EPRI Current Performance Fossil PP Monitoring Vol 06 Turbine Cycle PDF

crereli Topics: EPRI CSIEL-4415 Instrumentation Volume 5 Power plant availabilty Projects 1681, 2153 Fossil plant heat rate Interim Report Performance monitoring December 1987 Coalfired power plants Power Plant Performance Monitoring and Improvement Volume 5: Turbine Cycle Performance Analysis Propared by Potomac Electric Power Company Upper Marlboro, MarylandSUBJECTS Topics AUDIENCE BACKGROUND OBJECTIVES APPROACH RESULTS REPORT SUMMARY Fossil steam plant availability / Fossil steam plant systems and performance / Power system operations Instrumentation Performance monitoring Power plant availability CoaMfired power plants Fossil plant heat rate Generation and R&D engineers and managers / Power system operators Power Plant Performance Monitoring and Improvement Volumes 4-6 Utilities need improved instrumentation to optimize coalired power plant performance. This report presents results of an instrumentation and monitoring demonstration that includes boiler optimization, turbine cycle analysis, improved dispatch algorithms, and better communications between the plant and control center. Potomac Electric Power Company (PEPCO) is the host utility for this pertor- mance monitoring project, which has already produced a heat-rete improvement of 300 Btu/kWh at Morgantown unit 2. Volumes 1-3 of this report describe initial results of the work, including the relationship between component performance measurements, unit optimization, and determination of on-line incremental heat rate for unit dispatch. As a result of monitoring activities under the project, the Morgantown plant has realized significant cost savings. These savings are related to improved water-wall wastage measurement; on-line output/loss measurement of boiler performance; accurate measurement of turbine packing leakage flow; a reliable condenser back- pressure monitor; a high-accuracy, solid-state watthour metering system; and optimization of boiler operating parameters. To develop advanced instrumentation and monitoring systems for improving heat rate and plant availability and to demonstrate how the use of such systems can optimize plant operation and economic dispatch. Guided by a 26-member utility advisory committee, the project team monitors, and analyzes plant and component performance at Morgantown unit 2. Re- ‘search covers three areas: boilers and auxiliaries, including burner diagnostics, combustion optimization, and water-wall wastage; instrumentation te monitor the turbine, condenser, and feedwater heaters, including overall turbine cycle; and power system studies to improve economic dispatch and unit commitment. Volume 4 of this report describes performance measurement in the boiler. It discusses trade-offs between excess air level, coal size distribution, and stack gas temperature, and describes the HEATAT computer code for estimating the effects of fireside parameters on plant performance. The EPRI CSIEL44i5s Vols 4-6EPRI PERSPECTIVE PROJECTS volume also presents a refined method for the direct measurement of heat rate and improved techniques for the ultrasonic measurement of ‘water-wall tube thickness. Volume 5, on turbine oycle performance analysis, desoribes advanced instrumentation and computer programs for ‘online display of unit condition. Included are techniques for monitoring the performance of feedwater heaters and the main condenser, proce- ‘dures for planning turbine maintenance based on an analysis of pre- ‘outage testing and performance history, and an overview of the project's ‘computerized data handling and display systems. Volume 6 discusses incremental heat-rate analysis and presents the results of research on candidate dispatch algorithms. It includes recommendations for application of an automatic generation control system that includes a shortterm load predictor, a dynamic dispatch algorithm, and a coordinated controller between the load frequency controller and the dispatch algorithm. ‘Accurate measurement and analysis of power plant performance play a vital role in heat-rate improvement. The EPRI project at PEPCO’s Mor- gantown unit 2 is measuring component performance to identify areas of deterioration or nonoptimum operating conditions. Although work is still under way to improve the techniques developed in the project, uti ties should find the instruments, procedures, and models described in this multivolume report beneficial in the accurate measurement, calculation, and display of on-line power plant performance and in the determination of incremental heat rate for economic dispatch. For example, the dispatch algorithms developed in the project have the potential of returning annual savings of several milion dollars to a utility. Major proj- ‘ect emphasis is now on further use of Morgantown unit 2 in fine-tuning the performance analyses, refining field test methods, and researching areas such as on-line measurement of pulverized coal flow, coal fine- ness, and unburned carbon in ash. P1681, RPZ153 EPRI Project Managers: Robert H. Leyse; Neal Balu Coal Combustion Systems Division; Electrical Systems Division Contractor: Potomac Electric Power Company For further information on EPRI research programs, call EPRI Technical Information Specialists (415) 855-2411.Power Plant Performance Monitoring and Improvement Volume 5: Turbine Cycle Performance Analysis CSIEL-4415, Volume 5 Research Projects 1681, 2153 Interim Report, December 1987 Prepared by POTOMAC ELECTRIC POWER COMPANY Production Service Center 8711 Westphalia Road Upper Marlboro, Maryland 20772 Principal Investigators and Authors H.G. Crim, Jr J.C. Westcott R. W. de Mello. R. E. Brandon C. Kone 1.G. Schmeh! JR. Reddington Prepared lor Electric Power Research Institute 3412 Hillview Avenue Palo Alto, California 94304 EPRI Project Managers RH. Leyse Fossil Plant Performance Program Coal Combustion Systems Division N, Balu Power System Planning and Operations Program Electrical Systems Division‘ORDERING INFORMATION Requests for copies ofthis report should be directed tothe EPRI Dictibution Center 207 Coggins Drive, FRO. Box 23205, Pleasant Hil, CA 94523, (510) 934-4212. There is no charge for reports requested by EPRI member utes. cre Power Ressarch insu and EPRI are agitred service marks of Electric Power Research Inte, nc Copyright ©1987 Bectic Paver Research Inte, In. Al igs eserves DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES “TH REPORT WAS PREPARED BY THE ORGANZATION() NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED ‘OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTTUTE, IN. EPR. NETHER EPR, ANY MEMBER ‘OF EPR, ANY COSPONSOR, THE ORGANIZATIONS) NAMED BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, () WITH RESPECT TO ‘THE USE OF ANY INFORNATION, APPARATUS, METHOD, PROOESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (i) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PANATELY OWNED RIGHTS, INCLUGING ANY PARTY'S INTELLECTUAL [PROPERTY OR (i) THAT THIS REPORT IS SUITABLE TO ANY PARTICULAR USER'S CRCUMSTANCE: OR (@) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (NCLUDING ANY ENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBLITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS REPORT OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SMILAR ITEM DISCLOSED IN THIS REPO. ORGANIZATION) THAT PREPARED THIS REPORT Potomac Electric Power Company Upper Marlboro, Marylandcontenrs Section 1 MARK I DEMONSTRATION SYSTEM DESCRIPTION 2 COMPUTER SYSTEMS AND INFORMATION FLOW Computer Configuration Project Software Data Base History Files 3 CALCULATION OF ON-LINE PERFORMANCE INDICES Calculation of Performance Indices summary References 4 PERFORMANCE MONITORING OF MAIN STEAM CONDENSER Condenser Specific Data Current Performance Monitor New Performance Monitor Model Results Summary Reference 5 FEEDWATER HEATER PERFORMANCE MONITORING PEPCO's Morgantown Unit #2 Performance Monitoring varying TTD and DCA Model Program Description 2-21 2-23 3-1 3-27 3-28 4-1 4-2 4-3 4-5 4-11 4-17 4-22 4-23 5-1 5-2 5-3 5-6 5-7 5-14Section 7 PRE-OUTAGE TURBINE EVALUATION AND ANALYSIS PROCEDURE Goals Approach Pre-Inspection Turbine Perfomance Evaluation Appendix 7A - Steam Path Appraisal Appendix 7B - Closing Steam Path Appraisal Appendix 7C - Post Outage Performance Measurements and Analysis Appendix 7D - Other Diagnostics @ FLOW SECTION Flow Section Calibration Flow Section Specification References Page 5-17 7C-12-7 2-8 a1 4-2 4-3 44 ILLUSTRATIONS Computer Interconnections PI Computer System Data Storage Data Retrieval Data Transfer To and From CPI Computer system Input and Timing Control for On-Line Performance Calculations Disposition of Performance Indices Calculated On-Line Structure of History File Schematic of Input/Output Parameters for Condenser Analysis Model Condenser Cleanliness Factor as Function of Unit Load LP Turbine Exhaust Pressure Versus Unit Load Condenser Log-Mean Temperature Difference Versus Unit Load Condenser Cleanliness Factor as Function of Unit Load Temperature-Length Diagrams Input and Output Parameters of Feedwater Heater Performance Monitoring System for PEPCO - MU #2 Input and Output Parameters for Feedwater Heater Performance Monitoring with Variable TTD and DCA Analytical Model of One Typical Feedwater Heater Showing Three Sections Schematic of Rankine Cycle Power Plant with Single Reheat and Regeneration ‘Temperature-Entropy Diagram for Rankine Cycle with Single Reheat and Regeneration Design Flow Actual and Design Flows Turbine Pressures Page 203 27 2-12 21d 2-16 2-18 2-19 2-24 4-18 4-19 4-20 4-21 5-3Figure 7-4 Restriction Factor 7-5 First Stage Pressure 7-6 Change in Flow Area 7-1 HP Section Efficiency 7-8 HP murine Efficiency 7-9. Pre-outage HP Section Bfficiency 7-10. TP section Apparent Efficiency 7-11 Measured IP Section Béficiency 7H12 Blowiow System 7-13 IP Turbine Bfficiency 714 Shatt Rubbing 7-15 Solid Particle Erosion 7-16 Internal Danage ‘7-17 Steam Path Deposits @-1 Assenbly Drawing @-2 Material List 6-3. Part No. 1 ~ Flow Straightener Section G4 Part No, 2 ~ Flow Straightener 8-5 Part No, 3 ~ Flow Nozzle 8-6 Part No, 4 ~ Diffuser section e-7 Part No. 5 - Diffuser 8-8 Tap Set #3 Calibration Curve 8-9 tap Set #3 958 Confidence Interval of Calibration Curve i ya 7-17 7-20 7-21 7-23 7-25 7-26 7-28 7-29 7-39 7-40 7-41 1-43 e-11 e-12Section 1 MARK I DEMONSTRATION SYSTEM DESCRIPTION ‘The system description which follows is undergoing initial utilization as this document: goes to press. Publication of detailed cycle or component analysis tech niques and results will evolve over the next reporting period (1986-1988) of the project. ‘The automated demonstration system implemented under RP 1681-2/2153-2 is the primary vehicle through which the project goals will be accomplished. These goals are to establish a state-of-the-art performance monitoring system and then to use it for the several functions of performance measurenent, performance optimization and on-line unit dispatch. ‘he systen includes the data acquisition and software analysis capability as necessary to measure overall unit heat rate as well as measurement of the perfor~ mance of the individual systems which contribute to it including: boiler, turbine, condenser, feedvater heaters, and boiler feed pump and turbine systan. ‘he foundation of all the performance calculation and analyses is a full cycle flow and energy balance. Initial component analysis capability continues to be upgraded as the system is exercised. 1‘The lowest Level of measurement detail will allow analyses at the component level 0 as to indicate the components of concern and thereby more quickly lead the Performance Engineer to the cause of the degradation. When this capability is utilized in relation to unit maintenance outage planning, the impact on overhaul scope, schedule and long lead part ordering can result in significant cost savings. Multiple approaches to boiler efficiency and cycle heat rate measurement are in the process of being implemented and tracked including input/output, ASME loss, total boiler stack loss (to the exclusion of coal or flue gas constituents), and cycle loss via condenser duty (to the exclusion of turbine cycle analysis). The high quality input sensors utilized in the data acquisition systen include pressure, temperature, flow, level and electrical generation with the accuracy and resolution required by the analysis performed. Primary turbine cycle condensate flow is measured by an ASME type throat tap nozzle and flow straightener section section which was calibrated by Alden Research Laboratory of Worcester Polytechnic Institute ani is read out by a quartz/frequency type differential pressure transducer which yields an overall floy measurement error of no more than +0.2%. Other systan flows and pressures are monitored via high pressure transducers housed in a temperature controlled rom for elimination of temperature drift errors and ease of calibration and maintenance. Critical temperatures are monitored by high accuracy chromel constantan ther- mocouples or 100 ofm platinun four-wire resistance temperature devices (RTDs). ‘Custom RID temination circuits maintain very constant RID excitation current which is also accurately measured at RID scan time to assure the best possible accuracy. 12To assure the accuracy of the conversion of analog inputs to digitized engineering units, the new A/D converter was specified to be 13 bits plus sign. ‘this assures adequate resolution for the more critical inputs even at times of minimm unit load or flow where the lowest percentages of usable sensor ranges are encountered. Blectrical generation measurement accuracy has been assured by a thorough analysis of the total watthour metering systen incluling potential transformer calibration by the U.S, National Bureau of Standards, use of a fully three-phase solid-state watthour meter accurate to within +0.018 error, and incorporation of overall sys~ ten calibration factors which account for burden and voltage drop in the circuits between the current and potential transfomers and the watthour meter. Several EPRI "First Use" documents have been issued on project products including the watthour metering systen, a condenser back-pressure tip sequencing and purging systen, an HP/IP turbine packing (N2) leakage flow measurement systen and a microprocessor-based systen for processing vatthour meter pulses, controlling the basket tip sequence ani for detemnination of equivalent turbine control valve position via pressure ratio across designate turbine sections. The overall computer system developed for the project is a cost effective hybrid of the original GE/Honeywell process monitor and a new Computer Products, Inc. distributed, highway type system which provides an expansion of input data ac~ quisition capacity and a CRT based Operator/Engineer display capability not avail— able in the original G&/Honeywell computer. These two major systems together with several snaller microbased special-purpose data collectors all transnit their scanned data to a comprehensive data storage and retrieval (DSR) system housed in a PRIME 750 mini-camputer donated to the project by PEPCO. DSR is the critical project tool which enables the project to store both raw and calculated data in an orderly manner. DSR provides historical trending over time and the ability to extract parametric relationships over the load or flow range as necessary for 13characterizing the mit for economic dispatch or for the various performance analyses. DSR also allows the implementation of a diagnostic philosophy which routinely computes, stores and prepares for display all preconceived relationships Which would be useful or necessary in the event of an operational incident. Thus a major portion of the preliminary date collection and manipulation effort is automatically accomplished for the Performance Engineer or operator when faced with the diagnosis of an incident. Extensive graphics generation and display systems are resident in both the Prime and CPI camputer systems. these tools are the primary vehicles for display of the results of the major analytical effort within this project. At least two uses of the graphics system are envisioned: on analytical tasks such as burner balancing, air heater optinis ‘ion and auxiliary power reduction, the gxaphics generation capability is available to and used by the task contractors to review and present task progress and to generate unique displays. On contxactor tasks whose main purpose is to develop routine monitoring and analysis techniques, the resident PSFCO project team creates and tests out, over time, the emineer and operator displays to best depict the current condition and perfomance history of the plant. The logical extension of perfomance measucanent is the optimization of the specific subeystan or total mit. Relative to the stean generation system, the parametric analysis capability and the monitoring systen installed will be utilized to continuously optimize the boiler on-line and will be updated as increments of improvement becone available from the various project actions. Major actions to date include: (1) combustion optimization via individual burner fuel/air ratio balancing using flue gas analysis. This allows more econanic operation at lower average excess air while minimizing carbon loss, stack loss, fan power and S03 generation; and (2) air heater optimization utilizing special instrmentation 14which allows the lowest econanic exit temperature involving consideration of steam heating, fan power, air heater pressure drop, ash deposition, acid dew point, and longer tem maintenance. Turbine cycle optimization capability will select throttle pressure and temperature, attemperation spray flow, control valve position and mumber of auxiliaries in operation, Controllable parameters for the boiler ani turbine cycle will be displayed via the graphics system and will emphasize those itans not under autmatic set point con— trol and therefore requiring more operator attention such as when and where soot should be blown. The measurement syste will be utilized to characterize the host unit's increnen— tal heat rate curve shape over the load range for implementing on-line updating of the curve shape in an initially constrained mode using continuous segments and positive, monotonically increasing coefficients. 1sSection 2 COMPUTER SYSTEMS AND INFORMATION FLOW Electric Power Research Institute (EPRI) project RP 1681/2153, “Power Plant Performance Instrumentation System," has the broad goals of monitoring power plant performance, pinpointing causes of unit deterioration, identifying solutions to deterioration of plant performance, and enhancing the econamic dispatch of generating units. Potomac Electric Power Company (PEFCO) is the host utility for the project and Morgantown Generating Station Unit 2 is the test unit. Meeting the goals of the project would be impossible without the use of digital computers. Computer requirenents fall into two broad cateyories: (i) the real-time requirements of monitoring instrumentation, alarming abnormal conditions and displaying the current state of the plant; and (ii) the general computing requirements of a research project where "new ideas" can be tested quickly and efficiently. the first category falls within the traditional boundaries of event-driven, real-time power plant computer systems. The second category is typical of multi-user, time-sharing computers with the added requirement that data monitored by the real-time computers must be collected and saved for subsequent analysis. 21‘The simaltaneous requirenents of real-time computer operation ania tine-sharing environment are difficult to implenent ina single computer systen. Intead, several campater systens are being used for the project. ‘These inclase the computer originally installed at Morgantown, a GE-PAC 4020, a new distributed computer sys~ tem purchased fron Computer Products, Inc. (CPI), a general purpose tine-sharing inicanpatex (PRIME 750) and several microprocessor based systens for data acquisition. ‘this document provides an overview of the computer configuration at the Morgantown plant and programs that have been developed for the PRIME 750 computer system to © Gather and save process measurements. © Calculate various performance parameters fram process measurements. © Review and analyze measurenents and/or calculations that have been saved. © ‘Transfer measurements and/or calculations to the CPI computer sys— tem for display, alarming, logging, etc. COMPUTER CONFIGURATION ‘The project requires computers to: (i) monitor the plant's instrumentation; (44) provide the camputational capability needed to implenent the monitors and diagnostic programs developed during the project; (1ii) provide the man/machine interface between the plant and the operators and performance engineers at the plant; and (iv) provide Links to remote computers for the exchange of information. ‘The computers used for the project and their interconnections are shown in Figure 2-1, 22Inte, ne, vB ce-pac ea ICRO cas cr Drstaigureo cruTuney ‘SYSTEM para ACOUISITION Paine 780 LU ¥ DATA STORAGE, CALCULATIONS & ENGINEERING FUNCTIONS CONTROL ROOH INFORMATION & GUIDANCE Figure 2-1. Disrarcn CENTER OO Disraten cenren TMFORMATION & GUIDANCE Computer Interconnections 23PRIME 750 Computer [A PRIME 750 computer, previously purchased by PEFCO, is being used to provide gen- exal computing resources to the project's engineers and to provide the interface among the other computers being used. the camputer is configured with: © Two 675 megabyte disks; © One 80 megabyte disk; ¢ 6.5 megabytes of main manory; © Asynchronous, bit-serial communications ports to accannodate user's ‘teminals; © Several low speed Line printar: © A high speed electrostatic printer/plotter; © Asynchronous, bit-serial communications ports to serve as links to © Magnetic tape for backup and archival purposes. ‘The PRIME 750 is also used to run performance calculations, to develop and test new algorithms, to generate reports and presentation graphics and to run various engineering analysis programs. GB-PAC_4020 Computer ‘The computer originally installed for units 1 and 2 at Morgantown, a @E-PAC 4020, scans over 400 instruments on each of the two units that measure the various pressures, temperatures, flows, speeds and levels. In addition it monitors the state of about 700 limit switches on each of the two units. The man/mchine interface of the computer is crude by today's standards, there are no CRT Furthemore, the ability to install new functions or expand the mmber of monitored points is limited by the lack of manory and free time. A computer to computer data link has been installed between the 4020 and PEPCO'S PRIME 750 computer to take advantage of the existing instrumentation, Honeywell was contracted to install the data link in the 4020 and provide the necessary programming changes 24to make it work. CPI Distributed Computer systen he CPI distributed computer system permits the implementation of new instrumentation, data transfer to and fron PEPCO's PRIME 750 computer and provides operator and engineer an interface to the process using color CRIs with keyboards. A distributed system is one that provides a functionally divided solution to a specific application, The functions, assigned to relatively autononous hardare components or "boxes," are divided as follows: © Input/oatput: All input and output functions are provided in one unit. For instance: the acquisition of raw counts for an analog inpat; correction; conversion to engineering wits; alam processing (not necessarily alam reporting); and transmission on the data highway. © Logging: Logging functions include the gathering of data to be logged, processing of data, archival of data in non-volatile menory ‘and printing of logs on output devices. © Operator Displays: Display functions should include the storage of all displays, processing of displays, alarm reporting, operator interface and acquisition of data from the data highway. ‘his functional division provides the ability to design the system using a build— ing block approach where individual functional units, or "boxes," are incluiel as required. In addition, expansion in the field is very easy and allows for an al— most bunpless transition between an old system and a new expanded systen. A distributed data highway system eliminates the problens of the past associated with central computer based systems vhere the computer was the bottleneck of the system. Expandability had been determined by the computer systen's free tine, In a distributed data highway system, the focal point of the systen is the data high way and its ability to pass data rapidly. If the band width and the communication schene cannot handle large anounts of data then the data highway becomes the Limiting item. It was of utmost concern to the project that the data highway not 25became a bottleneck in the system. The data highway of the CPI system operates at 10 megabaxd (10 million bits per second) . Major components of the CPI computer system are shown in Figure 2-2, ‘the data highway, these are: apart fran © Operator's Station: ‘The operator's station is used to monitor the plant and generate reports. It consists of a microprocessor, a 40 megabyte Winchester disk, a video generator, printer, black ani white CRT, two color CRTs and various keyboards. © Eagineer's Station: the engineer's station incluies all the functions of the operator's station and pemits modification of the data base describing the points being monitored. In addition, logs and graphic displays are generated at this station, ‘The station is configured like the operator's station with the addition of magnetic tape and high speed printer. ‘Input/Output Boxes: Bach input/output box controls the scanning, limit checking, alarm checking, conversion to engineering units and calibration correction of a ‘set of points, Use of a fiber optic Link pemits flexibility in the placenent of one of these boxes. © Gateway Box: The gateway box consists of a microprocessor and an interface compatible with the PRIME 750 computer. © DEC VAX 11/730: A minicomputer is used for system software development and maintenance, It can also be used to mm applications that require extensive computational capabilities, INTEL Microprocessor A microprocessor provided by PTT contains data acquisition equipment and software to implenent a: © Pressure Ratio Monitor - which monitors main steam and first stage conditions and, with sone manufacturer's design data, calculates the position of the unit's control valves. © Bxhaust Pressure Monitor - which controls the purging and sequen- tial reading of pressures at the turbine exhaust. Pressures read at individual basket tips are averaged to provide turbine back pressure. © Watthour Meter Interface - which receives pulses from the project's watt hour meter. wo minute and ten minute average generation is calculated for each phase. 26OPERATOR'S STATION ENGINEER'S STATION uo [C). Cy) cess race en 85, Ne ORM ION INFORMATION m o ‘center caren | | eH soe L GB corey oo ! oer aire rave ‘COMPUTER pox id FIBER OPTIC tine v0 v0 ferore 79] a tT PROCESS. MEASUREMENTS, Figure 2-2. CPI Corputer System 2-7Data monitored by the microprocessor system and results of its calculations are sent the project's PRIME computer. KVB Gas Analysis ters ‘wo KVB gas analysis computers are being used to monitor the concentration of C0, 02 and NO, at each of 24 probes in each of the furnace back passes (at the econamizer outlet duct). ‘The information being monitored will be used for burner fuel/air ratio monitoring and optimization, Chemical concentrations gathered by the KVB computers are sent to the project's PRIME computer. PROJECT: SOFTWARE Programs and other software developed for the PRIME 750 computer are listed below with a brief description of their purpose and function. © CPILNK: CPILNK is an interactive program used to transfer files between the PRIME and CPI computer systens. © DB: DB is an interactive program that automates the linkage of user written programs with the data collected by DSR. Linkage can be made to either live (real-time) data that resides in shared man— ory or to old data extracted fran a History File. © MSC MSC is a library of miscellaneous functions and subroutines that are used in other programs. MSC contains inpat, output and parsing software as well as sorting, searching, statistical and other useful subroutines. The inter-task communication software used by some of the programs to pass messages anong thenselves is part of the MSC software library. © PC: PC (aka TASK6) is a program that contains the real-time, peri— odic perfomance calculations. Data are collected by TASKL, TASK2, TASK4, etc. and placed in a shared memory area where it is acces— sible to EC. After placing data in shared manory each task senis a message (using the inter-tast camunication software) to inforn PC that new data is available. PC averages data as they are collected. At ten minute intervals performance indices are calculated, The results are placed in the shared memory area and sent to TASK3 to be saved in History Files. 2-8© TASKO: TASKO is an interactive program that is used to examine data captured by the data acquisition programs (TASK1, TASK2, TASK4, etc.) and saved by TASK3 in History Files. TASKO is the “retrieval” portion of DSR. @ TASKL: TASKL is a progran that receives messages fron the Honeywell 4020 computer. Each message contains value and status of the analog inputs, and state and status of the digital inputs currently being monitored by the 4020. © TASK: TASK2 asks for and receives data fram the custan microprocessor used to implanent the Pressure Ratio Monitor, the Exhaust Pressure Monitor and the Watthour Meter Interface. © TASK3: TASK3 has the job of saving data in History Files on disk. Data are passed to TASK3 in the form of messages fran data acquisition tasks. After data is received TASK3 writes it into a History File. © TASKS: TASK4 receives data from the VB gas analysis computer. Each message received contains the concentration of C0, 05 and NO, at one of 24 probes. The probe mmber is included in the “message. © TASKS: TASKS asks for and receives data fran the CPI computer system. TASKS also has the job of sending data to the CPI computer system so they may be displayed in the control roan. © TASK6: See FC above. Software functions fall into the categories of data acquisition, data storage, data retrieval, data movement and the calculation of performance paraneters. In. addition software designed to pass infomation (messages) among programs running concurrently on the PRIME computer system has been written, and some interactive utility programs have been written. Data Acquisition TSKL, TASK2, TASK4, TASKS and TASK? are data acquisition programs. Bach program gathers data fran an external computer systen and updates an area of shared menory with the data, The prograns also: © Sond a message to TASK3 with data to be saved in the History File.© Compress data as required (averaging, significant change, etc.) and send a message to TASK3 with data to be saved in the Compressed History File. © Signal PC, the on-line performance calculations, that new data have been placed in shared memory. Data Storage One of the goals of the project is to identify the information that should be collected and saved to be able to detect deteriorating performance and identify the causes. Other parts of the project, the on-line monitoring of performance or incremental heat rate for example, benefit fran being able to use historical as well as on-line data during development and testing of algorithns. Enormous amounts of data will be monitored throughout the course of the project and it is impossible to select at the outset the subset of these data that will be needed as the project progresses. Therefore a systematic means of saving all data is required. Data Storage and Retrieval (DSR) programs have been developed to meet the requirements of automatic data storage. ‘The collection of data is an autanatic function: once started, data is collected ‘and saved without the need for human intervention, It is estimated that neasure- ments fron as many as 1000 instruments mst be saved each minute of each day for the duration of the project. At this rate about 80 megabytes of information will be collected each week. Saving all data has the advantage that nothing is lost. It has the disadvantage however that it is difficult to locate a particular datun ‘because of the volume of information available. 2-10Data compression techniques can be applied to reduce the volme of infomation collected. te value of a monitored point may be saved only when it changes significantly. Alternatively a 10-minute or 30-minute average may be saved. For sone points, points unimportant to the performance monitoring of the plant, noth— ing may be saved. Searching and sorting the reduced data set is much easier. ‘Same data however are inevitably lost. DSR uses both techniques and saves everything in one file while saving a com pressed set of data in another file, Thus the compressed set of data can be quickly searched for events of interest ani used as an index to the full set of data. TASK3 acts as a funnel between the data acquisition tasks and the History Files where data is saved. ‘The data acquisition tasks send messages to TASK3 with data to be saved. Messages contain the data, status of the data and the time that the data was collected. Figure 2-3 illustrates the flow of infomation fron the data acquisition tasks through TASK3 for storage in History Files. Data Retrieval Retrieving data for review and/or analysis is a highly interactive process shere an engineer, using CRT and keyboard, instructs TASKO, the DSR program. An en— gineer may, for example, want to plot a histogram of turbine bearing vibration versus turbine speed Guring startup. A startup must be located in the volumes of information available and selected portions extracted for plotting. DSR provides the ability to interactively: © Select a data file (History File) to examine. this may be data that was gathered only moments previously or many weeks or months ago. auATA FROH MONETWELL 4020 ‘TASK ATA FRON Inter TASK MICROPROCESSOR TasKe — NESSAGE SOFTWARE ea carn ConpnesseD nara Figure 2-3. Data Storage 212© Select a subset of points to examine or output: DSR contains well over 1000 points. It would be cumbersome if they all had to be viewed regardless of the the current interest. @ ‘Search for events: For example, pressure Greater than 3200 PSIA. nd the next instance of throttle © Select time intervals of interest. © Calculate simple statistics: average, standard deviation, 95% confidence interval of the mean, maxinum value, minimum value and the number of samples in a given time interval. © Generate plots of points against time. © — Generate plots of points against one another. © Generate data files for external analysis programs or presentation graphics prograns. © Run performance calculations such as HP turbine efficiency, IP turbine efficiency or a complete heat balance. TASKO provides an interactive method of searching the saved data and selecting and retrieving relevant portions as needed. 0SR provides functions to find, print and plot data. It also contains statistical routines and performance calculations that can be run interactively using data extracted from History Files. Figure 24 illustrates the inputs and outputs of TASKO, these are: © CRT & KEYBOARD: TASKO is an interactive program. Users of the program interact with it using the CRT and Keyboard. With the Keyboard users tell the program what to do (input). The CRT is used for confimatory messages, error messages and listings of data (output) - Command input, that normally comes from the Keyboard, may also come fron a Command File. Listings of data may be redirected to a Listing File, however, error messages and confirmatory messages are always directed to the CRT. © COMMAND FILE: A Command File is a file that contains the comands nomally typed at the Keyboard. TASKO may be directed to read its commands from a Comand File rather than fron the Keyboard. © DATA BASE: The Data Base is a file that contains a description of the points that are being monitored and whose values are being saved in History Files, the Data Base resides in the file DSRDB>MORGANTOWN and is Used to initialize TASK. 2-13Een,HISTORY FILE: A History File is a collection of point values and status maintained in chronological order. TASKO may read data fron one History File at a time. Portions of one History File may also be copied to another History File. The History File being read is called the "FROM" History File, ‘he History File being written into is called the "10" History File. A data sample may be obtained by averaging the information in the FROM History File over a period of time. This data sample may then be used as input to various performance calculations, LISTING FILE: Listings of point values and status, average values, statistics, point nanes, point definitions, etc. that normally appear on the CRT may also be directed to a Listing File. Error mes- Sages and confirmatory messages are always directed to the CRT and are never written into the Listing File. SAVE/RECALL FILE: A data sample is obtained by averaging information taken over a period of time from the FROM History File. The data sample may also contain manually entered information and/or information calculated by some of the performance functions. The data sample may be saved in a file for later recall. Data Movement ‘two prograns, CPILNK and TASKS, transfer data between the PRIME and CPI computers. CPILNK is an interactive program used to transfer files between the two computers. TASKS, in addition to acquiring data fram CPI, periodically sends measurenents and/or calculations to the CPI camputer system. ‘The flow of measurements from the CPI computer to the PRIME and the flow of calcu— lated results from the prime to the CPI is illustrated in Figure 2-5. The inputs to and outputs of TASKS are DATA BASE: The Data Base is a file that contains a description of the points that are being monitored and whose values are being saved in History Files. The Data Base resides in the file DSROB>MORGANTOWN and is used td initialize TASKS. DATA FROM THE CPI SYSTEM: ‘The analog inputs scanned by the CPI system are periodically requested by TASKS and sent over an asynchronous, bit-serial (RS-232-C) data link. 215aTA To & FROK a a } a se Th Inten-rasx wessace SOFTHAAE, suane NEnORT ‘AREA | TRsK3 mistory FILE ‘ Figure 2-5. Data Transfer To and From CPI Computer System 216© DATA 10 THE CPI SYSTEM: ‘The analog outputs are periodically read fron the shared memory region and sent to the CPI systen over an asynchronous, bit-serial (RS-232-C) data link. @ HISTORY FILE: Information gathered by TASKS is placed in a History File for future use. Data are passed to TASK3 using the inter-task message Software described in the manual "MISCELLANEOUS SOFTWARE". ‘TASK3 puts the data into the History File. ‘here are actually two History Files containing raw data and con pressed data respectively. Faw data is data that is saved as it is read. Compressed data have been averaged or somehow compressed before being saved. TASKS performs data compression (for example, Lo-minute averaging) and sends these as well as raw data to TASK3, © SHARED MEMORY: ‘The current values and status of the AI points are placed in a shared maory region as they are received. The current values of the AO points are read fram a shared manory region before they are transmitted . Other programs, such as the Performance Calculation Program, may access this “live” data, Performance Calculations ‘wo programs, TASKO and FC (aka TASK6), contain performance calculations. " TASKO (the DSR program) is an interactive program that pemits data to be extracted fran History Files and used as input to performance calculations; calculation results are directed to the CRI. FC is a non-interactive program that calculates perfor~ mance indices using live (real-time) data. The results of FC's calculations are placed in a shared manory area where they will be accessible to other prograns and archived for later use. Figure 2-6 shows the flow of information into PC; Figure 2-7 shows the disposition of PC's calculated results. The following items are illustrated: @ DATA ACQUISITION INTERFACE TASKS: Each interface task receives process measurements fran one of the external computer systens. This data is placed ina shared manory area where it will be accessible to BC. Each time the shared memory area is updated with new data a message is sent to PC to alert it to the new data. © INTER-TASK MESSAGE SOPIWARE: he inter-task message software permits the various tasks (each an independent progran) to communicate with one another and to synchronize thamselves with one another. 217ara ara ACQUISITION AcouTsIT10N conpuren INTERFACE ‘SYSTENS, "TASKS Figure 2-6. ON-LINE STNCHRONIZATION [MERSURENENTS RESSAGES InTER=TASK ‘SHARED HENORY NESSACE SOFTRARE, <} cme FORMANCE LES ‘Input and Timing Control for On-Line Performance Calculations 2181 rarvate owe ne ‘nose cme ONS sLacK é checunreo PERFORMANCE INDICES ‘SHARED MEMORY crass) coo ntfftace wisroninw Inren-rask MESSAGE SOFTMARE Pr conputen svSTEH mtstonr FILE CONTROL ROOM DISPLATS, ARCHIVAL OF ‘SCAM, LOG € ALARM FUNCTIONS PERFORMANCE inorces. Figure 2-7. Disposition of Performance Indices Calculated On-Line 2-19SHARED MEMORY: ‘The current values and status of all the monitored data points are placed in a shared memory region as they are received. FC uses the shared mamory region as a source of data. Upon receiving messages fron the data acquisition interface tasks PC will draw data fran shared manory to update its private menory block. Upon completing its calculations PC puts its results back into the shared memory area where they will be accessible to other tasks, PRIVATE MEMORY BLOCK: This is the location of data collected by PC. This manory region contains running averages, maximum and min— imum values, the accumlated status of each data point and any preliminary calculation values which are to be used for the performance calculations. ‘The private mamory block is continuously updated by PC for a ten minute interval. The data accumulated during this ten-minute interval is used by FC to execute the perfomance calculations. Upon completion of the performance calculations, the private menory block is re-initialized to receive new data for the next ten-minute interval. HISTORY FILE: Selected calculation results are placed in a History File for future use. A message is sent to the History File Writer (TASK3) to signal the completion of a set of performance calculations. ‘The message delivers all the data to be stored in the History File. Performance indices may subsequently be recalled to look for trenis (or abrupt changes) and used as an aid in detemining the state of the various components in the plant. CPI COMPUTER SYSTEM: Results of the performance calculations are periodically removed fran the shared memory area and sent to the CPI computer systan. Calculated perfomance indices are therefore available for display to operating personnel and may be logged and/or alamed using the standard features of the CPI canpiter system, Proprietary software Software developed for the Prime computer system incorporates the proprietary gram packages described below: INFO, a data base package of Henco Incorporated, 35 Walnut Street, Wellesley, MA 02181, (617) 237-4156. INFO is used to maintain a data base of the points being stored, the frequency at which individual points are stored, description of points, etc. INDEPLOT, a graphics package of Power Technologies, _Inc., P.O. Box 1058, Schenectady, NY 12301, (518) 374-1220." INDEPLOT. provides device independent ‘graphics capability. pro-In addition, the following proprietary software products augment the osk's functions: © © MINITAB, a statistical package of Pennsylvania State University. MINITAB provides statistical capsbilities and regression analysis. © STP, Steam Tables Progran of Power Technologies, _Inc., P.0. Box 1058, Schenectady, NY 12301, (518) 374-1220. STP’ provides steam property functions. DATA BASE A data base of DSR points is maintained in the UFD named DSROB. INFO, a data base program, is used to add points to the data hase or change data describing existing points, When DSR programs (TASKO, TASK, TASK2, ...) are started they read the contents of the data base to initialize tharselves. Described below are the itens in the DSR data base. POINT-ID A 12-character point name that uniquely identifies each data point. DESCRIPTION A 40-character description of the point. EU ‘A 16-character text string containing engineering units of the point, for example 'DEG-F' or 'PSIA', For digital inputs text strings such as '0-ON, 1-0FF' or ‘O=CLSD, 1-NCLSD' may be used. GROUP ‘An integer defining the source of the point. The following GROUP numbers are recognized by DSI 1 = 4020 Digital Input Group No. 2 = 4020 Digital Input Grow No. 3 = 4020 Digital Input Grow No. st 100 101 = Pressure Ratio Monitor 102 = Exhaust Pressure Monitor 103 = Watthour Meter Interface (2-minute data) 104 = Watthour Meter Interface (10-minute data) 106 = KVB Carbon Monoxide concentration 107 = KVB Oxygen concentration 108 = KVB NOx concentration 109 = PIT Performance Analysis Systen 200 = Manually Input Data 221G-1NmeX CPI-FREQ PILE-IDENT FREQ PLOT-HIGH c-rvPE An integer identifying the point within the GROUP. The com bination of GROUP and G-INDEX must be unique for each point. An integer giving the tine interval, in seconds, that data for a point is to be sent to the CPI computer systen (inverse of frequency). If CPI-FREQ is 0 (zero) nothing will be sent. An integer (between 1 and 2048) that is used to identify the point within a History Pile. FILE-IDENT must be a unigue nun ber for each analog input, digital input or point calculated with DSR. Once FILE-IDENT is assigned to a point it mst not be changed. If PILE-IDENT is changed the data for that point previously saved in History Files will no longer be recognized by DSR. An integer giving the time interval, in seconds, that data for the point is to be saved in a History File (inverse of frequency). If FREQ is 0 (zero) nothing will be saved. Data are never saved more often than they are received. If value and status of a point are received every 15 seconds and FREQ = 1, the value and status will be saved in a History File every 15'seconds: every time anew datum is received. If FREQ = 30, the value and status will be saved every other time ‘that a value for the point is received. The frequency that data is received from the Honeywell 4020, ‘the Pressure Ratio Monitor, etc. cannot be changed by making adjustments in the data base. Only the frequency that data is saved in History Files cen be adjusted. A real nunber defining the default low plot limit to be used when plotting a point. A real mmber defining the default high plot limit to be used when plotting a point. An integer defining the type of data compression technique to use on the point, Allowable data compression techniques are: 0: None, nothing is saved in the Compressed History File; 1: A 10 minute average is saved; 2: The maximm value detected in a 10 minute interval is saved} 3: The minimun value detected in a 10 minute interval is saved; A 30 minute average is saved; ‘The maximum value detected ina 30 minute interval is saved; ‘The minimun value detected ina 30 minute interval is saved, The point is saved when its value changes by a significant amount. The amount of the change, in engineering units, is stored in C-AUX. 2-22c-ROX A real number containing auxiliary information needed for data ‘compression. HISTORY FILES Data that are monitored are written into History Files. ‘wo History Files are maintained: (i) one contains raw data; and (ii) one contains data that has been averaged or somehow compressed. History File (Raw Data) Each Raw History File contains one week's data starting Monday morning, just after midnight, and ending Sunday night, just before midnight. Weekly intervals were chosen because: (i) One week's data will fit onto one magnetic tape for permanent storage; (ii) Several weeks' data will fit on the computer's disks at one time; and (iii) Weekly data collection allows a convenient routine of magnetic tape backup. Each History File is a UFD (User File Directory) or a sub-JFD containing files TIME, DATA and TITLE, The TITLE file contains descriptive information about the data collected. the TIME file contains the time that each group of data were col~ lected and a pointer to the start of the data. ‘he DATA file contains the collected data. The name of the History File is actually the name of the UFD or sub-UFD that contains the TIME, DATA and TITLE files. Figure 2-8 shows the structure of a History File.Descniprive <—_—_——_. Text TITLE FILE ¥ ——— Arine——Prounten Figure 2-8. stars tyaue IDENTIFICATION Structure of History File 224 area D420‘A naming convention is used for the Raw History Piles that are generated autaati— cally by TASK3 as data is collected. Each of these History Files is a sub-UfD in the UFD = DSRHIS with a name that identifies the week and year of the data, for example: WEEK-31-1983 or WEEK-07-1984. ‘The full tree names of the TIME, DATA and TITLE files would be: DGRHIS>WEEK-31-1983>TIME or DSRHISDWEEK-07-1984>TIME DSRHISSWEEK-31-1983>DATA or DSRHISOWEEK-07-198420ATA DSRHISSWEEK-31-1983>TITLE or DSRHISWEEX-07-1984>TITLE, le (Compressed Data) Bach Compressed History File is a monthly version of the weekly History Files generated by DSR. Monthly infomation is compressed before being saved. One of eight categories can be selected. These categories are: 10-minute average, lo-minute maximum, 10-minute minimm, 30-minute average, 30-minute maximum, and 30-minute minimam, save on significant change, or save nothing. ‘The Compressed History File has the same structure as the Raw History File shich is detailed above and show in Figure 2-8. ‘The naning convention of the Compressed History Files is by month and year. Bach of these History Files is a sub-UFD in the UED = DSRHIS with a nae that iden tifies the month and year the data vas collected. the month is specified with three letters, for exemple: JUL8$ or MARB3, The full tree names of the TIME, DATA, & TITLE files would be: DSRHISDIULE4>TIME oF OSRHISDMARSS>TIME DSRHISDIULB490ATA or DSRHIS>MARE3>DATA DSRHISDJUL427ITLE or OSRHISIMARGI>TTTLE 2-25History File Structure ‘The three files (TITLE, TIME and DATA) that make up a History File are described below and shown in Figure 2-8, TITLE FILE: The purpose of the TITIZ file is to store descriptive information about a History File. By default, the TITLE file will contain the time interval of the data stored in the History File, for exemple: ‘WEEK 3, 1984 FROM: “MONDAY 1/16/84 00:00:06 TO: SUNDAY 1/21/84 23:5 ‘The TITLE file may be modified with the editor. TIME FILE: Each time data is vritten to the DATA file and entry is, made in the TIME file describing the time that the data were writ~ ten and the location of the data. ‘The TIME file is made up of a series of &-byte records containing (i) Time (4-bytes) in the form of seconds past 00:00:00 on January 1, 1983; and (ii) A pointer (4-bytes) to the record in the data file vhere data collected at that time starts. DATA FILE: Data are written into the DATA file as a series of @-byte records. Bach record contait 1) An identification (2-bytes) of the datum stored in the record. Identification is the FILE-IDENT obtained from the data base. 2) Status of the datun (2-bytes), the bits of the status word are used as follows: = Bit 0 (Least significant bit): 0 = good, 1 = bad. it 1: 0 = measured value, 1 = substituted valve. 0 = not High alarm, 1 = High alarm. not Low alam, ‘1 = tow alarm. not campressed, 1 = compressed - Bits 5-1: 3) The datum (4-bytes), a saved datun can be either an analog input or the 23-bits of a digital input group. 2.26Section 3 CALCULATION OF ON-LINE PERFORMANCE INDICES The generation of electricity involves the use of large complicated machinery which is required to operate under very denanding conditions. High temperatures and pressures are called for to meet the desixed power output requirements at maximum efficiency. the result is machinery that mist operate for long periods of ‘time under high mechanical stresses and constant chemical attack. The performance of this machinery under these operating conditions must be continually monitored in order to insure that the equipment is operating satisfactorily. Performance monitoring then serves two main purposes. First it is an indication of the efficiency with which the electricity is being generated. By having on-line performance information available, operators and dispatch personnel can more accurately gauge the unit or system's performance and fine tune the genera— tion line up where practical. ‘The second benefit of performance monitoring is to provide maintenance personnel with a “scorecard” describing the physical conditions of the equipment in operation. This is useful in optimizing scheduled outages. By having information about each major component's condition, schedules can be drawn up and replacetent parts ordered based on actual component requirements. 31While on-line performance monitoring offers fine-tuning capability, storing this on-line data into history files is necessary for optimization of maintenance concerns, It is necessary to look at long term performance data and trend characteristics in order to get a true picture of each camponents condition. A rapid loss in performance which remains consistently low, for example, is an indication ‘of component failure, while a slow steady decrease in performance over a ong pericd of time would indicate normal wear and tear. One single reading would not be relevant in either of these cases. ‘The goal in each of the performance calculations, to be discussed herein, is to provide on-line performance indices for the major components of PEPCO's Morgantown Unit #2. A quick glance at these indices will provide operators and engineers with concise information regarding the component's performance or change in performance, Detailed information describing the data used to calculate the performance index is then available upon request. Indices and all pertinent data is autonatically stored on tape for later retrieval by maintenance personnel or other interested parties. This section describes a series of performance calculations which hive been developed to date. In addition, new calculations are being developed ani will be implenented. For example, Section 4 describes a condenser analysis, Section 5 desoribes a feaiwater heater analysis, etc. Each of these sections describe a new performance calculation vhich is in development. ‘These calculations will be incorporated into PC upon completion of the development stage or as part of the development if deaned necessary. 32‘the overall scope of the entire set of calculations is to provide an accurate on-line heat rate, an incremental heat rate for system dispatch, a vehicle by which component physical conditions can be quickly monitored and to provide nore operator control where applicable, All of these concerns require accurate and reliable transfer of a large quantity of measured as well as calculated data. the following is a description of what is presently available at PEPCO's Morgantown unit #2. CALCULATION OF PERFORMANCE INDICES Actual Heat Rate can be calculated from knowledge of the unit's component efficiencies. The OnLine Performance Calculations are a series of programs intended to examine the performance of each of the major components in the stean plant. A performance index indicating that component's operating efficiency, is calculated for each major component, ‘hese indices describe how each component's performance contributes to the unit Heat Rate. The resulting actual Heat Rate is fa function of all these indices. Performance indices need to be calculated continuously and then stored for future retrieval. these calculations, however, require the collection of on-line data over a period of time. That is, data neads to be collected over a period of tine and statistically combined to get averages, standard deviations and the like. In response to this requirenent, a program was written to automatically collect data semples, perforn a statistical analysis on this sample, use the statistical re sults to calculate performance indices and then properly store the performance in- ices along with the pertinent data used to calculate each index. 33RC is the progran which calculates the real-time, periodic performance calculations. Data collected externally are accumulated, averaged and stored for subsequent performance calculation execution by PC. The performance calculations are executed every ten minutes using the data accumulated during each ten-minute interval. Selected results of the performance calculations are sent for inclusion into History Files. Table 3-1 is a list of calculated parameters which are stored in History Files to date. Table 3-1 DRTA CALCULATED BY FC AND STORED IN HISTORY FILES DATA POINT DESCRIPTION unrTs I. Shaft Leakage Calculations (SL) 642 Seal wear factor - ‘T601 1st valve sten leakoff BAR ‘T602 2nd valve stem leakoff 1B/ER 7603 3rd valve stem leakoff IB/HR ‘T604 ‘N2 packing flow IB/R ‘T60S ‘Nl packing HP leakoff 1B/HR 17606 NI packing LP leakoft BAR ‘T607 ‘N3 packing leakoff 1B/HR T608 Leakoff to exhauster B/R ‘7609 Leakoff to exhauster 1B/ER 341. Heat Balance Calculation (HB) 7614 Gross turbine heat rate BIU/KWH 7615 HP turbine efficiency % 7616 IP turbine efficiency a 7618 LP turbine efficiency 8 7617 IP-LP combined efficiency 8 S616 Main steam flow sR 3627 Hot reheat steam flow B/E 3646 Exhaust stean flow B/AR weld Final feedwater flow 1B/HR wend Feed pump AsB suction flow BAR WEL Htr 7 cond inlet flow B/E S647 Her 1 ASB extr flow 1B/ER $632 Htr 2 extr flow 1B/ER 3634 Htr 3 extr flow B/E 8635 Deaer extr flow B/HR 8636 Her 5 extr flow B/ER 8637 Htr 6 extr flow 1B/HR 8638 Her 7 extr flow 1B/ER woL Htr 1 ASB drain flow B/E D620 Htr 2 drain flow 1B/HR D610 Her 3 drain flow B/HR pels Her 5 drain flow 1B/HR wos Htr 6 drain flow 1B/HR wo7 Her 7 drain flow 1B/HR MaLoss Generator loss = MASHAET Turbine shaft power a UBEP Used energy end point BrU/B VAN LP annulus velocity FT/SEC [EXHLOSS Corrected exhaust loss BIUAB ‘LEP, Expansion line end point BIU/B IIL, B€ficiency - Boiler Feed Pmps and ‘Turbines (FP) 7636 BEP-A turb efficiency 8 7637 BEP-B turb efficiency 2 W667 BEP-A pump efficiency s W668 BEP-B pump efficiency 8 W665 BEP-A pump bearing loss BIU/ER W665 BEP-B punp bearing loss BrU/HR 7619 BEP-A turbine mech loss BIU/ER 7620 BEP-B turbine mech loss BIU/HR BEPAFR BEP-A fraction of flow -W680 wes, QBLR BP W669 W690 weal, ‘conrTa UHEIA UHEB 604 A605 come ‘TCOBC W270 wm w272 w274 W275 W276 W277 W278 W279 p219 220 D221. D223 D224 IV. Condenser Calculations (CC) Cond-A cleanliness fact Cond-B cleanliness fact Heat added by boiler Heat in make up water Heat added by booster pump Heat rejected Cond-A heat rejected Cond-B heat rejected Cond-a taro Cond-B IMD Cond-A heat trans coef Cond-B heat trans coef Cooling water velocity Cond-A GW velocity Cond-B GY velocity Cond-A HEI temp correct Cond-B HEI temp correct Cond-A clean tube coef Coni-B clean tube coef V. Air Heater Effectiveness (AH) Heater A effectiveness Htr A cold end avg tenp Htr A corr gas out temp Heater B effectiveness Htr B cold end avg temp Her B corr gas out temp VI. Feedwater Heaters: ITD & DCA (Wii) ‘Tenminal temp diff 1A ‘Tenminal temp diff 1B Temninal temp diff 2 Tenninal temp diff 3 Teminal temp diff 5 ‘Temminal temp diff 6a ‘Teminal temp diff 68 ‘Temninal temp diff 7A ‘Terminal temp diff 7B Drain cooler appr 1A Drain cooler appr 1B Drain cooler apor 2 Drain cooler appr 3 Drain cooler appr 5 36 MBTU/HR MBTU/HR MBTU/HR MBTU/HR MBTU/HR MBTU/HR F E BIU/HR-E-SqET BIUAR--SqFT FI/sec Pr/sac PT/SBC BIU/IR-F-SgFT BIU/HR-P-SqET r P my dy ama eyVII. Correction for Off-Design Makeup Flow au) DERM ‘Total heat rate change BTU /n wes4 Actual makeup flow w/ER piaBTy HR adj - RH duty chy BIU/WH DiBaTU HR adj - Htr 1 extr chg BTU/KWH D2aT0 HR adj - Htr 2 extr chy BIU/AWH D3BTU HR adj - Htr 3 extr chg ‘BIU/KWH DaBTO HR adj - Htr 4 extr chg BIU/AWH DSBTU HR adj - Htr 5 extr chg BTU /RWH DesTu HR adj - Htr 6 extr chg BIU/RHH D7BT0 HR adj - Htr 7 extr chg ‘BIU/KWH DELBTU HR adj - Exh loss chg BIU/WH VIII. Correction for Off-Design Heater Perf. W RETR HR adj - all heaters BTU AH, BIURL HR adj - Htr 1 perf BIU/WH BTUER2 HR adj - Htr 2 perf BIU/KWH BIUR3 HR adj - Htr 3 perf BTU/RHH BTUHRS HR adj - Htr 5 perf BIU/KWH BTUERS HR adj - Htr 6 pert BIU/AHH BIUER7 HR adj - Htr 7 perf ‘BTU/AWH IX. Corrections for Off-Design Turbine BEE. (rs) DHRHP HR adj - HP turbine eff BTU /KHH DARIP HR adj - IP turbine eff BIU/AWH DARE HR adj - LP turbine eff BIO X. Correction for Off-Design Steam and Exhaust () rare Achievable GTHR BTU /AH BTUC Controllable losses BIU/RWH DERMST HR adj - main stm temp BIU/RWH DERRAT HR adj - HRH stm temp BTU AH DARMSP HR adj - main stm pres BIU/EHH DHRH2 HR adj - gen H2 pres BIU/KWH DHRRS HR adj - reheat spray BIU/KWH DHRRHP HR ad} - RH pres drop ‘BTU/AWH X, Correction for Off-Design Steam and Exhaust (ce) DHREXH HR adj - exhaust pres BTU/KHH DHREE HR adj - power factor BTU/KWH 34XI. Steady State Determination (0C) SSCOND Steady state condition O=¥ES/1=NO Main Program Description (PC) ‘The main program gathers data and organizes it into the appropriate fom for use by the performance calculation subroutines. The main program decides shich subroutines to call based on the data it receives and the source of that data. It is also responsible for maintaining the contents of the Accumulated Menory. The main program performs a few minor calculations required to initialize the performance calculation sequence. Upon completion of the performance calculations, the main program sends selected calculated data points to be stored in a History File. ‘The main program (PC) currently calls thirteen separate subroutines to execute all of the performance calculations. Trey are describai here in the order of execution. FCACM - This subroutine collects data from the Shared Memory to fill the Accumulated Memory during each ten-minute interval. This subroutine is called by supplying @ group number (designating the data type and origination) ani the time in which this data was put in Shared Memory. It is called from the main program (PC), PCACIM will then search through Shared Memory for any data points pertain. ing to the group number designated and check that each data point update time is the same as that provided to FCACIM. The following operations are then per formed by PCACIM on each data point for the group specified: a) Update a running average of the value of each data point and store this average value in Accumulated Memory. b) Update the minimum and maximm value for each data point and store the maximm and minimum values in Accumulated Memory. 38c) Keep track of the latest time which each data point was updated and store this time in Accumulated Memory. a) Evaluate the status of each data point and store in Accumulated Memory an "accumulated" status and identify points which are known to be bad. ‘This subroutine is called each time that new data is inserted into Shared Memory. COC - This subroutine is called once every ten minutes by PC and evaluates key data points stored in Accumulated Menory to determine if the plant was running at. steady state over the previous ten-minute interval. The following list comprises ‘the PCOC input: POINT ENGINEERING NAME DESCRIPTION UNITS M407 BLR EXTRACT VALVE ‘L=OPEN/O=CLSD 2239 GN 2 mw ra $200 MAIN STEAM NORTH P 8201 MAIN STEAM NORTH E 8202 MAIN STEAM SOUTH E 8203, MAIN STEAM SOUTH F 5208 HOT REHEAT NORTH F 8209 HOT REHEAT NORTH E 5210 HOT REHEAT NORTH E S211 HOT REHEAT NORTH rg S281 BEPT-A HP STEAM RiBs/H 8282 BEPT-8 HP STEAM ELBS/H ‘$408 CREP SIM TO AIRETR 1-0PEN/O=CLSD 409 BOX STEAM TO REBLR 1sOPEN/O=CLSD ‘8410 AUX STEAM 70 DEAER L-OPEN/O=CLSD s4ll SEP STEAM TO DEAER 1-OPEN/O=CLSD S412 ROX STEAM TO BEPT L-OPEN/O=CLSD 3413 PERFORMANCE CALC L-OFF /0=RON W436 BEP-B RECIRC VALVE 1 -OPEN/O=CLSD W437 BEP-A RECIRC VALVE 1=OPBN/O=CLSD PCOC examines each of these data points and compares their maximm and minimum value over the ten-minute period. Tf the difference between maximum and minimum value is within prescribed limits, PCOC sets the data point SSCOND = 0 indicating that steady state was achieved during this sample period. the following data is output by FCOC and sent to Shared Memory ané the History File for storage. 398DATA POINT DESCRIPTION wits ‘ssconD Steady state condition 0-¥BS/1=NO PCSL - This subroutine calculates the steam leakage flow along the HP, IP and LP turbine shafts. The following data poiats are input to CSL. POINT ENGINEERING NAME, DESCRIPTION onrTs: s179 BAROMETER uc $200 MAIN STEAM NORTH Fr S201 MAIN STEAM NORTH FE $202 MAIN STEAM SOUTH FE 8203 MAIN STEAM SOUTH g $204 COLD REHEAT NORTH FE 8205 COLD REHEAT NORTH RP S206 COLD REHEAT NORTH E s207 COLD REHEAT NORTH FP s2i2 XOVER 10 LP TURB FE $213 XOVER TO LP TURB E s2i4 XOVER TO LP TURB F S215 XOVER TO LP TUB E 5250 MAIN STEAM NAREOW PSIG s251 MAIN STEAM WIDE PSIG 8253 FIRST STAGE SHELL PSIG 8254 COLD REHEAT NORTH PSIG 8255 COLD REHEAT SOUTH PSIG 8257 XOVER TO LP TRB PSIG 8258 XOVER TO LP TURB BSIG S265 EXTR TO IP ATR 2 PSIG 8267 EXTR TO IP HTR 3 PSIG 7249 TORB SEAL SIM MEA KLBS/H The input data is run through a series of flow calculations, Bach shaft leak flow has a specific loss coefficient which relates the leakage flow to the pressure @rop across the leak. A seal wear condition factor is calculated and applied to each loss coefficient. This seal wear condition factor is re-calculated each time this subroutine is called. 3210The following data is output by PCSL and sent to Shared Metory and to the History File for storage. It is also stored in Accumulated Memory so that other perfor— mance calculation subroutines may access this data. DATA POINT DESCRIPTION UNITS 7642 Seal wear factor - ‘7601 Ast valve stan leakoff AR 7602 2nd valve stam leakoff w/HR 7603 3rd valve stan leakoff w/HR T6048 N2 packing flow 1B/HR 7605 UL packing HP leakoft 1B/ER T606 NL packing LP leakoff 1B/HR 7607 N3 packing leakoff 1B/HR r608 Leakoff to exhauster w/AR ‘7609 Leakoff to erhauster BAR FCHB - This subroutine performs a steam roo heat balance to calculate the gross turbine heat rate, FCHB also determines the HP, LP, IP and a combined IP-LP tur— bine efficiencies. The following data points are input to FCB: POINT ENGINEERING NAME DESCRIPTION oNITS 200 HEATER-1A DRAIN E 201 HEATERS-18 DRAIN F D202 HEATER 2 DRAIN E 204 HEATER 3 DRAIN FE D205 HEATER 5 DRAIN F 206 HEATER-6A DRAIN Fr 207 HEATER-6B DRAIN F 208 HIR DRN POMP DISC EF D210 REBOILER DRAIN Eg pail. BEP-A HP LEAK OFF E p2iz BEP-B HP LEAK OFF E D213 STM TO AIR HEATER KIBS/H paid SIM AIR HIR DRN B P pais KUBS/H pais KLBS/H pai? KOBS/H pais E 225 E 226 NO6 FO REBLR DRIP KLBS/H 8238 GEN 2 MVAR MARS E239 GEN 2 Mw ro 216 GENERATOR HYDROGE PSIG 1200 HOIWELL 2A LEVEL, Er 1201 HOIWELL 28 LEVEL ET1.202 173 $200 $201 $202 $203 $204 $205 $206 $207 $208 $209 S210 s2i1 s2i2 s213 s214 $215 S216 $217 S218 S219 $220 5221 $222 $224 8225 8226 8228 $229 $232 $250 S251 $254 8255 8256 8257 5258 8259 8260 5261 8262 5263 5264 8265 8267 8268 8269 8270 S27. 272 8273 274 8275 S276 $280 8281 5282 8283 gag S30Hg2828 age ee qaey Af55553ev0y g B NEB eM ES ERE SEAL ST™ 3 a EAE et ST Te I IGE II Te Ie Je KJ] sn a sega Z Bas 838 PSIG PSIG PSIG PSIG BSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG PSIG KIBS/H IBS/Hs284 BEPI-B LP STEAM IBS/H 5289 B EXT 10 DEA HTR PSIG 5292 SEAL SIM TO HTR 7 KLBS/Ht ‘T601 First valve sten leaxoff B/HR ‘7602 Secon valve sten leakoff BAR 17603 Third valve stem leaxoff 13/HR ‘r604 N2 packing flow IB/R 7605 NL packing HP leakoft 1B/HR 7606 NI packing LP leakoft 1B/ER 7607 N3 packing leakoff 1B/HR w2i2 COND HOTWELL-28 F w2i3 (COND HOIWELL-28 P w2i6 COND HEATERS 7 IN Ee W217 COND LVG HIR 7a E was COND LVG HTR 7B E w219 COND LVG HIR 6a EF W220 COND LVG HIR 6B E W221 COND ENT HEATER 5 FP W222 COND LVG HEATER 5 E W223 BOOSTER A SUCTION Fr W224 BOOSTER B SUCTION E W225, BOOSTER C SUCTION F W226 (COND ENT HEATER 3 E w228 COND ENT HEATER 2 E W229 COND LVG HTR2 BEP e W230 EW ENT HEATER 1A E W231 Fa BN? HEATER 18 E W232, EW ING HEATER 1A E W233, EW LVG HEATER 18 FE W237 REHEAT SPRAY WATE Fr W238 ‘COND FROM HEATER PSIG W239 EW LVG HIR 2BFPS PSIG W240 RHT SPRAY WTR HDR PSIG W243 BW ENT ECONOMIZER PSIG w24e MAKE UP KLBS/H W246 ‘CONDENSATE KLBS/H W249 REHEAT SPRAY IBS/H W263 BEP INJECTION WIR KIBS/H WEPSL BEP SHAFT LEAKAGE TO ATMOS 1B /ER ‘his subroutine uses the above data to perform an energy and mass balance on each feedwater heater. This results in a series of simultaneous linear equations in Which the extraction flow to each of the feedwater heaters is unknown. The extraction flows are determined by solving this sytem of equations using a simple Gaussian elimination technique. This method was found to be more efficient than the iterative technique previously used by PEPCO. 313once the feedwater heater extraction flows are knowm, the hot reheat flow and LP turbine exhaust flow are calculated. ‘his leads to the calculation of Used Energy End Point, Next the subroutine evaluztes the LP turbine exhaust losses to deter mine the Expansion Line md Point. Having now detemined the enthalpies at every state point, the gross turbine themal output can then be calculated and compared to the gross turbine electrical output to detemmine the gross turbine het rate. Finally, all the turbine enthalpies are combined to determine the HP, IP and LP efficiencies. In addition, a combined IP-LP efficiency is calculated. ‘The following data is output by PCHB and sent to Shared Menory and to the History File for storage. these output data points are also stored in Accumulated Memory so that other perfomance calculation subroutines may access them. DATA POINT DESCRIPTION units rela Gross turbine heat rate Brunt P61s Hp turbine efficiency 8 7616 IP turbine efficiency 3 eis Lp turbine efficiency % 7617 IB-LP combined efficiency & S616 Main stean flow w/AR 3627 Hot reheat stean flow B/iR $646 Exhaust steam flow 1B/HR wel2 Final feedwater flow B/aR weld Feed pump AEB suction flow 1B/HR W616 Htr 7 cond inlet flow 1B/HR 3647 Htr 1 ASB extr flow 1 /iR 3632 Her 2 extr flow BAR 8634 Htr 3 extr flow w/aR 8635 Deaer extr flow 1B/HR $636 Htr 5 extr flow 1B /HR 8637 Her 6 extr flow w/a 8638 Htr 7 extr flow BAR wl Her 1 ASB drain flow 1B /HR 620 Htr 2 drain flow w/aR D610 Her 3 drain flow B/R D615 Htr 5 drain flow te /HR wos Her 6 drain flow B/AR wo7 Btr 7 drain flow BAR sntoss Generator loss mw MWSHAPT Turbine shaft power Mi UBEP Used energy end point BIUAB vAN LP annalus velocity FI/SBC EXHLOSS Corrected exhaust loss BIU/B ELEP Expansion Line end point BUA 3414RCEP - This subroutine calculates the boiler feed pump efficiency and the boiler feed pump turbine efficiency. ‘Te following data points are input to PCFP: POINT ENGINEERING NAME DESCRIPTION ourts pai BEP-A HP LEAK OFF EF pai2 BEP-B HP LEAK OFF e D216 BEP-A DA LEAK OFF KLBS/H D217 BEP-B DA LEAK OFF KLBS/H 8179 BAROMETER "HS s216 SIM TO BEPT 2AHP E S217 Sm 10 BEPT 23 HP E s2i8 SM TO BFPT 24 LP E S219 SIM 10 BFP? 28 LP FE $259 HP STEAM BEPT-2A PSIG 8260 HP STEAM BEPT-2B PSIG S261 EP STEAM BPPT-28 PSIG 8262 LP STEAM BFPT-28 PSIG 8268 SM TO DEA HTR 4 BSI 8274 ‘Con 2a VACUO HG 8275 COND 28 VACUUM HG S281 BEPT-A HP STEAM KLBS/H 8282 BEPT-B HP STEN IBS/H 8283 BEPT-A LP STEAM KIBS/H $284 BEPT-B LP STEAM KLBS/H r212 BEPT-A SPEED RBM 7213" —-BEPT-B SPEED REM W229 COND LVG HTR2 BFP EF W235, BEP-A DIFF TEMP M w W236 ‘BEP-B DIFF TEMP M wv W237 REHEAT SPRAY WATE FE W239 EW LVG HTR 2BFPS PSIG w240 RH SPRAY WIR HDR PSIG W241 BEP-A DISCHARGE PSIG W242 BEP-B DISCHARGE PSIG W247 BEP-A SUCTION KIBS/H W248 BEP-B SUCTION RLBS/H W249 REHEAT SPRAY KLBS/H wel2 Final feedwater flow BAR welé Feed pump AgB suction flow 1B/HR ‘This subroutine takes the above data and calculates the estimated pump bearing losses and the turbine mechanical losses. It then calculates the turbine used en— ergy and efficiency. Finally, it calculates the pump suction and discharge flow velocities which are used to calculate the pump internal efficiency. B15The following data is output by PCEP and sent to Shared Menory and to the History File for storage. DATA POINT DESCRIPTION UNITS 7636 BEP-A turb efficiency 8 7637 BED-B turb efficiency s we67 BEP-A pump efficiency 8 688 BEP-B pump efficiency 1 W665 BEP-A pump bearing loss BIU/HR wees BFP-B pump bearing loss BIU/R T619 BEP-A turbine mech loss BTU/HR 17620 BEP-B turbine mech loss BTU/HR BEPAFR BEP-A fraction of flow = POCC - This subroutine performs a heat exchanger analysis on the condenser to determine a condenser cleanliness factor. ‘The following data points are input to occ: POINT ENGINEERING NAME DESCRIPTION wITS MSHA ‘Turbine shaft power vi 8179 BAROMETER "ac 8200 MAIN STEAM NORTH e S201 MAIN STEAM NORTH FE 8202 MAIN STEAM SOUTH FE $203 MAIN STEAM SOUTH FE 8204 COLD REHEAT NORTH E $205 COLD REHEAT NORTH E 8206 COLD REHEAT NORTH E $207 COLD REHEAT NORTH g 3208 HOT REHEAT NORTH E 5209 HOT REHEAT NORTH ze 8210 HOT REHEAT NORTH r S211 HO? REHEAT NORTH E 18250 MAIN STEAM NARROW PSIG 8251 MAIN STEAM WIDE PSIG 8254 COLD REHEAT NORTH PSIG 8255 COLD REHEAT SOUTH PSIG 8256 HOT REHEAT STEAN PSIG 8268 SIM TO DEA HTR 4 PSIG S274 COND 2A VACUUM. HG $275 ‘COND 28 VACUUM 8G S616 Main steam flow 1B/HR 13627 Hot reheat steam flow B/R ¥200 COND-2A WATER IN FE W201 COND-2A WATER IN E W202 COND-28 WATER IN FE 3-16w203 (CoND-28 WATER IN e W204 COND-2A WATER OUT E W205, COND-2A WATER OUT E W206 COND-2A WATER OUT E w207 COND-2A WATER OUT F W208, COND-28 WATER OUT r w209 COND-28 WATER OUT F W210 ‘COND-28 WATER OUT E wall, CoND-28 WATER OUT E W223 BOOSTER A SUCTION E W224 BOOSTER B SUCTION FE W225 BOOSTER C SUCTION F W226 (COND ENT HEATER 3 Pr W234 EW EN? BCONOMIZER r W237 REHEAT SPRAY WATE F W239 EW LNG HTR 2BFPS PSIG W240 HT SPRAY WIR HDR PSIG W243, EW ENT BCONOMIZER PSIG W244 MAKE UP KLBS/H W245 DRAW OFF KLBS/H w249 REHEAT SPRAY KLBS/H W259 (CIRC WATER COND-A GPM W260 CIRC WATER COND-B KG W612 Final feedwater flow 1B/HR weld Feed pump AKB suction flow IB/AR ‘This subroutine uses the above data to estinate the overall heat transfer cpeffi- cient across the condenser. The IMTD method is used to estimate an overall heat transfer coefficient. In addition, correction factors are calculated and used to determine an HBI corrected heat transfer coefficient, The condenser cleanliness factor is defined as the heat transfer coefficient divided by the HEI corrected heat transfer coefficient. ‘he following data is output by PCOC and sent to Shared Memory and to the History File for storage: DATA POINT DESCRIPTION outs we80 Cond-A cleanliness fact - W681 Cond-B clean! iness fact - QBLR Heat added by boiler seT/ER ono Heat in make up water MBTU/AR Be Heat added by booster pump — MBTU/HR W569 Beat rejected wIUAR QREIA Cond-A heat rejected yBTu/HR ‘QREJB Cond-B heat rejected MBTUAR W690 Cond-A ETD F 347wool Cond-8 ‘ID F AA Gond-A heat trans coef BIU/HR-E-SgFT AB Cond-B heat trans coef BIU/HR-P-SqrT vAVG Cooling water velocity FI/SEC vA Gond-A GH velocity FI/SBC vB Cond-B GH velocity FT/SEC CORRTA Goné-A HET temp correct - CORRS Cond-B HET tenp correct - DHEIA Gond-A clean tube coef BIU/AR-F-SgFT OHEIS Coné-B clean tube coef BIU/AR-F-SqPT CAH - This subroutine performs a heat regenerator analysis on the Ljungstran Air Preheaters to determine the heater effectiveness. ‘The following data points are input to PCAH: POINT ENGINEERING NAME DESCRIPTION UNITS: A202 LIUNGS 2A AIR IW F 8203 TJUNGS 2A AIR IN Er A204 LUNGS 2A AIR IW F 18205 TSUNGS 28 AIR IN E 12208 LUNGS 28 AIR IN F 8207 LIUNGS 28 AIR IN F 18208 TJUNGS 23 AIR OOT F A209 LIUNGS 2A AIR OUT E A210 TIUNGS 2A AIR OUP EF a2il LUNGS 28 AIR OUT F a212 LUNGS 28 AIR OUT E A213 LUNGS 28 AIR OUT E C600 Ljungstron 2A leakage constant Fraction cool Tjungstron 28 leakage constant Fraction 6200 LUNGS 2A AS IN r 201 LIUNGS ZA @S TH g 6202 EJUNGS 2A @S IN g 6203 LING 2 GAS IN g 6208 LJUNGS 28 GS IN g 6205 LUNGS 23 GAS IN g 6206 LUNGS 23 GAS OUT Fg 6207 LIUNGS 24 AS OUT r 208 LJUNGS 24 AS OUT P 6209 TJUNGS 28 AS OUT g 210 LJUNGS 28 GAS OUT Eg Gal. TJUNGS 28 GAS OUT F aus‘his subroutine reads the above data and calculates the heater effectiveness based on an actual to theoretical temperature difference. account for air to gas leakage. Shared Memory and to the History File fot storage: DATA POINT DESCRIPTION onrTs A604 Heater A effectiveness - ‘Compa Htr A cold end avg temp F TGOAC Her A corr gas out temp F ‘A605, Heater B effectiveness - ‘coLDB Htr B cold end avg temp EF ‘7rG0BC Htr B corr gas out temp F PCWH - This subroutine calculates the Terminal Temperature Difference (rm) ‘Temperatures are corrected to Te following data is output by PCAH and sent to and the Drain Cooler Approach Temperature (DCA) for each closed-type feedwater heater. ‘The following data points are input to PCW: POINT ENGINEERING NAME DESCRIPTION osiTs D200 HEATER-1A DRAIN E D201 HEATERS~1B DRAIN e D202 HEATER 2 DRAIN E D204 HEATER 3 DRAIN gE D205, HEATER 5 DRAIN F s179 BAROMETER HG 8263 EATR TO HP HTR 1A PSIG 5264 EXTR TO HP HTR 1B SIG 8265 EXTR TO IP HTR 2 PSIG 8267 EXTR 10 IP HTR 3 PSIG 8269 B EXTR 10 LP HTR PSIG 8270 EATR TO LP HTR 6A. PSIG S271 EXTR 10 LP HTR 6B PSIG 8272 EXTR TO LP HIR 7A PSIG 8273 EXTR TO LP HTR 7B PSIG 8280 AEXTR 10 LP BTR 5 PSIG W217 COND LVG HTR 7A rg w2i8 COND LVG HTR 7B F wai9 COND LVG HTR 6A E W220 COND LVG HIR 6B Ee w221 COND END HEATER 5 F W222 COND LVG HEATER 5 E w226 COND ENT HEATER 3 E W228 COND ENT HEATER 2 FE W229 COND LVG HTR2 BEP EF W230 EW ENT HEATER 1A Fw231 EW ENT HEATER 1B E w232 EW LVG HEATER 1A F W233 EW LVG HEATER 1B F POW uses the above data to determine the saturated steam to feedwater terminal temperature difference (TTD) and the condensate to feedwater drain cooler approach (Oca). The following data is output by POWH and sent to Shared Menory and to the History File for storage: DATA POINT DESCRIPTION onrTs W270 ‘Teminal temp diff 1A FE wan ‘Teminal temp diff 1B FE w272 ‘Temninal tenp diff 2 FE w274 ‘Terminal temp diff 3 Fr W275 ‘Terminal temp diff 5 E W276 Terminal temp diff 6a FE W277 ‘Terminal tenp diff 6B F W278 ‘Teminal temp diff 7A E w279 ‘Teminal tenp diff 7B E p2i9 Drain cooler appr 1A E 220 Drain cooler appr 1B E p22i Drain cooler appr 2 F D223 Drain cooler appr 3 F 224 Drain cooler appr 5 F RCM - This subroutine calculates the actual plant make-up water requirenents and determines a correction to be applied to the gross turbine heat rate The input ata points for this subroutine are as follows: POINT BNGINEERING NAME DESCRIPTION wits 200 HBATER-1A DRAIY F 201 HBATERS-18 DRAIN F 202 HEATER 2 DRAIN F 204 HEATER 3 DRAIN E 205 HEATER 5 DRAIN F 206 HEATER-6A DRAIN FE 207 HEATER-6B DRAIN F 208 HR DRN PUMP DISC EF 239 GN 2 MH Ma 1200 HOTWELL 24 LEVEL ET 1201 HOIWELL 28 LEVEL Fr 1202 DEAER STORAGE TAN FT 8179 ‘BAROMETER "ac 3-20REHEAT SPRAY WATE COND FROM HEATER EW LVG BTR 2BFPS EW ENT ECONOMIZER KE UP DRAW OFF Expansion Line end point Htr #6 extraction enthalpy Her #7 extraction enthalpy LP turb exhaust flow Gross turbine heat rate Final feedwater flow 3-21 am ma sm aThis subroutine checks for any feedwater heater which may be out of service and performs a mass balance on the operating feedwater heaters to calculate a required condensate make up flow. The difference between the actual make up flow and the design make up flow is used to determine a heat rate correction factor. “this correction factor can be applied to the gross turbine heat rate calculated by FCHB to correct for off-design make up flow. ‘The following data is output by RMU and sent to Shared Manory and to the History File for permanent storage: DATA POINT DESCRIPTION onITs DERM Total heat rate change BTU/KWH Wess ‘Actual makeup flow 1B/ER DIABTO HR adj - BH duty chy BTU/KWH DIBBTU HR adj - Htr 1 extr chg BTU D2eTU HR adj - Htr 2 extr chg BTU/RWH DaBT0 HR adj - Htr 3 extr chy BTU/RWE DaBTU HR adj - Htr 4 extr chy BTU/RWH DSBTU HR adj - Htr 5 extr chg BTU/KWH DeBTU HR adj - Htr 6 extr chg ‘BTU/KWH. D7sTU HR adj - Htr 7 extr chg ‘BTU/KWH DELBTU HR ad} - Exh loss chg ‘BTU/KHH BCEW - This subroutine evaluates the feedwater heater performance and calclates a correction to be applied to the gross turbine heat rate to account for off-design heater performance. ‘The following data points are input to PCEW: POINT ENGINEERING NAME DESCRIPTION onrTs D200 HEATER-1A DRATN Fr 201 HEATERS-1B DRAIN r 202 HEATER 2 DRAIN F 204 HEATER 3 DRAIN Er 205 HEATER 5 DRAIN E D206 HEATER-6A DRAIN F 207 HEATER-6B DRAIN F 208 HTR DRN PUMP DISC F D610 Heater 3 drain flow 1B/HR Dé15 Heater 5 drain flow 1B/HR 620 Heater 2 drain flow 1B/HR £239 EN 2 Ma HE6 Htr #6 extraction enthalpy BIUAB HE? Htr #7 extraction enthalpy BIUAB 8179 ‘BAROMETER "ug 3-22s222 ATR TO IP HTR 2 Eg 8224 EXTR 10 IP HIR 3 E $225 SIM 10 DEA HTR 4 F 8226 B EXTR TO LP HIR E 8263 EXTR TO HP HTR 1A PSIG S264 EXTR 10 HP HTR 1B PSIG 8265 EXTR TO IP HTR 2 PSIG 8267 EXTR 10 IP HTR 3 PSIG 8268 STM TO DEA HTR 4 BSIG 8269 B EXTR TO LP BTR PSIG 8270 EXTR TO LP HTR 6A PSIG s27L EXTR TO LP BIR 6B PSIG 8272 EXTR 10 LP HTR 7A PSIG 273 EXTR TO LP HIR 7B PSIG s292 SEAL SIM TO HTR 7 REBS/Ht 8632 Heater 2 extraction flow 1w/HR 3634 Heater 3 extraction flow w/R 3636 Heater 5 extraction flow B/HR 8637 Heater 6 extraction flow w/ER 3638 Heater 7 extraction flow B/R 8647 Heater 1 A&B extraction flow 1B/HR 614 Gross turbine heat rate BTU /KWH w2ié ‘COND HEATERS 7 IN F w2i7 COND LVG HTR 7a EF was ‘COND LVG HR 7B E W219 ‘COND LVG HTR 6A F 220 COND LVG HTR 6B F w22) COND ENT HEATER 5 Fg W222 COND LVG HEATER 5 EF W226 COND ENT HEATER 3 E W228 COND ENT HEATER 2 E W229 (COND LVG HTR2 BEP e W230 EW ENT HEATER 1A F W231 EW ENT HEATER 1B E W232 EW LVG HEATER 1A F W233 ‘EW, LVG HEATER 1B E W238, ‘COND FROM HEATER PSIG W239, EW LNG HTR 2BFPS SIG W246 ‘CONDENSATE, KLBS/# PCEW checks first for any feedwater heaters that may be out of service, For those heaters in service, POFW compares actual inlet and exit enthalpies to design values to determine a heat rate correction for each heater. The corrections are cambined for application to the gross turbine heat rate. the following data is output by PCEW and sent to Shared Menory and to the History File for storage. 3-23DATA FOINT DESCRIPTION onrTs DHRHTR HR adj - all heaters BrU/WH ‘BTUARL HR ad} ~ Htr 1 pert BTU /AHH BTUHR2 HR ad} - Htr 2 perf BIU/RWH BTUHR3 HR ad} - Htr 3 perf BIU/AwH BIUHR5 HR ad} - Htr 5 perf BTU AWE BTUHRS HR ad} - Htr 6 perf BTU/H BTUAR7 HR adj ~ Htr 7 perf BTU /AWH RCTS ~ This subroutine determines the correction to be applied to the gross turbine heat rate to account for actual turbine efficiencies which may be lover than ‘the design values. The following data points are input to ECTS: porn ENGINEERING NAME DESCRIPTION WITS 2239 cen 26 oi PRAOS Calculated Valve Position si79 BAROMETER "ae 5200 MAIN STEAM NORTH F S201 MAIN STEAM NORTH E 5202 MAIN STERN SO7TH ze 5203 MAIN STEM SOUTH F 8208 COLD REXEAT NORTH E $205 COLD REHEAT NORTH F $206 (Coed REHEAT NORTH F 8207 COLD REHEAT NORTH F $208 HOT REHEAT NORTH oF 5209 HOT REHEAT NORTH P s210 HOT REHEAT NORTH E sail Ho? REHEAT NORTH z s212 XOVER 10 LP TORS F 5213 HOVER TO LP TURB F sala XOVER TO LP TURB F S215 XOVER 10 LP TUB F $250 SAIN STEAM NARROW BIG 5251 MAIN STEM WIDE PSIG 8254 (COLD REHEAT NORTH PSIG 5255, (COLD REHEAT SOUTH PSIG 8256 HOT REHEAT STEAM PSIG 8257 XOVER TO LP TURB PSIG 5258 XOVER TO LP TURB PSIG 5278 Coup 2a vacuous HG 5275 cond 25 vACuM 6 S616 Main stean flow wR $627 Hot reheat steam flow 1B/HR 601 First valve sce leakotf a/R ‘7602 Second valve stem leakoff 1B/HR 7603 ‘Third valve stem leakoff 1B/HR ‘7604 N2 packing flow 1B/ER 7614 Gross turbine heat rate BTU 324els, Hp turbine efficiency s 616 IP turbine efficiency $ 617 i carbine (IP 6 LP) efficiency 618 ip turbine efficiency UEEP Used energy end point burs van up turb annulus velocity FT/SEC This subroutine determines the actual efficiency for each turbine based on the above measured data. ‘the resulting efficiency is compared to the actual efficiency as calculated by PHB. the difference in efficiencies is used to determine a correction to the gross turbine heat rate for actual turbine efficiencies not equal to their design values. The following data is output by POTS and sent to Shared Memory and to the History File for storage: DATA POINT DESCRIPTION wirs DHRHP HR adj - HP turbine eff BTU/KHH DHRIP HR adj - IP turbine eff BTU/KHH DHRLP HR adj - LP turbine eff ‘BIU/KWH POC - This subroutine detemines a correcticn to be applied to the gross turbine heat rate to account for stean and exhaust conditions which may not be identical to design values. The following data points are input to POCF: POINT ENGINEERING NAME DESCRIPTION ovis E238 GEN 2 MVAR VARS 5239 GEN 2 Mw ma G26 GENERATOR HYDROGE PSIG s179 BAROMETER mac s200 MAIN STEAM NORTH F S201 MAIN STEAM NORTH E 8202 MAIN STEAM SOUTH F 8203, MAIN STEAM SOUTH FE 8208 HOT REHEAT NORTH F 5209 HOT REHEAT NORTH E $210 HOT REHEAT NORTH EF S21. HOT REHEAT NORTH E 8250 MAIN STEAM NARROW PSIG 8251 MAIN STEAM WIDE PSIG 8254 COLD REHEAT NORTH PSIG 8255 COLD REHEAT SOUTH PSIG 5256 HOT REHEAT STEAM PSIG 3425274 (ConD 2a VACUUM HG 8275 COND 28 VACUUM HG S616 Main steam flow 1B/HR veld Gross turbine heat rate ‘BIU/KWH W249 REHEAT SPRAY LBS/H ‘The above data is used to make corrections for off-design steam temperatures and Pressures, system pressure Grop, generator hydrogen pressure, reheat spray, and Power factor. The corrections are combined into a single correction to be applied to the gross turbine heat rate calculaved by FCHB. ‘These corrections (calculated by POCE) are considered controllable by plant operators so FCCP calculates an achievable gross turbine heat rate which can be obtained if the above factors are adjusted, The following data points are output by PCCP and sent to the History File for permanent storage: DATA POINT DESCRIPTION UNITS omaRC Achievable GrHR BIU/KWH BIC Controllable losses BIU/AWH DARMST HR adj - main stm temp BTU/NH DHRRET HR adj - HRH stm temp BTU/NH DARMSP HR adj - main stm pres BIU/KRH DHRAZ HR adj - gen #2 pres ‘BIU/RWH DHRRHS HR adj - reheat spray BIU/KWH DHRRHP HR ad} - RH pres drop ‘BTU/KWH DEREXH HR adj ~ exhaust pres BIU/WH DHRPF HR adj ~ power factor ‘BTU/KWH RCSUM - This subroutine sumarizes all the corrections to be applied to the gross turbine heat rate. The input data for PCSIM is as follows: POINT ENGINEERING NAME, DESCRIPTION UNITS DAREXH Heat rate change due to exhaust pressure ® DHRH2 Heat rate change due to gen H2 pressure 8 DHREP Heat rate change due to HP turbine efficiency 8 DHRHTR Heat rate change due to perf of heaters 8 DHRIP Heat rate change due to IP 3-26‘turbine efficiency 2 DARL Heat rate change due to LP turbine efficiency 3 DHRMSP Heat rate change due to main steam pressure e DARMST Heat rate change due to main steam temperature + DARMU Heat rate change due to makeup flow 8 DERPE Heat rate change due to gen Power factor 8 DHRRHP Heat rate change due to RH Pressure drop 8 DRRHS Heat rate change due to RH spray flow 8 DARRHT Heat rate change due to HRH steam temperature & T6l4 Gross turbine heat rate ‘BTU /KWH Note that the above data points are all calculated by some previous subroutine. PCSUM collects the calculated data and displays the results in a concise format. Note that PCSUM collects the corrections in tems of percent whereas corrections sent to History Files are delivered in units of BTU/KWH. Sumary on Performance Calculations ‘the performance calculation subroutines use plant data to calculate the actual plant heat rate. Corrections are also calcalated to account for components which are not up to design capacity. The calculations are made at ten-minute intervals and are intended to be run in the background. ‘the results are sent to a History File where they can be exanined through use of DATA STORAGE & RETRIEVAL prograns . SUMMARY In response to the need to accurately gauge a unit's perfomance, a series of pro~ grams have been written and new prograns will continue to be written vhich collect data, calculate performance indices based on this data, and store the data and calculated values for future reference. The means have been provided to calculate an actual unit efficiency which takes into account the actual operating conditions. 3-27of each of the major components in the plant. ‘These prograns are designed to provide concise performance indices so that an operator can view the unit's perfor~ mance at a glance. At the sane time, provisions have been made to allow an operator to easily obtain additional information, if desized, regarding the data used to calculate each performance index. REFERENCES [1] McClintock, R.B. and Silvestri, G.J., "Formulations and Iterative Procedures for the Calculation of Properties of Steam", ASME publication H-17, ASME, 345 B. 47th Street, New York, NY¥ 10017. [2] McClintock, RB, and Silvestri, G.J., "Same Improved Steam Property Calculation Procedures", Paper No. 69-WA/Pwr-2, ASME, 345 B. 47th Street, New York, Ny 10017. 3-28Section 4 PERFORMANCE MONITORING OF MAIN STEAM CONDENSER one of the major components of a Rankine Cycle power plant is the main stem condenser. Of the heat added to steam in the boiler only a portion is converted to useful work through the turbine. ‘the rest mist be rejected to anbient in order to complete the cycle. The process of rejecting heat is performed in the main condenser. Heat is rejected by causing the stean exiting fram the last stage turbine to be condensed to water. this latent heat is transferred to the condenser cooling water vhich in turn transfers the heat to the enviroment. the cooling water usi- ally originates and ends at a large natural body of water such as the sea or @ river. Some power plants incorporate cooling towers in which the cooling water transfers its heat to ambient air. ‘this heat rejection process is very important to Rankine cycle efficiency. Basic thermodynamic principles danonstrate how the efficiency of a power plant can be increased by reducing the temperature at which heat is rejected. This means that the saturation temperature of the condensing steam should be kept as low as prac- tically possible. This is accomplished by optimizing the heat transfer rate be ‘tween the condensing steam and the cooling water. a1Once a condenser has been designed ani built, the way to optimize heat transfer is to insure that the condenser's physicel condition and operation is maintained as Closely as possible to the original equipment. Sane of the things that affect heat transfer are: tube fouling, tube restriction, cooling water flow rate, unit load and cooling water inlet temperature. Sone things are easily controlled while other items require costly condenser disassembly. The purpose of the main steam condenser performance monitor is to measure a series ©f condenser operating parameters on an on-line basis and compile the readings into specific performance indices. ‘These indices can be viewed immediately to obtain on-line knowlaige of the condenser's operating condition. In addition, these indices can be stored for long tem trend monitoring. ‘he resulting information will be useful for dispatch service and also for optimizing outage schedules. CONDENSER SPECIFIC DATA ‘The condenser used to test the performance monitor is the main condenser for the Potomac Electric Power Company's Morgantown Unit 42 generating unit. this unit is rated at 575 MH with the main stem throttle set at 3500 psig and 1000°F. the unit is a super-critical unit with single reheat to 1000%. ‘the unit design is based on maintaining a condenser vacuum of 1.25 in Hg. ‘The condenser is a single tube pass, single pressure, surface type heat exchanger. A vertical baffle splits the incoming steam into two identical passes. the com denser is rated to condense 2,420,000 LBS/HR of steam with a cooling water inlet temperature of 60°F and a cooling water flow rate of 496000 GM. The heat transfer area is 157625 sq.ft. per pass and is formed by 27472 tubes. The tubes are made of 70-30 copper-nickel. Each tube has an outside diameter of 1.027 inches with a wall thickness of 0.049 inches. 4-2CURRENT PERFORMANCE MONITOR one of the existing perfomance calculations currently determines a condenser cleanliness factor. This calculation is part of DSR (ata Storage and Retrieval) as well as PC (On-Line Performance Calculations). This routine calculates a heat transfer coefficient and compares it toa variable standard. ‘The ratio of the calculated heat transfer coefficient to the standard heat transfer coefficient is defined as the cleanliness factor. ‘The heat transfer coefficient is calculated using the log-mean temperature difference method. That is: where 1) a - arin = —2—+ 7, - 7, iouier in eS paige Note: Definition of all symbols is given in Table 5- In this equation, the tenperatures are measured on-line and averaged over 10 min. while the heat rejected is calculated at ten minute intervals from data extracted fron the heat balance perfomance calculation (PCHB). ‘he heat transfer area is constant. 43‘the "standard" heat transfer coefficient is also a calculated value but is determined from an empirical relationship. this relationship canes fran tests run by the Heat Exchanger Institute and is the result of a wide variety of tests on various condensers. It is a general relationship which attempts to adjust a design eat transfer coefficient for changes in inlet tengeratures and cooling water fluid velocity. ‘the design heat transfer coefficient denotes the clean tibe heat transfer characteristics of the condenser ani represents the maximum attainable heat transfer capability of the heat exchanger. In brief, the relationship is: Unei = Ua FTV (42) ‘The temperature correction factor (Tq) is obtained fram a fourth-order polynomial curve-fit given the cooling water inlet temperature. The velocity correction factor is simply the square root of the average tube velocity. ‘The cleanliness factor is then defined as: CF = Ope a3) ‘This method is useful if the interest in condenser operating condition is Limited to the deviation of heat transfer fram an arbitrary benchmark. The resulting Cleanliness factor is simply a measure of how far the condenser's condition is from some reference point. It does not necessarily indicate the absolute cleanliness of the condenser or what an optimum cleanliness might be. For example, the cleanliness factor calculated in this manner is not necessarily a mmber between 0 and 1, Also it does not indicate bow the cleanliness factor might be improved by adjusting the operating conditions. a4NEN PERFORVANCE MONITOR Anew method for condenser performance monitoring is being designed with the in- tent of providing a comprehensive set of perfomance parameters. This means that a variety of indices will be provided which can be calculated several different ways to allow maximm flexibility and cross check capability. Three itens of pri~ mary concern are: Attainable vacuum pressure ‘Tube fouling ‘Tube restriction ‘The attainable vacuum pressure is that pressure which would be attained if the condenser were clean. ‘he prime objective of efficient condenser operation is to reject heat at the lowest practical saturation temperature. Condenser vacuum pressure is directly related to steam saturation temperature so that the at— tainable vacuum pressure then should correspond to a value which could realistically be achieved by cleaning the condenser. tube fouling is @ build up of deposits on the inside of the tube walls due to im purities entrained in the cooling water. Tenperature gradients and fluid velocity Fluctuations cause the impurities to cone out of solution resulting in a scaly build up. ‘These scaly deposits fom a themal barrier vhich hampers the capebil— ity of the condenser to transfer heat effectively. ‘ube restriction results from snall objects entrained in the cooling water flow. ‘hese objects are swept along by the flow of cooling water and eventually are carried to the entrance of the condenser tubes where they becane lodged and block flow. ‘This reduction in flow is accompanied by an increase in dynamic head as the cooling water pump works harder to push flow through the restrictions. Heat transfer will be impaired if whole tubes or groups of tubes are blocked off. 45‘The three items mentioned above require constant monitoring if condenser performance is to be optimized. Reliable indices would be valuable in determining whether an outage for condenser cleaning is justified. In addition to maintenance concerns, on-line knowledge of the condenser physical condition will indicate the effect of its condition on the mit heat rate. Redundant calculations are incorporated in which calculated parameters are based on different measured values. For example, the cooling water flow rate may be calculated if the cooling water inlet ani outlet tenperatures are know. Conversely, the cooling water outlet temperature may be calculated if the flow rate and inlet temperature are known. By back calculating some of the measured parameters a redundant system can be built which will provide expected results to compare against measured results. This will help to eliminate bias errors and Provide a cross-check for suspect data points. Cooling water flow rate neasure~ ment instrumentation is planned for installation in late 1986. ALL of the above mentioned features are being incorporated in a Condenser Analysis Model. ‘This model is being developed using standard condenser design tools and calculations based on historical data fron the PERCO plant. Historical data is being used to make trial nuns for progran verification and to gain insights into the best method of presenting results, As work progresses, new itens my evolve and will be incorporated as additional features. Currently, a core model has been developed and is the basis for future develoments. Condenser tnalysis Develogment will continue through 1986. 46Figure 4-1 shows the input and output parameters for the new condenser clean| iness monitor. Table 4-2 lists the specific input parameters while Table 4-3 lists the specific output parameters for the condenser core model as it currently exists. symbol Ay rg ae oD D, i Do PERE PEO ™ Neubes con re Peripheral Steam Flowrates Attainable Measured . |} Vacuum Temperatures Pressure Measured Condenser Cleanliness: Pressures Analysis |—— Factor Measured Miscellaneous Water | Performance Flowrates undices Figure 4-1. Schenatic of Input/Output Parameters for Condenser Analysis Model ‘Table 4-1 LIST OF SYMBOLS USED IN THIS REPORT Description Condenser Tube Flow Area Condenser Heat Transfer Area Condenser Cleanliness Factor ‘Average Specific Heat of Cooling Water Condenser Tube Inside Diameter Condenser Tube Outside Diameter HEI Condenser Effectiveness Gravitational Conversion Factor Condensing Steam Film Coefficient Condenser Tubes Heat Transfer Coefficient Cooling Water Film Coefficient Condensate Film Themal Conductivity Condenser Tubes Thermal Conductivity Cooling Water Themmal Conductivity Number of Condenser Tubes Heat Added by Boiler Heat Added by Booster Pun Heat in Make-up Water 47 unit £02 £02 ‘BIU/Ibn/F inches inches PT/sec/sec BTU/tr/£t2/P BTU/hr /£t2/F BIU/hx/£t2/P BIU/hr/£t/P BTU/hr /£t/F BIU/hr/£t/E BTU/hr BTU/hr BTU/hrPOINT 179 5200 201 8202 8203 8204 8205 8206 18207 $208 $209 8210 s2u $250 5251 8254 8255 S256 S268 S274 S275 S616 8627 S646 W200 Heat Rejected Through Condenser Overall Fouling Factor Clean Tube Oxide Film Resistance Non Cleanable Portion of the Thermal Resistance Cooling Water Bulk Fluid Temperature HEI Temperature Correction Factor Cooling Water Inlet Temperature Cooling Water Outlet Temperature Steam Saturation Temperature Attainable Overall Heat Transfer Coefficient HEI Design Overall Heat Transfer Coefficient HEI Corrected Overall Heat Transfer Coefficient Condenser Overall Heat Transfer Coefficient Cooling Water Average Velocity Condensing Stean Flowrate ‘Turbine Shaft Power Cooling Water Flowrate Log Mean Temperature Difference ‘Temperature Difference Across Condensate Film Cooling Water Viscosity at Bulk Temperature Cooling Water Viscosity at Film Temperature Latent Heat of Condensate Average Condensate Film Density Average Cooling Water Density ‘Table 4-2 CONDENSER ANALYSIS INPUT LISTING DESCRIPTION Turbine shaft power BAROMETER MAIN STEAM NORTH MAIN STEAM NORTH MAIN STEAM SOUTH (MAIN STEAM SOUTH COLD REHEAT NORTH COLD REHEAT NORTH COLD REHEAT SOUTH (COLD REHEAT SOUTH HOT REHEAT NORT HO? REHEAT NORTH HOT REHEAT SOUTH HO? REHEAT SOUTH MAIN STEAM NARROW MAIN STEAM WIDE ‘COLD REHEAT NORTH COLD REHEAT SOUTH HOT REHEAT STEM SM TO DEA BIR 4 (COND 2A VACUUM COND 28 VACUUM Main steam flow Hot reheat stean flow LP turb exhaust flow (COND-2A WATER IN 4-8 Bru/he hr-£t2-"/BTU hr-£t2-F/BTU hr-£t2-8/8TU moras ton BTU /he/£t2/F ‘BIU/hx/£t2/E BTU/hr/£t2/E ‘BTU /hr/£t2/E FI/sec ba/he i Lam/ar P F Tan/hr/Pr Eam/hr/FT BIU/LBa m/CuFT Lam /CueT ENGINEERING UNITS 5 msW201 w202 W203 w204 W205 W206 W207 W208 W209 w210 wail waz was W223 w224 W225, W226 W234 W237 W239 W240 W243 Wadd W245 w249 W259 W260 W612 weld é : gagggggzes CIRC WATER COND-B Binal feedwater flow Feed pump A&B suction flow 49 ah Pa gh a Pa Pp ha a a‘Table 4-3 ‘CONDENSER ANALYSIS OUTPUT LISTING Pass-A cleanliness factor Pass-B cleanliness factor Pass-A Log-mean temperature difference Pass-B Log-mean temperature difference Pass-A heat transfer coefficient Pass-B heat transfer coefficient Pass-A clean tube heat transfer coefficient Pass-B clean tube heat transfer coefficient Pass-A fouling factor Pass-B fouling factor Cooling water flow rate Cooling water Outlet Temp Cooling water velocity Attainable saturation temperature Attainable saturation pressure Heat rejected to condenser Pass-A heat rejected Pass-B heat rejected Heat added by boiler Heat in make up water Heat added by booster pup 4-10mooet the condenser analysis program is a series of subroutines which collectively provide a comprehensive report on condenser perfomance. ‘he calculations are standard condenser design calculations with mach of the methodology contained in the ASME Test Code PIC 12.2 [1]. Currently two redundant calculational schemes have been developed which are: Calculate cooling water outlet temperature given cooling water flow rate and inlet temperature. Calculate cooling water flow rate given cooling water inlet and outlet temperature. A third calculational schane is planned in which the heat rejected may be calax- lated fram knowledge of cooling water inlet and outlet temperature and flow rate. More calculations may be added as the need arises. ‘he purpose of redundant calculations is primarily to reduce instrument bias. For example, the cooling water outlet temperature is know to be highly stratified. It is difficult to decide exactly how to weight the average of the tenperature readings. By calculating an expected outlet tanperature a cross~check is provided which increases the reliability of the entize syste. Likewise, by calculating an expected cooling water flow rate, the measured value can be brought into perspective. ‘The two calculations mentioned are performed through a simple heat balance fran knowledge of the heat rejected. ‘The following relationship governs the calculation: pes = WOOP (To - Ty) (44)By knowing the heat rejected and the cooling water inlet temperature, the cooling water flow rate or outlet temperature may be calculated fran knowledge of the other. The specific heat, of course, is dependent on the cooling water temperature. This means that an iteration is required if the outlet tenperature is the desired parameter. All cooling water properties used in this mdel are based on bulk temperature which is defined as: 12 (tj +1) (4-5) ‘The heat rejected is currently obtained from data extracted fran the heat balance performance calculation (PCHB), specifically the following are drawn fran PCHB: 1. Turbine Shaft Power (KW) 2, Feedwater Flowrate (LB/HR) 3. Feedwater Pumps Suction Flow (LB/HR) 4. Main Steam Flow (LB/#R) 5. Hot Reheat Steam Flow (LB/HR) Low Pressure Turbine Exhaust Flow (LB/HR) Temperatures and pressures corresponding to each flow rate above are measured to obtain the fluid enthalpies. Fron these, the heat aided by the stem generator and booster Pimps can be calculated as well as the heat inherent in the make-up water. The turbine shaft power is then subtracted such that the remainder is the heat rejected. That is: Bry res * atx ~ Sap ~ Gyy) - Wy 3413 RARE) (4-6) 4-12Since the condenser is a split-pass shell design, the resulting heat rejected is divided anong the two condenser halves. ‘the sane is done with turbine exhaust low (as obtained from PCHB) so that: (Ggey)PASS = 1/2 (Gpp5) TOTAL (4-7) (wg) PASS = 1/2 (ug) TOTAL (4-8) The calculations described this far characverize the beat transfer actually occurring. That is to say that these calculations are based on the condenser's ‘actual physical condition, Fron these results, an "on-line" log-mean temperature difference ani heat transfer coefficient is calculated using Byuation (4-1). At this point, the "on-line" condenser condition is fully characterized. ‘The condenser cleanliness factor is defined as the ratio of “on-line” heat transfer coefficient to the attainable heat transfer coefficient: oF = Uy (9) ‘the attainable heat transfer coefficient corresponds to that value which could realistically be achieved if the condenser were to be cleaned. It is a function of the operating conditions and is adjustable to take into account as-built con denser characteristics. In order to calculate the attainable heat transfer coefficient, it is necessary to first determine the average cooling water velocity. This is obtained by simply dividing the cooling water flow rate evenly among all the tubes. ‘The flow rate per tube is then converted to velocity by virtue of mass continuity. That is (wy) TOBE = (Hy) TOTAL ge ps (440) 4-13and: (yy) TURE TOA, aay where Py is evaluated at the cooling water bulk temperature (7) With the cooling water velocity known, it is possible to calculate the heat trans fer film coefficient on the inside of the tubes, he modified Dittis-Boelter eR ew relationship is used: ‘The heat transfer coefficient on the outside of the tubes (i.e. stean side) is calculated by Nusselt's equation: aedg) 4 =0.725 | (423) y Brat, a4The heat transfer coefficient relating to the tube wall conductivity is calculated from: @aay The attainable heat transfer coefficient is then: 1 a= (15) ead BT RY Bac * Reet where hy and hy aze calculated in tems of the tube outside heat transfer area. ‘The above calculations require knowing the inside and outside tube surface tem peratures and the steam saturation temperature. the required temperatures are based on attainable performance and are obtained through iteration. A reasonable value is chosen for the stean saturation temperature to begin the iteration. A heat transfer coefficient is then calculated based on the resulting tube wall temperatures and water properties. The steam saturation temperature is then calculated using the log-nean temperature difference method and compared against the assumed value, ‘The assumed value is then adjusted and the iteration sequence repeated until a consistent saturation temperature is calculated. Convergence is essentially vhen the heat transfer coefficient calculated by sim ming the reciprocal of all the parts is equal to that calculated from the log-mean 415temperature difference. The resulting saturation steam temperature is then converted to steam pressure to obtain the attainable condenser vacum pressure. Finally a fouling factor is calculated, defined a (4-16) a Uy In the equation (4-15) for attainable heat transfer coefficient, the clean tube oxide film resistance (Ry,) was included to take into account clean tubes which are not pare. New tubes will build up an oxide film fran reaction with ambient air during fabrication and installation. the currently used value for Ry, is 0000263 (hr-SqFT-°r) /aTU as extracted fran Reference [1] for 70-30 copper-nickel tubes. Another factor in Bquation (4-15) is Rugya- This thermal resistance is included a5 a separate input which can be changed quite easily. This parameter is intended to be used as an overall clean tube fouling resistance. As a condenser ages, it becames more difficult to clean the tubes back to their original "as-built" Cleanliness. By examining data immediately following a contenser cleaning, a value for Ropyd May be beck-calculated for insertion in future calculations. In this way, the condenser clean-tube heat transfer coefficient may be calibrated so that a calculated cleanliness factor of 1.0 will refer to an actual attainable cleanliness. 4-16RESULTS, The aforenentioned model was programmed ani incorporated into DSR (Data Storage and Retrieval), Historical data was used to verify the program and obtain sample results which could be evaluated for reasonableness. Historical data was rm through the program with all resulting parameters being plotted as a function of unit load, The data used was obtained fran PEPCO during a turbine test run im mediately following an outage in May 1985. Figure 4-2 shows the results of a direct calculation of cleanliness factor fran raw data, The two dark lines show the calculated cleanliness factor for each steam pass while the dashed line is an overall average. It shows a large spread in cleanliness factor at low loads and very high loads but reasonable results near 550 MH to 600 MW. Tt also shows that the cleanliness factor for each pass appear to cross each other at 560 MW. A characteristic of many instruments is to yield good readings at the calibration point with a bias going in opposite directions on either side of the calibration point. This is especially true if only one deta point is used for calibration. Figure 4-2 therefore implies that a measurenent bias could be causing the large spread in data. Figure 4-2 was generated by measuring cooling water inlet and outlet temperature and calculating the cooling water flow rate. At present, PEFCO does not measure cooling water flow rate so a cross check on outlet temperature is not possible. A sensitivity study was then performed on this condenser program and it was found ‘that the cleanliness factor is most sensitive to the measured turbine exaust pressure and the measured cooling water outlet temperature. If these two parmeters could be improved, it was reasoned that the spread in cleanliness factor could be reduced. 4470.00} CONDENSER CLERMLINESS FACTOR 0.29} o.a0! "200 300 400 500 00 700 wT LAO amu Figure 4-2. Condenser Cleanliness Factor as Function of Unit Load Data Taken from Steady State Turbine Test. at PEPCO MJ #2 May 1985 Figure 4-3 shows a plot of the measured low pressure turbine exhaust pressure. It is evident that the spread in data is substantial at low loads and improves near 600 Mw. The data at 645 MW is an ancnaly which is difficult to explain. It is known, however, that data for the 645 Mi run vas collected almost 7 days following ‘the run which precedai it. pata for all the other runs were collected within 1 1/2 Gays. It is conceivable then that the 645 Mi run may have seen sone altered setting sich caused it to have an abnomally low exhaust pressure. The “assumed” values on Figure 4-3 show readings that were estimated assuming that the the calibration capability near 600 Mi was carried down to lower loads. 438Tana Ty = 20d 4 Hi : . : Rope x i ie i i i mom “200 300 400 ‘soo 600 700 Figure 4-3, LP Turbine Exhaust Pressure Versus Unit Load Composite: Measured Data with Valves Assumed for Constant Instrument Bias Over Load Range ‘A parameter which encompasses both exhaust pressure and cooling water outlet ter perature is the log-mean temperature difference, which is: (4-17) 49The stean saturation temperature is obtained by converting the LP turbine exhaust Pressure into tenperature (by virtue of the steam properties) and averaging this value with the Hotwell temperature. Figure 4-4 is a plot similar to Figure 4-3 except the log-mean tenperature difference is plotted. A curve is dram through the data points to obtain an estimated "best-fit! © aso! 5 i 8 at i i x. : Bis. i 104 ‘00 700 Figure 4-4. Condenser Log-Hean ‘Tenperature Difference Versus Unit Load Composite: Measured Data with Values Assumed for Constant Instrument Bias Over Load Range The condenser cleanliness factor can now be recalculated using the best-fit og-mean temperature difference curve. The resulting calculation yields a clean Liness factor which would result if the accuracy of the instrumentation at near 600 MW was catried dom to lower unit loads. Figure 4-5 is the resulting plot and shows a dramatic inprovenent in the calculation of cleanliness factor over the load range. This plot shows the capability of the model to calculate a consistent Cleanliness fector which is independent of load with proper instrumentation calibration. New instrunentation will be installed throughout 1986 and it is expected that plots similar to Figure 4-5 will be obtainable in the near future 4-20directly from measurenents. ‘Te cleanliness factor calculated thus far is a lumped parameter index and is not very useful in itself except that it does indicate the overall physical condition of the condenser. AS more data fran PEECO is examined it will become possible to extract out of this lumped index a separate index for tube fouling and tube Yestriction. This will be useful to maintenance personnel in deciding shich portion of the condenser requires the most cleaning. Further indices to be developed will incluie things such as cooling water pump perfomance, optimum hotvell levels, effects of inlet stem quality ani others. These features will be er amined for feasibility as new instrumentation is installed and the condenser analysis is further developed. CONDENSER CLEANLINESS FACTOR ‘200 300 00 500) oo 700 ‘Unt Lom Figure 4-5. Condenser Cleanliness Factor as Function of Unit Load Data Assunes Constant Instriment Bias Over Toad Rance for LP Turbine Byhaust Pressure and Cooling Water Outlet ‘Temperature 421somaRy Work was initiated in developing a condenser model which will be used to monitor condenser performance, Realistic performance parameters have been identified to indicate deviation from attainable performance. ‘The data will be useful for optimizing maintenance schedules and for providing on-line performance criteria to dispatch. , ‘A model was developed which uses the log-mean temperature difference method to calculate on-line performance. attainable performance is calculated using the modified Dittis-Boelter and Nusselt equation for calculating the film coefficients. A comparison of the two methods yields the condenser cleanliness factor. ‘This model is incorporated in DSR (Data Storage ani Retrieval) and some historical data has been used to examine the reasonableness of calculated parmeters. Results show that condenser cleanliness factor is independent of unit load and is fa reliable measure of condenser performance. Future studies will focus on inter- preting results from data obtained throwh improved instrumentation and to develop a series of perfomance indices vhich will give a clear and concise picture of the physical condition of the main steam condenser and its peripheral equipment. 4-22REFERENCE [1] Code on Steam Condensing Apparatus; ANSI/ASME PIC 12.2-1983; Society of Mechanical Engineers; August 31, 1983 4-23 The AmericanSection 5 FEEDWATER HEATER PERFORYANCE MONITORING Feedwater heaters are typically used in Rankine Cycle power plants to enhance the plant's operating efficiency. A portion of the main steam flow is extracted and used to heat the feedwater through a series of feedwater heaters. A closed feed water heater is a heat exchanger in which the incoming high pressure steam is not allowed to mix with the feedvater. An open feedwater heater allows for steat to feedwater mixing and is used typically not only to heat the feedwater but to allow non condensable gases to bubble out of the feedwater. When used for this reason, an open feedwater heater is called a deaerator. A closed feedwater heater is usually a shell and tube heat exchanger with the feedvater flowing through the tube side and the steam passing through the shell side. As the two fluids pass each other, the stean gives up sane of its energy to ‘the feedwater cousing the steam to condense while the feedwater temperature is increased. In some designs a desuperheater section is used in which the tenpera— ture of incoming superheated stean is lowered to saturation temperature by the passing feedwater. In addition, sone designs employ a drain cooler section in which the condensate is cooled to a temperature lover than saturation temperature by the incaming feedwater, This sub-cooling keeps the condensate from flashing into steam as it is transported back to the incaming drain of another feedater heater or to be mixed with the colder feedwater. SelFeedvater heater perfomance mist be monitored periodically to rate how adequately the heater is performing its task. However, performance measurenents of feedwater heaters are complicated by the variance of themal/fluid conditions that are typical of all heat transfer equirment. This section describes a program for cal~ culating a realistic perfomance index which accounts for actual feedwater heater heat transfer characteristics as well as changes in unit load. ‘Two parameters useful for characterizing feedvater heater performance are: 1, Terminal Temperature Difference (TTD) 2. Drain Cooler Approach (DCA) Temninal Temperature Difference is the difference between the feedwater outlet temperature and the stean saturation temperature. the drain cooler approach is the difference between the drain cooler exit temperature and the feedwater inlet temperature. These tems are illustrated in Figure 5-1. PEPCO's MORGANTOWN UNIT #2 The power plant being used for testing the performance monitoring algorithms is Morgantown Unit 2 of the Potomac Electric Power Canpany (PEPCO). It is a 575 MM supercritical fossil fired power plant. Main stean operating condition at rated load is 3500 psig and 1000°F with single reheat to 1000°r, ‘The feedwater heating system consists of six closed type feedwater heaters and one open feedwater heater. Three of the closed feedwater heaters are placed before the open feedwater heater (i.e. towards the main condenser). ‘These are the low Pressure feedwater heaters. Three are placed following the open feedwater heater (i.e. toward the steam generator). These are the high pressure feedwater heaters. ‘The feedvater heaters are arranged so that the condensate from the three high pressure heaters drain into the next lower pressure heater in a cascade 5-2« © Figure 5-1. Tenperature-Length Diagrams of: a) Low Pressure without Drain Cooler b) Low Pressure with Drain Cooler ¢) High Pressure Feedwater Heaters fashion until it reaches the open feedwater heater where the drain is mixed with feedwater. Fron here the feedwater is pumped by a booster pmp to the nearest high pressure feedwater heater on its way to the steam generator. the drains fran the three low pressure feeduater heaters are cascaded back to the last feedvater heater vhere it is then pumped forward to mix with the feedwater leaving no. 7 heter and entering no. 6 heater. PERFORMANCE MONITORING ‘The current method enployed at PEPCO's Morgantown Unit #2 (as described in Section 3 of this report) for feedwater heater performance monitoring is quite simple. The steam and feadater temperatures at the inlet and outlet of each feadwter heater are measured along with the corresponding static pressure. Steam extraction and feedwater flowrates are obtained fron the heat balance performance calculation (CHB). For each feedwater heater, an ideal DCA is added to the actual 53temperature to obtain a reference drain outlet temperature. To the actual TTD an ideal TID is added to obtain a deviation fran ideal 17. Feedwater heaters not in service are properly accounted for. ‘The deviation in heat rate is obtained from these parameters. Figure 5-2 illustrates the process. FW Heaters not in Service | Measured Temperatures = Measured Pressures ————>} Poe |B Flowrates ———>| [> eat iene “Rate Ideal DCA 3m Ideal TTD, ————2> Figure 5-2. Input and Octput Parameters of Feedwater Heater Performance Monitoring System for PEPCO - MU #2 ‘The process mentioned above is little more than comparing an actual messured TID and DCA to sone ideal value. rt is tempting to replace the tem "ideal" TTD with “design 1m but this would be an inaccurate substitution. the "ideal" TTD and DCA were specified by the plant designer and not the feedwater heater designer. typically, when a plant is under construction, the plant designer will specify a value of TID and DCA to obtain the optimm unit heat rate. It is then up to the feedwater heater manufacturer to provide a feedwater heater which mets this specification. The purchase of feedwater heaters is generally a carefully thought out process but it is hard to say whether the actual feedwater heaters will ever stdeliver the specified or "ideal Tm ad OCA. Normally, under these circumstances, a test could be run to obtain actual “as-built” performance parameters. Testing of fecdmter heaters is not a straightforward process, however, because all the flowrates are coupled, making it difficult to obtain spe cific test conditions. All test data must be correlated by analytical means and few guidelines exist for feedwater heater performance calculations. another problen is that the ideal TTD and DCA entered into the algorithn are com stant values. For exemple, the ideal TD for the high pressure heater is -1°F, irrespective of unit load. sven if the feedwater heater were carefully designed to actually deliver -1°F TD at rated load, it is doubtful that TTD will be -1°F fat other than rated load. In fact, at lover loads the feedvater will not ex perience as high @ temperature rise because of the reduced saturation temperatures in each of the feedwater heaters. It should be expected then that actual TID and DCA depend on the unit load and the reference values used to determine the devia tion of these parameters should also reflect part load conditions. Stated another way, heat exchangers are designed to provide specified perfomance at sone design condition (i.e, temperature, pressure, and flow). It is rarely possible to obtain this specified performance at other than design conditions due to the nature of themal/fluid processes. By not allowing for a variable TTD and OCA, it is uncertain whether the actual measured TTD and DCA deviates as a result of fouling (or equipment damage) or be~ cause of a unit load change. Deviations in inlet conditions affect the reference parameters as described above. In order to determine the deviation in heat rate due to fouling alone, it is imperative that an accurate reference TTD and DCA be provided. By accurate is meant a value which adequately characterizes the at— tainable performance and then is properly adjusted for unit load. 55VARYING TTD AND DCA A program has been written which incorporates variable 71D and DCA into the per formance calculations. this progran performs a heat exchanger heat balance to obtain TID and DCA as a function of inlet conditions. This program was spliced into the existing feedwater heater performance monitor and provides the existing monitor with reference values for TID and DCA Figure 5-3 illustrates the salient = mse, innit ate = Temoerature: jeasured (Inlet and Outlet) Pressures Fon LL, AU Flowrates Deviation +» “in " Heat Rate REF DCA (intet only) REF TTD (inlet only) "Clean" Feedwater Heater Characteristics Figure 5-3. Input and Output Parameters for Feedwater Heater Performance Mon‘toring with Variable TID and DCA The algorithm denoted schematically in Box II is unchanged fron that depicted in Figure 5-2. The new perfomance calculation is contained entirely in Box I. Temperatures, pressures and flowrates at the inlet of each feedwater beater are tased to calculate a TTD and DCA for a "clean" feedwater heater. ‘he clean feed water heater is characterized by input parameters which were based on nme plate data provided by PEPCO for each of the Morgantown Unit #2 feedwater heaters. 5-6MoDEr. ‘The feedvater heater perfomance calculation (Box I of Figure 3) involves using the -NTU method to detemine outlet conditions as a function of inlet parameters and heat exchanger perfomance characteristics. This method is a standard heat exchanger analysis tectnique [1] and is essentially a non-dimensional heat balance, For ease of calculation, each feedvater heater is divided into three separate heat exchangers: A desuperheating heat exchanger 2. A condensing heat exchanger 3. Adrain cooling heat exchanger. Figure 5-4 illustrates the model for each feedwater heater. hesP Ve, . ey > - inet Parameters > - Flowrates + = Calculated State Points Figure 5-4. Analytical Model of One Typical Feedwater Heater Showing Three Sections: DS - Desuperheater; C - Condenser; DC ~ Drain CoolerThe input paraneters are designated by a large shaded in arrow. the incoming drain is shom entering into the condenser region. All flowrates are obtained fran the heat balance perfomance calculation (PCB). ‘he paraneters show near the snall dots are calculated fran the input parameters and the feedvater heater performance characteristics. Awell established formula for defining the performance characteristics of a heat exchanger is the log-mean temperature difference equati a= Wann (S-) In this equation, q is the heat transferred and Tim is the log-mean temperature difference which for a counterflow heat exchanger, can be defined as follows: po ~ Tes) ~ Tat ~ Too? mn : i (5-2) T an Shoot (Ts ~ Too) where ‘Tho = Hot fluid outlet temperature Thy = Hot fluid inlet temperature ‘Tee = Cold Fluid outlet temperature 222 = Gold fluld inlet tempereture 5-8If all the temperatures are known and the heat transferred is known, then: 6-3) ‘this last equation provides a means for calculating the heat exchanger performance characteristics. The heat transferred may then be calculated from? a =o (47) 6-4) where = Fluid Flowrate (1bm/hr) op = Average specific heat (BTU/Iin°F) ‘T = Temperature rise of fluid (°F) so that: Se) 6-5) 59In this way, the attainable feedwater heater performance characteristics may be embodied in the UA tem. ‘A UA term was calculated for each section of each feedwater heater. Table 5-1 Lists the data which was provided by 2EPCO to calculate the feedwater heater heat transfer characteristics. Table 5-2 lists the resulting heat transfer coefficients for Morgantown Unit #2. It should be noted that in the above calculations there is no input data for tan peratures Tz, and Tyz. These temperatures were calculated by performing a heat balance and noting that the heat transferred by a reduction in steam temperature drop in the desuperheater is the sane as the heat transferred by an increase in the corresponding feedwater temperature rise. The same is true in the drain cooler section, ‘The -NTU method is used to calculate output conditions for any given inlet condi- ‘tions and for the heat transfer characteristics listed in Table 5-2. This method incorporates the following relationships: wu wm ca 6) 5-20err"o wor |e we ont | 9 0t"o 1 wn | s sut"o ost oer |e set'o ua our | oz zero ow get | ot (gore) cores) | cotyoum | + oe °a Sy | a 24 ZINN NYOINWSYON OOdEa YOd WING SONWMORA WALL WAIVEH URINNGEH CHSOID THs eraen SL° 20"605 ° - oouet - L 86°196 ostezo ° cont oczot voresy ereee ° 599 octs - ores erst 89°00 Tove oot ocet € a y 6e-90t et | tee Wiser oe"sus rerwse ose sone 999 q Notsous noricas norsoas norsoas sorcas noriaas ‘ 1000 A rma | wasieranco | wauvansansaa wsoco wivea | wasuaqnco | wauvaiwaanssa | an @o-pua-mY/ nse) (aa) arD1ag500 wasted IaH ane waasioens, ava 2H INN NMOINYEUON COdHd YO SONVMUORA WONLLAO YOX SOLISTAAONIVHO SAAGNWL INGH UEINGH WAIVE CSOT 2-5 erqen1 - ew Camu (1 - cH) e= 1 Ct exp Cam (1 = C42 ou e= Fideal — Cain (Trax ~ Tain) a) where C= op (for any given fluid and section) cin = Lowest © for a given section max = Largest C for a given section ct = Gnin/onax ‘max = Highest tanperature in section Tmin = Lowest temperature in section Here NTU is a non-dimensional heat transfer coefficient and ¢ is the effective ness of the heat exchanger. The effectiveness is defined as the ratio of the ac- ‘tual heat transferred to an idealized maximm heat transferable. ‘The relationship between € and NIU shown above is valid for a counterflow heat exchanger and allows for the calculation of the heat exchanger temperature rise if NIU is knom. 5413By caleulating © and NIU for each section (i.e. desuperheater, condenser ani drain cooler) an outlet temperature for each section is detemnined. of course, some iteration is required because the intemediate temperatures (Ty, Tye, T¢) are unknown. This iteration, however, converges very quickly (usually within four to six iterations with a 0.005F convergence criterion on ‘y) « Since WA denotes the unfouled tube characteristics of the feedwater heater, the calculated outlet temperatures will be those values expected for maximum attainable heat transfer. Fron the calculated outlet tanperatures, a reference TID and DCA are easily calculated ani can then be delivered to Box II of Figure 3 for incorporation in the deviation in heat transfer calculation. PROGRAM DESCRIPTION The sequence of calculations is to first evaluate the saturation conditions in the condenser portion of the feedwater heater, Feeivater specific heats are then evaluated. I£ a desuperbeater section exists, its parameters are evaluated. The same is true if a drain cooler section exists. Finally all the section parameters are lumped together and an iteration process begins using the ¢-NTU method to balance the inlet conditions with the heat exchanger characteristics. the result a series of outlet temperatures which are used to calculate a reference TTD and DCA. ‘The reference TTD ard DCA are then combined with the actual measurei feedwater heater inlet and outlet temperatures to determine the effect of this devia~ tion in performance on heat rate. S14The input requirements for this program are listed below: POINT [ENGINEERING NAME DESCRIPTION UNITS D200 HEATER-IA DRAIN FE p20 HBATERS-1B DRAIN EF 202 HEATER 2 DRAIN FE D204 HEATER 3 DRAIN E D205, HEATER 5 DRAIN E 206 HEATER-6A DRAIN E D207 HBATER-68 DRAIN E D208 IR DRN POMP DISC E D210 REBOILER DRAIN F D213 STM TO AIR HEATER KLBS/H D214 STM AIR HTR DRN B F pais REBOILER DRIPS KLBS/#t pais STM AIR BTR DRN A E D610 Heater 3 drain flow /R D615 Heater 5 drain flow 1B/ER 620 Heater 2 drain flow B/eR 2239 GEN 2 MH Me HES Htr #6 extraction enthalpy BrUAB HET Her #7 extraction enthalpy BIU/AB 8179 BAROMETER HG $220 EXTR 10 HP HTR 1A FE 5221 EXTR TO HP HTR 1B Er $222 EXTR 10 IP HIR 2 E $224 EXTR 10 IP HIR 3 E 8225 ‘SIM TO DEA TR 4 E 8226 B EXTR 10 LP HTR F 8232 ‘SEAL STM 10 HIR 7 E 8263 EXTR 10 HP HTR 1A PSIG S264 EXTR 10 HP HTR 1B PSIG 8265 EXTR TO IP HIR 2 PSIG 8267 EXTR TO IP AIR 3 BSIG 8268 STM TO DEA HTR 4 PSIG 8269 B EXTR 10 LP HTR PSIG 8270 BXTR 10 LP HTR 6A PSIG S27. EXTR TO LP HTR 6B PSIG s272 EXTR TO LP HTR 7A BSIG 8273 EXTR TO LP HTR 7B PSIG $292 SEAL SIM 10 HTR 7 KIBS/H $632 Heater 2 extraction flow B/E 3634 Heater 3 extraction flow 1B/R 3636 Heater 5 extraction flow B/HR 8637 Heater 6 extraction flow IB/R 8638 Heater 7 extraction flow B/ER 8647 Heater 1 AGB extraction flow 1B/HR T614 Gross turbine heat rate ‘BIU/AWH W216 ‘COND HEATERS 7 IN FE W217 COND LVG RTR 7A E w2is ‘COND LVG HTR 7B E w2i9 ‘COND ING HTR 6a E W220 COND LNG HTR 6B E W221, COND ENT HEATER 5 F 5-15w222 W226 W228 W229 W230 W231 W232 W233 W238 W239 w243 W246 wel A sample portion of the output listing is shown below. In addition to vhat is shown, the program will output the list of input parameters along with their respective values. Correction for Off-Design Heater Performance HR adj - all heaters: HR adj - all heaters: atx 3 Her 5 Htr 6 Htr 7 Her 1 Htr 2 Htr 3 Htr 5 Htr 6 reference DCA: Generation: 21,5082 -0.260383 0.7035048-01, 0.5760292-02 0.2237878-03 -0.1314325-01 01438448 0.114875, 1.49542 5.54395, =1.40875 4.70453 4.34909 3.73117 0.848145 3.73248 6.30658, 7.96835 742154 589.784 Gross turbine heat rate: 8260.23 5-16 BIU/RWH GE Tenn Hmm awwwee oSUMMARY ‘tthe preceding discussion outlines a method for calculating a variable reference TD and DCA to more realistically determine the deviation in heat rate for feed~ water heaters operating at off-design conditions. This methodology has been incorporated into DSR (Data Storage and Retrieval) as shown in Figure 5-3. ‘The incorporation of this method into the interactive perfomance calculations of DSR and eventually into the on-Line performance calculations will allow for a nore accurate estimate of the deviation in heat rate due to off-design beter performance. In addition, it should be possible to pinpoint trouble regions by comparing actual performance of each section to its attainable performance. output parameters include specific data which will be an aid to feedwater beater troubleshooting. The exact form of the output will be modified as more historical data is used for calculating real performance values. 5-17REFERENCE a Heat Transfer; 1974. Chapman, Alan J.; Macmillan Publishing Co.; 5-18 New York, NY;Section 6 RANKINE CYCLE EFFICIENCY ‘A basic measure of unit performance is Heat Rate. ‘The Heat Rate is a measure of a unit's efficiericy calculated in terms of BIU/WwH. It is simply the ratio of the amount of thermal energy (fuel) required to produce a given amount of electric energy. ‘The calculation of Heat Rate is a highly involved process which requires the efficiency of every cmponent which contributes to the electrical generation process be taken into account. ‘This is a very involved calculation and is discussed in a separate section (i.e. Section 3 - Calculation of On-Line Performance Indices). A very quick method of obtaining an ideal Heat Rate, however, is through the Rankine Cycle Efficiency calculation. she Rankine cycle Efficiency is a measure of the ultimate best that a unit can perform, It is @ benctmark with which to rate a unit with a minimum of ingut data, It allows an operator to see how the major governing themal parameters affect the mit's efficiency without obscuring the results by including all of the individual canponent efficiencies. It is the purest indication of a unit's performance. |A program has been written and is now a part of Data Storage and Retrieval (058) which will calculate Rankine Cycle Efficiency and Rankine Cycle Heat Rate. It is intended to be used as a rating tool which would indicate the ultimate best that a unit can perform. 61Model. ‘The Rankine Cycle Efficiency calculation is a standard themodynanic textbook calculation. In modelling PEFCO's Morgantown Unit #2 it was important to include a reheat process and a regeneration process. The regeneration process is provided for by supplying one open type feedwater heater which takes into account the en tire feedwater heating process. The model, while designed for PEPCO, is, in fact, applicable to any unit which utilizes reheat and regeneration. the model can be easily adjusted if no reheat exists or if a specific extraction pressure is desired. Also the progran listed at the end of this section will work equally well for a non supercritical uit. A schematic of the Rankine Cycle model is shown in Figure 6-1. ‘he process chart, plotted on a temperature/entropy diagram is shown in Figure 6-2. At state point 1 the stean enters the first stage turbine at its maximum temperature and pressure, It expands isentropically to a stean quality of 1.0 at state point 2 where a portion of steam is extracted for feedwater heater duty. The portion of steam extracted is obtained by performing 2 heat balance on the feedwater heater. The remainder is reheated and then allowed to expand isentropically through the second stage turbine to state point 4. Here the steam is condensed to state point 5 then heated to the saturation temperature of the feedwater heater by the extracted steam resulting in hot water at state point 6. This water is then pressurized to the required boiler pressure by a boiler feed pump and is heated to the tempera ture of state point 1. 6-2Ist stage 2nd stage 7 Ral Condenser Extraction oe heater Pigure 6-1. Schematic of Rankine Cylee Power Plant with Single Reheat and Regeneration Figure 6-2. Tenperature-Entropy Diagram for Rankine Cycle with Single Reheat and Regeneration 6-3Input /outpat The Rankine Cycle calculation is performed with a minimum amount of inputs. This is because it is the ideal thermodynamic processes which carry the steam and water through the cycle. the inputs simply put an upper and lower bound on the cycle calculation. This is similar to a Carnot efficiency calculation in vhich the only inputs are the highest and lowest temperatures in the cycle. The following is a list of the required input for the Rankine Cycle Efficiency calculation: POINT ENGINEERING NAME DESCRIPTION UNITS 5239 GN 2 i $179 ‘BAROMETER ac. 5200 MAIN STEAM NORTH E S201 MAIN STEAM NORTH E 8202 MAIN STEAM SOUTH E 8203 MAIN STEAM SOUTH E 5208 HOT REHEAT NORTH E $209 HOT REHEAT NORTH FE 8210 HOT REHEAT NORTH E s2iL HOT REHEAT NORTH FE 8250 MAIN STEAM NARROW PSIG s25L MAIN STEAM WIDE PSIG 8274 COND 2a VACUUM, 1G 8275 COND 28 VACUUM HG Tt should be noted that this input lists anounts to simply: yn 2) 3) 4 Main Steam Pressure Main Steam Temperature Hot Reheat Temperature Condenser PressureThe unit load input is included for reference. ‘Shown below is a sample output provided by the Rankine Cycle Bfficieny calculation: Rankine Cycle (Ideal) Efficiency Rankine Cycle Eff: 43.5027 8 Rankine Cycle Heat Rate: 7036.71 BTU/KWH ‘Total Turbine Work Out: 587.379 -BTU/LBn Total Pump Work In: W108 BTU/LEm Total Heat Addition: 168.1 BTU/LBn Final FW Temperature: 458.196 FB Extr Stm to Main Stm: 0.325392 FRAC LP Turb Exhaust Quality: 0.375694 FRAC Main Steam Enthalpy: 1024.87 BTU/Lan Cold Reheat Enthalpy: 1204.79 BTU/Lan Hot Reheat Enthalpy: 1521.79 BTU/LEm LP Turb Exhaust Enthalpy 977.339 BIU/LBn Condensate Enthalpy: 70.3733 BIU/LBn Pump Inlet Enthalpy: 439.503 BTU/Lan Pump Outlet Enthalpy: 450.614 BTU/LEm HP Turbine Entropy LP Turbine Entropy 1.47203 BTU/LBm-F 1.74744 = BIU/LB-E Unit Load: 589.784 Main Steam Press: 3525.13 PSIA, Main Steam Temp: 1004.59 F Hot Reheat Temp: 1000.48 B Condenser Press: 1.01950 PSI 65Section 7 PRE-OUTAGE TURBINE EVAWATION AND ANALYSIS PROCEDURE A systematic analysis of historical performance tests and inspection reports can be used to improve maintenance planning. ‘he objective of this paper is to out~ Line one such analysis system used to predict the internal condition of a turbine prior to being opened for inspection. ‘The evaluation of data fran a test prior to stutdown was used to anticipate the condition of the steam path of the turbine, enabling improved planning for spare parts and repair procedures, At a subsequent internal inspection of the turbine, the evaluation was also used for comparison with the actual physical condition of the steam path including calculations of loss due to deterioration of buckets, nozzles, packings and other flow path components. After completion of repair of the turbine, its condition was again evaluated to determine the expected performance based on the repaired condition. Performance tests after startup and analysis of the associated data was the final calibration step in the process of test-inspect-repair-test evaluation. ‘The Morgantown 2 unit was used to illustrate the use of the procedure. the evaluation of that unit both before and after the major outage of May 85 was a vital step in the procedure. mmcoats ‘The major purpose of the following analysis tecmiques is to enable reasonable pre-inspection predictions of the internal stean path condition of large steam turbines. Such predictions enable improved planning including: selection and ordering of high-cost, long lead time spare parts; probable outage repair procedures; modifications to operating methods; recomendations for improved instrmentation and test activities; optimization of outage timing; and scheduling on a cost/benefit basis. [APPROACH ‘The evaluation and analysis techniques are presented in four sections as outlined below: @ Summary of the specific analyses with individual goals and data requirements. For each of the analytical procedures, the Pre-outage Morgantown 2 test results was used as an example. @ Results of the actual stean path inspection and correlation to predicted findings. This enabled identification of problens not Properly predicted by the tests and analyses, with subsequent con— siderations of improved tecnique (Appendix 7A). © Inspection of steam path components prior to reassenbly to deter— mine where hoped-for improvements were not accomplished or only marginally completed (Appendix 7B). © analysis of post outage test results. This provides an overall calibration of the procedures, identifying the degree of success: and pinpointing zones where expected performance improvement did not occur (Appendix 7C). ‘These would suggest areas for additional work. 7-2PRE-INSPECTION TURBINE PERFORMANCE EVALUATION ‘The method for performance evaluation is based on the analysis of test information, normal operating records, design information such as heat bales, expected pressures versus flow, section efficiencies and records of previous tur bine inspections. full-scale heat rate tests, while very useful, are not required. ‘The procedure has been categorized into nine basic steps. Each of these will be reviewed in detail and are summarized in Table 7-1. Step 1: Check Unit Flow Capcity Throughout the evaluation procedures, an accurate value of throttle flow at valves wide open (vWO) conditions will be important. If a WO throttle flow measurement. does not exist, the method presented in Step 4 may be used to estimate the probable flow rate. Where other limitations prevent WHO conditions at rated pressure, tests at reduced pressure with Wi0 conditions may be attainable. Such tests will be very valusble, for identifying flow capacity, and also for determining HP sec— tion enthalpy drop efficiency. ‘The test flow, Westy should be corrected for deviations fron design pressure and ‘temperature as follows: Wenrottle corrected = “test * (ay where v = specific volume. The first use of the corrected Wyyr is to compare it. to design. It is fairly normal for turbines to have flow capacity about 2.5% greater than the VWO heat balance (Note: Be sure to use a WO heat balance - not a rated heat balance).Unit Flow Capacity. For Morgantown 2 - Pre-inspection tests (see Figure 7.1). Megat (correcte®) 1.155, or 13.58 excess flow ¥pesign Since turbine manufacturers typically exceed the design flow values by about 2.58, this represents about 10.7% greater flow capacity than nomal. Values outside the 2.5% greater-than-design flow rate should raise several questions: TE test flow is high — ¢ can the turbine first stage nozzle be eroded? (known to be a highly likely condition for this unit) © can 1st stage shell pressure be low due to excessive packing clear~ ance or enlarged 2nd stage area? © can leakages and bypasses exist in piston rings, bell seals, start-up drain valves, etc.? © can HP feed valves to BFP turbines and steam seal systems be open when they are supposed to be closed? © can the flow measurement be wrong? If the test flow is low - © can the Ist stage mzzle be battered closed? © can the Ist stage pressure be high? © can the flow measurement be wrong? ‘The above questions will be addressed as the subsequent sections are completed. Step 2: throttle Wi/P7F Across the Load Range Several flow relationships will provide inediate information regarding the inter— nal conditions of the turbine. the basic flow By. 7-2 is correct for cir— cunstances where local. stage pressure ratios are constant. If erosion, danaje or other problens cause changes to local areas, the equation is no longer accurate. However, it can be used to recognize that a change has occurred and to help es~ 75timate the magnitude of change. -KOA VET (7-2) = the flow to the following stage = a value proportional to effective flow area = throttle pressure (absolute) = throttle specific volume where a supe = The equation can be rearranged to solve for A. W A= —— (7-3) RT In this way, where W (£low to the following stage), P and t are known, it is very easy to determine whether changes in Area have occurred (for most ciremstances the proportionality constant, K, can be ignored). The purpose of this analysis is to determine the changes in Ist stage nozale area at the various valve arcs. The desirable starting point is a plot of design flow values (at valve points) versus W//E7V; Note that since P and v ahead of the throttle valves are normally constant, this will simply be a straight line plot as shown in Figure 7-1. Unfortunately, many of the heat balances are made for convenient percentages of throttle flow (i.e., for 40, 60, 8 and 1008 flow) and not at discrete valve points. In addition, experience has shown that MIC turbines conmonly have their control valves set to improperly follow the design throttle flow versus valve position criteria, This can have significant impact on response, high pressure turbine efficiency and diagnosis of early valve arc flow areas,4000p Turbine Throttle Conditions 4th Valve Ply 2000 3rd Valve ws 2nd Valve 0 20 40 60 80 100 Throttle Flow Figure 7-1. Design Flow When valve point heat balance data is unavailable, baseline test data obtained at the initial start-up or after turbine overhaul may be substituted. Figure 7-2 illustrates actual test data (Morgantown 2) plotted against the design data at valve points. ‘The corrected test Wpnrottze with design pressure and spe cific volume should be used in this plot (Bg. 7-1). It is apparent that the first valve arc flow has increased by about 37%. At full load, the flow increase over design appears to be about 13%. cyThrottle Flow Relationships at Vaive Points [From Morgantown 2] oo clic am L ° - Design 37% L x — Measured ° 20 40 60 80 100 120 % VWO Throttle Flow w/ /P/v Figure 7-2. Actual and Design Flows It should be recognized that the stage flow passing ability is affected not only by area but also by the stage pressure ratio. ‘he stage flow is also affected by off-angle flow conditions, changed reaction, damage to profiles and other physical changes to the steam path. Most stages of the turbine operate with constant pressure ratios across the load range, But the first stage has a highly varying pressure ratio on those turbines utilizing partial arc operation. ‘As long as the 1st stage nozzle pressure ratio approaches or is greater than acoustic (1.83) the throttle flow will be directly proportional to area. ‘The nozzle pressure ratio at valve points can be crujely determined as follows: Phowl (Pressure ahead of nozzle) = Pehrottie ¥ 0-96 (7-4) 18Pp (pressure domstrean of nozzle) = Pict stage shell (7-5) Note: This assumes that the Ist stage reaction is zero. This is an inaccurate assumption but will not significantly compromise the calculation being made here. ‘These relationships are roughly sketched in Figure 7-3 for a six-valve mit, ‘Throttle Pressure Bow! Pressure 4% Valve Drop pee Pressure, Psia Ist Stage Shel! Pressure 0 10 20 30 40 50 60 70 80 90 100 120 Throttle Flow, %VWO, Design Figure 7-3. Turbine Pressures Where test flow is different than design (or baseline) flow, the approxinate change in nozzle area for each valve point can be detemnined as follows: Meese _ Weest(correctes) , Ndesign (7-6) Agesign Maesign Ntest where N is the restriction factor, a function of the nozzle pressure ratio, as shown in Figure 7-4.Restriction Factor, N in Factor B 2 e Ne ° LO i! b2 13 14 15 16 1? 18 19 20 Pressure Ratio Pow) /Pist Stage Shell Figure 7-4. Restriction Factor Note that when sonic conditions exist (N=l), area is directly proportional to flow. In the case of Morgantown 2, the first ani second valve points do have critical nozzle pressure ratios and the change in area can be quickly determined fron the increase in measured flow over design flow: Age valve = 1+37 x design ast + 2nd valve * 1-17 x design Ag valves open ~ 52° Step 3 below 710Keeping in mind the possibilities of piston ring leakage or other bypassing flor conditions, it would appear that the first valve nozzle has increased 37% in area ‘and the second arc is unchanged, (The 378 change in the 1st arc alone is enough to explain a 17% increase when both 1 and 2 are open.) It is Likely that the third and fourth arcs are unchanged also, but this calculation should be held until the first stage shell pressure effects are discussed (next section) . Step 3: WO Flow Versus Ist Stage Shell Pressure Analysis Figure 7-5 shows a comparison of design versus measured pressure for various Design wo Morgantown 2 Pressure a a ° 20 40 60 80 100 120 % Throttle Flow Figure 7-5. First Stage Pressure throttle flow conditions as observed at Morgantown 2. Pressure is plotted instead of W//B7F since specific volume (v) can rarely be determined for this turbine Location. allIf Ist stage shell pressure is low for a given level of throttle flow, it can mean one or more of several things: © adjacent packings (such as N2, 2nd stage diaphragm or HP dummy) may be badly rubbea © 2nd stage nozzle areas may be enlarged © measured flow may be exaggerated © other leakages or bypasses my exist If the pressure is hight © 2nd stage nozzle areas may be diminished © throttle flow measurement may be underestimated Tt can be seen that for Morgantown, the measured first stage shell pressure is ap- Proximately 17% below design at any given throttle flow. This lower pressure can. not be explained by an increase in second stage area or by packing rubs alone, since neither can cause that great an effect. these data suggest that both are enlarged and, further, that the main steam inlet piston ring seals may be leaking as well. Using Bq. 7-6: Design Conditions: FRowl, Prot = 2660; Paoy) = 3374 — = 1.268 Pist N= 0.85 (fran Figure 7-4) ‘Test Conditions: re Fowl ast = 2550; Pyoyy = 3374 — = 1.323 Fist N= 0.89Atest. Mest Npesign resign “Design Nest 0.851 = 1.136 x——+ = 1,086 0.890 ‘his would appear to be a good check that only the first valve arc of nozzles is eroded, The total area appears to have enlarged about one-fourth the increase of the first valve arc. Determination of which conditions exist must be deduced from this and subsequent sections. Keep in mind that past internal problens often reoccur. Past inspec ‘tion reports that indicate heavy packing damage, for example, would increase the odds that rubbed packings have again contributed to low Ist stage pressure. Note that low Ist stage pressure, in addition to increasing turbine flow capacity will increase the available energy on the Ist stage. ‘this lowers its velocity ratio and efficiency while taking energy away fram the more efficient later stages. Step 4: W//P7U Analyses at Hot Reheat (HRH), Crossover and All Extraction Points As indicated previously the relationship W//P7V = K A is essentially a calculation for area. It can be used in three basic ways: 1. to recognize that flow areas have changed 2. to deduce approximate local flow rates (only where flow areas are known to be unchanged) . 3. to validate test data (i.e., plot should be flat over the entire Load range) - 7-13At the hot reheat point, both erosion and damage are frequently present. At the cold reheat point, pressure is usually lower than the design heat balance values due to lower actual reheater pressure drop, also due to Ist reheat stage erosion. ‘At locations below this point (extractions and crossover), design flow conditions usually (but not always) exist. One common complication is the physical location of test measuranents for P and t. These frequently provide different conditions fram those used on the design heat balance. For example, pressure is usually measured in the extraction pipe a dis- tance from the turbine instead of imediately after the stage as shown on the heat balance. A cortection of about +38 is nomally sufficient to correct for this, however if the measurenent is made further domnstrean such as at the heater, a #68 correction may be required. ‘The temperature (enthalpy) at extraction points is usually hotter than that in the steam path by about 15 BTU. ‘This occurs due to the temperature segregation (from spillstrip leakage) that causes a ring of hot steam to exist where the extraction steam is draw off (this would not occur for extractions at the cold reheat or crossover points). The magnitute of high enthalpy in the extractions should be checked by drawing a section expansion line from the bowl to the exhaust, then spotting the extraction enthalpy (corrected back to the pressure at the stage). ‘The specific volume used in the relationship W//P/V should be determined from the conditions at the steam path, thus better xepresenting the steam that passed through the following stages instead of the steam drawn out the extraction point (about 15 Btu colder than the extraction). By doing this, an improved yP7V can be obtained for use in the test W(/P7V calculation for comparison with design. Note that if the test extraction is significantly hotter than the 15 BTU mentioned above, then leakage into the extraction pocket should be suspected. maAnother complication is the desirability to know the flow to the following staje. A preferred basis for W//P7T is where W is flow to the following stage. there this is not known and vhere the feedwater systen is known to be operating normally, it is usually satisfactory to use throttle flow at all points for both design and test. For points at or below the reheat point, correction should be made for reheat attemperation. If the test spray flow is 38, for example, the value Of Wype (throttle flow) should be increased by 38 for calculations at the HRH point and below. As an example of using W{/P7T for determining probable throttle flows, the follow ing unit had zero spray flow and was known to be free of erosion or damage in the IP section (not Morgantowm 2). Table 7-2 DEDUCED THROTTLE FLOW Design WeneWP7 Deduced Throttle Flow Location Test VF77 (x 10-3) (rest VP7W) Wen: WP7 peg) BRE 19.434 212.2 4.1238 x 106 1b/he Sth Stage 10.59 392.5 4.158 x 106 1b/ne LP Bowl 6.68 606 4.047 «106 1b/ar ‘AVERAGE 4.1097 x 106 ib/nr 7s‘The average was used in subsequent calculations for comparing design and test flow; design and test output; and design and test first stage pressure. Use of W/E7¥ for Determining Changes in Flow Area. Where flow area has increased, for instance due to erosion, the value WP will increase. The relationship is far from linear, however, since the percentage change in area will be far greater than the percentage change in W(/P7V- Figure 7-6, below, can be used to interpret changes in flow area based on WA/P7 effects. This figure assumes that the area change is confined to one Stage; that the rotating blades are not affected; and, further, that the pressure ahead of the rotating blades has not changed. These assumptions are not precise, but usually can give reasonable results for estimating area changes prior to opening the unit. Note that if there is a flow exaggeration or a flow bypass, W//P7V will appear large and the deduced area enlargement will also be exaggerated. Several additional suggestions are listed below: ‘The lowest two extractions usually show serious deviation fron design and should not be used to detennine flow. They also should not be trusted to predict area. Patterns of test W/B7V deviation from design may have significance. For example, if the BFP turbine is extracting excessive flow fran the main steam pipe at VWO flow conditions and then discharging portions of it into the crossover area, the HRH and top Ip section W//P7¥ may appear high, while the crossover may appear normal. A general pattern of high WVP7¥ values may either indicate the throttle flow is exaggerated or that steam is bypassing the turbine. Checking KW output and first stage shell pressure my provide additional clues. 7-1640 €0 ‘20 100 % Increase in Area—= Approximate Relationship Between Nozzle Area ond W/VP/v JP Assumes; @ incompressible flow 20 3 no change in pressure 208 ng chenge in ecnomu s na : Numbers on curves are nozzle pressure woh # 40 Figure 7-6. Change in Flow Area 4. If a beater is out of service, or performing poorly, the next higher heater pressure will ‘be low and W(/B/V values will appear high. In the case of Morgantown 2, the pre-outage tests indicate the following: 7-17Design Test [pee hocation WaT Wen Design REL 162 193 0.19 42 Extraction 345, 379 0.10 Crossover 492 590 0.20 ‘The above suggests that the first reheat and first LP stages are greatly eroded open and that even the middle IP stages are significantly enlarged. Serious doubts were felt about these mutbers: © The crossover pressure measurenent Was suspect. ‘* Flow bypassing of the turbine was suspected. * _Exaggeration of the flow by the flow nozzle was suspected. © Increased demand for flow by the boiler feed pump turbine was believed likely. In spite of the above concerns, preparations for repairs of seriously eroded nozzles was recomended. In regard to the flow nozzle, it is generally comon in existing plants to find the flow measurement with a variety of faults: © No flow straightener © Wo calibration © Close-up turns or pipe fittings © Not inspectable 7-18The usual result is an exaggeration of the actual floy — even 3-5t can be expected. This tends to make heat rate appear poorer than actual and also compli- cates the diagnoses of internal problens. here maximm plant loading is established by the flow measurement, it also limits the mit output to a smaller value than was intended, At Morgantown 2, anew flow section was to be installed during the inspection to eliminate the above concerns. Step 5: Compare Test Versus Expected HP Section Enthalpy Drop Efficiency Across the Load Range ‘The most important single test point for evaluation is WWO, even where throttle pressure has to be reduced to enable this test. Comparison with WWO heat balance values can be made directly, but one should be aware that manufacturers expect some margin of HP efficiency with the heat balance values -- typically 2.5% when the unit is in near-perfect condition. Where: lighter load tests are available, even more can be deduced. The performance (both design and test) is easier to diagnose if plotted ayainst percent valve position, but is also effective if plotted versus throttle flow. Plotting against. Pomi/Penrottie OF Pist/Pthrottie often makes diagnosis difficult because both Fox an Pig deviate from design due to reheat nozzle changes, low reheater AP, and excessive leakage fran the 1st stage shell area. Figure 7-7 shows a hypothetical plot of HP section efficiency versus control valve position (only the valve point data is shown). TgSection Efficiency ‘Shape Comparison Design ‘Measured HP Turbine Efficiency (Valve Points) ° 50 100, % Control Valve Position Figure 7-7. BP Section Efficiency Note that the difference between design and test is greater at light load than at WO. This is generally a sign that the first stage performance has deteriorated more than the later stages. ‘This would not be true, of course, for a unit with throttling valve control or full-are aimission. Where the opposite occurs (less deterioration at light load), it usually indicates ‘the latter stages are more affected than the first stage.‘the turbine cross section should be carefully examined to see if significant leak~ ages are directed to the HP exhaust. or example, separate single flow HP sec tions usually have a major packing that allows leakage stean to mix with the cold reheat flow. Poor section efficiency and loy ist stage shell pressure may be the result of excessive leakage through that packing. Another example would be leaky main stean piston rings (or bell seals). The turbine cross section will indicate whether such leakage will go to the cold rebeat, hot reheat, or crossover areas. Such leakages (at throttle enthalpy) will vary across the load range depending on which rings leak and what valves are open. This can have shape effects on HP section efficiency. Note the actual example in Figure 7-8 which is believed caused by such leakage. We TURBINE EFFICIENCY 1985 ASME TESTS. Heat Belonce 80 E 10) ae 5 P< nmaipy OropEtticiener i osama 5 so] 7} ceamainy orp eticieney~ ‘Threttie to HARP 030 a4 080080 G70 080 080 100 Throttle Flow Retio 721‘The dotted curve, enthalpy drop efficiency from throttle to the HARP (heater above the reheat point), shows the most proper shape. It would be unaffected by piston ring leakage that has probably affected the measured entire section enthalpy drop efficiency. In addition, since its end point is based on extraction enthalpy, the deduced efficiency should be lower than that of the entire HP section. Since it is higher (at WO conditions), one should further suspect that hot leakage stean into the cold reheat area is affecting both the level and shape of the measured overall section efficiency. The HP section kw losses should be detemnined at WWO as illustrated in the following equations: encottie ~ Hora) + Wenrottn oe ae on % HP Loss x Kw Output Design Kw Loss = (7-8) 100 Be sure to correct HP losses to a percentag: Efficiencypes ~ Efficiencyne, % HP Loss = = a (7-9) BEficiencypes, Small flow corrections can be made for flows that pass through the Ist stage then leak into the reheat section; also for HARP flow that does not pass through the last few HP stages. Note that the Kw losses calculated here do not inclule the effect of excess throttle flow. A further correction will be made later for the Kw check analysis. 7-22In general the HP section performance will deteriorate much more than the other sections. Usually the greatest source of such loss will be rubbed seals and excessive leakage, with solid particle erosion the second largest loss. In the case of Morgantown 2, Figure 7-9 below shows tested, heat balance and expected (heat balance plus margin) high pressure performance. Morgontown 2 HP Section Efficiency (pre-outage) Heat Bolance Plus Margia eet Balance-Vaive Points 0 10 2.0 +30 +40 5.0 Throttle Flow Million Ib/br At WWO, the test efficiency is 7.5 points poorer than the heat balance and 10 points poorer than expected. At the Ist valve point, the tested efficiency is about 10 points poorer than heat balance and 12.5 points poorer than expected. ‘This suggests that more degradation is present at the first stage than the latter HP stages. This is consistent with the expected heavy erosion. 7-23‘The approximate expected HP output in kilowatts is: Des. Throttle Flow (Henr - Horh) 3412 3975497 (2424 - 1266.5) = 183,511 3412 HP loss (fram new condition): Efficiencyerp - Bfficiencytest, % Loss (at des. flow) Efficiencyery 0.875 - 0.75 x 183,511 = 20973 Kw 0.875, But note that the loss fran heat balance conditions for output comparison with tests is less by the anount of HP section margin. 0.85 - 0.775 x 183,511 = 16192 Kw 6.85 ‘Also, note that at the higher than heat balance test flow, the HP Kw losses will increase to: ¥ SOS x 16192 = 1.0704 x 16192 = 17331 Kw Yes step 6 re Test Versus Expected IP Section mthalj Bffici A-series of complications make diagnosis of most IP sections difficult. These include: 7-24Heat balances usually suggest poorer performance than is really expected. Figure 7-10, from the paper “A Method for Predicting the Performance of Stean Turbine-Generators... 16/500 KW and Larger," [1] can usually be belpful in determining what is a rea sonable expected level for units built since 1960. Note that Pigure 7-10 is for performance from the IP bowl to the IP exhaust. Subsequent corrections for valve and crossover pressure drops are needed to make direct comparison with test results. (0 Fiara_by caren er Prenore Ration Peecii nt Gono ol or oF 08 2 76 Volume Flow (1,000,000 ex 76/0) Figure 7-10. IP Section Efficiency N2_or IP dumy packing flows on ogposed flow turbines (combined HP and IP sections on a single shaft span) will cool the IP bowl. ‘These low enthalpy flows decrease the IP bowl enthalpy and cause an apparent increase in IP section ecthalpy drop efficiency. To the best degree possible, this effect must be eliminated so that the true IP efficiency can be detemmined. These methods will be discussed later. Many double flow IP sections have shaft cooling steam flows. ‘These, too, cause the IP enthalpy drop efficiency to appear higher than its actual performance level. where practical, the cooling flow rate ani enthalpy should be measured or estimated to help detemine the IP bowl enthalpy. ‘he crossover area may receive a variety of other flows, including: © excess flow from BFP turbine exhaust 7-25© leakage from dummy packing @ leakage fran bell seals, piston ring seals As above, these flows, with their enthalpy, need to be measured or estimated. ‘The most critical of the above complicetions is the N2 packing on GE opposed Flow units and the IP dumy flow on W units, ‘The N2 arrangement is somewhat easier to diagnose from test results. Excess N2 packing leakage will cause an apparent increase in IP section enthalpy drop efficiency as shown in Figure 7-11 below. This effect is even more notice Measured Enthalpy Drop Efficiency Probably Real Efficiency IP Section Efficiency % Output Figure 7-11. Measured IP Section Efficiency able at light load. Since real IP section efficiency is virtually constant across the load range, it is often possible, by trial and error, to detemine a nani tule of packing clearance that would explain the deviations of IP enthalpy drop efficiency fram the expected flat condition. 7-26‘The trials are based on multiples of heat balace N2 flow, with success being achieved when the corrected efficiency is constant actoss the load range. the method is marginally successful at best. One problen area is the N2 enthalpy, which may not be as shom on the heat balances and will deviate differently at nigh and low load conditions. he paper "Estimating the Leskage fron HP to IP Turbine Sections" by John A. Booth and David E. Kautamann (2} suggests three methods for estimating flows from HP. to IP sections. These are well presented ani worth consideration with the following comments: © The "Blowdown Systen" (System A) is not recomended due to its several disadvantag = Creates excessive stress on packing teeth. = Dumps more steam to the condenser than the desuperheater may be designed to handle. ~ Indicates an enthalpy of the flow that would not be identical with that of normal N2 flow. = Figure 7-12 is taken fran the Booth and Kautmnann paper, with P, (blowown annglus pressure) added for clarity. When large, high velocity flows are being passed through the blowdown valve, the pressure drop from P, to P; can be jarge. This is @ zone with fairly small flow aréas and tight turns. It can be understood, then, that P, will be higher than'P;. ‘The assumed condition of Q, (flow fram the IP bowl to th? blowdow annulus) may not exist even though P3 is lower than P2. © The “more elaborate” system acknowledged in the same section is being tested at Morgantown 2 and a report will be issued later on its results including a comparison with system (B). 7-27™ meri Figure 7-12. Blowdown system Morgantown 2 IP Section Tests. In the case of Morgantown 2, the measured (apparent) efficiency (intercept valve to crossover) is shown below on Figure 7-13. ‘The design values are based on Figure 13 of the reference 1 but corrected to intercept valve pressure. Tt was estimated that with the suspected level of nozzle erosion and seal wear, that the real IP efficiency would be no better than 87%. 7-28IP turbine Efficiency 7 | ese a | 1 : | i } 5 | Actual i e= g 8 8 g g 8 Hot Reheat Press. PSiA Figure 7-13. IP turbine Bfficiency Based on the large difference between test efficiency and assumed actual efficiency, estimates were male for the possible combination of excessive N2 leak~ age and mainsteam piston ring leakage that could cause such a great impact. Leakage rates up to 500,000 Ib/hr were considered. Such flows could explain both the low Ist stage shell pressure and the high IP section efficiency, but were felt to be unreasonably great. But recomendations were made to be prepared for (and inspect) all main steam piston rings, to replace all N2 packing,and to add an W2 blowiown bypass systen capable of measuring the effective N2 clearance during operation. 7-29Like the HP section, the design IP section output can be determined from the heat balance: Rebeat Flow (Hiy - Hyo) Kvoutpat ‘3412 where flow can be roughly established by averaging the flow entering the IP sec~ ‘tion and the flow leaving. In the case of Morgantown 2 IP section %w output = 161850 Kw, and The Fw losses in ‘the IP section at WO flow are: Bf ficiency, -BE ficiency, rarest design outpat BEficiency pyoected 0.06 —— x 161850 = 7100 Kw 0.91 (At test flow, the loss is increased to 7500 Hy.) Step 7: WMO Kilowatt Check Knowing test flow, test losses in the HP and IP sections and test kilowatts, it is appropriate to compare output to heat balance values. One should recognize that there are a lot of unknowns in this process. LP section performance is unknown. N2 or dummy flows at best have been estimated. Testing inaccuracies may be great. Still, the check can be extremely valuable. Note that proprietary analytical programs such as PEPSI can be used in similar fashion, 7-30Tt is desirable to correct the test output for unusual back pressure, rebeat spray, reheater AP, serious feedwater heater deviation from design conditions and reheat temperature. It is usually bast not to make corrections for initial pressure and temperature. If throttle flow is St higher than the heat balance, one would naturally expect about St greater output. However keep in mind that greater flow usually increases RH pressure, thus decreasing the HP available energy (AE) and used energy. For Large flow deviations it is desirable to correct for this effect using test CRH Pressure to compare AByesy With AEpog+ Often the CRH pressure will be low, due to sualler-than-design reheater AP and HRH nozzle erosion, ‘This will decrease the correction for reduced HP section AE caused by excessive flow. Note that the combined effect of excess flow ani low reheater AP can be simultaneously considered by comparing test CRH pressure ani AE with the heat balance. ‘he themal kit correction for RHTR AP would then be unnecessary. An example of a simple Kw check procedure (not Morgantown 2) is shown below: Effect on Output _(m) HP Turbine Throttle flow: down from HB by 4.7% 5.0 Low HP efficiency and excessive N2 Leakage -15.1 IP Turbine IP flow (including sprays): up by 4.5% 412.5 Low IP efficiency 4.2 7-31Other KW Effects § effect Low IP bowl enthalpy (mixing) ~=0.25 Low HRH Temp. 0.25 High B.P. 4.75 BEPT Excess Flow ad Total change in output Compare with heat balance and test ‘The above suggests that the losses have been approximately pinpointed by the tests and analyses. Some unknown or LP losses can be expected. However, emphasis should be placed on the necessity to confirm that the local losses are as Heat Balance - Test Output 42.2. predicted when the turbine is open. This is to provide assurance that no nificant cause of loss is overlooked. Note that if large unexplained losses had occurred, placed in a mmber of areas: Incorrect flow measurement - Consider questions raised in previous section regarding w//P7¥ analyses Poor LP section performance - Do past reports indicate deposits, damage or serious rubbing has been previously observed? Can deposits be checked on the L-0 nozzle and bucket during a weekend shutdown? Is there reason to suspect that the LP turbine is really Poorer than is expected by the heat balance? LP turbines usually are relatively free fran losses caused by solid particle erosion and foreign material damage. Even excessive rub~ bing does not normally cause serious losses. whenever an analysis indicates the probability of nore than 2% LP losses, suspicion should be raised that sanething else may really be wrong. Serious N2 or IP dumy packing leakages ~ Does the IP section efficiency seen artificially high and/or increase as load is decreased? Do past reports show heavy rubs to be a problen? Unit is not isolated - flow is bypassing. 7-32 suspicion would have beenKilowatt Check ~ Pre-out Morgantown 2 5% Overpressure WWO Heat Balance Kw 625,496 Measured main stean test flow is higher ‘than the WHO 5% OP HB by 2.28 Increased HP output due to 2.2% extra flow = + 47070 Increased HP output due to low CRH pressure +13,007 (combined effect of low RHTR P and eroded HRH nozzle) HRH flow is higher than HB by 7.8% Tncreased RH section output = 434,360 Loss due to 8-8% poor HP section 17,331, Loss due to excess N2 flow (100,000 1b/hr) ~ 3,957 Loss due to poor IP section (down 48) ~ 7,654 Other cycle corrections considered negligible Expected Test Output 647,991, Actual Test Output 621,000 Unexplained Loss 26,991 (4.38) This unexplained loss can be assigned to a combination of: © test flow exaggeration © cycle flow bypasses (turbine flow bypass, piston ring leakage, etc.) © deteriorated LP section efficiency © Greater N2 flow than assumed Te is becaming apparent that there must be a combination of flow error and a cycle bypass affecting the diagnosis and output of this unit. These will be discussed later. Step 8: Other Losses Experience has shown that a large variety of leakages can impact strongly on the real output and efficiency of the unit. Most of the time these leakages do not reveal themselves in poor section efficiency. Where such leakages exist, the analytical methods described above would generally indicate that there must be 7-33flow errors or that the LP section must have poor efficiency. An excellent reference on this subject is "The Best Buy in Heat Rate Recovery - Turbine Cycle Isolation Maintenance" by W.H. Hopson, J.C. Peyton and J.K, Legg of Southern Company Services [3]. Morgantown Examination of the previous sections shows a consistent pattern: Ast stage pressure is unusuaily low for the measured throttle flow. Explaining this lomess by normal methods (high N2 flow; high 2nd stage diaphragm packing flow) requires such extrane clearances that. they seem unbelievable. W/P7T analysis showed consistently high values that suggest the deduced flow must be exaggerated. ‘The kilowatt check shows unexplained losses that suggest the deduced section flows must be exaggerated. While no measurements existed to confirm the magnitude of such flows, it was con cluded that two bypass conditions existed that had reasonable probability: A start-up bypass around the turbine was known to be hot when it. should have been cold. A level of 100,000 1b/hr was assimed (2.48). ‘Tre mainsteam turbine inlet piston rings were believed the mst Likely candidate for 2 second leak, 117/000 Ib/hr was assumed €o bypass the high pressure stages and re-enter at the cold seheat point. ‘These two potential leaks, if true, would cause over 21,000 Kw of the 27,000 kw unexplained loss. 7034Step 9: Determine the Probable Causes of Section Losses ‘The previous sections generally identify efficiency levels, magnitudes of loss, probable changes in local flow areas. ‘The next section deals with the probable causes of such effects. ‘To use the techniques described below, it is desirable to plot parameters that deviate significantly fran design as a function of time. This will allow deter~ mination of whether changes have happened suddenly or slowly. Four major areas of common trouble can be expected to afflict large steam turbines including: © Excessive Leakage © Erosion (solid particle) © Internal Damage © Deposits Bach of these potential troubles has characteristics that will enable the diagnos- tician to digferentiate using the available monitoring results. Differentiating Characteristics. Table 7-3 is a summary of the symptoms normally present for each of the camon problens. This brief listing will help diagmse turbine internal conditions. ‘The losses fran the above causes are highly variable. A range of typical conditions found on inspection are shown below: 735‘Table 7-3 DIAGNOSTIC CHART OF LOSS CHARACTERISTICS ‘ubbing Oenege on Spfllstripe and Packing ode of Appearance: Hoppers sutenly - nove Likely on a first startup. oat effects: Increases flov capacity (this effect highest in HP action Decreases section efficiency (worst on ow velune flow stayes)« ‘Side Bffecta: Wrsens flov texperature sepregati Nomally has Little effect on thrust. Shage Effects: Ratio of § sstficiency/® slow usally ‘greater than 1. (Abeolute values) Solid marticle Erosion ede of iepearance: Ossally appears gradually. ose etects: Increases flow capacity. Decreases efficiency. Worst effects usually at turbioe inlets; ie first stage, erosion magaitale may be Soest at the inlet fod by the first valve. Changed thrusts changed /®77 Giseribution; changed Ear ‘istration. ‘Shape Betects: JER ettects may be greatest at Ligne oad. ‘Efficiency’ loss compared co guarantee ‘aay be geeatest at Light load; thrust foorense my be in the sme direction as flor, Special Dangers: Overloaded buckets; weakened tenons. 7-36Table 7-3, continued Mode of repearance: ‘Local fects: Decreased efficiency; decreased flow capecity. ‘Side Refecta: Changes theast; changed VET Gistribution. ‘Shage Effects: Section efficiency aay decrease 3-4 ‘ines. as mich a0 flow capocity. ‘Taruse changes may be opposite the Bitection of flow Broeasive trate. Internal Oamag Node of topearance: Ossally abrupt may have subsequent symptans. Local Befects Decreased efficiency; decreased flow capectey. Side effects: Increased vibration; changed /E77 Aisteiiuetons; changed thrust. ‘Shape Bétects: Wo consistent pattern. ‘Special bangers: | Weakened or Inosoned mechanical ewtures. 7-37‘Typical Loss Magnitudes Solid Particle Internal Rubbed Erosion Deposits Damage Seals up 0-2 0-108 0-3 2-12 IP 0-28 O-5 0-2 1-48 we 0. - 0.58 0-3 0-18 0-38 Excessive Leakage Caused by Shaft Rubbing (Figure 7-14). Supporting Infomation: @ Was there a vibration problem at time x? © Was there a severe thermal problem at time X (cold start-up, water induction, load swing, temperature control problem, etc.)? © Has VB/V or P at the first stage shell decreased for a given throttle flow? Brosion - Solid Particle (Figure 7-15). This subject focuses on the damaje done by oxide material which spalls off hot boiler ani pipe surfaces, then, carried by steam, cuts and scratches away turbine material in high velocity zones. the im mediate effects are to increase flow areas and decrease efficiency. This is in contrast to the closing effects caused by larger, more dense particles such as weld beads, as discussed later. The above exemple indicates that it is generally the admission stage (ie., the first high pressure or first reheat stage) where erosion effects are most severe. The danage declines at subsequent stages as the particles becne finer vhile passing through the turbine and as they are dravn off at extraction points. sometimes tthe danaging capability of the fine oxide is revived at the crossover so that the first LP stage will be strongly attacked even though the last IP stage is unaffected. 7-38Excessive Leakage Caused by Shaft Rubbing Measured Opposed Flow Type Units Design Measured Separate IP Sections Measured IP Section-Enthalpy Drop Efficiency HP Section-Enthalpy Drop Ef! x Time— x Time — Figure 7-14. Shaft Rubbing viaterial differences between one stage and another (epilistrips, packings and nozzle mterials, aleo physical size of nozzles), may also cause an unexpected 1o- cal change in erosion danaje. Softer materials or snaller profiles are more vol- nerable to erosion damage. Supporting information: a past history of erosion. 7-39An Example of Solid Particle Erosion 2nd Extraction Design Measured Hot Reheat -Measured wh/P/v IP Efficiency Note that both erosion and rubbing tend to increase flow areas, The key elements for differentiating which problem exists are (see Table 7-3): © Brosion occurs relatively gradually -- rubbing is sudden. Erosion can cause large effects in the IP section, P/V functions —~ rubbing cannot. © Leakage can sometimes cause an apparent increase in IP efficiency (for units with HP and IB sections combined in one outer shell). Internal turbine damage is camnonly Internal Damage (not erosion type damage) . caused by: © Steam-carried hard particles from the boiler. © Turbine components which have broken or vibrated loose. 7-40© Foreign material left in the turbine or boiler. In most cases the result will be reduced flow area and a loss of capacity. ‘The hypothetical example in Figure 7-16 shows a probable damage situation at tine X. Both efficiency and flow areas have been decreased by a sudden incident. = & Design a § Eg Measured 2 = 8 — X time Figure 7-16, Internal Danage In addition to the pictured effects on efficiency and flow area, internal dmage can be accompanied ty vibration (tamporary or permanent) and by a change in thrust. Not infrequently an incident will have Secondary perturbations. The pos— sibility of subsequent reliability loss is very difficult to appraise in the case of internal damage, but important to consider, especially where a vibration change indicates the rotating structure has change. 7-41Evidence of heavy erosion damage is occasionally a precurser to potential internal damage caused by tenon erosion and subsequent loss of bucket covers. Deposits. Chemical deposits (oxides, sulfates, carbonates, silica, ete.) can heavily impact performance and capacity. In recent years, improved control of feedwater chemistry has generally decreased the magnituie of such losses. still, losses of several percent for groups of stages is occasionally found. this is especially true where large anounts of reheat desuperheating flow are used, since the desuperheating water bypasses the boiler drum, carrying over more contaminants. Extreme cases of copper oxide deposits in high pressure turbines have soverely Limited flow capacity and increased thrust to a condition of forcing a shutdown. Deposits usually accumulate in the stean path in areas of pressure drop where Steam conditions have been lowered to the point vhere the chanical saturation point is no longer exceeded. For impulse stages the deposition will be primarily in the stationary blades, on the suction surface, in the vicinity of the throat. Im reaction stages, the deposit will accmulate in both rotating and stationary blades. Figure 7-17 shows a hypothetical deposition case where both efficiency ani £low areas have been reduced with time. ‘Thrust may change, however vibration is usually not affected. Note that in double flow sections deposition may occur equally in both directions with no net change in thrust. 7-42Section Efficiency Time Time Figure 7-17. Steam Path Deposits Sone deposits are sensitive to temperature and may spall off during a shutdown or ‘a large load swing. Some are also soluable in water and have been washed from the steam path by steam that includes water to either dissolve or break loose the chemical layer by impact. Water washing does aot normally provide a lasting value and includes some risk. Deposit presence can occasionally be confirmed by temporary improvements in efficiency and a rise in hotwell conductivity during transients or after a shutdom. 7-43Sumary: Pre-outage Predictions for Morgantown 2 At Morgantown 2, prior to the 1984 outage, historical test data were not available to permit xigorous application of the diagnostic chart techniques described above. This was not a severe handicap in this case since the presence of both erosion and seal wear could easily be determined fran the available test data an previous ‘test reports. Future operation will include the cepebility to plot, versus tine, ‘the critical functions of pressure, flow, W//EVT, #P efficiency, IP efficiency and @ variety of other turbine perfornance-related values. The following predictions were made: © Severe Ist stage erosion on the Ist valve arc nozzle (37% area increase) © 2nd stage nozzle erosion up to 25% area increase © Main steam piston ring leakage (estimated over 100,000 lb/hr) © Excessive N2 leakage of 100,000 to 300,000 1b/nr © Excessive clearance in the HP and IP packings and tip seals © Severe erosion on the 1st reheat and Ist LP nozzles © Moderate erosion in other IP nozzles © Leakage flow (estimated) of 100,000 1b/hr from startup blowiom Line to the condenser Beyond the nine-step diagnosis program described in the preceding pages, two other calculating methods are occasionally useful. these are: © Partial recovery of losses by following stages © Determination of effects on heat rate due to changes in KW output. Both methods are described in Appendix 7D. 7-44REPERINCES: [1] R.C. Spencer, K.C. Cotton, C.N. Cannon, "A Method for Predicting the Performance Of Steam ‘Turbine-Generators ... 16,500 1 and Larger," General Electric Company GER-2007C, July 1974, based on ASME paper 62-WA-209, ASME, 345 E. 47th Street, New York, NY 10017. [2] 3.A. Booth, D-E. Kautznan, "Estimating the Leakage form HP to IP Turbine Sections," Electric Power Research Institute, Power Plant Performance Monitoring Conference/orkshop, October 1984. [3] W.H. Hopson, J.C. Peyton, J.K. Legg, "the Best Buy in Heat Rate Recovery ~ ‘Turbine Cycle Isolation Maintenance," Electric Power Research Institute, Heat Rate Improvement Workshop, October 1985. 7-45‘Appendix 7A STEAM PATH APPRAISALWith the turbine open for inspection, the high pressure and intermediate pressure sections were carefully exanined for sources of loss. Table 7A-1 below summarizes the results of the inspection. Table 7A-1 ‘TURBINE LOSSES Loss From Recoverable HP Losses New Losses Tip Leakage 4335, 4335 Packing Leakage 3860 3560 lst Stage Nozzle Erosion 3790 3790 ‘Steam Path Roughness 1970 1670 N2 Packing Excess Leakage 3060* 3060" Total 17015 16415 Effect on Enthalpy Drop Bffic. 8.08** 1.7 *Does not effect enthalpy drop efficiency. *#1£ 117000 lb/hr piston ring leakage is included the known losses increase by 5500 kw and the efficiency loss by 2.6 points. IP Losses Tip Leakage u3 3 Packing Leakage 600 500 ‘th Stage Nozzle Erosion 2300 2300 Steam Path Roughness 560 500 ‘Total on IP Section ee ee Effect on IP efficiency 3.08 3.08 LP Losses 875 875, cycle Losses N2 Leakage Effect on RH Section 2800 2300 ‘Total Unit 3525814. 58) 24058 ( Effect on Heat Rate 332 (3.78) TlDISCUSSION OF RESULTS Excessive leakage due to seal rubs. Average packing clearance in the HP-IP sections were approximately 70 mils (tip seals were about 50% greater). Probable excess N2 leakage flow was calculated at 110,000 lb/nr. Added to. the normal flow (52,000 lb/hr) this totals 162,000 lb/hr. ‘This is high, but significantly less than the worst range estimated from the pre-outage test results. Excess 2nd stage packing leakage was 150,000 1b/he. Leakage due to piston rings ‘The number 1 and 2 inlet pips piston rings were found stretched and eroded. ‘The leakage area could not be accurately estimated, but the over 100,000 Ib/ar assumed prior to the outage seamed reasonable. Solid particle erosion on nozzles Approximate Ist stage Increase 1st valve port 328 2nd valve port 258 3rd valve port crs 4th valve port 08 2nd Stage 38 8th stage (Ist reheat) 498 HP Section BEficiency When corrected for probable piston ring leakage losses, the test oss in this section appeared accurately explained. IP Section Efficiency Only 38 of the expected 4% IP loss was found by the appraisal. This has a good effect of reducing the amount of N2 leakage flow needed to explain the high apparent IP efficiency of the test. ‘Ist State Shell Pressure ‘The combination of excess N2 and 2nd stage diaphragm packing leak— age would explain about 6% lowness of the Ist stage shell pressure. Estimated piston ring leakage and turbine bypass leakage would ex— plain another 5%. Finally, the low throttle enthalpy (about 30 Btu) would explain 4.2% more, thus providing an apparently satis— factory understanding of the low Ist stage shell pressure. Ta-2CONCLUSIONS FROM INSPECTION ‘The pre-outage diagnosis ani the inspection were in strong agreement. Only in the area of the N2 flow was there significant deviation fram the predicted levels, and even it was within the range predicted. a3CLOSING STEAM PATH APPRAISAL| Significant danage (seal rubbing, nozzle erosion, pockmarking, etc.) had been found when the unit was opened. It was deaned desirable to evaluate whether the extensive repair procedures had consistently achieved the quality assumed for the predicted recovery of los performance. RESULTS OF CLOSING APPRAISAL With the exception of itens discussed below, the repairs had been skillfully completed: © Reconstruction of eroded and danaged nozzle profiles did not fully provide the intended quality. = some discharge edges were excessively thick - some profiles were suspected of significant deviation fram design. = some surface finishes were poorer than desirable © ‘Time did not permit improvement to the profile or edge thickness problems. The surface roughness condition was greatly improved by PEPCO blade specialists.Appendix 7¢ POST OUTAGE PERFORMANCE MEASUREMENTS AND ANALYSISqhese results represent a basis for determining the success of the effort to establish a high Level of performance for the turbine cycle. he sane proce- Gure used at the pre-outage analysis was used. Note that these data will also be used for benchnark conditions from which deviations with time will be recognized and followed. STEP 1: FLOW CAPACITY ‘VHO Design Flow = 3,964,071 With normal 2.5% excess 4,063,173, Test WO flow (corrected) = 4,242,966 This is 4.428 night the following questions nay be raised: © Has nozzle erosion alzeady occurred? © Is first stage shell pressure low? © Is there still some flow bypass? STEP THROTTLE W//P7V ACROSS THE LOAD RANGS est data were not available at the first valve point (#2 valve crack point) and heat balance information near the #3 valve crack point was not believed correct for the valve point condition. Figure 7C-l shows the test data and design information. Only the WO point is believed fully trustworthy. Using the recam mended relationship at vwO: eee Mest * Mies 4282966 Agesign Wades ¥ Mest 0-9 4063173 = 1.084Morgantown Unit 2 : May 1988" ‘Thrette Conditions Des WANPTV o Test W/dB7V 35000 25000 20000 wi PT 15000 10000 eee a ee Main Steom Flow ~ 1000 LBN/HR 4500, Figure 7C-1. Throttle Conditions ‘The 4.4% extra area, if confined to the 1st arc would suggest it has increased by about 188. It is surprising that signs of heavy erosion appeared so quickly. ‘WO FLOW/IS? STAGE PRESSURE ANALYSIS Figure 7C-2 shows the pressure/flow results. Significant improvement has occurred. Pre-inspection tests showed this pressure to be low by 17%; low by 6.78. row it is But even 6.7% lomess is a source of concern, Several questions need consideration:

EPRI Current Performance Fossil PP Monitoring Vol 06 Turbine Cycle PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EPRI Current Performance Fossil PP Monitoring Vol 06 Turbine Cycle PDF

Uploaded by

Copyright:

Available Formats

You might also like