inspection, complete: IC complete = OS + DM + IC
C-inspection after 150,000 equivalent operating hours: IC 150,000 = OS + DM + IC + BS1
enclosed in tubing and piping, it often (e.g., proper draining and nitrogen
is difficult to detect problems associ- blanketing) need to be followed in
ated with water chemistry. Potential order to minimize the potential for
problems caused by poor water corrosion. By following the manufac-
quality include magnetite deposition turer’s recommendations for shutdown
on the control valves, copper deposi- and standby conditions, maintenance
tion in the high-pressure turbine, costs can be reduced over the longer-
caustic embrittlement in the inter- term operation of the plant.
mediate- and low-pressure turbines,
and pitting corrosion and cracking Maintenance and repair
in the low-pressure turbine. Cold- of Generators
end corrosion diagnostics allow the Modern generators are very reliable
operator to monitor the temperature and require limited maintenance. The
and humidity inside the low-pressure main areas of concern are the insula-
turbine during operation. The opera- tion of the rotor and stator windings.
tor can decide to shut down the unit This insulation ages with time
if corrosive conditions persist. During according to the electrical, mechani-
shutdown, proper “lay-up” procedures cal, and thermal stresses in service.
11-52 | Clean Combustion Technologies
cost of the wet FGD system and the lower H2O consumption. This con-
large amount of water required to figuration typically results in slightly
operate the FGD system. lower SO2 capture (95% vs. 98% with
A second configuration is becoming wet FGD), and it does not produce a
very important, particularly for firing saleable by-product, such as wall-
lower-sulfur fuels (Fig. 11-35). In this board gypsum.
configuration, NOx also is removed
by SCR, but the SO2 is removed by Electrostatic Precipitators
a dry process. The particulate control To effectively address maintenance and
is by a fabric filter. The advantages of repair of an ESP, the operator needs to
this configuration are lower capital be concerned with the following issues:
cost, control of SO3 as well as SO2,
• Physical integrity and safety. Physi-
very low particulate emissions, and
cally, the mechanical components
that make up the ESP must be struc-
turally sound and safe for operation.
• Particle charging. Fly ash particles
must be successfully charged by the
Selective Spray Dryer
Catalytic Absorber high-voltage system.
Pulverized Reduction (SDA) • Particle collection and cleaning.
Coal Boiler (SCR)
Fly ash particles must be entrained
Fabric Filter (FF)
on the collecting surfaces, which
together with the charging system
minimizes the effects of back
corona. Particles held on the
collecting surfaces must be periodi-
cally dislodged and moved to the
Low NOx Air-Preheater Induced Stack
Burners Draft Fans ash hopper.
• Ash transport. The ash hopper
Figure 11-35 | Power plant configuration with dry scrubber must move its contents periodically
to the ash disposal system. (See
Clean Combustion Technologies Chapter 6 for a detailed discussion.)
Figure: 11.35/ MH
PMS5415 K
20%. To achieve this, additional baffle Normally, there are three discharge
plates may be added on the screens electrodes per collector panel. The
during the final checking of the gas collecting plates are firmly connected at
Collecting flow pattern on site. The collecting the bottom to shock bars (anvil beams),
Plate
electrodes are hung from a separate which ensures maximum transfer of the
and adjustable roof structure. The rapping energy (Fig. 11-41). The shock
suspension bar can be adjusted verti- bars are properly guided on the sides to
cally for easy alignments of the whole eliminate swinging or misalignment of
system, if necessary. the electrode curtain. Misalignment can
Each collecting electrode is a 30 inch cause electrical close clearances, which
Shock Bar (750 mm) wide, profile roll-formed lower the efficiency of the filter. In severe
panel, in mild steel, with a special cases, it will destroy the collecting plates.
Figure 11-41 | Collecting plates with shock shape to provide satisfactory current In normal operation, the collecting
bar distribution and stiffness. The profile electrode (CE) plates will collect the fly
edges are folded to improve stiffness ash particles. Over time, the layer will
Clean Combustion Technologies and to eliminate possible sparks to build up and need to be cleaned by me-
Figure: 11.41/SPS sharp edges. The design is based on a chanically dislodging the agglomerated
PMS5415 K
concept that requires no welding. particles. Alstom’s standard method for
cleaning the CE plates is with “tum-
bling hammers,” which are mounted
End Elevation View on a staggered shaft with one hammer
Loose or Missing Bolts / Nuts Wear at Shaft for each shock bar. As the shaft slowly
rotates, each of the hammers in turn
tumbles, hitting its associated shock
Worn Inner bar. The shock bar transmits the blow
Cracked Pin Loose Inner
Arm
Welds Arm simultaneously to all of the collecting
Wear at plates on that row.
Pins Broken Outer The design of the rapping mecha-
Arm
nism gives a high, steeply ramped
Worn or acceleration to shear off the dust cake
Missing in large agglomerates. The rapping
Washers
Missing Hammer
intervals are adjustable and have to be
Cracked Pin “Wheel” optimized for each precipitator field.
Incorrect Hammer Welds The measured minimum rapping ac-
“Peened” or Deformed Hammer
Size Installed celeration for a 15 × 49 ft (4.5 × 15 m)
“Wheel”
Side Elevation View curtain, with firm connection between
the plates and shock bar, is approxi-
Seized Hammer “Wheel”
(will not rotate freely) mately 150g at the upper far corner
measured perpendicular to the plate.
Each of the components listed below
should be examined during periodic
Seized Outer Arm
(will not “drop” or rotate) plant overhauls to ensure mechanical
integrity. Figure 11-42 gives some detail
on the tumbling rapper hammer itself.
Figure 11-42 | Typical problems of rapper hammers
Fabric Filters
Types of Fabric Filters and
Safety Considerations
Temperature °C
Hydrolysis Oxidation Price Relative
Fiber Abbreviation Acid Alkali Abrasion
(H2O) (O2) to PES
Continuous Peak
Polypropylene PP 90 95 5 5 5 3 5 1
Polyester PES 135 150 3 2 1 5 5 1
Dolanit Ricem PAC 125 130 4 3 4–5 3 3–4 1, 3
Ryton Procon Toray PPS 180 210 4 4 5 3 3–4 4, 5
Nomex APA 200 220 2 4 2 3–4 5 5
P84 PI 240 260 3 3 2 — 4 6
Teflon® Fiber PTFE 230 260 5 5 5 5 3 15
Fiberglass GLS 240 280 4 3 5 5 1 2–3
each type of bag. Table 11-2 sum- agent. Combined with water, SO3,
marizes the properties of the most a strong oxidizing agent, becomes
common filter materials. The most sulfuric acid, a strong acid and
important selection criteria for filter also a strong oxidizing agent.
media are the flue gas temperature, Either of these sulfur compounds
both continuous and peak, and the might degrade the bag material.
flue gas and fly ash chemistry:
increase the average emission rate by The functions of the EFFIC include:
100 fold for the affected bag and by 5 to
10% for a complete filter of 1,000 bags. • Minimizing dust emissions from
In high-ratio, pulse cleaned filters the fabric filter.
with an outside-to-inside filtration • Optimizing the dust layer thick-
direction through the filter elements, ness on the bags for the best
the filter bags are attached to the possible cleaning.
separation plate between the inlet • Minimizing the pressure drop and,
and outlet side of the filter with a thus, the energy consumption.
cuff containing a spring-steel band. • Reducing the total operating costs
The grooved cuff fits into circular of the fabric filter.
holes with narrow tolerances cut in • Bag failure analysis and valve fault
the bag plate and can be installed and detection.
removed without tools.
In all designs, a bag cage is required
Reasons for Emission Increases
to prevent the bag from collapsing
After Fabric Filters
under the pressure differential across
the filter fabric. The bag fits on the A number of reasons exist for a slow
cage with a certain looseness so that increase in emissions in the exhaust gas
it flexes between its filtration position path after a fabric filter that eventually
and the inflated shape at cleaning. This results in excessive emissions:
flexing is essential for the effectiveness
of the pulse cleaning. It also, however, • A gradual increase of gas flow
is a potential cause of failure because over years. This could be caused
of concentrated fatigue. Figure 11-45 by a gradual degradation of boiler
shows the inspection of a bag cage. performance and, quite often, by
slowly increasing air in-leakage
Fabric Filter Electronic Controls upstream of the filter.
The main function of most fabric filter • A production increase or changes
controls is the control of the pulse in fuel or other chemistry changes
cleaning in a fabric filter using com- that result in an increased gas flow
pressed air. Alstom’s electronic fabric and/or dust, which is more difficult
Figure 11-45 | Inspection of a filter cage filter integrated controller (EFFIC) is to handle.
a cost-effective solution designed to • Chemical or thermal degeneration
Clean Combustion Technologies
be installed in new fabric filters or to of the textile in the fabric filter that
Figure: 11.45/AR replace and modernize control systems results in a slow emission increase
PMS5415 K in existing fabric filters (Fig. 11-46). over time.
Several EFFIC control units can work
together in a network to control larger There are several reasons for
fabric filters. In such cases, each EFFIC rapid surges in emissions from a
would control part of the filter, one fabric filter:
line, one or more chambers, or even an
individual pressure tank. One EFFIC • Holes made by mistake during
is then configured to be the master installation (e.g., by a knife in
controller for the others. unpacking the bags)
Figure 11-46 | Electronic fabric filter
integrated controller (EFFIC)
control cabinet
Construction and Maintenance | 11-63
Air
Gypsum
Spray Tower Absorbers From Mw Dewatering
Tank Ball Mill Reagent System
for SO2 Removal Preparation
System
Types of Open Spray Towers
and Safety Considerations Figure 11-47 | Wet SO2 scrubbing system with major subsystems
Selection of SO2 removal technology
Clean Combustion Technologies
involves many considerations, includ-
Reagent Preparation Circuit Figure: 11.47/SPS
ing initial capital cost, long-term cost of PMS5415 K
reagent (lime or limestone), by-product The reagent preparation system gener-
requirements, landfill requirements, ally is straightforward (Fig. 11-48).
current and future SO2 removal Usually, the reagent is limestone, which
requirements, maintenance costs, long- must be ground before addition into
term coal flexibility, and future emission the absorber vessel to adjust pH. Repair
controls (mercury, trace elements, and and maintenance activities usually are
others). These multiple considerations focused on maintenance of the ball mill.
lead to different site-specific equipment; Typical repair activities involve rubber
however, the overall process and chemis- lining replacement, gearbox overhaul, Figure 11-48 | Typical horizontal ball mill for
grinding limestone
try are similar. and classifier component replacement.
Safety concerns with this equip-
ment are similar to those with other Recirculation Pumps, Headers,
plant equipment. Confined spaces and Nozzles
potentially pose a suffocation or Recirculation pumps are large, typically
chemical exposure hazard. The physi- 50,000–65,000 gpm (3,200–4,100 l/s),
cal size of the equipment creates fall and have a pump efficiency of between
hazards, and the strong chemicals 87 and more than 90%. The pumps
in the process create potential for generally are centrifugal, with a rubber-
chemical burn hazards. All of these lined casing and either a hard metal
concerns need to be addressed in a or rubber-coated impeller. The power Clean Combustion Technologies
site-specific safety plan before repair required is 800 to 1,500 hp (~1 MW) Figure: 11.48/AR
and maintenance operations are con- per pump. Components of a typical PMS5415 K
sidered. As described previously, there pump are shown in Figure 11-49.
are three major subsystems for the Wear generally is experienced on
modern wet FGD process, as shown the seals, lining, and impeller. A typical
in Figure 11-47. impeller is shown in Figure 11-50.
11-64 | Clean Combustion Technologies
present, the main reaction will produce and formic acid. Performance impacts
100
calcium sulfite. Additionally, pH are shown in Figure 11-54.
and additives such as dibasic acid and that have a strong impact on SO2 96
adipic acid also strongly influence wet removal, other important events occur 94
FGD chemistry. that are critical to operation. Particle 92
As mentioned previously, flue gas size of the ground limestone has a
90
containing SO2 contacts alkaline strong effect on how easily the desired 5.0 5.2 5.4 5.6 5.8 6.0
(limestone) aqueous slurry. The SO2 alkalinity can be achieved. Fly ash pH
pH
alkalinity, and the slurry is again circu- in the absorber. Natural oxidation is 5.4
lated back into the absorber section of the term used when no additional
5.2
the scrubber. The main determinant of oxygen is added to the absorber ves-
5.0
SO2 removal from a wet FGD is pH, sel, whereas forced oxidation refers 0 20,000 60,000 100,000
as shown in Figure 11-52. to the use of large air compressors Chloride Concentration, ppm
systems to form a scale on the walls of that has been distributed on the vac-
the absorber. uum belt filters. From a maintenance
and repair point of view, particular at-
By-Product Removal tention needs to be paid to proper belt
and Dewatering System tracking and tensioning, control of cake
The SO2 removal process in the thickness, and periodic measurement
absorber produces by-products that of cake moisture content.
must be removed in order for the There are two distinct operations
system to prevent the reaction tank in the filtration cycle: cake formation,
from becoming filled. The by-prod- and cake dewatering. Cake formation
uct treatment systems are tailored to occurs as the free water is removed
the specific needs of that site. Some from the slurry. The end of the cake
Figure 11-55 | Hydrocyclones used for
applications simply send the slurry formation portion of the cycle is noted
dewatering
to a pond, whereas others include visually by the free water disappearing
primary dewatering with thickeners from the surface of the cake. The cake
or hydrocyclones before ponding. Still formation occurs very quickly after
others follow with a secondary system vacuum is applied. The cake washwater
of vacuum filters or presses for further displaces dissolved contaminants from
dewatering. The secondary systems are the formed cake. Cake washwater may
used when the output is saleable gyp- be either recycled cloth washwater or
sum or the solids are stabilized before make-up water. After cake formation
sending to a landfill. and cake washing, cake dewatering
Generally, the slurry primary dewa- begins and continues through the
Clean Combustion Technologies tering system at a modern wet FGD remainder of the cycle. In this portion
Figure: 11.55/AR
system consists of hydrocyclones (Fig. of the cycle, water is removed from
PMS5415 K
11-55). Hydrocyclones can take as between the gypsum particles or
input the slurry from the absorber crystals. The filter cloth and dewatered
that contains approximately 15% cake pass over a small-radius discharge
precipitated solids. Using a cyclonic roller that separates the dewatered
action, the liquid leaving the cyclone cake from the cloth for discharge. The
is reduced to approximately 3% solids, vacuum belt filter design can include a
whereas the lower discharge, often cloth wash after cake discharge.
known as “underflow,” has a much
higher solid loading (55%).
Dry ScrubbinG Systems
Because commercial gypsum gener-
for SO2 Removal
ally needs to have a moisture content
of less than 10%, and because landfill Typical Spray Dryer Arrangement and
gypsum is more economical if its Operational Issues
moisture content is limited to 15%, an A typical dry scrubbing system is
additional component, a vacuum filter, shown in Figure 11-56. From the
is required. The filtration cycle begins standpoint of maintenance and repair,
when vacuum is applied to the slurry the dry scrubbing system can be
Construction and Maintenance | 11-67
• Atomizer disk. The atomizer disk to form the lime slurry. The system
is located below the scroll plate and includes:
extends into the flue gas path at
• Bulk lime truck unloading station
the top of the spray dryer absorber.
• Bulk lime storage
The disk is made of titanium,
• Lime slakers and make-up water
hardened tool steel, and ceramic
controls
components.
• Grit removal
• Atomizer lubrication system. The
• Lime slurry storage and atomizer
bearings and gears of the gearbox
feed pumps
and flex shaft are oil lubricated.
• Atomizer lime slurry and dilution
water controls
Generally, periodic maintenance
• Centralized Programmable Logic
on this machine is required to clean
Controller (PLC) and controls
deposits from the scroll plate. An in-
spection is recommended every 8,000 Handling bulk lime requires care
hours of operation; this consists of an and attention. Concentrated solutions
internal inspection of the mechanical of lime have a very high pH and are
seals, bearings, O-rings, and hard- very caustic. The lime also tends to
ware inside the flex shaft. If needed, cake on exposure to moisture, making
a rebuild of the flex-shaft atomizer it difficult to handle. As with a solids
should be done by a trained techni- handling system, wear can be an issue,
cian at the supplier’s facility/shop. although less so with lime.
Every 16,000 hours of operation, the
flex-shaft atomizer should be over-
Paul Lafferty
hauled, with special attention to the Phil Lafave
speed-increasing gearbox. Bruce Carney
Jim Sutton
Lime Preparation Jim Geyer
Charlie Hart
The lime preparation system takes the Richard Stone
quick lime and mixes it with water Bill Herman
Retrofits
Chapter Twelve
100 40,000
through performance enhancements
or maintenance upgrades. As a result,
50 20,000
many plants have invested substantial
capital to improve plant efficiency, lower 0 0
plant emissions, increase output, switch 1940 1950 1960 1970 1980 1990
Decade
to lower cost fuel, or increase plant
reliability and availability. Retrofitting Number of Units MW
Figure 12-1 | Unit age of installed U.S. base for Alstom units
Clean Combustion Technologies
Figure: 12.01/ (JG)
12-
PMS5415 K
12- | Clean Combustion Technologies
exist in areas that have a larger popula- capital, operational, and maintenance
tion of older boilers. costs, today’s utility managers have
When retrofitting existing power increased their focus on total owner-
plants, compromises are required ship cost; that is, what the expected
between an ideal design and exist- cost is over the life of a component.
ing physical constraints. Different All of the ingredients of long term cost
techniques and equipment may be have to be considered: lost genera-
required when retrofitting an existing tion due to forced outages, inspection
design compared to a white paper costs, maintenance costs, outage cycles,
design. In a retrofit situation, the outage length, unit efficiency, heat rate,
existing envelope limits the space and environmental considerations.
available for new equipment. When Initial cost, although important, is
developing solutions for existing not necessarily the overriding consid-
plants, the boiler designer must also eration. Consequently, this has put
account for the age-specific design is- increased focus on providing solutions
sues and existing plant condition. to long-standing problems.
Tube failures are the number one The success of this approach relies
cause for unscheduled boiler outages on the support of operators and
and account for almost 3% of all of owners working closely with the
the lost generation availability in the mechanical and performance design
United States. All boiler components engineers. This cooperation will
must be designed and maintained to facilitate the collection of invaluable
ensure that the boilers are available service data as well as insight into
when required. By reducing tube maintenance and performance issues
failures, a power generation owner and operational goals and strategies.
can realize a positive return on invest- A systematic approach, using root
ment through reduced operation and cause analysis, will define the cause of
maintenance (O&M) costs, reliable failure(s). Continuous interaction with
and available generation capacity, and the customer complements the analysis
optimal selection of units to supply by providing operational information
power based on lowest cost. that goes beyond service time and
An important factor to ensure long temperature to include fuel character-
term reliability of boiler pressure parts is istics, possible fuel switches, and other
the correct design to eliminate problem operational variables. Evaluation of
areas, while considering current and pro- changes in operation and their impact
jected future operations. Boiler pressure on unit performance will allow utility
part component replacements should management to make economically
be designed to minimize detrimental driven decisions to better meet perfor-
effects of highly corrosive fuels, heavy mance expectations.
slagging fuels, cyclic (peaking) opera- Based on recent market trends
tion, longer run time between outages, and forecasts in the United States,
higher temperature operation, and sub- the most common retrofits fit into
stoichiometric tuning for environmental one of the following categories: unit
considerations, as well as upgrading the uprating, emissions control, fuel
mechanical design to modern standards. switch/co-firing, and maintenance
Even though the current competitive upgrades/performance improvement.
environment requires close attention to Note that each plant will have its own
Retrofits | 12-
unique situation and resulting driv- section will discuss the boiler aspects of
ers. In retrofit situations, the use of boiler and plant optimization (Fig. 12-2).
best practices means that most of the Plant specific turbine and boiler mod-
available mechanical or performance els are set up and run simultaneously
upgrades would be used where appli- to develop an accurate picture of the
cable and economically feasible when potential plant retrofit. The turbine sup-
components are replaced or redesigned. plier will lead the optimization process
by modeling a turbine retrofit design.
UNIT UPRATING The boiler turbine interfaces are then
The power generation market is calculated and the effects on the boiler
becoming increasingly competitive and are determined. The boiler turbine
most utilities are looking to maximize interfaces, main steam flow, temperature
the output of existing plants. The and pressure, cold reheat temperature,
increased output is used to increase and final feedwater temperature are
revenue or in some cases offset the recalculated. The process is iterated to a
parasitic power loss of new emission converged solution and evaluated. The
control technologies installed at the interface points are modeled in consul-
plants. There is also an emergence of tation with the customer during this
potential CO2 requirements, where iterative process in which combinations
avoided CO2 emissions can be traded of main steam flow and pressure, as well
in emission trading schemes. As an as cold reheat (RH) steam temperatures
example, for a 400 MWe coal-fired and final feedwater (FW) temperature,
power station, a 3% efficiency improve- are compared with relative costs, equip-
ment avoids approximately 100,000 ment changes, power output, and net
tons of CO2 emission per year. plant heat rate. This allows the plant
To address these issues, a retrofit owner to make the best technical and
opportunity obtains more electrical economic decision regarding the plant.
output from an existing plant by us-
ing new technology for the boiler, the
turbine-generator, or a combination of Design Boiler Design Original
Optimization Criteria
the two. Any attempt to increase existing Geometry and Turbine Heat
and Constraints:
Equipment Balance Turbine Specs:
generation capacity requires a system • Steam Flow
• Inlet Flow, Temp, Press.
• Q Fired
evaluation approach to ascertain and • Emissions • Ffwt
Test Data as per Turbine and
address all of the potential impacts.2 As Boiler Requirements
• Others… • Reheat P
Boiler Specs:
a result of original design tolerances built • SH / RH Surfaces
into the plant, many existing coal-fired • Design Tilts
• SH / RH Sprays
plants have a hidden existing capacity. In • RH Press. Drop
Proposed Turbine Design
addition, major boiler or turbine compo-
Turbine and Boiler
nents may need extensive maintenance. Proposed Boiler Performance
Customer Input Geometry Fixed
Modifications
A replacement component that increases
Generator and Transformer Balance of Plant
the plant output almost always makes Information Information
economic sense. By exploiting these Preliminary Assessment Followed by Joint Project Review Off-Design Operation:
hidden system capacities and retrofit • Mill Configuration
Turbine Elastic (variable swallowing capacity) • Heaters Out of Service
needs, more generation can be achieved Boiler Surfaces Not Fixed • Boiler Fouling
from an existing plant. Turbine specific • Condenser Pressure Changes
• Part Load Operation
aspects of the optimized plant retrofits
are discussed in the turbine section. This
Clean
Figure 12-2 | The iterative thermal Combustion
optimizing Technologies
process
Figure: 12.02/ SPS
PMS5415 K
12- | Clean Combustion Technologies
Increasing the main steam flow to the superheater. This criterion may be
the original boiler design maximum the limiting factor in the amount of
continuous rating (MCR) and beyond steam flow available through a unit.
needs to be considered carefully. The The same check is needed on the RH
potential impacts to the boiler, turbine, safety valves.
generator, and balance of plant must be
evaluated. A logical starting point is to Water Circulation
review the current operating flow and, An assessment is needed of the ability
if below design, investigate ways to of drum type boilers to operate with
restore it. In some cases, fuel changes adequate safety margin at the increased
to lower rank fuels have reduced the flow. The circulation ratio (CR) is
output to below original design. defined as the mass flow ratio of the
If the main steam flow is to be in- circulation fluid to steam leaving the
creased beyond the original rating with circuit. A certain minimum CR value
or without a steam turbine upgrade, the should be maintained in order for a
approach should focus on incremental waterwall circuit to operate in a safe
increases. The point at which a par- zone for a given drum pressure and
ticular plant item becomes the limiting circuit fluid mass velocity (flow rate per
factor can be more readily identified. unit area). Of course, this condition is
Increased flow, therefore, involves the very boiler specific and is not always a
identification of major components and main concern.
balance of plant items, the relationship
between maximum component output Increased Firing Rate
and the main steam flow (which can The issues associated with increased
be accepted with current equipment), firing cover boiler furnace, pressure
and, finally, the modifications needed to parts, burners, pulverizers, air supply,
increase the duty. emissions, etc. Evaluating the firing
increases in incremental amounts and
Safety Valves reviewing the margins in each of the as-
Safety valves must be checked for re- sociated pieces of equipment will allow
lieving capacity and operating pressure an assessment to be made of the extent
based on current code. If the relieving and cost of modifications necessary to
capacity is exceeded, larger valve inter- reach a particular flow. An important
nals, larger valves, or additional valves performance goal for the boiler is to
will be required. Note that if the new maintain main steam and reheat steam
operating pressure is within 3% of the temperatures after the retrofit.
closure pressure of that valve, excess In certain boiler specific cases, boiler
leakage and subsequent maintenance efficiency, flue gas temperature control,
can occur. In some cases the affected and firing rate issues may collectively
sections could be redesigned to reduce point out a need for review of the ther-
pressure drop. With a turbine retrofit, mal performance of the economizer. In
the best solution is to design the HP several cases, upgrading the economizer
turbine for the required swallowing has been a viable solution for (1) main-
capacity at a lower throttle pressure. taining exit flue gas conditions for flue
Another concern is that the low set su- gas cleaning equipment, (2) lowering
perheater safety valve should lift before boiler heat input (Q fired) rates to ac-
the low set drum safety valve to protect ceptable levels, or (3) increasing project
Retrofits | 12-
0.24
0.22 (low)
0.20 0.20 0.20
tangential low NOx firing systems is
0.18 0.18 the utilization of concentric firing. In
0.15 a tangentially fired boiler, the coal and
0.13
0.10
air are injected in the corners and are
0.09
aimed tangent to an imaginary circle.
0.05 In concentric firing, a portion of the air
is aimed at a second imaginary circle
0.00
Eastern High-Vol. B Sub-Bituminous Sub-Bituminous C that is larger than the primary circle.
Bituminous to High-Vol. C A to B to Lignite
Bituminous This promotes an oxidizing atmo-
sphere near the waterwalls and changes
the aerodynamics, which impacts the
Figure 12-3 | Low NOx fuel comparison (T-fired units with SOFA) particle combustion time/temperature
Clean Combustion Technologies
history and the ash characteristics.
Figure: 12.03/ (JG) This combination has been shown to
PMS5415 K design features of staged air combus- mitigate the potential for waterwall
tion, early fuel devolatilization, and corrosion or wastage and to reduce the
CFS™ concentric firing system for the incidence of excessive ash deposition
secondary air. The differences among the on the boiler waterwalls.
options occur in the tradeoffs between In combination with finer grinding,
the extent of NOx emissions reduction the cumulative effects of separated
and the complexity and cost of material overfire air and concentric firing lead
modification and retrofit requirements. to the TFS 2000® R system for retrofit
Overfire air, or staged combustion, applications. This system can achieve
is the key component of the low NOx the levels required for NSPS for the
tangential firing systems. The two types lower rank coals.
of overfire air utilized are the close-
coupled overfire air (CCOFA), where Wall Fired
the overfire air is introduced within the There may be several different burner
existing windbox, and separated overfire solutions for wall fired boilers. The cho-
air (SOFA), where air is taken from sen solution will be based on the fuels,
the windbox and introduced into the original boiler manufacturer, and other
furnace above the burner zone. Both considerations. The radially stratified
are effective NOx reduction techniques. flame core burner (RSFC™ burner) is a
CCOFA is more economical on an typical wall fired burner (Fig. 12-4).
installed cost with a moderate NOx This burner uses a three zone,
reduction, whereas SOFA provides swirling airflow to delay mixing in the
a more aggressive NOx reduction near-burner zone. This produces the
strategy. As the regulated levels of NOx combination of a near field, high-
emissions have decreased, most low temperature, fuel-rich core, followed
Retrofits | 12-
not conform to the original physical equipment. This increase in the pres-
Retrofits to the combustion dimensions, clearances, or tolerances. sure capability of the fan will require
system offer some of the lowest This will either require repair or that the furnace implosion capabilities
replacement of existing equipment. of the boiler and ductwork be ad-
cost opportunities for reducing
The decision to repair or replace this dressed. The implosion phenomenon
NOx emissions. existing equipment is best determined is discussed further in the Chapter 7.
by the end user based on age, condi- The protection requirements are set by
tion, and economic factors. the NFPA code, which are currently
Damper upgrades are another based on either an absolute value of
modification included in low NOx pressure (+/–35 inch WG [+/–6.5
system retrofits. These utilize newer cm Hg]) or the test block conditions of
bearing designs and current siz- the forced draft (FD) and induced draft
ing standards for optimal damper (ID) fans. This requires a total analysis
operation and air control. There are of the boiler pressure part support
now fabric seal designs to replace system, including the air and gas ducts.
older corrugated steel seals. Again, Commonly referred to as a furnace pres-
the existing design and condition will sure upgrade study, the analysis includes
determine what upgrades or replace- positive and negative pressure in the
ments are appropriate. evaluation. In most instances, the nega-
The furnace buckstay system tive pressure condition is the controlling
configuration is designed to maintain factor. The modification required to
the integrity of the waterwalls dur- satisfy the negative pressure condition
ing pressure excursions. For existing will normally satisfy the requirements of
plants, the buckstay design will be the positive pressure condition.
dependent on the waterwall construc- For the analysis, the engineer
tion (loose tube, tangent tube, or makes two basic assumptions. The
welded design), the furnace dimen- first requires the existing structure to
sions, as well as contract or code be very near the original design. This
requirements in place at the time of implies minimal degradation of the
the contract. The location of SOFA structure. The second assumption
elevations can be affected by the requires the existing material to be in
size or location of the buckstays. In a “like new” condition. Although there
general, SOFA locations are chosen to is some minuscule degradation to be
avoid buckstays due to the extra cost expected, this is usually acceptable due
involved in relocating or bracing them. to material and design safety factors.
The engineer will analyze every pres-
Post-Combustion (NOx, SOx, sure load transfer point. The analysis
Particulates) process is very labor intensive, with
The installation of post-combustion minimal material requirements. The
controls such as SCR modules, return to the utility is an increase in
scrubbers (wet or dry), electrostatic the ability of the furnace to resist the
precipitators, or baghouse particulate increase in furnace fan size.
equipment (collectively called “back- The duct system of the unit is de-
end equipment”) usually requires the signed to contain the air and gas of the
addition of a larger induced draft (ID) boiler. The duct’s flat steel plate sur-
fan to overcome the additional pressure faces are stiffened to resist the internal
drop incurred by the installation of this pressure forces, the self-weight of the
Retrofits | 12-11
duct, wind loads, seismic forces, ther- There are cases where the duct plate
mal expansion differential forces, and and/or stiffeners can be in resonance
weight due to fly ash accumulation. with the fan frequency. This typically
The duct system is supported vertically requires the frequency immediately
and laterally to resist these forces. The downstream from the fan to be 45
expansion joints are designed to allow hertz or greater for the plate and the
differential thermal expansion between stiffeners. For areas downstream of
the duct component and the support flow restricting devices, the design
of the duct component. requires the stiffeners to have a first
Due to the use of expansion joints natural frequency of 25 hertz or
within the duct system, the internal air greater. Frequency is considered inde-
or gas pressure acting on the duct in- pendent of pressure-induced stresses.
ternal surface may result in unbalanced
pressure forces, which must also be Buckstay Design Criteria. The buckstay
resolved. Each duct component section system is required to limit the deflec-
is reviewed to determine if unbal- tion of the furnace tube walls as a result
anced conditions exist. The unbalanced of the pressure in the furnace due to
conditions are resolved by determining operation. The buckstay system consists
an acceptable load path to transfer the of a series of structural shapes attached
forces to the supporting structure. to the furnace tube walls. The buckstay
The unbalanced condition may system bands the furnace tube walls at
introduce forces directly into the struc- vertically calculated intervals. The buck-
ture, or create overturning moments, stays minimize the tube wall deflections,
which must be coupled into a guide or maintaining the integrity of the furnace
the hanger rods. This can lead to rein- tube walls. During operation, the buck-
forcement of the hanger rods, the duct stay system provides horizontal support
guides, and/or the associated support for the furnace tube walls, maintaining
lugs and brackets. the integrity of the furnace tube walls
and allowing the buckstay stirrups and
Duct Design Criteria. The analysis of buckstay corners to function as designed.
the duct stiffener and plate stress is the A system of stirrups transfers the
most important aspect of the analysis. furnace pressure load from the furnace
The stress in these components is tube wall into the buckstay system.
limited to the values specified by the The stirrup system is designed to
American Institute for Steel Construc- allow horizontal expansion between
tion (AISC) Code and standards the hot furnace tube wall and the cold
developed by the manufacturer. The buckstay. The furnace pressure load
analysis reviews the deflection of the is resolved as an end reaction at each
stiffener, limiting the deflection to corner. The end reaction is canceled
1/464 of the length of the member. out across the width and depth of the
Depending on the vibration criteria, furnace to the buckstay on the opposite
deflections may exceed this limit. In side of the unit.
cases where the stiffener ends at the Furnace pressure imbalances can
corner without overlapping the stiffen- occur as the result of a fan malfunc-
er from the other side, manufacturing tion, a furnace implosion, or a furnace
design standards limit the reaction that explosion. Under positive pressure
is transmitted to the corner. conditions, the furnace tube walls are
12-12 | Clean Combustion Technologies
pushed outward requiring the buckstay Beyond these values the steel will
system to restrain the force of the theoretically take a permanent set.
pressure. The outboard flange of the Allowable Stress Values: Allowable
buckstay goes into a tension load and stress values are simply a percentage
the inboard flange of the buckstay goes of the yield values. This percentage
into a compression load to resist the provides a safety factor in the design.
pressure load. The inboard compres- Allowable design stress for the bend-
sion flange has a tendency to buckle if ing is:
adequate restraint is not provided at the .6 * Fy = Bending Allowable
required intervals. Alstom’s design uses
the stirrup system to provide the lateral Gas Temperature Control. As discussed
support to prevent the buckstay flange in Chapter 5, the SCR is designed for
from buckling. The outboard buckstay a specific gas temperature window to
flange does not have a tendency to optimize performance. Most of the
buckle, distort vertically, or twist as a first SCRs installed in the United
result of the tension load. This case does States were designed to be operated
not require any restraint system to limit within a narrow temperature window
the buckstay tension load. based on the May to September ozone
Under negative furnace pressure season, when units are running at
conditions, the outboard flange of the maximum capacity due to peak electri-
buckstay can experience compression cal demand.5 Subsequent regulation
loading. Under this condition, the changes required these SCRs to be
outboard flange can buckle. Alstom uses operated all year long. This requires
a system of vertical structural shapes to operation at off-design load and gas
stabilize the buckstay against the com- temperature conditions. As the exit gas
pression load condition. Alstom designs temperature decreases with load, the
the furnace buckstay system using cri- challenge is to keep the temperature
teria as defined by the AISC Code and at near the design point as the load
Alstom standards to minimize the stress is reduced. There are several options
and instability of the buckstay system. available to maintain the temperature
window including gas or water bypass
Beam Stress: Foremost among the
systems, economizer resurfacing, split
criteria is stress and stability in the
economizer, and hot water recircula-
buckstay components and is limited to
tion systems.
values as specified by the AISC code
When the gas temperature avail-
and Alstom standards.
able to the SCR is too low at full load,
Buckstay Deflection: Manufacturing
a reduction in economizer surface to
standards limit deflection to 1/360 of
increase gas temperature is a relatively
the length of the member. The 1/360
easy solution. This addresses the SCR
standard is based on 8 inches H2O of
requirements but also reduces boiler
furnace pressure loading.
efficiency at all loads. In addition, the
Yield Stress Values: Structural steel
reduction of economizer surface will be
has a designated yield value (Fy) as
limited to the upper end of the control
indicated in AISC specification. For
temperature range. This will result in a
example:
finite control range. If a larger control
Fy for A36 Steel = 36 ksi range is desired, then one of the fol-
Fy for A7 Steel = 33 ksi lowing options should be considered.
Retrofits | 12-13
River Basin (PRB) in the United States TABLE 12-2 | FOULING POTENTIAL OF CALCIUM AND SODIUM IN COAL ASH
all have high calcium contents and some
Fouling Potential None Medium High Severe
fields also include high sodium contents.
Na2O (%) <1 < 2.5 2.5 to 5 >5
These coals will rank in the medium to
high slagging index with some rank- CaO (%) <5 5 to 10 ≥ 10
ing as a high slagging index. For plants F = % Na2O × B/A < 0.1 0.1 to 0.5 >5
designed for a low to medium slagging
index coal, this is an issue. enters the convective section. For new
The higher calcium and sodium units, the design is based on maintain-
contents also lead to an increase in ing a gas temperature that is lower
the fouling propensity of these coals. than the ash softening temperature.
Table 12-2 shows the fouling potential Since the 1970s, Alstom has built
for these constituents. many units designed for lower rank
coals. In addition, there have also been
Boiler Performance and Operation. many retrofits of units designed for
The challenge for today’s boiler design- higher rank coals to the lower rank
ers is to competitively adapt steam sub-bituminous C rank coals. Based
generator systems for the low rank on this experience, the ideal furnace
coals and to obtain the fuel cost and size for a unit firing a typical high
environmental benefits without limit- calcium, low sodium, sub-bituminous
ing the output of the existing unit. The C would be on the order of 1.7–1.9 ×
biggest issue during a boiler retrofit to 106 BTU/hr-ft2 NHI/PA (Fig. 12-8).
a lower rank coal is to accommodate Units designed or modified based on
the fixed furnace size and heat transfer this criteria have operations that are
requirements versus the higher slagging not limited by the waterwall slagging.
and fouling characteristics typical of New units would be sized for an
lower rank coals. Detailed analysis of average residence time (with air-
furnace and boiler design is presented staging) of 1.3 seconds to the bottom
in Chapter 3, but the concepts impor- of the panels/platens. Another way of
tant for retrofits are included. expressing the height is as a multiple of
the average furnace dimension (width
Furnace Considerations. For a new + depth)/2 for relative comparisons
unit, the furnace sizing (width × depth) across the range of unit capacities. A
is based on the fuel heat input and new unit designed today for a typical
ash quality, or the potential for ash to sub-bituminous C coal would have
slag, which would inhibit waterwall a height that is 1.0 to 1.15 times the
heat transfer. In addition, the furnace average furnace dimensions, which
height from the centerline of the top results in the 1.3 second average.
fuel elevation to the horizontal plane For retrofit boilers that meet the
at the bottom of the panels (horizontal size and residence time requirements
furnace outlet plane [HFOP]) and described above, little additional
resulting furnace residence time is a equipment is required to maintain
key design factor. The combination optimal furnace conditions. For these
of NHI/PA and residence time will units, proper maintenance and opera-
determine the vertical furnace outlet tion of existing steam waterwall soot
plane (VFOP) gas temperature that blowers is usually sufficient. For units
12-18 | Clean Combustion Technologies
been a very successful application for Convective Pass Soot Blower Coverage.
units firing lower rank coals with low To compensate for the nonideal gas
sodium content. The SFS economizer temperatures entering the convective
is not recommended for coals with section and the limited space for
higher levels of sodium due to the surface in the backpass, convective soot
very high fouling tendency of the ash. blower coverage is usually 100% for the
A typical SFS economizer consists of low rank sub-bituminous C coals. This
inline tubes with a spiral fin pitch of is necessary to maintain or regain all
approximately two fins per inch and superheater (SH), reheater (RH), and
a maximum fin height of 0.75 inches economizer thermal performance. An
(1.9 cm) with a thickness of 0.060 example of the coverage is shown in Fig-
inches (0.15 cm). Clear space between ure 12-11. In many cases these blowers
fin tips is maintained at 2 inches (5.0 will be used more frequently than pre-
cm), while the assembly depth is lim- conversion, i.e., once or twice per shift.
ited to approximately 5 to 6 feet (1.5 to Due to the increased frequency, steam
1.8 m) to ensure complete soot blower or air sources should be evaluated for
coverage (Fig. 12-10). the increased duty and tube shields or
thicker wall tubing should be installed
to offset the possibility of increased
tube erosion.
Economizer
Air Heaters.During conversions to
lower rank coals, air heater performance
Clean Combustion
Figure 12-11 | Example of soot blower coverage Technologies
Figure: 12.11/ SPS
PMS5415 K
Retrofits | 12-21
usually deteriorates. Increased air and been required. The PA fans have to
gas flows increase their respective draft supply an increase in hot airflow to the
losses through the machine. Many of pulverizers for coal drying. Typically
the lower rank coals have a higher foul- this is within the margin of the fan(s),
ing tendency, which will increase draft but in rare cases, re-tipping along with
loss and decrease thermal performance. motor upgrades is a viable option.
Typical remedies to improve these
conditions include the installation Milling Systems. The switch to lower
of hot and/or cold end soot blowers, rank coals also has to be accommodated
if they do not already exist, as well by the coal handling and pulverizing
as off-line water washing devices for systems. The heating value of sub-
removal of long-term fouling/plugging bituminous coals can be as much as 30%
with minimal outage time. In vertical less than high-volatile bituminous coal.
shaft air heaters, loose pack baskets This requires that the fuel processing
can be installed. These baskets are con- equipment be modified to handle this
structed from wider-spaced plates that fuel quantity increase. This requirement
are slightly loose in the bundle to resist directly affects the pulverizers.
plugging and improve cleanability. Any given size coal pulverizer has
a finite maximum capacity expressed
Fans. When switching coals, if the in throughput, lbs/hr (Kg/hr). This
coal has higher moisture than the de- can be either a mechanical limitation
sign coal (as is often the case for lower or thermal limitation depending on
rank coals), the moisture content the coal characteristics. During any
increases the mass flow of the flue retrofit that involves increasing the
gas. In addition, the moisture content coal throughput requirements of the
decreases the boiler efficiency by as pulverizers or changing the coal source
much as 5 percentage points, requir- requires an analysis of the milling sys-
ing the same percentage increase in tem to determine the coal throughput
coal flow and higher air and gas flows. capacity and options for increasing the
Thus, forced draft (FD), primary air throughput if required.
(PA), and induced draft (ID) fans
must be evaluated for both flow and Mechanical Capacity. There are two
head capacity increases. primary characteristics of the coal be-
Often times, the FD and ID fans ing ground that are integral parts of
can sustain full load operation by determining the mechanical capacity
using the original design margin for of a pulverizer. They are grindability,
both head and flow. If the margins which is the measure of the difficulty
are insufficient or have been used with which a coal is reduced in size,
in unit upratings, then changes to expressed as Hardgrove Grindability
add head and flow will be required. Index (HGI), and product fineness,
Typically re-tipping the fan blades and which is expressed as the percent-
upgrading the motors will obtain the age of the pulverizer output that can
necessary uprate in fan capacity. In a pass through a 200 mesh sieve. In
few cases, new rotors and motors have some cases with bituminous coals,
12-22 | Clean Combustion Technologies
moisture is also a factor in determin- addition, plant practices over time may
ing mechanical capacity. differ from recommended pulverizer op-
When evaluating pulverizer perfor- erational or mechanical set points. These
mance, the required mill throughput changes from original design conditions
in lbs/hr (Kg/hr) at MCR is com- may occur over the course of many
pared to the expected maximum mill years and can result in mill throughput
throughput for the given HGI and limitations. Quite often these limitations
desired fineness. Alstom recommends are not noticed due to the spare capacity
that the throughput at MCR does not designed into the milling system. When
exceed 90% of maximum capacity to there is a fuel switch to a lower rank,
ensure that the pulverizer is capable of lower BTU, higher moisture coal, the
processing the required amount of coal required increase in throughput will
when in a worn condition. highlight any mill mechanical weakness-
For coal blends, all of the mechanical es due to nonstandard or optimum parts
capacity is done using the grindability as well as incorrect set points. To elimi-
of the lowest HGI fuel. Due to the nate these choke points, the mill can be
propensity of the lower HGI coal to “blueprinted” with OEM parts, usually
collect in the pulverizer and make up upgraded, and all original tolerances and
the majority of the recirculating load, spring tensions as per recommendations
HGI values are not averaged. for the current fuel being fired. This
will result in the maximum mechanical
Dynamic™ Classifiers. Dynamic™ throughput of the milling system.
classifiers use a variable speed conical
rotor in place of the existing dual cone Thermal Capacity Limitations. Pulver-
static classifier. The addition of a Dy- izer thermal capacity is the amount of
namic™ classifier will increase pulverizer raw fuel the pulverizer can effectively
capacity due to a reduction in the re- dry. The thermal capacity is primarily a
circulation load created by the classifier function of fuel rank, available primary
for a given coal flow and fineness. The air inlet temperature, available primary
reduction in recirculation load repre- air mass flow, and desired pulverizer
sents an increase in classifier efficiency, outlet temperature. During initial pul-
reducing the unnecessary over-grinding verizer selection the thermal capacity
of the very fine (200 mesh) particles and is typically higher than the mechanical
discharging the properly sized particles capacity and is rarely an issue. Thermal
quicker. The capacity increase can be capacity can become an issue at a later
used to increase coal flow for a given time, particularly when a fuel switch is
product fineness or increase product being considered. Thermal capacity is
fineness for a given coal flow. determined using a mathematical heat
balance model. Pulverizers are usually
Mill Blueprinting. Over the course sized up to 100% of the thermal capac-
of a pulverizer’s operation, regular main- ity value, since wear has no effect on a
tenance includes parts replacements, pulverizer’s thermal capacity. Standard
mechanical adjustments, and operational airflow, recommended outlet tempera-
changes for a variety of reasons. It is tures, and specific heats are used to
common for a variety of non-original generate the required primary air inlet
equipment manufacturer (OEM) parts temperature for each operating case.
to be installed during this period. In These temperatures are then compared
Retrofits | 12-23
to the available primary air tempera- Higher airflows increase the power
ture leaving the air heater to determine requirements from the fans. If the
if there is adequate thermal capacity. existing motors are to be used, they
Generally a 50°F (30°C) margin is de- must be checked. Sometimes they can
sired between the required primary air be rewound for increased power.
temperature and the available primary
air temperature. If the thermal capac- Biomass Co-Firing
ity of the pulverizer is lacking, there Fuel Characteristics. Biomass fir-
are several options available to increase ing often involves locally obtained
the drying capacity of the pulverizer material as the cost to transport the
and throughput. lower heating value fuel can be-
come economically untenable with
Higher Hot Air Temperatures. Hot air large distances. The economics of
temperatures to the mill can be biomass firing are currently closely
increased by the use of in-duct gas or tied to the CO2 constraints and
oil-fired heaters. Alternatively, duct economic incentives to reduce CO2.
modifications can be made to take hot- Biomass does not have classifi-
ter air off of the air heater. cations similar to coal due to the
wide variety and local nature of the
Larger Exhauster Fan Wheels. For fuels. An incomplete list of biomass
exhauster style mills, larger diameter includes wood and wood products,
exhauster fan wheels will add head bagasse, straws, agricultural wastes,
and airflow to the system. Often times, construction wastes, and animal
a larger, conventional open paddle waste including derivatives of these
wheel design can fit within the existing products. Table 12-3 contains an
housing. There is also an option of example project where the fuels were
using a newer style high efficiency fan
(HEF) within the existing fan casing.
Using the HEF has an added benefit TABLE 12-3 | RAW BIOMASS FUEL PROPERTIES
of higher operating fan efficiency, Fuel Wood Pellets Palm Kernels Olive Pellets Olive Cake
minimizing fan power consump- Proximate Analysis
tion at the higher head and airflow.
H2O% 8.38 6.78 11.45 ND
Higher airflow also helps with the
Ash% 1.43 5.06 10.12 ND
mechanical capacity of the mills.
VM% 68.19 70.79 59.62 ND
Fan Speed Increases. Increased fan FC% 22 17.37 18.81 ND
speed can provide additional airflow Ultimate Analysis
through the mill for greater drying C% 46.41 45.14 44.18 42.8
capacity by the following methods: H% 5.15 5.75 4.92 5.75
N% 1.87 2.6 1.71 1.88
1. Add a higher speed motor and
O% by diff 36.44 34.24 27.25
new gear for bowl, which requires
mechanical analysis for bearings, S 0.21 0.23 0.14 0.19
shafts, structure and lubrication CI% 0.11 0.2 0.23 0.31
2. Decouple the fan from the pulver- GCV kJ/kg 18,906 19,235 18,691 17,588
izer and add a second separately GCV = Gross Calorific Value
driven fan
12-24 | Clean Combustion Technologies
all cellulose-based and thus have a in total heat input required. In a 20%
calorific value on a weight basis that co-firing scenario, the air and gas
are similar. weights are not significantly different
However the volume fired varied and, therefore, the effect on fans and
dramatically as the bulk density of downstream equipment is minimal.
these fuels varied by a factor of two. As the blended ash increases in the
For comparable heat inputs, different alkali elements such as calcium and
volumes of fuel must be processed, sodium, the rate of tube deposition
even though the heating value on a can increase. In addition, the fusibility
weight basis may be the same. Other characteristics of ash can vary as the
differences from coal include the high chemical constituents change. By limit-
amount of volatile matter. As can be ing the heat input to a nominal 20%
seen, the volatile matter of the prod- co-firing, the slagging and fouling poten-
ucts in the table are all 60% or higher. tial, although increased, can usually be
Coal contains around 30% volatiles. managed. In some scenarios, additional
Biomass fuel ash often contains cleaning equipment may be required.
alkaline metals. When these inorganic
compounds are in the form of salts or Firing Systems. For new boilers firing
bound in the organic material, these ele- biomass, the boiler design would be
ments are easily released. This creates a based on the characteristics of the fuel.
concern for the heat transfer sections of Biomass fuels have different handling,
the boiler. In addition, the high chlorine transport, and combustion charac-
content in some biomass fuels represents teristics compared to coal. Based on
a concern for boiler tubing. Chlorine these characteristics, steam output and
attack can occur on the tubing metal and conditions, etc., the firing system will
create wastage issues. For these reasons, usually be either a stoker system or a
in retrofit situations, biomass is usu- fluidized bed boiler.
ally limited to a percentage of the total With existing plants, the choice of
fuel heat input. In limited cases where firing system is limited. If a unit is to
biomass is retrofitted to fire at 100%, be converted to 100% biomass firing,
the boiler usually must be derated to a stoker can be installed in the bottom
account for the operational issues. of the boiler and traditional stoker feed
systems utilized. The total heat input
Boiler Performance and Operation. to the unit will be directly related to
As with coal switching issues, the the grate heat release rate and will usu-
addition of biomass co-firing will ally result in a significant unit derate.
be highly dependent on the original For a co-firing installation, the new
design of the boiler, age, and main- biomass burners typically are placed near
tenance of the equipment. Biomass the top of the windbox between adjacent
co-firing effects on the boiler are burners. The design generally must ad-
dependent on the type of biomass dress the following considerations:
and the quantity fired. In general, the
higher hydrogen content of biomass, • Existing arrangement of windbox
as well as the moisture content, will and physical restriction
usually lead to a boiler efficiency • Ensuring burnout of larger particles,
decrease and a subsequent increase which might fall through the furnace
Retrofits | 12-25
to adjacent tubes. This design allows rubbing. The failure would result in a
the lugs to operate at a much lower tube leak.
temperature than earlier generations The current spacer support system
of attachments. Due to the operating replaces the four lugs on the spacer
temperature of the assemblies, there tube with two interlocking lugs. The
is an axial expansion of the tubing. interlocking lugs prevent the spacer
Adjacent tubes will operate at slightly tube and the assembly tube from
different temperatures, resulting in dif- coming into contact. The arrangement
ferent expansions. The design of these prevents the tube failures due to the
lugs allows the tubes to move axially tube-to-tube contact.
independent of each other, while at the The interlocking support lugs
same time, maintaining alignment. provide a positive restraint, preventing
Alignment attachments that are not the assembly side-to-side movement.
welded to the tubes must be selected Because of the positive restraint, the
to operate at the gas temperature of vertical portion of the spacer tube gen-
the component. In most applications, erally requires an expansion loop. The
a higher quality material must be speci- vertical assemblies may operate at a dif-
fied, therefore increasing the cost of the ferent temperature or be of a different
attachment. material than the spacer tube. This dif-
ference introduces a thermal expansion
Changing SH and RH Fluid Cooled differential between the assembly and
Spacers to Modern Design the spacer tube. The expansion loop
The preferred method of controlling resolves the differential.
side-to-side movement of vertical Additional modifications include:
assemblies employs the use of a steam-
1. Reduce number of tube changes
or water-cooled tube circuit. The
in components to reduce shop
main function of the tube spacer is to
welds, minimize cost, simplify
prevent the assemblies from swing-
design, and minimize customer’s
ing in the path of the gas flow. If the
inventory requirements
assemblies are not restrained, over time
2. Change pendant section sup-
the vertical assembly could experience
port arrangement to eliminate
failure between the assembly tube and
tie welds
the seal where the assembly penetrates
3. Upgrade to split ring casting
the roof. The failure could be a tube
alignment design
leak or the failure of the gas seal, al-
4. Upgrade economizer and
lowing superheated gas into the roof
horizontal SH and RH support
enclosure area. The early spacer sup-
systems
port system consisted of four lugs on
5. Modularize components (with or
the spacer tube for each assembly tube.
without new headers) to improve
The spacer tube was located between
constructability and reduce
the first and second tube of the as-
outage length and manpower
sembly. The system was a very flexible
requirements
system. The downfall of the system is
that the spacer tube and the assembly
tube are free to come into contact with Furnace Bottom Replacement
each other. Over time, failures have In addition to the daily wear due to
occurred in one of the tubes due to the erosion and corrosion, the furnace
Retrofits | 12-27
their expectations. An example of the TABLE 12-4 | RESULTS FROM A 600 MW RETROFIT PROJECT
total plant concept is shown in Table
600 MW Plant Example 1977 1985 2001
12-4.
11,000 BTU/lb 8,600 BTU/lb
In the United States, a 600 MW Coal heating value
(25,586 kJ/kg) (20,000 kJ/kg)
plant over the course of years imple-
Generation (MWhrs) 12,200,000 13,100,000 16,700,000
mented a fuel switch and a unit
uprating with a new HP turbine. Max. Unit Capacity (MWe) 580 580 630
During this time, the unit realized 0.7 lb/MBTU 0.6 lb/MBTU 0.115lb/MBTU
NOx
(0.3 kg/kJ) (0.26 kg/kJ) (0.05 kg/kJ)
an increase in output and availability
while firing a lower rank fuel. NOx 6.0 lb/MBTU 4.8 lb/MBTU 0.52 lb/MBTU
SOx
(2.6 kg/kJ) (2.06 kg/kJ) (0.22 kg/kJ)
and SOx emissions decreased and
Operating Availability 75% 77% 90%
the outage interval was increased
significantly. This multiyear ongoing Pulverizer Capacity t 90,000 90,000 120,000
improvement project consisted of an Outage Interval (months) 12 18 36
HP turbine optimized plant retrofit
(OPR), a PRB coal switch, a low NOx
firing system, and mill modifications. substantial advances made over the
past 10 to 20 years. Improvements in
Steam Turbine Retrofits guide blade profile design alone have
Technological advances have led to yielded improvements in the region
the potential for realizing substantial of 2% on stage efficiency, with even
improvements in older generation greater improvements coming from
plants by retrofitting steam tur- advanced 3-D profiles and improved
bines with blade path components design techniques in all parts of the
of advanced design. The principal turbine (Fig. 12-13). Efficient, modern
of retrofitting existing steam plants blade paths also benefit from larger
in order to achieve cost effective numbers of stages than were possible
increases in output, efficiency, and in the past, and, thus from lower root
reliability is becoming well estab- diameters, improved blade aspect
lished. This has been driven not only ratios, and lower steam velocities.
by the trend toward deregulation of
the supply markets and the associated
95 Retrofit Technology
increase in competition, but also by
94
the advances in technology that have 93
been made in recent years by OEMs. 92 ‘Sustained’
Cylinder Efficiency, %
9% Improvement
ensuring that the engineer is able to stage in the process, a consistent model
establish a reliable and optimal design. containing the available dimensional
In general, the most efficient design data is established. Should there be
will have the maximum number of any conflicts or omissions, they can be
stages, with the associated reduction identified and dealt with. The design,
in root diameter, leading to low steam thus, develops as a “virtual assembly” of
velocities and long, efficient blade the new components under develop-
profiles. At the concept stage, however, ment with the detailed model of the
the number of stages must be opti- components that already exist. The
mized against other considerations, existence of this assembly ensures that
such as available axial length, rotor there is no risk of the late discovery
dynamics, and the economic evalu- of a misfit or of any clash when the
ation of each additional stage. Most module is finally assembled.
of this “concept” work can be done The details of the design itself are
with minimal dimensional informa- based on well-established practices that
tion on the existing machine. This have been refined on new modules,
work will form the basis of an effective with some evolution to suit the needs
and efficient measurement survey. of the non-OEM retrofit market. The
“modifications” are made in order
Retrofit Design to provide the increased degree of
The basis of the retrofit design is the adjustment necessary for retrofitting,
concept design, which exists, not only but are applied in such a way that the
as a concept, but also as a reliable and elements of the proven design features
detailed model in the CAD system and are unchanged. Standard established
a detailed analytical model in the rotor features will correct known “type”
design suite. This will be married with problems with the original equipment.
final thermodynamic data and detailed For example, the commonly used inlet
dimensional data on the components seal arrangements with “high-low”
to be retained from the existing tur- rings are known to give problems due
bine. The detailed design is, therefore, to distortion and wear in service and
a process of “adding” detail to the con- are commonly replaced at outages. The
cept. As discussed above, dimensional design, which replaces this as part of
data will consist of 3-D data from the retrofit, consists of piston rings fit-
either Faro or 3-D laser CMM equip- ted into the inlet stub of the new inner
ment, which is imported into the CAD casing that work against a new stellite
system in use. In addition, the manual sleeve fitted into the existing outer
measurements are also transferred into casing. This arrangement provides a “fit
the CAD system. At this point, a thor- and forget,” reliable, proven, and totally
ough consistency check can be made leak-free arrangement (Fig. 12-18).
in appropriate areas. It is also worth
observing that data from other similar Turbine Retrofit Technology
machines can be cross-referenced Steam turbines fall into the category
quickly and easily, to provide additional of either “reaction” or “impulse” design.
confidence and make maximum use of Alstom is in the unique position of
previous design work. Thus, at an early producing designs that include the
Figure 12-18 | Casing and rotor fit up
Clean Combustion Technologies
Figure: 12.18/ (TJS)
PMS5415 K
Retrofits | 12-35
best of both types by selecting the combined with laboratory testing for
best components from the combined both aerodynamic and mechanical
technologies. This is made possible verification. For different machines,
by having retained the expertise of either impulse turbine blades (ITB)
design and manufacture of both types or reaction turbine blades (RTB) or
of turbine designs, and having created a combination of both can provide an
a culture of sharing of technology optimal solution (Fig. 12-19).
and know-how across product lines.
Steam turbines are not mass pro- Impulse (Low Reaction) Technology
duced. Failures are rare and expensive. Blading. The impulse technology
Statistical feedback from operations (Fig. 12-20) consists of a bladed disc
is limited. Therefore, the sharing of
design knowledge and available operat-
ing experience from both technology
backgrounds has led to an improved
understanding of the fundamental
physics of steam turbine design. Ana-
lytical tools are extremely sophisticated
today. When design teams apply such
tools to both impulse and reaction
designs, then issues which, in the past,
have relied on empirical correlation
can be understood directly. Only with
this combined experience base and
expertise is it possible to “mix and
match” components safely. Combina-
tions of features must be done with
care. These must be properly designed
and analyzed.
Design Features
Low Reaction Reaction
At the heart of the retrofit, and a (impulse)
fundamental requirement in its justifi-
cation, is a high efficiency steam path. Figure 12-19 | Impulse and reaction blades
Alstom has continuously invested in Clean Combustion Technologies
Figure: 12.19/ (AR)
airfoil development for both impulse PMS5415 K
(low reaction) and reaction blading
technology. This includes both HP/IP
blades and LP front stages together
with new, retrofit-specific, penultimate
LP (L-1) and last stage LP (L-0)
blades. This has been achieved through
extensive use of advanced compu-
tational fluid dynamics techniques,
including fully 3-D viscous analysis
Figure 12-20 | Impulse blading
and diaphragm construction. The machined from a solid bar and include
moving blades are machined from a the following features:
single bar and include the following
• 3-D aerofoil design, scaled for spe-
features:
cific application, maximizing height
• 3-D airfoil design, scaled for while minimizing root diameter
specific application, maximizing and axial width for performance
height while minimizing root • Robust hook root fastening and
diameter and axial width for parametric design with two or four
performance hooks
• Robust pinned root fastening, • Integral shroud, with pre-twisted
parametric design with two or assembly where applicable,
three pins, two to six fingers, and an providing row continuity with no
efficient load-carrying mechanism requirement for closing blade and
• Integral tip shroud, with pre- ideal for vibration control and seal-
twisted assembly, providing row ing design
continuity with no requirement for
For impulse and reaction tech-
closing blade and ideal for vibration
nologies, aerodynamic improvements
control and sealing design
continue. Mechanical innovations, for
The fixed blades are also machined example in root fastening design and
from a single bar and then welded in materials developments, are allowing
the diaphragm rings with the following longer blades to be employed. As a
features: result of improved components, im-
proved analysis methods, and greater
• 3-D airfoil design, scaled for specif-
flexibility in design, increased num-
ic application, applying “controlled
bers of stages relative to the original
flow” technology where appropriate
design can be accommodated within
• Robust section profiles for SPE
the retained casing equipment, thus
resistance
providing another lever to generate
• Integral root and tip platforms
performance improvements.
with integrated fillet radii features
and accurate control of steam chan-
Last Stage Blades (LSB). The last stage
nel definition
blading in operation is likely to be of
• Welded diaphragm construction,
an old vintage, developed without the
with low blade numbers per row.
benefits of the latest computational
Kinematic support enables small
tools and optimization techniques.
radial clearances as the component
These older blades can have efficien-
remains circular
cies that are several percentage points
lower than is achievable today. Equip-
Reaction Technology Blading (RTB). The ment often is not ideally optimized to
reaction technology consists of a the output and condenser pressure.
drum-type rotor and separate moving Often there is scope for increasing the
and fixed blades (Fig. 12-21). Both the exhaust flow area by fitting longer last
Fixed Blade Rotating Blade moving blades and fixed blades are stage blades, even within the existing
Figure 12-21 | Reaction blades
casings. There are many different de- equipment LSBs for best fit to retrofit
signs of last stage blading in operation applications. More recently, industry
around the world. In order to be able has recognized the benefit of devel-
to offer an efficient retrofit solution for oping LSBs specifically for retrofit
such a diverse technical market, it is applications in order to gain most
valuable to have a number of efficient advantage through optimization. For
designs available. Again, for any given example, Figure 12-22 gives the range
application it is appropriate to select of large, LSBs available at 1,800 rpm,
the optimum last and L-1 stage blades offered by suppliers. As explained
from the best of both impulse and above, scaled versions of these blades
reaction designs. may also be applied for 1,500 rpm
Steam turbine manufacturers build applications.
their range of LSBs by scaling them From the measurements carried
from one rotational speed to another. out on standard LP modules of
By doing this, the operational condi- non-OEM equipment, it has been
tions of one last blade of the range possible to define the maximum LSB
are identical to those of the derived length that the module can accommo-
blades and the references in operation date, taking into account the existing
are those of the full family of blades geometry of the outer casing. Again,
it generated. Traditionally, suppliers this is an example of how flexibility
have made use of their range of new and innovation have provided further
1,800 rpm
Exhaust Area per Flow
m2 ft2
Alstom Westinghouse GE
design, the retrofit design includes pack- This necessitates cutting and re-
ers to permit easy adjustment. welding the extraction pipes. To make
this easier, standard practice is to add
Casing Baffles short lengths of pipe that can be read-
Where small gaps with close tolerances ily cut to the required length during
are required for control of steam flows, the installation, (see Fig. 12-27). This
as in the case of casing inter-space approach also provides the ability to
baffles, the design accommodated a accommodate misalignment between
“bolt-on” baffle, (see Fig. 12-25), that the new connection and the existing
could be easily adjusted on site with pipe by bending the new pipe spool in
available machine tools. Experience the required direction.
over a number of retrofit projects has The process for dealing with the
shown that the extent of adjustment interfaces during installation is fully
provided by the bolt-on baffle has not detailed on the interface drawings,
proved to be necessary. including tolerances. The interface
As a result of this experience feed- drawings are specific to the retrofit Figure 12-25 | Bolted baffle design
Clean Combustion Technologies
back, the design for the casing baffle has being installed and are in the custom- Figure: 12.25/ SPS
been simplified (see Fig. 12-26), so that er’s preferred units. Site installation PMS5415 K
the restriction is achieved from a combi- is facilitated by the design practices
nation of axial and radial clearances that discussed earlier. These practices have
are not so critical, resulting in cost sav- proven very successful and retrofits
ings on original manufacture and time of two, three, and four units at a
and cost savings during installation. station have been completed without
problems, based on the measurements
Blowdown and Pressure Taps taken on one unit. In some cases, the
In some cases, it is not practical to only measurements taken were on
obtain certain measurements during spare components.
a routine maintenance outage. These
would typically be parts that require Alignment
the removal of the bottom half inner The Alstom alignment philosophy for
casings, such as blowdown connec- new units is to set all bearings on the
Figure 12-26 | Simplified design
tions on combined HP-IP cylinders, or rotor natural catenary such that, at all Clean Combustion Technologies
pressure taps and extraction connec- couplings, the coupling gaps are parallel. Figure: 12.26/ SPS
tions on HP cylinders. In such cases, a When changing from shrunk-on disc PMS5415 K
flexible design is used with final adjust- LP rotors to the much stiffer, welded
ment at the site during the installation rotors, the natural catenary is signifi-
outage. An example would be the small cantly different. Realigning the complete
bore pressure connection for stage one unit to match the new, natural catenary
pressure in the bottom half of a casing. would be a major undertaking, requiring
Parts are provided such that the length large bearing/pedestal, generator, and
of the connecting pipe can be adjusted exciter moves, with significant program
by cutting and welding. and cost implications. Experience
has shown that it is often possible
LP Extraction Connections to stay with the original shrunk-on
For LP retrofits it is often necessary disc catenary, or at least minimize the
to replace the complete inner casing. height changes required, provided this
Figure 12-27 | Short pipe lengths for
Clean Combustion
extractionTechnologies
connections
Figure: 12.27/ SPS
PMS5415 K
12-42 | Clean Combustion Technologies
is known in the design phase and rotor for Alstom) be on-site during the
stresses and bearing loads are confirmed installation to ensure that the installa-
to be acceptable. Where a new retrofit is tion is carried out in accordance with
being coupled to an existing rotor with the drawings and other instructions
a shrunk-on coupling (such as an LP/ provided. The cost of these advisors
generator), then the goal is to maintain is generally included in the hardware
the parallel coupling gap philosophy supply contract. The TFA’s role is
at this position. This ensures that the to provide technical advice on disas-
bending loads across this coupling are sembly, inspection, adjustment, and
minimized and there is no increase in re-assembly. This includes:
the risk of fretting between the shrunk-
• Providing advice and assisting with
on coupling and its rotor. Minimizing
the outage scheduling
alignment changes is a particularly
• Providing the technical link
important part of the “design for instal-
between the site and the home
lation” process, since alignment changes
office during the installation
will have a major impact on the installa-
• Providing analysis of all strip down
tion duration and cost.
measurements and non-destructive
Planning for Installation examination (NDE) results
• Providing approval of all reassembly
An essential part of any installation, as
settings and clearances
with any overhaul work, is the up-front
• Compiling and issuing the final
planning for the activity. Coordination
installation report
with the site is critical and should be
initiated at an early stage, to assist and There are usually clear advantages
advise on the installation requirements to the customer in carrying out the
and installation planning. A part of installation using the customer’s own
this process is to visit the site to meet labor force when this is available:
key personnel and run through the fea-
• It is invariably less expensive than
tures of the installation requirements
using a third party or OEM.
to ensure that all are familiar with the
• It familiarizes the supervision and
product and any special requirements
labor force with the product and
of the installation.
practices of the supplier .
Installation • It is without doubt the best form of
training.
On steam turbine retrofits, it is not
necessary that the equipment supplier
carry out the installation of the retrofit. Start-Up and Operation
The customer can decide, either when As a general rule, the retrofit should
placing the retrofit hardware contract be made transparent to the operators.
or during the supply phase, whether to A major part of this transparency is
install the new equipment, to contract to minimize changes to the existing
with a third party, or to contract with operating instructions and procedures.
the equipment supplier. The supplier This philosophy becomes particularly
will usually require that a representa- important on multiunit stations, where
tive (a technical field advisor, or TFA, operators move from unit to unit.
Retrofits | 12-43
For a time there will be some units combustion byproducts from electric
retrofitted and others not retrofitted. power plants have changed over the
If the customer has specific concerns years. Additionally, wear and tear on
with the existing instructions, these older equipment, after 20 years or
often can be addressed and corrected more of operation, creates situations
or improved as part of the retrofit where major renovation of the equip-
supply. For example, on some units, ment must occur. In both of these
the start-up is extended by the original situations, a major renovation and
requirement to hold for prescribed upgrade, as opposed to the routine
periods after achieving a required re- maintenance and repair described
heat steam temperature. By supplying in Chapter 11, must be considered.
additional instrumentation to monitor Typically major upgrades involve
rotor stresses during the start-up and particulate removal or SO2 removal
modifying the start-up procedure, the equipment, as only this type of emis-
time taken to start these units can be sions control equipment has been in
significantly reduced. service long enough to fall short of
current performance requirements.
Balancing When considering upgrades, the
All new rotors are high speed bal- customer is typically faced with a
anced to a high standard before complex analysis with the following
shipment to site. During installation, factors:
it is essential to ensure that the rotors
• Current and anticipated air quality
are coupled together so that they
legislation at the target plant
rotate on the same centerline as when
• System-wide strategy for minimi-
balanced in the factory (the centerline
zation of emissions
of the journals). The coupled con-
• Capital cost of procuring and
centricity of rotors is the single most
installing an upgraded system
important parameter that ensures
• Length of time to install the
rotors operate with a low level of
upgrade and its effect on lost elec-
vibration, without the need for in situ
tric generation
balancing. Alstom installation practice
• Future operation and maintenance
requires measurement of journal/
costs
journal coupled concentricity as well
as the more usual coupling/coupling Each of these items is unique to
coupled concentricity using high each plant and to each corporate
accuracy dial indicators (0.0001 inch/ entity, with the result that there is not
division). The acceptance standard a single “one size fits all” approach to
is primarily on the journal/journal the correct answer.
concentricity, since this is clearly the
most important parameter. Upgrades for Electrostatic
Precipitators
Emissions Control Most power plants are equipped with
Equipment Retrofits electrostatic precipitators. Over years
As previously discussed, the require- of service, this equipment has reli-
ments for minimizing emission of ably proven that it can remove more
12-44 | Clean Combustion Technologies
Migration Velocity, Wk
Where: component, Wk, involves both art, in
� = Collection Efficiency of the Precipitator interpreting historical data, and scientific
e = Base of Natural Logarithm = 2.718
Wk = Average Migration Velocity, cm/s (ft/sec)
understanding of electrostatic principles.
k = A Constant, Usually 0.4 or 0.6 Some equations from electrostatics can
A = Collection Area, m2 (ft2) be useful in understanding the opera-
Q = Gas Flow Rate, m3/s (ft3/sec)
tion of the equipment (see below).
Examining these equations in total,
Input Power
we can see the collection efficiency is
driven by gas flow, migration veloc-
Figure 12-28 | Collection efficiency of the precipitator
Clean Combustion Technologies ity (which is affected by charge on
Figure: 12.28/ (JG) the particle), electrical field strength,
PMS5415 K
than 99% of the particulates from the and collection area. These parameters
flue gas exiting the boiler. While the become the focus of upgrade efforts.
fundamental principles of electrostatic Typical upgrades to improve collec-
precipitation are unchanged, over time tion efficiency are listed in Table 12-5
some major improvements in design and described in greater detail in
concepts have occurred, especially in subsequent paragraphs. The table is
the area of high voltage controls and grouped from lowest to highest cost
plate spacing. Collection efficiency is for convenience.
generally predicted by the Modified
Electrical Sectionalization
Some useful equations for understanding precipitation An electrostatic precipitator (ESP) is
divided into a series of independently
Charge on a given particle of size dp > 1 micron is given by an equation from electrostatics:
energized fields in the direction of the
3ε gas flow. Precipitator performance
q= π ε0Ec d p2
ε + 2 depends on the number of individual
where εo is the permittivity (8.85 × 10–12 C/V .m), ε is the dielectric constant of the particu- fields installed. Figure 12-29 shows an
late (usually between 2 to 8) and Ec is the electric charging field strength (V/m ). ESP consisting of four fields, which
Electrostatic force on the particle: is typical of a commercial ESP in the
power industry.
3ε
Fe = q ⋅E p = π ε0Ec E p d p2 Figure 12-29 introduces the termi-
ε + 2
nology for cells and chambers. Each
where Ep is the field at collection plate (V/m). section can be considered an inde-
The drag force on the particle: pendent precipitator and be adjusted
1 and operated accordingly. As a rough
Fd = 3π µ g d p w
K0 rule of thumb, each field removes
Usually Ep and Ec are the same. roughly 80% of the particles. Most
applications have at least three fields
Equating and rearranging we get
2
in the precipitator. However, to attain
3ε ε0Ec d p a collection efficiency of more than
w = K0
ε + 2 3µ g
99%, some ESPs have been designed
with seven or more fields. Previous
Retrofits | 12-45
experience with a particular applica- TABLE 12-5 | TYPICAL UPGRADES TO IMPROVE COLLECTION EFFICIENCY
tion is required to determine how
Issue Type of Upgrade Typical Best Fit
many fields are necessary to meet the
Particulate emissions Most cost effective way of achieving
emission limits.
from ESP are greater ESP electric or 30–50% reduction in particulate emis-
Each field has individual transform- than allowed or limit controls upgrade sions. Cost roughly 5% of complete new
er-rectifier sets, high voltage controls, operational flexibility ESP with minimal to no downtime.
and electrodes that, when energized, Low capital cost method of adjusting the
Flue gas
establish a current flow in the section. collection efficiency of the ESP by modi-
conditioning
fying the surface resistivity of the fly ash.
Since the particle loading varies so
much across the ESP, the ability to ESP sectionaliza- 50–100% improvement in performance.
tion and mechanical Makes best sense when ESP casing,
adjust the power input requirements upgrades with wide foundation, and support steel is in good
of each field becomes very important plate spacing, new condition. Cost is 50–90% of a new ESP.
in achieving optimal performance. electrodes, in- Implementation takes 8 to 15 weeks
creased plate height depending on ESP size and crane access.
The first field of an electrostatic
precipitator has very high particle con- An additional section can typically result
Addition of an in an 80% removal efficiency improve-
centrations. As these particles become another field to the ment. The cost varies and is driven by
charged, they create an electric field ESP the mechanical complexity and interfer-
that inhibits the flow of current from ences with other plant equipment.
the discharge electrode to the collecting Considered when condition of ESP is
plate. This is known as space charge. poor or where major change in perfor-
mance is required. New ESP cost can
Space charge is overcome by applying Total ESP replace-
often be $5 million or more. Installation
ment
the highest voltage possible, without time is 20 weeks or more. Often the new
spark over, and by using electrodes ESP is built adjacent to the existing ESP
to minimize plant outages.
that maximize corona current. Moving
further downstream in the precipitator, Considered where lowest particulate
emissions are required and limitations
the particulate loading is usually much Conversion of ESP
of the FF operating envelope (pressure
to fabric filter (FF)
lower, allowing the corona current to drop, temperature sulfur content, among
be relatively easily established. In each others) are not prohibitive.
section however, the voltage and cur-
rent need to be adjusted for optimum
performance. From these principles, Bus Section (16 typical)
it is relatively easy to understand the
following upgrades.
experience have progressed, the new Owners of the SO2 removal equip-
system standards for SO2 removal ef- ment have to determine the most
ficiency and operational reliability have cost effective and technically feasible
dramatically increased. As discussed in method to achieve the desired perfor-
the new wet fuse gas desulfurization mance and reliability improvements.
(WFGD) sections, 100% of the flue gas The analysis can be somewhat complex
is now typically treated with operational and site specific. For example, we can
reliability of more than 99% and SO2 consider the options for the following
removal efficiencies of greater than 98%. older SO2 removal system:
This large gap between the perfor-
Year put into service........... 1982
mance of the earlier technology and
Reagent................................. Lime
current state-of-the-art technology
(90% avail-
has given rise to numerous upgrade
able CaO)
options. Improvements that are now
Absorber SO2
generally possible include:
Removal Efficiency........ 90%
• Increase in the percentage of flue Overall SO2
gas treated Removal Efficiency........ 81%
• Increased SO2 removal rate Flue Gas Bypass.................. 10%
• Improved operational reliability Design Coal Sulfur............. 3%
and reduced number of outages for Absorber Diameter............ 22 feet
cleaning Recycle Pump
• Lower operation and maintenance Quantity.......................... 16 (4 per
costs absorber)
• Creation of a saleable by-product to Liquid to Gas Ratio........... 100 gpm/
reduce or eliminate disposal costs kacfm
Spray Levels
These improvements to the existing
per Absorber................... 3
SO2 scrubbers are made possible by the
By-Product.......................... Fixated
following equipment change options:
landfill
• Modification of reagent supply On quick inspection, there are
system to use limestone, lime, or numerous possible improvements and
other reagent results that can be achieved. These
• Change in the internal components range from doing virtually nothing be-
of the scrubber including new yond normal maintenance to complete
spray headers, addition of trays, change-out of the entire absorbers as
and elimination of wall sneakage shown in Table 12-6.
with performance enhancing plates The upgrade options will be
• Use of organic acid to improve presented in further detail focusing
removal efficiency on upgrade options for the absorber
• Addition of air compressors to itself, the inputs to the absorber,
force higher oxidation levels in the and, finally, upgrades to by-product
reaction tank to produce gypsum removal. Prior to beginning this
• Complete teardown of the existing discussion, for the hypothetical unit
multiple absorber vessels and replace- described above, the following results
ment with a single larger vessel were obtained (Table 12-7).
Retrofits | 12-51
Based on the site specific issues, avail- TABLE 12-7 | UPGRADE OPTION RESULTS
ability of capital, need to keep outages Organic SO2 Removal
Type Option Reagent By-product
minimized, and customer long-term Acid Efficiency (%)
strategy, virtually any of the upgrades Refurbish 1A Lime Fixated sludge No 98
can be justified as optimum. In general, 1B Lime Wallboard gypsum No 98
where highest reliability and removal
1C Lime Disposal gypsum No 98
efficiency is required without the use
Refurbish 2A Limestone Wallboard gypsum Yes 98
of organic acids, customers should
consider complete replacement of older 2B Limestone Wallboard gypsum No 90
WFGD equipment. Where high reli- 2C Limestone Disposal gypsum No 90
ability and good removal efficiency is New 3A Limestone Wallboard gypsum Yes 98
required, cost-effective upgrades can be 3B Limestone Wallboard gypsum No 98
made to meet plant operating require- 3C Limestone Disposal gypsum No 98
ments. Typically, 98% removal efficiency
can be achieved, often without the use
of additional additives. have been evaluated, analyzed, and
redesigned as new products, materi-
Upgrade Options within als, and information become available.
the Absorber Vessel From these efforts new solutions and
Improved Slurry Spray System design standards were developed.
(Headers and Nozzles). Over the years, The recycle spray system has received
the design concepts of various com- a large amount of attention, since
ponents within wet scrubber systems improvements here can be translated
12-52 | Clean Combustion Technologies
into substantial improvements in per- have headers that are metal and
formance, operation, and maintenance. most likely in good condition. In this
The recycle system is the single largest situation, the typical upgrade does
power consumer in the wet scrubber not require many header laterals to
system and can require high levels be replaced. In any case, the headers
of maintenance. Starting in the mid and installation procedures are always
1990s, the new design standard be- designed to minimize field labor while
came higher velocity spray towers with ensuring full coverage. A final benefit of
a greater number of lower flow, lower upgrades to spray nozzles and header
pressure drop spray nozzles to ensure systems is the potential improvement in
optimal liquid/gas contact. The net slurry flow that can result from lower
result was much higher performance backpressure (Fig. 12-34).
at a lower pumping horsepower. The
lower flow nozzles, although smaller in Performance Enhancement Plates (PEP).
size and flow than the nozzles provided The PEP, U.S. Patent No. 6531104,
on earlier systems, have free passage was developed as a result of in-depth
through nozzles that are larger and flow pattern and absorption studies of
less prone to pluggage than the older open spray tower systems to combat the
nozzles, with higher pressure drop and well-known gas sneakage problem at the
smaller free pass areas. With a larger walls of spray towers. However, the full
number of nozzles, tower spray cover- effect on performance enhancement was
age is much better. There is sufficient not quantifiable until the development
overlap of the sprays so that should of a unique measurement method to
a nozzle plug, the effect is much less verify this phenomenon. Gas velocities
than a plug in one of the old style, large at the wall can be 2 to 3 times higher
volume nozzles. Applying this concept than that in the balance of the absorber.
to older wet scrubber systems yields This phenomenon is primarily a result
many benefits and allows the owner of the lower spray liquid density at the
some choices on how to take advantage absorber walls that, when combined
of those benefits, whether it be power with the lower liquid levels, yields lower
savings, additional SO2 removal ef- SO2 removal efficiencies at the walls.
ficiencies, or both (Fig. 12-33). The PEP is essentially a guide vane
Essentially the recycle system strategically placed along the perimeter
Figure 12-33 | Hollow cone nozzles
and headers to enhance
Ź}Ź~
header/nozzle upgrade consists of the spray contact zone to improve the
performance
ČŅņčŇŇĞŷčč of redesigning the spray elevation liquid/gas contact. PEP plate sizing and
ʼnňŅʼn
nozzle placement to maximize liquid placement is specific to each installation
gas contact. This upgrade typically to maximize process improvement. As-
includes all new spray nozzles and pects that differentiate PEP from other
replacing or modifying the spray in situ process improvement devices are:
elevations to accept the new nozzle
arrangement. The extent of spray • Easily installed—low labor costs
header replacement or modifica- and short down time
tion depends on the age, condition, • Minimal impact on absorber pres-
materials of construction, and layout sure differential (typical 0.2 to 0.5
of the existing headers. Many plants inches WG [0.4 to 0.9 mm Hg])
Figure 12-34 | Tapered header showing
placement of nozzles
Flow Distribution Tray. The flow dis- Mist Elimination. Typical mist
tribution tray is a well proven device to eliminator upgrades consist of a stage
enhance the SO2 removal efficiency in system made up of three to nine vane
existing scrubber modules. The counter assemblies designed to utilize the exist-
current gas/liquid flow through the ing scrubber mist eliminator (ME)
perforated sections of the baffled tray supports. This is a high performance
has resulted in a significant increase in system that is designed for the typical
SO2 removal at the expense of about a gas velocities ranging up to 12 ft/sec
2 inch increase in system pressure drop. (3.7 m/s), but is capable of continuous
Flow distribution trays are typically operation of gas velocities of more than
located above or below the lowest spray 16 ft/sec (4.9 m/s) without adverse
levels in the spray tower depending on effect on moisture carryover, should the
site specific issues including access and need arise (Fig. 12-36).
potential for pluggage (Fig. 12-35). These mist eliminators are of
The important benefits of tray de- proven design and are used extensively
sign can be summarized as follows: throughout the scrubber industry. Key
advantages of the high performance
• Easily installed—low labor costs ME include:
and short down time
• No impact on recycle pumps • High droplet breakthrough
• Incremental installation (one velocity—The gas velocity at
absorber at a time as time and which there is an onset of mist
Figure 12-36 | Mist eliminators
Ź}Ź~
ČŅņčŇŊĞŷčč
ʼnňŅʼn
FRP vanes would also be a suit- the reaction tank in order to prevent
able selection for this application. solids buildup in the lances.
process gypsum slurry containing 15– 85 wt% solids. A liquid ring vacuum
20 wt% solids. The slurry is pumped pump provides the suction needed at
from the reaction tank to hydrocy- the drum filter cloth. A small amount
clone classifiers, which split the slurry of fresh water is needed for vacuum
into a low density stream of fines (the seal water and for cloth washing.
overflow) and a high density stream
of coarse crystals (the underflow). In Chloride Purge. The hydrocyclone
so doing, the hydrocyclones also clas- overflow is divided into two streams as
sify the slurry chemically: unreacted it leaves the overflow collection laun-
limestone is relatively fine and reports derer. The majority of the overflow
to the overflow; the product gypsum returns by gravity to the absorber reac-
is a coarser material and it preferen- tion tank. A portion of the overflow
tially reports to the underflow. The is diverted to the FGD purge tank,
hydrocyclone underflow product where it is combined with a portion
flows by gravity to rotary drum of the filtrate from the belt filters.
vacuum filters for further dewatering. These combined streams make up the
The hydrocyclone overflow is directly system purge that is sent to disposal
returned to the absorber reaction for control of chlorides and fines. The
tank by gravity. A small portion of the relative amounts of the two inputs to
hydrocyclone overflow is directed to the FGD purge tank are determined
the FGD purge tank to maintain the to provide enough removal of fines
steady state chlorides at the desired and impurities (limestone inert mat-
level. The chloride purge stream is ter, fly ash, and calcium fluoride) to
transferred from the FGD purge ensure the desired gypsum quality,
tank to the ash pond by the FGD while not burdening the disposal
purge pumps. The gypsum slurry, system with excess solids.
hydrocyclone underflow contain-
ing approximately 40–50% solids
by weight flows by gravity to rotary Pat Jennings
Keith Berger
drum vacuum filters. The rotary drum
Jim Sutton
vacuum filters dewater the underflow Phil Lafave
slurry to produce a gypsum cake of Walt Lacey
REFERENCES
1. R. Fernald, and G. Orrok, Engineering of Power Plants. New York 5. “Project Shows Potential for Optimizing SCR and Boiler Operations.”
McGraw Hill, 1927. Lehigh Energy Update, August 2006, Vol 24(2), Lehigh, PA.
2. “Maximizing Economic and Environmental Performance of Existing 6. P. Jennings, Low NOx Firing Systems and PRB Fuel; Achieving as Low as
Coal-Fired Assets,” P. Bartley, et al, 2008, Alstom. 0.12 Lb NOx/MBTU. Institute of Clean Air Companies Forum 2002.
3. F. Gabrielli and S. Goodstine. Utility Boiler Steam Purity Considerations. 7. “Case Study of a Dedicated Biomass Firing System at Fiddlers Ferry
1991 ABB Combustion Engineering Systems, Inc. Power Station, UK,” 2007, Alstom.
4. R. MacWhinnie and N. Nsakala, et al. NOx Emissions Performance of 8. S. Mathis, et al. Efficient Gas Cleaning with SF Electrostatic Precipitators.
Powder River Basin (PRB) and Reference Coals. Alstom Power Internal Fläkten, 1963/4, 1–12, pp. 93–110.
Report, 2003.
suggested readings
A. Lord, Features of turbine retrofit design, Alstom Retrofit Conference, San J. McCoach, Steam turbine retrofits for improved performance, Power Gen
Francisco, 2003. Europe, Barcelona, Spain, May, 2004.
S. Glover, Conceptual Mechanical design of a steam turbine retrofit, Alstom A. Holmes, Retrofit Installation, Alstom Retrofit Conference, San Francisco,
Steam Turbine Conference, Prague, 2005. 2003.
P. Hemsley, Building on the best of both worlds, Alstom Retrofit Conference, M. Oates, Economic justification of a steam turbine retrofit, Alstom Retrofit
San Francisco , 2003. Conference, Chicago, 2006.
12-57
Performance Testing
of Power Plants
Chapter Thirteen
The performance testing and analysis of power plants is a vital tool in the ongoing process of
maintaining optimum performance, reliability, and profitability of the plant. Understanding
how the unit as a whole is performing and then examining the individual components is the
first step in maintaining the system at peak performance. Performance testing is the key to
gaining this insight. Performance testing can be utilized to meet several objectives:
• Monitor ongoing unit performance is repeatability more than absolute CONTRACTUAL GUARANTEES
• Evaluate the unit for potential accuracy. The objective of this effort is VERIFICATION
modification or upgrades to identify changes in performance and After first commissioning or the
• Evaluate the unit for potential component degradation. modification of a unit, a performance
maintenance needs guarantee test is required to demon-
• Demonstrate guaranteed improve- COMPONENT EVALUATION strate that the unit has achieved its goal.
ment as a result of modification or When the routine performance This type of test requires the most ac-
initial unit acceptance monitoring indicates a problem with a curacy, since the guarantees are absolute
specific component or in a certain area values. This type of test will usually
OBJECTIVE OF TESTS of the unit, component testing may follow closely the applicable Ameri-
Testing is used to gather informa- be required to correctly diagnose the can Society of Mechanical Engineers
tion about the current condition of problem. This type of testing is specific (ASME) Performance Test Code.
the plant, unit, or component. This to the component being tested and the
information may be used for a variety issues associated with that particular ASME PERFORMANCE TEST CODES
of purposes. component. For more than 100 years ASME has
been developing Performance Test
PERFORMANCE EVALUATION/ DESIGN MODIFICATION Codes (PTCs). ASME, in cooperation
ROUTINE MONITORING When major changes are being consid- with the American National Standards
Probably the most important type of ered for a unit, current operating data Institute (ANSI), has issued PTCs
testing is the ongoing performance is used to evaluate the unit. The pro- for overall plant performance as well as
monitoring of a unit. This type of posed modifications are then factored for many individual plant components.
testing enables the unit to stay at in and the new unit performance is These Codes are recognized as the
peak performance on a continual predicted. Common design modifica- American national standard for the
basis. This type of program combines tions are fuel change, uprating, turbine determination of performance. While
online monitoring with periodic tests upgrade, and environmental control many of these Codes are written from
at specific conditions. The key aspect equipment additions. the perspective of an acceptance test,
13-
13- | Clean Combustion Technologies
the basic methodology detailed in the the overall plant heat rate as deter-
procedure may be used for other types mined by PTC 46. The 1998 Code
of tests. An acceptance test to deter- also included procedures to address
mine absolute performance usually fluidized bed combustion, as well as
requires the most stringent attention other environmental equipment within
to measurement accuracy and test- the boiler envelope.
ing detail. Similar codes in Europe The primary method for determining
include EN 12952-15: 2003 for steam efficiency is referred to as the energy
generators and IEC 60953-2 for steam balance method, formerly called the heat
turbines. Much of the testing method- loss method. This method is based on
ology that is presented below is based the energy balance relationship:
on ASME PTCs.
Output
blr =
PTC 46 OVERALL PLANT Input
PERFORMANCE Output = Input – Losses + Credits
ASME PTC 46 Overall Plant Per- Thus,
Losses Credits
formance addresses the corrected net blr = 1 − +
Input Input
power output, corrected heat input,
and corrected net plant heat rate. The
corrections are applied for variations of The following Codes are included in
the boundary conditions such as ambi- the 1998 revision:
ent conditions, heat sink conditions,
PTC 4.2 Pulverizers
and fuel composition. PTC 46 covers
PTC 4.3 Air Heaters
both Rankine cycle plants as well as
PTC 4.4 Heat Recovery Steam
combined cycle plants.
Generators
PTC 4 FIRED STEAM GENERATORS
ASME PTC 4 Fired Steam Genera- PTC 6 STEAM TURBINES
tors covers the determination of boiler ASME PTC 6 covers the determina-
efficiency. This Code was revised in tion of steam turbine performance for
1998 to reflect fuel efficiency, i.e., the steam turbines in a regenerative cycle
percentage of the heat energy in the fuel where feedwater heaters are present.
that is converted to energy in the steam The performance parameters that
and water. The previous Code, which may be determined from a PTC 6 test
was replaced with PTC 4, ASME PTC include the following:
4.1, was based on total energy inputs,
• Heat rate
considering energy that was not part of
• Generator output
the energy in the fuel. Since most users
• Steam flow
of the Code used the “short form” and
• Steam rate
considered only the fuel energy, they in
• Feedwater flow
fact were calculating the fuel efficiency.
When the Code was revised, the com-
mittee evaluated more than 16 different PTC 6.2 STEAM TURBINES IN
definitions of efficiency, all technically COMBINED CYCLE APPLICATIONS
correct, and selected the current PTC 4 ASME PTC 6.2 covers the determina-
method. This method allowed the boiler tion of steam turbine performance for
efficiency determined using PTC 4 to steam turbines in a combined cycle or
be used to calculate the heat input for cogeneration application. This Code
Performance Testing of Power Plants | 13-
provides procedures for testing and cessive, resulting in higher reheater (RH)
calculating turbine generator output temperatures or an increased RH spray Reducing air leakage into
corrected to reference conditions as flow. Engineers at General Electric1 the boiler improves overall
a measure of overall steam turbine developed a simple test methodology to
efficiency, as this additional
performance. estimate this flow. The method is known
ASME PTC 19 series outlines the as the Booth-Kautzman Temperature air is merely heated by
requirements for instrumentation and Shift Test. It is important to keep this the boiler and pumped up
includes the following codes: testing option in mind when evaluating the stack by the fan.
the performance of the boiler’s reheater.
PTC 19.1 Test Uncertainty
Since the steam leaks from the HP to
PTC 19.5 Flow Measurement
the IP, it bypasses the reheater, reducing
PTC 19.10 Flue Gas Analysis
the flow to the reheater. With less flow,
the reheater outlet temperature rises
OTHER TEST METHODS
and may incorrectly be interpreted as an
U.S. ENVIRONMENTAL oversurfaced reheater.
PROTECTION AGENCY (EPA)
REFERENCE METHODS AIR HEATER LEAKAGE
The EPA has a number of standard The measurement of oxygen with the
methods for measuring emissions from corresponding computation of flue gas
power plant stacks. These methods weight provides a powerful tool in the
include the following: analysis and diagnosis of several boiler-
related issues.
Method 5 Particulate Matter
Method 6 SO2 FURNACE OUTLET TEMPERATURE
Method 7 NOx
The most common devices for measur-
Method 9 Opacity
ing temperatures, including furnace
Method 10 CO
outlet temperature, are thermocouples.
Method 25 Gaseous Organics
Information on the proper applications
Method 26 Halogens
of thermocouples is found in ASME
Method 29 Metals
PTC 19.3, Supplements on Instru-
In addition, the EPA has provided ments and Apparatus.
specifications for continuous emis-
sions monitoring systems (CEMS) for TESTING RESULTS
the above listed materials. For most Testing is more than just taking a
power plants subject to EPA emissions snapshot of data in time and running
regulations, such CEMS are required some calculations to determine the
to demonstrate continuous compliance performance. Testing involves a meth-
with the appropriate regulations. odology of establishing the proper unit
operating conditions, ensuring the unit
high pressure-Intermediate is operating in a steady-state condition,
pressure (HP-IP) PACKING and collecting representative data to suit
LEAKAGE the test objectives. There are a number
On steam turbine designs where the of methodologies available to the user
HP and IP turbines are of an opposed to evaluate various components or the
design within a single shell, there is a entire unit. While the ASME PTCs
packing seal that has a leakage of steam are the most prevalent, there are several
from the HP bowl to the IP bowl. If the others that are available for certain
packing is worn, this leakage may be ex- components or desired results. The
13- | Clean Combustion Technologies
methodologies listed previously should the grid, and is determined after the
be viewed as the test engineer’s toolbox, auxiliary loads have been taken. The
a variety of methods to determine oper- difference between turbine cycle and
ating characteristics of a particular unit. unit refers to where the heat input is
Some of the typical results that an engi- determined. A turbine cycle heat rate
neer may want to test for are described uses the heat supplied to the turbine as
below, with more detailed discussion in the heat input and, thus, measures the
the validation and analysis section. efficiency of the turbine feedwater heater
and condenser cycle. The unit heat rates
HEAT RATE include the boiler and, thus, the total
Heat rate is a measure of the efficiency heat input to the cycle. The net unit heat
of the cycle. The units for heat rate are rate considers the total fuel heat input
BTU/kWh, which represents how to the boiler divided by the net power
much thermal energy is required to sent to the grid and reflects the true
produce a kilowatt of electric energy energy conversion efficiency. Any one of
for a one-hour period. Heat rate is the the previously defined heat rates can be
inverse of efficiency, in that it measures corrected to reference conditions. These
the energy required to produce a unit corrections adjust the current measured
of electricity. If the ratio were inverted heat rate to the heat rate that would
and the thermal energy converted into be present if specified external bound-
kilowatt hours, the cycle efficiency ary conditions were present. Ambient
would result. conditions represent a good example.
Cycle efficiency: If the cooling water into the condenser
was colder than the reference condition,
P the condenser would perform better and
cycle =
3412.14 × Q in the calculated heat rate would be lower
(better) than the reference value. These
where corrections are useful when comparing
heat rates measured at different times,
P = power in kilowatts
when ambient conditions were different.
Q in = t hermal heat supplied to the
Another example is correcting boiler
cycle in BTUs per hour
efficiency for fuel variations such as
Q in moisture content or the ambient air con-
HR =
P ditions for the inlet to a gas turbine in a
combined cycle plant. These corrections
There are several different variations allow the user to determine changes in
of heat rate used in a typical power performance that have occurred with the
station: turbine cycle heat rate, gross unit. ASME PTC 6, ASME PTC 46,
unit heat rate, net unit heat rate, and and ASME PTC 4 all provide details of
corrected heat rate. The differences in the corrections to be applied and how
these heat rate calculations are based on to apply them. Refer to those Codes for
different boundary envelopes around further reference.
the cycle. The prefixes “gross” and “net” The primary equation for corrected
typically refer to where the power is heat rate is:
determined. Gross usually refers to the
Q corr
power at the generator terminals, while HRcorr =
Pcorr
net usually refers to the power sent to
Performance Testing of Power Plants | 13-
The Orifice Probe. Another type of well application, spring loading of the
TC Junction furnace gas temperature measuring element is more practical.
Plug device is the orifice probe. The hot
gas is drawn through two meter Surface-Temperature Measurement.
ing orifices in series, with deliberate Another method of fluid temperature
cooling of the gas occurring between determination is surface-temperature
them. By measuring the two orifice measurement of the tubing containing
flow rates and the cold gas tempera steam or water. This method is only
ture, the initial gas temperature can possible where there is no heat trans
be calculated, since the weight flow is fer, because of temperature gradient
constant and the orifice flow differen effects from the hot gas or the steam
tials indicate volume flow, which is a film. In these zero heat flux areas, the
0 1 1 3
1 2 function of temperature. The advan tube wall gradient usually presents no
4 2 4
Inches tage of this type of measurement problem in measurement, although
is that it eliminates thermocouple even here precautions should be taken
Figure 13-1 | Thermocouples are used
radiation error. against boundary effects. Thermo
for surface-temperature couples are commonly peened directly
measurement of tubing Measurements Related into the tube metal or attached by
containing fluids at high to Heat Absorption
Cleanpressure
Combustion Technologies welding the hot junction to the metal.
Figure: 13.01/A.R. Heat absorption values are used in In both cases, the thermocouple is
PMS5415 K developing the surface heat transfer rela grounded and measurement of a
tionships and are determined from test number of points must be made by
steam temperature and flow measure separate circuits since series connec
ments, together with gas temperatures tions can form unknown loops using
and flows. Additional performance boiler metal as a third wire.
items in the water-steam cycle are pres
sures, pressure drops, and quality/purity Chordal Temperature Measurement.
determinations. Circulation studies in Chordal thermocouples are used to
furnace waterwalls and flow distribution determine heat flux in a waterwall tube.
between circuits are other examples of A 1⁄16 inch (1.6 mm) diameter ther
water-steam cycle testing. mocouple with a tip surface ground
flat, is inserted into a hole drilled into
Thermowell Measurements. Some the chord of a tube (Fig. 13-1). The
aspects, however, present more serious junction of the thermocouple is placed
problems than in the gas-air work. at the crown of the tube a known
Due to the high fluid pressures, the distance below the surface. Knowing
measuring elements are rarely exposed the TC location in the tube wall and
directly to the fluid stream. Measure the inside wall temperature (due to the
ment is commonly made by bottoming saturated steam conditions), the heat
contact with a metal well that proj flux can be determined.
ects into the stream and provides the
necessary mechanical strength. Several The Electrical Resistance Mapping
well designs are shown in the ASME (ERM) Intelligent Sensor
Performance Test Code, Supplements on Waterwall temperature monitoring
Instruments and Apparatus,6 PTC 19.3 for slagging detection or wastage
and in a paper by J. W. Murdock.7 In a detection are two major applications
Performance Testing of Power Plants | 13-
using ERM. ERM has been installed electrical current to pass through and
at four plants in the United States increases the electrical resistance (just
to detect wastage. Plant personnel like running water through a smaller
at three large utility companies were pipe). Increasing the temperature of
interested in waterwall temperature the metal also increases the electri-
monitoring for slagging detection. cal resistance by making it harder for
Each ERM module provides a the electrical current to pass through
temperature map of a large area of the the metal. Increasing the electrical
waterwall, typically 40 feet by 40 feet resistance for any reason increases the
(12 m by 12 m). Within this measure- voltage differences between measure-
ment area, the ERM system provides ment locations, given that the current
a grid of approximately 120 waterwall is constant.
temperature measurements. The ERM Special software in the ERM system
system views the boiler waterwall as uses the grid of voltage measurements,
a large metal plate and measures its along with a few local temperature
electrical resistance. The electrical measurements from thermocouples
resistance of the tube sheet metal mounted on the waterwall. The
changes with temperature in a known, two-dimensional grid of volt-
predictable manner. Using online ages is processed using mathematical
resistance measurements, compared deconvolution. A reference set of
to a baseline measurement, the ERM measurements (voltages and thermo-
system calculates the average metal couple temperatures) is taken at plant
temperature of the waterwall tubes. start-up. The software computes the
When a fixed, known, DC electrical temperature profile or metal thickness
current is passed through the waterwall loss on the waterwall by comparing
panel, the electrical resistance of the each set of new measurements to the
metal generates a very small electrical reference measurements. Using the
field in the waterwall panel. This is the independent temperature measure-
two-dimensional version of Ohm’s Law. ments from several thermocouples,
Although the voltages are very small,
they can be measured with a microvolt-
meter (Fig. 13-2).
The ERM module injects an
electrical current into the waterwall
and measures the resulting voltage at
more than 100 locations. The data
acquisition modules are designed by
Alstom Power Plant Labs and built Power Micro-
Supply voltmeter
with sophisticated electronics that can
accurately measure voltages within a
few millionths of a volt. The electrical
resistance of the panel is proportional
to the metal thickness and to the elec-
trical resistivity of the metal. Erosion Waterwall Panel
the software separates the increased deadweight gages, the liquid pressure is
electrical resistance due to higher metal line pressure, its value determined from
temperature from the increased electri- the weight total. In deadweight testers,
cal resistance due to metal erosion. the liquid is in a closed system, with
Since the ERM detects voltages at the pressure acting on the instrument
the microvolt level, attention must be to be calibrated and on the weight
paid to electromagnetic interference. piston, from which the pressure value
ERM sensor wires may need to be is determined. Pressure measurements
shielded. A site ERM test procedure are typically made using electronic
should check for ambient electrical smart transmitters. Their sensitivity,
noise. Solid-state electronic mo- linearity, repeatability, and accuracy are
tor controls, such as those used on very good. Experience has shown that
variable speed pumps, may create these devices either work accurately or
noise that disturbs the ERM. When do not work at all. These devices come
an outage is still in progress, electri- in a variety of ranges to match the ap-
cal welding on the waterwall induces plication pressure.
stray currents that are detected by
the sophisticated and sensitive ERM
FLOW
circuitry. Therefore, reference testing
is often done during quiet hours in Water and Steam Flow
the outage schedule. Both steam and feedwater flow are
From a physical viewpoint, the obtained by pressure-differential
ERM system uses wires connected in a measurement across flow nozzle or
two-dimensional grid to the insulation thin-plate orifice sections. The flow
pins on a waterwall section. The ERM element accuracy is usually in the
power supply establishes a 50-amp range of 0.6-1% of full scale, depen-
direct current flow in the grid. The dent upon the flow element design.
resistivity curve of the metal waterwall Refer to ASME PTC 19.5 on flow
is known. The current is constant, and measurement accuracy. Some plants
thus the ERM sensor wires become have test flow nozzles installed on low
the terminals of a microvoltmeter. The pressure, low temperature condensate
typical grid is 11 pins by 11 pins, with points in the feedwater cycle to avoid
pins on four foot (1.2 m) spacing to undue extrapolation of the original
give a coverage area of 40 by 40 feet calibration data. Their calibration
(12 × 12 m). The ERM principle is is performed with laboratory weigh
patented by Alstom Technology, Ltd. tanks. Extrapolation of flow coef-
as U.S. Patent number 6,288,528. ficients to the condensate conditions
of flow, temperature, and pressure
PRESSURE is usually accomplished on the basis
The fundamental reference pressure of the Reynolds number and other
and pressure-differential measur- dimensionless criteria. While a few
ing device is the deadweight tester. central stations have reheat steam
Deadweight instruments employ flow meters, it is more common to
incremental weights acting on a given- determine this quantity from primary
sized piston that floats when the liquid steam-water flow, extraction flow to
and weight pressures are equal. In the high-pressure heater (heat bal-
Performance Testing of Power Plants | 13-11
ance), and the appropriate turbine of gas being handled has on the sizing
leakage corrections. and cost of the gas treatment equip- Flow measurements are often
ment. The determination of total crucial for understanding what
Calculation of Flow Through an Orifice quantities of gaseous and solid con-
is happening in a power plant.
or Nozzle. This procedure calculates stituents in the gases leaving a steam
flow through a differential metering generator depends on an accurate Errors in flow measurements
element in accordance with ASME quantification of the gas flow itself. For can lead to unnecessary
PTC 19.5-2004. instance, in the sampling of dust, the shutdowns or load reductions
amount of material caught in a sampler due to emissions constraints.
2(P )32.17405 is reduced to a dust concentration of,
Qm = 300 da2 ()C
4 (1 − a4 ) say, pounds (kg) of dust per 1,000
pounds (kg) of gas. The total quantity
where of dust being emitted by the source,
then, is the product of dust concentra-
Qm = c alculated mass flow rate in tion and total gas flow by weight.
lb/h
da = e lement bore diameter at Boundary Layer Effects. Measure-
actual flowing conditions ment of Solids in Flue Gases, by P. G.
= e xpansion coefficient for W. Hawksley, defines the approach
flowing fluid = 1 for water of the British Institute of Fuel.8 This
C = coefficient of discharge publication discusses the effects of
= fluid density in lb/ft3 nonsymmetrical gas flow distribution
∆P = differential pressure in psi and boundary layer profiles. It should be
a = b eta ratio at actual flowing noted that the equal-area zone method,
conditions as typically used, can easily miss the
D = tube diameter boundary layer flow profile. Very small
incremental areas would be required for
d the velocity probe device to “see” all of
=
D the boundary layer. However, traversing
d time can be reduced if the boundary
a = a
Da layer is evaluated in detail for one test
only and then used to correct coarser
subsequent traverses. This approach is
Da = D + p D(T − 68)
justifiable since each flue or stack will
da = d + e d(T − 68) have its own characteristic boundary
layer, which is a function of the specific
where system geometry and flow turbulence.
There have been occasions where
p = pipe thermal expansion factor boundary layers representing 5% of the
e = e lement thermal expansion flue diameter have produced flow errors
factor of about 10–20%.
T = temperature
Variation in Results Using Pitot Static
Flue Gas Velocity and Flow Measurement. Pitot static tube velocity
This information is included because measurements have been coupled with
of the profound effect that the quantity the equal-area technique for many
13-12 | Clean Combustion Technologies
years. The results of total gas flow can stacks and fans frequently display high
be quite varied and misleading, if care levels of vortex energy. A basic problem
is not given to the selection of the associated with vortex formation is that
measurement plane, or consideration vortices will persist for many diameters
not given to the selection of the proper downstream from the source. The vorti-
type of flow measurement device for ces have a very strong effect on dynamic
the flow directionality that is pres- flow measurement devices.
ent. Table 13-1 presents the results of Occasionally, reverse (or negative)
several total gas flow measurements flow may be encountered. A question
using the pitot static tube and equal- then arises as to how to calculate the
area techniques as called for in ASME total volume flow rate from this type
PTC 27. The results listed are derived of data. Fluid flow continuity requires
from a typical utility steam genera- that all velocity points, both normal
tor installation operating at constant direction and reverse flow, be arithmeti-
load. Observe the gross divergence of cally averaged and applied to the total,
measured values and the necessity for cross sectional area of the plane of the
careful choice of measurement point. measurement. If this is not done, the re-
As noted, the flow as measured can circulated gas flow (the negative values),
vary from 104 to 150% of rated value. which can represent 5 to 15% of volume
throughput, will increase the positive
Power Plant Flow Patterns flow values by the same amount and
Industrial ductwork and flues are de- produce a “measured” flow that is high.
signed to connect pieces of equipment at
minimum cost and minimum space re- Velocity Measurement
quirements. The resultant flow patterns The classical approach to the measure-
quite often deviate from the desired ment of fluid velocity has been by pitot
uniform and unidirectional condition tube. Several references are available
required for accurate quantification of which describe the developing history
flow. Flow separation and reverse flows, of this device. Appendix E of Combus-
often found in the major runs of flues, tion, 3rd edition, 1981, titled “Velocity
are corrected only by major pressure Calculations by Graphical Methods,”
drops, such as an air heater or excessively presents graphs that can be used for
long runs of ductwork. Flow patterns in quick determination of gas velocities
from pitot tube readings.
TABLE 13-1 | GAS FLOW MEASUREMENT
The Pitot Tube. The typical flow
Gas flow measured at different locations in accordance with methods given in ASME PTC measurement device is the pitot tube
27 and by using a pitot static tube. Observe the gross divergence of measured values and
the necessity for careful choice of measurement point. shown in Figure 13-3. The impact, or
Measurement Point Gas Flow CFM Percent Rating
total pressure hole, is the most useful
part of the probe. Hydraulically, the
Rated Value 300,000 100
impact pressure is the true sum of
Mechanical-Collector Inlet 383,000 128
the local static and dynamic (veloc-
Mechanical-Collector Outlet 449,000 150 ity) heads. Also, the impact hole can
Induced-Draft Fan Inlet (Precipitator Outlet) 313,000 104 have a cone of response of about 60
Induced-Draft Fan Outlet 424,000 141 to 70°. Therefore, accurate angular
positioning of the probe for total
CFM, cubic feet per minute
(impact) pressure measurements is
Performance Testing of Power Plants | 13-13
FUEL ANALYSIS
a prior version of ASME PTC 19.5
which specifies isotropic turbulence Analyzing Fuel Characteristics
and the RMS approach. It is relatively In its broadest sense, fuel analysis
easy to apply this guide if the signal is refers to determination of all physical
sinusoidal. However, if the signal is not and chemical properties of a fuel.
sinusoidal, turbulence effect is difficult
to evaluate. This problem can be resolved Solid Fuels. The basic type of chemi-
by using a true RMS voltmeter to pro- cal analysis for coal is the proximate,
duce a true RMS reading independent which describes the fuel in weight
of the applied wave form. percentage of fixed carbon, volatile
The effect of flow-stream vorticity material, ash, and moisture content.
has just the reverse effect on the velocity In the ultimate analysis, the fixed
readout. Vorticity represents a well or- carbon and volatile material contents
dered flow field of significant curvature. are reported in terms of total carbon,
Flow curvature, in turn, produces a very hydrogen, oxygen, nitrogen, and sulfur,
definite radial pressure gradient. If the this form being required for the calcu-
radius of curvature of the flow is of the lation of air quantities. Fuel analysis
same order of size as the measuring and higher heating value (HHV)
device, the device will not be measur- determination are made in accordance
ing either the correct dynamic or static with the Test Code for Solid Fuels
pressures. Vorticity of small size and PTC 3.2 and ASTM Standards.13 This
high intensity can, therefore, contain is a constant volume determination.
a significant amount of dynamic, or Because the fuel is burned at constant
pressure, energy that will not be read if pressure in the steam generator, the
the measuring device is physically larger HHV at constant volume, as deter-
than the size of the turbulent vortices mined in the bomb calorimeter, must
being produced. Unfortunately, there be converted to a constant pressure
is no easy way to evaluate the effect of HHV. Throughout the Codes, this
vorticity on measurements of dynamic HHV for constant pressure combus-
flow. Therefore, vorticity should be tion is referred to as the HHV. (When
minimized or eliminated by egg crates testing by the input/output method,
or screens if no other flow traversing only the HHV and the moisture con-
station without vorticity is available. tent of the fuel are required.)
Recommendations on reducing the Generally, fuel sizing refers to
major sources of error in obtaining total crusher-prepared coal sizing measured
gas flow rates, based on the above, are with screens of one inch (2.5 cm)
given in Quantification of Stack Gas Flow, opening and larger, and is of impor
by C. L. Burton.12 tance mainly for fuel bed distribution
problems in stoker firing. Coal fineness
POWER usually refers to suspension burning
Electric power is typically measured sizes and, for pulverized coal, involves
using a watt-hour meter connected to sieving of 50 to 200 mesh (openings per
metering class, current transformers, inch) particle sizes. In both cases, a series
and potential transformers. For a de- of screen sizes is used, with the fractions
tailed discussion of the measurement retained or passed by each size being re-
of three phase, electric power, refer to ported. Knowledge of the size consist or
any one of several ASME PTCs: PTC overall fraction relationship, rather than
6, PTC 6.2, or PTC 22. just one size, is required. For example, in
Performance Testing of Power Plants | 13-17
pulverized coal firing, combustion effi- calculated by using the standard heat-
ciency (carbon loss) can be more affected ing values of the constituents. Where
by larger size consist, whereas grinding gas samples of large volume can be
power is more closely related to the finer provided easily, the classical method
size percentages. Another physical prop- of heating value determination, by
erty related to pulverizer performance is calorimeter, is feasible. For instance,
ease of grinding, or grindability index. recording calorimeters are often found
This is determined from the amount of at transmission-line entry points to a
power required to pulverize a prepared user’s plants for billing purposes. Many
sample of the test coal in the laborato- plants have their own calorimeters.
ry. The results are then compared with
a standard sample. In the Hardgrove VALIDATION AND ANALYSIS
method, the standard is 100, with The most important part of the testing
decreasing numbers indicating progres- of a unit is the review and validation
sively harder grinding coals. of the test data. Validation of the data
is accomplished by evaluating the
Liquid Fuels. Oil fuels are reported data collected as a whole and by using
in an ultimate analysis together with engineering relationships and mass and
HHV and sample density, viscosity, and energy balances to confirm the data.
flash point. Density can be stated as
either standard specific gravity, referred COMBUSTION CALCULATIONS
to water, or degrees American Petro- The combustion calculations allow us
leum Institute (API), the two terms to determine the pounds of theoretical
being readily convertible. Additional air required to burn a fuel of a given
properties are pour point, sediment, composition per million BTUs, or per
and presence of solid impurities such pound of fuel. This information provides
as metallic salts. These properties may the ability to compute the majority of
influence pumping, storage, slagging, the heat losses for boiler efficiency on a
and corrosion problems. In general, percentage basis, based on the fuel fired.
the ultimate analysis is sufficient to
determine fuel-air relationships for O
11.51C + 34.3 H − + 4.335S
byproduct liquid fuels such as pulp-mill 7.937
TDA =
black liquor, refinery wastes, pitch, and HHV × 100
others. Operating problems may require
special tests in addition to the standard where
physical property determinations.
TDA = theoretical dry air lb air/
Gaseous Fuels. Gaseous fuels are million BTU
usually combinations of saturated C = carbon in fuel percentage
and unsaturated hydrocarbons. by weight
Their analysis is reported as a mole, H = hydrogen in fuel percentage
or volume, percentage of these by weight
constituents. Common analyses O = oxygen in fuel percentage
include low-temperature distillation, by weight
mass-spectrometer, and adsorption S = sulfur in fuel percentage by
methods. From this constituent analy- weight
sis, the HHV, at standard volumetric HHV = higher heating value of fuel
conditions and saturated or dry, can be BTU/lb
13-18 | Clean Combustion Technologies
where
If the carbon loss is significant (greater than 1%), Cb should be used in lieu of C
Coke
Natural Gas
Coke-Oven Gas
Refinery Gas
800
Blast-Furnace Gas
Calculating the theoretical air on a wet basis can be used as a check on the fuel
A, lb Air/106 BTU Fired
Oil
Bagasse
750 analysis.
700
O
Wood
937
650
7.
600
Awet =
HHV × 100
550
Increasing Hydrogen Content The calculated theoretical air should be within the dark blue areas as shown in
Figure 13-8.
Figure 13-8 | Combustion-air requirements
for various fuels at zero EXCESS AIR
excess-air—a range of values
as an approximate
Clean Combustion function of
Technologies 31.32C + 11.528S + 13.334 N
hydrogen
Figure: 13.08/ SPS content R=
HHV × 100
PMS5415 K
O2 R + 10.338 × TDA
EA% =
(
TDA 2.73 − .13068O2 )
where
O2 = p ercentage of oxygen measured in the flue gas on a dry volume basis.
C H S O
AO2 = + + −
12.01 4.032 32 32
N
AN 2 =
28.02
Performance Testing of Power Plants | 13-19
EA%
EA =
100
Mole fractions
O2dvmf = ( AO2 × EA)/ TotM
C
CO2dvmf = 12.01
TotM
S
SO2dvmf = 32
TotM
N 2dvmf = 1 − O2 dvmf − CO2 dvmf − SO2 dvmf
Weight fractions
32O2dvmf
O2dw =
TotP
44.01CO2dvmf
CO2dw =
TotP
28N 2dvmf
N 2dw =
TotP
Air to combustion
TDA × HHV
TDAlb =
1000000
DryAir = TDAlb (1 + EA)
MoistAir = SpecHum × DryAir
WetAir = DryAir + MoistAir
13-20 | Clean Combustion Technologies
Fuel in products
Wetprod = WetAir + 1 – Dr
MF = (M + 8.936H)/100
Moisture in products
Mprod = MoistAir + MF
Dry products
BOILER EFFICIENCY
The primary method for determining efficiency is referred to as the energy balance
method, formerly called the heat loss method. This method is based on the energy
balance relationship:
Output
blr =
Input
Output = Input – Losses + Credits
Thus,
Losses Credits
blr = 1 − +
Input Input
Performance Testing of Power Plants | 13-21
The ASME PTC 4 Code considers compare results on a unit from dif-
A good fuel analysis is the basis 16 losses: ferent time periods, a correction to
a reference fuel analysis or ambient
for all combustion and efficiency
1. Dry gas condition is possible when using
calculations. the energy balance method.
2. Water from fuel
3. Moisture in air • Identification and quantification
4. Unburned combustible of losses—Since each of the losses
5. Sensible heat in residue are determined, a comparison
6. Hot Air Quality Control System to the design condition quickly
(AQCS) reveals the source of degradation.
7. Air infiltration
8. NOx The detailed calculations for PTC
9. Surface radiation 4 follow.
10. Additional moisture
11. Calcination Boiler Efficiency Calculation
12. Water in sorbent Methodology by ASME PTC 4
13. Wet ashpit loss The methodology is as follows:
14. Recycled material
1. Use the ultimate analysis of the fuel
15. Cooling water
and the carbon in the ash streams
16. Heat internally supplied to air
to determine the carbon burned,
preheater (APH) coils
theoretical air, and theoretical air
corrected using stoichiometric
and 7 credits:
calculations. These calculations
1. Dry air are done on either a lb per million
2. Moisture in air BTU basis or a lb/lb fuel basis.
3. Sensible heat in fuel Most losses are determined on a
4. Sulfation percentage loss per lb of fuel. Some
5. Auxiliary power losses are determined on a total
6. Sensible heat in sorbent BTU lost basis and are converted
7. Energy in additional moisture to a percentage per lb of fuel by
dividing the fuel flow and heating
In PTC 4, a series of worksheets are
value. This will require iteration
provided that walk the user through
using an initial estimate of the fuel
the calculations. There are three main
flow and recalculation when the
advantages to using the energy balance
efficiency has been determined.
method for boiler efficiency versus the
input–output method: Coal Ultimate Analysis (mass
percentage in fuel):
• Greater accuracy—Since the losses MpCF = percentage carbon
and credits are measured and they MpHF = percentage hydrogen
represent only 10-15% of the total MpNF = percentage nitrogen
energy, the impact of measurement MpOF = percentage oxygen
errors are significantly less than MpSF = percentage sulfur
those made in measuring the input MpH2OF = percentage moisture
or output. MpAsF = percentage ash
• Ability to correct to reference HHV = higher heating value
conditions—When attempting to BTU/lb
Performance Testing of Power Plants | 13-23
Combustibles in refuse
CR = MpFA×MpCFA +
MpBA ×MpCBA*
where
Carbon burned
DR × CR
MpCb = MpCF −
100
Theoretical air
where
MFrThA = theoretical air lb of air per lb of fuel
MFrThA
MqThAf =
HHV
where
MqThA = theoretical air lb of air per BTU
MFrThACr
MqThACr = lb/ BTU
HHV
MFrThACr
MoThACr = moles/ mass fuel
28..963
13-24 | Clean Combustion Technologies
where
MFrWA = specific humidity lb H2O/lb dry air
Moles of wet products of combustion
MpH 2 F MpWF MFrWAdz
MoWPc = MoDPc + + + + MoWSb
201.6 1801.5 18.015
3. Use the measured percentage of oxygen in the flue gas at both the air heater
inlet and outlet to determine the excess air level using the ASME equation.
Excess air based on oxygen volume dry basis
DVpO2 (MoDPc + .7905 × MoThACr )
XpA = 100
MoThACr(20.95 − DVpO2 )
where
DvpO2= percentage oxygen dry volume basis
Excess air based on oxygen volume wet basis
VpO2 (MoWPc + MoThACr(0.7095 + MoWA))
XpA = 100
MoThACr(20.95 − VpO2 (1 + MoWA)
where
VpO2= percentage oxygen wet volume basis
4. Calculate the products of combustion and the flue gas composition at both
the air heater inlet and outlet using stoichiometric calculations. Subscript z
denotes location specific information.
Dry air entering the boiler ahead of location z:
XpAz
MqDAz = MqThACr(1 + ) lb/ BTU
100
XpAz
MFrDAz = MFrThACr(1 + ) lb/ lb fuel
100
MpWF
MqWF = lb/ BTU
100 × HHV
8.937 × MpH 2 F
MqWH 2 F = lb/ BTU
100 × HHV
Gas from sorbent (if applicable, such as for a circulating fluidized bed (CFB)
type unit) Product gas weight is an
MFrCO2b important boiler design
MqCO2b =
HHV parameter as well as a necessary
factor in efficiency calculations.
Water from sorbent (if applicable)
MFrWSb
MqWSb =
HHV
MrStz
MFrWADz =
MrF
where
MFrWADz = mass fraction of moisture added per pound of fuel fired
MrStz = mass flow rate of steam/water added
MrF = mass flow rate of fuel (estimated initially)
MFrWADz
MqWADz =
HHV
Steam/water
Residue
Tk = T + 459.7
HDFg = –0.1231899E3 + 0.4065568 × Tk + 0.5795050E–5Tk2 +
0.631121E–7Tk3 – 0.2924434E–10Tk4 + 0.2491009E–14Tk5
Flue gas
6. Calculate the temperature of the flue gas leaving the air heater corrected for
no leakage.
Mean specific heat of air
HATFgLv − HAEn
MnCpA = BTU/ lb F
TFgLv − TAEn
MnCpA MqFgLv
TFgLvCr = TFgLv + − 1 × (TFgLv − TAEn) ºF
MnCpFg MqFgEn
where
HATFgLv = enthalpy of air at flue gas leaving temperature
HAEn = enthalpy of air at air entering temperature
Performance Testing of Power Plants | 13-27
where
MqDFg = lb dry flue gas per BTU
HDFgLvCr = enthalpy of dry flue gas at the corrected temperature leaving
the air heater.
where
MqWH2F = moisture from combustion of H2 in fuel
HstLvCr = e nthalpy of steam at the corrected temperature leaving the air
heater
HW Re = enthalpy of water at reference temperature
HwvLvCr = enthalpy of water vapor at the corrected temperature leaving
the air heater
Loss due to moisture in air
where
MFrWA = specific humidity of air lb H2O/lb dry air
MqDA = lb of dry air per BTU
HWvLvCr = enthalpy of water vapor at the corrected temperature leaving
the air heater
where
MpUbC = lb of unburned carbon per lb of fuel
Correctly estimating the losses HHVCR = higher heating value of carbon in refuse = 14,500 BTU/lb
from the system provides a HHV = higher heating value of fuel
more accurate calculation of the (2) Loss due to unburned hydrogen in residue (not normally present)
efficiency. MrRs × MpH 2 Rs × HHVH 2
QpLH 2 Rs =
MrF × HHV
where
MrRs = mass rate of residue
MpH2Rs = percentage of H2 in residue (mass weighted for multiple streams)
HHVH2 = higher heating value of H2 61,100 BTU/lb
MrF = fuel flow lb/hr
HHV = higher heating value of fuel
(3) Loss due to carbon monoxide
HHVCO
QpLCO = DVpCO × MoDFg × MwCO
HHV
where
DVpCO = CO percentage dry volume
MoDFg = moles of dry flue gas per lb of fuel
MwCO = molecular weight of CO 28.01
HHVCO = higher heating value of CO 4,347 BTU/lb
HHV = higher heating value of fuel
where
Mr Pr = mass rate of pulverizer rejects
MrF = fuel flow
Performance Testing of Power Plants | 13-29
where
Vp Hc = total hydrocarbons percentage dry volume
Mo Fg = moles of dry flue gas per lb of fuel
Mw Hc = m olecular weight of hydrocarbons (use MW of reference gas used
for analysis typically propane 44.096)
HHVHc = higher heating value of reference gas used
HHV = higher heating value of fuel
where
MqFgEn = lb flue gas per BTU entering AQCS
MqFgLv = lb flue gas per BTU leaving AQCS
HFgEn = enthalpy of flue gas at entering temperature
HFgLv = enthalpy of flue gas at leaving temperature
HAAqLv = enthalpy of wet air at temperature of gas leaving the AQCS
HALvCr = enthalpy of wet air at corrected temperature of gas leaving the
air heater
Loss due to air infiltration between the normal boiler exit (typically the gas econo-
mizer outlet) and the air heater exit not included in the leakage into hot AQCS:
where
MqALg = lb air per BTU entering air heater
HALgEn = enthalpy of wet air at temperature of air entering air heater
HALvCr = enthalpy of wet air at corrected temperature of gas leaving the
air heater
13-30 | Clean Combustion Technologies
where
DVpNOx = NOx percentage dry volume
MoDFg = moles of dry flue gas per lb of fuel
MwNOx = molecular weight of NOx 30.8
HrNOx = heat of formation of NO 38,600 BTU/lb and heat of forma-
tion of NO2 35,630 BTU/lb
HHV = higher heating value of fuel
(
Hcaz = 0.2 TDi )1/3
or
Hcaz = 0.35VAz 0.8
Hraz = 0.847 + 2.637 E −3TDi + 2.94 E −6TDi 2 + 1.37 E −9TDi3
QpLSrc = ∑ ( Hcaz + Hraz )Afz (TMnAfz − TMnAz )
QrLSrc
QpLSrc = 100
MrF × HHV
where
Hcaz = convection heat transfer coefficient
Hraz = radiation heat transfer coefficient
TDi = temperature of surface–temperature of ambient assumed to be 50
unless for surfaces known to be higher
VAz = air velocity near surface, assumed to be 1.67
Note: TDi and VAz may be measured for the test and multiple areas or
they may be assumed as noted above.
(
QrLRyFg = MrRyFg HFgCr − HFgEn )
QrLRyFg
QpLRyFg = 100
MrF × HHV
(
QrLRyRs = MrRyRs HRsLv − HRsEn )
QrLRyRs
QpLRyRs = 100
MrF × HHV
(
QrLAc = MrSt36 HW 36 − HW 24 )
QrLAc
QpLAc = 100
MrF × HHV
8. Credits
where
where
MFrWA = specific humidity
MqDA = lb dry air per BTU
HWvEn = enthalpy of water vapor at air heat entering temperature
Credit due to sensible heat in fuel
100
QpBF = HFEn
HHV
where
HFEn = enthalpy of fuel entering
HHV = higher heating value of fuel
where
MFrSc = mass fraction of sulfur capture
MpSF = mass percentage sulfur in fuel
HRSlf = heat of formation for sulfation
HHV = higher heating value of fuel
HStEn − HStLv
QrBX = MrStX EX
100
QrBX
QpBX = 100
MrF × HHV
where
MrStX = mass rate of steam supplied to the equipment
HStEn = enthalpy of steam to the equipment
HStLv = enthalpy of the steam leaving at the exit pressure and the entropy
entering
EX = overall drive efficiency
where
QX = energy input to devices kWh
EX = overall drive efficiency
where
MrSb = mass rate of sorbent entering
Hsb En = enthalpy of sorbent entering
MrF = mass rate of fuel entering
HHV = higher heating value of fuel
Credit due to additional energy supplied by additional moisture Spray water needs to be
accounted for in the overall
QrBWAd = ∑ MrStz ( HStEnz − HW Re ) efficiency calculation.
QrBWAd
QpBX = 100
MrF × HHV
where
MrStz = mass flow rate of additional moisture
HStEnz = enthalpy of additional moisture entering
HW Re = enthalpy of water at reference temperature
MrF = mass rate of fuel entering
HHV = higher heating value of fuel
9. Calculate efficiency.
Mspray = MStmIn
(hstmOut − hstmIn )
(hstmOut − hspray )
13-34 | Clean Combustion Technologies
1. J. Booth and D. Kautzman, “Estimating the Leakage from HP TUIP 10. C. J. Fechheimer, “Measurement of Static Pressure,” Transactions of the
Turbine Sections,” EPRI Power Plant Performance Monitoring Confer- ASME, 48: 965–977, 1926.
ance, 1984, EDRF, Palo Alto, CA.
11. R. G. Folsom, “Review of the Pitot Tube,” Transactions of the ASME, 78:
2. ASME Performance Test Codes, PTC 19.3, “Temperature Measurement.” 1447–1460, Oct. 1956.
New York: American Society of Mechanical Engineers, 2004.
12. C. L. Burton, “Quantitation of Stack Gas Flow,” Reprint from Journal of
3. P. H. Dike, Thermoelectric Thermometry, 3rd ed. Philadelphia: Leeds and the Air Pollution Control Association, 22(8): 631–635, Aug. 1972.
Northrup Co., 1958.
13. ASTM Standards, Part 26, Gaseous Fuels; Coal and Coke; Atmospheric
4. Chromel and alumel are chromium-nickel and aluminum-nickel alloys Analysis. D 2015, “Test for Gross Calorific Value of Solid Fuel by the
respectively. Constantan is a copper alloy. Adiabatic Bomb Calorimeter”; D 2961, “Test for Total Moisture in Coal
Reduced to No. 8 Top Sieve Size (Limited Purpose Method)”; D 3172,
5. Annual Book of ASTM Standards, Section 4, “Magnetic Properties;
“Proximate Analysis of Coal and Coke”; D 3173, “Test for Moisture
Metallic Materials for Thermostatic; Electrical Resistance, Heating, Con-
in the Analysis Sample of Coal and Coke”; D 3174, “Test for Ash in
tacts; Temperature Measurement; Illuminating Standards.” Philadelphia:
the Analysis Sample of Coal and Coke”; D 3177 “Test for Total Sulfur
American Society for Testing and Materials, 1999.
in the Analysis Sample of Coal and Coke”; D 3178, “Test for Carbon
6. “Orifice Metering of Natural Gas.” ANSI/API 2560. AGA Report No. 3. and Hydrogen in the Analysis Sample of Coal and Coke”; D 3180,
New York: American Society of Mechanical Engineers, 1960. “Calculating Coke and Coal Analyses from As-Determined to Different
Bases”; D 3286, “Test for Gross Calorific Value of Solid Fuel by the
7. J. W. Murdock, “Power Test Code Thermometer Wells,” Transactions of Isothermal-Jacket Bomb Calorimeter”; D 3302, “Test for Total Moisture
the ASME. Journal of Engineering for Power, 81, Series A: 403–416, 1959. in Coal”; Philadelphia: American Society for Testing and Materials,
8. P. G. W. Hawksley, et al., Measurement of Solids in Flue Gases, 2nd ed. latest edition 2008.
London: The Institute of Fuel, 1977.
9. E. S. Cole, “Pitot Tube Practice,” Transactions of the ASME, 57: 281–294,
1935.
13-35
Maintaining Availability:
Condition Assessment and
Remaining Life Analysis
Chapter Fourteen
In many respects, the electric utility or independent power producer is unique. Its product,
the day-to-day output of electric energy, is supplied on demand, with no opportunity to
be inventoried or stored. Thus, the network of generation, transmission, and distribution
equipment through which the electricity is delivered must have high operating integrity to
ensure optimum reliability of supply, which is one of the power industry’s highest priorities.
In addition, because of the long lead U.S. Department of Energy (DOE), the Edison Electric Institute (EEI), and
time required to add or contract for new the Electric Power Research Institute (EPRI) to pursue programs that identify
capacity, it is necessary for the industry means of improving plant performance.1 Some state and federal laws concern-
to plan to build generating equipment ing electrical plant reliability already have been enacted. As world demand
years in advance of the required service for electricity continues to increase steadily, the industry places tremendous
dates to supply future electric loads. Fur- importance on power generating plant availability. Fluctuations in fuel costs are
ther, the power industry ranks among felt in virtually all energy sources with a significant impact on political, social,
the most capital intensive, in terms and economic activities.
of dollars of investment per dollar of
revenue. This chapter explores the impli- PRODUCTIVITY INDICES
cations of reliability of supply and plant The utility industry uses several indices to evaluate electric generating unit reli-
availability in the generation of steam ability, availability, and productivity, which, in essence, is a measure of the ability of
electric power by fossil fired plants. a plant to produce electricity on demand. The North American Electric Reliability
Corporation (NERC) defined the following performance indices as the prin-
AVAILABILITY ciple measurement values for the power industry. The performance indices for an
The availability of an electric power plant individual unit are based on the standard definitions by IEEE Power Engineering
is important to both system reliability Society Standards 762-2006.
and generating company profit. Improv-
ing availability only slightly can save Net Capacity Factor (NCF), percentage =
considerably on reserve generating capac-
Net Actual Annual Generation (NAAG) in MWh
ity and costly replacement power. As one ×100
Period Hours (PH) × Net Maximum Capacity (NMC)
measure of reliability, published availabil-
ity statistics are of considerable interest, as
the power industry emphasizes produc- Service Factor (SF), percentage =
ing the most energy for the least cost. Service Hours (SH)
The importance of high plant avail- ×100
Period Hours (PH)
ability and reliability has spurred the
14-
14- | Clean Combustion Technologies
where:
Net Maximum = the capacity that a unit can sustain when not restricted by
Capacity (NMC) ambient conditions.
Available Hours = the period of time during which a unit or major piece of
(AH) equipment is capable of providing service, whether it is actually
in service or not, and regardless of capacity level that is provided.
Period Hours = the number of clock hours that the unit is in the “active state”
(PH) (usually taken as one year). The active state includes both
the available condition (with the unit operating from zero
to full load) and the no-load condition during forced or
scheduled outages. It does not include any period in which a
unit is on inactive reserve, mothballed, or retired
Service Hours = the total number of hours the unit is electrically connected to
(SH) the transmission system and performing generating functions.
Forced Outage = the sum of all hours during which a boiler or other major
Hours (FOH) equipment is unavailable because of a forced outage
Maintaining Availability: Condition Assessment and Remaining Life Analysis | 14-
MEASURING AVAILABILITY
Each equation identifies only a portion of the performance measurement for a
generating unit. No single index tells the overall performance story for a unit, a fact
that is apparent from Figure 14-1. Capacity factor, for instance, is not a true indica-
tor of a unit’s reliability, because actual generation may be limited by economic or
environmental dispatch, as well as forced and scheduled outages and deratings. This
is particularly true with older units, which typically have higher heat rates (lower
thermal efficiencies) than more modern units. The annual availability factor (AF)
establishes only the percentage of time during the year that the unit was capable of
producing power. This factor includes time when the unit was capable of generating
power but was not in service because more efficient units were being used. Thus, the
AF does not measure the ability of a unit to operate at a specific power level when
called on by the dispatcher. Rather, it measures only the capability of the unit to
produce at a power level ranging from
0–100% of its rated capacity. The equiva-
lent availability index provides for an Rated Reserve
Capacity PSO
AF adjustment to account for the effect PFO
of partial deratings (losses in electrical
power output capability) from partial ESD FO SO
forced and scheduled outages. Essentially, Total Production
Power
established to predict life, taking into understand the reasons for this, it is
account safety factors in the Ameri- necessary to consider some of the phi-
can Society of Mechanical Engineers losophy behind their development.
(ASME) Boiler Code as well as First, the Codes are design tools.
realistic, long term, future operating They are also, in a very real sense, some
conditions. In instances where observ- of the earliest expert systems. Their
able distress is incident related (rather purpose is the enforcement of design
than time related), such as with caustic rules that ensure a low probability of
attack, hydrogen damage, or pitting, failure by specifying the minimum
life predictions are not practical. allowable thicknesses of components
The situation in the time dependent made of approved materials fabricated
regime is much different. Because of in a stated manner.
creep damage along grain boundaries, Second, the philosophy of the Codes
components in the time dependent is based on such concepts as:
temperature range can rupture with es-
• Allowed materials are ductile and,
sentially no wall loss. Such components
therefore, in many cases, forgiving.
have a finite life because their strength-
• Life, operational effects, mate-
retaining capability is diminishing with
rial variability, local stresses, and
time. Uncertainties in life continuation
other similar effects are not directly
studies often relate to predicting metal
considered but are expected to be
behavior in the time dependent regime,
accounted for by the conservatism
especially for thick walled, pressure
applied in setting the allowable
containing elements.
material properties.
Design Code Aspects Design rules are developed by con-
Before discussing what is involved in sensus within groups of acknowledged
the remaining life analysis of thick experts and are subject to multiple
walled components, it is important to levels of review. Being developed in this
understand what is not involved. Life way, there may be no precise physical
analyses are not a simple review of the justification for some of the rules.
design vis-à-vis the ASME/American The result of this type of approach
National Standards Institute (ANSI) is that the actual stress in a local region
Codes and a rubber stamp to say of a properly designed component
that the design meets the Code and may exceed the allowable stress given
is, therefore, satisfactory. The fact is in the Code being used. This does not
that, if the unit met the Codes when mean that the component is unsafe,
it was designed, then, unless there are but only that the published values of
such effects as thinning present, it will allowable stresses for the material are
meet the Codes now. The reason for sufficiently conservative to ensure that
this is that the Codes (to which thick the resulting component will be safe,
components are built) do not explicitly even without considering local effects
consider the effects of time, nature of that may occur in design, manufacture,
service, and fluctuations in load. To field construction, or operation.
14-10 | Clean Combustion Technologies
from the user’s point of view it is just this approach, the predicted life may be
as bad to underestimate the remain- underestimated by as much as a factor Figure 14-3
Clean|
Life versus temperature
Combustion Technologiesat
Figure:ASME
14.03/code
BAMallowable stress
ing life as it is to overestimate it. The of 25, purely as a result of the variabil- PMS5415 K
(material: 2¼% Cr, 1% Mo steel)
percentage of deviations from design ity of material properties. Figure 14-3
stress and design temperature required shows the minimum and mean lives of
to halve the life of a typical component a component as a function of operating
are shown in Figure 14-2 as functions temperature assuming the actual stress
of design temperature. is the allowable ASME Code value.
The third variable involved is the The above effects apply even if
material of the component itself. the component is operated purely in
The properties of any single material steady state, that is, with no varia-
specification are extremely variable, tion in thermal, mechanical, and/or
particularly those that affect the life of pressure loadings. In fact, the compo-
a component. For instance, at 1,000°F nent undergoes fluctuations in these
there is a factor of five between the loadings both in a gross fashion during
predicted life of a component obtained start-up and shutdown of the system
14-12 | Clean Combustion Technologies
and, in a minor way, during operation. damage resulting from creep occurs
Although the Code does not require it, throughout the life of a component
the manufacturer of a steam generator and is exhibited as permanent defor-
specifies a start-up rate to limit poten- mation. This permanent deformation
tial damage to the system. However, allows a highly stressed area to shed
this start-up is assumed to be smooth load to a less highly stressed area. Such
and, for stress analysis purposes, linear. a redistribution of load can be signifi-
In fact, the change in loadings during cant. In the case of a cylinder under
start-up can be extremely nonlinear internal pressure, if creep is not present
as various subsystems are modulated. the highest stress will be on the inside
In addition, although the shutdown is surface. If the cylinder is operating at
assumed to be a mirror image of the creep temperatures, however, the high-
start-up, the transients in this situa- est stress area will, after time, be on
tion are often even more severe than the outside surface and that is where
in the start-up. The effect of transient failure will originate.
loadings, particularly in temperature,
can be very severe and result in local Fatigue. Fatigue is the process in
cyclic plastic deformation that can lead which materials fail under cyclic load-
to fatigue failure in a relatively low ing. For thick walled components of
number of cycles.9,10 high pressure steam generators, the
As described earlier, there are two transient thermal effects that occur
principal types of situations that can be during start-up and shutdown produce
considered in the determination of the the principal loading that can cause
life of thick walled components. The damage. These events are severe but,
first, in the time-dependent regime, is fortunately, relatively infrequent. When
the high temperature pseudo-steady- the transient events within a start-up
state loading where the creep of the or shutdown are considered, the total
material is the phenomenon that number of cycles during the life of
leads to failure. The second, in the a unit tends to be below 10,000. If
time-independent regime, is the cyclic failure occurs in fewer than 10,000
plasticity that occurs during start-up cycles, it is termed “low-cycle fatigue.”
and shutdown that can lead to low- It is characterized by local strains in
cycle-fatigue failures. the region of failure that are in excess
of the yield point of the material, with
Creep. Creep is the degradation of ma- the result that the material undergoes
terial properties that occurs with time plastic flow.
at temperature. It happens even at low Whenever fatigue failure occurs,
temperatures and somewhat differently it is initiated at a free surface even
in the heat-affected zone of a weld though this free surface may be at a
than in the base metal. An analysis that subsurface flaw. In low-cycle fatigue,
does not correctly include the effects this initiation is followed by crack
of creep may well predict failure in an growth through a region of previously
incorrect location. This is because the plastically strained material. Finally, the
Maintaining Availability: Condition Assessment and Remaining Life Analysis | 14-13
crack propagates through the rest of the rate of heat transfer, scale conduc-
the component, through regions that tivity, scale growth rate, and the length
originally saw only elastic strains, and of service. Using such data, the past
the component fails. Crack initiation temperature exposure of a tube can be
occurs at a very early time in life (less approximated if the age and thickness
than 5% of the number of cycles to of the oxide scale are known. Esti-
failure). The surface defect is usually mates of this kind, of necessity, assume
located at a persistent slip band where unchanging heat input to the tube, and
material extrusions or intrusions occur constant steam-side conditions.11
as a result of cyclic plastic flow. These
slip bands occur on planes of maxi- Determination of Critical Components
mum principal shear stress although A life continuation study is not solely a
the direction of opening will be that of boiler inspection program, or just non-
the maximum principal normal stress. destructive testing of unit components,
Although slip bands occur adjacent to or only a prediction of remaining life Component assessment is not
the crack, as soon as the crack starts of critical items. Although elements a one-time event. Remaining
to open, the stress on these adjacent of these are present, timing is also
life evaluation is an important
defects is relieved and crack growth is an important issue in the plant life
generally limited to a single crack. This continuation concept. Also, component safety analysis as well.
is different from the creep situation assessment should not be considered
where gross material degradation oc- a one-time event. Life evaluation ef-
curs throughout the region adjacent to forts should be a continuing process,
the main crack. especially on units that are operating
in excess of their original design life.
Metal-Temperature Determination. Instead of inspecting all components
Without reasonably accurate estimates at the same time, a plan should be
of metal temperature, stress-rupture created to continually inspect critical
life predictions, whether based on min- components periodically as the plant
imum or average strength properties, continues to operate. This information
cannot be accurate. There is research should be documented, tracked, and
data on the growth of steam-side oxide obtained once again in a continuing
scales on the low alloy, ferritic steels assessment plan.
used in steam generators.11 At elevated It is not feasible, for economic
temperatures, the internal surfaces of reasons, to evaluate in detail the life of
pressure parts will slowly oxidize and every component in a steam generator;
form an indigenous oxide scale. The nor is this necessary. The first step in
rate of oxidation is a function of the any life continuation study, whatever
internal surface temperature, the type the level of complexity involved, is
of material, and the time at tempera- therefore the determination of what
ture. It is possible to infer a “historical” are the critical components. The initial
metal temperature of a tube sample step in the determination of criticality
by measurement of the oxide scale is an assessment of the consequences
thickness, when there is knowledge of of a failure. If the consequences of the
14-14 | Clean Combustion Technologies
design approach. The difficulty is that, component has many years of remain-
during operation at elevated tempera- ing life, the critical area may have
ture, the stress in the component is totally exhausted its useful life. Despite
redistributed by creep relaxation and these cautions, when applied properly,
as a result the operating stress will this approach does provide the highest
be significantly different from the level of confidence currently possible
design stress. Another aspect of this for a single-outage evaluation.
approach is that, in general, to avoid
creating problems with the compo- OnLine Monitoring and Analysis. The
nent, the material sample has to be ultimate level of assessment is similar
taken from an area removed from to that described previously, except
geometrical discontinuities. Unfortu- that online monitoring, analysis, and
nately, in most cases, the critical areas evaluation can be accomplished by a
are precisely those areas that have to boiler stress and condition analyzer
be avoided during sampling. As a re- described in Chapter 10. Such a de- Proper operation is an
sult, unless miniature samples can be vice monitors temperature, pressure, important factor in achieving
taken from the highly stressed areas, and flow and uses this data together
maximum life expectancy of
it is necessary to translate the results with material and geometry informa-
of the accelerated creep testing into tion to assess damage on a real-time the plant.
the remaining life at the critical areas. basis. It makes no assumptions about
This means that, by the use of plant repeatability of operations. An option
history and analytical techniques, the to the use of this device is to incor-
original creep properties of the mate- porate “what-if ” capabilities such that
rial have to be obtained. These can the effects of changing operational
then be applied to the analysis of the modes can be evaluated.
critical areas.
Finally, it is, unfortunately, a fact Assessment Documentation and
that high pressure components are Tracking. Once condition assess-
fabricated from substructures made up ment steps have been performed and
of different heats of material. Unless summaries provided, all the informa-
the material sample is taken from the tion should be kept in an available
critical substructure, the use of acceler- documentation and planning tool.
ated creep test results is inappropriate. Specialized software programs exist to
In the worst case, if the accelerated keep track of past assessment locations
creep test results indicate that the and results as well as to store recom-
material properties are at the top end mendations for future assessments.
of the scatter band, and the critical area This data tool can provide a living
is in another substructure where the document that can be used for main-
material properties happen to be at the taining continued plant reliability.
bottom end of the scatter band, then
the life estimate will be nonconserva- SUMMARY
tive by a factor of about 25. Although Life continuation of a steam generat-
an analysis might indicate that the ing system implies an objective of
14-18 | Clean Combustion Technologies
1. “Availability Patterns in Fossil-Fired Steam Power Plants,” EPRI Report 8. J. S. Baylor, “Acid Rain Impacts on Utility Plans for Plant Life Exten-
No. FP-583-SR. Palo Alto, CA: Electric Power Research Institute, sion,” Public Utilities Fortnightly, March 1, 1990.
November 1977.
9. J. D. Fishburn et al., “Approaches for the Determination of Remaining
2. Generating Availability Report (GAR), Princeton, NJ: North Life in High Energy Piping Systems,” Proceedings of the American
American Electric Reliability Council. Data compiled following IEEE Power Conference, Chicago: Illinois Institute of Technology, 1988,
Standard 762 – Revised 2006, (Revision of IEEE Std 762-1987. IEEE vol. 50.
Standard Definitions for Use in Reporting Electric Generating Unit
10. T. McColloch, J. D. Fishburn, G. E. Roberts, and G. Hunter, “Evaluating
Reliability, Availability, and Productivity, 15 March 2007.) Tabulated
the Structural Integrity of High Energy Piping Systems on Fossil Boil-
results calculated from the Annual Unit Statistics for Years 1990–2007,
ers,” presented at the Second EPRI Fossil Plant Inspections Conference,
Periods 01–12.
San Antonio, TX, November, 1988.
3. S. S. Blackburn and D. E. Lyons, “Design for Availability— An Update,”
11. R. P. Aubrey, B. A. Hawkins, and T. D. Jamison, “The Use of Oxide
Proceedings of the American Power Conference, Chicago: Illinois Insti-
Scale Thickness Measurements in Life Extension Analysis,” presented at
tute of Technology, 1975, vol. 39, pp. 349–368.
ASNT 1989 Spring Conference, Charlotte, NC, March 22, 1989.
4. C. H. Gilkey and J. D. Rogers, “A Summary of Experiences with Fan-
12. J. D. Fishburn and R. W. Loomis, “Life Extension of Thick Walled
Induced Duct Vibrations on Fossil-Fueled Boilers,” Proceedings of the
Components,” presented at 11th International Conference of the
American Power Conference, Chicago: Illinois Institute of Technology,
AMIME, Irapuato, Mexico, November 2–4, 1988.
1975, vol. 37, pp. 728–734.
13. F. V. Ellis, R. W. Loomis, and S. Tordonato, “Life Extension: The C-E
5. “Extending the Lifespan of Fossil Plants,” EPRI Journal, June 1983.
Approach to the Analysis of Thick Walled Components,” presented at
6. “Generic Guidelines for the Life Extension of Fossil Fuel Power Plants,” Conference on Life Extension and Assessment of Fossil Plants, EPRI,
EPRI Report No. CS-4778, Palo Alto, CA: Electric Power Research Washington, DC, June 1986.
Institute, November, 1986.
14. B. W. Roberts, F. V. Ellis, and R. Viswanathan, “Utility Survey and
7. D. E. Gelbar and S. J. deMello, “Assessing Boiler Life Continuation Inspection for Life Assessment of Elevated Temperature Headers,” Pro-
Needs,” Proceedings of the 1988 Joint ASME/IEEE Power Generation ceedings of the American Power Conference, Chicago: Illinois Institute
Conference, Philadelphia, PA, September 25–29, 1988, ASME Paper of Technology, 1985, vol. 47, pp. 259–301.
No. 88-JPGC-Pwr-37.
Suggested reading
14-19
Alternative Fuels
Chapter Fifteen
Whether because of the increased costs of conventional fossil fuels, environmental constraints,
or the recovery of valuable chemicals, many industrial and utility boilers use alternative
fuels, either to supplement their existing fossil fuel sources or to burn as a stand-alone fuel.
Alternative fuels can be defined as energy-containing materials that are not expressly produced
for energy production. Rather, they are the by-products and wastes from a variety of industries
and processes from all parts of the modern economy. They are the unconventional sources
of energy that can minimize the use of of industries and processes, they have biomass fuels, municipal waste sources,
premium fuels. Industries that produce a wide range of physical and chemical and fossil-fuel derived products. The
large quantities of these alternative fuels properties. As a result of the exten- types of boilers and combustion sys-
include agriculture, mining, forestry, sive number and diversity of these tems used vary with the characteristics
petroleum refining, primary metals pro- fuels, a detailed discussion regarding of type of solid alternative fuel.
duction, chemical manufacturing, pulp the characteristics and associated
and paper production, waste treatment, combustion issues of each is not Biomass
and construction, to name a few. practical. Therefore, a select number The most common category of
Alternative fuels can essentially be of alternative fuels will be discussed, alternative fuel is biomass. Biomass
broken into three distinct categories: along with general comments regarding encompasses all types of plant mate-
solid, liquid, or gaseous. Table 15-1 combustion issues. rial, paper, and wood products as well
shows a number of by-products and as bio-derived fuels, such as munici-
wastes commonly used as alternative SOLID FUELS pal solid waste, sewage sludge, and
fuels in boilers. Because these alterna- Solid alternative fuels can be further animal waste. Figure 15-1 shows some
tive fuels originate from a wide range divided into several subcategories: examples of biomass fuels.
Table 15-1 | Examples of alternative fuels Most biomass fuels are cellulose-
Physical Form
based, which makes them physically
Non-Fossil Fuel/Material Burned and chemically very different from
Gaseous Liquid Solid
Agricultural residues X other solid fossil fuels, such as coal.
Agricultural waste–cotton and corn X
Alcohol (ethanol) X Compared to coal, most biomass fuels
Alcohol solvent X
Almond shells X contain a higher proportion of hydro-
Almond tree prunings X gen and oxygen and typically are low in
Animal fats X X
Ash pile char X ash, high in volatile matter, and highly
Bagasse X
Balsa wood X variable in moisture content. Specific
Bark X constituents within the biomass fuel,
Barley dust and chaff X
Barley needles X especially nitrogen, sulfur, and chlorine,
Bio-liquids X
Bio-gas X give rise to variations in combustion
Biomass X and emission behavior unless con-
Bitumen X X
Bituminous coke X sidered in the design and operation
Black liquor X
Blast furnace gas X of the boiler. The lower sulfur and
Blond fiber X nitrogen contents of most biomass
Cardboard X
Carpet scrap X fuels, however, have a positive implica-
Clarifier residuals, primary X
Clarifier residuals, secondary X tion for gaseous emissions. Some types
Coal refuse, anthracite culm X of biomass contain relatively high
Coal refuse, bituminous gob X
Coal tar oil X chlorine levels, which result in high
Coal tar plus soil X
Coating residues X HCl concentrations in the boiler flue
Coke oven gas X gases and, possibly, in high temperature
Composite water X
Compressed paper X corrosion.
Conforming waste materials X
Construction/demolition wood/debris X Biomass ash contains a number of
Corn fiber X inorganic minerals (SiO2, CaO, MgO,
Corn gluten feed X
Corn mill dryer off-gas X K, Na, Fe2O3) and trace elements of
Corn starch residue X
Corn stover X heavy metals (Hg, As, Pb). The high
Cotton gin residues X proportion of alkali metals within the
Cotton stalks X
Cottonseed hulls X ash, such as K and Na, is a good indi-
Cow manure X
Crankcase oil X cator for potential slagging and fouling
Deinking residuals X X problems within a boiler. Another
Dewatered combustible residues X X
Diaper scraps X major mechanism for ash deposition is
Digester gas X
Distillers dry grains X dependent on the ash fusion tempera-
Distillers grains X ture of the particular biomass being
Distillers grains stillage X
Engineered wood plant scrap X burned and on the combustion condi-
Envirofuel Pellets X
Filters X tions. Table 15-2 shows some typical
Foam residues X ash compositions of biomass fuels.
Fruit pits X
Fuel cubes (paper diaper clippings/refuse) X Raw biomass fuels, such as energy
Fuel oil solids (tank clean-out residue) X
Fume filter oil X crops (switchgrass, miscanthus, and
Gear oil (virgin and used) X short-rotation crops), can contain
Gluten feed pellets X
Grain dust X up to 60% moisture. They also have
Glycerol distillation by-product X
Hardboard dust X significantly different properties with
Hardboard residues X respect to storage, bulk handling,
Hardwood pellets (hammer-milled, dried, pelletized; sold as fuel) X
Heavy recycle X volume flow, milling, combustion,
High caustic fuel X
Hog fuel, hard wood X slagging, corrosion, and gaseous emis-
Hog fuel, soft wood X sions. Biomass fuels can be delivered
Hog fuel, urban X
Hogged bark X to the site in a number of forms, rang-
Hydro-pulper refuse X
Industrial plastics X ing from pelletized fuels to chipped
Ink solvents X fuels and bales. The issues regarding
Alternative Fuels | 15-
delivery, storage, and preparation of Table 15-1 | Examples of alternative fuels —continued
the biomass fuels differ from those Physical Form
with other solid fuels, and this pres- Non-Fossil Fuel/Material Burned
Gaseous Liquid Solid
ents the first obstacle to overcome Knots and knotter rejects X
Laminate production scrap dust X
regarding the continuous, long-term Laminated wood X
Lamination finishing plant residues X
use of biomass. Biomass pellets typi- Lamination plant residues X
cally will have lower moisture contents Landfill gas X
Latex paint water X
(and, thus, higher calorific values) Lightweight asphalt X
Lignin X
as well as increased bulk densities Liquefied chicken fat X
compared to raw biomass, and they Livestock manure, bedding and yard residues X X
Log yard cleanup X
generally are easier to handle. Low carbonate corn fiber X
Lube oil (virgin and used) X
Biomass tends to be more reactive Lumber refuse (hogged pallet boards/other non-treated scrap) X
than other solid fuels, but in almost Malt sprouts X
Manure X
all cases, it requires milling in order Mechanical pulp mill rejects X
Medium density fiber board sander dust X
to achieve a size that is suitable for Medium density fiber board trim (some with water-based primer) X
transportation and complete burnout Mill trash X
Millfeed X
within the furnace. Because the cost of Mixed liquid residues X
Mixed wood residues X
milling biomass is high, knowledge of Municipal sewage residues X
the maximum acceptable particle size Neutralene X
New and used solvent, Acetone X
that will still burn is of great impor- Non-condensable gas (includes stripper offgas) X
Non-halogenated solvent X
tance. This size will depend on the Non-hazardous by-product solvent X
specific biomass being considered, be- Nut shells X
Oat hulls X
cause the milling behavior and ultimate Oil booms X
Oil residues/excesses/by-products (no. 6, asphalts) X X
shape will vary among the biomass Oilseed (rapeseed, canola) X
fuels. In addition, the furnace in which Oil spill cleanup residues X X
Oilseeds (canola, corn, beans) X
the biomass is burned, and how that Oily rags X
Old corrugated container rejects X
furnace is being operated, also will Olive cake X
affect the aerodynamic and combustion Olive kernel X
Orchard prunings X
effectiveness. The ignition process of Oriented strand board trim/dry residuals X
Orimulsion X
biomass is similar to that of coal except Other biomass X
that more volatile matter is available. Paint rags X
Paint residues X
In some biomass fuels, however, high Palm oil plantation by-products (palm kernel exfoliate, fiber, shell,
X
empty fruit bunch [EFB])
moisture and ash contents can cause Paper X
ignition and combustion issues. Paper broke X
Paper cores X
Firing of biomass fuels in large Paper fines X
Paper, office waste X
power boilers can be achieved by either Paper wrapper and packaging
X
co-firing with the conventional fossil trimmings
Particleboard sander dust X
fuel or by firing as the primary fuel Particleboard trim X
Paunch manure X
source in suspension, on a stoker, or Peat X
in a fluidized bed combustor. With Pecan hulls X
Petroleum coke X
co-firing, as the proportion of biomass Petroleum distillation solvent X
Petroleum-refining residue oil X
increases, the overall average fuel analy- Pine and cedar trees, chipping and residues X
sis to the furnace varies according to Pine and fir X
Pine tar X
the proportion fired. In general, the av- Pitch/woodwaste X X
Plywood trim X
erage calorific value and amount of ash Post-harvest biomass residues X
that is produced generally will decrease Poultry litter X
Process-derived liquid fuel X
as the quantity of biomass increases. Process gas X
Pulp liquor X
In addition, the lower the quality of Pulp mill gas X
the biomass fired (i.e., biomass with a Railroad ties X
15- | Clean Combustion Technologies
Table 15-1 | Examples of alternative fuels —continued higher moisture content), the lower the
Physical Form overall boiler efficiency.
Non-Fossil Fuel/Material Burned As the boiler efficiency decreases
Gaseous Liquid Solid
Railroad ties, chipped X
Reclaim ink solvent X
with lower quality biomass fuels, the
Recovered gaseous butane X flue gas flows will increase and lead
Rectified methanol X
Red oil (steam stripper, steam condensate, including terpenes, terpe-
X
to an increase in gas loading on the
noids, methanol, total reduced sulfur [TRS])
Redwood and fir X
induced-draft fans and back end equip-
Refinery gas X ment. For units that were not originally
Refuse X
Refuse-derived fuel X designed to burn higher-moisture-
Reinjection char X
Reprocessed oil X
content biomass and that have little or
Resin solid X no fan margins, load reductions can
Restaurant oils and greases X
Rice hulls X be expected. In addition, the desuper-
Sander dust X
Sawdust X heater spray flow, air heater gas outlet
Sawmill scrap X temperature, main steam to reheater
Scrap x-ray film X
Screen rejects X outlet steam temperature differential,
Sewage gas X
Shredded cloth X and reheater outlet steam temperature
Silvicultural wood X can be affected.
Sludge X X
Sludge without water X Corrosion can be another major
Solid paraffin X
Solvents X issue in co-firing applications. Ele-
Soybean hulls X ments of concern include alkalies (Na
Spent coffee grounds X
Spent oxide X and K), chlorides, sulfur, lead, and
Straw X
Stripper condensate X zinc. High concentrations of alkalies
Sulfur-free organic by-product X
Sunflower hulls X
lead to low-melting-temperature ash
Sunflower husks X components. Molten ash compounds
Sunflower pellets X
Sunflower seed hulls X tend to be more aggressive in attacking
Sunwax (diatomaceous earth with sunflower oil wax) X
Switchgrass X
metals because of the high ion mobility
Synfuel coal X in the liquid phase. Low-melting ash
Tall oil and tall oil derivatives X
Tallow X X components also lead to sticky deposits
Tar X X
Tire chips X
that attract other ash material, leading
Tire-derived fuel X to a high buildup of deposits on tube
Tires (whole) X
Toluene X surfaces. The stickiness makes the
Turkey brood woodwaste X
Turpentine X
deposit hard to clean. Potassium com-
Used hydraulic oil X pounds tend to be more corrosive than
Used motor oil X
Used oil X sodium compounds. Figure 15-2 shows
Used thermal oil X
Vegetable oil X
a deposit from the co-firing of wheat
Virgin hydraulic oil X straw and coal.
Waste alcohol X
Wastewater treatment residues X X Chlorides can combine with
Wax and cellophane wrapper and packaging trimmings X
Wet distiller grains X
alkalies to form compounds with
Wheat fiber X very low melting temperatures. These
Whole-tree chips/harvest residue X
Wood X compounds can cause very rapid cor-
Wood bark X
Wood briquettes X
rosion of boiler tubing. Lead and zinc
Wood char X can be contributing factors. These
Wood chips X
Wood chips, fines X materials can be absorbed from soils
Wood chips, old X
Wood fuels, bark X by the biomass during the growing
Wood pallets X process depending on the location,
Wood shavings X
Woodwaste, clean X types of soils, and types of fertilizers
Woodex pellets X
Woodworking residuals (cabinet manufacturing) X being used. Figure 15-3 illustrates
Yellow grease (used cooking oils; sold as commercial fuel) X some of the eutectic mixtures of
Alternative Fuels | 15-
chloride compounds and their melt- Table 15-2 | Typical ash compositions of biomass fuels
ing points.
Illinois Cordero Wheat Hybrid Sunflower Wood
In most cases, burning biomass Alfalfa
No.6 Rojo Straw Poplar Hulls Waste
produces lower emissions compared
SiO2 54.7 30.0 66.5 30.4 4.7 4.8 28.4
with other, conventional fossil fuels.
Al203 20.9 22.1 0.0 0.5 1.5 1.6 11.3
Therefore, many larger power boil-
ers look to offset some or even all of Fe203 14.6 7.4 0.4 0.7 2.0 0.7 10.3
their fossil fuel with biomass. Biomass TiO2 0.9 1.9 0.1 0.1 0.1 0.1 0.5
can reduce the amount of greenhouse P2O5 0.2 1.2 1.4 8.8 7.9 8.0 1.4
gases, particularly CO2, being gener- CaO 3.5 29.9 3.5 25.4 49.6 12.8 29.2
ated from the combustion of fossil MgO 1.6 5.5 1.8 4.4 8.6 9.9 6.1
fuels. As part of the process called Na2O 1.3 1.7 2.8 1.5 0.4 0.0 5.0
photosynthesis, living plants utilize
K2O 2.2 0.4 23.5 28.2 25.2 62.1 7.8
CO2 and water in the presence of light
CI (ppm) 340 57 2,120 2,340 93 588 649
energy to form glucose and oxygen.
And although CO2 is released into the
atmosphere during the combustion be controlled by near-field aerody-
of biomass, it also is taken from the namic changes, whereas the remaining
atmosphere during growth. Therefore, nitrogen in the char phase is released
biomass fuels often are considered to at a slower rate and is influenced by
be carbon neutral. the furnace environment. In some 0 5 mm
The lower sulfur contents of most cases, firing certain biomass materials
biomass fuels will result in lower SO2 may produce NOx emissions that are Figure 15-2 | Ash deposit of wheat straw and
coal blends
emissions compared with most coal- or similar to, if not higher than, those of
oil-fired boilers. If the biomass fuel some coal- or oil-fired units.
contains a high level of chlorine, this The unburned carbon from certain
can form HCl in the flue gas, which biomass-fired systems, such as older Clean Combustion Technologies
will impair the operation of the flue gas suspension-fired units or stoker-fired Figure: 15.02/JG
desulfurization system. An increased units, can be relatively high. This PMS5415 K
presence of heavy metals, such as results from the low-density and high-
mercury, arsenic, or lead, in the flue gas drag-coefficient char particles that
could be concentrated in the flue gas often are associated with biomass.
desulfurization residues.
The situation regarding NOx emis-
sions when firing biomass is more
complex than the situation regarding 600
SO2 emissions. Unlike the situation
with sulfur, which directly forms SO2, 500
Temperature, °C
technology for more than a century process to assist in the commercial uti-
to recover chemicals that are used in lization of these fuels is emulsification.
the pulping process. Chemical recov- A water–oil emulsion can be prepared
ery boilers burn black liquor (a slurry that allows the material to be pumped.
of lignin, unused pulp, and cooking The goal is to be able to burn the fuel
chemicals) to recover sodium and sulfur in a manner similar to oil. These fuels
compounds for reuse. Bark-fired boilers tend to be high in sulfur, so appro-
provide steam for processing. Deinking priate emission controls would be
sludge from the recycling of newspaper required for proper utilization. Several
provides fuel for some of these boilers. large utility units have been retrofitted
Heavy oils, greases, fats, and solvents to use this fuel.
also are hydrocarbons of a nature
similar to that of conventional oils. Alcohol Fuels
Typically, however, these hydrocar- Alcohol fuels can be produced from
bons are of higher molecular weight plant material (primarily ethanol) or
and heavier in nature. They may be other hydrocarbon materials via gasifi-
solid at room temperature and need cation (primarily methanol). Current
to be heated to remain as a pumpable production of ethanol is from corn
liquid. Atomization may be a problem. or sugar manufacture, but substantial
High asphaltene contents can lead research is underway regarding the
to soot formation and relatively high production of cellulosic ethanol. Most
unburned carbon levels. Because some plants (i.e., biomass) consist of ap-
of these materials may have been used proximately 25% lignin, 45% cellulose,
for other purposes, they often contain and 30% hemicellulose. The lignin is
other, non-combustible materials that not suitable for conversion to oils or
must be accounted for in the combus- alcohols; however, the cellulose and
tion and handling processes. Solvents hemicellulose can be converted to sug-
in particular tend to have an adverse ars, and then to alcohols, fairly easily.
impact on seals and other material- The life-cycle energy balance for this
handling systems. biomass fuel is favorable compared to
The majority of these fuels represent the conversion of corn to ethanol.
only a small amount of the total fuels Full utilization of biomass materi-
burned. Some fuels, however, either als that are currently left behind could
exist in potentially large quantities or provide as much as 40% of the current
can potentially be produced in large oil requirement of the United States.
quantities. Two such fuels are very Current programs include genetic engi-
heavy oils (e.g., orimulsion) and alcohol neering of microorganisms that convert
fuels (e.g., ethanol and methanol). the sugar to ethanol. Additional work
on improving the utilization of incident
Very Heavy Oils sunlight by photosynthesis also would
Very heavy oils are found in the Orinoco contribute to the potential of this pro-
fields well inland in Venezuela. The cost cess. Although alcohol fuels are similar
of producing, processing, and trans- to solvents, their current utilization in
porting these fuels is rather high. One gasoline provides some experience in the
Alternative Fuels | 15-11
types of materials that would be needed design features. Once the fuel has been
for wide-scale application. burned in the furnace, the resulting flue
gas is comparable to more conventional
GASEOUS FUELS gas firing. One area of concern would
Gaseous fuels include refinery gases, be with those gases having a high
blast furnace gas, coke oven gas, non- hydrogen content. Hydrogen combus-
condensable gases, and landfill gas. tion tends to be associated with higher
These gases currently are utilized as NOx production. For the more dilute
long as the heating value of the fuel can gases, the temperatures usually are
be maintained above approximately low enough that this is not a problem.
60 BTU/ft3 (2,250 kJ/m3). Gases For gases with some concentration of
with lower or more variable heat con- methane or ethane, the heating value
tents usually are thermally oxidized, tends to be higher, and measures need
either by flare or thermal oxidizer. to be taken to ensure that NOx regula-
Design considerations are con- tions can be met.
cerned primarily with the fuel delivery
and burner systems. The low heating
Doug Hart
content of the gases generally requires William Bailey
larger fuel piping and special burner Carl BOzzuto
Coals of the World
Appendix A
This appendix consists of three sections. The source of this statistical data is WORLD COAL RESERVES,
The first has reference data on the BP Statistical Review of World Energy PRODUCTION, AND CONSUMPTION
reserves, production, and consumption of 2007.1 The sources of analytical data Figure A-1 shows the total recoverable
the coals from producing and consuming in this chapter are fuel investigations coal reserves and changes in production
countries of the world. The approximate by Alstom’s Power Plant Laborato- and consumption patterns since 1995.
reserves are stated at the end of 2006, ries (PPL); publications of the U.S. The estimates of economically recover-
while production and consumption Bureau of Mines and the Canadian able coal are at best approximate, and
information shows the changes over the Department of Energy, Mines and Re- depend just as much on economic and
last ten 10 years ending in 2005. sources; the National Coal Association, political considerations as they do on
The second section contains analyses Washington, D.C.; the World Energy exploratory techniques and technical
of all the significant world coals and Conference Surveys of Energy Resourc- capabilities. There are no universally
their ash constituents. This data is to be es; the Report of the Commission of agreed upon definitions of reserves and
used only in the preliminary design of Inquiry into the Coal Resources of the resources, many of them being a func-
boilers to burn coals from these coun- Republic of South Africa (1975); and tion of quantity of overburden or depth
tries. The data are not intended nor are COAL—Bridge to the Future, sum- of mining. And, depending upon the
adequate for the actual design of a steam mary report of the World Coal Study. source of data and the purpose for which
generator to burn a specific local fuel. * * * * it was gathered, the values may represent
Mine or core samples, or detailed fuel In this Appendix A, gross calorific undue optimism or conservatism.
analyses, are needed for that purpose. values are given in both BTU per Figure A-2 shows the changes in coal
The international systems of coal pound and megajoules per kilogram. production and consumption patterns
classification are described in the third Tonnages of coal resources, coal pro- since 1995. Coal is the fastest growing
section. These systems result from duction, and consumption are given fossil fuel with the Asia/Pacific region
a significant effort to categorize in a in millions of metric tonnes (2,204.6 accounting for most of its growth
consistent, rational manner all the dif- pounds) of oil equivalent (1TOE = (Tables A-1 and A-2).
ferent coals found worldwide. This is 42 GJ). A coal equivalent is a metric
vitally important where engineers from tonne of coal with a heating value ANALYSES BY CONTINENT
one country are involved in the design of 12,600 BTU/lb, 7,000 kcal/kg, AND COUNTRY
of fuel preparation and burning equip- or 29.31 MJ/kg. Because coals vary Coal is found on all continents and in
ment, steam-generating equipment, significantly in calorific value, much most countries throughout the world.
or ash-handling systems for another more than 1 tonne of sub-bituminous, However, some countries, such as Italy
country where coals may be designated lignitic, or brown coal is required to and Sweden, have resources so meager
by local custom rather than a useful produce the energy of 1 tonne of coal in proportion to their energy demands
engineering basis. equivalent (1 tce). as to be inconsequential. By far, the
A-
A- | Clean Combustion Technologies
1,400
1,200 analyses. Table A-3 is such a correlation,
1,000 based on a study of analyses of many
800 typical coals.
600
Clean Combustion Technologies COAL DEPOSITS OF
400 Figure: App A-01/ (AR)
200 PMS5415 K NORTH AMERICA
This section describes the coalfields of
North South and Europe and Middle East Asia
America Central America Eurasia and Africa Pacific Canada, the United States, and Mexico.
Figure A-2 | Worldwide Coal Production and Consumption1 Canada (Fig. A-3) has more than
6,500 million tonnes oil equivalent
largest world coal deposits are located it easy to ascertain coal rank. Vola- (MTOE) of economically recoverable
in the Northern hemisphere. tile matter (VM) contents and high coal reserves, ranging from anthracite
Clean Combustion Technologies
heating
Figure: App A-02/ (AR) values are normalized to a to lignite.
BASIS FOR ANALYTICAL DATAPMS5415 K moisture- and ash-free basis; ash and Ninety-five percent of the recover-
A most important consideration in the moisture contents are given on an as- able reserves are in western Canada
use of coal analyses is the determina- analyzed basis. (British Columbia, Alberta, and Sas-
tion of coal type or rank from the data Ash analyses, some of them available katchewan), of which about 80% can
presented. The selection of pulverizing from U.S. Bureau of Mines and Ca- be open-pit mined. As in the United
equipment, for instance, depends upon nadian Department of Energy, Mines States, coals in western Canada are low
characteristics related to coal rank, and Resources publications, but mostly in sulfur content, rarely exceeding 0.6%
as does establishing the heat transfer done in Alstom’s laboratory, are given on a mass basis, but are high in ash.
requirement and size of an air heater. for use in furnace sizing, determining Eastern Canadian coals generally have
For the best use of world coal data, convection-pass fouling potential, and high sulfur, particularly those from
the tabulations in this Appendix make designing ash-transport and storage Nova Scotia.
Coals of the World | A-
Approximately 14% of the coal deposits are located in four (in the Crowsnest Pass area) and
Canadian coal produced is bitumi- definable regions. into northeastern BC (Figure
nous coal from underground mines; A-3). The main center of deposi-
the balance is bituminous, sub- 1. The Mountain Region: tion is located in southeastern
bituminous, and lignitic coals taken Most of the bituminous coal BC. Large resources are also
by strip-mining. More than 70% resources are in the provinces of being found in the Peace River
of the domestic consumption is Alberta and British Columbia area to the north.
for power generation; most of the (BC), extending in a belt up to 2. The Foothills Region of Alberta:
balance is for metallurgical uses. Pro- 31 miles (50 km) wide from the Coals from this area are mainly
duction of coal in the year 2006 was U.S. border, northwest into the of high-volatile bituminous B and
measured to be about 32 MTOE per East Kootenay region of south- C rank. Mining conditions are
year, down 5% from 2005. Canada’s eastern BC, north into Alberta not as severe as in the mountain
A- | Clean Combustion Technologies
Volatile Matter
15 to 24 25 to 36 34 to 46 38 to 49 42 to 54 41 to 46 42 to 48 43 to 50 45 and higher
(VM) MAF
C +H
O2 19.5 11.8 11.2 7.0 5.4 4.3 3.2 2.8 2.7 and lower
(Typical)
15,100 to 15,000 to 14,600 to 14,000 to 13,400 to 12,900 to 11,900 to 11,200 to 10,200 to
BTU/lb, MAF
16,000 15,700 15,400 14,800 14,200 13,800 13,300 12,900 12,700
MJ/kg, MAF 35 to 37 35 to 36.5 34 to 36 32.5 to 34.5 31 to 33 30 to 32 27.5 to 31 26 to 30 23.6 to 29.5
region, but eventually most of the 155 miles (250 km) farther east, facing the Cabot Strait and the
resources will have to be extracted and form an immense arc from Atlantic Ocean. The coal is most-
by underground methods. the U.S. border up to about ly of high-volatile bituminous A
3. The Plains Regions of Saskatch- 56°N latitude, passing east of the rank, varying from medium to
ewan and Alberta: cities of Calgary and Edmonton. high in sulfur content.
The four lignite deposits of 4. The Atlantic Region:
Saskatchewan are located along The Cape Breton Island coal- Canadian reserves of peat, which
the U.S. border and form the field in Nova Scotia is the have not been surveyed, are thought to
northern fringes of the main most important of the region, be approximately equal to those of the
lignite zone centered in North although a small quantity of United States.
Dakota. The sub-bituminous high-sulfur bituminous coal is Table A-4 gives the analyses of typi-
seams of Alberta underlie the produced by surface mining in cal Canadian coals. For the ultimate
southern part of that province the province of New Brunswick. analyses and ash constituents, see
from the Rocky Mountains in The Cape Breton field is located Analysis Directories of Canadian
the west to their outcrops about on the north shore of the island, Coals published by the Canadian
Department of Mines and Technical
Surveys, Mines Branch, Fuels Division.
United States
Arctic Circle
Coal is found in 36 of the states and is
currently mined in about 26. Six states
contain more than 75% of the estimat-
60 °N ed 247,000 MTOE of economically
recoverable reserves: Montana (28%),
Illinois (16%), Wyoming (13%), West
Virginia (9%), Pennsylvania (7%) and
Kentucky (6%), Figure A-4.
The reserves of peat, not detailed
High-Volatile Bituminous in this text, are estimated to be equal
Sub-Bituminous in heating value to about 30% of the
Lignite and Brown Coal
Bituminous and Semianthracite currently recoverable coal reserves.
Peat is located in low-lying coastal ar-
Figure A-3 | Coal Map of Canada
Coals of the World | A-
eas, in the flatlands of the Great Lakes, The U.S. Geological Survey has Plains, (5) Rocky Mountain, and (6)
and in Alaska. divided the coal-bearing areas of the Pacific Coast. The provinces are sub-
The electric utilities of the United contiguous U.S. into six main prov- divided into coal regions, coalfields,
States consume 78% of all the coal inces designated as (1) Eastern, (2) and coal districts. (Alaska is treated as
produced in the states. Interior, (3) Gulf, (4) Northern Great a separate province.)
Nova Scotia
Thorburn Hvab 14.4 3.0 33.4 0.5 15,180 35.3 2,440 60
Broughton Hvab 14.4 5.2 40.4 7.6 14,560 33.9 2,100 59
River Hebert Hvab 19.3 1.5 41.0 6.8 14,530 33.8 2,070 61
Phalen Hvab 14.4 1.7 34.9 6.6 15,200 35.3 2,180 84
Harbour Hvab 9.3 1.2 40.6 2.7 15,420 35.9 2,480 70
Westville Hvab 15.4 2.8 32.9 0.9 15,020 34.9 2,240 64
Sydney Hvab 10.2 3.8 42.7 6.4 14,660 34.1 2,130 63
Joggins Hvab 15.8 3.7 42.4 6.3 14,460 33.6 2,140 60
Inverness Hvcb 14.2 9.3 44.1 8.0 13,530 31.5 2,070 60
New Brunswick
Minto Hvab 14.0 3.4 39.0 8.4 15,150 35.2 1,980 66
Chipman Hvab 17.6 2.1 42.2 9.3 14,970 34.8 1,920 63
Coal Creek Hvab 18.3 1.8 41.7 10.3 14,780 34.4 1,960 57
Saskatchewan
Bienfait Lig A 6.0 34.1 49.9 0.8 12,390 28.8 2,300 63
Roche Percée Lig A 6.3 30.8 46.6 1.0 12,450 28.9 2,250 52
Alberta
Cascade Sa 8.5 3.1 13.9 0.8 15,580 36.2 2,750+ 88
Crowsnest Mvb 10.4 4.6 27.5 0.6 15,200 35.3 2,750+ 77
Coalspur Hvcb 10.5 6.3 41.5 0.1 13,680 31.8 2,170 52
Lethbridge Hvcb 10.5 10.2 45.4 0.8 13,590 31.6 2,330 40
Pembina Subc 8.8 19.6 41.3 0.3 12,440 28.9 2,450 49
Drumhelher Subb 9.1 17.2 42.2 0.7 13,150 30.6 2,270 33
Castor Subb 5.5 24.8 43.6 0.6 12,470 29.0 2,140 45
Edmonton Subc 13.7 23.5 44.3 0.3 12,870 29.9 2,370 38
Taber Suba 9.9 14.2 44.3 1.8 13,350 31.0 2,350 47
Camrose Subc 5.8 21.4 44.5 0.4 12,710 29.5 2,160 31
British Columbia
East Kootenay Mvb 9.5 2.4 25.5 0.5 15,350 35.7 2,740 88
Vancouver Island Hvab 12.9 1.4 38.6 1.6 14,910 34.7 2,320 57
A- | Clean Combustion Technologies
OR VT
those of North Dakota, having moisture
WI
SD
MI NH
MA
to 36% and sodium levels of 5 to 8%.
WY NY
CT RI
CA PA
NV IA
NE IL
UT IN OH NJ THE NORTHERN GREAT PLAINS PROVINCE.
MD DE
CO
KS
WV
VA
This includes all the coal fields of the
KY
NC
great plains east of the eastern front
AZ
NM
TX
OK
MO
AR TN
SC range of the Rocky Mountains. In this
AL GA
province are immense lignite areas of
LA the two Dakotas, and the bituminous
MS FL
and sub-bituminous fields of northern
Bituminous
Sub-bituminous Wyoming and northern and eastern
Lignite Montana. Lignites from this province
Anthracite and Semianthracite
can be of very low rank, having moisture
content to 45% and sodium oxide as
Figure A-4 | Coal Map of United States
high as 15% of the ash.
Clean Combustion Technologies
Figure: App A-04/ (AR)
THE EASTERN PROVINCE. This includes
PMS5415 K Western region embracing the coal THE ROCKY MOUNTAIN PROVINCE. This
the anthracite regions of Pennsylvania fields of Iowa, Missouri, Nebraska, comprises the coalfields of the
and Rhode Island; the Atlantic Coast Kansas, Arkansas, Oklahoma, and the mountainous districts of Montana,
region, including the Triassic fields southwestern region of Texas. With Wyoming, Utah, Colorado, and
of Virginia and North Carolina; and some exceptions, the bituminous coals New Mexico. It possesses a great
the Appalachian region extending of this province are of lower rank and variety of coals, ranging from lignite
from Pennsylvania through eastern higher sulfur content than those of through sub-bituminous and high
Ohio, Kentucky, West Virginia, the Eastern province. Much of the grade bituminous coals to anthracite.
western Virginia, and Tennessee into surface-mineable coal in the Illinois The sub-bituminous coals range
Alabama. The Appalachian basin basin having low chlorine and sodium in moisture from 10 to 30%, with
contains the largest deposit of the content has been mined or committed calorific-values (higher heating values
high-grade bituminous and semi-bi- to be mined. Much of the remain- [HHVs]) from 6,500 to 10,000
tuminous coals. The Triassic deposits ing coal, with high chlorine content BTU/lb (15.1 to 23.3 MJ/kg). Sev-
of central Virginia and North Caro- (above 0.15% as fired), will have to be eral sources in this province have high
lina occur in irregular pockets rather deep-mined and will comprise a large sodium-oxide in ash, ranging to 15%.
than in seams and are at present of proportion of the future production.
little commercial value, because of ex- THE PACIFIC COAST PROVINCE. This area is
pensive mining and rather low-grade THE GULF PROVINCE. This consists of the largely confined to the State of Wash-
quality of coal. Mississippi region in the east and the ington, which contains the largest coal
Texas region in the west. The Missis- fields on the Pacific Coast. It also em-
THE INTERIOR PROVINCE. This province sippi region includes the lignite fields of braces the small fields of Oregon and
includes all the bituminous coal area Alabama, Mississippi, and Louisiana. California. The coals in this province
of the Mississippi Valley region and The Texas region comprises the lignite range from sub-bituminous through
the coal fields of Texas and Michigan. fields of Arkansas and Texas. The low- bituminous to anthracite.
This province is subdivided into the est rank coals in the United States are The coals of Oregon and California
Northern region consisting of the found in this province, with equilibrium rank somewhat lower than those of
coal field of Michigan; the Eastern moisture contents as high as 55%. Washington. Those of California consist
region, or Illinois basin, compris- Lignites from Texas can have moisture mainly of noncoking bituminous coals
ing the fields of Illinois, Indiana, contents to 40%, low grindabilities, in the southern part of the state, lignite
and western Kentucky; and also the and heating values, (as received), below in the northern, and sub-bituminous in
Coals of the World | A-
the center. The fields are scattered and produces about 35% of the nation’s Table A-5 gives detailed analyses
limited in area. coal. The Midwest produces 15% and of U.S. bituminous coals, with free-
Estimates of U.S. coal production in the West 50%. Surface mining will swelling indices, the Audibert-Arnu
the year 2006 were about 1,200 mil- predominate and is expected to meet dilatation, and the ASTM and Interna-
lion tonnes per year. Appalachia now more than 60% of the total output. tional classifications.
Table A-5 | Source, Analysis and Classification of U.S. Bituminous Coals— continued
Moisture and Ash Free
As Received Classification
Province and Area or Seam HHV FSIA A-ADB
%VM %C %H %0 %N %S
%Ash %H2O BTU/lb MJ/kg ASTMC IC No.D
Pennsylvania—continued
Greene, Pittsburgh 8.2 1.8 40.6 84.1 5.6 6.4 1.7 2.2 15,160 35.3 9 250 Hvab 635
Indiana, Lower Freeport 9.7 2.3 24.6 88.5 5.1 3.7 1.5 1.2 15,680 36.5 9 150 Mvb 435
Lawrence, Brookville 10.1 5.2 42.9 81.5 5.7 7.3 1.7 3.8 14,830 34.5 41⁄2 75 Hvab 634
Washington, Pittsburgh 5.1 1.6 40.5 84.2 5.6 7.2 1.6 1.4 15,070 35.0 71⁄2 180 Hvab 635
Tennessee
Campbell, Jordon 3.8 4.7 38.1 83.2 5.4 8.6 2.0 0.8 14,820 34.5 3 –21 Hvab 622
Utah
Carbon, D 7.3 3.3 47.2 80.7 5.8 11.7 1.4 0.4 14,330 33.3 2 –45 Hvbb 711
Carbon, Lower Sunnyside 5.6 4.1 42.9 82.1 5.6 9.3 1.8 1.2 14,770 34.3 5 5 Hvab 633
Emery, Lower Sunnyside 6.4 5.2 43.2 79.8 5.6 11.8 1.7 1.1 14,260 33.2 3 –35 Hvbb 721
Washington
Kittitas, No.1 (Big) 8.8 3.7 42.4 81.8 6.0 9.7 1.9 0.6 14,710 34.2 4 5 Hvab 623
Kittitas, No.5 (Roslyn) 9.9 2.9 45.1 82.7 6.2 8.6 2.0 0.5 14,920 34.7 4 35 Hvab 623
West Virginia
Boone, Alma 9.8 2.3 42.5 83.5 5.7 6.3 1.5 3.0 15,110 35.1 6 120 Hvab 634
Boone, Chilton 7.8 4.2 40.2 83.8 5.5 7.5 1.7 1.5 14,930 34.7 6 10 Hvab 633
Boone, Hernshaw 7.2 2.0 38.0 86.0 5.5 6.0 1.6 0.9 15,320 35.6 61⁄2 140 Hvab 634
Fayette, Fire Creek 6.2 3.0 20.2 90.0 4.7 2.9 1.3 1.1 15,540 36.1 7 10 Lvb 433
Fayette, Lower Eagle 4.8 1.8 33.4 87.7 5.5 4.4 1.7 0.7 15,610 36.3 81⁄2 310 Hvab 635
Greenbrier, Fire Creek 10.2 3.6 24.2 88.5 5.0 3.6 1.6 1.3 15,540 36.1 81⁄2 150 Mvb 435
Greenbrier, Sewell 3.2 3.2 29.1 88.0 5.3 4.3 1.6 0.6 15,540 36.1 9 220 Mvb 535
Kanawha, Winifrede 9.5 4.8 41.5 83.3 5.6 8.6 1.7 0.8 14,880 34.6 4 –5 Hvab 622
Logan, Cedar Grove 9.3 2.4 39.2 84.2 5.5 7.1 1.6 1.6 15,180 35.3 7 150 Hvab 635
McDowell, Beckley 10.3 2.7 18.4 89.8 4.7 3.2 1.6 0.7 15,630 36.3 71⁄2 25 Lvb 333
McDowell, Bradshaw 6.7 1.9 26.5 88.9 5.1 3.4 1.7 0.9 15,690 36.5 81⁄2 240 Mvb 435
McDowell, Pocahontas No.3 6.2 3.3 18.7 89.4 4.6 4.1 1.3 0.6 15,690 36.5 81⁄2 10 Lvb 333
McDowell, Pocahontas No.4 7.3 2.8 26.1 89.4 5.0 3.7 1.3 0.6 15,610 36.3 9 170 Mvb 435
McDowell, Douglas 4.0 2.5 25.6 89.3 5.0 3.1 1.7 0.9 15,680 36.5 9 230 Mvb 435
Marshall, Pittsburgh 10.5 4.1 45.6 81.2 5.6 6.8 1.4 5.0 14,790 34.4 7 230 Hvab 635
Mercer, Pocahontas No.3 7.1 4.4 17.4 90.7 4.6 2.7 1.2 0.8 15,730 36.6 6 5 Lvb 333
Mercer, Pocahontas No.6 7.9 5.1 20.8 90.3 4.8 3.0 1.3 0.6 15.710 36.5 7 65 Lvb 434
Mingo, Lower Cedar Grove 5.0 1.7 37.5 85.6 5.5 6.4 1.6 0.9 15,390 35.8 8 160 Hvab 635
Mingo, Upper Cedar Grove 9.1 5.1 36.3 85.5 5.5 6.6 1.6 0.8 15,090 35.1 61⁄2 30 Hvab 633
Monongalia, Pittsburgh 6.6 2.3 37.5 85.5 5.5 5.8 1.8 1.4 15,310 35.6 9 250 Hvab 635
Monongalia, Redstone 12.4 2.6 36.6 84.5 5.5 6.3 1.7 2.0 15,090 35.1 71⁄2 180 Hvab 635
Nicholas, Eagle 6.8 4.0 36.2 85.6 5.5 6.4 1.7 0.8 15,270 35.5 71⁄2 140 Hvab 634
Raleigh, Beckley 5.3 1.5 20.0 89.4 4.8 3.5 1.5 0.8 15,660 36.4 81⁄2 60 Lvb 334
Randolph, Sewell 5.4 1.5 31.2 87.4 5.3 5.2 1.4 0.7 15,420 35.9 8 180 Mvb 535
Wyoming, Campbell Creek 2.0 2.2 31.7 86.8 5.4 5.5 1.5 0.8 15,570 36.2 81⁄2 280 Hvab 535
Wyoming, Eagle 3.6 1.9 29.9 88.5 5.3 4.0 1.5 0.7 15,530 36.1 81⁄2 230 Mvb 535
Wyoming, Sewell 2.4 1.5 23.8 89.8 5.0 3.1 1.6 0.5 15,700 36.5 9 170 Mvb 435
Wyoming
Sweetwater, Nos. 7, 7½, 9, 15 4.1 10.3 47.0 79.3 5.6 12.2 1.8 1.1 13,900 32.3 11⁄2 –22 Hvcb 811
A
—FSI = Free Swelling Index
B
—A-AD = Audibert-Arnu dilatation index
C
—ASTM Classifications: Lvb = low-volatile bituminous; Mvb = medium-volatile bituminous; Hvab = high-volatile A bituminous; Hvbb = high-volatile B bituminous; Hvcb =
high-volatile C bituminous
D
—IC = International Classifications
Coals of the World | A-11
Mexico
Mexico has coal deposits in 16 of the
38 Federal States (Fig. A-5). Estimated
production by the year 2000 was 55
million tonnes. Only three deposits
currently have economically recover-
Mexico
able coal reserves: Barrancas basin
(Central Sonora), Oaxaca basin, and
Coahuila basin. In the Coahuila basin,
the most significant coal deposits are
located within the two largest districts
of Sabinas and Rio Escondido.
Brown coals and hard coals are
deposited in the Sabinas basin, which
is about 35 miles (56 km) long and 15
miles (24 km) wide. The hard coals are Figure A-5 | Coal Map of Mexico and Central America
of medium rank and have ash con-
tents from 14 to 35%. These coals are Table A-6 | Analyses of Mexican Coals
cokable. High-volatile brown coals are
Nava District Rio Escondido
deposited in the Rio Escondido basin.
As Received:
Most of the coals in the Barrancas
% Ash 30.9 28.3 29.0
basin are of high rank; some of these
% Moisture 4.8 5.2 7.8
coals are anthracitic.
Moisture and Ash Free:
Table A-6 gives coal and ash
analyses of Mexican bituminous and % Volatile (VM) 50.7 48.3 46.7
sub-bituminous coals. % Carbon (C) Clean Combustion
76.2 Technologies 76.5 76.3
Figure: App A-05/ (AR)
% Hydrogen (H) 5.6 5.8 5.7
PMS5415 K
Central America % Oxygen (O) 14.8 15.0 15.8
The coal resources of Costa Rica, % Nitrogen (N) 1.4 1.2 0.9
Guatemala, Honduras, Panama, and % Sulfur (S) 2.0 1.5 1.3
Puerto Rico (Fig. A-5) comprise sub- HHV, BTU/lb 13,790 13,690 13,260
bituminous and lignitic coals. The HHV, MJ/kg 32.1 31.8 30.8
extent is generally unknown in most Fusibility Temperature, Reducing Atmosphere, °F:
countries, and the deposits are not cur- Initial deformation 2,700+ 2,700+ 2,700+
rently of economic importance. Ash Composition:
% SiO2 60.4 61.7 58.8
Coal Deposits of South
% AI2O3 25.8 27.7 29.0
America
% Fe2O3 5.0 2.1 2.8
Compared to those of North Amer- % CaO 3.1 1.4 2.6
ica, South America’s coal deposits % MgO 0.9 0.9 0.6
are very small (Fig. A-6). The only % Na2O 0.3 0.5 0.4
southern countries where deposits of
% K2O 0.7 0.8 0.8
any economic importance are known
% TiO2 0.8 0.7 1.0
to occur are Brazil, Chile, Colombia,
% P2O5 0.1 0.1
Peru, and Venezuela. Argentina has
% SO3 1.1 0.8 2.4
small resources of low-rank coal. Ow-
Base/Acid Ratio 0.11 0.06 0.08
ing to inaccessibility of most of the
Hardgrove Grindability 58 54 66
deposits and lack of transportation,
A-12 | Clean Combustion Technologies
Chile
Chilean deposits occur in the low
Argentina
lying land along the ocean and, hence,
are easily accessible. The average
thickness of the main coal seam is
59 inches (150 cm). The Province of
Concepción and the deposits near the
Bay of Arauco have thus far yielded the
best coal in Chile. Table A-9 gives the
proximate analyses of the representa-
tive coals of these regions. In addition,
considerable low-sulfur brown coal de-
posits exist in the south of the country,
in the province of Magallanes.
Figure A-6 | Coal Map of South America In 1971 the private Lota-Sweiger
Clean Combustion Technologies Coal Company was nationalized. The
coal mining has notFigure: App A-06/ (TB)
been developed. coal is made available for steelmak- Empresa National del Carbon S.A.
PMS5415 K
Lack of capital and skilled labor also ing. The Port of Rio Gallegos is used (ENACAR) was founded to operate
limits development. for shipping coal to other parts of the the state-owned coal mines. Annual
country. Today, the main consumer is production in 2006 was about 1 mil-
Argentina the Hydro and Electric Energy Agency. lion tonnes.
Argentina’s only major coal deposit is Annual coal production of about 9
located in the southwest of the country, million tonnes was estimated for the Colombia
in the Rio Turbio Basin, near the Chil- year 2000. Analyses of typical Argen- Colombia has the largest coal
ean border. The seams are 3 to 6 feet (1 tinian coals are given in Table A-7. resources in Latin America having
m to 2 m) thick. The coal is generally proved reserves of 6,600 MTOE. The
high in ash but, after treatment or Brazil coal-fields are situated mainly in the
blending with imported coal, it is suit- The principal explored and developed Eastern Cordillera although some
able for steam generation. Mixed with coal deposit of Brazil extends south coal deposits are located in the West-
high-grade coking coal, Argentinian from southern São Paulo, through the ern and Central Cordillera region. All
Coals of the World | A-13
coal ranks are available, from brown Table A-7 | Coals of Argentina – Rio Turbio Area
coal to anthracite. A considerable
Hvcb Suba
portion of the coal has good coking
As Received:
properties. Most of the coking coals
% Ash 10.9 15.4
have ash contents of less than 10% % Moisture 8.6 10.3
and low sulfur contents of 0.3 to 1%, Moisture and Ash Free:
and are exported widely for both % Volatile (VM) 47.1 50.1
steam and coking. % Carbon (C) 77.8 73.2
In the deposits of El Cerrejon, % Hydrogen (H) 6.3 5.9
open-cast mining produces coal low % Oxygen (O) 14.2 19.4
in ash and sulfur content, and high in % Nitrogen (N) 0.7 1.0
calorific value. This coal is well suited % Sulfur (S) 1.0 0.5
for power stations, but can also be HHV, BTU/lb 14,020 13,490
used for coking. The deposits produc- HHV, MJ/kg 32.6 31.4
ing the largest coal output are situated Fusibility Temperature, Reducing Atmosphere, °F:
in the district of Paz del Rio. Further Initial deformation 2,020 2,190
deposits occur in Cundinamarca, Softening (H = W) 2,280 2,300
Boyaca, the Cauca River basin, and Fluid 2,470 2,530
Samaca. Colombia plans to increase Ash Composition:
its present coal production, ensur- % SiO2 48.1 53.0
ing the domestic energy supply, and % AI2O3 25.6 23.7
anticipating future exports up to % Fe2O3 7.0 7.6
% CaO 7.6 6.2
30 million tonnes. Table A-10 gives
% MgO 1.6 1.6
analyses of typical Colombian coals.
% Na2O 0.9 0.3
Ecuador % K2O 0.3 0.4
% TiO2 1.7 2.0
Coal deposits occur in various sec-
% P2O5 0.1 —
tions of Ecuador. The bulk of the
% SO3 4.6 5.1
deposits contain sub-bituminous
Base/Acid Ratio 0.23 0.20
coal and low-grade lignite, high in Hardgrove Grindability 49 44
ash. In some localities, coals of better
quality, approaching anthracite, have
been found. However, the general low Table A-8 | Analysis of Typical BrAzilian Coals
quality, high cost of mining, and lack Moisture and Ash Free
As Received
of transportation deter exploitation of District or Mine HHV
% VM %S
Ecuadorian resources. % Ash %H2O BTU/lb MJ/kg
Butia 13.6 11.5 42.7 1.7 — —
Peru Jacuhy 21.2 9.6 28.0 1.3 — —
Coal deposits have been detected in Tubarão 28.2 1.9 41.6 1.6 — —
a number of places along the Andean Santa Catarina 27.0 1.3 43.1 5.3 15,060 35.0
ranges, mainly in the Oyon area. The Lauro Müller 18.0 2.0 38.8 2.5 14,500 33.7
coals are both anthracite and bitumi- Rio Deserto 13.8 2.0 35.6 1.2 15,320 35.6
nous. Two coal deposits in Peru have Rio Negro 39.5 7.0 36.1 2.6 12,600 29.3
been developed. Those of Rio Santa Cambiu 18.4 2.5 40.6 — 14,600 34.0
mainly consist of anthracite; production Rio De Peixe 15.5 6.3 35.7 — 14,710 34.2
ceased in 1962. At present, cokable São Jeronimo 29.0 7.3 36.3 — 13,500 31.4
A-14 | Clean Combustion Technologies
Table A-9 | Analysis of Typical Chilean Coals coals are worked in the district of
Cerro de Pasco.
Moisture and Ash Free
City or As Received The mines at Goyllarisquisga and
HHV
Mine % VM %S Quishuarcancha account for the bulk
% Ash % H2O BTU/lb MJ/kg of the Peruvian production. The seams
Lota* 15.2 2.6 44.6 1.5 14,650 34.1 are generally thick, but the quality of
Schwager 3.6 2.9 44.2 1.0 15,290 35.6 coal is low. The analyses of Table A-11
Lebu 7.8 2.2 48.9 4.4 14,700 34.2 are of typical Peruvian coals.
Rios 18.0 2.0 52.6 5.0 14,250 33.1
Venezuela
Arauco 14.3 3.4 50.2 6.6 14,500 33.7
The main Venezuelan coalfields are
Buen Retiro 7.7 1.7 51.5 6.4 15,080 35.1
in the Eastern Cordillera and in the
Mafil 11.8 12.7 47.1 0.8 13,380 31.1 region of Naricual, in the State of
Quila Coya 13.3 1.7 7.2 1.0 14,820 34.5 Anzoategui. The coal industry is
• Chemical Composition of Ash From Lola Mine government owned. The deposits are
SiO2 42.&% MgO 1.4% in detached basins, some of which
AI2O3 26.3% P2O3 0.2%
Fe2O3 14.6% SO3 4.8%
cover hundreds of miles. Although
CaO 7.2% most of the coal is of bituminous and
Ash Softening Temperature = 2,280°F
lignite rank, some deposits of semi-
anthracite occur in the Coro district
of the State of Falcon. Venezuela’s
total coal production in 2006 was 5.9
MTOE, most of which is used for do-
mestic needs. Table A-12 gives analyses
United of some Venezuelan coals.
Denmark
Kingdom
COAL DEPOSITS OF EUROPE
Netherlands
AND RUSSIA
Ireland
Belgium
Germany Most of the European countries,
Poland
excluding Scandinavia, possess coal
resources. Scandinavia (including
Czech Rep.
Iceland) has substantial quantities
Lux. of peat, which increases in value as
Slovakia Moldova
Austria oil prices rise and as there is more
France
Switz. Hungary uncertainty with regard to other fos-
Slovenia Romania
sil fuels.
Italy In continental Europe, the bulk of
Croatia Serbia Mont. coal is found in Germany, the Czech
Bulgaria. Republic, Poland, and the former
Mac.
Spain Bosnia Yugoslavia (Fig. A-7). Many other
Herz. countries with reasonable reserves have
Sardinia Greece
Albania allowed their mines to become idle.
Portugal
French, Belgian, and Spanish coal is
expensive to mine. Austria, Denmark,
Figure A-7 | Coal Map of Europe
Italy, Luxembourg, the Netherlands, and Table A-10 | Analyses of Typical Colombian Coals
Switzerland produce little or no coal.
Hvbb Subb
Austria As Received:
% Ash 8.8 7.4
Coal production from small bi-
tuminous coal deposits has been % Moisture 9.0 13.8
Table A-12 | Proximate Analyses of Venezuelan Coals Germany, the district of Nord and Pas-
de-Calais and some scattered deposits
Moisture and Ash Free
As Received in the district Centre-Midi. In the
Region Rank HHV
% VM long term, only coal production in the
% Ash % H2O BTU/lb MJ/kg Lorraine district will be maintained.
Naricual Hvbb 2.3 2.5 46.5 14,460 33.6 The smaller brown coal deposits, most
Santa of which are in the southwest of the
Hvab 2.4 1.5 40.1 14,520 33.8
Rosa country in Provence and near Arju-
Barcelona Hvab 3.4 3.0 38.6 14,620 34.0 zanx, have only local importance for
Simplicio Mvb 4.3 3.8 27.9 electricity generation.
Most of the French hard coals
Belgium produces about half of the deposits in the district Ostrava/ represented are not cokable. While
coal necessary for domestic consump- Karvina and the Lower Silesian the mines in the Pas-de-Calais still
tion; the balance is imported. Only deposit of Walbrzych, near Zaclér, are produce mostly high-rank types of
one-fourth of the coal consumed is being mined. In addition to these two coal and coking coal, the coals from
used for thermal power plants. large hard-coal deposits, there are the the Lorraine and the Centre-Midi are
two districts of Plzeui and Kladno/ usually low-rank and only part of them
Bulgaria Rakovnik and a small hard-coal de- can be used for coking. The low brown
The country has only small bitumi- posit near Rosice. In North Bohemia, coal production is exclusively used in
nous coal deposits. There is a deposit there are large and economically very power plants. Table A-13 gives analyses
of anthracite coal north of Sofia. Cok- important brown coal deposits. The of typical French coals.
able types of coal (Cretaceous coal) largest district is situated near Teplice
exist in the small district of Ojabrovo and Chomutov on the border of the Germany
and in the Dobrudsha. The brown Krusnéhory, with a further deposit of Germany has large deposits of bitumi-
coal deposits of Bulgaria are economi- economic importance in the district of nous (stein) coal and brown (braun)
cally more important than the hard Sokolov. Smaller brown coal deposits coal, the latter being unconsolidated
coal deposits, being estimated at exist in West Bohemia near Cheb, in lignite. About 60% of the currently
greater than 2,000 million tce. Depos- the south of Moravia, and in Slovakia economical, recoverable deposits have
its of glance coal and of dull brown near Handlova. sulfur below 1%, as received at the
coal (cannel coal), most of which are The coals from the districts of Plzeh power plant.
extracted by deep mining, occur near and Kladno are mostly low-rank types. The principal districts where bitumi-
Pernik, in the Pirin Mountains, and All types of brown coal, from glance nous coal occurs in greatest abundance
in Tscherno. Larger lignite deposits in brown coal to lignite, are produced in are: the Lower Rhine and Westphalia,
the district of Maritsa can be extract- the country; most of them are used for Silesia, the Rhenish district, and the
ed by open-cast mining. The brown electricity generation. Saar district. In central north Ger-
coal mines mainly serve the domestic many, brown coal occurs in the Rhine
market, while the lignite mines supply France Province between Cologne and Aachen
coal mostly to power plants. French coal deposits consist of a large (Aix-la-Chapelle). Further economi-
number of small fields distributed over cally important brown coal deposits
The Former Czechoslovakia the northern, central, and southern are in Lower Saxony (Helmstedt)
There are economically important parts of France. The deposits are gener- and Lower Bavaria (Schwandorf ).
bituminous and brown coal deposits ally deep, jagged, and difficult to mine. The Westphalian and Rhine district
in the Czech Republic and Slovakia. Usable deposits are between 2,300 and fields include the Ruhr Basin, the
Technically and economically recover- 4,000 feet (700 and 1,250 m) deep. most important coal field in Germany.
able reserves exceed 4,000 million tce. The most important are the district of These coal districts are connected with
The wide Upper Silesian hard-coal Lorraine adjoining the Saar in Western those of southern Holland, Belgium,
Coals of the World | A-17
and northern France. The coal ranges The recoverable hard-coal resources Greece
from medium to high volatile, and the of Western Germany amount to about Greece does not have any bituminous
ash varies considerably from mine to 183 MTOE, with most of that concen- coal deposits of economic importance.
mine. The bulk of the entire German trated in the Ruhr. The balance comes However, there are important lignite
output comes from the Lower Rhenish from the Saar, Aachen, and Ibben- deposits in the districts of Ptolemais
Westphalian fields. buren districts. All hard coal is deep and Megalopolis. The reserves of
The Silesian district extends mined. Production has been steadily brown coal was about 3,900 MTOE
through the province of Silesia, which, decreasing, as it was 70 MTOE in in 2006, and forms an important basis
geographically, is divided between Ger- 1996 and 50 MTOE in 2006. Coal for the production of electric energy.
many, Poland, and the Czech Republic. imports provide the difference between Analyses of Greek lignites are given in
The bulk of the coal in this district this slowing production and German Table A-15.
is of first-class coking quality, and is consumption of 82 MTOE.
widely used in steel industries. The Germany also has economically recov- Hungary
Saar district stretches through parts erable brown coal totals of about 6,556 Hungary has some bituminous coal
of Alsace-Lorraine. The coals contain MTOE, all from open-pit mining. deposits in the districts of Pécs and of
considerable ash, particularly in small Table A-14 lists typical analyses of Komlo, and several brown coal depos-
sizes, which are usually washed. Western German coals. its of national importance.
A-18 | Clean Combustion Technologies
A large brown coal deposit is situ- Ireland Peat bogs cover 15% of the country;
ated in the Budapest district. Glance There are 8 coalfield areas in Ireland more than 3 million tonnes of peat are
brown coal is exploited by deep min- (the Republic of Ireland and Northern used yearly for generating electricity and
ing in the districts of Tatabáanja and Ireland), all of limited size. Only two are domestic heating.
Oroszláany. Large lignite deposits, currently being worked; the Connaught Irish peat in its natural state contains
extracted by open-cast mining, exist (Arigna) field, which contains semi- about 95% moisture; by long-term air
in the district of Gyöangyös; in the bituminous coal with high ash content, drying, this can be reduced to 20 to 40%
Combinate Borsod, brown coal is provides fuel to an adjacent generating to produce a commercial fuel. Dried
extracted by deep mining. Most of station; the Leinster (Castlecomer) peat has a volatile content between
the coal produced is used for electric- field produces high-grade anthracite 55 and 70%, and ash from 2 to 10%.
ity generation. The use of Hungarian for domestic heating purposes. Brown Lower heating value at 30% moisture
hard coal for coke production is very coal occurs in two localities in north- is 6,000 BTU/lb (15 MJ/kg); ash
limited. Table A-16 gives analyses of east Ireland, with large reserves in the softening temperature is about 2,100°F
Hungarian brown coal. Thistleborough area. (1,150°C). Current state-owned
Coals of the World | A-19
reserves of peat are equivalent to 40 and Adamów districts, and near Table A-17 | Range of
million tonnes of coal. Turóôw, Legnica, and Belchatów. Analyses of Italian
Brown coal production is solely for (Sardinian) Coal
Italy consumption inside Poland, primarily (As-received basis)
Italy has no economically important for the generation of electricity. (See
Ash, % 15 to 20
hard coal deposits. The once-mined Table A-18.)
deposits of high sulfur coal in Sar- Economically mineable reserves of Moisture (H2O) % 12 to 15
dinia are now being used in a limited hard coal are said to be 14,000 MTOE. Volatile matter
32 to 45
(VM) %
fashion. Larger deposits of brown coal Production was 67 MTOE in 2006, a
and partly also of glance brown coal decrease of 6 MTOE since 1996. Sulfur (S) % 2 to 7
exist in Toscana, in Calabria, and in HHV, BTU/lb 9,960 to 11,500
the south-west of Sardinia. The larg- Spain HHV, MJ/kg 23.2 to 26.7
est open-cast mine is situated in the Spain has bituminous coal and lignite Ash softening
2,000
Valdarno Basin of the district Arezzo. deposits, although in several deposits temp., approx., °F
Table A-17 shows analyses of Sardin- geological conditions render exploi-
ian coals. tation difficult. The economically southern area embraces the fields
important hard coal deposits are con- known as South Wales, Forest of
Poland centrated in the province of Asturia Dean, Bristol, and Kent.
In Poland, there are several rich and in the northwest—the districts of The central area includes the fields
easily accessed hard coal and brown Oviedo and León account for nearly of Lancashire, North Wales, York-
coal regions. Hard coals, which are 90% of the total Spanish produc- shire, Derbyshire, Nottinghamshire,
all deep-mined, occur in three basins: tion of hard coal. A second large Staffordshire, and Warwickshire. The
the Upper Silesian basin has about hard coal deposit is in the district of northern area comprises the fields of
93% of the economically recoverable Mariannica, with individual mines Northumberland, Durham, Cumber-
reserves; the Lower Silesian has about in the provinces of Cordoba and land, and the fields of Scotland.
1%, found in thin seams (1 to 3 feet Sevilla. The lignite deposits are in Intensive exploration has opened
[0.4 to 1 m] thick) near the towns the northeast of the country between new areas for development, particularly
of Walbrzych and Nowa Ruda. The Teruel and the Pyrenees. With its in the Selby district (a major eastward
Lublin basin, with about 6% of the production dropping, Spain currently extension of the Yorkshire field) and
reserves, is under development; the supplies only 33% of its demand. North-East Leicestershire, with coal of
thickness of the seams ranges from 2 (See Table A-19.) geological disturbance and at relatively
to 4 feet (0.7 to 3.6 m). Most of these shallow depths. Also, over more than
coals are high-volatile bituminous in United Kingdom 200 million tonnes of workable coal
rank, are high in calorific value, and The coal fields of the UK may be have been discovered in an area of
have sulfur contents of about 1%. grouped into three principal areas Coventry and north of Kenilworth, at
Brown coal deposits, mined by designated as the southern, the central, depths between 1,000 and 2,500 feet
open-pit methods, occur in the Konin and the northern coal fields. The (300 and 760 m).
A-20 | Clean Combustion Technologies
Table A-18 | Analysis of typical Polish Coals of typical British and Welsh coals.
Coal production in the UK is only
High-Volatile B
Sub-bituminous A a fraction of domestic consumption,
Bituminous
with most being burned in power-
As Received:
generating stations. Total recoverable
% Ash 11.5 24.4
reserves are on the order of 220
% Moisture 8.8 17.4 MTOE with production of 11 MTOE
Moisture and Ash Free: in 2006, down from 30 MTOE in
% Volatile (VM) 37.1 41.9 1996.
% Carbon (C) 82.9 74.2
The Balkan States (Former Yugoslavia)
% Hydrogen (H) 5.2 6.0
% Oxygen (O) 9.9 16.7
There is only one hard coal deposit in
the former Yugoslavia, with limited
% Nitrogen (N) 1.0 0.9
economical importance—on the Istria
% Sulfur (S) 1.0 2.2
peninsula in the district of Rasa. Of
HHV, BTU/lb 14,550 13,020 higher economical importance to the
HHV, MJ/kg 33.8 30.3 former Yugoslavia are the brown coal
Fusibility Temperature, Reducing Atmosphere, °F: deposits in the districts of Zagorje,
Initial deformation 2,160 2,280 Trbovlje, Hrastnik, Kakanj, Jenica, and
Softening (H = W) 2,210 2,460
Nis, with glance brown coals as well
as several lignite deposits in Slovenia,
Fluid 2,460 2,690
such as near Velenje, as well as in
Ash Composition:
Bosnia and Serbia. Only part of the
% SiO2 46.8 55.0 hard coals of the country are suit-
% Al2O3 21.8 24.1 able for coking. The vast majority of
% Fe2O3 9.6 9.3 the coal output is used for electricity
% CaO 5.8 3.4 generation. Table A-22 gives complete
% MgO 3.5 1.5 analyses of several brown coals.
% Na2O 0.8 1.1 Russia
% K2O 3.1 1.7
Russia is divided by the Ural Moun-
% TiO2 0.7 1.1 tains into European Russia and Asiatic
% P2O5 0.3 — Russia, the latter consisting principally
% SO3 6.6 2.8 of Siberia. The accessible coal reserves
Base/Acid Ratio 0.33 0.21 of the former western and central
Hardgrove Grindability 49 44 U.S.S.R. have been largely depleted,
while the Siberian and eastern fields
are difficult and expensive to develop,
The coal of the southern area var- In the northern area, the Durham and and have high transportation costs to
ies, from bituminous coal through Northumberland fields are among the the using areas.
the well-known Welsh steam coal to most important in Great Britain. The Total economically recoverable
anthracite of notable purity. The bulk character of the coal varies consider- reserves are estimated at 157,000
of the central-area coal fields lies in ably. It is of exceptionally fine quality MTOE, of which 30% is bituminous
the south of Lancashire. Coals in this for steam, coking, household, manu- coal and the balance lignite and brown
field contain 30 to 35% volatile matter facturing, and gas-producing purposes. coal. Eighty percent of the currently re-
(VM). Some of them yield good coke. Tables A-20 and A-21 provide analyses coverable resources are located east of
Coals of the World | A-21
Tungus, Kansk-Achinsk, Irkutsk, and Turkey Erigli coal, when properly cleaned,
Minusinsk basins, and the Karagan- Considerable deposits of coal and is suitable for most purposes, and
dinsk and the Saghalin deposits. To lignite occur in Turkey, the most compares favorably with the bulk of
transport coal to European Russia important so far being those in European bituminous. It contains 40
from many of these fields requires the northwestern regions of to 45% VM and is utilized by power
journeys averaging over more than 900 Anatolia. Commercial develop- plants, cement plants, and steel pro-
miles (1,400 km), with some up to ment, however, has been practically duction. Table A-24 gives analyses of
1,860 miles (3,000 km). confined to the Erigli Basin. The two coals from Turkey.
A-24 | Clean Combustion Technologies
Africa
Both the geological coal resources
and the economically recoverable
Arctic circle coal reserves are concentrated in
the southern part of the continent.
They are preponderantly bituminous
coal. Technically and economically
Russia
60 °N recoverable reserves are estimated to
Ukraine be in excess of 50,000 MTOE; the
Republic of South Africa has nearly
Kazakhstan 98% of these (Fig. A-9).
Mongolia
Kyrgyzstan
Japan Republic of South Africa
Turkey Uzbekistan China
North Korea The Republic of South Africa (RSA)
India South Korea
has about 98% of all the coal reserves
Vietnam
of the African continent. Most of
Philippines
Afghanistan the coal deposits are situated at
Pakistan Thailand shallow depth; more than 10% of
the RSA coal output is obtained by
open-cast mining. South African
coal production was 145 MTOE in
2006, substantially greater than its 94
Figure A-8 | Coal Map of former U.S.S.R. and Asia MTOE consumption. South African
Clean Combustion Technologies
coals are exported widely and are the
Figure: App A-08/ (TB) base coals for many power plants on
PMS5415 K every continent in the world. Domes-
tic South African consumption for
Algeria the generation of electricity has grown
steadily, increasing approximately 15%
since 1996.
South African coal is deposited in
relatively small, isolated basins. The
Nigeria
coal seams are thickest and of best
quality in the middle of each basin,
but pinch out and increase in ash
Democratic content toward the periphery. The
Republic of
the Congo
biggest single drawback of the coals
Malawi
of the RSA, particularly for export, is
Mozambique that much of the inert mineral matter
Zambia
present is so intimately associated
with the organic matter that its reduc-
Madagascar tion to reasonable levels is generally
Zimbabwe
difficult and costly.
Botswana
The lowest rank coals in South
South
Swaziland Africa are found in the Orange
Africa
Free State; the fields there may be
regarded as the southern extension of
Figure A-9 | Coal Map of Africa
Coals of the World | A-25
the Transvaal fields. They have a dry prospected or its mining is yet to be be between 10 to 25 billion tonnes.
ash-free carbon content of 77%; the fully developed. Estimates of South Africa’s coal pro-
highest rank coals are the anthracites Economically recoverable reserves duction in 2006 are about 230 million
southeast of Vryheid, with a dry in the RSA are variously stated to tonnes annually.
ash-free carbon content of 91.6%.
The ranks of all the other coals fall
Table A-25 | Analyses of Typical South African (RSA) Coals
between these two limits. Aside
from a few minor peat and lignite Bituminous Coal and Ash Analyses
occurrences, all coals vary in rank As Received:
from about the lower limit of high- % Ash 13.9 16.6
volatile bituminous (or upper limit of % Moisture 2.9 6.3
sub-bituminous) coal to anthracite. Moisture and Ash Free:
The higher rank coals are largely % Volatile (VM) 33.8 31.8
confined to Natal, where the fields % Carbon (C) 84.6 82.4
are essentially an extension of the % Hydrogen (H) 4.9 4.4
Transvaal fields. % Oxygen (O) 7.2 10.6
The coal bearing areas of the
% Nitrogen (N) 2.2 1.6
Republic are first subdivided into
% Sulfur (S) 1.1 1.0
coal provinces, which in turn are
HHV, BTU/lb 14,070 14,250
split into coal fields, based either on
HHV, MJ/kg 32.7 33.1
physical separation of the areas or on
conspicuous or slight stratigraphical Fusibility Temperature, Reducing Atmosphere, °F:
Two bituminous coals from South bituminous coals are of medium and Chinese coals; Table A-27 includes
Africa, together with their ash analyses, high volatile rank, the medium volatile ash analyses; and Table A-28 shows
are given in Table A-25. being rather high in ash. Table A-26 coals currently in use.
gives the proximate analyses of many
Asian Pacific Region
Table A-26 | Proximate Analyses of Chinese Coals
The coal producing nations of the
Asia Pacific region include China, Moisture and Ash Free
District/ As Received
India, Japan, Korea, Australia, New Province HHV
Mine % VM
Zealand, Indonesia, the Philip- % Ash % H2O BTU/lb MJ/kg
pine Islands, and Korea, Vietnam, Liaoning Fushun 10.2 6.7 48.0 14,510 33.7
Thailand, and Indonesia. Of these Sian 1.7 6.6 42.1 15,190 35.3
countries, China produces 38% of Jilin Muling 10.0 3.0 32.2 15,460 36.0
the coal in the world and its share
Jehol Chaoyang 7.0 3.0 35.6 14,350 33.4
is growing quickly. By contrast the
Fusing 7.3 12.3 37.3 14,760 34.3
United States produced 19% of the
world’s coal in 2006, up only 20 Heilongjiang Tangyuan 6.2 2.0 38.2 14,290 33.2
MTOE since 1996. Australia and In- Lubing 3.7 20.9 48.2 12,660 29.4
dia are the next largest coal producers Hebei Lintsing 10.0 1.6 34.1 14,060 32.7
in Asia with production of about 200 Wanping 15.0 2.3 9.2 15,360 35.7
MTOE in 2006 followed by Indone- Henan Anyang 11.4 1.1 22.7 15,440 35.9
sia with approximately 120 MTOE. Liho 9.9 1.1 9.6 14,790 34.4
These four countries accounted Shen Xien 27.5 4.0 26.9 13,610 31.6
for almost 57% of the world’s coal
Shandong Yih Xien 10.0 0.7 30.4 15,500 36.1
production in 2006. These trends are
Poshan 8.5 0.3 12.2 15,630 36.3
likely to continue in the future.
Shanxi Dahdong 7.3 3.7 33.7 15,980 37.2
People’s Republic of China Gansu Yong Don 3.8 5.6 36.6 15,280 35.5
China consumed 1,191 MTOE Hunan Tzehsin 10.0 2.8 25.7 14,400 33.5
in 2006. The rate of Chinese coal Siangtan 25.2 3.5 39.8 14,760 34.3
consumption is growing fast with Ningshian 8.0 6.0 35.6 15,530 36.1
2006 consumption being 7.5% greater Hubei Tayei 12.2 1.3 12.0 15,660 36.4
than 2005. China’s rate of produc- Tzequai 14.7 4.8 31.9 15,610 36.3
tion has grown proportionately, with
Jiangxi Po-Loo 10.2 3.1 47.5 13,610 31.7
nearly all production being consumed
Gao Jeng 8.9 1.0 27.1 15,870 36.9
domestically. Coal deposits of China
Jian 6.2 1.5 33.0 12,030 28.0
are extensive, coal being found in
almost every province. Total overall Anhui Dahdong 8.9 3.3 41.7 14,600 34.0
reserves are estimated at 114,500 Wheizhou 13.6 0.4 32.1 15,110 35.1
MTOE. One very large area is in Jiangsu Xiaohsien 16.2 2.0 26.9 15,550 36.2
northern China, extending over Zhejiang Shangshin 16.6 1.2 44.1 15,530 36.1
most of the southern part of Shanxi, Sichuan Weiyuan 8.2 2.2 33.6 15,610 36.3
and one in the south extending over Pahsien 13.2 1.2 21.5 15,940 37.1
southern Hunan, Guizhou, Yunnan,
Guizhou Tungtze 12.2 1.0 22.0 15,730 36.6
and Sichuan. In variety, Chinese coals
Guiyang 15.4 1.8 24.3 14,420 33.5
range from hard anthracite to lignites
Yunnan Shuanwei 12.7 0.9 30.9 15,530 36.1
of pronounced woody structure. The
Coals of the World | A-27
India
Table A-27 | Chinese Coals with Ash Analyses
India’s coalfields are extensive. They
An-Shan Mine Guangdong Shang Shan District
contain economically recoverable Liaoning Province Province South China
reserves estimated to be 900,085 Hvbb Lvb Lvb
MTOE of hard coal, and 2,360 As Received:
MTOE of lower grades. The coals % Ash 1.5 22.5 16.7
vary greatly in quality, but are general-
% Moisture 2.3 0.5 1.9
ly low in sulfur—less than 1% sulfur,
Moisture and Ash Free:
as fired, for most of the production.
Ash contents are high, ranging from % Volatile (VM) 50.3 13.0 20.1
15 to 45% from most mines. % Carbon (C) — — 87.0
The main coal deposits are located % Hydrogen (H) — — 4.4
in a triangle, of which the western % Oxygen (O) — — 2.9
edge is in the center of India and the % Nitrogen (N) — — 0.8
eastern corner is about 150 kilometers
% Sulfur (S) 0.5 1.6 4.9
west of Calcutta. Most of the active
HHV, BTU/lb 14,290 14,870 15,100
mining areas are in Andhra Pradesh,
Maharashtra, Madhya Pradesh, Orissa, HHV, MJ/kg 33.2 34.6 35.1
Bihar, and West Bengal. The oldest Fusibility Temperature, Reducing Atmosphere, °F:
coalfields, those of Raniganj and Jharia Initial deformation — 2,700 + 2,470
in the east, produce both high-quality Softening (H = W) — 2,700+ 2,550
metallurgical and nonmetallurgical Fluid — 2,700+ 2,700+
coals. The trend now is toward more
Ash Composition:
development in the western part of
% SiO2 39.7 52.2 51.2
the Indian coal-belt triangle. Brown
% Al2O3 42.7 34.3 29.2
coals are found mainly in the district
of Neyveli and in Tamil Nadu. Sixty % Fe2O3 6.2 6.6 4.6
percent of the brown coal reserves have % CaO 2.7 0.8 4.3
sulfur contents below 1%; the rest can % MgO 1.6 1.1 0.7
be as high as 3% sulfur. % Na2O 1.3 0.4 0.7
Coal mines in India are gener- % K2O 1.6 2.5 0.8
ally shallow, with depths to within
% TiO2 1.8 1.2 1.8
1,000 feet (300 m) of the surface;
% P2O5 0.3 0.3 ...
the deepest mine is about 2,600 feet
(800 m). Between 35 and 40% of the % SO3 1.4 0.5 6.4
future production of bituminous and Base/Acid Ratio 0.16 0.13 0.14
sub-bituminous coal will be mined by Hardgrove Grindability — 95 120
open-cast methods, the balance being
by deep mining. Much of the coal is in MTOE, and growing at a rate of 7% and Hokkaido are by far the most im-
thick seams—75% appears in seams per year. Analyses of typical coals from portant for quantity and quality. There
above 16 feet (4.8 m) in thickness. Pro- several areas are included in Table A-29. are two minor deposits on the main
duction was measured at 210 MTOE island of Honshu.
in 2006, an increase of 54 MTOE Japan Reserves of coal in Japan are estimat-
since 1996. But even this large increase Coal is widely distributed throughout ed at 359 MTOE, nearly all extractable
in production has not kept pace with all the islands and territories of the by deep mining; nearly half of this will
consumption, measured in 2006 at 237 Japanese Empire. The islands of Kyusku come from the Ishikari coalfield on
A-28 | Clean Combustion Technologies
Hokkaido. Production in the year 2006 Korea a consumption increase of 70% since
was 0.7 MTOE, a decrease of nearly All of the coal mined in Korea is 1996. Like Japan, Korea imports coal
80% since 1996. Japan’s consumption of high-ash-content anthracite. South mainly from regional producers like In-
119 MTOE is from imports originating Korea produces 1.3 MTOE, but donesia and Australia. Table A-30 gives
mostly in the Asian Pacific region. consumed almost 55 MTOE in 2006, analyses of several Korean anthracites.
Coals of the World | A-29
* Hardgrove Grindability: 98 at 10% H2O; 134 at 18% H2O; 135 at 29% H2O; 148 at 38% H2O
As-received sulfur contents range from Eastern Queensland and the Sydney are of significance. In Australia, a
0.1 to 0.5%. Basin in New South Wales. Nearly depth of 3,300 feet (1,000 meters) is
half of the country’s production is generally taken as the current limit
Australia from underground mines in New for bituminous coal to be regarded as
Australia’s large coal resources are South Wales; another 30% is from economically extractable.
concentrated in the Eastern States. open-cut mines in Queensland. The Major brown coal deposits are
The largest and economically most Galilee Basin in Queensland also located in Victoria where the Latrobe
significant deposits of bituminous contains a large resource and the Valley contains proved recoverable re-
coal are found in the Bowen Basin of Surat and Clarence-Moreton Basins serves of some 35,000 MTOE within
A-30 | Clean Combustion Technologies
the limits of a maximum mining depth higher quality coals. Table A-31 gives bituminous field. The coal is of good
of 660 feet (200 m) and a minimum analyses of Australian coals from quality and makes an excellent steam
seam thickness of 50 feet (15 m). The several mining areas. fuel. It is coking, but the coke produced
coal is used predominantly for power is not rated high.
generation. The average moisture New Zealand
content is 62% and the lower heating Coal ranging from anthracite to Indonesia
value, as received, ranges from 2,580 to lignite occurs in many sections Coal production has grown almost
5,370 BTU/lb (6.0 to 12.5 MJ/kg). of New Zealand (Fig. A-10). The 4 times since 1996. Total Indonesian
The coal resources in other states Buller-Mokihinui is the best-known coal production was 120 MTOE in
are relatively small, in most cases
comprising isolated deposits of sub-
bituminous coal or lignite. The major
use for these coals in South Australia
and Western Australia, such as from
the Collie basin, is in power stations.
Australia’s economically recover-
Australia
able reserves are in excess of 78,000
MTOE, about 50% of which is
bituminous coal and the balance,
lower ranked types. About 20% of
the bituminous coal is capable of
being surface-mined. Production
New Zealand
of all types of coal was 203 MTOE
in 2006, a 52% increase from 1996. Tasmania
Most of the brown coal and lower
quality bituminous coal is consumed
domestically while exporting the
Figure A-10 | Coal Map of Oceania
Coals of the World | A-31
2006. Indonesian coals with very low lignite on these islands are 740 MTOE The deposits contain lignite, sub-
sulfur contents are widely exported in and 4,428 MTOE, respectively. bituminous, and bituminous coals. The
the region and the entire developed bulk of the lignite is black, and seldom
world (Table A-32). Coal deposits of Philippine Islands displays a woody structure or brown
economic significance are confined to Coal is mined in eight localities in the color. The bituminous coals are black,
the western part of Indonesia, on the Philippine Islands, namely: Bataan, hard, and lustrous. They are gener-
islands of Sumatra and Kalimantan. Cebu, Zamboanga, Politti, Masbati, ally noncoking. A typical Philippine
Estimates of measured reserves of Mindoro, Luzon, and Mindanao. The coal, from Liguan, has an as-received
bituminous coal and sub-bituminous/ basins are small and discontinuous. ash content of 11.5%, and moisture of
A-32 | Clean Combustion Technologies
9.4%. On a moisture- and ash-free tions that specify the boundaries vitrinite, a component derived from
basis, VM is 45.3%, sulfur is 0.8%, that apply to these descriptive terms, the original plant material. Vitrinite is
and HHV is 13,480 BTU/lb (31.3 which all relate to the geological the most predominant maceral in coals
MJ/kg). Ash softening temperature maturity (rank) of the coals. Further, of intermediate reflectance. Vitrin-
of such a coal, in a reducing atmo- there is no simple system for the clas- ite exhibits varying degrees of light
sphere, is approximately 2,190°F sification of coals that can provide, on reflectivity depending on the degree of
(1,200°C). a comparative basis, an indication of geological metamorphosis. Liptinite,
coal characteristics on a worldwide another maceral also originating from
Vietnam basis. ISO standard 11760 provides plant tissues, is much less reflective.
The only known coal deposits in Viet- a basis for addressing both of these Inertite is a highly reflective maceral
nam are anthracitic, chiefly from the issues.2 The development of this that is more predominant in hard an-
Nong-Son area. As-received ash con- ISO standard has been guided by thracite coals than vitrinite. If the coal
tent ranges from 15 to 27%, with sulfur the recently published “International samples are prepared in a standardized
from 2 to 3.5%. VM is between 8 and Classification of In-Seam Coals,” manner and the optical equipment is
11%, with high heating value (HHV) developed by the Economic Com- referenced against known standards,
from 13,800 to 14,600 BTU/lb (32 to mission for Europe, Committee on then a consistent measurement of
34 MJ/kg), both on a moisture- and Sustainable Energy (United Nations vitrinite reflectance can be made. The
ash-free basis. Hardgrove grindability Report, Energy 19, New York and measurement of vitrinite reflectance is
is usually between 32 and 53, with Geneva, Sept., 1998). governed by ASTM standard D2798
some samples as high as 73. in the United States and internation-
CLASSIFICATION OF COALS BY TYPE ally, by ISO 7404.
INTERNATIONAL COAL The U.S. system, defined by ASTM The ISO major rank categories
CLASSIFICATION method D388 classifies coal according are wholly defined by their vitrinite
Coals occur worldwide and vary to its calorific value on a moist mineral reflectance as:
significantly in their physical and matter-free basis or fixed carbon con-
chemical characteristics for a variety tent calculated on a dry, mineral-free 1. Anthracite, synonymous with
of reasons, both with respect to the basis, results in 13 classes of coals. The high-rank coal, having a mean
organic coal substance and to the as- international system defined by ISO random vitrinite reflectance, Rr,
sociated mineral matter that is always standard 11760:2005 classifies coal equal to or greater than 2.0 % but
present to varying extents. Coals are by its light reflecting properties. These less than 6.0 %, or, preferably, a
an important source of energy, as well differences are expressed as three major mean maximum reflectance, Rv,
as being essential for the production categories of coal: high rank, medium max, less than 8.0 % for geologi-
of metallurgical cokes, and are widely rank, and low rank coals. These major cally unaltered coal.
used as feedstock for other industrial categories are further broken down 2. Bituminous, synonymous with
processes such as in the production into subcategories, designated by the medium rank coal, having a mean
of gaseous fuels and synthesis gas. letters A, B, C, or and D. random vitrinite reflectance, Rr,
Hence, a wide range of procedures has Petrography is the study of miner- equal to or greater than 0.5%, but
been developed by the International als, in this case coal samples with a less than 2.0%.
Organization for Standardization microscope, specialized illumina- 3. Low rank coal, comprising both
(ISO) for the analysis and testing of tion, an optical detector system, and lignite and sub-bituminous
coals. These ISO procedures are vari- a recording device. Distinct organic coals, having a mean random
ously designated as being applicable substances, termed macerals, can be vitrinite reflectance, Rr, less than
to “hard coals,” “brown coals,” “lignite,” discerned and classified on the basis 0.5%. The terms “lignite” and
“bituminous coals,” and “anthracite.” of these optical microscopic proper- “brown coal” are typically used
There are, however, no ISO defini- ties. The most important maceral is interchangeably, but brown coal
A-34 | Clean Combustion Technologies
is the term now preferred by the 9. Low rank B, (lignite B) Rr High vitrinite > 80% vitrinite
International Commission for < 0.4% and bed moisture by volume, min-
Coal and Organic Petrology. < 35% on an ash-free basis eral free
10. Low rank C, (lignite C) Rr
These three categories are further < 0.4% and bed moisture The following is used to describe
broken down into 10 subcategories > 35% and < 75% on an coal by ash yield:
defined as: ash-free basis
Very low ash < 5% ash by mass, dry
basis
1. High Rank A, (anthracite A) 4.0% These rankings are typically accom-
< Rr < 6.0% (or Rv (max) < 8.0%) panied by a description of the coal’s Low ash > 5% < 10% ash by
2. High rank B, (anthracite B) 3.0% properties with respect to vitrinite mass, dry basis
< Rr < 4.0% content and ash fraction. Medium ash > 10% < 20% ash by
3. High rank C, (anthracite C) 2.0% The following is used to describe mass, dry basis
< Rr < 3.0% coal by petrographic composition: Moderately > 20% < 30% ash by
4. Medium rank A, (bituminous A) high ash mass, dry basis
1.4% < Rr < 2.0% Low vitrinite < 40% vitrinite High ash > 30% < 50% ash by
5. Medium rank B, (bituminous B) by volume, min- mass, dry basis
1.0% < Rr < 1.4% eral free
6. Medium rank A, (bituminous C) Medium vitrinite > 40% < 60% Table A-33 is a list of selected Unit-
0.6% < Rr < 1.0% vitrinite by vol- ed States medium rank coal samples
7. Medium rank A, (bituminous D) ume, mineral free showing their origin, ASTM rank, key
0.5% < Rr < 0.6% Moderately high > 60% < 80% fuel properties, petrographic data, and
8. Low rank A, (sub-bituminous) vitrinite vitrinite by vol- ISO category.
0.4% < Rr < 0.5% ume, mineral free
Table A-33 | Comparison of Rank and ISO Category of Selected U.S. Medium Rank Coals
(Rr)
Heating (Ro)
Volatile Calculated
U.S. ASTM Ash Sulfur Value Mean-Max %
Coal Seam Moisture Matter Random ISO Category
State Rank (Dry) (Dry) BTU/lb Vitrinite Vitrinite
(Dry) Vitrinite
(DRV) Reflectance
Reflectance(1)
PA Lower Kittanning lvb 1.73 15.98 23.61 1.63 11,562 1.7 89.7 1.59 Medium Rank A
PA Lower Kittanning lvb 1.56 17.42 8.59 0.74 13,997 1.7 91.5 1.59 Medium Rank A
WV Elk Lick lvb 1.32 16.61 16.68 3.32 12,811 1.69 93.6 1.58 Medium Rank A
PA Lower Kittanning lvb 1.06 15.69 29.42 1.93 10,632 1.68 94.2 1.57 Medium Rank A
WV Elk Lick lvb 1.65 15.8 20.3 2.82 12,192 1.67 92.6 1.56 Medium Rank A
OK Upper Hartshorne lvb 2.51 19.18 12.74 3.29 13,414 1.67 92.8 1.56 Medium Rank A
PA Upper Freeport lvb 2.93 18.94 6.52 1.62 14,608 1.65 92.1 1.54 Medium Rank A
WV Pocahontas #3 lvb 1.43 19.2 10.86 0.7 13,953 1.62 75.5 1.52 Medium Rank A
AR Lower Hartshorne lvb 2.67 19.64 11.16 0.8 13,612 1.61 89.2 1.51 Medium Rank A
PA Kelly lvb 3.22 19.25 15.73 1.08 12,894 1.57 92.1 1.47 Medium Rank A
PA Lower Freeport lvb 1.98 18.2 12.93 6.44 13,477 1.57 65 1.47 Medium Rank A
PA Barnett mvb 1.16 21.27 13.3 0.55 13,568 1.54 92.7 1.44 Medium Rank A
PA Lower Freeport lvb 1.14 20.59 12.92 4.04 13,233 1.47 85.8 1.38 Medium Rank B
OK Hartshorne mvb 0.94 20.86 10.18 1.05 13,841 1.46 90.5 1.37 Medium Rank B
PA Lower Kittanning lvb 1.32 20.62 11.95 4.45 13,411 1.45 89.4 1.36 Medium Rank B
CO Colorado Basin B mvb 4.43 20.22 23.39 0.76 11,986 1.45 94.6 1.36 Medium Rank B
PA Lower Kittanning mvb 0.61 22.87 5.68 1.71 14,860 1.43 87.6 1.34 Medium Rank B
Coals of the World | A-35
Table A-33 | Comparison of Rank and ISO Category of Selected U.S. Medium Rank Coals —continued
(Rr)
Heating (Ro)
Volatile Calculated
U.S. ASTM Ash Sulfur Value Mean-Max %
Coal Seam Moisture Matter Random ISO Category
State Rank (Dry) (Dry) BTU/lb Vitrinite Vitrinite
(Dry) Vitrinite
(DRV) Reflectance
Reflectance(1)
CO Colorado Upper B mvb 0.95 24.08 9.94 1.35 14,111 1.39 93.9 1.31 Medium Rank B
PA Mercer mvb 0.81 24.84 8.98 3.81 14,032 1.36 88.3 1.28 Medium Rank B
PA Upper Kittanning mvb 1.47 25.42 10.25 0.55 13,946 1.35 86.3 1.27 Medium Rank B
WV Fire Creek mvb 1.03 23.11 13.93 1.14 13,284 1.35 86.7 1.27 Medium Rank B
WV Sewell mvb 1.45 24.98 4.22 0.62 14,871 1.35 77.2 1.27 Medium Rank B
PA Clarion mvb 0.62 25.81 5.53 1.34 14,770 1.33 90.9 1.25 Medium Rank B
WV Beckley mvb 1.65 25.73 6.17 1.99 14,444 1.33 92.1 1.25 Medium Rank B
PA Lower Kittanning mvb 0.67 24.24 11.48 3.92 13,749 1.31 87.1 1.23 Medium Rank B
AL Pratt mvb 0.76 24.23 4.58 0.96 14,872 1.31 78.6 1.23 Medium Rank B
PA Bloss mvb 1.35 19.19 27.45 2.22 11,013 1.31 86.8 1.23 Medium Rank B
PA Pittsburgh mvb 2.03 21.72 15.78 1.16 13,033 1.3 92.5 1.22 Medium Rank B
WV Sewell mvb 2.46 23.17 18.83 1.17 12,355 1.28 81.1 1.21 Medium Rank B
CO Coal Basin M mvb 1.1 27.19 5.37 0.65 14,889 1.28 94.4 1.21 Medium Rank B
PA E mvb 3.14 23.97 13.47 3.84 13,449 1.28 86.2 1.21 Medium Rank B
PA Upper Freeport mvb 1.96 26.36 12.92 2.17 13,461 1.27 91.6 1.20 Medium Rank B
PA Clarion mvb 0.83 22.96 18.72 0.72 12,248 1.25 88.9 1.18 Medium Rank B
PA Brookville mvb 2.36 22.94 16.52 3.81 12,908 1.24 77.3 1.17 Medium Rank B
PA Upper Freeport mvb 6.13 24.39 10.78 2.58 14,006 1.24 88.5 1.17 Medium Rank B
OK Stigler mvb 2.95 26.45 9.65 0.46 14,149 1.24 91 1.17 Medium Rank B
WV Sewell mvb 1.82 27.29 3.66 0.46 15,075 1.21 65.5 1.14 Medium Rank B
PA Middle Kittanning mvb 3.74 20.68 34.78 3.81 9,715 1.18 89.6 1.11 Medium Rank B
PA Lower Freeport hvAb 1.16 30.64 9.99 1.87 13,969 1.17 91.4 1.10 Medium Rank B
MD Brush Creek mvb 1.04 25.83 15.07 4.76 12,988 1.17 87.8 1.10 Medium Rank B
VA Splash Dam mvb 2.06 30.13 3.89 0.79 15,060 1.16 76.8 1.10 Medium Rank B
WV Sewell hvAb 1.38 28.37 11.29 1.03 13,693 1.16 69.2 1.10 Medium Rank B
WV Sewell mvb 1.27 28.5 8.69 0.82 14,048 1.15 68.6 1.09 Medium Rank B
WV Upper Kittanning lvb 1.37 18.08 12.53 2.16 13,600 1.15 91.9 1.09 Medium Rank B
PA Lower Kittanning mvb 0.58 27.8 9.95 2.31 13,971 1.14 86.9 1.08 Medium Rank B
WV Gilbert mvb 1.43 24.84 20.02 0.58 12,226 1.14 76.9 1.08 Medium Rank B
VA Splash Dam mvb 1.62 27.4 12.42 0.73 13,588 1.13 74.8 1.07 Medium Rank B
WV Upper Kittanning hvAb 1.72 29.05 12.64 2.69 13,692 1.13 85.2 1.07 Medium Rank B
AL Mary Lee hvBb 1.88 27.83 20.96 0.8 11,838 1.12 78.5 1.06 Medium Rank B
OK Stigler hvAb 1.5 29.97 11.62 5.45 13,726 1.11 87.7 1.05 Medium Rank B
WV Gilbert mvb 1.3 24.67 22.46 0.55 11,817 1.11 74 1.05 Medium Rank B
VA Upper Banner hvAb 1.38 32.22 14.75 0.63 13,149 1.08 69.3 1.02 Medium Rank B
VA Lyons hvAb 1.21 32.78 5.09 1.46 14,696 1.07 75.9 1.01 Medium Rank B
VA Lower Banner hvAb 1.25 32.82 11.08 0.86 13,725 1.06 63.5 1.00 Medium Rank B
KY Imboden hvAb 1.14 30.74 16.73 1.03 12,711 1.04 81.2 0.99 Medium Rank C
KY Lower Elkhorn hvAb 1.36 31.22 10.84 0.51 13,741 1.04 76.6 0.99 Medium Rank C
KY Lower Elkhorn hvAb 1.25 31.24 9.78 0.5 13,897 1.03 71.2 0.98 Medium Rank C
WV Sewell B hvAb 2.59 31.55 3.37 0.63 15,169 1.03 64.3 0.98 Medium Rank C
VA Imboden hvAb 1.26 33.63 7.21 0.54 14,239 1.03 79 0.98 Medium Rank C
PA Upper Freeport hvAb 1.09 32.85 13.12 1.19 13,294 1.02 88.5 0.97 Medium Rank C
WV Middle Kittanning hvAb 1.06 29.07 30.35 0.66 10,151 1.02 73.3 0.97 Medium Rank C
WV Lower Kittanning hvAb 0.94 30.23 16.41 2.8 12,735 1.01 77.5 0.96 Medium Rank C
VA Lower Banner hvAb 0.98 32.36 8.81 0.71 14,268 1.01 55.5 0.96 Medium Rank C
KY Lower Elkhorn hvAb 1.22 32.01 6.98 0.59 14,389 1.01 72.6 0.96 Medium Rank C
A-36 | Clean Combustion Technologies
Table A-33 | Comparison of Rank and ISO Category of Selected U.S. Medium Rank Coals —continued
(Rr)
Heating (Ro)
Volatile Calculated
U.S. ASTM Ash Sulfur Value Mean-Max %
Coal Seam Moisture Matter Random ISO Category
State Rank (Dry) (Dry) BTU/lb Vitrinite Vitrinite
(Dry) Vitrinite
(DRV) Reflectance
Reflectance(1)
WV Sewell B hvAb 2.54 31.88 4.08 0.56 14,956 1.01 61.3 0.96 Medium Rank C
VA Clintwood hvAb 1.3 35.31 5.93 1.96 14,391 1.01 66.2 0.96 Medium Rank C
VA Upper Banner #3 hvAb 2.44 34.3 6.36 0.59 14,543 1 71.1 0.95 Medium Rank C
KY Pond Creek hvAb 2.32 33.29 10.89 0.6 13,689 0.99 72.9 0.94 Medium Rank C
WV Sewell B hvAb 1.71 30.83 4.58 0.58 15,035 0.99 53.3 0.94 Medium Rank C
AL Mary Lee hvAb 1.56 29.36 20.09 0.93 11,857 0.99 74.4 0.94 Medium Rank C
AL American hvAb 1.27 31.43 15.67 2.34 12,853 0.98 71.3 0.93 Medium Rank C
KY Elkhorn #3 hvAb 3.37 37.29 3.33 0.98 14,540 0.97 84.8 0.92 Medium Rank C
WV Upper Kittanning hvAb 1.75 26.87 21.44 1.03 11,707 0.97 48.1 0.92 Medium Rank C
WV Redstone hvAb 2.07 35 5.43 0.8 14,616 0.96 93.4 0.91 Medium Rank C
VA Banner hvAb 1.34 32.4 13.63 0.7 13,325 0.96 75.2 0.91 Medium Rank C
PA Pittsburgh hvAb 1.77 34.26 11.35 1.61 13,311 0.89 87.4 0.85 Medium Rank C
TN Pee Wee hvAb 2.77 36.66 7.23 1.29 13,801 0.88 73.6 0.84 Medium Rank C
KY Elkhorn #3 hvAb 2.79 37.53 5.47 0.97 14,244 0.87 77.7 0.83 Medium Rank C
PA Pittsburgh hvAb 2.4 36.02 10.25 1.12 13,532 0.87 83 0.83 Medium Rank C
KY Upper Elkhorn #3 hvAb 1.74 36.71 3.96 0.78 14,434 0.87 54.8 0.83 Medium Rank C
WV Pittsburgh hvAb 1.41 37.51 7.63 1.59 13,988 0.87 88.1 0.83 Medium Rank C
WV Pittsburgh hvBb 3.07 36.54 10.52 1.94 13,382 0.87 75.4 0.83 Medium Rank C
NM York Canyon hvAb 1.77 35.79 7.93 0.44 13,876 0.86 85.9 0.82 Medium Rank C
NM York Canyon hvAb 1.83 35.89 7.23 0.44 13,877 0.86 84.2 0.82 Medium Rank C
OH Ohio #5 hvBb 4.15 37.98 4.8 1.92 14,258 0.86 87.2 0.82 Medium Rank C
OK Croweburg hvAb 4.45 33.86 5.69 0.54 14,066 0.86 93.7 0.82 Medium Rank C
TN Sewanee hvAb 1.57 34.09 14.38 2.29 12,959 0.86 81.9 0.82 Medium Rank C
OK Croweburg hvAb 2.71 36.83 5.13 2.92 14,330 0.86 89.9 0.82 Medium Rank C
NM York Canyon hvAb 2.56 34.9 9.25 0.45 13,574 0.85 84.1 0.81 Medium Rank C
WV Waynesburg hvAb 1.57 35.28 15.24 2.45 12,740 0.85 82.2 0.81 Medium Rank C
WA Queen hvAb 2.33 33.16 21.14 1.41 11,876 0.85 85.2 0.81 Medium Rank C
PA Pittsburgh hvAb 1.64 36.48 7.07 1.36 13,958 0.85 82.5 0.81 Medium Rank C
NM York Canyon hvAb 1.45 36.56 9.82 0.51 13,687 0.84 74.7 0.80 Medium Rank C
PA Upper Freeport hvAb 2.91 37.26 8.78 2.06 13,594 0.84 78.5 0.80 Medium Rank C
KY Hazard #7 hvAb 3.32 36.39 5.41 0.55 13,994 0.84 68.3 0.80 Medium Rank C
KY Upper Elkhorn #3 hvAb 1.87 37.57 4.78 1.6 14,316 0.83 62.4 0.79 Medium Rank C
OH Ohio #5 hvAb 4.07 38.03 3.85 1.27 14,042 0.83 91.2 0.79 Medium Rank C
WV Pittsburgh hvAb 2.18 40.06 6.81 2.68 13,989 0.82 91.6 0.78 Medium Rank C
PA Lower Kittanning hvAb 5.01 33.95 12.91 1.35 12,942 0.82 75.3 0.78 Medium Rank C
WV Waynesburg hvAb 1.46 34.38 16.77 2.68 12,433 0.82 74.9 0.78 Medium Rank C
PA Upper Clarion hvAb 1.73 39.41 8.91 4.82 13,390 0.81 79.7 0.77 Medium Rank C
NM York Canyon hvAb 1.55 35.92 8.69 0.51 13,856 0.81 83.4 0.77 Medium Rank C
KY Upper Elkhorn #3 hvAb 2 40.93 4.73 2.02 14,076 0.81 69.4 0.77 Medium Rank C
WV Pittsburgh hvAb 2.1 40.97 6.43 2.48 14,025 0.8 84.6 0.77 Medium Rank C
WV Pittsburgh hvAb 1.82 40.55 8.24 3.37 13,697 0.8 83.1 0.77 Medium Rank C
UT Upper Sunnyside hvBb 3.38 38.8 7.54 0.77 13,640 0.8 86.9 0.77 Medium Rank C
UT Lower Sunnyside hvAb 2.71 37.89 10.13 1.67 13,276 0.8 76.7 0.77 Medium Rank C
KS Mineral hvAb 2.35 36.19 16.35 5.34 12,489 0.8 80.8 0.77 Medium Rank C
Coals of the World | A-37
Table A-33 | Comparison of Rank and ISO Category of Selected U.S. Medium Rank Coals —continued
(Rr)
Heating (Ro)
Volatile Calculated
U.S. ASTM Ash Sulfur Value Mean-Max %
Coal Seam Moisture Matter Random ISO Category
State Rank (Dry) (Dry) BTU/lb Vitrinite Vitrinite
(Dry) Vitrinite
(DRV) Reflectance
Reflectance(1)
KY Hazard #9 hvAb 2.76 34.31 14.16 1.11 12,650 0.8 75.8 0.77 Medium Rank C
OK Croweburg hvBb 5.47 33.29 9.61 2.4 13,200 0.8 81.9 0.77 Medium Rank C
KY Hazard #5A hvBb 3.79 36 9.48 0.74 13,369 0.8 60.3 0.77 Medium Rank C
PA Middle Kittanning hvAb 2.86 37.43 16.22 7.65 12,434 0.79 80 0.76 Medium Rank C
PA Lower Kittanning hvAb 2.58 35.54 15.23 3.73 12,548 0.79 83 0.76 Medium Rank C
MO Knifeton hvAb 3.59 34.81 16.48 5.87 12,590 0.78 83.5 0.75 Medium Rank C
IN Brazil Block—Top Vein hvBb 9.91 35.48 5.69 0.5 13,710 0.78 62.4 0.75 Medium Rank C
OH Ohio #6A hvAb 3.22 38.78 6.61 2.14 13,923 0.78 74.1 0.75 Medium Rank C
WA Big #2 hvBb 2.62 32.01 25.74 0.35 10,431 0.78 89 0.75 Medium Rank C
WV Waynesburg hvAb 1.56 33.76 18.46 2.51 12,242 0.78 78.4 0.75 Medium Rank C
MO Drywood hvAb 2.41 34.87 16.59 4.45 12,515 0.77 86.2 0.74 Medium Rank C
PA Lower Kittanning 1.88 38.22 13.04 3.87 0.77 83.6 0.74 Medium Rank C
UT Upper Sunnyside hvAb 3.07 39.03 6.83 0.54 13,638 0.77 86.1 0.74 Medium Rank C
PA Lower Clarion hvAb 3.09 39.57 10.08 5.64 13,557 0.77 84.6 0.74 Medium Rank C
PA Lower Kittanning 3.49 38.75 6.62 3.81 0.77 85.5 0.74 Medium Rank C
PA Brookville hvAb 5.31 37.86 5.05 0.78 14,165 0.77 80.5 0.74 Medium Rank C
PA Pittsburgh hvAb 2.54 34.43 13.67 1.37 12,855 0.77 87.2 0.74 Medium Rank C
MO Mulberry hvCb 6.42 33.98 22.71 5.5 10,655 0.76 88 0.73 Medium Rank C
OH Ohio #12A hvAb 4.04 36.41 19.29 2.73 11,755 0.76 87 0.73 Medium Rank C
WV Redstone hvAb 2.53 39.68 6.79 2.41 13,897 0.76 91.8 0.73 Medium Rank C
CO Colorado B hvBb 3.37 40.92 5.85 0.67 13,780 0.75 91.7 0.72 Medium Rank C
PA Clarion hvAb 3.51 38.82 11.49 4.52 13,518 0.75 68.1 0.72 Medium Rank C
UT Upper Sunnyside hvAb 1.93 39.66 8.35 0.67 13,551 0.74 87.4 0.71 Medium Rank C
OH Ohio #5 hvAb 4.18 36.21 18.19 0.98 12,157 0.74 41.4 0.71 Medium Rank C
CO Dakota hvBb 4.57 30.46 22.06 0.88 11,157 0.74 53.9 0.71 Medium Rank C
AK Deadfall Syncline #4 subA 10.12 36.68 3.32 1.46 12,107 0.73 86.1 0.70 Medium Rank C
MO Rowe hvAb 2.39 31.79 30.17 5.46 10,360 0.73 85.4 0.70 Medium Rank C
UT Upper Sunnyside hvAb 1.72 39.54 7.67 0.62 13,748 0.73 85.7 0.70 Medium Rank C
PA Upper Clarion hvAb 6.44 35.77 17.67 5.93 12,204 0.73 77.2 0.70 Medium Rank C
OK Mineral hvAb 3.89 37.38 10.46 3.34 13,564 0.73 87.5 0.70 Medium Rank C
CO Juanita C hvBb 5.8 40.68 5.55 0.64 13,579 0.73 92.4 0.70 Medium Rank C
WV Pittsburgh hvAb 1.93 40.42 8.71 3.43 13,733 0.73 77.3 0.70 Medium Rank C
KY Kentucky #6 hvBb 3.81 37.27 8.13 2.82 13,414 0.73 88.4 0.70 Medium Rank C
MO Baxter hvAb 2.25 37.79 13.13 6.07 12,860 0.73 79.8 0.70 Medium Rank C
PA Pittsburgh hvAb 2 39.42 9.44 3.87 13,614 0.73 79.4 0.70 Medium Rank C
IL Illinois #6 hvAb 4.19 35.15 15.86 3.98 12,125 0.72 90.3 0.69 Medium Rank C
MO Rowe hvAb 2.38 35.04 17.25 6.93 12,310 0.71 81.1 0.68 Medium Rank C
UT Rock Canyon hvBb 3.38 37.18 12.42 0.55 12,505 0.71 79.9 0.68 Medium Rank C
PA Lower Clarion hvAb 2.3 36.93 12.64 4.52 13,157 0.71 69.1 0.68 Medium Rank C
NM York Canyon hvAb 1.45 36.78 11.42 0.46 13,467 0.71 85 0.68 Medium Rank C
UT Lower Sunnyside hvBb 4.81 39.63 8.35 0.9 13,411 0.71 89.1 0.68 Medium Rank C
PA Brookville hvAb 2.5 36.13 17.9 4.7 11,857 0.71 77.8 0.68 Medium Rank C
1. BP Statistical Review of World Energy 2007. London, England, British 2. Classification of Coals. ISO 11760:2005. Geneva, Switzerland: Interna-
Petroleum Company, pages 32–37, 2007. tional Standards Organization, ISO 11760:2005, “Classification of Coals”,
Geneva, Switzerland, 2005.
A-38
BIBLIOGRAPHY
V. F. Parry, “Production, Classification, and Utilization of Western United “International Classification of In-Seam Coals”, Committee on Sustain-
States Coals,” Economic Geology, 45: 515–532, 1950. able Energy, United Nations Report, Energy, 19, New York and Geneva,
Sept., 1998.
W.A. Selvig and P.H. Gibson, “Analyses of Ash From United States Coals,”
U. S. Bureau of Mines Bulletin 567. “World Energy Council Survey of Energy Resources”, WEC, London, 2007.
Washington, DC: U.S. Government Printing Office, 1956. ASTM Standard D388-05, “Classification of Coal by Rank”, 2005, ASTM
International, West Conshohocken, PA, 2005.
R.F., Abernethy and E.M. Cochrane, “Free-Swelling and Grindability
Indexes of United States Coals,” U.S. Bureau of Mines Information Circular “Analysis Directories of Canadian Coals: Supplement No. 6,” Canadian
8025. Washington, DC: U.S. Government Printing Office, 1961. Department of Enegy, Mines, and Resources, Canada Centre for Energy
and Mineral Technology, Ottawa, Canada, 1985.
R F. Abernethy and E.M. Cochrane, “Fusibility of Ash of United States
Coals,” U.S. Bureau of Mines Information Circular 7923., Washington, DC: Penn State University Coal Sample Bank and Database. Available at
U. S. Government Printing Office, 1960. [http://datamine.ei.psu.edu/index.php]. Accessed Pennsylvania State
8/4/2004.University, College of Earth and Mineral Science, “PSU Coal
“Analyses of Tipple and Delivered Samples of Coal,” Reports of Investigations
Sample Database.”, 8-4-2004.
of the Bureau of Mines, U. S. Dept of the Interior; Washington, DC: U. S.
Government Printing Office, Washington, D.C. The reports comprise
analytical data showing composition and quality of coal samples collected
by the Bureau for nearly every year since 1948.
A-39
Properties of Coal Ash
Appendix B
As mined, coal contains varying extent that they interact with the other net effect on the boiler during a given
amounts of mineral matter. When minerals to produce low-melting-point period of time will be directly propor-
the coal is burned, the mineral matter compounds that create slagging and tional to the amount of coal ash that
is converted into an incombustible fouling problems. is fired in the furnace. On a theoretical
residue known as “ash.” This mineral basis, completely deashed coal could
matter complicates the utilization of Ash in Coal: Design Factors be burned in a furnace designed for oil
coal. It has very little heating value, if The management of coal ash is one of or gas firing. Without ash, all furnaces
any. Yet because it comes with the coal, the major considerations in the design could be designed on the basis of heat
ash carries a freight cost that must be of a coal-fired steam generator. The transfer only.
paid. It also must be removed from behavior of the mineral matter in coal, Predicting ash behavior is still an
the furnace as fast as it enters in order as it influences furnace wall slagging experienced-based endeavor. In the
to avoid buildup. Provisions must be during the combustion process, is a sig- following sections, various methods
made for the handling and disposal of nificant factor in terms of volume, plan of predicting the likely impacts of ash
ash, which adds to the cost of use. area, and fuel-burning zone. Mineral on furnace behavior will be presented,
The mineral matter of the coal matter formed as tenacious deposits from the fairly simple chemical ratios to
cannot be completely separated out on waterwall surfaces insulates the the various correlations concerning ash
by any physical method in order to heat-transfer portions of the furnace. properties. These values should serve as
identify the individual minerals fully. Thus, the ash properties influence the guides to the likely impacts of coal ash
Consequently, to assess the amounts disposition of both radiant and convec- on furnace operations. All such infor-
of the major elements present in the tive heat-transfer surfaces as well as the mation is of vital concern to engineers
coal, an analysis is made to determine number and location of soot-blowing making decisions about furnace size and
the percentages of the oxides of these devices. The behavior of the ash, as it heating surface arrangements.
elements that are present in the coal influences the fouling of convection
ash. This ash analysis does not indicate surfaces, affects the amount and spac- Composition of Coal Ash
the nature or the distribution of the ing of these heat-transfer surfaces and Coal ash consists almost entirely of
mineral matter. Rather, it provides the quantity of cleaning devices. metal oxides. The composition varies
information that is useful in the study The percentage of the mineral mat- over a wide range. There is no “typical”
and utilization of coal. Examination ter in the coal is significant because of ash analysis.
of thousands of coal ash analyses the amount of potential slagging and
indicates that in most cases, the oxides fouling it may cause, the burden that it Chemical Analyses
of silicon, aluminum, iron, calcium, places on the ash-handling equipment, Chemical analyses of coal ash provide
and magnesium account for 95% or the frequency and length of soot blow- data from which to estimate coal ash
more of the ash. Of the remaining ing, and the rate of wear on pulverizer and slag characteristics as well as to
components, the oxides of sodium and parts. Thus, for any given chemical and calculate various correlation param-
potassium are important only to the physical properties of coal ash, their eters. Analyses generally are performed
B-
B- | Clean Combustion Technologies
SiO2 + Al2O3 + Fe2O3 + CaO + MgO + Na2O + K2O + TiO2 + P2O5 + SO3 = 100%
Table B-1 | Coal Ash Chemical Because it usually is present in insignificant quantities, P2O5 sometimes is omit-
Composition ted. Trace quantities of many metals also can be present, but these are less important
for furnace design and more important for emissions requirements. Table B-1 gives
Oxide Component Percentage
ranges for coals found in North America.
SiO2 10–70
The chemical composition of coal ash prepared to the ASTM Standard is dif-
Al2O3 8–38 ferent from that of fly ash, bottom ash, or wall slag, because they occur in actual
Fe2O3 2–50 boiler furnaces. Short of building and operating a furnace, however, few adequate
CaO 0.5–30 alternatives have been established to generate actual furnace ash. Thus, coal ash
MgO 0.3–8 and slag behavior commonly are estimated from the reported chemical composi-
Na2O 0.1–8 tion of the ASTM-prepared ash.
K2O 0.1–3
Mineralogical Analyses
TiO2 0.4–3.5
Mineralogical analyses of coal ash attempt to identify the original forms of mineral
SO3 0.1–30 matter in coal. The thermal behavior of coal ash in an operating furnace largely
depends on the reactions between the ash-forming minerals. These are not readily
detected in a small-scale laboratory setting. Standardized procedures for extract-
ing the minerals from the coal matrix at low temperature in order to preserve the
Table B-2 | Typical Mineral
original material have not been developed. A low-temperature ashing technique
Species Found in Coal
can be carried out in an oxygen environment that gradually breaks down the coal
without excessive thermal decomposition of the inorganic mineral species. At least
Mineral
Formula 100 mineral species are associated with coal. Table B-2 lists the more abundant
Species
mineral forms.
Kaolinite Al2O3·2SiO2·H2O
Illite K2O·3Al2O3·6SiO2·2H2O Coal Ash Slagging and Deposition
Muscovite K2O·3Al2O3·6SiO2·2H2O A number of parameters are used for evaluating coal ash behavior, because they af-
Biotite K2O·MgO·Al2O3·3SiO2·H2O fect slagging and deposition on both the furnace walls and the convection surfaces.
Orthoclase K2O·Al2O3·6SiO3 Some of these parameters are:
Albite Na2O·Al2O3·6SiO2 • Ash fusibility temperature
Calcite CaCO3 • Base/acid ratio
Dolomite CaCO3·MgCO3 • Silica/alumina ratio
Siderite FeCO3 • Iron/calcium ratio
Pyrite FeS2 • Ferric percentage
• Alkali percentage
Gypsum CaSO4·2H2O
Quartz SiO2 There are other ratios and indices. Those cited above often are used to lend
Hematite Fe2O3 insight regarding the potential for slagging and fouling. They organize the ash
Magnetite Fe3O4 composition into building blocks for expected behavior. Slag refers to fused, or
resolidified, deposits from molten material. Slag forms primarily on the furnace
Rutile TiO2
walls and other surfaces predominantly exposed to radiant heat and high gas
Halite NaCl
temperatures. Deposits are bonded ash buildups that form primarily on convection
Sylvite KCl surfaces or other relatively cooler heat-transfer surfaces.
Properties of Coal Ash | B-
Ash fusibility temperature If ash arrives at a heat-absorbing sur- is tightly bonded and difficult to remove.
Ash fusibility has long been recognized face at a temperature near its softening Because this material builds up on itself,
as a tool for estimating the performance temperature, the resulting deposit its insulating properties continuously
of coals in relation to slagging and de- is likely to be porous in structure. increase the surface temperatures until
posit buildup. It is one of the most basic Depending upon the strength of the the fluid temperature is reached and
means for predicting ash performance. bond, it may fall off the metal surface slag runoff results. The temperature dif-
In general, high fusion temperatures because of its own weight, or it can ferential between the initial deformation
result in a low slagging potential in be removed by soot blowing. If such a and fluid temperatures gives an insight
dry-bottom furnaces. Likewise, low deposit is allowed to build up in a zone regarding the type of deposit formation
fusion temperatures are mandatory for of high gas temperature, its surface to expect on furnace tube surfaces. A
wet-bottom, or slag-tap, furnaces. Low (because of the insulating properties small temperature spread indicates that
ash fusion temperatures indicate a likeli- of the ash) can reach the melting point the wall slag will be thin. This type of
hood of slagging. and then run down the wall surfaces in slag is extremely difficult to control by
The ASTM Standard D1857 gives the furnace. If the ash particles arrive at soot blowing. As the range from the ini-
one experimental procedure for deter- the heat-transfer surfaces at tempera- tial deformation to the fluid temperature
mining the various melting temperatures tures below the softening temperature, increases, the resulting slag deposit will
of ash. The test is based on the gradual they will not form a bonded structure build up to thicker proportions before
deformation of a pyramid-shaped ash but, instead, will settle out as dust. the surface becomes sufficiently liquid
sample. Mounted on a refractory This dust is relatively easy to remove. in order to run. The tube-to-ash bond
substrate, the sample is heated at a pre- On the other hand, if the ash particles is less adhesive and, therefore, responds
scribed rate in a gas or electric furnace. have been subjected to temperatures better to removal by soot blowing.
A controlled atmosphere is maintained higher than the softening temperature
inside the furnace. During the heating for enough time to become plastic or liq- Base/Acid Ratio
process, changes in the shape of the uid, the resulting deposit will be a coarse, The acidic oxide constituents of SiO2,
pyramid are observed. Automated cam- fused mass condensed on the cooler Al2O3, and TiO2 generally are consid-
eras record the progress of the heating metal surface. The resolidified material ered to produce high melting
(temperature) and the shape of the pyra-
mid. The following four characteristic Table B-3 | Properties of Coal Ash Components
deformation temperatures are reported: Element Oxide Melting Temp. Acid or Base Compound Melting Temp.
3,120°F 1,610°F
1. Initial Deformation Temperature Si SiO2 Acid Na2SiO3
(1,720°C) (880°C)
(IT). The temperature at which 3,710°F 1,790°F
the tip of the pyramid begins to Al Al2O3 Acid K2SiO3
(2,040°C) (980°C)
show evidence of deformation. 3,340°F 2,010°F
Ti TiO2 Acid Al2O3·Na2O·6SiO2
2. Softening Temperature (ST). The (1,840°C) (1,100°C)
temperature at which the ash 2,850°F 2,100°F
Fe Fe2O3 Basic Al2O3·Na2O·6SiO2
sample has fused into a spherical (1,570°C) (1,150°C)
shape in which the height is equal 4,570°F 2,090°F
Ca CaO Basic FeSiO3
to the width at the base. (2,520°C) (1,140°C)
3. Hemispherical Temperature (HT). Mg MgO
5,070°F
Basic CaO·Fe2O3
2,280°F
The temperature at which the (2,800°C) (1,250°C)
sample has fused into a hemispher- 2,330°F
2,535°F
Na Na2O (1,280°C) Basic CaO·MgO·2SiO2
ical shape in which the height is (1,390°C)
(sub)*
equal to half the width at the base.
660°F (d)† 2,804°F
4. Fluid Temperature (FT). The tem- K K2O
(350°C)
Basic CaSiO3
(1,540°C)
perature at which the ash sample
*sublimates
has fused down to a nearly flat †decomposes
layer, or puddle.
B- | Clean Combustion Technologies
Because most elements combine with each other according to their acidic or
basic properties, the base/acid ratio reflects a potential for ash-containing metals
to combine during the combustion process to produce low-melting salts. Extremes
at either end of the base/acid ratio indicate a minimum potential for forming
combinations with low fusibility temperatures. Note that the alkali metals, so-
dium and potassium, are exceptions, in that they form low-melting ash regardless
of the base/acid combinations. Their presence in an ash composition at sufficient
concentrations significantly alters the base/acid influences.
The base/acid ratio is an indication of the fusion characteristics of a coal ash,
but it should not be used as the sole criterion of evaluation. Other influences
should be considered as well. It is a valuable supplement to the measured ash fus-
ibility data. With most ash, a base/acid ratio is the range of 0.4 to 0.7 manifests
low ash fusibility temperatures and higher slagging potential.
Silica/Alumina Ratio
The silica/alumina ratio (S/A) is defined as
Iron/Calcium Ratio
This iron/calcium ratio (I/C) is defined as
ratios of between 10 and 0.2 have a marked effect on lowering the fusibility tempera-
tures of coal ash. Extreme effects are evident between ratios of 3 and 0.3.
Ferric Percentage
The ferric percentage (FP) looks at the ferric oxide (Fe2O3) compared to the
equivalent Fe2O3. The equivalent Fe2O3 is defined as
These terms were introduced when it was observed that the ash fusion tempera-
tures and slag crystallization temperatures were lower in reducing atmospheres
than in oxidizing atmospheres. It was concluded that FeO and Fe are stronger
fluxing agents than Fe2O3 and result in depression of the ash fusion temperature.
The equivalent Fe2O3 and FP depend upon combustion conditions and usually
are determined by chemical analysis of actual furnace slags. High percentages of
Fe2O3 indicate stronger oxidizing conditions.
Alkali Percentage
Sodium and potassium metals can join with other elements in combinations that
produce low fusibility temperatures. The influence of alkalies on fusibility and slag-
ging potential in the furnace is proportional to the quantity of alkalies, or the alkali
percentage, in the ash. With sodium-containing coals, the rate of buildup and the
ability to control deposits effectively are related to the amount of sodium present.
This is particularly true of coals with lignitic-type ash (low iron, high alkali, and
high alkaline earth), as found in U.S. coals west of the Mississippi River. The ash of
such coals can contain as much as 40% CaO and from 1 to 15% Na2O, as extremes.
Sodium oxide vaporizes at 2,330°F (1,280°C). A widespread distribution of sodium
vapor in the combustion gases ensures contact with ash particles and metal surfaces
as the gas cools. The particle surfaces are “conditioned” for bonding by the condensa-
tion of the sodium, and the resulting deposit forms rapidly. Because the buildup is
rapid, tube spacing in convective passes must be maintained on wide centers in order
to minimize bridging. Alkali levels of more than a few percent can cause problems.
but in most cases, kmetal >> kd, and k reduces to the thermal conductivity of the
deposit, kd.
The properties of slag deposits that are useful in performing calculations, such
as those of furnace heat absorption and ash-hopper cooling water flow, include the
deposit emissivity, thermal conductivity, and specific heat.
Characterizing the deposit properties is quite difficult in the laboratory, be-
cause deposit properties vary with coal type and with combustion environment,
allowing the formation of a variety of deposits that range from porous, friable
structures to dense, glassy-like structures. Modeling the deposit is quite complex,
but a comprehensive review of the theory can be found in Zbogar and colleagues.1
The uncertainties in predicting accurate deposit properties lead the boiler
designer to use a surface effectiveness factor (SEF) instead of a deposit thermal
resistance. The SEF multiplies the actual heat-transfer area and “reduces” the avail-
able heat-transfer area according to cleanliness. A clean surface has an SEF of one,
and dirty surfaces have an SEF of less than one.
Thermal conductivity
The thermal conductivity of a deposit is strongly affected by its structure: the coal
ash particles and condensed mineral matter that make up the deposit, the gases
that reside in the interstitial spaces, the porosity distribution, and the sintering
state. Heat conduction is mainly through the solid material, because that conduc-
tivity is much larger than the gas conductivity. In Figure B-2, thermal conductivity
data collected by Rezaei and colleagues2 are shown.
As a deposit grows, the insulating nature of the deposit material increases the
deposit surface temperature and changes its thermal properties. Figure B-2 shows
Properties of Coal Ash | B-
Thermal Conductivity,
very difficult to estimate in practice. 12
BTU/Ft2 - Hr - °F
Intervals
The curves of Figure B-3 show 10
thermal conductivities as a function SURFACE EMISSIVITY 8
6
of temperature. Separate curves are Equation B-6 shows that emissivity 4
given for fused and particulate states, plays a key role in heat transfer through 2
because thermal conductivity differs deposit layers. The emissivity or emittance 0
500 1,000 1,500 2,000 2,500
with the physical structure of the (e) is defined as the ratio of the actual Temperature, °F
deposit. The given conductivities are emission to the emission from a black Fused Deposit Particulate Deposit
averaged values of composite data body (b) at the same temperature (T):
obtained from Australian, Soviet, Figure B-3
Clean| Combustion
Thermal conductivity of fused
Technologies
E(T ) and particulate coal ash deposits
and U.S. investigations. In the figure, (T ) = Figure: B.03/ (JG)
Eb (T ) PMS5415 K
the particulate deposit is a fine, dry
Equation B-8
powder that forms an uncompacted
layer on the metal surface. The outer Generally, the emittance varies as a
deposit layers vary considerably in ratio of the emissive power emitting at where
character, from being porous or sin- a wavelength and in a direction to that
tered to being dense, fused, or molten. of a black body. Practical calculations a = total absorbance
The conductivity, which will vary with average the spectral or wavelength de- r = total reflectance
the physical properties of the specific pendence and directional dependence t = total transmittance
deposit, is expected to increase as to obtain radiatively gray values3 and
the deposit density increases and the are denoted total emittance. When the flame is in steady state
deposit becomes more molten. An energy balance at the deposit sur- with the surface, Kirchoff ’s law applies
Practical boiler calculations employ face requires that all radiation incident (a = e), and
an SEF that adjusts the available heat- is absorbed, reflected, or transmitted:
transfer area based on cleanliness. This
+ r + t=1 e + r + t=1
obviates detailed models of deposit
Equation B-9 Equation B-10
B- | Clean Combustion Technologies
For opaque deposits, t = 0, and The experimental determination integrating their respective spectral
of the emissivity and absorptivity of emissivities and their absorptivities:
e=1 – r coal ash deposits is quite difficult and
involves complex instrumentation and ∞
Equation B-11
careful laboratory procedure that gener- ∫ (T )e
l bldl
The deposit structure and its ally are not suited to the field. Early (T ) = 0
∞
emissivity depend upon the surface investigations by Agababov,5 by Becker,6
temperature of the deposit. In Figure and by Mulcahy, Boow and Goard4 ∫e
0
bldl
B-4, local trends of emissivity are assumed that deposits were gray bodies.
shown. On Curve A, deposit emit- This assumes that the spectral emissivi- Equation B-12
tance decreases up to the sintering ties and absorptivities, el and l, did
∞ ∞
point, past which both sintered ash not vary with wavelength, l.
and fused ash have higher emittance. Later investigations by Khrustalev,7 ∫ (T )e
0
l bldl ∫ (T )e
0
l bldl
Below the sintering temperature, Krustalev and Rakov,8 and Smith (T ) = ∞
= ∞
and Glicksman9 proved the gray body
the included pores scatter or reflect
energy. As the deposit sinters, the assumption to be inaccurate and ∫e
0
bldl ∫e
0
bldl
Surface Emissivity
0.8
layer, and outer deposit) as well as
by parent coal source (eastern and 0.7
western United States). On the basis
of the limited data, it appears that 0.6
for initial deposits and inner layers, Vertical Bars Represent 95% Confidence Intervals
0.5
western U.S. coal ash emissivities and 600 1,000 1,400 1,800 2,200 2,600
Temperature, °F
absorptivities tend to be significantly
below those of eastern U.S. coal ash.
Figure B-6 | Surface emissivity ofClean
coal ash deposits Technologies
Combustion
This tendency is expected on the
Figure: B.06/ (JG)
basis of chemical composition; that is, PMS5415 K
materials composed mainly of oxides Table B-4 | Summary of Emissivities and Absorptivities
of iron, magnesium, and silicon tend of Ash Deposits
to be better emitters and absorbers
Physical state Parent Coal Region Emissivity Absorptivity
than those containing high sodium or
calcium.6 Initial deposit Eastern U.S. 0.75–0.76 0.71–0.76
As the physical state of the deposit Initial deposit Western U.S. 0.37–0.56 0.52–0.61
tends toward a molten condition, emis- Inner layer Eastern U.S. 0.79–0.93 0.67–0.84
sivity and absorptivity approach 1.0, Inner layer Western U.S. 0.68 0.62
and differences between eastern and Outer deposit Western U.S. 0.66–0.87 0.62–0.82
western U.S. coals would be expected
Molten inner layer Western U.S. 0.90 0.91
to become less significant. The data
also indicate that molten deposits have
higher emissivities and absorptivities compared with those of powdery deposits.
It can be generalized that the emissivity of the ash deposit of a given physical
state decreases with increasing surface temperature throughout a wide tempera-
ture range.
Equation B-14
B-10 | Clean Combustion Technologies
where
Cp = specific heat (cal/g °C)
t = temperature (°C)
S = SiO2 + Al2O3 + FeO + MgO + MnO
For the temperature range of 1,350 to 1,600°F:
CaO
C p = (0.15 × 10 −2 )t − (0.478 × 10 −6 )t 2 − 0.876 + 0.016(1 − )
∑
Equation B-15
where
Cp = specific heat (cal/g °C)
t = temperature (°C)
S = SiO2 + Al2O3 + FeO + MgO + MnO
Foerster and Weston13 showed that their experimentally measured values
for liquid and solid slags agreed well with the calculated values from Vosko-
boinikov’s equation. Experimental values for liquid and solid slags were slightly
higher than the calculated values. The densities of various forms of coal ash slag
deposits were averaged from several investigations and are listed in Table B-6.
SURFACE TENSION
Table B-6 | Density of
Surface tension forces are involved in many important stages of the slag deposition
Various Coal Ash Slag
process, such as ash fusion, slag adhesion, and slag flow. The methods for estimat-
Deposits
ing surface tension of oxide systems have been developed in the ceramic and glass
Density (lb/ft3
Deposit Form industry; they should be applicable, to some extent, to coal ash slags because of the
[g/cm3])
similarity in their chemical compositions.
Liquid slag 150–180 (2.4–2.9)
Surface tension of liquid slags can be estimated according to the following
Dense, solid slag 140–170 (2.2–2.7) procedure:
Solid ash 120–160 (1.9–2.6)
1. Surface tension of liquid slags at 2,550°F (1,400°C) in air can be approxi-
Loose, powder ash 15–35 (0.24–0.56)
mated from the chemical composition using the following equation
Fine, ground slag 60–90 (0.96–1.4)
developed by Lyon14:
Table B-7 | Surface Tension l = 3.24SiO2 + 5.85Al2O3 + 4.4Fe2O3 + 4.92CaO + 5.49MgO + 1.12Na2O – 0.75K2O
of Slags in Different
Equation B-16
Atmospheres
Reduction in where l = surface tension of slag (dyne/cm)
Atmosphere
Surface Tension (%)* 2. To estimate surface tension of slags at different temperatures, a temperature
Nitrogen 0 coefficient of –0.017 dyne/cm°C is recommended by Parmelee and Harman.15
Hydrogen 1 3. To estimate surface tension of slags in different atmospheres, Table B-7 has
Air 3 been derived using data from Parikh.16
Carbon dioxide 7 In general, the surface tension of slags does not vary greatly and is within
Sulfur dioxide 27 the range of 300 to 400 dyne/cm at 2,550°F (1,400°C) in air. A lower slag
Water vapor 35 surface tension usually will lead to more rapid fusion and flow properties and
*Based on nitrogen atmosphere.
better wetting characteristics. Good wetting is a necessary, but not sufficient,
condition for stronger adhesive properties.
Properties of Coal Ash | B-11
METHODS OF ESTIMATING
ASH-FUSIBILITY TEMPERATURES 1,800 1 SiO2
2
Many investigators have attempted to 3
Hemispherical Temperature, °C
4 Curve Al2O3 + TiO2
calculate characteristic fusion tem- 5 1 1.20
1,600 6
peratures of coal ash from its chemical 2 1.39
composition. Most of the methods 3 1.52
4 1.69
require assumptions that tend to 1,400
5 1.82
oversimplify the composition of the ash. 7 6 1.99
8 9
Also, a common inadequacy among 7 2.52
1,200
many of the relations is that their valid- 0 5 10 15 20 25 30
8 3.00
ity usually is limited to a certain range % CaO + MgO Content 9 3.51
Hemispherical Temperature, °F
applicability of the correlations. It is very 2,800
Equation B-18
unlikely that the calculation of ash- 2,600
where
softening temperatures will replace the
determination. Nevertheless, using cor- t2 = softening temperature 2,400
Method of Kovitskii, Karagodina, For t2: 1.8 ≤ Kfu ≤ 9.9 Figure B-8 | Combustion
Hemispherical temperature of
Clean Technologies
and Martynova For t3: 1.8 ≤ Kfu ≤ 7.5 coal ashFigure:
as a function of
B.08/ (JG)Rm
The method Kovitskii, Karagodina, PMS5415 K
And Martynova,17 which assumes
Soviet Standard experimental condi- Method of Majumdar, Banerjee, and Lahiri
tions, is used to estimate the softening The method of Majumdar, Banerjee, and Lahiri, which assumes British Standard
temperature (t2) and the temperature experimental conditions, is used to estimate the hemispherical temperature in a
of the beginning fused state (t3) in a reducing atmosphere of ash from Indian coals. The procedure is as follows:
reducing atmosphere of ash from Rus-
1. Calculate Rm:
sian coals. The procedure is as follows:
Rm = (3.3SiO2 + 1.96Al2O3)/(2.5Fe2O3 + 3.57CaO + 5.0MgO + 3.22Na2O + 3.22K2O)
1. Calculate Kfu, which is expressed as Equation B-19
SiO2 +Al 2O3 2. From Figure B-8, estimate the hemispherical temperature of coal ash.
Kfu =
Fe 2O3 +CaO+MgO
This relationship was independently tested on 75 ash samples from U.S. coals. The
Equation B-17
difference between the corresponding estimated and measured values is as follows:
2. Estimate the softening tempera-
ture and the temperature of the 1. In 66.6% of cases, the estimated value is within ±100°F (±56°C) of the
beginning of liquid fused state measured value.
from the following relations: 2. The average difference between the two corresponding values is ±84°F (±47°C).
B-12 | Clean Combustion Technologies
The original report describing this relation is not available. The curve in Figure
B-8 was derived from calculations carried out on ash from eastern and western U.S.
coals. The original relation was said to be based on molecular percentages and not on
weight percentages. The independent tests, however, indicated that the divergence
of the estimated values from the corresponding measured values was much smaller
when weight percentages were used.
2,600 1 60
The viscosity of coal ash slag in the complete liquid phase can be measured or estimat-
10 50
ed at any temperature above its crystallization temperature. Above this temperature
Poise, at 2,600 °F
2,400
Liquid Viscosity
where
= slag viscosity (poise)
T = temperature (K)
Properties of Coal Ash | B-13
The Hoy, Roberts, and Wilkins relationship was based on the following range of
coal ash chemical composition:
SiO2 31–59%
Al2O3 19–37%
Equiv. Fe2O3 0–38%
CaO 1–37%
MgO 1–12%
Na2O + K2O 1–6%
SP 42–75%
SiO2/Al2O3 1.2–2.3
Method of Watt and Fereday. This viscosity of coal ash slags from British coals
in the complete liquid phase can be estimated according to the method of Watt
and Fereday19:
1. Normalize the chemical composition of the ash such that
10 7 × M
log10 = +C
(T − 150)2
Equation B-25
where
= slag viscosity (poise)
T = temperature (°C)
The Watt and Fereday relationship was based on the following range of coal ash
chemical composition:
SiO2 29–56%
Al2O3 15–31%
Equiv. Fe2O3 2–28%
CaO 2–27%
MgO 1–8%
Na2O + K2O 1.5–5%
SP 40–81%
SiO2/Al2O3 1.4–2.4
B-14 | Clean Combustion Technologies
Flow Temperature
The flow temperature is defined as the temperature at which slag has sufficient flu-
idity to allow flow without difficulty. Normally, the flow temperature corresponds
to a slag viscosity of approximately 80 poise. For coal ash slags of known chemical
composition, the flow temperature can be calculated by letting = 80 poise and
then obtaining the corresponding T80 value (the temperature at which the viscosity
is 80 poise). It is used to predict the ease of slag removal in wet-bottom furnaces.
It also can be used to estimate the maximum steady-state thickness of furnace-wall
deposits, because no accumulation of slag is expected beyond this point.
When estimating slag viscosity in the complete liquid phase, one should consider
the existence of the crystallization temperature, which is referred to as the tempera-
ture of critical viscosity (Tcv). The relation between T80 and Tcv is as follows:
If T80 > Tcv, then flow temperature = T80
If T80 < Tcv, then flow temperature = Tcv
ash-fluid temperature (FT). For a coal ash slag with a ferric content of about
2,600
10%, an approximate relation is expressed as
2,400
Flow temperature = 1.2(FT – 470)
2,200
Equation B-26
2,000
0 0.2 0.4 0.6 0.8 1.0
where flow temperature is in degrees-Fahrenheit.
Base/Acid Fe2O3 + CaO + MgO + Na2O + K2O
=
Ratio SiO2 + Al2O3 + TiO2
Slag Removal Temperature
2,600 80 1,427 (1,430°C). For dry-bottom units, high values of T250 are recommended for easier
2,500 70 1,371
removal (soot blowing) of waterwall deposits.
2,400 60 1,316
SiO2 When estimating slag viscosity in the complete liquid phase, the existence of
2,300 Al2O3 1,260
50 the crystallization temperature should be considered. Tcv, which is referred to as
2,200 1,204
2,100 1,149
the temperature of critical viscosity, is found as follows:
2,000 1,093
If T250 > Tcv , then slag-removal temperature = T250
10 20 30 40 50 60 70 80 If T250 < Tcv , then slag-removal temperature = Tcv
% Base
Method of Sage and Mcilroy.The T250 for slags from eastern U.S. coals can
Figure Clean
B-11 Combustion of T250 according
| EstimationTechnologies
Figure: B.11/ (JG) percentage and
to dolomite be estimated from the curves in Figures B-10 and B-11. The method of
PMS5415 K
percentage base Sage and McIlroy20 is said to be valid for bituminous-type ash and for
Properties of Coal Ash | B-15
lignitic ash with acidic content of range (with Kv between 0.42 and 2.03): at which this almost discontinuous
more than 60%. SiO2 20.9–63.1% transition occurs is referred to as the
Al2O3 14.1–29.0% temperature of critical viscosity, or Tcv ,
Temperature of Normal Slag Removal. Equiv. Fe2O3 3.2–36.3% as illustrated in Figure B-12. The Tcv is
The temperature of normal slag remov- CaO 1.2–27.3% believed to be the temperature at which
al, Tns, is defined as the recommended MgO 0.8–7.6% solid phases begin to crystallize and
temperature for easy slag-tapping from P2O5 0.1–2.9% separate out from the liquid; hence, it
a furnace. Usually, it corresponds to a can be considered as a crystallization
slag viscosity of 200 poise. This concept Effect of Equivalent Fe 2 O 3 and Ferric temperature. Actual measurement of
has been widely used in Russia and is Percentage. It has been observed that Tcv is expensive and difficult. Several
identical, in principle, to the U.S. crite- coal ash slag viscosity in the com- methods of estimating Tcv from the
rion of the slag-removal temperature. plete liquid phase can differ between chemical composition have been devel-
For coal ash slags of known chemical reducing and oxidizing atmospheric oped and are described below.
composition, the Tns can be calculated conditions. This difference implies the The Tcv, then, provides the limiting
from the relations described above by effect of variation in ferric percent- temperature below which slag viscosities
letting = 200 and obtaining the cor- age. Usually, however, the difference in cannot be calculated using the methods
responding T200 value (the temperature viscosity is small. described earlier. It also allows the pre-
at which the viscosity is 200 poise). Nicholls and Reid21 initially diction of the boundary temperature at
Alternatively, it can be estimated by the observed that the entire slag-viscosity which molten slag ceases to flow on fur-
following method. profile was raised as the ferric percent- nace waterwall deposits. Typical cooling
When estimating slag viscosity in age decreased. Sage and McIlroy20 curves for four different coal ash slags are
the complete liquid phase, the existence indicated the opposite; that is, the shown in Figure B-13. The explanation
of the crystallization temperature viscosity profile was lowered as the fer- of the four curves is as follows:
should be considered. The relationship ric percentage decreased. In either case,
Curve 1. Represents a true glass,
between T200 and Tcv the temperature the effect of ferric percentage on slag
which does not have a distinct
of critical viscosity, is as follows: viscosity in the complete liquid phase
crystallization temperature. Very few
seems to be small and, usually, can be
If T200 > Tcv , thenTns = T200 coal ash slags are of this type.
neglected. Accordingly, the original
If T200 < Tcv , thenTns = Tcv Curve 2. Represents a slag that is
Fe2O3 content of the ash can be sub-
close to glass in behavior. Separa-
stituted for the equivalent Fe2O3 when
tion of solids begins after a long
According to the method developed estimating slag viscosities according
by Kovitskii, Karagodina, and Mar- to the methods described above. To ob-
tynova,17 the T200 for coal ash slags tain results that reflect actual furnace
from Russian coals can be estimated by conditions more closely, it is suggested
Solid
the following procedure: that estimations be carried out at a Phase
Freezing Temperature
ferric percentage of 20%.
1. Calculate Kv, which is expressed as Tfr
cooling period. Tcv is usually very low. Slags of this type frequently are referred
Freezing to Solid to as “long” slags.
Critical Viscosity Curve 3. Represents a slag with a long freezing range. Separation of solids
Freezing Temperature
begins at high temperatures at low rates, with freezing to essentially a solid at
100,000 1 23 4 much lower temperatures. Tcv is usually high.
Curve 4. Represents a slag with rapid crystallization and freezing characteristics. Tcv
10,000 can be either high or low. Slags of this type frequently are referred to as “short” slags.
Viscosity, Poise
1,000
Method of Sage and McIlroy. The Tcv (°F) of coal ash slags from eastern U.S. coals at
a ferric percentage of 20% can be estimated using the method of Sage and McIlroy20
100
with the following equation:
10
1,600 1,900 2,200 2,500 2,800 Tcv = HT + 200
Temperature, °F
Equation B-29
Figure Clean
B-13 Combustion
| Typical cooling curves for
Technologies
Figure: B.13/ (JG)types of coal ash slag
different where HT is the hemispherical temperature of coal ash (°F).
PMS5415 K
Method of Watt and Fereday. The Tcv of coal ash slags from British coals can
be estimated according to the method of Watt and Fereday19 using the
following procedure:
1. Normalize the chemical composition of the ash, such that
Equation B-30
to = 0.75t2 + 480
Semi-Molten Deposit
Furnace-Wall Tube
Equation B-32
Molten Deposit
Initial Deposit
Solid Deposit
Heat Flow
where t2 is the softening temperature of coal ash according to the Soviet standard (°C).
Kalmanovitch and Frank24 have clean waterwall tubes, because upon surface temperatures allows the adher-
described a prediction methodology approaching the relatively cooler tube ence of rigid plastic secondary deposits.
using thermodynamic equilibrium be- surface, the slag particles become less
tween the mineral phases to calculate adhesive as a result of rapid cooling MECHANISM OF COAL
the mixture viscosity. in the wall-adjacent area. Accordingly, ASH SLAG DEPOSITION
coal ash deposition generally is consid- The following two types of primary
COAL ASH SLAG DEPOSIT ered to be a two-stage process. First, a deposits are most commonly observed:
ACCUMULATION AND STRUCTURE primary layer of deposit forms on the
Molten to semi-molten coal ash slag waterwall tube surface. Subsequently, 1. Primary deposits that result from
deposits usually will not form on the resulting rise in the surrounding the settling of the finer fractions
Radiant Wall
Deposits
Reflective 960-1,320 °C
Ash
Flame 650-980 °C
Radiation
Ash
Buildup
0.1-5
microns
Coal + Air 1,100-1,650 °C Coal + Air
Early Combustion
Initial Coal Products
Gas
Phase
Minerals
ns
cro
Rapidly Receding
O– Na+
mi
C Char Surfaces
OH Slag
76
Solids
O– Ca++OH Inorganic
C
OH Droplets
Carbon and Inorganic 0.1-5
Fly Ash microns
Components (Liquids)
(<30 microns) of fly ash. This type of primary deposit is loose in structure
and does not provide strong cohesive or adhesive bonds.
2. Primary deposits that result from the selective deposition of certain reac-
tive components of the ash (iron, calcium, or alkalies). These components
can be present in the deposit in high concentrations as oxides and/or sulfur
compounds, leading to the formation of low-melting eutectic mixtures. This
type of primary deposit is more dense in structure and has stronger cohesive
and adhesive bonds. The properties of the primary layer, which provides the
link between the rigid secondary deposits and the tubes, have an appreciable
influence on the ease of removal for the deposits by soot blowing. During the
deposition process, a transitional stage occurs during which the plastic sec-
ondary deposits begin to stick on the primary layer. These secondary deposits
are strengthened by time and increasing temperatures.
The transition from primary to secondary deposits has been shown to be a func-
tion of the immediate gas temperature outside the deposit. The temperature at
which the secondary deposits begin to form has been referred to as the initial slag-
ging temperature, tis. Methods of estimating tis are described below.
Equation B-33
Equation B-34
Clean Coal
Selected Waste
Coals
Anthracite Lignite
Bituminous Sub-Bituminous
Bench-Scale Full-Scale
Testing Boiler
Pilot-Scale
Boiler
Data Evaluation
and Correlation
Computer
Expert
Model
temperature because of the insulating composition and distribution of the of the model has been presented by
properties of the ash (their low relative mineral inclusions from Computer Zygarlicke and McCollor.33
thermal conductivity). The high deposit Controlled Scanning Electron Micro-
temperature increases both the amount scope (CCSEM) and the amount and A second set of tools is designed to
of liquid in the deposit and in the type of organically bound minerals in aid the user in interpreting the labora-
ash-sticking potential. Depending upon the coal. These models predict the size tory data from a specific coal samples:
the temperature, a variety of deposits distribution and composition of the
may form, ranging from porous, friable fly ash, the viscosity of the fly ash per PARTCHAR©. This software uses
structures to dense, glassy-like struc- the size ranges and temperature, and the size, area, perimeter, and chemi-
tures. These deposit structures can the deposited fraction of the fly ash. cal composition. PARTCHAR
protect heating surfaces, if controlled manipulates raw CCSEM data to
by soot blowing, or they can impede LEADER©, Low-Temperature classify mineral grains based on
heat transfer and adversely affect boiler Algorithm for Deposition Risk. This elemental composition and size.
performance, if left uncontrolled. algorithm predicts deposition of
The mechanisms controlling depo- entrained ash in the lower-tem- MINCLASS©. This classification
sition change throughout the boiler, perature convective region of a utility program uses the quantitative chem-
making modeling of the processes boiler.27 The software also predicts a ical data to classify data into mineral
difficult. In Figure B-15, pictorials are growth and shedding index.29 phases. The mineral phase criteria
given to represent possible scenarios were based on phase stoichiometry
in regions of the boiler. ADLVIC,30 Ash Deposit Local and laboratory experience.
Traditionally, boiler designers have Viscosity Index of Refraction and
characterized slagging and fouling de- Composition. This model from Sandia VISCALC©. This is a viscosity
posit potential using empirical indices. Laboratories makes quantitative pre- prediction, thermodynamic equilib-
Recent years have seen an emphasis dictions of the elemental composition rium program and runs in parallel
on computer technology to catalogue of ash deposits and deposition rates. with MINCLASS. The program
slagging and fouling experience and to uses either mineral phases that are
refine predictive capability. The newest PHOEBE©, Phase Ordering and silicon-rich or silicon mineral phases
and most exciting of these tools are the Equilibrium Evaluation. This thermo- as input to calculate viscosity.24
computer-based models. These tools chemical equilibrium calculation
are designed to provide plant designers package is tailored to large, com- More recent predictors include the
and operators information regarding plex, multi-phase systems found in Coal Quality Impact Model (CQIM),
ash behavior as a strategy to improve pulverized coal-fired systems.31 The which is a sophisticated model devel-
plant operability. software has the capability to evaluate oped by the Electric Power Research
Specific tools that have been blends of coals. Institute.34 It is available commercially
reported in the literature are designed and is one element of a comprehensive
for predicting the particle-size and PCQUEST©, Predictive Coal Quality Coal Quality Expert (CQE) system.
composition distributions of ash slag- Effects Screening Tool. This model uses Inputs to the model include all design
ging or fouling. These tools include: detailed mineralogical and chemical data for the plant components, includ-
fractionation analyses in addition to ing schematics of various systems,
ATRAN, an Ash Transformations the conventional bulk measurements electrical motor capacities, and design
Model. This model predicts particle- of coal moisture, principal elements, points for operation of the compo-
size and composition distributions of and ash oxides.32 Eight indices are nents, based on a design coal or a test
ash, slagging, or fouling deposition.27 calculated by PCQUEST to predict coal. Input data can be very detailed
furnace wall slagging, high-tem- and specific to a particular plant or
MIT, Mineral Transformation Code. perature fouling, low-temperature only a minimum, in which case default
This collection of models predicts a fouling, slag tapping, opacity, tube values are used. The software includes
relative fouling tendency (RFT).28 erosion, coal grindability, and soot aspects of pulverization, fuel burnout,
The input includes size and chemical blower requirements. A validation ash deposition, heat transfer, NOx
Properties of Coal Ash | B-21
emissions, and particulate emissions. Computational fluid dynamics transform into char, and then burn to
Evaluators claim that generally, the (CFD) has been used as a platform to ash. As particles impact the boiler walls,
model is very useful, but it also is develop and apply deposition models a predetermined, temperature-depen-
cumbersome to use (as is gathering the for slagging and fouling.35,36 In other dent sticking coefficient is used, and the
input information). studies, the commercial code Fluent deposition potential is inferred by ac-
The model (see Figure B-16) was has been used to model slagging and cruing only that fraction of the sticking
developed by jointly by an industrial fouling with user-defined routines mass of the reacting coal particle. For
(including Alstom) and government that model the impact of particles on example, all of the particle sticks above
partnership (“Software systems in a surface and the turbulence and ther- Tfluid, and none of the particle sticks
Clean Coal Demonstration Projects,” mophoretic enhancement of deposition below Tdeformation (Figure B-17).
DOE Topical Report 17, 2001). The (see, e.g., Fluent News, “Simulating Slag- Typical CFD results are shown in
CQE software is composed of various ging Films in Coal Combustors,” Fall Figure B-18, with high mass fluxes
technical models (derivatives of models 1999). For boiler designers, CFD tech- colored in red and low fluxes in blue.
developed over the last decade and nology, although new and promising, These CFD studies have been
described above) that evaluate perfor- needs further development of the basic useful for determining qualitative
mance issues; environmental models physics and the constitutive relations. differences in wall accretion for
that examine emissions and regula- Today, CFD is an indicative tool for use variations in burner tilt in Separated
tory issues, and financial models that in boiler studies to assist with solving Overfire Air (SOFA) and non-SOFA
relate performance to costs. Advanced operational deposition problems. As applications. Similar studies have
computer algorithms, such as neural the verification and validation of such
network analysis, are used to optimize models is achieved, deposition modeling
results for unit-wide decision making: will find a place in the design process. 1
operating strategies, fuel blending, and Alstom routinely uses the Fluent
soot blower cycles. code to model coal combustion for new Sticking
Coefficient
The CQE software uses the afore- and retrofit boiler applications. For coal
mentioned CQIM model as well as combustion, fuel is injected into the 0
modules for NOX prediction, such as boiler, and the particle phase is mod- Tdeformation Tfluid
NOxPERT, Acid Rain Advisor (ARA), eled using the discrete phase model,
and CQE Boiler Expert, which consists which tracks particle trajectories as Figure B-17 | Example of sticking coefficient
of a slagging module (SLAGGO) and a the coal is modeled to release volatiles, versus particle temperature
fouling module (FOULER): Clean Combustion Technologies
Figure: App B.17/ (GM)
SLAGGO. This module simulates PMS5415 K
the ash formation cycle, includ-
ing deposit initiation, growth, and Deposition
(kg/m^2-s)
removal, based on coal proper- 1.00e-02
7.36e-03
5.41e-03
ties, boiler design, and operating 3.98e-03
2.93e-03
2.15e-03
parameters. SLAGGO has internal 1.58e-03
1.17e-03
8.58e-04
6.31e-04
modules for ash formation, ash 4.64e-04
3.41e-04
2.51e-04
transport, deposit growth, thermal 1.85e-04
1.36e-04
1.00e-04
7.36e-05
properties, and removal models. 5.41e-05
3.98e-05
2.93e-05
2.15e-05
1.58e-05
1.17e-05
8.58e-06
FOULER. This module predicts 6.31e-06
4.64e-06
3.41e-06
convective pass fouling based on 2.51e-06
1.85e-06
1.36e-06
boiler design, temperature, and gas 1.00e-06
Logarithmic
distributions. The time intervals Scale
1. A. Zbogar, F. J. Frandsen, P. A. Jensen, and P. Glarborg, P, “Heat Transfer 18. H. R. Hoy, A. G. Roberts, and D. M. Wilkins, “Behavior of Mineral
in Ash Deposits: A Modeling Toolbox,” Progress in Energy and Combus- Matter in Slagging Gasification Processes,” Institution of Gas Engineers
tion Science, 31:371–421, 2005. Journal, 5(6):444–469, 1965.
2. H. Rezaei, et al., “Thermal Conductivity of Coal Ash and Slags and 19. J. D. Watt and F. Fereday, “The Flow Properties of Slags Formed from
Models Used,” Fuel, 79:1697–1710, 2000. the Ashes of British Coals: Part 1. Viscosity of Homogeneous Liquid
Slags in Relation to Slag Composition,” Journal of the Institute of Fuel,
3. F. Incropera and D. Dewitt, Introduction to Heat Transfer. Hoboken, NJ:
42(338):99–103, March, 1969.
John Wiley and Sons, 1985.
20. W. L. Sage and J. B. McIlroy, “Relationship of Coal Ash Viscosity to
4. M. F. R. Mulcahy, J. Boow, and P. R. C. Goard, “Fireside Deposits and
Chemical Composition,” Combustion, 31(5):41–48, 1959.
Their Effect on Heat Transfer in a Pulverized-Fuel-Fired Boiler, Part
1: The Radiant Emittance and Effective Thermal Conductance of the 21. P. Nicholls and W. T. Reid, “Viscosity of Coal-Ash Slags,” Transactions
Deposits,” Institute of Fuel Journal, 39:385–394, 1966. of the ASME, 63:141–153, 1940.
5. S. G. Agabobov, “Emissivity of Boiler Furnace Slag,” Teploenergetika, 22. Y. L. Marshak and A. V. Ryzhakov, “Flow of Slag Film When Affected
8:56–60, 1958. by Gas Flow,” Thermal Engineering (Translation of Teploenergetika),
16(101):12–16, 1969.
6. H. Becker, “The Effect of Ash Fouling on Heat Transfer to Boiler Tubes, Part
2: Thermal Properties of Particulate Brown Coal Ash,” Report No. 252, State 23. W. T. Reid and P. Cohen, “The Flow Characteristics of Coal-Ash Slags
Electricity Commission of Victoria–Scientific Division, Australia, 1972. in the Solidification Range,” Transactions of the ASME, 66:83–97, 1944.
7. B. A. Khrustalev, “Spectral Radiation Properties of Some Materials at 24. D. P. Kalmanovitch and M. Frank, “An Effective Model of Viscosity
High Temperatures and Their Influence on the Integral Absorption and for Ash Deposition Phenomena,” in Proceedings of Mineral Matter and
Radiation Properties,” Heat Transfer–Soviet Research, 5(2):60–64, 1973. Ash Deposition from Coal (R. W. Bryers and K. S. Vorres, eds.). Santa
Barbara, CA: Engineering Foundation, 1988, pp, 89–101.
8. B. A. Khrustalev and A. M. Rakov, “Methods of Determining the Integral
and Spectral Radiative Properties of Materials at High Temperatures,” 25. A. N. Alekhnovich, V. V. Bogomolov, N. V. Novitskii, and N. I. Ivanova,
Heat Transfer–Soviet Research, 1(4):163–177, 1969. “A Study of the Slagging Properties of Kuznetsk Coal Ash,” Soviet Power
Engineering (Translation of Elektricheskie Stantsii), 6:509–513, 1977.
9. R. A. Smith and L. R. Glicksman, “Radiation Properties of Slag,” Report
No. DSR-70332-61. Cambridge, MA: Massachusetts Institute of Tech- 26. E. P. Dik, and R. A. Sikhora, “Study of the Conditions of Formation
nology, 1969. of Ash Deposits When Burning Nazarova Coal,” Thermal Engineering
(Translation of Teploenergetika), 16(10):23–27, 1969.
10. G. J. Goetz, et al., “Development of a Method for Determining Emissivi-
ties and Absorptives of Coal-Ash Deposits,” ASME Paper 78-WA/FU-6. 27. S. Benson, J. Hurley, C. Zygarlicke, E. Steadman, and T. Erickson, “Pre-
New York: American Society of Mechanical Engineers, 1978. dicting Ash Behavior in Utility Boilers: Assessment of Current Status,”
Presented at EPRI Coal Quality Conference, La Jolla, CA, August, 1992.
11. N. B. Vargaftik and O. N. Oleschuck, “Thermal Capacity of Slags of
Various Fuels,” Teploenergetika, 2(4):13–17, 1955. 28. J. Beer, A. Sarofim, and L. Barta, “From Coal Mineral Matter Properties
to Fly Ash Deposition Tendencies: A Modeling Route,” Proceedings of
12. V. G. Vaskoboinikov, “The Heat Capacity of Blast-Furnace Slags at
the Engineering Foundation, Inorganic Transformation and Fly Ash
High Temperatures,” Teoriya Probtiba Metallurgii, 12(10):3–5, 1940.
Deposition During Combustion (S. Benson, ed.), New York, ASME,
13. E. F. Foerster and P. L. Weston, Jr., “Heat Content of Some Blast-Fur- March 10–15, 1992, pp 81–94.
nace and Synthetic Slags,” U.S. Bureau of Mines Report of Investigation
29. J. Williamson, The Impact of Ash Deposition on Coal Fired Plants. New
6886. Washington, D.C.: U.S. Bureau of Mines, 1967.
York: Taylor and Francis, 1994.
14. K. C. Lyon, “Calculation of Surface Tension of Glasses,” Journal of the
30. L. L. Baxter, “Ash Deposition During Biomass and Coal Combustion: A
American Ceramic Society, 27:186–189, 1944.
Mechanistic Approach,” Biomass and Bioenergy, 4(2):85–102, 1993.
15. C. W. Parmelee and C. G. Harman, “Effect of Alumina on Surface
31. M. Ramanathan, S. Ness, and D. Kalnamovitch, “New Techniques for
Tension of Molten Glass,” Journal of the American Ceramic Society,
Thermophysical Phase Equilibrium Predictions in Coal Ash Systems,”
20:224–230,1937.
Symposium on Ash Deposition: Mineral Definition and Characteriza-
16. N. M. Parikh, “Effect of Atmosphere on Surface Tension of Glass,” tion, American Chemical Society, Spring Symposium, Dallas, TX,
Journal of the American Ceramic Society, 41:18–22, 1958. Volume 34, Issue 2, 1989, pp. 311–317.
17. N. V. Kovitskii, N. V. Karagodina, and M. 1. Martynova, “Investigation 32. S. Benson, E. Steadman, C. Zygarlicke, and T. Erickson, “Ash Forma-
of the Influence of the Chemical Composition of the Ash of Power tion, Deposition and Erosion in Conventional Boilers,” in Applications of
Coals in Its Fusibility and Viscosity,” Solid Fuel Chemistry (Translation Advanced Technology to Ash-Related Problems in Boilers (L. Baxter and R.
of Khimiya Tverdogo Topliva), 9(3):59–63,1975. DeSollar, eds.). New York: Plenum Press, 1996, pp. 1–15.
B-23
33. C. Zygarlicke and D. McCollor, “Validation of Fireside Performance Lignites,” Presented at the Sixth Conference on Technologies and Com-
Indices,” EERC Report for the Department of Energy, 1997. bustion for a Clean Environment, Porto, Portugal, July 9–12, 2001.
34. J. U. Pavilsh, P. R. Miller, N. C. Craig, and A. K. Mehta, “CQIM—An 38. B. Epple, R. Leithner, and W. Linzer, Numerical Simulation of Power
Analytical Tool Used to Evaluate Performance and Economic Issues,” Plants and Energy Systems. Berlin: Springer, 2008 (in German).
Presented at the Pittsburgh Coal Conference, Pittsburgh, PA, Oct.
39. Z. Ma, F. Inman, P. Lu, R. Sears, K. Lingbu, A. Rokanuzzamann, D.
12–16, 1992.
McCollor, and S. Benson, “A Comprehensive Slagging and Fouling Pre-
35. F. Lockwood and F. Lee, “Modeling Ash Deposition in Pulverized diction Tool for Coal-Fired Boilers and Its Validation and Application,”
Coal–Fired Applications,” Progress in Energy and Combustion Science, Fuel Processing Technology, 88:1035–1043, 2007.
25:117–132, 1999.
40. A. Tortosa-Masiá, F. Ahnert, H. Spliethoff, J. C. Loux, and K. Hein,
36. U. Schnell, S. Richter, and K. R. G. Hein, “Numerical Simulation of “Slagging and Fouling in Biomass Combustion” Thermal Science,
Slagging and Fouling in a Pulverized Coal–Fired Utility Boiler,” Proceed- 9(3):85–98, 2005.
ings of UEF Conference on Heat Exchanger Fouling, Davos, Swizerland,
41. T. Heinzel, et al., Evaluation of Biomass Combustion in a Cyclone Slag Tap
July 8–13, 2001, pp. 98–105.
Furnace. Stuttgart: IVD Stuttgart, 1998.
37. B. Epple, B. Krohmer, A. Hoppe, H. Müller, and R. Leithner, “CRFD
Studies for Boilers Fired with High Ash Containing and Slagging
B-24
SUGGESTED READINGS
S. Allan, T. Erickson, and D. McCollar, “Modeling of Ash Deposition in and Metallurgical Investigations, Carnegie Institute of Technology, Mining
the Convective Pass of a Coal-Fired boiler,” in Applications of Advanced and Metallurgical Advisory Boards, Cooperative Bulletin 62, 1934.
Technology to Ash Related Problems in Boilers (L. Baxter and R. DeSollar,
T. G. Estep, H. Seltz, and W. J. Osborn, “Determination of the Effect of
eds.). New York, Plenum Press, 1996, pp. 451–470.
Oxides of Sodium, Calcium, and Magnesium on the Ash Fusion Tem-
S. Benson, J. Hurley, C. Zygarlicke, E. Steadman, and T. Erickson, “Predicting perature by the Use of Synthetic Coal Ash,” In Mining and Metallurgical
Ash Behavior in Utility Boilers,” Energy and Fuels, 7:746–754, 1993. Investigations, Carnegie Institute of Technology, Mining and Metallurgical
Advisory Boards, Cooperative Bulletin 74, 1937.
B. K. Dutta, B. Rai, and K. R. Chakravorty, “Fusion Characterization of
Coal Ashes: Softening Temperature of Coal Ashes in Oxidizing Atmo- A. W. Gauger, Progress During the First 12 Years of Cooperative Research on
sphere,” Journal of Scientific and Industrial Research, 19B:206–212, 1960. Bituminous Coals, Report from Mineral Industries Exp. Station, Pennsyl-
vania State College, 1951.
B. K. Dutta, et al., “Study of the Fusion Characteristics of Coal Ashes: Soft-
ening Temperature of Coal Ashes in Mild Reducing Atmosphere,” Journal R. L. Schaefer, The Relation Between the Chemical Constituents and the Fus-
of Scientific and Industrial Research, 21D:44–48, 1962. ibility of Coal Ash. Columbus, OH: Ohio State University, 1933.
A. F. Duzy, “Fusibility–Viscosity of Lignite-Type Ash,” ASME Paper 65- G. Wilemski, S. Shrinivasachar, and A. Sarofim, “Modeling of Mineral Mat-
WA/FU-7. New York: American Society of Mechanical Engineers, 1965. ter Redistribution and Ash Formation in Pulverized Coal Combustion,”
in Inorganic Transformation and Ash Deposition During Combustion (S.
T. Erickson, S. Allan, D. McCollar, J. Hurley, S. Srinivasachar, S. Kang,
A. Benson, ed.). ASME Engineering Foundation Press, New York, NY,
J. Baker, M. Morgan, S. Johnson, and R. Borio, “Modeling, Fouling
1992, pp. 545–562.
and Slagging in Coal-Fired Utility Boilers,” Fuel Processing Technology,
44:166–171, 1995 M. Xu, X. He, J. L. T. Azevedo, and M. G. Carvalho, “An Advanced Model
to Assess Fouling and Slagging in a Coal-Fired Boiler,” International
T G. Estep, H. Seltz, H. L. Bunker, Jr., and H. S. Strickler, “The Effect of
Journal of Energy Research, 26:1221–1236, 2002.
Mixing Coals on the Ash Fusion Temperature of the Mixture,” In Mining
B-25
Combustion Calculations
by Graphical Methods
Appendix C
This appendix presents a unique 4. Products of combustion requirements and products of combus-
method of accomplishing detailed 5. Moisture in the combustion air tion of each constituent, gross errors
calculations of combustion products of 6. Moisture from fuel may result if, say, the heating value used
many fuels. The appendix supplements 7. Dry gas content of the combus- is not correct for, and correspondent to,
information from Chapter 2 on the tion products the given analysis.
production, transportation, and physi- 8. Combustion products The method employed here is based
cal condition of these fuels as they are on the concept that the weight of air
received in a steam-generating facility. In the method outlined here, the required in the combustion of a unit
Many of these factors (particularly tem- first four items are needed for the cal- weight of any commercial fuel is more
perature and water content) influence culation of gas and air quantities. Items nearly proportional to the unit heat
the calculation of air and gas weights, as five to seven form the basis of heat bal- value than to the unit weight of that fuel.
well as the boiler efficiency itself. ance calculations in either the design Consequently, the weights of air, dry gas,
or the testing of a steam generating moisture, wet gas, etc., are expressed in
THE GENERAL METHOD unit, when the ASME Performance lb/106 BTU (kg/GJ) fired. In the case
FOR ALL FUELS Test Codes, Short Form, is used. The of solid fuels, it is difficult to burn 100%
Combustion calculations by the designer carbon dioxide in the last item, and its of the combustible, so a correction for
are one of the first steps in proportion- relationship to excess air, is important solid combustible loss must be made.
ing a steam boiler. For example, the gas in combustion calculations, because Before considering in detail the
weight that results from fuel combustion in test work the CO2 in the gases is method and the sample calculations, a
is required to determine proper arrange- measured and from it, the excess air quick review of the combustion process
ment and extent of heating surface in is calculated. The equipment designer of a fuel is necessary. Take, for example,
the furnace, boiler, economizer, and air works with excess air even though it a fuel with no ash. When this fuel burns
heater. The gas weight is also neces- may appear to have been based on completely, the weight of the fuel is sim-
sary for proportioning gas ducts, dust estimates on the percentage of CO2. ply added to the weight of atmospheric
collection equipment, emission control The method outlined here is air supplied for its combustion to obtain
equipment, induced draft fans, and complementary to the presentation in the wet products of combustion, or
stacks. Air weight is used to set the size Chapter 2 and can be easily adapted total wet gas. Thus: fuel (F) + air (A) =
of air heaters, air ducts, fuel burning to desktop analysis with a PC. The products (P). If some of the fuel is ash,
equipment, and forced draft fans. charts make allowance for all the or, if because of incomplete combustion,
This appendix will discuss the fol- important variables in the analysis. For some of the fuel does not pass out of
lowing items: estimating purposes, therefore, they are the furnace with the gases, then F in the
as accurate as more laborious methods above equation will be less, and both A
1. Fuel in products of calculation. On the other hand, even and P will also be reduced. This concept
2. Atmospheric air for combustion in working with a complete ultimate of adding F and A to get P is the basis
3. Effect of unburned combustible analysis and calculating the combustion of the method.
C-
C- | Clean Combustion Technologies
FUEL IN PRODUCTS, F
By definition, fuel in products, F, is that portion of the fuel fired that reappears in
gaseous form in the products of combustion, as separate elements or in chemi-
cal combination with other elements. Since with the method of this appendix
all quantities are to be those required for or resulting from firing 106 BTU
(1.055GJ), F must also be calculated on that basis. If a fuel contains no ash and
if, in addition, it deposits no carbon in the furnace or on other heating surfaces,
F is simply obtained by dividing 106 by the “as-fired” heating value of the fuel in
BTU/lb ( J/kg). For the other cases where ash and/or solid combustible loss must
be considered, Equation C-1A may be used.
where:
F = lb/106 BTU (kg/GJ)
% ash = percentage by weight in fuel as fired
% solid combustible loss = percentage by weight in fuel as fired
fuel heating value (higher = higher heating value as fired, BTU/lb ( J/kg)
heating value [HHV])
Figure C-1 is a graphical solution of Equation C-1A, and applies to all fuels.
In the United States and many other countries of the world, the HHV is the
accepted standard and is obtained by
HHV, BTU/lb as Fired calorimetric analysis of the fuel in a
10,000 20,000 30,000 laboratory. Producers often sell fuel
1,000 100
900 90 on the basis of its heating value and
800 80 users generally check it periodically.
700 70 In any event, manufacturers of steam
600 60 generating equipment need to know
it. If it is not furnished, they must
500 50
make an independent test or calculate
it from constituents. For solid and
F, lb/106 BTU as Fired
400 40
Ash Sediment and Solid Combustible liquid fuels, empirical formulas must
Loss % by Weight as Fired be used, but for gaseous fuels the sum
300 30
0 of the heating values of the various
5
combustible constituents may be em-
10
15 ployed. Considerable difference may
200 20
180 20 18
exist between the analysis “as received”
160
25 and “as fired,” even though nothing is
16
30 intentionally done to the fuel between
140 14
120
the time it is received and the time it
12
is burned. On the contrary, certain
100 10 fuels are purposely dried before they
1,000 2,000 4,000 6,000 8,000 10,000
HHV, BTU/lb as Fired are fired and others take on moisture
hygroscopically, or are purposely
Figure C-1 | Weight of fuel in products of combustion
Clean Combustion Technologies
Figure: C.01/ (JG)
PMS5415 K
Combustion Calculations by Graphical Methods | C-
“tempered.” In the storage system of pulverized coal firing, the moisture removed
from the coal may actually be vented directly to a stack. In the direct-fired
system, the moisture, although removed from the coal, is fired with the coal and,
therefore, does not truly alter the combustion quantities.
In some cases the lower heating value (LHV) may be reported. The HHV may
be calculated from the following:
The lower heating value is denoted by LHV, and the difference in the higher- and
lower-heating value is the latent energy of vaporization. The parameters, w and hfg,
are the fraction of moisture derived from the fuel and the heat of vaporization. At
77°F (25°C), the latent heat of vaporization is 1,050.4 BTU/lb (2443 kJ/kg).
The heating value and analysis of gaseous fuels are generally reported on the
volumetric basis. Therefore, the temperature and the pressure conditions of the
fuel are an essential part of the analysis. Standard conditions are commonly stated
as 60°F (16°C) and 30 in. Hg (1.0 bar).
ATMOSPHERIC AIR, A
All combustion requires oxygen, which in commercial practice must be supplied
from the atmosphere. The theoretical quantity of oxygen may be calculated ac-
curately from the fuel analysis and simple reaction formulas, such as
C + O2 = CO2
H2 + ½O2 = H2O
S + O2 = SO2
Equation C-2A
The corresponding weight or volume of dry air required to supply this oxygen may
then be calculated, knowing that air contains approximately 23.14% oxygen by weight
(20.9% by volume). Throughout this appendix, the notation labeled “A” gives the rela-
tion of air quantity to percentage excess air calculated for the several different fuels.
The curves, or the theoretical estimate of A, are based on “atmospheric air,” which
means that an arbitrary amount of moisture has been added to the air. The American
Boiler Manufacturers Association (ABMA) suggests a moisture level of 0.013 lbs of
water vapor per lb of dry air (60% humidity at 80°F [27°C]). (To neglect this would
introduce an error of approximately 1%.) Thus, the theoretical air can be calculated
from the methods of Chapter 2 with the assumptions that 23.14% of the dry air is
oxygen and there is moist air with 0.013 Kg H2O/Kg dry air.
f f f f MO2 (1 + w)10 4
Awet = C + H − O + S − C fo
MC 4 MH MO MS aO2 hhhv
Equation C-2B
The values (f) are a percentage from the ultimate analysis. The Ms are
the molecular weights. The parameters aO2 , w, and hhhv are the fraction of
C- | Clean Combustion Technologies
oxygen in the dry air, the humidity ratio, and the HHV, and the values stated above the theoretical air requirement
respectively. With the molecular weights of Chapter 2, follows as
10 4
Awet = (11.68%C + 35.03%(H − %O / 8) + 4.38%S) •
h hhv
Equation C-2C
% Eair
Aactual = Awet 1 +
100
Equation C-2D
Natural and manufactured gaseous fuels contain becomes a time-consuming task for which shorter meth-
varying proportions of several different combustible ods are desirable if accuracy can, at the same time,
constituents, such as C, H2, S, and various hydrocarbons. be maintained.
Some also contain inert constituents, such as ash, The fuel analysis is usually reported on a volume basis.
N2, H2O, and CO2, as well as oxygen in varying quanti- The total weight of the fuel, considering all species in the
ties. Obviously, determination of air for combustion then fuel, is
7 9
Moles = 2(CH4 ) + (C2 H 6 ) + 5(C3 H8 ) + ... + (C2 H 6 ) + .....
2 2
Equation C-2F
EFFECT OF UNBURNED COMBUSTIBLE The method used to determine the solid combustible loss
consists of collecting fly ash and refuse from various hoppers.
With a well-designed burner and furnace, it is commer- From a dry sample burned in a muffle furnace, weight loss
cially possible to burn liquid and gaseous fuels completely. due to combustion is measured. This loss can be expressed
No allowance ordinarily is made for loss of combustible as percentage by weight of fuel as fired.
in gas and air weight calculations. In the combustion of In the combustion of a fuel having pure carbon as the only
solid fuels, even in pulverized form, it is not commercially combustible constituent, the air required may be accurately
feasible to burn all the available combustible material in obtained by multiplying the excess air (e.g., Curve A reading)
the fuel. by the factor
In this case, air for combustion is MOISTURE IN AIR, Wa zero or a trace in lampblack to 100
directly proportional to carbon in the Since the ASME Performance Test lb/106 BTU (43 kg/GJ) fired in the
fuel. If, however, all heat in the fuel Codes require that heat loss due case of some refinery gases.
does not come from carbon alone (so to moisture in air be reported as a
that the air is not strictly proportional separate item in the test heat balance 10 4 lbH O
to carbon burned), the factor C will of a steam generating unit, engineers Wc = fH2 O • 6 2
hhhv 10 BTU
not be exact. For high-carbon, low- customarily include it in a predicted
volatile fuels, it will be nearly exact and heat balance. In an actual test, the Equation C-7
will result in only a small error, even moisture in air can be determined
for fuels low in fixed carbon and high from wet- and dry-bulb temperature Fossil fuels will also yield water from
in hydrogen. The error involved by readings. For a predicted heat balance the burning of the hydrogen content.
using Equation C-3 in all cases is quite it must be assumed. It may vary from It is possible to determine this water,
within the limits of accuracy of all day to day even in the same locality, Wh, in lb/106 BTU (kg/GJ), from the
other combustion calculations. but the ABMA suggested value of percentage of hydrogen and HHV of
For a heat balance, the combustible 0.013 lb of water per lb of dry air is the fuel.
weight loss must be converted to per- included in the atmospheric air, A, as
centage of heat loss. This can be done read from the air-weight curve. When fH 10 4 lbH2O
Wh = MH2 O = 9 f H •
conveniently by dividing the percentage required as a separate item for heat 2MH hhhv 10 6 BTU
of solid combustible weight loss by the balance calculations, an assumed value
Equation C-8
heating value of the fuel as fired and from the following equation will be
multiplying by 14,500 BTU/lb (3.372 sufficiently accurate. Groups of curves or special charts,
107 J/kg), which is the heating value are used to calculate Wf for each fuel.
for combustible in the refuse recom- Wa = 0.013A
mended by ASME Performance Test DRY GAS, Pd
Equation C-5
Codes, PTC 4. The need to calculate the “dry gas”
or “dry products,” in addition to the
PRODUCTS OF COMBUSTION, P MOISTURE FROM FUEL, Wf
total products, is due to the ASME
Having calculated the foregoing quan- This is another item that is reported Performance Test Codes. Dry gas loss is
tifies F and A, the gaseous products of separately both in an ASME Perfor- a separate item of the heat balance.
combustion may readily be determined mance Test Codes heat balance and in The dry gas may be determined by
by the addition of F and A, as previ- a predicted heat balance. In the case subtracting the water vapor from the
ously corrected. Thus: of some fuels (such as natural and total products, thus
refinery gases), the heat loss due to
P = F + CA this moisture may be the largest single Pd = P – (Wa + Wf)
Equation C-4 item in the heat balance. Wf includes
Equation C-9
where: the combined surface and inherent
moisture, Wc from a fuel plus the
P = t otal gaseous products of moisture formed by the combustion of where:
combustion, lb/106 BTU fired hydrogen, Wh.
(kg/GJ) Pd = dry gas, lb/106 BTU fired
F = fuel fired exclusive of ash or Wf = Wc + Wh (kg/GJ)
solid carbon loss, lb/106 BTU Equation C-6 P = total products of combustion,
fired (kg/GJ) lb/106 BTU (kg/GJ)
A = atmospheric air required, Wc will vary from zero or a mere Wa = moisture in air, lb/106 BTU
lb/106 BTU fired (kg/GJ) trace in fuel oil to more than 115 fired (kg/GJ)
C = combustible loss correction lb/106 BTU (49.4 kg/GJ) fired in the Wf = moisture from fuel, lb/106
factor case of green wood. Wh will vary from BTU fired (kg/GJ)
C- | Clean Combustion Technologies
fC kg CO2
′ 2=
XCO MCO2 = 3.667 fC
MC kg fuel
f w
′ 2 O = H MH2 O = 9 f H + C
XH Aactual
2MH 1+ w
f
′ 2 = S MSO2 = 2 fS
XSO
MS
0.2314 Eair
′ 2 =C
XO Aactual
1+ w
0.7686
′ 2 = fN + C
XN Aactual
1+ w
Equation C-10
Values on a kg/kg product basis can be obtained by dividing the X′ by the sum
of the X′. For example, the CO2 composition follows as
′ 2
XCO kg CO2
XCO2 =
′ 2 + XH
XCO ′ 2 O + XSO
′ 2 + XO
′ 2 + XN
′ 2 kg mix
Equation C-11
Combustion Calculations by Graphical Methods | C-
′ 2P
XCO lbCO
ZCO2 = 2
′ 2 + XH
XCO ′ 2 O + XSO
′ 2 + XO
′ 2 + XN
′ 2 10 6 BTU
Equation C-12
Molar values are obtained by dividing the X′ by the molecular weights to obtain
the following for CO2.
′ 2
XCO
′ 2=
YCO
MCO2
′ 2
YCO MoleCO2
YCO2 =
′ 2 + YH′ 2 O + YSO
YCO ′ 2 + YN′ 2 + YO′ 2 Mole
Equation C-13
The other species (H2O, SO2, O2, N2) follow a similar calculation. If a dry basis
is desired, exclude the moisture from the denominator in the above.
COAL
Chapter 2 and Appendix A give the properties of the many types of coal found
throughout the world. Particularly relevant to the calculation of the combustion
products of coal are the methods of laboratory analysis described in Chapter 2,
and found in the ASTM Standards referenced there.
PROXIMATE ANALYSIS
This quick and simple method is used to determine the moisture, volatile matter,
fixed carbon, and ash of a given coal sample, by weight. The moisture content is
obtained by noting the loss of weight in the sample when it is heated to a tem-
perature of about 220°F (104°C) for an hour. The moisture determined includes
surface as well as inherent, or hygroscopic, moisture. The volatile matter is the
loss in weight, which follows additional heating of the sample to a temperature of 7
% H2 in Coal, H2O- and Ash-Free
about 1,750°F (955°C) for a half-hour, with all air excluded. The fixed carbon is 6 12,500 BTU and Up
calculated from the further reduction in weight of the sample, which takes place 5
when the coal is completely burned in the presence of air, while the ash is the 4
12,000 – 12,500 BTU
3
residue at the end of complete combustion.
2
Since this laboratory method is approximate in nature, it is well to remember
1
that combustion calculations, which are based on this method, may be consider-
0
ably in error. For example, it is necessary to know the hydrogen in the coal to 0 10 20 30 40 50 60
evaluate the moisture loss in a heat balance. Figure C-2 shows two curves, taken % VM, H2O- and Ash-Free
from DeBaufre’s original correlation,1 which relate the percentage of volatile mat-
ter in the proximate analysis to the percentage of hydrogen in the fuel. Figure C-2 | Determination
Clean of hydrogen,
Combustion Technologies
Actual values of hydrogen for certain coals will differ from these average Figure:moisture, and ash free (W. L.
C.02/ (JG)
PMS5415De Baufre,
K “Composition in Heating
curves by as much as 1.7%. This represents a variation of 12 lb/106 BTU fired (5.2 Value of Fuels,” Combustion,
kg/GJ) in the water formed by combustion of the hydrogen, which means that 2(11): 29–35, May 1931)
C- | Clean Combustion Technologies
the moisture loss estimated from known. It is possible to obtain a better are the principal reasons for preference
a proximate analysis may be 1.5% correlation between the percentage of of the ultimate over the proximate
greater, or smaller, than it should be, if CO2 in the dry products of combus- analysis. For purposes of reporting or
such loss is based on 500°F (260°C). tion and the ratio correlation, it is often convenient to ex-
This discrepancy alone may, then, be press proximate and ultimate analyses
sufficient to absorb an entire manufac C on a basis other than as fired.
turer’s margin and unaccounted for Thus, an analysis on the as-received
H 2 –0.1 O2
loss, commonly 1.5%. or as-sampled basis serves to define the
Equation C-14
The character of the volatile matter condition of the coal at the time it was
is not the same for all coals, because tested. As dry, the fuel is reported with
some contain more carbon dioxide for all coals, including lignite and sub- the moisture item omitted; as ash free,
than others. This fact accounts for the bituminous. This is shown in Figure it is without ash, and as moisture and
two distinct groups of points occurring C-4, where no CO2 value plotted devi- ash free, it has had both moisture and
on Figure C-3. A similar correlation of ates from the solid line by more than ash eliminated from the analysis.
the percent CO2 in the dry products 0.30%. A similar correlation to that
versus volatile matter in Figure C-3 in Figure C-4 can be established with HEATING VALUE
can be established with Table C-1 at Table C-1. It is customary to report the HHV
the end of this section. Thus, lignites Sometimes, instead of being re- of coal, in BTU/lb ( J/kg), along with
and sub-bituminous coals form, on ported as a separate item, the moisture the analysis, whether proximate or
the average, about 1% more CO2 than in the ultimate analysis is divided into ultimate, and on the same basis as the
bituminous coals of the same volatile eight parts oxygen and one part hydro- analysis. Burning a small sample of
content. The kind of hydrocarbon in gen and then added to the oxygen and coal in a bomb calorimeter immersed
the volatile also has a direct bearing on hydrogen of the combustible in the in water, and measuring the tem-
the position of points in Figure C-3, coal. For combustion calculations, this perature rise of the water, accurately
since different hydrocarbons generate is not a good practice, because in com- determines the HHV at constant
different amounts of carbon dioxide. puting a heat balance, the surface and volume. If an LHV is reported, use of
Plotting percentage of CO2 against the inherent moisture in the fuel require Equation C-1B is recommended.
ratio of fixed carbon to volatile matter different treatment from the water Various formulas have been pro-
does not help to line up the points. formed when hydrogen is burned. posed for evaluating the energy per
Since an ultimate analysis includes unit mass BTU/lb ( J/kg) of coal from
ULTIMATE ANALYSIS the percentage of hydrogen, the its proximate or ultimate analysis. All
As discussed in Chapter 2, it is pref- amount of water formed by burning yield only approximate results because
erable to use the ultimate analysis, if it can be accurately calculated. These they do not take into account the com-
plex thermodynamic changes occurring
when coal is burned. If only the coal
20 Lignites and 21 rank is known, it is best to assume the
Sub-Bituminous Coals
CO2 in Dry Products,
CO2 in Dry Products,
20
heating value from a reasonably close
% by Volume
% by Volume
*A is the air required for combustion under stoichiometric conditions (no excess air), with 0.013 lb H2O per lb dry air.
**Orchard Bed, ***Mammoth Bed, ****Holmes Bed, RANK KEY: Ma-Meta-anthracite, A-Anthracite, Sa-Semi-anthracite, Lvb-Low-Vol. Bituminious, Mvb-Med.-Vol. Bitumi-
nous, Hvab-High-Vol. Bituminous A, Hvcb-High-Vol. Bituminous B,Hvbb-High-Vol. Bituminous C, Sub-Subbituminous, LigA-Lignite A
C-10 | Clean Combustion Technologies
of the coal, the method of firing, and ATMOSPHERIC AIR, A the surface moisture, which produces
the furnace design and operation. With A theoretical calculation can be made “wetness” in the coal, and the inher-
proper furnace design and operation, with Equations C-2B and C-2C. A ent or hygroscopic moisture, which
the loss due to unburned volatile, as graphical estimate of the theoretical air being intimately held by the coal is not
determined by the presence of CO in is shown in Figures C-5 and C-6, which readily sensed. Table C-1 reveals that,
the flue gas, is generally negligible. On reveals that the maximum deviation in in some high-volatile bituminous coals,
the other hand, the loss due to solid the calculated weight of air from the re the total moisture may be as low as
combustible in the ash, as measured by spective average curve is less than ±2.5%. 1.4%, while it may increase to 34.8%
the reduction in weight that a sample Figure C-5 shows the variation of the and higher in lignite. The portion of
of ash refuse undergoes on being atmospheric air at zero excess air with Wc that is inherent moisture, likewise,
completely burned, may vary from a volatile matter on a moisture and ash- varies from negligible amounts in
negligible amount in burning pul free basis, while Figure C-6 indicates the super-anthracites to high values in
verized coal to 20% of the coal fired, or change in atmospheric air with the ratio, lignites. The percentage of moisture in
more, with certain sizes of anthracite. the fuel is shown in Figure C-7D or can
If K1, K2, etc., are known or [C/(H2 + 0.1 O2)] be computed with Equation C-7. The
assumed constants representing, moisture formed by combustion of the
respectively, the fractions of the The correlations of the data in Fig- hydrogen is shown in chart C of Figure
total ash in a pound of coal that is ures C-5 and C-6 are used to establish C-7 or can be computed with Equation
deposited in the ash hopper, as fly the values at 0% excess air and to C-8. If only a proximate analysis of the
ash, etc., W. S. Patterson2 has shown parameterize the chart A of Figure C-7, coal is available, its hydrogen content
that the total combustible loss may be using Table C-1. must first be read from Figure C-2.
expressed by Equation C-15.
MOISTURE IN FUEL Wf PRODUCT COMPOSITION
% solid combustible weight loss = The total moisture in the fuel, Wf, is With an ultimate analysis the product
C K C K the sum of moisture in the fuel, Wc and composition can be computed with
100 R1 1 + R 2 2 K
1– CR1 1– CR 2 that due to the hydrogen, Wh, Equation C-10.
If a graphical analysis is desired, the
× ash in coal as fired
Wf = Wc + Wh percentage of CO2 by volume in the dry
Equation C-15 Equation C-16 products may be read from Curves B,
Figure C-7, for excess-air values up to
where CR1, CR2, etc., are the respective The moisture Wc is usually reported 100% of theoretical requirements for
weights of combustible per pound of in an ultimate/proximate analysis as the ratio [C/(H2—0.1 O2)] from 12.5
dry refuse from siftings, ash hopper, fly shown in Table C-1. It is the sum of to 50. Figure C-4 was the basis for the
ash, etc. For a given coal burning unit,
the sum of all Ks is, then, equal to 1.
The results of Equation C-15 may 850 850
A, lb/106 BTU as Fired
pears in the products of combustion, F, Solid Line was Used for Correlating A and
VM for Curves on Figure C-7A Solid Line was Used for Curves on Figure C-7A
is taken directly from Figure C-1.
Figure C-5 | Variation
Clean Combustion of theoretical air with
Technologies Figure C-6 | Variation
Clean Combustionof theoretical air with
Technologies
Figure:volatile
C.05/ matter
(JG) (VM) Figure:C.06/C(JG)
PMS5415 K PMS5415 H2 +K0.1 O2
Combustion Calculations by Graphical Methods | C-11
where C, H2, O2 and S are weight fractions from the ultimate analysis.
COMBUSTIBLE LOSS
In burning coke or coke breeze in pulverized form or on stokers, a certain amount
of the combustible in the fuel is lost as siftings with the fly ash, etc., in the manner
C-12 | Clean Combustion Technologies
30
= 34% where CR1, CR2, etc., are the respective
1– 11.7 weights of combustible per pound of
100
dry refuse from siftings, ash hopper,
For this value of volatile matter and 20% excess air, read from Curves A Figure C-7, etc., and K1, K2, etc., are constants
A = 920 lb/106 BTU (396 kg/GJ).
representing fractions of the ash in a
3. Unburned Combustible Factor, C. Since the solid combustible loss is 0.8%, from
Equation C-3 obtain
pound of coke found, respectively, in
siftings, ash hopper, etc.
% solid combustible weight loss 0.8
C = 1− = 1 − = 0.992
100 100 FUEL IN PRODUCTS, F
Knowing the HHV of coke, its ash
4. Total Products, P. From Equation C-4, P = F + CA = 66 + 0.992 × 920 = 979 lb/106 BTU
(421 kg/GJ). content, and its combustible loss factor
5. Moisture in Air, Wa. From Equation C-5, Wa = 0.013 × 920 = 12 lb/106 BTU (5.2 kg/GJ). C, that portion of fuel, F, which reap-
6. Moisture from Fuel, Wf. This is the sum of Wc, the moisture in the coal as fired, and Wh, pears in the products of combustion, is
the water formed in combustion. From Curves D, Figure C-7, for 2.6% moisture in the fuel calculated from Equation C-1 or read
and a HHV of 13,610 BTU/lb (31.7 MJ/kg), read Wc = 2 lb/106 BTU (0.86 kg/GJ). from Figure C-1.
Next, convert the HHV as fired to a moisture-and ash-free basis, as follows:
ATMOSPHERIC AIR, A
13.610
= 15,400 BTU / lb Analyses of coke are reported on
1– 11.7
100 either the proximate or the ultimate
and use it, together with the moisture-and ash-free volatile determined for A, to read from
basis. The analysis of typical cokes
Figure C-2, H2 = 5.4% moisture and ash free, or reverting to the as-fired basis in Table C-2, lists the theoretical air
at zero excess using Equation C-2D.
11.7
1– 100 = 4.8% hydrogen Figure C-8 shows the empirical relation
between A and the volatile matter on
From Curves C, Figure C-7, for a HHV of 13,610 BTU/lb (31.7 MJ/kg) and H2 = 4.8%, read a moisture- and ash-free basis and
Wh = 32 lb/106 BTU (13.8 kg/GJ). Wf = Wc + Wh = 2 + 32 = 34 lb/106 BTU (14.6 kg/GJ). Figure C-9 indicates how A varies with
7. Dry Gas, Pd. From Equation C-9, Pd = P – (Wa + Wf) = 979 – (12 + 34) = 933 lb/106 the ratio C/H2. It will be seen that
BTU (401 kg/GJ).
no calculated point deviates from the
8. Percentage of CO2 in Products. For a moisture-and-ash-free volatile matter of 34% and
solid, average lines on these figures by
20% excess air, read from Curves B, Figure C-7, CO2 = 15.3%.
more than ±2.0%.
(Note: Many of the foregoing figures will be somewhat different if the ultimate, instead of the
The chart, Figure C-10, was prepared
proximate, analysis in line PA, Table C-1, is used.)
so that values of atmospheric air, A,
lb/106 BTU as fired may be obtained
after knowing either the moisture- and
ash-free volatile matter in the coke or
Combustion Calculations by Graphical Methods | C-13
EXAMPLE B
850
A, lb/106 BTU as Fired
With the Pennsylvania bituminous coal in the previous Example A and the ultimate analysis of
800
line 25, Table C-1, the coal is burned with 20% excess air and the expected solid combustible
loss is 0.8% by weight. Then
750
1. Fuel, F. The sum of the ash plus solid combustible loss is 9.1 + 0.8 = 9.9% by weight.
700 With this sum, and a HHV from line PA, Table C-1, of 13,610 BTU/lb (31.7 MJ/kg) read
0 4 8 12 16 20 from Figure C-1, F = 66 lb/106 BTU (28.4 kg/GJ).
% VM, H2O- and Ash-Free
2. Atmospheric Air, A. Calculate the ratio C/(H2 – 0.1 O2)
Solid Line was Used for Correlating A and
VM for Curves on Figure C-10A C/(H2 – 0.1 O2) = 76.6/(4.9 – 0.1·3.9) = 16.98
Figure C-8
Clean| Theoretical
Combustionair versus volatile
Technologies For the value of this ratio and 20% excess air, read from Curves A Figure C-7
Figure:matter
C.08/ (VM)
(JG) of coke A = 930 lb/106 BTU (400 kg/GJ).
PMS5415 K 3. Unburned Combustible Factor, C. Since the solid combustible loss is 0.8%, from Equa-
tion C-3 obtain
850
% solid combustible weight loss 0.8
A, lb/106 BTU as Fired
C = 1− = 1 − = 0.992
800 100 100
4. Total Products, P. From Equation C-4, P = F + CA = 66 + 0.992 × 930 = 989 lb/106 BTU
750 (425 kg/GJ).
5. Moisture in Air, Wa. From Equation C-5, Wa = 0.013 × 930 = 12.1 lb/106 BTU (5.24 kg/GJ).
700 6. Moisture from Fuel, Wf. This is the sum of Wc, the moisture in the coal as fired, and Wh,
0 100 200 300 400 500 600
C/H2 in Coke, Dry or as Fired the water formed in combustion. The ultimate analysis indicates a value of hydrogen 4.9%.
From Curves C, Figure C-7, with hydrogen 4.9% and an HHV of 13,610 BTU/lb (31.7 MJ/kg),
Solid Line was Used for Curves on Figure C-10A the Wh = 34 lb/106 BTU (14.6 kg/GJ). From Curves D, Figure C-7, for 2.6% moisture in the
fuel and a HHV of 13,610 BTU/lb (31.7 MJ/kg), read Wc = 2 lb/106 BTU (0.86 kg/GJ).
Figure C-9 | Theoretical Then Wf, = Wc,+ Wh, = 2 + 34 = 36 lb/106 BTU (15.5 kg/GJ).
Clean Combustionair versus C/H2
Technologies
Figure: C.09/ (JG) 7. Dry Gas, Pd. From Equation C-9, Pd = P – (Wa + Wf) = 989 – (12 + 36) = 941 lb/106
PMS5415
its C/H K solid lines from
ratio. The BTU (405 kg/GJ).
2
Figures C-8 and C-9 were used in label- 8. Percent CO2 in Products. For a ratio of C/(H2 – 0.1 O2) = 16.98 and 20% excess air, read
from Curves B, Figure C-7, CO2 is also approximately 15.3%.
ing Curves A of Figure C-10.
9. Combustion Products. With the Pennsylvania bituminous coal from line 25, Table C-1,
the combustion products can be computed using graphical analysis or analytically. Using
TOTAL PRODUCTS, P
Figure C-7, Chart B, for 20% excess air and [C/H2 – 0.1 O2] = 16.98, the CO2 is approxi-
The unburned combustible factor is mately 15.3%. The product composition using Equations C-2 through C-11 is
obtained from Equation C-3. This
XCO2 = 14.07%; XSO2 = 0.09%; XH2O = 7.65%; N2 = 75.02%; O2 = 3.17%
factor C is then used in correcting F,
the fuel in products, and A, the atmos The CO2 on a dry basis is 15.2%.
pheric air when determining the total
products P from Equation C-4.
MOISTURE IN FUEL, Wf which is this moisture in lb/106 BTU, nature. Knowing the HHV and the
The moisture reported in a coke may, therefore, be a small quantity percentage of moisture in the coke
analysis, just as for coal, is the loss in its if the quenching is done rapidly, or as fired, Wc may be calculated from
weight when dried at 220°F (104°C). may amount to 12 lb/106 BTU (5.2 Equation C-7 or read from Curves D,
Any moisture present in coke kg/GJ) or more in cases where storage Figure C-10.
results either from the quench- conditions enabled the coke to absorb The lb of water per 106 BTU fired,
ing process or outdoor storage. Wc, moisture, readily because of its porous Wh, formed by combustion of the
C-14 | Clean Combustion Technologies
EXAMPLE C
Assume that a high-temperature coke breeze, having the typical proximate analysis shown in Table C-2 is burned with 35% excess air and that the
expected combustible loss is 10% by weight. Then
1. Fuel, F. The sum of the ash plus solid combustible loss is 18.0 + 10.0 = 28.0% by weight. With this sum, and an HHV from Table C-2 of 10,200
BTU/lb (23.7 J/kg), read from Figure C-1, F = 70 lb/106 BTU (30.1 kg/GJ).
2. Atmospheric Air, A. The sum of the ash plus moisture in the coal is 18.0 + 12.0 = 30%, and the moisture-and ash-free volatile matter is
4.2
VMdaf = = 6%
30
1−
100
For this value of volatile matter and 35% excess air, read from Curves A, Figure C-10, A = 1,065 lb/106 BTU (458 kg/GJ).
3. Unburned Combustible Factor, C. Since the solid combustible loss is 10%, from Equation C-3 obtain
10
1 − % solid combustible loss = 1 − = 0.9
100
4. Total Products, P. From Equation C-4, P = F + CA = 70 + 0.90 × 1,065 = 1,028 lb/106 BTU (442 kg/GJ).
5. Moisture in Air, Wa. From Equation C-5, Wa, = 0.013A = 0.013 × 1,065 = 14 lb/106 BTU (6.0 kg/GJ).
6. Moisture from Fuel, Wf. This is the sum of Wc, the moisture in the coke as fired, and Wh, the water formed in combustion. From Curves D,
Figure C-10, for 12% moisture in the fuel and a HHV of 10,200 BTU/lb (23.7 J/kg) read Wc = 12 lb/106 BTU (5.2 kg/GJ).
Next, with 6% moisture- and ash-free volatile matter determined for A, read from Figure C-11 Wh = 10 lb/106 BTU (4.3 kg/GJ). Then
7. Dry Gas, Pd. From Equation C-9, Pd = P – (Wa + Wf) = 1,028 – (14 + 22) = 992 lb/106 BTU (427 kg/GJ).
8. Percentage of CO2 in Products. For a moisture- and ash-free volatile matter of 6% and 35% excess air, read from Curves B, Figure C-10, CO2 =
14.7%.
(Note: Many of the quantities calculated in the example will be different if the ultimate, instead of the proximate, analysis of coke breeze is used.)
Combustion Calculations by Graphical Methods | C-15
30
% by Volume
20
20
20
19
10 10
18
0 0 4 8 12 16 20
0 2 4 6 8 10 12 14 16 18 % VM, H2O- and Ash-Free
0 % VM, H2O- and Ash-Free
0 2 4 6 8 10 12 14 16 18
% VM, H2O- and Ash-Free Solid Lines were Used for Correlating
Solid Line was Used for Curves on Figure C-11 CO2 and VM for Curves on Figure C-10B
20
analyses are shown in Tables C-3 and C-4. moisture content in wood or bagasse
19 Moisture is the most variable single is, therefore, necessary. To obtain this,
item in the composition of wood and a weighed sample is placed in a steam
18 bagasse, its value ranging from 6% in a or electric oven and heated to 212°F
0 200 400 600 800
C kiln-dried wood to more than 60% in (100°C) until no further loss of
in Coke, Dry or as Fired
H2 – 0.1 O2 green wood or bagasse and up to 80% in weight occurs. The difference between
Solid Line was Used for Curves on Figure C-10B
bark. It is of importance in combustion the original weight and that of the
calculations because it directly influences dried sample divided by the original
Figure C-14
Clean| Variation ofTechnologies
theoretical CO2 with
Combustion the heating value of the fuel burned by weight is the moisture fraction as
Figure: [C/(H – 0.1 O2)]
C.14/2(JG) increasing the proportion of inert matter received. In the lumber industry, it
PMS5415 K
present. Furthermore, it requires some is customary to speak of moisture in
of the heat released by the dry substance wood on an oven-dry basis, whereby
to evaporate and superheat it. the loss of weight in the sample tested
* Calculated from reported high heating value HHV) of kiln-dried wood assumed to be 8% moisture.
† The terms “hard” and “soft” wood, contrary to popular conception, have no reference to the actual hardness of the wood. According to the Wood Handbook, prepared by
the Forest Products Laboratory of the U.S. Dept. of Agriculture, hardwoods belong to the botanical group of trees that are broad leaved, whereas softwoods belong to the
group that have needle or scalelike leaves, such as evergreens. Cypress, larch, and tamarack are exceptions.
Combustion Calculations by Graphical Methods | C-17
% as Received
dry must then be converted to the 15
as-received basis before using it in The HHV of wood or bagasse is de-
10
combustion calculations. termined by a bomb calorimeter, as in
The oven-drying method of the case of other solid fuels. From the 5
determining the moisture content similarity of the ultimate analyses of 0
is not accurate with woods such as different woods, it might be reasonable 20 30 40 50 60 70 80 90 100
% Relative Humidity (ambient)
southern yellow pine, which con- to assume that equal weights of dry
tain oils that are easily volatilized, wood will release the same amount of
Figure C-15
Clean| Moisture content
Combustion of wood at 70°F
Technologies
because these oils will distill with heat, regardless of species. Actually, as
Figure: (21°C)ambient
C.15/ (JG) temperature
the moisture. It is then necessary to Table C-3 indicates, this is not the case, PMS5415 K
use the more elaborate distillation the heating value of wood depending in a lower heating value for the bagasse.
method, wherein the water and the to some extent on the physical struc- As a rule, larger quantities of trash and
oil are measured separately. ture and form of the wood tissue, and dirt are picked up when the cane is har
Other less reliable, but quicker, still more on the presence of resins, vested and loaded by mechanical means
ways of judging the moisture in wood gums, tannins, essential oils, or pig than when this is done by hand.
consist of measuring its electrical resis- ments. Resin alone, for instance, may Owing to their complex composi-
tance or using a wood hygrometer. increase the heating value of the same tion, it is not possible to make use
As a rule, kiln-dried woods contain wood as much as 15%. Table C-3 also of a Dulong-type formula to predict
from 6 to 10% moisture, although shows that the BTU/lb of dry soft- the heating value of either wood or
softwoods will show a much higher per- woods is a little higher, on the average, bagasse. The formula appears to give
centage unless properly handled in the than that of hardwoods, due perhaps values of available hydrogen in the fac-
kiln. Air-dried woods have had enough to the fact that most softwoods are tor H2 – O2/8 which are too low and,
moisture evaporated to be in equilib- conifers containing resins and oils. as a result, the heating values calculated
rium with the surrounding atmosphere. As mentioned, however, the chief with it are invariably low.
The actual percentage present depends factor affecting the heating value of the In sugar-mill practice, bagasse analy-
on both the temperature and relative wood or bagasse as fired into a furnace ses are frequently reported in terms
humidity of the ambient atmosphere. is its moisture content. In the case of of its fiber, sucrose, glucose, moisture,
Figure C-15 shows how the moisture wood refuse, the heating value may also and ash content. If these are given, the
in wood varies with different relative be influenced by the amount of trash HHV of bagasse in BTU/lb may be
humidities at 70°F (21°C). In arid cli- and dirt gathered in processing the computed from
mates, it may be as low as 5%, whereas wood. In harvesting sugar cane, a certain
HHV = 8,550(F) + 7,119(S) +
in humid regions it may reach 20% amount of trash, leaves, cane tops, and
6,750(G) – 972(M)
or more. Green wood, wood waste, grass may be included that would result
Equation C-19
C-18 | Clean Combustion Technologies
750 A B
A, lb/106 BTU as Fired
700 1,600 21
650 1,500 20
Wood C C
Bagasse 1,400 19
600 H2 + 0.1 O2 H2 – 0.1 O2
FUEL IN PRODUCTS, F C D
100 150
The combustible loss in burning wood
or bagasse is generally assumed to be
90 130
zero. Therefore, if the HHV and the H2, % by
ash content of these fuels are known, Weight as
Wh, lb/106 BTU as Fired
the ultimate analysis of wood or bagasse constituents of their analyses are the
21
as fired is known, as well as the HHV, same, although the relative quantity of
% by Volume
20
and Wh from Curves C of Figure C-18. tabulation in his article on “Typi-
Wood
cal Solid and Liquid Fuels”3 shows 19 Bagasse
FUEL OILS that the combustible fraction chiefly
The fuel oils ordinarily burned in consists of carbon and hydrogen with 18
10 20 30 40
steam generating units are derivatives small amounts of sulfur. Negligible C
in Wood or Bagasse, Dry or as Fired
H2 – 0.1 O2
of petroleum comprising the so-called percentages of nitrogen and oxygen
“bunker” grades, diesel-engine oils, as well as water and sediment are also Solid Line was Used for Curves on Figure C-17B
gas oils, kerosene, and gasoline. As frequently found. DeBaufre’s table is, in
Figure C-18
Clean |
Variation ofTechnologies
theoretical CO2
they all have a common source, the part, reproduced as Table C-5. Combustion
with [C/(H
Figure: C.18/ (JG) 2 – 0.1 02)]
PMS5415 K
EXAMPLE D
Assume that sawmill refuse from Western hemlock, containing 55% moisture as fired and having the ultimate analysis shown in Table C-3, is burned
with 45% excess air. Then
1. Fuel, F. Convert both the ash and the HHV of Table C-3 from the dry to the as-fired condition, or
100 − 55
Ash = 2.2 = 1%
100
and
100 − 55
HHV = 8,620 = 3,880BTU / lb
100
as fired. With those values for ash and HHV, read from Figure C-1, F = 255 lb/106 BTU (110 kg/GJ).
2. Atmospheric Air, A. Calculate the ratio
C 50.4
= = 5.07
H2 + 0.1 O2 5.8 + 0.1⋅ 41.4
and with this ratio read from Curves A, Figure C-17, for 45% excess air, A = 1,040 lb/106 BTU (447 kg/GJ).
3. Unburned Combustible Factor, C. The usual assumption when burning wood in steam boiler furnaces is that the combustible loss is zero;
consequently C in Equations C-3 and C-4 may be taken as one.
4. Total Products, P. From Equation C-4, P = F + CA = 255 + 1 × 1,040 = 1,295 lb/106 BTU (557 kg/GJ).
5. Moisture in Air, Wa. From Equation C-5, Wa = 0.03A = 0.013 × 1,040 = 14 lb/106 BTU (6.02 kg/GJ).
6. Moisture from Fuel, Wf. The total moisture in the products of combustion of wood, Wf, is the sum of Wc, the moisture in the wood as fired, and Wh,
the water formed by its combustion. Since the wood contains 55% moisture as fired and its HHV as determined for F is 3,880 BTU/lb (9.02 MJ/kg) from
Curves D of Figure C-17, Wc = 142 lb/106 BTU (61.5 kg/GJ). Next, convert the percentage of hydrogen in Table C-3 from the dry to the as-fired basis, or
100 − 55
H2 = 5.8 = 2.6%
100
as fired, and with a HHV of 3,880 BTU/lb (8.33 MJ/kg) read from Curves C, Figure C-17 W6 = 60 lb/106 BTU (25.8 kg/GJ). Then
7. Dry Gas, Pd. From Equation C-9, Pd = P – (Wa + Wf) = 1,295 – (14 + 202) = 1,079 lb/106 BTU (464 kg/GJ).
8. Percentage of CO2 in Products. Compute the ratio
C 50.4
= = 30.4
H2 − 0.1 O2 5.8 − 0.1 × 41.4
and for 45% excess air read on Curves B, Figure C-17 = 14.2%.
C-20 | Clean Combustion Technologies
15
60
ATMOSPHERIC AIR, A 14
The atmospheric air in lb/106 BTU as 13
fired may be calculated from Equation 50 12
C-2D or taken from Curve A of Figure 11
C-19 for any value of excess air up to 40 10
100%. The values obtained are aver- 9
ages for a wide range of fuel oils; no
30
chemical analysis is necessary in using 17 18 19 20 21 22
HHV, 1,000 BTU/lb
Curve A.
CO2, values may be obtained after 8. Percentage of CO2 in Products. From Curves B, Figure C-19, for 10% excess air and a
C/H2 ratio of 7.5, read CO2 = 14.4%.
calculating the ratio C/H2 from the
ultimate analysis of the fuel.
NATURAL GAS applied to natural gas refer to its gasoline the depth of the well and may vary from
Characteristics of typical natural gases vapor, and not to its moisture content. 32 to 165°F (0 to 74°C). The general
are given in Table C-6, with the volu- In fact, the only time natural gas has any rule among geologists is to assume an
metric heating value on a moisture-free moisture at its point of origin is when increase in temperature of 1°F (0.6°C)
basis. The largest percentage of the it is next to salt water, but this is a rela for every 50 feet in depth, although this
volume contains hydrocarbons (CH4, tively unusual occurrence. rule will not always check with actual
C2H6, C3H8, C4H10, C5H12). measurements taken in the field.
From the standpoint of combustion TEMPERATURE It is also apparent that gas, which
calculations, it is important to remember The temperature of natural gas as it has flowed through many miles of
that the terms dry or wet as commonly issues from the ground is dependent on pipeline, will have whatever ambient
TABLE C-6 | CHARACTERISTICS OF TYPICAL NATURAL GASSES AT 60°F (16°C) AND 30 IN. HG (1 bar), DRY
% by Volume HHV A at Zero CO2 at
Density, Excess Zero
lb/Cu Ft BTU/ BTU/ Air, lb/ Excess
CO2 N2 H2S CH4 C2H6 C3H8 C4H10 C5H12
CuFt** lb 106BTU Air, %
5.50 — 7.00 77.73 5.56 2.40 1.18 0.63* 0.05622 1,060 18,880 738 12.1
3.51 32.00 0.50 52.54 3.77 2.22 2.02 3.44* 0.06610 874 13,220 729 12.3
26.2 0.7 — 59.2 13.9 — — — 0.06747 849 12,580 732 15.3
0.17 87.69 — 10.50 1.64 — — — 0.07120 136 1,907 732 6.9
0.20 0.60 — 99.20 — — — — 0.04491 1,006 22,410 732 11.7
— 0.60 — — 79.40 20.00 — — 0.08812 1,935 21,960 735 13.3
— 0.50 — — 21.80 77.70 — — 0.11079 2,389 21,560 738 13.7
* All hydrocarbons heavier than C5H12 were assumed to be C5H12 for combustion calculations.
** If gas is saturated with moisture at 60°F (16°C) and 30 in Hg (1 bar), reduce by 1.74%.
C-22 | Clean Combustion Technologies
air temperature prevails in the locality TABLE C-7 | COMBUSTION CONSTANTS OF DRY GASES AT 60°F (16°C)
where it is burned, most probably from AND 30 IN. HG (1 bar)
40 to 80°F (4.4 to 27°C). In combus-
Chemical Density of Dry HHV of Dry Gas
tion calculations for natural gas, it is Gas
Formula Gas, lb/cu Ft BTU/Cu Ft* BTU/lb
deemed sufficiently accurate to assume
the standard 60°F (16°C), when its Oxygen 02 0.08461 … …
actual temperature is not reported. Nitrogen (atmospheric) N2 0.07439 … …
Air … 0.07655 … …
MOISTURE
Carbon dioxide CO2 0.1170 … …
As already noted, the only time that Water vapor H20 0.04758 … …
natural gas can have any moisture on
Hydrogen H2 0.005327 325 60,901
leaving a well is when it previously has
Hydrogen sulfide H2S 0.09109 647 7,100
been in contact with salt with moisture
at its temperature in the well. Carbon monoxide CO 0.07404 321 4,323
However, gas, which is delivered from Saturated Hydrocarbons;
a pipeline, has often been “rehydrated;” Methane CH4 0.04246 1,014 23,896
that is, saturated with water vapor by Ethane C2H6 0.08029 1,789 22,282
means of steam jets to lower the cost Propane C3H8 0.1196 2,573 21,523
of maintaining pipe gaskets. Since the
Butane C4H10 0.1582 3,392 21,441
steam ordinarily is added to the gas
Pentane C5H12 0.1904 4,200 22,058
in the high-pressure line, when the
pressure is lowered for local distribution Unsaturated Hydrocarbons or Illuminants;
the relative humidity of natural gas will Ethylene C2H4 0.07421 1,614 21,647
also drop. Still another complication is Propylene C3H6 0.1110 2,383 21,464
introduced when a wet displacement Butylene C4H8 0.1480 3,190 21,552
meter is employed to measure the gas Pentylene C5H10 0.1852 4,000 21,600
consumption. In this meter, the gas may
Acetylene C2H2 0.06971 1,488 21,344
be saturated with water.
Benzene C6H6 0.2060 3,930 19,068
In view of the variable and uncertain
moisture content in natural gas, and in the Toluene C7H8 0.2431 4,750 19,537
absence of more definite determinations, * If gas is saturated with moisture at 60°F and 30.0 in. Hg, reduce by 1.74%.
it has become practice in industry to as-
sume natural gas to be dry at 60°F (16°C)
and 30 in. Hg (1 bar). Vol. cu ft Density of
Weight lb/cu ft
% by Vol. constituent/ constituent,
fuel
cu ft fuel lb/cu ft
DENSITY
CO2 5.50 0.0550 × 0.1170 = 0.00644
The density of a gaseous fuel may be
obtained by adding the weights of the H2S 7.00 0.0700 × 0.09109 = 0.00638
constituents in the fuel. The following is CH4 77.73 0.7773 × 0.04246 = 0.03300
an example of a density calculation for C2H6 5.56 0.0556 × 0.08029 = 0.00446
the natural gas listed first in Table C-6. C3H8 2.40 0.0240 × 0.1196 = 0.00287
Densities of the individual con- C4H10 1.18 0.0118 × 0.1582 = 0.00187
stituents are listed in the table to the
C5H12 0.63 0.0063 × 0.1904 = 0.00120
right and have been extracted from
Total = Density of Dry Gas = 0.05622
Table C-7, which lists the “saturated”
and “unsaturated” or “illuminant”
hydrocarbons that are most frequently
Combustion Calculations by Graphical Methods | C-23
found in gaseous fuels. Saturated Wc = entrained water found only calculated gas density can be no more ac-
hydrocarbons do not unite directly with in the primary gas, lb/ft3 at curate than the volumetric analysis from
hydrogen; i.e., they are stable in the temperature, t, °F which it is derived. A natural-gas analysis
presence of hydrogen, while unsaturated that groups all hydrocarbons as CH4 and
hydrocarbons readily take on more Before going to a great deal of refine- C2H6, or C2H6 and C3H8 may indicate
hydrogen. The latter are also called ment, however, remember that the a lower density than the actual value.
illuminants because they burn with a
bright luminous flame, as distinguished
from the saturated hydrocarbons, which X Y
have a blue flame.
.022 1.0
Extreme accuracy in calculations of
.020 1.2
density and heating value of gaseous Gas Temp.,
.018 °F
fuels is not warranted; the precise 1.4
values of the individual constituents .016 60
1.6 100
Density Correction, X
Density Correction, Y
continue to change as more accurate .014
1.8 150
experimental procedures are de- .012
200
veloped, presenting somewhat of a .010 2.0 250
“moving target.” .008 300
2.2
Figures C-20 and C-21 and Equation .006 350
C-13 may be used whenever it is desired 2.4
.004 400
to correct for moisture content or for a 2.6 450
.002
temperature higher than 60°F (16°C). 500
.000 2.8
0 0.10 0.20 0.30 0 0.10 0.20 0.30
d60 + X WS, lb/lb of Dry Gas WS, lb/lb of Dry Gas
dt = ( Dt + wc )
Y
Note: If the Gas Pressure, p-in. Hg, is Other Than 30��, Y as Read Here Must Be Multiplied
Equation C-21 by the Ratio (30/p) Before Being Used in Eqs. 21, 23, and 26.
where
Figure C-21 | Density correction Clean
for useCombustion
in EquationTechnologies
C-21
dt = d ensity of gas at tempera- Figure: C.21/ (JG)
ture, t, lb/ft3 (kg/m3) PMS5415 K
M N
d60 = density of dry gas at
1.00 140 °F
60°F (16°C), lb/ft3
500
X and Y = correction factors from
120
Figure C-21 0.95 450
Dt = d ust content of the gas,
100 400
lb/ft3 at temperature t, °F
Heating Value Correction, M
0.90
350
80
0.85 300
0.30
Gas Pressure, In. Hg Abs. 60 250
WS, lb/lb of Dry Gas
29
0.20 30 0.80 200
31 40
150
0.10 0.75
20
100
60
0 0.70 0
60 80 100 120 140 160 0 0.10 0.20 0.30 0 0.10 0.20 0.30
Dew Point of Gas, °F WS, lb/lb of Dry Gas WS, lb/lb of Dry Gas
Figure C-20 Figure C-22 | Heating value corrections for use in Equation C-22
Clean| Combustion
Water vaporTechnologies
in gas Clean Combustion Technologies
Figure: C.20/ (JG) Figure: C.22/ (JG)
PMS5415 K PMS5415 K
C-24 | Clean Combustion Technologies
HEATING VALUE
% by Volume Heating
Heat Evolved by Com
As with all other gaseous fuels, it is Volume, cu ft/ Value, BTU/
ponents, BTU/cu ft
customary to compute the HHV of Dry cu ft of Gas cu ft
natural gas, in BTU/ft3 (MJ/m3) at H2S 7.00 0.0700 × 647 = 45
60°F (16°C) and 30 in. Hg (1 bar), by CH4 77.73 0.7773 × 1,014 = 788
adding together the heat evolved by the C2H6 5.56 0.0556 × 1,789 = 99
combustible components reported in C3H8 2.40 0.0240 × 2,573 = 62
the gas analysis. Thus, for a gas such as
C4H10 1.18 0.0118 × 3,392 = 40
the first one in Table C-6, the procedure
C5H12 0.63 0.0063 × 4,200 = 26
indicated to the right may be employed,
using the heating values in Table C-6. Heating Value of Dry Gas = 1,060
BTU/cu ft
The HHV in BTU/lb is then ob-
tained by dividing the BTU/ft3 by the
density at 60°F (16°C) and 30 in. Hg
(1 bar). To correct the BTU/lb ( J/kg)
for the effect of higher temperature A B
and moisture content, refer to Equa- 1,400 HHV, BTU/Cu Ft at 60 °F and 30�� Hg
tion C-22 and Figure C-22. 160
600 800 1,000
1,300 1,200
1,200 150
HHVt = M (HHV60) + N
1,100 Atmospheric Air 140
14
FUEL IN PRODUCTS, F 13
12
After determining the HHV in BTU/ 11
lb, F, is easily computed from Equation 10
C-1 or read from Figure C-1. 9
8
ATMOSPHERIC AIR, A 7
The atmospheric air in lb/106 BTU as 6
A( HHV )60
At =
HHVt (1 + Ws + Dt / dt )
Equation C-23
TOTAL PRODUCTS, P
The weight of the products of combustion in lb/106 BTU as fired is given by P =
F + CA, where C is a factor to correct A for the combustible loss due to imperfect
combustion of the fuel. When burning natural gas, C is assumed to be one for any
properly designed furnace supplied with the correct amount of excess air.
Wf = Wsat + Wc + Wh
As explained before, natural gas is usually assumed to be dry, and since it sel-
dom carries any entrained moisture, the first two terms generally may be taken as
zero. If the gas is not dry, but contains Ws pound of water vapor per pound of dry
fuel, the lb of water vapor per 106 BTU as fired are given by
Wsat = 150
HHV 100 0.30
80 0.25
Equation C-24 60 0.20
0.16
40 0.12
where HHV is the high heating value of the gas in BTU/lb as fired. Figure C-24 is 0.10
0.08
a convenient plot of Equation C-24. 20 0.06
The water produced by the combustion of natural gas is given by Equation C-25
which follows. 10
8
6
4
(H2S+2CH4 +3C2H6 +4C3H8 +5C4H10 +6C5H12 ) 0.05
Wh = 473 × 0.04
BTU/ cu ft at 60°F and 30 in. Hg, dry 2 0.03
0.02
Equation C-25 0.01
1
1 2 4 6 8 10 15 25
For dry natural gas at 60°F (16°C) and 30 in. Hg (1 bar), Wh may be HHV, 103 BTU/lb as Fired
conveniently read from Curves B of Figure C-23. When it is desired to
Figure C-24 | Water vapor in fuel gas
Clean Combustion Technologies
Figure: C.24/ (JG)
PMS5415 K
C-26 | Clean Combustion Technologies
correct Wh for higher temperature and for moisture content, Equation C-26
may be used.
Wh (HHV)60
(Wh )t =
(HHV)t (1 + Ws + Dt / dt )
Note : If t = 60°F, the equation reduces to (Wh )t =Wh ,
regardless of the moisture andd dust content.
Equation C-26
1. Fuel, F. For a high heat value of 18,880 BTU/lb (43.9 MJ/kg), refer to Figure C-1 and REFINERY AND OIL GAS
from it read F = 53 lb/106 BTU (23 kg/GJ).
The range in composition for refinery
2. Atmospheric Air, A. From Curve A, Figure C-23, for 10% excess air read A = 807 lb/106
BTU (347 kg/GJ). gas and oil gas is shown in Table C-8.
3. Unburned Combustible. The general assumption when burning natural gas in stationary Refinery gas is predominantly made
boiler furnaces is that the combustible loss is zero. Consequently, C in Equations C-3 and up of saturated gases of low boil-
C-4 may be taken as one. ing point, such as methane (CH4),
4. Total Products, P. From Equation C-4, P = F + CA = 53 + 1 × 807 = 860 lb/106 BTU (370 ethane (C2H6), propane (C3H8), bu-
kg/GJ).
tane (C4H10), and pentane (C2H12)
5. Moisture in Air, Wa. From Equation C-5, Wa = 0.013A = 0.013 × 807 = 10 lb/106 BTU
with small amounts of illuminants.
(4.3 kg/GJ).
6. Moisture from Fuel, Wf.. Since Wc, the entrained moisture, is zero and Ws, the saturation
Minor quantities of carbon dioxide
moisture, is negligible, from Curves B, Figure C-23, for a HHV of 1,061 BTU/ft3 (40.2 (CO2), carbon monoxide (CO), and
MJ/m3) and H2S + 2CH4 + 3C2H6 + 4C3H8 + 5C4H10 + 6C5H12 = 7.0 + 155.5 + 16.7 + 9.6 hydrogen (H2) may also be present
+ 5.9 + 3.8 = 198.5% by volume, read Wf = Wh, = 88 lb/106 BTU (38 kg/GJ). with oxygen (O2) and nitrogen (N2)
7. Dry Gas, Pd . From Equation C-9, Pd = P – (Wa + Wf) = 860 – (10 + 88) 762 lb/106 BTU introduced to lower the hydrocarbon
(328 kg/GJ).
vapor pressure, in some cracking
8. Percent C02 in Products. For CO2 + CH4 + 2C2H6 + 3C3H8 + 4C4H10 + 5C5H12 = 5.5 +
77.7 + 11.1 + 7.2 + 4.7 + 3.2 = 109.4% by volume and 8.5 CO2 + N2 – 5.5H2S = 46.7 + 0
processes. If the original crudes
– 38.5 = 8.2% by volume, from Curves C, Figure C-23, read 12.1% CO2 at zero excess air. contained sulfur, the ensuing gases
Following Curves D as guidelines to 10% excess air, read CO2 = 10.9%. include hydrogen sulfide (H2S) un-
less the sulfur compounds have been
Combustion Calculations by Graphical Methods | C-27
TABLE C-8 | CHARACTERISTICS OF TYPICAL REFINERY AND OIL GASES AT 60°F (16°C) AND 30 IN. HG, (1 bar) DRY
% by Volume HHV CO2 at
A at Zero
Density, Zero
BTU/ Excess Air,
lb/Cu Ft Excess
O2 N2 CO2 CO H2 H2S CH4 C2H6 C3H8 C4H10* C5H12† C3H6‡
Cu Ft**
BTU/lb lb/106 BTU
Air, %
— — — — — 2.18 41.62 20.91 19.72 9.05 6.52 — 0.08676 1,898 21,880 722 13.3
— — — — — — 4.30 82.70 13.00 — — — 0.08377 1,858 22,170 725 13.4
— — — — — — 92.10 1.90 4.50 1.30 0.20 — 0.04845 1,136 23,460 723 12.1
— — — — — — 5.0 12.0 30.0 34.0 19.0 — 0.13760 2,988 21,720 717 13.9
— — 3.3 1.5 5.6 — 30.9 19.8 38.1 0.6 — 0.2 0.08102 1,696 20,930 725 13.4
2.3 8.7 — — — — 30.3 13.4 19.1 14.7 1.8 9.7 0.09232 1,844 19,970 715 13.6
0.9 8.4 2.2 14.3 50.9 — 15.9 5.0 — — — 2.4 0.03631 519 14,300 648 12.2
0.1 2.7 1.0 6.8 59.2 — 25.4 — — — — 4.8 0.02756 586 21,270 657 10.6
removed by purifying operations. In on natural gas, with appropriate amounts of excess air up to 70% of
Table C-8, the last two analyses rep- correction factors for temperature the theoretical requirements. Due
resent oil gas, whereas the others are and moisture as derived for blast-fur- to the relatively large percentages
representative of refinery gas. Note nace gas. The BTU/lb is the ratio of of carbon monoxide and hydrogen
that the hydrogen content of oil gas is BTU/ft3 and the density. Although present in some of the oil gases, the
higher than that of refinery gas, as is there is no doubt that some refinery magnitude of A in Table C-8 varies
also its carbon monoxide content. gases contain little or no moisture, from 648 to 725 lb/106 BTU (279
others are saturated with it. The to 312 kg/GJ). The correlation of A,
DENSITY values in the table are given on a at zero excess air, with the term (CO
In the section on natural gas, a method dry basis. + H2), is shown in Figure C-25. The
was outlined for calculating the density line on this figure was used as the
of a gaseous fuel, given its analysis by FUEL IN PRODUCTS, F basis for spacing Curves A of Figure
volume. The same method can be used With the BTU/lb (HHV) already C-26. The atmospheric air as taken
to find the density of refinery gas, or determined, it is only necessary from Figure C-26 can be corrected for
oil gas, either at 60°F (16°C) or at a to refer to Figure C-1 for the value temperatures other than 60°F (16°C)
higher temperature. of F which may also be calculated
No correction for gas pressure, other directly from
than the standard 30 in. Hg (1 bar), is 750
required in the calculation of density, 10 6
F= 700
A, lb/106 BTU
1,100
120 1,800
1,000 0 BTU fired, by the addition P = F + CA.
40
60 110
2,000
900 80 MOISTURE FROM FUEL, Wf
100
2,200
800
90 The total amount of moisture in the
2,400
700 2,600 products of combustion, Wf, which is
80
derived from the fuel itself, is the sum-
600 70
mation of the entrained moisture Wc,
500 100 150 200 250 300 350 400 450 the water vapor held by the gas Ws,
0 10 20 30 40 50 60 70 H2 + H2S + 2CH4 + 3C2H6 + 4C3H8 + 5C4H10
% Excess Air + 6C5H12 + 3C3H6, % by Volume as Fired and the water produced by burning
hydrogen, or hydrogen compounds
C D
making up the fuel, Wh.
N2 + 4CO2 + 3CO + C3H8 in Fuel
Unless the refinery gas has under-
0 20 40 60 80 15
gone a scrubbing or washing operation
14 to remove objectionable elements from
13 the distillate, it contains hardly any
CO2 in Dry Products, % by Volume
When, for more accuracy, it is required to correct Wh for the effect of tempera-
ture and moisture content, Equation C-23 can be utilized.
The guide Curves D of Figure C-26 make it possible to obtain the percentage of
CO2 in the products for an excess-air value up to 70%.
TABLE C-9 | CHARACTERISTICS OF TYPICAL COKE-OVEN GASES AT 60°F (16°C) AND 30 IN. HG (1 bar), DRY
% by Volume HHV A at Zero CO2 at Zero
Density,
BTU/ BTU/ Excess Air, Excess
CO2 O2 N2 CO H2 CH4 C2H4 C6H6 lb/Cu Ft
Cu Ft* lb lb/106 BTU Air, %
1.8 0.2 3.4 6.3 53.0 31.6 2.7 1.0 0.0298 596 20,010 674 11.1
1.4 0.5 4.2 5.1 57.4 28.5 2.9 — 0.0263 539 20,490 664 10.0
2.6 0.6 3.7 6.1 47.9 33.9 5.2 — 0.0316 603 19,070 676 11.3
3.13 — — 11.93 42.16 37.14 4.76 0.88 0.0359 663 18,500 684 12.7
0.1 — 2.4 6.8 27.7 50.0 13.0 — 0.0393 829 21,100 700 12.3
0.75 — 12.1 6.0 53.0 28.15 — — 0.0291 477 16,390 668 9.4
* If gas is saturated with moisture at 60°F and 30 in. Hg, reduce by 1.74%.
(1,760 Pa). If the gas is saturated with A, at zero excess air with the term (CO
720
moisture at 60ºF (16 ºC) and 30 in. + H2) is shown in Figure C-27.
Hg (1 bar), then the partial pressure of The atmospheric air in lb/106 BTU 700
A, lb/106 BTU
all its dry constituents is 30.0 – 0.52 = may be read from the curve family A 680
29.48 in. Hg (0.9983 bar). in Figure C-23. The spacing of these 660
A density correction is made as follows: curves is based on the average line
640
drawn in Figure C-27. 30 40 50 60 70
29.48 (CO + H2) in Fuel, % by Volume
0.02977 × = 0.02925
30.0 TOTAL PRODUCTS, P
Add to this the density of the water The weight of the products of combus- Figure C-27
Clean|
Variation ofTechnologies
Combustion theoretical air with
vapor (from the steam tables) 0.00083 tion is calculated from P = F + CA. Figure: C.27/ (JG) combustion of
(CO – H 2 ) in
PMS5415coke-oven
K gas
Density of saturated gas 0.03008
MOISTURE FROM FUEL, Wf one-half volume of oxygen to produce
It can be seen that the density correc- The moisture, Wc, in a fuel gas satu- one volume of water. Similarly, one vol-
tion for moisture content will be 1%. rated at 60°F (16 ºC) may be neglected ume of CH4 will form, on burning, two
as it amounts to less than 2 lb/106 volumes of H2O, etc. Since a coke oven
FUEL, F BTU (0.85 kg/GJ) as fired. Wh, the gas may contain H2, CH4, C2H4, and
The portion of the fuel, which reap- water formed by the combustion of hy- C6H6, with the aid of Table C-7 it is seen
pears in volatile form in the products drogen and the hydrocarbons present that the total volume of H2O evolved by
of combustion can be read from Figure in the coke-oven gas, is obtained from these constituents is proportional to (H2
C-1 when HHV (BTU/lb of fuel) Curves B, Figure C-28. Since Wc is + CH4 + C2H4 + 3C6H6). The follow
is known. If HHV is not given, it is neglected, Wf which is the total mois- ing exact relation will convert the volume
found by dividing the BTU/ft3 by the ture in the flue gases derived from the of H2O to lb/106 BTU,
fuel density. No moisture correction is fuel, is equal to Wh. In the simple case
necessary, because F is determined to of hydrogen, the volumetric chemical (H2 +2CH4 +
two significant figures only. equation for combustion is 2C2H4 +3C6H6 )
Wh = 473
BTU/ cu ft at 60°F and
1
ATMOSPHERIC AIR, A H2 + O2 = H 2 O 30 in. Hg, dry
2
The atmospheric air required for Equation C-32
Equation C-31
combustion will vary with the carbon
monoxide and hydrogen content of the Equation C-31 shows that one which is also plotted as Curves B,
gas. A correlation of atmospheric air, volume of hydrogen combines with Figure C-28.
Combustion Calculations by Graphical Methods | C-31
A B
1,200 Fuel HHV
BTU/Cu Ft, Sat. at 60 °F, 30.0�� Hg
1,100 400 450 500
150 550
140
1,000 600
A, lb/106 BTU as Fired
130
C D
(100 + C2H4 + 5C6H6) - (H2 + N2)
% by Volume
14
80 70 60 50 45
40 13
12
35
11
CO2 in Dry Products, % by Volume
30 10
4
900 800 700 600 500 400 0 10 20 30 40 50 60 70
HHV, BTU/Cu Ft % Excess Air
Equation C-33
C-32 | Clean Combustion Technologies
Curves C, Figure C-28 are a plot of BLAST FURNACE GAS MOISTURE FROM CLEANING
Equation C-33, while Curves D serve The range in composition of steel Blast furnace gas is classified as raw
as guidelines for determining the per- plant, blast furnace gas is given in Table gas that had a primary cleaning, or
centage of CO2 in the dry products C-10. Note that the volumetric heating gas that has had a final cleaning. For
of combustion for any other value of value is based on the gas being mois- purposes of combustion calcula-
excess air than zero. ture free, and that a correction must be tions, however, it is more convenient
made for moisture of saturation. to speak of washed and unwashed
gas, and these will be the terms most
EXAMPLE H frequently employed in the discussion
that follows.
Assume a coke-oven gas with the typical analysis of the first gas listed in Table C-9 and satu-
rated with moisture at 60ºF (16ºC) and 30 in. Hg (1 bar)to be burned with 10% excess air. Washed gas is blast furnace gas that
has gone through either primary or
1. Heating Value. The high heat value given as 596 BTU/ft3 (22.2 MJ/m3) must first be final washers. It may be considered
corrected for moisture content by deducting 1.74%; 596 – 10 = 586 BTU/ft3 (21.8 MJ/m3)
saturated fuel. This value is divided by the fuel density (density = 0.0298 lb/ft3), neglecting
saturated with moisture at whatever
the moisture correction) to obtain the approximate BTU/lb. temperature it leaves the washer
2. Fuel, F. For a high heat value of 19,700 BTU/lb (45.8 MJ/kg), read from Figure C-1, F = because of the intimate contact
51 lb/106 BTU (22 kg/GJ). between cooling water and hot gas. If
3. Atmospheric Air, A. From Curves A, Figure C-28 for 10% excess air and (CO + H2) = 6.3 + the gas has only been given a primary
53.0 = 59.3, read A = 742 lb/106 BTU (319 kg/GJ).
wash, it will contain about 4 grains
4. Unburned Combustible. For stationary boiler furnaces, the general assumption when
(0.00057 lb/ft3 or 0.0091 kg/m3)
burning coke-oven gas is that there is no combustible heat loss. Therefore, C in Equations
C-3 and C-4 is equal to one. of entrained moisture in addition to
5. Total Products, P. From Equation C-4, P = F + CA = 51 + 1 × 742 = 793 lb/106 BTU (341 the water vapor required to saturate
kg/GJ). it. This entrained moisture is carried
6. Moisture in Air, Wa. From Equation C-5, Wa = 0.013A = 0.013 × 742 = 10 lb/106 BTU along by the gas in the form of sus
(4.3 kg/GJ). pended droplets, which are effectively
7. Moisture from Fuel, Wf. Since Wc = 0, from Curves B, Figure C-28, for a high heat value removed by a final cleaner.
of 596 BTU/ft3 (22.6 MJ/m3)and H2 + 2CH4 + 2C2H4 + 3C6H6 = 53.0 + 636.2 + 5.4 + 3.0
= 124.6% by volume, read Wf = 99 lb/106 BTU (43 kg/GJ).
Unwashed gas, on the other hand,
8. Dry Gas, Pd. From Equation C-9, Pd = P – (Wa = Wg) = 793 – (10 + 99) = 684 lb/106
may exist as raw gas or as gas that has
BTU (294 kg/GJ). had a primary, or even a final, clean-
9. Percentage of CO2 in Products. For a heating value of 596 BTU/ft3 (22.6 MJ/m3) and ing by means of dry separators. Its
(100 + C2H4 + 5C6H6) – (H2 + N2) = (100 + 2.7 + 5.0) – (53.0 + 3.4) = 51.3% by volume, temperature is much higher than that
from Curves C, Figure C-28, read 11.1% CO2 at zero excess air. Following Curves D as of washed gas and because of this it is
guidelines to 10% excess air, read CO2 = 10%.
capable of holding much more mois-
ture, since it will contain all the water
TABLE C-10 | CHARACTERISTICS OF TYPICAL BLAST-FURNACE GASES AT 60°F (16°C) AND 30 IN. HG (1 bar), DRY
% by Volume Density, HHV A at Zero Excess CO2 at Zero
CO2 N2 CO H2 CH4 lb/Cu Ft BTU/Cu Ft* BTU/lb Air, lb/106 BTU Excess Air,%
* If gas is saturated with moisture at 60° and 30.0 in. Hg, reduce by 1.74%
Combustion Calculations by Graphical Methods | C-33
vapor that it had on leaving the blast furnace regardless of its degree of cleaning.
Using a dew point of 140°F (60°C) has been recommended for calculating the
moisture in unwashed gas, but it is obviously preferable to know the actual dew
point of a particular gas than to assume it. Figure C-20, then, offers a convenient
curve for determining Ws, the saturation moisture, for any dew-point tempera-
ture from 60 to 160°F (16 to 71°C). There is, of course, no entrained water in
unwashed gas.
DENSITY
The density of dry blast furnace gas at 60°F (16 °C) (d60) is easily computed from
its volumetric analysis by the method explained in the section on gas, which assumes
that the gas is clean. The true density (dt) is obtained by correcting d60 for the actual
temperature (t), moisture content, and dust according to Equation C-21.
As an example, assume that the blast furnace gas in the first line of Table C-10 is
unwashed and goes to a burner with 15 grains/ft3 (0.034 kg/m3)of dust, at 500°F
(260°C) with a dew point of 140°F (60°C) and at a pressure of 30 in. Hg (1 bar).
What is its actual density if there is no entrained moisture?
For a saturation temperature of
140°F (60°C) and a pressure of 30 in. % by Volume, cu ft/ Density Ib/cu ft Weight Ib/
Hg (1 bar), read from Figure C-20, Volume, Dry cu ft of Gas at 60°F, Dry2 cu ft
Ws = 0.152 lb per lb of dry gas, and CO2 14.5 0.145 × 0.1170 = 0.01697
from Figure C-21. Obtain X = 0.0115 N2 57.5 0.575 × 0.07439 = 0.04277
and Y = 2.29. Then the density of clean CO 25.0 0.250 × 0.07404 = 0.01851
gas at 500°F (260°C) and a dew point
H2 3.0 0.030 × 0.005327 = 0.00016
of 140°F (60°C) will be expressed as
Density of Clean Dry Gas at 60°F = 0.07841
shown in the table to the right.
HEATING VALUE
The BTU/ft3 of clean and dry blast furnace gas at 60°F (16°C), may be calculated
from its analysis and the heating value of its combustible components as obtained
from Table C-10. The BTU/ft3 thus determined, when divided by the density at
60°F (16°C), will give the BTU/lb (HHV)60. However, in calculating the true
heating value (HHV)t of a gas as fired, we must again take into account its dust
content, its temperature, and its moisture content.
The dust in blast furnace gas is made up chiefly of particles of iron ore, lime-
stone, and coke. Since its structure is sandy or granular, there is reason to believe
that iron ore is the main component in this mixture, and is, therefore, more easily
lifted by the blast than coke or limestone, which comes in relatively large lumps.
When the dust burns, the coke in it may be thought of as supplying the heat
required to reduce ore and limestone. In the absence of actual tests showing the
calorific value of the dust, it may be assumed that the heat evolved by the coke is
C-34 | Clean Combustion Technologies
just equal to that required to reduce the BTU, may be read from Curves A in TOTAL PRODUCTS, P
ore and the limestone. If this is true, the Figure C-29 for any value of excess air As with other fuels previously dis-
presence of dust will not influence the up to 70%. This value must be corrected cussed, the weight of the products of
heating value of blast-furnace gas. for the dust and moisture content and combustion is given by P = F + CA
It should also be clear that any increase temperature of the actual blast-furnace lb/106 BTU as fired. This equa-
in the temperature of the gas above 60°F gas. After finding A from Figure C-29 tion holds true for either washed or
(16°C) will enhance its heating value by and the heating value and density as unwashed gas, provided F and A are
an amount equal to the sensible heat of explained before, the true mass of calculated as explained in the previ-
the gas above 60°F (16°C). atmospheric air per 106 BTU as fired is ous sections.
Furthermore, assume, as an example, calculated using Equation C-23.
that an unwashed gas is saturated at
140°F (60°C). It will then have 0.152
lb of moisture for each pound of dry
gas. As an inert vapor, this moisture A B
will reduce the combustible fraction 1,200 HHV
and, therefore, the heating value of BTU/Cu Ft of Dry Gas
CH4 80 100 120
blast-furnace gas, at the same time that 1,100
by Volume
50
it augments the heating value by the CO + H2 140
1,000 0.2
sensible heat due to its high temperature. 40
A, lb/106 BTU as Fired
90 23
22
ATMOSPHERIC AIR, A 100
21
The air required for the combustion of 20
blast-furnace gas is a function of the 19
ratio CH4/(CO + H2) where CH4, 18
CO, and H2 represent the percentage 17
fractions by volume of methane, carbon 16
Wf = Wh + Wc + Ws
First, there is the moisture formed in burning the fuel and previously designated
by Wh. Equation C-34 shows that for clean gas at 60°F (16°C). Curves B of Fig-
ure C-29 are plotted from Equation C-34
473(H2 +2CH4 )
Wh=
BTU/ cu ft at 60°F and 30 in. Hg, dry
Equation C-34
If the actual gas is at a temperature higher than 60°F (16°C) and contains dust
and moisture, the readings from Curves B must be multiplied by the same factor
used in correcting the atmospheric air, to obtain the true moisture (Wh)t in lb/106
BTU as fired. This factor is calculated using Equation C-26.
The next moisture source called “entrained moisture” is carried along by the gas in
the form of suspended globules. Designated as Wc, it is considered only when the gas
has been given a primary wash and averages 7 lb/106 BTU (3 kg/GJ).
The last source is in the water vapor required to saturate the blast-furnace gas
at any temperature from 60°F (16°C) up, as the case may be, and is designated as
Ws. For blast-furnace gas saturated with moisture at 60°F(16°C), 8 lb/106 BTU
(3.4 kg/GJ) may be assumed for Ws without serious error. This moisture must be
considered separately from Wh and Wc in a heat balance since it exists as a vapor
requiring no latent heat of vaporization.
Wf = Wh + 7 + 8
Of course, Ws must be subtracted from the total products, along with Wh and
Wc, to obtain the dry gas.
100(CO2 +CO+CH4 )
%CO2 (at zero excess air)=
100 + 0.94(H2 +2CO+8CH4 )
where H2, CO2, CO, and CH4 are the symbols for percent by volume of these
constituents in the fuel. This CO2 must not be confused with the percentage of
carbon dioxide in the dry products of combustion.
Woody Fiveland
C-36 | Clean Combustion Technologies
EXAMPLE I
Assume a blast-furnace gas with the typical analysis of the first gas listed in Table C-10 and
saturated with moisture of 60°F (16°C) and burned with 15% excess air after going through a
primary washer.
1. Density. The density given as 0.0779 lb/ft3 for the dry gas must first be corrected for
moisture content. The saturation moisture for a dew point of 60°F (16°C) is read from
Figure C-20 as Ws = 0.02 lb moisture per lb of dry gas. The values of X and Y are read
from Figure C-21 for Ws = 0.02 and a temperature of 60°F (16°C) as X = 0.0015 and Y
= 1.03. The entrained moisture in a gas given a primary wash is assumed as
Wc = 0.00057 lb/ft3.
d +X 0.0779 + 0.0115
dt = 60 + (Dt + w c ) +
Y 1.03
(0 + 0.00057) = 0.0771 + (0.00057) = 0.077 lb / cu ft
1. W. L. De Baufre, “Composition and Heating Value of Fuels,” Combustion, 3. W. L. De Baufre, “Typical Solid and Liquid Fuels,” Combustion, 3(2):
2(11): 29–35, May 1931. 23–31, August 1931.
2. W. S. Patterson, “Calculations of Air Requirements and Combustion 4. W. L. De Baufre, “Typical Gaseous Fuels,” Combustion, 3(6): 26–33,
Products of U.S. Coals by Simple Graphical Methods,” Combustion, 3(8): December 1931.
9–17, February 1932.
C-37
Thermodynamic
Properties of Steam
Appendix D
The following steam tables have been (SI) units. Tables D-4 and D5 are in and specific entropy are each made
condensed from the ASME Interna- American Customary Units (ACU). exactly zero. It is defined as 32.02°F
tional Steam Tables for Industrial use The nomenclature used in the tables is (0.01°C), corresponding to a pressure
(Copyright 2000 by The American listed below. of 0.006117 bar (611.657 Pa). The
Society of Mechanical Engineers) and The SI units have been modified in temperature at 0.0°C is the point at
are based on the International Associa- that the pressure is presented in bar which the thermodynamic equilib-
tion for the Properties of Water and rather than Pascal (Pa), note 1bar = rium state is solid.
Steam (IAPWS) industrial formulation 105Pa, and the temperature is given in The critical point is defined as a tem-
(IAPWS-97). More information can degrees Centigrade and not in degrees perature of Tc=373.946°C and a critical
be found on the Web sites www.asme. Kelvin (°K=°C + 273.15). pressure of 220.64 bar.
org and www.iapws.org. Alstom Power, For convenience, Mollier charts
Inc. has permission from ASME to re- REFERENCE STATE have been provided (see Figures D1
print the tables and figures in this form. The liquid phase at the triple point and D2) to illustrate the complexity
Tables D-1 through D-3 are of water substance is the state for of the tabular data presented on the
presented in Standard International which the specific internal energy following pages.
NOMENCLATURE
Variable Description SI units ACU units
P Pressure bar psi
T Temperature °C °F
h´ Specific enthalpy of saturated liquid J/kg BTU/lb
h˝ Specific enthalpy of saturated vapor J/kg BTU/lb
hfg Latent heat (h˝– h´) J/kg BTU/lb
s´ Specific entropy of saturated liquid J/kg/°K BTU/lb/°F
s˝ Specific entropy of saturated vapor J/kg/°K BTU/lb/°F
v´ Specific volume of saturated liquid m3/kg ft3/lb
3
v˝ Specific volume of saturated vapor m /kg ft3/lb
D-
D- | Clean Combustion Technologies
1,000
100
Pressure, bar
10
.1
0.01
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
Enthalpy, kJ/kg
Table D-2 | Properties of Saturated Water and Steam (Pressure) in SI Units— continued
Table D-3 | Properties of Compressed Liquid and Superheated Steam (pressure) in SI Units
Pressure,
(Sat Temperature
Temp)
°C 5°C 25°C 50°C 100°C 150°C 200°C 250°C 300°C 350°C 400°C 500°C 600°C 700°C 800°C 900°C 1,000°C
0.01 v 0.001 137.5362 149.0961 172.1934 195.2789 218.3599 241.4391 264.5173 287.5949 310.6722 356.8261 402.9796 449.1327 495.2857 541.4386 587.5914
(6.969632) h 21.01948 2547.554 2594.4 2688.536 2783.646 2879.996 2977.734 3076.953 3177.716 3280.076 3489.77 3706.341 3929.96 4160.659 4398.415 4642.824
s 0.076252 9.092122 9.243004 9.513831 9.752997 9.968187 10.16453 10.34564 10.51418 10.67217 10.9625 11.22583 11.46824 11.69384 11.90562 12.10551
0.05 v 0.001 0.001003 29.78216 34.41862 39.04306 43.6631 48.28123 52.89837 57.51497 62.13121 71.36306 80.59442 89.8255 99.05642 108.2872 117.518
(32.87549) h 21.02347 104.8401 2593.36 2688.046 2783.368 2879.816 2977.606 3076.856 3177.639 3280.014 3489.727 3706.31 3929.937 4160.641 4398.401 4642.812
s 0.076252 0.367256 8.497559 8.769989 9.009692 9.225102 9.421551 9.602718 9.77129 9.929301 10.21966 10.48301 10.72542 10.95103 11.16281 11.36271
0.1 v 0.001 0.001003 14.86738 17.19672 19.51357 21.82599 24.13649 26.44601 28.75497 31.06359 35.68018 40.29628 44.9121 49.52776 54.14332 58.7588
(45.80755) h 21.02844 104.8447 2591.994 2687.431 2783.02 2879.59 2977.446 3076.734 3177.543 3279.936 3489.673 3706.271 3929.908 4160.619 4398.383 4642.797
s 0.076252 0.367255 8.174144 8.448771 8.689152 8.904837 9.101417 9.282656 9.451271 9.609311 9.8997 10.16307 10.40549 10.63111 10.8429 11.04279
0.5 v 0.001 0.001003 0.001012 3.418781 3.889851 4.356265 4.820687 5.284113 5.746976 6.209486 7.133872 8.057764 8.981378 9.904826 10.82818 11.75146
(81.31674) h 21.06827 104.8818 209.3687 2682.397 2780.201 2877.774 2976.158 3075.761 3176.777 3279.316 3489.245 3705.96 3929.674 4160.438 4398.238 4642.681
s 0.076251 0.367244 0.703776 7.695201 7.941201 8.159143 8.356789 8.538605 8.707567 8.865832 9.156488 9.420002 9.662509 9.888174 10.09999 10.29992
1 v 0.001 0.001003 0.001012 1.695959 1.936727 2.172495 2.406191 2.638868 2.870972 3.102722 3.565583 4.027949 4.490037 4.95196 5.413787 5.87554
(99.60592) h 21.11804 104.9281 209.4118 2675.767 2776.592 2875.475 2974.537 3074.54 3175.817 3278.54 3488.709 3705.571 3929.383 4160.212 4398.057 4642.535
s 0.07625 0.367231 0.703753 7.360999 7.614673 7.83556 8.034576 8.217124 8.386524 8.545073 8.836064 9.099759 9.342371 9.568101 9.779961 9.979912
2 v 0.001 0.001003 0.001012 0.001043 0.959894 1.080519 1.198909 1.316231 1.432964 1.549336 1.781438 2.013041 2.244367 2.475526 2.706591 2.937581
(120.2115) h 21.21759 105.0207 209.498 419.1732 2769.089 2870.779 2971.259 3072.084 3173.891 3276.984 3487.635 3704.794 3928.799 4159.76 4397.694 4642.244
s 0.076248 0.367206 0.703707 1.306937 7.280913 7.508113 7.709992 7.894046 8.064335 8.223457 8.51512 8.779177 9.021999 9.247858 9.459798 9.659807
3 v 0.001 0.001003 0.001012 0.001043 0.634032 0.716445 0.796452 0.875339 0.953622 1.031538 1.186721 1.341405 1.49581 1.650049 1.804193 1.958261
(133.5254) h 21.31712 105.1133 209.5843 419.2483 2761.182 2865.952 2967.933 3069.608 3171.957 3275.424 3486.561 3704.016 3928.215 4159.307 4397.332 4641.952
s 0.076247 0.36718 0.703661 1.306859 7.079054 7.313187 7.518097 7.70371 7.874907 8.034609 8.326948 8.591367 8.8344 9.060389 9.272408 9.472475
4 v 0.001 0.001003 0.001012 0.001043 0.470887 0.534345 0.595199 0.654884 0.713947 0.772637 0.889363 1.005587 1.121532 1.23731 1.352993 1.468602
(143.6125) h 21.41663 105.2059 209.6705 419.3234 2752.78 2860.992 2964.556 3067.112 3170.013 3273.86 3485.485 3703.238 3927.631 4158.855 4396.97 4641.661
s 0.076245 0.367154 0.703614 1.30678 6.930479 7.172366 7.380475 7.567702 7.739826 7.900114 8.193134 8.457916 8.701161 8.927279 9.139378 9.339503
5 v 0.001 0.001003 0.001012 0.001043 0.00109 0.425034 0.474429 0.522603 0.570138 0.617294 0.710947 0.804095 0.896964 0.989667 1.082274 1.174806
(151.8362) h 21.51614 105.2985 209.7568 419.3985 632.2663 2855.896 2961.13 3064.596 3168.061 3272.292 3484.408 3702.459 3927.046 4158.402 4396.607 4641.369
s 0.076243 0.367128 0.703568 1.306702 1.841925 7.061071 7.272553 7.461448 7.634518 7.795401 8.089105 8.354251 8.597708 8.823955 9.036134 9.236317
6 v 0.001 0.001003 0.001012 0.001043 0.00109 0.352116 0.393899 0.434409 0.474262 0.513731 0.592002 0.669768 0.747253 0.824571 0.901794 0.978942
(158.8324) h 21.61563 105.391 209.843 419.4736 632.328 2850.663 2957.653 3062.06 3166.1 3270.719 3483.33 3701.679 3926.461 4157.95 4396.245 4641.078
s 0.076242 0.367102 0.703522 1.306624 1.841813 6.968353 7.183382 7.374002 7.548038 7.709522 8.003915 8.269426 8.513094 8.739472 8.951731 9.151972
8 v 0.001 0.001003 0.001012 0.001043 0.00109 0.260868 0.293199 0.324151 0.354411 0.384273 0.443321 0.501857 0.560113 0.618202 0.676195 0.734112
(170.4135) h 21.81457 105.5762 210.0155 419.6239 632.4514 2839.77 2950.543 3056.924 3162.15 3267.561 3481.169 3700.118 3925.291 4157.043 4395.52 4640.495
s 0.076238 0.367051 0.703429 1.306467 1.84159 6.817616 7.040293 7.234531 7.410557 7.573266 7.86905 8.135294 8.379387 8.606023 8.818443 9.018801
10 v 0.001 0.001003 0.001012 0.001043 0.00109 0.206004 0.232739 0.257979 0.282492 0.306595 0.35411 0.401111 0.447829 0.49438 0.540835 0.587214
(179.8856) h 22.01345 105.7613 210.1879 419.7742 632.5749 2828.268 2943.222 3051.703 3158.163 3264.385 3479.004 3698.556 3924.119 4156.137 4394.795 4639.912
s 0.076234 0.366999 0.703337 1.306311 1.841366 6.695488 6.926623 7.124712 7.302809 7.466772 7.763965 8.030946 8.275464 8.502361 8.714941 8.915416
15 v 0.000999 0.001002 0.001011 0.001043 0.00109 0.132441 0.151999 0.169699 0.186578 0.203012 0.23516 0.266781 0.298117 0.329284 0.360355 0.391351
(198.2952) h 22.51043 106.2239 210.619 420.1499 632.8839 2796.016 2923.959 3038.266 3148.028 3256.366 3473.566 3694.639 3921.185 4153.867 4392.981 4638.454
s 0.076224 0.36687 0.703106 1.30592 1.840808 6.453652 6.711114 6.91989 7.103528 7.270766 7.571559 7.8404 8.085987 8.313536 8.52652 8.727286
20 v 0.000999 0.001002 0.001011 0.001042 0.001089 0.001156 0.111484 0.125501 0.138594 0.151208 0.17568 0.199614 0.22326 0.246737 0.270116 0.293419
(212.3845) h 23.00708 106.6864 211.0499 420.5256 633.1931 852.5725 2903.231 3024.252 3137.641 3248.227 3468.093 3690.709 3918.244 4151.593 4391.166 4636.996
s 0.076214 0.36674 0.702874 1.30553 1.840252 2.330092 6.547418 6.768515 6.958244 7.128962 7.433474 7.704199 7.950863 8.179067 8.392458 8.593516
30 v 0.000999 0.001002 0.001011 0.001042 0.001089 0.001155 0.070622 0.081175 0.090555 0.099377 0.116193 0.132445 0.148403 0.164189 0.179876 0.195488
(233.8584) h 23.99938 107.6107 211.9116 421.2774 633.8126 852.9781 2856.548 2994.349 3116.062 3231.571 3457.041 3682.807 3912.341 4147.034 4387.532 4634.08
s 0.07619 0.36648 0.702412 1.304752 1.839142 2.328507 6.289287 6.541211 6.74492 6.923259 7.235592 7.510169 7.759009 7.988531 8.202742 8.404388
40 v 0.000998 0.001001 0.00101 0.001041 0.001088 0.001154 0.001252 0.058868 0.066474 0.073432 0.086441 0.098857 0.110973 0.122915 0.134757 0.146523
(250.3575) h 24.99036 108.5342 212.773 422.0295 634.4334 853.3874 1085.686 2961.651 3093.318 3214.374 3445.837 3674.848 3906.41 4142.46 4383.895 4631.163
s 0.076163 0.366219 0.70195 1.303976 1.838037 2.326932 2.793329 6.363826 6.584264 6.771192 7.091896 7.37044 7.621488 7.85234 8.067383 8.269619
Table D-3 | Properties of Compressed Liquid and Superheated Steam (pressure) in SI Units— continued
Pressure, Temperature
(Sat
Temp) 5°C 25°C 50°C 100°C 150°C 200°C 250°C 300°C 350°C 400°C 500°C 600°C 700°C 800°C 900°C 1,000°C
°C
50 v 0.000998 0.001001 0.00101 0.001041 0.001087 0.001153 0.00125 0.045347 0.051971 0.05784 0.068583 0.078703 0.088515 0.098151 0.107686 0.117144
(263.9429) h 25.98003 109.457 213.634 422.7819 635.0554 853.8004 1085.662 2925.644 3069.294 3196.592 3434.476 3666.831 3900.453 4137.871 4380.253 4628.245
s 0.076134 0.365956 0.701489 1.303202 1.836936 2.325367 2.790892 6.21094 6.451529 6.648127 6.977804 7.260437 7.513727 7.745923 7.961806 8.164636
100 v 0.000995 0.000999 0.001008 0.001038 0.001084 0.001148 0.001241 0.001398 0.022442 0.026439 0.032813 0.038377 0.043594 0.048624 0.053546 0.058389
(310.9995) h 30.90891 114.0599 217.934 426.548 638.1836 855.9179 1085.717 1343.097 2923.958 3097.375 3375.058 3625.845 3870.274 4114.733 4361.998 4613.658
s 0.075942 0.36463 0.699186 1.299363 1.831498 2.317683 2.779094 3.248366 5.945796 6.213929 6.599323 6.904513 7.169571 7.408675 7.628898 7.834734
150 v 0.00099288 0.00099636 0.0010056 0.0010361 0.0010810 0.0011435 0.0012330 0.0013783 0.011481 0.015671 0.020828 0.024921 0.028619 0.032118 0.035502 0.038808
(342.16) h 35.806 118.64 222.22 430.32 641.34 858.12 1086.0 1338.1 2693.0 2975.5 3310.8 3583.3 3839.5 4091.3 4343.7 4599.1
s 0.075679 0.36328 0.69689 1.2956 1.8262 2.3102 2.7679 3.2275 5.4435 5.8817 6.3479 6.6797 6.9576 7.2039 7.4287 7.6376
800
0 MPa
80
60
v = 10
v = 50 m-1 kg-1
40 50
700
30
15 20
3,800
8 10
6
1.0 1.5
3
0.6 0.8
0.3 0.4
0.10
0.15 0.2
0.02
600
4
2
0.05
3,600
0.05
0.02
0.1
500
0.2
0.5
1.0
0.01
2
25 MPa 0.005
3,400
Temperature, °C
5
10
20
0.002
0.0015
Pa
0.001 M
400
3,200
MPa
0.0012
300
0.005
Saturated Vapor
100
Saturated Liquid
3,000
MPa
0.002
200
90% 80% 70% 60% 50% 40% 30% 20% 10%
2,800
m -1 k -1
a
g
01 MP
100
v = 50
v = 0.0
1,200
1,000
1,400
1,600
1,800
2,000
2,200
2,400
400
600
800
2,600
200
0
0 1 2 3 4 5 6 7 8 9 10
Entropy, kJ•kg–1•K–1
Table D-4 | Properties of Saturated Water and Steam (Temperature) in ACU Units
Table D-4 | Properties of Saturated Water and Steam (Temperature) in ACU Units— continued
Table D-5 | Properties of Saturated Water and Steam (Pressure) in ACU Units
Table D-5 | Properties of Saturated Water and Steam (Pressure) in ACU Units— continued
Tables E-1 through E-8 (units of previous numbers as 4.0469(3) and Obtain column, and then multiply the
length; area; volume; mass; force; 3.9370(1). Negative exponentiation quantity to be converted by the ap-
pressure; energy, heat, and work; is represented as follows: 0.00001 = propriate factor in the By column to
and power) are designed to provide 1.0(–5) and 0.000254 = 2.54(–4). obtain the desired units. For example:
accurate conversion of units between To facilitate formatting, in all to convert U.S. statute miles to kilo-
various equivalent unit forms. The To cases four digits are shown after the meters, multiply the number of miles
Start unit is located on the hori- decimal point. Rounding was done in by a factor of 1.609347. From the
zontal and the final Obtain unit is the last digit and all conversions are table: 2.5 miles × 1.609347 = 4.02
located on the vertical. For example, consistent both ways: meters → feet kilometers. Please note the scientific
to convert a measurement with a and vice versa. notation: 1 × 101, 1 × 102, 1 × 103
starting unit of foot to inches, the Table E-9 is provided for a number are equivalent respectively to 10, 100,
factor 12 is located at the intersection of selected mixed unit conversions. 1,000. For negative powers, represen-
foot and inches: 1 foot = 12 inches. This table is organized alphabetically tations such as 1 × 10–1, 1 × 10–2,
These tables are for reference only in categories: density, energy, etc. In 1 × 10-3 represent the numbers 0.1,
and present conversions with five sig- that table, to convert a unit, locate the 0.01, 0.001. Therefore, the number
nificant figures, for example, 4,046.9 unit in the Multiply column of the 2.345 × 103 is 2,345.0 and 2.345 ×
and 39.370. The tables format the table, select the desired unit in the To 10–3 represents 0.002345.
Obtain
Mils 1 1.0000(3) 1.2000(4) 3.6000(4) 6.3360(7) 3.9370(1) 3.9370(2) 3.9370(4) 3.9370(7)
Inch 1.0000(–3) 1 1.2000(1) 3.6000(1) 6.3360(4) 3.9370(–2) 3.9370(–1) 3.9370(1) 3.9370(4)
Foot 8.3333(–5) 8.3333(–2) 1 3.0000(0) 5.2800(3) 3.2808(–3) 3.2808(–2) 3.2808(0) 3.2808(3)
Yard 2.7778(–5) 2.7778(–2) 3.3333(–1) 1 1.7600(3) 1.0936(–3) 1.0936(–2) 1.0936(0) 1.0936(3)
Mile 1.5783(–8) 1.5783(–5) 1.8939(–4) 5.6818(–4) 1 6.2137(–7) 6.2137(–6) 6.2137(–4) 6.2137(–1)
Millimeter 2.5400(–2) 2.5400(+1) 3.0480(2) 9.1441(2) 1.6093(6) 1 1.0000(1) 1.0000(3) 1.0000(6)
Centimeter 2.5400(–3) 2.5400(0) 3.0480(1) 9.1441(1) 1.6093(5) 1.0000(–1) 1 1.0000(2) 1.0000(5)
Meter 2.5400(–5) 2.5400(–2) 3.0480(–1) 9.1441(–1) 1.6093(3) 1.0000(–3) 1.0000(–2) 1 1.0000(3)
Kilometer 2.5400(–8) 2.5400(–5) 3.0480(–4) 9.1441(–4) 1.6093(0) 1.0000(–6) 1.0000(–5) 1.0000(–3) 1
E-
E- | Clean Combustion Technologies
Obtain
in2 1 1.4400(2) 1.2960(3) 6.2726(6) 4.0145(9) 1.5500(–3) 1.5500(–1) 1.5500(3) 1.5500(9) 1.5500(7)
2
ft 6.9444(–3) 1 9.000(0) 4.3560(4) 2.7878(7) 1.0764(–5) 1.0764(–3) 1.0764(1) 1.0764(7) 1.0764(5)
yd2 7.7160(–4) 1.1111(–1) 1 4.8400(3) 3.0976(6) 1.1960(–6) 1.1960(–4) 1.1960(0) 1.1960(6) 1.1960(4)
Acre 1.5942(–7) 2.2957(–5) 2.0661(–4) 1 6.4000(2) 2.4710(–10) 2.4710(–8) 2.4710(–4) 2.4710(2) 2.4710(0)
mi2 2.4910(–10) 3.5870(–8) 3.2283(–7) 1.5625(–3) 1 3.8610(–13) 3.8610(–11) 3.8610(–7) 3.8610(–1) 3.8610(–3)
mm2 6.4516(2) 9.2903(4) 8.3613(5) 4.0469(9) 2.5900(12) 1 1.0000(2) 1.0000(6) 1.0000(12) 1.0000(10)
2
cm 6.4516(0) 9.2903(2) 8.3613(3) 4.0469(7) 2.5900(10) 1.0000(–2) 1 1.0000(4) 1.0000(10) 1.0000(8)
m2 6.4516(–4) 9.2903(–2) 8.3613(–1) 4.0469(3) 2.5900(6) 1.0000(–6) 1.0000(–4) 1 1.0000(6) 1.0000(4)
2
Km 6.4516(–10) 9.2903(–8) 8.3613(–7) 4.0469(–3) 2.5900(0) 1.0000(–12) 1.0000(–10) 1.0000(–6) 1 1.0000(–2)
ha 6.4516(–8) 9.2903(–6) 8.3613(–5) 4.0469(–1) 2.5900(2) 1.0000(–10) 1.0000(–8) 1.0000(–4) 1.0000(2) 1
To Ounce Quart Gallon Barrel Cubic Inch Cubic Foot Cubic Yard Cubic Cubic Liter
Start (oz) (qt) (gal) (bbl) (in3) (ft3) (yd3) Millimeter Meter (dm3)
Obtain
oz 1 3.2000(1) 1.2800(2) 5.3760(3) 5.5412(–1) 9.5751(2) 2.5853(4) 3.3814(–5) 3.3814(4) 3.3813(1)
qt 3.1250(–2) 1 4.0000(0) 1.6800(2) 1.7316(–2) 2.9929(1) 8.0790(2) 1.0567(-6) 1.0567(3) 1.0567(0)
gal 7.8125(–3) 2.5000(–1) 1 4.2000(1) 4.3290(–3) 7.4805(0) 2.0197(2) 2.6417(–7) 2.6417(2) 2.6417(–1)
bbl 1.8601(–4) 5.9524(–3) 2.3810(–2) 1 1.0307(–4) 1.7811(–1) 4.8089(0) 6.2897(–9) 6.2897(0) 6.2897(–3)
in3 1.8047(0) 5.7750(1) 2.3100(2) 9.7020(3) 1 1.7280(3) 4.6656(4) 6.1024(–5) 6.1024(4) 6.1024(1)
3
ft 1.0444(–3) 3.3420(–2) 1.3368(–1) 5.6146(0) 5.7870(–4) 1 2.7000(1) 3.5315(–8) 3.5315(1) 3.5315(–2)
yd3 3.8681(–5) 1.2378(–3) 4.9511(–3) 2.0795(–1) 2.1433(–5) 3.7037(–2) 1 1.3079(–9) 1.3079(0) 1.3079(–3)
3
mm 2.9574(4) 9.4635(5) 3.7854(6) 1.5899(8) 1.6387(4) 2.8317(7) 7.6460(8) 1 1.0000(9) 1.0000(6)
3
m 2.9574(–5) 9.4635(–4) 3.7854(–3) 1.5899(–1) 1.6387(–5) 2.8317(–2) 7.6460(–1) 1.0000(–9) 1 1.0000(–3)
dm3 2.9574(–2) 9.4635(–1) 3.7854(0) 1.5899(2) 1.6387(–2) 2.8317(1) 7.6460(2) 1.0000(–6) 1.0000(3) 1
Units (ACU), the unit pound is used or lbf ); in ACU, pound is always
as either a mass unit (pound-mass, or pound avoirdupois, not troy pound.
lbm) or as a force unit (pound-force,
Obtain
grain 1 4.3750(2) 7.0000(3) 1.4000(7) 1.5680(7) 1.5432(–2) 1.5432(1) 1.5432(4) 1.5432(7)
oz 2.2857(–3) 1 1.6000(1) 3.2000(4) 3.5840(4) 3.5274(–5) 3.5274(–2) 3.5274(1) 3.5274(4)
lb 1.4286(–4) 6.2500(–2) 1 2.0000(3) 2.2400(3) 2.2046(–6) 2.2046(–3) 2.2046(0) 2.2046(3)
ton-sh 7.1429(–8) 3.1250(–5) 5.0000(–4) 1 1.1200(0) 1.1023(–9) 1.1023(–6) 1.1023(–3) 1.1023(0)
ton-lg 6.3776(–8) 2.7902(–5) 4.4643(–4) 8.9286(–1) 1 9.8420(–10) 9.8420(–7) 9.8420(–4) 9.8420(–1)
milligram 6.4799(1) 2.8350(4) 4.5359(5) 9.0719(8) 1.0161(9) 1 1.0000(3) 1.0000(6) 1.0000(9)
gram 6.4799(–2) 2.8350(1) 4.5359(2) 9.0719(5) 1.0161(6) 1.0000(–3) 1 1.0000(3) 1.0000(6)
kilogram 6.4799(–5) 2.8350(-2) 4.5359(–1) 9.0719(2) 1.0161(3) 1.0000(–6) 1.0000(–3) 1 1.0000(3)
metric ton 6.4799(–8) 2.8350(–5) 4.5359(–4) 9.0719(–1) 1.0161(0) 1.0000(–9) 1.0000(–6) 1.0000(–3) 1
Obtain
oz-f 1 1.6000(1) 3.2000(4) 1.6000(4) 3.5969(0) 3.5969(–5) 3.5274(–2) 3.5274(1) 3.5969(3)
lb-f 6.2500(–2) 1 2.0000(3) 1.0000(3) 2.2480(–1) 2.2480(–6) 2.2045(–3) 2.2045(0) 2.2480(2)
ton-f 3.1250(–5) 5.0000(–4) 1 5.0000(–1) 1.1240(–4) 1.1240(–9) 1.1023(–6) 1.1023(–3) 1.1240(–1)
kip 6.2500(–5) 1.0000(–3) 2.0000(0) 1 2.2480(–4) 2.2480(–9) 2.2045(–6) 2.2045(–3) 2.2480(–1)
Newton 2.7801(–1) 4.4484(0) 8.8964(3) 4.4484(3) 1 1.0000(–5) 9.8067(–3) 9.8067(0) 1.0000(3)
dyne 2.7801(4) 4.4484(5) 8.8964(8) 4.4484(8) 1.0000(5) 1 9.8067(2) 9.8067(5) 1.0000(8)
gram–f 2.8350(1) 4.5361(2) 9.0722(5) 4.5361(5) 1.0197(2) 1.0197(–3) 1 1.0000(3) 1.0197(5)
kg–f 2.8350(–2) 4.5361(–1) 9.0722(2) 4.5361(2) 1.0197(–1) 1.0197(–6) 1.0000(–3) 1 1.0197(2)
kN 2.7801(–4) 4.4484(–3) 8.8964(0) 4.4484(0) 1.0000(–3) 1.0000(–8) 9.8067(–6) 9.8067(–3) 1
Obtain
in-H2O 1 1.2000(1) 1.3595(1) 5.3525(0) 2.7681(1) 5.3525(–1) 4.0680(2) 4.0148(–3) 4.0148(2)
ft-H2O 8.3333(–2) 1 1.1330(0) 4.4606(–1) 2.3067(0) 4.4604(–2) 3.3900(1) 3.3456(–4) 3.3456(1)
in-Hg 7.3554(–2) 8.8265(–1) 1 3.9370(–1) 2.0360(0) 3.9370(–2) 2.9921(1) 2.9530(–4) 2.9530(1)
cm-Hg 1.8683(–1) 2.2419(0) 2.5400(0) 1 5.1714(0) 1.0000(–1) 7.6000(1) 7.5006(–4) 7.5006(1)
2
lbf/in (psi) 3.6126(–2) 4.3352(–1) 4.9116(–1) 1.9337(–1) 1 1.9337(–2) 1.4696(1) 1.4504(–4) 1.4504(1)
mm-Hg 1.8683(0) 2.2419(1) 2.5400(1) 1.0000(1) 5.1714(1) 1 7.6000(2) 7.5006(–3) 7.5006(2)
atm 2.4582(–3) 2.9499(–2) 3.3421(–2) 1.3158(–2) 6.8046(–2) 1.3158(–3) 1 9.8692(–6) 9.8692(–1)
2
N/m 2.4908(2) 2.9890(3) 3.3864(3) 1.3332(3) 6.8948(3) 1.3332(2) 1.0133(5) 1 1.0000(5)
Bar 2.4808(–3) 2.9890(–2) 3.3864(–2) 1.3332(–2) 6.8948(–2) 1.3332(–3) 1.0133(0) 1.0000(–5) 1
Obtain
Btu 1 1.2850(–3) 2.5444(3) 3.9683(–3) 9.47817(–4) 3.4122(3) 3.4122(0)
ft-lbf 7.7817(2) 1 1.9800(6) 3.0880(0) 7.3756(–1) 2.6553(6) 2.6553(3)
Hp-hr 3.9302(–4) 5.0505(–7) 1 1.5596(–6) 3.7251(–7) 1.3410(0) 1.3410(–3)
Calorie 2.5200(2) 3.2383(–1) 6.4119(5) 1 2.3885(–1) 8.5990(5) 8.5990(2)
Joules 1.0551(3) 1.3558(0) 2.6845(6) 4.1867(0) 1 3.6000(6) 3.6000(3)
kW-hr 2.9307(–4) 3.7661(–7) 7.4569(–1) 1.1629(–6) 2.7778(–7) 1 1.0000(–3)
W-hr 2.9307(–1) 3.7661(–4) 7.4569(2) 1.1629(–3) 2.7778(–4) 1.0000(3) 1
Obtain
BTU/s 1 2.7778(–4) 1.2851(–3) 7.0680(–1) 3.9683(0) 9.4781(–4) 9.4781(–1)
BTU/hr 3.6000(3) 1 4.6261(0) 2.5445(3) 1.4286(4) 3.4121(0) 3.4121(3)
ft-lbs/s 7.7817(2) 2.1617(–1) 1 5.5000(2) 3.0883(3) 7.3760(–1) 7.3760(2)
Hp 1.4148(0) 3.9301(–4) 1.8182(–3) 1 5.6150(0) 1.3410(–3) 1.3410(0)
k–cal/s 2.5200(–1) 6.9988(–5) 3.2381(–4) 1.7809(–1) 1 2.3885(–4) 2.3885(–1)
W 1.0551(3) 2.9307(–1) 1.3557(0) 7.4600(2) 4.1867(3) 1 1.0000(3)
kW 1.0551(0) 29307(–4) 1.3557(–3) 7.4600(–1) 4.1867(0) 1.0000(–3) 1
F-
F- | Clean Combustion Technologies
P = xxxx
T = xxxx
H = xxxx
M = xxxx
Q = xxxx
xx xxxx xx xxxx
xx xxxx xxxxx
xx xxxx xx xxxx
xx xxxx xxxxx
Superheat
Cavity
M = xxxxxx
Q = xxxxxx
M = xxxxxx
Q = xxxxxx
SH
xx xxxx xx xxxx
x xxxx x xx xxx xx xxxx xxxxx
x xxxxxxx
xx xxxxx
xx xxxx xx xxxx
xx xxxx xxxxx
P = xxxx
Upper T = xxxx
Furnace RH xx xxxx xx xxxx H = xxxx
xx xxxx xxxxx M = xxxx
Q = xxxx
P = xxxx
T = xxxx
H = xxxx
M = xxxx
Q = xxxx
M = xxxxxx
Q = xxxxxx
P = xxxx
LTRH T = xxxx
H = xxxx
M = xxxx
Q = xxxx
x xxxx x xx
x xxxxxxx
xx xxxxx
M = xxxxxx
Q = xxxxxx
x xxxx x xx xxx
x xxxxxxx
xx xxxxx P = xxxx
T = xxxx
Furnace H = xxxx
� xx xxxx xx xxxx
xx xxxx xxxxx
M = xxxx
Q = xxxx
Economizer
PTC =
xx xxxx xx xxxx
xx xxxx xxxxx
Boiler Efficiency
PULVERIZER MODELING
Emissions reduction projects with
Figure F-2 | Typical boiler with OFA
OFA often include improvements to
Clean Combustion Technologies
the coal pulverization equipment. The Figure: App F.02/ (GM)
study of the coal pulverizer flow pat- PMS5415 K
terns with CFD has been an effective
tool to improve the coal fineness levels.
CFD analysis has aided the develop-
ment of Dynamic™ Classifiers. The
geometry inside the mill including the
vane wheels, rolls, and grinding table are
important aspects of the geometry that
need to be modeled. In Figure F-8, the
flight patterns of the particles have been
illustrated by particle tracks colored by Detailed mesh
generation of
velocity (blue is low and red is high). windbox zone
Velocity distribution plots shown enter-
ing the classifier vanes are important
aspects of a mill CFD model.
CORROSION MODELING
CFD software has the available
(a) Furnace Surface Mesh (b) Windbox and SOFA Compartments
“hooks” to model corrosion, slagging,
Temperature
°C °F
1,927 3,500
1,860 3,380
1,793 3,260
1,727 3,140
1,660 3,020
1,593 2,900
1,527 2,780
1,460 2,660
1,393 2,540
1,327 2,420
1,260 2,300
1,193 2,180
1,127 2,060
1,060 1,940
993 1,820
927 1,700
860 1,580
793 1,460
727 1,340
660 1,220
593 1,100
527 980
460 860
393 740
327 620
260 500
Burner zone
A Center line of
water wall tubes
Section A-A
175
168
161
154
147
140
133
126
119
112
105 Pulverizer vane wheel
98 designed to maximize
91
84 pulverization throughput.
77
70
63
56
49
42
35
28
21
14
7
0
Path Lines Colored by Velocity Magnitude (ft/s)
1,230
1,220
1,210
SOFA
1,200
1,190
CCOFA
H
G
F 1,180
Post LNCSF™ III
E Wastage Rate
D 1,170 (mils/10K-hr)
C 65+
B 1,160 50 to 65
A 35 to 50
20 to 35
1,150 5 to 20
-10 to 5
1,140 -25 to -10
0 100 200 300 400 500 600 0 100 200
Front Wall Tube Number Right Side Wall Tube Number
DUCTWORK TO TOWERS
Low-Pressure Bank
Some boiler systems use dry scrubbers
High- and Intermediate-
to reduce sulfur emissions. Boiler sys- Pressure Economizers
tems may have one or more towers. It is Intermediate-Pressure
often important to determine the inter- Dust Burner Bank
action between towers. In Figure F-13, High-Pressure Superheater High-Pressure Bank
a complex ductwork is modeled. CFD
models can be run to predict the flow
Inlet: Flow Vectors
splits to multiple towers or to examine
the flow distribution with a tower out of
Figure F-12 | HRSG: Development of flow controls
service. Figure F-13 shows the velocity
distribution through the complex flow
Clean Combustion Technologies
passages, with higher values shown as Figure: App F.12/ (GM)
the flow turns via the flow controls. As Velocity ft/s
PMS5415 K
shown in Figure F-14, the new features 65
62
of CFD model software are capable of 60
57
representing detailed geometry with 55
52
very large meshes, best utilized by 49
47
running the calculations on parallel 44
42
computer clusters. These large meshes 39
36
34
are crucial in resolving the complex flow 31
29
in these devices. 26
23
21
CIRCULATING FLUIDIZED BED 18
16
(CFB) POWER STATION 13
10
8
The simulation of the CFB hydro- 5 Scrubber Towers
3 (center out-of-service)
dynamics represents a significant 0
technical challenge. Until recently,
Figure F-13 | Dry scrubbers: Inlet ductwork to towers
Tee Cracks
Ligament Cracks •Crotch or External
•At and Between Bore •Piping Flexibility
Holes •Thermal Fatigue
•Axial or Circumferential and Creep
•Thermal Fatigue and
Creep
profiles can then be used to deter- the rapid changes in exhaust tem- models must be used. Complete mod-
mine the associated thermal stresses perature that occur during start-ups els of the tube and header assemblies
which, when combined with the and shutdowns. The load profile of are needed to capture the effect of
normal mechanical stresses, can cause combined cycle plants also means overall temperature profiles that cre-
stress cycles that can result in fatigue that the HRSG has to endure many ate expansion mismatches and loads
damage. The stress redistribution such cycles, often on a daily basis. To between parts of the assembly. These
and creep strain accumulation due ensure appropriate durability, designs models are normally constructed from
to high temperature service can also are verified and optimized using a beam and shell elements that provide
be predicted using the finite element combination of simulation tools an accurate and computationally ef-
method, which can incorporate com- including boiler transient models ficient representation of these parts.
plex constitutive equations. that solve the appropriate mass and To give more accurate stress estimates
energy balances to accurately predict at features such as tube to header con-
HRSG SUPERHEATER HARP the time variation of metal and steam nections, local solid models are used.
ASSEMBLY ANALYSES temperatures. This information can The boundaries of these models can
HRSGs incorporate modules of then be used as the boundary condi- be driven by global structural models.
closely spaced tubes to capture tions for structural analyses using the This approach permits accurate and
exhaust heat from a gas turbine to finite element method. To accurately efficient analysis across a wide range
create superheated steam (Fig. F-18). predict the local stresses at features of length scales.
The pressure parts experience sig- that might be vulnerable to fatigue
nificant thermal cycling because of damage, a variety of finite element NATURAL FREQUENCY AND
DYNAMIC RESPONSE
Power plants incorporate a number of
pieces of rotating machinery, includ-
HS10 1 - 16"
110
100
GE - 109FA HRSG Start-up (with purging)
HP Predicted Performance - Warm Start HS1 3 - 10" ing mills and fans. The finite element
HP method is frequently used to under-
90
HP Desup 80
GT Exhaust Temp HP Steam Temp
RS8 1 - 24"
Percent of Full Range
70
Drum
GT Exhaust Flow
60
HS RH Desup HW14
40
1- HP Steam Press Full Range Values :
1-
1- 1- 1- 30 24"
GT Exhaust Temp = 605 °C
GT Exhaust Flow = 653.63 kg/s 10"
16" 24" 16" 20 HP Steam Flow = 82.96 kg/s
HP Steam Pressure = 95.5 bara
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 10" HW13
RS9 RS6 Time (minutes)
16" HS11 24" 16" HS8 24" 16" HS2
HW12
RS10 RS5
HS
17 4"
9-
4"
9-
4"
9-
RS
15
15
4"
15
4"
15
4"
HS
19 4" 4" 4"
RS
17
4" 4" 4" 4" 4" 4"
HS
21 4" 4" 3" 3" 3" 3" 3" 3" 3" 3" 3"
HW
16 3" 3" 3" equations can be transformed from
9- 9- 9- 9- 9- 9-
20"
natural frequencies at which damaging,
RS3
HP HS
15
SH
RS
Woody Fiveland
Paul Chapman
Utility Boiler HRSG
This appendix describes some of unusual fuels represent some of the • Combustion testing for fuel prop-
the laboratory facilities that Alstom specific R&D objectives that support erties, firing system performance,
Power has assembled for conducting the overall goal of cost savings. Some and fuel handling issues
research and development (R&D) work, such as improved fuel utiliza- • Mechanical analysis and testing for
on steam power plants. The facili- tion, can enhance both cost savings structural, stress, heat transfer, vibra-
ties range from small, sophisticated and environmental performance. tion, integrity, and corrosion analyses
devices for examining the microstruc- Alstom Power strongly believes that of common and advanced materials
ture of various fuels and materials to well-equipped laboratories, staffed • Testing and analysis of environ-
large furnaces for testing full-scale, with highly skilled personnel, are man- mental controls
commercial components, such as datory in order to stay at the forefront • Chemical analyses of fuels, boiler
firing systems. The focus of Alstom of power generation technology. The materials, ash, deposits, and boiler
Power’s R&D is quite broad, cover- Alstom Power laboratory facilities are water chemistry issues
ing such areas as troubleshooting for used for conducting company spon- • Solid fuel-processing test
customers who are experiencing prob- sored projects; programs sponsored equipment for determining size-
lems, improvement or optimization of by private organizations, including its reduction performance
present designs, and investigation and customers; and those funded by the • Controls and operational software
development of new concepts. U.S. Department of Energy, European development for optimizing plant
A common goal for most of the work Union Framework programs, the Elec- performance
conducted in these laboratories—and tric Power Research Institute, the U.S.
of significant concern to Alstom Environmental Protection Agency, Key Alstom Power boiler labora-
Power’s customers—is the reliable the Gas Research Institute, and other tories include Windsor, Connecticut,
generation of electrical power while similar governmental and institutional USA; Birr, Switzerland; Vaxjo, Swe-
minimizing environmental impact. organizations. den; Chattanooga, Tennessee, USA;
Over time, the quantities of accept- The implementation of most R&D and Naperville, Illinois, USA. These
able emissions levels have dropped, programs requires the integration of laboratories have played a significant
and the number of regulated efflu- efforts from many disciplines within role in the development of Alstom
ents has increased. The ongoing goal Alstom. For example, if the owner of a Power’s products over the last century
of continuously improving future large steam generator is contemplating and continue in that role today.
emissions performance has remained. a switch to a new fuel, know-how is The Alstom Power laboratories use
Of course, cost is also of primary required in the areas of fuel combus- a range of techniques to address R&D
concern to Alstom’s customers. If tion and fireside effects, heat transfer, issues. Some of these, such as stress
their costs are unacceptably high, and environmental controls. Alstom analysis, overlap to a certain degree with
then there is no business. Greater Power laboratories around the world capabilities within the business units.
component availability, longer have recognized expertise in each of What makes the Alstom Power R&D
component life, and utilization of the following disciplines: laboratories unique is their modeling
G-
G- | Clean Combustion Technologies
and facility experimentation capa- are employed. An example of a physical particulate controls for selective catalytic
bilities. These range from advanced model of an electrostatic precipitator reduction inlet duct systems, cyclone
physical and numerical modeling, (ESP) system is shown in Figure G-1. development for circulating fluidized
such as computational fluid dynamics, Qualitative flow visualization, such bed boilers, and correction of fan-related
to fundamental-scale and pilot-scale as smoke injection, plays a dominant problems in several utility boilers.
facilities. Additionally, the laboratories role in analyzing many process-
provide a variety of support services to equipment situations. This technique COMPUTATIONAL MODELING
Alstom’s businesses. is used in conjunction with two- and Computational fluid dynamics (CFD)
three-dimensional physical modeling. support the analysis and development
The two-dimensional technique is a of many types of equipment. In this
PHYSICAL AND NUMERICAL
valuable tool in reducing the design computationally intensive process,
MODELING
effort necessary to achieve optimum the basic equations of heat transfer,
PHYSICAL MODELING equipment performance. Based on thermodynamics, chemistry, and so on
Physical modeling requires the design, qualitative results and the complex- are repeatedly solved throughout a flow
construction, and testing of prototypi- ity of the process configuration, field field until a stable result is achieved.
cal models using an appropriate fluid. modifications may follow; alternatively, Alstom has been using CFD for
Working fluids can be air, water, or quantitative tests may be required. more than 25 years, and it has had an
some other medium—including multi- Quantitative component optimiza- important role in many product-
phase fluids or components such as solid tion is achieved through evaluation of related improvements pertaining to coal
particles or liquid droplets—depending changes in flow fields resulting from combustion, heat transfer, and particle
on the problem under study. Typically, geometric alterations within process tracking dynamics. CFD has been used
model materials are wood, sheet metal, equipment. Flow-field changes are for many applications, including:
and clear plastic (for visualization). Scales measured using pressure and tempera-
• Oil- and gas-fired furnace systems
are a small fraction of actual size. Flow ture measurements from the physical
• Coal combustion, emissions, and
modeling is useful for analyzing issues model combined with gas sampling.
Loss on Ignition (LOI) predictions
such as flow maldistribution, separation The results are used to optimize and
• Tangential and wall-fired, low
of two component flows, and erosion. select improved equipment geometries.
NOx burner and Overfire air
Depending on the specifics of the prob- Recent work with these techniques
(OFA) system optimization
lem, two- or three-dimensional models has been in the area of flow and
• Stoker and Chemical Recovery
Unit (CRU) OFA systems
• Circulating Fluid Bed (CFB) boilers
• Selective Catalytic Reduction
(SCR) systems, including complex
ammonia injection grids
• Electrostatic Precipitator (ESP)
• Baghouses
• Ductwork flow and Differential
Pressure (dP) evaluation
• Heat exchangers
• Process heater and refinery furnace
simulation
• Erosion control screens and baffles
• Multiphase flow, including particle
separation
In most situations, proper use of
CFD requires a specialist user. The
Figure G-1 | Physical model of an electrostatic precipitator (ESP) and ESP heat exchanger
Alstom R&D centers share CFD ex- can influence the quality of the final FUNDAMENTAL FACILITIES
perience and knowledge both within result. Experience with both the METALLOGRAPHIC AND
and beyond Alstom Power. Boundary product and the model is essential. MATERIALS EXAMINATIONS
conditions, calibration data, sub- Often, other modeling techniques Occasional failures of highly stressed
model selection, and even location are combined with CFD in order power plant components are inevitable.
of the calculation points (gridding) to obtain the most complete results. When they do, a detailed analysis of the
Specialized computational systems component usually reveals exactly why
also are required. A typical system to and how the failure occurred. Alstom
support 10 users includes 50 dedicat- Power R&D laboratories perform state-
ed CPUs distributed in a combination of-the-art analyses to assist in preventing
of desktop and Beowulf-style central recurrence of the problem.
processing arrays. Examples are Figure G-5 shows the microstructure
shown in Figures G-2, G-3, and G-4, of a tube sample. These analyses re-
as well as in Appendix F. quire comprehensive determination of
Temperature
Figure G-2 | Gas temperature distributions in 4.500e+02
a coal-fired boiler 4.369e+02
4.238e+02
Clean Combustion Technologies 4.107e+02
3.980e+02
Figure: App G-02/ (MH)
PMS5415 K
Velocity Figure G-3 | Gas velocity distributions in an electrostatic precipitator and ID fan inlet duct
(Vector 1)
2.500e+01
Clean Combustion Technologies
Figure: App G-03/ (MH)
1.875e+01
PMS5415 K
1.250e+01
6.250+00
0.000+00
[m s^–1]
z
x
200 microns
Figure G-4 | Electrostatic precipitator/dry Figure G-5 | Documenting the condition of the microstructure at the outside diameter (OD)
flue gas desulfurization gas surface of a sample in an area of high wastage (~9:30 position). Evidence of
recirculation and mixing carburization at the surface of the tube is readily visible
Clean Combustion Technologies
G- | Clean Combustion Technologies
failure modes and solutions, combining Figure G-6 shows a large header • Inductively coupled plasma
operational experience with analytical section arriving at the laboratory for spectrometry
and metallurgical methods such as: analysis. Although a rare occurrence, • Brunauer, Emmett, Teller (BET)
such pieces are sent from operating surface area measurement
• Sample preparation, polishing, etc.
units to determine the root causes of • Particle size analysis
• Optical and scanning electron
observed problems. • Ultimate and proximate analysis
microscopy to reveal sample
• Resistivity measurements
mechanical condition CHEMICAL ANALYSES • Spectrophotometry
• Mechanical properties
Alstom Power uses modern chemical • Scanning electron microscopy
determinations
laboratories to provide comprehensive (Fig. G-7)
• Metallurgical analyses
chemical and physical testing of fuels
• Use of any of the chemical analyses
and boiler-related structural materials, ADVANCED MATERIALS
listed in the next section
ash- and filtering-related materials, and
metallurgical analyses. The prime func- The key factor limiting the improve-
tions of these groups are: ment of power plant efficiency is the
selection of high temperature materials
• To support relevant R&D activities
and their temperature/pressure limitia-
• To support field start-up, per-
tions. Materials are available that could
formance testing, and ongoing
dramatically increase boiler efficiency.
operational problems
Their costs are high, however, and the
• To support all aspects of coal and
quantities required for a boiler are so
ash analysis
great that the choice is not economi-
• To support all aspects of detailed
cal. Therefore, significant efforts have
ash and filter media properties
gone into the development of low-cost,
• To support all aspects of metallur-
improved materials suitable for boiler
gical analysis
application.
The analytical laboratories have a Alstom Power R&D has tested the
Figure G-6 | A large header section arriving full complement of state-of-the-art suitability of many new materials for
for analysis
instruments to provide accurate boiler service. Test capabilities include
data on a whole range of sample bench-scale cycling, creep, and other
Clean Combustion Technologies analyses, including: mechanical tests via banks of dedicated
Figure: App G-06/ (MH)
PMS5415 K
testing machines. Bench-scale simula-
• Graphite furnace atomic absorption
tion of corrosive environments and
• Thermogravimetric analysis
exposure to actual combustion condi-
• Differential thermal analysis
tions in “slipstream” type test rigs at the
• Ion chromatography
site of utility boilers also are performed
• X-ray diffraction and fluorescence
to detail material properties (Fig. G-8).
Through detailed testing, the ap-
plication limits of each material are
determined. With these new materi-
als, a continual increase in the steam
temperature and pressure capabilities
of Alstom boilers will be achieved.
Subcritical boilers will be gradually
phased out, and higher-temperature
supercritical and ultrasupercritical
boilers will be phased in.
Figure G-7 | Scanning electron microscopy Figure G-8 | Material testing via creep
machines
Development of new boiler materi- DTFS. Knowledge gained from this Multi-Component
als and manufacturing processes has facility has been used in conjunction Experimental Test Stand
a long history at Alstom. Some past with Alstom Power’s computational The multi-component experimental
achievements include: boiler models to predict successfully test stand is used to study highly
the carbon loss in large, coal-fired reactive species, such as mercury and
• Production of the first welded
boilers (see Chapter 3). The DTFS sulfur trioxide, created as by-products
boiler drum in America fabricated
also has provided basic data regarding of fossil fuel combustion. Hot flue gas
to American Society of Mechani-
the formation and control of emis- generated from coal combustion in the
cal Engineers (ASME) Code
sions such as NOx, dioxins (from 4-inch (100-mm) fluid bed combuster
specifications
firing refuse-derived fuel), and organic reactor is fed into the test stand. The
• Development of the low-hydrogen
hydrocarbons. stand is then used to measure phe-
welding electrode, which has since
nomena such as mercury speciation,
become universally used in industry FLUIDIZED-BED TEST FACILITY mercury capture on carbon-based
• Successful welding of the first stabi-
Fuel and sorbent performance during sorbents, sulfur trioxide formation and
lized, stainless-steel pressure vessel
fluidized bed combustion are charac- absorption, poisoning of mercury sor-
• Design and fabrication of the first
terized in a 4-inch (100-mm) diameter, bents by the presence of sulfur trioxide,
commercial reactor for nuclear
fluidized bed reactor. This apparatus and biases created by flue gas sampling
power generation
provides highly controlled conditions methods for various species. Figure
• Development of the large, welded-
for assessment of devolatilization rates, G-10 shows a schematic of the system.
wall, waterwall panel—a design
char reactivity, and fragmentation and After cleaning the particulates from
innovation later adopted in boiler
attrition properties of various solid the entering gas stream, various solid
manufacture throughout the world
fuels and additives. The reactivity and gaseous reactants are fed into the
and attrition attributes of different stream, depending on the needs of
DROP-TUBE FURNACE SYSTEM sorbents for sulfur capture also can the particular experiment. Interac-
Several fundamental facilities are used be established under actual fluid bed tions between sensitive species and
to characterize fuel properties and operating conditions. The ability to the walls of the reactor are reduced
the impacts of these properties on test under highly controlled conditions by injecting nitrogen or air through
the design and performance of steam in a fluidized regime at representa- microperforations in the reactor walls,
generating equipment. Fundamental tive combustion temperatures allows reducing the buildup of solids on
studies of the reaction kinetics of generation of information regarding the walls, and discouraging contact
solid fuels during combustion and fuel and sorbent behavior that is not between the walls and the main gas
gasification processes are conducted in possible to obtain through bench-scale stream. At the exit of the test stand
a drop-tube furnace system (DTFS) analytical techniques. This facility is is an advanced, miniature ESP that
(Fig. G-9). routinely used to determine the follow- removes any remaining sorbent or
Two DTFS facilities are available to ing information on fuels and sorbent: particulate in order to extract clean,
provide controlled, high-temperature bias-free gas samples for analysis.
environments that are representative of • Fuels
the range of suspension-fired com- Combustion kinetic parameters
mercial systems. The devolatilization, Emissions (unburned carbon, SO2,
combustion, and gasification reactions NOx, N2O, CO, and CO2)
that take place in the furnace can be Fragmentation, attrition, and
detected and quantified by analysis decrepitation (FAD) index
of the solid and gaseous reaction • Sorbents
products. The combustion kinetics of Sorbent decrepitation and attri-
a broad range of coals, petroleum coke, tion indices (SDI and SAI,
coal-derived synthetic fuels, and refuse- respectively)
derived fuel have been studied in the Sorbent utilization
Figure G-9 | Drop-tube furnace system
through a water-cooled dipleg into a (15 MW) in a pulverized coal-design of the Low NOx Tangential Firing
seal pot. The hot solids in the seal pot furnace configuration and a maximum Systems began at the Industrial Scale
may return directly to the combustor permitted firing rate of 90 MMBTU/h Burner Facility. (Fig. G-14).
through an insulated, stainless-steel, (26 MW) for gas- and oil-designed
solids return pipe, or a portion of the furnace arrangements. BOILER SIMULATION FACILITY
solids may be bypassed through one The furnace walls and heat-transfer The Boiler Simulation Facility (BSF)
of two water-cooled, fluid bed heat surfaces of the ISBF are cooled by a has many of the same capabilities
exchangers before returning back to surrounding, atmospheric-pressure as the ISBF described above. It is of
the combustor. The gases flow through water jacket. Selective refractory lining similar size and capacity as the ISBF,
a water-cooled heat exchanger, then of the inside furnace walls and control and it has similar refractory-lined,
through a fabric filter/FDA test sys- over the fuel firing rate are used to water-cooled walls. The BSF, however,
tem. The FDA system is a dry process maintain an appropriate furnace gas is designed to replicate the shape and
based on the reaction between SO2 time–temperature history as compared flow patterns of tangentially fired
and Ca(OH)2 in humid conditions. to that of the commercial furnace or utility boilers as well as their time-
Additional equipment for the FDA process being evaluated. temperature-stoichiometry history. In
test system include an FDA mixer/hy- The ISBF is used primarily for fuel many respects, with tangential firing,
drator, additive feed system, and FDA and near-field burner evaluation. Solid the boiler itself is the actual burner. In
reaction duct. fuels tested have included a full range most wall-fired burners, swirl is inter-
of sub-bituminous to bituminous nally generated in order to mix air and
INDUSTRIAL SCALE BURNER coals, plus petroleum cokes. Liquid fuel. In tangential firing, the swirling
FACILITY fuels have included standard indus- flow field takes place across the boiler
Alstom Power’s Industrial Scale trial oils, heavy oils such as asphalt, cross section and is generated by the
Burner Facility (ISBF) is a balanced- heavy oil/water emulsions, coal/wa- interaction of many fuel- and air-ad-
draft, front wall-fired combustion test ter fuels, and coal/oil fuels. Burners mission assemblies.
facility designed to replicate the time/ tested have been both wall fired, such The ISBF uses a single, large,
temperature/stoichiometry (mixing) as Alstom’s Radially Stratified Flame air- and fuel-admission assembly and
history of a typical industrial steam Core (RSFC™) burner, and many thus cannot generate this swirling flow
generator. All major aspects of an variants of the typical tangential firing field. It is intended only for optimiz-
industrial boiler are duplicated in the fuel- and air-admission assemblies. ing near-field performance. The BSF,
ISBF, including the radiative furnace For tangential firing, a scaled version in contrast, uses many, much smaller
cavity and simulated superheat/re- of a single-corner, single-elevation assemblies in a multi-elevation, tan-
heat and convective (economizer) assembly is tested in order to allow gential configuration. Its scale makes
heat-transfer surfaces. The ISBF has a comparison of different firing system near-field evaluation impossible, but
nominal firing rate of 50 MMBTU/h designs at near commercial size. Many it is very useful for testing large-scale
firing system variations, such as swirl
angle, overfire air location and quantity,
fuel biasing, and so on. The BSF has
been used to develop and optimize the
overfire air arrangements and flow rates
for the Low NOx Tangential Firing
System and TFS 2000™. Other test
programs have looked at fuel reburn-
ing, SO2 capture by furnace sorbent
injection, and NOx control by chemi-
cal injection (Fig. G-15).
Current plans are to modify the
ISBF and BSF for oxygen-fired test
Figure G-14 | Industrial Scale Burner Facility
Alstom Power Research and Development Facilities | G-
virtually any mineral, chemical, food, analyses. Large material lots are tested Efficient removal of these materials
or other material. Equipment also is on this full-size equipment to reduce requires substantial test and develop-
available to perform pre- and post- much of the uncertainty of scale- ment work, followed by extensive
processing chemical and physical up between bench- and full-scale field demonstrations.
equipment. Managing large produc- Sulfur emissions, for instance,
2.80e+01 tion quantities of test materials often typically are removed via washing of
2.66e+01
2.52e+01 uncovers handling difficulties and the flue gas with an aqueous, high-pH
2.38e+01
2.24e+01
2.10e+01
other problems that might otherwise slurry in order to capture and neutral-
1.96e+01
1.82e+01 be overlooked. ize the sulfur acids. The required
1.68e+01
1.54e+01
1.40e+01
Other pulverizer test facilities spray-tower equipment originally
1.26e+01
1.12e+01 include a 5 ton/h (4.5 tonnes/h) unit was developed in conjunction with
9.60e+00
8.40e+00
7.00e+00
in Windsor, Connecticut, based on the large combustion test facilities to
5.60e+00
4.20e+00 the HP mill product line (Fig. G-20). provide a source of combustion gases.
2.60e+00
1.40e+00
0.00e+00
This facility has been used to develop The laboratory test programs that
improved product classification sys- were performed provided informa-
tems, both static and dynamic, as tion needed to quantify gas-to-liquid
Figure G-17 | Optimized steam–water
separator circulating fluidized well as other improved components. absorption, product crystallization,
bed model It also has been used to support the absorbent reactions, and product
Clean Combustion Technologies pulverized fuel needs of the large properties affecting byproduct dis-
Figure: App G-17/ (MH) combustion test facilities (BSF, ISBF, posal. This information was combined
PMS5415 K
and MTF). with descriptive mathematical models,
which became part of the design
FLUE GAS EMISSION procedure now used for commercial
TEST FACILITIES flue gas desulfurization systems. Ad-
Undesirable emissions from power ditionally, again through testing and
plants take many forms, includ- modeling, this equipment has evolved
ing gaseous emissions (e.g., NOx over the years into other forms be-
and SOx), particulate matter, trace sides spray towers, including packed
elements (e.g., mercury), and CO2. towers and bubble beds.
Figure G-18 | Steam–water separator
flow test For ash removal, ESPs were
developed. These apply a charge to
Clean Combustion Technologies entrained ash and then use an op-
Figure: App G-18/ (MH) posing charge to attract the ash to a
PMS5415 K
collector. The material collected is used
in gypsum or cement manufacture or is
landfilled, where appropriate.
More recent particulate control
developments, such as fabric filters,
were developed to allow ash and sulfur
removal at minimum cost. These
were developed in test rigs such as
that shown in Figure G-21. Work to
improve the existing processes and
develop new methods continues on an
ongoing basis.
Figure G-19 | Naperville test mill Figure G-20 | Windsor HP test mill
many boiler operators. Alstom Power operating problems, or to be brought ELECTRICAL SYSTEMS FOR
chemists develop water treatment up-to-date on current technology in OPTIMIZING PLANT PERFORMANCE
procedures that are essential for maxi- the above areas. Alstom power plants are controlled
mizing the availability of fossil fueled As described in Chapter 9, steam by many levels of hardware and
steam generators. Their experimental generator tubes collect internal deposits software, both of which require
work includes the investigation of during operation (Fig. G-23). Because continuous improvement. Alstom
ion-exchange technology, operational these deposits may promote corrosion Power’s Electrical and Computer
and out-of-service corrosion protec- or cause tubes to overheat, it is neces- Engineering groups provide expertise
tion, chemical cleaning, and water sary to remove the deposits periodically for research and technology studies,
treatment methods. Boiler owners can by chemical cleaning. Based on sample new product or application develop-
take advantage of consulting services to analysis, Alstom chemists can determine ment, engineering consulting, and
have tube failures analyzed, to resolve the cleaning solution composition. general problem solving services.
These groups provide full capability
to perform new product/application
or concept studies and to develop
new products. They also provide
services, including field test instru-
mentation (Fig. G-24), measurement
and data acquisition and remote
monitoring, an in-house calibration
laboratory providing measurement
Figure G-23 | Extensive internal deposition and test equipment calibrations
traceable to the National Institute of
Standards and Technology (NIST),
Clean Combustion Technologies
Figure: App G-23/ (MH)
custom electronic circuit design
PMS5415 K and prototype boards, and software
engineering.
Sensors and Instrumentation mathematical modeling, control simula- time systems development, including
Sensors, instrumentation, and tions and transient analysis, advanced safety critical software application
diagnostic systems are routinely control methods (e.g., model based/ development in multiple languages
developed for unique power industry predictive, Linear Quadratic Regulator and operating systems; internet/web
applications. These capabilities include [LQR], Neutral Network [NN], and software technologies; collaboration
measurements and data acquisition; fuzzy), process identification and state software; artificial intelligence applica-
instrument calibration; sensor and estimation, optimal and non-linear tions; multimedia from video source
instrument product development; control, and control system simulations. through format conversion; scientific
vibration, acoustic, and stress/strain Advanced controls developed in this visualization; high-performance com-
measurements and analysis; machine area are applied in boilers as well as for puting; mathematical programming;
diagnostics and condition assessment; combustion optimization and control, information systems; and technical
instrumentation for experimental and nuclear control and protection systems, software upgrades (e.g., FORTRAN
field testing; digital signal processing and gas turbine dynamics. to JAVA/C++). This area also
and analysis; fiber-optic sensors; and supports Alstom Power Plant
real-time optical signal processing and Computer/Software Engineering Laboratories’ Scientific Computing
image processing. This area has a wide range of technical Environment (SCE) (with a Linux
capabilities, from real-time software/ cluster, 100-Mb local area network,
Process Control Systems systems, e-business, and Internet appli- several multi-processor systems, and
and Analysis cations to scientific computing. Areas virtual reality capability).
State-of-the-art processes require ad- of experience include computer system
vanced process control systems. Control configuration; network architectures; John Marion
system technology is advanced, using security; performance issues; real- David Towle
subject Index