PDF The Reciprocating Pump Theory Design and Use John e Miller 2nd Editio DL

THE RECIPR CATIN PU P
THEORY, DESIGN, ANO USE
Second Edition
John E. Mmer
White Rock Engineering, Inc.
Dallas, Texas
KRIEGER PUBUSHING COMPANY

MALABAR, FLORIDA
1995
Second Edition 1995
Printed and Published by

KRIEGER PUBUSHING COMPANY
KRIEGER ORIVE
MALABAR, FLORIDA 32950
Copyright © 1995 by Krieger Publishing Company
Ali rights reserved. No par! of this book may be reproduced in any form or by any means, electronic or mechanical,
including informa!ion storage and retrieval systems without permission in writing from the publisher.
No liability is assumed with respecl lo the use of the iriformation contained herein.
Printed in the United States of America.
FROM A DECLARATION OF PRINC!PLES JOINTLY ADOPTED BY ,A COMMITTEE OF THE AMERI-

CAN BAR ASSOCIATION ANDA COMMITTEE OF PUBLISHERS:
This publication is designed to provide accurate and authoritalive information in regard to the subject matter
covered. lt is sold with the understanding that !he publisher is not engaged in rendering legal, accounting, or
other professional service. lf legal advice or other expert assistance is required, the services of a competen!
professional person should be sought
Library of Congress Cataloging-in-Pubiication Dala
Miller, John E. (John Evans), 1909-

The reciprocating pump : theory, design, and use / John E.
Miller.-2nd ed.
p. cm.
lncludes index.
ISBN 0-89464-599-4 (allc paper)
L Reciprocating pumps. L Titie.
TJ901.M55 1995
62L6'5-dc20 94-46260
CIP
!098765432
PRE CE T
SEC NO EDITI N
In the four years following the First Edition, 'The Reciprocating Pump; Theory
Design and Use' it became evident that the confüsion between the Centrifuga!
Pump and the Reciprocating Pump had greatly increased because of the increased
interest in reciprocating pumps for highly technical applications and high to ex-
tremely high discharge pressures (l0,000 to !00,000 PSI) for such services as
water purification by reverse osmosis, cutting and cleaning materials, food pro-
cessing, etc.
High discharge pressure requirements prompted a discussion of high-pressure
pump design and the method of Autofrettage of pump liquid-ends in order to
reduce corrosion-fatigue failures in high pressure use.
Further experience with the use of multiple pump applications and how to
these for mínimum cost is included.
Additional experience in the complicated effects of suction requirements has
led to a completely new approach to the problems.
This Edition is dedicated to the memory of my loving Margaret, who
away on April 16, 1991, after 56 years of companionship of the kind that
gave me the encouragement to complete this book and who, with the ravages of
Alzheimers, sat for hours with an autographed copy of the First Edition clasped
to her breast.
JOHN E. MILLER
PREFACE
The intent of this book is to bring together rnost of the aspects of reciprocating
purnps, keeping in rnind the requirernents of designers, rnanufacturers, and users.
For rnany years there has been sorne confusion in the rnatter of the effects of
liquid dynarnics (ftow variation and acceleration) on the performance of recipro-
cating purnps. One possible reason is the great difference between reciprocating
and centrifuga! purnps; those dealing with the two types are usually not confronted
with the sarne types of disturbances. Another reason is the neglect that recipro-
cating purnp theory has experienced in the midst of an increase in the problerns
resulting frorn high-speed operation as the result of rnanufacturers' frequent speed
upratings applied over the years dueto the pressure of cornpetition and econornics.
Many subjects in this book are covered by rneans of discussion, allowing the
reader to better understand the cause and effect. In rnany cases, examples of cal-
culations and derivations are given to support the explanation.
Aside frorn the casual interest associated with the past history of purnps, such
history is sornetirnes used to draw attention to discarded ideas so that reinvention
of the wheel will be avoided. (And they rnay stirnulate new ideas.)
A rather new and wide application of reciprocating purnps is in the transporta-
tion of solids in the form of a liquid-rnixed slurry. This subject is covered in sorne
detail.
In order to provide a cornprehensive encyclopedia of reciprocating purnps,
Chapters 13 and 14 contain many useful tables, charts, and conversions.
After expressing appreciation to Oilwell Division of U.S. Steel Corp. for their
sharing of knowledge to the industry, the author wishes to thank several experts
and institutions in the field of purnping who have rnade valuable contributions,
acknowledged throughout this text. Special thanks to Mike Rizzone, Bob Crane,
•• ¡¡
Viii PREFACE
and my son Jim Miller, professional engineers, and my son Ben Miller. All are
well experienced in the field of reciprocating pumps and have reviewed the man-
uscript and offered many worthy suggestions. My wife, Margaret, supplied the
encouragement.
JOHN E. MILLER
Dallas, Texas
e NTENTS
1. PUMP TYPES
l.l Definitions I l
l .2 Nomenclature and Definitions I 13
1.3 Double-Acting vs. Single-Acting Pumps / 23
l .4 Plunger vs. Piston Pumps I 33
1.5 Interna!. Gears / 34
2. DYNAMICS 36
2.1 Introduction I 36
2.2 Standard Definitions / 36
2.3 Flow Variation I 37
2.4 Acceleration / 38
2.5 Derivation of Acceleration Pressure / 40
2.6 Derivation of Acceleration Pressure at Pump Inlet / 41
2.7 Critica! Suction Conditions I 42
2.8 Discharge Acceleration I 42
2. 9 Hydraulic Flow and Pressure Waveforms / 44
2.10 Discussion of Ideal Waveforms I 44
2.11 Mean or Average Flow Rate I 46
2.12 Pump Displacement / 47
2.13 Graphics of Pump Performance / 51
SUCTION REQUIREMENTS 55
3.1. l Introduction / 55
3.1.2 Analysis of Suction Requirements / 57
ix
X CONTENTS
3.1.3 More on Suction Requirements / 58

3.1.4 Suction Requirement Factors / 63
3.1.5 Valve Leakage and Slip / 65
3.1.6 Typical Suction Pressure Waveforms / 67
3. l. 7 Relation of Pressure Waves to Pump Cycle / 67
3. 2 Performance Curves I 69
3.2. l Method of Recording Actual Mechanical End-of-Stroke / 69
3.2.2 Performance-POSIVA and RPM / 69
3.2.3 Performance-Reduced Plunger Diameter / 72
3.2.4 Volumetric Efficiency Statistics / 72
3.2.5 Short vs. Long Suction Line / 73
3.2.6 Acceleration / 74
3.2. 7 Computer Calculations / 75
3.2.8 Suction Stabilizer / 75
3.3 Testing Pumps for Suction Requirements / 80
3.3.1 Introduction / 80
3.3.2 TCP Required, (TCPR) ! 81
3.3.3 Procedure for Testing ! 82
3.3.4 The Myth of NPSH ! 82
3.5 Air and Gas Saturation of Liquids / 86
3.6 Dissolved Gas at Pump Inlet / 87
3. 7 Notes Regarding TCP ! 88
3. 8 Methods of Increasing TCP / 88
3.9 Cavitation / 89
3.10 Suction Systems / 89
4. PULSATION ANO SURGE CONTROL 95

4.1 Pulsation Types / 95
4.1.1 Type DV - Discharge-Velocity / 95
4.1.2 Type DO - Discharge-Valve Open / 100
4.1.3 Type DR101 - Discharge-Acceleration (P,ac) / 101
4.1.4 Type SV - Suction Velocity / 101
4.1.5 Type SA - Suction Acceleration / 101
4.1.6 Type SD - Suction Valve Opening (P,v0 ) I 101
4.1. 7 Type VA - Vertical Acceleration / 10 l
4.1.8 Type PS - Pulsation-Surge ! 101
4.1.9 (WH)-Water Hammer ! 102
4.2 Dampener Types / 103
4.2.1 Type G / 103
4.2.2 Type GCAE / 103
4.2.3 Type GUAE ! 104
4.2.4 Type GCDE ! 104
4.2.5 Type LNFA ! 104
CONTENTS xi
4.2.6 Type GFFE I l05

4.2.7 Type DCFR I 105
4.3 Reporting Degree of Pulsation / l 08
4.4 Discharge Dampening / 109
4.5 Gas-Type Dampener Sizing I l 11
4.6 Gas-Type Dampener Sizing Equations I 112
4.7 Derivation of K ! 114
4.8 PASAFE Pulsation Control Sizing / ll4
4.9 Multiple Dampeners I 117
4.lO Precharging Gas-Type Dampeners I l l 8
4.11 Vaporization or "Heated" Dampeners / 118
4.12 Manifolded Multiple Pumps (Phasing) I l l 9
4.13 The Suction Stabilizer / 120
4.14 The Discharge System I 120
4.15 Dampener Facts / 121
4.16 Dampener Performance í 121
4.17 Acoustic Filters / 123
4.18 The Suction Stabilizer / 125
4.19 Surge Control in Water Systems I 127
4.19.l Methods of Surge Control in Water Systems / 131
4.19.2 Sizing of Surge Suppressors / l 35
4.19.3 Conclusion I 137
4.20 Sample Problem I 138
5. PUMP DESIGN 140

5.1 Rating Standards I 140
5.2 Family Planning I 142
5.3 Windows of Non-utilization I 143
5. 4 Piston Rod Load i 144
5.4. l Unusual Pressure in the Pump Cylinder / 145
5.5 Maximum Piston Diameter i 146
5.6 Stroke Length / 146
5.7 NumberofCylinders I 147
5.8 Speed I 147
5.9 Speed of Duplex and Similar Pumps 149
5.10 Speed of Multicylinder Pumps I 149
5. 11 Direction of Rotation I 150
5.11. l Rotation "Overrunning" / 150
5.11.2 Rofation "Underrunning" I 152
5.12 Offset Crankshaft I 152
5.13 Connecting Rod Forces / 153
5.13.1 Introduction I 153
5. 13. 2 Equations / 154
5.14 Crankshafl Bending Moments-Calculations I 159
XII CONTENTS
5.15 Crankshaft Bending Stress-Calculations I 161

5.16 Crossheads I 162
5.17 Bearings / 163
5.17.1 Roller Bearing B-10 Life I 164
5.17.2 Roller Bearing Average Life / 164
5.17.3 Journal Bearings / 165
5.18 Lubrication / 166
5.18.1 Pressure Lubrication / 166
5.18.2 Oil Cleaning I 167
5.18.3 Oil Temperature / 167
5.19 Volumetric Efficiency (Eq. 2.26) / 167
5.20 Mechanical Efficiency / 168
5.21 Intermittent Service I 169
5.22 Continuous Service / 170
5.23 Reciprocating Pumps with High Suction Pressure / 171
6. LIQUID ENDS 172

6.1 Typical Liquid-end Manifold Configurations / 172
6.2 S-N Curve I 172
6.2.1 The Goodman Diagram / 174
6.3 Stress I 176
6.4 Stress Reduction Methods I 177
6.4.1 Strain Bolts / 177
6.4.2 Autofrettage Procedure for Liquid Ends / 178
6.4.3 Shot Peening I 186
6.4.4 Coating and Plating / 186
6.5 Ultra-high Pressure Pump Design / 187
6.6 Abrasive Jet Cutting / 190
6.7 The Intensifier / 191
7. EXPENDABLE PARTS 193

7 .1 API Standards for Slush Pump Components / 193
7.2 Pump Liners / 206
7 .2.1 Liner Development / 207
7 .3 Pistons I 207
7.4 Plungers I 211
7.4.1 Metal-to-metal Plungers / 212
7.4.2 Ceramic Plunger Construction / 213
7.5 Packing / 215
7. 5 .1 Introduction and History / 215
7.5.2 Packing Types I 216
7.5.3 Split Packing Rings I 218
7.5.4 Spring Loaded Packing / 218
7.5.5 Segmenta! Metal Packing / 218
CONTENTS xiii
7.5.6 Multiple Material Assemblies / 219

7.5.7 Double Stack Height Seal Rings / 220
7.5.8 Spiral Packings / 220
7.5.9 FLAT-BAK Vee Ring I ·222
7.5.10 Knitted Wire Mesh / 222
7.5.11 Packing "Dos" / 223
7.5.12 Packing "Don'ts" / 224
7.6 0-Rings / 225
7.7 Controlled-Compression Gaskets / 225
7.8 Basic Elastomers / 226
7.9 Elastomers in Carbon Dioxide / 227
7.10 Stuffing Box Brass or Trim / 228
7.10.1 Stuffing Box Wear / 228
7.10.2 Stuffing Box Design Criteria / 230
7.10.3 Gland Tightening (Screwed Gland) / 231
7.10.4 Packing Lubrication / 231
7.10.5 Lubrication by Lantern Ring / 232
7 .11 Jacoby Leakage I ·232
7.12 Glossary of Terms / 234
7.13 Guidelines for Material Selection / 236
7.14 Titanium as a Pump Material / 239
8. VALVES 240
8.1 Valve-Type Classification / 240
8.2 Pump Valves / 241
8.3 Valve Seats / 245
8.4 Valve Springs / 248
8.5 Valve Spring Design / 251
8.6 The Effect of Valve Design on Suction Requirements / 254
8.6.1 Determining Valve Efficiency / 255
8.6.2 Valve Tests / 256
8.6.3 Valve Combinations / 256
8.6.4 Velocity Through Valves I 256
8.6.5 Valve Through Area / 259
8.6.6 Required Valve Lift / 260
8.6.7 The Effect of Valve Weight / 260
8.6.8 Effect of Spring Load I 262
8.6.9 Summary / 262
8. 7 Derivation of Val ve Velocity / 263
8.8 Unbalanced Valve Area I 264
8.9 Power-operated Valves / 265
8.10 Valves in Series / 266
8.11 Valves in Parallel / 266
8.12 Valve Flutter / 266
xiv CONTENTS
8.13 Steady-State Flow Through Valves / 267

8.14 Valve Delay / 268
9. SLURRY PUMPING 270

9.1 Slurry Properties / 270
9.2 Pumps for Slurry St>rvice / 271
9.2.2 Packing I 274
9.2.3 Plunger Flushing Methods / 276
9.2.4 Flushing Details / 278
9.2.5 Pistons I 280
9.2.6 Liners I 280
9.2.7 Piston Membrane Pumps / 280
9.2.8 Switch-Loop Pumping / 281
9.2.9 Other Methods / 281
9.3 Horizontal vs. Vertical Pumps for Slurry Service / 282
9.4 Suction Pressure for Slurry Puinps / 283
9.5 Coal Slurries / 284
9.5.1 Concentration of Solids l 284
9.5.2 Particle Size / 286
9.5.3 Concentration and Particle Size / 286
9.6 Valve Service for Slurry Pumping / 287
9.7 Slurry Erosion / 287
9.7.2 Slurry Wear Modes / 287
9.7.3 Effects of Wear / 289
9.7.4 Dry Abrasivity / 289
9.7.5 Conclusions / 289
9.8 Slurry Abrasion Testing / 293
9.8.1 Miller Number / 293
9.8.2 SAR Number (Slurry Abrasion Resistance) / 294
9.8.3 Test Equipment / 294
9.8.4 Procedure / 295
9.8.5 Calculation of Results / 295
9.8.6 Miller Number / 296
9.8.7 SAR Number / 296
9.8.8 A Miller Number System Overview I 297
9.8.9 Slurry Concentration, Particle Size, and Particle Shape / 299
9.8.10 Oil-Mixed Slurries / 300
9.8.11 Corrosion / 300
9.8.12 Corrosive Effect of Slurries / 301
9.9. The Gold Number for Low Abrasivity / 302
9.10 A Method for Locating a Plug in a Slurry Pipeline / 303
9.11 Black Mesa Pipeline / 304
CONTENTS XV
9 .12 Savage River Mines I 305

9 .13 Slurry Tables I 305
1 PARTS WEAR AND LIFE 316

10. i The Mechanics of Wear in Pumps I 316
10.Ll Introduction / 316
10. l.2 Wear Modes / 316
10.l.3 Effects of Wear I 320
10.1.4 Effect of Pump Stroke Reversa! Rate I 323
10.l.5 Piston Liner Clearance I 333
10.2 Phmgers I 333
10.3 Ceramic Plunger Pitting / 334
l0.3. l Causes I 335
10.3.2 Cures I 336
10.4 Theory of Ceramic Plunger Failure Mode I 336
10.5 Other Means of Reducing Wear Rate I 337
10.6 Slurry Abrasivity I 338
10.7 Examples of Parts Life in Hours I 338
10.8 Reducing DriHing Pump Parts Cost I 338
10.9 Danger of "Strainers" or "Filters" in the Suction Line I 340
10. lO Slurry Particle Size I 340
11. APPUCATIONS 342

1 Ll Effect of Liquid Compressibility I 342
11.Ll Pump Hook-up / 343
11.1.2 Charging or "Booster" Pump I 345
l l. 1.3 Pump Valve Unloading System I 345
11.l.4 Suction System Loops I 345
11.2 Hydraulics Institute Standards of Application I 347
ll.3 The Dynamics of Liquid Piping Systems I 367
l l.3.1 Where does Vibration Come From? I 368
11.3.2 How Does the Pulsation Couple into the Piping? 368
11.3.3 Cakulating Mechanical Natural Frequency / 369
11.3.4 Making the Application to Actual Piping / 37 l
11.3.5 What About Generalized Finite-Element Analysis? I 372
11.3.6 Design Philosophy / 373
11.4 High Suction Pressure / 375
11.5 Sizing Pumps for Pipelines / 376
11.5.l Pump Cost / 377
11.5.2 Expendable Parts Cost I 378
11.5.3 Methods of Cost Projection I 378
11.5.4 Standby Pumps I 379
11.5.5 Procedure for Pump Selection I 380
11.5.6 Stations I 381
xvi CONTENTS
11. 5. 7 Calculations I 383

11.5.8 Study Results I 384
11.6 Bolt Tightening Specifications I 384
12. INSTRUMENTATION 387

12.1 Pressure Measurement Methods I 387
12.2 Vibration Measurement I 389
12.3 Typical Waveforms I 389
12.4 Miscellaneous Waveforms (Oscillographs) I 390
12.5 Optical Phaser I 402
12.6 Positioning Strain Gages to Monitor Torsional Loads I 403
12.7 Damped Pressure Gauge I 405
12.8 Measuring Pressure Drop by Oscilloscope / 405
13. THEORY OF FLOW IN PIPE 408

13. l Introduction I 408
13.2 Physical Properties of Fluids / 409
13.3 Nature of Flow in Pipe-Laminar and Turbulent / 412
13.4 General Energy Equation-Bernoulli's Theorem / 414
13.5 Measurement of Pressure I 415
13.6 Darcy's Formula-General Equation for Flow of Fluids I 416
13.7 Flow Through Nozzles and Orífices I 419
13.8 Pressure Drop and Velocity in Piping Systems I 420
13.9 Pipe Line Flow Problems I 422
13.10 Flow Through Orífice Meters I 424
13.11 Reference Figures and Tables I 427
13.12 Summary of Formulas / 445
13.13 Nomenclature / 450
Appendix 1 Symbols and Nomenclature-Conversion / 451
Appendix 2 Subscripts I 459
Appendix 3 Abbreviations / 460
Index / 462
1
PUMPTYPES
DEFINITIONS
RECIPROCATING PUMP. A mechanical device used to impart a pulsating, dynamic
flow to a liquid and consisting of one or more single- or double-acting positive-
displacement elements (pistons or plungers). The elements in the liquid end are
driven in a more or less harmonic motion by a rotating crank and connecting rod
mechanism. The liquid flow generated by this reciprocating motion is directed
from the pump inlet (suction) to the pump outlet (discharge) by the selective op-
eration of self-acting check valves located at the inlet and outlet of each displace-
ment element.
HORIZONTAL PUMP. A pump in which the axial centerline ofthe cylinder, piston,
piston rod, and CTO!'\Shead is horizontal. See Figure 1.1.
VERTICAL PUMP. A pump in which the axial centerline of the cylinder, plunger,
extension rod, and crosshead is vertical. See Figures 1.2-1.3.
SIMPLEX SINGLE·DOUBLE-ACTING PUMP. This arrangement allows, in effect, a sin-
gle-acting pump to perform like a double-acting, thereby eliminating the usual
half-stroke period of no delivery. The delivery is divided into two reduced and
equal flow rates per stroke, thereby permitting improved pulsation control with a
dampener of smaller size. This is accomplished by making the ratio of diameters
(1.1)
The uneven displacement generated by the piston rod of a conventional double-

acting pump is avoided. See Figure 1.4.
Figure 1.1. 1700 BHP duplex double-acting piston pump. (Courtesy Oilwell Division,
U.S. Steel.)
Figure 1.2. Vertical triplex single-acting plunger pump. (Courtesy Worthington Pump Di-
vision, Dresser Industries.)
2
1.1 DEFINITIONS 3
Vertical triplex single-acting plunger pump. (Courtesy Wo1thington Pump Di-

S!NGLE-ACTING PUMP. A pump in which the liquid in each cylinder is discharged

only during a head-end or crank-end stroke during one half of a revolution. See
Figure 1.5.
DOUBLE-ACTING PUMP. A pump in which the liquid in each cylinder is discharged
during both a head-end and a crank-end stroke during one full tevolution. See
Figure 1.6.
1
ON - + - - - • - .....,_._
1
Simplex single-double-acting pump. (Courtesy American Spray Industries,

American Power Equipment Co.)
.¡:.
Figure 1.5. Horizontal quintuplex single-acting plunger pump. (Courtesy Oilwell Divi-
sion, U.S. Steel.)
Fabricated Steel Power Frame
Double Extended
Pinion Shaft
Screw-Type
Valve Covers
/
Hl-HARD Piston Rods
Roller Bearings
Liner Retention Super Dl-HARD Liners Throughout
Abrasion-Resistant Slurry-Type Pistons
Figure 1.6. 1700 BHP duplex double-acting piston pump. (Courtesy Oilwell Division,
U.S. Steel.)
en
6 PUMPTYPES
Figure 1.7. 1700 BHP triplex single-acting piston pump. (Courtesy Oilwell Division, U.S.
Steel.)
PISTON PUMP. A pump in which the liquid is displaced by pistons. See Figure 1. 7.
PLUNGER PUMP. A pump in which the liquid is displaced by plungers. See Figures
1.8 and 1.9.
Figure 1.8. Horizontal single-acting plunger Pump. (Courtesy Worthington Pump Divi-
sion, Dresser Industries)
Figure 1.9. Horizontal triplex single-acting plunger pump. (Courtesy Oilwell Division,
U.S. Steel.)
....,¡
8 PUMPTYPES
1" to 2" Plunger by 2'h" Stroke Simplex
Figure 1.10. Simplex horizontal single-acting plunger pump. (Courtesy Kerr Machine Co.
Ada, OK.)
SIMPLEX PUMP. A pump consisting of a single cylinder. Contains one piston or its
equivalent, that is, a single- or double-acting piston. See Figure 1.10.
DUPLEX PUMP. A pump consisting of two cylinders. Contains two pistons or their
equivalent, that is, single- or double-acting pistons. See Figures 1.1 and 1.6.
MUL TICYLINDER PUMP. A pump consisting of three or more cylinders. Contains
more than two pistons or their equivalent, that is, single- or double-acting. See
Figures 1.11 and 1.12.
Figure 1.11. Horizontal triplex single-acting plunger pump. (Courtesy Oilwell Division,
U.S. Steel.)
1.1 DEFINITIONS 9
Figure 1.12. Horizontal quintuplex single-acting plunger pump. (Courtesy Oilwell Divi-
sion, U .S. Steel.)
ARTICULATED PUMP. A double-acting design attained by driving an opposed set

of single-acting cylinders by means of articulated connecting rods that drive the
opposed crossheads, giving the character of a single-acting pump in a double-
acting action. See Figure l. 13.
OPPOSED PUMP. A doubie-acting design attained by driving an opposed set of
single-acting cylinders by means of crankshaft-straddling connecting bars that drive
the opposed crossheads, giving the character of a single-acting pump in a double-
acting action. See Figures 1.13 and 1.14.
MEMBRANE PUMP. The membrane pump, (sometimes erroneously referred to as a

"diaphragm" pump) differs from the diaphragm pump in that clean liquid dis-
placed by a conventional piston pump is used in tum to ''pulse'' an isolated
membrane of rather large diameter and of great placed in a chamber
between the piston liquid end and the "dirty" or abrasive pumped liquid on the
opposite side. See Figure l.15 and l.16.
Figure 1.13. Horizontal triplex articulated piston pump. (Courtesy Worthington Pump Di-
10 PUMP TYPES
Figure 1.14. 3200 BHP opposed triplex piston slurry pump. (Courtesy Worthington Pump
Division, Dresser Industries.)
DIAPHRAGM PUMP. A single-acting or double acting reciprocating pump with the

displacing piston being replaced with a fixed-edge flexible diaphragm being re-
ciprocated by a piston-rod connected to the center of the diaphragm. This ar-
rangement eliminates the actual wear and leakage path between a moving piston
and the liner. Positive displacement is obtained by the alternating "pulses" of
the diaphragm. These pumps were developed for low-pressure high-volume han-
dling of abrasive liquids. See Figure l. !A. Neither the membrane nor the dia-
phragm pump will isolate the valves from the liquid being pumped.
DUAL-DISC PUMP. Two tough rubber discs reciprocate in opposition to each other.
A Iarge cavity is created between them at one end of the stroke, producing a
positive suction and discharge sequence, resulting in a smooth positive pumping
action. See Figure l .2A.
HEAD END (HE) OF LIQUID END. The end of the liquid end farthest from the crank-
shaft. Sometimes called the cylinder head end in horizontal pumps.
CRANK END (CE) OF LIQUID END. The end of the liquid end closest to the crank-
shaft.
HEAD-END STROKE. Travel of the piston toward the head end.
CRANK-END STROKE. Travel of the piston toward the crankshaft.
UPPER OR OUTSIDE CROSSHEAD. An additional crosshead located outside the
power end of the pump at the head end and usually driven by rods connected to
the interna! crosshead. In such pumps the plunger travels toward the crankshaft on
the delivery stroke. Most vertical pumps use this feature. The upper crosshead is
"'(¡;
O>
~"'
-~~
Cl >
11
DISCHARGE
1 FLEXIBLE DIAPHRAGM
VAL VE
RECIPRDCATING RDD
VAL VE
SUCT ION
Figure l.lA Diaphragm pump
Figure l.2A Dual-disc pump (Courtesy MPL Pumps Limited, Victoria Road, Feltham
Middlesex TW13 7DS)
1.2 NOMENCLATURE AND DEFINITIONS 13
,,,.----$>-
'
1
1
!_\__~~~~~~~~~~~~-~~~~~~
-------"--
Figure 1.16. Piston membrane pump, sectional, Hquid end. Wearing parts: (6) valve
eones, (7) valve seats, (8) valve rubbers, (9) pump membranes. (Courtesy GEHO Pumps,
Holthuis b.v., Venlo, Holland.)
noted for its ability to keep leaking liquid from entering the power end and tends
to minimize plunger misalignment that can result from intemal crosshead mis-
alignment.
OUTSIDE-PACKED. A term usually used to refer to plunger pumps in which the

displacement element (plunger) runs through a stuffing box extemally accessed.
1.2 NOMENCLATURE ANO DEFINITIONS
The text and illustrations of this section are taken from Hydraulic lnstitute Stan-
dards, 1983, by courtesy ofthe Hydraulic Institute. The Institute is an organization
of pump manufacturers that has been in existence in the U nited States since 1917.
14 PUMP TYPES
It has been successful in setting up engineering standards for pumps of ali types
and continues to contribute to the industry through close association with most
pump manufacturers.
Purpose
The nomenclature and definitions in these Standards were prepared to provide a
means for identifying the various pump components covered by these Standards
and also to serve as a common language for all who <leal with this type of equip-
ment.
The following definitions and drawings illustrate typical construction of power
pump components but do not necessarily represent recommended designs. Varia-
tions in design may exist without violating the intent of these Standards.
RIGHT ANO LEFT HAND SHAFT EXTENSION OF POWER PUMPS. "Right" or "left
hand" designates the side of the power end from which the crankshaft or pinion
shaft extends. (lt <loes not designate in which direction the shaft rotates).
Horizontal Power Pumps are termed right hand or left hand as viewed when
standing behind the power end with the liquid end being the most distant part. A
left hand pump has the shaft extending out of the left side of the power end. A
right hand pump has the shaft extending_ out of the right side of the power end.
Vertical Power Pumps are termed right hand or left hand pumps as viewed when
standing at and facing the suction manifold of the pump. A left hand pump has
the shaft extending out of the left side of the power end. A right hand pump has
the shaft extending out of the right side of the power end.
LIQUID END. That portion of the pump which handles the liquid. It consists of a
liquid cylinder, valves, and other components.
LIQUID END PARTS

LIQUID CYLINDER. A chamber(s) in which the motion ofthe plunger(s) or piston(s)
is imparted to the liquid. The cylinder can be made integral with a suction and
discharge manifold or can be made with separate manifolds. (See Figure 1.17.)
Figure 1.17. Liquid cylinder.

1.2 NOMENCLATURE ANO DEFINITIONS 15
Figure 1.18. Manifold.
CYUNDER UNER. A replaceable liner which is placed in the cylinder of a piston

The piston reciprocates within the liner. (See Figs. 1.36 and 1.37.)
MANIFOLDS. A suction manifold is a chamber which accepts liquid from the suc-
tion port(s) and distributes it to the suction valves. (Fig. 1.18.)
A discharge manifold is a chamber which accepts liquid from the individual
discharge valves and directs it to the discharge port(s). {Fig. L 18.)
VALVE CHEST COVER. A cover for the valves within the cylinder. (Fig. 1.37 .)
VAi.VE PLATE (VAlVE DECK). A plate that contains the suction or discharge valves.
(Fig. 1.37.)
PISTON. A cylindrical body which is attachable to a rod and is capable of exerting
pressure upon a liquid within the liquid cylinder. A piston usually has grooves for
containing rings which seal against the cylinder or cylinder liner.
A piston in a reciprocating pump is usually double-acting.
The pistons in Figs. 1.19 and 1.20 have followers which retain the packing
Other types of pistons are shown in Figures 1. 21-1. 23.
Figure 1.19. Piston assembly.

Figure 1.20. Bull and snap ring piston.
Figure 1.21. Cup type piston.
Figure 1.22. Slush piston.
16
1.2 NOMENCLATURE AND DEFINIT!ONS 17
Figure 1.23. Individual ring piston.
PLUNGER. A smooth rod which is attachable to a crosshead and is capable of ex-

pressure upon a liquid within the liquid cylinder. Sealing rings for a plunger
are stationary, the plunger sliding within the rings.
A plunger is normally single-acting, requiring a double-acting pump to have
two plungers on each crosshead axis. (See Figure l
STUFFING BOX. A cylindrical cavity through whích the plunger or piston rod re-
ciprocates and in which liquid leakage is controlled by means of packing. (Fig.
1.25.)
A follower ring and throat bushing are used to guide the plunger or rod-as it
reciprocates. The throat bushing and follower ring contain the packing within the
stuffing box.
PACKING. A material used to provide a seal around the plunger, píston rod, or
piston. (Fig. 1.26.)
GLAND. A part which retains the packing in the stuffing box. (Fig. 1.27.)
LAlllTERN RING (SEAL CA.GE). A ring located in the stuffing box to provide space
for the introduction of a lubricant ora barrier liquid. (Fig. 1.28.)
VAL.VE ASSEMBLY. Usually consists of a seat, valve, spring, and spring retainer.
It allows liquid to enter and leave each pumping chamber of the cylinder. Each
pumping chamber has one or more suction and discharge valve(s). (Figs. 1.29 to
1
Figure 1.24. Plunger.

Figure 1.25. Stuffing box.
Figure 1.26. Packing.
Figure 1.27. Gland.
Figure 1.28. Lantern ring.

18
Figure 1.29. Plate valve assembly.
Figure 1.30. Disc valve assembly.
Figure 1.32. Ball valve assembly.
19
20 PUMP TYPES
UPPER CROSSHEAD. U sed in vertical plunger pumps to transmit the reciprocating

motion of the side rod to the plunger. (Fig. 1.33.)
uaum ENDS. Figures 1.34-1.37 show the typical liquid end configurations for
most types of reciprocating pumps, as described.
POWER END. That portion of the pump in which the rotating motion of the crank-
shaft is converted to a reciprocating motion through connecting rods and cross-
heads.
POWER END PARTS
POWER FRAME. That portion of the power end which contains the crankshaft, con-
necting rods, crosshead, and bearings used to transmit power and motion to the
liquid end. It may consist of one or two pieces. If two, one upper and one lower
half. (Fig. 1.38.)
CRANKSHAFT. The stepped shafi which transmits power and motion to the con-
necting rods. Main bearings and connecting rods are fitted on this member. (Fig.
1.39.)
MAIN BEARING. The bearing which supports the crankshaft. Main crankshaft bear-
ings may be sleeve or antifriction type, mounted at each end ofthe shaft or located
elsewhere to provide proper support. These bearings absorb the liquid and inertia
loads which are developed by the plunger as it displaces the liquid. (Figs. 1.40
and 1.41.)
CONNECTllllG ROD. Articulates the motion of the crankshaft to the crosshead. Power
is transmitted thru compression and/or tension. (Fig. 1.42.)
Figure 1.3;3. Upper crosshead.

1.2 NOMENCLATURE ANO DEFINITIONS 21
CRANKPIN BEARING. Transmits the oscillating reciprocating load transmitted by

the connecting rod to the crankshaft. (Fig. 1.43.)
POWER CROSSHEAD. Creates a linear reciprocating motion derived froµi the

crankpin rotary motion thru the connecting rod. The reciprocating motion of the
crosshead is applied to the plunger or piston via the side rods or crosshead exten-
sion. (Fig. 1.44.)
WRIST PIN. Connects the connecting rod to the crosshead. (Fig. 1.45.)
WRIST PIN BEARING. Transmits the reciprocating load of the crosshead into the
connecting rod. (Fig. 1.45.)
CROSSHEAD EXTENSION (PLUNGER EXTENSION). Connects the crosshead to the
plunger. (Fig. 1.46.)
FRAME EXTENSION. Connects the liquid end to the power frame when the liquid
end is not bolted directly to the frame. A horizontal extension is sometimes called
a cradle. (Fig. 1.47.) Diagrams ofpower ends are shown in Figures 1.48-1.50.
Figure 1.34. Liquid end, horizontal plunger power pump.
Cylinder, Liquid 309 Manifold, Suction 321 Deftector

Head, Liquid 310 Box, Liquid Stuffing 322 Ring, Lantem
Cylinder 311 Gland, Liquid 323 Cover, Valve
Plunger Stuffing Box 331 Flange, Discharge
Valve Assembly, 312 Fing, Follower Companion
Discharge 314 Bushing, Throat 332 Flange, Suction
Valve Assembly, 315 Packing Companion
Suction
Figure 1.35. Liquid end, vertical plunger power pump.
301 Cylinder, Liquid 308 Manifold, Discharge 312 Ring, Follower

303 Plunger 309 Manifold, Suction 314 Bushing, Throat
306 Valve Assembly, 310 Box, Liquid Stuffing 315 Packing
Discharge 311 Gland, Liquid 316 Crosshead, Upper
307 Valve Assembly, Stuffing Box 324 Ring, Gland
Suction
22
1.3 DOUBLE-ACTING VERSUS SINGLE-ACTING PUMPS 23
323
306
302
321
317
1.36. Liquíd end, horizontal side pot piston pump.
Cylinder, Liquid 310 Box, Liquid Stuffing 317 Follower, Piston

Head, Liquid Cylinder 311 Gland, Liquid 321 Deflector
Piston, Liquid Stuffing Box 323 Cover, Valve
Rod, Liquid Piston 313 Ring, Liquid Piston 325 Liner, Cylinder
Valve Assembly, Discharge 315 Packing
1 DOUBLEmACTING VERSUS SINGLEmACTING PUMPS
The design of the common double-acting piston pump (Fig. 1.6) is such that the
piston-rod displacement reduces the theoreti.cal volumetric piston displacement on
the crank-end stroke. The larger the piston-rod diameter, the greater the reduction.
When a double-acting pump is uprated in pressure, it is necessary to increase the
piston-rod diameter to provide the increased strength demanded by the increased
pressure. If this process is continued to the extremes of high pressure, it can be
seen that the piston rod becomes excessively large in dfameter, eventually ap-
proaching the diameter of the piston with complete loss of any liquid delivery from
the crank-end stroke. The double-acting pump then becomes a duplex single-acting
pump with about the greatest flow variation and pulsation that could be experi-
enced. As shown in Chapter 5, Pump Design, the flow variation of a typical triplex
single-acting pump has about one-half the flow variation of a typical duplex dou-
pump. The single-acting multicylinder type was developed to obtain
even greater improvement in fiow variation of reciprocatí.ng pumps.
320 306
302
326
327
305 317
Figure 1.37. Liquid end, horizontal valve plate piston pump.
301 Cylinder, Liquid 311 Giand, Liquid 319 Plate, Valve

302 Head, Liquid Stuffing Box 320 Plug
Cylinder 312 Ring, Follower 325 Liner, Cylinder
304 Piston, Liquid 313 Ring, Liquid Piston 326 Nut, Piston
305 Rod, Liquid Piston 315 Packing 327 Nut, Piston Jam
306 Valve Assembly, 317 Follower, Piston 333 Foot, Liquid Cylinder
Discharge 318 Cover, Liquid Valve
307 Valve Assembly, Chest
Suction
310 Box, Liquid Stuffing
Figure 1.38. Power frame.
24
Figure 1.39. Crankshaft.
SLEEVE
BEARING Figure 1.40. Sleeve bearing.
Figure 1.41. Roller bearing.
25
Figure 1.42. Connecting rod.
Í CRANK PIN BEARING

\ f --1
Figure 1.43. Crank pin bearing.
Figure 1.44. Power pump crosshead.

26
PIN SEARING
Figure 1.45. Wrist pin bearing.
Figure 1.46. Crosshead extension.
Figure 1.47. Frame i:;xtension.
27
102 119 107
115 106 103
110 114 104 101 112
Figure 1.48. Power end, horizontal plunger power pump.
101 Frame, Power 106 Bearing, Crankpin 112 Cover, Crankcase

102 Crankshaft 107 Bearing, Main 114 Box, Wiper
103 Rod, Cennecting Crankshaft 115 Breather
104 Crosshead, Power 109 Bearing, Wrist Pin 119 Housing, Crankshaft
105 Pin, Wrist 110 Extension, Crosshead Bearing
28
Frame, Power 107 Bearing, Main 115 Breather

Crankshaft Crankshaft 117 Rod, Side
Rod, Connecting 109 Bearing, Wrist Pin 118 W ay, Crosshead
Crosshead, Power 111 Extension Crankshaft 124 Cover, Crankshaft
Pin, Wrist 112 Cover, Crankcase Extension
Bearing, Crankpin
In selecting the type of pump for a given application, the features of each design
'must be considered. The single-acting pump offers:
1. Greater mechanical efficiency.

2. No piston-rod displacement reduction.
3. No wearing piston rod.
4. Fewer valves and piston rubbers to replace.
5. Possibility of conversion to plunger pump.
6. Simpler piston or plunger connection to crosshead (Fig. 1.51).
7. Lower dampener cost in view of lower ftow variation.
8. Operation at greater speed than is possible with the double-acting pump.
9. Ease of liquid-end servicing.
30 PUMP TYPES
Figure 1.50. Power end, horizontal duplex power pump with integral gears.
101 Frame, Power 108 Bearing, Pinion Shaft 119 Housing, Bearing
102 Crankshaft 109 Bearing, Wrist Pin Crankshaft
103 Rod, Connecting 112 Cover, Crankcase 120 Pinion Shaft
104 Crosshead, Power 113 Cover, Cradle 121 Pinion
105 Pin, Wrist 114 Box, Wiper 122 Gear
106 Bearing, Crankpin 115 Breather 126 Housing, Bearing,
107 Bearing, Main Crankshaft Pinion Shaft
Sorne advantages of a double-acting purnp, on the other hand, are:
l. Pistons can be of srnaller diarneter.

2. Sorne operators have clairned that because the liner is cooled by the
purnped liquid on both sides of the piston, there rnay be less piston and
liner wear.
Catastrophic failure of a piston does not flood the cradle with liquid that
might enter into the power end.
:gure 1.51 compares other features of the single-acting vs. the double-acting
mp.
ble 1.3 tabulates certain pump parameters as related to pump power, such data
ing been taken from a current manufacturers catalog. Note that the most sig-
cant factor is the inherent greater weight-per-horsepower of the Duplex Dou-
-acting type, partly contributed by the longer stroke.
e
A
11
11
(a) 11
A Liner
11
Extension
rod
Crosshead
11
(b)
Features of (a) double-acting versus (b) single-acting pumps.

(a) Double-Acting (b) Single-Acting
Two taper-thread connections requir- Two flanged and clamped connec-

ing high-torque tightening (and tions. Cap screw fastening with
loosening). hand tools.
Long piston rod subject to high re- Short nonwearing piston rod with
versing stress and column load- one-half stress range. All loads in
ing. compression only.
Taper piston fit requires high-pres- Straight piston fit for easy removal.
sure pulling tool.
Stuffing box and packing required. No rod or packing wear. More toler-
Piston rod and packing subject to ant to crosshead or other rod mis-
wear. Sensitive to misalignment. alignment or inherent lateral rod
motion.
(a)
N
TABLE 1.3 Pump Parameters Related to Pump Power
Triplex BHP Stroke DIA GPM RPM PSI WT, LB PRL LB/HP GPM/HP RPM/HP PSl/HP PRL/HP
1700 12 7.5 J033 150 2540 41997 112214 25 0.6 0.09 1.5 66
1400 JO 7.5 861 150 2540 413JO 112214 30 0.6 0.11 1.8 80
1100 JO 7.5 861 150 1992 38839 88004 35 0.8 0.14 1.8 80
850 9 7.5 826 160 1584 36469 69979 43 1.0 0.19 1.9 82
600 8 7 700 175 1325 19600 50992 33 1.2 0.29 2.2 85
400 8 7 700 175 882 18090 33943 45 1.8 0.44 2.2 85
300 8 4.25 332 225 14JO 8300 20003 28 1.1 0.75 4.7 67
200 6 4.25 354 320 881 3952 12498 20 1.8 1.60 4.4 62
100 4 4 294 450 525 2300 6597 23 2.9 4.50 5.3 66
Duplex
1700 18 7 683 65 3640 68311 140084 40 0.6 0.09 1.5 82

1400 18 7 683 65 3000 64000 115454 46 0.6 O.JI 1.8 82
JOOO 18 7.75 870 65 1675 47690 79015 48 0.8 0.14 1.8 79
1.4 PLU~JGER VERSUS PISTON PUMPS 33
PISTON
importance is the often overlooked fact that the packing action of an

t:;A;.,-.. ... w
ne-mu..:r._cu single-acting plunger pump is

1 opposite to that of
""'«U'~"-'" all multicylinder single-acting piston pumps.
purpose of packing is to close up the clearance gap between the
ger and its associated in the box or between the piston and its
r between the piston rod and its box With ordinary packing
using a material with considerable """"·'""' ~
e mechanics of all packing are such that uoi~a1uno:s~
the pressure acts to force that member the clearance
j\ccordingly, all sealing and subsequent wear or extrusion take place at a
1 comer or "heel" of that packing member.
a pump 1.52) (with both and piston-rod packing),
the rod is traveling on its pressure stroke, the tends to be dragged
friction into the dearance gap, thus that force ~-•'Y"'~~
""·'ª~'~'- and creating excessive extrusion (destmction). When the plunger
,_,-,~'""''ª" pump (Fig. l is traveling through the hydraulically loaded
the friction generated by the tends to counteract the effect of
.raulic pressure and reduce the tendency for extmsion. Thus the single-acting
pis inherently a ' ' pump.
ubrication of is extremely in service. It can
that only with the single-acting plunger pump can a lubricant be applied
plunger as it is entering the pressure-loaded when lubrication is
needed. (See Figures 1.52 and 1 Any attempt to lubricate a piston or
jLUBRICATION
PISTON ROD TRAVEL
HEEL
(extruded) r
DOUBLE ACTING PISTON PUMP ON PRESSURE STROKE
Note travel-of-piston and rod drag is reinforcing
tendency lar pressure to extrude pack1r.g into clearance.
pump packing action on pressure stroke. Note travel-

rPintrirr.in"tendency for pressure to extrude into clear-
34 PUMP TYPES
LUBRICATION ffi
1 .. · HYDRAULIC
.__--llll·~:·:~ ~ PRESSURE
SINGLE ACTING PLUNGER PUMP ON PRESSURE STROKE.
Figure 1.53. Single-acting plunger pump packing action on pressure stroke. Note travel-
of-plunger drag is counteracting tendency for pressure to extrude packing into clearance
(clearance exaggerated).
piston rod is not effective because the lubricated parts move into the packing on
the unloaded suction stroke.
Because of the typical packing mechanics, a plunger pump is not <;>rdinarily
as satisfactory as a piston pump for pumping abrasive liquids. Since the packing
is relaxed on the suction stroke of a plunger pump, the "dirty" plunger can readily
load the packing with abrasive particles; the packing subsequently acts asan effi-
cient lapping tool to wear the plunger. But this problem can be overcome by ad-
equate flushing of the internal packing with clean liquid, even to the exterit that a
flushed plunger pump is more desirable for pumping abrasive liquids than a piston
pump if the dilution from flushing can be tolerated. For a discussion on plunger
packing flushing, see Chapter 9, Slurry Pumping, Section 9.2.4.
1.5 INTERNAL GEARS
Because manufacturers have incorporated an internal gear reduction as part of the

power end of sorne of their pumps, a general discussion of the application of pump
drive connections is in order.
In Chapter 5 a discussion on pump speed will show that there are physical and
practica! limitations to the speed at which a reciprocating pump can be operated.
Most prime-mover speeds are much greater than those of typical large reciprocat-
ing pumps. There must always be a speed-reducing device between the pump and
the prime mover. These permanent speed-reducing devices are usually single or
double reduction gear sets, pulleys and belts, or sprockets and chains.
The larger pumps with their inherent lower speed require a greater speed re-
duction. Therefore, sorne pumps above 100 BHP incorporate partial speed reduc-
tion by the installation of a main or "bull" gear on the crankshaft anda pinion
1.5 INTERNAL GEARS 35
gear on an integral input shaft, with a reduction of about 5 to l. For example, oil
well drilling mud pumps of 500-1700 BHP (Figs. 1.1 and 1 have intemal gears
and operate at up to 125 RPM, being driven by internal combustion engines or
electric motors of 900-1200 RPM. A belt or chain drive from the rig to the pump
then requires a ratio of only 3 to l instead of 15 to 1 if the pump has no intemal
reduction.
Because drilling rigs must be capable of being moved from one site to another,
partial reduction is most suitable, because a chain or belt extemal drive is readily
dismantled and reassembled and is tolerant of sorne misalignment. For permanent
installations, a single external reduction gear can be used. The overaU cost of the
extemal gear reducer is about 60% of the internal-external reduction ar-
One of the most important reasons for omitting the gear from the interior of the
power end is the oppmtunity to elirninate an oil-bath type of lubrication with its
propensity to pick up abrasive or corrosive debris from the bottom of the crankcase
and direct it to the running parts, A "dry sump" can be used that will allow all
circulating oil to be cleaned and filtered before it is recirculated to the running
2
DYNAMICS
2.1 INTRODUCTION
The positive displacement pump generates very specific dynamic liquid ftow and
pressure conditions, but the response of the system to the pump dynamics is a
result of the system operating conditions and specific piping arrangements.
There are three steady-state pressure conditions that affect system liquid dy-
namic responses: friction-generated pressure from the average liquid ftow, vertical
head from a column of liquid, and pressure applied by a gas blanket. There are
also three sources of dynamic pressure disturbances generated by the positive dis-
placement pump: frictional pressure drop, acceleration from the liquid flow vari-
ation of the pump, and low-amplitude waterhammer-type pressure disturbances
that occur each time a pump valve opens or doses.
System response dynamics will be discussed in Chapter 4, Pulsation and Surge
Control.
2.2 STANDARD DEFINITIONS

SUCTION. The entire piping system that connects the liquid supply to the pump
inlet.
INLET. The actual connecting point ofthe suction to the pump, usually in the form
of a ftanged, screwed, clamped, or welded joint on a manifold that connects all
cylinder inlet valves to a common point.
DISCHARGE. The entire piping system that connects the liquid delivery to the pump
outlet.
36
2.3 FLOW VARIATION 37
TLET. The actual connecting point of the discharge to the pump, usually in the
nn of a flanged, screwed, clamped, or welded joint that connects all cylinder
tlet valves to a common point.
FLOW VARIATION
·eciprocating pumps exhibit certain typical flow variations in the discharge (out-
t) and suction (inlet) caused by the rotary motion of the power end driving the
placement elements (pistons and plungers). This is demonstrated by the ge-
etry in Figure 2.1. The formulas for instantaneous velocity are:
Vp(ft/s) 0 = 0.042sw[sin () + (s/4Lc) sin W] (2.1)
w = 27rN /60 (2.2)

vi
Vs (ft/s)o = :¿ vp Vs
~ (2.3)
Vpi
Vd (ft/s)o = :¿ vp Vd
-i (2.4)
e general Equations for Velocity and Velocity Pressure in pipes are;
V = 0.4085 x GPM
~~~~~~ (2.5)
vi
Pv = 0.0001137 x (GPM)i/D4 (2.6)
.e Symbol Notation in Appendix.
Stroke s, in pipe síze, m
270°
Pisten diameter, in
Head
end
lnlet connection
pipe size, in
Figure 2.1. Pump geometry.

38 DYNAMICS
where
. . 2
AP = area o f p1ston, m
Ad = area of discharge pipe, in2
As = area of suction pipe, in 2
= piston velocity, ft/s
Vs = suction {Jiquid) velocity ft/s
Vd = discharge (liquid) velocity, ft/s
E = summation (algebraic) of individual cylinder functions
The type of pump is responsible for large differences in the shape of the ftow
pauem (Fig. 2.2). This shape depends on both the ratio of connecting rod length
Le to crank radius r and the loss of displacement on the piston rod end (crank end)
of a double-acting pump.
Conventional pump design is such that the arrangement and number of cylinders
generate different maximum and minimum ftow rates. The flow rates are expressed
as a percentage of the average as shown in Figure 2. 3, which illustrates one typical
pump geometry with specific Le/ r ratio. For instance, a triplex single-acting pump
has six points of maximum flow rate per revolution. These points are duplicated
in both the discharge and the suction of a pump.
Because frictional pressure drop in a pipeline is a function of the square of the
velocity, these flow variations will be converted into pressure variations (pulsation)
following the same shape as the flow pattern.
2.4 ACCElERATION
In reciprocating pumps, again because of the typical harmonic motion, the liquid
entering and leaving each pump cylinder experiences a "start-and-stop" flow with
velocities ranging from zero at the beginning and end of a stroke to a maximum
at midstroke.
The velocity variation and its contribution to the instantaneous value of accel-
eration is determined by the following:
(2.7)
The accelerations of the liquid at the suction and discharge connections are the
result of the cylinder-generated acceleration and are cakulated by
Dp2
as (ftlsz)o = :¿ aP Dsz (2.8)
Dp2
ad (ft/s 2)e = 2.: ªr Dd 2 (2.9)
2.4 ACCELERATION 39
Single-acting
1
Double-acting
360°
re 2.2. Reciprocating pump mechanisms. CE = Crank end; HE = Head (cylinder)

; RH, LH = Right hand and left hand of opposed-cylinder pumps.
liquid, having mass, is subject to Newton's law of and acceleration

converted to pressure according to the basic formula Force = mass X acceler-
P",(PSl) 8 0.0069Lpas (2. l 1)

Pª.;(PSl)a 0.0069Lpa,¡ (2.12)
0.0069 is a constant used to convert to PSI units.

40 DYNAMICS
Duplex double-acting
Maximum +243
Minimum -223
Total 463
Triplex single-acting
Maximum + 63
Minimum -173
Total 233
Quintuplex Singel
Maximum +23
Minimum -5%
Total 73
Septuplex single-
acting
Maximurn +l.2%
Minimum -2.63
Total 3.83
Figure 2.3. Flow variations for multicylinder pumps.
2.5 DERIVATION OF ACCELERATiON PRESSURE
Newton's second law of rnotion is the basis of the calculation of pressure due to
acceleration.
The acceleration of a body (of liquid) is directly proportional to the resulting

force acting upon it and is inversely proportional to the mass of the body.
2.6 DERIVATION OF ACCELERATION PRESSURE AT PUMP INLET 41
The proportionality factor K is made unity by choosing a proper unit of mass,

the slug.
a= KF/m (2.13)
Set K = 1; then
m = Fla slugs (2.14)

anda = lb1/slug (2.15)
Where slug = (lb/ft3)/g (2.16)
= lb · s2 /ft4 (2.17)
The weight of a slug on earth is about 32.17 lb.

Density p is expressed in slugs per cubic foot. p = 1.94 for water.
2.6 DERIVATION OF ACCELERATION PRESSURE AT PUMP

INLET
Basic premise: Pressure is due to acceleration; F = ma, or (2.18)
Pª, = 0.0069aLp for one cylinder (2.19)
Pa, = 0.0132aL (2.20)
Transforming Pac into commonly used pump terms, we have
Pac = CLNQS! D 2 (2.21)

where C, = 0.0132 d 2/NQS (2.22)
To derive C 1 set (2.13) equal to (2.12) and solve for C;;
C 1,LNQS 0.0132a.d 2
d2 = 0.0132aL so C2 = NQS
ubstitute the value for maximum acceleration (a) as computed for, for example,
triplex single-acting pump; use that pipe diameter (d); and solve for C:
C, = 0.132 X 38.7 X 3.8262 /150 X 147 X 1 = 0.000339
r any type of triplex single-acting pump.

42 DYNAMICS
Solve for C 1 for all other pump types in the same manner. See Table 2.1.
Wright equation for Suction Accelleration Head in feet
(2.23)
TABLE 2.1. Constant C1
Pump e Pump e
Simplex DA 0.001660 Triplex SA 0.000339
Duplex SA 0.001032 Sextuplex 0.000237
Duplex DA 0.000624 Quintuplex SA 0.000190
Quadruplex DA 0.000434 Septuplex SA 0.000146
Triplex DA 0.000338 Nonuplex SA 0.000107
DA = double-acting; SA = single-acting.
TABLE 2.2. Constant C2
Pump Type C2 Pump Type C2

Simplex DA 0.200 Triplex SA 0.066
Duplex SA Sextuplex SA 0.055
Duplex DA 0.115 Quintuplex SA 0.040
Quadruplex DA Septuplex SA 0.028
Triplex DA 0.066 Nonuplex SA 0.022
2. 7 CRITICAL SUCTION CONDITIONS
Two conditions in the suction system of a reciprocating pump that affect pump
performance are (1) the pressure drop between the source and the pump due to
frictional loss of liquid flow velocity and (2) the instantaneous pressure, which
must always be of such magnitude as to accelerate the mass of liquid in the suction
pipe at that instant. In both cases, the pressure reduction created at the pump inlet
may vaporize the liquid and cause cavitation in the cylinder. Of these conditions,
acceleration problems in the suction are by far the most severe.
2.8 DISCHARGE ACCELERATION
Acceleration can occasionally contribute to problems in the discharge of the pump.

Reciprocating pumps of ordinary design, with their distinctive discharge flow char-
acteristics, can generate pressures due to the variations of flow that vary linearly
with pump speed. But the acceleration of the liquid at the beginning of each in-
crease in flow rate generates a pressure proportional to the square of the pump
2.8 DISCHARGE ACCELERATION 43
~peed. Accordingly, acceleration effects from high pump speed can be the greatest
~ause of unsatisfactory performance in the discharge.
In typical hydraulic systems, where the discharge pressure is generated by such
pump(s), and is the result of friction losses from a rather long pipeline or through
some sort of restriction, the hydraulic pressure variations (pulsations) dueto ftow
are far greater than those generated by acceleration. However, when the discharge
pressure is generated by a direct vertical head or other system with very small
friction losses, the acceleration pressure variations can become predominant.
For example, Figure 2.4(a) shows a pump forcing water out of a mine shaft at
'ª depth of 2000 ft through a vertical pipeline that has a fridion loss of only 1 PSI
,. r 1000 ft. This would result in a discharge pressure of 2000 X .43 + 2, or 862
SI. lfthis same line were laid horizontally, as in Figure 2.4(b) it would have to
e 163 mi long to show a similar pressure. lt is obvious that the two systems
resent completely different conditions to the pump. Also considera pump feeding
~(ir\
T
2000 ft water = 860 PSI head
+ 2 PSI due to friction
ººº
l
~--~ Total pressure at pump = 862 P 1
(b)
s
163 mi; Friction = 1 PSl/1000 ft = 862 PSI

.1
0001~~~~~~~~\·,
. Total pressure at pump = 862 PSI
~
¡ ,\\ 1
(a)
Short connecting pipe

Pressurized system
000 Total pressure at pump
860 PSI
= 862 PSI
(e)
ure 2.4. Dynamic differences in pump discharge systems. (a) Vertical discharge. (b)
orizontal discharge. (e) Pressurized systems.
44 DYNAMICS
an existing pressurized pipeline with a short, low-friction connecting line, as shown

in Figure 2.4(c).
Sorne of these applications may exhibit discharge pressure pulses of the water-
hammer type due to acceleration as discussed, with amplitudes far in excess of
those normally generated by ftow variation. Experience shows that these high-
frequency pulses cannot be attenuated by the use of conventional gas-bladder de-
vices simply because such devices cannot react quickly enough to absorb much of
the pulse. See Chapter 4, Pulsation and Surge Control.
Since such acceleration disturbances are related to high pump speed (to the
second power), unexpected problems from these excessive disturbances in vertical-
lift applications could certainly be minimized by limiting the pump speed.
2.9 HVDRAULIC FLOW ANO PRESSURE WAVEFORMS
A centrifuga! pump differs profoundly from a reciprocating pump in that it always

has hydraulic communication between the suction and the discharge so that dy-
namic disturbances in the suction can greatly affect the discharge and vice versa.
The discontinuity and isolation between the suction and discharge hydraulics of a
reciprocating pump are the result of the valve action. A suction valve cannot begin
to admit liquid until the discharge valve has completely closed and the cylinder
pressure drops to below the liquid suction pressure minus the spring loading of the
valve. Because of this discontinuity, suction deviations can be studied apart from
the discharge deviations, or vice versa.
The basic hydraulic pressure variation from the positive-displacement pump is
generated by the varying velocity resulting from the mechanical geometry of the
pump. As angular momentum of the pump's crank (Fig. 2.1) is translated into a
sliding linear motion, a varying velocity or instantaneous ftow rate is generated
based on the mechanical design of the pump (Fig. 2.2).
A graphical presentation of one revolution of several pump types is presented
in Figure 2. 3 to demonstrate how the instantaneous discharge velocity or ftow rate
varies with pump type. As the number of pump cylinders is increased, the range
of instantaneous velocity or ftow rate decreases.
The ways in which hydraulics and pressures affect system and pump perfor-
mance will be discussed in Chapter 4.
2.10 DISCUSSION OF IDEAL WAVEFORMS
Before beginning a discussion on pump generated waveforms, the three distinctly

different discharge systems encountered in practice, namely those shown in Figure
2.4, should be considered.
a. A long horizontal pipeline, about one mile or longer, where the principal
pressure generated by the pump is due to the frictional liquid ftow losses in'
the pipe.
2.10 OISCUSSION OF IDEAL WAVEFORMS
b. A vertical discharge pipe with very little frictional loss but with enor-
mous acceleration disturbances resulting from the forces to accel-
erate the slug of in the column.
c. A short low-resistance line connecting an already pressurized such
to a pipeline. Unless other pulsations are present in the
this anangement produces the least amount of disturbance.
ín each of these cases, one should consider the unpredictable acoustic disturb-
ances that may exist in sections of tuned to a harmonic of the basíc pump-
-M.~r~"""1 frequencies.
2.5 is the waveform of a triplex single-acting pump for one revolution .
.The upper chart shows the instantaneous flow at various of crankshaft
Discharge flow rate
t\ \
\
j
\
\
'
o 30 60 90 120 150 180 210 240 270 300 330 360
Crank a11gle, degrees
Figure 2.5. Ideal

44 DYNAMICS
an pressurízed pipeline with a short, low-friction connecting line, as shown

in Figure 2.4(c).
Sorne of these applications may exhibit discharge pressure pulses of the water-
hammer type due to acceleration as discussed, with amplitudes far in excess of
those nonnally generated by fl.ow variation. Experience shows that these high-
pulses cannot be attenuated by the use of conventional gas-bladder de-
because such devices cannot react quickly enough to absorb much of
See Chapter 4, Pulsation and Surge Control.
Since such acceleration disturbances are related to high pump speed (to the
second power), unexpected problems from these excessive disturbances in vertical-
lift applications could certainly be minirnized by limiting the pump speed.
HYDRAUUC FLOW ANO PRESSURE WAVEFORMS
A centrifuga! pump differs profoundly from a reciprocating pump in that it always

has hydraulic communication between the suction and the discharge so that dy-
namic disturbances in the suction can greatly affect the discharge and vice versa.
The discontinuity and isolation between the suction and discharge hydraulics of a
reciprocating pump are the result ofthe valve action. A suction valve cannot begin
to admit liquid until the discharge valve has completely closed and the cylinder
pressure drops to below the liquid suction pressure minus the spring loading of the
valve. Because of this discontinuity, suction deviations can be studied apart from
the discharge deviations, or vice versa.
The basic hydraulic pressure variation from the positive-displacement pump is
generated by the varying velocity resulting from the mechanical geometry of the
pump. As angular momentum of the pump's crank (Fig. 2.1) is translated into a
sliding linear motion, a varying velocity or instantaneous ftow rate is generated
based on the mechanical design of the pump (Fig. 2.2).
A graphical presentation of one revolution of several pump types is presented
in Figure 2.3 to demonstrate how the instantaneous discharge velocity or flow rate
varies with pump type. As the number of pump cylinders is increased, the range
of instantaneous velocity or flow rate decreases.
The ways in which hydraulics and pressures affect system and pump perfor-
mance will be discussed in Chapter 4.
Before beginning a discussion on pump generated waveforms, the three distinctly

different discharge systems encountered in practice, namely those shown in Figure
2.4, should be considered.
a. A long horizontal pipeline, about one mile or longer, where the principal
pressure generated by the pump is due to the frictional liquid flow losses in
the pipe.
b. A purely vertical discharge pipe with ver¡ little frictional loss but with enor-
mous acceleration disturbances resulting from the forces to accel-
erate the slug of liquid in the column.
c. A short low-resistance line connecting an such
to a pipeline. Unless other are in the
~,,,.,~,,.y,,,. produces the least amount of disturbance.
In each of these cases, one should consider the unpredictable acoustic disturb-
that may exist in sections of pipe tuned to a harmonic of the basic pump-
----~~n•ª·" frequencies.
2.5 is the wavefonn of a triplex pump for one revolution.
upper chart shows the instantaneous flow at valious degrees of crankshaft
Discha rge flow rate
175 ?""" !"-

- -~~ ...... ,.,.
/ '
165
A /"""1' j /
Ave
137
\/ \ 1/ ~\ \
\ '
/
\ 1
\
\
\
\
One cylinder /
I
I
t I
I
o
o
\
Suction flow rate - Sarne purnp
/
'
The discontinuity
i \
\
\
\ I
I
I
I
I '
\ ¡
\ !
\ 1
í1
137 \\ I
I
__ L_¡_ 1
'
\ J
\
'-
J\ u ~ J \ "
' V
J
,,,., 1 "- !.r
7
o 30 60 90 120 150 180 210 240 270 300 330 360
Crank angle, degrees
Figure 2.5, Ideal single-acting pump.

46 DYNAMICS
rotation. Note how the three single-cylinder displacements add to generate a typ-
ical six-node flow curve. Almost an exact mirror image of the discharge flow rate
is seen in the suction system of that pump (see lower chart).
The typical ftow pattem, shown in Figure 2.5, will generate a pressure _"pat-
tem" of the same shape but of an intensity proportional to the frictional pipe or
stricture presented to the pump.
Experience has shown that a nondampened discharge system whose discharge
pressure is the result of frictional drop due to liquid velocity will experience pres-
sure variations of approximately double the ftow variation of the pump times the
average operating pressure. This pressure variation can be explained by the steady-
state equation for frictional pressure drop in a pipeline where the. pressure differ-
ential is a constant times the square of the velocity.
A reciprocating pump discharging through a short frictionless pipe generates no
"flow-induced" pressure pulses.
Figure 2.6 shows another source of pressure generation dueto the acceleration
of the liquid as it moves through the system at varying velocity. At each change
in velocity of flow an acceleration pressure will arise, and this pressure component
must be added to the existing flow-induced pressure, drastically modifying the
resultant waveform. This is shown in the upper chart of Figure 2.6.
The lower chart of Figure 2.6 also shows that the pressure waveform in the
suction is almost completely controlled by the intensity of the acceleration pres-
sure. All reciprocating pumps exhibit this difference in performance.
Pump systems that have a substantial discharge pressure due to the acceleration
of gravity on a vertical column or vertical lift will experience ª!1 amplification
effect on the liquid acceleration pressure component. Acceleration pressures are
neutral from a steady-state standpoint-they only cause pressure ftuctuation and
do not contribute to discharge pressure Fig. 2. 7.
Figure 2.8 shows the discharge pressure as affected by the valve operation delay
inherent in reciprocating pumps. Figure 2.9 shows the nonuniformity of ftow rate
in a duplex double-acting pump. It is obvious that the pres~ure waveforms ob-
served in real pumps present a myriad of shapes that must be objectively analyzed ..
2.11 MEAN OR AVERAGE FLOW-RATE
Referring to Figure 2.5 one may ask "What factors tell one to place the mean
flow-rate at only about six percent below the peak flow, from 175 to 165 GPM,
but about 16 percent above the mínimum ftow-rate of 137 GPM?''
Such a decision need not be made-the average or mean delivery rate of 165
GPM has already been exactly determined by the pump displacement per time,
or an amount controlled by the pump piston diameter, stroke length, number of
active cylinders and speed in RPM. Around this basic ftow-rate of 165 GPM the
dynamic geometry of the pump generates the typical instantaneous ftow-rate re-
lated modified sine wave expected for that pump.
2.12 PUMP DISPLACEMENT 47
These departures above and below the average ftow-rate become less in value
and of different relation to each other as the number of active cylinders are
increased. See Table 2.3
2 PUMP DISPLACEMENT
The actual rate of volume output of a reciprocating pump, Q", is called Displace-
ment and is described as the rate-of-flow in US Gallons per minute. A
Discharge: Resulting pulsation, 513

1
üi
1195
Ave
1028
V,
,I
·\, 11,...., :-.., / /
,. \, ~ ~, /.
/' ~' 7' 1\
~ II '
I
i\1
Q..
"' . I V
......
"""'
ro:::; \
1 \ \ /
.i::"'
u V>
Normal "floJ" Plus "acceleration"
pressure pressure
"'"'5.
i5 669 1 1
' , ,,
.....
',\
'' ~,
'\
'\ \
'\
\
\
'\ ~,
\
'
' \
\
<:¡¡¡ \ \
\
oc..
·~ ~ '' '' ' \ ''
~~
"'"'
u O>
u ...
..; a.
t
o
Acceleration
1 1
'
The discontinuity
Suction: Same pump ("Abs O" not applicable to lower trace)
45
Cñ
Q.. Static
35
/ // / / j / j
V
e:;
"' V
V V Lf
/
!/ / ~J
o if¡
-
gea. /
·-
IJ)
25
"Acceleration"
ma jor ettect
,,
e¡¡:;
oO..
~e
t
Abs
o ".
// J/1 I"
// /
/
,/'
// /
[/
-"'"'
"'
"'<Jl / /
I
.,,i' / ,,,,/ /
/
u"'
u .... _,/
< a.
o 30 60 90 120 150 180 210 240 270 300 330 360
Crank Angel, Degrees
2.6. Ideal waveform-triplex single-acting pump. Effects of acceleration on suc-

tion and discharge.
48 DYNAMICS
cating pump seldom delivers one hundred percent of the theoretical displacement
volume of the swept stroke length, due to slip through the valves and to de-
compression of the liquid remaining in the cylinder at the end of the delivery
stroke (Clearance Volume 1). Accordingly, Q0 must be determined by test or actual
measurement of the liquid volume delivered for a definite number of strokes.
~
t
~
=>
::¡
eo.
1400
815
Ave
o
-- -· - v-- ----= - rr- --
:;;,,-= ~ r-- ;;;.;;,;:::
/
-......;;; ~ ;;;.;:;;..:: ::::,,. t-;.;::;: ~ ---
--~
- ---
Acceleration Discharge
~ pressure pressure
"':;¡
.r:.
o 30 60 90 120 150 180 210 240 270 300 330 360

Figure 2.7. Ideal waveform-triplex single-acting pump. Discharge pressure. 1800 ft ver-
tical head, 77% pulsation.
DISCHARGE PRESSURE - 10º VALVE DELAY, 300 RPM, 933 PULSATION

1217r--=~-r~-r-~-r-~r---.r---,.~-r~-r-~-r-~r---,
Ave
966 ...............~..................._...,...,,,,_--+-......-+----+--<r---+-.........~-+-l~+---+->--+-'.........- - I
\ 1
/
' ,,...
1
Acceleration
pressure
o 30 60 90 120 150 180 210 240 270 300 330 360

Figure 2.8. Ideal waveform-triplex single-acting pump. Discharge pressure-10º valve

delay, 300 rpm, 93% pulsation.
2.12 PUMP DISPLACEMENT 49
The theoretical displacement is used to evaluate the Volumetric Efficiency and

is ca!culated by the following formulas;
For Single-acting Purnps;
Q, OJXJ34D;snN (2.24)
For
0.0034(2D; - (2.25)
DISCHARGE FLOW RATE ~ 62 GPM 46 PERCENT FLOW VARIATION
86
::¡¡;
o..
(.!)
62
2
r:
3
o
cyiio~l\
:;::
"'E:'
"'
.;:;
u I
"'
i5 \
J°' I
I \
\
to I
I \
\
\
The discontinuity
o
t
:a;
o..
('J
2
e
:;:
o
;;::
51.
e
o
~
:>
U)
o 30 60 90 120 150 180 210 240 270 300 330 360

Figure 2.9. Ideal double-acting pump. Discharge flow rate 62

gpm, 46% flow variation.
g: TABLE 2.3. Constant C:i
Numb Crank Total(3) Max Press
Pump Type Cyls Angle A(I) 8(2) Percent K(4) Percent(5)
Simplex SA 1 360 0.58 1 158 0.684 250
Simplex DA 1 180 0.29 1 129 1.368 166
Duplex SA 2 180 0.29 1 129 0.558 166
Duplex DA 2 90 0.24 0.22 46 0.199 93
Triplex SA 3 120 0.06 0.17 23 0.100 43
Triplex SA (0) 3 120 0.062 0.17 23 0.100 44
Triplex SA (30) 3 120 0.061 0.12 18 0.078 35
Triplex SA (60) 3 120 0.05 0.09 14 0.061 27
Triplex DA 3 120 0.06 0.17 23 0.200 43
Quadruplex SA 4 90 O.JI 0.22 33 0.143 62
Quadruplex DA 4 90 O.JI 0.22 33 0.286 62
Quintuplex SA 5 72 0.02 0.05 7 0.030 14
Quintuplex DA 5 72 0.02 0.05 7 0.076 14
Sextuplex SA 6 60 0.06 0.17 23 0.100 43
Sextuplex DA 6 60 0.06 0.17 23 0.200 43
Septuplex SA 7 51.4 0.012 0.026 3.8 0.016 8
Septuplex DA 7 51.4 0.012 0.026 3.8 0.032 8
Octuplex SA 8 45 0.026 0.052 8 0.035 15
Octuplex DA 8 45 0.026 0.052 8 0.070 15
Nonuplex SA 9 40 0.006 0.002 2.1 0.009 2
Nonuplex DA 9 40 0.006 0.002 2.1 0.018 2
--
NOTES;
(1) - "A" Flow variation above average, decimal.
(2) - "B" Flow variation below average, decimal.
(3) - Flow variation, total percent
(4) - K Factor for calculation of dampener size (Equation 4.5)
(5) - Maximum pressure variation (pulsation) without dampener, percent.
(0) - Fixed phase angle of two pumps compounded.
2.13 GRAPHICS OF PUMP PERFORMANCE 51
Volumetric Efficiency is then calculated as follows;
VE = Qa/Qt X 100 (2.26)
To calculate the instantaneous piston/plunger position in the cylinder, in inches

from the head-end dead-center, per degree of crankshaft rotation.
X = 0.5 S(l - cos 8) Le ( l -

+ -; ¡ - r: 'Sin 8))
Le
(2.27)
x = Position of piston or plunger from HE dead center,

per degree of crank shaft rotation (8), in inches.
S = stroke, inches.
8 = Crank angle, degrees.
Le = Length, con rod, inches.
r = Crank radius, inches.
(1) See Section l l for discussion on Clearance Volume
GRAPHICS OF PUMP PERFORMANCE
Following a discussion of pump dynamics, it is proper to cover the different

means of recording and displaying pressure and flow-rate waveforms.
Val ve Delay and Compressibility of the Liquid tend to distort and complicate.
the flow patterns and pressure waveforms of reciprocating pumps. For example,
it is difficult to picture the Suction function being affected by the plunger Dis-
charge stroke and vice-versa, keeping in mind the discontinuity between Suction
aod Discharge.
· F-A is the actual or effective volume of output liquid leaving the pump.
Figure 2. lA a typical flow-rate pattern, shows that Valve Opening or Closing
elay and Liquid Compressibility, both being time-dependent, and while both
generally exacerbated by high pump speed, the graphics tend to present a
{ightly distorted view of the relative time involved when the data is plotted on
'degree of rotation' base, resulting in an equal-time unit base.
Figure 2.2A is interesting in that it shows a completely different shape when
he same data is plotted on an equal stroke-length unit basis with the discharge
troke being folded back as in actual operation.
Figure 2.3A is also presented in the normally used continuous pressure display
here the full cycles are repeated over many revolutions on a constant time unit
¡tsis.
Figure 2.4A shows the Pressures or Flow-rates vs. Equal Stroke Length,
olded back on itself as in an "indicator card".
lle:
(.¡.) ~~
...CIJo f'll
><
1 1 1 1 1 11 1 l 1'.11
N
~
llro'2
co
~
o: -
E--
::i
o
-<
z)>
-
Q
:1 1
1:1.. 2: ;¡::
~
o !11 1 (i
o (/)
o
' ~
1
Q
M
¡... a:
<
¡¡i::
-
o
ro
~
a)
..'l
"" ·~
rnz
-o
- ~
>
..l
<
~
~
g:¡,,
;:¡¡
ot.) 1
'
-o
....
fo¡
a) z
QO > Qo
soo 330 o so 60 90 120 uso 180 219 240 270 300 339 360 30 80
H A :a e D E F Q H A' B' C'
SEQUEN CE OF EVENTS
A Start of Mechanical Suction Stroke. -E-F Suction Valve
A-8 Discharge Valva Closing Delay. Back-flow from D-E
Discharge Liquid back-flows into F Suction Valve Closes.
Cylinder. F-G Compression of trapped liquid in
8 Discharge Valve Closes. Cylinder,
8-C Decompression of Trapped Liquid. G Discharge Valva Opens.
e Suction Valve Opens, G-H Liquid delivered to System.
C-E Suction Cylinder filling with liquid, H-A' Equivalent Liquid back-flow to A'-B'.
E Start of Mechanical Discharge Stroke.
SIMPLEX SA PUMP
30 DEGREE DELAYS
zoo
150
100
50
-50
-100
-150
-zoo .....;-..... ·=· ·-=· ~-·~·-~· -=· ~-· ~L=·-~· -¡· e=-·=· 9· -~· =-··-··-··-··-··-··-··-··1····-··-··-··-··-··-··-··-
- -t"~ l
-250 '--~~~-'-~~~~-'-~~~~'--~~~-'-~~~~..i.....~~~--1
o 2 3 4 5 6
STROKE - INCHES
-e- SUCTION -e- DISCHARGE
Figure 2.2A
SIMPLEX SA PRESSURE
WITH 30 DEGREE DELAYS
250 ¡
A
-·~·-··1 -·-..··t···-··-···t···-··-··t····-..···I--·-··-··- ...........¡........... i·-·--·-r--··r.··-..·r-

1 F
zoo
=tt---~ t:t~=t=
150
~lfFF
100
1 1 1 ! 1
50 ............. ··-· ··-l·-..···-··-+·-··-··-·4··-··-··-..··-·--·-·
!
i
1 !i
i ¡
!
o
-5
-10
-15
-zo
-25
o 30 60 90 1ZO 150 180 210 240 270 300 330 360 390
DEGREES ROTATION
Figure 2.3A
53
SIMPLEX SA INDICATOR CARO
WITH 30 DEGREE DELAYS
250
200
150
100
~ 50
u.i
a:
:::1 o
g¡
..
w
a: -5
-10
-15
-20
-25
o 2 3 4 5 '6
STROKE, INCH
Figure 2.4A
54
3
SUCTION
REQUIREMENTS
FOR RECIPROCATING
POWER PUMPS
3.1.1 lntroduction
The definition of a reciprocating power pump is "A mechanical device used to
impart a pulsating, dynamic fiow to a liquid and consists of one or more single-
acting or double-acting positive displacement elements (pistons or plungers) re-
ciprocating in a liner or stuffing box in the liquid end of the pump. These ele-
ments are driven in a more-or-less harmonic motion by a rotating crank and
connecting rod mechanism. The liquid fiow generated by the displacement of these
reciprocating,elements is directedfrom the inlet (suction) to the outlet (discharge)
by the selective operation of self-acting check-va/ves located at the inlet and the
out/et of each displacing element." While there are pumps of many combinations
of multiple cylinders, from one to nine or more, and single-acting or double-
acting cylinders, this discussion will be confined to a popular triplex (three cyl-
inder) single-acting pump as a model.
The term NPSHR for a Centrifuga! Pump is a precise, absolute, and ideal and
almost the only point on the suction-head performance curve at which the pump
will operate at maximum efficiency and performance, and the system must pro-
vide those exact requirements. Figure 3.1.
The present so-called 'NPSHR' ('NPIPR') value for a Reciprocating Pump,
(whether given in geodetic head or pressure, makes no difference), published by
the manufacturer is usually not the ideal point on the performance curve at which
to run the pump-it is .most times a poor minimum that may allow the pump to
barely operate at ali! Fi~ure 3.2.
55
O 1 2 3 4 6 6 7 8 9 10 11 12 13 14 15 16 17 HI 19 20 21 22 23 24
CAPACITY, 100 GPM
Figure 3.1 Centrifuga! Pump NPSH vs. Capacity and Head
100 100
9!1 ·····················-····· ....................................................................................................................
100 l'PH
+...........................i .............................+...........................
...... ,. ........................... ,. ..... 97
ll3 ........................................ ,. .. .
92 ............................................................................................................................................................... .
91>--~~~--<~~~--+~~~~--+~~-~-+-~-~~-+--~~~~91
-Hi -10 -6 O 6 10 16
SUCT!Oli PRESSURE, PSIG
Figure 3.2 Reciprocating Pump-Volumetric Efficiency vs. Speed and Suction Pressure
56
3.1 INTRODUCTION 57
Suction Valve
Opening Delay
¡e-¡- Decompression
.+-Discharge Val11e
Seal ing Deloy
ROUNDED BOTTOM Mechanical Stroke
OF WAVEFORM AS TCP·P 0 +P 5 t+Psz-P 5 f-P 50 c-Pp -Pvp
IT EXTENDS INTO
VAPOR PRESSURE Psvo • Psac + PP
REGION.
Figure 3.3 Pump TCP
e adoption of the Centrifugal Pump term NPSHR for Reciprocating Pumps

caused more technical confusion in the industry than it deserves and it should
immediately eliminated and replaced with a performance curve from which
operator can select the best operating point for the project at hand.
f NPSHR figures are bandied about, it is possible that one reciprocating pump
'be unfairly rated over another. The general design of the liquid ends is not
uch different that variation in suction requirements would be encountered-
e (spring) design and pump operating speed contribute to any great differ-
s, both independent of basic liquid-end design.
Analysis of Pump Suction Requirements

these facts in mind, one should not stop the analysis of the reciprocating
p suction requirements at the pump "inlet" flange or connection (as Jegiti-
ly allowed with Centrifugal Pumps)-it should continue through the inlet
. ection and through the suction valve into the pump cylinder chamber at the
terline of the cylinder. See Figure 3.3. Accordingly, this gives further support
58 SUCTION REC!UIREMENTS FOR RECIPROCATING POWER PUMPS
that the so-called 'NPSH' or 'NPIP' be discarded and be replaced with a new
term, TCP (Total Cylinder Pressure). Or better still, just a Performance Curve of
TCP vs VE from which the operating point can be judicially selected. Equations
for suction requirements as listed in Section 3.2.6 will show that valve spring
POSIVA (installed valve spring load of 'POunds per Square lnch of Valve Area')
must be involved in ali calculations of suctíon requirements, for example;
Neglecting valve spring of POSIVA 4 PSI;
Pre TCP
Pre TCP 14.7 +O+ 4.3 - l.69 + 0.14 - 0.13 - 10 = 7.32 PSI
However, in order for the liquid to enter thc cylinder it must now pass through
that val ve having 4 PSI POS IVA;
So;
Pre TCP pª +
Pre TCP 14.7 +o+ 4.3 - 1.69 + 0.14 -0.13 10 - 4 3.32
Pump would rim (how reliable and efficient?)

For valve spring POSIVA = 8, as required for high speed;
TCP = 14.7 +O+ 4.3 - 1.69 + 0.14 - 0.13 - !O - 8 -0.68 PSI
pump would not run with riegative TCP.
Then, TCP, available or required, is that absolute pressure ex1stmg in the

pump cylinder at the instant of Suction Valve Opening-always after an unpre-
dictable delay of a few degrees from start of the mechanical Suction Stroke.
3.1 .3 More on Suction Requirements
In order to better understand sorne of the mysteries surrounding the dynamic

suction pressure alignment at the inlet and cylinder of reciprocating pumps, a
discussion of these pressures is in order.
Starting with the basic equations for TCP mentioned above, ali of the pres-
sures involved are contained in these equations-(3. l) through (3.10), Section
3.2.6.
Figure 3.4 shows the actual simultaneous cylinder and suction pressure wav- .
eforms along with the corresponding theoretical flow-pattern for a 3 X 6 Triplex
Single-Acting pump. Note a slight amount of "rounding" (cavitation) of the suc-
tion waveform even in a "good" suction system.
Equation (3. 7), where it can be used, is the most accurate since the suction
gauge pressure reading presents an integration of most pressure losses existing at
ACTUAL PUMP DISCHARGE PRESSURE TRACE
CYL
'
Discharge
TilE OVERWHI:LMING TBI:ORETICAL su=oN ACCll.I:RATION Plll&.TION (ll)

SUPDUMPOSI:D ON THI: TBI:ORETIC..U. fl.QW-INDUCI:D l'ULU.nDN (SV).
sucnoN V AL VE OPENING DELA y

(SA)
SO) 1
------------~.....-----------------1-9
ACTUAL PUMP INLET PRESSURE 'ffiACE
Figure 3.4 Composite Pump Dynamics
59
KEY TO FIGURE 3.4 - COMPOSITE RECIPROCATING PUMP DYNAMICS
A-Real Oscillograph of Discharge Pressure Waveform of a 3 X 6 Triplex Single-

acting Pump.
Note the tendency to follow the theoretical flow pattern 'C'.
1-Maximum Discharge Pressure, 1644 PSI.
2-Average Discharge Pressure, 1400 PSI.
3-Minimum Discharge Pressure, l 151 PSI.
B-Cylinder Pressure Waveform per stroke.

4-0vershoot Pressure, l 944 PSI.
Seen only in Cylinder.
5-Average Discharge Pressure 1400 PSI.
6-Suction Pressure 15 PSI.
C-Theoretical Flow-rate or Velocity.

Note similarity in shape to actual.
A-Points of Maxirnum Flow-rate or Velocity.
B-Points of Maximum Plunger Acceleration.
C-Points of Minor Acceleration.
D-The Absolute Discontinuity between Suction and Discharge.

Non-existant in Centrifuga! Pumps.
E-Theoretical Flow-rate in Suction.
A, B and C Points same as C above.
F-Actual Suction Pressure Waveform (Typical). Note how Acceleration

Pressure Overwhelms Flow-induced Pressure.
7-Peak Suction Pressure, 25 PSI.
8-Average Suction Pressure, 15 PSI.
9-Liquid Vapor Pressure, PSIA.
PULSATION AND CONTROL CLASSIFICATION

PULSATION PRESSURES-TYPES
Note; Pulsation Pressures in PSI:

(DV) Discharge Liquid Velocity (SA) - Suction Acceleration
(DO) Discharge Valve Open (SO) - Suction Valve Open
(DA) Discharge Acceleration (VA) - Vertical Acceleration
(SV) Suction Liquid Velocity (PS) - Pipeline Surge
(WH) - Water Hammer
60
3.1 INTRODUCTION 61
the pump inlet, excluding and the cylinder pressure resulting from PP, at the
beginning of the plunger suction stroke-these must be included in Equations
and 3. 7. A pressure transducer placed in the cy!inder head will revea! the
pressure, top of Figure 3.4, for the entire cycle but the low pressure
variation is difficult to observe because the device must be of
amplification of both extremely low then high pressures.
In the figures that follow, the pressures ínvoived can be classified as; "static"
that remain somewhat constant, namely; Ps,, P,.,, PP; then those that vary
rotation, and Psac· The least understood is probably Psac and PP.
In order to accelerate the column of liquid in the suction pipe at the start of
each plunger suction there must be present in the pump cy!inder sufficient
potential pressure to overcome the inertia of that column of liquid. lf sufficient
is not available in the cylinder the acceleration wave (always in the order
of 50 to 150 Hz and not related to pump speed) may extend into the vapor
pressure region, resulting in cavitation. It is interesting that high suction pressure
does not necessarily reduce the effects of acceleration-it just moves the gener-
ated waveform higher and away from the vapor pressure region. The high initia!
acceleration pressure, Psac, in 3.5 is the peak pressure demanded to
accelerate the slug of liquid in the suction piping system. H should not be con-
sidered a loss since it generates a standing wave and most of the energy expended
acceleration is returned to the liquid during the deceleration.
The maximum acceleration, cakulated by Eq. 3.4, occurs at the start of each
plunger stroke (at points "C") for a triplex single-acting pump, Figure 3.4 and
3.5. It is imperative that the value of Psac (which is actually initiated in the
cylinder) must fall rapidly to a lesser value until the suction valve sees a greater
'-'F"'"''""' force or pressure unbalance. Because it requires a finite time (and an
apparent more or less constant time for any type and size valve, probably con-
tributing to the consistent frequency of 50 to 150 frequency of the typical suction
standing waves noted) for this pressure to fall so that the valve can open, a valve
vv"'"'""' delay occurs which in effect reduces the displacement or volumetric ef-
ficiency, Incidentally, Psvo is shown as a negative value because that degree of
pressure was required to accelerate the column of liquid, leaving an instantaneous
low pressure in the chamber that quickly responds to the reaction to form a sharp
spike or surge, resulting in a standing wave with an average value of
pressure. Note that in Figure 3.3. the decompression to a deceleration
wave-'retuming' energy) adds impetus to the further drop in pressure after the
suction valve opens, ending up at a value of Ps"º' Assuming that the pump is
at slow speed the delay time is a small percentage of the total stroke but
at high speed that delay time becomes pronmmced, creating a significant drop in
volumetric efficiency. High speed operntion, in terms of reversals per minute,
greatly increases such disturbances as cavitation and loss of efficiency. Over a
long period of incipient cavitation, actual damage to the Hquid plungers and
va!ves, with deep pitting, can quietly take
The high frequency standing waveform described above dominates the
of the entire suction pressure trace in reciprocating pumps. Because of the rela-
CHART 1 - Slmplex SA Pump wltll JO degree d•layii
~
250
200
150
100
~
a.
CJ 50
~IC o
1
~
50
100
150
200 . D
D:E 1 E:J<'
E F
F:A
A
A
250
1
o 30 60 90 120 150 180 210 240 270 300 330 380 390 420
DEGREES ROTATION
A- Start of Mechanical Plunger Suction Stroke;
D:E - Suction Valve Closure Delay. Suc~
A:B - Oischarge Valve Closure Delay, Discharge Liquid trapped in Chamber;
tion Liquid returned to manifold;
B- Discharge Valve Closes
E. - Suction Valve Clases;
B:C- Decompression of Trapped Liquid. in Clearance Volume;
E:F-Compression of Trapped Liquid;
C - Suction Valve Opens;
F - Discharge Valve Opens;
C:D- Suction, Cylinder filling with Liquid;
F:A- Liquid Volume delivered to outlet;
O - Start of Mechanical Discharge Stroke
3.1 INTRODUCTION 63
·ively low friction pressure drop at the inlet as compared to the discharge, be-
ause of the usual short length and relatively large diameter of the suction pipe,
he ftow-generated wave has little chance against the acceleration disturbances. It
as been claimed that the liquid velocity in any suction pipe should not exceed
.305 mis ( 1 ft/s), very good advice if it can be obtained.
Suction Requirement Factors

ne of the most important factors in the suction requirements of reciprocating
mps, high pump speed (over about 200 fpm),-requires high valve spring PO-
YA of about 8 in order to reduce valve delay and still retain a good degree of
olumetric Efficiency, and which in turn then demands equivalently higher suc-
n pressure at any speed. This is where most pump problems in suction require-
ents usually originate. Normally, a good rule of thumb would be; one (1)
SIVA for each (50 fpm) piston speed.
Sorne applications of pumps at extremely low speed, 50 fpm or less, one can
metimes get by with no suction valve springs, obviously with very low suction
essure.
The effect of pump speed and spring POSIVA is dramatically shown in Table
l. The table lists ali operating data with pumps of different speed but with
unger diameters that permit equal flow-rate of 135 GPM average.
Figure 3.6 shows the same c;lata plotted and it reveals that the pump speeds of
and 360 RPM would result in marginal or complete failure of performance
cause the TCP in those cases are close to or below zero (negative).
It is undesirable to use different POS IVA springs in suction and discharge
ves because of the chance of interchanging such springs with possible drastic
ults in performance.
Experience has shown that high POSIVA suction valve springs are usually
sirable (and TPC affordable) for high suction pressure applications. If the high
ction pressure is turned in to the main pump equipped with .low POSIVA
rings, before the main pump is started, continuous ftow may be established
ough the main pump with the valves being held open. At the start-up of the
ain pump, it may be difficult for it to ''pick up prime.'' When possible, the
.ain pump should be started just before the application of high suction pressure.
e Chapter 11, Applications, for other dangers of extra-high suction pressure.
The characteristic curve of inlet pressure vs volumetric efficiency (perfor-
nce), Figure 3.2 is such that if one tries to operate near the low end of that
essure, the deterioration of the performance with low VE rapidly increases
"th slight change in inlet pressure.
During regular operation an unpredicted drop in suction pressure by, say, in-
ase in liquid temperature, obstructions in the suction pipe such as scale or
ttled particles, more than usual dissolved or entrained gas, tolerance of replace-
ent parts, especially valve spring rate and lnstalled load, change of valve type,
c. may force the operation into an untenable point on the curve with possible
mp delivery stoppage.
TABLE 3.1 Pump Speed vs TCP
Pst = O, Psz = O, Pa = 14.7 10 FT 4" SUC Pipe, 135 GPM
Degrees Vel Accel GPM P.,.. P.if P_,.,A TCP
3.375 X 6 TRIPSA SA 200 RPM

1 2.89 43.2 117 0.06 0.06 5.66 15.35
39 3.65 0.312 148 0.09 0.10 0.04 9.73
40 3.65 -1.13 148 0.09 0.10 -0.15 9.54
60 3.38 -31.4 137 0.08 0.09 -4.12 5.57
61 3.41 29.9 138 0.08 0.09 3.92 13.61
80 3.65 1.13 148 0.09 0.10 0.15 9.84
81 3.65 -0.313 148 0.09 0.10 -0.04 9.65
120 2.86 -44.8 116 0.05 0.06 -5.87 3.82
121 2.89 53.2 117 0.06 0.06 6.97 16.67
3 x 6 TRIPSA 250 RPM
1 2.86 53.4 116 0.05 0.06 7.00 16.69
39 3.61 0.385 146 0.09 0.10 0.05 9.74
40 3.6 -1.4 146 0.09 0.10 -0.18 9.51
60 3.34 -38.8 135 0.07 0.08 -5.09 4.60
61 3.36 36.9 136 0.08 0.09 4.84 14.53
80 3.6 1.4 146 0.09 0.10 0.18 9.87
81 3.61 -0.387 115 0.05 0.06 -0.05 9.64
120 2.82 -54.3 114 0.05 0.06 -7.12 2.57
121 2.86 53.4 116 0.05 0.06 7.00 16.69
2.625 x 6 TRIPSA 320 RPM
1 2.8 66.9 114 0.05 0.06 8.77 18.46
39 3.53 0.483 143 0.08 0.09 0.06 9.75
40 3.53 -1.76 143 0.08 0.09 -0.23 9.46
60 3.27 -48.7 133 0.07 0.08 -6.38 3.30
61 3.3 46.3 134 0.07 0.08 6.07 15.76
80 3.53 1.75 143 0.08 0.09 0.23 9.92
81 3.53 -0.485 143 0.08 0.09 -0.06 9.63
120 2.76 -68.1 112 0.05 0.06 -8.93 0.76
121 2.8 66.9 114 0.05 0.06 8.77 18.46
2.5 x 6 TRIPSA 360 RPM
1 2.86 76.8 116 0.09 0.06 10.07 19.79
39 3.61 0.55 146 0.09 0.10 0.07 9.76
40 3.6 -2.82 146 0.07 0.10 -0.37 9.31
60 3.34 -55.8 135 0.08 0.08 -7.32 2.38
61 3.36 53. l 136 0.09 0.09 6.96 16.66
80 3.6 2.01 146 0.09 0.10 0.26 9.95
81 3.61 -0.56 146 0.05 0.10 -0.07 9.58
120 2.82 -78.2 114 0.05 0.06 -10.25 -0.56
121 2.86 76.8 116 0.00 0.06 10.07 19.70
64
3.1 INTRODUCTION 65
3 x 6 TRIPLEX SA - TCP
VS SUCTION PRESSURE
30--~~~~-..-~~~~---,..--~~~~-.-~~~~--....~~~~---.
1
1
...............·-··r··-··-··-......... --··¡-.1 .· -·-· -· -· -· . ··--··-··1·-··-··-··..··-·-··-··..·-··--
-...... !
1 ¡
,
. 1
................... 1....................................._...
10 5 o -s -10
SUCTION PRESSURE, Ps
Figure 3.6
he data for all of the charts to be discussed in this Chapter was taken from
ies of over 300 tests on a 3 x 6 Triplex Single Acting pump at all speeds
360 RPM (360 fps). The single low point at 82 percent indicates the mini-
limit of VE at which the pump would run-below that, complete stoppage
livery occurred .
. ost typical suction systems present a seldom recognized fact that the source
ergy for the pump suction is at the supply tank, a finite distance from the
, thereby introducing a time factor.
consider an example with water as the liquid (Fig. 3.7), assume that the
required to accelerate a slug of liquid at the beginning of a stroke is a generous
Üisecond (1 ms = 0.001 s). Even with a normally "short" suction pipe of
, it requires about 4 ms (O. 004 s) for a pressure wave to travel from the source
· ergy (tank head) to the pump inlet, about 3 ms too late! Because of the com-
tions and variations of the acceleration-generated waveforms in the suction,
difficult to make precise predictions of how a pump will perform in actual
·cations.
Valve Leakage and Slip

rn valves may continuously leak liquid under high pressure and all valves
ience "slip" whereby a small amount of liquid trapped between the valve
66 SUCTION REOUIREMENTS FOR RECIPROCATING POWER PUMPS
t Location of transducer
14 f t used to record oscill-
TANK shown below.
000
TRIPLEX
PUMP
Least póssible time in which

pressure (head) at tank can be
transmi tted to pump is at the
speed of sound in water, or
20/4725= 0.004 s
CENTER CYLINDER
PRESSURE
SUCTION PRESSURE
Acceleration ''spike''
generated by start of
suction stroke shown in
cylinder trace above. One revolution,
0.162 s at 370 RPM
At points of maximum acceleration, the

time involved is almost infinitesimal-
in this case not over 0,001 s.
Figure 3.7. Energy transit time. Showing why higher tank pressure or head alone v/
not cure acceleration problems in pump suction.
and seat immediately before closure must be displaced, part toward the dischar
and part backward into the cylinder, tending to reduce VE.
Moreover, Ieaking valves can cause an unusual effect in medium and hi
pressure pumps. Despite the typical usually downward turn of the VE Curve w·
Speed increase, as shown in Figure 3.2, it is sometimes noted that the Cut
will go through a slight rise early in a Test. This is explained by the fact t
3.1 INTRODUCTION 67
when pump speed is increased while the leakage rate remains constant, the over-
all effect is such that the increased delivery becomes an asset over the constant
leakage until such speed reaches the point where the detrimental effects of speed
begin to show up in the performance. This "hump" appears in many of the
charts.
Typical Suction Pressure Waveforms

he Oscillographs shown in Figure 3.8 each depict, on' the bottom trace, the
ction pressure waveform for one revolution with one cylinder pressure wave-
rm immediately above it. They were selected from the records of an extended
st program on a 3 x 6 (and a 2 x 6) Triplex Single-acting Plunger pump,
nsisting of over 300 separate runs with water under varying conditions.
Note the complex and variable nature of the suction waveforms (actually "sig-
tures" because each waveform differs in shape) which always exhibit an un-
edictable high frequency in the range of 50 to l 00 Hz regardless of the speed
d size of the pump. But on each test they invariably retrace themselves exactly
er severa! cycles. These complex frequencies are generated by the shock wave
sulting from the suction valve opening and the high acceleration surge, system-
oustic-related and excited by, but not related to, the pump RPM. The fre-
ency, amplitude and shape of these waveforms have no relation to the power
the pump from which they were obtained-3 BHP or 3000 BHP.
In each of these Graphs the actual average of the complex wave with its
cursions, which mostly extend for ten to 50 psi above and below that average
always in position at or near the actual suction pressure, P,. A rather large
tension of these excursions can appear in a "good" suction waveform, Graph
, for example-in fact, a badly cavitating pump may produce a waveform with
ry small excursions. Significant in Graph 17 is the trace that begins to show
existence of cavitation occurring in the pump inlet as indicated by the typical
nded bottom negative excursions due to formation of vapor (cavitation) and
ensive sharp upward spikes as the result of surges generated by the collapse
the cavitation voids.
Relation of Pressure Waves to Pump Cycle
Figure 3. 8 observe in these waveform charts the obvious relation of suction

turbances to the stroke position as revealed by the upper trace of the pressure
the third cylinder, for example, of a triplex pump. If a suction disturbance
urs nearly beneath the start of the suction stroke in the cylinder pressure
eform (at B) as shown in the pressure traces, it can be attributed to an ac-
ration disturbance. If it occurs near (A) the center of the delivery stroke it
be construed to be a :flow-induced disturbance. It has been observed that
Y one or the other type of disturbance can appear in any one trace because
(11
m 1 Dlschorgo 3 X 6 TRIPLEX SINGLE ACHNG PUHP
(tyliOO.rl DECREASING SUCH!lN PRES:S:URE. f'SIG _ ___...,.
~~:~~G
riJ 'r-L_j-
+ 5 + 0 +5 -5
g n 1 r ,---·L._j- 9
le f ~
..,
_¡- ,- r' , ~I· - ' .
{"Tll"""'I- " ·-- 1¡¡.,
11~••
,.,_,.,.¡1..:....JL:...........o .............__..-' . . llo•.... Jii
~1/11-"" •ir"'
111. 1 1
~-...-
!. '·
º ..
-: :M - - ----·---------
] ~ V . E. 90 . 4 9!L 9 97 . B 98 . 6 91:1 . 4
.~ i!! 6 10 -L__J-lt 1!1
~ g ]!] r--· r----i
N §ll J J j
1 ~j t \L •• ,¡.\.¡,_ 11.
1 1· ~~'
~ .:J 1 91 . O 1 97 . 3 1 9!Ul 1 96 . 4 1 93 . l
• ~·!
~ g íi~
~ ~
!i
-!i l
1 1 96.3 1 96.0 1 96.3 1 96.0
j~
~ JU !Jl\]~nli~-¡j'1··1 i\f•
1
!A'! tu' V
' ~
vy
• •
;.A¡¡Jj~A'fJiJ•.1\JJHI
' \¡
tfl"·•r~A
V•·111!¡¡1 nn.i,, l!ll ~-";~·
'\fi/'FVfJO
;¡ 95.5 94.11
Figure 3.8 Actual Suction Pump Pressure Waveforms

3.2 PERFOFlMANCE CURVES 69
one disturbance (acceleration) destroys the ability

recover and initiate another ( flow-related) disturbance in the
interval between them in high speed pumps and vice versa.
PERFORMANCE CURVES
3.4 and 3.5 show how valve opening and closing delays, caused by the
compressibility of the liquid or the presence of dissolved or entrained air or gas,
· the elasticity of the system, and how such delays reduce the Volumetric
of the pump by "robbing" part of the actual stroke. However, in
rleiffee-c1t-rot2tt1cm ' based charts, it should be well recognized that these delays
exaggerated as a proportion of the pump revolution. For example, the total
degree delays shown only about 0.5 inches of a 6 inch stroke pump.
real VE would be (6" - OS')/6 11 = 92 percent, not (!88º - / 180º = 83
Method of Rec:ording Actual Mechanlcal

;,End~ofsStroke
is almost impossible to determine the mechanical end-of-stroke from the re-

oscillographs and for precise evaluation it is necessary to provide sorne
of marking the graph at these points. One simple means is by the use of
"optical phaser", a simple solid-state device that detects the presence of light
an electrical output.
The device is temporarily mounted and aimed at sorne part of the reciprocating
say the exposed plunger in the pump cradle. A stamp-sized piece of
paper taped to the plunger will enter the field of the phaser at the end of
e generating a reflected light pulse on the graph (on a multi-channel
ope) that will precisely mark the desired position. The other extreme of the
roke can then be scaled from the pressure wave data. The device will normally
ernte from light in a well lit room but it may be necessary to provide more
ght on the "target", preferably by a flash-light with 'clean' DC light. See Fig-
Pertormance-POSIVA and RPM
Figure 3.10 note that at the relatively slow speed of 200 RPM (200 fprn piston
suction valve spring characteristics are of little consequence and an ex-
low suction pressure can be tolerated, and less discussion is required.
pump suction valve spring POSIVA, regardless of spring rate, has an enor-
us effect 011 the performance of a pump at high Figure 3 º l l shows the
extreme effect at the upper speed limit of this particular pump {360 RPM) with
70 SUCTION REQUIREMENTS FOR RECIPROCATING POWER PUMPS
1
·~·-······-····r .. ·-·
Pump stroke, s (CE) Pump stroke, s
1
1
1
blip, end-of-stroke 1
Oelay, suction 1 1 1 1 t Oelay, discharge
l 1 ·- 1 1 1 1
valvs closing ,........_,... valve closing
ki.T Oelay, compression

: Delay,
decompression
1
1 i
•••• L • - • ... ·~ ..... .l. - .,.
¡. 1
1
1
1
¡ 1
1
1
1 1 1
]------.. ¡----· - •. ,. •· r -
1
1
' j - - - - •• -- - - ,
1
1
1
1
1
J______ _L.- ------

1
.:.... _,
l ¡
1
Figure 3.9 Use of Optical Phaser to Upper Trace Mark End of Mechanical Stroke.
one type of valve with various POSIVA springs. The author has observed that
this spring effect is most always ignored in pump testing.
By any suction test procedure the pump in question would show different and
widely separated "required" suction pressures, depending solely upon the com-
bined effects of spring POSIVA and pump speed.
The use of a higher POS IVA results in a higher VE but at the expense of ~
higher suction pressure, a typical compromise. Also shown is this typical como
promise associated with spring POSIVA; i.e. lower volumetric efficiency with.
lower suction pressure or higher efficiency with higher POS IVA-but never with
cross results. Pointed out is also the old rule-Use 'light' (POSIVA) springs foi;
low pressure and low speed and 'heavy' (POSIVA) springs for high speed anq
high pressure.
3 x 6 TRIPSA - 200 RPM
INLET PRESSURE VS VE AT VARIOUS POSIVA
90~··························· .. ····+······· .. ··········· .. ·············+·· .. ··················· .......... +................... ..
BB•· ..................................................................... +················ .................. +···················
66 ................................................................
84 ................................................................ +········........................... +.. ·············· ..······""''"""l""'""""""
-5 o
INLET PRESSURE, PSIG
.._ 0.15 POSIVA ..,.. 2 POSIVA
·4'· 4 POS!\/ A •.,. 8 POSIVA
Figure 3.10

INLET PFIESSURE VS VE A T VARIOUS POSIVA
9!l ............................................................... +..................................+ .................................¡................................ .
88 t ................................ t
84 , ...................................t .. ·• ............................................................................................................................
821--~~~~~ ......~~~~~+-~~~~~-1-~~~~~-1-~~~~~--l
-10 -5 o 5 16
INLET PRESSURE, PSIG
-+- 0.5 POSl\IA ·• 2 POSIVA
·•· 4 POSiVA _.,_ 8 POSIVA
Figure 3.H.
71
SUCTION REOUIREMENTS FOR RECIPROCATING POWER PUMPS

iNLET PRESSURE VS VE AT VARIOUS POSIVA
98 t'""""º'''º'''ºº"'"ººº"ºººº'+•••""''''ºº'ººº'ºº'"""º"ººf'ºº'ººº"'ºº'''"ººº"º'ºº""'' ' " ' " ' " " " ' " º " " ' ' ' ' º ' ' º ' ' º •••o••o•••oo••••oo•oo•o•••••••o••
96 ~ .....................-...........-¡ ................................ +·············..············..··· ································ ...,.............................
94 r------·---·------,==:=:=~··~~-··-~"·r=-=-~~.:~:~~-~·=·-..::-:-~=~::J;::~:::::::;;~;;;;,;;;;;;;;;:t:::;;;;;;;;;~;;::::::~~:::::i
------ .. ~::·····
92 ~ .................................~:::::::.::::::::::::::::::::::::::t..:.::.:.:.:.:.:..::.:.:..::.:..::.:.:...::..::.:..: .:.:..~.1~:-::::~::.::.~~~~··-~uuJ
........... ,
.............................,..
90 ................................ ,................................................................................................................................
116 ................................................................................................................................................................ .
as ................................................................................................................................................................
84 .................................................................................................................................................................
a21--~~~~-+~~~~~-+-~~~~--1r--~~~~-t-~~~~--1
-10 ~ o 5 10
11\!LET PRESSURE, PSIG
....,_ 0.5 POSIVA ·• 2 POSl\IA
-+· 4 POSIVA -11- 8 POSIVA
Figure 3.12.
3.2.3 Performance-Reduced Plunger Diameter

Figure 3. 12 shows the performance of the same pump as in Figure 3. 1O and at
the same conditions, but with the pump equipped with 2 inch diameter instead
of 3 inch diameter ph.mgers. Particularly in such pumps designed for the use of
replaceab!e pistons or plungers from a range of different sizes, a very common
practice, the increased clearance-volume resulting from the use of a smalier
plunger combined with the decompression from the higher discharge pressure
results in a lower Volumetric Efficiency but the reduced flow-rate contributes to
a very "flat" curve at extremely low suction pressure.
3.2.4 Volumetric Efficiency Statistics

In a particular series of over 300 test records, the mínimum practica! operating
Volumetric Efficiency was noted to be never less than about 85 percent. Below
that minimum value complete loss-of-prime is usually experienced, indicating
that a critica! point exists and it would be folly to try to operate near that point-
it is in the critica! position where even a slight drop in suction pressure would
result in a rapid and dangerous drop in performance.
3.2 PERFORMANCE CURVES 73
SHORT SUCTION LONG SUCTION

6 ft of 6"p1pe 90 ft of 4· pipe
100 J ,_,__r l_J \ J 1
#13E VE 97.8
1~-P·v'~·V
#1005 VE 98.1
1
360
RPM
#16E VE 96.7 #1008 VE 96.0
ALL AT t4.9 PSIG SUCT\ON PRESSuR~
Figure 3.13. Short vs. Long Suction Line
3.2.5 Short Versus Long Suction Line

Figure 3. 13 shows selected graphs taken from a series of tests on suction require-
ments of the same pump with water under the same operating conditions but
with a long (90 feet long, 4 inch pipe) versus a short (4 feet long, 6 inch pipe)
suction line on the 3 x 6 Triplex Single-acting pump.
Another factor to be considered is the effect of liquid and mechanical reso-
nance of that length of suction pipe involved. lt appears that in this case at 100
RPM (Figure 3.13, Graph 1005) such resonance that appears in the waveform as
standing waves is beneficial to the Volumetric Efficiency, a paradox that contrib-
utes to complications. See discussion on Pipe Vibration in Section 11.3.
TABLE 3.2 TCP by Degrees of Rotation, 10 Degree Delay
.
.. .
DEGREE Pa + Pst: + Psz: Pp1 flvp 1 Psf lACCElJ¡ Psác TCP
1! 14.7 Oi o: 4! 1.5, 0.23 -36.6' -4.80 13.77
5: COr-tSTANT-· - - 9.2; 0.22 ·-38.9 -5.10 14.08
91 9.21 0.21 -40.9 -5.36 14.35 **
'
101 1 1 9.21 0.261 291 3.801 5.14
39. 1 9.21 0.3 0.26' 0.03 8.87
40, 1 1 9.21 0.3 -0.91 -0.12 9.02
65 ! i 9.21 0.27 -31.9 -4.18 13.11
66: 1 9.2! 0.29 18.2 2.39 6.52
so: ! 9.2: 0.3 0.91 0.12 8.78
81' 1 1 1 9.2! 0.3 -0.26! -0.03 8.93
129 1 : 1 9.2! 0.21 -40.9 -5.36 14.35
130: ! 9.2' 0.26 29 3.80 5.14
159· ¡ i 9.21 0.3 0.26 0.03 8.87
160' ! 9.21 0.3 -0.9 -0.12 9.02
1801
'
! 9.2i 0.28 -25.7 -3.37 12.29
*(ACCEL) shown for reference only.
**Maximum Ps~c (required) at 10, 130 and 250 Degrees.
Note; Minus (-) sign of Psac by Eq (3.4) used in Eq (3.1)
becomes Plus(•).
A somewhat redeeming feature in long suction lines is the fact that pipe fric-
tion pressure losses are out of phase with acceleration pressure and they are not
added to the already high value of acceleration pressure. This is also shown in
Table 3.2.
3.2.6 Acceleration
The so-called acceleration head or pressure in a pump suction is sometimes er-

roneously considered as a "constant" demand of the pump for an entire cycle or
revolution (shades of Centrifugal terminology) but, the maximum demand for
high acceleration head or pressure occurs for only an infinitesimal instant three
times per revolution in the case of a triplex pump, for example, as shown in
Table 3.2. For a large portion of the stroke-time, the energy imparted to the
liquid mass at the beginning of the suction stroke is being returned to the mass
during the deceleration process, representing a low average demand. The average
energy represented by the Psac values is extremely low, demanding little energy
from the source.
To demonstrate how rapidly the present concept of acceleration of a pump
loses validity and viability with long suction lines, the pump represented in Fig-
ure 3.13, Chart 1008, with a 90 foot long suction pipe would require a calculated
pressure of over 137 PSI (Eq. 3.3) to overcome acceleration pressure. However,
the pump ran with good efficiency with only 4.9 PSIA suction pressure at the
pump inlet. lt is evident that the concept neglects such factors as pipe resiliency;
dissolved gas or air, liquid compressibility, energy conservation and probably
3.2 PERFORMANCE CURVES 75
others. Accordingly, to convert TCP test data to an actual complex suction sys-
tem great caution must be exercised.
It seems that the effects of acceleration in the suction pipe are limited to a
maximum equivalent of about ten feet of pipe, even though the standing wave
generated by the pump action is seen for the entire length of the pipe system.
A more accurate and realistic TCP can be determined by the described test
roethod using the pump inlet pressure reading, Ps Formula (3.7), because most
of the energy factors, excluding P,.c and PP are integrate!f into a final figure by
the damped suction pressure gauge. The final figure must include a further de-
mand, PP, to open the spring-loaded suction valve.
Figure 3.13 supplies overwhelming evidence that "acceleration" is not the
demon for which it has been accused and that the presently accepted NPSH For-
mulas should be used with caution and that Formula (3.5), as used on actual
suction pressure measurement is most accurate.
IMPORTANT
It is suggested that in ali calculations of Acceleration Pressure, the value of L,
length of suction pipe, be assumed to be 10 feet regardless of the actual length.
Many tests show that beyond that length the validity of the Equations begin to
become questionable.
3.2.7 Computer Calculations
On account of such a large number of calculation points ( for 360 degrees Qf

rotation) involved, the data can be programmed giving outputs like Table 3.2 and
3.4.
Suction Stabilizer
Since it has been inferred that additional head alone may not always "cure"
acceleration problems of a long suction line, by the simple expedient of placing
a "lumped" volume in the form of a small closed flow-through vessel, as close
to the pump inlet as possible, will negate most of the acceleration effects by
fooling the pump into thinking that, for repeated instances at the moments of
maximum demand, the supply tank or source is at the pump inlet. The introduc-
tion of such a relatively large impedance, called a Suction Stabilizer, Figure 3.14
into the system also prevents the pump-generated acceleration disturbances from
being transmitted to the suction line by an acoustic filtering effect.
Experience has shown that if the Suction Stabilizer "liquid volume" is de-
signed with adequate liquid volume greatly improved pump performance can be
expected in marginal systems. See Chapter 4 for additional information on Suc-
tion Stabilizer design.
TABLE 3.3 EQUATIONS
Piston-Plunger Velocity, ft/s;

VP 9 = 0.042sw [sin 0 + (sl4LJ sin 20]
Model (9°) = 0.042 X 6 X 20.9 X [0.156 + 0.083 X 0.309] = 0.956
(10°) = 0.042 X 6 X 20.9 X [0.174 + 0.083 X 0.342] = l.066
Suction Velocity, ft/s; 4" Suction pipe.

V,.9 = L VPaCD~ID;)
Model (9º) = 3.64 x 0.563 = 2.05
(IOº) = 4.64 X 0.563 = 2.61
Piston-Plunger Acceleration, ft/s 2

ªPa = 0.042swz
+ (s/2LJ cos 20]
[cos 0
Model (9°) = 0.042 X 6 X 20.9'2 [0.988 + 0.167 X 0.951] = 126
(10°) = 0.042 X 6X 20.9'2 [0.985 + 0.167 X 0.940] = 125
Suction Acceleration, ft/s 2 ; 4" Suction pipe

ª•a = 2.ap9 (D~ID";)
Model (9º) = -53. l x 0.563 = -29.8
(10°) = 51.5 X 0.563 = 29.0
Suction Acceleration Pressure, PSI;

Psaca = 0.0069La,9 p
Model (9°) = 0.0069 X lO X (-29.8) X 1.9 = -2.59
(10°) = 0.0069 X lO X 29 X l.9 = 3.80
TCP by Degrees Crank Rotation, PSIA;
Model (9º) = 14.7 + O + O - l.5 -4 -(-2.59) = 6.61

(10º) = 14.7 + o + o - 1.5 -4 -3.8 = 5.40
76
TABLE 3.3 EQUAT!ONS {Contínued)
Inlet Pressure (Test), PSI;

TCPe = Pª + Ps -
Model (9º) = 14.7 - 0.2 - 1.5 -4 - (- 14.35
(10º) = 14.7 - 0.2 - LS -4 -3.8 = 5.2
Pressure at Pump Inlet, PSIG

= ± ± Ps: -
= o+ o- 0.2 =- 0.2
w = 2nRPM/60 (3.9)
Model = 2 X 3.1416 X 200/60 = 20.9
Valve opening surge;

(3.
Model (9°) -5.36 + 4 = -1.36
Mode! (10º) = 3.80 + 4 = 7.80
SYMBOLS
Description Units
length, suction pipe ft
length, pump connecting rod in
revolutions per minute RPM
pressure, acceleration, suction PSI
pressure, friction loss, suction PSI
pressure, suction PSIG
pressure, pump suction (inlet) PSIG
pressure, velocíty, suction PSI
pressure, liquid vapor PSIA
pressure, tank head FT TO PSI
POSIVA (Pounds per Square Inch off Valve Area)
pump stroke in
specific gravity, liquid
TABLE 3.3 EQUATIONS (Continued)
Symbol Description Units
V velocity of flow ft/s

Ds Diameter, suction pipe, m
N Revolutions per Minute, RPM
NPSH Net Positive Suction Head, ft
NPIP New Positive Inlet Pressure, PSI
TCP Total Cylinder Pressure, PSIA
p Density, (Water l. 9)
0 Crank Rotation Degrees
TABLE 3.4A Pump C-3" x 6" Triplex Single-Acting 300 RPM, 1650 psi
Selected degrees
PUMP ANALYSIS SUMMARY
PUMP SPECIFICATIONS
PUMP MODEL triplex
NUMBER OF PUMPS 1
PHASE ANGLE BETWEEN PUMPS o
PUMP TYPE-NUMBER OF CYLINDERS 3
PISTON DIAMETER-IN 3.0000
DOUBLE ACTING PUMPS ONLY
PISTON ROD DIAMETER-IN 0.0000
CONNECTING ROD LENGTH-IN 18.0000
STROKE-IN 6.0000
CRANK ANGLE-DEG 120.0000
CRANK ROTATION OVERRUN
PISTON/VALVE PUMPS ONLY
LIFT-IN 0.0000
DELAYED SUCTION VALVE CLOSING-DEG 0.0000
DELAYED DISCHARGE VALVE CLOSING-DEG 0.0000
COMPRESSIBILITY DELAY-DEG 0.0000
INLET PIPE DIAMETER-IN 4.0000
OUTI.ET PIPE DIAMETER-IN 2.0000
OPERATING SPECIFICATIONS
FLUID-SPECIFIC GRAVITY 1.0000
ATMOSPHERIC PRESSURE-PSIA 14.7000
SUCTION STATIC PRESSURE-PSIG 20
SUCTION FLOW FACTOR-FT 20
SUCTION ACCELERATION FACTOR-FT 20
DISCHARGE STATIC PRESSURE-PSIG 1.ooci
DISCHARGE FLOW FACTOR-FT 3,000
DISCHARGE ACCELERATION FACTOR-FT 20
ALLOWABLE PULSATION-PSIG 50
78
TABLE 3.4A Continued
OAMPENER PRECHARGE-PSIG 600
OPERATING SPEED-RPM 300
PUMP ANALYSIS RESULTS
TIIEORETICAL R.OW-OPM 165.2354
MAXIMUMS
R.OW-GPM 17S.42S7
R.OW-% 6.1672
DISCHARGE PRESSURE-PSIG 1,720.8189
DISCHARGE PRESSURE-% 104.4938
MINIMUMS
R.OW-GPM 137.2JS6
R.OW-% -16.94S4
DISCHARGE PRESSURE-PSIG 1,463.0SlS
OISCHARGE PRESSURE-% 88.8413
OIFF BETWEEN MAX ANO MIN
FLOW-OPM 38.1902
FLOW-% 23.1126
FLOW-GALLONS O.OM687
EFF FLOW OUR DISCH OF PEAK-GPM 12.S8S8
OISCHARGE PRESSURE-PSIG 257.7674
OISCHARGE PRESSURE-% IS.óS2S
PUMP STABILIZER ANO OAMPENER SIZING
PASAFE LIQUIO VOLUME-GALLONS 16.S23S
PASAFE GAS VOLUME-GALLONS 0.S-lS8
TABLE 3.4B Pump C-3" X 6" Triplex Single-Acting 300 RPM, 1650 psi
PUMP SUCTION ANALYSIS
CRANK PUMP-INLET CYLINDER·HF.AD END CYLINDER-CRANK END
ANGLE
VELOCITY ACCEL FLOW VELOCITY ACCEL FLOW VELOCITY ACCEL FLOW
DEO FPS FPS2 GPM FPS FPS2 GPM FPS FPS2 GPM
SUCTION DISCHARGE
1 3.55E ()() 7.96E 01 l.39E 02 l.60E-Ol 2.88E 02 3.52E ()()
40 4.4KE ()() -2.09E ()() l.75E 02 5.70E ()() l.96E 02 l.26E 02
61 4.IKE ()() 5.50E 01 l.64E 02 7.43E ()() 9.78F. 01 l.64E 02
81 4.4KE ()() -5.76E-OI 1.75E 02 7.96E ()() -l.02E ()() 1.75E 02
121 3.55E ()() 7.96E 01 l.39E 02 6.ISE ()() -l.46E 02 l.35E 02
160 4.4RE 00 -2.JJ9E 00 l.75E 02 2.26E 00 -2.00E 02 4.99E 01
DISCHARGE SUCTION
181 4.18E 00 5.50E 01· l.64E 02
201 4.4KE 00 -5.76E-Ol l.75E 02
241 3.55E 00 7.96E 01 l.39E 02
280 4.4KE 00 -2.09E· 00 1.75E 02
301 4.lRE 00 5.SOE 01 1.64E 02
321 4.4KE 00 -5.76E-01 l.75E 02
1 l.42E 0.1 3.18E 02 1.39E 02

40 l.79E 01 -8.34E 00 1.75E 02
60 l.66E 01 2.31E 02 1.62E 02
81 l.79E 01 -2.30E 00 l.75E 02
121 l.42E 01 3.18E 02 l.39E 02
160 l.79E 01 -8.34E 00 l.75E 02
180 l.66E 01 2.31E 02 l.62E 02 2.06F. 02
DISCHARGE SUCTION
201 1.79E 01 -2.30E 00 1.75 02 2.38E 00 2.00F. 02 5.23E 01
241 1.42E 01 3.18E 02 1.391! 02 6.31E 00 l.41E 02 l.39E 02
280 1.79E 01 -8.34E 00 1.75E 02 7.96E 00 -3.71E 00 1.75E 02
300 1.66E 01 2.31E 02 1.62E 02. 7.37E 00 - l.03E 02 1.62E 02
321 1.79E 01 -2.30E 00 1.75E 02. 5.59E 00 -2.0lE 02 l.23E 02
79
Figure 3.14. Suction Stabilizer (White Rock Engineering, Inc.)
3.3 TESTING PUMPS FOR SUCTION REQUIREMENTS
3.3.1 lntroduction
For reciprocating pumps ali reference to suction requirements as a discreet "num-
ber" should be replaced by a TCP (Total Cylinder Pressure) Performance Curve
(and possibly a Liquid Performance Curve for liquids other than water). The TCP
performance test will be performed under a strict set-up and procedure, using
3.3 TESTING PUMPS FOR SUCTION REQUIREMENTS 81
aerated water as the test liquid at room temperature. This curve will be a
of the calculated TCP on the X-axis versus measured Volumetric
on the Y-axis, using Equation 3.7.
This method does not require an absolute knowledge of the acceleration
. of whatever magnitude) because they and ali other losses the
valve POS IVA are integrated in the suction pressure,
If the test is performed to if the pump would at an ex1stmg or
at a system suction pressure, the of expected operating conditíons
the TCP Performance Curve constructed from the test data will give a clue
performance. This TCP Curve can also be used for comparing
of different pumps or, most important, for a yard-stick to mea-
sure the effect of the design of expendable (more-or-less rapid wearing) liquid-
such as valves, pistons and plungers, on pump performance. It is al-
to always use this performance data to predict the
to be expected on another complex system, particularly with liquids
other than water. General application data could be compiled for any pump for
conditions listed previously. Because of the general similarity in the design
, of ali nearly the same performance would hold true for any make
size of pump if ali of the previously listed ·conditions were equally and pre-
established. It is possible that a so-called "published TCPR" for one pump
>could be discriminatory and unfair to the reputation of another particular pump.
The Liquid Performance Test should follow the same procedure as the water
but will be run with any other liquid in question. The information thus
will serve as a guide to be applied to an actual system.
TCP Required, (TCPR)

If it would be deemed desirable or necessary that a specific and verified value of
be had, the performance curve could be generat~d, as stated, but instead
selecting a point on that curve at three below the maximum VE value
has been recommended ( following centrifuga! pump select a point
as close to the "knee" in the curve where rate-of-change of VE begins to
appear. It will be seen in sorne of the plotted charts that at high speeds and low
pressure a distinct is observed where there is a increase in
rate of Volumetric Effi.ciency drop or slope of the line increases rapidly. At
high speed operation, where suction pressure is extremely a prescribed
drop is to extend into the dangerous area so if a value is wanted, use that
conservative point as described as the value for TCPR. The definition for
becomes; That pressure in PSIA of water (above vapor pressure), as de-
termined test, required at the pump cylinder to fil! the cylinders sufficiently
· so that no more than a specified in Volumetric Efficiency from maximum
to that point on the curve where the first indication of a rapid change or "knee"
on the VE curve appears. See Figure 3.2.
It behooves a system designer or user to first select the lowest praciical pump
and next, to provide the greatest suction pressure possible as
~-------------------
PROXll.llTY
'-SLIP-RINO FOR li'PUT
SYITCH
ITALICS INDICATE SHAFT S TRA 1N GAUOE
DATA POINTS.
tT~-
TORQUE OETECT 1ON
REAOOUT.
Figure 3.15 Test Set-up
delineated in high head and short, ample pipe size, straight suction piping system
and with adequate pulsation control and suction stabilization. In an existing sys-
tem, the use of a suction stabilizer or even a centrifugal "booster" or suction-
charging pump may be required to solve sorne obvious problems.
3.3.3 Procedure for Testing
Sorne means of measuring the quality of performance of a reciprocating pump

must be adopted. Volumetric Efficiency, the relation of actual flow rate to theo-
retical, is most reliable and test data can be very accurately obtained without
sophisticated instrumenting.
In order to produce data of a viable nature that can be used to compare pump
performance, or as a basis for measuring the efficiency of liquid end and ex-
pendable parts design, the tests must employ a system that will provide equal
external conditions for each test. In order to accomplish this the test system must
a
be "standardized" as muchas possible and Figure 3.15 shows such complete
set-up that can best be described by the following features;
A. A closed-system liquid supply tank with variable air-pressure supply and
suction throttle valve to permit a wide choice of practical suction pressures re-
quired to develop a useful performance curve. The variable air (or hydrogen)
blanket described is entirely satisfactory as an artificial means of increasing the
effective head. The amount of air or hydrogen absorbed in room-temperature
2960
DISCHARGE --2"""6"'"'9_6,___ 2 8 2 8
PRESSURE, PSI
3175
2828
CYLINDER
PRESSURE, PSI
SUCTION
PRESSURE, PSI
(a)
SUCTION
PRESSURE AT ~--- 80 llP=38
INLET, PSI
66
(b)
Figu;e 3.H:í.High-frequency pulsations. The trace of the discharge pressure in (a) shows
that only a remnant of the flow-induced pulsations appear in the well-dampened pump. Had
this pump not been dampened, the trace would have shown the ftow-induced pulsations of
low frequency to be predominant. The acceleratíon-induced pulsations are always predom-
inant in the suction as shown in (b).
Pump Data .l<'requencies Present
4i x 9 quintuplex Rotational, 3. 1 Hz
184 RPM Cylinder, 15.3 Hz
2800 PSI discharge Flow pulse, 31 Hz
80 PSI suction Acceleration, discharge, 138 Hz
Pump well dampened Acceleration, suction, 77 Hz
water is so miniscule that the effects are barely noted over the obvious improve-
ment in performance resulting from the additional head.
B. As opposed to a centrifuga! pump with a suction pressure that remains
almost constant (with constant conditions), a reciprocating pump generates a rap-
idly varying and repeating pressure through each revolution that results in com-
waveforms as shown in Figure 3.16. Tests have verified that a judicially
damped bourdon-tube pressure-vacuum gauge will respond to an accurate average
suction pressure of P_,. An integrating amplifier used with a solid-state pressure

transducer will also show a true average of a complex waveform.
C. A suction pipe of a next-size larger than the pump inlet connection pipe
size with a length no longer than five feet presents an ideal configuration without
any complications of a complex system.
D. A thermometer well will be provided near the pump inlet to monitor the
liquid temperature. Fast response of the temperature measuring device is also
importan t.
E. NO SUCTION DAMPENER OR ST ABILIZER of any description will be
installed on the suction system (unless such device is an intimate part of the
pump) because of the unpredictable effects that would give distorted information
on the hydraulic characteristics of that pump. One could, if deemed necessary
and so noted, insert such a device to compare its effect on the pump.
F. A gas-type energy-absorbing discharge dampener will be installed close to
the pump outlet in order to suppress the pulsations on the ftow meter and the
discharge pressure gauge.
G. The displacement type ftow meter will be placed on the downstream side
of the pressure-regulating choke. The specific-volume of the metered liquid will
then match the specific-volume of the suction liquid.
H. By returning the discharge liquid to the suction tank, a desirable constant-
state suction head will be maintained for the duration of the test.
l. The use of a stroke (revolution) counter, rather than a tachometer, allows a
more precise calculation of the displacement of the pump per period of time,
recognizing that a slight variation of crankshaft speed may occur during the test
period.
J. Ali pressure-access taps will be installed on the horizontal centerline of the
pipe to eliminate the need for datum correction. Top centerline locations present
problems from separated and trapped air or gas that "cushions" the true value.
Bottom locations tend to collect sediment that could plug the passage.
3.3.4 The Myth of NPSH

Perhaps "NPSH" was overrated as a precise engineering requirement for recip-
rocating pumps, and in every application it behooves one to provide as much
suction head as possible guided by the principies mentioned and for reasons ex-
plained in the following remarks.
"NPSH" was a somewhat mythical figure that infers that a new pump may
operate at a fair volumetric efficiency with a minimum suction head. But it is a
fact that any suction pressure above that "NPSH" will:
1. Result in higher volumetric efficiency.

2. Reduce the effects of dissolved or entrained air or gas in the liquid.
3. Allow higher temperature liquid to be pumped.
4. Offset the effect of instantaneous pressure drop the valves.

5. Minimize effects introduced in multiple pump installations.
6. Reduce the arnount of air leakage through plunger or rod packing.
The following discussion deals with each of these factors in tum.
L The simple definition of suction head pressure) for a

is that it is that pressure, above vapor pressure, required to completely fil!
each cylinder with on each suction stroke.
Theory and experience show that a pump can actually lift liquid from a level
. below the pump inlet. However, the term "completely" in the definition should
be modified to indicate sorne degree of completeness. With fed pmnps,
the cylinder never reaches the perfection of 100% As the suction pressure
on a pump is decreased, there is a gradual decrease in volumetric be-
ginning at a rather slow rate wíth a fairly high head but finally a region
where the efficiency begins a rapid drop. In fact, the "NPSH" of a pump was
based on an arbitrary acceptable value of lower volumetric efficiency,
usually 3% below maximum.
On the above basis alone it is clear that up to a limit the greater the suction
pressure, the better the hydraulic performance of the pump.
2. The theory of NPSH is based on a "tme" liquid. Unfortunately, most liq-
uids being pumped contain dissolved or entrained air or gas, which makes it dif-
ficult to predict what additional suction pressure will be required to overcome these
e:ffects. Air or gas may break out in the vacuum created in the cylinder on the
suction stroke and cause erratic performance.
3. Temperature of the liquid has a decided effect on the suction pressure re-
quirement of a pump. Even water, with its relatively low vapor pressure compared
to sorne hydrocarbons, requires increasingly greater suction pressure with increase
in liquid temperature. In many pump applications there is a possibility that an
increase in operating temperature will result in performance if suction
pressure is not also increased.
4. Recent studies show that the pressure drop through the suction valves, par-
ticularly at the instant of maximum flow rate in the cycle, can be considerably
greater than the available suction pressure without having a measurable effect on
the volumetric efficiency of the pump. Cavitation, plunger pitting, poor packing
life, etc., can result, and this is another case in favor of higher suction pressure.
5. More information is needed regarding the effect of suction pressure in the
case of multiple pump installations. There is strong evidence that two or more
pumps being fed with separate suction lines from the same tank somtimes expe-
rience detrimental interaction. The interaction of multiple pumps operating from
a common header is even greater. It is logical to assume that greater suction pres-
sure would help to overcome such effects.
6. The possibifüy of air leakage through the stuffing box packing on the suction
stroke is minimized by greater NPSH.
TABLE 3.5 Solubility of Gases in Water
ft3 Free Gas/ 100 Water

Gas 32ºF 83ºF 140ºF
Air at 14.7 PSIA 2.8 l.6 1.2
Air at 29.4 PSIA 5.7 3.2 2.4
N2 at 14.7 PSIA 2.3 1.4 1.0
N2 at 29.4 PSIA 4.6 2.8 2.0
H2 at 14.7 PSIA 2.1 1.8 1.6
H2 at 29.4 PSIA 4.2 3.6 3.2
C02 at 14.7 PSIA 171.3 67.5 36.3
C02 at 29 .4 PSIA 342.6 135.9 72.6
3.5 AIR AND GAS SATURATION OF UQUIDS

In regard to the effects of dissolved air or gas on the performance of reciprocating
pumps, the following conclusions have been developed on the basis of actual ex-
perience and analysis.
l. Water exposed to any gas blanket will become saturated with dissolved gas
in direct proportion to the partial pressure of the gas. At any given temperature
and pressure, the water will hold no more than a certain amount of gas in a true
dissolved state.
The encouraging fact is that the maximum amount of elementál gas (for ex-
ample, hydrogen, nitrogen or oxygen) that can be dissolved in the water is actually
very small. (This is not true, however, of gaseous compounds. For example, car-
bon dioxide will dissolve in water, and methane in hydrocarbons, in large amounts.)
Table 3.5 shows the cubic feet of common free gas (STP) dissolved in 100 ft 3
of water at various temperatures and pressure. Note that according to Henry's law
the concentration is directly proportional to the pressure.
2. The TCPA can be improved in direct proportion to the blanket pressure
increase. One generally accepted method of measuring a pump's TCPR is to use
a closed tank with a low water level and then measure the effect of volumetric
efficiency of increased head or TCPA by applying a gas (air) pressure to the tank.
Head is decreased by either applying a vacuum to the tank or by throttling a
valve in the suction pipe.
3. The detrimetnal e:ffect of dissolved gas on the performance of the pump
cannot be improved by increasing the blanket pressure. If hydrogen-saturated water
from a closed tank at 83 ºF is lowered at the pump inlet to, say, 5 PSI below the
tank pressure, the amount of free gas that can flash is in direct relation to the drop
in pressure, or about 1.2% by volume, whether the blanket pressure is at 14.7
PSIA, 29.4 PSIA, or any other pressure. Therefore, the pump has to cope with
the same amount of free gas in either case. Statement 3 then does not mean that
the pump wm not perform better at the higher pressure, as will be explained.
3.6 DISSOLVED GAS AT PUMP INLET 87
The boron-charging systems encountered in most nuclear power plants consist

a dosed suction system with a well-elevated supply tank with a hydrogen gas
of about 15 PSI. The theoretical TCPA would be about 35 PSI, more
for a normal pump, even considering the acceleration head and com-
fl.ashing of the 2 % of free hydrogen. Once the free gas flashes, the plunger
continue its remaining stroke with "solid" water.
Natural gas is sometimes used as a gas blanket to exclude oxygen and to
a little additional head on oil-field water flood or disposal projects. The
ingredients of natural gas are methane and ethane, which are only slightly
more soluble than air in water, and the effects should be similar. There is evidence
that improved performance can sometimes be had with such gas blankets.
DISSOLVED GAS AT PUMP INLET

Given: Water at 83ºF, 15.3 PSIG hydrogen gas blanket on supply tank. V 1 = 3.6
frorn Table 3.5. At constant temperature.
P 1 = 30 PSIA P2 = 25 PSIA = 0.5067 PSIA

J 1 = J2 = l. l X 106 at constant temperature (3.11)
w
Af = 100 X . . 100 X 62.4/18 = 346.7 (3.12)
mol. wt. hqmd
Mx 1 = M P 1( -
pvp')
---¡: = 346.7(30 - 0.5067 /1.1 X 10 6 ) = 0.009296 (3.13)
Af<i = M( P2
\
- pvp)
12
= 346.7(25 - 0.5067/1.1X106 ) = 0.007720 (3.14)
Since the quant;ty of dissolved hydrogen is proportional to Mx, the calculated mols
per 100 ft 3 of water, it can be assumed that the volume of hydrogen flashed in the
pump suction at a llP of 5 PSI would be related to the ratio of Mx 2 / Mx 1 at the two
pressures.
Therefore, if the quantity of dissolved hydrogen, V1 , is 3 .6ft3 /100 ft 3 of water
3.5) at P 1
VM
= ___!__E = 3.6 X 0.007720 /0.009296 = 3.0 ft 3 (3.15)
Mx1 .
V = V1 - V2 = 3.6 - 3.0
(3 .16)
= 0.6 ft3 free hydrogen per 100 ft3 water, or 0.6%
where
1 1 , 12 = modified Henry's constant = 14.7 X H

H = Henry's constant for gas
M = moles of liquid per 100 ft3 liquid (solvent)
Mx 1 = initial moles of gas per 100 ft3 liquid
Mx 2 = secondary moles of gas per 100 ft3 liquid
P 1 = initial pressure, PSIA
P 2 = pump inlet pressure, PSIA
Pvp = vapor pressure of liquid
V = volume of free gas, ft3
w = specific weight liquid, lb/ ft3
3. 7 NOTES REGARDING TCP
When calculating the TCPR for liquids other than water, it is recommended that
at least 10% additional head pressure be provided to compensate for slight varia- ,
tions in the given properties of that liquid.
With high-vapor-pressure liquids such as ethane, the temperature _of the liquid
entering the pump cylinders must be maintained at the supply tank temperature.
The term P,.c in the TCPR formulas should be used with caution, recognizing
that if the calculated TCPR is Iess than the pump manufacturer's TCPR addi-
tional geodetic head, suction charging, or the use of a gas blanket may not solve
suction problems. lt behooves one to install the most efficient suction system
consistent with recommendations in this text.
3.8 METHODS OF INCREASING TCP
If it is necessary to increase the TCPA on an existing system, the following

approaches should be considered, in the order of effectiveness:
1. Install a centrifuga! pump in the suction system. The capacity of this cen-
trifuga! suction charging pump should be 150% of the reciprocating pump capac-'
ity. The existing TCP for the reciprocating pump must obviously be sufficient
for the proposed centrifuga! pump. A suction stabilizer at the reciprocating pump
inlet provides a favorable transition from the steady-state delivery of the centrif-
uga! pump to the variable demand of the reciprocating pump.
2. Reduce pump RPM.
3. Increase suction head (raise level in tank or raise tank).
4. Provide gas blanket.
5. Install larger diameter suction pipe as short as possible and with fewer bends.
3.10 SUCTION SYSTEMS 89
6. Install suction stabilizer at pump inlet (if problem is attributable to acceler-

ation). (See Chapter 4.)
7. Reduce valve spring POSIVA. (See Chapter 8, Valves.)
\ 3.9 CAVITATION
Cavitation in the liquid end of a reciprocating pump is described as the rapid con-
version of the liquid to a vapor, followed by a sudden collapse of the vapor bubble
to the liquid phase. lt is this sudden collapse that generates microscopic but intense
blasts of high-velocity liquid. lf cavitation occurs near a metal surface, damage to
that surface can be expected, usually in the form of pits or honeycombed surfaces
in local areas. These areas are usually on the upper surface of horizontal bores
(above the plunger travel) in the liquid end. lf pitting occurs at the intersection of
bores, the stress risers created can induce early corrosion-fatigue failure. Plungers
will sometimes show cavitation erosion on the packing area or on the end of the
plunger exposed to the liquid. A ring of pits around a plunger is a sure sign of
cavitation. Such conditions result in rapid packing wear.
Pressure deficiency at the pump inlet can be the result of a poor suction system
that can be described by "long" or "small diameter" or "sharp bends" or "low
head. '' Pressure reduction can also be generated by the acceleration of the liquid
at the beginning of a plunger or piston suction stroke. In that case, increasing the
head on the supply tank or source alone may not cure the problem. The installation
of a properly designed suction stabilizer will usually be the only remedy short of
a newly designed suction system. lt is also believed that air or gas dissolved in the
liquid can cause cavitation effects in the sense that sorne gas can break out with
the same results as vapor formation.
Cavitation is usually manifested by liquid knocking or pipe vibration. Exami-
nation of the instantaneous pressure at the pump inlet, measured with sensitive
electronic equipment and displayed on an oscilloscope screen, will show the typ-
ical rounded-bottom, sharp-upward-spikes type of waveform instead of the usual
symmetrical shape. Figure 3 .17 shows a typical waveform for cavitation in a
pump suction. In this case, a great improvement was achieved, as shown in (b)
by the use of a suction stabilizer as described in Chapter 4.
3.1 O SUCTION SYSTEMS
Cases 1 through V presented in Figures 3.18 through 3.21 show the physical
arrangement of the five possible suction piping and equipment configurations that
will be encountered. The appropriate formulas with examples are given in each
case.
(a)
(b)
Figure 3.17. Suction pressure trace showing typical cavitation waveform. (a) 63 PSI
to peak. Pump not equipped with suction stabilizer. Note typical "cavitation" wavef1
with sharp upward spikes and rounded bottoms. Also note extreme pressure excursio,Q
an upward or positive direction while the bottom of the trace is prevented from exte11' .·
into the negative region by the formation of vapor. (b) 13 PSI peak to peak. Same pu
with suction stabilizer, same operating conditions. Note the "clean" sine wave and h
frequency typical of "good" suction. The low-frequency cycles, over which the high
quency is imposed, are the remnant of pump rotation-generated cycles. While this
induced pulsation is not sertous, it could have been reduced by the use of a properl
charged bladder in the stabílizer.
90
Open
tank
p
T z
vp
1--+------'---l
D
ooo
Pump
Example
Water 120 degrees z 10 ft

Pump RPM 150 L 10 ft
Atmos 14.7 f 0.5
VP 1.5 POSIVA 4
TCP = 14.7 - 4.3 - 4 - 0.5 - 8.6 = +5.9
Figure 3.18. Typical suction system, Case 1: open suction tank, suction head.
91
Pump
L z
l
Open
tank
p
vp
Example
(3.18)
Water 120 degrees Z IO ft

Pump RPM 150 L IO ft
Atmos 14.7 f 0.5
VP l.5 POSIVA 4
TCP = 14.7 - 4.3 - 4 - 0.5 - 8.6 = -2.7

Pump will not run with minus TCP.
Figure 3.19. Typical suction system, Case U: open suction tank, suction lift.
92
Closed
tank 1 D
z
p
vp
Example
TCP =Pu+ P.,,+ P.,, - Pp- P4- P.'"c - P,"P (3.19)
Atmos 14.7 f 0.5
VP 1.5 POSIVA 4
TCP = 14.7 + 15 + 4.3 - 4 - 0.5 - 8.6 - 1.5 = + 19.4
re 3.20 Typical suction system, Case III*: Closed suction tank, suction head.
for Cases 111 and IV:
hydrocarbons, saturated pressure = vessel pressure, but vapor pressure < vessel
93
D
.--------'-----t ooo
Pump
z
_l
Closed
tank
Example
TCP =P.+ P.,, - P.,, - P" - P.>f- P...... - P,-. (3.20)
Atmos 14.7 f 0.5
VP 1.5 POSIVA 4
TCP = 14.7 + 15 - 4.3 - 4 - 0.5 - 8.6 - 1.5 = + 11.3
Figure 3.21 Typical suction system, Case IV:* Closed suction tank, suction lift.
* Notes for Cases 111 and IV:

For boiling liquid, P.P = P. + P
94
PUMP PULSATION
ANO CONTROL
PULSATION TYPES
ciprocating pumps are responsible for the generation of many types of pressure
turbances (pulsations) that may require special attention, Figure 4. lA. Follow-
are the types of pulsation encountered and the Dampener Types usually re-
ended.
There are so many different Pulsation Types and features in the selection of
ices for Pulsation Control, the following consolidated grouping of such will
of assistance:
Type (DV)-Discharge-Velocity
erent and predictable pump ftow-variation generated pressure disturbances pro-
ed by any type of restriction to that ftow, more pronounced in the long pipe-
s usually encountered in discharge systems. The frequency of these
urbances are exactly related to the pump speed and the number of active
linders, being in the order of about 1 to 50 Hz in most cases shown in Figure
, ftow-rate patterns of various pumps. These types of pulsation are usually the
st common and the most damaging because they are closely related to typical
e support span lengths that have natural vibration frequencies, Figure 4.3.
ch pulsations can generally be almost eliminated, both in the pump and in the
em, by the proper installation of gas-filled, energy-absorbing, dampeners
pe G, Figure 4.6, 4.11, and the judicious placement of pipe supports. The
ree of pipe vibration is related to the driving force and any reduction of such
ces by dampening will usually be effective in reducing that degree.
95
00 PUMP PULSATION AND CONTROL
TilI OVLRwm:L.'UNG T.BIORETICAL SUCTION ACC!l.I:RATION PULSA.TION (U)

SUP!JUMl'OSl:D ON TilI Tm:ORETIC.U. >LOW-rnJJUCED YUI.SA.TION (SV).
sucnoN V AL VE OPEN!NG DELA y

( SA)
SO) 1
.................-........~.......----........-..........-...--.........;..................1--9
AcruAL PUMP INll'T PRESS1JRE TRACE
Figure 4.lA Composite Pump Dynamics

4.1 PULSATION TYPES 97
KEY TO FIGURE 4. lA - COMPOSITE RECIPROCATING PUMP DYNAMICS
A-Real Oscillograph of Discharge Pressure Waveform of a 3 x 6 Triplex Single-

acting Pump.
Note the tendency to follow the theoretical flow pattern 'C'.
1-Maximum Discharge Pressure, 1644 PSI.
2-Average Discharge Pressure, 1400 PSI.
3-Minimum Discharge Pressure, 1151 PSI.
B-Cylinder Pressure Waveform per stroke.

4-0vershoot Pressure, 1944 PSI.
Seen only in Cylinder.
5-Average Discharge Pressure 1400 PSI.
6-Suction Pressure 15 PSI.
C-Theoretical Flow-rate or Velocity.

Note similarity in shape to actual.
A-Points of Maximum Flow-rate or Velocity.
B-Points of Maximum Plunger Acceleration.
C-Points of Minor Acceleration.
D-The Absolute Discontinuity between Suction and Discharge.

Non-existant in Centrifuga} Pumps.
E-Theoretical Flow-rate in Suction.
A, B and C Points same as C above.
-Actual Suction Pressure Waveform (Typical). Note how Acceleration

Pressure Overwhelms Flow-induced Pressure.
7-Peak Suction Pressure, 25 PSI.
8-Average Suction Pressure, 15 PSI.
9-Liquid Vapor Pressure, PSIA.
PULSATION AND CONTROL CLASSIFICATION

PULSATION PRESSURES-TYPES
ote; Pulsation Pressures in PSI:

DV) - Discharge Liquid Velocity (SA) - Suction Acceleration
DO) - Discharge Valve Open (SO) - Suction Valve Open
DA) - Discharge Acceleration (VA) - Vertical Acceleration
SV) - Suction Liquid Velocity (PS) - Pipeline Surge
(WH) - Water Hammer
98 PUMP PULSATION ANO CONTROL
TABLE 1A. Dampener Classification

·······························································.·····················
IDENT TYPE FEATURES
G Gas Use of Gas Conpression, usually Nitrogen.

L Liquid Ali Liquid. Use of Compressibility of Liquid.
D Dual Device using both Gas and Liquid theory.
BLADDER
e Confined Bladder Confined.
u Unconfined Bladder Unconfined.
N None Bladder None
F Foam Baldder-Cellular Foam
STYLE
A Appendage lnstalled on Tee in Pipe.
F Flow-Thru Ali Flow passing through device.
D Di verter Gives effect of flow-thru by deflecting Liquid t
Bladder containment.
PRINCIPLE
E Energy Absorbing Compressibility of Gas.
A Acoustic A device designed for the frequency to be filtere
out by filter networks.
R Reactive Similar to Acoustic but requiring somewhat greate
volume of Liquid. No Choke used.
COMBINATION OF FEATURES
Four-letter Groups: (Type-Bladder-Style-Principle)
**GCDE (A3 4.11) DUFR (D 4.9) LNAR (NS) LNFA (B 4.7)
GCAE (Al 4.6) GUFE (NS) LNFR (NS) LNFA (NS)
GUAE (A2 4.3) GNAE (NS) GFFE (C 4.4A) DUFR (NS)
**Numbers refer to Type and Figure Number.
(NS) Not shown.
4.1 PULSATION TYPES 99
DUPLEX DOUBLE-ACTING
Maximum +247.
Minim1.llll -227.
Total 467.
o 60 120 180 240 300 360
- - - - 17
- - TRIPLEX SINGLE-ACTING
V
"'
/
V I
\
\
I
I
"" \ ,,
\
I
I
'\
\ I
I
"""' Maxim= +67.

Minim= -1 77.
I
"
I
' Total 237.

I
I " \
\ I
I
\
I ' ' \
'
I \
\
I ' \ I \
\
o 60 120 180 240 300 360

QUINTUPLEX SINGLE-
ACTING
~ "V
MaximUlll +27.
_... _ _.__
,. "'-... - -"... --
...
t'). ',-
_.1.-_r-... , Minimum -57.
' '
-r-.., Total 77.
;-'. I .._
' /
/
/ ' ... / '
I ,,, .(
... '· /
/-
'
'I
I
I
I /
' ' I ' '
60 120 180 240 300 360 SEPTUPLEX SINGLE-
ACTING
Maximum +1.27.
Minimum -2.67.
Total 3.87.
360 NONUPLEX SINGLE-

ACTING
Maximum +0.67.
Minimum - l. 5 %
Total 2 .1%
o 60 l~O 180 240 300 360
Figure 4.2. Discharge Flow Variations (Velocity) of Multicylinder Pumps.

Frequency, Hz
Duplex Triplex
RPM Pump Pulseª Pulseª
50 0.8 3.2 2.4
100 1.7 6.8 5.1
150 2.5 10.0 7.5
200 3.3 13.2 9.9
ªPulse= pump RPM x number of cylinders/60
(a)
30' 2", 5 lb Pipe
l 2.9Hz 8.5Hz
4.3Hz 12.5 Hz
l (b)
l 1.9 Hz 5.5Hz
Figure 4.3. Typical pump and pulse frequencies and natural frequency of pipe spans.
4.1.2 Type (DO)-Discharge Valve Open

A pressure disturbance with a period unrelated to pump speed of a higher bu
with a very close range of frequency of about 50 to l 00 Hz. These pulsation
are caused by the acceleration pressure generated at the sudden opening of t
pump valves but the frequency-determining factor is unpredictable. The ampl
tude of these pulsations increase at the power of two of pump speed but they
4.1 PULSATlON TYPES 101
usually too serious and may be neglected. They can be alleviated by the use
any of the listed dampener types Type G.
(DA)-Discharge-Acceleration (Psac)
:ot::u"'""" the Acceleration pressure at the pump discharge is so miniscule, as com-
to the Velocity Pressure, it can usually be ignored. (Cases where it should
considered are cited.)
(SV)-Suction Velocity
velocity related disturbance but usually in exceptionally long or small
diameter suction lines. They can be alleviated with much smaller Type G Dam-
(SA)-Suction Accelerafüm
far the most serious disrupting disturbance encountered, for the reason that
overwhelms any Velocity effects. A type D Dampener (Suction Stabilizer, (Fig-
te 4.4), will tend to relieve the pump suction of any acceleration disturbances.
(SO)-Suction Valve Opening {Psvo)

Pulsation caused by the surge of the suction valve opening, being displayed as a
· ¡ipidly dying almost pure pressure sine wave of a frequency not related to the
p rotational frequency, always about 80 Hz, near the start of each plunger
stroke. Probably the most important contributor to poor pump perfor-
because the minus excursions may 'dip' into the vapor pressure, resulting
(V.A)-Vertical Acceleration
in Figure 4.5(a) that when the system involves a short vertical discharge,
as a mine dewatering system, the pump "sees" a purely gravitational-in-
outlet with little of the frictional losses of a long pipe line. In this case a
pe D dampener is a must for proper alleviation. This also applies to pumps
ith short connecting lines to an already pressurized system, Figure 4.5(c).
(PS) Pulsation-Surge-Not shown

hile not recognised as a pump related problem, the high amplitude instant pres-
re surges caused by sudden opening or closing of valves downstream of the
3 7 8 6
11
~ilt;;~~~~~l}-~--5 .
o
o Liquid
o
Figure 4.4. Type Suction Stabilizer (Courtesy, Greer, Inc.)
pump can result in excessive damage. A properly sized Type G Dampener wi'.I
protect the system, See Section 4.13.
4.1.9 (WH)-Water Hammer

So-called 'Water Hammer,' a pipe-rattling, continuous vibration usually cause
by small, loose parts in the system such as control valve buttons, etc., finding ..
resonant part of the system to match.
4.2 DAMPENER TYPES 103
::\(0
1í
2000 ft water = 860 PSI head
+ 2 PSI due to friction
ººº
l
Total pressure at pump = 862 PSI
(a)
163 mi; Friction = 1 PSl/1000 ft = 862 PSI
'ººº' Total pressure at pump = 862 PSI
(b)
Short connecting pipe

Pressu rized system
000 Total pressure at pump
860 PSI
= 862 PSI
(e)
Figure 4.5. Dynamic differences in discharge systems .

a. Vertical discharge .
b. Horizontal discharge.
c. Pressured Systems.
4.2 DAMPENER TYPES
4.2.1 Type G
Gas (Nitrogen) filled, bladder type, energy absorbing, adjustable pre-charge dam-
pener, most commonly used.
4.2.2 Type GCAE

Provided with a bladder anti-extrusion device (button or poppet valve) to allow
an almost unlimited high pre-charge pressure which extends the effectiveness of
the device but precludes any protection below the pre-charge pressure. The use
Elastomeric bladder
Figure 4.6. Type GCAE Appendage Gas-type (Energy absorbing) Dampener. Restrained
bladder. (Courtesy, Greer, lnc.)
of multiple-dampeners on a system, one of them precharged to a low pressure

minimizes this objection. Figure 4.6.
4.2.3 Type GUAE

A "free-bladder" type dampener usually accepts a lower pre-charge which re
duces its effectiveness at high pressure but it does provide attenuatíon down
zero discharge pressure. Figure 4. 7.
4.2.4 Type GCDE

A so-called "ftow-through" type of dampener equipped with a "diverter" baffl
to provide an additional extremely effective filter action to prevent many of t
high frequency disturbances from passing into the discharge or suction pipin
system. Figure 4.8.
4.2.5 Type LNFA

Acoustic ( filter) dampener that filters out and prevents pump-generated pressu
disturbances from entering the piping systems. For the low frequency of PV ty
pulsations an abnormally large volume dampener is required and due to the. ·
tentionally inserted chokes an actual back pressure is placed upon the pump. A,
acoustic filter, per se, only "filters" and does not reduce the pulsations occurrin
VENT"" /
PRE-CHARGE VALVE
REMOVABLE
..__..-......-1-~~,_, CAP
BLADDER
GAS
U QUID
Figure 4.7 Type GUAE Appendage Gas-type (Energy absorbing) free-bladder Dampener.
Limited pre-charge pressure.
in the pump as does any of the gas-type devices. Such acoustic dampeners must
be designed for only one pump operating speed-at other speeds their effective-
ness falls off rapidly. Figure 4. 7. The acoustic fil ter can be compared to the
electrical analogy as shown in Figure 4. 9.
4.2.6 Type GFFE

Cellular type, fixed low pre-charge, gas type. Requires no pre-charging or blad-
der replacement and is effective for both flow-induced and acceleration-induced
pulsations. lt is also effective from minimum discharge pressure but somewhat
less effective at extremely high maximum pressure because of the limited but
permanent so called "pre-charge". Figure 4. IO.
4.2. 7 Type DCFR

Combination Type G and Type L dampeners are almost mandatory for a recip-
rocating pump suction inlet, Figure 4.4. and for a pump outlet discharging into
a vertical pipeline (such as a mine de-watering system) or long pipelines passing
through mountainous country with exceptionally high and low elevations.
Anti·extrusion rin¡¡
Bladder button
Plu¡¡
Figure 4.8 Type GCDE Appendage, gas-type (energy absorbing) Dampener. Restrained
b!adder. Diverter. "Pulse-tone" (TM Greer.)
In Volume2 Out
Chokes
Figure 4.9. Type LNFA Acoustic Dampener

Figure 4.10. Type GFFE Flow-through Closed-cell foam elastomer, energy absorbing,
impedance dampener. (White Rock Engineering.)
Pressure gauge
Patented reinforced
flexible membrane
Outlet
1
i-+--------Facetoface•--------__..
1
Figure 4.11. Type GUFA PASAFE Dampener (White Rock Eng.)

LIOUID
REACTANCE
BAFFLE
Figure 4.12. Type LNFR Dampener
!\.
..
MAX1140
~: ~ MIN 680
~
-+~-------'IJ------+- AVE 1
~~iyc1021
MIN 95í
t:.P 70
t:.P 460 ZERO
ZERO
(a) (b)
Figure 4.13. Methods of reporting degrees of pulsation and control. (a) Case 1: Non.
dampened waveform (from flow variation). (b) Case 11: Dampened waveform (from fla
variation).
4.3 REPORTING DEGREE OF PULSATION
Method A: Used throughout this book and recommended ·as standard. Percent residu
pulsation pressure:
Case 1: liP/Ave X 100 = 460/1000 X 100 = 46%

Case 11: liP/Ave X 100 = 70/100 X 100 = 7%
4.4 DISCHARGE DAMPENING 109
~tbod B: AP or change in pressure, min to max:
b.P = 460 PSI
D.P = 70 PSI (4.2)
C: Percent attenuation or suppression:
I and Case II:
AP1 - APn/ AP1 X 100 = 460 - 70/460 X 100 = 853 (4.3)
Percent transmission ratio:
I and Case II:
!::.Pu/ AP1 X 100 = 70/460 X 100 = 15% (4.4)
ortant: ANY REFERENCE TO DEGREE OF PULSATION SHOULD APPLY TO THE. TOTAL EX-
!ON IN TERMS OF PRESSURE OR PERCENTAGE. fOR EXAMPLE, IN CASE l, THE TOTAL PUL-
N OF 460 PSI (46 %) INFERS THAT THE EXCURSION IS FROM 680 PSI (32 % BELOW THE
ERAGE) TO 1140 PSI (14% ABOVE THE AVERAGE).
DISCHARGE DAMPENING
e application of gas-filled, bladder-type dampeners (Fig. 4.6) is the generally

epted practice in most applications in "normal" service, particularly where
eral pumps are discharging into the same system. Normal service includes long
lines, oil-field water injection, drilling-mud circulation, and process systems
high-pressure letdown, etc. Any practical degree of ftow-induced discharge
tion attenuation can be had by selecting the proper dampener size and using
correct gas precharge pressure.
Experience shows that the high-frequency acceleration-generated pulses cannot
attenuated by the use of conventional gas-bladder devices because these devices
not react quickly enough to absorb much of the surge.
There are about 30 different devices that can be used for pulsation control, ali
them variations of the three basic types (Figs. 4.14 and 4.15). Most of them
Id perform on a theoretical basis, but many are inefficient.
The trouble-free gas-bladder type of dampener must be ruled out for pumps at
uid temperatures above 300ºF because of the effects high temperatures have on
elastomers, and also where there is the possibility of chemical attack on the
stomer bladder. However, the gas or vapor-over-liquid type is justas efficient
s of dynamic aspects, and perhaps something can be done to counteract two
ections: (1) the difficulty of maintaining proper gas volume and (2) loss of gas
y absorption in the liquid.
2 3 4
Long
bladder
~~r-
_}-
-------------
lnverted bladder Bourdon tube

9 10 11
Gas Gas
-----------
Metal bellows Gas bottle Manual charge Autocharge
13
Water
vapor Water
16
n ontinuous
........ or as
- - -¿¡- other
500 F 1. .d
9 u___ 1qu1
s-=-=-=-=-=Í}fi~
Vapor bottle Cellular bottle Sparger
17 18 19 20
Gas
Vaporization Piston
Figure 4.14. Gas-type dampener variations.

4.5 GAS-TYPE DAMPENER SIZING 111
23 24
w v. Wave
trap
•
Acoustic Choke Tuned trap
27Q~~~~ter 28
26
wall ~
shell ~
Turbulent vapor
Multicylinder
Turbulent vapor Low modulus pumps
31 32 Gas
tn:: Gas c_-_Lt~fJÚ:-~9LL~

--i -5x1-rT1r11-{1-¡--·
-----------------
-
Perforated Elastomer
000 sleeve sleeve
Sleeve or
Offset pump "flow-through"
Figure 4.15. Acoustic and special methods of dampening.
Gas volurne could be rnaintained by the use of a liquid-level control or contin-

ous injection of hydrogen or other gas as part of a process. The bottle filled with
as would provide the required darnpening effect. This would provide rnuch better
ulsation darnpening than sparging gas into the systern downstrearn frorn the purnp.
f course, continuous· introduction of gas into the process would naturally elirni-
ate objections to the continuous absorption of the gas in the liquid. As to the
.roblern of gas absorption by sorne liquids, the problern is notas severe as it was
first thought to be. For exarnple, the absorption of nitrogen in water is lirnited,
.and sorne gas-over-liquid darnpeners at low pressure use an insignificant arnount
of rnake-up gas.
GAS-TYPE DAMPENER SIZING
ormula (4.5) in Section 4.6 can be used to calculate the size (volume in U.S.
allons) of darnpener by inserting the degree of residual pulsation allowed (gen-
erally 3-6%) and the value of precharge pressure. Precharge pressure of a gas-
bladder type of dampener is the charge of gas, usually dry nitrogen, injected into
the bladder through sorne type of connector and valve.
The most desirable precharge pressure is 60-70 % of the average working pres-
sure or the limit allowed by the bladder design. If the precharge is higher, there is
danger of the anti-extrusion valve (if so equipped) being destroyed by continua!
closing on the seat at each pulsation.
A precharge pressure below about 25 % of the average discharge pressure should
be avoided to prolong bladder life by decreasing the degree of flexure. Low pre-
charge pressure also reduces the effective size ofthe dampener. With the gas-over-
liquid type of dampener, the same sizing calculations can be used as for the bladder
type, using a theoretical precharge of 100 %.
4.6 GAS-TYPE DAMPENER SIZING EQUATIONS
a. Volume of dampener, U.S. gallons:*
V8 = KsD 2Pd/TP ..
where K = (aQt%)!23l
*Based on Nitrogen charged Bladder with n = 1.4.
b. Equivalent volume of multiple dampeners (a, b, e, etc.) with different pre-

charge pressures, where Pea > Peb > Pee:
To calculate equivalent effect by eq (4.1) use Ve precharged to Pea·
c. Pulsation pressure, tú', without dampener =

(1 + A)\Pd + 14.7) - (1 - B)\Pd + 14.7)
where
A = maximum instantaneous flow % (Table 4.1)

B = mínimum instantaneous flow % (Table 4.1)
TABLE 4.1 Pump Type Constants

Pump Type Cyls Angle A(l) 8(2) Percent K(4) Percent(5)
Simplex SA 1 360 0.58 1 158 0.684 250
Simplex DA 1 180 0.29 1 129 1.368 166
Duplex SA 2 180 0.29 1 129 0.558 166
Duplex DA 2 90 0.24 0.22 46 0.199 93
Triplex SA 3 120 0.06 0.17 23 0.100 43
Triplex SA (0) 3 120 0.062 0.17 23 0.100 44
Triplex SA (30) 3 120 0.061 0.12 18 0.078 35
Triplex SA (60) 3 120 0.05 0.09 14 0.061 27
Triplex DA 3 120 0.06 0.17 23 0.200 43
Quadruplex SA 4 90 O.ti 0.22 33 0.143 62
Quadruplex DA 4 90 O.ti 0.22 33 0.286 62
Quintuplex SA 5 72 0.02 0.05 7 0.030 14
Quintuplex DA 5 72 0.02 0.05 7 0.076 14
Sextuplex SA 6 60 0.06 0.17 23 0.100 43
Sextuplex DA 6 60 0.06 0.17 23 0.200 43
Septuplex SA 7 51.4 0.012 0.026 3.8 0.016 8
Septuplex DA 7 51.4 0.012 0.026 3.8 0.032 8
Octuplex SA 8 45 0.026 0.052 8 0.035 15
Octuplex DA 8 45 0.026 0.052 8 0.070 15
Nonuplex SA 9 40 0.006 0.002 2.1 0.009 2
Nonuplex DA 9 40 0.006 0.002 2.1 0.018 2
--
NOTES;
(1) - "A' Flow variation above average, decimal. (5) - Maximum pressure variation (pulsation) without dampener, percent.
(2) - "B" Flow variation below average, decimal. (0) - Fixed phase angle of two pumps compounded.
....
.... (3) - Flow variation, total percent (30) - Fixed phase angle of two pumps compounded .
w (4) - K Factor for calculation of dampener size (Equation 4.5) (60) - Fixed phase angle of two pumps compounded.
4.7 DERIVATION OF K
flow
Flow rate, Variation
GPM Q, 3 K
Maximum
158 .67
2 Average
E
;!: Minimum (zero)
o Single-acting simplex
¡;:
1
1
Maximum
Average
~ 129 .55
E
;!: Mínimum (zero)
o Single-acting duplex
¡¡:
Double-acting simplex
Maximum
Average
Minimum 23 .098
2
E
;!: Zero
o Single-acting triplex
¡¡:
1
Figure 4.Hí. Derivation of constan! K in dampener sizing Equation (4.5).
Vx = KsD 2PjTP, (4.5)
K = .6.Q%!231 (in 3 /gal) (4.6)
4.8 PASAFE PULSATION CONTROL SIZING
The PASAFE Model selected for a given application is based on pump flow rate.
Generally, the suction and discharge units are sized to be equivalent to 10% of ·
the flow rate in volume per minute delivered by the pump.
PASAFE Model (Gallons) 0.1 * Flowrate (gal/min)

PASAFE Model (Liters) l. 7 * Flowrate (m 3 /hr)
4.8 PASAFE PULSATION CONTROL SIZING 115
MODEL
Liquid Volume Gas Volume Max Gas Charge
Gallons Liters Gallons Liters PSI Bar
20 75 2.6 10 300 20.7
40 150 5 19 250 17.2
80 300 13 49 200 13.8
120 450 28 106 150 10.3
240 900 53 201 125 8.6
400 1500 91 344 100 6.9
600 2250 112 424 100 6.9
Calculate volume displacement of one pump cylinder using the pumps bore and
stroke.
VcvL (Gallons) = 0.7854 * Dia2 * Stroke/231 (cu.in.)

VcvL (Liters) = 0.7854 * Dia2 * Stroke * 10-6(mm.) (4.9)
Pump ftow volume variation is the· quantity of liquid the pulsation control unit
must be able to accommodate and then dispense to stabilize pipeline ftowrate. A
Factor for each pump type has been determined by integration of a computer
generated ftow curve. Model assumptions are 10 degrees valve sealing delay and
10 degrees liquid compression delay.
The resulting factors will be appropriate for 95% of the pump market. The
exception are high pressure pumps with large clearance volumes or highly com-.
pressive liquids.
dVPuMP (Gallon or Liter) VcvL * FACTORPT (4.10)
where FACTORPT is:
Pump Type
Simplex SA 0.600
Simplex DA 0.280
Duplex SA 0.240
Duplex DA 0.160
Triplex SA 0.026
Quintuplex SA 0.016
Calculate gas volume available in the dampener at system operating pressure:
(4.11)
where:
Ve PASAFE Gas Volume at System Operating Pressure

P0 System Operating Pressure
Vep PASAFE Initial Gas Volume
PCP (psi or bar) = PASAFE Nitrogen Charge
Pressure 50% of System Pressure
or Maximum PASAFE Charge
Pressure whichever is less.
Residual pressure variation resulting from changing frictional pressure drop as a

result of downstream flow variation when using the selected PASAFE Unit is
cakulated using formula (4.12). Gas volume provided by the PASAFE Unit al-
lows down stream flow variation to be reduced because of its ability to accom-
modate peak flow and dispense fluid during periods of low pump flow. Pressure
pulsations when operating below normal system pressure wil! always be less than
the pulsations created at normal system pressure. This is not true for appendage
dampeners when system pressure is below the dampener charge pressure.
(4.12)
where:
Po (psi or bar) System Operating Pressure

Ve (gal or liter) PASAFE Gas Volume at Operation Pressure.
Target residual pressure variation (peak to peak) is figured as a percent of system

operating pressure. The following table lists industrial recommended standards.
However, system requirements or economics may require a different Target o/odP
to be used.
Type Pump Target %dP

Duplex DA 10%
Triplex SA 3%
Quintuplex 2%
Percent dP = IOOdP / P0 (4.13)
Adjustment of residual pressure pulsations can be obtained by selecting a larger

or smaller PASAFE Unit or by the addition of an appendage pulsation dampener
to augment pulsation control at high system operating pressures. The addition of
gas volume from another dampener requires an adjustment to the total gas vol-
4.9 MULTIPLE DAMPENERS 117
ume and dampener charge pressure to recalculate the residual hydraulic pressure
variation at design operating pressure. Since the PASAFE Unit has a relatively
tow nitrogen charge pressure, the PASAFE Unit provides pulsation control over
a broad operation pressure range. An appendage dampener used to augment the
PASAFE performance can be charged toan optimum higher charge pressure. Use
the following equation to determine a new gas volume available at system oper-
ating pressure.
(4.14)
where:
Vap PASAFE Initial Gas Volume

PcP PASAFE Nitrogen Charge Pressure
VGA Appendage Dampener Initial Gas Volume
PcA Appendage Dampener Nitrogen Charge Pressure
4.9 MUL TIPLE DAMPENERS
For reciprocating pumps in long pipelines, the use of two or more discharge dam-
peners (with a total gas volume as required for one dampener) should be considered
for the following reasons:
1. Because pipeline start-up pressure rise is brought about overa relatively long
period (severa! minutes to an hour or more}, there is a period during which the
pump pressure is less than the optimum precharge pressure of about 60 % of the
discharge pressure for one dampener. During that time, the pumps would operate
without pulsation control. With two dampeners (the second of which is precharged
to about half of the first}, for example, pulsation control is extended well into the
critica! start-up period.
2. In sorne cases, it is necessary to pump altemate batches ofliquids with widely
differing viscosities. lt is obvious that the pump pressure required to displace the
less viscous liquid is less than that for the more viscous-sometimes less than the
optimum precharge pressure for one dampener at maximum pipeline pressure. A
second dampener precharged to a lower pressure would give protection for a wide
range of pipeline pressures.
3. The use of multiple discharge dampeners provides sorne redundancy and
protection in case of the loss of precharge or bladder failure in one of the other
dampeners.
Figure 4.18 shows why multiple dampeners are desirable on long pipelines or
in other applications where there are long periods of operation at various pres-
sures.
Theoretical
600
/ without
dampener
;/
500
~
/
¿ 400
/
~,/
.2
~
~ "'
~
~ ~
!?::! 300
~
Q.
(..)
a;
iQ,/ / "'1
"'
<
Q,
76 200
~
i /~
~/ ie
ie ,,,,..,,,.. .....
...,,,,,.. One 30·¡allon
dampener
• e!!! ,,,,,.,,,..
! ..,..,,,.. Thr111 lO·pllon
100 dampener~
1000 1200 1400

Oper11tíng pipeline pr111ur11, PSI
Figure 4.17. Multiple-discharge dampeners. 9 by 12 in triplex pump. Maximum operat-

ing pressure, 1400 PSI.
4.10 PRECHARGING GAS-TYPE DAMPENERS
Figure 4.16 can be used to estimate the volume of high-pressure nitrogen re-.
quired to precharge gas-type dampeners of various sizes. The chart is based on thé
amount of gas contained in the popular T cylinder of 300 ft 3 capacity.
Remember the basic rule for precharging:
1. There should be no pump or system pressure on the dampener during pre-

charge.
2. Never pre-charge a dampener in excess of 60-70% of the working pres-
sure.
4.11 VAPORIZATION OR "HEATED" DAMPENERS
A simple method of generating a continuous "gas" charge for a bottle-type dam-

pener (no bladder) is by the expediency ofplacing an externa! heatíng coil (electric; '·
or steam-trace) around the upper part of the bottle. With sufficient heat applied,
theHquid in the bottle will be converted to vapor, thereby providing the necessary
gas for cushioning.
The limiting factor in such an application is the system pressure. At high .
sure an exceedingly large amount of heat would be required, and above the critical
pressure of the pumped liquid no vapor would ever be formed. Also, on account
of the lower ratio of specific volume of vapor to liquid at high pressure, such
4.12 MANIFOLDED MULTIPLE PUMPS (PHASING) 119
Dampener size, gal
Example:
1One cylinder of N2 would precharge about
1
1one 40-gal (or tour 10-gal) dampeners to
¡ 540 PSIG (always 603 of discharge or
1 system pressure).
200 1 1
1 Example: Cylinder
¡ pressure reduced
1 to precharge
1 of aboút 540 PSIG.
100
V i 234cuft N2(STP)
------ avaiiitl1e--
Maximum
cylinder
pressure
500 1,000 1,500 2,000 2,500
Nitrogen cylinder andlor precharge pressure, PSIG
Figure 4.18 Nitrogen cylinder and/or precharge pressure, PSIG
l. Draw a line from the desired precharge pressure on the base of the chart, upward to
the dampener size line at the top ofthe chart. Read the numberof gaUons of dampener
size that can be precharged from the top scale. For this example, dampener size is
about 45 gal. Accordingly, four 10-gal, or one 10-gal four times, or eight 5-gal
dampeners could be precharged.
2. Draw a line horizontally through the point of intersection of the diagonal line and
the vertical. This shows how much nitrogen is left (left side) or how much was
removed (right side) in cubic feeL The remaining low-pressure nitro.gen can be used
to precharge low-pressure suction stabilizers.
3, The precharge pressure should be about 60-70% of the average discharge or suction
pressure or' to the limit allowed by the manufacturer .
.devices are more efficient at lower pressure. Accordingly, such dampeners are
satisfactory for pump suction systems with reasonably low suction pres-
sures, but for discharge systems they are in most cases impractical.
MANIFOLDED MULTIPLE PUMPS (PHASING)
n the case of multiple pt¡mps discharging into a common system, there are two
onsiderations;
l . Ali pumps coupled with fixed phase angle maintained.

2. Pumps with independent drives with no regard to fixed phase angle.
When pumps are arranged as in No. l, the fixed phase angle should be selected
for each pump by the following rule so as to result in minimum ftow-rate induced
pulsations;
Phase angle, degrees = crank angle/number of pumps
For example, two triplex single-acting pumps would require a fixed angle of 6
degrees.
Pumps with independent drives will normally seek an in-and-out phase re!
tionship resulting in maximum and minimum degrees of pulsation, unless a
elaborate precision prime mover speed control is used.
With either selection the required pulsation dampener size will not be ef
fected-each pump will require the size based on a single pump.
Table 4.1 includes a set of data on the effects on the pulsation degree
different phase angles of two pumps.
4.13 THE SUCTION SYSTEM
The friction losses in a suction system are usually low because of the relative
short length and large diameter of piping involved. Accordingly, the ftow-induc
A-type pressures generated are of low magnitude compared to the accepted p
centage of change. For example, if the suction pressure is a static 20 PSI, the 23
variation of a triplex single-acting pump would generate a theoretical A-type p
sure variation of only 9 .2 PSI. This is hardly enough energy to set pipes in motio
compared to the A-type pressure variation of 644 PSI in the discharge at 1400 P
at the same ftow variation.
But the forces of acceleration become the overwhelming disturbances in
suction. Pressure pulses of more than 25 PSI are encountered in pumps even
systems with short suction pipes. A small amount of dampening of the ftow-i
duced PSI can reduce it to a negligible amount, leaving the 25-PSI C-type acc
eration pulsations present for any damage they can cause by possible cavitation:
4.14 THE DISCHARGE SYSTEM
To carry the example to the discharge, the same forces are at work, but the pre
at A due to ftow-induced pulsations becomes overwhelming at 460 PSI. The
PSI contribution from acceleration at C is a small percentage (2.5%) of the~t
discharge pressure.
An exception is when the pump is delivering into a low-friction, high-press
4.16 DAMPENER PERFORMANCE 121
system such as a short vertical discharge system in mine dewatering (Fig. 4.5a);
to an already pressurized system such as a pressurized pipeline through a short
connecting pipe, or to an already pressmized system such as hydraulic press ac-
cumulators and similar systems (Fig. 4.5c). In those cases, the acceleration pres-
sures can become the overwhelming disturbance, particularly if the piping system
is relatively long compared to a suction system (but considerably shmter than a
''pipeline'
4.15 DAMPENER FACTS
A gas-bladder energy-absorbing type dampener of a specified size is not

to the effects of pump speed-the reduced pressure resulting from the reduced
speed still provides the same percentage of alleviation.
Multiple pumps operating in parallel without precise or fixed phasing demand
a maximum degree of dampening because of the adverse effects of the frequent
"in-phase" or simultaneous delivery of severa! pumps. Individually driven
pumps are almost impossible to be driven in fixed phase. If so, a permanent!y
maintained 60 degree out-of-phase with two triplex pumps, for instance, would
require less dampening than two triplex pumps. See Table 4.1.
Any dampener equipped with elastomeric (rubber) bladders or are tem-
perature limited to about 140 degrees F. Viton can be used up to about 200
degrees F but any elastomeric bladder suffers reduced life at higher temperatures
and in adverse chemical environment.
4.16 DAMPENER PERFORMANCE
The most important function of the gas-type dampener is to prevent the genera-
tion of the most destructive low-frequency pulses such as those generated by the
pump rotary motion and the combination of flow from each of the pump cylin-
ders, basically RPM times number of cylinders. Accordingly, on the basis of a
maximum of 500 RPM for most small pumps, the maximum frequency involved
should not be over about 50 Hz.
Above 50 Hz, such as at frequencies generated by the acceleration of the liquid
and the system acoustics, the strictly gas-type dampeners become less efficient.
From 50 Hz to around 200 Hz, the acoustical type is rather efficient. But, inter-
estingly, it becomes exponentially more efficient and less complex with even higher
frequencies.
It is therefore evident that there may be a place for both types of dampeners.
Figure 4.8 compares the performance ofthe most popular types on a typical pump.
It is evident that in this particular case the gas-type dampener offers better perfor-
mance then the acoustic type.
An analysis of reciprocating pump pulsations is based on the premise that pres~

sure or pressure variation (pulsation) due to ftow velocity or ftow-velocity variation
is a function of the square of such velocity. And the system must present enough
resistance to ftow, due to pipe friction, bends, fittings, chokes, etc., to allow á
pressure or pulsation to be generated in the first place. ·The magnitude of that
pressure is directly related to the resistance. The analysis is also based on the
premise that the acceleration of the liquid at the beginning of an increase in ve~
locity (start of stroke) generates a pressure _pulse, and the magnitude is directly
related to the length of pipe connectd to the pump suction or discharge. lt is alsc>.
inversely related to the diameter of the pipe and directly related to the rate q
change of velocity.
Because the inherent ftow variations of a reciprocating pump are of the sa
intensity (but out of phase) in both the suction and discharge, for many years
was believed that a pulsation dampener of the same size as the discharge shoul
be used on the suction, and many such installations can be found. However, fro"
the standpoint of ftow-induced pulsations iilone, it is logical that a smaller da
pener can be used on the suction. Por example, a typical triplex single-acting pum
has a ftow variation of about 23 % in both the suction and the discharge. If all
this ftow variation were converted to pressure pulsations, the pulsations woul
show a maximum of 46% of the average pump pressure. In our example, .·
maximum change in pressure at the discharge would be about 460 PSI at 1000 P
average pressure. Now if there is, say, a 50-PSI suction pressure, the change i
pressure there would be only 23 PSI; that is, there would be about 20 times le
energy available to set the piping in motion.
lt is also now known that the acceleration due to velocity variations is the sa
in the suction and discharge systems of the pump. In the suction system, however,
such disturbances usually overwhelm the disturbances due to flow variation, an
conversely, in the discharge system the ftow-induced pressure pulses usually ovell"
whelm the acceleration pressure pulses.
There is evidence that only a flow disturbance or an acceleration disturbance
(not both) can exist during any suction or discharge stroke of a pump. One will
overwhelm and reduce the effect ofthe other. A plausible explanation is the theorSi
that in pumps of greater than about 300 FPM piston speed, a disturbance of either
type will momentarily reduce the volumetric efficiency to such an extent that .it
cannot recover fast enough to allow the ftow rate to reach an adequate value ta.
affect another disturbance in the short period of time, about 3 ms, between distur-
bances. There is no worry with pumps of slow speed because few if any distur
bances are generated.
lt is conceivable for a pump with high inherent flow variations to actually de
liver an almost pressure-pulseless liquid into an open-ended (in effect), short, refa•
atively large-diameter horizontal discharge pipe. Sorne typical cases are pumps'
with short connecting pipes feeding a large-diameter already-pressurized syste
such as an existing pipeline or hydraulic press supply. If the discharge pipe is no
made vertical, as in a mine dewatering system, the acceleration of the liquid col~
umn due to gravity becomes the predominant generator of pressure pulses. This
4.17 ACOUSTIC FILTERS 123
dynamic and hydraulic di:fference is shown in Figure 4.5, where the pump
' the same discharge pressure under three di:!ferent physical arrnngements:
Pure vertical discharge-pump pressure is predominantly due to gravity head
liquid; (b) pure horizontal discharge-pump pressure is predominantly due
·to velocity friction losses; and (e) discharge into pressmized system-pump pres-
nre is predominantly dueto the constant pipeline or system pressure.
Energy-absorbing or gas-type dampeners may not solve pulsadon problems in
(a) and (e), and filter or acoustic-type dampeners should be considered. In
cases, combination energy-absorbing and acoustic-type dampeners may be
Jt is impossible to speculate on the cause of disturbances of an obviously de-

nature in existing systems. A pressure survey of the system by the use of
and oscilloscope readout will reveal the causes and allow proper
·on control to be applied. See Chapter Instrumentation.
ACOUSTIC FilTERS
filters (dampeners) opernte on distinctly different pri.nciples than gas-type

dampeners. Acoustic filters, as the name implies, do not alle-
flow variations but they do filter out the high-amplitude pressure waves
enerated by the pump befare they trnvel into the downstream system of piping,
whereas gas-type dampeners prevent the generation of such waves by the pump
actual reduction of the flow-variations.
Because reciprocating pumps are the main interest of those this book, it
only fair to emphasize that the acoustic filter, reacting to the deliberate insertion
of flow restriction in the device, can impose greater discharge pressure and con-
º"'"""r1t1" higher pressure pulses in the pump liquid end itself, even beyond those
an undampened pump. This is weH demonstrated in Figure 4. where the
electrical analogy shows that the insertion of resistance alone reduces the down-
0stream pulsations but increases the amplitude of the upstream pump-generated
_pulses.
Acoustic filters are most efficient at only one particular design pressure wave
which militates against their acceptance for systems that require vari-
in speed or displacement, liquid type, and temperature. They also become
more efficient at higher frequencies, while gas-type dampeners ex-
an inverse reaction to frequency. (See Fig. 4 .18.) This feature precludes the
use of acoustic filters on large, low-speed pumps, such as slurry pipeline triplex
at 75 RPM, which have a basic frequency of 7.5 Hz. This low frequency
.,v,uuu1<ou with the great flow rate would dictate an abnormally large and costly
acoustic device.
It has been shown that reciprocating pumps usually generate pressure pulsations
of several different frequencies in both the suction and the most violent
,and damaging being that related to the basic pump pulse frequency determined by
the pump RPM times the number of cylinders and usually times a number rela
to the crank arrangement. Figure 2.3 illustrates various pump-type ftow peaks
stroke. These basic frequencies seldom exceed about 50 Hz. Secondary high-
quency pulsations result from the effect of acceleration discontinuities in both;
suction and discharge. These are on the order of 75-150 Hz, having no relatio
pump speed. Unpredictable pressure pulses are also generated when the acou
properties ofthe system result in "noise" from flow-through orifices, valves,
ulators, etc. These latter high frequencies can certainly be reduced by the us
a properly designed acoustic filter.
As previously mentioned, combining the features of both types is often e
tive. The suction stabilizer shown in Figure 4.9, the PASAFE* (pump accelera
stabilizer and flow equalizer) in Figure 4.10, and the PULSETONEt gas-type
ener shown in Figure 4.11 are examples.
An excellent example of how such a combination can improve the contro
pulsation is shown in Figure 4.8(d), where the PULSETONEt [Fig. 4.11) re
in improved downstream control by "filtering out" the secondary high-frequ
pulsation as compared to that of Figure 4.8(d), a typical gas-type appen
dampener, Figure 4.8(c). Note that the "pure" acoustic filter [Fig. 4.8(b)]
mediocre low-frequency relief, both on the pump and downstream.
It is beyond the scope of this book to include the design of such devices. I
ested readers should contact the manufacturers or suppliers directly.
The following acoustic filter formulas are intended only to show how al
parameters of an acoustic filter are related. Refer. to Figure 4. 7 for a typical a
tic filter configuration.
Acoustic Filter Formulas tt
K', effective bulk modulus, PSF = 144 K/[1 + (KD/Eh)(l - µ 2 )]

(
a, acoustical velocity, ft/s = (K'g/p)º· 5

C, acoustic capacitance of volume, ft5 /lb = Vg /pe?
C, acoustic capacitance of volume, ft5/lb = V/ K'
L, acoustic inductance of tube, lb-s2/ ft5 = pL 1/ gB
RA, acoustic resistance of pipeline, lb-s/ft5 = P/ Q

P, pressure of rapid valve closure, lb/ft2 = paQ/gA
*PASAFE™, White Rock Engineering, Inc.

tpuLSETONE™, Greer Hydraulics, lnc.
ttModified from E. J. Hicks, T. R. Grant, "Acoustic Filter Controls Recip Pump Pulsatio
and Gas Journal, Jan. 15, 1979.
4.18 THE SUCTION STABILIZER 125
RA = paf gA (4.24)
PdP0 , transmission ratio = ../1/(X + Y) (4.25)
X= [(wL/RA) 2 (CR~/L)- 1 ] 2 (4.26)

Y= (wL/RA{( CR~/L) 2 (4.27)
A = flow area of pipeline, ft2

B = cross-sectional area of tube, ft2 , (tria!).
D = pipe inside diameter, in
E = modulus of elasticity of pipe, PSI
f = frequency of pulsation, Hz
g = gravitational constant, ft/ s2
h = pipe wall thickness, in
K = bulk modulus, uncorrected for pipe, PSI
L 1 = tube length, ft, (trial).
P = pressure, lb/ft2
PL = peak pulsating pressure with filter, lb/ft2
p o = peak pulsating pressure without filter' lb/ ft 2
Q = volume of fl.ow, ft3/s
V= capacitance volume, ft 3 , (trial).
p = liquid density = lb/ft3 x 0.53
µ = Poisson ratio
w = 27rf
THE SUCTION STABILIZER
the most economical and efficient pump performance, particularly with the
ent of higher pump speeds for greater capacity, the suction system deserves
ful consideration.
small energy-absorbing, gas-type dampener on the suction will prevent the
eration of flow-induced "A" pulsations. As for acceleration-induced high-fre-
ncy "B" and "C" pulsations, it was discovered sorne years ago that the pump
with a long suction system could be made to act as though it had its suction
close to the inlet. This was done by installing a lumped volume (vessel) wit
and out connections in a flow-through configuration, a baffle to further interfü
with passage of certain pressure waves, and a small gas-type dampener as pre
ously described. Such a device, called a suction stabilizer, was patented seve
years ago (Fig. 4.9). Additional advantages of a suction stabilizer will beco
evident from the following discussion.
Most liquids contain varying amounts of dissolved or entrained air or gas.
iow-pressure area is created in the pump cylinder, resulting in "breakout" of so
of this gas with subsequent lowering of volumetric efficiency. Another problem
the fact that these gases break out of the liquid with greater ease than they d
solve-it takes a longer time and higher pressure to redissolve them in the liq
Consequently, even though the pressure in the cylinder quickly builds up to
discharge pressure on the delivery stroke, sorne of the gas remains to account
sorne loss of displacement.
Since a low-pressure region exists near the pump inlet, most of the air or
tends to break out there. If space is provided in the upper part of the stabilizer
the fluid velocity is reduced to give time for good separation (by its large volu
most of this gas will migrate upward and accumulate there instead of procee
into the pump cylinder. Somewhat the same problem is introduced by liquids
inga high vapor pressure. Vapors can break out with the same results.
A secondary, and perhaps more noticeable, effect of air or gas in a suction 1i
is the tendency for the gas to accumulate in small pockets at the high spots in
system. The pockets eventually grow large enough to move through the pump
slugs and cause either momentar¡ or long-lasting air lock or loss of prime,
associated noise and knocking. Again, if a sufficiently large intemal space is
vided in the stabilizer where these slugs of air or gas can accumulate, they
never reach the pump.
What happens to all the gas accumulated in the stabilizer over a period of ti
Of course, if there is an excess it should be bled off through a vent. However
most cases, the pressure-smoothing ability of the stabilizer minimizes the l
pressure disturbances in the pump suction, which in tum minimizes breakout ·
small amount of gas will redissolve slowly and consistently and will be ca
through the pump without slugging.
Reciprocating pumps are inherently good pressure wave generators. All wav
tend to add or subtract to produce undesirable effects and loss of efficiency.
produced by one pump can be reflected back on each other to produce the
amplified effect as multiple pumps. As with electric devices, such interlering wa
can be prevented by' the use of simple impedance-transforming devices. In
hyraulic aspect, such transforming can be done by the shape, size, and arran
ment of connections and baffies. By flowing through the stabilizer at the Pif
suction, maximum filtering of interfering waves is had, particularly from o.
pumps in the system.
Again, the belief that additional suction pressure alone, provided by a charg
4.19 SURGE CONTROL IN WATER SYSTEMS* 127
pump or by greater head, will negate all of the problems of a ''poor:' suction
system is erroneous. On a long suction line the source of additional potential en-
ergy is so far away from the main pump that its effect is not fully realized because
of the time delay encountered. Al so, if a centrifugal charging pump is placed close
to the main pump, there is little chance for the constant delivery from the charge
pump to transform to the varying demand of the main pump. At those recurring
mome:ms of maximum demand of the reciprocating pump, the "constant" supply
of the centrifugal pump may actually act as a restriction. Sorne fluid "fiexibilty"
must be introd_uced. In both cases a properly designed stabilizer will provide the
necessary effect.
The advantage of using a gas-filled bladder in the suction stabilizer is that it
provides an energy-storing device to take care of the inherent low-frequency flow
characteristics of a reciprocating pump, and it is most desirable on multiple-pump
Remember, these are actual ftow variations and can be handled by
such a device. As previously mentioned, the large-volume ftow-through features
are required to control the high-frequency pressure wave disturbances resulting
from the effects of liquid acceleration disturbances.
·On pumps handling sand-laden mud or highly concentrated slurries, the need
for suction stabilization is just as important or more so. But a legitimate fear is
sometimes expressed regarding the use of a stabiiizer. Such devices have been used
successfully, and if the function is analyzed, fears of settling will be dispelled.
Even though the velocity through the device is reduced, the loss in velocity is
replaced by an increase in turbulence, which enhances the solids-carrying ability.
Also, even with a stabilizer volume of about 10 times the pump displacement per
revolution, all ofthe liquid in the device, say at 120 RPM pump speed, is displaced
every 5 s, hardly time for much settlement.
Finally, assumíng that there is sorne settlement, the increased velocity due to
reduced space would soon induce reentrainment of the settled solids.
SURGE CONTROL IN WATER SYSTEMS*
Pressure surge in liquid pipelines, or ''waterhammer,'' as this phenomenon is more

commonly known, is probably the most undesirable feature of any liquid piping
system. Waterhammer has been the greatest cause of pipeline failure, weakened
joints, loosened fittings, extemal leakage, severely damaged valves, and hammer-
noises followed by vibration of the piping. Noise and vibration are transmitted
to the entire structure of the building housing the piping and in sorne instances
even to the municipal piping system.
*This entire section is from the publication "Surge Control in Water Systems" by comtesy of the
author, Zeke Zahid, president of Zemarc Corp., Los Angeles, CA.
...
!f11\jllflfVt
~
163 PSI 78 PSI
20% PULSATION 10% RESIDUAL
l
/11/\lf\jl\ft\
163 PSI 98 PSI
(a) 1(b)
--~""'i",t"'I~~ 55 PSI
7% RESIDUAL
20 PSI ~
2.5% RESIDUAL
81 PSI
81 PSI
(e) (d)
Figure 4.19 Performance of different dampener types, 3 by 6 in triplex pump, 800 PSI
discharge. Top trace: Downstream of dampener. Bottom trace: Upstream of dampener. (a)
No dampener. (b) Acoustic-type dampener. (e) Gas-type dampener. (d) Greer tlow-through
dampener (Pulse-tone T• ).
No dampener
Acoustic filter
Combination of above volumes
o 50 100
Frequency of pressure disturbance, Hz
Figure 4.20 Attenuation characteristics of gas-type dampeners vs. acoustic filters. (Cour-
tesyWhite Rock Engineering, Inc.)
Waterharnmer is caused by the sudden stoppage or rapid deceleration of liquid

flowing in a pipeline, which results from the quick closure of a valve in the line.
This instantaneous ftow stoppage generates a pressure wave that propagates at the
speed of sound upstream from the valve until it reaches a larger-diameter riser or
tank; the wave is then reflected back to the valve, causing an increase in the line
pressure. This cycle is repeated ata regular frequency (depending on the length of
the pipeline and speed of pressure transmission) until the wave's energy is finally
dissipated as friction, heat, and elastic energy in the pipeline.
In order to select proper methods to alleviate the waterhammer, it is essential
to know its magnitude. The fundamental knowledge of the waterhammer phenom-
enon presented above can be applied in determining the intensity of shock or pres-
sure rise it causes.
Assume a simple water system consisting of a pipe, a water tank, anda quick-
closing shutoff valve. Under a free-ftow or steady-state condition (Fig. 4.19) the
water ftows from the reservoir through the piping at an initial constant velocity of
V. The water pressure at the valve is P. Now, if we consider the valve closing
completely and instantaneously, the following sequence of events takes place (Fig.
4.20). The ftow ofthe water mass immediately adjacent to the valve is completely
ecelerated to zero, and the unconsumed kinetic energy in the water is converted
o a pressure rise that compresses the water and expands the pipe walls. The trans-
ormation of kinetic energy takes place, generating a pressure wave that travels
pstream to the end of the pipe at sonic speed of propagation designated by a in
igure 4.21.
If the length of pipe is assumed to be L, then it will take L /a s for the pressure
Valve=~
===E =e 3
Figure 4.21 Steady-state condition.
wave to reach the end of the pipe. At this moment, water will be at rest, and the
water density, pressure, and pipe diameter will tend to be greater than those of the
normal free-ftow condition. When the pressure wave reaches the reservoir, it re-
bounds back into the pipe, traveling toward th<;: valve and arriving at the valve in
2L /a s after the valve closure. The water pressure is reduced to its normal level,
and the pipe retums to its original diameter. The time 2L /a is considered the
critica! valve closure time.
This surge phenomenon repeats with decreasing pressure amplitudes until the
total original kinetic energy is absorbed.
If the valve closure is accomplished in time t ~ 2L /a the closure is assumed
to be instantaneous. The pressure rise, therefore, can be expressed as
Pressure rise, 6.P = wavl l44g (4.26A)
where
!:J..P = pressure rise, PSI

v = initial velocity of water in the pipe before valve closure, ft/s
w = specific weight of water, 62.4 lb/ft3
a =
speed of pressure wave in water, which also depends on the pipe
material. However, an average value of 4000 FPS can be used.
g = acceleration dueto gravity, 32.2 ft/s 2
Therefore, for the water system, the maximum waterhammer pressure can be cal-
culated by modifying the pressure rise formula to
6.P = 53.8v (4.27A)
The maximum system pressure is
Pmax = 6}J + P, (4.28)

Pressure wave
(Propagating at speed A)
Quickly closed valve
~----- ......
-· 1 ;( .$ _,}--
1
1 -
--
..........
---___ -- --
_,,.
1 ""' """"' ' --
Figure 4.22 Sudden stoppage of water ftow (creation of waterhammer).
where
P5 = normal static water pressure, PSI

Pmax = 53.8v + Ps
For example, in a water system with normal pressure of 60 PSI and velocity of
5 ft / s, the maximum system pressure, as a result of quick closure of valve, would
be
Pmax = (53.8 X 5) + 60 = 329 PSI
This example clearly demonstrates that waterhammer can easily create a pres~
sure rise of more than five times normal system pressure. Figure 4.22 is a chart
that can be used to determine the maximum system pressure of a water system
once the flow velocity (flow rater divided by pipe area) and pressure have been
determined.
Once a basic understanding of the phenomenon of waterhammer has been
achieved, an examination of various methods of control is in order.
4.19.1 Methods of Surge Control in Water Systems

Several methods can be employed to reduce waterhammer pressure.
Waterhammer is generated by the rapid deceleration of water flow, and the
resultant pressure increase is proportional to the rate of deceleration of the water
flow in the pipeline. Consequently, if moving water is stopped gradually, the pres-
sure rise will be considerably smaller than if the water flow is stopped suddenly.
Thus, an obvious method to reduce waterhammer is to close the valve gradu-
ally. A valve can be closed rapidly during the initial closing stages but slowly
toward the end to bring the moving water to a more gradual stop. The length of
the pipe and the velocity of the water determine the rate and degree of closure
necessary to prevent waterhammer. The longer the pipe and the higher the veloc-
4760.6
4000
3500
"'
~
ro
(¡)
> 29.5
~ 3000
!!'. 24
:::>
"'
"'e
D.
o 2500 15
~
·g 12
03 11
>
10
2000
1500
3.4
25 50 75 100 125 150

Pipe inside diameter/Pipe wall thickness
Figure 4.23 Pressure wave velocity, water. Numbers to right of curves indicate modulus.
of elasticity in millions of PSI units. C, = 0.91.
the slower the valve closure should be. However, one can readily see that this.
does not o:ffer a very practical or accurate solution.
The second method to reduce waterhammer is a simple bypass or a relief device
(Fig. 4.23). These devices basically relieve the excessive pressure by discharging
the decelerating water volume either to the atmosphere or into the piping down- •
stream ofthe rapidly closing valve. In sorne cases, the water is discharged into the
supply tank through a line connecting the relief valve to the tank. This method
not very reliable and requires frequent inspection and maintenance.
Air chambers also offer a solution for the waterhammer problem. An air -·.,~·-· .. ., .•
ber is simply a tank or a large pipe with its top closed to prevent the release of
entrapped above the water level (Fig. 4.24). Air within the chamber occupies only
4.19 SURGE CONTROL IN WATER SYSTEMS 133
wlOOO~~~-.-~~~.,-~~--.-~~~·..-~~--.
:;
Vl
ti)
~
Q.
~ 800F=~~--+-~~~+-~~--1-~~---1
:;¡
"'e
!JI)
'5
:;¡
] 6001--~~-+~-=""""......=='---+~~~-i--~~--1
iñ
~
\':'.
~ 400r-~~-r-~~-=-¡-~~-t~~~T-~~--i
"'
¡::
Q.
E
"'
tí
~ 2001--~~-+-~~~+-~~-1-~~~+-~~--1
E
:::l
E
·x
"'
::¡;;
o 30 60 80 120
Water ilow pressure, PSI
Figure 4.24 Maximum system pressure. A = 15 ft/s, B = IO ft/s, C = 5 ft/s.
25 or 20% of the total chamber volume, depending on the water system pressure.
The air inside the aír chamber accommodates the kinetic energy i.n the system by
compressing. Kinetic energy is thus converted to potential energy. However, if
the system pressure reduces below the design pressure, trapped air in the air cham-
ber will discharge into the system, thus rendering the entire device ineffective.
In addition, the compressed air in direct contact with the water tends to slowly
dissolve into the water, which also diminishes the device's surge-dampening ca-
pability. Then, when the system pressure is reduced, the dissolved air resumes its
gaseous fonn, causing undesirable sponginess in the system and possible damage
to the system components.
The shortcomings and deficiencies of air chambers are e:ffectively eliminated in
gas-loaded, flexible, separator-type surge suppressors (Fig. 4.25). Gas confined
inside a flexible bladder separator provides an efficient means of transfonning the
system's kinetic energy into potential energy. A gas-filled separator-type surge
suppressor is installed upstream and close to the valve, which is the source of the
waterhammer (Fig. 4.26). In the event ofthe valve's sudden closure, the suppres~
sor accommodates the abruptly stopping column of water immediately adjacent to
the valve (Fig. 4.27).
As the system pressure increases after quick valve closure, the gas, which is
indirectly in contact with the water through the flexibie bladder, is compressed,
thereby absorbing and suppressing the high-pressure surges that would otherwise
be dissipated only after they had been detrimental to the system's piping, com-
ponents, and ultimate performance.
Relief
val ve
By-pass
line
Tank
Quick-Closing valve
Figure 4.25 Relief valve upstream of the shutoff valve.
Air-pressurizing
valve connection Gas charging valve
Bladder:
BUNA N
Shell:
Alloy steel
Figure 4.26 Nonseparator-type Figure 4.27 Gas-loaded flexible separator~

surge suppressor. type surge suppressor.
4.19.2 Sizing of Surge Suppressors

The size of a surge suppressor must be properly calculated, taking into account all
the effective system parameters, if the unit is to perform with optimum efficiency.
The following simplified explanation illustrates the method of determining the surge
suppressor capacity for water systems.
The analysis is based on two assumptions: (1) friction losses are small and may
be disregarded and (2) the energy absorbed by compressing the column of water
and expanding the pipeline is very small compared to the energy stored and ab-
sorbed by the surge suppressor and, consequently, is also a negligible factor. The
sizing of the surge suppressor, based on these assumptions, offers a conservative
answer, which is desirable since all the system variables cannot be known or taken
into account.
The water ftowing in a pipeline possesses a finite amount of kinetic energy (KE)
which can be expressed as
KE wALv 212g (4.29)
where
w = specific weight of water, 62 .4 lb/ ft 3

A = effective area of pipe, ft2
v = initial velocity of water at normal ftow, ft/s
L = length of pipe, ft
g = acceleration due to gravity, 32.2 ft/s
When the system ftow is stopped abruptly, this kinetic energy has to be transferred
to the gas inside the separator. The gas follows Boyle's law:
(4.30)
P1 = initial gas pressure (precharge), the same as system pressure, PSIA

P2 = allowable surge pressure, PSIA
V1 = surge suppressor volume (gas volume inside suppressor), in3
V2 = gas volume at the allowable surge pressure, in3
n = polytropic exponent of expansion of the gas ( nitrogen = 1.4)
Therefore, to achieve energy balance in the system before and after the valve
closure, the k:inetic energy of the system before the valve closure is equal to the
energy stored in the suppressor gas volume V1 between the pressure limits P 1 and
Pz.
Hence,
(4.31)
Surge
suppressor
Valve open
Figure 4.28 Free-flow surge suppressor installed in line.
Surge
suppressor
Quickly
closed valve
Figure 4.29 Sudden stoppage of flow (suppression of waterhammer).

4.19 SURGE CONTROL IN WATER SYSTEMS 137
1.4
l
1.2
\
1.0
i
.8
\
e
.6 \
.4
~
.2
""'!--_ !----__
o
1.00 1.251.50 1.75 2.00 3.00 3.50 4.00
Figure 4.30 Pressure constant C curve.
For the water system,
4.65ALv 2
(4.32)
To reduce mathematical computations, this formula can be further simplified to
(4.33)
The value of C, based on the pressure ratio Pi/ P¡, can be readily determined from
Figure 4.28. A cakulation illustrating the use of the above formula is given at the
end of this discussion.
4.19.3 Conclusion
One aspect of surge control not directly connected with maintenance cost savings
bears mention. Because of new safety codes and noise control laws and regula-
tions, surge control in water systems may soon become a necessity. Thus, water
system engineers and designers should become thoroughly familiar with pending
legislation a:ifecting their current and future efforts. Such legislation is an expres-
sion of the growing concern over noise pollution and industrial safety, and costly
retrofit programs may be avoided by careful consideration of surge control during
the initial system design.
4.20 SAMPLE PROBlEM
Problem. Determine the surge suppressor capacity required to Hmit the maximum
surge pressure to 125 PSIA in a water supply system with the following parame-
ters:
Pipe length, 1200 ft; pipe size, 8 in, schedule 40

Pump flow rate, 1500 GPM
System pressure prior to valve closure, 80 PSIA
Liquid pumped, water at ambient temperature
Valve closure time, 0.4 s
Solution. The necessary surge suppressor size is determined by the equation (4.10):
4.65ALv 2 ·
where
V1 = surge suppressor capacity, in3

L = 1200 ft
A = 0.348 ft 2
v flow velocity
=
= flow rate / ftow area = 1500 X 232 / 50 X 720
= 9.67 FPS
P 1 = normal system pressure = 80 PSIA
Pmax = maximum shock pressure = 53.8v + P 1
P2 = maximum allowable surge pressure = 125 PSIA
te= critical time= 2L/a = 2 X 1200/4000 = 0.6 s
Since the valve closure time is less than the critical time, this condition can be
treated as an instantaneous valve closure (the valve doses before the pressure wave
retums to the valve).
p max = 53.8 X 9.67 + 150 = 670 PSI
Therefore,
4.65 X 0.348 X 1200 X (9.7/
286
80[(125/80)" - 1]
Vi= 16,690 in 3 (72 gal)
An 80-gal capacity surge suppressor is required.
4.20 SAMPLE PROBLEMS 139
lternate Method. The size of the necessary surge suppressor can also be deter-
mined by equation (4 .11)
(4.35)
bere C is the pressure constant determined from Figure 4.28 for a corresponding
tío of P2 /P 1 • When Pi/P 1 = 1.56, C = 0.54. Therefore,
V¡ = 62.4 X 0.348 X 1200 X 93.5 X 0.54 /80

= 16,690 in3 (72 gal)
n 80-gal capacity surge suppressor is required.
FINAL IMPORTANT NOTES
word about the most severe type of pulsation in the discharge of a reciprocat-
g pump-the 'flow-variation-induced' pulsations inherent in the pump itself.
gure 4.2. Gas-type dampeners work on the principie that they prevent the for-
ation of pulsations in the entire system, including the pump, by the simple
cess of 'smoothing' out the entire discharge system ftow-variations themselves
that the cause of pulsation is removed.
Pure 'acoustic' type dampeners not only allow the pulsations to be formed,
n to be partly 'filtered' out of the downsteam side of the system only-they
metimes add resistance to that flow in the form of chokes ·and 'tu bes' to be
ded to the power load of the pump. Such devices must be large with respect
the pump.
PUMP DESI N
5.1 RATING STANDARDS*
The following definitions are reprinted from Hydraulic Institute Standards, 1985,
by courtesy of the Hydraulic Institute. The purpose of this section is to define terms
used in pump ratings. These ratings are characteristics of pump design and not
conditions of the specific application.
STROKE. One complete uni-directional motion of piston or plunger. Stroke length
is expressed in inches.
PUMP CAPACITY (Q). The capacity of a reciprocating pump is the total volume
through-put per unit of time at suction conditions. It includes both liquid and any
dissolved or entrained gases at the stated operating conditions. The standard unit
of pump capacity is the U. S. gallon per minute.
PUMP DISPLACEMENT (D). The displacement of a reciprocating pump is the volume
swept by all pistons or plungers per unit time. Deduction for piston rod volume is
made on. double-acting piston-type pumps when cakulating displacement. The
standard unit of pump displacement is the U.S. gallon per minute.
For single-acting pumps:
Asnm
D=-- (5.1)
231
*Nomenclature in Section 5.1 may not be consistent with that in the main text.
140
5.1 RATING STANDARDS 141
For double-acting piston pumps with no tail-rod(s):
(2A - a)snm
D = 231 (5.2)
where
A = plunger or piston area, square inch

a = piston rod cross-sectional area, square inch (double-acting pumps)
s = stroke length, inch
n = RPM of crankshaft
m = number of pistons or plungers
PLUNGER OR PISTON SPEED (v). The plunger or piston speed is the average speed
of the plunger or piston. lt is expressed in feet per minute.
ns
V= - (5.3)
6
PRESSURES. The standard unit of pressure is the pound force per square inch.
Discharge pressure ( Pd ). The liquid pressure at the centerline of the pump

discharge port.
Suction pressure ( Ps ). The liquid pressure at the centerline of the suction port.
Differentia/ pressure ( Ptd ). The difference between the liquid discharge pres-
sure and suction pressure.
Net positive suctiOfJ head required (NPSHR). The amount of suction pres-
sure, over vapor pressure, required by the pump to obtain satisfactory volumetric
efficiency and prevent excessive cavitation.
The pump manufacturer determines (by test) the net positive suction head re-
quired by the pump at the specified operating conditions.
NPSHR is related to losses in the suction valves of the pump and frictional
losses in the pump suction manifold and pumping chambers. Required NPSH does
not include system acceleration head, which is a system-related factor.
SLIP ( s ). Slip of a reciprocating pump is the loss of capacity, expressed as a frac-
tion or percent of displacement, dueto leaks past the valves (including the back-
fiow through the valves caused by delayed closing) and past double-acting pistons.
Slip. does not include fluid compressibility or leaks from the liquid end.
POWER (P). Pump power input (P¡ )-The mechanical power delivered to a pump
input shaft, at the specified operating conditions. Input horsepower may be cal-
culated as follows:
142 PUMP DESIGN
P; = --~-
Q X Pu1
(5.4)
1714 X Y/p
Pump power output ( P0 ) - The hydraulic power imparted to the liquid by the
pump, at the specified operating conditions. Output horsepower may be cakulated
as follows:
p = Q X P1c1
(5.5)
() 1714
The standard unit for power is the horsepower.

EFFICIENCIES ( 11 ). Pump efficíency ( r¡P), (also called pump mechanical effi-
ciency)-The ratio of the pump power output to the pump power input.
pº
(5.6)
P;
Volumetric efficiency ( r¡ v ) -The ratio of the pump capacity to displacement.
Q
r¡,. = /5 (5.7)
PLUNGER LOAD (SINGLE-ACTING PUMP). The computed axial hydraulic load, act-
ing u pon one plunger during the discharge portion of the stroke is the plunger load.
It is the product of plunger area and the gauge discharge pressure. It is expressed
in pounds force.
PISTON ROD LOAD (DOUBLE-ACTING PUMP). The computed axial hydraulic load,
acting upon one piston rod during the forward stroke (toward head end) is the
piston rod load. It is the product of piston area and discharge pressure, less the
product of net piston area (rod area deducted) and suction pressure. It is expressed
in pounds force.
5.2 FAMllY PLANNING
It is usually anticipated that a family of pumps of any one type wiU be produced
in a series of horsepower sizes. The éhoice of sizes should follow sorne order of
progression, and the geometric series seems desirable. For example,
n, nr1 , nr 2 , nr 3 , nr4 ; etc. (5.8)
where
n = base horsepower (BHP)

r = progression ratio
5.3 WINDOWS OF NONUTILIZAT!ON 143
TABLE 5.1. Geometric Si:ze Progression
n = 100 = 100 BHP n4 = 100 X 5.1 = 510

n 1 = 100 X 1.5 = l 50 n5 = 100 x 7.6 = 760
n 2 = 100 X 2.3 = 230 n6 = 100 X 11.4 = 1140
n3 = 100 X 3.4 = 340 n1 = 100 X 17. l = 17 10
TABLE 5.2. Arithmetic Si:ze Progresslcm
n = = BHP n4 = 400 + 100 = 500

n 1 = 100 + 100= 200 n5 = 500 + 100 = 600
!12 = 200 + 100 = 300 11.5 = 600 + 100 = 700
n3 = 300 + 100 = 400 !17 = 70.0 + 100 = 800
With a base of 100 BHP anda progression ratio of 1.5, each term is 50% larger
than the previous term (see Table 5.1).
An arithmetic series may sometimes be desirable. This is
n; n 1 = n + d; n2 = n1 + d; = n2 + d; n4 = n3 + d; etc. (5.9)
where
n = base horsepower (BHP)

d = progression difference
Table 5.2 illustrates the first seven members of an arithmetic series with a basen
of 100 BHP anda difference d of 100.
5.3 WINDOWS OF NONUTIUZATION
.Predictable pumping requirements include a wide order of displacement and pres-

sure pararneters. This unavoidably large variation of and mandatory economic limit
on the number of pumps in a family contribute to the complexity of the family-
size selection process.
Sometimes a project may warrant the design and manufacture of a pump of a
particular size to exactly fit that project. This is true in the long slurry pipeline
that entails a great number of large pumps and a long lead time from conception
to completion.
Nevertheless, there are unavoidable gaps or "windows of nonutilization" in
the overlapping pump specifications, as shown in Figure 5.L These are regions in
the typical pressure vs. displacement curve for each pump of a family where op-
eration is impossible on account of the pump overload. The chart shows the in-
144 PUMP DESIGN
Maximum pump pressure (minimum piston diameter)

1500
1400
1200
800
600
5001----L~~--:'-_.__._..___.__,___;,,._....L...l-,--~""""~....i.~~-L.::,,,_--I
Minimum pump pressure (maximum piston diameter)
100 200 300 400 500 600 700 800 900

GPM
Figure 5.1. Performance characteristics of a typical family of reciprocating pumps. Win

dows of nonutilization (shaded areas).
herent disparity in displacement vs. pressure characteristics of a family of fo*

similar pumps of any practica! progression.
For example, if a pump requirement falls at point A (280 GPM at 1300 PSÍ
just outside the 210-BHP envelope, one would be forced to choose a 320-BH
pump ( point B) and accept a certain amount of first-cost penalty of excess installe
~; f;
power.
5.4 PISTON ROO LOAD
There is no absolute rule for selecting the piston rod load (PRL) for a particul
pump design. However, the nature ofthe mechanism associated with reciprocaf
pumps of all types along with past practices dictate that the PRL be directly rela
to the hydraulic horsepower (HHP) in about the following ratio:
192 X BHP
PRL=----
n
where n = number of cylinders

An exception to this rule is the case of multicylinder plunger pumps where
same reciprocating parts, both power-end and liquid-end, are used for econom
advantage in all of the cylinders of the triplex, quintuplex, septuplex, and non
plex. In those instances, the PRL is based on the original triplex design and i
5.4 PISTON ROD LOAD 145
TYPICAL
SMALL PIN FDR RETURN STRDKE
FASTENING
r----".--~~~~~~--~~~~------¡
/ 1
DIRECTIDN DF LOAD - - - -
_J
CONNECTI NG ROD
END OF CONNECTING RDD

GROUND TO BEARING BEARING
FINISH.
Means of Increasing Crosshead Pin Bearing Load on Single Acting Pumps
used for the other multicylinder pumps. The horsepower increase in each succeed-
ing increase in the number of cylinders comes from this addition of cylinders
rather than from an increase in PRL.
The PRL is a function of piston area AP and the discharge pressure:
(5.1 l)
where Pd = maximum allowable pressure, PSI
In double-acting pumps, the piston-rod diameter reduces the effective area of

the piston on the cmnk-end stroke. The pump must be designed on the maximum
PRL seen on the head end of the piston where the full area is subjected to the
discharge pressure.
A high suction pressure on a double-acting pump will tend t0>reduce the PRL,
but this should not inftuence the design of the pump, because the actual suction
pressure cannot be predicted or guaranteed.
Unusual Pressure in the Pump Cyiinder
Naturnlly, any excessive pressure in the pump cy!inder will show up in excessive
bearing loads and the question arises as to what maximum pressures can nor-
mally be expected. An investigation started about 35 years ago on the failure of
íluid ends on pumps due to corrosion-fatigue, with sorne statements that, "Pres-
sure as high as seven times the average discharge (design) pressure have been
measured". This led first to a study of the best means of measuring these pres-
sures and the conclusion was that the strain gauge type of pressure transducer
146 PUMP DESIGN
with a small, ftush, sensing diaphragm mounted directly into the pump cylinder
head without any connecting pipes, valves, etc., is required. Then the pressure
read-out should be by means of an oscilloscope to eliminate any mechanical (in-
ertial) effects associated with any strip-chart type of recorder. With literally thou-
sands of oscillographs taken in the cylinder of 1· BHP to 1700 BHP pump in
every conceivable service, it can be said that the normal cylinder over-pressure
seldom exceeds 10% at the beginning of each pressure stroke.
Now there are occasions when a poor suction system or other contributions
can cause an increased over-pressure but ·even then, the worst that has been ob-
served is about 100% and in those cases corrective measures had to be taken.
While on this subject, it is important to note that a typical cylinder pressure
is theoretically a "good" square wave which means that the cylinder (and the
associated bearings) "see" the full discharge pressure for almost ali of the 180
degrees of discharge stoke. This is a departure from the less severe pressure
waveform seen in gas compressors and interna! combustion engines.
lt is also of interest that the "overshoot" pressure, sometimes seen in the
cylinder, is usually not seen in the discharge pressure waveform. A good example
is shown in Figure 4.3, Parts A and B. Never-the-less, the crankshaft and bear-
ings "see" this elevated pressure load.
5.5 MAXIMUM PISTON DIAMETER
There must be sorne size limit to the piston diameter, but here again there is no
absolute rule. One guideline is that the practica! diameter of the piston should be
no greater than the pump stroke length s, thus setting the ratio of maximum dis-
placement to maximum pressure limits:
Dmax =S
A pump with such a piston diameter is known as a '' square'' pump. With pump~
having a range of replaceable and multisize liners and pistons, the PRL /MAP
relation must be maintained. The design approach in that case is to assume an
arbitrary piston diameter of approximately half-stroke length:
Dave = s/2
Maximum limits of PRL dictate that for extremely high pressure pumps (abov~
10,000 PSI MAP), a drastic reduction in the plunger size/stroke length ratio is
required.
5.6 STROKE LENGTH
The selection of a design stroke length is roughly related to accepted past practices,
dictated by the mechanisms associated with reciprocating pumps, usually resulting
5.8 SPEED 147
TABLE 5.3. Pl1mp Stroke length
PumpBHP Stroke, in Pump BHP Stroke, in

----~--
50 4 400 10
100 5 800 13
200 7 1600 18
in the "square" pump limitation previously with the maximum stmke

about equal to the piston diameter. St.roke length is more closely
related to pump horsepower in the arbitrary manner:
s = 0.7(BHP)°" 44 (5.14)
But sorne departure will be encountered.

This relation is illustrated in Table 5.3 for sorne common pump horsepower
values.
See also the relation of pump RPM to stroke in Sections 5.9 and 5.10.
5. 7 NUMBER OF CYUNDERS
Duplex double-acting pumps with four pressure strokes per cyde would naturally
require 90º crank angle spacing of the two throws to generate a tolerable flow
variation pattem.
Multicylinder pumps can be designed with any number of cylinders, but for
most efficient distribution of flow variations per cycle, the use of an odd number
of cylinders, namely 3, 5, 7, or 9, is desired. Even numbers of cylinders result in
superjacency of two or more flow pattems per cylinder, which in turn result in
emphasized peaks that generate greater pressure pulsations.
Sometimes the main reason for using a multicylinder pump is its more desirable
flow pattem, with lower values of flow valiation and subsequent pulsation. As
illustrated in Table 5.4, even-numbered cylinders have an inherently greater vali-
ation.
5.8 SPEED
There is a great temptation to increase the horsepower rating of pumps by simply

increasing the speed. An existing pump can be uprated only slightly in speed. At
a higher speed the liquid velocity through the valves may be greatly exceeded, to
the extent that hydraulic difficulties would be encountered. The. valves in an ex-
isting pump were probably designed for a practica! maximum velocity as deter-
mined for its original rated pump speed and piston size.
... TABLE 5.4. Pump Type flow Variation
t
Pump Type Cyls Angle A(I) B(2) Percent K(4) Percent(5)
Simplex SA 1 360 0.58 1 158 0.684 250
Simplex DA 1 180 0.29 1 129 1.368 166
Duplex SA 2 180 0.29 l 129 0.558 166
Duplex DA 2 90 0.24 0.22 46 0.199 93
Triplex SA 3 120 0.06 0.17 23 0.100 43
Triplex SA (0) 3 120 0.062 0.17 23 O.lOO 44
Triplex SA (30) 3 120 0.061 0.12 18 0.078 35
Triplex SA (60) 3 120 0.05 0.09 14 0.06! 27
Triplex DA 3 120 0.06 0.17 23 0.200 43
Quadruplex SA 4 90 0.11 0.22 33 0.143 62
Quadruplex DA 4 90 0.11 0.22 33 0.286 62
Quintuplex SA 5 72 0.02 0.05 7 0.030 14
Quintuplex DA 5 72 0.02 0.05 "7
1 0.076 14
Sextuplex SA 6 60 0.06 0.17 23 0.100 43
Sextuplex DA 6 60 0.06 0.17 23 0.200 43
Septuplex SA 7 51.4 0.012 0.026 3.8 0.016 8
Septuplex DA 7 51.4 0.012 0.026 3.8 0.032 8
Octuplex SA 8 45 0.026 0.052 8 0.035 15
Octuplex DA 8 45 0.026 0.052 8 0.070 15
No1mplex SA 9 40 0.006 0.002 2.1 0.009 2
Nonuplex DA 9 40 0.006 0.002 2.1 0.018 2
--
NOTES;
(1) - "A" Flow variation above average, decimal. (5) - Maxirnum pressure variation (pulsation) without dampener, percent.
(2) - "B" Flow variation below average, decimal. (0) - Fixed phase angle of two pumps compounded.
(3) - Flow variation, total percenl (30) - Fixed phase angle of two pumps compounded.
(4) - K Factor for calculation of dampener size (60) - Fixed phase angle of two pumps compounded.
(Equation 4.5)
5.10 SPEED OF MULTICYLINDER PUMPS 149
TABLE 5.5. RPM of Dup!ex Pumps
Stroke, in RPM Stroke, in RPM

,.,,_ 600 12 100
4 300 14 86
6 200 16 75
8 150 18 67
10 120
However, dueto competition and economic reasons, there has been a tendency
to design greater speed into pumps, and the hydraulic limit has probably been
reached. Greater speeds will no doubt require drastic design approaches, such as
positively actuated valves, tending to offset any economic advantages associated
with the speed.
The greater the pump speed (RPM), regardless of stroke length, the greater the
detrimental effect of high acceleration pressure at the pump suction and discharge.
In other words, maintaining a constant piston speed (FPM) by choice of RPM vs.
stroke length results in constant displacement in GPM, but the short-stroke high-
speed pump suffers most from acceleration problems.
It has also been shown that liquid-end parts life is exponentially reduced by
increases in RPM and/or in number of reversals. See Chapter 10, Parts Wear and
Life.
5.9 SPEED OF DUPLEX AND SIMILAR PUMPS
Aside from the hydraulic limiting factors affecting maximum RPM, the imbalance
of pump crankshafts with nonsymmetrical throws, such as the duplex double-act-
ing with two throws at 90º separation, creates intolerable rotary imbalance. Such
pumps must be limited to a relatively low RPM, because the e:ffects of the unbal-
anced forces (centrifuga!) increase to the power 2 with rotating speed. Accord-
ingly, a practical speed rating for such pumps is
(5.15)
where FPMd = 200

Table 5.5 was constructed on that basis.
5.10 SPEED OF MULTICYLINDER PUMPS
Pumps with symmetrical crankshafts (equally crank angles) such as a du-

plex single-acting and all multicylinder pumps can be operated at much higher
150 PUMP DESIGN
TABLE 5.6. RPM of Multi-Cylinder Pumps
Stroke, in RPM Stroke, in RPM
2 600ª 12 160
4 450 14 130
6 300 16 112
8 225 18 100
10 180
ªNote: An arbitrary limit has been set at 600 RPM for stroke of 3-in. or less.
speeds because of the inherent static balance of the crankshaft, but now being
limited by the hydraulics.
Liquid flow or friction pressure increases directly with velocity. Liquid accel-
eration pressure increases as the second power of crankshaft rotating speed. Both
of these factors affect the volumetric efficiency of the pump, and there must be
sorne limit to the maximum speed at which the pump can operate.
Due to economic factors and competition, the industry has seen a gradual in-
crease in the maximum allowable speed (MAS) for multicy linder pumps. lt appears
that a limit has been reached beyond which the problems associated with high
speed cannot be tolerated. In fact, sorne manufacturers have followed a trend to
lower speeds. Speed reduction is desirable in many cases where suction conditions
are minimal. See Chapter 2, Dynamics.
An acceptable range of basic speeds for multicylinder pumps is determined by
the following formula, based on a constant piston speed of 300 FPM:
RPMMax. = FPMs/(2s/12) (5.16)
where FPM5 = 300

Table 5.6 illustrates this relationship.
For further discussion of the effects of pump speed see Section 10.1.4 Chapter
10. Parts Wear and Life.
5.11 DIRECTION OF ROTATION
5.11.1 Rotation "Overrunning"

In an overrunning pump the crank rotation is such that the crank approaches the
crosshead from the top of the rotation circle as shown in Figure 5.2. This motion
is preferred for horizontal pumps because the connecting rod force component R
and the weight component W are always directed downward. To resist the normally
high crosshead forces, more rigidity can be built into the lower crosshead guide,
Crank Head
end end
(a)
(b)
Figure 5.2. Direction of rotation. (a) Overrunning. (b) Underruning.
l'lhl -
PIN .. LIJAD?~···················
BEARING
CRDSSHEAD PIN ANO BEARING ASSEMBLY VITH

CLEARANCE DBTAINED VITH PARTS OF DIFFERENT
SHDVING CEXAGGERATED J THE ºLINE CDNTACTº
PRESSURE RESULTING IN EXTREMELY HIGH
CDMPRESSIDN STRESS FRDM APPLIED LDADS.
THE MDDULUS DF ELASTICITY DF THE METALS
DF BDTH PARTS CTHAT OF THE BEARING BEING
THE LDVEST J ALLOVS AN ºAREA CDNTACT .
TO BE GENERATED
Figure 5.2A. Oscillating Crosshead Bearing Arrangement.
151
152 PUMP DESIGN
since it is tied into the base of the power end. Overrunning also minimizes cross'
head slap or knock.
5.11.2 Rotation "Undem.mning"

Rotating in the opposite direction, an underrunning purnp, in Figure 5.2(b) has
force component R that is always directed upward while the weight component
is always downward. These altemating forces can cause crosshead slap or kno
Extremely close crosshead and guide clearances must therefore be incorporated.
5.12 OFFSET CRANKSHAFT
As described in Section 5 .11, all crosshead forces or loads are usually direct
downward in a horizontal reciprocating pump. The offset crankshaft (Fig. 5.
causes a modification of these loads in such a manner that the maximum downw ·
load at the center of the stroke is reduced by an amount that is transferred to ·
upper guide at the beginning and end of the stroke (Fig. 5.4). A slight impro
ment in mechanical efficiency should be expected, but there is the possibility'
crosshead slap or knock at the points of load reversal.
The offset crankshaft offers little if any advantage with respect to the
standpoint. The acceleration at the ends of each stroke is not altered, and the
pattern change is almost undiscemible.
The following special formulas apply to the offset crankshaft:
X= r(l - cos 8) + (r sin()+ h) 2/2Lc
a = arctan [ ( r sin () + h) / ~L~ - ( r sin () + h) 2 ]
90°
Figure 5.3. Offset crankshaft.

5.13 CONNECTING ROD FORCES 153
UPPER GUIDE
~UNNING CLEARANCE
SH1JE
LO\.JER GUIDE
UPPER CLEARANCE
¡ EXAGGERATED J
IJRDNG RIGHT
SHOES TURNED TO SH!JES TURNEO TO
D!AMETER LESS THAN Sl\ME D l AME TER AS
GUIDE Dll\METER CR!JSSHEAD !lDRE .
RESULTS IN L!NE RESUL TS lN 1007.
CONTAtH BEAP.INu CONTACT.
'"-----'
Figure 5.3A. Shim-Adjusted Crosshead Shoe and Guide Fits
w = 27íN /60 (5. l 9)

r sin ( () + o:)
V=-w . (5.20)
12 sin (1.57 - a)
r h \
a = 0.084rw 2 ( cos () + - cos 2e + - sin fJ) (5.2l)
Le Le /
R = (tan a)PRL (5.22)
CONNECTING ROO FORCES
lntroduction
fundamental mechanism of the power end of a reciprocating pump is the crank-
nnecting rod-crosshead system. The design of a crankshaft provides single or
154 PUMP DESIGN
2" Offset crankshaft

Underrunning
~ c.
.!+1--~~,.......+~--+~-+-....,...'"+--~
~
al to e
....~+-~+--~I----+~--+~~
l=----l~---l-...::::,,....¡.~::::..¡_~--1--~-t=-~J-~-l------ll-----l-~---+~-I
"' ;::
b8
No offset
Overrunning
180 210 240 270 300 330 180 210 240 270 300 330 360
Crank angle, deg
Figure 5.4. Offset crankshaft, crosshead loads.
multiple throws or crank pins, the axis of which are on the desired distance (radiu
from the axis of the main shaft.
Many forms of crankshaft configuration are used, most representing differen
methods of manufacture such as forged, cast and machined-from-billet and eve
bolted assemblies. Figure 5.5 shows the general shape of several crankshaft de
signs, all having opinionated features.
The basic geometry of a crank-connecting rod-crosshead system is shown i
Figure 5.6. The pump mechanism forces are shown in Figure 5.7 for a conven
tional crank mechanism and in Figure 5.8 for an articulated mechanism.
5.13.2 Formulas
The following formulas are used to calculate the crank mechanism forces and a
example is shown in Figure 5. 9.
Example: Triplex Single-Acting; Type 111 Crankshaft

See Figure 5. 7.
r = 3 in; Le = 18 in; PRL = 10,000 lb
Pd = 1415 PSI; Ps = 100 PSI

V
~ Bearing Gear
~ Frame
Cast eccentric
Multicylinder Barrel type
Single- or double-acting Duplex only
11 VI Gear
Shaft-eccentric Overhung crank

Multicylinder Duplex only
Single- or double-acting
111 VII
Cast marine type Modular eccentric

Multicylinder Multicylinder
Single- or double-acting Single- or double-acting
Optional full or VIII
half "back-up"
journals
Forged marine type lntermediat.e journal

Multicylinder
Single- or double-acting
Figure 5.5. Crankshaft types.
155
StrokeS, in Outlet connection
pipe size, in
270°
Piston diameter, in
180°
90°
pipe size, in
Figure 5.6. Pump geometry.
<"O ::o
CD O CD
B
g:~~
Q) o-·
-:::i:::i
c -
"O O -
::T
3., u;·
"O Discharge press
!"
A :.._,,,.,...,.~~:-:---:--::-~~::.!!.~~-=¡:;:;o¡R't--t-HE:""j'P~:'re
Piston diameter
Oº - 180° Cycle:
Suction: lnside crosshead
Discharge: Outside crosshead
(Double-acting pumps: Subtract suction
pressure on CE from discharge pressure on HE.)
Figure 5. 7. Pump mechanism forces.
156
5.13 CONNECTING ROO FORCES 157
a1 = arcsin (r sin 8/Lc) = arcsin (3 sin 86º)/18 = +9.57º (5.23)

PRL = 0.7854 X D 2 X PSI = 0.7854 X 32 X 100 = +707 lb (5.24)
CRL = PRL/cos a = 707 /0.986 = +717 lb (5.25)
XHL = CRL sin a= 717 x 0.166 = +119 lb (5.26)
Nonnal load = CRL cos (8 +a) = 717 (-0.097) = -70 lb (5.27)
Tangent load = CRL sin (8 +a) = 717 (0.995) = +714 lb (5.28)
where
Symbols:
a-rod to crosshead axis angle, deg.
{3-rod to crank angle, deg.
0-crank rotation, deg.
L,.-connecting rod length, in
r-crank radius, in.
Displacement
-- -- - -- Displacement
Articulated end Regularend
Figure 5.8. Articulated pump mechanism forces. a = angle between rod and crosshead
angle, deg; fJ = angle between articulated rod and crosshead angle, deg; O = crank angle,
deg; Le = connectiQg rod length, in; L 0 = articulated rod length, in; r 1 = crank radius, in;
r2. = effective radius of articulated rod, in; e = connecting rod extension, in;
r2 = sin a(Lc + e)
fJ = arcsin (ri/L0 )
a = arctan (ri/Lc)
Suction,
PRL = +7071b 100 PSI
(a)
Discharge,
1415 PSI
o2 = o1 + 120º = 206º
(b)
PRL = 10,000 lb
(e)
Figure 5.9. Connecting rod geometry offorces for Example A: triplex single-acting, Type
III crankshaft. See Section 5.13 for calculation. (a) Forces on connecting rod l. (b) Forces
on connecting rod 2. (e) Forces on connecting rod 3. Consider crosshead weight if signif-
icant.
158
5.14 CRANKSHAFT BEND!NG MOMENTS-CALCULATIONS 159
D = piston diameter, in.

PRL = Piston rod load, lb
CRL = Connecting rod load, lb
XHL = crosshead load, lb
PSI = pressure, PSI
Rod 2, 82 = 86º + 120º = 206º
XHL = -733 lb
PRL = 10,000 lb Normal load = - 9309 lb
CRL = 10,027 lb Tangent load = - 3725 lb
Rod 3, 03 = 206º + 120º = 326º
a 3 = -5.34° XHL = -935 lb

PRL = 10,000 PSI Normal load = + 7768 lb
CRL = 10,044 lb Tangent load = -6367 lb
Note: Consider each crosshead weight if significant.
5.14 CRANKSHAFT BENDING MOMENTS-CALCULATIONS
See Figure 5.10 for bending moment diagram construction.
Example: Triplex Single-Acting Pump, Type 111 Crankshaft

Moment signs: Clockwise, plus ( + ); counterclockwise, rninus (- ).
Cakulate forces on rods 1, 2, and 3 at any degree of (} for rod 1, at (1 + 120º
for rod 2, and át O + 240º for rod 3. See Section 5.13.
To solve for RAA, take moments about RBA:
M = -(D4 X PRL3) - [(D4 + D3)x PRL2) - [(D4 + D3 + D2) X PRLl]

+ [(D 1 + D2 + D3 + D4) x RAA] = O Solve for RAA. (5.29)
To solve for RBA, take moments about RAA:
M = +(DI x PRLl) + [(Dl + D2) x PRL2] + [(Dl + D2 + D3) X PRL3]

-[(DI + D2 + D3 + D4) X RBA] = O Solve for RBA. (5.30)
(+)(=)
cw ccw
MA
(a)
(b)
(e)
Figure 5.10. Crankshaft bending mornent diagram for Example A: triplex single-acting;c:\
Type III crankshaft. See Section 5 .15 for calculation. (a) Axial plane. (b) Coaxial plane.
(e) Resultant.
5.15 CRANKSHAFT BENDING STRESS-CALCULAT!ONS 161
Calculate bending moments (axial), MA, and plot on Figure 5.10:
i>AA at rod l = (Dl X RAA) (5.31)
MA at rod 2 = ( D l + D2) x RAA - ( D2 X rod 1 force) (5.32)
MA at rod 3 = (Dl + D2 + D3) X RAA - (D2 + D3) X rod 1 force)

- (D3 x rod 2 force) (5.33)
Calculate bending moments (coaxial), MC, and plot. Repeat as for MA but
forces from coaxial plane.
Cakulate bending moments (resultant), MR, and plot:
MR = ..JMA2 + MC 2 (5.34)
CRL = connecting rod load, lb

RAA = reaction A, axial, lb
RBA = reaction B, axial, lb
RAC = reaction A, coaxial, lb
RBC = reaction B, coaxial, lb
RAR = reaction A, resultant, lb
RBR = reaction B, resultant, lb
MA = moment, axial, in-lb
MC = moment, coaxial, in-lb
MR = moment, resultant, in-lb
D = spacing between rods, in
M = moment, in-lb
5.15 CRANKSHAFT BENDING STRESS-CALCULATIONS
Example
Assume 5-in uniform diameter. Select:
4340 oil-quenched and tempered steel, 360 BHN

Ultimate strength, 170,000 PSI
Yield, 156,000 PSI
Allow combíned stress of 18 % of ultimate, or 30,600 PSI. Reduce by 25 % for
stress risers, or allowable stress, 23,000 PSL
162 PUMP DESIGN
l = 17/4 x r 4 = n/4 X 2.5 4 = 30.7 in4
O"max = Mr/l = 150,480 X 2.5/30.7 = 12,254 PSI
BHP pump = 145; RPM = 300
Torsional stress 'Y = Tr / J
J = 2 X [ = 2 X 30.7 = 71.4
T = 63,000 X BHP/N = 63,000 X 145/300 = 30,450 in-lb
'Y = 30,450 X 2.5 /71.4 = 1066 PSI
Maximum Principal Stress:
s = a/2 + Y(a/2) 2 + y2
12,254/2 + \/(12,254/2) 2 + 1/36,356 12,631
Nomenclature
l = moment of inertia, area, in4

J = moment of inertia, polar, in4
M = bending moment, in-lb
N= RPM
r = shaft radius, in
s = stress, PSI
T = torque, lb-in
a = normal stress, PSI
'Y = torsional stress, PSI
5.16 CROSSHEADS
From the mechanical and lubrication viewpoint, the crossheads

bearings of reciprocating pump power ends are the most critica!, and too, they
are a most inefficient mechanism, contributing to a significant lowering of the
pump's mechanical efficiency.
Because of the reversa! or intermittent loading and the oscillatory motion of ,
the crosshead bearing, it is impossible to depend upon much of a dynamic-oil-f~
film and an oil-pressure-film is hard to come by because of the difficulty
making a reliable oil-pressure connection to the reciprocating part.
5.17 BEARINGS 163
'ee Figure 5. 1A for one method of increasing crosshead pin bearing capacity on
·ngle acting pumps.
Figure 5.2A shows the diffic~lty of obtaining sufficient bearing area in typical
xed-diameter pin-and-bearing fits.
The author's opinion of full-compliment needle-type crosshead bearings is
ewhat negative because of the tendency for the rollers to skew or cant and
ult in difficult assembly or damage to the bearing. Caging the rollers to pre-
t this, drastically reduces the number or rollers (and bearing capacity).
Very good performance has been had with plain sleeve bronze or bi-metal
sshead pin bearings, properly grooved to conduct oil around the outside di-
eter to at least three equally spaced (120 degree) closed-end inside longitudinal
ooves, running on a hardened and ground steel pin. In horizontal pumps sat-
actory lubrication can be had by use of an oil gallery to supply a flow of oil
, a funnel-shaped oil-hole in the eye of the connecting rod or by use of a sepa-
te oil line from the oil pump. In either case, ample ftow should be provided to
ercome any possibility of capillary attraction diverting ali of the oil along the
'per shoe surface to the outer edge. A system of upper shoe oil grooves that
not extend beyond the edge& or ends of the shoe, thereby preventing the
ersion of oil from the connecting rod through open-ended grooves, is also
ommended.
In vertical pumps, the only possible means, other than "splash", of lubricat-
g the crosshead bearing is by the use of a rifte-drilled connecting rod, sharing
il pressure delivered through a drilled crankshaft.
In horizontal pumps the bottom crosshead shoe should also incorporate closed-
nd oil grooves and it is imperative that a pressured oil supply be furnished
brough the bottom guide on heavily loaded pumps. The closed-end grooves will
,ben provide a desirable hydraulic lifting action on the crosshead near the center
;f- the stroke where the load is maximum. The closed-end grooves at each end
''( the longitudinal oil groove should be allowed to slightly over-stroke beyond
e guide at both ends of the stroke so as to provide a momentary but repeated
'flushing" effect for rejection of any unwanted trapped foreign matter at each of
e two zero-load positions.
With vertical pumps, both crosshead shoes should be supplied with pressured
)1 in heavily loaded pumps.
' Because of the vulnerability of crosshead problems, it is wise to use compat-
ible materials for the shoes and guides that tend to minimize the generation of
destructive debris from wear, particularly in an oil bath lubricated power end.
Bronze shoes running on cast iron guides is most satisfactory.
Figure 5.3A illustrates a method of obtaining full bearing area for crossheads
y using shim-adjusted shoes turned to diameter of ways.
BEARINGS
In the matter of the seiection of bearing types, roller or joumal, for the power end
of a pump, both types seem to be widely used with complete satisfaction.
164 PUMP DESIGN
It cannot go without notice that pump manufacturers in the Unit'ed States have
completely converted to roller bearings for drilling mud pumps, mostly in the 500-
1700 BHP size. Of course, sorne plain bearings are still used at such locations as
crosshead pins, eccentrics, and sorne crankpins.
Roller bearings are thought to be rather immune to occasional oil contamina-
tion, particularly to contamination by solids, which can be flushed out. All journal
beaiings are intolerant of any contamination.
The question of whether a roller bearing can be readjusted seems moot, since
there is only one recommended clearance for the application at hand, and if a
bearing wears to the point of requiring readjustment it is probably worn beyond
its useful life.
A practice that should be avoided in the installation of bearings is the provision
for a dam or reservoir at the bottom of that bearing to retain a small quantity of
oil in place for start-up. Because of their tendency to accum.ulate water and debris,
such pockets do more harm than good.
At certain minimum speeds, plain or journal bearings depend upon the dynamic
generation of a high-pressure load-carrying and separating oil film, and the me-
chanical efficiency is equal to that of roller bearings.
For assurance against accidental damage from a start-up and the lack of lubri·
cation on crosshead guides, sacrificial bronze crosshead shoes running on cast-iron
guides are usually used. Should any scuffing of the shoe take place, the generated
debris is not damaging to the other bearings in the pump if it is mixed in the main
oil bath.
5.17.1 Roller Bearing B-10 Life

The life of a bearing is expressed as the number of revolutions or the number of
hours at a given speed for which the bearing can be used before any evidence of
fatigue develops on the rolling elements or the raceways. Bearing life may va11
from one bearing to another, but it stabilizes into a predictable pattern when .~
large group of bearings of the same size and type are considered. The rating lije
of a group of similar bearings is defined as the number of hours or revolutions (at
a given constant speed and load) that 90 % of the tested bearings will exceed before
the first evidence of fatigue develops. This is called "B-10 life" or "minimuíll
life."
5.17.2 Roller Bearing Average Life

The results of testing a large group of ball or roller bearings may be graphicall
illustrated. The distribution curve is obtained by plo~ing relative life versus per
cent of bearings tested. From this curve it is determined that the average life i
approximately five times the minimum life. About 50% ofthe bearings will exc
the average life. Since it is not possible to predict the exact life of a single bearing
5.17 BEARINGS 165
a safety factor must be allowed to minimize the risk of early failure. The cost of
replacing a bearing plus the expense of machine downtime may greatly exceed the
relatively low cost of the bearing. Therefore, most designers prefer to use mini-
mum life as a design basis. In sorne applications where safety or maintenance
economy is not critica! and low initial bearing cost is desirable, the average life
value may be used.
5.17.3 Journal Bearings

Sorne of the problems to be avoided with split joumal bearings are:
l. Distortion or pinching-in of split bearings by poor mating or distortion of

the bearing housing.
2. Improper selection or installation of shims, where used, that allows the shims
to rub against the joumal with resultant oil scraping.
3. Lack of proper backing for the bearing shells due to improper machining of
the housing or effects listed above. Anything that interferes with good metal-to-
metal contact of the shell to the housing interferes with heat transfer and allows
, local destructive "hot spots" to develop on the bearing.
4. Ductile iron joumals rotating in bearings in the opposite direction to that at
which the joumals were originally tumed and polished may result in rapid wear of
the bearings.
The ability of joumal bearings to generate a dynamic oil film that reduces friction
is a very powerful argument in their favor. The following formulas show the fac-
tors that affect such friction:
T= FLr (5.42)
F = 27rµ,pr / C (5.43)
U= 7rDN/60 (5.44)
T = [27rµ,e( 7rDN /60) r/C] Lr (5.45)
Example:
D = 8 in, L = 8 in, C = 0.006 in, RPM = 85, ~RL = 30,000 lb

SAE 70 oil at 150ºF = 100 cP, /1-e = 0.00209 lb-s/ft2
T = [6.28 X 0.00209(3.1416 X 8 X 85/60) X 4/.006]

X 8 X 4 ·= 9974 lb-in
166 PUMP DESIGN
The relationship ZN / P is used to classify joumal bearing applications, since it

contains the three most important limiting factors. Any value below about 15 is
considered marginal.
85
ZN /P = lOO X 30000/(8 X 8) 18
where
e = radial bearing clearance, in

D = bearing diameter, in
F = tangential friction force per unit axial length, lb
L = bearing length, in
N= RPM
P = force per square inch of projected bearing area, PSI
PRL = piston rod load, lb
r = bearing radius, in
T = torque, friction, lb-in
Z = absolute viscosity, cP
µ, = absolute viscosity, cP
µ,e = absolute viscosity, lb-s / ft2
U = tangential velocity, in/ s
See Tables 13.17 and 13.18 for viscosity equivalents.
5.18 LUBRICATION
5.18.1 Pressure Lubrication

With pressure lubrication applied to the bearings, the actual pressure is not a~:
critica! as the amount of oil or rate to and through the system. Most lubricatin ·
systems use positive-displacement oil pumps (gear, vane, piston, etc.), and unles
excessive bearing clearances exist somewhere in the system the required amoun('
of oil should be circulating to the proper places. In other words, as the oil increases'
in temperature, the pressure drops radically but the volume or rate probably in-
creases due to self-improved volumetric efficiency of the oil pump.
In a power pump, the lube oil pressure drop dueto temperature increase presents
a problem in monitoring because the great change in viscosity prevents the pre-•
diction of an exact operating pressure. Pressure ranges of 20-40 PSI at operatin
temperature are usually acceptable and give good reference for monitoring.
In many applications, the lube oil pump is driven by the main pump and at;·
reduced speeds below the recommended, a separately driven oil pump and perhaps
cooling equipment are recommended.
5.19 VOLUMETRIC EFFICIENCY 167
Oíl Cleaning
The natural accumulation of airbome moisture, dirt, sand, dust, and even metal
""''w,~~ carried into the crankcase through the "breathers" or the diaphragm
packing can be injurious to bearings, patticularly joumal bearings, so continuous
removal by proper filtering is rnost desirable. (In some humid, corrosive, and dirty
~~~,n'"'"'Q it is the practice to bleed an inert dry gas such as nitrogen or clean
drv air into the crankcase of any pump so as to maintain a slight positive pressure
0 { a few o unces to minimize entrance of such materials.)
Oil Temperature
The viscosity of oil is affected by ambient temperature, and in certain geographical

!ocations this can be a problem. In extreme cold, the oil has a tendency to congeal
and the pump may run far several minutes before oil circulation is established. In
such cases it is usually the practice to install electric crankcase heaters so the oil
can be heated before start-up. An annoying problem with cold operation is the
excessive momentary pressure buildup that may damage the pressure-sensing or
pressure-controlling devices.
In cold climates, sorne operators follow a dangerous practice of diluting the oil
with diesel fuel. Heaters or special low-temperature oils should be specified.
VOLUMETRIC EFFICIENCY (Eq. 2.26)
Volumetric efficiency is defined as the ratio of the actual displacement of Hquid to

the calculated theoretical (100% VE) displacement based on pump piston diame-
ter, stroke, and speed, on a percentage basis.
Accurate volumetric efficiency is difficult to calculate on account of the unpre-
dictable value of valve slip, a term used to describe a combination of conditions,
namely, dynamic delay in val ve closing and opening (a function of val ve spring
load, valve weight, dimensional proportions of the valve, life limitations, liquid
viscosity, and specific gravity-none of these being the result of pump speed or
pressure). Another important factor is the physical condition of the val ve and seat.
A worn valve or valve insert will allow a certain amount of leakage, and of course
the amount of leakage is related to the discharge pressure and the time of subjec-
tion.
Subjection time leads to a phenomenon that generates a curve for volumetric
efficiency versus RPM that shows a tendency to peak early on the RPM base,
showing that ata low speed the val ve will leak for a longer period of time at closure
due to "wear-clearance" leakage. The effect becomes less apparent at slightly
· higher speeds, but then the dynamic effects of NPSH take over, causing a drastic
rate of VE reduction as shown in Figure 5 .11.
168 PUMP DESIGN
95
l>~
;,:;
u
e 90
"'
·¡:¡
i:
<l.>
u
·;::
'$
E 85
:::¡
o
>
80
o 100 200
Figure 5.11. High-pressure valve performance ·plex single-acting pump at

3000 PSI, with 2, 4 and 8 POSIVA valve sprin of a 3 by 6 in tri
1':\S.
This chart shows the desirability of proper val ve . ¡0 w speed, the 4 POSIV A
spring load is superior, but at high speed the ;Pn.ng load. fo~ns to show its advantage.
High-speed requires high POSIV A, obviously r .Pos1v A be~¡.¡R for that pump. It also
ª
shows the "peaking" of volumetric effi.ciency . 1sing the NP pM, where the time period
for valve slip is relatively shortened. From tha~t about 200 ~ ¡he effects of insufficient
cylinder filling (cavitation) due to increased Sf>e 5Peed upwar ~tribute to a rapid drop in
volumetric efficiency. ect starts to ca
5.20 MECHANICAl EFFICIENCy
Introduction
Mechanical Efficiency (ME) of a pump is ge of input horsepower
(BHP) or energy imparted to the liquid be¡ the percenta
ng pumped;
ME = (HHP/BHP) (5.46)
X: 100
HHP = (gpm X PSi)/(17¡4 X ¡v1J3) (5.47)
With modem pump design and construct¡0 ·cal efficiency of 85% in
double-acting pumps and 90% in single-act~' mechalll 0 be expected.
ª
The horsepower required for a double-a. l~g pumps cahould be calculated from
formula (5.39), Section 5.22. Cting pump 5
For the same reason that a single-actin~ . bigh suction pressure has
higher stresses in the power end, all the be P~mp with ore heavily loaded in re-
lation to the work being done and therefore ªn.ngs are '1'.cal efficiency is reduced.
the mechalll
5.21 INTERMITTENT SERVICE 169
Q.,"
;;:~ 0.6 , _ _ - - - - + - - - - + - - - - + -........- - - - - - - -
º
~
~
:::J
"' 0.4 t-----+-----+----+-----i\-------1
~
Q_
o 0.2 0.4 0.6 0.8 0.9 1.0

Mech¡mical efficiency ME
Figure 5.12. Single-acting pump mechanical efficiency vs. suction pressure / discharge
pressure. To determine BHP: Calculate pressure ratio; determine mechanical efficiency from
curve; substitute in:
BHP = (Pd - P,) x GPM/1714 X ME (5.48)
where Pd = discharge pressure, PSI; and P, = suction pressure, PSI.
lt has been determined by test that the relation shown in Figure 5.12 exists. There-
fore, for a single-acting pump the required horsepower is derived from formula
(5.40) Section 5.22.
Since the manifestation of low mechanical efficiency is sensible heat generation,
higher power-end temperatures will be expected with high suction pressure.
5.21 INTERMITTENT SERVICE
Any pump can be overloaded by a reasonable amount for a few minutes of oper-
ation at rare time intervals by increasing either the pressure or speed or both. And
any pump can probably be operated at extremely slow speed for a short period
without harm.
Spurred by the need for the greatest amount of power packed in a small space
for portable use in the oil fields for sand fracturing, acidizing, and cementing,
most manufacturers offer such pumps for this and other services. By increasing the
170 PUMP DESIGN
GPM
o 30 60 90 120 150 180 210
0 01--~2~5~-..Jso'--~7~5~-1~00~~12~5~-1~5-0~1~7-5~--' 0
RPM
Curve A- lntermittent service, constant horsepower
Curve B- Continuous service, proportional horsepower
Curve C- lntermittent service, proportional pressure
Curve O- Continuous service, constant pressure
Figure 5.13. Continuous vs. intennittent duty.
design piston rod load (PRL) three- or fourfold and increasing the maximum speed
about twofold, pumps with a threefold reduction in weight per horsepower have
been produced. These so-called constant-horsepower pumps are rated for inter-
mittent service and typically Iimit the duty to l 1/2 hours per day. Figure 5.13
shows the performance curve of two pumps of this type with equal design horse" ·
power. lt is obvious that such pumps will require more frequent replacement of ·
bearings and parts.
5.22 CONTINUOUS SERVICE
Continuous service denotes operation at rated conditions of speed and pressure for
24 hours a day. Basically, this is a constant-torque operation with a 'speed turn-
down of about 50% allowed, such turndown usually being limited by the types of
bearings and the lubrication system provided. Antifriction (roller) bearings
throughout the power end will allow a greater turndown than journal or plain bear-
ings.
The design criteria incorporated into this text apply to continuous service ºP"
eration.
5.23 RECIPROCATING PUMPS WITH HIGH SUCTION PRESSURE 171
5.23 RECIPROCATING PUMPS WITH HIGH SUCTION

PRESSURE
For Double-Acting Pumps
BHP = (Pd - P,) X GPM/1714 X 0.85 (5.49)
For Single-Acting Pumps
BHP = (Pd - P,) X GPM/1714 X ME* (5.50)
Triplex pumps p = pd + !P, (5.51)

Quintuplex pumps P = Pd + ~P, (5.52)
Septuplex pumps P = Pd + iP, (5.53)
Nonuplex pumps P = Pd + ~P, (5.54)
P = artificial discharge pressure upon which plunger/piston is selected

when suction pressure is in excess of 5 % of discharge pressure
Pd = actual discharge pressure
P, = actual suction pressure
BHP = input brake horsepower
GPM = actual U.S. gal/min
*ME = mechanical efficiency from Figure 5.12 ·
Example
A certain triplex pump rated at 1000 PSI with 3-in plungers ( PRL 7200 lb) would
have to be equipped with 2i-in plungers ( 1250 PSI at PRL 7200 lb) to pump at
actual 1000 PSI with 500 PSI suction pressure.
P = 1000 + 500 /2 = 1250 PSI

6
LIQUID ENDS
6.1 TYPICAL LIQUID-END MANIFOLD CONFIGURATIONS
The valve-over-valve configuration for multicylinder pumps is probably the m

efficient from the standpoint of suction problems and efficiency and ease of mai
tenance (Figure 6. la).
Pumps with long suction and discharge passages, called surge legs (b) and is
lation leg (e), where the valves are placed a great distance from the cylinder ·
the dubious reason of isolating the piston from abrasive materials in the liqti
often exhibit poor hydraulic performance. These long passages tend to mag
the acceleration and ftow problems because of the greater mass of liquid that m
be moved from a "dead-stopped" condition. The large volumetric clearance
the isolation-leg design results in low volumetric efficiency and contributes to ni?
and rough running. This is particularly true if air, gas, or vapor is present in
liquid.
The inverted configuration (d) has been considered for slurry services on
contention that the solids would "naturally" fall through the liquid end with
danger of plugging and stacking. But its shortcomings include the inability to p
itself of air or gas, and valve replacement would be ext~emely awkward.
6.2 S-N CURVE
As shown in the S-N curve (stress vs. number of cycles) of Figure 6.2, if
maximum stresses are kept lower than 35 ,000 PSI there should never be failu
air. In water, a stress of about 15,000 PSI would be the limit.
172
Minimal
length
Sign ificant
(a) NORMAL - VALVE-OVER-VALVE length
(b) SURGE-LEG
(d) INVERTED
(e) ISOLATION LEG
Figure 6.1. Liquid-end configurations. (a) Normal, valve-over-valve. (b) Surge leg. (e)
Isolation leg. (d) Inverted.
50
40
.te
§...
-= 30
rr.ie
en
en
...
Q)
ti 20
E
:::1
E
·;;e
...
:E
B
10
105 106 107 loB

Fatigue life N, cycles
Figure 6.2. S-N curve. Exposure of steel to water greatly reduces fatigue life. Salt water
and hydrogen sulfide still further reduce life .. Curve A, dry fatigue life; curve B, wet fatigue
life.
173
174 LIQUID ENDS
MOD IFI ED GOODMi\N DIAGRAM
11l0 •....•... ........ . .:......... .. ;..... .. .. .. .: .. ........ .: ...........:.... ...... .: ..

. ... . .. ..
... . .. .. .
180 . ...•••..
140 ...••••. ··~ . ... •....
~
•••••
..
~
. .
•••.•••••• j •••••• . . •• • ..•.••. •• • ;. ••.•..• . • ·:. . . . . . . . . •
... ...
• ••.. • • •••
.. ... ..
. .
~ 100 .. . ....... ; . ..... ·r·· ······ ··········!··········:··········:···········¡···········: ................ ..
80 8á~¿··[ ..........:······ ·· ·· ·········· >·······[··········:··········[···········¡·········· ··········
80 . . .•..... · [ .. ..... .. ·\·. .. . . . . .. . .. ~w.1.m.<tM ..-!r.1r~s .... .. .¡ ..........¡.......... j· ...................
. : . : : :
40 ·· ·· · ·····:········· i ········· .......... j··········:······· . ··:··········:-·········¡· .................. .
20 · · · ···· ·· >·······-r········ ........ ·1 · · ·····r·······-:--······· :·······. r........··········
80 100 120 140 160 180 200
JNJTIAL STRESS, 1000 PSI
Figure 6.IA
The results to be expected with the material being stress cyded in even a slightly
corrosive liquid (pure water with dissolved air, for example) differ greatly; the
stresses that would allow extended life are far below those in air. In figure 6-2, a
stress level of about 17 ,000 PSI would offer a fair amount of additional life, but
it is now believed that the corrosion-fatigue curve never flattens out and that such
stressed parts will have a finite life (hopefully beyond the life of the unit of which
they are a part).
Liquid ends and expendable parts of all pumps are subject to corrosion-fatigue
failure, and therefore it is important to consider this in their design.
Where corrosion fatigue is expected, the logical approach would be to design
for the lowest possible stress. In parts with associated stress-risers (intersecting
bores in liquid ends, notches, pitting, machining marks, sharp shoulders, and dras-
tic change in cross section being the most important), steps can be taken to reduce
stresses or protect the point of stress concentration.
6.2.1 The Goodman Diagram

In the design of repeatedly-stressed parts or devices, such as valve springs, with
a material in question one can take advantage of the fact that most materials
exhibit a trait that if the stress range is limited by a certain minimum stress of
higher value than zero, fatigue strength can be substantially increased as shown
in Figure 6.2A.
6.2 S-N CURVE 175
YIELD
115.000 psi
'
-
--210.000 psi
CD1nens1on X" .. ) RADIUS 1\LL BDRE INTERSECTlDHS t 1/4' l.

ADD THAf RADIUS/4 1 .061 TD BDRE
RADIUS "ID OBTAIN RI.
DIA AD
LDCATIDN INCH R2/Rl LAME R
DIA AB '{. 77
DIA BC '"/.88
DIA CD '/107
DIA DA
"·º J,,,
AlRl
A1R2
2.0,,.
7.5' '·''
D2R2 A1R2 A2Rl
AlRI A A2R2
2.fl,,f
"'·'º },O'f
cu
...
o::
CI
D
o::
CI
-
o::
cu
<
cu
o::
cu
<
111
<
BIRI
BIR2
¡,/5
B2Rl .J,or,, /. 2 'I

cu
o::
cu
u
Q:
cu
u
-
o::
111
cu
o::
111
--i:
ot ....
wi::i B2Ri~
C!Rl J. l'f
e CIRI
B2Rl CIR2
3. '(t.
CIR2
B2R2 C2Rl '(, l J. I/
1 w
aj
a: C2R2
lLI DIRI e .ocs S.'lt. /.o
e
"'<"!.J DIR2 f2.S/
IJIA 1 BC D2Rl 2. "(1 9.'I!. J. /'f
D2R2 8.'f
EIRI 2.53 .S. 4 I
El '2 fj.13
'·' r
Figure 6.2A. Liquid End Data for Autofrettage.
176 LIQUID ENDS
For example, a material in question with zero mm1mum stress may show
fatigue limit a of 80,000 PSI but that could be raised to about 115,000 PSI ·
the minimum stress never was allowed to extend lower than about 60,000 PSI.
6.3 STRESS
Maximum stresses in typical pump liquid ends occur at the intersections of bores
since stresses there are magnified by a stress concentration factor. Experience h
shown that this factor has a value of 2. 75 for all configurations. For instance,
the simple T-block illustration of Figure 6.3, the maximum stresses occur at A
Strain-bolt holes
Figure 6.3. Stress concentration in liquid ends. Strain gauges were installed in a
block and then pressured. Two l~-in diameter strain bolts, 400 ft-lb torque,
in-lb 4800
Bolt tension = - - = = 19,200 lb
0.2D 0.2 X 1.25
Without strain bolts, using stress concentration factor of2.75, theoretical hoop stress
by Lamé formula, is
R1
2
+ R2
1 X
S = P X 2.75
R~ - Rt
3.5 2 + l. l2
X 4000 X 2.75 13.411 PSI
3.5 2 - l. 12
(Compare this to actual stress of 14 ,400 PSI.) Note: W ith bores of different diameters, :u
largest diameter for source of R 1 .
6.4 STRESS REDUCTION METHODS 171
BORE 11111ERSECTION STRESS

RADIUS RA·. 10 VS TOTAL MUlTiPLIEA
2.oo+--'""-1-t------i----;l--+---1---+----l----+----l
z.oo 4.oo e.oo e.oo 10.00 12.00 14.00 16.00 18.00 20.00
3,7 + li.=3 + k~:/R\. k"5
Figure 6.3A. Bore Intersection Stress.
stress resulting from hydraulic pressure applied to the interior. Using the
wall thickness at B to calculate R 2 , the calculated stress, by Lamé's
Stress = ( R;
2
+
R 2~ ) X P X 2.75 PSI (6.1)
R2 - RI
different size bores, use mínimum R2 and maximum R 1 •

·'-'"''"''~""'
of these stresses, compounded by the cyclic nature of the liquid pres-
liquid ends tend to have a finite life. It is at the point denoted A in Figure
.that practically all liquid-end failures occur, beginning with an infinitesimal
ck that propagates slowly but defiantly to the outer surface. (An interesting
··light of the above calculation is the revelation that as the wall thickness ap-
hes infinity, the inner stress can never be reduced below the actual hydraulic
STRESS REDUCTION METHODS
Strain Bolts
·imple but effective method of reducing the stresses in liquid ends of adaptable
pe (Fig. 6.3) is to use strain bolts in the locations shown. Tests have shown
178 LIQUID ENDS
With strain bolts

400 in/lb torque,
Stress = 260 x 30 = 7800 PSI
4000
Compare this to calculated

data of 13,389 PSI
using 2.75 stress
concentration factor
-300 -200 -100 +100 +200 +300 +400

Compressive strain Tension strain
µin/in
Figure 6.4. Effect of strain bolts on liquid end. Plotted strain gauge test data showing
critica! point A is forced into compressive stress by strain bolts. Maximum stress du
pressure of 4000 PSI is reduced from 14,400 PSI to 7800 PSI.
that the critica! stresses can be reduced by as much as 60 %. This method c

used to decrease the weight of Iiquid ends or to improve the integrity of exi.
parts if the configuration of the part allows its application.
Figure 6.4 shows the details of a test on a liquid end with strain bolts and
graphic results.
6.4.2 Autofrettage Procedure for Liquid Ends*

6.4.2.1 INTRODUCTION
The rather simple process of Autofrettage ("self-strengthening") is fairly
known and it has been used extensively with simple thick-wall cylinders s~
gun barreis. ··
But high-pressure reciprocating pump Iiquid ends by design must have, .;
secting cylindrical bores, usually in solid metal blocks, Figure 6.3. The st
generated by interna! hydraulic pressure during regular use of the pump ar~
that they are multiplied in the order of two to three times the stress calcÜ
by the Lame formula for straight cylinders, caused by an intensification of s
at the intersection of cylinder bores. This intensification, (factor k) along
*Note; It is most desirable to employ the help of a person well experienced in the appl
instrumentation and interpretation of data in the strain-gage procedure described herein for me
strain. Otherwise, one may find this an expensive and drastic learning enterprise.
the pulsating nature of a reciprocating pump discharge combine to produce early

fstress-corrosion-fatigue failure of the liquid end, always initiating at one of the
fintersections.
·· There are several methods of "pre compression-stressing" these intersections
y placing the metal around the "hot spot" into permanent residual compressive
tress that will tend to prolong the life of the liquid end, but autofrettage is
onsidered superior to any of the others because of the ability to make a greater
nd deeper "compression" area that is more damage proof, permanent and ef-
ective. Figure 6. lA. Coatings to protect the metal from chemical (corrosion)
· ffects lack permanency and are easily damaged.
Because of the described Stress Concentration Factor, a blessing in disguise,
utofrettaging a pump liquid end becomes a very practical and enduring method
f prestressing. By reason of the natural effect of the stress concentration at the
ntersections, a predetermined internal high pressure hydraulic application will
Jectively overstress (beyond the yield) those "hot spots" without affecting the
raight bores or other passages in the liquid end. But for reasons of this same
Jessing, great care must be exercised in pre-analysis and execution of the pro-
ess.
When this temporary "high" autofrettage pressure is released, the yielded hot
ot will acquire a compressive stress from the contraction of the non-yielded
etal around the spot. Thus the stress at the hot spot generated by a much Jower
orking pressure will never force the spot into tension, the secret of overcoming
ress-corrosion fatigue.
Because metals differ in character and the encountered stresses are a function
f both the metal and the configuration of the liquid end and the bores, it is
ather difficult to precisely predict the autofrettage pressure and the following
rocedure is recommended. ·
4.2.2 PRELIMINARIES
the actual demonstration example included, ali values referred to the given
xample are placed in parenthesis.
l. Determine the R2/R1 ratio for each intersecting bore. See Figure 6.2A and
ssociated Table 6.1. Note that where there are radiused edges the R 1 dimension
ould include the edge radius and represents the actual radius at the point of
eatest strain. To the bore radius add edge radius divided by 4.
These Ratios (R2/ R 1) range from about 1 to 6 in typical Iiquid ends and natu-
lly one is encouraged to design around the greatest value practical, generally in
e order of 3, the greatest advantage coming from the higher number. Values
ater than 5 or 6 tend to offer Jess advantage and conversely, values below
the stresses tend to skyrocket. See Figure 6.3A.
2. Start with the preparation of Table 6.1, Calculated Data, including the R2!
1 followed by a determination of the Lame' equation for thick-wall cylinders.
uation 6.2.
3. Calculate and add to Table 6.1, a MULTIPLIER "M" which, when mul-
Iied by the applied pressure results is the stress at that point. Equation 6.4.
180 LIQUID ENDS
TABLE 6.1. Celculatec:I (Theoretlcal) Straln and Stress at Fivepoints
A B e D E F
Rl Rl R2 R2 R2 R Lame M
AlRl 2.065 AlR2 7.56 3.66 1.16
A2Rl 2.565 A2R2 12.31 4.80 1.09
BlRl 2.438 BlR2 9.13 3.74 1.15
B2Rl 2.47 B2R2 7.56 3.06 1.24
ClRl 2.47 ClR2 8.44 3.42 1.19
C2Rl 2.065 C2R2 9.13 4.42 1.11
DlRl 2.065 DlR2 12.31 5.96 1.06
D2Rl 2.47 D2R2 8.44 3.42 1.19
ElRl 2.53 ElR2 9.13 3.61 1.17 1.17
E =
LAME RATIO,R2/Rl, EQ 6.1 H =STRESS AT 10000 Pa x G
F =
LAME MULTIPLIER, EQ 6.2 I = STRESS at 15000 Pa X G
G =
INTERSECTION MULTIPLIER, EQ 6.4 L = STRAIN at 50000 STRESS/
H =
STRESS AT 10000, E 6.3
H I J
POINT STRS lOK STRS.15K STRS.50K STRS. 65 STRN.50K
A 31934 47902 159672 207574 6919
B
e
D
E
RULE OUT MINIMUM FOR EACH POINT

Because of the design features of this Ji.quid end, the most
hi.ghly stressed point is 'B' as indicated.
Stress at Stress at Strain at Hoop at

15,000 65,000 65,000 10,000
47902 207574 6919 12774
44994 194976 6499 11999
47584 206199 6873 12689
51109 221472 7382 13629
48978 212237 7075 13061
45698 198024 6601 12186
43639 189101 6303 11637
48978 212237 7075 13061
Because of the design features of this liquid end, the most highly stressed in
section is R1B2 as indicated here.
SECTION THROl.Di
INTERSECTION. GAGE
RI / LOCATICPN ON CENTER OF
/
/ PERPENOICLLAR 2
RADIUSEO EOGE WITH AXIS
PAGE.
··································51···
1NSET SHOWS REo..NJANT
GAGES AS ALTERNATE.
SINGLE ELEMENT STRAIN GAGE MOUNTEO WJTH AXIS

R - BCIRE RAOIUS. IN
SHOWN BY ARROW BELDW
RI • 80RE R + d~ IN
R2 - OUTER WALL AT LEAST SINGLE ELEMENT [XJMMY GAGE M!JU'ITEO ON STEEL
OISTANCE FRDM BDRE CENTER. IN BLOCK ANO LA ID 1N SOME CONVEN 1ENT PLACE 1N
r = IBTERSECTION RAOIUS. IN CYLJNOER.
ALL GAGES. INCLUOING DUMMY. SHDLLO BE PREPAREO
WITH 3 FOOT BEELOEN PAAIREO WIRE. TAG ALL
LEAOS WTH LOCAT 1ON MARK.
Figure 6.4A. Placement of Strain Gages.
4. Complete Table 6.1 by calculating all of the theoretical stress and strains
ed for the points in question.
5. Short single-element strain gages (l/16"), while difficult to install, are rec-
.mended and it is worthwhile to obtain gages with short factory-soldered leads
ached. On the workbench, build up gage units by soldering the gage leads to
~rminal strip. Then to the terminal strip solder paired lead (Belden 8430) about
feet long which will be finally soldered to the feed-through fitting. Be sure
tag all leads for identification. lt goes without saying that all gage assemblies
tested with an ohm-meter for continuity and proper resistance. Just before
allation dip each gage assembly in a cup of acetone for a few seconds and Jet
in a clean environment. For maximum accuracy make all leads from each
e to the instrument of equal length.
6. Follow exactly the suppliers procedure for cementing strain gages and apply
(preferably two, for redundancy, end-to-end with leads facing away so that
sing elements _will be as close together as possible) gages on the center of the
iused intersections. Figure 6.4A The time and expense involved for redun-
cy will more than pay for the time involved due to "loss" of a gage during
high pressure application. The axis of the gage(s) should be parallel to the
p direction, which would be represented on the drawing as vertical to the
face, on Figure 6.2A'.
182 LIQUID ENDS
7. Mount a 'dummy' gage on a small block of steel and lay it in a convenient

place in the cylinder. One 'dummy' can provide the second arm of the four-arm
Wheatstone Bridge Circuit for ali test locations by switching it in to the circuit.
This dummy will serve to provide 'pressure' and 'temperature' compensation on
the active gages.
8. As stated, one or two gages should be cemented to a section of the largest
diameter straight bore as shown in position E, Figure 6.2A.
9. Prepare the high pressure pump to handle water-free transformer oil instead
of water for the high pressure liquid medium. This will preclude the need for
any difficult water-proofing of the gages.
6.4.2.3 HIGH PRESSURE APPLICATION PROCEDURE
One must follow this procedure on a step-by-step basis because of the unknown
reaction of a newly designed liquid end to exceptionally high pressure. Once the
pressures for one design of liquid end are determined, and if there are no appar-
ent anomalies, subsequent liquid ends could be autofrettaged with just one or two
strain gages at the most highly stressed point for monitoring.
In order to maximize the autofrettage effect a certain area around the critica(
point must be kept in a non-yielding condition so as to provide the re-compres-
sion forces expected at the critica! point. Experience has shown that the maxi-
mum autofrettage effect will be had where the critica! point is placed in its
desired state with a strain of no more than 10,000 microstrain.
To avoid the possibility of bursting or damaging the liquid end by excessive
pressure during the autofrettage cycle, a safety rule should be observed; "Never
expose the liquid end to pressures that would generate yielding in the straight
bore sections of the largest straight bores, as at (E= 111,"810 PSI)". One is
concerned with over stressing those areas at the points of high stress only at the
bore intersections. In fact, as stated above, the under strained area beyond the
point of protection is required to generate the re-compression stresses required
that makes the theory of autofrettage acceptable as an "improver" and points
like "E" may limit the maximum pressure to be applied.
There is little danger from over-stressing typical liquid ends but in those
shapes where the straight bores are of considerably greater diameter than the bore
at the intersection or with materials of a low modulus, care should be taken.
6.4.2.4 PRESSURE APPLICATION
1. Provide a safe location for the liquid end involved, · preferably an under-
ground pit. Closed circuit television can be used to observe the pressure gauge
readings, etc. Prepare a Test Data Report Form similar to Table 6.2.
2. It is good practice to first zero ali gages and then apply a pressure of
approximately the design or working pressure of the liquid end, (10,000 PSI).
This will provide a check on the reliability of the gages and the use of the
recording instruments. Incidentally, if re-zeroing is necessary on any gage, those
gages should be so noted and viewed with suspicion. Any change in zero of the
gage installation is usually caused by improper cementing technique where air is
trapped under the gage with thick layers of cement, unclean surfaces or lack of
BLE 6.2. Test Data and Calculated Stress

8 e o G H
TARGET PRESSURE STRAIN 8 STRESS 8 LAME' B STRAIN A STRESS A STRAIN E STRESS E
o o o o o o o o
2 5000 5000 931 27930 5964 701 21030 611 18330
3 10000 10000 1590 47700 11927 1316 39480 1136 34080
15000 15000 2260, 67950 17691 1885 56550 1626 48780
20000 20000 2911 87330 23854 2406 72180 2063 61890
25000 25000 3608 lüi 240 29818 2964 88920 2536 76080
30000 30000 4431 13; 930 35782 3566 106980 3215 96450
32000 32000 4809 144270 38167 3847 115410 3021 90630
9 34500 34500 5269 15i070 41149 4295 128850 2510 75300
36000 36000 5593 16. 790 42938 4464 133920 2418 72540
38000 38000 598:' 17' .\60 45323 5046 151380 2746 82340
40000 40000 645(. 19 ~90 47709 5415 162450 2845 65350
42000 42000 6932 20. 160 50094 5910 177300 3082 92460
44000 44000 7461 22. J30 52480 6333 189990 3295 98850
46000 46000 797S 23' 250 54865 6688 200640 3536 106680
48000 48000 8531 25'. ?30 57251 7179 215370 3629 108870
50000 50000 8988 269640 59636 7587 227610 3727 111810
55000 10000 318~ 9$670 11927 2789 83670 1270 38100
60000 1318 41.\98 o 1273 48390 o o
- Run Number
- Pro>osed Pressure Range
- Act• al Pressure Applied
- Instrument Styrain Reading
- Calculated Stress (O x 30)
- Lam<.' Stress - ( R2'2+Rl '2 )/( R2'2-Rl '2 j x Pressure
- Inslrument Strain Reading
·~ Calculated Stress ( G x 30)
- Strain at Plain Cylinder E
- Stri·ss at E
ficient application pressure. Sometimes a slight shift of zero on a dubious gage

1 "heal" with the first tria! application of pressure. .
.3. Re-zero ali gauges and apply increments of about 5000 PSI and record
in. When ~he plotted data shows a distinct departure from a straight line
oportional limit),. that point represents the instant of yield and from that point
ard, the relation of strain to applied pressure or stress has no apparent mean-
so the upper limit of pressure must be decided. (See above text).
4. The STRAIN reading at a point can be used to determine the actual Stress
ncentration Factor. k, (for what it is worth) at, for example Point B, apparently
highest stressed point, by comparing it to the strain at [E],. (8988/
7 = 2.41) .
•2.5 INTERPRETATION OF PRESSURE-STRAIN CURVE
le 6.2 shows the total autofrettage pressure excursions for a typical liquid end.
te that the applied pressure of (50,000 PSI) generated a residual strain of about
83) microstrain of residual compression. It is felt that reading the residual
in after release of pressure is not necessarily the most accurate method be-
se of danger of a damaged or altered gage after being subjected to the high
184 LIQUID ENDS
pressure. The following interpretation also precludes the accuracy of strain mea
surement-regardless of the value of the readings, they are relative to each othe .
and if they ali differ, the results will always be the same. The use of rosett~
strain gages is considered superftuous because of the complicated stresses exist-
ing at the point of two radii when only the hoop stress is significant. See Figure
6.5A and 6.6A. Accordingly, single-element gages can be used.
As shown in Figure 6.5A, plot the test data of the highest strained intersectio
stress (usually at the intersection with the least metal section as noted in Figur
6.2A) on the X axis and applied pressure on the Y axis. Construct a curv
through the average of the data points from zero ending at maximum applie
pressure.
6.4.2.6 AUTOFRETTAGE-INTERSECTING BORE EQUATIONS
E = Young's Modulus of Elasticity (steel = 30) actually 30 X 106 • (Multipli
106 can be ignored since strain is given in microinches.)
Stress (lbf/in 22 or PSI) = 30 X Strain
Strain (Microstrain) = Stress/30
Pressure, PSI
Rl = lnside cylinder or intersection radius, in.
R2 = Outside cylinder, in. or shortest distance to outside wall.
Lame' Ratio of Wall Thickness = R2/Rl
Lame' Multiplier for Thick Wall Cylinders
ML = (R2 2 + Rl 2)/(R2 2 - RI2)
Stress in ID of Thick Wall Cylinder;
Stress (PSI) = ML X Pª
Modified Lame' Equation for Intersecting Bore Stress;
Where MT = ML + Me
SYMBOLS
ML = Lame' Multiplier
Me = Concentration Multiplier (sometimes called k)
(can vary with design-about 2.75)
MT = Total Multiplier
PA = Pressure, applied, PSI
AUTOFREITAGE
SINGLE ELEMENT HOOP STRESS
..: .............................,..............................,....... .;.............................................

.......................................................... 0000 . . 0000 ................
50,000 100,000 150,001) 200,000 250,000 300,000

STRESS, PSI 216,500
51,140 07,330 269 ,·>40
Figure 6.5A. Autofrettage Analysis-Piotted Data
AUTOFRETTAGE
ROSETIE CALCULA T!ON
::::::::::::~:~:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::~:~::::..:::::.. ....... .,
/
.................................. j···· .. ·+·····.. ····............... j
:::::::::::::::~::::::::::::::::::::::::··7·!...... :::::. ::::::::::. :::::::::::::: ................................
···:::::::::. .~:::::::::t:::::::::::: :::::::::::::::::::::::::::::::::::::::::::::::::::::·:::::::::::::::::::::::::::::::·:::::: ::::::::::::::::::::::::
60000 100000 150000 200000 250000

40,664 68.,445 STRESS, PSl 171 ~ 000 211,664
Figure 6.6A. Autofrettage Analysis-Rosette Gage Data

186 LIQUID ENDS
k = Concentration Factor (Approx. 2. 75, Cakulated 3. 7)

Rl Bore Radius, in.
R2 = Outside radius or shortest distance to outside wall.
6.4.2.7 CAlCUlATIOlllS FROM TEST DATA (See Figure 6.5A)

This method is preferred because it utilizes strain gage differential readings
averaging, superior to a single actual residual strain gage reading at fin;¡
with the pressure released.
[ l] Calculate SLOPE of Proportional Curve; Se!ect an Applied Pressure

low the yield and positively on the Curve, for example, at (20,000) PSI, ,
could apply greater accuracy by averaging severa! points on that line.
Slope = Stress/Pressure = (2911) x 30 (20,000) = (4.37)
[2] Project Proportional Curve to Maximum Applied Pressure;
Projected Stress = Slope x Maximum Applied Pressure

= 4.37 x 50,000 = 218,500 PSI
[3] Select Actual Indicated Stress at Maximum Applied Pressure, (269,640,

[4] Determine Stress Difference;
Delta Stress = [3] - [2] = 269,640 - 218,500 51,140 PSI
[5] Calculate New Allowable Working Pressure;

Project line from Maximum Applied Pressure Stress (269,640) par¡i
Proportional (Slope = 4.37), through [4] (51, 140) to intersect zero (Ó
sure abscissa at the New Working Pressure;
New Working Pressure = (51,140)/(4.37) = 11,750 PSI without
tensile stress at the critica! point in the working pressure range.
6.4.3 Shot Peening

Shot peening is usually scientifically applied to the critica! areas in a liquid
bombarding them with glass beads. The shot particles in effect "peen" the
layer, compressing the "skin" surface and in tum protecting against có
fatigue.
6.4.4 Coating and Plating

Coating and plating have been used as protection against corrosion (subse
against corrosion fatigue), but because of the strict requirements for comple
erage (no "holidays") and the fragility of such coatings, they are seldom ~,
6.5 ULTRA-HIGH PRESSURE PUMP DESIGN 187
WATER COMPRESSIBILITY
2·9 r·\·T . . . :....... .......... :. · · ."[' ...... f.. . T..... ·: ................ .
~
FACTOR ll
1 ...., : : : . . : :
r::¡ :~~; T ! TI < , ., .

'·ut··········¡···········'~···>···················: ···.-·········:·····················:··········:
:J l I t?;LJ~~l~J__· .•
"!·········· ........ -······· ....................................................•..........·
2.2
o 10000 2000!l 30000 40000 50000 6000!) 70000 60000 90000 100000
PRESS\JRE, PSI
Water Compressibifüy
Compressibility Factor jj, x I0- 6 Contraction in Unit Volume per PSI Pressure
tCompressibility from 14.7 PSIA at 68º F.
UlTRAmHIGH PRESSURE PUMP DESIGN
requirements for hydraulic pressures extending into the 100,000 psi range
'~"·u~'"'"' and the most severe problem in pumping at high pressure liquids
resulting from excessive clearance-volume in a pump, that volume of liq-
.aining in the cylinder between the suction and discharge valves after the
ement element (piston or plunger) has completed its delivery stroke. Typi-
mp design involves significant ammmts of clearance volume, as a percent-
f displacement, in the order of l 00 to 200 percent, an ammmt that does not
t a serious problem below about 5000 PSI.
basic reason that a high clearance volume contributes to the poor perfor-
of a reciprocating pump is because of the compressibility of liquids, Fig-
Picture a cylinder ful! of high pressure "compressed" just as
unger starts its return (suction) stroke. Before the suction valve can open,
ble movement of the plunger is required to allow the compressed liquid
to the suction pressure. A part of the useful stroke is forever lost as a
of completely filling the cylinder and the ful! capability of the pump is
re 6.8A shows proposed suction valve design which takes advantage of a

feature of reversing the suction valve spring from the cylinder side, or
188 LIQUID ENDS
LIOUIO ENO
/777777:_:/
f\IN!f\Uf\ CLEARANCE VOLUf\E
NOTE STYLE OF VALVE ANO PLACEf\ENT
Figure 6.SA. Ultra High Pressure Suction Valve.
UHP PUf\P
!ULTRA HIGH PRESSUREJ
LIOUID ENO BLOCK
FEATURES:
ASSISTED SUCTIDN VALVE CLDSURE
'ZERD ' CLEARANCE VOLU/\E
Figure 6.9A.
top, of the valve to the suction or bottom side, thereby removing a

volume occupied by the spring and its associated mechanism from t
side, resulting in the ability for the displacing element to sweep almos
cent of the cylinder volume.
Another feature, not formerly recognized, that will improve hig
pump performance is the "augmented" closing force on the suction
sulting from the pressure surge created by the reversal of the plunger
of its delivery stroke, by placing the suction valve directly in front of:
6.5 ULTRA-HIGH PRESSURE PUMP DESIGN 189
nger. That pressure surge is aimed directly toward the open suction valve,
the closing force of the valve spring at the instant when rapid valve
s desired to minimize delay and leakage through a partially closed valve.
mp design places the suction valve in positions whereby such assisting
greatly attenuated by turns and restrictions in the path from the end
lunger to the valve. (Quick-closing of the discharge valve is also desirable
an be obtained by the use of an extremely strong spring load without
g the pump cylinder filling problems). See Figure 6.9A.
understood that a single-acting multi-cylinder (probably a triplex) plunger
ould be selected because of the inherent and desirable anti-extrusion ef-
packing friction counteracting the hydraulic pressure extrusion effect. See
'NGER VERSUS PISTON PUMP.
s 6. lOA and 6.1 IA are examples of "high-pressure" pumps now on the
A power pump liquid end for operation at pressures up to 30,000 psi.

ght Kobe, /ne. Used with permission.)
190 LIQUID ENDS
11 = Suction water
1 = High-pressure water
IJ = Suction valve
= Pump pistan
Courtesy
Paul Hammelmann
Maschinenfabrik GmbH
Zum Sur.dem 13-21 · Post Box 3309
0·4740 Oelde · W.-Germany
Phone 02522176-0 ·Telex 89455
Figure 6.HA. Ultra-high Pressure Pump
6.6 ABRASIVE JET CUTTING
"NOTHING ON EARTH IS SO YIELDING AS WATER, BUT FOR

BREAKING DOWN THE FIRM AND STRONG IT HAS NO EQUAL"
-Lao Tzu 600 BC.
Many industries are looking at the use of a high-velocity water jet as a cutti
tooL In order to generate the high velocity required, high pressure and low vq
ume are the parameters needed. Pressure in the order of 60,000 PSI or more
used. For cutting metal, an abrasive like fine silica sand is introduced into t
jet stream at the sapphire cutting tip or nozzle.
Several years ago a drilling contractor ran a test oil weU drilling rig whereb
the formation rock was disintegrated by the use of 10,000 PSI drilling m
charged with chilled steel shot as the abrasive medium. The high pressure m
was pumped through typical roller bearing rock bits equipped with special sm
diameter nozzles, the disintegration of the formation being purely the result
impingement of the shot-laden mud. Here again, intensifier type of pressure ge
eration was used. This system compromised on the pressure-volume relations
on account of the relatively high circulating rate of mud to carry the rock cutti.
to the surface. Incidentally, the drilling rate in hard rock was increased about
fold but the overall operating cost was found to be prohibitive.
6.7 THE INTENSIFIER 191
POVER-OPERATED
FOU.-VAY VALVEI
_,. .....- ......

J X 1 11UM.D' llta.I ..
10· X 24º P~ C'Yl.IN>ER 7• X 24• HIOH·PRESSlRE C'Yl.IN'.>ER HIGH·PRESSlRE LIQUID Tf.H<

100 GPW 5000 PS 1 43 OPM 11.:1110 PSI
Figure 6.12A. Typical Single-Cylinder Intensifier
THE INTENSIFIER
e intensifier is an ingenious but rather costly way of generating low-volume

'gh-pressure services for such applications as cutting materials, liquid sand-
lasting, autofrettaging and oil well formation fracturing and cementing, or other
uirements for extremely high pressure to about 100,000 PSI.
The principie usually involves the output of a medium-high pressure recipro-
ting pump of about 5000 PSI to operate a differential tandem direct-acting
draulic pump-a large diameter piston and cylinder connected by a common
·ston-rod to aoother small diameter piston in a cylinder. The pressure magnifi-
tion being directly. related to the square of the two piston diameters.
Figure 6.12A shows the basic principie involved in a typical intensifier. lt
ould be noted that the delivery or ftow-variation generated pulsations of the
imary pump will .not be reftected and magnified by the same percentage. lt is
portant that the primary pump be well dampened in order to protect the system
om shock-pressure waveforms generated at the intensifier reversals.
In order to eliminate the inherent zero delivery pressure occurring at each end
the piston stroke, sometimes two or three such units are operated in parallel,
e stroking being phased by sequential operation of the four-way valves being
grammed by end-of-stroke detecting switches (micro or proximity). A severe
rge problem can arise from improperly timed valves.
Why not generate the ultra-high pressure directly with a typical multi-cylinder
wer pump? The problem of handling a high-concentration of abrasive solids,
ch as used for "sand fracing" oil well formations (the high pressure separating
192 LIOUID ENDS
the formation layers with the sand grains acting as a "propping" agent
stimulating the rate of oil flow to the well bore), takes advantage of th~
long-stroke and low stroke-reversal-rate of the intensifier. Por other appli
with clean liquids, properly designed multi-cylinder pumps could be use
conserving cost and space. A discussion on the design of ultra-high p
liquid ends is covered in Section 6.5.
7
EXPENDABLE PARTS
API STANDARDS FOR SLUSH PUMP COMPONENTS*
The American Petroleum lnstitute (API) has rendered a valuable service to the oil-
well drilling industry in promoting standards for mud (and sorne slurry) pumps,
pistons, and piston rods, liners, valve pots, and extension rods. Standardization
has made it possible to use such parts interchangeably in any malee of pump.
Such standards also serve as a guideline in the design of nonstandard parts.
The API Standards on Parts in the following pages are reprinted here by per-
mission from the American Petroleum Institute, 211 N. Ervay, Suite 1700, Dallas,
Texas 75201. Figure, table, and paragraph numbers have been changed to be con-
sistent with the rest of this volume.
S/ush Pump Piston Rod and Piston Body Bore, Fluid End
7. 1 Sizes and Dimensions. Fluid ends of slush-pump piston rods and piston
body bores shall be in accordance with Table 7 .1 and Figs. 7 .1 and 7 .2 for double-
acting pumps and Table 7 .2 and Fig. 7 .3 for single-acting pumps.
7.2 Threads. Threads on rod ends and in retainer nuts shall conform to the
dimensions given in Tables 7 .1 and 7 .2, and shall be controlled by class X gages
conforming to the stipulations in ANSI Bl.2: Screw Thread Gages and Gaging.
If supplementary production or working gages are used, they shall be accurate
copies of the master gages.
*Nomenclature used in Section 7 .1 may not agree with that used in the main text.
193
194 EXPENDABLE PARTS
TABLE 7.1. Fluid End of Double-Actlng Slush Pump Plston Rods and Plston Body
Bores
(All dlmenslons are in lnches. See Flgs. 7.1and7.2.)
2 3 4 6 7 8 9 10 11 12 13
Piaton Rod. ~iat.ou Pilt:on
.. Rod
Taper,
In. Per
Plotoa Rod Ft. on
.. Rod .Diamet.er Diam•
a......
Ta-
No.
1 1 -lñ
A
3'!!! 1.000 1'1.i 0.979 u 1%
±•····
1.000
K
'Á
2 l'Á - lH 5'1.i 1.250 2'1.i 1.229 1,i. 1 '!!i 1.000 'Á
a 1'1.i - lH 7'1.i 1.500 2% 1.474 ls\ 2tt 1.250 'Á
4 l'!!i-2ñ 8 1.875 4 1.854 l/r 2U 1.000 'Á
5 2'Á - 211 8% 2.250 4 2.229 lU 2U 1.000 'Á
6 2114 - 2H 9% 2.750 4'1.i 2.729 2% 2U 1.000 'Á
5HP** 2114 - 3'1.i 8% 2.225 3% lU 2.229 lU 2U 1.000 0.041 0.113
6HP 3 - 3'1.i 9% 2.725 4'Á 2n 2.729 2% 2U 1.000 0.041 0.113
·~.~jjrfn:~~meter tolerancea for API rod numbers 1 and 2:+0.0l0¡-().005 inch. For rod number 3 and larger: +
••Recommended aa a substitute for API 6HP piston for reduced liner sizes only.
H•Dimeneion G, column 8, relates to dimension S, min. only (column 12).
J ~t PISTON (HANOTIGHT)
/¿R OR
~~------ _ TAPER K !.........,_.................,~.~·

~~O~L~tiNOCCURS _J __ 1 1 ~
t-4-----D~ r--E
i---------- 8---
Figure 7.1. Tapers 1 through 6.
7.3 High Pressure Pistons and Rods. Shoulder faces M and N of pisto
rods numbered 5HP and 6HP shall be square to the taper within 0.001 inc
indicator reading (TIR). Shoulder face P shall be square to the taper within .
inch TIR.
7.4 High Pressure Piston lnstallation lnstructions.
a Clean rod and piston tapers and assemble (oíl free).

b. Piston must stand off from rod shoulder when made up handtight.
c. Apply lubricant on thread and nut faceto prevent galling.
d. Draw piston to rod shoulder with nut.
e. After initial shoulder contact, mark relative position of both nut and p
with punch marks or paint stripe. Continue tightening 60 to 72 deg
7.1 API STANDARDS FOR SLUSH PUMP COMPONENTS 195
'-._BREAK b, MAX
~ MAX-j
-~1
FlLLETS a ÜÑDERCUT DIA TO BE PRE-

STRESSED BY COLDWORK!NG
UNDERCUT DETAil
. . ,._________ ª
Figure i .2. Tapers 5HP and 6HP.
TABLE 7.2. Fluid End of Single-Acting Slush Pump Pistan Rods ami Pistan Body
Sores
(A!I dimensions are in inches. See Fig. 7.3.)
2 3 4 5 6 7 g
Piston Rod Piston
..-----...
Piston Start
andRod Length of Thread
Con- Connection Rod from Shoulder Bo:re
nection Diamete:r~ Rod End, Shoulder, Diameter Thread
No. nominal Diamete:r, ±,¡.. maximum ±-h Designation
A B e D
SA-2 1 0.997-0.999 4i'. 1~ 2 l-8UNC-2A 1.000-1.003
SA-4 l'h 1.497-1.499 5t. 2% 3% l'h-8UN -2A 1.500-1.503
2~ splines for API 5 locknut with 12 splines and 2~ to 3 splines for API 6
locknut with 15 splines).
Markíng
a. Pistons, Double Acting. Pistons with a taper confonning to this specification

shall be marked with the manufacturer's name or mark, the API monogram,
and the taper number. High pressure pistons number 5HP and 6HP are di-
Figure 7 .3. Fluid end of single-acting slush pump piston rod and piston body bore. See
Table 7.2.
mensionally interchangeable with pistons 5 and 6. It is permissible to stamp

both tapers on shoulder P.
b. Pistons, Single Acting. Pistons with straight bores conforming to this spec~.
ification shall be marked with the manufacturer' s name or mark, the API
monogram, and the connection number.
c. Rods, Double Acting. Piston rods conforming to this specification shall be
marked with the manufacturer's name or mark, the API monogram, and the
taper number. The crosshead extension end of the piston rod shall be marked ·
with the API monogram and the taper thread number or the straight thread
number from Table 7.3 or 7.4.
d. Rods, Single Acting. Piston rods conforming to this specification on the flui
end shall be marked with the manufacturer's name or mark, the API mono
gram, and the connection number. If the crosshead extension end of the·
piston rod conforms to Par. 7 .6 or 7 .9, this end shall be marked with the
API monogram and the taper thread number or the straight thread numbe
from Table 7.3 or 7.4.
Slush Pump Crosshead, Crosshead Extension, and Piston Rod

Connections-Tapered Thread Type
7.6 Sizes. Tapered thread type connections between crossheads, crosshead ex-
tensions, and piston rods shall be 8 TPI, Series UN, Class 2A-2B modified, in th
sizes given in Table 7 .3.
TABLE 7.3. Crosshead, Crosshead Extenslon, and

Plston Rod Connectlons-Tapered Thread Type
(All dlmenslons are In lnches. See Flg. 7.4.)
1 2 3 4 5
Length Length of Lock-
Taper of Straight nut
Thread Nominal Taper Thread, Thick-
Number Size,* Thread, Min ness,
A B e D
Tl 1 1% 1 %.
T2 1% lU 1 %.
T3 1% 1-h 1 %
T4 1% lit 1 %
T5 11h 1% H~ 1
T6 1% 2n 1% 1
T7 1% 2-ilr 1% 1%
TS 1% 2H 1% 1%
T9 2 21h 11h 1%
TlO 2% 2U 1% 1%
Tll 2% 3% 1% 1%
T12 2% 3* 1% 1%
Tl3 3 3% 2 1%.
T14 3% 4ñr 2 1%
T15 31h 4% 2% 2
Tl6 4 5 2% 2
T17 4% 5% 2% 2
T18 5 6% 21~ 2
T19 51h 6% 2% 2
T20 6 7% 2% 2
*Ali threads are 8 TPI, Series UN, Class 2A-2B modified.
7. 7 Thread Dímensíons and Tolerances. Tapered thread type connections shall

conform to dimensions given in Table 7 .3, Figs. 7.4 and 7 .5, and the following
tolerances:
a. Taper. Tapered threads shall have a taper of 2 in. per ft on pitch cone di-
Figure 7 .4. Crosshead, crosshead extension, and piston rod connections-tapered thread
type. See Table 7 .3.
TABLE 7.4. Crosshead, Crosshead Extension, and

Piston Rod Connections-Straight Thread Type
{All dimensicms are in inches. See Fig. 7.7.)
1 2 3 4 5
Lock-
Straight
Length
of In-
Le~¡th nut
Thick-
Thread Nominal ternal Externa! ness
Number Size Thread Thread Min.
A* B e D
Sl 1 1% 2% %.
82 1% lH 2ii
S3
S4
1%
1%
l/ir
lH
2/ir
2H
*%%
S5 11h 1% 3% 1
S6 1% 2ñ 3S\i 1
S7 1% 2-h 3i« 1%
SS 1% 2U 3U 1%
89 2 21h 4 1%
S10 2% 2ii 41'\r 1%
S11 21h 3% 4% 11h
S12 2% 3.[g 5i\- 1%
S13 3 3%. 5% 1%
$14 3% 4-fr 6~ 1%
S15 3'.lh 4% 6% 2
S16 4 5 7% 2
S17 41h 5% 7% 2
S18 5 6% 81h 2
819 51h 6% 9% 2
S20 6 71h 9% 2
*All thr!'ad111 are 8 TPI, Series UN, Class 2A-2B.
ameter with a tolerance of +0.000, -0.020 in. for intemal threads and
+0.020, -0.000 in. for externa! threads.
b. Concentricity. Within limitations of good practice, threads shall be concen-
trie with rod design axis. Angular misalignment of thread axis with rod de-
sign axis shall not exceed 0.0005 in. per in. of length.
---0-
Figure 7.5. Tapered thread form. See Par. 7.7,

7.1 API STANDARDS FOR SLUSH PUMP COMPür~ENTS 199
c. Length. Total length of externa! threads = B + B = 1.25.A.

d. Perpendicularity. Face of intemal thread member shall be perpendicular to
thread axis within 0.001 in. per in. of face diameter.
e. Lead. Lcad tolerance shall be ±0.0022 in. per in. Cumulative lead tolerance
shall be ±0.0022 in.
f. Thread Angle. Half angle tolerance of thread angle shall be ± l deg.
g. Truncation. Crest on both intemal and externa! threads shall be truncated
parallel to taper to produce a fiat 0.030 in. wide. Root on both interna! and
external threads shall be truncated paraUel to thread axis to produce a flat
0.015 in. wide. Root of interna! threads may be trnncated parallel to taper
of thread at option of manufacturer. Straight threads truncated same as ta-
pered threads.
h. Pitch Diameter. Pitch diameter and pitch diameter tolerance of straight
threads shall be as designated in [111] ANSI B 1.1: Unified Screw Threads.
L Standoff. In gaging tapered threads, stando:ff of product from plain and
threaded plug and ring gages shall be maintained within a tolerance of ± 1/16
in.
CAUTION TO USER: Threads must not be damaged, as damage will
cause misalignment and failure.
TB Lock Nuts. Crosshead extension and piston rod lock nuts shall be fumished
in accordance with Par. 7 .12.
Slush Pump Crosshead, Crosshead Extension, and Piston Rod

Cormections-Straight Thread Type
7. 9 Sizes. Straight thread type connections between crossheads,' crosshead ex-
tensions, and pistan rods shall be 8 TPI, Series UN, Class 2A-2B modified, in the
sizes given in T~ble 7 .4.
7. 10 Thread Dimensions and Tolera.neas. Straight thread type connections

shall conform to the dimensions and tolerances given in Table 7.4, Figs. 7 .6 and
7. 7 and ANSI B 1.1: Unified Screw Threads, and shall be gaged in accordance with
ANSI Bl.2: Screw Thread Gages and Gaging. The following requirements are
also applicable:
a. Concentricity. Within limitations of good practice, threads shall be concen-

tric with rod design axis. Angular misalignment of thread axis with rod de-
sign axis shall not exceed 0.0005 in. per in. of length.
b. Length.
Interna!: B = 1.25A
Extemal: C = B + D + 0.25
INTERNAL
THREAD
EXTERNAL
THRE'AD _J'v'~I" /
TH~EAD AXIS7 . ~ltr' PITOH--i •
-.___¡_______ -.J.-__L Figure 7.6. Straight thread form. See Par. 7.10.
c. Perpendicularity. Face of interna! thread member shall be perpendicular to

thread axis within 0.001 in. per in. of face diameter.
7. 11 Lock Nuts. Crosshead extension and piston rod lock nuts shall be fumished
in accordance with Par. 7 .12.
7. 12 Lock Nuts. Crosshead extension and piston rod lock nuts shall be fumished
in accordance with Fig. 7 .8.
7. 13 Taper Threads. Locknut threads for the taper type connection shall con~
form to the requirements of Par. 7. 7.
7. 14 Threads. Locknut threads for the straight type connection shall confo
to the requirements of Par. 7. 1O.
} ct-L·:
CROSSHEAD EXTENSION LOCKNUT PISTON
Figure 7. 7. Crosshead, crosshead extension, and piston rod connections-straight threa

type. See Table 7.4.
7,1 API STANDARDS FOR SLUSH PUMP COMPONENTS 201
Figure 7 .8. Crosshead extension and piston rod lock nut.
Slush Pump Valve Pots

7. 15 Sizes and Dimensions. Slush pump valve pots shaH be fumished in the
sizes and dimensions given in Table 7.5 and Fig. 7.9, as specified on the purchase
order. API valve pots for caged valves shall provide a minimum G dimension (see
Table 7 .5) for cage clearance.
A.BLE 7.5. Slush Pump Valve Pots

;(All dimensions are in inches. See Fig. 7.9 for explanation of dimensional symbols.
'oimensions for pot sizes 1, 2, and 3 are tentative.)
1 2 3 4 5 6 7 8 9 10 11 12
Sp:ring Mounting
Valve Pot Dimensions Dimensions
,--------- r---~
A B e D E F G J L M N
2% 2 1 314 So lid 1% Solid 21,4 1 2% 2 112
3% 2 Pis 3% u 2% 3% 21,~ 1% 3 314
3% 2 1%. 41,4 u 2% 33114 2% 1% 3 3%
4% 2 1% 4%. u 2* 41/s 2%. 2 3 3%

5 2 1 \/2 5% lf. 3 4% 3% 2% 3%. 4%
5% 2 1% 5% Ho 3% 514 3% 2%. 3% 4%
5% 2 1% 6 1·i'c. 3%. 51,4 3% 2%. 3% 4%
6% 2 2 6% l1i1 3% 5% 3% 2%, 3%. 4%,
7 2 2',i 7~~ 11:;1.i 3%. 6 3% 2% 3% 5
7% 2 2% s~;g lf. 4 6"',g 41/g 2% 3% 5%
8'11. 2 2Vs
31,1,
8% u.
H,
41,4 6% 4% 2% 3%. 5 1h
9% 2 9% 4% 7'/s 5% 2% 3% 5%.
*Dimensions for these pot sizes are tentative.
,jr ~
JT,n· ,i
CM•• .//, r---B lTAPER r'ER FOOT
t L .:!_:': ON DIAME+-~[R) 1
L_.___ .L ___ -- ______ JMINIMUM

CLEARANCE
Figure 7.9. Slush pump valve pot. See Table 7.5 for dimensions.
7. 16 Spring Mounting Dimensions. Valve pot spring mounting dimensions ·

shall conform to dimensions L, M, and Nin Fig. 7.9 aild Table 7.5.
7. 17 Marking. Slush pump valve pots fumished to this specification shall b

marked with the manufacturer's name or mark, the API monogram, and the valv
pot size number. Markings shall be cast or die stamped on the fluid cylinder
applied to a plate securely affixed to the fluid cylinder. Markings shall be applie
in a location visible after installation of the fluid cylinder on the pump and ma
be applied to either pot. For pumps having divided fluid ends, each section sh
be marked.
Slush Pump Pistons

7. 18 Sizes and Dimensions. Slush pump pistons shall be bored to fit the st .
dard taper of piston rods as given in Fig. 7 .1 and Table 7 .1. Piston outside di
ameters shall be suitable for use in liners or cylinders having increments of di
ameter change noted in Fig. 7 .10.
7. 19 Marking. Pistons conforming to this specification shall be marked with th

manufacturer's name or mark, the API monogram, the corresponding API
number, and standard bore. Markings shall be stamped in letters ~-in. high on t
end face of the piston core at the large end of the piston-rod hole.
Slush Pump Liners

7.20 Uner Bares. Bores of slush pump liners shall be gíven in one-fourth inch
increments and with tolerance as noted in Fig. 7 .10 and as specified on the pur-
chase order.
7.21 The inside edge of the piston end of slush pump liners
be chamfered as shown in Fig. 7.10.
Marking. Slush pump liners confonning to this specification shall be marked

with the manufacturer's name or mark, the API monogram, and the size (standard
bore) of the liner. Markings shall be stamped in letters ~ in. high on the retainer
end (outer end) of the liner.
Slush Pump Gear Ratings

Provisions. Gear ratings as given herein are derived from AGMA Std
424.0l · Standard Practice for the of Helical and Gearing far
Oilfield Mud Pumps. Ratings are based on surface durability (which is mn'""''n"
of pítch). However, the gear manufacturer shall as sume responsibility for selecting
a pitch sufficiently coarse to provide adequate tooth strength.
7.24 Design. Gears shall be single reduction, either helical or herringbone. Gear
materials to be in accordance with AGMA Std 241.01: Gear Materials-Steel.
While field experience in the use of nodular iron as a gear material in slush pumps
PIS TON
~:W4-1
ENTERING
END
__\_A
/ /.
!'
1/8" MIN
A= Noml1111I Diameter in one-fourth ind1 incremenh, tol-

eruc:e plus O.O 1O inchas, minus 0.000 iru:luu.
Figure 7.10. Slush pump Hner.

is limited, it does, to indicate this matelial can be used. Tentative use

nodular iron is permissible for gears only (not pinion), providing it is in accordanc~
with AGMA Std 244.01: Nodular !ron Gear Materials. Use the steel hardne
curves of Fig. 7 .12 to obtain Kr values. Any practica! combination of tooth height;
pressure angle, or helix angle may be used. However, American Gear Manufac:
turers Association standards are recommended. The slush pump manufacturer shaU
be responsible for adequate shafting and support to maintain proper alignment
derload.
7. 25 Rating Formulas. The horsepower rating for surface durability shall

determined from the following formula:
F;KrDs
P=--
where
F;.65F. Combined factor for face width and inbuilt factor (where F =
=
face width in inches.)
F, = rating factor, see Fig. 7 .11.
Kr = combined factor for materials, tooth form, and ratio, see Fig. 7 .12.
Ds = combined factor for pinion rpm, pitch diameter, and velocity factor,
using Fig. 3 of AGMA Std 211.01: Surface Durability of Helical and ·
Herringbone Gears, or the following formula:
D =---
v;cvn
s 126,000
1. 7 l-----+---+---f--__,l-----'----1---1----l
-
._ r,= 16
·- HP
5000
"-.i.6 le:-'-'!---+--_,, USE RATING FACTOR OF 1.4 ON
~ 1 PUMPS 1000 HP a HIGHER
t; 1 1 1
e;: 1.5 1
~ i
z
Si 1.4 >----+----+-,--1---J--~¡.....-.¡....._...¡.._....¡
a:: 1. 3 1---~-L_,__,___ _,__--+-----+----<
1 1
o 2 4 6 8 10 12. 14 16
INPUT HORSEPOWER RATING OF PUMP, IN HUNDREDS
Figure 7.11. Gear rating chart for mud pumps. See Par. 7.25.
600
500
......."'
o
u
,.
400
300 STANDARD HARDNESS COMBINATIONI

AS SHOWN ARE llU:COMMENDED.
WHEN NON•STANDARD HAADNEll

COMllNATIONS .U:E USED 1 THAT
CURVE SATISFVING IOTH GEAR a
PINION HARDNESIES SHALL APPLV •
NO INTERPOLATION IS PERNITTED.
3 4 5 6 7 8 s 10 11
RATIO OF GEAR TO PINION IEiiJ
Figure 7.12. Variation of K, factor with gear ratio. See Par. 7.25.
DP = pinion pitch diameter, inches. With enlarged pinions, a value equal to

outside diameter minus two standard addendums, may be used.
Cv 78/(78 + ./V)
v = pitch line velocity in fpm (do not use enlarged value of Dp).
n ·= pinion rpm.
7.26 Name Plate Ra.ting. The name plate rating of a slush pump shall not ex-
ceed the API rating of the gear.
7.2 PUMP LINERS
In the 1920s, the cylinders of reciprocating pumps were either an integral part of
the liquid end or employed a pressed-in bronze or cast-iron sleeve that requirecf
time-consuming and usually difficult rernoval using hammer and chisel to split the
sleeve.
The use of double-acting reciprocating steam pumps and then power pumps for ~
circulating drilling mud (and later slurries), usually a severely abrasive service
requiring frequent replacement, led to the adoption of a "quick-change" renew-
able cylinder or ''liner.'' The loose-fitting liner was equipped with a shoulder that
engaged with a short rubber sleeve to serve as a slug packing, being forced against
the packing in its recess by mechanical means such as set screws in an internal .
cage or through the cylinder head. Such an arrangement did not necessarily lend
itself to easy removal of liners. Corrosion and packed sand and mud between the
loose-fitting Iiner and the liquid-end bore often made necessary the use of me-
chanical or hydraulic jack arrangements either with a puller head that gripped the
inside of the liner by expanding serrated jaws or with a toggled head that engaged
the rear end ofthe liner. Such a puller required two jacks working against a "strong
back'' spanning the two jacks. As a matter of historical interest, the jenny jack (so·.
named after the female jackass) was developed with an open center hole that could
be slipped over the single puller rod.
Another objection to this liner-retention arrangement was the lateral movement
or "working" of the liner against the relatively compressible packing rubber,
caused by the pulsating pressure on the exposed end areas of the metallic liner.
The relative motion between liner and liquid-end bore in the presence of abrasive
sands resulted in rather rapid destructive wear of the liner bores in the liquid end.
This was overcome by the use of an additional shoulder on the liner to enable it
to make metal-to-metal contact with the liquid end. Opposed lip-type packing rings
were placed in the original packing space with the lips facing each other and with
a metal lantern ring between them to provide a small amount of compression for
an initial seal, the final seal being provided by the pump pressure. An improved
liner packing arrangement was developed in 1970 whereby the opposed pressure-
activated lip-type packing rings were installed in a packing space that extended
throughout the length of the liner, this space being filled with pressurized oil. This
resulted in complete. protection against corrosion and packed sand and also pro-
vided the desirable pack-o:ff against clean oil rather than abrasive mud or slurry.
With the .introduction of the single-acting piston-type pump in drilling and slurry
service, the means of packing and clamping the renewable liner required a di:fferent
approach. The usual practice was to install the liner through the cradle opening ·
and force the liner end against a rather hard, fabric-reinforced packing ring in a
recess provided in the liquid end. The bolted or screwed clamp sometimes works
against a shoulder or fiange as an integral part of the liner. At other times the
clamp may work against the outside end of a straíght cylinder liner.
Ease of installation and simplification of the clamping arrangement can be had
7.3 PISTONS 207
designing the liquid end to receive a flanged liner through the cylinder head.
This precludes the need for clamping against the high hydraulic ram effect working
on any since the hydraulic forces tend to hold the liner against its shoulder.
7.2.1 Liner Development
The value of any liner is the wear resistance of the inside diameter against which
the piston runs. Thís surface must be compatible with the chemical constituents of
the pumped liquid and the abrasivity of the solids that may be deliberately added
as in the case of drilling mud and slurry.
The first typical "quick-removable" mud-pump linern were of single-metal
construction-cast iron, hardened carbon steel, and chrome-plated steeL Bimetal
liners with centrifugally castor spun hard metals were introduced later. However,
about 1950, a trend to 27% chrome iron was recognized by all pump liner manu-
facturers. Sometimes this hard metal is centrifugally cast into a liner shell, but
other liners consist of a sleeve of the hard metal centrifugally cast and machined
and shrunk into a mild steel or carbon steel shell. This construction results in a
liner with greater hoop strength because of the prestressed liner and shell. On the
other hand, the sleeve construction limits the maximum size of liner to a diameter
considerably less than a spun liner because of the loss of thickness of the sup-
porting shell. A spun-in hard facing is usually much thinner than a shrunk-in sleeve
and contributes little to the strength of the unit.
Of significance is the "rebirth" of the chrome-plated liner. Tests show that a
properly plated liner will surpass one of 27% chrome iron. The reason for the
demise of the early chrome-plated liner was the poor plating technique and appli-
cation of a minimal thickness of chromium. Old liners (and piston rods) hada limit
of about 0.010-0.015-in thickness of plate. As this relatively thin coating wore
away, the base metal would become exposed, and the edge of the wom chrome
plate acted as a knife to quickly decimate a piston rubber (or rod packing). The
improved plating technique and the thicker plating (which allows the acceptable
maximum w.ear off¡, in in diameter) results in a superior product.
7.3 PISTONS
The following discourse on the introduction of an "improved" piston for drilling

mud service will provide information on the physical action of a piston and reveal
sorne of the pitfalls that should be avoided. Examples of current piston designare
shown in Figures 7 .13 and 7 .14.
A piston with a solid steel back-up plate bonded to the fabric heel was marketed
with the contention that it offered the ability to use several sizes of rubbers on a
single size piston body (a dubious advantage) and it al so provided a clearance-
renewable feature in that with each new rubber installation the liner clearance was
Figure 7.13. Mud and slurry pistons. (Courtesy Fluid King.)
brought back to the "new" piston condition (Fig. 7 .15). Sorne detrimental effe
of having a "floating" back-up plate were observed. (Even though it was bond
to the rubber, it was rather free to float radially.) The first evidence of trouble w
the galling that appeared in 27% chrome-iron liners. No thought was given at t
time to the possibility of excessive plate "shifting" (Fig. 7 .16), but there w
concem for the damage to the liner. lt was determined that the galling produced
"fish-scale" surface on the liner that was extremely hard on rubber pistons. T
(a) (b)
Figure 7.14. Typical pistons. (a) Double-acting. (b) Single-acting. (Courtesy Fluid Kin
7.3 PISTONS 209
Pistan body
Figure 7.15. Piston with "back-up" plate.
fact that any steel piston Figures 7 .17 and 7 .18, without plates could still result in
the same injury with the steel body galling on the liner tended to dispel any worry
about back-up plate shift.
during the development of a large piston ( 12-in diameter) for coal
· slun-y pumping service, it was dramatically shown that the back-up plate shift was
undesirable. Figure 7.16 shows how the shifüng of the back-up plate generates
liner and piston wear on the opposite side from the clearance gap due to the high
unit loading of the back-up plate against the liner.
The suggestion that the back-up plate be split into segments and unbonded so
that it would expand equally with the rubber is not practical, because the rule
-E-
Pressure
Fabric extrusion into clearance

results in rapid failure o! heel,
and wedging efíect
produces force on back-up plate
t
This generates wear
on back-up plate,
and the failure
Hydraulic pressure
mode regene~ra.,.,te~s~.-;-1,
bulges rubber to help
11111 center piston but
¡:11¡ 1
provides no centering
to ring.
Figure 7.16. Piston with "back-up" plate. Reaction to pressure.

'i¡i
111,111
111111
Figure 7.17. Typical piston.
applying to any packing system is that the packing ring next to the clearance gap
performs all of the packing effects-the remaining multiple rings become pressure ::
balanced and perform no work except to wear out from the initial interference buil{
into the ring.
The argument that any piston (Fig. 7 .17) can be ''kicked'' over by the wedging
action of the fabric into the clearance gap (Fig. 7 .18) is true, but the ability of
solid piston to resist this force is so great that there is hardly any regeneration. ,
The solid piston takes this force on both rubbers and a much wider metal face
(body fiange). The hydraulic pressure on the working rubber tends to "bulge" it
out to liner size for the entire circumference, thereby acting as a centering forc.'
Fabric extrusion into clearance

produces force
But force is counteracted by "rigid"

rubbers and steel flange in contact
with liner. Otters greater bearing
area with reduced wear rate.
Hydraulic pressure
bulges rubber, provides
~ additional force to
"center" piston.
Figure 7 .18. Typical piston reaction to pressure.

7.4 PLUNGERS 211
11111!1ll1 51 1 l l l 1!11111
Figure 7 .19. Improved piston (patented). Controlled expansion (by bonding or vulcanizing
to fabric section), high-modulus, gap-closing, compatible-with-liner, anti-extrusion ring.
(Courtesy Chromium Corporation.)
for the entire piston, overcoming the side force produced by extrusion. Such a
centering force is lacking in intensity in the "back-up plate" piston. Accordingly,
it appears that the use of any one-piece metal or high-modulus back-up plate is of
doubtful value.
lt is surprising to observe that in the mid- l 950s, drilling mud pump piston
construction including the calandered 18 X 18 cotton-duck fabric plied into
about 20 plies (later improved by stacking the plies with the warp and woof at
random angles) is practically the same as its original design. This is great testi-
mony to the sanctity of the fabric with its special properties-probably due to
the ability of the fibers to absorb liquid and provide lubrication against the liner
on the high pressure stroke, not an attribute of most synthetic fabrics.
Of interest is a new concept in piston design, Figure 7 .19, whereby the clear-
ance gap is kept at zero by the controlled expansion of a high-modulus segmenta},
step-cut anti-extrusion ring. Because failure of a piston begins with the deteriora-
tion of the fabric heel at the clearance gap, this means of reducing the gap will
extend the life ofthe piston. See Chapter 10, Sec. 10.1.5.
7.4 PLUNGERS
Plungers are made in a wide variety of materials, the most common being solid
ceramic, sprayed ceramic, and spray-welded Colmonoy 6. For clean water service,
solid ceramic offers the best performance, lasting many years with no discernible
wear. However, it is subject to damage from abrasive liquids, rough handling, and
thermal shock. Spray-welded Colmonoy 6 is probably the most popular plunger
Figure 7 .20. Plungers, val ves, and packing for typical multicylinder pump. (Courtesy Fluid
King.)
material, resisting corrosion and abrasion fairly weH, and having none of the ce-
ramic disadvantages. Refer to Figure 7 .20.
There are many other plunger materials, including sprayed ceramic, stainless
steel, and even hardened carbon steel.
Plunger construction varies with size. For plungers up to about 3 in in diameter,
solid base metals are usually used. Those of larger diameter are usually of hollow
or sleeve construction, which reduces weight and cost.
In horizontal pumps, most plungers have short, quick-connect ends to allow
plunger change without opening the power end for access to the crosshead. An
extension rod working through the diaphragrn packing has sorne sort of grooved
clamping arrangement to allow the plunger to be fastened by working through the
cradle opening, using a clamp of sorne sort.
In order to ensure plunger alignment, the stub end should use a pilot-fit pin and
socket. Early attempts to use a so-called self-aligning connection without a pilot
fit resulted in disaster because of the care needed to assure that the plunger was
aligned before clamping. Sorne clamps actually forced the plunger out of alignment
during the tightening process.
The outside crossheads used on most vertical pumps contribute to good plunger
alignment because of their freedom of movement.
7.4.1 Metal-to Metal Plungers

0
With clean oil or with water and soluble oils, metal-to-metal close-füting plungers
running in an appropriate sleeve, both parts being honed and lapped to extremely
close clearance, are rather popular in smaller high-pressure pumps. Plungers can
7.4 PLUNGERS 213
be of hardened (nitrided) steel or chome An 0-ring is sometimes placed in

a groove in the sleeve near the atmospheric end. This prolongs the useful life of
the plunger by the additional seal afforded as clearance develops from wear. For
an estímate of the leakage rate to be expected with such a close-fitting plunger,
use the formula below. Of course, the leakage into the cradle would have to be
disposed of by drainage ora scavenger pump.
where
Q leakage rate per plunger, GPM

=
Dp = plunger diameter, in
Pd = discharge pressure, PSI
e = diametric clearance, in
µ = absolute viscosity, cP
L = length of plunger, sleeve contact, in
Example
For a triplex single-acting pump with
DP= l in
= 5000 PSI
e= 0.001 in
µ = 3 cP
L = 6 in
Q = 29,308 X 1 X 5000 X 0.001 3 X 6 X 3* = 0.024 GPM
7.4.2 Ceramic Plunger Construction (See Fig. 7.21)

The original supplier of solid ceramic plungers utilized the typical constru... : · 'Jn of
a solid ceramic body with a reduced diameter neck or shank epoAy-c.:.:nepted into
a metal adapter for fastening to the extension rod (Fig. 7.21a). After 's 1arge num-
ber of "pullouts" or bond failures, an effort to the cemer.tL.: connection
by deepening the socket of the adapter to provide greater bonding area was tried
(Fig. 7.2lb). Loss of adapter strength '°rmn such removal of metal resulted in
numerous cases of adapter failure from fatigue. Another manufacturer obtained
better results by following the same design except that the adapter was heat-shrunk
onto the ceramic shank.
A superior ceramic plunger construction is the hollow "thimble" of solid ce-
*For three cylinders.

cjfl:"' &,
"Pullout" due
to failure
of cement
Breakage of cerarnic
due to mishandling
or misalignment of
1
1
Breakage of ceramic
due to thermal shock.
Craze-cracking
pump usually evident
Heat - shrink adapter

b ---+-
Adapter fails by fatigue Ceramic failure same

at either location as above
e
c[J- _<_c_em-en_t>__ )
(
/=(~)'" ~ T\./O BOL T CLA/"\P
.
---~
d
PLUNGER
..
¡ .
l )
• () 1
·~··
Plunger clamp connection
Figure 7.21. Ceramic plunger construction. (a) "Old" design. (b) "New" design. (
"Improved" design. Adapter with integral stinger provides strength of bond and reduc.
stresses. (d) Typical Groove-and-Clamp Connection.
ramic cemented toan all-metal adapter with an extended "stinger" that provid
strength and reduction of stresses (Fig. 7.21c).
Aside from the screwed method of attaching the plunger to the extension r
another popular method is the 'groove-and-clamp' scheme shown in Figure 21~
Note that a pilot-fit boss assures absolute axial alignment.
7.5 Pl\CKING 215
7.5 PACKING
7.5. 1 lntroduction and History
The terms "packing" or "seals" are sometimes used to describe two distinctly
different applications; namely, "dynamic" packing that usually refers to tha!
used on a moving piston or plunger and to a stuffing box through which a plunger
or piston-rod reciprocates (or rotates in the case of a centrifuga! or rotary pump).
Then there is the term "static" packing, generally called "gaskets", that are
used to sea! the gap between two fixed or stationary parts, typically pipe flanges,
pump cylinder heads and valve pot covers or any other opening that must be
permanently or temporarily sealed.
Many packing types or shapes (lip-type rings, homogeneous or braided, flat,
0-rings, metal-segmenta!, etc.) are used in either application with al! types in
both applications being subject to extrusion into the gap between the two parts
but it will be seen that the mechanics of sealing are entirely different in the two
applications-the dynamic packing being subjected to the additional friction
forces and wear not encountered in static application and the greater ''gap'' nec-
essary between moving parts.
A discussion of plunger pumps and packing must be preceded by an expla-
nation of why they differ from piston-type pumps, particularly in the matter of
stuffing boxes and packing.
With the common duplex double-acting piston pump, typical design requires an
increase in piston rod strength, in both tension and compression, in sorne propor-
tion to the increase in pump discharge pressure. Therefore, a point is reached
where the piston rod diameter theoretically becomes so large that the pump, in
eftect, approaches a single-acting duplex, the discharge characte1istics of which
are extremely ''rough. '' Accordingly, single-acting pump design dictates three,
five, or more cylinders, and multiplex single-acting pumps inherently have
smoother discharge characteristics, even over a small-piston-rod duplex double-
acting pump.
Of extreme importance is the often overlooked fact that the packing action of
an outside packed plunger pump is completely opposite to that of any piston type,
including a multiplex single-acting piston pump.
In a plunger pump (Fig. 9.3), the plunger, during the pressure stroke, is trav-
eling to the right out of the pressure-loaded packing into the liquid, and during the
suction stroke the plunger is traveling to the left out of the dirty liquid into the
relaxed packing. Conversely, in a piston-type pump (Fig. 9.4) (with both the pis-
ton and the piston rod packing), on the pressure stroke the piston is traveling to
the left into_the pressure, and on the suction stroke it is tmveling to the right away
from the liquid. With the piston rod packing, the same action is seen: on the
pressure stroke the rod is traveling into the packing, which is loaded by hydraulic
pressure.
The purpose of packing ·is simply to close up the clearance gap between the
moving plunger and it associated parts, particularly the gland bushing, in the
stuffing box or the piston and its cylinder, and the pistan rod and its stuffing box'
parts. With ordinary packing this is accomplished by the use of material with con~
siderable resiliency.
The mechanics of ali packing are such that regardless of the general shape
the sealíng member, the hydraulic pressure tends to force the member through the,
clearance gap. Accordingly, practically ali of the sealing and subsequent wear
extrusion take place at the "heel" (Figs. 9.3 and 9.4).
It can be seen that the action in a plunger pump (Fig. 9 .3) is such that on the
pressure stroke the heel is being "dragged" away from the clearance gap, thereb
greatly overcoming the force produced by the hydraulic pressure that causes
trusion through the clearance gap, a benefit in high-pressure service. With
piston-rod packing (Fig. 9.4), the heel is being dragged into the clearance gap
both the motion and the hydraulic pressure, accelerating wear of the packing.
Lubrication of packing is extremely important in high-pressure service. It
be seen that only with the plunger pump can a lubricant be applied to the plung
as it is entering the hydraulically loaded packing, when it is most needed. This i
another benefit in high-pressure service.
Any attempt to lubricate a piston or piston rod is not as effective, since th.
lubricated moving parts enter the packing only on the unloaded or suction strok
when lubrication is not required.
The l 940s saw an intensive search for an improved packing. Early styles of
ing went through a period of popularity due to a false impression that one packi
was better than another, when in reality a "different" unworn packing would seal
for a while in a "different" location in the stuffing box, away from the wash'-~
boarded area caused by the previous packing. is justas important for a packi ·
to sea! on the inside surface of the stuffing box as for it to seal on the rod
plunger.) ...
All of the packings at that time were more or less adjustable in that gland
ening caused the packing to squeeze with great force against the rod. An excepti·
was packing with altemate metal spacer rings with pins extending through hol
in the packing rings, the gland force being transmitted through these pins. Thi
was a fairly successful packing, but trouble was encountered in high pressure
the pins would crush and distort under the high load.
At that time a development program was carried on, and as a result the füsf
truly "nonadjustable" packing was made available. This is a lip-type packing wit~
a fairly hard, fl.at center section that will withstand considerable gland load withou.t
affei:-~ing the lip load on the rod (Fig. 7 .22).
7 .5.2 Packing Types

The type of service in which a pump operates determines the type of packing t
be used. The choice is not always simple, but in order to have a guide to t
selection of packing, a discussion of each type follows.
General Servíce. A nonadjustable packing set composed of rather hard phenoH

bottom and top adapters in combination with nonadjustable seal rings, The
7.5 PACKING 217
• J
Figure 7.22. Nonadjustable packing style 0740. (Courtesy Utex Industries, Inc.)
ring has a phenolic core, and the sealing lipis composed of Buna-N and fabric or
bther materials. It is recommended for general service on water, oils, hydrocar-
bons, alcohols, glycols, and amines. Depending on local service conditions and
maintenance, it will perform satisfactorily at pressures up to 5000 PSI and tem-
peratures to 200ºF. See Figure 7.22.
Acid Service. Same design and configuration as general service except that the
sealing lipis homogeneous Buna-N. TJ:iis packing must be lubricated to give sat-
isfactory service in well service pumps. Again rock drill or steam cylinder oil is
recommended. Since most well service applications are intermittent, the packing.
will perform up to 15,000 PSI and will give satisfactory service in the fluids com-
monly used in well service operations. Occasionally it is necessary to use this
packing in services that are highly abrasive due to slurries or sand content.
Organic Service. Same design and configuration as general service. The sealing
lipis Buna-N, nylon, and Teflon. Organic service packing is recommended when
problems are encountered dueto corrosion or organic action on the composition.
The addition of Teflon offers sorne resistance to corrosive and organic action in
such services as lean oil, sulfur concentrations in water and oil, and sorne amines.
This is recommended only in isolated instances.
A molded duck and synthetic oil- and water-resistant composition

packing, recommended for fluid rods on pumps handling oil and water-base muds.
The packing is designed with noncrushable features, and thus excessive gland or
fluid pressure will not cause the usual crushing action of the packing with resultant
undue wear on packing and rods. However, the packing may be adjusted to adopt
to rods that are undersize.
High-Pressure Service. An adjustable packing composed of die-formed lead,

die-formed flax, and molded phenolic bearing rings. This is recommended as an
alternative to general service packing in local areas where combinations of pres-
sures, temperatures, and maintenance practices dictate a change in styles. It will
be serviceable up to 8000 PSI and at temperatures up to 240ºF. Since it is adjus

able, it is subject to human error. lt is recommended for the same broad classi
cation as general service packing and can be used as an altemative in applicatio
where the customer cannot use lubrication.
Power Oil Service. This is a die-formed Tefton filament packing with bro
end rings. The classification is misleading, since the packing can be used on
myriad of services. The Tefton is very susceptible to adjustment due to the
expansion, and it has not been too successful due to this. Continua! and judi
gland adjustment is required. ·
7.5.3 Split Packing Rings

An age-old practice is the use of cut-ring packing that can be installed in a pu
without removing the rod or plunger. Whenever possible, pumps should be ..
signed with readily removable rods or plungers. The use of solid or uncut pac . .•.
rings will repay the user in extended life. Split rings present a leakage path t.
sometimes requires excessive gland tightening to stop the leakage.
7.5.4 Spring-Loaded Packing

The use of springs to load or activate packing rings is not new. However,
advent of new materials coupled with a better understanding of the mechanical
hydraulic functions of the stuffing box have regenerated considerable interes
this concept. ...
Sorne of the advantages are: There is no need for additional adjustment; des·
and assemblies are possible to accommodate corrosive, hot, and extreme p
applications. The main disadvantage is that the packing cannot be adjusted to
a small leak for a short time until repacking can be accomplished. However,
ing failure is not catastrophic, so this is not too severe a problem. lt is the opi
in sorne circles that eventually the federal OHSA regulations will not allow
ing to be adjusted while the pump is moving. In this eventuality, the spring-loa¡
concept will be very practica!. See Figure 7.23.
7.5.5 Segmenta! Metal Packing

In an environment of hot or cold and surgically clean liquids and steam, seg
tal metal packing, Figure 7. IA, has performed with outstanding service.
The piston-ring principie is utilized to allow the liquid to force the segm
activated by a gentle garter spring, against the piston rod. The tangential-cut
(A) performs the sealing, the tangential cut allowing the segments to mov
wards the rod to compensate for rod and packing wear. The leakage gap a
segmental-cuts are sealed by a radial-cut segmental ring (B) positioned ove
above gap. Typical arrangements used for example, three pairs of rings, a a
enclosed in solid cages ali stacked on each other. lt can be seen that precl
machining is required to produce this packing and obviously adds to its cost
7.5 PACKING 219
Spring-loaded packing style 0805-4. (Courtesy Utex Industries, Inc.)
THREE
RADIAL CUTS ~
~;::Y
//c2...__,___ __,____,9
GAIHER SPRlNG A
9..._~_ ___,___,(;
B /
ÓA ??,,
91-~B'-'-----==9-'---l
R!NG A B B STACKED
Figure 7.1A. Segmenta! Metallic Packing
Multiple Material Assemblies

ea! assemblies using real rings of different materials is a common practice.
me reasons for this practice are abrasive conditions, corrosive materials, ex-
reme temperature environments, and local requirements on ieakage or emissions.
One of the most common assemblies is that of alternating rings of homoge-
ous and fabric reinforced rings. These are generally used for abrasive condi-
ns or low pressure leakage; however, multiple materials and designs can be
sembled for myriads of service conditions. They combine the strengths and
of various materials and designs into a synergy that improves perfor-
on the whole.
For instance, the use of an ali TFE set can be enhanced with the addition of
itable homogeneous or fabric reinforced rings. Obviousiy, the additional ma-
terial needs to be generally compatible with the pumpage, but it can be plac
in the set so as to perform without exposure to the ful! extent of the strea
The addition of this ring will keep the TFE pressure rings from reforming/
molding or "slugging" up and allow the rings to perform as individual
rings. The same type of situation can be used for high temperature, low t
perature, and other applications that require additional engineering to prov·
assemblies for satisfactory service.
7.5.7 Double Stack Height Seal Rings

Sea! rings and/or pressure rings in vee packing sets do not have standard heig
that are common between manufacturers in the sea! industry. There is sorne
fort into this with a ne standard for homogeneous and fabric reinforced ri
for hydraulic cylinders and the hydraulic industry. These rings and JIC dim
sions are generally very thin or short so as to provide a short assembly for t
hydraulic cylinder market.
The high pressure, industrial pump service market requires sea! rings and ;
semblies of a different design and/or dimensions. The generally accepted m
of failure is a wearing or fretting away of the pressure ring material until m
of it is gone, and the pressure ring cannot deform and ''heal'' itself. When t
happens, failure occurs; therefore, the design of a "double stack" height or "t
stack" height rings adds material bulk to the sea! ring.
This additional material must be added in such a manner as to enhance t
performance rather than hinder. Also, additional attention should be given ·
lubrication since the thicker or taller rings have more material rubbing on t
rod shaft and this must be compensated for by more lubrication. Generally
rings are recommended; however, one double stack or tall stack ring will suf
in most cases.
7.5.8 Spiral Packings

The use of braided materials cut into rings and formed around the shaft of
pump or the stem of a valve is a common practice. The features of this practí
are the ease of repacking and less inventory since the bulk material is on a s
or in a coi!. It can be used in an emergency. The disadvantages of cut rings
incorrect size due to inaccurate measurements or stretching, ragged and
matched joints, and in general dirty or poor housekeeping associated with ins
lation. Also, each joint provides a leak path for fluid or gas to escape.
The use of dieformed rings of braided materials will remedy sorne of
problems associated with cut rings. Measurements can be made correctly. Joi.
can be cut cleanly and the packing kept clean; however, the joint problem
the leakage problem through the joint still remains. For instance, in four
sets there are four joints or potential leak paths for fluid to escape.
The spiral combines the features of the ease of manufacture from brai
stock with the dieformed features and goes further in that it eliminates the jó
problem. The spiral is made from one continuous length of braided material s
7.5 PACKING 221
ed around a plug the proper ID, then cut and dieformed. The ends are properly
riented so as to provide a spiral with no excess or slight of material. It provides
n endless packing ring with no joints or Ieak paths and can be installed much
e same as one ring of packing.
Cut rings can be installed either spring Ioaded or hand adjusted. In either case
e joints must be staggered. When spring loading cut rings, it is necessary to
e the spring in a compressed manner so the cut rings can be installed into the
x. lf not, it is very difficult due to the cut rings not holding shape. The spiral
liminates this problem since it is endless and can be installed as a simple one
·ng packing set. lt can be used with either a metal coil spring or with the new
astomer spring concept.
FEATURES OF THE SPIRAL PROBLEMS OF THE SPIRAL

l . Endless construction-no leak l. Must be made in the factory.
path.
2. Ease of installation. 2. Un-spirals at times unless special
packing design.
3. Precision height and length. 3. Customer resistance due to ap-
pearance.
STYLE 242
SPIRAL UNITIZED PACKING SET
Figure 7.2A
CONVENTIONAL MULTIPLE RING SET

Figure 7.3A
7.5.9 FLAT-BAK Vee Ring
Vee ring technology, with the exception of material improvements, has remai
unchanged for years. One manufacturer has now significantly changed the seali
capabilities of the vee ring with the creation of the FLAT-BAK vee ring. T
FLAT-BAK vee ring is applicable to any area where vees are used especially~
reciprocating sealing applications.
The ordinary vee ring design with its vee shape on the face and heel side
the ring has to have sorne type of adapter that is vee shaped on one side and
on the other. The FLAT-BAK vee ring replaces this adapter with the additio
capability of sealing pressure.
Another significant improvement which results in increased seal life is t
heavy duty construction as well as the optimum anti-extrusion capability.
Flat-Bak Vee Ring is molded to twice the height of the standard vee ring. T
along with the flat backed feature allows for a significant increase in the volu
of material available for wear and extrusion resistance. The standard vee rí
with its shallow height, has a much shorter distance for an extrusion or w
zone to travel before a completed Jeak path occurs (See Fig. 1). The FLAT-B
vee ring increases this distance over two times (>2X), resulting in significan
increased sea! set life.
The flat back feature now permits the full utilization of fabric reinforcem
capabilities. The standard vee forces a flat fabric into the vee shape quite oft
causing interna! or external folds in the ring. With the FLAT-BAK vee, t
fabric plies are now in the optimum wear and extrusion resistant position in t
heel of the ring. An additional feature built into the FLAT-BAK vee as well
many other Utex vee rings is horizontal bias fabric construction (See Fig. 7.5
This creates a vee ring that has ali fabric reinforcement plies with equal fi
Jengths, allowing uniform wear performance. The horizontal bias elimina
short and long fiber zones in the vee ring.
With the FLAT-BAK vee, a single sea! can be used with greater effectiven
than multi-vee ring sets, providing Jower cost sealing and higher pressure seali
capabilities.
7.5.10 Knitted Wire Mesh
Knitted Wire Mesh has many uses such as back-up rings, scrapers, filters, a
high temperature seals. A patented process ( #4219204) whereby knitted wi
mesh is molded in conjunction with homogeneous rubber or other materials
act as an anti-extrusion device is available.
Knitted Wire Mesh is also available with fillers such as graphite, PTFE, a
aramid yarns. Standard materials for the mesh itself include stainless steel, co
per, inconel, and phosphor bronze. Other special alloys are available, and no
standard densities can be produced from ali materials.
7.5 PACKING 223
s ;t;
/ /,¡
1 : :
' ' 1
! ' .
1
11
1
11
¡ 11
l 11
111
i 11
¡
_J1
X
>c:::x
!
1
k.i-
1
¡. PATH LENGTH l<:J----"';c__-c:>!
LEAK PATH LENGTH
Figure 7 .4A.
HORIZONTAL
SHORT & LONG FIBERS BIAS FABRIC
(usual construction)
Figure 7 .5A.
7.5.11 Packing "Dos"

l. "Drip-lubrication" of the plunger on the atmospheric or "outside" side
of the gland is highly recommended.
2. A high volume flush of lubrication oil tends to cool and wash the plunger.
3. A screwed gland provides uniform loading of the packing.
4. See the récommended stuffing box trim ciearances shown.
5. Limit number of packing rings to three or four.
6. Refer to packing manufacturer for oil type for the service involved.
o o"
o o
o o " 00
o o
1 1
n o o
o o "
o o
. . . ..
o o o o
~
o:: ~
UJ UJ
o::
o o
m m ~
Figure 7 .6A.
Figure 7.7A.
7.5.12 Packing Don'ts

1. Excessive trim clearance contributes to short packing life.
2. Long stacks of many rings of packing are not requiréd for long life.
3. Bolted glands can result in "cocking" and plunger scoring. Excessiv
packing compression by inexperienced operators is sometimes dangerous.
4. Lantern rings, with drain boles on the bottom of the stuffing box, invite,
channelling of oil around the circumference of bore ·or rings by capillar
attraction. Oil may never reach bottom of plunger.
5. A seal ring on the atmospheric side of a lantern ring is not desirable be
cause it carries the entire sealing load without benefit of lubrication. (Be
cause of the distinct mechanics of plunger packing, the lubricated plunger
would be travelling away from the packing ring.)
6. So-called "force-feed" lubrication always requires an undesirable lantern ring.
7.7 CONTROllED-COMPRESSION GASKETS 225
Metal-to-metal
(a)
.,
( D¡
Figure 7.24. (a) Conventional gasket. (b) 0-ring seal.
0-RINGS
The concept of an 0-ring seal carne from the realization that in any packing or
gasket set, only a small portion of one resilient part of that set-an extremely small
theoretical ''bead'' nex.t to the clearance gap or leakage path-is effective in bridg-
that gap. See Fig. 7.24(a).
By replacing the bulk of a conventional sealing ring with a small ring of resilient
material as in Figure 7.24(b), an effective seal could be obtained. Thus the 0-
ring becomes an effective bridge.
Probably the shape of the cross section need not be circular. There are O rings
on the rnarket that are somewhat square with rounded comers.
0-rings are inherently most satisfactory as static seals in close-clearance füs
(metal-to-metal) such as flange faces where the clearance gap can be reduced to
almost zero. Close-fitting static plug-in-cylinder applications are a little less reli-
able because of the rather large clearance required for assembly.
0-rings do not serve well in dynamic conditions such as on pistons or in piston-
rod packing on high-pressure, fast-moving devices (unless well lubricated as when
handling clean oil). Multiple rings in the same groove or space on a piston, for
instance, tend to pressure-lock and result in excessive ring wear. They also do not
serve well in static cases where there are pulsating pressures or minute mechanical
motion from vibration. They tend to ''work'' or move in the groove, causing wear
on ring and metal parts. Double seals in such applications tend to pressure-lock if
the space between them is not vented.
As with any dynamic packing sets, two or more 0-rings or packing rings stacked
in the same space in a stuffing box serve no purpose in extending the total life. It
has been contended that in such an arrangement the rings are sharing the packo:ff.
They are not; the ring next to the atmosphere perfonns the complete seal, and the
others are redundant.
7. 7 CONTROLLED-COMPRESSION GASKETS
Controlled-compression gaskets are superior to 0-rings in applications where pul-

sating pressure or mechanical vibration is encountered. A controlled-compression
Controlled
(a)
..
Metal-to-metal
(b)
Figure 7 .25. Controlled compression gasket.
gasket is a square elastomeric gasket with dimensions such that it must be de-
formed (but not necessarily confined) at assembly of the parts, as shown in Figure
7.25(a) before tightening and 7.25(b) after tightening.
There must be space provided into which the elastomeric deformation can move,
allowing complete metal-to-metal contact of the parts. The "clamping" effect on
the elastomer provides an initial seal and minimizes movement from pulsations.
Any discussion on packing should stress the importance of designing all device~
with the least clearance gap practica!, in either static or dynamic applications. ·
7.8 BASIC ELASTOMERS

;:t
Following is a list of the most popular elastomers along with their recommended
usage.
NITRILE OR BUNA N (NBR). Can be used for ethylene glycol-base liquids, general
purpose, petroleum oíl and grease, water, silicone greases and oils, and dieste
base lubricants. Cannot be used for carbon tetrachloride, trichloroethylene, nit
benzene, aniline, phosphate ester hydraulic liquids, MEK, acetone, ozone, autoi
motive brake fluid.
SBR RUBBER (BUNA SOR GRS). Can be used for water, alcohols, automotive brak!.
fluid. Cannot be used for sunlight, petroleum oils, ozone.
BUTADIENE RUBBER (BR). Used for tires. Similar to natural rubber.
BUTYL RUBBER (llR) Can be used for phosphate ester-type hydraulic liquids, ME
acetone, silicone liquids and greases. Cannot be used for petroleum oils, dieste
base lubricants.
NEOPRENE RUBBER (CHLOROPRENE, CR) Can be used for freon, ammonia, hi
aniline point petroleum oils, mild acid resistance, silicate ester lubricants. Cami
be used for phosphate ester liquids, MEK, acetone.
CHLOROSULFONATED POLYETHYLENE (CSM). Can be used in acid.
ETHYLENE PROPYLENE RUBBER (EPM, EPDM). Can be used for phosphate este
7.9 ELASTOMERS IN CARBON DIOXIDE 227
base hydraulic liquids, steam (400ºF), water, silicone oil and greases, dilute acids,
dilute alkalies, MEK, acetone, alcohols, automotive break fluid. Cannot be used
for petroleum oils, diester-base lubricants.
FLUOROCARBON RUBBER (VITON, FPM). Can bé used for petroleum oils, diester-
base lubricants, silicate ester-base lubricants, silicone liquids and greases, carbon
tetrachloride, trichloroethylene, selected phosphate ester liquids, acids. Cannot be
used for MEK, acetone, amines, anhydrous ammonia, low molecular weight esters
and ethers, hot hydrofluoric or chlorosulfonic acids.
ISOPRENE RUBBER, SYNTHETIC (IR). Uses same as natural rubber.
NATURA!:- RUBBER, NATURAL POLYISOPRENE (NR). Can be used for automotive
brake fluid. Cannot be used for petroleum products.
POLYACRYLATE RUBBER (ACM). Can be used for petroleum fuel and oil, oxida-
tion, ozone, sunshine, hot oil to 350ºF, type A fluid. Cannot be used for hot water.
POL YSULFIDE RUBBER (T). Can be used for specific applications only.
POLYURETHANE RUBBER (AU, EU). Can be used for petroleum oils, hydrocarbon
fuels, oxygen, ozone. Cannot be used for acids, ketones, chlorinated hydrocar-
bons, water, humidity.
SILICONE RUBBER (SI). Can be used for low-temperature ( -175ºF) and high-tem-
perature ( 400ºF), high aniline point oils, dry heat, chlorinated biphenyls. Cannot
be used for most petroleum liquids, MEK, acetone.
FLUOROSILICONE RUBBER (FSI). Can be used for fuel systems up to 350ºF.
EPICHLOROHYDRIN RUBBER (CO, ECO). Can be used for hydrocarbon oils, fuels,
ozone, low-temperature properties.
LEATHER. Leather, probably being one of the first packing materials, should still
be on the list. It possesses very good characteristics for sorne applications. Its
resistance to extrusion must be recognized, because leather antiextrusion or
''backup'' rings for use with 0-rings in high pressure have been available for many
years. Leather <loes not swell after subjection to high-pressure gas, but the possi-
bility of chemical damage should be considered.
VITON, KAZRAZ ANO ZALAK. AU DuPont products, offer superior high temperature
performance along with other characteristics.
7 .9 ELASTOMERS IN CARBON DIO XI DE
Most elastomers tend to absorb carbon dioxide when exposed to high pressure for
a period of time, sorne more than others. Then upon exposure to lower pressure
the absorbed gas causes the elastomer to momentarily swell up to as much as 200 %
in volume, gradually returning to almost its original volume after the gas has been
deleted. Sorne elastomers will have been physically damaged by "blistering" and
"splitting" during this process. lf. an elastomer is confined in a static seal, an

resistance to such damage will result in its retum to its original integrity.
However, in dynamic sealing (piston and piston-rod packing)·, the swell due
gas absorption will tend to destroy the packing in a catastrophic manner by
chanical failure dueto the extreme friction generated by such swelling. In that c
it matters not what damage, if any, is done to the elastomer by escape of the g
alone.
In sorne dynamic applications, if sorne space is provided into which the ela
tomer can expand, mechanical damage will be minimized. Then if the material
resistant to damage by gas, it may provide fair service.
Little published data seems to be available describing the resistance of elast
mers to the nonchemical reaction of gas absorption. Certainly, experience sho
that most elastomers do not react in this manner in the air, hydrogen, or nitro
that is frequently used with them.
In the selection of an elastomer for a particular application, a simple gas-bo
test of severa! elastomer samples will show ·which material will perform best. T
selection then would rest on a compromise with minimum swell and minimu
damage.
7.10 STUFFING BOX BRASS OR TRIM
1t is important to use stuffing box bushings, lantem rings, followers, etc., m ·

of a material that is compatible both with the plunger or rod material and with
fluid being pumped. For water service, bronze has been found to be the rn.
compatible with all plunger and rod materials. For liquids other than water bro
may also be considered.
Bushing material of Ni-Resist, used in sorne special applications, is incom
ible with dry rods or plungers, resulting in extreme galling to both with sli
. contact, especially when a pump runs dry due to loss of prime.
After severa! swings ofthe pendulum from "loose" to "tight" trim-to-plun
clearance, the closer clearance of0.005-0.008 in for all trim is now recommend
Such close fits are very important, particularly in high-pressure, high-speed pu
(see Fig. 7 .26).
7.10.1 Stuffing Box Wear
Any resilient packing material-and that includes the great majority-will move ·
the stuffing box when subjected to pressure. Sorne types, of course, move less th"
others. The repeated movement, even of infinitely small magnitude, produced.
the pump pressure pulsation in the cylinder (from suction to discharge) will t
to wear out any metal stuffing box bore with which it is in contact. The so-cal
nonadjustable packing (pressure rings) is held rigidly in the stuffing box by
hard center, but both the inner and outer lips are free. There is also a tendency.
7.10 STUFFING BOX BRASS OR TRIM 229
BoreF ~:gg~;;~1
-'- 00011
Turn E .'._:002" 1
+.002"~
Bore D _ .OOO,,
Turn C
' 000"
=:002 ,, --=1
e
Concer.tricity toleran ce .002" TIR
Nominal
~__..____..¡
Basic Diameters
D
Nom. + .005
E
Nom. - .002
F
Nom.
IJ
Clearance, in
A B
Minimum .005 .002
Maximum .009 .007
Figure 7.26. Recommended stuffing box cleamnces.
accelerate wear of a stuffing box if it is not kept properly tightened. This type of
wear, exhibiting rings of wear corresponding to the of the packi.ng, is called
''washboarding.''
There is a misunderstanding as to how tight the nonadjustable packi.ng should
be maintained. This packi.ng tolerates and requires much more gland pressure than
any other packing. In seeki.ng a solution to stuffing box wear, this would be the
first thing to look into.
Another approach to the problem of stuffing box wear is the use of a reduced
number of packi.ng rings. Even though the nonadjustable packing is faidy hard, it
does respond to increased gland pressure, and the fewer the rings the greater the
"clamping" load for each ring, resulting in reduced movement.
With "filler" or "junk" rings in a reduced stack height packing, one would be
able to extend stuffing box life by altemating the position of the rings so as to
present an unwom portion of the bore.
Corrosion always more or less accelerates stuffing box wear, even in supposedly
corrosion-resistant metals. The products of corrosion usually tend to protect the
exposed metal surface, but in the case of the stuffing box the packing movement
continually removes the products of corrosion to keep a fresh surface exposed.
Then, too, the products of corrosion can be held ih the packi.ng to actas a lapping
compound for further wear.
Gland, adjustable
Throat
bushing
i---~
Follower
Plunger or piston rod
Gland, metal-to-metal
Figure 7.27. Screwed stuffing box gland.
7 .10.2 Stuffi ng Box Design Criteria

The following criteria should apply to new stuffing box designs for all pumps:
Glands: Stuffing box glands should provide uniform tightening to the entire cir-
cumference of the packing set. The undesirable bolted-gland design (the gland
being tightened by the use of two or three bolts or studs), with its propensity to
"cocking" by uneven tightening ofthe bolts, has been almost completely replaced
by the more desirable ' 'screwed'' gland with internal threads engaging the threaded
outer surface of the stuffing box (see Fig. 7.27). Such gland design allows abso-
lutely uniform tightening, but it should be recognized that these glands require
greater tightening torque than bolted glands to provide the same packing load.
Packing Cross-Section. 3/8-1/2 in. is desirable for all diameters. Do not ex-
ceed 3/4 in.
Packíng Arrangement. The complete set of packing rings should be placed in

the bottom (liquid end) of the stuffing box, with the lubricating lantem ring and/
or follower on the atmospheric side. Better still, use no lubricating lantern ring but
drip oíl on the plunger or rod on the outside. Avoid placing the lubricating lantern
ring between rings of packing; the contention that it will confine and distribute the
lubricant evenly is questionable.
Stuffing Box Trim and P!unger Dimensions. As shown in Figure 7 .26.
Throat Bushing. There is no need for a ''top hat'' or ftanged throat bushing
unless it is decided to use one to obtain "long" plunger or rod contact for flushed
7.1 O STUFFING BOX BRASS OR TRIM 231
packing or to "move" pítting of solid-ceramic plungers forward to prevent the

damaged area from entering the packing or to improve the dearance volume.
Chrome-Plated Stuffing Box Bore. Will extend the life of packing and box.
Al! Trim Parts. Should be at least as long as one-fourth plunger or rod diameter
to mi nimize "cocking" tendency.
10.3 Gland Tightening (Screwed Gland)

The following torque values should be followed in tightening all so-called non-
adjustable packing, one man on a wrench length indicated:
Packing Size Length of Approx. Torque,

(Diameter), in Wrench, in ft-lb
l~ 18 225
2!-3! 24 300
3i-4~ 36 450
4i-6~ 48 600
Nonadjustable packing should be tightened at installation, then retightened after

24 h, and then retightened at weekly intervals. Even daily tightening is highly
recommended. With the introduction of strict safety laws, it is no longer recom-
mended that packing be tightened while the pump is nmning so that advantage can
be taken of the pulsating pressure to help seat the packing.
Die-formed packing should be tightened very lightly, and then after the pump
has run severa! minutes the gland should be tightened to reduce the amount of
leakage. With this type of packing, if a small amount of leakage can be tolerated,
there will be better lubrication of the packing and less chance of buming up pack-
ing.
Tefion packing has a tendency to expand with heat. It is quite common for it to
start off satisfactorily, but with only a slight amount of overtightening a build-up
of temperature can take place if it is not watched and the gland backed off some,
if necessary, until it ''takes a set.''
10.4 Packing Lubrication

It stands to reason that any packing wm perform better if it is lubricated. Even
"nonlubricated" packings will show improved performance with lubrication. These
die-formed, compression-type packings are not usually used with lantem rings,
because of their "soft" nature, but lubrication, drip or force-fed (metered), can
be applied to the atmospheric side just in front of the gland. In fact, this method
is now considered by sorne to be superior to the lantem-ring introduction of lubri-
cant.
For water service, rock drill oíl is recommended. For hot water, steam cylin
oil is recommended.
A detailed discussion of "flushed" stuffing box packing for slurry service w
be found in Chapter 9, Section 9.2.3.
7 .10.5 Lubrication by Lantern Ring
Lubricating lantem rings located at the atmospheric end or in the center of t

stuffing box with a bottom drain hole can rob the plunger of oíl since capill
action causes the oíl to cling to the stuffing box wall or the lantem ring recess
ftow to the bottom without any oil actually reaching the plunger. The most positi
lubrication method is to drip the oil directly on the plunger on the atmosphe ·
side of the stuffing box.
7.11 JACOBY LEAKAGE
Jacoby leakage is described as that characteristic tendency of resilient packings

leak slightly at the reversa! of stroke on a reciprocating pump or static seals wh.
there is any slight movement between the sealed elements generated by change
pressure in the system. A pressure-seal, lip, or 0-ring packing or piston mus~
involved. (There is sorne evidence of such leakage even with "slug" or ho
geneous packing.)
Such a packing ring seals by virtue of the resilience or compliance of the el
tomer or material allowing it to be forced (extruded) into the clearance gap exis ·
between the parts in question, for instance, the extrusion gap between the pist
and the liner. This action drastically deforms the original shape of the ring or s
Once a seal is affected, the only useful part of the packing is that small seg
damming the gap against the escape of liquid. The entire remaining parts of·
packing set, including a major portion of the actual sealing ring and success
multiple rings or lips, become hydraulically balanced, and liquid will appe
the space around the piston rubber or packing rings up to the point of seal at,
"heel" damming the gap.
Upon reversa! of the piston or plunger or rod (or upon a lowering of syst
pressure), the clearance-gap seal is momentarily broken by the action of the
duction in friction and/or the relaxatíon of the deformed packing to its orig~.
shape. The action permits momentary leakage of the liquid trapped around 1
packing ring to be followed shortly by a resealing action from the hydraulic p
sure until another reversa! or pressure change is encountered.
Figure 7 .28 shows the sequence of events at the reversa! of a piston stroke,
Figure 7 .29 a piston rod stroke. At A, the relaxed piston rubber maintains
original shape. At B, the piston and the rod packing are distorted by the H
pressure. When the end of the stroke is reached at C, the distorted piston or p
ing retums to its normal relaxed shape but the liquid trapped around the rub
.displaced to the atmosphere end.
7.11 JACOBY LEAKAGE 233
A B
e
Figure 7 .28. Jacoby leakage in pistons.
LJmir 111111 I
EQUIVALENT POSITION TO B , FIG. 7.28 EQUIVALENT POSITON TO C , FIG. 7.28
t?11 )11111
\
Figure 7.29. Jacoby leakage in piston rod packing.
A clue to the early recognition of Jacoby leakage is found in an old 0-ring

manual, which st11tes: "A single 0-ring in a groove is used for both single- and
double-acting pumps. Two or more of 0-rings in the same groove (or in separate
grooves on the same piston) are not recommended. There is danger of locked
pressure between the 0-rings unless the space between them is vented." This
statement implies that there must be sorne leakage past the first ring in order for
the described phenomenon to exist (see Fig. 7.30c).
1t is evident that the amount of leakage involved is dependent upon several
factors such as seal resiliency and the effect of local temperature on that resiliency.
Three early experiments, (Fig. 7.30a, b; ande), clearly demonstrated the ex-
istence of the Jacoby leakage effect.
Figure 7 .30a shows a test setup designed to apply a high pressure to experi-
mental test pistons. A standard off-the-shelfpiston was used on the bottom. During
the test, making use of an available steam-heated rubber molding press, the pres-
Press ram Gauge
Gauge
~ Plunger
Opposed
packing
rings
/ /'.
Steam-heated
press platen
a. PISTON SEAL TEST b. EXPERIMENTAL PACKING
Figure 7 .30. Jacoby leakage examples. (a) Piston sea! test. (b) Experimental packing.
Cylinder head.
sure was raised in !000-PSI increments to I0,000 PSI, where it held at eac11.
incremenL Upon releasing the pressure slightly, a puff of steam would appear as
shown, indicating liquid leakage past the piston. At each pressure lowering, the
puff of steam would appear for an instant and then the system would set for any
length of time at an absolutely perfect sea!. After checking ali parts for integrity;,
it was possible to repeat this at will. Not much thought was given to the unusual,
action until sorne time later when an experimental packing was tested (Fig.
7.30b). The opposed packing rings were supposed to expend ali of their extende
life sealing off clean lubricated liquid instead of dirty unlubricated liquids. The;
test was supposed to be run at 2500 PSI, so the oil was pressured to 3000 PSI.
The packing held the pressure perfectiy. But when the pump was started the oiL
pressure immediately fell to the pump pressure. Repeated efforts to make it worl(
met with failure-even when a continuous supply of high-pressure oil was used:.
The results were valuable in a study of an extraordinary problem related to the'
original piston test
In the third experiment, Figure 7 .24c, a two-0-ring cylinder head seal that was.
found to be difficult to remove because of "stored" pump peak pressure betwee!l>
the rings tumed out to be crowning proof of the existence of Jacoby leakage. ...
7.12 GLOSSARY OF TERMS

SPRA YWELDING. Application of a hard facing layer to steel to increase resistance
to wear and corrosion.
In the spraywelding process, a special type of alloy in extremely fine powdet·
7.12 GLOSSARY OF TERMS 235
torro is sprayed through an oxyacetylene flame, impinging on a clean work pi.ece

sticking by a mechanical interlocking of the semiplastic metal particles. to
point, it is exactly like wire metallizing. In a second step, the sprayed deposit
together with another torch-heating operation to form a solid
wear-resistant coating that is bonded rrietallurgically to the base material.
These special materials have a happy combination of properties. They have
wear and corrosion resistance because of the alloys and rnicrostmctures
xm~s(;'ll•· They have a matrix of soft nickel containing a variety of very hard par-
(up to 80 R) including chromium bmides, chromium carbides, anda mixture
complex cross-combinations. The result is a very low coefficient of friction and
vA\NUVU• corrosion resistance. The soft matrix allows sorne ' for the hard
u"'""'v~ to fioat on a microscopic scale. Because of the low temperature
about 1900ºF), a variety ofthese base materials can be used. The macro hardness
the deposit is uniform throughout.
, One of the spraywelding materials contains particles of tungsten carbide in the
soft matrix, imparting an effective extreme hardness. This material should be
;'avoided for plunger coating for the reason that the "scrubbing" action of an elas-
~Jomeric packing will remove sorne of the soft matrix, leaving a sandpaper finish
·:.that is in tum extremely damaging to subsequent packing. While the surface just
described may appear smooth, a simple test will reveal such microscopic "sand-
paper" surfaces. Just stroke the edge of a copper coin (penny) along the surface;
.the "rough" surface will generate a visible streak of bright copper color.
CQUENCHING. Obtaining higher strength in steel by heating above the transfonna-

range (above 1450ºF) and cooling rapidly in a liquid bath. (Exception:
~Quenching an austenitic stainless steel results in annealing or sofiening.)
c.TEMPERING. Heating hardened steel to an íntennediate temperature ( 800 to

1200ºF) to decrease hardness and increase toughness.
PRECIPITATION HARDENING. Increasing the strength of a group of special alloys
heating at intem1ediate temperatures (800-llOOºF) and allowing a microcon-
·tuent to precipitate from a metallic solid solution.
FLAME HARDENING. Increasing the hardness and wear resistance of a medium to
high carbon steel (with or without alloy) in a localized area by applying flame
'heating to the steel smface and quenching.
INDUCTION HAFIDENING. Increasing the hardness and wear resistance of a medium
carbon steel (with or without alloy) in a locaiized area by generating heat
electrical induction in the steel surface and quenching.
CARBURIZING. Making low-carbon steel more wear-resistant by increasing the sur-
face carbon content through di:lfusion from a carbonaceous medium and subse-
quenching.
· CHROME PLATING. Increasing resistance to wear and coITosion by depositing hard
chromium on a surface in an electroplate bath. Hardnesses up to 70 R (750BHN)
can be obtained.
NORMALIZING. Strengthening steel by heating above the transformation range

(above 1450ºF) and then air cooling.
PLASMA COATING. Increasing resistance to wear and corrosion by applying a sur-'
face layer of hard facing materials, utilizíng the plasma are principle. A plasma-
forming gas is passed through a de are in a small nozi:le and is superheated, its.
molecules breaking down into ionized atoms with a high energy content. Working
temperatures of 10,000-40,000ºF are obtained, and gases such as argon, helium,
or nitrogen-hydrogen or argon-hydrogen mixtures. Materials to be deposited are<
in powder form and are introduced into the are, heated, and blown onto the work
piece, where they melt on the surface and form a solid bond. Since the are
not contact the work piece, it remains relatively cool.
7 .13 GUIDELINES FOR MATERIAL SELECTION
Piston Pump Part Material

Piston Rods Quenched and tempered 4150 alloy
steel, induction hardened.
Same, with 0.020-0.030-in hard ehro-
mium plate on surfaee.
Carbon steel with spraywelded hard
nickel alloy on surface.
Pisto ns
Body Forged, quenched and tempered 4140
alloy steel.
Snap ring Shot-peened spring steel.
Retainer ring Medium earbon steel.
Support rings Medium carbon steel.
Rubber Natural rubber.
Liners Chrome-plated.
27 % chrome-iron lining in carbon or
low allow steel jacket.
Flame-hardened medium carbon or low-
alloy 8620 steel.
Valves and Seats
Valve Quenched and tempered 4150 alloy steel
forging, flame hardened on wear sur-
faces.
Seat Same as valve.
Nut Normalized and shot-peened medium
carbon steel forging.
Insert Urethane (Buna Nin sorne sizes).
Spring Oil-tempered spring steel or titanium.
7.13 GUIDELINES FOR MATERIAL SELECTION 237
Plunger Pump Parts Material

Plungers Spraywelded nickel-base alloy on low-
carbon steel. (Special-same alloy on
304 stainless steel.)
Solid 85 % alumina with 304 stainless
steel shank.
Sprayed ceramic coating on low carbon
or 304 stainless steel, all with 304
stainless steel shank'.
Valves and Seats
Disc-type Seat-cast 316 stainless steel.
Guard-same.
Disc-Delrin, titanium, 316 stainless
steel.
Guide bushing-316 stainless steel,
nickel aluminum bronze, 316 stain-
less steel with delrin sleeves.
Spring-Titanium, 17-7 precipitation
hardened stainless steel Inconel.
Capscrew-K-Monel.
Externally mounted Valve-17-4 precipitation hardened
· stainless steel wing-guarded.
Seat-17-4 precipitation hardened stain-
less steel.
Spring-17-7 precipitation hardened
stainless steel.
Wing-guided Valve-410 stainless steel, n1ckel-alu-
minum bronze casting.
Seat-410 stainless steel, nickel-alumi-
num bronze casting.
Spring-17-7 precipitation hardened
stainless steel.
Wing-guided slush Valve-410 stainless steel, carburized
type 8620 alloy steel.
Seat-410 stainless steel, carburized
8620 alloy steel.
Nut-medium carbon steel.
Insert-Urethane.
Table 7 .6 gives the chemical analysis of many of the popular metals used in pump
construction.
TABLE 7.6 Nominal Chemic:al Analyses, percent by welght"
Material e Mn Si Cr Ni Mo Cu Other %
N
c.>
ti>
----
Low carbon steel 0.20 0.45 0.25
Medium carbon steel 0.45 0.75 0.25
8620 alloy steel 0.20 0.80 0.25 0.50 0.55 0.20
4140 aHoy steel 0.40 0.88 0.25 0.95 0.20
4150 alloy steel 0.50 0.88 0.25 0.95 0.20
Sp1ing steel 0.65 LOO 0.20
27% Cr-iron 2.30 0.60 0.50 27
410 stainless steel 0.10 0.70 0.60 12.5
304 stainless steel 0.06 0.70 0.60 19 10
316 stainless steel 0.05 0.70 0.60 19.5 10.5 2.5
17-4 precipitation hardened 0.04 0.30 0.50 16 4 2.7
stainless steel
17-7 precipitation hardened 0.05 0.30 0.50 17 7 Al 1
stainless steel 0.65 4 15 13 Fe 4
Sprayweld
B 3.5
Ni-Al bronze 1.0 5 81 Al 10
Fe 5
Ni-Al bronze casting 5 79 Al 10
Fe 5
Titanium 0.08 Ti 99
Fe 0.25
Inconel 0.15 1.0 0.5 16 72 0.5 Fe 8
K-Monel 0.15 0.60 65 29.5 Fe 1.0
Al 2.8
Ti 0.5
gD Aluminum Bronze 4 80 Al ll
Fe 4
ªOther common materials: Urethane-polyurethane elastomer.

'1
7.14 TITANIUM AS A PUMP MATERIAL 239
7.14 TITANIUM AS A PUMP MATERIAL
Titanium possesses the most desirable characteristics for use in pump liquid-ends
and parts handling waters of ali sorts, including sea water and oil-field brine.
Commercially Pure Titanium (RMI-70) has excellent corrosion resistance and has
been used for severa! years giving excellent service.
This extreme resistance to corrosion results in reduced liquid-end failures
caused by corrosion-fatigue at the highly stressed region at the intersection of
bores where a stress-concentration of 2. 75 may be had. In recent years, titanium
has become more plentiful and it is well within the acceptable cost-range for
pumps in special service.
Titanium must not be used in strong reducing acids or chlorine.
8
VALVES
8.1 VALVE TYPE CLASSIFICATION*
Valve and seat configurations can be classified as to type as shown in Figure 8 .1.
It is obvious that there can be numerous combinations of most of these features.
Probably the main division of greatest difference is between the self-contained
( Q, R) and the in situ (H-P). Configurations of the self-contained type have the
advantage of dimensional and assembled accuracy, but they are confined to the
smaller pump sizes, about 100 BHP and smaller. Large ph)'.sical size of valves
limits the ability to use the self-contained principie.
In situ valves require guiding and spring support to be a part of the liquid end.
Self-contained valves can be reclassified into center-guided ( Q) versus cage-
guided ( R). Center-guided valves seem to be preferred because of their simplicity
and ease of assembly. Caged valves are not recommended for abrasive liquids.
The next logical division would be the metal-to-metal (A) versus the rubber
seal or "slush" type (B). As the name implies, the rubber seal valve is mandatory
for most liquids containing even small amounts of solid materials and slurry.
The matter of valve-guiding methods is very broad in that all five (S, T, U, V,
W) and their combinations are widely used. Wing-guided valves (U) are not usu-
ally accepted for slurry service because there is rapid wear between the ribs and
seat and there is a tendency for solid particles to interfere with the guide motion.
The most popular method of mounting the valve in the deck is the taper fit, but
the sandwich or ftange method is widely used. The screwed method is no longer
used in high-pressure pumps.
*Letters in parentheses refer to keyed drawings in Figure 8.1.
240
8.2 PUMP VALVES 241
Valve Types Seat Types
A \ Metal
Bevel
2i
B (
Rubber seal
Metal
Slush
ria H
C 1 METAL ! !PLASTIC 1 Three-rib Four-rib
Discorflat
J K
Double-port
Full-flow Suspended-guida
Bonded Disc Valve for "Dirty" Liquids
' , 1 1 1 1
1--írr¡-¡ \ 1 1 1
L 1
1
1 I 11
11 l 1
1
1
o 'l...,) 1
1
1
1
t
Ball Bevel-rib Flat-rib
' 't:==~:;::;=,.:---1 / t----1 : : r--1

1 1 l_J 1 1
M 1 1I 1 1 1 p 1 1 I 1
1 11 11 1 1 ~ / 1
Bevel recessed-rib Flat recessed-rib
GW?
Spherical
_._
N \'¡--------¡}
Durable disc Full-flow
Figure 8.1. Valve classification.
8.2 PUMP VALVES
A liquid valve is a rather simple device in that it is a freely movable plug that is
forced open when there is liquid pressure under it and is forced to close and seal
when there is pressure on top of it. The simplest valve is the leather ftapper used
in the pitcher pump and in air bellows for centuries. But with the advent of high-
pressure, high-speed pumps and the presence of abrasive solids, even in small
amounts, simplicity has given way to extreme sophistication.
242 VALVES
Self-Contained Valves
Q R
Center-hole guided Caged
Valve Guiding Methods
J~L
~
s
GAUGE-LINE LIMITS
t 1·. ~
SHCULDER ; , Ef"FECTIVE THRDUGH ,
Top stem
T CPTI CNAL " " ' ..,: ....,..,.,;~AR__;;E.....
A _--+i,D;:..;;l.;.;;.AM;::.ETc::;ER"-i.~..,·
T A
1
TAPER
rn
Bottom stem
B STRAIGHT
1
1
1 1
V
Wing-guided
1
e : : .
THREADED
1 1.
-
Center-hole
w
Cage-guided
Figure 8.1. (Continued) Figure 8.lA. Valve Seat Retention Methods
Two types of high-pressure valve-sealing methods are used (see Fig. 8.1): (A)
metal-to-metal seals for clear liquids and (B) elastomeric seals for dirty liquids.
(For lack of better terminology, the latter is referred to as the "slush"-type.) The
elastomeric seal performs the function of providing a '' dam' ' between the metal
parts of the valve and seat even though these parts never come into intimate con-
tact, being held apart by the solid particles in the liquid. The elastomer must be
resilient enough to deform over a solid particle so that the valve will not be held
open by that particle. However, it must not be so resilient as to allow excessive
8.2 PUMP VALVES 243
extrusion into the gap at the valve and seat contact or to allow "puncturing" by
the solid particles. Unfortunately, most elastomers have low strength at maximum
resilience and vice versa.
Many val ves incorporate the rib-supportéd design ( H, l) for the following ob-
vious reasons. Consider the use in abrasive liquid service of a non-rib-supported
slush type (K) seat with a bevel sealing face where both the valve metal and seal
contact the seat bevel. Since the guidance of such a valve, either stem or wing ( S,
T, U), rnust have sorne clearance, the valve can close on each stroke in a slightly
"cocked" position. When the full hydraulic load is imposed on top of the valve,
the valve tends to seek its true center, and a small amount of movement on each
stroke grinds solids between the metal faces, resulting in rapid wear. If flat ribs
are úsed in the valve seat upon which the valve rests, it always lands on a flat
surface and cannot shift when the hydraulic load is imposed. A sharp bevel angle
((J) of 55º induces better centering as the valve doses and also reduces ftow tur-
bulence since the flow path is closer to a straight line.
All rib-supported valves should use the superior three-rib design (H). Since
the entire hydraulic load must be supported by the ribs, the three-rib concept ab-
solutely allows equal loading on eac;h rib. Since the three equally spaced ribs al-
ways carry equal shares of the load, regardless of solids trapped on the ribs or
dimensional discrepancies due to machining and wear, no one rib is ever over-
loaded. This also means that the valve body is always uniformly loaded between
the 120º segments.
Now in the case of four ribs, solids on one or more of the ribs or dimensional
discrepancies in the parts could cause support on only two ribs, thereby overload-
ing both the ribs and the val ve body.
With a wing-guided valve (U) or with a non-rib-supported stem-guided seat
( K), flow-restricting metal is placed in the through-area path and serves no other
purpose than to guide the valve. When the same amount of metal is used in ribs
that support the valve, a stronger and longer-lasting valve results without sacrifice
of through area. Several types of mud or slurry valves are shown in Fig. 8.2.
The failure mode of any elastomeric seal valve follows the pattern of pinch-off
of the elastomeric seal at the thin circumferential area between the seal and the
metal. Since the elastomer is in effect a liquid, the pressure above is transmitted
through the seal to the point where it "wants" to be extruded or squeezed past the
gap, if any, between the metal valve and seat. The minute pinching off progresses
until a rather large void in the elastomer occurs. Fluid trapped in this void is then
subjected to high pressure from above and low pressure from below. The liquid is
suddenly squirted out through the relatively close metal-to-metal contact, resulting
in "hair-line" or "wonn-eating" washing of metal parts. Corrosive drilling fluid
(salt mud, etc.) accelerates the advancement of "worm eating" and reduces valve
life due to the more rapid loss of metal. This failure mechanism is regenerating in
that rapid failure and complete washout of the val ve and seat soon follow. There-
fore, the longer one can delay the above-described pinch-off by replacing the seal,
the longer valve life can be expected.
Rapid pinch-off results from:
244 VALVES
Figure 8.2. Mud and slurry type valves. (Courtesy Fluid King.)
1. Use of a low tensile strength elastomer.

2. Reduced tensile strength of any elastomer in hot liquids.
3. Presence of large particles or lost circulation material, the material lodging
under the elastomer seal, allowing trapped liquid to be squirted out in much
the same manner as a pinched-o:ff seaL
4. Use of a new seal on a wom valve, which allows excessive seal overhang
and results in the seal bending upward, again allowing trapped fluid to be
squirted out. Not only does this cause hairline washing, but it also results
in rapid seal wear because the elastomer at the sealing point is placed in
extremely high tensile stress even before the full hydraulic load is imposed.
5. Chemical reaction to the seal, which causes it to swell, resulting in excessive
overhang anda reduction in tensile strength.
These points are discussed in tum below.
1. The inherent high tensile strength of polyurethane (about two times that of
Buna N) and its chemcial resistance, particularly in oil, make it the most desirable
seal available.
2. High temperature is an enemy of all elastomers, particularly of polyure-
thanes. The ·effect of temperature on any elastomer is such that it is misleading to
apply a fixed limitation to them. For instance, a generally accepted temperature
limitation for polyurethane has been placed at 140ºF by the industry, but since
elastomers undergo a gradual reduction in physical characteristics, particularly ten-
sile strength, as temperature is increased above room temperature, it is reasonable
8.3 VALVE SEATS 245
to assume that even polyurethane can work ata temperature above 140ºF if other
factors such as lower pressure are favorable. There are many cases where poly-
urethane seals perform satisfactorily above this temperature. Buna N has the de-
sirable characteristic that it retains its physical strength ·over a higher temperature
range--to such an extent that Buna N seals may better at high tempera-
tures.
Remember that all elastomers are subject to deterioration, and
shorter life must be expected at higher temperatures.
3. Solid abrasive particles are a necessary evil in slurry pumping.
4. The obvious remedy is not to try to use a badly worn valve.
5. Chemical reaction becomes a problem. No doubt certain muds and slurries
contain chemicals that are detrimental to either polyurethane or Buna N, and these
are sometimes difficult or impossible to track down. It behooves one to try one or
the other seal when a chemícal reaction is suspected. As previously stated, poly-
urethane seals will stand up in most slurries," and are particularly suited for
oil or oíl-base muds. Buna N, although called "oil resistant," is not completely
so. AH seals of Buna N swell and deteriorate in any oil and are not recommended
for oils with an analine lower th.:.n 150ºF. (The lower the analine point, the
greater the content of aromatics, which are hard on Buna N).
Experience and extensive testing have shown that the desirable amount of lift
of a valve should be such as to provide a lift area equal to the through area of the
valve. Lift in excess of this is not normally harmtul, since the valve stiil will only
lift the required amount. However, a stop is desirable at this point since erratic
pump operation due to air, gas, or hot mud causes abnormal valve action with
pounding and bouncing. It should be stated that in normal operation a valve does
not slam shut; it follows a smooth, gradual opening related to the displacement of
the piston, which is inherently a slightly distorted sine wave.
The velocity of the abrasive liquid through a valve should be limited to avoid
erosional effects on the valve parts. For most abrasive materials, including drilling
mud, a limit of 12 FPS for uncharged suctions and 16 FPS for charged suctions is
generally recommended. Slurry valves should be limited to a maximum of 12 FPS.
In investigation of a pump valve lift and liquid velocity, keep in mind that the
recommended limits are based on maximum liner or plunger diameter and maxi-
mum operating speed.
In order to minimize excessive "dash-pot" effect in retarding the valve mo-
tion, resulting in delay of closing, the pot area should be at least 1.5 times the
valve disc area to provide ample "escape" area.
8.3 VALVE SEATS
Table 8. l shows recommended valve seat taper diameter or straight-bore diameter

(gauge-line) for an extended number of valve sizes the present API Range
(Table 8.1).
¡
TABLE 8.1 Valve Seat Taper Dimensions (See Chapter 7, API Standards) (Proposed expansion)
Size Gauge Line Approx. Taper GPM/Valve Approx. Through "Size" Number
Depth, in Depth, IN. At IO FPS Area, sq. in. Proposed
-15 0.45 0.27 1.4 0.09 1
-14 0.5 0.29 1.7 O.JI 2
-13 0.57 0.32 2.2 0.14 3
-12 0.64 0.35 2.8 0.18 4
-11 0.72 0.38 3.6 0.23 5
-IO 0.81 0.41 4.4 0.28 6
-9 0.92 0.45 5.5 0.35 7
-8 1.03 0.49 6.9 0.44 8
-7 1.16 0.53 8.6 0.55 9
-6 1.3 0.58 I0.7 0.69 10
-5 1.47 0.63 13.4 0.86 1
-4 1.66 0.69 17 1.09 12
-3 1.87 0.75 21.2 1.36 13
-2 2.1 0.81 26.5 1.7 14
-1 2.37 0.88 33.2 2.13 15
o 2.75 0.96 41.6 2.67 16
2.875 1 51 3.3 17
API 1 3.375 1.125 65 4.2 18
API 2 3.875 1.25 81 5.2 19
API 3 4.375 1.375 I03 6.6 20
API 4 5 1.5 128 8.2 21
API 5 5.625 1.625 160 I0.3 22
API 6 6.25 1.75 202 13 23
API 7 7 2 249 16 24
API 8 7.75 2.25 312 20 25
API 9 8.5 2.5 389 25 26
API 10 9.5 2.875 467 30 27
API 11 12 11 3.25 623 40 28
13 12.5 3.75 779 50 29
14 14.125 4.25 935 60 30
15 15.875 4.75 1246 80 31
16 17.875 5.25 1558 100 32
17 20.125 5.875 1869 120 33
18 22.75 6.5 2336 150 34
19 25.5 7.25 2960 190 35
20 28.75 8.25 3583 230 36
- - -,-~,·= . U a. "2&$,,L,,~,,.,"iiiii __ ~w-

248 VALVES
Aside from clamping a "ftanged" valve seat into the deck with a washer or
gasket under the ftange (for pressure), there have been three other fairly popular
methods of securing the seats into the deck: Taper fit (L), ftanged or "sandwich"
fit (X), and screwed fit (Y) (Fig. 8. IA). Table 8.1 shows taper dimensions and
Figure 8.3 shows the terminology applied the valve and seat dimension. Figure
8.4 gives details of the design of taper fits for valve seats.
Tapered valve seats should be provided with a shoulder to help prevent com-
plete pump-through, which would result in damage to the pump valve deck bore
if the pump were overpressured by accident.
The installation of new valve seats should be preceded by a thorough cleaning
of the deck bore with emery cloth and wiping dry and free of oil or grease. The
seat should have any protective coating removed with solvent and wiped clean and
dry. A preliminary check can be made by dropping the seat into the deck bore
with a slight ''thunk. '' If the fit is proper and clean, the seat should be difficult to
break loose by hand.
With the seat in place, install a valve. Then place a pipe over the stem of the
valve and give one or two solid blows with a sledgehammer. With a firmly seated
seat, the pipe should ring with the characteristic sound.
The widely used taper-fit valve seat seems to be the most practica! for high-
pressure service. However, such seats are sometimes difficult to pull. A hydraulic
valve seat puller using a jack of 150 tons capacity and 2i-in diameter puller bolt
is required. In stubbom cases it is sometimes helpful to break a seat loose by
sledgehammering the top of the puller bolt (a driving face is provided) or by even
sledgehammering the side of the valve pot after the jack has been pressurized.
Should it ever become necessary to remove a seat with an acetylene torch,
follow this procedure: Cut two or three of the ribs completely free. Heat one spot
ofthe seat to a dull red (do not flame cut) while frequently playing the torch around
the entire seat to heat it. Then quench the heated seat rapidly with a stream of cold
water. This procedure reduces the danger of pump damage from cutting. Heating
one spot dull red places the metal in a plastic state. Heating the rest of the metal
forces it to expand, and since it is confined in the deck bore the expansion is forced
into the plastic region. Upon cooling, the seat diameter becomes smaller than the
original, and it can usually be bumped out with a hook or pipe wrench.
8.4 VALVE SPRINGS
A generally accepted rule for approximate valve spring load is:
2 POSIVA for ''poor'' suction conditions such as lift from a dug pit and for
low speed.
4 POSIVA for "normal" suction conditions, such as from an elevated tank.
6 POSIVA for ''charged'' suction conditions of 20 to 40 PSI.
'-----Da----Aa----
i.E----¡j v-----.1
i.E----Av-----+i
i.E----D.-----.i
1
\ TPF on diameter /
~-----Dg Ag _ _ _ ___.
Valve dimensions
Figure 8.3. Valve dimensions. Symbols are as follows:
ex = angle of taper, one side, deg FL = force on valve, maximum lift, lb

(:J = angle of coefficient of friction, deg P = pressure on valve, PSI
Aa = disc or valve area, in2 L = valve lift, in
Ag = deck-bore area at gauge line, in2 LL = spring space, maximum lift, in
AL = valve lift area, in2 L¡ = spring length, installed, in
AL = Area of "belt": L¡ = spring, free length, in
AL = 1íDv (or D.) X L Na = spring, number of active coils
Av = valve through area, in2 N, = spring, total number of coils
Apo1 = valve pot area at Dpoo in2 (=Na + 2), closed end
Da = disc or valve diameter, in Q = force on deck bore due to seat
De = seat opening, effective diameter, in .drive, lb
Dg = gauge-line diameter at bóre, in s¡ = spring, installed stress, PSI
D, = spring mean diameter, in sL = spring, lift stress, PSI
Dv. = seat opening diameter, in TPF = taper per foot on diameter, in/ ft
d = spring wire diameter, in W= Wahl factor
F = force on valve, lb x = valve seat depth, in
F; = force on valve, installed spring, lb (} = bevel angle, deg
A,,., = 1.5 X Ad
249
Figure 8.4. Taper valve seat bore fit.
f = coefficient of friction
TPF = taper per foot, on diameter, in/ft
D 8 = gauge-line diameter, in
D,, = effective through diameter, in
F = force on seat, lb
P = pressure on seat, PSI
s = allowable stress, PSI
Q = force on seat bore, lb
Qc = force per circumferential inch, lb
X= seat depth, in
a = angle of taper, one side, deg
(3 = angle of coefficient of friction, deg
Rule: For seat to be self-locking in bore, a < (3.
Sample Calculation: For D 8 = 7 in, P = 2000 PSI, f = 0.1, s = 75,000 PSI, and TPF
= 2 in.
a= arctan TPF/24 = arctan 2/24 = 4.76º (8.1)

(3 = arctanf= arctan0.1=5.71º (8.2)
F = 0.7854D~P = 0.7854 X 72 X 2000 = 77,000 lb (8.3)
Q = (F/2) tan (a+ (3) = (77,000/2) tan 10.47º = 208,000 lb (8.4)

Qc = 2Q/ 7r = 2 X 208,000 /3.1416 = 132,600 PSI ( 8.5)
X= Qc/Zs = 132,600/2 X 75,000 = 0.9 in (8.6)
Note: API Standard TPF is 2 in/ft; API Standard for X when Dg = 7 in is 1.25 in. [For-
mulas do not include requirement for set strength to support the forces to which it is sub-
jected.]
250
8.5 VALVE SPRING DESIGN 251
The greater the pump speed, the heavier the spring required for maximum ef-
ficiency and smooth operation. Accordingly, a higher suction pressure is required.
8.5 VALVE SPRING DESIGN

Formulas
POSIV A = F;/ Av (8.7)
De = .JAv/0.1854 (8.8)
AL (See valve dimensions, Fig. 8.3)
d = ;¡8kD;Na/G (8.9)
f = 8FD;Na/Gd 4 (8.10)
Kc = 1 + (0.615d/D,) (8.11)
s = (2.55FD,/d 3 )Kc ( 8.12)
k = Gd 4 /8D;Nª ( 8.13)
Na = fGd 4 /8FD; (8.14)

where
Av = valve through area, in2

AL = valve lift area, in2
d = spring wire diameter, in
De =effective valve seat diameter, in
D, =mean diameter spring, in
f = spring deftection, in
F = force, lb
F; = installed force, lb
FL = force at full lift, lb
G = spring torsion modulus, lb /in2
k = spring rate, lb/ in
Kc = Wahl factor (Figure 8.5)
L¡ = free length, in
L; = installed length, in
LL = spring space length, in
Na = number of active coils
N 1 = Total coils = Nª + 2 (closed & ground)
s = stress, PSI
252 VALVES
2.0
1.8
\
::.::
...
o 1.6
\
tí
"'
LL
1.4
\
1.2
~ ¡-..,..._
-¡-._
- -
1.0
2 4 6 8 10 12
Ratio: Mean diameter
Wire diameter
Figure 8.5. Total valve spring stress according to Wahl formula. K x S = total stress.
See Figure 8. 6 for required val ve lift to pro vide lift area of 100 % of through area.
Example * (iterative solution)
Av = area, through valve = 1.3 in2

Ds = mean spring diameter = 1.9 in
k spring rate = Try 30 lb/ in
=
LL = length, spring space = 1.0 in
Nª = number of active coils-Try 4
G = torsional modulus, spring wire = 11,500,000 lb/in2
POSIV A (pounds per square inch valve area) = 4
l. Determine F;, installed force, lb:
F; = POSIVA X Av =4 X 1.3 = 5.2 lb (8.18)
2. Determine De, effective diameter from valve through area·(Av):
De= ,JAv/0.7854 = ,Jl.3/0.7854 = 1.3 in (8.8)
*See Figure 8.6.

8.5 VALVE SPRING DESIGN 253
r-:nd---i
,..-,1
f-- Dd__,,.., I \ ~Dd----4
1 \ r------,
- - r-------
-~
- - L _ _ _ _ _ _J
"....__ - - -
L
-t11 ~ ILJ
__!:_ ,_.,,,
-'-OIID
~Dv~ ~Dv-1 f--nu-j
~
(a)
@ @ BEVEL SEAL
WING OR STEM
(b) BALL
SPHERICAL
(e) DISC
OR PLATE
8.6. Required valve lift for 100% of valve through area (shaded area). (a) Bevel
seal, wing, or stem
L, in = A,,/1r cose (ni + cos f) sin IJ) 15)
(b) Ball or spherical
L, in = A,,/0.707 1íDv (8.16)
(e) Disc or plate
L, in= A,,/1íD,, (8.17)
where L = lift, in; Dv = seat opening diameter, in; = seat effective diameter, in; A,,
= valve through area, in2 ; e = seat bevel angle, deg.
3. Determine lift to provide lift area (AL) of 100% of Av:
L=Av/r.De = 1.3/1.31!' = 0.32in (8.19)
4. Determine d, spring wire diameter, in:

4 . 4~---~------~
d = :;/8kD;Na/G = :;/g X 30 X l.9 3 X 4/11,500,000
=0.155in (8.9)
*Or D ..
254 VALVES
5. Determine L1 , free length of spring, in:
L¡ = L¿ + L + (F;/k) = 1 + 0.32 + (5.2/30) = 1.49 in (8.20)

6. Check for spring space:
4 active coils + 2 (closed end) = 6

6 x d =6 X 0.155 = 0.93 [OK, less than LL (1.0 in)]
7. Check stress at F;:
S; = 2.55KJ;D./d 3
= 2.55 X l* X 5.2 X 1.9/0.155 3 = 6765 PSI (8.21)
8. Check stress at FL:
F¿ = (L¡ - L¿) k = ( 1.49 - 1.0) 30 = 14.7 lb

S¿ = 2.55 X l* X 14.7 X 1.9/0.1553 = 16,106 PSI (8.22)
Val ve spring life can be improved by taking advantage of the Goodman effect
by designing within the stress ranges shown in Figure 8. 7
8.6 THE EFFECT OF VALVE DESIGN ON SUCTION

REQUIREMENTS
Section 8.6 reports on an extensive investigation of the valve performance in a

six-inch stroke Triplex Plunger Pump.
The choice of liquid valves for a plunger pump is dependent upon many fac-
tors, the most important of which are
Pressure
Liquid temperature
Suction pressure
Liquid corrosion
Pump speed
The ideal valve would be weightless, springless, corrosion and temperature re-
sistant, andas large as possible (to provide maximum through area).
A brief look into these requirements will immediately show the tug-of-war that
exists in valve design. One cannot have lightness with strength or large size.
*Shot-peen spring and use Wahl factor of l.

8.6 THE EFFECT OF VALVE DESIGN ON SUCTION REQUIREMENTS 255
1201--~--J.--~--+~~-+-~~-+~..........,,.¡..r::--,r-¡
Anea of additional stress
possible dueto shot·peening
~"' soi--.c..~r--~~r--~~r--~.,,._+--~~+-~--1
ro
"'::>o
..;:
~ 601---~~+--~~+--~-r-+--~~+-~~+-~--!
~
1ií
o;
e:
~ 40,..._~~i--~-----~~+--~~--~~+--~·--<
ASTM A230
ASTM A232
Wire 0.207 in or smaller
o 20 40 60 80 100 120
lnitial stress, thousands of PSI
Figure 8. 7. Goodman diagram for stresses. Spring stress range corrected by Wahl
factor.
Strength and temperature requirements dictate a metallic valve, which introduces

corrosion problems. Heavier springs are required for high suction pressures and
higher speed, and this in turn demands a higher TCP requirement.
8.6.1 Determining Valve Efficiency

Before one can begin the search for the "best" valve, a method of determining
the "best" valve must be adopted. It is logical to assume that a pump operating
at a high volumetric efficiency with a low NPSH has the "best" valve from the
point of view of hydraulics as well as endurance. Therefore, since valve design
affects the pump's NPSH requirement, it was decided that NPSH tests would be
the basis for studying a variety ofvalve designs, through areas, spring loads, valve
lifts, and valve weights.
The standard for establishing a reciprocating pump' s TCP Figure 3. 7 and a
curve Figure 3.8 relating volumetric efficiency wit.h decreasing suction head should
be the best way to present the data. By using such a curve, one can select the
volumetric efficiency that is satisfactory and then compare all tests at this same
figure.
256 VALVES
Throughout this discussion, certain terms will be used:
NPSHR. Net positive suction head required at the pump suction flange, in PSI
of liquid being pumped
Velocity ( V). Average velocity through val ve (based on plunger displacement
and val ve through area), FPS
Through area (Av). Net area through valve (ribs and wings deducted), in2
Lift ( L). Lift of val ve, such that the circumferential area is a percentage of the
through area, in
Spring loads. Installed spring load divided by valve through asea, pounds per
square in of valve area (POSIVA)
actual oallons
Volumetric efficiency (VE) = e. x 100
theoretical gallons
8.6.2 Valve Tests

Figure 3.8 in Chapter 3 shows a schematic ofthe test setup, along with the formula
used to cakulate and construct a family of curves of TCPA versus VE for dif-
ferent pumps.
8.6.3 Valve Combinations

Table 8.2 lists all the valve combinations used in the investigation. Note that for
each combination of valve type, through area, weight, and lift, there is a number
assigned for easy cross-reference.
The first and simplest analysis was to list valve combinations in increasing order
of TCP requirement for a selected VE. This is shown in Table 8.3. At 250 and
350 RPM, it is evident that valve combination 28 is the "best" This is a single-
port plate valve with maximum through area, maximum lift, minimum weight,
and 4 POSIVA load. Valve 18 appears to be the "worst" in this series. This is
the same val ve but with a through area of 2 in2 •
At 625 RPM, valve types or combinations surprisingly reverse their relative
positions in that the ''best'' valve at 350 RPM becomes one of the ''worst'' val ves
at 625 RPM. This leads to the conclusion that speed has more effect on valve
performance, and consequently on TCP requirements, than was first thought. At
one time, it was assumed that if velocity through the valve could be kept down,
high e:fficiency could be maintained.
8.6.4 Velocity Through Valve

To continue this analysis, the next step was to compare pump speed with valve
velocity, as tabulated in Table 8.4. lt is evident that for equal valve velocity,
greater TCP is required for higher speeds.
TABLE 8.2. Valve Combination Numbers
Spring
Comb. Through Weight, Load, Lift,
Valve Type No. Area, in2 oz POSIVA %
Double-port, plate, metal 1 2 13.5 2 125
2 100
3 50
3-1 4 138
4 100
5 50
6 2 6.8 2 125
7 100
8 50
9 4 125
10 100
11 50
Double-port, plate, Delrin 12 2 3 2 125
13 100
14 50
15 4 125
16 100
17 50
18 7 125
Single-pon, plate, Delrin 19 2 2.75 2 125

20 100
21 50
22 4 125
23 100
25 6 100
26 3.3 2.75 2 125
27 100
28 4 125
29 100
Wing-guided 30 3.3 17.3 2 125
31 100
32 50
33 4 125
34 100
35 50
36 3.3 8.5 2 125
37 100
38 50
39 4 125
40 100
41 50
42 2 11 2 100
43 4 100
TABLE 8.3. Valve Peñormance at Various Speeds with 3-in Plungersª
75% VE 80% VE 85% VE 90% VE 95% VE

N
UI
co Valve Valve Valve Valve Valve
Comb Comb Comb Comb Comb
TCP No. TCP No. TCP No. TCP No. TCP No.
250RPM
9.8 28 10.5 28 11.3 28 12.3 28 14.0 28
10.4 22 11.5 29 12.4 29 13.8 29 17.1 22
10.9 29 11.7 22 13.1 22 14.9 22 17.5 29
11.5 15 12.4 15 13.8 15 15.8 15 19.2 15
11.8 23 13.1 23 14.9 23 16.9 23 19.7 23
13.8 18 15.0 18 16.5 18 18.3 18 20.5 18
350RPM
15.1 28 16.2 28 17.7 28 20.5 28 25.5 28
16.1 29 17.2 29 19.4 29 22.5 29 28.0 29
17.8 22 19.7 22 22.7 22 28.3 22 X 22
18.6 15 20.0 15 22.7 15 28.1 15 X 15
19.0 18 20.5 18 23.3 18 28.8 18 X 18
X 23 21.7 23 24.9 23 X 23 X 23
625RPM
35 29 39 29 45 29 58 29 83 18
47 15 50 15 50 28 64 15 89 15
49 22 55 22 55 15 68 18 X 23
? 18 55 18 60 18 80 23 X 22
54 23 60 23 65 22 82 22 X 28
? 28 ? 28 67 23 95 28 X 29
TABLE 8.4. TCP as Function of Velocity, Velocity Obtained by Plunger Size, Speed, and
Valve Through Area
Velocity, Plunger Valve Through TCP

FPS Diameter, in Area, in2 RPM at 85% VE
7.3 1~ 3.3 625 21.0
8.3 2! 2 250 9.1
16.0 3 3.3 450 29.0
14.7 3 2 250 13.1
20.6 3 2 350 28.3
21.4 3 3.3 625 95.0
TABLE 8.5 Effect of Through Area on TCP with Valves of Same Type. Through Area
lncreased 65%
(TCP at 350 RPM, 3-in Plungers, 90% VE)
TCP with
Valves of:
Spring Lift, Percent
Valve Type Load, lb % 2 in2 3.3 in2 lmprovement
Wing-guided 2 100 26.0 20.2 22%

Single-port, plate 2 100 30.0 25.0 17%
4 100 31.0 22.5 27%
2 125 30.8 22.8 26%
4 125 28.3 20.5 28%
Average 24%
Rather than a direct relationship between velocity and speed, it can be seen that
speed has a greater effect than valve velocity. It should be noted that these data
are based on a 6-in-stroke triplex plunger pump, where various velocities through
the valve were obtained by varying either plunger size or valve through area.
8.6.5 Valve Through Area

As previously stated, val ve through area heretofore has been considered one of the
most important features of a pump valve. It is a generally held belief among pump
users as well as manufacturers that the greater the through area the lower the TCP
required. The tabulated data show that this is not necessarily true, because larger
valve through areas necessitate larger, heavier, and stronger valves. This, in turn,
may increase TCP requirements if carried to the extreme.
Table 8.5 Iists the effect of increase in valve through area on TCP require-
ments. An increase of 65 % in through area results in an improvement of only 24 %
in TCP requirement for all types and combinations of valves. It is rather obvious
that the increase is beneficia! but that it will have to be weighed with other factors
to be discussed later, such as weight.
260 VALVES
TABLE IUi Effect of Lift on TCP with Valves of Same Type

(TCP at 350 RPM, 3-in Plungers, 85"/o VE)
TCP at Lift of:

Through Spring
Valve Type Area, in2 -Load, lb 50% 100% 125%
Wing-guided 3.3 2 30.0 30.0 17 .8
Síngle-port, plate 3.3 2 19.2 19.0
Double-portª 2 2 29.0 21.0 20.8
Single-port, plateb 3.3 4 45.0 50.0
ªAt 80% VE (could not obtain 85% VE with 50% lift).
bAt625 RPM.
TABLE 8.7. Effect of Valve Weight on TCP with Same Type of Vaive
(NPSH at 85"/o VE)
Valve Through Spring Lift, Weight,

Type Area, in2 Load, lb % RPM oz TCP
Wing- 3.3 2 125 350 17.3 19.6
guided 3.3 2 125 350 8.5 17.8
Double- 2 2 100 350 13.5 27.8

port 2 2 100 350 6.8 24.0
2 2 100 350 6.8 29.5
2 2 100 350 3.0 23.6
Wing- 3.3 5 68 625 12.2 55.0
guided
Single-port 3.3 4 125 625 2.8 50.0
8.6.6 Required Valve lift

The question of required valve lift has been mentioned in the literature, and an old
rule of thumb was to provide lift sufficient to equal at least 50% of the through
area. A bevel-seal valve (wing-guided) naturally requires a greater lift than a plate-
type valve in order to provide equal "escape" area. The evidence shown in Table
8.6 indicates that 50% lift is undesirable even at slow speed and cannot be tolerated
at high speed. ("Slow speed" in this case is 250 RPM.)
The effect of excessive lift on valve life was not evaluated. While the tests
indicate that unlimited lift is desirable for best pump performance, it is known that
this would result in shorter valve life due to the greater impact on closing.
The tests show that for all operating conditions a lift of l 00 % is satisfactory.
8.6. 7 The Effect of Valve Weight

Valve weight, surprisíngly, has little to do with TCP requirements, even at high
speed. Table 8.7 shows that at 625 RPM a 12.2-oz valve required 23 PSI TCP
while a 2.75-oz valve required 21 PSI TCP, a loss of only 10% in TCP for a
gain of 350% in weight ..
Va/ve Lift Formulas
(8.23)
Q for pow_er pumps = total GPM X ?!" / nn 1 (8.24)
Q for simplex steam pumps = total GPM/0.75n (8.25)
Q for duplex steam pumps = total GPM / n (8.26)
L for bevel-face values = (F; + W - 0.3Fu/k)
{JM[l + (2M/IOO)] + 1 - 1) (8.27)
L for disc-type valves = '(F; + W - l.3Fu/k)
· ( v'M[l + (2M/IOO)] + 1 - 1) (8.28).

where
M for wing valves = kQ../S/40.5(F; + W - 0.3Fu) 312 (8.29)

M for disc valves = kQ..fS/58.5(F; + W - 1.3Fu) 3 12 (8.30)
2
.1Pforwingvalves = S[Q/72(D - L)L] ( 8.31)
2
.1Pfordiscvalves = S[Q/lOl(D - 2.4L)L] (8.32)
Nomenclature:
Av = valve through area, in2

D = diameter of seat opening, in
F; = installed spring load, lb
L = Iift of valve, in (at maximum flow)
.1P = pressure loss through valve, PSI
Q = maximum flow rate, U.S. GPM
k = spring rate, lb/in
S = specific gravity of Iiquid
W = weight of valve, lb
n = number of single-acting plungers or total number of working sides of
a double-acting piston (for triplex single-acting, n = 3)
n 1 = number of valves per section (normally 1)
Fu = upward force on valve in deflecting flow, lb
262 VALVES
TABLE 8.8 Effect of Spring Load on TCP with Same Type of

Valve
(TCPA at 90%)
TCP at
Spring Load
of:
Through Lift, Weight,
Valve Type Area, in2 2 lb 4 lb % oz
Wing-guided 3.3 25.5 29.4 100 17.3
Double-port 2 28.3 28.1 125 3.0
Conversion of Flow Rates
GPMave = 2GPM10131 / n (8.33)
GPMmax = ?rGPM1018i/ n (8.34)
8.6.8 Effect of Spring Load

In general, the spring load can be added to the TCP requirement; in other words,
the heavier the spring load, the greater the TCP requirement.
Previous tests indicated that a general rule for spring load would be 2, 4, a,nd
8 PSI (POSIV A) of valve through area: the 2 POSIVA for low-speed, low-suction
heads; the 4 POSIVA for "good" suction conditions and high speed; and the 8
POSIVA for high suction pressure (charged suction) and high speed. See Table
8.8.
These tests indicate that the need for a 2 POSIVA spring is doubtful, except
on marginal, low-speed applications where an absolute mínimum TCP is required.
However, there is something subtle about valve spring selection that is not brought
out in these tests. In many cases of "noisy" pumps, particularly those operating
in multiple units with a common suction header, a trial of different spring loads
will often produce a satisfactory spring that does not necessarily conform to lab-
oratory findings as determined on a single pump.
8.6.9 Summary
To summarize, these tests show that:
1. Pump speed has the greatest effect on valve performance.

2. Velocities through the valve of up to 35 FPS can be used if sufficient TCP
is provided.
8.7 DERIVATION OF VALVE VELOCITY 263
3. Improvement in TCP requirement is not proportional to the increase in

valve through area; therefore, a compromise through area must be adopted
based on operating pressure.
4. Valve lift should be sufficient to provide a circumferential Iift area to
the through area of the valve.
5. Valve spring load is still subject to question, but for general operation a load
of at least 4 POSIV A is desirable.
6. Valve weight has only a slight effect on valve performance. Generally, it
should not be a factor in valve design, except as it affects impact on closing
and opening and contributes to rapid wear.
8. 7 DERIVATION OF VAL VE VELOCITY
Valve velocity in a reciprocating pump is liquid velocity through the individual

pump valves and is based on the average pump ftow rate during that half-cycle of
a revolution for which the valve should be open.
Suction Discharge
Hall
revolution
One revolution
V = 2 ( for half rev) *

~~~~~~~~~~~~~~~~
x 231 in 3/gal x GPM
v 60 s/min X ] 2 in/ft X nAv
or
Vv = 0.642 X GPM/nAv (8.35)
where
Vv = liquid velocity through valve, ft/s

A,, = valve area, in2
n = number of discharge valves active during a revolution. For example,
for triplex single-acting pump, n = 3
*Beware of formulas that use the entire revolution for the calculation, resulting in an unrealistic ve-
locity of half of the velocity obtained by the formula.
264 VALVES
1~~~~~D2~~~~
1
1 '
1 1
r----D1----l
1 1
1
(a) (b)
¡..,;-~~-D2-~~-.,.¡
(d)
Figure 8.8. Unbalanced valve area.
Experience indicates that valve velocity should be limited to about 16 FPS for
dean liquid and about 12 FPS for slurries.
Note: AH GPM and gallons are in U.S. gallons, where 1 gal = 231 in 3 .
8.8 UNBALANCED VAL VE AREA
For many years a theory has been held that a liquid valve in a reciprocating pump
has an "unbalanced area" that results in requiring a much greater pressure in the
cylinder to open the valve, which is held closed by the discharge pressure above
it, creating a "high" pressure at the start of the stroke. For example, with a prac-
tical valve (Fig. 8.8(a)) the area at D 2 could be two times that at D 1 • In a 2000-
PSI pump, this could amount to a 4000-PSI opening pressure. Magnitudes as high
as seven times the discharge pressure have been reported. However, an excess of
about 10% overpressure in a pump cylinder that is operating normally has never
been seen.
In order to investigate the unbalanced val ve area theory, a "visible" valve
chamber was constructed to observe and measure valve action. Among other find-
ings it was discovered that a valve made with a large amount of unbalanced area
showed little, if any, difference in pressure required to open the valve in excess of
the pressure holding the valve shut.
The theory of "unbalanced valve area" is convincing, but if practical cases are
examined.it will be seen that it does not apply. For instance, considera plate- or
disc-type valve (Fig. 8.8(a)). Any material has a modulus of elasticity, which
means that it will defiect under load. In this case, the disc defiects so that the seal
8.9 POWER-OPERATED VALVES 265
is a "line contact" at point A (Fig. 8.8(b)). The so-called unbalanced area at Bis
actually balanced by hydraulic communication to the discharge pressure, even
though the communication path is extremely small. The same deflection takes place
with any type of valve, as shown in Figure a bevel wing-guided and
the same explanation applies.
A val ve with an elastomeric seal is sometimes accused of having a large ''un-
balanced area'' because the outside diameter of the seal is used to calculate the
"topside" area versus the diameter of the seat, usually resulting in a exces-
sively large difference. Figure 8.8(d) shows that here too there is line contact. The
elastomer itself is a high-viscosity fluid, and it transmlts the pressure to the usual
point of seal at B. If point A did seal (if there were rí.o solids in the liquid), the
entire unbalanced area to point A would become balanced and the trapped liquid
from B to A would become
8.9 POWERmOPERATED VALVES
It is possible that the use of reciprocating pumps for handling "coarse" coal or
other soft materials will require the use of positively driven or power-operated
liquid-end valves in order to actually crnsh the that become trapped be-
tween the valve and seat upon closure. There are three basic methods of driving
the valves: (1) mechanical, pneumatic, and hydraulic.
Mechanical. The intemal combustion engine type of cam-dliven valves could

be considered, allowing cam opening and spring closure. This would limit
the closing force to the spring load. But would the fixed spring load always
be sufficient to crush the particle? "Heavy" springs would tend to increase
cam and tappet wear and would affect mechanical Positive me-
chanical closure would be ruled out because of possible damage from metal
or hard objects caught in the valve. Cam operation also dictates the type of
valve motion available.
Hydraulic. Hydraulically operated valves (power cylinder) would possess a
computer-controlled advantage of any desirable and closely controlled valve
motionº The crushing force required for for example (and limited by
supply pressure), would build upas necessary.
Pneumatic. Pneumatically operated valves might be considered because of
their "softness." They have the same force as the hydraulic but
have the desirable feature of opening and closing more rapidly.
Both pneumatic and hydraulic valves are less complicated and less costly than
mechanically operated valves. However, the complexity and delay of valve action
associated with any of these methods of operation may introduce insum1ountable
problems.
266 VALVES
8.10 VALVES IN SERIES
Sorne reciprocating pumps employ multiple suction and discharge valves in series
(one or more above the other), and sometimes hydraulic systems are designed with
two or more spring-loaded check valves in series, always in an attempt to obtaín
maximum assurance of preventing back-flow. Statistically, the degree of assurance
is in direct proportion to the number of valves. However, in both applications the
magnitude of the pressure required to open all the valves is equal to the sum of
the POSIV As (spring pounds per square inch of valve area) of each of the valves.
This should be realized in any discussion of TCP, since the total valve opening
pressure (POS IVA) is directly related to TCP.
8.11 VALVES IN PARAllEl
Sorne years ago several pumps appeared on the market whose design took the
desire for low valve liquid velocity to the extreme by using double, side-by-side
(in parallel) suction and discharge valves in the liquid end on the theory that flow
velocity is a val ve' s worst enemy. Elsewhere in this book it is explained that the
worst moment in a valve's life is that at the instant of closure. High velocity
does result in a greater pressure drop through the valve, adding slightly to the
TCP requirement but not nearly as much •as the relatively heavy valve spring
POSIVA required for efficient pump operation.
Experience has shown that valves in parallel do not perform as one would ex-
pect-they do not open simultaneously and equally as required. Manufacturing
tolerances, particularly in valve springs, do not offer the precision necessary to
obtain exactly equal installed loads. The valve with the least POSIVA will open
first, and once it is opened a pressure equilibrium will be established which de-
stroys any opening differential for the second valve (until by possible opening at
maximum flow rate later in the stroke). Single-valve opening is especially apparent
in such pumps being operated, as is usual, ata slower speed or with smaller pistan
diameter than the design maximum. Therefore, it is generally noted in such pumps
that one valve of the pair always shows more wear. It has been found that large-
capacity pumps perform well with seemingly large single valves scaled up in size
to suit the pump.
8.12 VAL VE FLUTTER
Free-falling valves (usually with the absence of springs or little or no installed

spring load), both ball and disc types, have a tendency to flutter at the moment of
closure, which contributes to lower volumetric efficiency and noise. Such flutter
is a sort of planetary oscillation best described as similar to the motion resulting
from dropping a coin flat on a hard surface: usually it will roll in an oscillatory
motion until it comes to rest, whereas at other times it will "plunk" to rest im-
8.13 STEADY-STATE FLOW THROUGH VALVES 2o7
mediately. This motion in a val ve is set in operation when the free-falli.ng ball
strikes an edge of its seat in a slightly off-center position or when the disc strikes
the flat seat in a slightly camed position.
Many times when ball valves are used without springs, the reason is to allow
the ball to seat in an infinite number of positions in order to distribute wear
over the entire surface of the balL
Even though in sorne cases it may appear advantageous to have a val ve ''rotate''
in order to distribute wear, experience has shown that normal spring loads will not
allow the ball or disc to rotate, in spite of attempts to force them to do so by means
of skewed ribs or roller-bearing spring seats.
Subsurface or bottom-hole pumps used in pumping crude oil wells employ free-
falling ball valves, and it is quite common to hear the noíse of flutter at the of
the well.
Sorne proponents of light installed-valve-spring load on disc valves often go so
far as to provide zero load, which would not only invite valve fl.utter but, more
important, would also fail to provide the high spring load on closing required to
reduce the delay as much as possible.
Conventional vertical pumps with a horizontal valve axis cannot accept spring-
less val ves because of lack of any closing force from gravity. Such pumps must
be equipped with 90º suction and discharge elbows, with the springless vertical-
axis valves at the manifold ends of the elbows. This design greatly increases the
volumetric clearance and consequently contributes to lower volumetric e:fficiency
and possibly greater noise.
There is no known method of preventing such valve flutter, and if pumps must
be run without valve springs, they should be operated at reduced speeds.
8.13 STEADY-STATE FLOW THROUGH VALVES
Steady-state flow of liquid through a valve does not duplicate the dynamics of the
varying nature of fiow in a reciprocating pump. However, an experiment with
steady-state fiow in a test setup, whereby a "free" disc-type valve was installed
in a chamber with pressure taps above and below the valve, led to sorne interesting
results.
Figure 8.9 shows that without the disc, the pressure drop through the openings
(1-, 2-, and 3-in2 seats) followed the expected orifice laws of flow with low pres-
sure drops. However, with a "free" disc placed on top ofthe seat, sorne surprising
results were noted. First, the Bemoulli effect that restricts the free lift of valves
(where the high velocity of flow under the disc generates a low pressure that tends
to force the disc against the valve) was well demonstrated. It is obvious that the
Bemoulli efl:ect is greatly exaggerated when the area of the disc is large compared
to the seat opening area.
A critica! point was discovered with the two square inch area seat; at a par-
ticular fl.ow rate, a change of equilibrium took place and a constant-pressure flow
condition was established. One should find this chart of extreme interest.
268 VALVES
Loose !itting pin to keep

0
1 in 2 with disc in place over seat. No 3"Dia. Delrin disc

disc contact with pin during te.st.
~
~ 301-.~~-l-~--iJ-1..~~~+-~4---f-~~-¡ 2
3 in with disc encouraged
~ to hug seat by temporarily
¡;¡ _._ _......¡_ _ _ lorcing against seat
~ 201--~~-l----,!~-l-~~-!-~--'\.-l\---~---; 3 in 2 seat only
\
~ and with disc
' 2 in 2 only
o 10 20 30 40 50 60 70 80
Flow rate, GPM
Figure 8.9. Steady-state flow through valve.
8.14 VALVE DELAY
The term valve delay specifically refers to the finite time delay encountered from
that instant of the mechanical end of the delivery stroke until the actual full closure
of the discharge valve. Just prior to full closure, the liquid trapped between the
val ve and the seat must have work done on it to displace part of it into the discharge
manifold space and part back into the cylinder. This will involve an infinitely small
time period but the ratio of that time to the stroke time becomes significant at
higher pump speed. Viscous liquids and the presence of solid or semi-solid parti-
cles hinder rapid valve closure.
An addition to this delay is another contributed by the high pressure liquid
remaining in the space between the discharge and suction valve (Clearance Vol-
ume) that requires sorne "useless" movement of the piston in order to reduce the
pressure of the trapped liquid to a value somewhat below the suction pressure
before the suction valve can open.
Delay as discussed exists in all reciprocating pumps and it is impossible to
predict the degree. Delay does contribute to lower volumetric efficiency and noise
and it can be determined after the fact from pressure waveforms as described in
Chapter 12, Instrumentation. A signal picked up from any moving part ofthe pump
by the use of a proximity switch can serve as a marker for the absolute end-of-
8.14 VALVE DELAY 269
stroke on a dual-trace A simultaneous trace of the cylinder pressure

will enable one to determine the degree of delay.
Valve delay can be minimized by the use of valve springs with a greater
POSIVA. See figure 8. IO for the effect of valve on the shape of the pressure
waveform.
For further discussion of Valve Delay, see Chapter 3, SUCTION REQUIRE-
MENTS.
NOTE
In the preceeding Chapter 8 the !erms NPSH or NPIP have heen changcd to the
new terminology, "TCP."
9
SLURRY PUMPING
9.1 SLURRY PROPERTIES
A slurry is a mixture of solid particles in water or other liquid, the mixture being
of such a consistency that it can be pumped like a liquid. The most ancient slurry
pumping system in existence is the animal blood-circulating system, where a pos-
itive displacement pump (the heart) circulates a slurry of particles (blood corpus-
cles) in a liquid (the serum) through a complex pipeline (the viens).
The rules for fiow of slurries differ from the Darcy notion of "clean" liquids
because the rheology differs. In all fiow problems the viscosity of the liquid has a
greater effect than any other property. However, while most liquids behave like
water or oil, with the viscosity-fiow relation being Newtonian, slurries represent
a new phase, Bingham plastic, where the relation of shear stress to shear rate takes
on new meaning due to the Fanning friction factor f, the ratio of frictional forces
to inertial forces:
f = 144Dg i1.P /2pLV2 (9.1)
Figure 9. 1 shows the various types of fiow behavior (shear rate vs. shear stress)
encountered in slurry service. Each type has its own friction factors.
lt is beyond the scope of this book to describe the technology involved in slurry
pumping. For those interested in the subject a good reference is Solid-Liquid Flow,
by Wasp, Kenny, and Gandhi. *
*E. J. Wasp, J. P. Kenny, and R. L. Gandhi, "Solid-Liquid Flow, Slurry Pipeline Transportation,"
Trans Tech Publications, Clausthal, Germany, 1977.
270
9.2 PUMPS FOR SLURRY SERVICE 271
Rate ol shear
Figure 9.L Viscosity, shear rate vs. shear stress. A, Bingham plastic; B, yield pseudo-
plastic; C, dilatant; D, Newtonian; E, pseudoplastic.
9.2 PUMPS FOR SLURRV SERVICE
9.2.1 lntroduction
The abrasivity of a slurry is mainly a function of the hardness and shape (sharp-
ness) of the particles. However, there are several modes of abrasivity, one of
which is the action on metal or elastomer caused by erosion resulting from velocity
of fl.ow and another the reaction of the metal or elastomer caused by mechanical
abrasion. In the first instance, particle size and density are additional characteris-
tics that must be considered. In the second case, density probability has no effect
on wear rate. Mechanical abrasion as opposed to erosion is unpredictable, and few
data have been heretofore available. It stands to reason that there is sorne general
relation between the wear generated by both modes and others. Considerable work
has been done in this area, and as a result a standard method of measurement of
abrasivity of slurries has been developed; this is ASTM Standard G75.82, known
as the Miller number (see Section 9.7).
The movement of liquids in pipelines requires the use of pumps. Those who
have had experience with the petroleum industry, where the movement of slurries
is an everyday matter, are familiar with the transport of solids with mud pumps
and cementing and fracturing pumps.
Over the years, mud pumps have been improved so that rather long life can be
obtained from liquid-end parts subject to the effects of abrasive liquids. Recipro-
cating-type slurry pumps are one result.
272 SLURRY PUMPING
Centrifugal pumps have also been used to transport solids, but they have been
used where low heads are required, typically up to 100 PSI or so, for short pipe-
lines. R~aders may recall that a centrifugal pump is a hydrokinetic device where
the velocity of a liquid stream is converted to pressure, and to achieve high pres-
sures a large change in velocity is necessary. Abrasive liquids have a deteriorating
effect on the impellers and casings through which they flow, as a result of the
erosion caused by the liquid and the suspended particles. Therefore, these pumps
can be used only where the intemal velociry of flow i.s relatively low.
· Traditionally, the advantage of centrifuga! pumps is that they have a high ca-
pacity for a relatively low capital cost and usually require relatively little space.
One of their disadvantages when they are used in sluny pipelining can be seen in
the typical pressure-volume performance relationship, which tends to work against
the application (Fig. 9.2). If an increase in pressure (head) is occasioned by flow
restriction in the pipeline, say by the dropping out of solids, a desirable charac-
teristic of a sluny pump would be an ability to develop increased pressure to over-
come the restriction. The centrifugal pump provides the increased pressure only at
the expense of considerably reduced volume, which then results in a lower flow
velocity. With the reduced flow rate, the velocity might not be adequate to hold
the material in suspension and keep it fiowing in the line. On the other hand,
reciprocating pumps maintain a constant flow rate regardless of pressure, thereby
tending to "purge" any plugging effect.
Electrical power must be transmitted to each pumping station of a long sluny
pipeline systern, such stations at most times being many miles from the main trans-
mission lines, and this adds to the already rather high cost of power. Any savings
resulting from the more efficient conversion of energy during the life of the project
Performance limit íor

abrasive service
AH
Capacity, U.S. GPM
Figure 9.2. Centrifugal pump characteristics.

is desirable, and the greater mechanical efficiency of reciprocating pumps, on the

order of 85 to 90%, should certainly be considered in the design of such projects.
Reciprocating pumps have the desirable characteristic of maintaining a high
volumetric efficiency at any desired flow rate. This allows greater flexibility in
system design. Becuase of the positive displacement and high-efficiency features,
these pumps can be used for metering station throughput. Reciprocating slurry
pumps are so designed that liquid-end parts that are subject to the deteriorating
effects of slurries can be easily and quickly replaced. Other pump designs usually
require complete dismantling and overhaul. Figures 1.1 and 1. 7 in Chapter 1 show
a typical duplex double-acting piston pump and single-acting plunger pump, re-
spectively.
Because of their inherently larger capacity at lower speeds, duplex double-act-
ing piston pumps may appear to be a "natural" for ali abrasive pumping appli-
cations. When high pressures are considered (above 2000-3000 PSI), sorne are
quick to point out that such pumps have been used for years in the oil-well
drilling industry for pressures up to 4000 PSI or more, so why not apply them
in abrasive slurry service?
A. Practically all pumps in the higher horsepower range sold to the drilling
industry in recent years have been of the triplex single-acting type, regardless of
anticipated pressure.
B. Triplex (plunger) pumps were considered for drilling service because:
1. They have an inherently a high pressure pump.
2. They are light in weight per horsepower (important in a transportable
rig of any type).
3. A flushed stuffing box can prolong parts life.
C. When single-acting pumps were introduced to the drilling industry, they
were of the plunger type. Sorne attempts were made to flush the stuffing box, but
it was found impractical because (1) dilution of drilling mud is usually undesirable,
and (2) there is no source of clean flushing water on a drilling rig. (So point 3
above is no longer valid, and all pumps now use pistons.)
D. Drilling mud per se is not an abrasive liquid in the sense of present-day
slurry concepts. Typical drilling mud has an abrasivity of about Miller number 10.
(The reputation of drilling mud for being "abrasive" comes from the fact that it
picks up sand from the drilled formation.) It is ironic that the drilling industry
goes to great pains to reduce the sand content of the drilling mud to less than 2 %
in order to obtain greater pump parts life while in the slurry pipeline industry the
battle to increase the percentage of solids to the ideal 99.9% goes on.
E. It is improper to directly compare drilling mud pumping to the pumping of
most slurries. Many muds contain oil and special chemicals that are corrosive and
detrimental to elastomers. Muds run at relatively high temperature, 130ºF being
common. Chemistry, corrosion, high temperature, and high pressure combine to
overshadow abrasivity.
274 SLURRY PUMPING
9.2.2 Packing
A discussion of plunger pump packing must be preceded by an explanation of why
plunger purnps are sometimes used for abrasive slurry service and how they differ
from piston-type pumps, particularly in the matter of stuffing boxes and packing.
With the common duplex double-acting piston pump, typical design requires an
increase in piston rod strength, in both tension and compression, in sorne propor-
tion to the increase in pump discharge pressure. Therefore, a point is reached
where the piston rod diameter theoretically becomes so large that the pump, in
effect, approaches a single-acting duplex, the discharge characteristics of which
are extremely "rough." Accordingly, single-acting pump design dictates three,
five, or more cylinders, and multiplex single-acting pumps inherently have
smoother discharge characteristics, even over a small-piston-rod duplex double-
acting pump.
Of extreme importance is the often overlooked fact that the packing action of
an outside packed plunger pump is completely opposite to that of any piston type,
including a multiplex single-acting pistan pump.
In a plunger pump (Fig. 9.3), the plunger, during the pressure stroke, is trav-
eling to the right out of the pressure-loaded packing into the liquid, and during the
suction stroke the plunger is traveling to the left out of the dirty liquid into the
relaxed packing. Conversely, in a piston-type pump (Fig. 9 .4) (with both the pis-
ton and the piston rod packing), on the pressure stroke the piston is traveling to
the left into the pressure, and on the suction stroke it is traveling to the light away
from the liquid. With the piston rod packing, the same action is seen: on the
pressure stroke the rod is traveling into the packing, which is loaded by hydraulic
pressure.
;t
"' " "' s
Lubrication
~ pressure
Plunger travel
Figure 9.3. Single-acting plunger pump on pressure stroke-packing action. Note that
travel-of-plunger drag is counteracting tendency for pressure to extrude packing into clear-
ance (clearance exaggerated).
9.2 PUMPS FOR SLURRY SERVlCE 275
Lubrication
~:Frj
Hyd-ra-u,.,-lic----,
Piston~
\ 1 1
pressure ~
Heei
(extruded)
/
Heei
(extruded)
Figure 9.4. Double-acting piston type pump on pressure stroke-packing action. Note travel
of piston and rod drag are reinforcing tendency for pressure to extrude packing into clear-
ance.
The purpose of packing is simmply to close up the clearance gap between the
moving plunger and its associated parts, particularly the gland bushing, in the
stuffing box or the piston and its cylinder, and the piston rod and its stuffing box
parts. With ordinary packing this is accomplished by the use of material with con-
siderable resiliency.
The mechanics of all packing are such that regardless of the general shape of
the sealing member, the hydraulic pressure tends to force the member through the
clearance gap. Accordingly, practically all of the sealing and subsequent wear or
extrusion take place at the "heel" (Figs. 9. 3 and 9 .4).
It can be seen that the action in a plunger pump (Fig. 9.3) is such that on the
pressure stroke the heel is being "dragged" away from the clearance gap, thereby
greatly overcoming the force produced by the hydraulic pressure that causes ex-
trusion through the clearance gap, a benefit in high-pressure service. With the
piston-rod packing (Fig. 9.4), the heel is being dragged into the clearance gap by
both the motion and the hydraulic pressure, accelerating wear of the packing.
Lubrication of pacldng is extremely important in high-pressure service. It can
be seen that only with the plunger pump can a lubricant be applied to the plunger
as it is entering the hydraulically loaded packing, when it is most needed. This is
another benefit in high-pressure service.
Any attempt to lubricate a piston or piston rod is not as effective, since the
lubricated moving parts enter the packing only on the unloaded or suction stroke
when lubrication is not required.
Of extreme importance, because of the opposite packing mechanics, a plunger
pump in itself is not as satisfactory as a piston-type pump for pumping abrasive
material. Because the packing is relaxed on the suction stroke of a plunger pump,
the "dirty" plunger can readily load up the packing with abrasive particles and
subsequently act as an efficient lapping tool. But this objection can be overcome
by fiushing the packing internally with clean liquid-even to the extent that a
276 SLURRY PUMPING
plunger pump is more desirable for pumping abrasives if the dilution from ftushing
can be tolerated.
9.2.3 Plunger Flushing Methods

There are two fundamentally different methods of ftushing packing for slurry ser-
vice: synchronized and nonsynchronized ftushing.
Synchronized flushing is the positive injection of an exact volume of clean ftush
Iiquid, preferably during the suction stroke of the main pump plunger. This is
accomplished by directly coupling a reciprocating pump with individual cylinders
to the main pump crankshaft (Fig. 9.5).
Nonsynchronized flushing is the continuous injection of a certain amount of
clean ftushing liquid by an independently driven reciprocating pump (Fig. 9.6).
Main pump
Flush pump
directly
driven by
main pump
Figure 9.5. Synchronlzed flushing, typical hookup.
Main pump
~-~11
-Btt
11
Flush pump
separately
driven
Manifold
Gas-bladder
dampener
(Accumulator)
Figure 9.6. Nonsynchronized flushing, typical hookup.

9.2 PUMPS FOR SLURRY SERVICE 2n
TABLE 9.1 Flushing Methods for Slurry Pumps
Method
A B e D E F G H
Synchronizedª X X X X X
Nonsynchronized X X X
High-pressureª X X X X X
Low-pressureb X X X
Orifice used X X X X
Check valve used X X X X X X
Flush on suction stroke X X X X X X
Flush on discharge stroke X X X X
ªVariations in timing-Usually timed to ful! suction or discharge stroke of main pump.
There are several versions of both of these methods as shown in Table 9. 1. Of

these, the most popular are A and H.
Method A has been most popular in the past, but experience has shown that it
has many shortcomings. Inadequate flushing fluid is injected at the end of each
plunger stroke due to the sinusoidal shape of the ftow pattem of a reciprocating
pump. See Figure 9.7.
At the present state of the art, it is evident that method high-pressure non-
synchronized, is the most practical and efficient if properly designed flushing bush-
ings are used and if a sufficient supply pressure is available to produce the presently
acceptable mínimum of 3 % of main pump displacement. In any method, the clean-
liness of the fl.ushing liquid is paramount. A pressure of about 500 PSI is required
to provide a ftow of 3 % through the ftush system of a triplex pump. Accordingly,
in H, a pressure of 500 PSI above the main pump operating pressure would be
required. The advantages of method H are:
1. With continuous flushing the chance of any lack of fl.ushing due to phasing
is minimized. There is assurance that positive flushing is always achieved well
befare the start and well after the end of the main pump plunger suction stroke.
2. Sorne flushing during the pressure stroke is desirable to compensate for any
slight packing leakage which, if not flushed, would allow slurry to enter the pack-
ing space.
3. It allows prestart and post-stop flushing for sorne time before and after the
main pump is started and stopped. This provides added assurance that abrasives
are well fl.ushed out of the packing space before the plunger makes a stroke.
4. A single standard flush pump can be used to flush one or several main pumps.
5. A change in flushing rate can be more readily implemented by a simple
change of flush pump speed.
278 SLURRY PUMPING
Main pump
suction
-- Flush flow rate

----.1..-
----~ ...
Flush pump
discharge stroke
r---- Main pump valve lag
¡----------- -t 3GPM
l'tttt'b'~wt--1'\l~"""'H+
/ 1 Flush pressure 2500 PSI
' 1 1
Figure 9. 7. Flushing analysis. (a) Synchronized flushing on suction stroke of main pump.
Flush pressure is not important, as flow is positive as shown. Note that at beginning and
end of main-pump suction stroke, flush flow rate has diminished to zero. Dueto variations
such as valve lag, there could be periods of zero flush during sorne part of the main-pump
plunger stroke. (b) Nonsynchronized high-pressure flushing. Admits a constant flow-max-
imum during the main-pump suction stroke-with overlap at each end and even slight flush-
ing during the discharge stroke. There can be no period of starved flushing.
9.2.4 Flushing Details

A rate of about 3 % of the total main pump displacement is required to guarantee
sufficient flushing. The actual requirement should be determined by test, and the
rule that ''the more the better'' applies up to a reasonable limit. lt bears repeating
that the flushing liquid should be as "clean" as possible.
The following features are desirable in a fl.ushed slurry stuffing box:
l. Spring-loaded main and auxiliary packing.

2. Water-cooled stuffing box.
3. Dimensions such that "wetted" portian af plunger never enters main pack-
ing.
4. Clase-fitting stuffing box trim (cast iran or ductile iron preferred). Figure
9. 8 shows the appraximate dimensions and general design af a caal conversion
type stuffing box. See Figure 7 .6, Chapter 7, for recommended stuffing box trim
clearance for ali applications.
Flushing liquid velocity should depend upon the flushing rate and the plunger-
flush bushing díametric clearance. The following formula can be used to approx-
imate the carrect velacity:
percent flush
V= 0.4085 X GPM X D2 _ D2 X number af cylinders (9.2)
b p
where
Db = stuffing box diameter, in

DP = plunger diameter, in
The "flush" bushing should be of the design shown in Figure 9.9. This design
provides uniform fiush by directing the fl.ush liquid to the bottom of the plunger.
--------2.5S------
Tell· tale Coo! Flush
.5S -----r-1
- - - - - + - - - _ .n_I_
.._________,______. . __ _____ l_J
~s
1
,,,,
JI
¡.o---------------4.5S-------------____,.,
:1
Figure 9.8. Generalized coa! conversion stuffing box.
280 SLURRY PUMPING
¡ji
Y."-;;.j¡11¡...__
11
11
See Fig, 7 ,26 lor

recommended diameters,
,..._____ ,¡, stroke-----
Figure 9.9. Typical flush bushing for horizontal pumps. R 1 and R 2 offset as shown so as
to result in feathering out of grooves at the 90º limits, as indicated at A = A.
9.2.5 Pistons
The piston is so constructed as to have three basic elements, as shown in Chapter
7. Under hydraulic loading, the ''rubber'' is pushed back against the fabric section
and out against the liner to form a seal. The fabric section provides extrusion
clearance control, and the metal back-up plate or piston body provides the struc-
tural capacity to hold the piston load.
9.2.6 liners
The liners are made of abrasion- and corrosion-resistant metals that have been
found in extensive service to resist wear for specific slurries. Piston rods and
plungers are usually coated with similarly suited materials for the same objectives
while retaining base-metal characteristics for the required mechanical loads.
9.2.7 Piston Membrane Pumps

There is increasing interest in pis ton membrane pumps (described in Chapter 1)
for the pumping of abrasive slurries, particularly with slurries of abrasivity above
Miller number 50.
There have been many improvement in engineering concepts, metals, and
elastomers, and perhaps sorne of the old ideas may flourish. The membrane
pumps deseribed in Chapter 1 are in service in many applications with extreme
success.
While membrane pumps will protect the pistons, liners, or plungers against
Leg 1 filling
Clean
Main
liquid Leg 2 displacing
pump
Figure 9.10. Switch-loop pumping.
abrasive slurry, no design has ever offered protection for the liquid-end valves of
a pump.
9.2.8 Switch-Loop Pumping

Switch-loop pumping is a method of isolating pipeline pumps from the destruction
of abrasion, allowing "standard" reciprocating pumps to operate at ali times in
an environment of "clean" liquid, thus prolonging the life of parts, including
valves. One version of the system is shown in Figure 9 .10. Legs 1 and 2 are
reasonable lengths of pipe, either "legs" or "loops," of sufficient volumetric
capacity between the directional valves A and C or B and D to allow a decent
slurry pumping period of several minutes before switching to the other leg.
In operation, Figure 9.10 shows that the slurry previously introduced into leg
2 is being forced into the pipeline with "clean" high-pressure liquid through di-
rectional valves B and D. While this event is occurring, leg 1 is being filled with
low-pressure slurry, displacing the clean liquid remaining from the last high-pres-
sure cycle through valves A and C.
Timing devices can cause the directional valves to operate on a predetermined
cycle, and such timing can allow the liquid to continue through the valve for a
short interval of time after the slurry passes, permitting the valve to close in clean
liquid. The slight amount of slurry dilution caused by this operation could probably
be compensated by an original higher concentration of solids.
The cost of maintaining at least four large high-pressure, high-cost directional
valves must be weighed against the saving in pump parts life.
9.2.9 Other Methods

Aside from the present 50% by weight coal/water slurry pumping with recipro-
cating pumps directly into the pipeline, other versions are mentioned;
282 SLURRY PUMPING
l. Lock-Hopper. A system similar to the Switch Loop method, alternate

hoppers are arranged at the inlet of the pump whereby one hopper is
being filled with slurry while the other is being pumped.
2. Capsules. Is a method of sealing water or liquid sensitive materials,
(grain for instance) in a plastic capsule and pumping through the pipeline
with a water vehicle.
3. For pipelining, there has been a proposal that grain be coated with and
protected by sorne material stable in water until time for removal.
4. Sewage of course, can usually be handed by one of the lock-hopper meth-
ods.
5. Pneumatic air or gas as the carrying medium has been considered.
6. Liquid C0 2 has been successfully used to transport coal in a pipeline.
7. Proposals have been made to withdraw sorne of the pipeline coal and
water to generate steam for station pumping power.
8. Sulphur-oil slurries have been considered.
9. Slack-flow energy recovery methods have been considered. Slack-flow is
typical in hilly country where there are substantial down-hill runs to the
station or destination.
10. In areas of scarce water supply, double pipelines have been considered-
One to pump slurry and the Qther to return the separated water.
l l. Coal "logs"-coal molded into logs and pumped through the pipeline
like a pig.
High concentration coal (80% by weight) systems are being considered, even to
the extent of burning the slurry as received without water separation.
There is extreme interest in pumping course coal (run of the mine) but little
success has been had. Present high pressure reciprocating pumps will not tolerate
large particles. Pumps with power driven valves may be developed for such ser-
vice.
9.3 HORIZONTAL VS. VERTICAL PUMPS FOR SLURRY

SERVICE
Any claim that a vertical plunger pump requires less plunger flushing than a hor-
izontal pump is based on the conjecture that gravity aids in keeping solid particles
away from the packing. However, only large particles can be kept from the pack-
ing by gravity. It is an axiom that large particles do not affect packing and plunger
wear as muchas small particles; it is much easier for the packing to exclude large
particles than small ones. The smaller particles can be readily carried into the
packing, become embedded, and actas a lap on the plunger. The turbulence that
exists inside a fluid cylinder of a high-speed plunger pump is so severe that it is
impossible to rely on the force of gravity to keep particles from the packing of any
typical pump. Unsuccessful attempts have been made to create "permanent" bar-
9.4 SUCTION PRESSURE FOR SLURRY PUMPS 283
riers by the use of grease, oil, etc. Sorne highly specialized pumps successfully
maintain an oil barrier between the fluid and the plunger but run at very low speed.
It has been stated that plungers or pistons in horizontal units in such service
tend to undergo greater wear than in vertical untis. There is no evidence that wear
is concentrated on the bottom of any plunger or piston in a propedy designed
horizontal pump. (Remember Henry Ford's critics who said the pistons of a V-
type engine would wear out rapidly because they were lying down?) The wear
pattem of all plungers is the expected hourglass shape, with uniform wear around
the center portion of the plunger, and the typical liner and rod in pumps
exhibit the familiar wear at each end of the stroke.
The mechanics of packing are such that the effect of piunger weight is trivial.
As the hydraulic pressure works on the packing, it forces the resilient members
against the plunger for its entire circumference, thereby centering the plunger in
the stuffing box with a greater force by far than the opposing force of gravity. It
is during that part of the stroke that the plunger is experiencing its greatest rate of
wear due to abrasion.
The horizontal design has the distinct advantage of having a lower fluid end
with a gain of several feet of actual suction head. Sorne think the horizontal design
is more accessible for servicing. Because valves should operate in a vertical mo-
tion, vertical pumps must be equipped with surge legs or elbows of considerable
length to which the valve pots can be attached, thus adding extensively to the
volumetric clearance of the pump. Horizontal pumps allow vertical motion of the
valves, a mandatory requirement.
9.4 SUCTION PRESSURE FOR SlURRY PUMPS

It is an axiom that slurry pumps must have a "charged" suction, usually provided
by the use of a centrifugal pump or by the controlled pipeline pressure at the
downstream stations. Other methods of fumishing high suction pressure can be
used. For instance, where conditions will permit, elevated conical-bottom supply
tanks have been used with success. One objection to this system is the drastic drop
in suction head as the tank liquid leve! is lowered with vv ithdrawaL
Slurry-pump suction pressure requirements are affected by the viscosity of the
liquid, greater viscosity resulting in greater pressure drop between the source and
the pump inlet. The greater density of rnost slurries also affects the losses due to
liquid acceleration. These losses, exacerbated by unusually long and tortuous suc-
tion systems associated with slurry-pumping installations, demand that such pumps
have their suctions charged, usually with centrifugal pumps. (With very high vis-
cosity slurries, such as sorne coal-oil mixtures, it may be necessary to consider
limiting the maximum pump speed.)
Slurry pumps require heavier than normal (greater POSIV A) valve springs for
severa! reasons. A closing val ve cannot take advantage of the applied (discharge)
pressure to effect a seal until the valve is almost completely closed against its seat.
If the apparent high viscosity of the slurry and the presence of rather large solid
particles delay the final valve closure, poor volumetric efficiency and a rough-
running pump are the result. The use of heavier springs helps to overcome these
284 SLURRY PUMPING
deficiencies. Installed valve spring loads as high as 8 POSIVA are used. Of course,
the penalty of heavy valve springs is the required higher suction pressure or
NPSHR.
Single-acting piston-type pumps usually require high suction pressure to allow
proper filling of the cylinders on each suction stroke because the piston design is
such that a "suction" cannot be created with the "one-way" piston. Efforts to
supply a piston seal on the suction stroke by the use of a "double-acting" piston
have not been successful because of the early failure of the back-seal owing to
lack of lubrication in a partiaily dry liner. (It should be realized that with a double-
acting piston, the seal generated by the pressurized rubber also provides the "suc-
tion" required on the back side.)
· Rather than the need for calculating or testing for the NPSHR for each slurry-
pumping application, experience has shown that a suction pressure of at least 50
PSI should be provided, either with an appropriate centrifugal pump or by regu-
lating the upstream pressure on in-line station pumps to that value.
9.5 COAL SLURRIES
Currently, the most widely used liquid phases for coal slunies are water, oil, tol-
uene (or other solvents), and liquid carbon dioxide.
9.5.1 Concentration of Solids

Frequently, the question is asked, ''What is the maximum concentration of coal
solids that can be pumped?" Confining the answer to the pump only, a general
statement can be made: The rheology of the slurry is dictated by the pipeline re-
quirements, and the pump can usually handle anything that can be transported
through a pipeline.
In the case of a coal-water slurry, Figure 9 .11, the chart of a typical coal slurry,
indicates that the apparent viscosity increases with concentration. In a reciprocat-
ing pump, no unusual operating conditions are encountered with viscosities up to
8000 SSU for true Newtonian liquids. It will be seen that in this "coal B" slurry
the equivalent value of 8000 SSU was reached at about 70% by weight solids.
Note also that at 73% the mixture had no fluidity (was pasty). So again, as far as
the pump is concemed and with that particular coal, there is good reason to believe
that a concentration of 60% by weight could be readily pumped.
Coal-oil slurries are more critical in that they increase in viscosity at a more
rapid rate and are highly affected by temperature and the viscosity of the mixing
oiL Figure 9.12 shows that for the particular case illustrated, concentrations above
55 % result in very high viscosity. This phenomenon must be taken into account
in arriving at a concentration for such slurries. From this, it is evident that the
limiting concentration could be determined only by actual test.
Coal-toluene mixtures would probably follow coal-water characteristics, but lit-
tle information is available on this.
8
::¡
IJ'l
IJ'l
§.....
6
¡;:
·¡¡;
o
u
vi
·;;;
.,_,
e: 4
...
Q)
~
c.
<
2
O_!.-~-,,!~~~~~-,,,~~~i,,,.........,,,,~
30 40 50 60 70 -So
Concentration, percent by weight
Figure 9.11. Coal-water slurry viscosity. A, fine coa! ( 80% pass 200 mesh); curve B,
coarse coal ( 80 % pass 100 mesh).
8
:::::>
IJ'l
VJ
§..... 6
¡5.
·¡¡;
ou
-
(/)
·:;;
e: 4
f:'.
"'c.a.
<
2
Concentration, percent by weight
Figure 9.12. Coal-oil slurr)1 viscosity. Oil is No. 6 fuel oil. A, 200ºF, B, 160ºF.
285
286 SLURRY PUMPING
9.5.2 Particle Size
Again, there is sorne relation to particle size vs. pipeline requirements, and it can
be generally stated from experience to date in actual slurry projects that the pump
has not been the limiting factor on particle size.
In a reciprocating pump, the particle size only affects the operation ofthe valves.
In pumps of larger capacity (larger valves), no trouble is encountered with particles
up to 8 mesh. (Fortunately, crushed coal has a good distribution of smaller parti-
cles below this maximum size.) Statements appear in the literature that l-in di-
ameter particles of coal have been pumped. These were cases where the percentage
of l-in particles was very small, and there is no doubt that the pump will tolerate
a few large particles.
Types of valves vary in their ability to handle large particles. The elastomeric
seal slurry-type valve is far superior for normal coal-water slurries at low temper-
ature. But dueto the wide metal-to-metal and elastomer-to-metal contact surfaces,
large particles have more of a tendency to hold the valve open. Such valves have
no difficulty in handling coal with 8 mesh maximum particle size.
Tests show that particles of coal up to !-in in diameter can be handled with
properly designed spherical valves. The high loading of a line bearing contact
easily crushes those particles caught between the edges when the valve closes.
Particles on each side of the sealing line are readily displaced away from the seal-
ing line by the rapidly diverging shape of the spherical portion of the valve, which
minimizes the tendency of large particles to hold the valve open.
Spherical valves are also valuable for hot or chemically active liquids and slur-
ries that are detrimental to elastomeric seals.
Spherical valves should be considered only in the applications listed above.
They lack the advantage of a renewable elastomeric seal. This, combined with the
inherently high metal-to-metal bearing loads, results in shorter life than a slurry-
type valve. Spherical valves will not tolerate highly abrasive slurries, of a Miller
number of above 50. A slight reduction in volumetric efficiency may be expected
dueto lack of the more efficient elastomeric seal.
9.5.3 Concentration and Particle Size
In the milling process, coal and other minerals appear to follow a rather consistent
pattem of particle size distribution. In other words, a coal reduced to the extent
that the largest particles have a dimension smaller than another milled coal will
have a larger percentage of "fines." These fines tend to control the apparent vis-
cosity of a water mixture, and a slurry with a large percentage of fines will have
a higher viscosity than another slurry of equal concentration of solids by weight
but with a smaller percentage of fines.
In view of this, it is difficult to predict the pumpability of coal slurries of a
given concentration without knowing the approximate apparent viscosity at that
concentration.
'V SLURRY EROSION 287
9.6 VALVE SERVICE FOR SLURRY PUMPING

Valve service for slurry pumping is covered in Chapter 8, but this statement serves
as a waming that limited valve lift caused by reduced fiow through the valve open-
ing when a lower than rated pump speed (RPM) is maintained may create a re-
striction to the passage of solid particles. Extended pump operation at signí:ficantly
reduced speed may result in an accumulation of solids in the pump cylinder with
drastic results.
9.7 SLURRY EROSION
9.7.1 lntroduction
For the purpose of this discussion, a "slurry" is described as a mixture of solid
particles in a liquíd (usually water), of such a consistency that it can be pumped
like a true liquid.
The term "slurry erosion" is strictly defined as that type of wear or loss of
mass of material when exposed to a high velocity stream of slurry, whether the
material is moving at a certain velocity through the slurry or whether the slurry
is moving past the material at a certain velocity. However, one should include
other forms of wear encountered in handling slurries-a process seeing increas-
ing interest in industry, especially with the rather new method of transporting
minerals, principally coa!, and other solids (even with the possible indusion of
grain), for long distances through pipelines at pressure in the order of 2,000 PSI
(13,790 kPa).
Dry Abrasive Wear, another mode of wear, is mentioned in this discussion
but it is a type of wear seldom if ever encountered in slurry handling. The com-
bination of Abrasion-Corrosion of a typical slurry system is responsible for the
most severe form of wear and it leads the list of severa! other important modes.
9. 7 .2 Slurry Wear Modes

Sorne of the most common wear modes are listed below and described in Figure
9. IA. Note that a few of these basic modes can be extended into sub-modes,
according to whether they are responding to the following motions; reciprocating,
oscillating, circular or continuous; oi! or liquids other than water; and ratio of
areas of mating parts as shown in Figure 9. lA.
Mode A. Abrasion-Corrosion wear is the result of any metal-to-metal rubbing
in the presence of abrasive solids in a liquid. Aside from high-velocity erosion,
this is the most destructive and most misunderstood mode encountered in han-
dling slurry. Typical parts of reciprocating pumps involved are; metal-to-metal
valves and seats (upon each closure), metal piston parts rubbing on metal liners
and plungers, or piston-rods rubbing against metal stuffing box parts or trim.
In the case of parts of different metals exposed to the slurry, it is obvious that
electro-chemical {electrolysis) effects are included in this mode.
288 SLURRY PUMPING
A
-- D F t.
--
' 'f
"~~'
7-~""""'º~
1
PIPE 11~
V 'Jb : , ~,'.;D,
q> ,... ~
LARGE TUMBLING ROCKS
ABRASIDN-CORRDSIDN HIGH VELOCITY ERDSIDN SALTATIDN ERDSIDN
Bl B2 B3
SCDURING llEAR SCOURING llEAR

llEAR AREAS ECUAL llEAR AREAS UNECUAL llEAR AREAS UNEQUAL
e E G
~
':"o "::.-/o:c:_;é)_":_.
/ -c:aLLArsrn'u VAPOR
PIPE \IALL
BUBBLES.
CRUSHING ~ GRIND LO\/ VELOCITY ERDSIDN CAVITATION
Figure 9.lA. Slurry Erosion Wear Modes
Mode B. Scouring Wear is encountered with elastomer-to-metal rubbing with

abrasive so lid particles becoming embedded in the softer elastomer or rubber,
Le. pistons, packing and valve inserts.
Mode C. Crushing and Grinding in abrasive metal-to-metal contact. For ex-
ample, a valve repeatedly closing with great force against solid particles trapped
between the valve and seat at closure.
Mode D. High-velocity Erosion, while not a usual mode of wear in recipro-
cating pumps handling slurry, it can become a very destructive one. For exam-
ple, when a valve seat or piston wears to the extent that a slight leak develops,
the extremely high velocity of slurry leakage through that small gap can result
in catostrophic and rapid failure of the parts and even the costly liquid-end of
the pump. See Figures 2 and 3. Also slurry-throttling valves and parts down-
stream experience this type of rapid wear. The impellers and cut-water of cen-
trifuga! pumps are subject to this type of wear.
"High" velocity is usually considered as that greater than 20-30 ft/s (65-
100 mis).
Mode E, Low-Velocity Erosion is usualiy a low rate wear mode that takes place
where there is flow of slurry at regular low velocities. In a pipeline with laminar
flow, the velocity profile (in the shape of a parabola) is such that the velocity
near the wall of the pipe is nearly zero, and minimum wear takes place. Also
the impellers and cut-water of centrifuga! pumps are sometimes subjected to this
mode.
9.7 SLURRY EROSION 289
Mode F. Saltation Wear comes about in pipelines handling a great number of

larger than usual particle size-actually "chunks"-which tend to tumble along
the bottom of the pipe, resulting in rapid wear. An example is the transportation
of phosphate rock from slurry-pit to the processing plant. The pipe requires fre-
quent turning to distribute the wear.
Mode G. Cavitation can result in damage to the metal in the liquid-end of the
pump or to the parts of a reciprocating pump through the microscopic but intense
liquid pressure blasts against the metal near the cavitation area following the
repeated collapse of the vapor bubbles.
9. 7.3 Effects of Wear

The effect of solids concentration on abrasivity should be recognized. Figure 10.6
shows how the abrasivity increases very rapidly from zero to about 10 percent
concentration, then begins to ftatten. lt is an interesting fact that in the oil well
drilling industry the well known slurry called "drilling mud" in itself is not
abrasive-the abrasivity comes from the solid particles of crushed and broken
rock or "cuttings" generated in the drilling process. Contaminated mud is usu-
ally "cleaned" by; settling, screening or centrifuging.
Curve 10.6 is supported by the fact that the sand content of drilling mud must
be reduced to less than 2 percent before effective improvement in pump parts life
can be realized.
9. 7.4 Dry Abrasivity

As a result of further work it was decided that ali Miller or SAR Number Tests
would be run in duplication, one with the "as received" material and another
with an inhibited slurry obtained by adding a <lose of strong NaOH solution to
raise the pH to the order of about 13. Thus a "number" approaching the dry
abrasivity could be obtained by the reduction of corrosion effects. See Table
9. lA. Note that the increase in pH of the slurry by the addition of NaOH (as is
sometimes done in actual systems) in most cases reduces the Miller or SAR
Number.
9.7.5 Conclusions
Typical pumpable slurries possess inherent "~pparent abrasivity" which must be
determined by test to enable one to predict the cost of pump replacement parts
or equipment run in those slurries. The "apparent abrasivity", without inhibi-
tion, is the synergistic reaction of many factors as shown in Figure 9.2A, such
reaction or effects being described as The Morrison-Miller effect(2): Due to the
complex synergism involved, the wear response of a material (A) in a certain
(2) A Modified G-75 Abrasion Test for Corrosive Environments, Morrison and Corbett-Corrosion
Testing Laboratories, Inc. Wilmington, DE 19804 and Jenkins, Savannah River Laboratory, Aiken,
se 29802
~
co
o
TABLE 9.1A. Slurry Abrasivity Reduction BT PH lncrease
A B e D E F G H
PH MILLER NO PH MILLER NO PCT REDUCE MN/PCT
TEST NO SLURRY PH AS REC MN AS REC PH AS INI-:1 MN AS INH
651 FLY ASH 9.6 85 13 54 36 2.33
648 FLY ASH 12.4 18 12.5 14 22 0.81
647 TAILINGS 9 34 12.4 26 24 1.45
602 A 6.5 776 13.2 477 39 20.14
603 B 6.1 825 13.3 576 30 27.33
600 MINE WAT 6 122 12 85 30 4.02
596 e 7.4 17 12.5 15 12 1.45
599 D 7 60 11 50 17 3.60
595 CRYSTALS 7.2 0.2 12.1 0.1 50 0.00
593 COMPOSITE 7.8 53 12.7 40 25 2.16
592 COMPOSITE 7.7 48 12.6 39 19 2.56
594 COMPOSITE 8.3 47 12.5 39 17 2.76
563 LIME 11.4 0.9 13.1 4.8 -433 0.00
589 E 8.3 69 12.5 42 39 1.76
548 F 5.7 5.5 12.l 3.1 44 0.13
549 G 4.1 20 12.9 12 40 0.50
561 BASALT 7.3 249 13.2 131 47 5.25
560 DOLERITE 215 12.8 182 15 14.01
543 REFACTORY 68 53 22 3.08
541 COMPOSITE 7.6 113 12.2 49 57 2.00
539 TILL 6.8 8! 12 67 17 4.69
540 TILL 11.8 42 11.8 48 - 14 -2.94
537 PHOSPHATE 7.5 81 12.9 21 74 2.09
502 PHOSPHATE 6.8 75 12.9 30 60 1.25
525 REFACTORY 6 100 13 46 54 1.85
523 PHOS. TAIL 7.9 80 12 67 16 4.92
517 H 7.9 86 12 62 28 3.08
518 I 6.5 94 12 82 13 7.36
519 J 7.6 89 12 63 29 3.05
516 GROUT MUO 6.1 61 12 49 20 3.!0
506 SAND 8.6 55 12 51 7 7.56
502 MARBLE 5.5 13 4.2 24 0.23
503 FLY ASH 12.1 11 12.2 13 - 18 --0.61
Ali the above were inhibited with a strong dose of
The following were inhibited with CaOH.
601 COPPER TAL I0.4 162 ,,,

lLoL
" 50 69 2.34
601 COPPER TAL 8.2 560 12.1 55 90 6.21
.:1
.
t:I
,.............................., ,................................................................., .........................................
l 1. CORROSIVE l ¡2. RESISTANCE OF PROTECTIVE l l3. DISSOLVED AIR j
¡ LIQ ID ¡ 1 FILM OF CORROS ION PRODUCTS 1 l COXYGEN OR l
"·················· ··········" l ENVIRONMENT 1 l
"······································ .......................... •...................... ···············•
~
¡·4·:···E:t."E:c=rRai.:v·s1s·;···1;::·········¡ :-··························· ······································•

¡ 5. TRUE ABRASIVITY OF SOLIDS: ¡
j TIJO METALS INVOLVED j ¡ CPARTICAL HARDNESS. SIZE. j
•·····················································• l SHAPE AND CONCENTRATION. 1 ¡
·································~··································.Z
,....................................................... r·······························•·····•··••••••••••••••••••••••••••.,
l::7. RELEASED CORROSIVE j !6. SOLUBLE ELEMENTS IN SOLIDS j
CONNATE IJATER FROM f°"""""1 FDRMING CORROSIVE SOLUTION j
¡ ORE PARTICLES. l
i. •••••••••••••••.••••••••••••.••.••••••••••••••••••••••••••••••••••• .i
i. ••• ...................................................
Figure 9.2A. Synergistic effects of seven factors in slurry abrasivity
MATERIAL
A e
MATERIAL
B pH y
OF THE FDUR CDMBINATIDNS CAND MANY MDREI INDICATED ABDVE

NDNE OF THE RESPONSES DR EFFECTS ARE PREDICTABLE AND
THERE USUALLY IS Na NUMERICAL RELATIDNSHIP DF MILLER DR
SAR NUMBER BETVEEN ANY PAIR. INDIVIDUAL TESTS MUST BE RUN.
Figure 9.3A. Morrison-Miller effect in slurry abrasivity
292
9.B SLURRY ABRASION TESTING 293
THE GOLD NUMBER

FOR LOW ABRASIVITY
JOOO ........... "' ......... .
2500 ..................... .
~ 2000 ....................... . 00•00000••••••••••,.•o••O••••••••••••••>•
~
1500 .................... .. ••••••i•••O•ooooo•••••••O••••• •••••••••••••••••<•••••
~ 1000 ........... .
500 •••••• ' ' ••••••••••••••• ~ •••••••••• o •••••••
0+--''--~~~~-1-~~~~~~¡..........~~~~~-+-~~~~~--1
o 10 20 30 40
STANDARD MILLER NUMBER
l''igure 9.4A.
slurry (C) gives no clue as to how material (A) would respond to another Slurry
(D) and, vice versa, the effects of Slurry {C) on material (A) gives no clue as
to the effects of Slurry (C) on material (B). Figure 9.3A.
The elusive combination of erosion-corrosion was actually recognized as early
as 1967 as shown in Figure 10.7 where the mass-loss rate is the reverse of what
would be expected.
9.8 SLURRY ABRASION TESTING
9.8.1 Miller Number*

AH solid minerals possess a degree of hardness, from soft tale used in face powder
to aluminum oxide with a hardness next to that of diamond. Such hardness imparts
abrasivity to the slurry, and if one is concemed about the type of pump and the
life of both the parts and the pipeline, the degree of abrasivity must be known.
The greater the abrasivity, the shorter the life of pump parts and pipeline. The
Miller number describes the relative degree of abrasivity of slurries as based on
the mass loss of a "standard" 27% chrome iron wear block when run in a partic-
ular slurry for a period of time.
*See ASTM G75-89, "Determinalion of Slurry Abrasivity by Miller Number," Standard of ASTM,
1916 Race Street, Philadelphia, PA, 19!03.
294 SLURRY PUMPING
Figure 9.13. Miller number machine.
9.8.2 SAR Number (Slurry Abrasion Resistance)

To relate the effects of any slurry abrasivity and corrosivity on materials other than
the standard wear block, the SAR number was developed, using the same test
equipment and procedure as for the Miller number. The SAR number of a material
is based on the mass-loss rate (converted to volume-loss rate) of a wear block of
the material in question when ron in any particular slurry for a period of time.
Therefore, the following procedure will apply to both tests, the difference being
applied in the final part of the computations, where an additional step is required
to convert the final results into a number representing the volume loss.
9.8.3 Test Equipment

The device used to measure the relative abrasivity of various slurries (Miller num-
ber) or to determine the abrasion resistance of materials to various slurries (SAR
number) consists in general of a standard 1.27 mm X 2.54 mm<! in X 1 in) metal
wear block, driven at a rate of 48 strokes per minute, with a 200-mm ( 8 in) stroke,
riding in the bottom of a tray containing a 50 % by mass slurry of the solids mixed
in distilled water. A dead weight of 22.24 N (5 lb) is applied to the block. For
each test, the tray is fitted with a new piece of neoprene rubber sheet to act as a
lap. The interior of the tray is of a ftat-bottomed V shape so as to confine the path
taken by the wear block through the actual solids.
At the end of the stroke, the wear block is lifted 1 mm (.Ji in) off the lap to
allow fresh slurry to ftow under the block. The block holder is made of plastic, as
are the trays, so that electrolysis in certain slurries is eliminated. Figure 9 .13 il-
lustrates machine principies.
9.8 SLURRY ABRASION TESTING 295
TABLE 9.2. Sample of Recorded Test Data
Weights, g Loss, mg Loss, mg

Block 1 Block 2 Block 1 Block 2 Cum. Ave.
Initial 16.4746 16.4069 o.o o.o o.o
After 4 h 16.4478 16.3809 26.8 26.0 26.4
After 8 h 16.4274 16.3613 20.4 19.6 46.4
After 12 h 16.4085 16.3427 18.9 18.6 65.2
After 16 h 16.3924 16.3266 16.1 16.1 81.3
9.8.4 Procedure
The wear block is weighed to 0.1 mg, after being scrubbed in detergent and
dried. The block is aligned in the holder placed in the slurry tray, and the recip-
rocating motion is started. At the end of the 2 h, the block is removed, washed,
and weighed, and the loss of metal recorded. This is repeated three times for a
total of 6 h. Duplicate samples are run simultaneously so as to provide a verifi-
cation and averaging.
9.8.5 Calculation of Results

The mass loss of the wear block or specimen, the average of two runs in a typical
slurry, is recorded. Por example, see Table 9.2.
The basic mathematical equation for a curve of the data listed in Table 9.1 is
Mass loss, mg =A X hours8 (9.3)
Using the least-squares method, the values of A and B are calculated for a best-fit
curve of the test data. In this case, the values were determined to be
A = 8.65 and B = 0.81
The equation for the example becomes
Mass loss, mg = 8.65 X hours0 · 81 = 15.17 mg

The Miller and SAR numbers are described as being related to the rate at which
the wear block loses mass at 2 h into the test, which can be calculated by using
the first derivative of equation (9.3) at 2 h. (This is the slope of the line tangent
to the curve at 2 h.)
Por example, for MLR = Mass loss rate, mg/h,
MLR, mg/h =A X B X 2B-t (9.4)

296 SLURRY PUMPING
So
MLR = 8.65 X 0.81 X 2º· 81 - 1 = 6.14 mgfh
9.8.6 Miller Number

It is desirable to have a meaningful whole number for the expression of abrasion-
corrosion effects. It was decided that the range of the whole number would range
from 1 for powdered sulfur to 1000 for 220 mesh alundum, and this is accom-
plished by the use of a scaling factor C determined to be 18.18. Using the mass
loss rate of 6.14 mg/h previously determined, the equation for the Miller number
can thus be written as
Miller number = C X MLR where e= 18.18 (9.5)
Then
Miller number = 18.18 X 6.14 = 111
A computer program listing (digital basic plus) for calculating results is published
as part of ASTM G-75-89, "Determination of Slurry Abrasivity by Miller Num-
ber," available from ASTM, 1916 Race St. Philadelphia, PA 19103.
9.8. 7 SAR Number

Because the SAR number is described as an index related to the rate at which the
specimen loses volume at 2 h into the test, the volume loss rate (VLR) can be
calculated by dividing the mass loss rate (MLR) by the specific gravity of the
specimen material. For example,
VLR, ml/h = M~R
where S = specific gravity of the specimen material.

For the example, again using the specific gravity of 27 % chrome iron and the
mass loss rate of 6.14 mg/h previously determined,
VLR = 6.14
7 _74 = 0.793 m1
/h
lt is desirable to have a meaningful whole number for the expression of the abra-
sion-corrosion effect of the slurry on the specimen. The range of these numbers is
arbitrarily selected to be 1-1000, the same range as for the Miller number for slurry
abrasivity. The equation for SAR number therefore includes the same constant C
(18.18) to force the SAR number for standard 27% chrome iron in standard sand
sluny to .be equal to the Miller number of a standard sand sluny as in the example
used here. It is also necessary to always multiply by a constant of 7. 74, the specific
gravity of standard 27 % chrome iron.
The equation for SAR number can then be written:
SAR number = Cs X VLR
where 18.18 X 7.74 = 140.7.

In the example, the SAR number for 27 % chrome iron in standard sluny is
SAR number = 140.7 x 0.793 = 111
9.8.8 A Miller Number System Overview
Even though the Miller number system uses a mbbing type of abrasion, experience
has shown that the results can be used to evaluate wear from impingement. The
abrasivity of a sluny is a function of the following characteristics of the solid
particles:
l. Hardness
2. Size
3. Shape
4. Size distribution
5. Friability
6. Concentration
probably in that order. Now all of these factors also contribute to metal attrition
by impingement as well as by rubbing. The only missing character not revealed in
the Miller number test is the effect of specific gravity or density. However, since
the force of impact is proportional to mv, it can be seen that the effect of mass is
rather small, particularly in most pumpable slurries where the particles are rela-
tively small.
The Miller number is of value in predicting pipe erosion in pipelines as dem-
onstrated from reported experience in a large coal (Miller number 10-20) pipeline
in Arizona as compared to a magnetite ( Miller number 67) pipeline in Tasmania.
The Miller number sysiem is usefül in predicting the life in a reciprocating
pump by correlating Miller numbers and parts life in existing sluny service to the
Miller number of a proposed sluny application.
The variation in Miller numbers in certain generic minerals such as coal can be
considerable. Coal, for instance, can have from 5 to 25% ash (the most abrasive
constituent), and even the type of ash can vary from soft calcareous to hard and
sharp quartz and pyrite. The same holds true for other numerals such as bauxite.
Table 9.3 lists the Miller numbers for a few generic materials.
TABLE 9.3. Typlcal Mlller Numbers•
Alundum 400 mesh 241

Alundum 200 mesh 1058
Aragonite 7
Ash 127
Ash, fly 83, 14
Bauxite 9, 33, 50, 76, 134
Calcium carbonate 14
Carbon 14, 16
Carborundum 220 mesh 1284
Clay 34, 36
Coal 6, 10, 21, 28, 47, 57
Copper concentrate 19, 37, 58, 68, 111, 128
Detergent 6, 8
Dust, blast furnace 57
Gilsonite 10
Gypsum 41
Iron ore (or concentrate) 28, 37, 64, 79, 122, 157,
234
Kaolin 7,30
Lignite 14
Limestone 22,30, 39,43,46
Limonite 113
Magnesium hydrate 4
Magnetite 64, 71, 134
Microsphorite 76
Mud, drilling 10
Nickel 31
Phosphate 68, 74, 84, 134
Potash 1, 2
Pyrite 194
Quartzite 99
Rutile 10
Salt brine 11
Sand and sand fill 51, 68, 85, 116, 138,
149, 246
Sea bottom 11
Shale 53, 59
Serpentine 134
Sewage, digested 15
Sewage, raw 25
Sodium sulfate 4
Soda ash tailings 27
Sulfur 1
Tailings (all types) 24,61, 91, 159,217,
480, 644
Tarsand 70
Waste, nickel 53
Waste, coal 22,28
ªNote the wide range of abrasivity of sorne types of materials, such
as coal or iron ore.
298
120
110
1 -
100
~~
-e-
"'
90
/
V-
~
~ 80
;;;¡
:: 70
/
~
~ 60
I
f 50
I
J
·¡;;
e
.o
40
<(
30
20
10
o 12.5 25 50
Solids concentration, percent by mass
Figure 9.14. Solids coilcentration vs. abrnsivity of 70 mesh um sand. Curve shows the
abrupt change in the reiationship of solids concentration to in the below
about 10-12 % solids.
9.8.9 Slurry Ccmcentration, Particle Size, and Particle Shape

The Miller number is meant to be used to compare the relative abrasivity of slurries
as caused by the solids that make up those slurries. Since that is the object, it can
be seen that certain standards had to be adopted so that the comparison would be
realistic. For this reason, an arbitrary concentration of 50% by weight for the
Miller nurnber test sample was chosen-partly because most slurries deal with
concentrations of that order and partly because the higher concentration reduces
the error of measurement.
Early in the development of the test, the question of concentration was consid-
ered and preliminary tests were run with variations. Figure 9 .14 shows that above
a certain value the concentration ofthe solids has little effect on the Miller number.
This can be readily understood when it is realized that one is looking at the effect
of particle size, shape, hardness, and distribution. These are factors that affect the
relative abrasivity of the solids in a slu:rry, and it is generally accepted that above
a certain low minimum concentration of solids, reciprocating pump parts life is
related not so much to concentration as to the other physical characteristics men-
tioned. For instance, the sand content of drilling rnud must be reduced to less than
2 % before an appreciable savings in pump parts life can be realized.
The Miller number test is inherently a ''high-abrasivity'' test as indicated
Figure 9.14. Note that the relative loss of abrasivity from the standard 50% test
concentration to 12 % test concentration is only about 32 %. However, the abrasiv-
ity then abruptly drops to zero at zero concentration. Therefore, it is logical to
300 SLURRY PUMPING
assume that a more realistic estimation of the erosional effects of concentrations

from zero to about 12 % should be based on the relation
Concentration percent (decimal) X 7 X Miller number ( 9. 7)
For example, a 6 % concentration of Miller number 110 material would have a

relative erosional effect of
0.06 X 7 X 110 = 46
Accordingly, it is useless to ron Miller number tests with low concentrations. In

the case of such slurries as mine water or mill water, it is desirable to ron a 50%
concentration of the dry solids and then apply the above correction.
9.8.1 O Oil-!Vlixed Slurries

OH-mixed slurries ron on the Miller number system exhibit a lower abrasivity than
the same solids in a water-mixed slurry. For example, 70 mesh sand run for l h
with 25 % chrome iron wear blocks showed the following results: With water-
mixed slurry, 13.4 mg loss; with oil-mixed (No. 6 fuel), 0.8 mg loss. In another
case, a ''spent'' industrial waste containing diatomaceous earth mixed with oil
showed no wear-block loss at the end of 4 h, but the same material washed in
solvent and remixed to the same concentration in water showed 4.2 mg loss.
Most slurries are water-mixed, and there is no doubt that the relative abrasivity
of other water-mixed slurries as determined by the Miller number system are valid.
However, there is sorne question as to how oil-mixed slurries would react in actual
pumping situations. Does the higher viscosity, ''lubricity,'' and corrosion protec-
tion of the oil actually contribute to a lower abrasivity with respect to pump parts
and pipeline wear? Until such a relationship can be further determined, all mate-
rials should be run in water-mixed slurries.
9.8.11 Corrosion
The effects of corrosion abrasion must be considered in the selection of materials
for pumps and other equipment for slurry pumping. There is no doubt that the
combination of corrosion and abrasion is much more severe in terms of metal loss
than either alone. The insidious cycle is that the products of corrosion, usually
protective, are rapidly removed by abrasion. This presents a "clean" surface to
further corrosion, thereby exaggerating the situation.
The 27 % chrome iron used for the Miller test wear block is in itself somewhat
corrosion-resistant, but in certain ores, particularly those containing copper, a great
deal of metal loss can be attributed to pitting corrosion, no doubt due to dissolved
air from agitation. However, such conditions can be expected in actual slurry
pumping, for the process of slurry mixing invites the absorption of air.
From one standpoint, the Miller test could be run without regard for corrosion,
but for practica! reasons it is thought best to try to separate the effects if possible.
Accordingly, it has been found that the effects of corrosion can be greatly inhibited
by a strong dose of NaOH, to a pH of about 13. If corrosion is suspected, it is
best to run two different samples, one as mixed and the other as inhibited. The
results will.give a clue as to the true abrasivity. For example, the Miller number
for different samples of a particular ore were as follows:
Sample pH Miller Number

No. 1 6.3 92
No. 2 1.3 553
No. 3 5.9 117
No. 3 NaOH inh. 13.+ 33
It is evident that the high weight losses in the first three cases are due to the classic
combination of corrosion and abrasion.
9.8.12 Corrosive Effect of Slurrles

Because of the synergistic interactions it is almost impossible to separate the
individual effects or to predict the "apparent abrasivity" of a slurry. Thus it is
imperative that laboratory testing be involved in order to evaluate any slurry
handling equipment or material for use in such slurry. Figure 9.3A.
The predictability of the "apparent abrasivity" of a slurry is further compli-
cated by an unexplainable phenomenon shown in Figure 10. 7 whereby the rate
of metal loss during undisturbed immersion or agitated erosion is completely re-
versed.
l . The ''true abrasivity'' of the solid particles is a dry and purely mechanical
function of the particle hardness, size shape and distribution-each' con-
tribution to purely "abrasive" wear.
2. Corrosive effects of the mixing liquid, many times being the result of
dissolved air-the usual slurry-mixing methods are notorious air mixers.
3. The effects of the product resulting from the solution of soluble elements
in the solid particles reacting with the mixing liquid as a solvent to form
a corrosive liquid.
4. The physical and chemical properties of the metal or material specimens,
combined with the slurry properties, that determine the characteristics of
the protective film of corrosion products that may form. Such film char-
acteristics can affect the abrasive wear-rate of the metal or material.
5. Electrolytic corrosion if dissimilar metals are involved.
302 SLURRY PUMPING
9.9 THE GOLD NUMBER FOR LOW ABRASIVITY
The original Miller Number concept was conceived as a method of comparing

the relative abrasivity of various pumpable slurries such as limestone, coa!, baux-
ite and various iron ore concentrates. The nature of minerals is such that they
may contain variations of "tramp" material of greater abrasivity-like silica ash
in coa!. In order to compare the relative abrasivity of such a great range of
materials, it was necessary to select a "standard" metal wear-block that would
exhibit reasonable rate-of-mass loss for the duration of the test. Research deter-
mined that 27% chrome iron would provide the desirable mass losses in materials
of the range of about 1 or 2 on Moh's scale, with a Miller Number of 1, to 1000
for 220 mesh alundum with a Moh's scale of 9. However, in dealing with ma-
terials in the range of Moh's scale of 2, or less. (tale, for instance) the-rate-of-
mass-loss of the 27% chrome iron blocks was not sufficient to give consistent
and reliable results.
There has always been a need to measure the relative abrasivity of less ag-
gressive materials such as polishes, cleaners, baby powder, tooth paste, hand
cleaners, etc.
While the Miller Number, ASTM F75-89, has served the slurry industry very
well with Numbers above about 10 (and 50 or below at which degree materials
are considered "non-abrasive") for slurry pipelining, the test is not sufficiently
sensitive to revea! accurate differences in materials below about Miller Number
10. Severa! years of research work in the matter of a "standard" block material
that would enable one to detect differences and extend the Miller Number range
into the decimal values has been somewhat frustrating. With many wear-block
materials in low abrasivity slurries, the corrosive effects, even in water, can
greatly overwhelm the abrasivity, giving misleading information. Passing years
of experience with the Miller Number has revealed that because of so many
factors affecting the Miller or SAR Number measurements (The Morrison-Miller
effect*), it could not be used to determine the reaction of other metals in different
slurries, or, different metals in the same slurry.
After running the gamut of seores of different "soft" metals or plastics, we
arrived at the conclusion that the perfect metal would be pure gold and many
tests have shown it to be ideal for the purpose. See Table 9.2A for results in
comparison with 27% chrome iron. Note the drastic non-relationship of 27
chrome iron to gold. Because the purity of gold affects it's specific gravity, the
standard source of gold is the Swiss Credit, an oblong block of metal from which
three wear-blocks can be cut. This is a .999 pure material with a specific gravity
of 18.4, which is used to calculate the Gold Miller Number on a mass loss basis.
*Note; The Morrison-Miller Effect( 1): Because of synergistic effects, the Abrasion-Corrosion re-
sponse of one metal in a specific slurry cannot be reliably used to predict the response of another
metal in that slurry or of the same metal in a different slurry. Accordingly, a specific test must be
run for the candidate metal and slurry.
(1) Paper WSRC-Rl>-89-511, William S. Morrison, Richard A Corbett: Corrosion Testing Labo-
ratories, lnc. Wilmingron, DE 19804. Charles F. Jenkind: Westinghouse Savannah River Laboratory,
Aiken, SC 29802.
9.10 A METHOD FOR LOCATING A 'PLUG' IN A SLURRY PIPELINE 303
TABLE 9.2A.
··----
MILLER GOLD RATIO
SO UDS MOHS SCALE U QUID NUMBER NUMBER G/M*
J. Tale water I .4 19 !4
S. Tale ¡ water !.6 i 14 71
CaCo3 3 water 2.2 116 53
Gun Cleaner water 4.l 394 96
Fine coa! oi! 5 479 96
Fine coa! water 17 751 44
Fine coa! condensate 2.6 926 356
Fly ash water 28 1328 47
Pum ice 7 water 34 2806 83
*Note wide difference in Ratio.
9.10 A METHOD FOR lOCATING A 'PlUG' IN A SLURRY

PIPELINE
It is almost inevitable that an accumulation of solids wil! sometime occur at a
certain point on a long slurry pipeline, such an accumulation turning into a
"plug" of such an extent as to cause permanent stoppage of ftow with further
pressuring in an attempt to move the plug, resulting in an increase of the tenacity
of the plug due to fürther pressure "wringing-out", through the plug's inherent
permeability, what little liquid was left in the plug.
The following method for precicely locating such a plug is offered;
Experience with operation of the pipeline will have revealed the fact that a
pressure wave generated at the station by a rather sudden build-up of pressure
following a sudden shut-down, will generate a pressure surge or wave that will
be reflected back to the station from any change in pipeline impedence resulting
from such parameters as an abrupt change in pipe diameter, open end or restric-
tion at the end of the pipeline, sharp-bend elbows, valve c!osure or a solids plug,
From observation of this recorded pressure wave on a time basis, the speed of
wave travel in that particular slurry would have been established. See Section
4.13, Surge Control in Water Systems.
By the use of such wave velocity study, the approximate location of the plug
can be determined from the relation
L t/2a 18)
Where;
L ft (m)
time, s
a - speed of pressure wave, ft/s (mis)
Once the approximate location is determined, the exact location of the plug can
be determined in the following manner: Select a location in the region of the
304 SLURRY PUMPING
Figure 9.15. Black Mesa pipeline pumps.
plug and, rather than making severa! time-consuming hot-taps for a pressure-
sensing device, simply apply a quick-cement strain gauge to the pipe or, better
still, clamp a special extensiometer (a clamp-band with a strain gauge attached)
and cabled up to a battery-operated bridge amplifier and strip-chart recorder. A
radio message to the station operator would ask him to generate a pressure build-
up by starting a pump. If the strain gauge is located beyond the plug, no pressure
rise will be detected-if on the pump side, the pressure rise will naturally be
detected. By repeating that procedure the exact location of the plug can be brack-
eted.
A suit-case sized kit, containing the strain gauge equipment or extensiometer,
a bridge amplifier and recorder and necessary cables can be assembled or pur-
chased.
9.11 BLACK MESA PIPELINE
The Black Mesa pipeline (Fig. 9 .15), one of the world' s longest and largest coal
pipelines, has been in operation for over 12 years. The capacity is about 4.5 mil-
lion tons per year of solids. Four pump stations on the lineare used to pump the
coal-water slurry through an 18-in line from a mine located at Black Mesa in
northeastem Arizona for a distance of 275 mi to a power-generating station on the
Colorado River west of Kingman.
All main pumps are 18-in stroke, double-acting piston-type rated at 1700 BHP.
A total of 13 pumps are used. The pumps are assigned to four stations, with three
pumps at each of three stations and four at the other.
Each pump is driven by an ac induction motor through a variable-speed fluid
9.13 SLURRY TABLES 305
Figure 9.16. Savage River Mines pipeline pumps.
drive anda reduction gear directly coupled to the pinion shaft ofthe pump. Normal
operation is 65 RPM, with one pump at each station on standby. Complete auto-
mation of the line through microwave radio is used.
9.12 SAVAGE RIVER MINES

The long-distance pumping of heavier materials such as iron ore concentrates is
admittedly more difficult than that of coal and similar light materials. However, a
54-mi pipeline was placed in service in Tasmania, Australia, in October 1967,
pumping iron ore concentrate, and it has proveo highly successful. This line has
a capacity of 2.5 million tons per year.
The actual installation is a single pumping station comprising four triplex
plunger pumps (Fig. 9.16), pumping iron ore sluny of 1.92 specific gravity through
a 9-in pipeline 54 mi long. The sluny is 60% magnetite by weight with 85% minus
200 mesh particle size. The four pumps are electrically driven, two at 175 RPM
fixed speed and two with variable-speed liquid coupling drive at 173 RPM maxi-
mum speed. All pumps have 5!-in diameter boron-alloy-coated sleeve-type plung-
ers and are rated 560 BHP input.
9.13 SLURRY TABLES

Tables 9 .4 through 9 .12 list the important properties of solids and slurries and how
to calculate the specific gravity of slurries. A set of precalculated data is also
included.
TABLE 9.4. Specific Gravity and Hardness of Minerals
Spec. Gr. Hardness

Albite 2.6 6-6.5
Anhydrite 2.9 3-3.5
Apatite 3.2 2-3
Aragonite 2.9 3.5-4
Asbestos 2.5 1-2.5
Azurite 3.8 3.5-4
Barite 4.5 2.5-3
Bauxite 2.5 1-3
Becyl 2.7 7.5-8
Calcite 2.7 3
Chalcopyrite 4.2 3.5-4
Coal 1.3 2-2.5
Corundum 4.0 7-9
Cyanite 3.6 5-7.3
Dolomite 2.9 3.5-4
Feldspar 2.7 6-6.5
Fluorite 3.1 4
Galena 7.5 2.5
Gamet 4.0 6.5-7.5
Gypsum 2.3 1.5-2
Hematite 5.1 5.5-6.5
Ilmenite 4.7 5-6
Kaolin 2.6 1-2.5
Lignite 1.3 2-2.5
Limes tone 2.7 3
Limonite 3.8 5-5.5
Magnetite 5.2 5.5-6.5
Olivine 3.3 6.5-7
Phosphate 3.2 2-3
Potash 2.7 3.5-4
Pyrite 5.0 6-6.5
Quartz 2.7 7
Rutile 4.2 6-7
Sulfur 2.0 1.5-2.5
TABLE 9.5. Screen Size
U.S. No. Mesh Opening, in Opening, 1-tm

4 4 .187 4760
5 5 .157 4000
6 6 .132 3360
7 7 .111 2830
8 8 .0937 2380
10 9 .0787 2000
12 10 .0661 1680
14 12 .0555 1410
16 14 .0469 1190
18 16 .0394 1000
20 20 .0331 841
25 24 .0278 707
30 28 .0234 595
35 32 .0197 500
40 35 .0165 420
45 42 .0139 354
50 48 .0117 297
60 60 .0098 250
70 65 .0083 210
80 80 .0070 177
100 100 .0059 149
120 115 .0049 125
140 150 .0041 105
170 170 .0035 88
200 200 .0029 74
230 250 .0025 63
270 270 .0021 53
325 325 .0017 44
400 400 .0015 38
TABLE 9.6. Relatlve Size of Mlcronic Particles
Relative Sizes
Lower limit of visibility (naked eye) 40 microns
White blood cells 25 microns
Red blood cells 8 microns
Bacteria (cocci) 2 microns
Linear Equivalents
1 micron = 1 micrometer (1-tm)
1 inch (in.) = 25.4 mm = 25,400 1-tm
1 millimeter (mm) = 0.0394 in = 1000 1-tm
1 micrometer (1-tm) = 1 /25,400 in = 0.001 mm
1 micrometer (1-tm) = 3.94 X 10- 5 in = 0.000039 in.
307
Rhombohedron stacking of spherical particles
Figure 9.17 Bulk properties of Mineral solids.
A-Principie solid particle material. (i.e. Coal)

B-Porosity, as percent of volume of particle (may be occluded air, gas or connate water).
C-Voids, between particles as percent of bulk volume as function of size, shape, size
distribution and stacking mode.
O-Water-soluble material.
E-"Tramp" material such as ash in coal and silica in limestone.
BULK VOLUME-Volume of container, ft3.
BULK DENSITY-Weight (mass) of Bulk Volume, lb/ft3.
BULK SPECIFIC GRAVITY-Bulk Density/62.3.
APPARENT SPECIFIC GRAVITY-Density of solids, including A, B, D and E, com-
pared to Water.
TRUE SPECIFIC GRAVITY-Density of Solids A compared to Water.
Void percentage is independent of particle diameter. One cubic foot of baseballs or marbles
has the same void percentage; 47.6% for cubic stacking and 25.9% for rhombohedrol
stacking. Random particle size solids result in much lower void percentage simply because
the smaller particles tend to fill the voids between larger particles.
308
TABLE 9.8. Calculation of Specifü: Gravity of a Slurry of one Solid
l. Water-Mixed ll. Other Liquids
S = _ _ _ _S_1- - - (9.11)
Sm = 1 - Cw(: - 1/Ss) ( 9 .S) m l - Cw(l - S¡/Ss)
C = CvSs ( 9 .9 ) (9.12)
w sm
C = Sm - l ( 9 .10) C = Sm - S1
(9.13)
" SS - 1 " S,, - S1
where
Cv = volume fraction of solids

Cw = weight fraction of solids
sm = specific gravity of mixture (sluny)
Ss = specific gravity of solids
S1 = specific gravity of Hquid
TABLE 9.9. Calculafü:m of Specifü: Gravity of a Slurry of a Mixture of Solids
l. lf Given in Weight Fraction
sm = -----------
wt. frac. A wt. frac. B
(9.15)
SA + SB +
EXAMPLE
0.4 wt. frac. A SA = 5.0

0.3 wt. frac. B SB = 3.0
0.3 wt. frac. C Se = 4.0
sm = 0.4 0.3 0.3 = 3.92 (9.16)

-+-+-
5.0 3.0 4.0
ll. lf Given in Volume Fraction

Sm = vol. frac. A X SA + vol. frac. B X SB + 17)
EXAMPLE
0.4 vol. frac. A SA = 5.0
0.3 vol. frac. B SB = 3.0
0.3 vol. frac. e = 4.0
Sm = 0.4 X 5.0 + 0.3 X 3.0 + 0.3 X 4.0 = 4.1

309
w
....
o
TABLE 9.10. Pipeline GPM per Mllllon Short Tons per Year of Sollds
Solids Percent Concentration by Weight with Water

Solids
Sp. Gr. 30 35 40 45 50 55 60 65 70 75
1.1 1,482 1,264 1,101 974 872 789 720 662 611 568
1.2 1,447 1,230 1,066 939 838 755 686 627 577 533
1.3 1,418 1,200 1,037 910 809 725 656 598 547 504
1.4 1,393 1,175 1,012 885 783 700 631 573 522 479
1.5 1,371 1,153 990 863 762 679 609 551 501 457
1.6 1,352 1,134 971 844 743 660 590 532 481 438
1.7 1,335 1, 118 954 827 726 643 573 515 465 421
1.8 1,320 1,103 939 812 711 628 559 500 450 406
1.9 1,307 1,089 926 799 698 614 545 487 436 393
2.0 1,295 1,077 914 787 686 602 533 475 424 381
2.1 1,284 1,066 903 776 675 592 522 464 413 370
2.2 1,274 1,056 893 766 665 582 512 454 404 360
2.3 1,265 1,047 884 757 656 573 503 445 395 351
2.4 1,257 1,039 876 749 647 564 495 436 386 343
2.5 1,249 1,032 868 741 640 557 487 429 379 335
2.6 1,242 1,024 861 734 633 550 480 422 372 328
2.7 1,236 1,018 855 728 626 543 474 415 365 322
2.8 1,230 1,012 849 722 620 537 468 409 359 316
2.9 1,224 1,006 843 716 615 531 462 404 353 310
3.0 1,219 1,001 838 711 609 526 457 398 348 305
3.1 1,214 996 833 706 604 521 452 393 343 300
3.2 1,209 992 828 701 600 517 447 389 339 295
3.3 1,205 987 824 697 595 512 443 385 334 291
3.4 1,201 983 820 693 591 508 439 380 330 287
3.5 1,197 979 816 689 588 504 435 377 326 283
3.6 1,193 976 812 686 584 501 432 373 323 279
3.7 1,190 972 809 682 581 497 428 370 319 276
3.8 1,187 969 806 679 577 494 425 366 316 273
3.9 1,184 966 803 676 574 491 422 363 313 270
4.0 l, 181 963 800 673 571 488 419 360 310 267
4.1 1,178 960 797 670 568 485 416 358 307 264
4.2 1,175 958 794 667 566 483 413 355 305 261
4.3 l,173 955 792 665 563 480 411 352 302 259
4.4 1,170 953 789 662 561 478 409 350 300 256
4.5 1,168 950 787 660 559 475 406 348 297 254
4.6 1,166 948 785 658 556 473 404 345 295 252
4.7 1,164 946 783 656 554 471 402 343 293 250
4.8 1,162 944 781 654 552 469 400 341 291 248
4.9 1,160 942 779 652 550 467 398 339 289 246
5.0 1,158 940 777 650 548 465 396 337 287 244
5.1 1,156 938 775 648 547 464 394 336 285 242
5.2 1,154 937 773 646 545 462 393 334 284 240
5.3 l, 153 935 772 645 543 460 391 332 282 239
5.4 1, 151 933 770 643 542 459 389 331 280 237
5.5 1,149 932 769 642 540 457 388 329 279 235
5.6 1,148 930 767 640 539 456 386 328 277 234
507 l ,147 929 766 639 537 454 385 326 276 233
5.8 l,145 928 764 637 536 453 383 325 275 231
5.9 1,144 926 763 636 534 451 382 324 273 230
6.0 1,143 925 762 635 533 450 381 322 272 229
-
For water-mixed slurries only:
457
GPM* = - - , = U .S. GPM of slun-y for each million short tons per year of solidsº
CwSm
......
w
Use pump volumetric efficiency to calculate pump size.
w
....
NI
TABLE 9.11. Fraction Volume of Solids in Slurry Mixture, c.

Solids
Sp. Gr. 30 35 40 45 50 55 60 65 70 75
1.1 0.280 0.329 0.377 0.427 0.476 0.526 0.577 0.628 0.680 0.732
1.2 0.263 0.310 0.357 0.405 0.455 0.505 0.556 0.607 0.660 0.714
1.3 0.248 0.293 0.339 0.386 0.435 0.485 0.536 0.588 0.642 0.698
1.4 0.234 0.278 0.323 0.369 0.417 0.466 0.517 0.570 0.625 0.682
1.5 0.222 0.264 0.308 0.353 0.400 0.449 0.500 0.553 0.609 0.667
1.6 0.211 0.252 0.294 0.338 0.385 0.433 0.484 0.537 0.593 0.652
1.7 0.201 0.241 0.282 0.325 0.370 0.418 0.469 0.522 0.579 0.638
1.8 0.192 0.230 0.270 0.313 0.357 0.404 0.455 0.508 0.565 0.625
1.9 0.184 0.221 0.260 0.301 0.345 0.391 0.441 0.494 0.55J 0.612
2.0 0.176 0.212 0.250 0.290 0.333 0.379 0.429 0.481 0.538 0.600
2.1 0.169 0.204 0.241 0.280 0.323 0.368 0.411 0.469 0.526 0.588
2.2 0.163 0.197 0.233 0.271 0.313 0.357 0.405 0.458 0.515 0.577
2.3 0.157 0.190 0.225 0.262 0.303 0.347 0.395 0.447 0.504 0.566
2.4 0.152 0.183 0.217 0.254 0.294 0.337 0.385 0.436 0.493 0.556
2.5 0.146 0.177 0.211 0.247 0.286 0.328 0.375 0.426 0.483 0.545
2.6 0.142 0.172 0.204 0.239 0.278 0.320 0.366 0.417 0.473 0.536
2.7 0.137 0.166 0.198 0.233 0.270 0.312 0.357 0.408 0.464 0.526
2.8 0.133 0.161 0.192 0.226 0.263 0.304 0.349 0.399 0.455 0.517
2.9 0.129 0.157 0.187 0.220 0.256 0.296 0.341 0.390 0.446 0.508
3.0 0.125 0.152 0.182 0.214 0.250 0.289 0.333 0.382 0.438 0.50Ó
3.1 0.121 0.148 0.177 0.209 0.244 0.283 0.326 0.375 0.429 0.492
3.2 0.118 0.144 0.172 0.204 0.238 0.276 0.319 0.367 0.422 0.484
3.3 0.115 0.140 0.168 0.199 0.233 0.270 0.313 0.360 0.414 0.476
3.4 0.112 0.137 0.164 0.194 0.227 0.264 0.306 0.353 0.407 0.469
3.5 0.109 0.133 0.160 0.189 0.222 0.259 0.300 0.347 0.400 0.462
3.6 0.106 0.130 0.156 0.185 0.217 0.253 0.294 0.340 0.393 0.455
3.7 0.104 0.127 0.153 0.181 0.213 0.248 0.288 0.334 0.387 0.448
3.8 0.101 0.124 0.149 0.177 0.208 0.243 0.283 0.328 0.380 0.441
3.9 0.099 0.121 0.146 0.173 0.204 0.239 0.278 0.323 0.374 0.435
4.0 0.097 0.119 0.143 0.170 0.200 0.234 0.273 0.317 0.368 0.429
4.1 0.095 0.116 0.140 0.166 0.196 0.230 0.268 0.312 0.363 0.423
4.2 0.093 0.114 0.137 0.163 0.192 0.225 0.263 0.307 0.357 0.417
4.3 0.091 0.111 0.134 0.160 0.189 0.221 0.259 0.302 0.352 0.411
4.4 0.089 0.109 0.132 0.157 0.185 0.217 0.254 0.297 0.347 0.405
4.5 0.087 0.107 0.129 0.154 0.182 0.214 0.250 0.292 0.341 0.400
4.6 0.085 0.105 0.127 0.151 0.179 0.210 0.246 0.288 0.337 0.395
4.7 0.084 0.103 0.124 0.148 0.175 0.206 0.242 0.283 0.332 0.390
4.8 0.082 . 0.101 0.122 0.146 0.172 0.203 0.238 0.279 0.327 0.385
4.9 0.080 0.099 0.120 0.143 0.169 0.200 0.234 0.275 0.323 0.380
5.0 0.079 0.097 0.118 0.141 0.167 0.196 0.231 0.271 0.318 0.375
5.1 0.078 0.095 0.116 0.138 0.164 0.193 0.227 0.267 0.314 0.370
5.2 0.076 0.094 0.114 0.136 0.161 0.190 0.224 0.263 0.310 0.366
5.3 0.075 0.092 0.112 0.134 0.159 0.187 0.221 0.259 0.306 0.361
5.4 0.074 0.091 0.110 0.132 0.156 0.185 0.217 0.256 0.302 0.357
5.5 0.072 0.089 0.108 0.129 0.154 0.182 0.214 0.252 0.298 0.353
5.6 0.071 0.088 0.106 0.127 0.152 0.179 0.211 0.249 0.294 0.349
5.7 0.070 0.08(í 0.105 0.126 0.149 0.177 0.208 0.246 0.290 0.345
5.8 0.069 0.085 0.103 0.124 0.147 0.174 0.205 0.243 0.287 0.341
5.9 0.068 0.084 0.102 0.122 0.145 0.172 0.203 0.239 0.283 0.337
6.0 0.067 0.082 0.100 0.120 0.143 0.169 0.200 0.236 0.280 0.333
w
....
w
w
....
....
TABLE 9.12. Specific Gravity of Slurry Mixture, Sm

Solids
Sp. Gr. 30 35 40 45 50 55 60 65 70 75
1.1 1.028 1.033 1.038 1.043 1.048 1.053 1.058 1.063 1.068 1.073
1.2 1.053 1.062 1.071 1.081 1.091 1.101 1.111 1.121 1.132 1.143
1.3 1.074 1.088 1.102 1.116 1.130 1.145 1.161 1.176 1.193 1.209
1.4 1.094 1.111 1.129 1.148 1.167 1.186 1.207 1.228 1.250 1.273
1.5 1.111 1.132 1.154 1.176 1.200 1.224 1.250 1.277 1.304 1.333
1.6 1.127 1.151 1.176 1.203 1.231 1.260 1.290 1.322 1.356 1.391
1.7 1.141 1.168 1.197 1.227 1.259 1.293 1.328 1.365 1.405 1.447
1.8 1.154 1.184 1.216 1.250 1.286 1.324 1.364 1.406 1.452 1.500
1.9 1.166 1.199 1.234 1.271 1.310 1.352 1.397 1.445 1.496 1.551
2.0 1.176 1.212 1.250 1.290 1.333 1.379 1.429 1.481 1.538 1.600
2.1 1.186 1.224 1.265 1.308 1.355 1.405 1.458 1.516 1.579 1.647
2.2 1.196 1.236 1.279 1.325 1.375 1.429 1.486 1.549 1.618 1.692
2.3 1.204 1.247 1.292 1.341 1.394 1.451 1.513 1.581 1.655 1.736
2.4 1.212 1.257 1.304 1.356 1.412 1.472 1.538 1.611 1.690 1.778
2.5 1.220 1.266 1.316 1.370 1.429 1.493 1.563 1.639 1.724 1.818
2.6 1.226 1.275 1.327 1.383 1.444 1.512 1.585 1.667 1.757 1.857
2.7 1.233 1.283 1.337 1.395 1.459 1.530 1.607 1.693 1.788 1.895
2.8 1.239 1.290 1.346 1.407 1.474 1.547 1.628 1.718 1.818 1.931
2.9 1.245 1.298 1.355 1.418 1.487 1.563 1.648 1.742 1.847 1.966
3.0 1.250 1.304 1.364 1.429 1.500 1.579 1.667 1.765 1.875 2.000
3.1 1.255 1.311 1.372 1.439 1.512 1.594 1.685 1.787 1.902 2.033
3.2 1.260 1.317 1.379 1.448 1.524 1.608 1.702 1.808 1.928 2.065
3.3 1.264 1.323 1.387 1.457 1.535 1.622 1.719 1.828 1.953 2.095
3.4 1.269 1.328 1.393 1.466 1.545 1.635 1.735 1.848 1.977 2.125
3.5 1.273 1.333 1.400 1.474 1.556 1.647 1.750 1.867 2.000 2.154
3.6 1.277 1.338 1.406 1.481 1.565 1.659 1.765 1.885 2.022 2.182
3.7 1.280 1.343 1.412 1.489 1.574 1.670 1.779 1.902 2.044 2.209
3.8 1.284 1.348 1.418 1.496 1.583 1.681 1.792 1.919 2.065 2.235
3.9 1.287 1.352 1.423 1.503 1.592 1.692 1.806 1.935 2.086 2.261
4.0 1.290 1.356 1.429 1.509 1.600 1.702 1.818 1.951 2.105 2.286
4.1 1.293 1.360 1.434 1.516 1.608 1.712 1.830 1.966 2.124 2.310
4.2 1.296 1.364 1.438 1.522 1.615 1.721 1.842 1.981 2.143 2.333
4.3 1.299 1.367 1.443 1.528 1.623 1.730 1.853 1.995 2.161 2.356
4.4 1.302 1.371 1.447 1.533 1.630 1.739 1.864 2.009 2.178 2.378
4.5 1.304 1.374 1.452 1.538 1.636 1.748 1.875 2.022 2.195 2.400
4.6 1.307 1.377 1.456 1.544 1.643 1.756 1.885 2.035 2.212 2.421
4.7 1.309 1.380 1.460 1.549 1.649 1.764 1.895 2.048 2.227 2.442
4.8 1.311 1.383 1·.463 1.553 1.655 1.771 1.905 2.060 2.243 2.462
4.9 1.314 1.386 1.467 1.558 1.661 1.779 1.914 2.072 2.258 2.481
5.0 1.316 1.389 1.471 1.563 1.667 1.786 1.923 2.083 2.273 2.500
5.1 1.318 1.392 1.474 1.567 1.672 1.793 1.932 2.094 2.287 2.519
5.2 1.320 1.394 1.477 1.571 1.677 1.799 1.940 2.105 2.301 2.537
5.3 1.322 1.397 1.480 1.575 1.683 1.806 1.949 2.116 2.314 2.554
5.4 1.324 1.399 1.484 1.579 1.688 1.812 1.957 2.126 2.328 2.571
5.5 1.325 1.401 1.486 1.583 1.692 1.818 1.964 2.136 2.340 2.588
5.6 1.327 1.404 1.489 1.586 1.697 1.824 1.972 2.146 2.353 2.605
5.7 1.329 1.406 1.492 1.590 1.701 1.830 1.979 2.155 2.365 2.621
5.8 1.330 1.408 1.495 1.593 1.706 1.835 1.986 2.164 2.377 2.636
5.9 1.332 1.410 1.497 1.597 1.710 1.841 1.993 2.173 2.389 2.652
6.0 1.333 1.412 1.500 1.600 1.714 1.846 2.000 2.182 2.400 2.667
...w
UI
1
PARTS WEAR ANO LIFE
10.1 THE MECHANICS OF WEAR IN PUMPS
10.1.1 lntroducUon
The word wear has many connotations, but in the pumping industry it means the
gradual deterioration of any part in the system to the point of danger or uselessness.
The word has no quantitative meaning, and a value must be applied before it can
be included in any discussion. The obvious approach is to apply a mass loss rate.
Then the related term life can be derived from the time function.
There are times when the point of uselessness or failure is reached without mass
loss, as in the case of fatigue failure. Nevertheless, for the work at hand, the point
of uselessness, reached by whatever means, determines the life of that part.
10.1.2 Wear Modes

An extraordinary number of wear modes may be encountered in the pumping of
liquids, and a list and description of them would be of great interest. Twelve of
the most common are listed below and shown in Figure 10. l. Note that most of
these basic modes can be classified into submodes, according to whether they are
affected by the following factors: reciprocating, oscillating, circular, or continuous
motion; dry, oil-wet, or water-wet; and ratio of areas of mating parts.
Mode A. Adhesiva Wear. Usually metal-to-metal rubbing in any of the stated

motions and with oil lubrication. For example, plain joumal bearings, crosshead
pin bushings, crosshead shoes and guides.
316
10.1 THE MECHANICS OF WEAR IN PUMPS 317
Al Applied
load
A2 A3 1
"""'~~~~~~LL.¿LL.; - - - - - ~~1-~~
Adhesive wear Adhesive wear 1 Adhesive wear
Wear areas equal Wear areas unequal 1 Wear areas unequal
Material posilion elected Material position reversed
82 83
- .J.=
Motion:
Reciprocating
-oscillating
-continuous
/'""/.'.'.~'' 7 7 ? 7 )
'77TT7:'7/"'
///,:;, //'.
Abrasive wear Abrasive wear Abrasive wear

Wear areas equal Wear areas unequal Wear areas unequal
Material position elected Material position reversed
Cl
~
rr~~"""~-'-k'f-4~~ ~
Polyrner
with
Seo u ri ng wea r embedded Seo u ri ng wea r Scouring wear
Wear areas equal abrasive Wear areas unequal Wear areas unequal
particles Material position elected Material position reversed
Figure 10.1. Wear modes.
Mode B. Abrasiva Wear. Same as mode A except in the presence of abrasive

solids or particles. Figure 10.2 shows abrasive particles of various sizes and shapes.
Mode C. Scouring Wear. Elastomer-to-metal rubbing with abrasive particles

embedded in the elastomer. For example, rubber pistons, packing, and valve in-
serts.
Mode D. Abrasion-Corrosion Wear. Any metal-to-metal in the pres-

ence of water-mixed slurries. Aside from high-velocity erosion, this is the most
severe and -destructive mode encountered in slurry pumping. Parts involved are
metal-to-metal valves and seats (upon each closure), metal piston parts rubbing on
metal liners and plungers, or pistan rods rubbing against metal stuffing box parts
or trim. Wear in this mode also accelerates scouring wear. Normal flow in pipe-
lines also causes this type of wear.
Mode E. Crushing and Grindíng. In abmsive metal-to-metal contact. For ex-

ample, a valve repeatedly ciosing with great force against solid partides trapped
between the valve and seat at closure.
318 PARTS WEAR ANO LIFE
D E F
Abrasion-corrosion
Crushing and grinding High-velocity eros ion
G H
..- -·-_,_.. -
~~~~~~ Pipewall
- • --.• -.::;;• !-
.
~~~~~ Pipewall
-·;:::;·~·
~
--
~~~ ;~·~-
Low-velocity erosion · Sa ltation eros ion Corros ion
L~lor
J K
roller
beari
~
_. _- , ;o-
0
_ ' _
1,
, O _'
Collapsing
vapor
· - bubbles Fatigue and
Cavitation corrosion fatigue Fretting wear
Figure 10.1. (Continued)
Mode F. High-Velocity Erosion. Forexample, when a valve seal orpiston rub-

ber is worn to the extent that there is a slight leak, the extremely high velocity of
the leaking slurry results in catastrophic failure of both mating parts. Such a con-
dition existing for a matter of minutes and not attended to will result in washout
of the liquid end of the pump, as shown in the photographs of Figures 10.3 and
10.4.
Mode G. Low-Ve/ocity Erosion. This is a slow rate of erosion that takes place
in the pipeline or in any passage where near-normal velocities are maintained.
Mode H. Saltation Wear. Comes about in pipelines handling unstable slurries

with larger than usual particles, which tumble along the bottom of the pipe. An
example is in the transportation of phosphate rock from pit to processing plant,
where the pipe wears rapidly on the bottom and requires frequent tuming to dis-
tribute wear.
(a) (b)
(e) (d)
Figure 10.2. Abrasive particles. Note 1 mm division. (a) Bunker Hill sand, Miller number
218. (b) Saskatchewan sand, Miller number 149. (e) Los Angeles sewage, Miller number
77. (d) 50-70 "standard" test sand, Miller number 136.
F - HIGH VELOCITY
EROS ION
B - ABRASIVE WEAR
E - CRUSHING WEAR
I - CORROSION WEAR
Figure 10.3. Typical wom and washed out sluny valve.
319
320 PARTS WEAR AND LIFE
G - HIGH VELOCITY
EROS ION
C - SCOURING WEAR
E - CRUSHING WEAR
B - ABRASIVE WEAR
Figure 10.4. Typical wom and washed-out sluny valve seat. (*Note washed-out liquid
end.)
Mode /. Corrosion. Listed as a mode of wear since it can still contribute to

metal loss in the absence of abrasive materials.
Mode J. Cavitation. Can result in damage to the metal in the liquid end or other
parts of a reciprocating pump through the microscopic but intense liquid blasts
against the metal near the cavitation area following the collapse of the vapor bub-
bles.
Mode K. Fatigue. While not a true ''wear'' mode, fatigue does result in early
failure of pump parts, particularly the liquid end.
Mode L. Fretting Wear. A minor cause of failure, particularly in roller and ball
bearings and other loose-fitting parts that tend to generate a form of chafing, as
between the rollers and the mating cup or cone when the bearing is loaded but not
rotating.
10.1.3 Effects of Wear

Sorne of the aspects of wear of pump parts have been briefly covered. Most im-
portant, the regular replacement of these wom parts over the life of a pipeline can
often amount to a much greater cost than the first cost of the pumps and becomes
a significant consideration in the economics of any project. See Figure 10.5.
Of the modes of wear just listed, mode D, abrasion-corrosion, is most destruc-
tive. This type of wear can be encountered with many "clean" liquids. For ex-
ample, most clean and even potable waters originating from rivers, lakes, canals,
c::::J Incremental power cost

~ Partscost
~ Dampener cost
P:::'''''''''::':''i':J Pump cost
2.0
1.0
4 5 6
Number of pumps per station
Figure 10.5. Number of pumps far a given station. Costs based on 20-yr life of project.
(Courtesy Worthington Pump Division-Dresser Industries)
aqueducts, and wells often accumulate and carry dissolved air, airbome dust, and
minute amounts of particles of sedim~nt, usually silica sand, which have a ten-
dency to become concentrated in the close clearances of parts and in the elasto-
meric (rubber, leather, etc.) parts of the pump.
In this respect, the effect of solids concentration on abrasivity should be rec-
ognized. Figure 10.6 shows that abrasivity increases very rapidly from zero to
about 10% with an increase in the concentration of abrasive particles. It is inter-
esting to note that in the oil-well drilling industry, drilling mud is in itself not
abrasive-it actually inherits its bad abrasivity reputation from the particles picked
up from the geologic formation and recirculated through the mud pump. Various
"desanding" methods are employed, such as screens, cyclones, centrifuges, and
settling, and it is significant that the percentage of sand must be reduced to below
2 % befare effective life-of-parts improvement can be realized.
Of course, in slurry pumping it is obvious that one is dealing with the insidious
contribution of high concentration and corrosion. It is evident that the combination
of abrasion and corrosion results in greater destruction than the sum of the indi-
vidual effects. This phenomenon is shown in Figure 10. 7 where the introduction
of agitation that continually removes the protective products of corrosion results
in a fantastic increase in metal loss rate and reverses the effects of pH for sorne
unexplained reason.
120
~
110
-----
~
100
90
l./"
-;::-
CI)
..Q 80
/
E
:::J
e:
.... 70 /
~
60
I
! I
·f
·¡¡;
50 1
E 40
..Q
<
30
20
10
o 12.5
-
25 50
Solids concentration, percent by mass
Figure 10.6. Solids concentration vs. abrasivity for 70 mesh um sand. Showing the abrupt
change in the relationship of solids concentration to abrasivity in the region below about
10 to 12% solids.
400
>
D.
::¡; 300
~e:
o
·¡¡;
e.... 200
o
(.)
7
pH of distilled water
Figure 10.7. Effect of pH of distilled water on erosion-corrosion of carbon steel at SOºC

(velocity, 39 FPS). A, erosion-corrosion disc. B, specimens immersed in tank, (From
M. G. Fontana and N. D. Greene, "Corrosion Engineering," McGraw-Hill, New York
1967. Copyright© 1967 McGraw-Hill.)
322
10. 1 THE MECHANICS OF WEAR IN PUMPS 323
. Sr1aded area-Liner wear

_)Metal (piston body)
\ ~-
Piston
! Piston
rod 6"
Figure 10.8. Wom liner profile (opposite sides) showing "end-of-stroke scmbbing wear."
This typical liner wear pattern in a single-acting pump is explained: The stored energy of
high-pressure compression of the elastomer parts of the piston rubber is suddenly reduced
to near-zero pressure at the end of the pressure stroke. With abrasive material trnpped be-
tween the rubber and the liner wall, this sudden (explosive) change of shape of the
to its natural shape generates a greater rate of wear for that instant than at any other position
of the piston in the liner. Note that even at the midstroke of the piston, where maximum
velocity is seen, there is no tendency for greater wear as one would expect. This bears out
the contention that number of reversals is the greatest cause of wear.
10.1.4 EFFECT OF PUMP STROKE REVERSAl RATE

A reciprocating pump and piston rod follows a more or less distorted sine
wave shape of instantaneous velocity by the crank mechanism,
the maximum velocity occurring about mid-stroke but with zero velocity at each
end-of-stroke. One would expect maximum liner and rod wear to take place at
such mid-stroke. However, many years of observation shows that maximum wear
takes place at each end of a pressure stroke. This can be explained from an
analysis of the action of any sealing element, such as resilient pistons and rod
packing or even metallic ring-type sea.Is. Sealing must rely on a pressure-gener-
ated distortion of the sealing elements in order to seal the gap between the mov-
ing parts and this distortion occurs in a rather short period of time (high energy)
at each end of the stroke, with the sudden application of discharge pressure.
The violent generation of such distortion at one end-of-stroke and return from
such distortion at the other end-of-stroke, results in a concentrated wearing con-
dition between the mating parts, being exacerbated the presence of abrasive
material as in the case of drilling-mud and slurries. In an interna! combustion
engine the effects are also shown in the "shoulder" on the cylinder wall of a
worn engine at the top of the piston stroke where there is zero piston velocity
but with explosive forces acting on the rings in an extremely short instant.
See Figure 10.8 for a typical mud-pump liner wear profile.
Figure 10.9. Wom slurry pistori.
"Relaxed" piston High

atinstallation pressure
Piston motion _ .
Zero
· pressure
-+-- Piston motion
Liner
t::::=::===========~::::::::~::j wear
Figure 10.10. Liner wear at reversa!. (a) Liner. (b) Deformed piston, caused by pressure
and friction. (e) Scouring of liner by violent retum to natural shape of piston.
Also a critical examination of the pump sequence of operation will revea! that
the worst moment in the life of a valve in a pump cycle is that instant when the
valve doses, again at the end-of-stroke of the piston or at the moment of rever-
sa!. The valve at these moments is subject to every wear mode known: Abrasion,
Crushing, Veiocity erosion, Scouring, etc. Once the valve is closed and sealed
there is no effective wear for the entire remaining time of a stroke when the
valve opens at end-of-stroke. An exception is the case of a rapid (a matter of
10,1 THE MECHANICS OF WEAR IN PUMPS 325
High Zero
c.
pressure pressure
,l. J
Jlilll
11------'j
Rod wear
(a)
ol Jlllll
(b)
Figure 10.11. Piston rod wear at reversal. (a) Sealing ring deformed by pressure and fric-
tion. (b) The sudden retum to original shape tends to scour the metal rod.
minutes) complete deterioration of a valve and seat with the high velocity of a
smaH leak rapidly cutting away nearby metal. Valve guiding requires considerable
clearance for proper action and such clearance allows a valve to seat in a slightly
'cocked' or canted position in its seat. Then when the high discharge pressure
applies a high load to the top of a beveled valve, it forces the closed valve to
slide a short distance as it seeks its natural mating position with its seat-all a
severe wearing mechanism at each closure and ali at the end-of-stroke instant.
So again, valve wear occurs at the end-of-stroke, or at reversa!, exacerbated by
the presence of abrasive materiel such as drilling mud and slurry.
In view of these facts it is obvious that most rapid wear of expendable parts
occurs at each end of the stroke, or at each stroke reversa!. Then the life of parts
becomes a function not of speed, per se, but of the reversa[ rate, again a matter
of stroke-length and RPM.
10.1.4.1 PUMP SPEED REPORTING
Because pump RPM is design-related to pump stroke length, the term Feet Per
Minute (FPM) becomes a factor that includes stroke-length and speed, the use of
FPM must be involved in the relation to pump reversal-rate and parts life:
FPM s x RPM/6 (lO.I)
RPM 6 X FPM/s (10.2)
R'PM 2 X RPM (10.3)
Where s pump stroke, inches

R'PM = Reversals per minute
Now ali punips can be limited to a constant maximum speed by the term FPM.
For example, Table lO. l will show;
TABLE 10.1 Relafü:m of RPM, FPM and Reversa! Rate

(R'PM)
Stroke 100 FPM 200 FPM 300 FPM

RPM R'PM RPM R'PM RPM R'PM
6" 100 200 200 400 300 600
8" 75 150 150 300 225 450
10" 60 120 120 240 180 360
1211 59 IOO 100 200 150 300
14 11
43 86 86 172 129 258
16 11
38 76 76 152 113 226
IO" 33 66 66 132 99 190
20 11
30 60 60 120 90 180
22" 27 54 54 108 82 164
24 11
25 50 50 100 75 150
Thus it can be seen that the Rate of Reversa! is a function of both RPM and
stroke length.
Experience has shown that piston speeds below 200 FPM should provide trou-
ble-free hydraulic performance. See Chapter 3.
Apparently many drilling mud pump manufacturers failed to recognize this
'reversa!' mode of wear because recently introduced Triplex SA mud pumps ap-
peared with short stroke (sorne 1700 BHP pumps with a 12 inch stroke compared
to the 18 inch stroke in a Duplex DA pump of the same power). Note that 1211
falls in the worst part of the scheme in Figure 10.4A. Also over the years,
smaller pumps saw an arbitrary increase in speed rating, resulting in increased
power per pound and power per dollar. There are appropriate needs for such
design where portability or space is at a premium, but for the greater number of
stationary installations, long stroke and low speed should be the requisite.
The argument that 'long stroke pumps cost money' is not conclusive as shown
in Figure 10. lA where even a double capital cost would still remain well below
the long-term expendable parts cost.
Accordingly, the wearing process on pump 'expendabie parts' is responsive
mainly to the 'reversal rate'. Simply stated, the following Cases will result in
exponentially extended expendable parts life by putting these facts to work.
CASE 1, Figure l0.2A. By slowing the main pump(s) down below rated speed
and complimenting with the other or 'standby' pump(s) at the required reduced
speed. Multiple pumps in such applications as pipelines, particularly abrasive
slurry pipelines, tend to garner the most from the 'ali pumps rmming' applica-
tion. Table 2 compares such multiple pump usage. This scheme should perform
well on oil well drilling rigs that invariably make use of two pumps, one main
pump and one stand-by, alternating the position to distribute the wear between
both pumps over the long time period.
PARTS LIFE VS PISTON SPEED
AS FUNCTION OF STROKE LENOTH
40000 •••••••••••• ····-·· ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• •• ,..t:. ••••••••
30000 ••••••••••••• ····-·· ••••••• ....... ....... ....... ;-;"'·

'
······
:::;::• .~:::::........
.
,,,,~' 1
,
J.······· 1(··": 1 •
~
20000 •••••• ••••••
1
111······· ....... ······· ........ ······~¡;.,
,,
A~(J .•... J. . . 10000000 !0000000
·······
1
!,....·¡ ' 1 1 1
,...::. ~~---·· '·······1······· ·······l:;:;:::.' ••:::.'
......... ..........
,,,,,·,
······· ······· ...........
/
10000 -··-/- •••••• ····-·· ··::: ,.,,..:. •••••••
/ ·······
::::::: ::::::: ,.,. ...! .:::::: ::::::: :;:::::
······•······
--:::.:::::: ..........................................
... .. .. ::::: ::::· ·····- ···-· .......
/ ,,..11'" ••••••• ................ ..
;;; :::::: :::::.,

o'P"-~+---t,.....-+~-+~-+~-+-~+-~-t-~t---t,.....-+~-t-~-+~-+-~+-~-t-~t---t~-t~-t-~-t
3 8 10 11 12 13 14 16 18 17 18 19 20 21 22 23 24
STROKE LENGTH - INCHES
. . 100 fpm ·•· 200 fpm
+ 300 fpn ... 400 lpm
Figure 10.lA
PUMP SPEED VS PARTS LIFE:: INCREASE
6.00
·········r········r·······T········r········¡··········· ··········· ........... ··········r········· ........... ··········· ..........
1 1 1
5.00 ····················l···················-r········l·········1·································································
1 1 !
.......... ¡...........•........•...........•..........
! i 1
11.•.•....... 1........... 1•••••••••••1; ••••••••••• 1...•••..•.•1..............................
4.00
1 1 i ¡ ¡ ! i • i 1 . 1
~
~
.,w
:;\
IC
u
<;
3.00
-l 1l-1-l--l--1---1-i _J__
i 1 1 1 1 1 l
1
1
,,.---
~
:::;
2.00 .....................1 .................... J. . . . . l........... .......... 1 ~~::: •••: : : :
1.00
0.00-1-~--t~~-1-~~1--~-+~~-t-~--t~~-1-~~1--~-+~~-t-~--t ......~-1-~~-1-~-+~~~
150 145 140 135 130 125 120 115 110 105 100 95 90 eo 76
PUMP SPEEO - OECftEASE
+ ONE PUMP ·•· 2 PUMPS
Figure 10.2A
327
CASE 2, Figure lO. lA. By selecting a pump(s) with long stroke-length, resulting
in a lower RPM and R'PM for the same delivery.
Note that in either Case the stroke-reversal rate has been reduced.
CASE 3, If there is an unusual application where both Case l and Case 2 can
be applied, double extended life can be expected.
10.1.4.2 CALCULATIONS FOR PARTS LIFE EXTENSION
For individual pumps involved in the above Cases, note that with pump slow-
down, another factor affecting wear comes into play, namely, pump power is
reduced, also contributing to increased parts life. However, when another pump
in the system is placed in service to compensate for the loss, substantial benefits
are still to be gained in spite of a greater number of pumps operating. (The
advantages become less as the number of pumps in a system extend beyond three
or four).
Case 2 can only be obtained by the original selection of (or replacement by)
pumps with a longer stroke and equal power. The relation of parts wear to pump
speed and reversal-rate, confining the effects to one pump, is given in the Equa-
tion:
(l0.4)(1)
For the Case l, where an additional pump is required to maintain a certain
through-put, the Equation becomes;
(10.5)
Where
L1 =
Life l, Present, hrs
L 2 = Life 2, Expected, hrs
N 1 = RPM l, Piston speed, Present
N 2 = RPM 2, Proposed
10.1.4.3 EFFECT OF PRESSURE ON WEAR
Although the effects of discharge pressure on the life of parts has nothing to do
with this discussion, any discharge pressure affects a decrease in life according
to the Equation. See Figure l0.3A.
(l0.6)m
Where
P1 =
Pressure, present, PSI
P 2 = Pressure, expected, PSI
"'Reference, The Economics of Positive Displacement Slurry Pumps. Holthuis and Simon. Holthuis,
B. V. The Netherlands. Slurry Transport Assoc. 19_ _ ,
' 21Reference. Pumps and Prime Movers. Huthnance. World Oíl. April 1964.
.................................. ,.................................. ,......................................, ................................... ., .............•.•. §·····--···-······
_///
//
•····························-··-···•;--································+···············-~·-···········-~/·'f·····································i··--·······························-~
_/V
/
t····························-·-···:!;>111"'··············--············t·····································I···················-··············· ·····················•·••••••••••••
~I
f'AESSURE, 1000 PSI
Figure 10.3A Mud pump pressure vs. parts cost
10.1.4.4 INCREASING PUMP HORSEPOWER AVAILABLE

On applications where multiple pumps are used, the demand for extra power
allows one to run both pumps for a demand for more power than allowed with
one pump. Even if this is carried to the extreme of using the ful! power of both
pumps, the improved parts life would still be had over single pump of the re-
quired BHP. but when maintenance of one pump is required the "standby" is nil
and the system would have to operate at half through-put, if permissible. For
normal two-pump operation in a drilling rig see the following Section.
10.1.4.5 STANDBY PUMPS
The following discussion should dispel any idea that one is trying to operate
without a standby pump and the idea is not as careless as it may seem.
With two pumps (one running and one standby) the standby capital cost is
two times that for a single pump to perform a certain function but with two
pumps running, as described in CASE l, the capital investment pays a greater
dividend in the form of greater parts life. With two pumps running, the need for
standby is greatly reduced because of the ability to extend the operating life of
both pumps. But there may be understandable objection to the apparent loss of
'standby'. However, it is not a complete loss for one of the two pumps can be
speeded up to the rated RPM while the second one is being maintained. If the
described method of obtaining more total power by combining two pumps, and
c.!
(,)
Q
TABLE 10.2 Operating Data
TYPE NUMBER STROKE SPEED BHP GPM TOT PARTS COST COST
COMB PUMP PUMPS INCHES FPM RPM PUMP PUMP BHP COST POWER TOT
l DUPDA l 18 195 65 1706 682 1706 $2020 $4743 $6529

2 DUPDA 2 18 98 33 855 342 1706 1129 4743 5872
3 TRIPSA l 12 240 120 1612 682 1612 1692 4480 6172
4 TRIPSA 2 12 120 60 806 341 1612 1065 4480 5545
5 DUPDA 2 18 144 48 1009 403 20!9 2034 5611 7645
6 TRIPSA 2 12 176 88 1009 428 2018 1819 5273 7092
7 TRIPSA 2 18 98 33 806 342 1706 499 4480 4979
8 DUPDA 4 18 195 65 1706 682 5118 23924 14209 38133
9 DUPDA 4 18 98 49 1280 512 5118 18371 14209 32580
COMBINATIONS OF ABOYE DATA PARTS COST > REDUCTION
%
A One DUPDA full speed vs two DUPDA at half Speed;
l 1700 BHP DUPDA, One, ful! speed 2020
2 1700 BHP DUPDA, Two, half speed 1129 44%
B One TRIPSA full speed vs two TRIPSA at half speed;
3 1700 BHP TRIPSA, One, full speed 1692
4 1700 NHP TRIPSA, Two, half speed !065 37%
C One DUPDA ful! speed vs one TRIPSA at ful! speed;
l 1700 BHP DUPDA, one at ful! speed 2020
3 1700 BHP TRIPSA, one at full speed 1692 16%
D Greater than normal one-pump power output
5 1700 BHP DUPDA, (two) running above half speed 2034
6 1700 BHP TRIPSA, (two) running, above half speed- 1819 11%
E 12" TRIPSA (NORMAL) vs 18" Theoretical long-stroke TRIPSA
3 1700 BHP TRIPSA, 12" stroke, two running 1692
7 1700 BHP TRIPSa, 18" stroke, two mnning 499 71%
F Four-pump slurry station, slurry abrasivity Miller Number 80.
8 l 700 BHP DUPDA, Three running full speed 23924
9 1700 BHP DUPDA, Four running, reduced speed 18371 23%
Ali COSTS are in $1000 for 20 years pump life.
PUMP PARAMETERS EXCEPTAS NOTED;
DUPDA-1700 BHP Duplex Doubie Actíng Pump-18 inch Stroke.
TRIPSA-1700 BHP Triplex Single Acting Pump-12 inch Stroke.
Normal System Through-put-683 GPM, 3640 PSI, 1450 HHP. 2049 GPM for Slurry Station.
~~~~--~~~.~~~~~~~~
TABLE 2 reveals that;

1-Running both pumps 011 a drilling rigor al! pumps in a pipeline station, at reduced for equal throughput, results
in a great saving in pump expendable parts.
2-A long-stroke pump offers superior parts saving over a short-slroke pump.
3-A TRIPSA pump is more Mechanically Efficient than a DUPDA purnp.
c.>
w
....
98 ························-·· •••••••••.••..•...••••••••• ••••....••••.•....••••••••• ·····························t····························· .........................•..

~ 1 1
~ 1 i
~ 97 ························-··· ····························f·············································· ...•.... ···························· ········•••··•·•••····•·····
i 98 ························-·· ••••••.•••••••••.••...•••••• •••••••·•·••·••••·•·••••••••·•••••••••••••••••·•••••••••••••••••• •••••••••••••••••••••••••••••••••••••••••••••
96 ························-··· ••••••••••••••••••••.••••••• •••••••••••••••••••••••••••• •••••••••••••••••·••••••••••• •••••••••••••••••••••••••• ••••••••••••••••••••••••••••
94.l..-~~~-l-~~~~+-~~~-+~~~~-+-~~~-+-+~~~
100 160 200 260 300 400
PISTON SPEEO • FEET PEA MINUTE
Figure 10.4A Pump speed (test data) vs. volumetric efficiency
if the power difference is great (up to the full power of both pumps) then one
would have to consider a period of time of half power if the system would allow
it. For example, with two 1700 BHP 120 RPM rated TRIPSA pumps running at
75 RPM each, the total BHP would be 1700. With both pumps running at, say,
90 RPM the output would then be 2040 BHP at which the system could probably
be operated for a short period of time at 1700 BHP.
This Standby reasoning discussion also applies to multi-pump systems such as
long slurry pipelines where as many as seven or more large pumps are used, but
with less advantage in parts life extension. However, if the slurry in question is
highly abrasive, such as sorne iron ores, the extension of parts life takes on more
importance. Table 2 shows such reasoning on a slurry application, for example,
pumping 80 Miller Number phosphate slurry.
10.1.4.6 VOLUMETRIC EFFICIENCY
Any speed reduction below the pump's rated speed will usually result in im-
proved Volumetric Efficiency as a result of improved hydraulic performance. See
Figure l0.4A that shows actual test data for a Triplex Single Acting pump, typi-
cal of all pumps.
With such manipulation of speed as above, piston diameter need not be con-
sidered.
10.2 PLUNGERS 333
--~ ~ :::..:~'~·~
1003 • '''"""'""'
Q) 903
:!:: 803 '· ~ ~ .,.. ........ ,,,,~
.~ 703 ..... ~~~
~ o '(A ' ' "'i¡¡;':•~
a>
"'603 ~ •,. "' "'~:;!!!I
.~.....
1500 PSI
1¡
.tl 503
3000 Ps1'Xí}
·~·
·\.\.
'
°'U\
':'
2 2000 PSI
~:~=~-t---+----t~+--t--+~+--t---+----t~+--+---+-~--1
203~-t---+----t~+--t--+~+--t---+----t~+--+---+-~--I
103~-t---+----t~+--t--+~+--t---+----t~+--+---+-~--I
.010 .020 .030 .040 .050 .060 .070 .080

Piston-liner diametrical clearance - inches
Figure 10.12. Importance of piston-liner clearance. (© TRW Inc. 1957. Reprinted by

permission of TRW Inc.)
10.1.5 PISTON-LINER CLEARANCE
Figure 10.12 shows the effect of piston-liner diametrical clearance, as the result
of wear of the liner and piston body, on the life of piston rubbers. For example,
if the service life of a piston operating at 2000 PSI is 300 hours when the initial
clearance is 0.01 in, then the expected service life of the next rubber, when the
clearance is 0.04 in, would be 80% ór 240 hours.
10.2 PLUNGERS
Liner and piston rod wear modes and the fact that they wear more rapidly at the
end of each pressure or delivery stroke have been discussed. However, because of
the distinct difference in the mechanics of packing in a single-acting pump, at
reversal ofthe pressure stroke the packing has already been ''dragged'' by friction
forces away from the severe extrusiOn-gap seal to its "relaxed" condition, thereby
minimizing the scrubbing wear at that point.
Plungers invariably contradict the end-of-stroke type of wear and exhibit a
typical ''hourglass'' or necking-down at midstroke where the velocity is greatest.
As to the fear of greater wear on the plunger bottom due to its own weight in
horizontal pumps, that fear can be dispelled, as no wom plunger has ever been
observed without more or less uniform wear around its circumference. This can
be explained by the fact that the forces of pressure on the packing circumference
tend to force the plunger toward the center of the stuffing box and generate unifonn
wear.
A phenomenon exhibited by solid ceramic plungers is that under the ideal con-
ditions of almost surgically clean liquids, with no abrasive materials or particles,
they exhibit no reduction in diameter from wear as such. They take on a discol-
ored, polished appearance and would probably run ''forever'' but for other factors.
These factors are: (1) Abrasive material in the liquid, (2) drastic temperature
changes that produce thermal shock and fracture (a typical case is where a pump
runs without prime for sorne time and is suddenly subjected to a dose of cold
water), (3) misalignment in the pump causing breakage of the rather fragile ce-
ramic, pitting due to cavitation as described earlier, and (5) rough handling
and striking with tools.
Certain sprayweld coatings that consist of a rather soft matrix with the addition
of extremely hard particles, such as tungsten carbide, should be avoided for plunger
service. Scrubbing wear or selective chemical attack will remove sorne of the soft
matrix, leaving a sandpaper surface that is extremely damaging to packing. This
process is almost microscopic, and a used plunger may appear to be in excellent
condition but a magnification will show the effect described. A simple test to de-
termine if such a condition exists is to rub a copper coin (penny) along the plunger
surface. A bright copper-colored streak will indicate a damaged surface.
Sprayed ceramic coatings for plungers have performed satisfactorily in many
applications, but certain corrosive liquids may tend to penetrate the porous coating
and attack the base metal, resulting in spalling of the coating. It is reported that
such plungers have been improved by the application of pressure-applied sealant.
Because the bond between the ceramic particles and the sealant may be doubtful,
the repeated application of high to low pressure (discharge to suction) could have
a detrimental e:ffect by early fatigue of the bond with subsequent loss of seal.
10.3 CERAMIC PLUNGER PITTING
Ceramic (alumina) plungers have a propensity to pit and there are three distinct
types of pitting: ring, local, and end.
Type I or ring-type pitting occurs in a circumferential pattem usually com-
pletely around the plunger diameter in a single row of pits at a uniform distance
back from the end of the plunger. Measurements show that this type of pitting
usually occurs at the last ring of lip-type packing at the end of the forward or
pressure stroke as shown in Figure 10 .13.
Type II or local pitting occurs in a single patch or area from ~ in to as large as
1 in in diameter, always on top ofthe plunger and always in the regionjust forward
of the stuffing box throat bushing on the end of the forward or pressure stroke.
Type m or end-type pitting occurs on the very end of the plunger, usually at
th.e top, and is an "eating out" of a large portion of the plunger, resulting in holes
as Iarge as a walnut.
10.3 CERAMIC PLUNGER PITT!NG 335
Type 1 - Ring Pitting Type 11

Local Pitting
/
/ Type 11!
End Pitting
I
Area of Local Pitting

moved to non-contributing
Longer throat location by use o! longer
bushing. throat busr1ing.
Plunger at end oí pressure stroke.
Figure 10.13. Ceramic plunger pitting.
Types I and II, ring and local pitting, <;ire extremely hard on packing, since the
rough surfaces run through the packing on each stroke. End pitting does not affect
packing as the pit does not enter the packing.
10.3.1 Causes
Ring pitting always occurs at the very tip of a 45 º sealing lip on pressure-seal
packing, ahead of which is an annular void created by the large inside diameter of
a phenolic bottom adapter. It is possible that water (and no doubt mixed air) under
pressure is suddenly released at the end of the stroke, creating extreme turbulence
that mechanically works on the ceramic. Ring pitting has been cured by the use of
a close aüapter, removing cavities where liquid could accumulate.
Local pitting was first reported when an operator told that he was able to move
this type of rather rare pitting toward the fluid end of the plunger by the use of a
longer throat bushing. The pitted area then would be in such a position that it
would never enter the packing and would do no harm. This has been proved to be
a facL
Local pitting is associated with the increased operating speed of pumps as prac-
ticed in recent years. This has led to speculation that suction problems introduced
by high-speed operation was responsible.
There is evidence that heavy valve springs (required for high-speed operation),
among other conditions, may cause an excessive pressure drop in the cylinder.
Recent tests show that the valve springs of even medium load will result in an
instantaneous pressure drop through the valve greater than the suction pressure
without any measurable effect on the volumetric efficiency of the pump. Perhaps
at the instant of low pressure in the cylinder, air or gas will break out, and since
gases tend to break out of a fluid with rnuch more ease than they go back into
solution, their presence provides the pocket of air to ''explode'' on the end of the
pressure stroke, giving the extreme turbulence and cavitation necessary to act on
the plunger.
"Soft" spots in the ceramic were at one time held responsible for local pítting
but are now ruled out. This did lead to the idea that perhaps a denser aluminum
oxide material would not pit, but tests have since shown that even 99 % aluminum
oxide plungers pit.
End pitting is probably caused by turbulence of the fluid either at the end of the
pressure stroke or at the end of the suction stroke. Little concem has been given
to this type of pitting since it usually causes no trouble. However, it should be
warning that suction conditions may be inadequate and could result in reduced
packing, plunger, and valve life. These pitting problems occur on both solid and
sprayed ceramic plungers. They do not occur as such on metal plungers. Some-
times a pit does show up in the same location as ring or local pitting, but this is
either typical random pitting or the coincidence of a defective spot susceptible to
pitting occurring in the region of high turbulence.
The extreme brittleness and the nature of the ceramic material can account for
the fact that only ceramic plungers are subject to pitting.
10.3.2 Cures
It presently appears that nothing can be done to the ceramic plungers themselves
to prevent pitting. Since the occurrence of pitting is infrequent (considering the
large number of ceramic plungers in service), local handling of the problem on an
individual basis would be the best approach to the problem.
Ring pitting may be eliminated by the use of one clase-fitting bottom adapter
(the first adapter in the bottom of the stuffing box). This could be made of bronze
and should have an inner diameter about 0.010-0.020 in larger than the nominal
plunger diameter. Dimensions could be taken from the present phenolic bottom
adapter, or the information could be obtained from the manufacturer.
In the case of local pitting, the practice of installing a long throat bushing could
be employed. Here again use could be made of bronze, using the old bushing for
dimensional data. Before attempting to use a longer bushing, it would have to be
determined if the pump in question will accept one. In sorne pumps the bore be-
yond the present bushing space is smaller than the stuffing box bore, which pre-
cludes the use of a long bushing. Then if pitting remains a serious problem, an
altemative step would be to use a metal plunger.
10.4 THEORY OF CERAMIC PLUNGER FAILURE MODE

Solid ceramic (sintered alumina) plungers are paradoxaly useless in liquids con-
taining abrasive solids, probably for the following reason; these plungers are ex-
tremely hard and brittle and are usually polished with diamond dust to an appar-
ently smooth finish. However, such ceramic has a microscopic porosity which,
when the plunger is polished, leave minute weak-walled cavities as shown exag-
gerated in Figure 10.14(a). If a grain of sand, for instance, is trapped under one
10.5 OTHER MEANS OF REDUCING WEAR RATE 337
Elastomeric
packing ring
I
V \
(a) "Weak" Trapped sand (b) Cavities Progressive
shelí particle íailure of shelf
(e) (d)
Figure 10.14. (a) Ceramic plunger failure mode. (b) Progressive recession of Niagara Falls.
of the elastomeric pressure rings, as shown in Fig. 10.14(a), and if high hydraulic
pump pressure is applied to the ring at the instant when the grain of sand is located
at the precarious edge of the cavity as shown in (a), failure dueto high force will
fracture that minute shelf of ceramic.
With millions of repeated strokes, the remaining edge of the lengthening cavity
is continually being fractured as shown in Figure 10.14(b) and develops into a
groove equal to the length of the stroke, thereby appearing as a score. The effect
is much like the receding of Niagara Falls, where the hard (brittle) top layer of
limestone loses its support of softer underlying shale in a progressive manner,
thereby gradually moving the falls upstream [see 10.14(c) and (d)]. The hard-
ness of alurninum oxide would certainly not allow a direct scoring of the material
by a foreign particle or particles of normally encountered materials.
10.5 OTHER MEANS Of REDUCING WEAR RATE

Sorne means of reducing the abrasive-corrosive wear losses in a slurry system can
be considered. AH of the following recommendations must be balanced by the
savings of rate-of-wear improvement relative to the cost of the effort:
1. Decrease the corrosivity of the slurry by increasing the pH, or inhibit it by

use of appropriate chemicals.
2. Decrease the maximum particle size of the solids.
3. Use a "ftushed stuffing box" plunger pump. (Such flushing can extend the
life of plungers dramatically if the 3 % or so dilution with clean liquid can
be tolerated.)
4. Reduce concentration of solids to gain sorne reduction. (Slight reduction in

concentration is not too effective.) See Figure 10.3.
The remaining items apply to any pumping system and should become man-
datory rules.
5. Reduce the speed of pumps by using maximum piston-plunger diameter

commensurate with the pump rating.
6. Replace parts as they wear.
7. Provide a "good" suction system with at least 50 PSI suction pressure.
8. Provide effective suction and discharge pulsation control on reciprocating
pumps to reduce wear-generating high-pressure peaks.
10.6 SLURRY ABRASIVITY
Slurry abrasivity is covered in depth in Chapter 9. Accordingly, the reader is di-

rected to Table 9.3 for typical Miller numbers for slurries.
10.7 EXAMPLES OF PARTS UFE IN HOURS
Table 10.1 is included to show the comparison of the approximate typical lives in
hours of pump parts in various services.
10.8 REDUCING DRILLING PUMP PARTS COST
Because of the phenomena of Pump Stroke Reversa! Rate on the life of pump
expendable parts (See Section 10. l .4) significant saving in parts cost can be had
by:
l. Running both pumps on a drilling rig (or any multiple pump application
where non-operating standby pump(s) are used, by running both pumps
(usually of the same size and type) at half speed. In this case the "stand-
by" feature is still had by the simple expedient of running one pump at
fuB speed while the other is being maintained for a relatively short period
of time.
2. Use of pumps designed with maximum stroke (and with usually slower
speed) offer the same saving in parts cost because of the lower reversa!
rate.
3. A Triplex Single Acting pump of the same stroke length and equivalent
displacement as a Duplex Double Acting pump will contribute to a lower
parts cost and to better mechanical efficiency because of the non-wearing
piston rod and its absence of friction. See Table 10.4.
______._.__._..............
--~--~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-~
TABLE 10.3 Examples of Parts Life in Approximate Hours
Miller
Service PSI RPM Valve Seat Liner Rubber Rod Packing Plunger Packing Number
Drilling 2000 80 394 555 48 14
Triplex mud pump 2500 100 974 974 1415 523
Duplex mud pump 2000 65 71 1072 886 303
3000 65 1202 1363 1109 889 1109
2000 - 1200 1200 400 300
Duplex test mud 2000 65 1000 1000 500 300 700 400
pump
Ohio coal duplex 800 65 2000 2000 4000 600 1300 1500 - - 37
Black Mesa duplex
Estimated 1000 65 2000 2000 3600 600 1400 700 - - 27
Actualª 1000 65 4800 4800 8000 1200 1200 324 - - 27
Savage River magne-
tite
560-hp plunger 1500 75 700 780 - - - 720 425 64
Cerro Paseo tailings
Duplex 970 60 250 250 1323 590 523 448
Plunger pump 970 218 278 - 2900 448
Water ftood Plunger 3000 100 8000 8000 8000 4000 o
x Phosphate Tripsa 1400 120 1200 420 800 400 - - - 80
y Ironore 1300 - 400 400 2000 2000 160
Failings (copper) - - 300 300 400
ªSee J. G. Montfort, "Operating Experience of the Black Mesa Pipeline," Proceedings of the Seventh Jnternational Technical Conference, Lake Tahoe,
NV, 1982.
w
w
co
TABLE 10.4. Drilling Rig Mud Pump Combinations
Pump BHP Parts Power Cost

Comb Make Type BHP RUN PSI RPM GPM Act. Cost Cost Total
One 1700 BHP Duplex Double-Acting Pump Vs Two Running At Half Speed;
A A DUPDA 1700 1 3640 65 683 1706 1877 4773 6650
DITTO 2 3640 32 341 853 929 4773 5702
Obtaining More Total BHP By Running Two Duplex Double-Acting Pumps At Greater Than
Half Speed For Each
B A DUPDA 1700 1 3640 65 683 1706 1877 4773 6650
DITTO 2 4000 38 400 1000 1460 6!05 7565
One 1700 BHP Triplex Single-Acting Pump Vs Two Running At Haif Speed;
C OWS DUPDA 1700 1 3640 65 683 1706 1877 4773 6650
DITTO 2 3640 60 342 806 1049 4480 5229
Obtaining Greater Power From Two Low-Power Pumps
D A TRIPSA 1700 1 3640 120 683 1612 19!0 4480 6092
DITTO 2 4000 106 400 1037 2690 5766 8445
One Pump Running Vs Two Running At Half Speed-Other Sizes
E B TRIPSA 2000 1 4000 100 400 2000 2321 5755 8076
DITTO 2 4000 50 200 1000 1309 5755 7075
18" Stroke Duplex Double-Acting Vs 12" Stroke Tripiex Single-Acting Pump;
F A DUPDA 1700 1 3640 65 683 1706 1877 4773 6650
A TRIPSA 1700 3640 120 683 1612 1910 4480 6398
18 11 Stroke Duplex Double-Acting Vs 18" Stroke Triplex Single-Acting Pump

G A DUPDA 1700 1 3648 65 683 1706 1877 4773 6650
A TRIPSA 1700 3640 65 683 1459 1612 743 2355
10.9 DANGER OF "STRAINERS" OR "FllTERS" IN THE

SUCTION LINE
The use of strainers or filters in the suction line of a reciprocating pump should
be viewed with apprehension. If there is an excess of particulate matter of rela-
tively large size in the supply liquid there could be danger of complete plugging
or "strangulation" of the pump with complete loss of delivery with little or no
chance of recovery.
10.1 O SlURRY PARTICLE SIZE
Single-stage centrifuga! purnps can handle a larger percentage of particles of

large size but reciprocating pumps are limited by their inability to effectively seal
over large particles as they pass through the valves, the size being closely related
10.10 SLURRY PARTICLE SIZE 341
to the pump and valve size-the larger the valve the larger the particles it can
handle. A very general equation for particle size limitation is:
Maximum Particle Dimension = Approximate Valve Through-area/80.
lt has been proposed that power-operated valves be used to handle large coal
particles, the power-operated valves being forced to closure by crushing any large
particles caught under the valve but this idea has not been given much chance of
satisfactory performance. However, power-operated valves are common on most
slow-speed direct-acting cement-aggregate slurry pumps.
High valve spring POSIVA tends to allow a "straining" effect on the larger
particles during the last moment before complete closure when the clearance
between the valve and seat becomes less than the particle dimensions. At low
pump speeds this may lead to complete clogging of the pump chamber accom-
panied with total failure of delivery.
With pumps used in hydraulic high-pressure press operation in steel milis, very
few pump problems of the sort described, from the return water containing the
usual contamination of mili scale, dust, etc. are experienced. Reduced valve life
is the usual result.
11
APPLICATI NS
11.1 EFFECT Of LIQUID COMPRESSIBILITY AND

CLEARANCE VOlUME ON THE VOLUMETRIC EFFICIENCY OF
RECIPROCATING PUMPS
There are three main reasons why a pump falls short of delivering a volume of
liquid equal to the displacement of the piston or plunger.
1. All liquids are more or less compressible in direct proportion to the pressure.
2. A reciprocating pump, because of practica! limitations, has a certain volume
of liquid remaining in the cylinder between the suction and discharge valves,
even when the piston or plunger has completed its delivery stroke.
3. Pump valves experience a small amount ofleakage (slip), particularly at the
moment of closure. Since this is a function of valve condition, it is difficult
to predict, and it is the usual practice to select an arbitrary decimal value of
0.02.
On each stroke the cylinder fills (almost) with liquid under suction pressure (1).
On the delivery stroke the liquid is compressed to the discharge pressure, and then
delivery is made through the discharge valve (2). The remaining liquid in the
cylinder must reexpand to suction pressure and accept a small volume of slip liquid
through the discharge valve before the suction valve can open (3). Therefore, sorne
effective piston or plunger travel is "lost" on the suction stroke, resulting in a
reduction in the amount ofliquid being taken into the cylinder.
342
11.1 EFFECT OF LIQUID COMPRESSIBILITY 343
Method of Measuring C/earance Volume
V = volume of cylinder between the suction and discharge valves with the
piston or plunger at the end of its suction stroke to right, in 3
D = displacement of piston or plunger, to left, in3
e= clearance volume, that liquid volume remaining in the cylinder be-
tween the suction and discharge valves with the piston or plunger at
the end of its discharge stroke, in3
C=V-D
e
C(percent) = DX 100
C (percent) is clyarance volume as percent of displacement. V, the volume of the

cylinder, can be measured by first removing the valve cover; rotate the pump so
that the piston or plunger is at the end of its suction stroke. Grease suction valve
(to obtain a positive seal), and insert it into its seat. With the discharge valve
removed, meter water into the void to fill the cylinder to the top of the discharge
valve seat :and record the cubic inches. This method allows rapid calculation of
clearance volume for any specific liquid end by subtraction of displacement for
any size piston or plunger.
11.1.1 Pump Hook-up

The idealized pump installation depicted in Figure 11.1 is self-explanatory and
includes details of all the desirable features that should be incorporated into the
system for maximum efficiency, safety, and successful operation.
~=~~=~~va
HJRIZONTAL PLUNGER PlW IUTH OUAl Sl.ETION ANl
OISCtwtGE cotKCTICl"5 ti4S TI-E AO\IANTAGES OF:
A. LOW SU:HOH CC»IECTION MINIMIZES SU::TION FOR ACCEPTABLE f"ERf'1 TIE PlW MJST
PRESSURE CHEAO) REQUIROENTS. BE PROVREJ wrrn AI'E TE SU::TION PRESSl.H:.
B. DUAL CfHECTICffi SI!if'LIFIES "tE CE:SIGN CF A MINIW'i OF 4 TO 5 T Of WATER (00 HS
t sutTION ANJ DISCHARGE PIPIN; ~ EQUIVAlENT CF OTIER F UlD) AOOVE TlE PlW
\
C. LCll CE:NTER f.F GRAVITV REm.ICES VIBRATICN.5. SUCTION IS USUALLY FICXENT. HIGH TEMP-
O. LOR PfU"ILE EASES MAINTEmNCE Atl1 REPAIR. ERATlflE, AL TITWE, V PRESSOOE, LitE FLUID SlPPLY TAf-.1<:
FRICTI~, PlW SPEED tD .l\CCELERAT!Of'l t-EAIJ
~\
*00 001' ~ &Cl'ICE" PIPEr BY MENi5 OF
A LOOP, ro IDl'H P'tMP su::.T!Clif aHm::!'.r!CfiS .. lfl!IU.. Itt;REASE THIS MI DU4 REQUIREf.ENT .
PRESSLAE RELIEF VllJ...VE SET NO HIG&-ER \ f l l . L OPENING VALVE
LI~'
11iAN 1.25 Ht.ES TtE MAXIM..fl! FLtMGER LOCATED NEAR TAtf<.\
LUBAICATICtJ OF PACKING IS RECCKE~D WORK!NG PRESSlJi'E-, \ BY -PASS LINE Af'ID RELIEF
TO REDUCE FR!CTION At('l EXTEKl THE LIFE RETURNING TO THE StPPLY TANK
OF TI-E PLUNGEAS ANJ PUJNGER PACKil\G CliARGil<l Pl>P - >IJTE 4 \~
~
,,,.-g~\ f'RESSIJiE GAUGE
AIDID STRAINER5 m FTIJl'ERS IN T!lE su::noo LINE
FLEXIBLE HJSE - t<ITE 2
AS CLOSE AS POSSIBLE TO THE PlW

STRAINS Am MINIMIZE VIBRA TIONS
BLOCK VALVE
NOTES GE~RAL RECC!f4\EMJATlONS
l. A SEPARATE Sl.CTION LINE IS RECGKNJEO FOO EACH PLM'. I . PROVIOE SAfETY Sl-UT DOWN CDNTROLS FOR ALL "'AL -
TtE: LINE SH:llLD BE A MINitJlM OF Of\E SIZE LARGER THAN FU!\CTIONS OF HE SYSlEM, PLW At-ll PRil"E M!JVl;R .
OISCHARGE UNE - HE FUf> SUCTION FLAN:;E OR IYEN!NG TO PROVIl:X:: A Mo\X- If'K:LUOING:
IJl.U.I Of l TO 2 FEET /SECCHJ FLUID VELOCITY IF BEt{)S A. UJW SUCTION PRESSURE OR LO't'I TAJIK LEVEL
IN Tl.f'. SUCTION L!J'.E AS CLOSE ARE tECESSARY USE CtL Y 45º LOM; RADIUS ELLS. A SLCPE 8 LOW OISCHARGE PRESSLRE
AS POSS!BLE TO ThE PlM9 CF 1/4" PER FOOT FROM TAf<I< TO PlM' IS RECo.M.t«D. C . HIGH DISCHARGE PF!ESSURE
2. A FLEXIBLE HJSE Af'il/00 EXF'ANSION .JJINT IN TI-E SUCTION O. LOl'I OIL PRESSU1E 00 LOO' OIL LEVEL
L!NE (ALSO DISCHMGE LINE) PROVIDES FOFI TH:.RMAL EXPAN- E. HXGH TEMPERATURE, EXCESS!vE VIORATXONS.
SIIM ANJ TEMJS TO OA14'EN VIBRATIONS. OVERLOAOif<J3 GF PRIM:: 1-0VER, ETC.
3. OIRECT HE OISCHARGE LIN::, USil'XJ 45" LONG RAOitlS ElLS, II. FOA SERVIC!NG AHJ SAFETY PROVIOE AOEQ'JATE
TO GRADE 0R BELOW GRADE LEVEL AS SO~ AS POSSIBLE TO WORK A.REA AROUN.) M Pt.J>IP UNIT.
RELIEVE STRAIN Afl4J REOLCE V!BAATIONS. HE UNE SilE III. CARE AM:l Of'ERATIONS SHOUL!J INCLUOE CLEAJ'.i!...I/\'ESS.
SHJll.O BE SUFFICIENT NOT TO EXCEEO A ~XIMUM FLUID VELO- DAILY INSPECTIONS, PERIODIC INSPECTIONS. ROUT.
CITY OF a TO 10 FEET /SECONO. INE MAINTEN~NCE AHJ PREVENTTVE M,.UÍt"EN.4Ni:E
4. IF CHARGING PtW IS USED !T SHJl.A.0 BE LOCATED NEAR THE
Slff'LY TAN<. IN ITS SELECTION GIVE CONSIOERATION TO
THE FLUID Pl.WED, VELOCITY, Pll.N>GER PL.f.f> VAL VI~. LlhE IDEALIZEO PUMP HOOK-UP
FRICTION. ne .
11.1 EFFECT OF LIQUID COMPRESSIBILITY 345
Sorne important precautions with respect to pump installation are:
1. Never retum the by-pass or relief-valve outlet back to the main pump suc-
tion. The turbulence created results in erratic pump operation.
2. Never use 90º elbows in the discharge system; long-sweep 45º elbows are
desired.
3. Never use a suction header for multiple pump installations from the same
supply tank. Use separate suction lines as shown.
4. Avoid high points in the suction line that form "pockets" for the accumu-
lation of air or gas. Note the use of eccentric ftanges on the inlet.
11.1.2 Charging or "Booster" Pump

A suction charging pump (usually centrifugal) is recommended for the main pump
if there is any doubt about NPSHA and to allow pump-down of the supply tank to
a very low level. However, it should be kept in mind that the charging pump itself
requires a certain NPSH.
A simple rule for sizing the charging pump is to use a pump with a rated ca-
pacity of about 150% of the main pump displacement. It is most desirable to drive
the charging pump with a separate prime mover. Those driven from a power take-
off from the main pump suffer extreme loss of capacity at low main pump speed.
11.1.3 Pump Valve Unloading System

Sorne pump manufacturers fumish pump unloaders that operate on the principie of
selective disablement of the suction valves, thereby allowing the pump to run at
full speed under no liquid delivery or pressure. The device consists in general of
a rod, operating through a stuffing box in the suction manifold, that holds the
suction valve open as long as a power device (hydraulic cylinder or electric sole-
noid) is energized and holds the rod against a suction valve (Fig. 11.2). To load
the pump, the cylinder is deenergized to allow normal valve operation. All ofthese
operations are controlled by a distributor on the crankshaft or by proxirnity switches
located near any moving pump part.
While such a system is rather complicated, it supplies the almost ideal means
of loading and unloading a pump in such service as press and die operation.
11.1.4 Suction System Loops

Sometimes "loops" are intentionally introduced into the suction of a pump by
1. U sing both suction inlet connections on a pump equipped with a suction

manifold that has optional connections on each end of the manifold, forming the
loop by bringing the two suction lines back to a T or Y on the end of the main
346 APPLICATIONS
Solenoid or
hydraulic
cylinder
Figure 11.2. Pump valve unloading mechanism.
suction pipe. This in effect provides double inlets on the assumption that ''better''
suction conditíons will result.
2. Pumps equipped with a suction charging pump, directing liquid through a
system of piping that branches off to the main pump inlet but continues back to
the tank with a submerged retum. This is sometimes used in slurry service where
a higher velocity in a long suction line is desired, such velocity being much greater
than that generated by the main pump.
In both cases, the liquid in the loop is prone to "oscillate" at the same fre-
quency excited by the pump RPM and the physical dimensions of the suction pip-
11.2 HYDRAULIC INSTITUTE STANDARDS OF APPLICATION 347
ing system. The amplitude of the oscillations will be multiplied if there frequency
is resonant with the piping system, to the extent that the suction of the main pump
will be greatly disturbed. Such loops should be avoided where possible.
11.2 HYDRAULIC INSTITUTE STANDARDS OF APPLICATION
Section 11.2 contains recommendations of the Hydraulic lnstitute in the matter of

pump applications, covering such subjects as pump arrangement, hook-up, prac-
tices and precautions. The nomenclature and terminology may differ from that used
in the main text and there may be sorne duplication of subject matter.
Type of Duty
Continuous duty-8 to 24 hours per <lay, fully loaded.
Light duty-3 to 8 hours per <lay, fully loaded.
Intermittent duty-Up to 3 hours per <lay, fully loaded.
Cyclical operation-1 minute loaded out of every 3 minutes.
Maintenance level-attended or unattended operation. Skill, training, and tools
of operating and maintenance personnel.
Medium Speeds
Power pump seeds at ornear the manufacturer's published "rated" or "normal"
curve will include those applications when clean, cold liquids are involved, and
will provide long life and economical operation, if all important application details
are carefully handled and regular, skilled maintenance is provided.
Medium speed selection requires excellent piping layout, good environment,
adequate NPSHA, periodic preventive maintenance and lubrication, rigidly fixed
piping, and solid pump and prime mover foundations (or bases). lt may require,
in sorne cases, automatic safety shut-down devices, suction and discharge dam-
peners, and may also require plunger or piston rod packing lubrication.
Medium speeds may be too fast for slurries, marginal NPSH situations, or un-
attended operation.
SlowSpeeds
Selection of an operating speed below the manufacturer's "rated" or "normal"
speed curve is often desirable when any strongly adverse factor is present, such as
the following:
Abrasive liquid (slurry) Unattended operation

Hazardous liquid Poor maintenance
Extreme pressure No spare parts, or no standby pump
Corrosive chemical High liquid temperature
High viscosity High ambient temperatµre
348 APPLICATIONS
Extremely long life desired Radioactive liquid

High-cost down-time of related facili- Dissolved gas in liquid
ties Borderline suction (NPSHA) situation
Extreme isolation of site
Operation at extremely slow speeds may require supplementary power end lu-
brication. Cooling of the power end oil may be necessary when hot liquids or
ambients occur. Always consult the manufacturer when very hot or very cold liq-
uids are involved. Revisions may be required in construction for these types of
applications.
High Speeds
Selection of speed above manufacturer's "rated" or "normal" curve and/or near
bis "maximum" or "intermittent" curve (if any) is sometimes merited when in-
termittent, attended service is involved. High speed selection requires very close
attention to all application details, skilled operators, and proper pump design. A
suction booster pump may be required to obtain sufficient NPSHA.
High speeds imply only optimum application factors are present, and that re-
duced life may occur. Sorne pumps are inherently designed for high-speed, short
duration, and infrequent usage. All conditions of such service should be well
understood by all parties prior to the sale. Oil well fracturing, acidizing, and ce-
mented plunger pumps are examples of this type of high-speed, intermittent ap-
plication.
Starting Power Pumps

Pump Torque Characteristics
Selection of pump driver type can be infiuenced by the cyclical torque character-
istic of a reciprocating power pump. This is especially true when slow-speed pumps
are employed. Such equipment often lacks enough mechanical mass to smooth-out
any torque variations imposed by the pump.
Torque fluctuations imposed by reciprocating power pumps on driving equip-
ment vary according to the kind or type of power pump, its number of cylinders,
and on the inertia of the pump and driver rotating masses. Thus, sorne pumps are
inherently ''smoother'' than others b.ecause less cyclical variations in driving torque
occur.
Peak torque requirements of power pumps can often be dampened by the use
of large diameter drive pulleys, or sprockets with high torsional inertia. These act
as "fiywheels."
Reduction of peaks in power torque may be possible by reducing discharge
pressure surge peaks, since torque and discharge pressure are closely related.
Hence, pulsation dampeners which effectively <lampen liquid surging will also help
smooth out torque variations.
Single-acting simplex power pumps have the most uneven torque requirement
of all the types of reciprocating power pumps. As the number of cylinders and
pumping strokes-per-revolution increase, the smoother is the torque. Thus, a quin-

tuplex pump is said to be very "smooth."
Objectionable amperage ftuctuations and heating in polyphase induction motors
driving simplex or duplex power pumps can often be minimized by choosing
NEMA Design "C," or Design "D" motors. These types of A.C. induction mo-
tors provide a "soft" driving torque which reduces line voltage disturbances which
Design "B" motors would cause.
Large double-acting duplex power pumps particularly tend to cause driver speed
surging. This type can also cause heavy amperage surging in an induction type
electric motor, unless a special motor type, or other preventive steps are taken.
Pump Torque Criteria

Application of power pumps requires careful consideration of their starting and
running torque demands. These affect the selection of driver rnotors, motor start-
ers, engines, gear reducers, belts or chain drives, couplings, and universal joints.
The effect of such loads on an electrical distribution system· requires thought, es-
pecially if the pump is large.
The starting torque required by a power pump, i.e. the twisting effort (moment)
applied to pump shaft or reducer shaft, falls usually into one or two general ap-
plications, as follows.
Starting With Liquid By-Pass

Operating personnel manually opens by-pass valve. Or, a power-actuated dump
valve (programmed to open automatically) by-passes the liquid during the start,
and the stop function.
A check valve is employed in the dump discharge line. lt remains shut as long
as the by-pass (dump) valve remains fully or partly open. (See Fig. 11.3.)
Thus, the liquid pressure exerted on the plungers (or pistons) is largely that
caused by liquid mass and friction as it passes through the by-pass valve and pip-
ing. When correctly sized, the by-pass valve and piping cause low back-pressure.
Only a relatively small torque is required while by-passing the liquid to tank.
With liquid by-pass, the total starting torque requirement is mainly related to
the mechanical inertia of the pump, couplings, gears and motor rotor. These items
are heavy, and substantial starting torque may result. AH the liquid in the pump
suction line and in the by-pass line must be accelerated from stand-still to full
liquid velocities.
The torque needed to accelerate the entire mechanical hydraulic inertia system
depends on the inertia of all the moving parts (including liquid) and the rate of
acceleration, plus total system friction.
The rate of acceleration is very important. The starting torque is directly pro-
portional to it. (Stated another way: Accelerating peak torque is inversely propor-
tional to the time duration of acceleration.)
A few pumps are provided with mechanical suction valve unloader devices.
These devices reduce the total starting torque requirement to that needed to accel-
erate the mechanical items, such as pump crankshaft, gears, couplings,. etc. They
Figure 11.3. Schematics of liquid bypass systems.
350
11.2 HYDRAUUC INSTITUTE STANDARDS OF APPUCAT!ON 351
stop liquid pumping action by mechanically holding open the suction

valves, allowing the liquid inducted into each liquid cylinder to be delivered back
into the pump suction. Because no liquid pumping occurs, the driving machinery
need provide no torque to accelerate liquid. Only the torque to overcome mechan-
ical inertia and friction is needed during the start.
After the pump and driver have reached ful! speed, the suction valve unloading
devices are retracted and normal pumping action then commences. At this point,
the driver must supply additional torque to accelerate the liquid system and also
to meet the total running torque requirement caused by discharge pressure.
Another design used to facilitate starting is the use of a hydraulic drive coupling
or eddy-current drive coupling. These cause the pump to accelerate more slowly,
while the driver quickly reaches füll speed. Such a drive coupling permits a re-
duced rate of pump and liquid acceleration, reducíng the torque required.
Reduced voltage staiting, with less shock to the pump, the gearing, and the
electrical system may be used. Reduced voltage starters include the primary resis-
tor, primary reactor, auto-transformer, part-winding, and wye-delta types. AH re-
duce the motor torque (and amperage) developed, since all reduce the starting
voltage effectively applied to the motor. Hence, the pump always must be un-
loaded, or the motor may not accelerate to full-load speed.
Reduced voltage starting causes the least disturbance to line voltage and avoids
the high torsional stresses, and gear tooth loadings which accompany full-voltage,
full-load starting.
Experience indicates that a normal-starting-torque electric induction motor [see
Table 11.1] provides adequate starting torque when a by-pass valve, or suction
unloader device is used. (Locked-rotor torque is that developed at standstill by an
induction motor, when full line voltage is applied.)
Special conditions, such as starting against discharge pressure, may require a
special dliver with high-starting-torque capability.
Starting Without Liquld By-Pass

Starting a power pump against discharge pressure may reduce the life of power
end bearings and other pump and drive components, and may prevent complete
priming ofthe liquid end (resulting in rough operation and reduced capacity). This
practice should therefore be avoided except in emergency situations.
Starting without liquid by-pass may be divided into two categories:
First, there are multiple-pump applications where the pump must start against
full line pressure.
This discharge pressure already exists, having been developed by the other
pumps and the starting pump must accelerate against it. No liquid by-passing, or
suction valve unloading is provided, and it may be termed "full-load" starting.
With full-load starting, the torque requirement is high, since the driver must
accelerate itself, couplings, gears, pump crankshaft, rods, crossheads, and plung-
ers. Additionally, it must accelerate all the liquid in the pump's suction and dis-
charge lines. It must also develop the torque required to move the plungers (or
pistons) against the line pressure, already present.
352 APPLICATIONS
If the pump is engine-driven, a clutch or drive coupling of adequate torque and

thermal capacity to meet these demands is chosen.
Full-load across-the-line motor starting can cause a prolonged in-rush of high
current. Serious motor over-heating (or damage) may result unless the motor and
its starting equipment have been generously selected and sized to meet this. An
A.C. induction motor provides a fixed starting torque, related to the applied volt-
age.
Since the starting torque developed by an induction motor is related only to the
applied voltage, and to the size and design of the motor, an across-the-line starter
will provide the maximum starting torque. And, a given induction motor will de-
velop the same locked-rotor (starting) torque and amperage at this voltage, regard-
less of the nature of the driven load. High inertia loads (within motor capability)
will símply require longer accelerating time than will low inertia loads.
No advantage is really provided by specifying an induction motor with an ex-
tremely high locked-rotor torque rating. Such motors do accelerate faster, but they
draw more amperage, and cause more power system disturbance during the start.
An A.C. induction motor with a locked-rotor torque rating of 150% of full-load
torque is usually sufficient for full-load, across-the-line pump starting.
Secondly, another full-load starting situation may occur when a single pump
starts with no liquid by-passing provided. Then, pump discharge pressure will be
related very largely to pump speed {discharge rate) and acceleration rate.
Consider a single pump which forces fluid through a long piping system which
is nearly level. Pump discharge pressure is zero at the instant of starting. Because
resistance to flow is caused by liquid inertia and by pipe friction and fitting restric-
tions, pump pressure increases in relation to liquid velocity, often such that pump
pressure is proportional to the square of ftow rate. Thus, it is proportional to the
square of pump speed, and to the rate of acceleration.
Analysis of the starting torque requirement is complex, and depends on the
inertia of the accelerating liquid, the size and length of piping,.liquid viscosity and
density, and on the elasticity of the piping. If the mass of liquid is very large, a
discharge ~ampener and a check valve may be advantageous, since these permit
the pump and its driver to accelerate faster by first delivering liquid into the damp-
ener, rather than into the line.
No general rules may be given as to the motor locked-rotor torque needed, since
each liquid system is different. However, if a large liquid dampener is provided,
pipeline pump induction motors with locked-rotor torques of 125% of rated full-
load torque will usually be adequate. An alternate is the automatic by-pass valve
and check valve arrangement, with dampener as shown in Fig. 11.3.
The arrangement shown in Fig. 11.3 is suitable for both single and multiple
pump applications. It largely relieves the motor of load from liquid inertia (except
suction), and isolates the pump from the discharge system pressure and inertia.
The arrangement affords a convenient means of expelling any air trapped in the
pump cylinders before placing pressure load on the pump. This is desirable, es-
pecially for multicylinder pumps which sometimes become rough and "air bound"
after servicing or prolonged idleness. Simply open the liquid by-pass valve to al-
TABLE 11.1. Minimum Locked-Rotor Torque Ratings
Rating l!!OORPM 1200RPM

(HP) Motora Moton;
1 275% 170"/o
1-l/2 250",4 165%
2 2350/o 160"/o
3 215% 155%
5 1850/.> 150"/o
7-112 175% 150"/o
10 165% 150\'lo
15 160% 1500/o
20 150"A. 135%
25 150".4 135%
30 1500/o 135%
40 140'!/ó 135'%
50 140"/o 135%
60 i40"/o 135"/o
75 140"/o 135%
100 12.SOA 125"/Ó
125 110".4 125"/o
150 110"/o 1200/o
200 100% 120%
250 80";(, lOO"Ai
300 80"/o 100"/o
350 80"/o HJOO/o
400 80"/o
450 80".4.
500 80".4
Note: In the range from 1 through 75 horsepower, the 1800
RPM motors show higher locked-rotor torque ratings than do the
1200 RPM motors. However, from 125 through 350
horsepower, the 1200 RPM motors have larger NEMA ratings.
Locked rotor torques ol large motors must be carefully evalu-
ated befon.! final selection.
low the liquid to discharge back to the tank thereby expelling the air. When run-
ning smoothly, close by-pass valve and thus load the pump.
Electric Motor Locked~Rotor Torques

Table 11.1 summarizes minimum locked rotor torque ratings for standard NEMA
Design "B" 60 Hertz squirrel-cage induction motors expressed as percent of fuU-
load torque.
lnlet System for Power Pumps

An inlet system for a reciprocating power pump must provide a flow of liquid, at
a relatively constant pressure, to the pump at a pressure sufficiently above vapor
pressure to prevent flashing as the liquid enters the pump chambers. If gas bubbles
are entrained in the liquid, or if flashing occurs in the pump, damaging vibrations
may occur in both inlet and outlet lines, volumetric efficiency will drop, and var-
ious pump and system components may foil. Small amounts of gas or cavitation
can reduce Hfe of packing, valve springs, valves, seats and gaskets. Larger quan-
354 APPLICATIONS
tities of gas, or more severe cavitation, can cause pitting of liquid end components
and catastrophic failure ofthe liquid cylinder, crankshaft, bealings, and drive train
components.
It is recommended that the design of the inlet system for a power pump follow
these guidelines:
1. The liquid source shown as a tank in Fig. 11.4, should be designed with
the following features:
a. Sufficient size to allow entrained gas bubbles to rise to the surface.
b. Lines which feed liquid into tank below minimum liquid level.
c. Completely submerged baffle plate separating incoming from outgoing
liquid.
d. Vortex breaker at outlet connection (to pump).
2. Each pump should be provided with a separate inlet line from liquid source
to pump, rather than connecting two or more pumps to a common manifold. Mu-
tually reinforcing pulsations are thus avoided.
3. Inlet pipe diameter should be at least equal to, and preferably larger than,
pump inlet connection.
4. Inlet pipe should be as short and direct as possible with a minimum of
tums, bends, and restrictions. All turns should be made with long-radius elbows
or laterals. Pulsations resulting from a long inlet line {:an sometimes be partially
reduced by a pulsation dampener and sometimes by raising the liquid level at the
source, but these changes seldom provide results as satisfactory as a short, direct,
large-diameter line.
5. The inlet system must provide NPSH that exceeds the sum of the NPSHR
of the pump, all friction losses, and acceleration head. Additional head must be
BAO OES!GN GOOD DESIGN
Figure 11.4, Suction tanks.

provided if the liquid contains dissolved gases. It is recommended that a margin

of at least 7 feet be provided.
6. The inlet system should contain no high points that would conect gas. An
"horizontal" runs should slope up toward the pump. The pipe reducer at the pump
inlet should be of the eccentric type, instaned with the flat side up.
7. A strainer, if used, should have a free flow area at least three times the
flow area of the inlet line. If there is doubt about its regular maintenance, a strainer
should not be used. (A plugged strainer may cause more damage to a pump than
solids.)
8. The inlet line valve should have a flow area equal to that of the inlet line.
9. If a foot val ve is used (for a source liquid level below the pump inlet open-
ing), the net flow area should at least equal the flow area of the inlet line.
10. An inlet pressure gauge should be located adjacent to the pump.
If a system will not provide sufficient NPSH, and cannot be redesigned, it shan
be necessary to do one or more of the fonowing:
1. Install pulsation dampener in inlet line adjacent to power pump liquid cyl-
inder. A dampener, properly instaned and charged, may significantly reduce the
length of pipe used in the acceleration head equation (see Pulsation Dampener,
fonowing).
2. Reduce the power pump NPSHR by selecting a larger, lower-speed unit.
The lower speed will also reduce acceleration head.
3. Instan a booster (charge) pump.
A booster pump for a power pump is normany a centrifuga! pump, but may be
a positive displacement pump under special conditions. Care must be exercised in
the selection and installation of a booster pump, because improper selection and/
or instanation can result in increased pulsations and attendant problems. In addi-
tion to the recommendations contained in the appropriate section of these Stan-
dards, the fonowing are recommended:
1. Instan booster pump as close to inlet source as practica!.

2. The booster pump must add enough pressure to the system to provide suf-
ficient NPSH to the power pump allowing for the acceleration head and friction
losses.
3. Install pulsation dampener in inlet line adjacent to power pump liquid cyl-
inder (or if of proper construction, on the opposite side of cylinder). The dampener
is often omitted, though, between a centrifuga! booster pump and a low-speed-
power pump under the fonowing conditions:
a. Diameters of inlet and outlet connections of booster pump are equal to,
or larger than, inlet connection on power pump.
356 APPLICATIONS
b. Diameters of all piping between liquid source and power pump are equal
to, or larger than, inlet connection of power pump.
c. The booster pump is sized for ffi<l;Ximum instantaneous capacity of the
power pump. The following tabulation gives the percentage that the max-
imum instantaneous capacity exceeds the mean capacity for each type of
power pump.
% Over
Mean
Type of Power Pump Capacity
Simplex (1), Single-Acting 220%
Duplex (2), Single-Acting 60%
Duplex (2), Double-Acting 27%
Triplex (3), Single or Double-Acting 7%
Quintuplex (5), Single or Double-Acting 2%
Septuplex (7), Single or Double-Acting 1%
Nonuplex (9), Single or Double-Acting 1%
d. Acceleration head is calculated, not only between booster and power

pump, but also between liquid source and booster.
4. If the booster pump is a constant-speed positive-displacement pump (such
as a motor-driven rotary), a self-regulating by-pass valve is required between
pumps. The booster pump must be sufficiently over-sized to provide the minimum
ftow required through the by-pass valve.
Suctlon Tarll<
It is recommended that a suction tank, if used, be arranged as shown [in Fig. 11.4]
for retum and make-up lines.
Pulsation Dampener
A pulsation dampener is a device which reduces liquid pulsations in the suction or
discharge piping. It is also sometimes called one of the following:
Suction Chamber Alleviator

Discharge Chamber Damper
Cushion Chamber Suction Bottle
Surge Chamber Inlet Bottle
Suction Stabilizer Stand Pipe
Desurger Air Chamber
Pulsation Suppressor Accumulator
1·1.2 HYDRAULIC INSTITUTE STANDARDS OF APPLICATION 357
Where the suction or discharge line, or both, are of considerable length, or if

the suction is under a static head of poor design, or where the liquid handled is
hot, a desurging device of suitable size for the suction or discharge lines, or both,
may sometimes be necessacy to insure smooth, quiet operation of the system. The
size of the pulsation dampener will depend upon the type, size, and speed
the liquid, and the layout of the piping systems. Recommendations as to size and
type of pulsation dampener should be obtained from the pulsation dampener man-
ufacturer.
Pulsation dampeners, if used, should be considered as a part of the piping sys-
tem, rather than as a pump accessocy.
In general, suction pulsation dampeners are more frequently required than are
discharge pulsation dampeners. In the following list, pump types are arranged in
order of their relative smoothness of fiow.
Nonuplex Power Pump-Single Acting

Septuplex Power Pump-Single Acting
Quintuplex Power Pump-Single and Double Acting
Triplex Power Pump-Single and Double Acting
Duplex Steam Pump-Double Acting
Duplex Power Pump-Single and Double Acting
Sirnplex Power Pump-Double Acting
Simplex Steam Pump-Double Acting
Simplex Power Pump-Single Acting
For pulsation dampeners which require a gas charge, provision should be made
to keep them charged with nitrogen or a similar inert gas. A liquid level gauge is
desirable to permit a check on the amount of air in a conventional air chamber.
Pulsation dampeners, particularly on the suction, should be located as close as
possible to the pump and in such position that they will absorb the impact of the
moving liquid column and thus cushion the pulsations in the most efficient manner.
On high speed power pumps, the chamber air volume can be at least 1 to l~
times the pump displacement per revolution.
Discharge Piplng
The facilitate starting and eliminate air, a by-pass valve should be installed close
to the pump. Also, to protect the pump, a stop valve and a check valve should be
employed. If an increaser is used to increase the size of the piping, it should be
placed between check valve and pump.
The vibration in pipe lines must be minimized, with lines as short and as direct
as practicable. When a change in direction is required, long radius elbows and
trees ora 45 degree bend in the pipe itself should be utilized. "Dead ends" must
be avoided.
Hydraulic systems, using quick closing valves or similar mechanisms, must
358 APPLICATIONS
provide sorne means for absorbing the shock resulting from the sudden opening or
closing of valves.
Adequate provisions should be made for anchorage of high pressure piping.
Relief Va/ve
The insertion of a discharge relief valve of suitable size for the capacity of the
pump, set to open ata pressure above the operating discharge pressure required of
the pump, is mandatory because of the safety it affords. The relief valve should
be placed in the discharge line close to the pump and ahead of any other valves.
Full-opening, pilot-operated, or shear pin relief valves, or burst-discs, which
require little or no over-pressure to develop fully-open ftow capacity are often
preferable to spring-loaded relief valves. Spring types may require considerable
over-pressure to compress the spring, before becoming fully open, creating pump
overload.
Pressure-relief valves have a "set" pressure, which may be described as the
pressure at which the relief valve cracks and begins to open, allowing sorne ftow
to pass through. As additional pressure is applied, above this "set" or "crack"
pressure, the spring-loaded relief valve will gradually increase its port area until
the valve is fully open.
When fully open, the relief valve must have sufficient capacity so it will relieve
the full capacity of the pump without excessive over-pressure. Available spring-
loaded relief valves differ among manufacturers in the extent of over-pressure
needed to open from the barely cracked to the fully open, fully relieved position.
This range is generally 10 percent to 25 percent above the set pressure, depending
on spring design. By choosing a larger valve, this increase may be reduced.
Calculating Vo/umetric Efficiency For Water

The volumetric efficiency of a reciprocating pump, based on capacity at suction
conditions, using table of water compressibility, shall be calculated as follows:
1 - [Ptd,81 (1 + c/d)]
Vol. Eff. =------- - S
1 - Ptd.81
where
,81 = Compressibility factor at temperature t (degrees Fahrenheit or centi-

grade). (See Tables 11.2 and 11.3.)
e = Liquid chamber volume in the passages of chamber between valves
when plunger is at the end of discharge stroke in cu in [see Fig. 11.5]
d = Volume displacement per plunger in cu in [see Fig. 11.5]
P1d = Discharge pressure minus suction pressure in psi
S = Slip, expressed in decimal value
TABLE 11.2. Water Compressibiiity
Compressibility Factor p, x 1o- 6 :::: Contraction In f.!nit Vol u me per PSI Pressure Compressibility from 14. 7 PSIA, 32 to
212ºF and from Saturatlon Pressure above 212ºF
Tt1mp11rntur11
Pressu"' oc l!OC 40C 60C ~c~cmc~c~c~c~c~c~c~c~c~c~c~c~c
Na 321' lli3 I' 104 F 140 F 176 F 212 f 248 F 284 F 320 F 356 F 3!12 F 428 F 464 F 500 F 536 F 572 f 6011 f 644 F 6110 F
200 3.12 3.06 3.06 3.12 3.23 3.40 3.66 4.00 4.47 5.11 6.00 7.27
400 3.11 3.05 3.05 3.11 3.22 3.39 3.64 3.99 4.45 5.09 5.97 7.21
600 3.10 3.05 3.05 3.10 3.21 3.39 3.63 3.97 4.44 5.07 5.93 7.15 8.95
800 3.10 3.04 3.04 3.09 3.21 3.38 3.62 3.96 4.42 5.04 5.90 7.10 8.85 11.6
1000 3.09 3.03 3.03 3.09 3.20 3.37 3.61 3.95 4.40 5.02 5.87 7.05 8.76 11.4 16.0
1200 3.08 3.02 3.02 3.08 3.19 3.36 3.60 3.94 4.39 5.00 5.84 7.00 8.68 11.2 15.4
1400 3.07 3.01 3.01 3.07 3.18 3.35 3.59 3.92 4.37 4.98 5.81 6.95 8.61 11.1 15. l 23.0
1600 3.06 3.00 3.00 3.06 3.17 3.34 3.58 3.91 4.35 4.96 5.78 6.91 8.53 10.9 14.8 21.9
1800 3.05 2.99 3.00 3.05 3.16 3.33 3.57 3.90 4.34 4.94 5.75 6.87 8.47 10.8 14.6 21.2 36.9
2000 3.04 2.99 2.99 3.04 3.15 3.32 3.56 3.88 4.32 4.91 5.72 6.83 8.40 10. 7 14.3 20.7 34.7
2200 3.03 2.98 2.98 3.04 3.14 3.31 3.55 3.87 4.31 4.89 5.69 6.78 8.33 10.6 14.1 20.2 32.9 86.4
2400 3.02 2.97 2.97 3.03 3.14 3.30 3.54 3.85 4.29 4.87 5.66 6.74 8.26 10.5 13.9 19.8 31.6 69.1
2600 3.01 2.96 2.96 3.02 3.13 3.29 3.53 3.85 4.28 4.85 5.63 6.70 8.20 10.4 13.7 19.4 30.5 61.7
2800 3.00 2.95 2.96 3.01 3.12 3.28 3.52 3.83 4.26 4.83 5.61 6.66 8.14 10.3 13.5 19.0 29.6 57.2 238.2
3000 3.00 2.94 2.95 3.00 3.11 3.28 3.51 3.82 4.25 4.81 5.58 6.62 8.08 10.2 13.4 18.6 28.7 53.8 193.4
3200 2.99 2.94 2.94 3.00 3.10 3.27 3.50 3.81 4.23 4.79 5.55 6.58 8.02 10.l 13.2 18.3 27.9 51.0 161.0
3400 2.98 2.93 2.93 2.99 3.09 3.26 3.49 3.80 4.22 4.78 5.53 6.54 7.96 9.98 13.0 17.9 27.1 48.6 138.1
3600 2.97 2.92 2.93 2.98 3.09 3.25 3.48 3.79 4.20 4.76 5.50 6.51 7.90 9.89 12.9 17.6 26.4 45.4 122.4
3800 2.96 2.91 2.92 2.97 3.08 3.24 3.47 3.78 4.19 4.'14 5.47 6.47 7.84 9.79 12.7 17.3 25.8 44.5 110.8
4000 2.95 2.90 2.91 2.97 3.07 3.23 3.46 3.76 4.17 4.72 5.45 6.43 7.78 90 70 12.5 17. l 25.2 42.8 101.5
----"---·---
4200 2.95 2.90 2.90 2.96 3.06 3.22 3.45 3.75 4.16 4.70 5.42 6.40 7.73 9.62 12.4 16.8 24.6 41.3 93.9
4400 2.94 2.89 2.90 2.95 3.05 3.21 3.44 3.74 4.14 4.68 5.40 6.36 7.68 9.53 12.2 16.5 24.1 40.0 87.6
4600 2.93 2.83 2.89 2.94 3.05 3.20 3.43 3.73 4.13 4.66 5.37 6.32 7.62 9.44 12.1 16.3 23.6 38.8 82.3
4800 2.92 2.87 2.88 2.94 3.04 3.20 3.42 3.72 4.12 4.64 5.35 6.29 7.57 9.36 12.0 16.0 23.2 3706 77.7
5000 2.91 2.87 2.87 2.93 3.03 3.10 3.41 3.71 4.10 4.63 5.32 6.25 7.52 9028 11.8 1508 22.7 36.6 73.9
5200 2.90 2085 2.87 2.92 3.02 3.18 3.40 3.69 4.09 4.61 5.30 6.22 7.47 9.19 11.7 15.6 22.3 35.6 70.3
5400 2.90 2.85 2.86 2.91 3.01 3.17 3.39 3.68 4.07 4.59 5.27 6.19 7.41 9.12 11.6 15.3 21.9 34.6 66.9
w
(11
fD
TABLE 11.3. Water Compressibility
Compressibility Factor '3t x 10- 6 = Contraction in Unit Volume per PSI Pressure
Compressibility from 14.7 PSIA at 68ºF
Temperalure Temperalore
Pre1111UR 20C lOOC 200C Pressure 20C lOOC 200C

PSIA .. 68F 212F 392F PSIA 68F 212F 392F
6000" 2.84 3.14 S.20 22000 2.61 2.42 3.7S
7000 2.82 3.10 5.09 23000 2.59 2.38 3.68
8000 2.80 3.05 4.97 24000 2.S8 2.33 3.61
9000 2.78 3.01 4.87 25000 2.S7 2.29 3.SS
10000 2.76 2.96 4.76 26000 2.56 2.24 3.49
11000 2.75 2.92 4.66 27000 2.6.5 2.20 3.43
12000 2.73 2.87 4.57 28000 2.SS 2.1.5 3.37
13000 2.71 2.83 4.47 29000 2..54 2.11 3.31
14000 2.70 2.78 4.38 30000 2.53 2.06 3.26
15000 2.69 2.74 4.29 31000 2.52 2.02 3.21
16000 2.67 2.69 4.21 32000 2.51 1.97 3.16
17000 2.66 2.6.5 4.13 33000 2.50 1.93 3.11
18000 2.65 2.60 4.05 34000 2.49 1.88 3.07
19000 2.64 2..56 3.97 35000 2.49 1.84 3.03
20000 2.63 2.51 3.89 36000 2.48 1.79 2.99
21000 2.62 2.47 3.82
Figure 11.5. Plunger movement when calculating volumetric clearance.
360
Example. Find the volumetric effi.ciency of a reciprocating pump with the fol-
lowing conditions:
Type ofpump 3 in diam plunger X 5 in stroke triplex .

Liquid pumped Water
Suction pressure Zero psig
Discharge pressure 1785 psig
Pumping temperature 140 F
e 127.42 cu in
d 35.343 cu in
s .02
Find (j1 from Table of Water Compressibility (Table 11.2).
(j1 = .00000305 at 140 F and 1800 psia
Calculate volumetric effi.ciency:
1 - [Ptd[:J,(1 + c/d))
Vol. Eff. ~~~~~~~~- - s
1 - Ptd(:Jt
1 - [(1785 - 0)(.00000305)[1 + 127.42/35.343])

~~~~~~~~~~~~~~~~~~~-- 02
1 - (1785 - 0)(.00000305) .
= .96026 = 96 per cent
Calculating Volumetric Efficiency For Hydrocarbons

The volumetric effi.ciency of a reciprocating pump based on capacity at suction
conditions, using compressibility factors for hydrocarbons, shall be calculated as
follows:
Vol. Eff. = 1 - [S - ~(1- ::) J

where
e = fluid chamber volume in the passages of chamber between

valves, when plunger is at the end of discharge stroke, in cubic
inches
d = volume displacement per plunger, in cubic inches
P = pressure in psia (Ps = suction pressure in psia; Pd = discharge
pressure in psia)
Pe = critcal pressure of liquid in psia (See Table 11.4.)
362 APPLICATIONS
TABLE 11.4. Physical Properties of Hydrocarbons
Te Pe rilo 1/
sigma 1
Carboil Degrees lb. per grams
Atoms Na me Ra11k111e 5q.ln. per ce
1 Metlia11e 343 673 3.679
2 E tila ne 550 717 4.429
3 Peo pan e 666 642 4.BOJ
4 Bu ta ne 766 544 5.002
5 Pentane 847 482 5.128
6 He:urne 915 433 5.216
7 Heptlrne 972 394 5.285
8 Octane 1025 362 5.349
9 Nonane 1073 332 5.3112
10 Decane 1114 308 5.414
12 Dodecane 1185 272 5.459
14 Tetradecane 1248 244 5.483
16 Hexadecane 1900 221 "5.48
18 Octadeca11e 1345 202 "5.49
20 E.lcosane 1380 187 •s.5
25 Pentacosane 1465 156 "5.5
30 Triacont.ane 1545 133 "!5.5
35 Pentatrlacontane 1610 120 "5.5
40 Tetracontane 1675 108 •5_5
45 Pentat.etracontane 1740 100 "5.4
*Based on experimental density, questionable because
near melting point.
Pr = reduced pressure
actual pressure in psia P
critical pressure in psia Pe
p rs = reduced suction pressure = p s /pe
P,d = reduced discharge pressure = Pd/ Pe
S = slip expressed in decimal value
t = temperature, in degrees Rankine
= degrees F + 460 (ts = suction temperature in degrees Rankine;
td = discharge temperature in degrees Rankine)
Te = critica! temperature of liquid in degrees Rankine (see Table
11.4.)
T, = reduced temperature
= actual temp.· in degrees Rank:ine = - (see Fig. 11.6)
critical temp. in degrees Rankine Te
T,s = reduced suction temperature = t5 /Te
T,d = reduced discharge temperaturt: = td/Tc
Vol. Eff. = volumetric efficiency expressed in decimal value.
p = ( p 1 / w1 ) X w X 62.4 = density of liquid in lb per cu ft
Ps = density in lb per cu ft at suction pressure
11.2 HYDRAULIC INSTITUTE STANDARDS OF APPUCATION 363
Pd = density in lb per cu ft at discharge pressure

w = expansion factor of liquid (see Fig. 11
p 1 / w = characteristic constant in grams per cubic centimeter for any
one liquid which is established by density measurements and
the corresponding values of w (see Table 11.4)
Example. Find volumetric e:fficiency of the reciprocating pump exam-

ple with the following new conditions:
Type of Pump 3 inch dia. plunger x 5 inch stroke triplex

Liquid pumped Propane
Suction temperature 70 F
Discharge temperature 80 F
Suction pressure 242 psig
Discharge pressure 1911 psig
1.0
F-
z
otL
-'
<(
(.) 0.9
¡:::
¡:;::
(.)
0,8
¡;..."
.,
'-
~
.,c.í:.'
0.7 E
.l!!
.,u
u
.,
::J
u
o::
0.6
'--~...1-.~.....;_~-'-~--'l...-~..__~......._~_,_~ .......~--'~--'0.4
.14 .13 .12 .11 .10 .09 .08 .07 .06 .05 .04
Expansion factor w
Figure 11.6. Thennal expansion and compressibility of liquids.

364 APPLICATIONS
Find density at suction pressure:
t5 460 + 70
T
rs
= -
Te
=
666
= .795
Ps 257
P,s = - = 642 = .4
P1
- = 4.803 {From Table 11.4, propane)
W¡
w = .1048 (From Fig. 11.6)
P1
Ps = - X w X 62.4
W¡
= 4.803 X .1048 X 62.4 = 31.4 lb per cu ft
Find density at discharge pressure:
T = td = 460 + 80 = 81
rd Te 666 .
pd 1926
p rd =p- = - =30
642 .
e
w = .1089 [from Fig. 11.6]
Pi
Pd = - X w X 62.4
W¡
= 4.803 X .1089 X 62.4

= 32.64 lb per cu ft
Therefore
Vol. Eff. = l - [ S - ~(1 - ::) J

= l - [ .02 - ~~~~~~ ( 1 - 332¡~) J
= .8376 = 83.76 per cent
Friction Head (h, ). Friction head is the hydraulic pressure required to overcome
frictional resistance of a piping system to liquid flow.
Static Head (hs)· Static suction head is the vertical distance from the centerline
of the pump inlet port to the liquid level at the source of supply.
Velocity Head (hv)· Velocity head is the hydraulic pressure needed to move
fluid from rest to the average velocity.
Submerged Suction. A "submerged" suction exists when the centerline of the

pump inlet port is below the level of the liquid in the supply tank. However, the
absolute pressure of the liquid entering the centerline of the pump inlet port may
be below atmospheric pressure when the pump is operating at the specified speed.
This will occur whenever friction head exceeds the static suction head (submer-
gence) of the pump.
Flooded Suction. "Flooded" suction implies that the liquid will ftow from an
atmospheric source to the pump without the average pressure at the intake port of
the pump dropping below atmospheric pressure with the pump operating at spec-
ified capacity.
Thus, the static suction head must always exceed friction head in a "flooded"
suction situation.
Static Suction Us ). Static suction lift is a hydraulic pressure below atmospheric

at the intake port of the pump with the liquid at rest. It is usually expressed in
inches of mercury vacuum (in Hg vac). To convert, use the formula, psi = 0.49
X inches Hg. Suction lift may be thought of as "negative" static suction head.
Net Positive Suction Head Available (NPSHA). Net positive suction head
available is the total suction pressure, including allowance for acceleration head
available from the system at the pump suction connection, minus the vapor pres-
sure of the liquid at the pumping temperature. NPSHA for a reciprocating pump
is normally expressed in pounds per square inch (psi) or feet.
Total Suction Lift. Total suction lift is the difference between the absolute op-
erating inlet pressure at the pump inlet port centerline and atmosphere pressure. It
is also the sum of suction system frictional losses and the static suction Hft.
Acceleratlon Head (h 2 )
Total suction lift, as defined in the preceding paragraph, represents the average
without reference to the fluctuation above and below this average due to the inertia
effect of the fluid mass in the suction line. With the higher rotative speed of pres-
ent-day pumps or with relatively long suction lines, this pressure ftuctuation or
acceleration head must be taken into account if the pump is to fill properly without
separation and pounding or vibration of the suction line.
With the slider-crank drive of a power pump, maximum plunger acceleration
occurs at the start or the end of each individual stroke, and this is refl.ected in a
366 APPLICATIONS
similar discontinuity in the cyclical pattern of the combined ftow curve correspond-
ing to each piston or plunger and crank arrangement. The head required to accel-
erate the fluid column is a function of the length of the suction line, the average
velocity in this line, the rotative speed, the type of pump, and the relative elasticity
of the fluid and the pipe and may be calculated as follows:
LVnC
h =--
ª Kg
where
ha = Acceleration head in feet

L = Length of suction line in feet
V = Velocity in suction line in fps
n = Pump speed in rpm
e = 0.200 for simplex double-acting
= 0.200 for duplex single-acting
= 0.115 for duplex double-acting
= 0.066 for triplex single or double-acting
= 0.040 for quintuplex single or double-acting
= 0.028 for septuplex, single or double-acting
= 0.022 for nonuplex, single or double-acting
K = A factor representing the relative compressibility of the liquid ( K =
1.4 for hot water; K = 2.5 for hot oil)
g = Gravitational constant (32.2 ft/sec 2 )
Note: The constant C will vary from these values for unusual ratios of
connecting rod length to crank radius.
The pulsation dampener properly installed near the pump with a short, full-size
connection to the pump or suction pipe can absorb the cyclical fiow variation and
reduce the pressure ftuctuation in the suction pipe to that corresponding to a length
of 5 to 15 pipe diameters, if kept properly charged.
There is a similar pressure fl.uctuation on the discharge side of every power
pump, but it cannot be analyzed as readily because of the great infiuence of liquid
and piping elasticity and the smaller diameter and much greater length of the dis-
charge line in most applications. However, a pulsation dampener can be just as
effective in absorbing the flow variation on the discharge side of the pump, as on
the suction side, and should be used if pressure-fluctuation or piping vibration is
a problem.
Example. Given a 2" x 5" triplex pump running at 360 rpm and displacing 73
gpm of water with a suction pipe made up of 4 feet of 4-inch pipe and 20 feet of
6-inch pipe:
11.3 THE DYNAMICS OF LIQUID PIPING SYSTEMS 367
Average velocity in 4-inch pipe
.321 X 73
V4 = 12.73 = 1.84 fps
Average veloclty in 6-lnch pipe
.321 X 73
V6 = 28 . 89 = 0.811 fps
Acceleration head in 4-inch pipe
h04 =4 X 1.84 X 360 X .066 = 3 .88 ft

1.4 X 32.2
Acceleration head in 6-inch pipe
h = 20 X .811 X 360 X .066 = 8 55 ft

ª6 1.4 X 32.2 .
Total acceleration head
ha = 3.88 + 8.55 = 12.43 ft
11.3 THE DYNAMICS OF LIQUID PIPING SYSTEMS*
Piping systems of ali kinds have exhibited vibration due to pulsation produced by
compressors or pumps. The problems and the solutions are slightly different in
liquid systems than in gas systems. Piping vibration in liquid systems has been
traditionally associated with a sudden change in ftow and is usually called water-
hammer. Although waterhammer does representa problem area, most piping vi-
bration problems in liquid systems today are caused by excitation from recipro-
cating pumps. Reciprocating pumps are generally used when the pressure ratio is
high. The high pressure on the discharge side of the pump makes safety a major
factor in the design and operation of the system. This sensitivity to safety and
reliability has caused the pump industry to increase its awareness of piping vibra-
tion problems. An understanding of three areas is essential to the proper design
and operation of a reciprocating pump system:
1. Pulsation generation by the pump
*Section 11.3 is included here by courtesy of the authors, C. L. Bates and L. E. Blodgelt, Department
of Applied Physics, Southwest Research lnstitute, December 1985.
368 APPLICATIONS
2. The acoustics of the piping system

3. The mechanical response of the piping system
This section will discuss techniques that the piping engineer can incorporate in
the design process to ensure that the installed system will not suffer excessive
piping vibration due to pulsation excitation. In order to properly apply the me-
chanical techniques it is necessary to understand the nature of the problem and the
mechanisms that cause excessive vibration.
11.3.1 Where Does Vibration Come From?

Excessive piping vibration in liquid systems is almost always due to the coinci-
dence of pump-induced pulsation and mechanical resonance. In like manner, ex-
cessive pulsation problems usually occur due to the coincidence of acoustic reso-
nance in the piping and the excitation of the pump plunger process. A reciprocating
machine will produce a degree of pulsation by the vary nature of the process in-
volved in moving the fluid. In plunger pump systems the pulsation frequency pro-
duced by the pump will be a multiple of the basic plunger frequency. The basic
plunger frequency is ( RPM / 60) X number of plungers. If the acoustic resonances
of the piping are coincident with the pump excitation, this pulsation can be mag-
nified by 5-50 times the original pulse. This acoustic coincidence points out the
need to have a knowledge of the acoustic response of the piping system. The most
efficient technique for determining the acoustic response of the piping system is to
use an electroacoustic analog such as that sponsored by the Southern Gas Asso-
ciation.
If the mechanical resonances of the piping system are coincident with this same
frequency, further magnification can occur in the 5-50 times range. It is easy to
see how a normal pump can produce enough pulsation to lead to a serious problem
if both mechanical and acoustical resonances coincide with a multiple of the pump
plunger excitation. Experience has shown that when both of these coincidences
occur, excessive vibration and excessive dynamic stress are certain to be a prob~
lem. Because the forcing function is coming from the pump through the acoustics,
it is instructive to understand the coupling mechanism by which this process oc-
curs.
11.3.2 How Does the Pulsation Couple into the Piping?

Dynamic pressure (pulsation) in a piping system can couple into the mechanical
system in two ways. The first is through simple pressure-areaforce. \\'."hen a pres-
SU!e acts against an area such as the end of a pipe section, a simple force is pro-
duced. When the pressure is dynamic (changing), the resultant force is also dy-
namic. The frequency of the dynamic force will be the same as the pulsation
frequency. The second mechanism by which energy can transfer is through the
change in momentum that occurs due to change in direction. It is a common prac-

tice to pipe liquid systems with long sweeping elbows. This will lessen the effect
of momentum coupling and reduce the dynamic shaking force coupled at the el-
bow. However, such elbows can be expensive and may not always be required.
This type of coupling also occurs at tees and other devices where a change in
direction of ftow occurs. It is important to realize that the point of coupling may
not be the actual location of the maximum vibration. The vibrational mode shape
(the vibration pattem at resonance) will usually control the location of maximum
vibration points. The forces in a piping system enter through coupling and can
transfer along the system by structural conduction. An area of piping that is very
susceptible to structural condition is piping racks where there is very little stmc-
tural rigidity to restrain the movement of forces in the system.
11.3.3 Calculatlng Mechanical Natural Frequency

In calculating the mechanical natural frequency of a piping system it is important
to include all significant factors and realize that the calculated natural frequency
is approximate at best. For example, it is possible to cakulate the natural fre-
quency and forget that the weight of the fluid inside the pipe is significant. De-
signers are warned to beware of piping natural frequency calculations where the
weight of the fluid has not been properly included. It is not uncommon to be off
by 30% if the weight of the fluid is not considered. Restraint points in a piping
system are difficult to estimate, and therefore the end conditions must be assumed
and a possible range of the natural frequency is the best that can be achieved.
With the proliferatíon of programmable hand-held calcuiators and microcom-
puters, it is most convenient to give the basic mathematical equations and allow
the engineer the choice of how to make the calculation. The basic relationship can
be summarized as
f = _µ,_ r;EI (11.1)

27rL2 ~~
where
f = natural frequency, Hz
µ, = frequency factor
7f = 3.14159
L = length of span, in
I = moment of inertia, in4
g = 386.088, inís2
E = pipe material modulus of elasticity, lb/ in2
w = weight of pipe per unit length, lb Í in
370 APPLICATIONS
The weight of the pipe is determined from
[1r(OD2 - ID 2 ) X Ps] + [1r X ID2 X PL]

w= ( 11.2)
4
where
OD = outer pipe diameter, in

ID = inside pipe diameter, in
p s = density of pipe material, lb/ in3
PL = density of liquid, lb/ in 3
The basic expression has been derived from simple beam theory, and the frequency
factors have been derived from either a theoretical end condition or from finite
element calculations. Frequency factors for idealized pipe span and bends are in-
cluded in Figure 11. 7. The ftequency factors reflect the accuracy generally re-
quired for design calculations.
Frequency
Factor
Beam Type µ.¡ P.2
Cantilever 3.52 22.0
Simply supported 9.87 39.5
Fixed-supported 15.8 50.0
Fixed-fixed 22.4 61.7
Free-free
= 22.4 61.7
Out of In
Plane Plane
L-bend, equal legs 3.74 15.4
U-bend, equal legs 2.00 3.1
Z-bend, equal legs 2.26 2.8
Figure 11. 7. Natural frequency factors and coefficients for idealized pipe spans and bends.
11.3 THE DYNAMICS OF LIOU!D PIPING SYSTEMS 371
10 1---91--+-
8 ,__..,__-+-+---+--+-
¡;. ~dlh
6 >--t---l--4--l--->--
51--1---+-+-+--+--+"""~-+-+-.¡._j~I--+--+---~
41--t---f-+-+--+--+--'~~~'+-'--+-+-+- - + - - - + - - - - - - t
31--+--1--1-~---+---t-~--''"""
"- 2 l--+--1--1--1--+---+--~
....
o
tí
~
(;'
:¡¡ l 1--t---t---+--+---t---t--·-+--+-
"'
~ .81--+--+--+--+---t--t---+-+---11~~..--'k---+-t----~
u..
.41--t--t--t---+---t---+---+-+--+-+--flr\ct-T---.
Out-of-plane vibration
.31--+--+--+--+---t---t----+-+--+-+-+-'<-+"<---+-~.-'«'---~
.l ........._._...__.__.___.~___,~-~-'-........-'-_...__,__.._~......._._.......~
108 6 5 4 3 1 .8 .6 .5 .4 .3 .2 .1
RatioL/h
Figure 11.8. Frequency factor ( µ) vs. ratio L / h for uniform U-bend.
Piping bends raise another unique problem in the calculation of natural fre-
quencies. To compensate for the added fiexibility of elbows in piping systems, a
plot of frequency factor versus configurational parameters has been included in
Figure 11.8. The information given on the plot can be used to calcula.te the me-
chanical natural frequencies of L-bends and U-bends with variable dimensions.
The effect of concentrated masses in a system can be obtained through the use
of Rayleigh's method. Applying this method, it can be shown that it is first nec-
essary to determine the natural frequency of a beam without the mass, and then
apply a correction factor to obtain the approximate solution. The equations and
correction factors are contained in Figure 11. 9.
11.3.4 Making the Application to Actual Piping

In making the calculation of mechanical response it is first necessary to define the
assurnptions that cause the natural frequency to deviate from the ideal.
372 APPLICATIONS
Correction Factor C for

Natural Frequency Equation
Beam Type
fp =fj~I + C~ (11.3)
Cantilever, load at end F===O 3.9
Simply supported, load at center 2.0
Fixed supported, load at center 2.3
Fixed-fixed, load at center 2.7
L-bend, load at center of leg,

equal legs (L)
U-bend, load at center,

, 0.28
O. 7 Out of plane
equal legs (L)
Z-bend, load at center,

n 0.61 In plane
0.68 Out of plane

equal legs (L)
~ 0.66 In plane
Figure 11.9. Concentrated weight-correction factors for ideal piping spans.
1. The restraint stiffness of a clamp may not be fixed or simply supported but
may be somewhere in between.
2. The beam will probably be continuous beyond the restraint points.
3. The pipé may not be uniform in the span.
4. The configuration may be more complex than allowed for in the simple cal-
culations.
5. Added masses are usually not ideally concentrated but are distributed over
sorne finite length of the pipe.
11.3.5 What about Generalized Finite-Element Analysis?

There are many finite-element computer programs available to make calculations
of complex three-dimensional systems. Such programs as STARDYNE, ANSYS,
NASTRAN, MARC, and ADINA are available through most computer systems.
In many cases these programs can be overkill from the standpoint of efficient de-
sign. If all simple spans were calculated as if they were complex, the cost to
dynamically design piping systems would more than triple. And no matter how
complex and rigorous the piping model, the same questions of end condition and
restraint effectiveness limit the accuracy of the calculations. lt is important to keep
in mind that a simple calculation with the proper assumptions is far superior to a
rigorous calculation with poor assumptions.
11.3.6 Design Philosophy

In designing a pump piping system to avoid or minimize vibration problems, the
desired goal is to avoid having any piping spans or piping components that are
mechanically resonant to any of the major excitation frequencies in the system. lf
this is accomplished, then the vibrations should be within acceptable criteria, as-
suming the basic supporting mechanisms are adequately designed. Safe vibration
limits in piping systems are a function of the dynamic stresses introduced by the
particular vibratory mode shape. However, experience with numerous piping vi-
bration and failure problems at existing installations and design experiences with
over 5000 pump and compressor installations have developed the criteria presented
in Figure 11.10. These simplified criteria are intended to provide rough guidelines
200
100
~'
80 '
60
40
-- -.....
.... -¡,;
'~·SI'
• /,o
,r--....~
~a,,
~
.."'
.><
c.
30
~ .....~ ........
C'o,.,,
.........
r-.......,
.$ 20 ...
·~~ ........
~r--....
""~
.><
al r-....
c.
.!!!
.E
~ ""i-..
.... ~
4'a,,. ~
ai
'O
10
8 - ll111a¡
......
- ...
:::1
:!: 6 ...... Des·
... , 1'-
a. -......;!1111 !'.,.
' ""'
~
E
4 ,r-..... ,r-......
"'o
e:
~"e~.,,
.... ""~ .............
3
~ ........
.a 2
...,~~efq
"" ~e
r-.... ~
~
~
> i"' ....
r--......
lfoe,,
~.Ot,·
~~
1
.8
.!Ot¡
,
...
' . ... -
.6
... ....
i"' ....
.4
.......
...............
.3
1 2 34 6810 20 30 40 60 80 100 200 300
Vibration frequency, Hz
Figure 11.10. Standardized piping vibration criteria. Note: Indicated vibration limits are
for average piping system constructed in accordance with good engineering practices. Make
additional allowances for critica! applications, unreinforced branch connections, etc.
374 APPLICATIONS
for evaluating piping vibrations to help determine if additional analysis or modi-

fications should be considered.
The design philosophy that has been successfully utilized for controlling piping
vibrations for reciprocating pump and compressor installations is to make the me-
chanical natural frequencies of the piping spans and components at least two times
the highest excitation frequency of concern. Thus, based on a knowledge of the
pulsation energy in the piping from an electroacoustic analog analysis or other
techniques, mechanical resonant frequency calculations can be made to determine
the spacing, location, and sizing of piping supports or restraints to minimize vi-
bration problems. From a dynamic standpoint, the primary function ofpiping sup-
ports should not be to hold the piping by "brote force" but to detune and dampen
the mechanical natural frequencies of concern.
One of the most important design considerations that can be used to help prevent
vibration problems is simply to eliminate all unnecessary piping bends, since they
provide a strong coupling point between pulsation excitation forces and the me-
chanical system. All bends that are necessary should be designed and located so
that they can be effectively supported. Another guideline is that supports should
be located at all large concentrated masses and discontinuities in the piping system.
This includes in-line valves such as block valves, regulator valves, and check
valves, and other components such as flanges.
Another area of particular concern is the design of small auxiliary piping con-
nections such as vents, drains, and pressure test connections. These are the most
common locations where vibration-induced piping failures are experienced in in-
dustrial plant systems. The typical design is such that relatively large valve and
flange masses are cantilevered out from the main piping on small piping niP.ples.
This results in a high-Q (amplification factor) mechanical system with a natural
frequency in the range of typical excitation forces and with a high stress concen-
tration factor in the connection area. In order to prevent this type of problem, the
number of small auxiliary branch connections should be minimized and the nec-
essary connections should be designed and supported to effectively eliminate rel-
ative vibration between the branch connectio~ mass and the main piping.
Design guidelines for controlling piping vibrations and failures for reciprocating
pump and compressor installations are summarized as follows:
Piping Dynamic Design Guidelines

1. The excitation forces in the system should be determined utilizing the best
suitable techniques available. For reciprocating pump and compressor installa-
tions, the use of the electroacoustical analog has been the most effective method
for designing satisfactory pulsation control.
2. For reciprocating pump and compressor piping, it is desirable to have the
mechanical natural frequencies of all piping spans and components at least two
times the highest pulsation excitation frequency of concern.
3. Eliminate as many piping bends as possible.
11.4 HIGH SUCTION PRESSURE 375
4. Add effective pipe supports at all necessary bends.

5. Add pipe supports at all large concentrated masses and piping discontinu-
ities (block valves, regulator valves, flanges, etc.).
6. Design and support small auxiliary piping components (vents, drains, in-
strument connections, etc.) to eliminate relative vibrations between the branch
connection masses (valves and flanges) and the main piping.
7. Do not use unreinforced branch connections.
8. Do not use piping fittings or reinforcements that are characterized by high-
stress intensification factors.
9. Do not use threaded connections.
10. Do not restrain the piping by directly welding to it.
11.4 HIGH SUCTION PRESSURE
With double-acting pumps, a high suction pressure actually helps to "unload" the
power end because the suction pressure is at all times reacting to the force on the
piston dueto the discharge pressure. This results, as far as the power end is con-
cemed, is a net hydraulic horsepower based on the difference between the dis-
charge and suction pressures, and the power end is "unloaded" with respect to
stress in proportion to the applied suction pressure. Accordingly, the horsepower
requirement can be calculated on the pressure difference and with a basic mechan-
ical efficiency of 85%, using formula (11.4).
In a single-acting pump, an entirely different condition exists. In such a pump
the configuration is such that a high suction pressure acting on the piston or plunger
is added to that imposed by the discharge pressure, this total load being transmitted
to the crankshaft and associated bearings, etc. In a triplex pump (one with three
cylinders), two plungers or pistons always have either suction or discharge pres-
sure at the same time while the third has discharge or suction pressure, respec-
tively.
A quintuplex pump (five cylinders) can have discharge pressure on three plung-
ers or pistons and suction pressure on the other two, or vice versa.
In view of these facts, when the suction pressure exceeds about 5 % of the
discharge pressure, it becomes necessary to take this into account so that the power
end of the pump will not be overloaded. On this basis, formulas (11.5)-(11. 9) are
used to arrive at an artificial discharge pressure for determining a reduced plunger
or piston diameter.
For Double-Acting Pumps:
GPM
BHP = (Pd - Ps) X 1714 X .85 ( 11.4)
376 APPLICATIONS
For Single-Acting Pumps:
GPM
BHP = (Pd - P,) X -- X ME* ( 11.5)
1714
Triplex pumps: (11.6)
Quintuplex pumps: (11.7)
Septuplex pumps: (11.8)
Nonuplex pumps: (11.9)
where
P = artificial discharge pressure upon which plunger/piston is selected

when suction pressure is in excess of 5 % of discharge pressure
Pd = actual discharge pressure
Ps = actual suction pressure
BHP = input brake horsepower
GPM =actual U.S. GPM
*ME = mechanical efficiency from Figure 11.11
Example
A certain triplex pump rated at 1000 PSI with 3-in plungers ( PRL 7200 lb) would
have to be equipped with 2~-in plungers ( 1250 PSI at PRL 7200 lb) to pump at
actual 1000 PSI with 500 PSI suction pressure.
P = 1000 + 500 /2 = 1250 PSI
11.5 SIZING PUMPS FOR PIPELINES"
In designing a proposed pipeline that involves a large number of pumps and sta-
tions, several factors are influential in determining the optimum relationship be-
tween station capacity and reciprocating pump size. Total cost, including pump
*This section is an excerpt from a paper entitled "Sizing Pumps for Long Slurry Pipelines" prepared
by J. E. Miller and Will Smith, Worthington Group, Dresser Industries. The paper was presented by
J. E. Miller at the English lnternational Conference on Slurry Transportation, March 15-18, 1983 in
San Francisco, CA, and .is excerpted by permission of Worthington.
115 SIZING PUMPS FOR PIPELINES 377
0.8
r:i..."'
'Q, ~
o 0.6
:;::;
l.':
~
::i
"'~
V>
0.4
c..
0.2
0.2 0.4 0.6 0.8 0.9 1.0

Mechanical etticiency ME
Figure 11.11. Plunger pump mechanical effi.ciency vs. suction pressure/discharge pres-
sure. To determine BHP: Calculate pressure ratio. Detennine mechanical effi.ciency from
curve. Substitute in:
BHP = (Pd - P,) X GPM/1714 X ME* ( 11.5)
where Pd = discharge pressure, PSI, and P, = suction pressure, PSI.
first cost, installation costs, and cost of operation and m.üntenance, is obviously
an important consideration.
Certainly the number of operating pumps in a station related to pump
size) must permit practical operation over the full range of flow rates expected for
the pipeline. Thus, a relatively clear-cut solution can be developed to establish the
appropriate pump size for a given pipeline.
11 .5.1 Pump Cost

To construct a pump-cost model, it is first necessary to have or assign viable values
for the original cost of new pumps, dampeners, and expendable parts, as well as
power costs. It is desirable to include the cost of dampeners in a study of this kind
because the pulsation-control equipment required for reciprocating pumps repre-
sents a considerable portion of the first cost of an installation.
To study the fundamental effect of purnp sizing, models for such a study can
be selected from an existing family of pumps or they can be of odd sizes and types,
or, if the project warrants, newly designed sizes and types can be considered.
378 APPLICATIONS
To apply weighing factors to the costing of pumps, it is assumed that the cost
is closely related to the pump BHP regardless of type. (See Section 11.5.3.) The
cost of a larger pump is based on the generally accepted formula where the scale-
up factor is raised toan exponent of O. 75:
Cost = s0 ·75 X basic pump cost (11.10)
where Sis the scale-up factor.

For example, if our model is an 800-BHP pump that costs $160,000, a 1600-
BHP pump should cost
( 1600/800)°"75 X $160,000 = $269,000

There may be a departure from the constant dollar-per-horsepower with pumps
of different type, but, without supporting data and in the interest of simplicity, this
possibility has not been considered.
11.5.2 Expendable Parts Cost
The estimated expendable parts cost for larger pumps is based on a projection from
presently available catalog prices for similar pump parts. Parts usage is directly
related to pressure, and this is included in the equation and is based on available
data from existing usage. (See Section 11.5.3.)
The parts cost over the life of the pipeline is probably the second most important
factor. The study reveals sorne differences in the long-term parts and power costs
with different types of pumps. These differences must also be considered.
11.5.3 Methods of Cost Projection
Pump cost = (BHP /800)º·75 X $160,000 (11.10)

0.75
gallons
Dampener cost = ( --::¡o- ) X $22,000
x .
station pressure )0.75
( (11.11)
1500
Parts cost = (~)

hfe
X number per pump X number of operating pumps
X ( size )n x base cost x (station1000

pressure)
x PV
base size
(11.12)
11.5 SIZ!NG PUMPS FOR PIPELINES 379
For 20 years and 10% interest
PV ( Present Value) [1 -- (1 + .1)- 2º]/o.1 = 8.5 (11.13)
where
n = exponent dedved from Published Prices.
Parts Cost Factors:
Part Lifeª Base Size Base Cost Exponent ( n)

Pisto ns zb 1 in $ 24 1.0
Liners 2b 7 in 1003 0.85
Valves and seats 1 API#9 310 2.0
Dampener bladders 2 40 gal 2575 0.5
Piston rods 1 3.5 in D 798 1.0
Packing 0.5 3.5 in D 64 1.0
ªLife, in years, selected for comparison uses at l 000 PSI.
bFor doub!e-acting pumps, use half value.
Power cost - /KWH (11.14)
Usage based on mechanical efficiency of pump type:
Single-acting O. 90
Double-acting 0.85
All costs in 1983 dollars.
11.5.4 Standby Pumps

It is logical to proceed with the general assumption that a minimum number of
pumps is desirable, since the fewer the pumps the fewer the parts and the less
maintenance required. Carried to the extreme, this assumption would lead to using
two pumps, one with sufficient power for the station requirement and another of
equal size-an expensive l 00 % of standby power. It would result in a loss of
ability to maintain continuous pipeline operation with the breakdown of an addi-
tional pump.
In the other extreme, a large number of smaller pumps would result in lower
standby cost and would contribute to less upset in flow with the breakdown of one
of the pumps, but would involve a higher parts and maintenance cost. Somewhere
between these extremes should lie an optimum number of pumps per station.
Accordingly, this study is based on the assumption that the station equipment
380 APPLICATIONS
OPTIMIZATION OF PUMPS/STATION
1
1.5
1.4
1(Minirnum at 3 pumps)
1
'I
1
1
tí
o
1 Relative first
u 1.1 1 cost of pumps oniy
<I> 1
>
·.;:;;
"'
-.;;
o::
1.0
1(Minimum at 4 pumps)
1
0.6 1
1
Relative cost of
0.5 expenda ble
parts only
(Present value of
0.4 20 yr requirements
at 103 interest)
0.3
o 2 3 4 5 6
Number ol pumps in a station
(including one spare)
Figure 11.12. Optimization ofpumps per station. (Courtesy Worthington Pump Division-
Dresser Industries)
would consist of a certain number of pumps that would supply the total displace-
ment requirements, with the addition of one pump of the same size for standby.
Figure 11.12 shows the optimum number of total pumps per station as four, on
the basis of the previously mentioned exponent of O. 75 for pump cost alone.
The effect of parts cost tends to shift the optimum cost toward a smaller number
of pumps, in this case three. But due to "windows of nonutilization," size of
pumps available, and the diversity of station requirements, this ideal can seldom
be realized.
11.5.5 Procedure for Pump Selecfüm *
l. List available pumps with specifications: Type, piston range, RPM, BHP,
pump cost, number of parts per pump, and parts cost.
2. List number of stations and requirements: Station number, GPM, pressure,
pumping BHP.
*AH costs in thousands of dollars.

Parts include val ves and seats; liners and rubbers; piston rods and packing.
11.5 SIZING PUMPS FOR PIPELINES 381
For each station and for each pump type:
3. Determine total number of pumps per station, Ns
N = (. station pumping BHP ') * + 1 ( standby) (11.15)

s , pump BHP ,
4. Calculate excess BHP, E percent:
N, x pump BHP - station pumping BHP

E= . X 100 (11.16)
station pumping BHP
5. Calculate station pump cost, CP:
CP = N, X cost of 1 pump (11.17)
6. Cakulate station expendable pump parts cost, CP 1 :
CP 1 = (Ns - l ) X parts cost per pump t (11.18)
For entire pipeline system:
7. Calculate and combine totals and arrange in ascending order of total cost
(see Table 11 Pump, total number of pumps, percent excess, pump cost,
parts cost, total cost, parts per changeout, power cost
11.5.6 Stations
In order to show the fundamental relationship between horsepower and number of
pumps, a hypothetical station of 3290 GPM at 1500 PSI (2880 BHP) is selected.
The pumps listed in Table 11.5, Group l (single-acting) and Group 2 (double-
acting) are specially designed pumps with HHP ratings exactly divisible into the
station HHP, resulting in maximum utilization of the pump's capability.
From the practical standpoint and to show the effects from available and ran-
domly selected pumps, Groups 3 and 4 indude a probable family of geometrically
sized pumps.
*With this fraction increased to next whole number.

tFor any period of time, 20 years in the example.
TABLE 11.5 Costs for all Permutations
Pump Number Pump Dampener Parts Power Piston Diam,

w BHP of Pumpsª Cost Cost Total Cost Total Cost Total in
Oll
N
Group l. Single-Acting Triplex, Divisibleb

3200 2 905 199 1104 154 1258 8892 10150 16.4
1600 3 807 177 985 192 1177 8892 10069 11.6
1067 4 794 174 969 223 1191 8892 10084 9.5
800 5 800 176 976 248 1224 8892 10116 8.2
640 6 812 178 990 271 1261 8892 10154 7.3
Group 2. Double-Acting Triplex, Divisible
3387 2 944 209 1154 297 1450 9415 10866 12.0
1694 3 843 187 1029 400 1430 9415 10845 8.5
1129 4 829 184 1012 482 1494 9415 10909 6.9
847 5 835 185 1020 552 1571 9415 10987 6.0
678 6 848 188 1036 614 1650 9415 11065 5.4
Group 3. Single-Acting Triplex, Randomc
3400 2 947 208 1155 161 1316 8892 10208 16.9
2200 3 1025 225 1250 238 1488 8892 10380 13.0
1750 3 863 190 1053 204 1257 8892 10149 12.1
850 5 837 184 1021 258 1279 8892 10171 8.5
560 7 857 188 1045 300 1346 8892 10238 6.9
Group 4. Double-Acting Duplo:, Random
3400 2 947 210 1157 297 1454 9415 10870 12.0
2200 3 1025 227 1252 463 1715 9415 11131 9.7
1750 3 863 191 1055 408 1462 9415 10878 8.6
850 5 837 185 1023 553 1575 9415 10990 6.0
560 8 980 217 1196 779 1975 9415 11390 4.9
Station: 3290 GPM, 1500 PSI, 2880 HHP
Ali values in thousands of 1983 dollars.
"Number of pumps includes one standby.
b¡}ivi.sible: station. HHP divisible by whole number of p~mps from 2 to 6 (Pumps designed to exact station requirements)
1i .5 SIZING PUMPS FOR PIPELINES 383
11.5. 7 Calculations
With the above data on hand, the calculations are carried out in the order shown
in Section 11.5.5.
In order to dramatize these results, Figure 11.13 is provided. There is only a
slight difference in total pump cost vs. the number of groups, but the diíferences
in parts cost and power cost (incremental) between the single-acting and the dou-
ble-acting pumps are pronounced.
Once the combining mechanics have been completed, selection logic can be
applied using the data at hand. First, let's assume that five pumps per station will
be desirable. As shown in Table 11.5, Group l and 2, the least costly would be
the 800-BHP, special divisible, single-acting triplex.
On the other hand, if available or random pumps are considered, Group 2 and
3, then the 850-BHP single-acting triplex would be the next choice. The absolutely
least-cost random pump would be the Group 3, three-per-station 1750-BHP pump.
There are other subtle factors that could temper t.he choice. A parameter that is
difficult to evaluate is the sanctity of the proven unit, particularly physical size
scale-up. For example, the lowest-cost triplex single-acting pump might require
the use of uncommonly large diameter pistons and a high piston rod load.
C::::J Incremental power cost

~Partscost
~ Dampener cost
c::::::J Pump cost
2.0
1.0
2 3 4 5 6
Number of pumps per station
Figure 11.13. Number of pumps per station. Based on 20-year life of project. Shows the
long-term cost in constant dollars of pump operation, stressing cost of parts. Note differ-
ences in cost for pump type and number of pumps per station. (Courtesy Worthington Pump
Division-Dresser Industries)
384 APPLICATIONS
Another factor difficult to evaluate is the requirement for a wide variety of

different piston sizes (and associated liners and valves) for each pipeline. To obtain
the advantage of a lower inventory of parts and possible cost reduction on a quan-
tity basis, one could choose the option of more pumps with one fixed piston di-
ameter. A redeeming feature of this choice is the fact that the parts cost, even with
more pumps, would remain about the same as with fewer pumps because of the
direct relationship of parts life to actual horsepower. A compromise of, say, only
two different piston diameters could be considered.
11.5.8 Study Results

Perhaps the most startling revelation of this study is the fact that the pump cost
alone for a diverse pipeline system may have little relation to any particular pump
type or size (see Table 11.5).
Long-term operating costs are greatly affected by the type of pump because of
the variation in liquid-end parts requirements and incremental power costs.
lt is impossible to design an individual pump of a size that would provide ab-
solute optimization for all pipelines. In general, the optimum pump size or horse-
power can be directly related to the station horsepower on the basis of four, five,
or six pumps per station (including standby).
Reiterating, maximum optimization could be gained by designing a specially
sized pump to best fit the proposed system. This approach is not necessarily out
of the question for long pipelines because of the long lead time from system con-
ception to completion. The magnitude of sorne projects would certainly warrant
this approach.
11.6 BOLT TIGHTENING SPECIFICATIONS
All bolted assemblies require proper torqueing in order to obtain maximum effi-
ciency and to prevent eventual failure dueto uneven tightening. Table 11.6 gives
the torque specifications for most bolt sizes and materials.
TABLE 11.6 Bolt Torque Specificatlons•
Tensile Stress Area

Size* Area, in2 Size* Area, in2
!(20) .0318 iCIO) .3340
-fu (18) .0524 ~ (9) .4620
iC16) .0775 1 (8) .6060
~(14) .1063 1!(7) .7630
!(13) .1419 1!(7) .9690
i(l l) .2260 Ii(6) 1.1550
*Bolt diameter, in; threads per inch given in parentheses.
TABLE 11.6 (Continued)
Tightening Torquefor ASTM A-307 OR SAE Grade l; low-carbon bolts, heads not
marked
Dry Threads Lubricated Threads

Bolt
Diam. Torque, Force, Arm, Torque, Force, Arm, Bolt Clamp
in lb-ft lb ft lb-ft. lb ft Load, lb
1 1 1
4 2.6 5.2 2 1.95 3.8 2 630
5 1 1
T6 5.4 10.8 2 4.05 8.1 2 1,040
3
8 9.6 9.6 1 7.2 7.2 1 1,535
7
T6 15.3 15.3 1 11.5 11.5 1 2,100
1
2 23.5 23.5 17.6 17.6 2,810
5
8 46.5 31.0 1! 34.9 23.3 1! 4,470
3
4 82.7 41.3 2 62.0 31.0 2 6,620
7
8 133 53 2! 100 40 2! 9,410
1 200 67 3 150 50 3 12,000
lk 282 71 4 211 53 4 15,100
ªBased on tensile stress = 60% yield strength.
T = 0.2DL/12 L = EYA
where T = torque, ft-lb; D = nominal diameter, in; L = clamp load, lb; E = 0.6; Y = yield, PSI;
and A = stress area of bolt, in 2
Tightening Torquefor ASTM B-7 or SAE Grade 7

Alloy steel studs, head marked letter H

Bolt
1 1 1
4 8.3 16.6 2 6.2 12.4 2 2000
5
T6 17.2 17.2 1 12.9 12.9 1 3,300
3
¡¡ 30.5 30.5 23.9 23.9 1 4,880
7
T6 48.8 48.8 36.6 36.6 1 6,700
1
2 74.5 37.2 2 55.8 28 2 8,930
5
8 148 48 3 111 37 3 14,200
i7 262 87 3 196 66 3 21,000
¡¡ 425 106 4 318 79 4 29,100
1 635 159 4 477 119 4 38,200
lk 900 180 5 675 135 5 48,000
1! 1270 254 5 955 191 5 61,000
li 1660 276 6 1245 207 6 72,500
385
386 APPLICATIONS
TABLE 11.6 (Continued)
Tightening Torque for ASTM A-325 OR SAE Grade 5 or A325

Heat-treated bolts; capscrews, head marked with ticks at each of
si.x points and!sometimes ''A325 ''.
QoRe
Bolt
l l l
4 6.7 13.5 2 5.06 10.1 2 1,620
~ 13.9 13.9 1 10.4 10.4 1 2,670
3
8 24.7 24.7 16.5 16.5 3,950
7
ITí 39.4 26.3 1! 29.6 19.7 1! 5,400
l
'i: 60.3 30.2 2 45.2 24.6 2 7,240
5
8 120 60 2 90 45 2 11,500
3
4 212 71 3 159 53 3 17,000
7
8 315 79 4 236 59 4 21,600
1 472 118 4 354 88 4 28,400
lk 633 127 5 475 95 5 33,800
1! 900 180 5 675 135 5 43,000
li 1170 195 6 877 146 6 51,000
12
INSTRUMENTATION
12.1 PRESSURE MEASUREMENTS METHODS

1. To obtain uniformity of data collections, standard transducer locations and
identification (A, B, C, D, E) have been adopted as shown in Figure 12.1, which
also illustrates the use of the terms "upstream" and "downstream."
2. The pressure taps should be on the horizontal centerline of the pipe so as to
eliminate air entrapment if on the top and sediment entrapment if on the bottom,
and to provide a zero datum point.
3. The transducers should be located as close to the pump inlet and outlet as
possible (within 1 or 2 in).
4. All taps should be !-in NPT. In preference to using a welded-in coupling or
other fitting, it is best to drill and tap the pipe, when possible, so as to make a
''close-to-the-liquid'' connection.
5. Whenever possible, a cylinder pressure waveform (location C) should be
taken simultaneously (through a tapped cylinder head) to obtain a "marker" that
will help in the analysis as shown in most of the oscillographs included in this
text.
6. A strain-gauge type of transducer (of 'flush" construction that places the
sensing element or diaphragm directly into the liquid stream) is recommended,
since there would be no cavity to introduce resonance and frequency response
problems. It is recommended that the transducers be left installed for the duration
of the test only because most transducers have a finite life when subjected to pul-
sating pressure.
7. No nipples, valves, or other fittings should be interposed to create resonant
cavities.
387
388 INSTRUMENTATION
Transduc:er locations
Upstream - . - Downstream
Figure 12.1. Pressure-measuring points.
READ IN THIS DIRECTJON---

SCOPE VERTICAL GAIN FROM
WHICH PRESSURE IS CALCULATED GAUGE PRESSURE AT LlMITS MARKED
TRANSDUCER LOCATION CALCULATED FROM TRANSDUCER SPECS.
AND f1EASURED (Z) •
MINIMUM TO
MAXIMUM
PRESSURE
(Af')
20082 HIGH
O, 2 E
2,oos LOW
PRESSURE IN
ONE CYLINDER 0.2 C3 2.r270
AS "MARKER"
44 SUCT ION
ONE
REVOLUT!ON
BETWEEN o.oso 235 RPM ~ CALCULATED FROM SCOPE SWEEF
MARKS- % THEORET!CAL DISCHARGE CYC. : 99.3~IME FOR ONE REVOLUTION (Y)
~T ~
MEASURE (Y)
SCOPE SWEEP 80016-S RECIP. PUMP PP-49, READINGS PUMP 1

RATIO OF DISCHARGE
TEST NUMBER PUMP DESCR l PTI ON SPECIAL TATEMENTS LENGTH ()() TO
LENGTH OF STROKE (Y)
(RELATED TO V. E. l
Figure 12.2. How toread oscillographs.
8. In making a survey it is wise to always use the same transducer before and
after any changes are made to the system.
9. The use of an oscilloscope for pressure waveform readout is considered the
rnost reliable method, since it removes the problems of frequency response, iner-
tia, resonance, and other mechanical doubts associated with recording chart de-
vices.
Figure 12.2 shows the method of interpreting a typical oscillograph.

12.3 TYPICAL WAVEFORMS 389
Calculation of frequency
Pump No. 4 9-8-79 20 cycles/125 mm
Natural frequency of 11:28 AM Chart speed - 100 mm/s
suction pipe Pump shut down 20/(125/100) = 16 Hz
Amplitude - 50 mV /div
Figure 12.3. Oscillograph-pipe vibration.
12.2. VIBRATION MEASUREMENT
Complete instrumentation of a pump for the purpose of analyzing problems should

include (in addition to pressure transducers) vibration transducers (accelerometers)
placed on the piping system in order to determine the degree and frequency of
vibration of the particular section of piping that may be subject to excessive vi-
bration. The frequency can be related to the pump-generated pulse or rotational
frequency to determine if a critica! resonance is present.
In all of these measurements it is assumed that the recording device (oscillo-
scope or oscillograph) has an accurate time-base trace.
One can determine the natural frequency of a section of pipe by striking the
pipe with a wooden ball bat, for example, while the accelerometer is mounted in
the proper location and with the pump stopped but with the piping filled with
liquid. (See Figure 12.3.)
Figure 12.4 offers a good example of how the vibration of pipe spans can be
measured and related to the pump rotation by the second simultaneous trace of the
pump cylinder pressure. lt will be seen that the pipe vibration is related to the
second harmonic of the pump pulsations.
Figure 12.5 is handy to convert acceleration to velocity or displacement.
12.3 TVPICAL WAVEFORMS
The series of 35 oscillographs shown in Figures 12.6 and 12.7, obtained during
NPSH tests, show the typical shapes generated under a myriad of conditions. (An
apology for the poor quality of this series of oscillographs is in order. They are
part of a series of hundreds of oscillographs taken during an extensive test that
would be most difficult and expensive to rerun and record. They all show the
important shape of the suction waveform.) Note the variation of the suction pres-
sure waveform shapes, particularly the "rounded-bottom, sharp upward spikes"
at 360 RPM (and others) indicating cavitation.
390 INSTRUMENTATION
Discharge pipe vibration
=Ft=J~~r=icF'cF'c
~
Cylinder pressure / 100 mm 1
~ st~oke 1 One
Figure 12.4. Pipe vibration vs. pump pulsations. Chart data: Chart speed, 100 mm/s;
upper trace, discharge pipe vibration; lower trace, center cylinder pressure, PSI. This cha1t
is an actual trace taken on a 7 X 10 in triplex single-acting pump pumpíng water at 1000
PSI. The lower trace is the cylinder pressure obtained from a transducer located in the center
cylinder head. The upper trace is the output of an accelerometer attached to the discharge
pipe at the point of maximum vibration. From this chart it was determined that the basic
speed of the pump was 133 RPM, generating the following numerical and pressure fre-
quencies: Basic pump speed, 133 RPM (2.2 Hz). Discharge pipe víbration frequency, 25
Hz. Pulse frequency of triplex single-acting pump, 13.3 Hz (Almost second harmonic of
pump pulsation frequency. Slight difference provides phasing as shown in the periodic na-
ture of the pipe vibration frequency.)
The test involved a 3 X 6 in triplex single-acting pump with a valve spring

POSIVA of 4 anda discharge pressure of 1400 PSI. The controlled suction pres-
sure was measured by a damped Bourdon tube pressure gauge, and the suction
pressure was controlled by a throttling valve.
In order for these oscillographs to be presented in a form that will allow direct
comparison, they have been rather congested, and sorne further explanation may
be needed. The graphs are arranged vertically by decreasing suction pressure and
horizontally by increasing speed as shown at the top of each column. Each graph
is labeled with its associated volumetric efficiency. Figure 12.6 is for a 90-ft long
suction pipe and Figure 12.7 is for a 10-ft suction.
12.4 MISCEllANEOUS WAVEFORMS (OSCILLOGRAPHS)
The typical self-described oscillographs included here as Figures 12. 8-12 .17 show
the variations of pressure waveform shapes generated by different operating con-
ditions. It should be noted that in many of these waveforms the suction and dis-
12.4 MISCELLANEOUS WAVEFORMS (OSCILLOGRAPHS) 391
Frequency, Hz
Figure 12.5. Vibration nomograph. Arrows indicate direction of lines.
charge traces are aligned with the cylinder trace, which is usually shown. In other
words, any disturbance in the suction or discharge can be related to a particular
position of the crank rotation by reference to the cylinder trace.
For example, in Figure 12.8 the marked disturbances (pressure spikes) in the
suction can be related either to a flow or acceleration peak as shown. Without a
suction stabilizer and with a "good" suction head (upper left), an acceleration
spike is identified as such because it occurs immediately at the start of a suction
stroke in the cylinder trace immediately above. To show how the disturbances can
change when the suction head is reduced to 2 ft (lower left), the acceleration spike
has disappeared and a flow variation spike has replaced it as a major disturbance
occurring at a point of maximum flow.
100 RPM 200 RPM 300 RPM 360 RPM
1 1 1 1 98% VE , 1
91.2%V - - ·-"--. 97% VE, ~: 96% VE
1 ,. ! . i
1 1 1 1
;
, , .
1 1
r u ~ ~ .. '
'·i 1 1 '
;1. F ·-t--=:d : •
\: _..,. n: 1 1 '
¡ 1 ; 1 1 1 1 ¡j,. . i. ¡ ! /.i 1 ,, ; ' ' '! . . . ¡ i ;
1 l , L.1. 1 1 1 ¡ 1>. • .f1tv-I; ºiJ i.J7!'H'l lllt. . ·-,- i ··~-~ ' . ,_j ¡ lJ.,,J
YI I~ :S7I ~.. !J f 1 t·ttff.~ ¡·¡ f"tl_Jfi- _f¡ iJi·tn
1 ¡ . •
: ~·97.5% VE1-,1~~;:::t.' 97.5% vE. :·-~ 97.5% vr..· ' ; 93.BXV,
! I' ¡¡ 1 , 1 frr 1 ! : • .. I· ~ ! ' ! I~ . ¡., . , 1 ,; 1 1 ;
: 1 ; 1 ·t 1 . 1 1 !- : 1 ¡ 1 1 ! : 1" ¡, 1: 1 1 1 . • .; : ¡ 1 .
i 1 ;. 1 i ' 1 1 1 1 1 1 1 ! ¡ 1 1 1 h 1'"' l. 1 . i i 1 ~
il..1 1ll: i 1~ 1- 1 ,1 ni: 11.~ ~ .1f., 1 1 11 1 1 i . ! 1 ..1 1111- 1: l:. ¡ l L..:..J
·O PSI
---1. . 1. 1.....1_....
1· ....·--·i.....
· 1J_1. WlJ trrr ~r-r.1 rJti
99.6% : ~...:....'
v __ 98.1% VE _·_:"'......!. 97 '. 4% vt ~~' 80% VE
,~.. .r. "T
~-' 1: • • ((..¡_ ·-'- · .-.
-r--·
'
- -
·1- - - -~
. ! . - .___j_ =-·
\
----==11
,_._ -· - . -~ - - .; -
l==~,·=·=-;~:~·-:......,.._r:::::.::=: l=d!~-.t.J.L·-~i....:.._,_...:...::=I==-;_.;__.· _;..._~:;;_¡:~;"-·-,___/\,.___./._;_
;-1: ,,,.._1¡__,
- :-·t 1 -~-.:...·~t--""11'! 1 i 1 ¡-,- ~¡ ¡ 1 ~ - -

-5 PSI N
- - -L9f.5X VE -·- - :.. 95% VE;-:,.--;-· ;
--
_ _c1
·:
T ~-
.--~~--
--- ·- '.- ......:.
· - L--··
COMPLETE
CAVITATION
Figure 12.6 Long Suction Pipe. Oscillographs at various RPMs and suction pressures.
Typical cylinder and inlet (suction) pressure waveforms for 3 x 6 in triplex single-acting
plunger pump. 4 POSIVA valve spring load. Vertical scale (suction): JO PSI per division.
Suction system: 90 ft of 4-in pipe. RPM, suction pressure, and volumetric efficiency
noted.
392
100 RPM 200 RPM 300 RPM
SPACE
COMPLETE
CAVITATION
N - NOISE
1 1
TOP TRACE -
CYLINDER
PRESSURE
- - 7 • 5 PSI· :f 11 ¡ tf)l
N : ·: /
Figure 12. 7 Short Suction Pipe. Oscillographs at various RPMs and suction pressures.
Typical cylinder and inlet (suction) pressure waveforms for 3 X 6 in triplex single-acting
plunger pump. 4 POSIVA valve spring load. Vertical scale: IO PSI per division. Suction
system: 8 ft of 4-in pipe. RPM, suction pressure, and vo!umetric efficiency noted.
393
Figure 12.8. Oscillographs of pressure waveforms. Left: No suctioh stabilization. Right:
with 30-gal suction stabilizer. Top: Suction head, 14 ft. Bottom: Suction head, 2 ft. This
series of oscillographs show how maximum disturbance (arrow A) is caused by "acceler-
ation" at start of suction stroke as marked by cy linder trace above that trace and at "valley"
in the theoretical fiow pattem plotted below each card. After lowering the suction head to
2 ft, lower left, the arrow B shows the predorninant spike is now a "flow" -induced dis-
turbance because it occurs at a peak in the plotted ftow. Upper-right card (arrow C) shows
how the "acceleration" spike has been drastically reduced by the use of a suction stabilizer.
The "flow" spike at lower right (arrow D) showed that the stabilizer reduces the "flow"
disturbances somewhat also.
394
(a)
(b)
Figure 12.9. Oscillographs of pressure waveforms. (a) Unrealistic "butterfty" type of

trace produced by natural frequency of vibration or cavity resonance of the particular trans-
ducer, havihg a small cavity between the connection and the sensing element. lt can be
recognized by the symmetrical shape above and below the average and the fuzziness pro-
duced by the extremely high frequency (in this case about 3000 Hz). (b) Actual suction
pressure waveform in the same pump, operating under the same conditons, but obtained
with another (nonresonant) transducer with the sensing element flush with the liquid.
395
(a)
(b)
Figure 12.10. Oscillographs of pressure wavefonns. Suction pressure trace of 5 X 8-in

quintuplex plunger pump. (a) Pump not equipped with suction stabilizer. Note typical
"cavitation" wavefonn with sharp upward spikes and rounded bottoms. Also note extreme
pressure excursion in an upward or positive direction while the bottom of the trace is pre-
vented from extending into the negative region by the fonnation of vapor. 63 PSI peak-to-
peak. (b) Same pump with suction stabilizer, same operating conditions. Note the "clean"
sine wave and high frequency typical of "good" suction. The low-frequency cycles, over
which the high frequency is imposed, is the remnant of pump rotation-generated cycles.
While this flow-induced pulsation is not serious, it could have been reduced by the use of
a properly gas-charged bladder in the stabilizer. 13 PSI peak-to-peak.
396
AT PLW INLET
WITl-OUT "BEFORE"
STABlllZER
(GA!N ·• 57 PSI/IN)
UPSTREN1 OF
STABIL!ZER
11AFTER"
AT PLW INLET
WITH
STABIL!ZER
(GAJN - 126 PSI/IN) 1
Figure 12.11. Oscillographs of pressure waveforms. "Before" and "after" test results
showing the effects of a suction stabilizer that is almost ali flow and acceleration
disturbances in the suction system of a reciprocating pump. This particular test was made
with a 60-gallon stabilizer of the type shown in Fig. 4.9, Chapter 4 on a 6 X 8-in triplex
pump at 3000 PSI discharge and 30 PSI suction.
3" :X 6" 28% PULSATION

TRIPLEX
800 PSI
NOTE SIX FLOW DISTURBAN~ES PER REVOLUTION
ONE PUMP
REVOLUTION
3n X 5t~ 10% PULSATION

QUINTUPLEX
NOTE 15 FLOW DISTURBANCES PER REVOLIJTION
Figure 12.12. Oscillographs of pressure waveforms. Nondampened discharge pressure

waveforms. Not shown is the duplex double-acting pump with four flow-disturbances per
revolution.
397
PSI
DISCHARGE
---,,..,..,..?-psr 2828 PSI
PRESSURE,PSI 264 PSI
PSI
CYLINDER 2828 PSI
PRESSURE,PSI
SUCTION
PRESSURE, PSI.
(a)
t.P = 38 PSI)
94 PSI ~
SUCTION 80 PSI
PRESSURE AT
INLET, PSI 66 PSI
(b)
Figure 12.13. Oscillographs of pressure waveforms, showing the presence of high-fre-

quency pressure waves in the suction, cylinder, and discharge of a quintuplex plunger pump.
(a) Cylinder and discharge pressure waveforms. Only remnants of flow-induced, low-fre-
quency pulsations appear in the discharge. Major pressure excursions in both suction and
discharge are high-frequency, acceleration-induced. ( b) Cylinder and suction pressure
waveforms. He re the predominant pulsations are of the high-frequency, acceleration-in-
duced type.
PUMP DATA: FREQUENCIES PRESENT:
4! x 9-in quintuplex Rotational, 3.1 Hz
184 RPM Cylinder, 15.3 Hz
2800 PSI discharge Flow pulse, 31 Hz
80 PSI suction Acceleration, discharge, 138 Hz
Acceleration, suction, 77 Hz
398
!~
_, r- .6"
Figure 12.14. Oscillographs of pressure wavefonns. Typical pressure wavefonn of high-

vapor-pressure liquid (ethylene). Pressure wavefonn of cylinder and suction pressure. Of
extreme interest is the shape of the cylinder wavefonn. The effect of vaporization of the
liquid in the cylinder due to heat transmitted through uninsulated cylinder drain fittings,
being manifested by the low volumetric efficiency indicated by the length of the delivery
portion of the wavefonn being considerably shorter than the suction portion. Note the pla-
teau at the critica! pressure of 700 PSI, which prevails until the cylinder pressure builds up
to the discharge pressure of 1000 PSI.
This demonstrates how a pressure wavefonn can be used to estimate the volumetric
efficiency of a pump. Determine the full-stroke length by scaling from one obvious point
on the wavefonn to another, in this case, 2 in. Accordingly, a full delivery stroke would
theoretically be 1 in, but the actual delivery is measured as 0.6 in. Therefore, the volumetric
efficiency is about 60 percent.
399
DISCHARGE PULSATION
PRESSURE PERCENT
llP
114 PSI 3.8
68 PSI 2.3 %
46 PSI 1.5 %
Figure 12.15. Oscillographs of pressure waveforms. Oscillographs taken at intervals to

show the "phasing" or in-and-out of step of multiple pump pulsations, whereby the pulses
tend to "add" for a short period of time and then tend to "subtract" for another period of
time. It is impossible to prevent pumps from such phasing. Regardless of the pump drive
ratios, there will be repeated periods of phasing.
400
294 PSI
94 PSI or 47% Overshoot

,----0
Rise time; .0052 sec,
.109 sec/rev
(a)
PSI
PSI
106 PSI or 4,5% Overshoot
_ _ _ _ __._..3~4...,3 sec/rev
(b)
Figure 12.16. Oscillographs ofpressure waveforms. (a) Triplex single-acting pump, .708
X .945 in, 550 RPM. (b) Triplex single acting pump, 6 X 8 in, 175 RPM. These show
actual typical cylinder pressure "overshoot" amplitude and rise time of two triplex pumps
differing greatly in size and speed. The overshoot is related to time rise (pump speed) but
can also be present at slow speed with low NPSHA.
401
402 INSTRUMENTATION
1/2
t
STROKE
•I!!!~ -~~
ONE STROKEJ
:1
~~il6~R~ALF
92% VE
STROKE
1111••1•••••1111 -·---
CYLINDER
:1111•·-····
.llJl'•••r•••llli---22_so_p_sr_A_vE_.
(a)
DELIVERY 81% OF r l
HALF STROKE_. . .
ONE STROKE
~~ ••
HALF STROKE
1111.iM.• • ·--·--CY-LI-ND_ER_
-~··•L••··~ 1000 PSI AVE.

~-·•llliiiiiiilll--.- -
(b)
Figure 12.17. Oscillographs of pressure waveforms. (a) Typical cylinder pressure wave-
form of liquid containing little dissolved gas. Note approximate square wave with abrupt
rise in pressure at beginning of delivery stroke. Single-acting quintuplex pump pumping
water (boronated with hydrogen blanket), 105 ºF, 226 RPM. ( b) Typical cylinder pressure
containing a large amount of "free" gas. Note compression type of pressure rise at begin-
ning of delivery stroke. Free gas in cylinder must be compressed to. discharge pressure
befare liquid can be delivered. Note low VE caused by dissolved gas. 2ft X 10-in double-
acting duplex pumping crude oil with gas. Ambient temperature; 82 RPM. Suction-charged
with centrifuga! pump, 30 PSI.
12.5 OPTICAL PHASER
A useful device to be used in conjunction with oscillograph displays of waveforms

generated by a reciprocating pump is a simple light-sensitive pickup or phaser
positioned to view a white sticker placed on any visible reciprocating part of the
pump. The end-of-stroke blip can be related to any secondary trace recorded on
a dual-trace oscilloscope as shown in Figure 12. IA.
12.6 POSITIONING STRAIN GAGES TO MONITOR TORSIONAL LOADS 403
: One pump revolution

1 1 1
....... ¡. P~~·~· ;t~~k!e-: -s- CC·E-) ...

Pump
1
1
1
Harker blip, end-o'-st;oke 1
1- 1 1 ~ l . ¡ ¡ 1; : 1 1 • •• 1 1 1 1 ir 1 1 1 1
· De ay, suct1on ' Delay, discharge
1 valvs closing -1,..._,,.valve closing
kt- 1
;
Delay,
1
co~pressior. __
~
JI Dela y ,1 •
decompress1on
1 ..... ...L . . . . '••.•
l 1
1
1
1
1
·r---·- -.. ·-¡- --. -

¡
1 1
¡ 1
1
1
·¡· - • - ......... I
1
1
-
~
,. _. r -
1 1
1
1
¡. :
_J _________,l ...
.:......
l ~
1
Figure 12.IA. Use of Optical Phaser to Mark End of Mechanical Stroke.
12.6 POSITIONING STRAIN GAGES TO MONITOR

TORSIONAL LOADS
To measure minute strains one must be capable of measuring minute resistance

changes. The Wheatstone Bridge configuration as shown in Figure 12.2A is ca-
pable of measuring these small resistance changes. Note the signs associated with
each gage numbered 1 through 4. The total strain is always the algebraic sum of
the four strains.
The total strain is represented by a change in "V" out. If each gage had the
same positive strain, the total would be zero and "V" out would remain un-
changed. Bending, Axial, and Shear strain make up the most common types of
404 INSTRUMENTATION
V In
Reguloted
DC
V out +
Figure 12.2A. Wheatstone bridge
strain measurements. The actual arrangement of the strain gages will determine
the type of strain measured and the output voltage change.
TORSIONAL STRAIN (y) equals torsional stress (T) divided by torsional
modulus of elasticity (G). See Figure i 2.3A and Equations (12.1) and (12.3).
Where torsional stress (r) equals the torque (M,) multiplied by the distance
from the center of section to outer fiber (d/2) divided by (J) the polar moment
of inertia. The polar moment of inertia is a function of the cross sectional area.
For solid circular shafts only see Equation (12.7).
The modulus of shear strain (G) has been defined in the preceding discussion
on shear stress. Strain gages can be used to determine torsional moment (M,),
from which can be calculated in Equation (l 2.6).
T
"'=2Xe = -
' ' G
Where e, = e 1 - E~ + e, - E4 ( 12.2)
1' = MT X (~)/J (12.3)
MT = r(J) (l>) = '}'G(J) (~) yG( 1TD 3 / ! 6) (12.4)
L
<P = MT G(J) (12.5)
BHP = 2nNMT/33,000 (12.6)
J = 7T(d) 4 /64 ( 12.7)

12.8 MEASURING PRESSURE DROP BY OSCILLOSCOPE 405
180 DEGREE ROSETTE STRA!N GAGES
11 y
.---~4--~li\-
_-. ~Jb-~=-~)=-~--- z
4 3
( 7 1
2
L \~
MT
L
y
Figure 12.3A. Torsional strain
y-Torsional strain, uinch/inch

E-Normal strain, uinch/inch
r-Torsiona! Stress, lbf/inch 2
G-Modulus of Elasticity, Shear, lb/inch 2
M-Moment, Torque, lb.ft
D-Diameter shaft, inches
J-Polar Moment of Inertia, inches 4
<f>-Angle, radians
L-Length of shaft 1.mder stress, inches
BHP-Brake Horsepower
N-RPM
12.7 DAMPED PRESSURE GAUGE
The series of oscillographs in Figure l2.4A serve to show that a properly

"damped" pressure gauge of the bourdon tube type wil! show the approximate
"average" pressure of a complex pressure wave. This aspect is important in the
measurement of suction pressures with the presence of standing waves generated
by acceleration, where the "average" only of that pressure, is indicated.
12.8 MEASURING PRESSURE DROP BY OSCILLOSCOPE
The ability of sorne dual-trace oscilloscopes to invert one of the pressure traces
allows one to display the pressure drop of two complex pressure waves as shown
in Figure 12º5Aº
------.------ -------- -----·-;
STATIC OR: INLE'T GAUGE :
MAXIMUM M!:Al< M'.NIMIJM
_JER!J....... . PS_IG ,
- - - - - - ; - - - - - - - - - - - - : - O. 7 PS ¡-;.- O. 75 PS IG ~
L,__ _ _ __;__ _ _ _ _ _ _
-'--------'--------
l SUCTI!JN SCOPE - 5 PSI P':':F J:.:s:mi
0.7 PSi-:- O PS!G
- - - - - - ·-------.:..- - - - - - - - - - · - -
2 SUCTION
200 PSIG
O PSI
3 DISCHARGE SCOPE - 16~ "'S: PER ::~ISIDN

'
1
- 5.6 PSI
'3.0 "'SIG
1
-·--------'-· - - - - - - - ·-----~ - - -----;!

4 SUCTHJN SCCPE • .4 PSi PER ClVISIDN 1
1
7.7 PSI SEE 7
BELD~' 3.0 PSIG

O PSI 1
--'--- ---'--------'
5 :SUCTIIJN SCDPE - 1.4 PSI PER DIV!S!DN
!
1
1
1
1
2.0 PSI 2..0 PSIG i
~~---=J
6 :SUCTIDN SCDPE l .4 PSI PER DIV!SIDN
!NTEGRAT _r_N~-. A_MP~l F!ER ... ¡
3. 4 P S I - - - - - - - - - - - - - 3.0 PSIG
'7 SUCTHJN SCDPE - 1.4 PSI PER DIV!SION
Figure l2.4A. Accuracy of Damped Pressure Gauge Readings
406
----------....----
B
.....
e
LDV 6
4
2 1------'····A ....+.... 8 ............~~~~1
o
DUAL
TRACE
SCOPE ···················¡:,-Rfs.SúRE .. riR.i:if=>·················
3 ADD A + B ! A - Bl
·························~--················· .. ········· 8
......................... A..................................... . 6 ._._._._._.__ ._._._._·_·_·_-_._._._._._._._._._._E:::::::::.: ..:"""""""-"""".·"""""""""".
4
t------C.
...... -e ......... 7"-----t
... 2
o
-2 t-----<:.:······B-·······,_,..
.. - - - - - ·
-4
-6
RECORD A AND B AS POSITIVE -8 2 INVERT B
PRESSURE
Figure 12.SA. To Measure Pressure-Drop of Complex Waveforms
407
1
THEORY OF FLOW IN
PIPE
AH of the material in Chapter 13 is from Flow of Fluids, 1985, by courtesy of

Crane Company. Because the nomenclature in this section may not agree with that
used by the author in the rest of the text, a nomenclature list for this chapter is
included as Table 13.10.
13.1 INTRODUCTION
Many so-called pump problems are not associated with the pump itself but with
the system connected to it from both the supply and the delivery. Accordingly, it
behooves the system designer to employ the best engineering practices in the pip-
ing system. This chapter provides up-to-date information on the theory of fiow in
pipe.
The most commonly employed method of transporting fluid from one point to
another is to force the fluid to fl.ow through a piping system. Pipe of circular section
is most frequently used because that shape offers not only greater structural strength
but also greater cross-sectional area per unit of wall surface than any other shape.
Unless otherwise stated, the word pipe in this book will always refer to a closed
conduit of circular cross section and constant intemal diameter.
Only a few special problems in fluid mechanics-laminar fiow in pipe, for ex-
ample-can be entirely solved by rational mathematical means; all other problems
require methods of solution which rest, at least in part, on experimentally deter-
rnined coefficients. Many empirical formulas have been proposed for the problem
of flow in pipe, buúhese are often extremely limited and can be applied only when
408
13.2 PHYSICAL PROPERTIES OF FLUIDS 409
the conditions of the problem closely approach the conditions of the experiments
from which the formulas were derived.
Because of the great variety of fluids being handled in modem industrial pro-
cesses, a single equation which can be used for the flow of any fluid in pipe offers
obvious advantages. Such an equation is the Darcy* formula. The Darcy formula
can be derived rationally by means of dimensional analysis; however, one variable
in the formula, the friction factor, must be determined experimentally. This for-
mula has a wide application in the field of fluid mechanics and is used extensively
throughout this chapter.
13.2 PHYSICAL PROPERTIES OF FLUIDS
The solution of any flow problem requires a knowledge of the physical properties
of the fluid being handled. Accurate values for the properties affecting the flow of
fluids ... namely, viscosity and weight density ... have been established by many
authorities for all commonly used fluids and many of these data are presented in
the various tables and charts.
Viscosity
Viscosity expresses the readiness with which a fluid flows when it is acted upon
by an extemal force. The coefficient of absolute viscosity or, simply, the absolute
viscosity of a fluid, is a measure of its resistance to intemal deformation or shear.
Molasses is a highly viscous fluid; water is comparatively rimch less viscous; and
the viscosity of gases is quite small compared to that of water.
Although most fluids are predictable in their viscosity, in sorne, the viscosity
depends upon the previous working of the fluid. Printer's ink, wood pulp slurries,
and catsup are examples of fluids possessing such thixotropic properties of viscos-
ity.
Considerable confusion exists conceming the units used to express viscosity;
therefore, proper units must be employed whenever substituting values of viscosity
into formulas. In the C.G.S. (centimeter, gram, second) or metric system, the unit
of absolute viscosity is the poise which is equal to 100 centipoise. The poise has
the dimensions of dyne seconds per square centimeter or of grams per centimeter
second. It is believed that less confusion conceming units will prevail if the cen-
tipoise is used exclusively as the unit of viscosity. For this reason, and since most
handbooks and tables follow the same procedure, all viscosity data in this paper
are expressed in centipoise.
The English units commonly employed are "slugs per foot second" or "pound
force seconds per square foot"; however, "pound mass per foot second" or
*The Darcy formula is also known as the Weisbach formula or the Darcy-Weisbach formula. lt is also
known as the Fanning formula, sometimes modified so that the friction factor is one-fourth the Darcy
friction factor.
410 THEORY OF FLOW IN PIPE
"poundal seconds per square foot" may also be encountered. The viscosity o
water ata temperature of 68 F is:
0.01 poise
ll = 1 centipoise* = [ 0.01 gram per cm second
0.01 dyne second per sq cm
_ [0.000 672 pound mass per foot second
lle - 0.000 672 poundal second per square foot
, [0.000 0209 slug per foot second
lle = 0.000 0209 pound force second per square ft
Kinematic viscosity is the ratio of the absolute viscosity to the mass density. In.
the metric system, the unit of kinematic viscosity is the stoke. The stoke has diÍ
mensions of square centimeters per second and is equivalent to 100 centistokes.
. ll ( centipoise) ll
v ( centistokes) = p , (grams per cub"1c cm ) = -8
By definition, the specific gravity, S, in the foregoing formula is based upon watef
at a temperature of 4 C (38.2 F), whereas specific gravity used throughout thi~
paper is based upon water at 60 F. In the English system, kinematic viscosity Ífa1$
dimensions of square feet per second. 1
.,,
Factors for conversion between metric and English system units of absolute al\'~
kinematic viscosity are given in Tables 13.1 and 13.2.
The measurement of the absolute viscosity of fluids (especially gases and va~'
pors) requires elaborate equipment and considerable experimental skill. On the
other hand, a rather simple instrument can be used for measuring the kinemati~*
viscosity of oils and other viscous liquids. The instrument adopt~d as a standW-et;
in this country is the Saybolt Universal Viscosimeter. In measuring kinematic vis~'
cosity with this instrument, the time required for a small volume of liquid to flow.
through an orifice is determined; consequently, the "Saybolt viscosity" of tli~i
liquid is given in seconds. For very viscous liquids, the Saybolt Furol instrumenr
is used. ' .
Other viscosimeters, somewhat similar to the Saybolt but not used to any extent'
in this country, are the Engler, the Redwood Admiralty, and the Redwood. Thei
relationship between Saybolt viscosity and kinematic viscosity is shown in Tables:··.·
13.3 and 13.4; equivalents of kinematic, Saybolt Universal, Saybolt Furol, árid'
absolute viscosity can be obtained from the chart of Figure 13.8. ·
The viscosities of sorne of the most common fluids are given in Figures 13.9 ..
lt will be noted that, with a rise in temperature, the viscosity of liquids decreases;·
whereas the viscosity of gases increases. The effect of pressure on the viscosity of
*Actually the viscosity of water at 68 F is 1.005 centipoise.

13.2 PHYSICAL PROPERTIES OF FLUIDS 411
liquids and perfect gases is so small that it is of no practica! interest in most ftow
problems.
Weight Density, Specific Volume, and Specific Gravity

The weight density or specific weight of a substance is its weight per unit volume.
In the English system of units, this is expressed in pounds per cubic foot and the
symbol designation used in this paper is p (Rho). In the metric system, the unit is
grams per cubic centimeter and the symbol designation used is p' (Rho prime).
The specific volume V, being the reciproca! of the weight density, is expressed
in the English system as the number of cubic feet of space occupied by one pound
of the substance, thus:
1
V=-
p
Computations in the metric system are not commonly referred to in terms of

specific volume; however, the number of cubic centimeters per gram of a substance
can readily be expressed as the reciproca! of the weight density, that is:
1
p'
The variations in weight density as well as other properties of water with changes
in temperature are shown in Table 13.5. The weight densities of other common
liquids are shown in Table 13.6. Unless very high pressures are being considered,
the effect of pressure on the weight of liquids is of no practica! importance in ftow
problems.
Specific gravity is a relative measure of weight density. Since pressure has an
insignificant effect upon the weight density of liquids, temperature is the only con-
dition that must be considered in designating the basis for specific gravity. The
specific gravity of a liquid is its weight density at 60 F (unless otherwise specified)
to that of water at standard temperature, 60 F.
S = =-P_{,._an---=-y_I_iq=-u_id_at_6_0_F_,_unl_e_ss_o_t_he_rw_is_e_s....:.p_e_ci_fi_ed__._}
p (water at 60 F)
A hydrometer can be used to measure the specific gravity of liquids directly. Three
hydrometer scales are common in this country-the API scale which is used for
oils-and the two Baumé scales, one for liquids heavier than water. and one for
liquids lighter than water. The relationship between the hydrometer scales and
specific gravity are:
For oils,
S
( 60 / 6
F F
)
° = 131.5 + deg.
141.5
API
For liquids lighter than water,
140
°
S( 60 F/ 6 F) = 130 + deg: Baumé
For liquids heavier than water,
145
°
S ( 6 F / 60 F) = 145 - deg. Baumé
For convenience in converting hydrometer readings to more useful units, refer to

Table 13.7.
13.3 NATURE OF FLOW IN PIPE-LAMINAR ANO

TURBULENT
A simple experiment [illustrated in Figs. 13.1 to 13.3] will readily show there are
two entirely different types of flow in pipe. The experiment consists of injecting
small streams of a colored fluid into a liquid flowing in a glass pipe and observing
the behavior of these colored streams at different sections downstream from their
points of injection.
If the discharge or average velocity is small, the streaks of colored fluid flow
in straight lines, as shown in Figure 13.1. As the flow rate is gradually increased,
these streaks will continue to flow in straight lines until a velocity is reached when
the streaks will waver and suddenly break into diffused pattems, as shown in Fig-
ure 13.2. The velocity at which this occurs is called the "critica! velocity." At
velocities higher than ''critica!,' ' the filaments are dispersed at random throughout
the main body of the fluid, as shown in Figure 13.3.
The type of flow which exists at velocities lower than "critica!" is known as
Figure 13.1. Laminar flow. Actual photograph

of colored filaments being carried along undis-
turbed by a stream of water.
Figure 13.2. Flow in critica! zone between lam-

inar and transition zones. At the critical velocity,
the filaments begin to break up, indicating flow
is becoming turbulent.
13.3 NATURE OF FLOW IN PIPE-LAMINAR AND TURBULENT 413
Figure 13 .3. Turbulent ftow. This illustration

shows the turbulence in the stream completely
dispersing the colored filaments a short distance
dowhstream from the point of injection.
laminar fl.ow and, sometimes, as viscous or streamline flow. Flow of this nature
is characterized by the gliding of concentric cylindrical layers past one another in
orderly fashion. Velocity of the fluid is at its maximum at the pipe axis and de-
creases sharply to zero at the wall.
At velocities greater than "critica!," the flow is turbulent. In turbulent fl.ow,
there is an irregular random motion of fluid particles in directions transverse to the
direction of the main flow. The velocity distribution in turbulent flow is more
uniform across the pipe diameter than in laminar fl.ow. Even though a turbulent
motion exists throughout the greater portion of the pipe diameter, there is always
a thin layer of fluid at the pipe wall, known as the "boundary layer" or "laminar
sub-layer," which is moving in laminar flow.
Mean Velocity of Flow

The term "velocity," unless otherwise stated, refers to the mean, or average,
velocity at a given cross section, as determined by the continuity equation for
steady state flow:
q w wV
v=-=-=- ( 13.1)
A Ap A
Reynolds Number
The work of Osbome Reynolds has shown that the nature of flow in pipe-that is,
whether it is laminar or turbulent-depends on the pipe diameter, the density and
viscosity of the fl.owing fluid, and the velocity of fl.ow. The numerical value of a
dimensionless combination of these four variables, lrnown as the Reynolds num-
ber, may be considered to be the ratio of the dynamic forces of mass fl.ow to the
shear stress due to viscosity. Reynolds number is:
R
e
= Dvp (13.2)
For engineering purposes, flow in pipes is usually considered to be laminar if

the Reynolds number is less than 2000, and turbulent if the Reynolds number is
greater than 4000. Between these two values lies the "critica! zone" where the
fl.ow-being laminar, turbulent, or in the process of change, depending upon many
possible varying conditions-is unpredictable. Careful experimentation has shown

that the laminar zone may be made to termínate at a Reynolds number as low as
1200 or extenqed as high as 40,000, but these conditions are not expected to be
realized in ordinary practice.
Hydraulic Radius
Occasionally a conduit of noncircular cross section is encountered. In calculating
the Reynolds number for this condition, the equivalen!: diameter (four times the
hydraulic radius) is substituted for the circular diameter. Use friction fractors given
in Figures 13.11and13.12.
cross sectional flow area
R = -~~~~~~~-
8 wetted perimeter
This applies to any ordinary conduit (circular conduit not flowing full, oval, square
or rectangular) but not to extremely narrow shapes such as annular or elongated
openings, where width is small relative to length. In such cases, the hydraulic
radius is approximately equal to one-half the width of the passage.
To determine quantity of ftow in following formula:
q = 00438d
.
2ffLD-fL
The value of d 2 is based upon an equivalent diameter of actual flow area and.4R8
is substituted for D.
13.4 GENERAL ENERGY EQUATION-BERNOULLl'S

THEOREM
The Bemoulli theorem is a means of expressing the application of the law of con-
servation of energy to the flow of fluids in a conduit. The total energy at any
particular point, above sorne arbitrary horizontal datum plane, is equal to the sum
of the elevation head, the p~ssure head, and the velocity head, as follows:
144P v2
Z+--+-=H
p 2g
If friction losses are neglected and no energy is added to, or taken from, a piping
system (i.e., pumps or turbines), the total head, H, in the above equation will be
a constant for any point in the fluid. However, in actual practice, losses or energy
increases or decreases are encountered and must be included in the Bemoulli equa-
13.5 MEASUREMENT OF PRESSURE 415
id_ ·~~Energy Grade 1-;- -r

2ut- I HYdra ·
~lic Grade•·
~ L
-~ 2~
'i,';
1
1
Arbitrary Horizontal Datum Plane

--------'-
Figure 13.4. Energy balance for two points in a fluid. (From R. A. Dodge and M. J.
Thompson, "Fluid Mechanics," McGraw-Hill, New York, 1937. Copyright © 1937
McGraw-Hill.)
tion. Thus, an energy balance may be written for two points in a fluid, as shown
in the example in Figure 13.4.
Note the pipe friction loss from point 1 to point 2 is hL foot pounds per pound
of fiowing fluid; this is sometimes referred to as the head loss in feet of fluid. The
equation may be written as follows:
Z1
144P1 vi
+ - - + - = 2i + - - +
144P2
+ h, (13.3)
Pi 2g P2 2g ~
All practica! formulas for the flow of fluids are derived from Bemoulli's theo-
rem, with modifications to account for losses due to friction.
13.5 MEASUREMENT OF PRESSURE
Figure 13.5 graphically illustrates the relationship between gauge and absolute
pressures. Perfect vacuum cannot exist on the surface of the earth, but it never-
theless makes a convenient datum for the measurement of pressure.
Barometric pressure is the leve! of the atmospheric pressure above perfect vac-
uum
"Standard" atmospheric pressure is 14.696 pounds per square inch, or 760
millimeters of mercury.
Any Pressure Above Atmospheric

~ 1¡;
"' :!l
~
o':
e:
O..-
"'
~
..,
u
0
~~~~§1--~~A~t~A~tm~o~sp~h~er~ic~P~re~ss~ur~e~Le~v~el_-~Va~r~iab~I•~-~~_._
¡ :;;
ro
E
+ "'
~ >
~-
"'
11
Any Pressure Below Atmospheric
"'
~
o:"
"'
Oí
s¡
~
Absolute Zero o! Pressure-Perfect Vacuum
Figure 13.5. Relationship between gauge and absolute pressures.
Gauge pressure is measured above atmospheric pressure, while absolute pres-

sure always refers to perfect vacuum as a base.
Vacuum, usually expressed in inches of mercury, is the depression of pressure
below the atmospheric level. Reference to vacuum conditions is often made
by expressing the absolute pressure in inches of mercury; also millimeters
of mercury and microns of mercury.
13.6 DARCY'S FORMULA-GENERAL EQUATION FOR FLOW

OF FlUIDS
Flow in pipe is always accompanied by friction of fluid partides rubbing against

one another, and consequently, by loss of energy available for work; in other
words, there must be a pressure drop in the direction of flow. If ordinary Bourdon
tube pressure gauges were connected to a pipe containing a flowing fluid, as shown
in Figure 13.6, gauge P 1 would indicate a higher static pressure than gauge P2 •
The general equation for pressure drop, known as Darcy's formula and ex-
pressed in feet offluid, is hL = fLv 2 / D2g. This equation may be written to express
pressure drop in pounds per square inch, by substitution of proper units, as fol-
~h_______-,.-___,9 Figure 13.6.

13.6 DARCY'S FORMULA-GENERAL EQUATION FOR FLOW OF FLUIDS 417
lows:
tlP = pfLv2
( 13.4)
144D2g
The Darcy equation is valid for laminar or turbulent flow of any liquid in a
pipe. However, when extreme velbcities occurring in a pipe cause the downstream
pressure to fall to the vapor pressure of the liquid, cavitation occurs and calculated
flow rates will be inaccurate.
Equation (13.4) gives the loss in pressure dueto friction and applies to pipe of
constant diameter carrying fluids of reasonably constant weight density in straight
pipe, whether horizontal, vertical, or sloping. For inclined pipe, vertical pipe, or
pipe of varying diameter, the change in pressure dueto changes in elevation, ve-
locity, and weight density of the fluid must be made in accordance with Bernoulli' s
theorem. For an example using this theorem, see (Sections 13.2-13.5).
Friction Factor
The Darcy formula can be rationally derived by dimensional analysis, with the
exception of the friction factor, f, which must be determined experimentally. The
friction factor for laminar flow conditions (Re < 2000) is a function of Reynolds
number only; whereas, for turbulent flow (Re > 4000), it is also a function of the
character of the pipe wall.
A region known as the "critica! zone" occurs between Reynolds numbers of
approximately 2000 and 4000. In this region, the flow may be either laminar or
turbulent depending upon severa! factors; these include changes in section or di-
rection of flow and obstructions, such as valves, in the upstream piping. The fric-
tion factor in this region is indeterminate and has lower limits based on laminar
flow and upper limits based on turbulent flow conditions.
At Reynolds numbers above approximately 4000, flow conditions again become
more stable and definite friction factors can be established. This is important be-
cause it enables the engineer to determine the flow characteristics of any fluid
flowing in a pipe, providing the viscosity and weight density at flowing conditions
are known. Forthis reason, Equation (13.4) is recommended in preference to sorne
of the commonly known empirical equations for the flow of water, oil, and other
liquids, as well as for the flow of compressible fluids when restrictions previously
mentioned are observed.
If the flow is laminar (Re < 2000), the friction factor may be determined from
the equation:
J= 64 = 64 µe = 64 µ
Re Dvp 124dvp
Ifthis quantity is substituted into Equation (13.4), the pressure drop in pounds per
square inch is:
µLv
íiP = 0.000 668 -¡j2 (13.5)
which is Poiseuille's law for laminar flow.

When the flow is turbulent (Re > 4000), the friction factor depends not only
upon the Reynolds number but also upon the relative roughness, e/ D, the rough-
ness of the pipe walls (e), as compared to the diameter of the pipe ( D). For very
smooth pipes such as drawn brass tubing and glass, the friction factor decreases
more rapidly with increasing Reynolds number than for pipe with comparatively
rough walls.
Since the character of the interna! surface of commercial pipe is practically
independent of the diameter, the roughness of the walls has a greater effect on the
friction factor in the small sizes. Consequently, pipe of small diameter will ap-
proach the very rough condition and, in general, will have higher friction factors
than large pipe of the same material.
The most useful and widely accepted data of friction factors for use with the
Darcy formula have been presented by L. F. Moody* and are reproduced in Fig-
ures 13.11-13.13. Professor Moody improved upon the well-established Pigott
and Kemlert friction factor diagram, incorporating more recent investigations and
developments of many outstanding scientists.
The friction factor ,J, is plotted in Figure 13 .12 on the basis of relative rough-
ness obtained from the chartjn Figure 13.13 and the Reynolds number. The value
off is determined by horizontal projection from the intersection of the e/ D curve
under consideration with the calculated Reynolds number to the left hand vertical
scale of the chart in Figure 13 .13. Since most calculations involve commercial
steel or wrought iron pipe, the chart in Figure 13.12 is furnished for a more di-
rection solution. It should be kept in mind that these figures apply to clean new
pipe.
Effect of Age and Use on Pipe Friction

Friction loss in pipe is sensitive to changes in diameter and roughness of pipe. For
a given rate of flow anda fixed friction factor, the pressure drop per foot of pipe
varíes inversely witli the fifth power of the diameter. Therefore, a 2 % reduction
of diameter causes a 10% increase in pressure drop; a 5% reduction of diameter
increases pressure drop 23 %. In many services, the interior of pipe becomes en-
*L. F. Moody, ''Friction Factors for Pipe Flow,'' Transactions of the American Society of Mechanical
Engineers, Volume 66, pages 671-678, November 1944.
tR. J. S. Pigott, "The Flow of Fluids in Closed Conduits," Mechanical Engineering, Volume 55,
No. 8, page 497, August 1933; E. Kemler, "A Study of Data on the Flow of Fluids in Pipes,"
Transactions ofthe American Society of Mechanical Engineers,.Volume 55, page HYD-55-2, 1933.
13.7 FLOW THROUGH NOZZLES AND ORIFICES 419
crusted with scale, dirt, tubercules or other matter; thus, it is often prudent
to make allowance for expected diameter changes.
Authorities point out that roughness may be expected to increase with use (due
to corrosion or incrustation) at a rate determined the pipe material and nature
of the fluid.
13. 7 FLOW THROUGH NOZZLES AND ORIFICES
The discharge of fluids through nozzles and orifices has been subject to continued
investigation and, as a result, well-established data are still being supplemented.
A portion of the subject is covered on these facing pages but more complete ref-
erences will be found from the data supplied by meter manufacturers.
The rate of flow of any fluid through an orifice or nozzle, neglecting the velocity
of approach, may be expressed by:
( 13.6)
Velocity of approach may have considerable effect on the quantity discharged

through a nozzle or orifice. The factor correcting for velocity of approach,
may be incorporated in Equation (13.6) as follows:
(13.7)
The quantity
is defined as the flow coefficient C. Values of C for nozzles and orifices are shown
in Figure 13 .14 and 13 .15. U se of the ftow coefficient C eliminates the necessity
for calculating the velocity of approach, and Equation (13.7) may now be written:
r;;:-;- [2;( 144) L:i. p

q = CA v2ghL = CA~ -"'-·~P~- (13.8)
Orífices and nozzles are normally used in piping systems as metering devices
and are installed wíth fiange taps or pipe_ taps in accordance with ASME specifi-
cations. The values of h¿ and ll.P in Equation (13.8) are the measured differential
static head or pressure across flange taps when values of C. The fl.ow coefficient
C is plotted for Reynolds numbers based on the intemal diameter of the upstream
pipe.
Flow of liquids
For nozzles and orifices discharging incompressible fl.uids to atmosphere, C values
may be taken from Figure 13.14 if hL or !lP in Equation (13.8) is taken as the
upstream head or gauge pressure. For most conditions of flow of fl.uids having a
low viscosity, i.e., water, gasoline, etc., the Reynolds number need not be cal-
culated since it will fall in the range of the values in Figure 13 .14, where the flow
coefficient C is a constant.
13.8 PRESSURE DROP ANO VELOCITY IN PIPING SYSTEM
Example 13.1 Bemoulll's Theorem-Water

Given: Water at 60 F is fl.owing through the piping system, shown in [Fig. 13.7],
at a rate of 400 gallons per minute.
Find: The velocity in both the 4 and 5-inch pipe sizes and the pressure differential
between gauges P 1 and P2 •
Solution:
1. Use Bemoulli's theorem:
Z¡
144P vf
+ - -1 + - = Z2 + --- + - +
144P? v~ h
L
Pi 2g P2 2g
5" Welding Elbow 5" Schedule 40 Pipe P,
4" Schedule
40 Pipe
P,
FLOW 5" Schedule 40 Pipe
-!-,....__;;;_,;11.-..ll..,..<:..:1C..-- _ _ _ _ _ _ Eltva.!i2!Jl::1.:. O
5" x 4" Reducing Welding Elbow
Figure 13. 7,
13.8 PRESSURE DROP ANO VELOCITY IN PIPING SYSTEM 421
Since, p1 = p2
h _ 0.00259KQ 2
2. L - d4
50.6Qp
Re = dµ,
L
K=f-
D
K= JL
D{34
K = l4fr
K = 141" + (1 - (32)2
JT (34
Note: In the absence of test data for increasing elbows, the resistance is conserv-
atively estimated to be equal to the summation of the resistance due to a straight
size elbow and a sudden enlargement.
3. p = 62.371 (Table 13.5)

µ, = 1.1 (Fig. 13.10)
d1 = 4.026 ( 4" Sched. 40 pipe, Table 13.8)
d2 = 5.047 (5" Sched. 40 pipe, Table 13.8)
fr = 0.016
4. (3 = 4.026 = o 80
5.047 .
Z2 - Z1 = 75 - O = 75 feet
v 1 = 10.08 ( 4" pipe, Table 13.9)
v2 = 6.42 (5" pipe, Table 13.9)
V~ - vi = 6.422 - 10.082
2g 2 X 32.2 = -0.94 feet
5. For Schedule 40 pipe,
R. = 50.6 X 400 X 62.371 = 2 .85 X l05 (4" pipe)

4.026 X 1.1
Re = 50.6 X 400 X 62.371 = 2 _27 X l05 ( 5" pipe)

5.047 X 1.1
f = 0.018 (4 or 5" pipe)
0.018 X 225 X 12
6· K or K 9.6 for 225' of 5" Sched. 40 pipe
= 5.047
0.018 X 110 X 12
or K 5.9 for 110' of 4" Sched. 40 pipe
K = 4.026
With reference to velocity in 5" pipe,
K2 = 5.9 + 0.84 = 14.4

K = 14 X 0.016 = 0.22 (5" 90º elbow)
0.362
K = 0.22 + 0 _84 = 0.54 (5 x 4" 90º elbow)
7. Then, in terms of 5-inch pipe,
KTOTAL = 9.6 + 14.4 + 0.22 + 0.54 = 24.8
8.
h = 0.00259 X 24.8 X 4002 = 15 8
L 5.0414 •
62.371
9. P1 - P2 =---¡¡¡-- (75 - 0.94 + 15.8) = 39.0
13.9 PIPE LINE FLOW PROBLEMS
Example 13.2 Sizing of Pump for 011 Pipe Lines

Given: Crude oil 30 degree API at 15.6 C with a viscosity of 75 Universal Saybolt
seconds is flowing through a 12-inch Schedule 30 steel pipe at a rate of 1900
í3.9 PIPELINE FLOW PROBLEMS 423
barrels per hour. The pipe line is 50 miles long with discharge at an elevation of
2000 feet above the pump inlet. Assume the pump has an efficiency of 67 per cent.
Find: The brake horsepower of the pump.
Solution:
fL B2
l. !:.P = 0.000 1058 ~
d
t = l.8tc + 32
R = 35 4Bp
e . dµ
144LiP
p
brake QHp
horsepower 247 OOOeP
2. t = (1.8 X 15.6) + 32 = 60F

3. p = 54.64 (Table 13.6)
S = 0.8762 (Table 13.6)
4. d = 12.09 (Table 13.8)
d 5 = 258 304
5. 75 USS = 12.5 centipoise Fig. 13.8
35.4 X 1900 X 54.64
6. R - - 24 100
e - 12.09 X 12.5 - -
7. f = 0.025 (Fig. 13.11)
!:.P = 0.000 1058 X 0.025 X 50 X 5280 X 54.64 X 19002

~~~~~~~~~~~~~~~~~~-
8.
258 304
b..P = 533
9. HL = 144 X 533 = 1405

54.64
1O. The total discharge head at the pump is:
H = 1405 + 2000 = 3405

11. Q = ( 1900 bbl) (42 gal) (~) = 1330

hr bbl 60 mm
12. Then, the brake horsepower is:
1330 X 3405 X 54.64 = 1496

or say 1500
247 000 X 0.67 '
13.1 O FLOW THROUGH ORIFICE METERS
Example 13.3 Liquid Service

Given: A square edged orifice of 2. 0-inch diameter is installed in a 4-inch Schedule
40 pipe having a mercury manometer connected between the ftange taps.
Find: (a) The theoretical calibration constant for the meter when used on 60 F
water and for the ftow range where the orifice ftow coefficient C is constant and
(b), the ftow rate of 60 F water when the mercury deftection is 4.4 inches.
Solution (a)
l.
50.6Qp
Re= dµ (Fig. 13.15)
2. To determine differential pressure across the ftange tips,

íiP = íihmP
12 X 144
where: íihm = differential head in inches of mercury

3. The weight _density of mercury under water equals Pw(SHg - Sw), where (at
60 F):
pw = density of water = 62.371 (Table 13.5)

Sg = specific gravity of mercury = 13.57 (Table 13.6)
Sw = specific gravity of water = 100 (Table 13.5)
4: And p of Hg under H 20 = 62.371 (13.57 - 1.00)

= 784 lb/ft3
13.10 FLOW THROUGH ORIFICE METERS 425
5. AP = Ahm ( 784 ) = 0.454Ah

12 X 144 m
6. d1 = 4.026 (Table 13.8)
7. d0 = 2.00 = 0.497
d¡ 4.026
8. e= o.625 (Fig. 13.15)
2 0.454Ahm
9. Q = 236 X (2.0) X 0.625
62.34
Q = 50 .4 .JA hm ( calibration constant)
Solution (b):
10. Q = 50.4 .JAhm = 50.4 J4A = 106

11. µ, = 1.1 (Fig. 13.10)
12.
R = 50.6 X 106 X 62.371
e 4.026 X 1.1
Re = 75 500 or 7.55 X 104
13. C = 0.625 is correct for Re = 7 .55 X 104 ; therefore, the flow rate through
the pipe is 106 gallons per minute.
14. When the C factor on Figure 13.15 is incorrect, for the Reynolds number
based on calculated flow, it must be adjusted until reasonable agreement is
reached by repeating Steps 9, 10, and 12.
Example 13.4 Laminar Flow

In ftow problems where the viscosity is high, calculate the Reynolds number to
determine the type of ftow.
Given: SAE 10 Lube Oil át 90 F is ftowing through a 3-inch Schedule 40 pipe and
produces 0.4 psi pressure differential between the ftange taps of a 2.15-inch l. D.
square edged orifice.
Find: The flow rate in gallons per minute.
Solution:
l.
R = _50_._6Q_p
e dµ
2. µ = 38 ( suspect laminar flow; Fig. 13.16)

3. d¡ = 3.068
4.
d0 = 2.15 = 0 _70
d¡ 3.068
5. C = 0.8 (Fig. 13.15)

6. S = 0.876 at 60 F
S = 0.87 at 90 F
7. p = 62.4 X 0.87 = 54.3
r;:;
8. Q = 236 X 2 l'i 2 X 0 8 / 0. 4 = 75
.- . '\J 54.3
50.6 X 75 X 54.3
9. Re= ~68 X 3S-- = 1768
10. e= o.n for Re= 1768 (Fig. 13.15)
Since the assumed C value of 0.8 is not correct, it must be adjusted by

repeating Steps 5, 8, 9, and 1O.
11. e= o.91 (Fig. 13.15)
12. Q = 236 X 2.15 2 X

lo4
0.91 /-·- = 85.3
'\J 54.3
50.6 X 85.3 X 54.3
13. R = = 2010
e 3.068 X 38
14. e= o.91 for Re = 2010 (Fig. 13.15)
Since C = O. 91 is correct for the flow, the flow through the meter is 85. 3
gallons per minute.
13.11 REFERENCE FIGURES ANO TABLES 427
13.11 REFERENCE FIGURES ANO TABLES
1000
700 The empirical relation between Saybolt Universal Viscosity

and Saybolt Furo! Viscosity at 100 F and 122 F, respectively.
600 µe µ. and Kinematic Vi'scosity is taken from A.S.T.M, D2161-63T.
50~ .05
1000 At other temperatures, the 3aybolt Viscosities va1·y only
soo .04 2000 slightly.
~ 800
g 700 .03 Saybolt Viscosities above those shown are given by the rela-
~ 300 tionships:
600
=
¿. 500 .01 1000 Saybolt Universa! Seconds Centistokes x 4.6347
90~
§ 400
-80
700
Sayboit Furo! Seconds = Centistokes x 0.4717
> 600
~ 300 .01
.009
.008
.007
100
.006
.005
10--~-
200 35
60---. s
! ¡¡¡
(_)
10
20 o:
<C
-;j'. .0004 10
20
.0003
30
1
~
=
.0002 50 ·¡;;
!?
lO
liO "'
9 0.7 70
8
.0001 Problem 1: Determine the absoiute viscosity 80
7 .00009 of an oil which has a kinematic viscosity of 82
•00008 4 centistokes and a specific gravity of 0.83 . 00
.00006 3 Solution. 1: Connect 82 on the kinematic vis- 100
cosity scale with. 0.83 on the .specific gravity 0.6
.00005 sea.le; read 6i centipoise at the intersection on
""'I 2 the absolute viscosity scaie .

.00004 -
.00003 Problem 2: Determine the absolute viscosity of

an oil having a speciflc gravity of 0.83 and a
Saybolt Furo! viscosity of 40 seconds . 0.5-
.00001 l
Solution 2: Connect 0.83 on the specific gra.vity
scale with 40 seconds on the Sayboll Furo! sca1e;
read 67 centipoise at the intersection on the
abso!ute viscosity scaie.
Figure 13.8. Equivalents of kinematic, Saybolt Universal, Saybolt Furo!, and absolute
viscosity.
10
6.0
~ \ 1¡ ~
5.0
4.0
3.0
1,
\'~ \ \\ ~
\
11~ ~\ 16
\'
2.0
~
~ ~\
'
\ I\'\
"~
~ '"~"~ i"\
'
'\
~~
"'
1.0 .........5" ~
12
.........
' ' ''

' '
•9
.8 \.' 6 1\..
' ....
"'"
'
"" " " ~
5l .7 "'<''- '\. ''"'\."lo.
·g_ .6 ...... "'" ........... ..............
...........
~ " ["-...
"' ~~
""
e:
8 .5
4 í'..." '\... .............
....... ............. r-....._
.5
¡¿. .4
l \.
.......
-~
~ .3
~
~
~ r':: '- ~ ~
~
~ ......_
~
"' .............
1
"""" --
""
::t.
~~
.2
~-
'\.
_...;:
~
- r--
0.1
1\ ~ "-
.09
.08 '\ .........
.07
.06
\ '" ~
\
.05
.04
\ ' "'t'-....
"""
.03
-40 o
\ 40 80 120 160 200 240 280 320 360
t - Temperature, in Degrees Fahrenheit
1. Car bon Dioxlde. , CO,
2. Ammonia •.•• , ••• NHa 9. Ethyl Alcohol 16. 103 Sodium Chlorlde Brine ••• 103 NaCI
3. Methyl Chlorlde. , CHaCI 1O. lsopropyl Alcohol 17. 203 Sodium Chlorlde Brine ••• 203 NaCI
4. Sulphur Dloxlde •• so, 11. 203 Sulphuric Acid •••.•• 203 H,so, 18. 103 Calcium Chlorlde Brlne •• 103 CaClt
S. Freon 12 ........ F-12 12. Dowtherm E 19. 203 Caldum Chloride Brlne •• 203 CaClt
6. Freon 114 ....... F-114 13. Dowtherm A
7. Freon 11 ........ F-l 1 14. 203 Sodium Hydroxide •• 203 Na OH Example: The viscosity of am-
8, Freon 113 ....... F-113 15. Mercury monia at 40 F is 0.14 centipoise.
Figure 13.9. Viscosity of various liquids (data from "The Dowthenn Handbook," Dow
Chemical Company, Midland, MI, 1954; R. F. Steams, R. M. Jackson, R. R. Johnson,
and C. A. Larson, "Flow Measurement with Orifice Meters," D. Van Nostrand, New
York, 1951; and "Air Conditioning Refrigerating Data Book-Design," 9th edition, Amer-
ican Society of Refrigerating Engineers, New York, 1955).
428
¡]
1\
1S
\
\
119
\
\
\
1
21
\ ·--
1 --- +ffffi
1
'
L--~
·-
\ \
' \
\ \ 4. Naturc;l Gaso/ine
\ \ \ \ \
l6r--. 1 \\ \ \ \ 1
1
6. Water
\ \ \ \ \ 1
13 i\ \ \ \ \ \ e. Disti!!ot®
' ' 9. 48 Deg. API Crudlij
'' l O. 40 Oeg. API Crude
11. 35.6 Dog. API Crndo
l 2. 32.6 Dog. API Crud•
l 3. Soif Creek Crucle
14. ruel 3 IMo•.)
15. ~uel 5 (MinJ
SAE 1O lubo 1100 V.I.)
!7. SAE 30 lube (100 V.I.)
18. Fue! 5 (Mox.J or

Fuol 6 (Min.I
6 ¡-....._ ""' \ :-.~' ~,, \ \ \ \ \ \
19. SAE 70 Lub• (100 V.I.)
20. Bunktir C Fue! {Mali:.) and

M.C. Residuum
~l. A•phcl!
Data extractcd in part

by pcrmission from thc
Oil and Gas Journal.
''
1 "
1 '
1
30 40 60
'
80 100
1
200 300 400 600 800 1000

-Temperalure, in Degrees Fahrenheit
Ex11mple: The viscosity of water at

125 F is 0.52 centipoise (Curve No. 6).
Figure 13.10. Viscosity ofWater and Liquid Petroleum products (data from R. F. Stearns,
R. M. Jackson, R. R. Johnson, and C. A. Larson, "Flow Measurement with Orifice Me-
ters," D. Van Nostrand, New York, 1951; W. L Nelson, "Petroleum Refinery Engineer-
ing," McGraw-Hill, New York, 1949; E. C. Bingham and R. F. Jackson, Bureau of Stan-
dards Bulletin 14, Special Publication 298, August 1916, pp. 58-86; data extracted in part
by permission from Oil and Gas Journal).
429
VALUES OF (vdJ FOR WATER AT 60º F {VElOCiTY IN FT./SEC. ){ O!AMETIER IN iHCHESl
·~" ,.f ,?'

~
400 000 !iOO
.os
.04
.06' '' •• '1 ' ' ' ' -.-..._,, I\.
.03
.05
.02
.015
.04
.Ol
.008
f
.006
Friction .03 Relative
Factor = .004 Roughness
h '· .015
=
(~\'"'. .002 D
DI ~g
.01
·8&\s
.0006
.0004
.015
.IJfJ1l1
.0001
.IJOOIJS
-:m111111111111111111111111111111111111111111•
. 10' 2 3 4 5 6 B lil' 1 3 4 5 6 8 !O' 1 3 4 5 6 8 !O' • . . . . .
Dvp
R,. Reynolds Number = -
µ,
For olher fornu of th~ Re equation, ieii po:.igft 3·2.
Problem:
Determine the friction factor for io-inch cast iron Solution: The relative roughness (see page A-23) is
pipe (10.16" l.D.) ata Reynolds number ftow of 30.000. 0.001. Then, the friction factor (j) equals 0.010.
Figure 13.11. Friction factors for any type of commercial pipe (data from L. F. Moody, "Fric-
tion Factors for Pipe Flow,'' Transactions of the American Society of Mechanical Engineers,
Volume 66, pp. 671-678, November 1944).
VALUES OF (ud) FOR WATER Ar GOº F í\fEl..OCl"rY IN FT./SEC. X OIAMETER IN iNCHES)
'),~ "'~~i?~~~~
Nominal
Pipe
Size,
lnches
f_E==t::::; ___ . ._~1/8
f
Friciion
Factor =
--t 11·1J 111 H-ttt= H+~HH f-H#--++ttt+ft+l-Hl º·'º -P
---r
(;;_)~
D 2g . 02 h-1+1--!--1 1 1\1111 11111 1 1 111111
llllJJJII:· [!a
4 6 8 1 1 1
000024
o o o
S<:hedule Numher
36
46
.009 LLLU_L_LJ.......Ll...1 .LLLLU. '---l-'---..L.L..LLLLLLLL----L--....!........!.....LJ ...L.LL .. l 1

2 3 4 5 6 8 10' 2 3 4 5 6 310' 2 3 4 5 6 810' 2 3 4 5 6 810 7 2 3 4 5 G 8 JOB
DvP
R, ·· Reynolds Number = =µ;- For other fo:-m$ of the fl~ equatlon, ~ page J.2.
Problem: Determine the friction factor for 1 2-inch Sched- Solution:

ule 40 pipe at a flow having a Reynolds number of 300,000. The friction factor (f) equals 0.016.
Figure 13.12. Friction factors for clean commercial steel pipe (data from L. F. Moody, ''Friction
Factors for Pipe Flow, '' Transactions American Society of Mechanical Volume
66, pp. 671-678, November 1944).
Pipe Diameter, in Feet -D
.1 ? .3 .4 .5 .6 ' . B , 1 ' 3 4 5 ,10 lO 25
.o 5 ' ' 1
' ~
07
.04
.03 '
-" ' - .06
.02 "' 1\... "' ~ " 1\.. ~ .05

1\... l'I""
"',,
.... "'
- '"
"'r-. ..... 04
.01
.008
~
-- ,'
' .... RV ED ''
~
035
.006 ' ....
T L
.005 1'-'.... .... ,.... ...

:004 .... ....'
.... !'-
"' -.03
.003
'
,...,
''
N R
W< O
E
' '
""
' "'-'.,.
.002 ' ' !"\"'" ""'"~, I" r-,1'
T/W ~
' ' "'ro. ~ ...

025
8.
'' "r-. 'ro. ,ro.
' '"' -..
1\..
', ' e:
--' 018 ¡;f

..,¡~ ro. '!\.. ro., ...... 02 ~
1 .001
:!! .0008
"' .0006
'- I'
....
...
,'
.,:1.s->' ....
tr'~-
<>::
~ ' ' ' Q)

:;
'lo.." (1'º''
a .0005
'!; .0004
l'I
'
~...
' -\~'llrrrr. ~""
' ''
' '" - 016 .=-e
.:: ~ 0,,._
'''\, <~'?"'
' ~
'"<_. ~<~~"-. "
~ .0003 -~-s- . .,..""""'
\,.
-....[,,
~
o::
(~ ~º~' 1, " ,
.0002
rt<-r
r-e.1> ..">-"'' r-..'
~ "'---- - ....
'
,~,
"IP..
.....
- .014 8
.f
1
~o º+
r-..,' ~ ....
.0001
.00008
~-\.
'<~o
~
'
~'
"" ' -·~-
,.,. - -
"',...,.'11 !'-
' 012
'-.
.00006 '
.00005 " ...."º ' ,.... '"'
.... '~
,.- -
,.,. ............
.00004 ~-
""' -' ..... 01
,...
.00003 "1~, .......
.00002
N
..... , ~
' "of
' .....009
', ~~
.00001
.000008
'
""'1
- ~
008
.000006 1 ......t '

.0000051 2 4 5 8 10 20 30 40 50 60 80 100 200 300
Pipe Diameter, in loches -d
Data extracted from Priction
Faclors for Pipe Flow by L.F. Problem: Determine absolute and relative roughness, and friction fac-
Moody, with pcrmission of thc tor, far fully turbulent flow in 10-inch cast iron pipe (l.D. = 10• 1 6').
publishcr, The American Soci-
cty of Mechanical Enginccrs. Solution: Absolute roughness (•) = 0.00085 ..... Relative roughness
29 Wcst 39th Strcct, New York. (•/D) = 0.001 ..... Friction factor at fully turbulent flow (/) = o.otQ6.
Figure 13.13. Relative roughness of pipe materials and friction factors for complete tur-
bulence (data from L. F. Moody, "Friction Factors for Pipe Flow," Transactions ofthe.
American Society of Mechanical Engineers, Volume 66, pp. 671-678, November, 1944).
432
d,/d2=~
1.20 ~75
¡...-- ~
u
¡....- ..... ,,725
u ./ '-
/ .... ~ ~
d.
1.1
1.1 , ~
,,.,...... ....
"""' .....
¡...--
-- 70
675 a..
!ij
a
-~
-
./ .s
1.11 1 -~
......... .65 Qj
1.081
/ ..,...v L.-- 1il
1ij
......... .625
/ _,..
---
~
i-- e
Flow- 1.116; """' 1.&0
C1>
/ ..........
/ """' .......
M75 Ñ
z::l
,_,,_ ..
o.55
1:041
-- ........,..i-
cd
--
~- ..,.t'7 ,.,... 1.50 o
1.02! """'
........... _,... ...
~
~
c=-v1-f14 , ~ ....... _
~
-
45
C for nozzles
1.001
0.981
;
,......
'.:¿ .....:::;: ....... _........ ----
40
JI
0.20
"'
o::
0.96;
,,-~ ~
Example: The ftow coeffi- ... ~-
e
cient for a diameter ratio 0.94
',,' ,
{3 of 0.60 at a Reynolds 0.92
number of 20,000 (2 x 1a4) 2 6810' 2 4 68105 2 4 68106
equals 1 .03. /l,. - Reynolds Number based on d2
Figure 13.14. Flow coefficient C for nozzles (data from "Fluid Meters, Part I," 4th edition,
American Society of Mechanical Engineers, 1937).
e
1 - 132
Korifice ~ C2 4 /3
Flow- e for square-edge

e
1.3
1
l. 2
17
'
11
l.
,.._ ....__ =~·;¡;-=.SO
d,
/
./
1.0
_.__
·--
=.75
I\\ ,/ ..-- --.......... \
I~ v ... "' __..
1/
/
1\ \~
-.70
¡- =.65 -r--..._
o. 9 =.60
I~ ~ v"
~
__..
r--."f'....."- !"-....
=.50 -i- !'--...._
l./~
o. 8 1
¡ t:-¡,.... ,...
0.7
1
,/ ~ -- t-- r-- ~ r-- ........
t:'t--
-~
r-- t-- ,.... ....

0.6
.~ ~
175 r--;;::: ~- f.-._
11
~
,._ ...
v"~ "" ~
~¡,,
,, [A ~o-d;
d,
= 401- V/1
0.5
8
\:: ,____
~
= .301- (_/
0.4 ~ ~ = o ... :
~
6 8 10 '¿Q 4Q60BOJ02 4 6 8 10' 4 6 8 10•
Rr - Reynolds Number based on d 2
--
0.78
---
0.76
-
¡...._
..._ j
·º.75 º.,
0.7 4
r-- ¡...._
- ,_ ---
L
0.71 o.115f
t-- -¡..
0.7o o.70 ~
1---~ .,
~
e16
0.68
...._ o.65
0.66 fl
r--..__ o.60 ~
0.64 o
~ ...___ o.55 'O
-
.0.61
0.60
0.58
-
1
-~
6 8 10• 6 8 10' 6 8 10•

o.50 ~
o.45 &;
~.40
0.30
I' 0.10
R,. - Reynolds Number based on d 2
Figure 13.15. Flow coefficient C for square-edge orífices (data from "Fluid Meters, Part
1," 4th edition, American Society of Mechanical Engineers, 1937; G. L. Tuve and R. E.
Sprenkle, "Orífice Coefficients for Viscous Liquids," Instruments, p. 201, November
1933).
434
r
q Q w d
20 000 Jnternal Pipe
D1ameter, lnches .5
lndex
'~
8 000
~f
.6
·1
10
10
-JO 000
·ª 000
6 000
.
6
000
4 000
3 000
.4
.5
p
6 ~3000
.6 3/4
B 4 000 R, .7
lOOO .8 70
10 000
6000 3 l.O
2 llOO
4 000 1
!• -~m
1000
000 1 fü 65
¡ 000 l 000 \
3/4 l.5
llOO 400 600 lll
300 ~ 400 1/1
GOO ~
©
w o
200 ©
~
~
~ "-
400-¡¡ 200 ~
6IJ
e
100 " o
~
"
50 ·- ,_,
~o
w w ;,;: 2\\ 1.5 :;.
100 40 o
llO ~ \' •\ "-
~
10 ·-
~
~ "' -;; ~
50 ~
o
lO IH-~o : 3
w
55
40 ~ ~ 3* ©
"-
e
30 ¡:: ~
;;; ¡¡¡ ·-
~~ "' b
- 10
·-
io ~
.O! .02 .03 .04 .o 11l-di ';;
w
-;;;
!lJ ~
·;;
f
"
N
¡;: ~ Friction Factor !or Clea11 Z.i
©
';;
.6
.4 Stoel Pipe 30-'l~ -;;¡ ~
w
lO ·¡¡;
8
w
.2 ' ' ""
"' .! "' ~ "" 1
~
' ·8
<>.
.02 45
9
100 10 JO
.O!
11
.008
! 14
.006 .8 :!JO 16 15 40
.004 .6
300 18
.003 .4 1{)
.3 '!JO 20
.002l1
.8 500 24
.2 -14
.6 6lllJ
t
U"I
Figure 13.16. Reynolds number for liquid flow friction factor for clean steel pipe.
TABLE 13.1. Equlvalents of Absolute (Dynamic) Vlscoslty
TO OllTAIN - - ¡ Centipoi1e Poise
~e
tPoundm
Ft Sec
'.>lULTlPLY
Grsm *Pound¡ Sec Poundal Sec
Cn;i. Sec Ft' Ft'
Dyne Sec
~
(µ) (100 µ) (µ',) (µ,)
1 '\"'"
C•ntipoi•• (µ) 11'. O.O! 2.09 (I0- 5) 6.72 (10-•)
Poi se
1
Gram
CmSec (100 µ) 100 l 2.09 (!O-•) 0.0672
Dyne Sec
~
1
Slugo 1
Ft Sec
(µ'e) 47 90ll 479 1 g or 32.2
•:Pound1 Sec
Ft•
1
tPoundm
Ft Sec 1
Poundal Sec 1 (µ,) 1487 14.87 l.or .0311 1

fJ
Ft'
1 1
*Pound¡= Pound of Force

1
tpoundm = Pound of Mass
To convert absolute or dynamic viscosity from one Asan example, suppose a given absolute viscosity of
set of units to another, locate the given set of units 1 poise is to be converted to slugs 1 foot second. By
in the left hand column and multiply the numerical referring to the table, we find the conversion factor
value by the factor shown horizontally to the right to be 2.og ( I0-3). Then, 2 (poise) times 2.og ( 1 o-3)
under the set of units desired. = 4.18 (10-3) = 0.00418 slugs. foot second.
TABLE 13.2. Equivalents of Kinematic Viscosity

TO OBTAI\;-- Centistokes Stokes Ft 2
Sec
\!LLTIPLY Cm'
l BY
~
(,)
Sec
(IOO ,) (,')
Ceñtistokes (,)
11' 0.01 1.076 (I0-5)
Stokeo
Cm' (100 ,) 100 !.076 (10-')
Sec 11
Ft 2
(/) 92 900 929
Sec
11
To convert kinematic viscosity from one set of units As an example, suppose a given kinematic viscosity
to another, locate the given set of units in the left of o. 5 square foot.isecond is to be con verted to centi-
hand column and multiply the numerical value by stokes. By referring to the table, we, find the con-
the factor shown horizontally to the right, under version factor to be g2,goo. Then, 0.5 (sq ft/sec)
the set of units desired. times q2,goo = 46,450 centistokes.
436
TABLE 13.3. Equlvalents of TABLE 13.4. Equlvalents of
Klnematlc and Saybolt Universal Klnematlc and Saybolt Furol Vlscoslty
Vlscoslty Kinematic Equivalent Saybolt
Kinernatic Viscosity, Furo! Viscosity, Sec
Equivalent Saybolt
Viscosity, Universal Viscosity, Sec Centistokes At 122 F At 210 F
V
Centistokes At IOOF At 210 F
V Basic Values
48 25.3
50 26.1 25.2
1.83 31.01 32.13 60 30.6 29.8
2.0 32.62 32.85
4.0 39.14 39.41 70 35.1 34.4
80 39.6 39.0
6.0 45.56 45.88 90 44.1 43.7
8.0 52.09 52.45
10.0 58.91 59.32 100 48.6 48.3
15.0 77.39 77.93 125 60.1 60.1
20.0 97.77 98.45 150 71.7 71.8
175 83.8 83.7
25.0 119.3 120.1
30.0 141.3 142.3 200 95.0 95.6
35.0 163.7 164.9 225 106.7 107.5
40.0 186.3 187.6 250 118.4 119.4
45.0 209.1 210.5 275 130.1 131.4
300 141.8 143.5
so.o 232.1 233.8 325 153.6 155.5
55.0 255.2 257.0 350 165.3 167.6
60.0 278.3 280.2 375 177.0 179.7
65.0 301.4 303.5
70.0 324.4 326.7 400 188.8 191.8
425 200.6 204.0
75.0 347.6 350.0 450 212.4 216.1
80.0 370.8 373.4 475 224.1 228.3
85.0 393.9 396.7
90.0 417.1 420.0 500 135.9 240.5
95.0 440.3 443.4 525 247.7 252.8
550 259.5 265.0
100.0 463.5 466.7 575 271.3 277.2
120.0 556.2 560.1
140.0 648.9 653.4 600 283.1 289.5
160.0 741.6 625 294.9 301.8
180.0 834.2 650 306.7 314.1
675 318.4 326.4
200.0 926.9 700 330.2 338.7
120.0 1019.6 725 342.0 351.0
240.0 1112.3 750 353.8 363.4
260.0 1205.0 775 365.5 375.7
280.0 1297.7
800 377... 388.1
300.0 1390.4 815 389.1 400.5
310.0 1483.1 850 400.9 411.9
340.0 1575.8 875 412.7 425.3
360.0 1668.5 Saybolt Seconds
380.0 1761.1 equal 90Ó 414.5 437.7
Centistokes 925 436.3 450.1
400.0 1853.9 times 4.óó73 950 448.1 461.5
420.0 1946.6 975 459.9 474.9
440.0 2039.3
460.0 1131.0 1000 471.7 487.4
480.0 2224.7 1025 483.5 499.8
500.0 1317.4 1050 495.1 512.3
1075 507.0 524.8
Saybolt Seconds 1100 518.8 537.1
Over 500 equal 1125 530.6 549.7
Centistokes 1150 542.4 561.2
times 4.6347 1175 554.1 574.7
Note: To obtain the Saybolt Universal viscosity 1200 566.0 587.1
1225 577.8 599.7
equivalent to a kinematic viscosity determined at t, 1250 589.5 611.1
multiply the equivalent Saybolt Universal viscosity 1175 601.3 614.8
at 100 F by 1+ (t - 100) 0.000 064. 1300 613.1 637.3
For example. 10 v at 210 F are equivalent to 58.91
multiplied by 1.0070 or 59.32 sec Saybolt Universal
Over 1300 • t
at 210 F.
*OVER 1300 CENTISTOKES AT 122 F:
Saybolt Fluid Sec = Centistokes x 0.4717
These tables are reprinted with the permission of
the AmericanSociety for T esting Meteríais (ASTM) toVER 1300 CENTISTOKES AT 210 F:
The table at the left was abstracted from Table 1, Log (Saybolt Furo! Sec - 2.87) -
02161-ó3T. The table at the right was abstracted 1.0276 ll.og (Centistokes)] - 0.3975
. from Table 3, 02161-óH.
437
TABLE 13.5. Physical Properties of Water
Temperature Saturation Specific Weight Weight
of Water P;ressure Volume Density
1 1 1 1
t P'
1
v p
Pounds per 1
1
Degrees Square Jnch
1
Cubic Feet 1 Pounds per Pounds
Fahrenheit Absolute Per Pound Cubic Foot Per Gallon
1 1 1
32 ' 0.08859 0.016021 62.414

1 8.3436
40 0.12163 0.016019 62.426 8.3451
50 0.17796 0.016023 62.410 8.3430
60 0.25611 0.016033 62.371 8.3378
1
70 0.36292 0.016050 62.305° 8.3290
80 0.50683 0.016072 62.220 8.3176
90 0.69813 0.016099 62.116 8.3037
100 0.94924 0.016130 61.9% 8.2877
110 1
1.2750 0.016165 61.862 8.2698
120 1.6927 0.016204 61.7132 8.2498
130 2.2230 0.016247 61.550 8.2280
140 2.8892 0.016293 &1.376 8.2048
150 3.7184 0.016343 61.188 8.1797
160 4.7414 0.016395 60.994 8.1537
170 5.9926 0.016451 60.787 8.1260
180 7.5110 0.016510 60.569 8.0969
190 9.340 0.016572 60.343 8.0667
200 il.526 0.1)16637 60.107 8.0351
210 14.123 0.016705 59.862 8.0024
212 14.696 0.016719 59.812 7.9957
220 17.186 0.016775 59.613 7.9690
240 24.%8 0.016926 59.081 7.8979
260 35.427 0.017089 58.517 7.8226
280 49.200 0.017264 57.924 7.7433
300 67.005 0.01745 57.307 7.6608
350 134.604 0.01799 55.586 7.4308
400 247.259 0.01864 53.648 7.1717
450 422.55 0.01943 51.467 6.8801
501) 680.86 0.02043 48.948 6.5433
550 1045.43 0.02176 45.956 6.1434
600 1543.2 0.02364 42.301 5.6548
650 2208.4 0.02674 37.397 4.9993
700 3094.3 0.03662 27.307 3.6505
Specific gravity of water at 60 F = 1 .oo

Weight per gallan is based on 7.48052 gallons per cubic foot.
All data on volume and pressure are abstracted from ASME Steam
Tables (1967), with permission of publisher, The American Society of
Mechanical Engineers, 345 East 47th Street, New York, N. Y. 10017.
TABLE 13.6. Weight Density and Specific Gravity of Various liquids

Liquid Temp. Weight 1Specific Liquid ¡Temp.' Weight Specific
Density Gravity Density Gravity
1 p s p s
1 i
Deg
Fahr. 1cb~.- P~.r I Deg
Fahr·. 1c~: ~tr
Ace tone óO
1
49.4 0.792 Mercury 20 849.74
Ammonia,. Saturated 10 40.9 0.656 Mercurv 40 848.03 113.623
13.596
Benzenc 32 Sb.I 0.899 MercurY 60 846.32 !3.568
Brinc, IOC;~ Ca Cl 32 1 68.05
1 !.091 Mercury 1 80 1 844.62 13.541
Brine, 10% Na Cl 1 32 67.24 1.078 Mercury 100 842.93
BunkcrsCFue!Max. 60 63.25 1.014 Milk ... t l '3:'.l.4
Carbon Disulphide 32 80.6 1.292 OliveOil 59 57.J 0.919
Distillate 60 1 52.99 1 0.850 Pentanc 1 59 1 38.9 ' 0.624
Fue! 3 Max. 60 56.02 0.898 SAE 10 Lubej 60 54.M 0.876
Fue! 5 Min. 60 60.23 0.966 SAE 30 Lube! 1 60 56.02 0.898
Fue! 5 Max. 60 1
61.92 0.993 SAE 70 Lubet 60 57.12 0.916
Fue! 6 Min. 1 óO 1 61.92 1 0.993 Sal t Creek Crude 1 60 1 52.56 1 0.843
Gasolinc 60 46.81 0.751 32.6º API Crude 60 53.77 0.862
Gasoline, Natural 60 42.42 0.680 35 .6' API Crude 60 52.81 0.847
Kerosene 60 50.85 0.815 40' API Crude 60 51.45 0.825
M.C. Residuum 1 60 1 58.32 1 0.935 48' AP! Crude 1 óO 1 49.16 1 0.788
438
TABLE 13.7. Equlvalents of Degrees API, Degrees Baumé, Speclflc Gravity, Welght
Density, and Pounds per Gallon at 60 F /60 F
Degrees Values for API Scale Valuea for Baumé Scale
on
API Oil Liquida Lighter Than Water Liquida Heavier Than Water
or
Baumé Specific Weight Pounds ~cific Weight Pounda ~ific Weight Pounda
Gravity Density, per- ravity Density, vity Density, per
Se ale
Lb/Ft• Gallon Lb/Ft• G~n Lb/Ft• Gallon
s p s p s p
o ... . .. ... ... ... ... 1.0000 62.36 8.337

2 ... . .. . .. ... ... . .. 1.0140 63.24 8.454
4 ... ... ... ... . .. . .. 1.0284 64.14 8.574
6 ... . .. ... ... ... . .. 1.0432 65.06 8.697
8 ... ... . .. ... ... ... 1.0584 66.01 8.824
10 1.0000 62.36 8.337 1.0000 62.36 8.337 1.0741 66.99 8.955
12 0.9861 61.50 8.221 0.9859 61.49 8.219 1.0902 67.99 9.089
14 0.9725 60.65 8.108 0.9722 60.63 8.105 1.1069 69.03 9.228
16 0.9593 59.83 7.998 0.9589 59.80 7.994 1.1240 70.10 9.371
18 0.9465 59.03 7.891 0.9459 58.99 7.886 1.1417 71.20 9.518
20 0.9340 58.25 7.787 0.9333 58.20 7.781 1.1600 72.34 9.671
22 0.9218 57.87 7.736 0.9211 57.44 7.679 1.1789 73.52 9.828
24 0.9100 56.75 7.587 0.9091 56.70 7.579 1.1983 74.73 9.990
26 0.8984 56.03 7.490 0.8974 55.97 7.482 1.2185 75.99 10.159
28 0.8871 55.32 7.396 0.8861 55.26 7.387 1.2393 77.29 10.332
30 0.8762 54.64 7.305 0.8750 54.57 7.295 1.2609 78.64 10.512
32 0.8654 53.97 7.215 0.8642 53.90 7.205 1.2832 80.03 10.698
34 0.8550 53.32 7.128 0.8537 53.24 7.117 1.3063 81.47 10.891
36 0.8448 52.69 7.043 0.8434 52.60 7.031 1.3303 82.96 11.091
38 0.8348 52.06 6.960 0.8333 51.97 6.947 1.3551 84.51 11.297
40 0.8251 51.46 6.879 0.8235 51.36 6.865 1.3810 86.13 11.513
42 0.8155 50.86 6.799 0.8140 50.76 6.786 1.4078 87.80 11.737
44 0.8063 50.28 6.722 0.8046 50.18 6.708 1.4356 89.53 11.969
46 0.7972 49.72 6.646 0.7955 49.61 6.632 1.4646 91.34 12.210
48 0.7883 49.16 6.572 0.7865 49.05 6.557 1.4948 93.22 12.462
50 0.7796 48.62 6.499 0.7778 48.51 6.484 1.5263 95.19 12.725
52 0.7711 48.09 6.429 0.7692 47.97 6.413 1.5591 97.23 12.998
54 0.7628 47.57 6.359 0.7609 47.45 6.344 1.5934 99.37 13.284
56 0.7547 47.07 6.292 0.7527 46.94 6.275 1.6292 101.60 13.583
58 0.7467 46.57 6.225 0.7447 46.44 6.209 1.6667 103.94 13.895
60 0.7389 46.08 6.160 0.7368 45.95 6.143 1.7059 106.39 14.122
62 0.7313 45.61 6.097 0.7292 45.48 6.079 1.7470 108.95 14.565
64 0.7238 45.14 6.034 0.7216 45.00 6.016 1.7901 111.64 14.924
66 0.7165 44.68 5.973 0.7143 44.55 5.955 1.8354 114.46 15.302
68 0.7093 44.23 5.913 0.7071 44.10 5.895 1.8831 117.44 15.699
70 0.7022 43.79 5.854 0.7000 43.66 5.836 1.9333 120.57 16.118
72 0.6953 43.36 5.797 0.6931 43.22 5.778 ... ... ...
74 0.6886 42.94 5.741 0.6863 42.80 5.722 ... ... ...
76 0.6819 42.53 5.685 0.6796 42.38 5.666 ... ... ...
78 0.6754 42.12 5.631 0.6731 41.98 5.612 ... ... ...
80 0.6690 41.72 5.577 0.6667 41.58 5.558 ... ... ...
82 0.6628 41.33 5.526 0.6604 41.19 5.506 ... ... . ..
84 0.6566 40.95 5.474 0.6542 40.80 5.454 ... ... . ..
86 0.6506 40.57 5.424 0.6482 40.42 5.404 ... ... . ..
88 0.6446 40.20 5.374 0.6422 40.05 5.354 ... ... ...
90 0.6388 39.84 5.326 0.6364 39.69 5.306 ... ... . ..
92 0.6331 39.48 5.278 0.6306 39.33 5.257 ... ... ...
94 0.6275 39.13 5.231 0.6250 38.98 5.211 ... ... ...
96 0.6220 38.79 5.186 0.6195 38.63 5.165 ... ... ...
98 0.6166 38.45 5.141 0.6140 38.29 5.119 ... ... . ..
100 0.6112 38.12 5.096 0.6087 37.96 5.075 ... ... ...
439
TABLE 13.6. Commercial Wrought Steel Pipe Data (Schedule Wall Thickness-per ASA B36. Hl-
1950)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - .;¿··
Schedule Wall 'l'hicknes.-Per ASA 836.10-1950 ''~~]
Nominal Outside Thick- Inside ¡ Inside Diameter Functions 1 T:ransverse ~,~

Pipe Diarn- ness Diamete• (In lnches) lnternal Area >5
Size eter
lnches lnches Inches ln:hes 1 F~et 1 d' 1 d' 1 d' 1 d' 1 Sq.ª In. 1 Sq ~t. • ~;
14 14 0.250 13.5 1.125 182.25 2460.4 33215. 448400. 143.14 0.994
16 16 0.250 15.5 1.291 240.25 3723.9 57720. 894660. 188.69 l.311)
18 18 0.250 17.5 1.4583 306.25 5359.4 93789. 1641309. 240.53 l.670
20 20 0.250 19.5 l.625 380.25 7414.9 144590. 2819500. 298.65 2.074
24 24 0.250 23.5 1.958 552.25 12977. 304980. 7167030. 433.74 3.012 , •.
30 30 0.312 29.376 2.448 862.95 25350. 744288. 21864218. 677 .76 4.707
108 1 8.6251
Hl.75
0.250
0.250
18.12510.6771166.02
10.25 0.8542 105.06
1 536.38
1076.9
111038.
4359.3 1 35409.
113141.
151.8510.3601
82.52 0.5731
...
12 12.75 0.250 12.25 1.021 150.06 1838.3 22518. 1 275855. 117 .86 0.8185 ••
14 14 0.312 13.376 l.lll 178.92 2393.2 32012. 428185. '140.52 0.9758 ·;:
16
1!1
Hi
18
0.312
0.312
15.376
17.376
1.281 236.42
1.448 301.92
13635.2
5246.3
55894.
91156.
859442.
1583978.
185.69
237.13
1.290
l.647
;·,,.
..i.>
20 20 0.375 19.250 l.604 370.56 7133.3 1373i7. 2643352. 29!.04 2.021 ;':../
24 24 0.375 23.25 1.937 540.56 112568. 292205. 6793832. 424.56 2.948 •w
30 1 30 0.500 29.00 2.417 841.0 24389. 707281. 20511149. 660.52 4.587 :":!
8
10
1 8.6251 0.2771 8.07110.6726165.14
10.75 0.307 10.136 0.8447 Hl2.74
1 525.75
1041.4
4243.2
10555.
34248.
106987.
1 51.1610.3553
1 80.69 0.5603
1
. 1
~ 1
12 12.75 0.330 12.09 1.0075 146.17 1767.2 21366. 258304. 114.80 0.7972
14 14 0.375 l
13.25 li.1042 175.56 2326.2 30821. 408394. l 137.88 0.9575
-~--+----7---~---'----'-"-----'------+------!---------'---'----!---.:...c:.__ ::;}'
16 16 0.375 15.25 1.2708 232 ..% 3546.6 54084. 824801. 182.65 1.268
18 18 0.438 17.124 293.23 5021.3 l.4270
85984. 1472397. 230.30 1.599
20 20 o.soo 19.oo 1.5833 36i..oo 6859.o 130321. 2476099. 283.53 1.969 ·.;~
24 24 0.562 22.876 1.9063 523.31 H971. 273853. 6264703. 411.00 2.854 .··~~-·-~¿;
30 30 0.625 28.75 2.3958 826.56 23764. 683201. 19642160. 649.18 4.508 .:.
--;~-,-,~-;--,.-,.,..,......,....,~~;--~~;--~-;-~~--;~~~~..;--~~------;-~~~~~..;--~~-,-~~
%
%
1 0.405
o.540
0.0681
o.oss
0.26910.02241
o.364 0.0303
0.07241
0.1325
0.01951
o.o4&2
0.0052421
0.01756
0.00141
0.00639
1' 0.05710.00040 :;:;i
0.104 o.ooon,:\¡;.i
% 0.675 0.091 0.493 0.0411 0.2430 0.1198 0.05905 0.02912 0.191 0.00133 "~~j
!1 1 ~:~~ ~:!~~ 1~::i¡ 1~:~:~~1 ~:~~:91 ~:~~~~ 1 ~:!ii~ 1 ~:~~~~º U~j¡g:~~i~! .·~-~-·
l 1.315
1'
0.133 1.049 0.0874 l.100 1.154 1.210 1.270

11
0.864 0.00600 ;¡¡

114, l.660 0.14() l.380 0.1150 1.904 2.628 3.625 5.005 1.495 0.01040 ';;.í
~112 1 t~ 1g:~~: 1~:~!~ 1g:~~~i1 !:~ii 1 ::~;: 1 ,g~~ 1 i~:~i5 1 ;:~~~ 1g:g~jl~ ~1.4'~ .•
i1h i::~ g:~~~ ;:¿:: g:i~~~ ::~~ i~:~~~ ~~:!~! 2ii:~ ~:;~~ ~:g~i;~ ~)
!Yz
5
1 !:=
5.563
1g:i;~
0.258
!:~~: 1g:5j~~1 ~~:~i 1 11.8.56
1
5.047 0.4206 25.47
!~:~~~ 1 ~~t~!
648.72
1 i~~~:2
3275.
l 1i:~~~ g:g::~~ '.F;~I%~
1
20.00610.1390
6 li.625 0.280 6.065 0.5054 36.78 223.10 1352.8 8206. 28.891 0.2006 ~t;
~! 1~H~5 ltili1~u~;1rm~l!~n: 1tfüT 1i~¡~v 1i~I~~~ !iUi;IH~~i :~!~

~:
i~
1 ~::g
i~:g
1g:~~ 1~:::: 1~:!~~3 1i~~:~ 1!~~::~
g:~:~ i~::~: ~ ::~;: ~r~:~ 1~~¡j: 5
1~~~i~:
i~~~!~:
l 1;~:~;~:
~~~~~:::
1~~tig 1
402.07
um c'~~l!~
2.7921
8
HJ
1 8.62510.40617.81310.6511
10.75 0.500 9.750 IJ.8125 61.04
95.06
476.93
926,86
1 3725.9
9036.4
1 29113.
88110.
147.9410.332' 1 ,p,
74,66 0.51!15 ,c_,q¡
0
i! ~!:~ ~:~~; g:~i: ~:~~~: ~~:~~ ~~~~:¿

5 m:~: ;¡;~:¿: i~~:~~ g:~!~~ ·~\ll
16
18
116.0
18.0
0.656114.68811.22401215.74
0.750 16.500 1.3750 272.25 13168.8
4492.l
146544
74120:
11222982.
683618. 1169.441 i.1766 'ti':j
213.83 1.4849·;~'..~:
20 20.0 0.812º 18.376 l.5313 337.68 6205.2 114028. 2095342. 265.21 l.8417 ,,,~·'.
24 24.0 0.%8 22.064 1.8387 486.82 10741. 23&9'!4. 5229036. 382.35 2.6552;.~•.;
l/s
1/.i 0.405
0.540 1 o.0951
O.H9 0.302 10.01791
o.21s 0.0252 o.04621
0.0912 0.009941
0.0275 0.0011341
IJ.008317 0.0004591
ll.002513 o.0361 o.¡002s.·~·.·.~·.·.
0.072 O.
% il.675 0.126 0.423 0.0353 0.1789 0.0757 0.03200 0.01354 0.141 o.
1
~ 1 ~:~:g
1.315
10.1471 0.54610.04551 0.29811
~:~~: ~:~:; ~:~~~ ~:~~~
~:!~~~
0.8765
~:~~~~61
1 0.8387 ~:~i:~21 0.719
0.8027
~:!~~1~:003oq~·.·.·.i._,_·
0.00499'}'°
1% 1.660 0.191 1.278 0.1065 l.633 2.087 2.6667 3.409 1.283 0.0089í%'¡i::
' "\:~..,,~.:
(confinued on rha next poge} -~'<i;<.j
440
TABLE 13.8. (Contínued)
Schedule Wall Thickness-Per ASA 836.10-1950
Nominal Outside Thick- Inside 1 Inside Diameter Functions 1 Transverse

Pipe Diam- neas Diameter (In Inches) Internal Area
ln~es F~et
Si:e eter 1 1 1 A
lnches lnches lnches 1 d' 1 d' 1 d' 1 d' Sq.ªIn. Sq. Ft.
1% 11.900 0.200 11.500 1 0.1250 2.250 1 3.3751 5.062 1 7 .594 1.76710.01225

2 2.375 0.218 1.939 0.1616 3.760 7.290 14.136 27.41 2.953 0.02050
2% 2.875 0.276 2.323 0.1936 5.396 12.536 29.117 67 .64 4.238 0.02942
3 3.5 0.300 2.900 0.2417 8.410 24.389 70.728 205.1 6.605 0.04587
.. 3% 1 4.0 1 0.3181 3.3641 0.2803111.32 1 38.0691 128.14 1 430.8 1 8.88810.06170
~~1-~4-o-~7--4~·~5~+-oº-·3=3~7-+--3~·~8~26=-i-~º-·3~1~88-,---;-1~4~.64~-¡--;-:5~6~.006~-;---=21~4~.3~3~-¡--~=81~9~;8~~~1~1~.4~97~~º~·º~7~98::-6
r 5
6
5.563 o.375
6.625 0.432
4.813 o.4011 23.16
5.761 0.4801 33.19
111.49
191.20
536.38
1101.6
2583.
6346.
18.194 0.1263
26.067 0.1810
-3 1 ~ l 1t;~5 1 ~:~~ 1 ~:~!: 1 ~:;~~ 1 ;t!~ 1 ~!::i 1 ~::: 1 ~;~: 1 ~::31 ~:~~~~
l,clS ,-__1412-o-~7--12~·~7-5-i-~º-·68~7-+-~ll~·~37~6-i-~º-·94-=80~7l2~9~.4~1,--i-=-14~772~.2,---;--1~67747~.~--;-1~90~5~2~3-.~~71~071~.64~~º-·7~0~58~
14.0 0.750 12.500 1.0417 156.25 1953.1 24414. 305176. 122.72 0.8522
16
18 116.0
18.0 1 0.843114.31411.19281204 89
o.937 16.126 t.3438 260:05 12932.8
4193.5 141980.
67626. 1600904.
1090518. 160.92
204.24 11.1175
t.4183
20 20.0 1.031 17.938 1.4948 321.77 5771.9 103536. 1857248. 252.72 1.7550
24 24.0 1.218 21.564 1.7970 465.01 10027. 216234. 4662798. 365.22 2.5362
8 1 8.625 0.5931 7.439 0.6199 55.34 411.66 3062. 22781. 43.46 10.3018
10 10.75 0.718 9.314 0.7762 86.75 807 .99 7526. 69357. 68.13 0.4732
12 12.75 0.843 11.064 0.9220 122.41 1354.4 14985. 165791. 96.14 0.6677
14 14.0 0.937 12.126 1.0105 147.04 1783.0 21621. 262173. 115.49 0.8020
16
18
20
116 o
18:0
20.0
11 031 113.938 'i
1:156 15.688
1.281 17.438
1.16151194.27 12707.7
t.3057 246.11
1.4532 304.0!1
3861.o
5302.6
137740.
60572.
92467.
1526020.
950250.
1612438.
193.30
238.83
¡
1152.58 / 1.0596
1.3423
1.6585
24 24.0 1.531 20.938 1.7448 438.40 9179.2 192195. 4024179. 344.32 2.3911
4 1 4.50 1 0.4381 3.6241 0.302 113.133 47.5951 172.49 625.1 10.31510.07163
5 5.563 0.500 4.563 0.3802 20.82 . 95.006 433.5 1978. 16.35 0.1136
6 6.625 0.562 5.501 0.4584 30.26 166.47 915.7 5037. 23.77 0.1650
~.. 8
10 1 10.75
8.6251 0.7181
0.843 7.1891 0.5991 151.68
9.064 0.7553 82.16 1 744.66
371.54 ·1 6750.
2671. 119202.
61179. 140.59
64.53 ,0.2819
0.4481
'3 12 12.75 1.000 10.750 0.8959 115.56 1242.3 13355. 143563. 90.76 0.6303
~ 1 _ _14~~,__14_._o~_,__1_.0_9_3~_11_._8_14_,__o_.9_8_45__._13_9_.5_7__,,__1648~_·9~_,_1_9480~-·~__,_23~0_13_7_.~~~1_09~.6_2~-º-·1_6_12~
Ul 16 16.0 11.218113.56411.1303 183.98 12495.5 33849. 1459133. 1144.50 11.0035
18 18.0 1.375 15.250 . 1.2708 232.56 3546.6 54086. 824804. 182.66 1.2684
20 20.0 1.500 17.000 1.4166 289.00 4913.0 83521. 1419857. 226.98 1.5762
24 24.0 1.812 1 20.376 1.6980 415.18 8459.7 172375. 3512313. 326.08 2.2645
8 8.625 0.8121 7.001 0.5834 49.01 343.15 2402. 16819. 38.50 ,0.2673
10 10.75 1.000 8.750 0.7292 76.56 669.92 5862. 51291. 60.13 0.4176
12 12.75 1.125 10.500 0.8750 110.25 1157.6 12155. 127628. 86.59 0.6013
14 14.0 1.250 11.500 0.9583 132.25 1520.9 17490. 201136. 103.87 0.7213
16
18 116.0
18.0 11.438113.12411.09371172.24
1.562 14.876 1.2396 221.30 12260.5
3292.0 129666.
48972. 1389340.
728502. 1135.28
173.80 10.9394
1.2070
20 20.0 1.750 16.5 1.3750 272.25 4492.1 74120. 1222981. 213.82 1.4849
24 24.0 2.062 19.876 1.6563 395.06 7852.1 156069. 3102022. 310.28. 2.1547
1h 1 0.840 1 0.1871 0.4661 0.0388 0.21721 0.1012 0.047161 0.021971 0.170610.00118
3,4 1.050 0.218 0.614 0.0512 0.3770 0.2315 0.1421 0.08726 0.2961 0.00206
1 1.315 0.250 0.815 0.0679 0.6642 0.5413 0.4412. 0.3596 0.5217 0.00362
11,4 1.660 0.250 1.160 0.0966 1.346 1.561 1.811 2.100 . 1.057 0.00734
1112
2 11.900
2.375 1 0.281
0.343 11.3381
1.689 0.111511.790
0.1407 2.853 1 2.3951
4.818 3.205
8.138 1 4.288
13.74 11.40610.00976
2.241 0.01556
2% 2.875 0.375 2.125 0.1771 4.516 9.596 20.39 43.33 3.546 0.02463
! 3 3.50 0.438 2.624 0.2187 6.885 18.067
1---o-~-i--=-~-i-~=o-+--o--=-i-~~~-:-:~~¡--~-==-+~'77'~~-¡--~~~~-;--=--=,,-,~~=-
47.41 124.4 5.408 0.03755
-3'O 4
5
1 4.50 1 0.531 1 3.4381 0.2865111.82
5.563 0.625 4.313 0.3594 18.60
40.6371
80.230
139.7
346.0
1 480.3
1492.
1 9.283 10.06447
14.61 0.1015
j1-~:o----il--0:~::~~7:~-~~:7~~8-r-:~:7!~~:-+-7~~:~~6~==-r~:~:~:~o--rl-!~~7::~~~2 -+l-=-2~~!~::-º~-rl-l~~~76~!~:~~rl7!6~1:~~.,--7,~~:7~=o-~-
10
12
14
10.75
12.75
14.0
1 1.125
1.312
1.406
8.500
10.126
11.188
1 1
0.7083 72.25 614.12
0.8438 102.54 . 1038.3
0.9323 125.17 1400.4
5220.
10514.
15668.
44371.
106461.
175292.
56.75
80.53
98.31
0.3941
0.5592
0.6827
16
18 116.0
18.0 1.593112.81411.06781164.20
1.781 14.438 1.2032 208.45 12104.0
3009.7 126961.
43454. 1345482.
627387. 1128.96
163.72 ,0.8956
1.1369
20 20.0 1.968 16.064 1.3387 258.05 4145.3 66590. 1069715. 202.67 1.4074
24 24.0 2.343 19.314 1.6095 373.03 7204.7 139152. 2687582. 292.98 2.0346
441
TABLE 13.8. (Continued)
Nom- Outside ldentification Wall lnside Are a Transverse Moment Weight Weight Externa! Section
in al Diam. Steel Stain- Thick- Diam- of Interna! Area of Pipe Water Surface Modulus
Pipe less ness e ter Metal Inenia
(a) (A)
Size lron Sched. Steel (t) (d) (/) Pounds Pounds Sq. Ft.
Pipe No. Sched. Square Square Square per per foot per foot
In ches loches Size No. Inches loches Inches loches Feet lnches4 foot of pipe of pipe
SS .156 13.688 6.78 147.lS 1.0219 162.6 23.07 63.77 3.665 23.2
lOS .188 13.624 8.16 145.78 1.0124 194.6 27.73 63.17 3.66S 27.8
10 .2SO 13.SOO 10.80 143.14 .9940 2SS.3 36.71 62.03 3.66S 36.6
20 .312 13.376 13.42 140.S2 .9758 314.4 45.61 60.89 3.665 4S.O
STO 30 .37S 13.250 16.0S 137.88 .9S7S 372.8 S4.S7 S9.7S 3.665 S3.2
40 .438 13.124 18.66 13S.28 .9394 429.1 63.44 58.64 3.665 61.3
14 14.00 XS .soo 13.000 21.21 132. 73 .9217 483.8 72.09 S7.46 3.665 69.1
60 .S94 12.812 24.98 128.96 .8956 562.3 85.0S SS.86 3.665 80.3
80 .7SO 12.SOO 31.22 122.72 .8522 678.3 106.13 S3.18 3.665 98.2
100 .938 12.124 38.45 11S.49 .8020 824.4 130.8S S0.04 3.665 117.8
120 1.094 11.812 44.32 109.62 .7612 929.6 150.79 47.45 3.66S 132.8
140 1.250 11.SOO 50.07 103.87 .7213 1027.0 170.28 45.01 3.665 146.8
160 1.406 11.188 SS.63 98.31 .6827 1117.0 189.11 42.60 3.665 IS9.6
SS .16S IS.670 8.21 192.8S 1.3393 2S7.3 27.90 83.S7 4.189 32.2
lOS .188 lS.624 9.34 191.72 1.3314 291.9 31.7S 83.08 4.189 36.S
10 .2SO lS.SOO 12.37 188.69 1.3103 383.7 42.0S 81.74 4.189 48.0
20 .312 IS.376 IS.38 185.69 l.289S 473.2 S2.27 80.SO 4.189 S9.2
STO 30 .37S IS.2SO 18.41 182.65 1.2684 562.1 62.S8 79.12 4.189 70.3
16 16.00 xs 40 .soo 13.000 24.3S 176.72 1.2272 731.9 82.77 76.58 4.189 91.S
60 .656 14.688 31.62 169.44 1.1766 932.4 107.50 73.42 4.189 116.6
80 .844 14.312 40.14 160.92 l.117S 11SS.8 136.61 69.73 4.189 144.S
100 1.031 13.938 48.48 1S2.58 l.OS96 1364.S 164.82 66.12 4.189 170.S
120 1.219 13.S62 S6.56 144.SO l.003S lSSS.8 192.43 62.62 4.189 194.S
140 1.438 13.124 6S.78 13S.28 .9394 1760.3 223.64 S8.64 4.189 220.0
160 l.S94 12.812 72.10 128.96 .8956 1893.S 245.2S SS.83 4.189 236.7
SS .16S 17.670 9.2S 245.22 1.7029 367.6 31.43 106.26 4.712 40.8
lOS .188 17.624 10.S2 243.9S 1.6941 417.3 3S.76 105.71 4.712 46.4
10 .250 17.SOO 13.94 240.S3 1.6703 549.1 47.39 104.21 4.712 61.1
20 .312 17.376 17.34 237.13 1.6467 678.2 58.94 102.77 4.712 7S.S
STO .37S 17.250 20.76 233.71 1.6230 806.7 70.S9 101.18 4.712 89.6
30 .438 17.124 24.17 230.30 l.S990 930.3 82.lS 99.84 4.712 103.4
18 . 18.00 xs .soo 17.000 27.49 226.98 l.S763 1053.2 93.45 98.27 4.712 117.0
40 .S62 16.876 30.79 223.68 l.SS33 1171.S 104.67 96.93 4.712 130.1
60 .7SO 16.SOO 40.64 213.83 1.4849 1Sl4.7 138.17 92.S7 4.712 168.3
80 .938 16.124 S0.23 204.24 1.4183 1833.0 170.92 88.50 4.712 203.8
100 1.156 IS.688 61.17 193.30 1.3423 2180.0 207.96 83.76 4.712 242.3
120 l.37S 1S.2SO 71.81 182.66 1.2684 2498.1 244.14 79.07 4.712 277.6
140 1.562 14.876 80.66 173.80 1.2070 2749.0 274.22 7S.32 4.712 30S.S
160 1.781 14.438 90.7S 163. 72 1.1369 3020.0 308.SO 70.88 4.712 33S.6
SS .188 19.624 11.70 302.46 2.1004 S74.2 39.78 131.06 S.236 S7.4
lOS .218 19.S64 13.SS 300.61 2.0876 662.8 46.06 130.27 S.236 66.3
10 .250 19.SOO lS.Sl 298.65 2.0740 765.4 S2.73 129.42 S.236 7S.6
STO 20 .37S 19.2SO 23.12 290.04 2.0142 1113.0 78.60 12S.67 S.236 111.3
xs 30 .soo 19.000 30.63 283.S3 1.9690 1457.0 104.13 122.87 S.236 145.7
40 .S94 18.812 36.IS 278.00 l.930S 1703.0 123.11 120.46 S.236 170.4
20 20.00 60 .812 18.376 48.9S 265.21 1.8417 2257.0 166.40 114.92 S.236 22S.7
80 1.031 17.938 61.44 2S2.72 l.7S50 2772.0 208.87 109.Sl S.236 277.1
100 1.281 17.438 7S.33 238.83 1.6585 331S.2 256.10 103.39 S.236 331.S
120 l.SOO 17.000 87.18 226.98 l.S762 3754.0 296.37 98.3S S.236 37S.S
140 l.7SO 16.SOO 100.33 213.82 1.4849 4216.0 341.09 92.66 S.236 421.7
160 1.969 16.062 111.49 202.67 1;4074 4585.S 379.17 87.74 S.236 458.5
SS .188 21.624 12.88 367.2S 2.SS03 766.2 43.80 1S9.14 S.760 69,7
!OS .218 21.S64 14.92 365.21 2.S362 884.8 S0.71 158.26 S.760 80.4
10 .250 21.SOO 17.08 363.0S 2.S212 1010.3 S8.07 1S7.32 S.760 91.8
STO 20 .37S 21.2SO 2S.48 354.66 2.4629 1489.7 86.61 153.68 S.760 13S.4
xs 30 .500 21.000 33.77 346.36 2.4053 1952.S 114.81 150.09 S.760 117.S
22 22.00 60 .875 20.250 S8.07 322.06 2.2365 3244.9 197.41 139.56 S.760 29S.O
80 1.125 19.7S 73.78 306.3S 2.1275 4030.4 250.81 132.76 S.760 366.4
100 1.375 19.2S 89.09 291.04 2.0211 47S8.S 302.88 126.12 S.760 432.6
120 1.625 18.7S 104.02 276.12 l.9!7S 5432.0 3S3.61 119.65 S.760 493.8
140 l.87S 18.2S 118.SS 261.S9 1.8166 6053.7 403.00 113.36 S.760 SS0.3
160 2.125 17.7S 132.68 247.45 l. 7184 6626.4 451.06 107.23 5.760 602.4
ldentllcdon, w•ll thlckneu and weight1 are extractad from ANSI 836.10 and 836.19. The notatlons Tranavarte lntemal arH values listed in "square feet" also
STO, XS, and XXS indica1e Standard, Extra Strong, and Oouble Extra Strong pipe respectlvely. represent volume in cubic teet per foot of pipe length.
442
TABLE 13.8. (Continued)
Nom- Outside Identification Wall lnside Area Transverse Moment Weight Weight Externa! Section
inal Diam. Steel Stain- Thick- Diam- of Interna! Area of Pipe Water Surface Modulus
Pipe less ness e ter Metal lnertia
Size lron Sched. Steel (t) (d) (a) (A) Pounds Pounds Sq. Ft.
(/)
loches loches
Pipe
Size
No. Sched.
No. Inches loches
Square
loches
Square
Inches
Square
Feet
per
loches• foot
per foot per foot
of pipe of pipe
(20.~J
... . .. 5S .218 23.564 16.29 436.10 3.0285 1151.6 55.37 188.98 6.283 96.0
... 10 lOS .250 23.500 18.65 433.74 3.0121 1315.4 63.41 187.95 6.283 109.6
STD 20 ... .375 23.250 27.83 424.56 2.9483 1942.0 94.62 183.95 6.283 161.9
xs ... ... .500 23.000 36.91 415.48 2.8853 2549.5 125.49 179.87 6.28S 212.5
... 30 ... .562 22.876 41.39 411.00 2.8542 2843.0 140.68 178.09 6.283 237.0
24 24.00 ... 40 . .. .688 22.624 50.31 402.07 2.7921 3421.3 171.29 174.23 6.283 285.l
... 60 ... .969 22.062 70.04 382.35 2.6552 4652.8 238.35 165.52 6.283 387.7
... 80 ... 1.219 21.562 87.17 365.22 2.5362 5672.0 296.58 158.26 6.283 472.8
... 100 ... 1.531 20.938 108.07 344.32 2.3911 6849.9 367.39 149.06 6.283 570.8
... 120 ... 1.812 20.376 126.31 326.08 2.2645 7825.0 429.39 141.17 6.283 652.1
... 140 ... 2.062 19.876 142.11 310.28 2.1547 8625.0 483.12 134.45 6.283 718.9
... 160 ... 2.344 19.312 159.41 292.98 2.0346 9455.9 542.13 126.84 6.283 787.9
... 10 ... .312 25.376 25.18 505.75 3.5122 2077.2 85.60 219.16 6.806 159.8
26 26.00 STD ... ... .375 25.250 30.19 500.74 3.4774 2478.4 102.63 216.99 6.806 190.6
xs 20 ... .500 25.000 40.06 490.87 3.4088 3257.0 136.17 212.71 6.806 250.5
... 10 ... .312 27.376 27.14 588.61 4.0876 2601.0 92.26 255.07 7.330 185.8
STD ... ... .375 27.250 32.54 583.21 4.0501 3105.1 110.64 252.73 7.330 221.8
28 28.00 XS 20 ... .500 27.000 43.20 572.56 3.9761 4084.8 146.85 248.11 7.330 291.8
... 30 . .. .625 26.750 53.75 562.00 3.9028 5037.7 182.73 243.53 7.330 359.8
... ... 5S .250 29.500 23.37 683.49 4.7465 2585.2 79.43 296.18 7.854 172.3
... 10 JOS .312 29.376 29.10 677. 76 4.7067 3206.3 98.93 293.70 7.854 213.8
30 30.00 STD ... . .. .375 29.250 34.90 671.96 4.6664 3829.4 118.65 291.18 7.854 255.3
XS 20 ... .500 29.000 46.34 660.52 4.5869 5042.2 157.53 286.22 7.854 336.1
... 30 ... .625 28.750 57.68 649.18 4.5082 6224.0 196.08 281.31 7.854 414.9
... 10 ... .312 31.376 31.06 773.19 5.3694 3898.9 105.59 335.05 8.378 243.7
STD ... . .. .375 31.250 37.26 766.99 5.3263 4658.5 126.66 332.36 8.378 291.2
32 32.00 xs 20 ... .500 31.000 49.48 754.77 5.2414 6138.6 168.21 327.06 8.378 383.7
... 30 ... .625 30.750 61.60 742.64 5.1572 7583.4 209.43 321.81 8.378 474.0
... 40 . .. .688 30.624 67.68 736.57 5.1151 8298.3 230.08 319.18 8.378 518.6
... 10 ... .344 33.312 36.37 871.55 6.0524 5150.5 123.65 377.67 8.901 303.0
STD ... ... .375 33.250 39.61 868.31 6.0299 5599.3 134.67 376.27 8.901 329.4
34 34.00 XS 20 ... .500 33.000 52.62 855.30 5.9396 7383.5 178.89 370.63 8.901 434.3
... 30 ... .625 32.750 65.53 842.39 5.8499 9127.6 222.78 365.03 8.901 536.9
... 40 ... .688 32.624 72.00 835.92 5.8050 9991.6 244.77 362.23 8.901 587.7
... 10 ... .312 35.376 34.98 982.90 6.8257 5569.5 118.92 425.92 9.425 309.4
STD ... ... .375 35.250 41.97 975.91 6.7771 6658.9 142.68 422.89 9.425 369.9
36 36.00 xs 20 ... .500 35.000 55.76 962.11 6.6813 8786.2 189.57 416.91 9.425 488.1
... 30 ... .625 34.750 69.46 948.42 6.5862 10868.4 236.13 417.22 9.425 603.8
... 40 ... .750 34.500 83.06 934.82 6.4918 12906:1 282.35 405.09 9.425 717.0
ldentlftc•tlOill, wall thlckneH and wefghts are extracted from ANSI 836.10 and 836. 19. The notations TrmnevePM lntemal are• values llsted in "aquare feet" also
STO, XS, and XXS indicate Standard, Extra Strong, and Double Extra Strong pipe respectively. represent volume in cubic leet per foot of pipe length.
443
T ABlE 13.9. Flow of Water through Schedule 40 Steel Pipe
Prenure Drop per 100 feet and Velocity in Schedule 40 Pipe for Water at 60 F.
Discharge
Veloe- Pre... 1 Velo<- P.c... 1 Veloc- P,eu. 1Vcloc- p,.... 1 Veloc- p,..., 1Velo~ p,..u, 1 Veloc- p,.... 1 Voloe- p,.,.,
ity Drop ity Drop ity Drop ity Drop ity Drop ity Drop ity Drop ity Di:op
Ga!lons 1 Cubic Ft. Feet Lbs. Feet Lbs. Feet' Lbs. Fe~t Lbs. Feet Lbs. Feet Lbs. Feet Lbs. Fe1:t Lbs.
- ~1- - - - ~ - - - - - - - - -
pc11 per
Minute Seo>nd &cond sq. In. &X:ond SQ. In. S..:cond sq, In. &icond Sq. In. Second Sq. in., Second Sq. In. Seoond Sc]. In. Seconcl Sq. In.
.. ..-..
.2
.3
0.000446
0.000891
l.13
l."9
2.26
V&
.... 11.· %' 'h' .,..
4.12
ó.98
O.ól6
0.924
l.23
0.359
0.903
1.61
0.504
0.672
0.159
0.345
0.317
0.422
0.061
0.08&
1
..
.5
.6
1
o.ooau.
0.00134
0.00178
2.82
3.39
4.52
5.M
10.5
14.7
25.0
l.54
l.B:í
2.46
l.39
3.l"J
5.44
0.840
1.01
1.34
0.539
6.751
l.25
0.528
0.633
0.844
0.167 O.JO!
0.140 0.Jól
0.400 0.481
0.033
0.041
0.102 ¡• !•¡(¡'
' 11/z"
l
0.00213
º·""* ll .29
37.2
134.~
3.08
6.16
8.28
30.1
i.óB
3.36
5 .04
1.85
6.58
1.06
2.11
0.600
2.10
0.602
1.20
0.155 0.371
0.526 ü.74) º·""'
0.164 0.42.9 0.044
3
•
5
0.00668
0.00891
0.01U4 1'
9.25
12.JJ
64.!
111.2 ó.72
8.40
13.9
lJ.9
36.7
3 .17
4.n
5 .28
4.33
7.42
11.2
l.81
2.41
3.01
l.O'l
l.83
i.114
1.49
2.75 i.8ó
0.336 0.644
O.SllS 0.858
0.835 J.073
º·""" 0.47J
0.150 O.DJO
0.213 o 788
0.043
0.071
0.104
,.
10
•• O.OJ.337
().01782
0.02228
o.oun
0.574
0.765
0.956
lAJ
0.044
0.013
0.100 0.670
0.214 1.01
2%"
0.046
0.094
10.08
13.44
3'
51.9
91.l
ó.33
8.45
10.56
15.S
27.7
42.4
3.ól
4.81
6.02
9.0J
....
3.84
9.99
11.6
2.23
2.97
J .71
5.57
1.17
1.99
2.99
6.36
1.29
1.72
2.15
3.22
0.309
0.518
0.7i4
!.63
0.946
l .26
1.58
2.37
0.145
0.241
O.JM
0.755
31iz,,
20
25
0.""456
0.05570
l.91
2.39
0.375 l .34
. 0.561 1.ó8
0.158 0.868 0.056
0.234 1.0'l 0.083 0.812 0.041
12.03
4'
37.8
9.28
7.43 10.9
16.1
4.29
:í.37
l.78
4.12
J.16
3.94
'·""
1.')3
30 o.06684 2.87 0.786 2.01 0.327 1.30 0.114 0.974 0.056 ll.14 13.8 6.44 5.92 4.73 1..72
...
3S
45
0.01798
().08911
0.1003
3.35
3 .83
4.30
1.05
!.3S
1.67
2.35
l.ó8
3.02
0.436 1.52
0.556 l.74
0.661! l .95
0.151
0.192
0.239
1.14
1.30
1.46
0.704 0.882
0.095 l.01
0.117 1.13
0.041 12.99
0.051 14.85
0.064
32.2
41.5
7.51
8.59
9.67
7.90
10.24
ll.80
5. 52
ó.JO
7.09
3.64
4.65
5.85
.
Sil
""
70
90
0.1114
0.1337
0.1560
0.1782
0.:2005
4.78
5.74
6.70
7.65
S.óO
2.03
2.87
3.8'
4.97
6.20
3.35
4.02
4.69
5 .36
6.03
{).839 2.17
l.18 2.óO
1.59 J.04
2.03 J.47
1.53 3.91
o.wi
0.406
o.54o
l.62
l.95
2.27
0.687 2.óO
0.861 2.92
0.142
0.204
0.261
0.334
0.416
! .'!ó
l.51
l.76
2.02
2.27
0.076
0.107
0.143 1.12
0.JSQ 1.28
O.U4 l.44
5'
0.047
º·°""
0.074
I0.74
12.89
6'
15.66
22.2
7 .88
9.47
11.05
12.62
14.20
1Ul
13.71
17.59
l2.0
100 0.1128 9.56 7.59 6.70 3.0'l 4.34 LOS J .25 0.509 2.52 0.212 1.60 0.090 l.ll 0.036 15 .78 26.9
125 0.1785 ll .97 11.76 8.38 4.71 5.43 1.61 4.06 0.769 J .15 0.415 2 .O! 0.135 1.39 0.055 19.72 41.4
""" .....
!50 0.3342 14.36 16.70 l0.05 6.69 6.51 2.24 4.87 !.08 3.78 0.580 2.4! (Ll90 1.67 0.077
175 0.3899 16.75 ll.3 ll .73 8.97 7.60 3.00 4.41 0.774 2.81 0.253 1.94 0.1021
200 0.4456 19.14 l8.S 13 .42 ll.68 8.ó8 3.87 6.49 l.85 5.04 0.985 3 .21 0.323 2.22 0.130 B'
125 0.5013 15 .09 14.63 9.77 4.83 7 .JO 1.32 5 .67 1.23 J .61 0.401 2.50 0.162 1.44 0.043
250
175
300
325
......
0.557
0.6127
o.nu
" 10.85
il .94
13.00
14.12
5.93 8.12
7.14 8.CJJ
8.36 9.74
9.89 10.5J
2.84
3.40
4.02
4.0'l
ó.30
6.93
7.56
8.19
!.46
1.79
2.U
l.47
4.01
4.41
4.81
5 .21
(.l.495 2.78
0.583 J .0)
0.683 3 .JJ
0.797 J.i:il
0.195
0.234
0.275
0.320
1.60
1.76
1.92
2.08
0.051
0.061
c.on
0.083
360 0.7198 l 1 .Jó 5.41 8.82 2.84 5.62 0.919 3.89 0.367 2.24 0.095
....
··-
375 0.8355 12.17 &.18 9.45 J.25 ó.02 !.05 4.16 0.416 2.40 O.i08
0.8911 12.98 7oOJ 10.08 3.68 6.42 l.19 4.44 0.471 2.56 0.111
425 13.80 7.89 I0.71 4.12 ó.82 1.33 4.72 0.519 2 .73 0.136
450
475
500
550
1.003
1.059
l.114
1.225
I.93
2.03
2.24
10'
.....
0.05'9
0.071
14.61 8.80 li.H
11.97
12.60
IJ.85
4.60
5.12
5.65
6.79
7 .62
8.02
8.82
7.22 1.48
i.64
L81
l.17
5.00
5.27
5. 55
ó.11
0.590 2.89
0,653 3.04
o.no J .21
0.861 J. 53
0.151
0.166
0.182
0.119
600 1.337 2.44 Oo083 15.12 8.04 c:i.63 2.55 6.66 1.01 3.85 0.258
650 1.448 2.64 0.097 12' I0.43 2.98 7.22 1.18 4.17 (1.301
700 1.560 2.85 O.Ul 2.01 0.047 11.23 J.43 7.78 1.35 4.49 0.343
750 l.671 3.05 0.121. 2 .15 14" 12.0J J.92 4.81
...
0.054 8.33 l.55 0.392
800 1.782 3.25 0.143 2.29 0.061 12.SJ 4.43 8.88 1.75 5.13 0.443
850 1.894 J.4ó 0.160 2.44 0.068 2.02 0.042 IJ.64 5.00 9.44 l.% 5.45 0.497
l.005 J.66 0.179 2.58 0.075 2.13 0.047 14.44 5.56 9.99 2.18 5.77 0.554
950 2.117 Uló 0.198 2.72 0.083 2.25' 0.052 15 .24 6.21 10. 55 2.41 ó.O'l 0.613
¡ 000 2.128 4.07 0.1!8 2.87 0.091 2.37 0.057 16' ló.04 6.84 ! ¡. IO 2.68 ó.41 0.67&
1100 2.451 4.48 0.260 J.15 o.uo 2.61 0.068 17 .65 8.23 12.22 3.22 7.05 0.807
1200 2.674 4.88 0.306 J.44 0.116 2.85 o.oso 2.18 Oo042 13 .33 3.8! 7.70 o.948
1300 l.8% 5 .29 0.355 3 .73 0.150 J.08 0.093 2.36 0.048 14.41 4.45 s.:n 1.11
1400 3.!19 S.70 0.409 4.01 C.171 J.32 0.107 2.54 0.055 15.H 5.13 8.98 1.28
1500 3.342 6.IO 0.466 4.30 0.195 3.56 0.122 2.72 0.063 18' ló.66 5.85 9.62 1.46
1600 3.565 6.51 0.527 4.59 0.219 J.N 0.138 2.'lO 0.071 17.77 6.61 I0.26 1.65
1800 4.010 7.32 0.663 5.16 0.276 4.27 0.17l 3.27 0.088 2.58 0.050 ¡q_qq 8.37 11.54 2.08
2000 4.456 8.14 0.808 5.73 0.339 4.74 0.209 3.óJ 0.107 2.87 0.060 22.21 !0.3 12.82 l.55
20·
....
2500 5.570 10.17 J..24 7.17 o.515 5.93 0.321 4.54 0.163 J.59 0.091 \ó.OJ 3.94
3000 6.684 12.20 1.76 8.60 0.731 7.11 0.451 5.45 0.232 4.JO 0.129 3.46 0.075 24" 19.24 5.59
3500 'J.i98 14.24 2.38 10.03 0.982 8.JO 0.607 ó.35 0.312 5 .02 0.173 4.04 0.101 22.44 7.56
8.912 16.27 3.08 11.47 1.27 9.48 0.787 7.26 0.401 5 .74 0.122 4.62 0.129 J .19 0.052 2Ló5 9.80
HOO 10.03 18.Ji 3.87 12.90 1.60 10.67 0.990 8.17 0.503 6.46 0.280 5.20 0.162 3.59 0.065 28.87 !2.2
5 000 H.14 20.H 4.71 14.33 1.95 11.85 1.21 9.08 0.617 7.17 0.34-0 5.77 0.19'9 J.99 0.079
....
6000
7000
• 000
13.37
15.60
17.82
20.05
24.41
28.49
6.74 i"/ .20
9.U 20.07
22.9J
25 .79
2.'71
3.74
4.84
6.09
14.23
ló.60
18.%
21.34
1.7! 10.8()
2.3! 12.71
2.99 14.52
3.76 16.34 ,, .
0.877 8.61
t.i8 10.04
1.5! 11.47
12.91
0.483 6.93
0.652 8.08
0.839 ().23
1.05 10.39
0.281)
0.376
0.488
0.608
4.7CJ
5.59
6.)8
7. i8
O.lU
0.150
0.192
0.242 "
...
. ..
10000
11000
14000
16000
22.28
26.74
31.1~
35.65
...
...
28.66
34.40
7.46 23 .71
10.7 28.45
33 .19
4.6i 18.15
6.59 21.79
8.89 25.42
29.05
2.34
3.33
4.49
S,83
14.34
17.21
20.08
22.95'
l.28
1.83
2.45
3.18
11.54
13 .85
16.16
18.47
....
0.739 7.98
9.58
l.4J 11.17
1.85 12.77
0.294
0.416
0.5&2
0.723
18000 40.10 32.63 7.31 25 .82 4.03 20.77 2.32 14.Jó 0.907
20000 <4.50 J6.Jl '9.03 28.69 4.93 21.08 2.86 15.% 1.12
For pipe lengths other than 100 feet, the pressure drop the 1s proport1onal to Veloc1ty 1s a funct1on of the cross sect1onal
length. Thus, for 50 feet of pipe, the pressure drop is approximately one-half flow area; thus, it is constant for a given
the value given in the table ... for 300 feet, three times the given value, etc. flow rate and is independent of pipe length.
444
13.12 Summary of Formulas
To eliminate needless duplication, formulas ha~ e Head loss and pressure drop
been written in terms of either specific volume V In stralght pipe:
or weight density p, but not in terms .of both, since Pressure loss due to flow is the same in sloping, a
one is the reciproca! of the other. vertical; or horizontal pipe. However, the dif-
ference in pressure due to the difference in head
v =-'- P
p=
1
=
V must be considered in pressure drop calculations:
see page 1-5.
These equations may be substituted in any of the
formulas shown in this paper whenever necessary.
JLv'
0.1863 -d-
JLq' JU!'
e Bernoulli's theorem: fqualJen .1· I hL = Ó200 --¡¡;- = 0.03 1 1 -r
z+ 144 P +
p
-7!._
2g
= H
hL = 0.015 24
JLB'
"""'d"' = 0.000 483 --d'--
JLW•V'•
z1 + -P-,
144?, v', z.+ 144?,+~+h, ~ JLV•
- + 2¡ = P¡ 2g • L:i.P = 0.001 294 d = 0.000000 3 5 9 +
e Meon velocity of flow In pipe: L:i.P = 4J.5 f~"! = o.ooo 216 JL;Q'
(Continuity Equation) fq ..110• 3·2
fLpB' jLWSV
V =
q L:i.P = o.ooo 1058 ----¡¡o- = o.ooo 003 36----¡¡¡-
A
V
B
= 0.286 {j2 = 183.3 (j2
wV
0.0509 T
wv L:i.P = o.ooo ooo 007 26
JL T(q' .)•S,
d'P'
¡L(q' )'S'
o.oo 1 44 P'd'
q',T
= 0.003 89
q',S,
-¡;¡¡:- L:i.P = o.ooo ooo 019 ;9 - rPp '
V 2.40
wv
-ª- 3.06 (12
wv For simplltl.O co111preui•le lflild
· formulo, ·~ poge 3·22.
\/ 0.233 q;i.,
'S
• Head loss .and pressure drop
with laminar flow in straight pipe:
Far laminar flow conditions (R, < 2000). the friction
e Reynolds number factor is a direct mathematical function of the
of flow in pipe: fquatien .1-3
Reynolds number only, and can be expressed by
R, = Dvp = Dvp, the formula: J = 64/ R,. Substituting this value of
µ., 31.2µ, fin the Darcy formula, it can be rewritten:
µLv fquaflon .1-•
R, = 22 700<J..e.=1 7 NP = 50.6 Q!'_ h L = 0.09Ó2 · "Ci'P
dµ. RHµ. dµ.
R,
w 0.482 q
'S
dµ.' Bp µLq
h,, = 17.6; ~ = 0.0393
µLQ
dJP
6.31 dµ. 3Hdj;"
Dv dv µLB µL\l?
R,
7 7740 7 hL = 0.027; CFP 0.004 90. d' P'
-~ ~ T
µl I' µLq
R, 1 419000 ·= 3100 L:i.P = 0.000 668 = 0.1225 ~
µLQ µLB
L:i.P = o.ooo 273 ··([i- 0.000 191 ~
e Viscosity equivalents:
µ µLW
• = L:i.P = o.ooo 0340 ~
s
445
13. 12 Summary of Formulas - continued
Umílalion• of Darcy formula $ Empirical formul"• fer ih• fiow

of water, 1team, and gas
Nen·ccmpr&uiblit flow; liquidti:
Although the rational method (using Darcyºs for-
~rhe Darcy formula may be u.sed without restriction
mula) for solving flow problems has been recom-
for the ftow of water, oíl, and other liquids in pipe.
mended in this paper. sorne engineers prefer to use
However, when extreme velocities occurring in pipe
empírica! formulas.
cause the downstream pressure to fall to the vapor
pressure of the liquid, cavitation occurs and cal-
c\:J.lated flow rates are inaccurate.
Haun and Williams
Compreuihie flow; QC!H\l c;nd VA pon: formuiA for fh~w of wat111r: , Equation 3-9
When pressure drop is less than 103 of P,, use por
V based on either inlet or outlet conditions. Q = º.442 d'·" e (Pi ~ p,)°·"
When pressure drop is greater than 1 o3 of P, but
where:
less than 40';/( of P,, use the average of p or V e = 1 40 for new steel pipe
based on inlet and outlet conditions, or us.:: Equ~ 1 30 far new cast iron pipe
tion 3-20. 1 1 o far ri,·eted pipe
When pressure drop is greater than 40% of P,, use
the rational or empirical formulas given on this
page for compressible flow, or use Equation 3-20
(far theory, see page r -q). Equation 3-JO
(de!eted)
O lsothermal flow cf gas

Equafion 3-7
in pipe lines
Spitzglan formula fer low prenure gas:

(pressu;e len thon one pound gcuge) Equafion 3- l 1
q'h =
Flov..ing tcmpcraturc is 60 F
l& Simplifled comprenible flow EquaHon 3 _70

for long pipe line• Weymouth formulti
J( ~14/ ~') ( for high preuure gtu: Equafion 3- i 2
w =
2
(P'r)' ;;;, <P'2) )
q', = 28.0
/-(CP',)~J')--(-
d'·"'\j \ · S, Lm T
520)
w 0.1072 I(~)
\) V,JL
( (P',)'-:P, _(P',)')
Panhandlo formulal fer narural gas
pipe linH 6 to 24~inch diamelu
and RP = (5 x 10") to (14 x 106): Equation 3- J 3
q', 114.2
q,
' = 68E
3 .
el'·º"'( (P'r)' Lm- (P',)')º""•
19 Maximum (sonic) velocily of where: gas tempera tu re = 60 F
ccmpre..ible fluids in pipe
S, o.6
'fhe maximum possible velocity of a comprcssible E flow efficiency
1uld in a pipe is equivalent to the speed of sound E 1.oo ( 100%) for brand new pipe without
n the fluid; this is expressed as· any bends, elbows, valves, and change
of pipe diameter or elevation
v, Eq1.1ation 3-8
E 0.95 far very good operating conditions
v, ../ k g 144P' V E 0.92 for average operating conditions
E 0.85 for unusually unfavorable
v, 68. 1 ../ k P' V operating conditions
13.12 SUMMARY OF FORMULAS 447
13.12 Summary of Formulas - continued

e Head loss and pressure drop e Reslstance coefficient, K, for sudden and gradual
through valvas and fitting• enlargements in pipes
Head loss through valves and fittings is generally
given in terms of resistance coefficient K which Ir, o" 45º,
2.6 sin~ (1 -
indicates static head loss through a valve in terms
of "velocity head", or, equivalent length in pipe
K, = 13') 2 *Eqwfloo 3-17
2
diameters L/ D that will cause the same head loss
as the valve. Ir, 45º <o" 18oº,
K, =(1 - fJ')' *fquatloa 3-17.1
From Darcy's formula, head loss through a pipe is:
L v'
hL = f D 2i fquGlion 3-5 e Resistance coefficient, K, for sudden and gradual
contractions In pipes
and head loss through a valve is:
Ir, o< 45°,
hL = J< V2 fquation 3-14
2g K 1 -o.8sin!.(1 13')
2
therefore: K = f .!:::... fquation 3-15
D
Ir, 45º <O" 180º,
To eliminate needless duplication of formulas, the
following are ali given in terms of K. Whenever
necessary, substitute (/ L/D) for (K).
K 1 = 0.5 Vsin~ (1 - fJ') *Equal/oo 3-18.1
522 Kq' KQ'

-d-,- = 0.002 59 ---;;¡¡- fquation 3-14 *Note: The values of the resistance coefficients (K)
in equations 3-17, 3-17.1, 3-18, and 3-18.1 are
KB' KW'V• based on the velocity in the small pipe. To de-
0.001 270----¡;¡< = 0.000 0403 --d-,- termine K values in terms of the greater diameter,
divide the equations by (34.
0.000 1078 Kpv' = o.ooo ooo 0300 Kp V'
Kpt KpQ'
3.62 ----¡;¡< = 0.000 017 99 --r¡¡-
KpB'
e Discharge of fluid through valves,
0.000 008 82 -¡¡¡- flttings, and pipe; Darcy's formula
Lfquld flow: fquation 3.. 1'
=
K
0.000 000 280 --d-·-
w•v q 0.0438 d' ~~ = 0.525 d' ~ ~:
K(q',)'TS,
o.ooo ooo ooo 005 d' P'
K (q',)'S,'
l!:.P = o.ooo ooo 001 633 d' P
For compressible flow with hL or .6.P greater than approxi-

mately 10% of inlet absolute pressure, the denominator
should be multiplied by Y•. For values or Y, see page A-22.
ComprH1lltle flow:
/6PP',
• Pressure drop and flow of liquids of law q'11 40 700 Yd2 '\J KTi S, fquafion 3-20
viscosity using flow coefficient Yd' /6PP,

q'. 24700 s;-v~
6P ( Q )' p fquatiP 3-1'
Cv 62.4
q',. = 6 8 Yd' /6PP'1 = Yd' / 6PP,
• 7 "\JKT,S, 412 S,"\JK.
Q Cv~ 6P 6 : 4 = 7.90Cv ~-P-
6P
/6PP', Yd' ftJ5P,
q' 11.30 Yd' "IJ KT, S, = 6.87 S, '\J----¡("
Cv - Q J 6P (6q)
29.9 d'
,¡ / L/D
= 29.9 d'
,¡J< w = 0.525 Yd' ~
\}KV;
W= 1891 Yd' ~
\}KV;
891 d'
K Values of Y are shown on page A-22. For K, Y, and
(Cv)' t::.P determination, see examples on pages 4-13 anci 4-14.
13.12 Summary of Formulas - concluded
e Flow through nonios ond oriflcos e Speciflc gravity of liquids

(h and il/' measured across pjpe taps Any llqultl: lqNtion 3--2.5
al 1 diameter and 0.5 diameter)
( any liquid at óo F, )
Llquid: fquofion 3-21 p unless otherwise specified
8 = p (water at óo F)
q = AC..J2g hL = AC~2g (14:) t:,. P
0111:
- /6P 141.;
q 0.0438 d,• e ..J hL = 0.525 di' e 'V-P- S (óo F/óo F)
131.5 +DegAPI
- /6P Ll11ultl1 ll9hter than water: fquofion 3-27

Q 19.65 di2 C ..J hL = 236 di'C '\i-p-
S (óo F/óo F) = 140
1JO + Deg Baumé
w o.0438d12 C..J hLP' = 0.525.d,'C-VD.Pp
Liqultls hoavlor than wotor: fquation 3-21
W 157.6d1'CvhLp'= 1891d1'C..J6Pp
Values of C are shown on page A-20 S (óo F/óo F) = 145 - ~:~ Baumé
Ce1npres1lble flultls: Equ•ion 3-22
q'. ~6PP' e Spociflc gravity of gases

40 700 Y d,' C T, S, '
Yd,'C - - R (air) 5J.J
q'. 24 700 -s;-
v6P P 1 S, R (gas) = R (gas)
M (gas)
q'm 618 Y d,' e ~6; %," s, M (air)
Yd,•C --
q' m 412 ---:s;- ..J6PP 1
e General gas laws for ·porfod gases
q' 11.30 Yd'
, e ~6PP',
T, s, 141 ...,;... 3-30
p'V, w.RT
q' 6.87 -s.-

Yd1 C - -
2
V6PP, p w.
v.
p'
RT
144 P'
7fT""
1,llClfÍOlt WJ
w 0.525 Y d,' e ~t:,.p

~
v, R 1545 144 P'
pr Ecruation 342
M
w 1891 Yd,•C ~t:,.p
~
v, ··-'"" 3-33
w. 1545 T
p'V. n,MRT = n. 1545T = M
Values of C are shown on page A-20
Values of Y are shown on page A-21
fquation 3-34
w, p'M P'M _ 2.70 P'S,

• Equivalants of hoad loss
and pressure drop fquotion 3-23
p = V, = 1545 T 10.72 T - --T--
where:
hL =
144 6P
6P = ~
n, = w,/M
-
p 144 = number of mols of a gas
e Changos in resistance coefficiant, K,

required to compensa!•
e Hydraulic radius • ..-,..
for difforont pipo l. D. R cross sectional flow area (sq. feet)
H - wetted perimeter (feet)
K.-K,, (d.)'
J.
fqNfion 3-24
<- pago A·30) Equivalent diameter relationship:

D = 4RH
Subscript a refers to pipe in which valve will be installed.
Subscript b refers to pipe for which the resistance coefficient d = 48f{n
K was established. •See page 1-4 for limitations.
13.12 SUMMARY OF FORMULAS 449
Velocily cf l.lquicb in Pipe
The mean velocity of any flowing liquid can be caiculated

frorn the following formula. or. from the nomograph on the
opposite page. The nomograph is a graphicai solution
of the formula.
(For values of d, see pages B-16 to &.19)
The pressure drop per 100 feet and the velocity in Sched-
ule 40 pipe, for water at 60 F, have been calculated for
commonly used flow rates for pipe sizes of Ys to ·i4-inch;
these values are tabulated on page B- •4.
Exampl., 1
Given: No. 3 Fue! Oil at 6o F flows through a 2- Given. Maximum ftow rate of a liquid will be 300
inch Schedule 40 pipe at the rate of 45,000 pounds gallons per minute with maximum veiocity Jimited
per hour. to 1 ?, feet per second through Schedule 40 pipe.
Find: The rate of ftow in gallons per minute and Find: The smailest suitable pipe size and the
the mean veloc1ty in the pipe. velocity through the pipe.
Solution: Solution:
l. p page A-7 Con:nect Read
Connect Re ad l. Q V= 12 d = 3.2
1 300 1
2. W= 45 000
1
p = 56.02
1
Q = 100 2. 3%" Schedule 40 pipe suitable
J. Q = 100
1
2" Sched 40 1 V = 10 J. Q = JOO 1 J ~1Í" Sched 40 1 V = 10
Re<1•"""ble V8l@<iliea
for th" Flow of ""'"' !l"ough l'ip@
SeI"vice Condition Reasonable Velocity
Boiler Feed. 8 to 1 5 feet per second

Pump Suction and Drain Lines . .. 4 to 7 f eet per second
General Service. .4 to 1 o feet per second
City. to 7 feet per second
Nomendature
A cross sectional area of pipe or orilice. in RH = hydraulic radius. in feet
square feet re critical pressure ra(IO for c0mpressible flow
a cross sectional area of pipe or orifice, or íl.ow S = specific gravity of liquids at specified temper-
area in valve, in square inches ature relative to water at standard tempera-
B rate of flow in barreis (42 gallons) per hour ture (6o Fl
C flow coefficient for orifices and nozzles .S, specihc gravity of a gas relative to air =
= discharge coefficient corrected far vel- the ratio of the molecular weight of the
ocity oí approach = e,!,.;~¡ gas to that of air
e, discharge coefficient far orífices and nozzles T absolute temperature, in degrees Rankine
(46o + t)
Cv flow coefficient far valves: expresses tlow
rate in gallons per minute of 60 F water temperature. in degrees F ahrenheit
with 1.0 psi pressure drop across valve V specifíc volume of fluid, in cubic feet per
D interna! diameter of pipe, in feet pound
d interna! diameter of pipe. in inches \/ mean velocity of flow. in feet per minute
base of natural iogarithm = 2.718 \/, volume. in cubic feet
f friction factor in formula hL = f Lv'/D 2g mean velocity of flow. in feet per second
fr = friction factor in zone of complete turbulence v, sonic (or critica!) velocity of Aow of a gas,
in feet per second
g acceleration of gravity = 12.2 feet per
second per second W rate of flow, in pounds per hour
H total head, in feet of fluid w rate of f!ow, in pounds per second
h = static pressure head existing at a point, in U\i weight, in pounds
feet of fluid x percent quality of steam = 1oo minus per
h, total heat of steam. in Btu per pound cent of moisture
hL loss of static pressure head due to fluid Y net expansion factor far compressible flow
flow, in feet of fluid through orinces, nozzles, or pipe
static pressure head, in inches of water Z potential head or elevation above reference
level, in feet
res1stance coefficient or velocity head loss
k
in the formula. h 1, = Kv'/ig
ratio of specilic heat at constant pressure
....
G-k IAll..,.s
fJ ratio of small to large diameter in orifíces

to specific heat at constant volurne =
e.ple" and nozzles, and contractions or enlarge-
ments in pipes
L length of pipe, in feet
Lí D = equi valent length of a resistance to Aow,
in pipe diameters 6 differential between two points
Epsilon
length of pipe, in miles
molecular weight absolute roughness or effective height of
pipe wall irregularities, in feet
universal gas constant = 1545
Mu
n exponent in equation for polytropic change
(p'\/: = constant) µ. absolute (dynamic) viscosity, in centipoise
p pressure, in pounds per square inch gauge µ, absolute viscosity, in pound mass per foot
second or poundal seconds per sq foot
P' pressure, pounds per square inch absolute
(.set page 1-5 for diagram showing relalion-
µ'. absolute viscosity, in slugs per foot second
ship betwun gauge and absolute pres.sure) or pound force seconds per square foot
p' pressure. in pounds per square foot absolute Nu
Q rate of flow, in gallons per minute kinematic viscosiry, in centistokes
q rate of tlow, ih cubic feet per second at v' kinematic viscosity, square feet per second
flowing conditions ftho
q' rate of flow, in cubic feet per second at p we1ght density oí fluid, pounds per cubic ft
standard conditions ( 14. 7 psia and 6oF) p' density of fluid. grams per cubic centimeter
'
q' rate of ftow, in millions of standard cubic
feet per day, MMscfd angle of convergence or divergence in enlarge-
q'. rate of flow, in cubic.feet per hour at stand- rnents or contractbns in pipes
ard conditions (14.7 psia and boF). scfh
rate of llow, in cubic feet per minute at
flowing conditions Sllileerlpla lo< Dl11moter
q' .. - rete of flow, in cubic feet per minute at ( 1) ... defines smaller diameter
std. conditions (14.7 psia and óoF). scfm ( 2) ... defines larger diameter
R individual ¡¡as constant MR/M llullocrlpls lor Fluid Property
154s/M (1) ... defines inlet (upstream) condition
R, Reyno!ds number (2) ... defines outlet (downstream) condition
PPENDIX 1
SYMBOLS ANO
NOMENCLATURE-
CONVERSION
Symbol Quantity us X Muiti = Metrk

(x) Found in Text. Units Units
(a) acceleration, linear ft/s 2 3.0482E - Ol m/s 2
(a) speed of wave in liquid ft/s 3.0482E - 01 mis
(ap) acceleration, piston (plunger) ft/s 2 3.0482E - Ol m/s 2
(as) acceleration, suction ft/s 2 3.0482E -- Ol m/s 2
(ad) acceleration, discharge ft/s2 3.0482E - 01 m/s 2
a wave amplitude in 2.5400E + 01 mm
a wave amplitude ft 3.0480E - 01 m
(a) area, cross section, orifice in 2 6.4521E + 02 mm 2
(a) area, piston-plunger in 2 6.4521E + 02 mm 2
atmos pressure atmos l.0133E + 00 bars
atmos pressure atmos l .Ol33E + 05 N/m 2
atmos pressure atmos l.0133E + 05 Pa
atmos pressure atmos !.013 E + 02 kPa
atmos pressure atmos l.4696E + 01 psi
A are a in 2 6.4516E - 04 m2
(A) constant
A are a ft 2 9.2903E - 02 m2
452 APPENDIX 1
Symbol Quantity us X Multi = Metric

(Av) area, valve through in 2 6.4521E + 02 mm 2
A amplitude in 2.5400E + 01 mm
A amplitude ft 3.0480E - 01 m
·b breadth or width in 2.5420E + 01 mm
b breadth of width ft 3.0480E - 01 m
(b) coefficient of compressibility, liquid
bar pressure, barometric bar l.4504E + 01 psi
bar pressure, barometric bar 9.8692E - 01 atmos
bar pressure, barometric bar l.OOOOE = 02 kPa
barrel (42 gal) bbl l.5899E + 02 liters
B constant
B area, acoustic tube ft 2 9.2903E - 02 m2
B rate-of-flow 42 gal bbl/hr l.5899E + 02 L/hr
(BHP) Break Horse Power
e speed of surge wave ft/s 3.0482E - 01 m/s
e neutral axis to extreme fiber in 2.5400E + 02 m
e concentration, decimal or percent
e radial clearance, journal bearing in 2.5400E - 02 m
e acoustic capacitance ft5 /lb 5.797 E - 03 m5/kg
e fraction of solids (volume or weight)
e clearance volume, percent
(C..) volume fraction of solids
(C.,) weight fraction of solids
(CP) cost, pump
d diameter in 2.5400E + 01 mm
d displacement volume, in 3 l mm3
(D) diameter in 2.5400E + 01 mm
D displacement, liquid GPM 3.7854E + 00 lit/mio.
D displacement, linear in 2.5400E + 01 mm
(Dp) diameter, piston or plunger in 2.5400E + 01 mm
(D,) diameter, suction pipe in 2.5400E + 01 mm
(Dd) diameter, discharge pipe in 2.5400E + 01 mm
e pipe wall thickness in 2.5400E + 01 m
e base nat. log 2. 718
(E) Young's modulus of elasticity lbf/in 2 6.8948E + 03 N/m 2
(E) bulk modulus of elasticity lbf/in 2 6.8948E + 03 N/mm 2
E energy, kinematic lbf.ft l.3558E + 00 N.m
E energy, potential lbf.ft l.3558E + 00 N.m
E efficiency, decimal or percent
if) friction factor
if) frequency s - Hz
(j) spring displacement in 2.5400E + 01 mm
~~"'
APPENDIX 1 453
Symbol Quanfüy us X Multi = Met:ric

(x) Found in Text. Uriits Urdts
(j) coefficient of friction
(ft) foot ft 3.0480E - Ol rn
ft foot ft 3.0480E + 02 mm
( ft 2) foot 2 ft 2 9.2903E - 02 m2
foot 3 ft 3 2.8317E - 02 m3
( ft4 ) foot4 ft 4 8.63IOE - 03
(ft 5) foot 5 ft 5 2. 6307W -- 03 ms
(ftlsec) ft/s 3.048 E - Ol mis
( f.t/min) ft/min 3.048 E - 01 m/min
(ft-lb) force ft.lbf l.356 Joules
foot water ft 3.048 E + 02 kg/m 2
(F) force, load lbf 4.4482E + 00 N
F Froude Number
{F) valve spring force, installed lbf 4.4482E + 00 N
(F) valve spring force, at max. lift lbf 4.4482E + 00 N
g wavelength ft 3.0482E -- 01 m
(g) acceleration due to gravity ft/s2 3.0480E - 01 m/s 2
(G) modulus of elasticity, shea lbf/in 2 6.8947E + 03 N/m 2
G mass flow slug/s.ft 4.7880E + 01 kg/s.m
(G) torsional modulus lbf/in 2 6.8948E + 03 Nlm 2
(GPM) Gallons Per Minute
(gal) US (231 in 3) gal 3.7854E + 00 liters
gal US (231 in 3) gal 3.7854E - 03 m3
gal US (gallons per minute) gal/min 2.2712E - 01 m 3 /hr
(GPM) US (gallons per minute) gal/min 3.7853E + 00 lit/min.
gpm US (gallons per minute) gal/min 3.7854E - 03 m 3/min
grains (troy) grains 6.4800E - 02 grarns
h depth m 2.5400E + 01 mm
h height in 2.5400E + 01 mm
h head ( friction) ft 3.0480E - 01 m
h thickness, pipe wall m 2.5400E + 01 mm
(hcJ accelleration head ft 3.0480E - 01 m
H head, total Bernoulli ft 3.0480E - Ol m
(H) Henery' s gas constant
(H) head, liquid elevation ft 3.0480E - 01 m
H wave height ( = 2a) ft 3.0480E - 01 m
(HHP) Hudraulic Horse Power
HP horsepower HP 7.4570E + 02 watts
(HP) horsepower HP 7.4567E - Ol kw
horsepower-hours HP-hr 7.457 E - 01 kw-hrs
in inch in 2.5400E - 02 m
in inch in 2.5400E + Ol mm
454 APPENDIX 1

in 2 inch2 in 2 6.4516E - 04 m2
in 2 inch 2 in 2 6.4516E + 02 mm 2
in3 inch3 in3 l.6387E - 05 m3
in3 inch3 in3 l.6387E + 04 mm 3
in4 inch4 in 4 4.l623E - 07 m4
in 4 inch4 in4 4.lq23E + 05 mm4
in5 inch5 in 5 l.0572E - 08 ms
in 5 inch5 in 5 l.0572E + 07 ms
in Hg in Hg 3.3860E + 03 kPa
(in Hg) in Hg 3.342 E - 02 atmos
in H20 inches water in H20 2.54 E - 01 mm wat
(i) head loss gradient ft/ft m/m
(l) inertia, mass moment of slug.in 3.7069E - 01 kg.m
(J) constant derived from Henry's 14.7 X H
(J) inertia, polar mass moment of in4 4.l623E + 05 mm4
J pressure gradient psi/ft 2.2620E - 01 bar/m
J radius of gyration m 2.5400E - 02 m
(JI) modified Henry's constant 14.7 X H
(12) modified Henry's constant 14.7 X H
(k) concentration factor
(k) spring constant lbf/in l.7513 + 02 Nlm
k permeability of sand, etc.
(k) Wright factor for acceleration head
(K) bulk or vol. modulus of elasticity lbf/in 2 6.8948E - 03 N/mm 2
(K) stress concentration factor
(K) dampener sizing constant
(K') effective bulk modulus lbf/in 2 6.8948E - 03 N/mm 2
(Kc) Wahl' s factor for springs
In natural logarithm
Lm length pipe in miles miles l.6093E + 00 kilo
l length m 2.5400E + 01 mm
l length ft 3.0480E - 01 m
(lb) pound (weight) lb 4.5360E - 01 kg
lb pound (weight) lb 4.53600 + 02 g
(lb) pound (force) lbf 4.4482E + 00 N
(L) length in 2.5400E + 01 mm
L length ft 3.0482E - 01 m
(L) length of shaft in 2.5400E + 01 mm
(Ll) life hr hr
(L2) life hr hr
(Lv) lift, valve m 2.5400E + 01 mm
(LJ length, connecting rod m 2.5400E + 01 mm
APPEND!X 1 455
Symbol Qmm.tity us X Multi = Me trie

(x) Fouml. in Text. Units Units
(L,) length, suction pipe ft 3.0482E - Ol m
(Ld) length, discharge pipe ft 3.0482E - Ol m
(L) length pipe ft 3.0482E - Ol m
(L¡) length, free, spring in 2.5400E + O! mm
(L;) length, installed spring m 2.5400E + 01 mm
m mass (F!a) slug iA594E + Ol kg
(m) mass lb 4.5359E - 01 kg
m molality
(M) Moles of liquid per IOO ft of solvent
M Mach Number
(M) moment of force, bending lbf. ft l.3558E + 00 N.m
(M) torque lbf.in Ll298E - 01 N.m
(ME) Mechanical Efficiency
(ML) Lame' Multiplier
(MLR) mass log rate mg/hr mg/hr ''I'
(MR) universal gas constant (1544)

(M X 1) initial moles of gas per l 00 ft cu 3 liquid
(M X 2) secondary moles of gas per l 00 ft cu 3 liquid
n revolution per unit time
n number
(n) number of cylinders
n.. number of valves 9multiple)
n polytropic exponent for gas
n factor of safety
n roughness factor
(n') number pressure strokes per revolution
ns specific speed
N number
N normal force lbf 4.4482E + 00 N
N safety factor
(N) revolutions per minute RPM RPM
(Nª) number of active coils, spring.
(N,) number of total coils, spring (usually Na + 2).
oz ounces (avdp) oz 2.8349E + Oi grams
(P) normal force per unit area lbf/in 2 6.8948E + 03 N/m 2
(P) pressure PSI 6.8948E + 03 N/m 2
(P) pressure PSI 6.8948E - 02 bars
p pressure PSI 6.8948E + 03 Pa
p pressure PSI 6.8948E + 00 kPa
p pressure PSI 6.897E - 03 N/mm 2
p pressure atmos l.4696E + 01 PSI
p pressure, bars bars 9.8692E - 01 atmos
456 APPENDIX 1
Symbol Quantity us X Multi = Metrk

p pressure, bars bars l.4504E + 01 psi
p pressure, atmospheres atmos l.Ol32E + 05 N/m 2
p pressure, atmospheres atmos L0132E + 05 Pa
p total load lb 4.4482E + 00 N
(P) power (HP)ft.lb/s l .3558E - 03 kW
p peri meter
(Pl) primary pressure PSI 6.8948E - 02 bar
(P2) secondary PSI 6.8948E - 02 bar
(PA) applied pressure PSI 6.8948E - 02 bar
(Pu) atmospheric pressure PSI 6.8948E - 02 bar
(P.,.J suction tank head, ft to PSI 6.8948E - 02 bar
(P.,") liquid vapor pressure PSI 6.8948E - 02 bar
(P,,) suction tank gauge pressure PSI 6.8948E - 02 bar
(P,,) suction velocity pressure PSI 6.8948E - 02 bar
(Ps¡) suction pipe friction loss press. PSI 6.8948E - 02 bar
<Psac) pump valve spring POSIVA PSI 6.8948E - 02 bar
(PV) Present Value
(PRL) Piston Rod Load, lbf 4.4482E + 00 N
(POS IVA), val ve, lb/in 2 of val ve area lbf/in 2 6.8948E - 03 N/mm 2
q load per unit distance lb/ft l.488lE + 00 kg/m
q flow rate ft 3/s 2.8316E - 02 m 2/s
q' flow-rate, std. cond. ft 3/s 2.8316E - 02 m 3/s
(Q) flow rate US gpm 3.7854E + 00 lit/min
(Q) flow rate US gpm 2.27l2E - 01 m 3/hr
(Q,) ftow-rate, theoretical US gpm 3.7854E + 00 L/min
(Qa) fiow-rate, actual US gpm 3.7854E + 00 L/min
(Q) force or load, concentrated lbf 4.4482E + 00 N
(Q) moment of area in 3 l.6308E + 04 mm 3
(Rl) Radius, bore, liquid end in 2.5400E + Ol mm
(R2) Outside, bore, liquid end in 2.5400E + Ol mm
(Rª) acoustic resistance lb.s/ft 5 l.7242E + 02 kg.s/m 5
(r) radius in 2.5400E + 01 mm
(R) Radius pipe in 2.5400E + 01 mm
(Re) Reynold's number
(R) Gas constant (MR/M) 1544/M
(R) radius (conduit, hydraulic) in 2.5400E + 01 mm
(s) stroke length in 2.5400E + Ol mm
(s) stress lbf/in 2 6.8948E + 03 N/m 2
(slug) slug(32.17 lb) lb 4.5359E - 01 kg
(S) specific gravity, with respect to water.
s slip, pump loss, decimal or percent
(S1) specific gravity of liquid
APPENDIX 1 457
Symbol Quantity us X Multi = Me trie

(Sm) specific gravity of slurry (mixture)
(S,) specific gravity of solids
(T) temperature, degrees as specified; F., C., R., K.
t temperature degF degK
t thickness ID 2.5400E + 01 mm
(t) time s s
(ton) long (2240 lb) ton l.Ol60E + 03 kg
ton long (2240 lb) l.12 tons
(short)
(ton) short (2000 lb) ton 9.0718E + 02 kg
(T) torque lbf.ft l.3558E + 00 N.m
(T) tangent force lbf 4.4482E + 00 N
(T) temperature, absolute deg R deg K
(T) pulsation, percentage
T period s s
(TCP) Total Cylinder Pressure PSI 6.8948E - 02 bar
u displacement, linear in 2.5400E + 01 mm
u x velocity component
u peripheral velocity of runner
V velocity, linear ft/s 3.0480E - 01 mis
V y velocity component
V Poissons ratio
(V) Volume of free gas ft 3 2.8317E - 02 m3
(V) specific volume ft3/lb 6.2428E - 02 m3/kg
(V) velocity, mean ft/s 3.0480E - 01 mis
. (V) Volume ft 3 2.8317E - 02 mJ
(VE) Volumetric Efficiency
(Vd) Volume discharge GPM 3.7850E + 00 lit/mio
(V,) equivalent volume GPM 3.7850E + 00 lit/mio
(Vg) Volume dampener gal 3.7850E + 00 lit
(Vp) Volume piston GPM 3.7850E + 00 lit/mio
(V,) Volume suction GPM 3.7850E + 00 lit/mio
w displacement, linear ID 2.5400E - 02 m
w specific weight lb/ft3 l.6018E + 01 kg/m 3
w load per unit distance lbf/ft l.4594E + 01 Nlm
w rate of ftow lb/s 4.5359E - 01 kg/s
w weight per foot, pipe lb/ft l.4882E + 00 kg/M
w load, total lb 4.4482E + 00 N
(W) weight, mass lbf 4.4482E + 00 N
w rate of ftow lb/h 4.5359E - 01 kg/h
(W) work ft.lbf l.3558E + 00 m.N
X distance in 2.5400E + 01 mm
458 APPENDIX 1

X displacement in 2.5400E - 02 m
y depth of flow
y net expansion factor for compressible ftow
z distance in 2.5400E - 02 m
(Z) distance ft 3.0480E - 01 m
z section modulus in3 l.6388E - 05 m3
alpha angle, con. rod to crosshead deg deg

alpha coefficient of expansion, linear deg F deg K
theta angle, crank rotation deg deg
BETA angle, connecting rod to crank deg deg
beta compressibility factor, liquids
tau angle rad rad
tau torsional stress lbf/in 2 6.8948 + 03 N/m 2
DELTA differential between two values
DELTA deftection of.beam, maximum in 2.5400E - 02 m
delta elongation, total in 2.5400E - 02 m
epsilon strain, normal
gamma strain, shear
epsilon roughness, pipe ft 3.0480E - 01 m
gamma specific weight lb/ft3 N/m3
epsilon roughness
eta efficiency
theta angle, crank rotation deg deg
theta angle deg deg
lambda wavelength in, ft m
lambda ratio; stroke to connecting rod length.
mu viscosity, dynamic lbf.s/ft2 4.7881E + 00 N.s/m 2
rho density slugs/ft3 5.1538E + 02 kg/m3
nu Poisson ratio
nu frequency factor, pipe spans
upsilon viscosity, kinematic ft2/s 9.2903E - 02 m2/s
SIGMA summation
sigma stress, normal lbf/in 2 6.8948E + 00 kPa
delta surface tension
tau stress, shear lbf/in 2 6.8948E + 00 kPa
phi angle
psi angle
omega angular velocity (2 f) rad/s rad/s
APPENDIX 2
SUBSCRIPTS
a area g gauge
a acceleration gs gauge, suction
a atmospheric i installed
ac acceleration L lift
ad acceleration, discharge l liquid
ap acceleration, piston m mixture
as acceleration, suction p pis ton
e precharge s suction
e connecting rod s solids
d discharge u upward
d diameter V velocity
e equivalent V valve
f free V volume
f friction vp vapor pressure
g U.S. gallon z head, feet
ACft
APPENDIX 3
ABBREVIATIONS
RECIPROCATING PUMP TYPES

Type Name n
SIMSA Simplex single-acting 1
SIMDA Simplex double-acting 2
DUPSA Duplex single-acting 2
DUPDA Duplex double-acting 4
TRIPSA Triplex single-acting 3
TRIPDA Triplex double-acting 6
TRIART Triplex articulated 6
TRIPOP Triplex opposed pushrod 6
QUASA Quadruplex single-acting 4
QUADA Quadruplex double-acting 8
QUINSA Quintuplex single-acting 5
QUINDA Quintuplex double-acting 10
SEXSA Sextuplex single-acting 6
SEXDA Sextuplex double-acting 12
SEPSA Septuplex single-acting 7
SEPDA Septuplex double-acting 14
NONSA Nonuplex single-acting 9
NONDA Nonuplex double-acting 18
Other
ATM Atmosphere
BHP Brake or input horsepower
460
APPENDIX 3 461
Other
BPD Barreis per day
CFM Cubic feet per minute
cv Clearance volume, percent
FPM Feet per minute
FPS Feet per second
GPM U.S. gallons per minute
GPR U.S. gallons per revolution
HHP Hydraulic horsepower
MAP Maximum allowable pressure
ME Mechanical efficiency
NPSHA Net positive suction head available
NPSHR Net positive suction head required
PRL Piston rod load
PSIG Pounds per square inch gauge
PSIA Pounds per square inch absolute
PSI Pounds per square inch
PTP Peak-to-peak
RPM Revolutions per minute
VE Volumetric efficiency
R'PM Reversal Rate, per Minute
INDEX
Abrasion, 316 Bending stress, crankshaft, 159

Abrasive wear, 317 Bernoulli' s equation, 414
Abrasivity: Black Mesa Coal Pipeline, 304
definition, 271, 321 Blanket, gas, 93
dry, 289 Bolts, 384
measurement of, 293 tightening specifications, 384
Acceleration: Boosterpump, suction, 345
definition, 38, 74 B-lObearinglife, 164.
head, 41, 365 Bulk properties of solids, 308
piston,41 By-Pass, pump, 349
pressure, 41
Acoustic filters (dampeners), 123 Calculation:
Adhesive wear, 316 free gas, 66
Air and gas saturation of liquids, 86 mechanical natural frequency, 369
American Petroleum lnstitute (API) standards, MillerNumber, 296
193 parts life extension, 328
Application, pump, 342 SAR Number, 296
Articulated pump, 9 volumetric efficiency:
Attenuation, 108 hydrocarbons, 361
Autofrettage, 178 water, 358
Carbon dioxide, effects, 227
Basic elasatomers, 226 Cavitation, 89, 320
Baume degrees, 439 Centrifugalpumps,55
Bearings, 163 Ceramic plunger, 211
journal, 165 failure mode, 336
life, average, 164 pitting, 334
life, B-10, 164 Charging pumps, suction, 345
roller, B-10 life, 164 Clearance volume:
Bending moments, crankshaft, 159 method of measuring, 343
INDEX 463
Coal slurries, 284 Duty:

Coal slurry viscosity, 284 continuous, 170
Coating and plating, 186 intermittent, 169
Compressibility, liquid, 342 Dynamics, liquid piping system, 36, 60, 367
Concentration, slurry, 284, 286
Connecting rod, 153 Effectsofwear, 289, 320
forces, 153 Efficiency:
Continuous service, 169 mechanical, 168
Controlled-compression gaskets, 225 volumetric, 274
Corrosion, 320 Elastomers:
Cost, pump, 338 basic, 226
Crank end (CE), 10 in carbon dioxide, 227
Crank-end stroke, 10 Energy, transmit time, 66, 87
Crankpin bearing, 21 Equations suction system, 76
Crankshaft: 20 Equivalents, Symbols and Nomenclature,
bending moments, 161 APPENDIX 1
bending stress, 161 Erosion, 287
offset, 152 high-velocity, 318
types, 155 low-velocity, 318
Critical suction conditions, 42 Expendable parts, guideline, 236
Crosshead, 21, 162 Iiners,206
extension, 21 pistons, 207
forces, 158 plungers, 211
outside (upper), 10 packing, 215
Crushing and grinding, 317 standards, API, 193
Cylinderliner, 15 Extension, crosshead, 21
Cylinders, numberof, 147
Family planning, 142
Damped pressure gauge, 405 Fatigue, 320
Dampener (See Pulsation dampener) Filter, acoustic, 124
Darcy's Formula, 416 Final important notes, 139
Definitions, pumps, 1 Finite element anaysis, 372
Delay, val ve, 58 Flow:
Derivation: ofliquids, 420
acceleration pressure, 40 nature offlow in pipe, 412
inlet pressure, 41 through nozzles and orifices, 419
Design philosophy, 373 through orifice meters, 424
Diaphragm pump, 10 variation, 37
Direction, rotation, 150 Auids, physical property of, 409
Discharge: Aushing, plungers, 276
acceleration, 42 Aushing velocity, 279
dampeners, 95 Autter, valve, 266
definition, 36 Forces:
system, 103 connecting rod, 153
velocity (liquid), 95 crosshead, 162
Discontinuity, 97 pumpmechanism, 156, 157
Discussion of ideal waveforms, 44 Fraction volume of solids, 309
Dissolved air and gas, 56 Frequency:
Double-acting pump, 3 natural, pipe, 99
Double-acting vs. single-acting, 23 pulsation, 99
Duplex pump: response, 387
double-acting, 8 Frettingwear, 320
single-acting, 8 Friction factor, 417
464 INDEX
Gas, blanket, 86 ends, configurations, 20, 172

Gaskets, controlled compression, 215 ends, stress reduction, 177
Gas saturation, 86 ultra-high pressure design, 187
Gas type dampener, 109 Load, piston rod (PRL), 144
sizing, 112 Loops:
Gear rating, 203 pumphook-up, 345
Gears, interna!, 34 suction system, 345
Geometry, pump, 37 Lubrication, 156
Gland, 18
tightening, 231 Main bearing, 20
Glossary ofterms, 234 Manifolds, 15
Gold Number, 302 Material selection, guidelines, 236
Guidelines for material selection, 236 Maximum piston diameter, 146
Meaos of reducing wear rate, 337
Harmonic frequency, 99 Measurement:
Head: TCP(ex NPSH), 80
end stroke, 10 pressure, 405, 415
static, 365 vibration, 389
velocity, 365 Mechanical efficiency, 167
Head end (HE), 10 Mechanical forces, 154
High-frequency pulsations, 57 Mechanics of wear, 316
High suction pressure, 375 Mechanisms, pump, 37
Hookup, pump, 343 Metal-to-metal plungers, 121
Hydraulic flow and pressure waveforms, 44 Micronic particles, 307
Hydraulic Institute: relative size, 307
Nomenclature and definitions, 13 MillerNumber, 293, 296
Standards of application, 347 calculation, 295
Hydraulic radius, 414 system, overview, 297
MillerNumbers, typical, 298
Increasing TCP, 88 Minerals, specific gravity, 306
Inlet, definition, 36 Modes, wear, 316
Intermittent service, 169 Multicylinderpump, 8
Intemalgears,34 Multiple dampeners, 117
Myth ofNPSH, 84
Jacoby Leakage, 232
Joumalbearings, 165
Nature offlow in pipe, 412
Laminarflow, 412 Net Positive Suction Head, 55
Lantem ring, 17 Nitrogen cylinder, capacity, 119
Life, expendable parts, 339 Nomenclature, 13.
Lift: Numberofcylinders, 147
head, 365
suction, 365 Offset crankshaft, 152
valve, required, 260 Oil:
Liner development, 207 cleaning, 167
Liners, 280 temperature, 167
Liquid: Oil mixed slurries, 300
by-pass, 349 Opening, sieve mesh, 307
compressibility, 342 Opposed pump, 9
Liquid Ends, 172 Optical phaser, 402
cylinder, 14 0-Rings, 225
end parts, 14 Oscillographs. reading of, 390
end stresses, 176 Outlet, definition, 37
INDEX 465
Outside-packed, 13 (POSIVA) and RPM, 248

Overrunning, 151 Power:
Overshoot, 60 costs, 363
end parts, 20
frame, 20
Packing, 17, 274 Power-operated valves, 265
elastomers, 226 Precharging gas-type dampeners, 118
gaskets, 225 Pressure:
leather, 227 disturbances, 97
lubrication, 231 drop in pipe, 420
non-adjustable, 216 dueto acceleration,
0-rings, 225 high suction, 375
spring-loaded, 218 lubrication, 166
tightening, 231 measurement methods, 387
Packing history, 215 pulses, 95
Particles, slurry, characteristics, 316 suction, 91
Particle shape, 316 survey, 387
Particle size, 286 waveforms, 390
Parts wear and life, 316 Properties, slurry, 299
PASAFEdampener, 107, Pulsation, 60
sizing, 114 degree of, 106
Phase, 117 dueto flow-variation, 98
Phaser, optical, 402 Pulsation and control, classification, 97
Pipe: Pulsation dampener, 356
data,441 classification, 102
friction factors, 418 discharge, 120, 357
relative roughness, 413 gas-type, 102
Pipeline flow problems, 422 manifolded pumps, 119
Pipeline GPM per Million Tons Solids, 31 O multiple dampeners, 117
Pipe vibration, 367 PASAFE control sizing, 11
Piping: precharging gas type, 118
acoustics, 367 sizing, 111
design guidelines, 374 suction, 356
natural frequency, 367 types, 103
Piston, 15, 280 vaporization, 118
acceleration, 38 Pulsation dampener Facts, 121
liner clearance, 333 Pulsation and surge control, I 09
maximumdiameter, 146 water hammer, JO I
membrane pumps, 280 Pulsations, high-frequency, 57
rod load, 144 Pulse, pressure, 57
speed, 69 Pump:
velocity, 37 acoustic filters (dampeners), 123
Piston pump, 6 application, 342
Piston, bearings, 163
maximumsize, 146 connecting rod forces, 153
Plunger: cost, 377
ceramic, 334, 336 crankshafttypes, 154
flushing methods, 276 crankshaft bending moments, 159
flushing details, 278 crossheads, 162
vs. piston pump, 33 design, 140
pump,6 family planning, 142
Plungers, 17, 333 hookup, ideal, 344
POSIVA, valve spring, 248 lubrication, 156
466 INDEX
maximum piston size, 146 properties, 270

mechanical efficiency, 168 pumping, 270
numberofcylinders, 147 service, pumps, 271
offset crankshaft, 152 suction pressure, 283
performance waveforms, 69 tables:
pis ton rod load, l 44 fraction volume, 312
POSIVA and RPM, 69, 248 GPM per mili ion tons, 3 JO
pressure survey, 387 specific gravity, 305
rating standards, 140 S-N curve, 172
reduced plunger, 72 Solids, bulk properties, 308
service, intermittentorcontinuous, !69 Solubility, air or gas, 86
for slurry service, 27 l Southwest Research Institute, 367
speed, 149 Specific gravity, 411
stroke length, 146 Specific volurne, 411
stroke reversa!, 69 Speed of duplex pumps, 149
torque, 348 Speed ofmulticylinder pumps, 149
unusual pressure in cyiinder, 145 Speed fRPM), 325
volumetric efficiency statistics, 72, 167 Split packing rings, 218
windows ofnonutulization, 143 Spring loaded packing, 262
Springs, valve, 248
Stabilizer, suction, 102
Rating standards, 140
Standard definitions, dynamics, 96
Reciprocating pump, J, 342
Standards of Application, HI, 347
Reducing wear, 337
Standby pumps, 329, 379
Reliefvalve, 358
Starting pumps, 348
Reynolds Number, 413
Strain bolts, 384
Roller bearings, l 64
Strain gages, 403
Stress reduction:
Saltation ftow, 318 autofrettage, 178
Sample problem, surge control, 138 coating and plating, 186
SAR Number, 294, 296 shot peening, l 86
calculation, 295 strain bolts, 384
Saturation, air and gas, 86 Stress reduction methods, 177
Savage River iron ore pipeline, 305 Stress-strain curve, 172
Screen sizes, 307 Stroke length, 146
Seats, valve, 245 Stroke reversals, 323
Shotpeening, 186 Stuffing box, 17
Simplex single-double-acting pump, 1 design, 230
Single-acting pump, 3 trim, 228
Size, particle, 307 wear, 228
Sizing pumps for pipelines, 376 Suction, critica! conditions, 89
Sizing of surge suppressors, 135 Suction charging (booster) pump, 345
Slurries, oil-mixed, 300 Suctiondampenersuctionpressure: high, 171
Slurry: Suction requirements, 55
abrasion testing, 293 Suction requirements, analysis, 57
abrasivity, 338 Suction requirernents, factors, 63
concentration, 299 Suction stabilizer, 75, 125
corrosion, 300 Suctíon system !oops, 345
corrosive effect, 301 Suction system(s), described, 92
erosion, 287 short vs. long, 73
horizontal vs. vertical pumps, 282 Surge control, 127
partirle shape, 299, 340 methods, 131
particle size, 299 sample problem, 138
INDEX 467
Surge control sizing, 135 seats, 245

Surge supressor, 135 seat taper, dimensions, 245
Switch-looppumping, 281 in series, 266
Synchronized flushing, 276 for slurry service, 287
spring:
TCP, definition, 58 installed load, 248
TCP, by degrees rotation, 74 POSIVA,248
TCP testing, 80 spring design, 251
Testing, pump suction requirements, 89 springs, 248
Test procedure: steady state flow, 267
MillerNumber, 294 tests, 256
Notes regarding, 88 through area, 259
procedure for testing, 82, 295 types, 241, 242
SAR Number, 294 unbalanced area, 264
TCP,80 unloading system, 345
TCP required, 57 weight, 260
Tests, valve, 256 Vaporization or heated dampener, 118
Tests, valve combinations, 256 Velocity:
Theory offlow in pipe, 408 head,445
Thread types, 194-98 liquids in pipe, chart, 448
Through area, valve, 259 inpipe,420
Tightening, gland, 231 piston, 37
Tons per year, solids, 3 JO valve, 256, 263
Torque: Vertical pump, 1
gland tightening, 231 Vibration, pipe, 368
pump, 348 Vibration measurement methods, 389
Transducers, pressure, 387 Vibration nomograph, 391
Transit time, suction, 87, 66 Viscosity, 409
Transmission ratio, 108 Volume fraction of solids, 31 O
Turbulent flow, 412 Volumetric efficiency, 72, 167, 332
Typical Miller Numbers, 298 calculation, 358
Typical waveforms, 392-93 hydrocarbons,361
Unbalanced valve area, 264

Underrunning, 150 Waveform, pressure, 69
Uppercrosshead, 20 Waveforms, 44
performance, 69
Valves, relation to pump cycle, 67
assembly, 17 typical, 67, 389
Ball, 253 Wear:
chestcover, 15 abrasion-corrosion, 317
classification, 240 abrasive, 317
combinations, 256 adhesive, 316
delay, discharge opening, 99, 268 crushing and grinding, 317
efficiency, determination, 255 liner profile, 323
flutter, 266 mechanics of, 316
leakage and slip, 65 modes,287,316,317
lift required, 260 saltation, 318
in parallel, 266 scouring, 317
plateordisc, 15 Wearblock, 293
plate(valvedeck), 15 Weight density, 411
poweroperated, 265 Weight fraction of solids, 309
seat, taper fit, 250 WindowsofNonutilization, 143

PDF The Reciprocating Pump Theory Design and Use John e Miller 2nd Editio DL

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

PDF The Reciprocating Pump Theory Design and Use John e Miller 2nd Editio DL

Uploaded by

Copyright:

Available Formats

THE RECIPR CATIN PU P

THEORY, DESIGN, ANO USE

KRIEGER PUBUSHING COMPANY

Printed and Published by

Copyright © 1995 by Krieger Publishing Company

FROM A DECLARATION OF PRINC!PLES JOINTLY ADOPTED BY ,A COMMITTEE OF THE AMERI-

Library of Congress Cataloging-in-Pubiication Dala

Miller, John E. (John Evans), 1909-

3.1.3 More on Suction Requirements / 58

4. PULSATION ANO SURGE CONTROL 95

4.2.6 Type GFFE I l05

5. PUMP DESIGN 140

5.15 Crankshaft Bending Stress-Calculations I 161

6. LIQUID ENDS 172

7. EXPENDABLE PARTS 193

7.5.6 Multiple Material Assemblies / 219

8.13 Steady-State Flow Through Valves / 267

9. SLURRY PUMPING 270

9 .12 Savage River Mines I 305

1 PARTS WEAR AND LIFE 316

11. APPUCATIONS 342

11. 5. 7 Calculations I 383

12. INSTRUMENTATION 387

13. THEORY OF FLOW IN PIPE 408

The uneven displacement generated by the piston rod of a conventional double-

Vertical triplex single-acting plunger pump. (Courtesy Wo1thington Pump Di-

S!NGLE-ACTING PUMP. A pump in which the liquid in each cylinder is discharged

Simplex single-double-acting pump. (Courtesy American Spray Industries,

1" to 2" Plunger by 2'h" Stroke Simplex

ARTICULATED PUMP. A double-acting design attained by driving an opposed set

MEMBRANE PUMP. The membrane pump, (sometimes erroneously referred to as a

DIAPHRAGM PUMP. A single-acting or double acting reciprocating pump with the

Figure l.lA Diaphragm pump

OUTSIDE-PACKED. A term usually used to refer to plunger pumps in which the

1.2 NOMENCLATURE ANO DEFINITIONS

LIQUID END PARTS

Figure 1.17. Liquid cylinder.

Figure 1.18. Manifold.

CYUNDER UNER. A replaceable liner which is placed in the cylinder of a piston

Figure 1.19. Piston assembly.

Figure 1.21. Cup type piston.

Figure 1.22. Slush piston.

Figure 1.23. Individual ring piston.

PLUNGER. A smooth rod which is attachable to a crosshead and is capable of ex-

Figure 1.24. Plunger.

Figure 1.26. Packing.

Figure 1.27. Gland.

Figure 1.28. Lantern ring.

Figure 1.30. Disc valve assembly.

Figure 1.32. Ball valve assembly.

UPPER CROSSHEAD. U sed in vertical plunger pumps to transmit the reciprocating

POWER END PARTS

Figure 1.3;3. Upper crosshead.

CRANKPIN BEARING. Transmits the oscillating reciprocating load transmitted by

POWER CROSSHEAD. Creates a linear reciprocating motion derived froµi the

Figure 1.34. Liquid end, horizontal plunger power pump.

Cylinder, Liquid 309 Manifold, Suction 321 Deftector

301 Cylinder, Liquid 308 Manifold, Discharge 312 Ring, Follower

1.36. Liquíd end, horizontal side pot piston pump.

Cylinder, Liquid 310 Box, Liquid Stuffing 317 Follower, Piston

1 DOUBLEmACTING VERSUS SINGLEmACTING PUMPS

Figure 1.37. Liquid end, horizontal valve plate piston pump.

301 Cylinder, Liquid 311 Giand, Liquid 319 Plate, Valve