Industrial Valves Calculations_Wiley

Industrial Valves
Industrial Valves
Calculations for Design, Manufacturing,

Operation, and Safety Decisions
First Edition
Karan Sotoodeh
University of Stavanger
Oslo, Norway
Copyright © 2023 by John Wiley & Sons, Inc. All rights reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise,
except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to
the Publisher for permission should be addressed to the Permissions Department, John Wiley &
Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at
http://www.wiley.com/go/permission.
Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John
Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be
used without written permission. All other trademarks are the property of their respective
owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.
Limit of Liability/Disclaimer of Warranty

While the publisher and author have used their best efforts in preparing this book, they make no
representations or warranties with respect to the accuracy or completeness of the contents of this
book and specifically disclaim any implied warranties of merchantability or fitness for a particular
purpose. No warranty may be created or extended by sales representatives or written sales materials.
The advice and strategies contained herein may not be suitable for your situation. You should consult
with a professional where appropriate. Further, readers should be aware that websites listed in
this work may have changed or disappeared between when this work was written and when it is
read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial
damages, including but not limited to special, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please
contact our Customer Care Department within the United States at (800) 762-2974, outside the
United States at (317) 572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print
may not be available in electronic formats. For more information about Wiley products, visit our
web site at www.wiley.com.
Library of Congress Cataloging-in-Publication Data

Names: Sotoodeh, Karan, author.
Title: Industrial valves : calculations for design, manufacturing,
operation, and safety decisions / Karan Sotoodeh.
Description: First edition. | Hoboken, New Jersey : Wiley, [2023] |
Includes bibliographical references and index.
Identifiers: LCCN 2023002088 (print) | LCCN 2023002089 (ebook) | ISBN
9781394185023 (cloth) | ISBN 9781394185030 (adobe pdf) | ISBN
9781394185054 (epub)
Subjects: LCSH: Valves–Design and construction–Mathematics. | Fluid
dynamics–Mathematics. | Engineering mathematics–Formulae.
Classification: LCC TS277 .S68 2023 (print) | LCC TS277 (ebook) | DDC
621.8/40151–dc23/eng/20230302
LC record available at https://lccn.loc.gov/2023002088
LC ebook record available at https://lccn.loc.gov/2023002089
Cover Image(s): Rawf8/Adobe Stock Photos; Who is Danny/Adobe Stock Photos

Cover Design: Wiley
Set in 9.5/12.5pt STIXTwoText by Straive, Pondicherry, India

v
Contents
1 Flow Capacity 1
1.1 Introduction 1
1.2 Flow Coefficient Chart and Flow Curve 8
1.3 Rangeability and Turndown 12
1.4 Valve Authority 14
1.5 Valve Gain 15
Questions and Answers 16
Further Reading 20
2 Valve Sizing 22
2.1 Introduction 22
2.2 Isolation Valve Sizing 22
2.3 Nonreturn (Check) Valve Sizing 26
2.4 Control Valve Sizing 34
2.4.1 Control Valve Sizing for Liquids 34
2.4.1.1 Specify the Variables Required to Size the Valve 35
2.4.1.2 Determine the Equation Constant (N) 37
2.4.1.3 Determine Piping Geometry Factor (FP) 37
2.4.1.4 Determine the Maximum Flow Rate (qmax) and Maximum Pressure
Drop (ΔPmax) 39
2.4.1.5 Solve for Flow Coefficient 44
2.4.1.6 Select the Correct Valve Size 44
2.4.2 Control Valve Sizing for Gas and Steam 47
2.4.2.1 Specify the Variables Required to Size the Valve 47
2.4.2.2 Determine the Equation Constant (N) 48
2.4.2.3 Determine Piping Geometry Factor (FP) 48
2.4.2.4 Determine the Expansion Factor (Y) 48
2.4.2.5 Solve for the Required Flow Coefficient (Cv) 50
2.5 Safety Relief Valve Sizing 56
vi Contents
2.5.1 Sizing for Gas or Vapor Relief 59

2.5.1.1 Critical Flow 59
2.5.1.2 Subcritical Flow 73
2.5.2 Sizing for Steam Relief 75
2.5.3 Sizing for Liquid Relief 79
2.5.3.1 Sizing for Liquid Relief with Capacity Certification 79
2.5.3.2 Sizing for Liquid Relief Without Capacity Certification 84
2.5.4 Sizing for Two-Phase Liquid/Vapor Relief 85
2.5.4.1 Sizing for Saturated Liquid and Saturated Vapor, Liquid Flashes 88
2.5.4.2 Sizing for Subcooled at the Pressure Relief Valve Inlet 91
2.5.5 Sizing for Fire Case and Hydraulic Expansion 93
2.5.5.1 Hydraulic Expansion (Thermal Expansion) 95
2.5.5.2 Sizing Safety Valve for the Fire Case 96
Further Reading 110
3 Cavitation and Flashing 112

3.1 Introduction 112
3.2 Cavitation 112
3.2.1 What is Cavitation? 112
3.2.2 Cavitation Essential Parameters 113
3.2.3 Cavitation Analysis 115
3.3 Flashing 116
Further Reading 123
4 Wall Thickness 125

4.2 ASME B16.34 Minimum Wall Thickness Calculation 125
4.2.1 Conservation Approach (Mandatory Appendix A) 125
4.2.2 Nonconservation Method 129
4.2.3 ASME Sec. VIII Div. 02 Wall Thickness Calculation 134
4.3 Wafer Design Thickness Validation 136
Further Reading 147
5 Material and Corrosion 149

5.2 Carbon Dioxide Corrosion 150
5.2.1 Corrosion Mechanism 150
5.2.2 Corrosion Mitigation 151
5.2.3 Corrosion Rate Calculation 152
Contents vii
5.2.3.1 Basic CO2 Corrosion Rate 152

5.2.3.2 Corrective CO2 Corrosion Rate 154
5.2.3.3 Final CO2 Corrosion Rate 161
5.3 Pitting Corrosion 162
5.4 Carbon Equivalent 165
5.5 Hydrogen-Induced Stress Cracking (HISC) Corrosion 167
5.5.1 HISC and Vulnerable Materials 168
5.5.2 HISC and Stress 168
5.5.3 HISC and Cathodic Protection 168
5.5.4 HISC and DNV Standard 169
Further Reading 184
6 Noise 185
6.1 Introduction to Sound 185
6.2 Introduction to Noise 186
6.3 Noise in Industrial Valves 189
6.3.1 Mechanical Noise and Vibration 190
6.3.2 Fluid Noise 190
6.3.2.1 Aerodynamic Noise 191
6.3.2.2 Hydrodynamic Noise 191
6.3.3 Noise Control Strategies 191
6.4 Noise Calculations for Pipes and Valves 192
6.4.1 Acoustic Fatigue Analysis 192
6.4.1.1 Sound Power Level Calculations 193
6.4.1.2 Mach Number 198
6.4.2 Noise in Control Valves 203
6.4.2.1 Aerodynamic Noise in Control Valves 203
6.4.2.2 Hydrodynamic Noise in Control Valves 208
6.4.3 Noise in Pressure Safety or Relief Valves 215
6.4.3.1 Calculation of Noise Emission According to ISO 4126-9 216
6.4.3.2 Calculation of Noise Emission According to API 521 218
6.4.3.3 Calculation of Noise Emission According to VDI 2713 221
Further Reading 231
7 Water Hammering 233

7.2 Water Hammering and Pressure Loss in Check Valves 233
7.3 Water Hammering Calculations 243
Further Reading 256
viii Contents
8 Safety Valves 258

8.2 Safety Valve Parts 259
8.3 Safety Valve Design and Operation 259
8.3.1 Design and Operation Parameters 259
8.3.1.1 Overpressure Criteria 277
8.3.2 Principle of Operation 278
8.3.3 Safety Valve Reaction Forces 282
8.3.4 Safety Valve Capacity Conversion 294
Further Reading 302
9 Safety and Reliability 304

9.2 Safety Standards 305
9.3 Risk Analysis 308
9.4 Basic Safety and Reliability Concepts 312
9.4.1 System Incidents and Failures 312
9.4.1.1 Failure Rate 313
9.4.1.2 Repair Rate 317
9.4.1.3 Mean Time to Failure (MTTF) 317
9.4.1.4 Mean Time Between Failure (MTBF) 318
9.4.1.5 Mean Time to Repair and Recovery (MTTR) 319
9.4.1.6 Mean Time to Detection (MTTD) 319
9.4.2 Reliability and Unreliability 319
9.4.3 Availability and Unavailability 331
9.5 Safety Integrity Level (SIL) Calculations 336
9.5.1 SIL 336
9.5.2 Probability of Failure on Demand (PFD) 338
9.5.3 Mean Downtime 339
9.5.4 Diagnostic Coverage 342
9.5.5 Safe Failure Fraction (SFF) 342
9.6 Condition Monitoring (ValveWatch) 347
Further Reading 354
10 Valve Operation 357

10.2 Valve Torque 358
10.3 Stem Design 363
10.3.1 MAST Calculations 363
Contents ix
10.3.2 Buckling Prevention 369

10.3.3 Torsional Deflection Prevention 374
10.3.4 MAST Limitation for Quarter-Turn Cryogenic Valves 376
Further Reading 384
11 Miscellaneous 385
11.2 Joint Efficiency 386
11.2.1 Weld Joint Efficiency 386
11.2.2 Bolted Joint Efficiency 388
11.2.2.1 Bolted Bonnet or Cover Joints 388
11.2.2.2 Bolted Body Joints 392
11.2.3 Threaded Joint Efficiency 394
11.2.3.1 Threaded Bonnet or Cover Joints 394
11.2.3.2 Threaded Body Joints 395
11.3 Stem Sealing 395
Further Reading 405
Index 407
1
Flow Capacity
1.1 Introduction
Valve flow capacity, also called flow coefficient or capacity index, is defined as the
valve’s capacity for a liquid or gas to flow through it. The flow coefficient is tech-
nically defined as the flow rate of water in US gallons per minute at the temper-
ature of 60 F with a pressure drop of 1 psi across the valve. Flow coefficient, as
shown with parameter Cv, increases by opening the valve to the maximum value
when the valve is 100% open. Pressure drop, also called differential pressure, is
defined as the difference between the inlet and outlet of the valve; for a valve with
a specific size, the greater the differential pressure, the greater the flow rate. These
two parameters, flow rate and differential pressure, are brought together by a flow
coefficient, which allows the performances of different valves regarding flow rates
to be compared. The differential pressure across the valves is calculated by know-
ing the flow rates, and finally, the flow rates are determined for given differential
pressure values. Furthermore, the flow coefficient is an essential parameter for
sizing valves. The flow coefficient (Cv) for liquids is calculated from Eq. (1.1).
Flow Coefficient (Cv) Calculation for Liquids
SG
Cv = Q 11
ΔP
where:
Cv: Flow coefficient (US gal per minute [gpm]/psi);

Q: Flow rate (US gpm);
SG: Fluid-specific gravity (dimensionless) calculated by Eq. (1.2);
ΔP: Differential pressure across the valve (pound per square inch [psi]).
Note: Water-specific gravity is equal to one.
Industrial Valves: Calculations for Design, Manufacturing, Operation, and Safety Decisions,
First Edition. Karan Sotoodeh.
© 2023 John Wiley & Sons, Inc. Published 2023 by John Wiley & Sons, Inc.
2 1 Flow Capacity
Specific Gravity (SG) Calculation
ρfluid
SG = 12
ρwater
where:
ρfluid: Density of the fluid passing through the valve;

ρwater: Water density equal to 1 g/cm3 or 1000 kg/m3.
To summarize, the flow coefficient calculated by Eq. (1.1) is a parameter

obtained by experiments that indicate the flow capacity of incompressible fluid
(liquid) during nonchoked and nonflashing flow conditions with a unit of differ-
ential pressure across the valve. Choked flow is fluid with a limited mass flow rate
due to restrictions like nozzles, orifices, and reducers. Flashing flow refers to the
significant evaporation of liquid due to a dramatic pressure drop.
Example 1.1 Calculate the flow coefficient of a control valve that passes 20 US
gpm of water with a 4 psi pressure drop in a fully open position.
Answer
SG 1
Cv = Q Cv = 20 = 10 gpm psi
ΔP 4
Two flow coefficients are commonly used in theory and practice: one American
flow coefficient (Cv) and the other metric (Kv) also called the flow factor. The met-
ric version of the flow coefficient or flow factor offers the measurement in terms of
cubic meters per hour (m3/h) of flow in a temperature range of 5–30 C for a dif-
ferential pressure of 1 bar. The flow factor (Kv) for liquids is calculated by Eq. (1.3).
Flow Factor (Kv) Calculation for Liquids
SG
Kv = Q 13
ΔP
where:
Kv: Flow coefficient (m3/h bar);

Q: Flow rate (m3/h);
SG: Fluid-specific gravity (dimensionless) calculated through the use of Eq. (1.2);
ΔP: Differential pressure across the valve (bar).
1.1 Introduction 3
The flow coefficient (Cv) for air and gases is calculated based on either Eq. (1.4)
or Eq. (1.5), depending on the criticality of the pressure drop. A critical pressure
drop indicates an outlet pressure (PO) that is less than or equal to 50% of the inlet
pressure (Pi). A noncritical or subcritical pressure drop is where the outlet pressure
is higher than 50% of the inlet pressure.
Flow Coefficient Calculation for Air and Gases (Critical Pressure Drop)
SG T + 460 PO
Cv = Q ≤05 14
660Pi Pi
where:
Cv: Flow coefficient (Standard cubic feet per minute [SCFM]/psi);

Q: Gas flow rate (Standard cubic feet per minute [SCFM]);
SG: Specific gravity of flowing gas relative to air at 14.7 psi and 60 F (Note 1);
T: Flowing air or gas temperature ( F); and
Pi: Inlet gas absolute pressure (psia) (Note 2).
Note 1
The specific gravity of some common gases can be found in Table 1.1. SG of air
is equal to one.
Table 1.1 Specific gravity of some common gases.
Gas Specific gravity
Acetylene (ethyne) – C2H2 0.90

Air (Note 1) 1
Alcohol vapor 1.601
Ammonia – NH3 0.59
Argon – Ar 1.38
Benzene – C6H6 2.6961
Blast furnace gas 1.02
Butadiene – C4H6 1.87
(Continued)
4 1 Flow Capacity
Table 1.1 (Continued)
Butane – C4H10 2.0061

I-Butene (butylene) – C4H8 1.94
Isobutene – C4H8 1.94
Carbon dioxide – CO2 1.5189
Carbon monoxide – CO 0.9967
Chlorine – CL2 2.486
Digestive gas (sewage or biogas) 0.8
Ethane – C2H6 1.0378
Ether vapor 2.586
Ethylene (ethene) – C2H4 0.9683
Fluorine 1.31
Helium – He 0.138
Heptanes 3.459
Hexane 2.973
Hydrogen 0.0696
Hydrogen chloride – HCL 1.268
Hydrogen sulfide – H2S 1.1763
Hydrofluoric acid 2.370
Hydrochloric acid 1.261
Isobutane 2.01
Isopentane 2.48
Krypton 2.89
Mercury vapor 6.940
Methane – CH4 0.565
Natural gas 0.6–0.7
Neon 0.697
Nitric oxide – NO 1.037
Nitrogen – N2 (pure) 0.9669
Nitrogen – N2 (atmospheric) 0.9723
Nitrous oxide – N2O 1.530
Nonane 4.428
Octane 3.944
Oxygen – O2 1.1044
1.1 Introduction 5
Ozone 1.660
Pentane 2.487
Propane – C3H8 1.5219
Sulfur dioxide – SO2 2.264
Water gas 0.71
Water vapor 0.6218
Xenon 4.53
Note 2
Absolute pressure measurement considers zero as the reference point, and it is
the total pressure of gauge and atmospheric pressure values. Atmospheric
pressure represents the pressure within the earth’s atmosphere.
Flow Coefficient Calculation for Air and Gases (Noncritical

Pressure Drop)
Q SG T + 460 PO
Cv = >05 15
1360 ΔP × PO Pi
where:
PO: Outlet gas absolute pressure (psia);

ΔP: Differential pressure between the inlet and outlet (psia) equal to Pi − PO;
T: Fluid temperature ( F).
Example 1.2 Air with a pressure of 1000 psia is flowing through a globe valve
that reduces the pressure to 400 psia. Considering the flow of air through the valve
is 400 SCFM, and its temperature is 30 C, what is the valve’s flow coefficient in
SCFM/psi?
Answer
Figure 1.1 is a schematic of the valve, inlet, and outlet air including values of
pressure, flow rate, and specific gravity.
6 1 Flow Capacity
Air Gas inlet Gas outlet
P1 P2
SG
QG 400 psia
1000 psia
1.0
400
SCFM
Figure 1.1 A globe valve schematic including inlet and outlet air service, values of
pressure, and flow rate.
The service is air, and the outlet pressure is less than half of the inlet pressure
indicating a critical pressure drop, so the flow coefficient is calculated using
Eq. (1.4). The air temperature is 30 C, which equals to 86 F.
SG T + 460 1 × 86 + 460
Cv = Q = 400 = 0 014 SCFM psi
660Pi 660 × 1000
Example 1.3 Methane (CH4) with a flow rate of 100 SCFM and a pressure of 140
psia is passing through a control valve, which reduces the pressure to 110 psia.
Assuming the gas temperature is equal to 60 F, calculate the flow coefficient Cv.
Answer
The outlet pressure is 78.5% of the inlet pressure, so Eq. (1.5) for gases with non-
critical pressure drop should be used. The specific gravity of methane is 0.565, per
Table 1.1.
Q SG T + 460 100 0 565 60 + 460

Cv = =
1360 ΔP × PO 1360 30 × 110
= 0 0735 × 0 2984 = 0 022 SCFM psi
The flow factor (Kv) for air and gases is calculated based on either Eq. (1.6)
or Eq. (1.7), depending on the criticality of the flow and pressure drop.
A supercritical flow or pressure drop indicates an outlet pressure (PO) that is less
than or equal to 50% of the inlet pressure (Pi). A noncritical pressure drop is where
the outlet pressure is higher than 50% of the inlet pressure.
1.1 Introduction 7
Flow Factor Calculation for Air and Gases (Critical Pressure Drop)
Q PO
Kv = ρ × Ti ≤05 16
257pi Pi
where:
Kv: Flow factor (m3/h bar);

Q: Gas flow rate (Standard cubic meter per hour [m3/h]);
Pi: Inlet gas absolute pressure (bara);
ρ: Gas density (kg/m3);
Ti: Flowing air or gas inlet temperature ( K).
Flow Factor Calculation for Air and Gases (Noncritical Pressure Drop)
Q ρ × Ti PO
Kv = >05 17
514 ΔP × PO Pi
where:
PO: Outlet gas absolute pressure (bara);

ΔP: Pressure loss in valve (bara).
Example 1.4 Isobutane with temperature and pressure values of 20 C and

10 bar, respectively, passes through a control valve where the pressure is reduced
to 2 bar. Calculate the flow factor of the valve, assuming the flow rate of isobutane
is 2000 m3/h.
Answer
PO 2
= = 0.2 < 0.5 It is a critical pressure drop, so the flow factor is calculated
Pi 10
according to Eq. (1.6). The specific gravity of isobutane is 2.01 (refer to Table 1.1).
Considering the air density equal to 1.225 kg/m3, the isobutane density is calcu-
lated as follows:
ρisobutane ρisobutane
SGisobutane = 2 01 = ρisobutane = 2 46 kg m3
ρair 1 225
The isobutane temperature is converted from C to K as follows:
T i K = T i C + 273 15 = 20 + 273 15 = 293 15 K
Q 2000
Kv = ρ × Ti = 2 46 × 293 15 = 20 90 m3 h bar
257pi 257 × 10
8 1 Flow Capacity
Conversions between the flow coefficient and flow factor are calculated from
Eqs. (1.8) and (1.9).
Relationship Between Flow Coefficient and Factor
C v = 1 16 × K v 18
K v = 0 853 × C v 19
Example 1.5 Calculate the flow coefficient (Cv) for the control valve in
Example 1.4.
Answer
C v = 1 16 × K v = 1 16 × 20 90 = 24 24 US gpm psi
1.2 Flow Coefficient Chart and Flow Curve
A flow coefficient chart or flow curve shows the relationship between the opening
percentage and the amount or percentage of Cv or flow rate. Figure 1.2 is a flow
characteristics chart or curve showing the relationship between the opening ratio
of valves and the percentage of maximum flow or flow coefficient. Three flow
Inherent control valve characteristics

100
90
80
Percent of maximum flow, Cv
70
Quick opening
60
50
Linear
40
30
20
Equal percentage
10
0
0 10 20 30 40 50 60 70 80 90 100
Percent of rated travel
Figure 1.2 Flow characteristics of valves.

curves or characteristics are shown in the figure: quick opening, linear, and equal
percentage. Quick opening means that a large amount of flow passes through the
valve as it begins to open. A small valve opening, especially in the beginning, can
lead to a high amount of flow. Linear flow means that the change in the flow is
equal to the opening percentage. For example, 40% of opening a valve passes
40% of the fluid through the valve. An equal percentage means that the valve
releases a relatively low amount of fluid at the beginning of the opening stage.
For example, a valve that is getting opened by 35% to just start releasing the flow
has an equal percentage flow characteristic. Control valves have typically equal
percentage and linear flow characteristics, and the quick opening flow character-
istic is most commonly used for on/off valves.
Example 1.6 A control valve with a linear flow characteristic has a flow coeffi-
cient of 700 at a fully open position. Calculate the Cv value of the valve at 5% and
95% opening percentages.
Answer
Flow rates and Cv values are increased linearly by increasing the valve opening
percentage for a linear flow characteristic. Thus, values of flow rate and flow coef-
ficient at any valve opening percentage for a linear flow characteristic are calcu-
lated according to Eqs. (1.10) and (1.11), respectively.
Valve Flow Rate Calculation Based on Valve Opening Percentage for a

Linear Flow Characteristic
qvp = qmax × vp 1 10
where:
qvp: Valve flow rate linked to the valve opening percentage;

qmax: Maximum flow rate passing through the valve at the fully open position;
vp: Valve opening percentage.
Valve Flow Coefficient Calculation Based on Valve Opening Percentage

for a Linear Flow Characteristic
C vvp = C vmax × vp 1 11
where:
Cvvp: Valve flow coefficient linked to the valve opening percentage;

Cvmax: Maximum flow coefficient of the valve at the fully open position.
10 1 Flow Capacity
C vvp = Cvmax × vp C v5 = 700 × 0 05 = 35 and

C v95 = 700 × 0 95 = 665
Therefore, the flow coefficients of the valve at 5% and 95% opening percentages
are 35 and 665, respectively.
Example 1.7 A control valve with an equal percentage flow characteristic has a
flow coefficient of 640 at a fully open position. Calculate the Cv value of the valve at
10% and 90% opening percentages, respectively.
Answer
Flow rates and Cv values are not increased linearly by increasing the valve opening
percentage for an equal percentage characteristic. Thus, values of flow rate and
flow coefficient at any valve opening percentage for an equal percentage flow char-
acteristic are calculated according to Eqs. (1.12) and (1.13), respectively.
Valve Flow Rate Calculation Based on Valve Opening Percentage for an

Equal Percentage Flow Characteristic
qvp = qmax × αvp−1 1 12
where:
α = 50
Valve Flow Coefficient Calculation Based on Valve Opening Percentage

for an Equal Percentage Flow Characteristic
C vvp = C vmax × αvp − 1 1 13
where:
Cvvp = Cvmax × αvp−1 Cv10 = 640 × 500 1 − 1 = 640 × 0 02957 = 18 93

Cvvp = Cvmax × αvp−1 Cv90 = 640 × 500 9 − 1 = 640 × 0 676 = 432 80
Therefore, the flow coefficients of the valve at 10% and 90% opening percentages
are 18.93 and 432.80, respectively.
Valves are divided into two major categories based on the closure member
motion: linear and quarter-turn. Linear motion valves set off have a closure mem-
ber that moves up and down with a linear motion to change the flow rate through
the valve. Quarter-turn like ball and butterfly valves in which a closure member
and the stem rotate 90 between the open and closed positions. The valve flow
coefficient chart or flow characteristic for a quarter-turn valve could be provided
as a relationship between the flow rate or Cv and the angle of the stem or closure
member. It should be noted that the angle of stem and closure member for quarter-
turn valves is equal. For example, Figure 1.3 illustrates a couple of flow coefficient
charts or curves for a 38 CL1500 (equivalent to 250 bar nominal pressure) ball
valve. The chart here shows the relationship between the angle of stem and flow
coefficient, whereas the lower chart illustrates the connection between the opening
percentage and the flow coefficient of the valve. Even when the stem angle is 33 ,
no flow is passing through the valve, meaning that the valve has an equal percent-
age flow characteristic. In light of the fact that 90 rotation of the stem and closure
member corresponds to 100% or full valve opening, using interpolation, a 33 stem
angle corresponds to 36.6% opening. The full open Cv of the valve is the one typ-
ically provided by valve manufacturers on general arrangement (GA) valve draw-
ings, which is 118,179 gpm in this case. GA valve drawings deliver the dimensions
Tables of estimated Cv values

Flow Estimated stem Estimated
ratio (%) angle (°) Cv Full open Cv 118,179
0 33 3
5 36 666 1,40,000
Cv = f (stem angle)
10 39 1750
1,20,000
15 42 3169
1,00,000
20 44 4797
80,000
25 47 6756
60,000
30 50 8938
35 53 11,543 40,000
40 56 14,454 20,000
45 59 17,964
0
50 62 21,945 00° 10° 20° 30° 40° 50° 60° 70° 80° 90°
55 64 26,836 1,40,000
Cv = f (%flow ratio)
60 67 32,496
1,20,000
65 70 39,595
1,00,000
70 73 47,962
80,000
75 76 58,557
80 79 70,896 60,000
85 81 85,623 40,000
90 84 1,00,306 20,000
95 87 1,12,664
0
100 90 1,18,179 0% 20% 40% 60% 80% 100%
Figure 1.3 Flow curves for a 38 CL1500 ball valve.

12 1 Flow Capacity
Table 1.2 Flow coefficient (Cv) values in gpm/psi for gate, globe, and check valves
in the size range of 1/4 –3 from one manufacturer.
Reduced Full Piston Ball Swing

NPS DN port gate port gate Globe check check check
1/4 8 1.7 1.7 1 0.9 0.8

3/8 10 4.2 5.7 1.2 1 0.9
1/2 15 5.7 8.2 1.5 1.3 1.2 11
3/4 20 8.2 22 2.4 2.1 1.9 17.6
1 25 26 34 5.6 5 4.4 31.4
1 1/4 32 37 60 15
1 1/2 40 60 92 21 12.6 10.5 57.6
2 50 92 200 29 17.4 14.5 80.1
3 80 200
of main valve attributions, such as face-to-face or end-to-end, and valve height.

Process engineers require Cv values to ensure that valves provide expected flow
capacities. It is important to know that the flow coefficient depends on many factors
such as the size and type of valves, type of design (e.g. full bore or reduced bore), and
valve manufacturer. Table 1.2 provides Cv values in gpm/psi for gate, globe, and
check valves in the size range of 1/4 –3 from one manufacturer. Gate and globe
valves are used for stop/start and flow regulation, respectively. Check valves are
nonreturn valves that are opened by the fluid flow in one direction and prevent
the flow from returning to the upstream side of the valve. NPS and DN in
Table 1.2 represent nominal pipe size and dimension or diameter nominal, respec-
tively. NPS and DN are two ways to express the size of the piping and valves. DN is
expressed in metric units of a millimeter. For example, DN25 and DN50 imply
dimensions/diameters of 25 mm and 50 mm, respectively. DN is the metric equiv-
alent of NPS, and DN is calculated by multiplying the pipe or valve size in inches
(NPS) by 25. For example, a 1 NPS pipe or valve corresponds to DN25.
1.3 Rangeability and Turndown
Rangeability is another flow rate definition for industrial valves used for valve siz-
ing. Rangeability is relevant for the valves used for flow control or throttling, such
as control, globe, butterfly, plug, and V-notch ball valves. A valve with higher flow
1.3 Rangeability and Turndown 13
rangeability enables control of the flow over the wider flow range. A control valve
is an actuated globe valve used to control the fluid by altering the size of the fluid
passage. Direct flow rate control leads to changes in major process quantities such
as pressure, temperature, and liquid level. An actuator is a mechanical or electrical
device installed on top of industrial valves to operate (open and close) the valves
automatically without any need for an operator. Rangeability is defined as the ratio
of maximum flow to a minimum controllable flow of a valve installed in a piping
system, which is calculated by Eq. (1.14). Thus, if the minimum controllable
flow is 10% of the maximum controllable flow, the rangeability of the valve is
100/10 = 10. It is important to note that the minimum controllable flow is not
the flow leakage through the valve when it is closed.
Valves are often not required to handle the maximum controllable flow; as an
alternative, the maximum operation flow passes through the valve. Thus, the term
turndown is equal to or smaller than rangeability and is calculated according to
Eq. (1.15). For example, the valve for which the rangeability is calculated may need
to have a maximum normal operation flow equal to only 70% of the maximum
controllable flow, so the turndown is 70/10 = 7.
Rangeability Calculation
Maximum controllable flow

R= 1 14
Minimum controllable flow
Turndown Calculation
Maximum normal operating flow

T= 1 15
Minimum controllable flow
Example 1.8 The maximum and minimum controllable flows by a control valve
are at 95% and 5% of their full opening positions, respectively. Calculate the valve
rangeability in two cases of the linear and equal percentage flow characteristics.
Answer
Maximum controllable flow Flow at 95 valve position
R= =
Minimum controllable flow Flow at 5 valve position
95 qmax 95
R = Linear flow characteristic = = = 19
5 qmax 5
qmax × α0 95 − 1 500 95 − 1
R = Equal percentage flow characteristic = =
qmax × α 0 05 − 1 500 05 − 1
0 8223
= = 33 84
0 0243
14 1 Flow Capacity
Example 1.9 The flow coefficient of a valve in the fully open position is 640. The
pressure drop of the fluid service, which is water, is 1 psi when the valve is fully
open. Calculate the valve rangeability, assuming that the valve’s maximum and
minimum controllable flows are adjusted at 90% and 10% of their full opening posi-
tions, respectively, and the valve flow characteristic is linear.
Answer
SG 1
Cv = Q 640 = Q Q = qmax = 640 gpm
ΔP 1
10 opening position qvp = qmax × vp = 640 × 0 1 = 64 gpm
90 opening position qvp = qmax × vp = 640 × 0 9 = 576 gpm
Maximum controllable flow Flow at 90 valve position
R= =
Minimum controllable flow Flow at 10 valve position
576
= =9
64
1.4 Valve Authority
Valve authority is another identification for selecting and sizing the valves used for
flow control, such as control valves. It expresses the ratio between the pressure
drop across the valve used for flow control and the total pressure drop across
the whole system as per Eq. (1.16).
Valve Authority
ΔP1
N= 1 16
ΔP1 + ΔP2
where:
N: Valve authority (dimensionless);

ΔP1: Pressure loss in a valve in a fully open position (bara);
ΔP2: Pressure drop across the remainder of the system (circuit);
ΔP1+ΔP2: Pressure drop across the whole system (circuit)
The conventional definition of a control valve authority (N) measures how much
the system’s pressure drop is related to the control valve. A valve authority value
1.5 Valve Gain 15
lower than 0.25 indicates that the valve for fluid control has unstable fluid control
characteristics, so it is not recommended. Conversely, a valve authority value from
0.5 to 1 provides good to excellent fluid control but with a high pressure drop. So, it
results in high energy consumption, which is undesirable. A valve authority
between 0.25 and 0.5 provides fair to good control with a reasonable pressure drop.
The best engineering practice for control valve selection and sizing is to keep the
value of N near 0.5 but not greater than, e.g. 0.4.
Example 1.10 A system has a total pressure drop (ΔP1+ΔP2) of 1.25 bar. If the
control valve in the system has a valve authority (N) equal to 0.45, what is the pres-
sure drop across the valve? Assuming that the valve fluid is water with a flow rate
of 15 m3/h, what is the valve flow factor?
Answer
ΔP1 ΔP1
N= 0 45 =
ΔP1 + ΔP2 1 25
ΔP1 = 0 5625 Pressure drop across the valve
SG 1
Kv = Q K v = 15 = 20 m3 h bar ;
ΔP 0 5625
1.5 Valve Gain
Valve gain is the ratio of flow change to the valve closure member travel (stroke) or
valve opening percentage. This parameter is calculated from Eq. (1.17). Gain
should never be less than 0.5 to avoid any problem for the valve to control the flow.
The gain same as the pressure drop is constant in the linear valve, whereas it
increases as the valve opens in equal percentage flow characteristics.
Valve Gain
Flow rate change

Valve gain = 1 17
Valve opening percentage change
Example 1.11 Three flow rates of 35 gpm, 100 gpm, and 170 gpm correspond to
stroke or opening percentages of 42%, 69%, and 90%, respectively. Calculate gain
values, and can this valve perform flow control?
16 1 Flow Capacity
Table 1.3 Valve flow rates and corresponding opening percentages.
Flow (gpm) Stroke (%) Change in flow (gpm) Change in stroke (%)
35 42 100 − 35 = 65 69 − 42 = 27
100 69
170 90 170 − 100 = 70 90 − 69 = 21
Answer
All provided information, including the valve flow rates and corresponding
opening percentages, is summarized in Table 1.3:
Flow rate change #1 65
Valve gain #1 = = = 2 41
Valve opening percentage #1 27
Flow rate change #2 70
Valve gain #2 = = = 3 33
Valve opening percentage #2 21
The difference between the two gains should be less than 50% of the higher
gain value.
3 33 – 2 41 = 0 92, 50 × 3 33 = 1 65
Since 0.92 is less than 1.65, there should be no problem using the valve for flow
control. Also, it should be noted that both gain values are larger than 0.5. Thus, the
selected valve can perform flow control nicely without any operational problems.
Questions and Answers
1.1 Which answer is correct for a given size of a valve?

A The greater the flow, the less pressure drop
B The greater the pressure drop, the greater the flow
C The lesser the flow, the greater the pressure drop
D The greater the flow, the lesser the pressure drop
Answer
Option B is the correct answer. Increasing the pressure drop or differential
pressure across a valve leads to a higher flow rate through the valve.
1.2 What statement is correct about the flow coefficient in the form of a
CV value?
A Cv is a theoretical value that an experiment cannot measure.
B Cv value can only be measured when the valve is fully open.
C Cv value is independent of the type of fluid passing through the valve.

D Metric units are not used for Cv value calculation.
Answer
Option A is incorrect because an experiment can measure Cv. Option B is not
correct either since Cv can be measured from fully closed to fully open at any
valve position. Option C is wrong because the type of fluid passing through
the valve affects the fluid’s density, specific gravity, and flow coefficient.
Option D is the correct answer.
1.3 Which sentence is correct about valve flow characteristics?

A Opening a valve by 30% leads to passing 30% of the total flow capacity of
the valve. This valve has a linear flow characteristic.
B A valve is opened 50%, while the fluid passage from the valve is just 20%
of the total flow. This valve has the flow characteristic of quick opening.
C If it is needed to have a small flow passage at the beginning of the opening
of the valve, a linear flow characteristic trim is required.
D A valve releases 25% of the total flow capacity at a 27% opening. This
valve has a quick opening characteristic.
Answer
Option A is correct, and the valve-opening percentage is equal to the percent-
age of the flow capacity. So the characteristic is linear. Option B explains an
equal percentage flow characteristic since the amount of flow at the half open-
ing is relatively low. Therefore, option B is not correct. Option C is wrong
because having a small flow at the beginning of the valve opening is an equal
percentage and not linear. Option D is very similar and close to the linear flow
characteristic. So it does not address the quick opening characteristic.
1.4 Water with a flow rate of 20 m3/h is circulated in a piping system, including
the valve. Calculate the pressure drop across the valve with a Kv value of 20.
A 0.5 bar
B 1 bar
C 2 bar
D 0.25 bar
Answer
The fluid service is water and liquid. So, Eq. (1.3) is applicable to fluid factor
(Kv) calculation as follows:
SG 1
Kv = Q 20 = 20 ΔP = 1
ΔP ΔP
Thus, option B is the correct answer.

18 1 Flow Capacity
1.5 Find the wrong statement about flow coefficient and flow factor.
A Conversion between the flow coefficient and flow factor gives a larger
value to the flow coefficient than a flow factor.
B The value of the flow coefficient on a valve GA drawing is not given for
the valve’s fully open position.
C The flow factor is the metric version of the flow coefficient.
D A flow coefficient of 2 US gpm/psi is equal to a flow factor of 1.706 m3/
h bar.
Answer
Except for option B, all options are correct since the flow coefficient values
on the GA drawings are provided for the valve’s fully open positions.
1.6 Find the correct choice about rangeability and turndown.

A The maximum normal operating flow is 66 gpm, and the minimum con-
trollable flow is 3 gpm. The rangeability is equal to 22 in this case.
B Turndown is always smaller than rangeability.
C If the minimum controllable flow is 2% of the maximum controllable flow,
the turndown is 50.
D It is impossible to have a flow rangeability of 1 : 1 for a control valve.
Answer
Option A is incorrect because 22 is turndown and not rangeability. Option
B is not correct because turndown is equal to or smaller than rangeability.
Option C is wrong because the value of 50 is for rangeability. Option D is
the correct answer since globe valves usually provide a flow rate from 5%
to 95%, giving a rangeability of 1 : 19.
1.7 Which parameter is related to the valve flow factor?

A Cv
B Kv
C Av
D SG
Answer
Option B is the correct answer.
1.8 Figure 1.4 illustrates a couple of flow coefficient charts or curves for a 20
CL1500 (equivalent to 250 bars’ nominal pressure) ball valve. Which state-
ment is correct for the valve in this example?
A The valve has a quick opening flow characteristic.
B The maximum flow rate passing through the valve in the fully open con-
dition is 22,970 US gpm.
Tables of estimated Cv values

Flow Estimated Estimated
ratio (%) stem angle (°) Cv Full open Cv 22,970
0 33 1
5 36 166 25,000
Cv = f (stem angle)
10 39 430
20,000
15 41 775
20 44 1171 15,000
25 47 1645
30 50 2192 10,000
35 53 2798
5000
40 56 3521
45 59 4328
0
50 61 5301 00° 10° 20° 30° 40° 50° 60° 70° 80° 90°
55 64 6443 25,000
Cv = f (%flow ratio)
60 67 7740
20,000
65 70 9325
70 73 11,126
15,000
75 76 13,291
80 79 15,725 10,000
85 81 18,186
5000
90 84 20,537
95 87 22,202
0
100 90 22,970 0% 20% 40% 60% 80% 100%
Figure 1.4 Flow curves for a 20 CL1500 ball valve.
C The flow coefficient value on the valve GA drawing is 22,970 US gpm/psi.

D Half opening position of the valve provides a flow coefficient equal to
2192 US gpm/psi.
Answer
Option A is incorrect because the flow curves indicate an equal percentage
flow characteristic. An equal percentage implies that the valve discharges a
relatively low amount of fluid at the start of the opening stage. The curve on
the top illustrates the relationship between the stem angle and flow coeffi-
cient, demonstrating that the flow starts passing through the valve after
the stem angle of 33 , equal to 36% valve opening approximately. Option
B is not correct either because there is no information available about the
type of fluid and its specific gravity as well as the pressure drop across the
valve. So, it is not possible to convert the flow coefficient to the flow rate.
If the fluid is water and the pressure drop is one psi, option B would be correct.
Option C is correct because valve manufacturers provide the value of the flow
20 1 Flow Capacity
coefficient in the valve’s fully open position on the valve’s GA drawings.

Option D is not correct because a 50 stem angle corresponds to a 55.55 valve
opening percentage, giving a flow coefficient of 2192 US gpm/psi. Thus,
option C is the right answer.
1.9 What is the optimum range of valve authority?

A 0.1–0.25
B 0.25–1.0
C 0.5–0.75
D 0.25–0.5
Answer
Option D is the correct answer.
1.10 Calculate the valve authority if the pressure drop across the control valve
is 40 kPa and the total pressure drop in the remainder of the system is
60 kPa.
A 0.4
B 0.33
C 1.5
D 0.25
Answer
ΔP1 40
N= = =04
ΔP1 + ΔP2 40 + 60
where:
N: Valve authority (dimensionless);
ΔP1: Pressure loss in a valve in a fully open position (bara);
ΔP2: Pressure drop across the remainder of the system (circuit);
ΔP1+ΔP2: Pressure drop across the whole system (circuit).
Thus, option A is the correct answer.
Further Reading
Bahadori, A. (2014). Natural Gas Processing, 1e. Austin, TX: Elsevier (Gulf Professional
Publishing).
Bolton, W. (2021). Instrumentation and Control Systems, 3e. Austin, TX: Elsevier (Gulf
Professional Publishing).
Boyes, W. (2010). Instrumentation Reference Book, 4e. Oxford: Butterworth-
Heinemann.
Further Reading 21
Nesbitt, B. (2007). Handbook of Valves and Actuators: Valves Manual International,

1e. Oxford: Elsevier.
Skousen, P.L. (2011). Valve Handbook, 3e. New York: McGraw-Hill.
Smit, P. and Zappe, R.W. (2004). Valve Selection Handbook, 5e. New York: Elsevier.
Sotoodeh, K. (2021). Subsea Valves and Actuators for the Oil and Gas Industry,
1e. Austin, TX: Elsevier (Gulf Professional Publishing).
22
Valve Sizing
2.1 Introduction
Valve sizing is an essential engineering task that must be done accurately since
improper valve sizing is both expensive and inconvenient. If a valve is too small,
it cannot pass the required flow and compromises flow assurance and process
integrity. On the other hand, an oversized valve increases the cost. Valve sizing
is typically conducted based on both theoretical and experimental methods: Valve
sizing is split based on the valve’s applications, such as isolation valves, nonreturn
or check valves, control valves, and safety valves.
2.2 Isolation Valve Sizing
A ball, plug, or gate valve that stops and starts fluids normally has the same size as
the pipe it is connected to. A 3 isolation valve, for example, is typically connected
to a 3 pipe. It is possible for ball valves to have a full or reduced bore, and the bore
is defined by the size of the hole inside the ball. Full-bore ball valves in 3-in. sizes
have a nominal pipe size (NPS) of 3 . Reduced bore ball valves in 3 in. sizes have a
NPS of 3 × 2 . As an example, a 3 refers to the size of the valve end connected to
the piping, whereas a 2 refers to the size of the valve bore. Engineers intend to use
a reduced bore ball valve as a first choice because of its lower cost unless the
reduced bore causes flow assurance problems or undesirably very high-pressure
drop. As an example, ball valves on subflare lines which are installed before
and after pressure safety valves (PSVs) shall be a full bore to allow a sudden release
of fluid from the piping system to the flare line upon overpressurization. Further,
some ball valves are subject to the piping injected gadget (PIG) running for clean-
ing or maintenance. So these valves must have an internal diameter equal to the
internal diameter of the connected piping. Using the following example, it is

possible to look at how to select a ball valve for stopping/starting the flow in a
piping system. This will satisfy the process engineers’ need to set the flow rate.
Example 2.1 A ball valve is installed on the water piping that is 4 in. in diameter
and 60 F. According to the process engineer, the valve has a flow rate of 8000 US
gallons per minute (gpm). Valve inlet and outlet pressures are 600 psi and 585 psi,
respectively. Tables 2.1 and 2.2 provide the flow coefficient values provided by the
valve manufacturer for full bore and reduced bore ball valves in size ranges ran-
ging from 1/2 to 20 . Select the appropriate size for the valve.
Table 2.1 Flow coefficient (Cv) for full-bore ball valves (size ranges: 1/2 –18 ).
Valve size
in. 1/2 3/4 1 1.5 2 3 4 6 8

mm 12 19 25 37.5 50 75 100 150 200
Flow 26 50 94 560 480 1300 2300 5400 10,000
coefficient
(Cv)
in. 10 12 14 16 18
Mm 250 300 350 400 450
Flow 16,000 24,000 31,400 43,000 57,000
coefficient
(Cv)
Table 2.2 Flow coefficient (Cv) for reduced bore ball valves (size ranges: 3 –20 ).
Valve size
in. 3 4 6 8 10 12 14 16 18
mm 75 100 150 200 250 300 350 400 450
Flow coefficient 420 770 1800 2500 4500 8000 12,000 14,000 18,000
(Cv)
in. 20
mm 500
Flow coefficient 22,000
(Cv)
Ball valves in the size range of 2 and below are not reduced bore typically.
24 2 Valve Sizing
Answer
SG 1
Cv = Q Cv = 8000 = 2066 gpm psi
ΔP 15
The first ball valve choice is a reduced bore 4 × 3 ball valve, which is a
cheaper alternative. However, a 4 reduced bore ball valve has a flow coefficient
of 770 gpm/psi as per Table 1.2 which does not meet the required flow coefficient
of 2066 gpm/psi. Thus, a reduced bore ball valve is not appropriate in this case.
Alternatively, the flow coefficient of a 4 full-bore ball valve is 2300 gpm/psi,
which is compatible with the required flow coefficient. Thus, a 4 full-bore ball
valve is the best choice.
Butterfly valves have gained popularity for flow isolation since they are more
compact and lighter than gate and ball valves. A butterfly valve is a reduced bore
valve, which is helpful to know. Butterfly valves, which are quarter-turn valves,
have advantages over ball valves, gate valves, and plug valves, such as saving
weight, space, and costs, as well as the amount of torque required to open
and close a butterfly valve. A quarter-turn valve has a valve closure member
and stem that are rotated 90 between open and closed positions. A butterfly
valve may also have a wafer-type design or a flangeless design, which enables
the installation of the valve between two flanges. Based on the required flow
coefficient and the inlet and outlet pipe sizes, the following example shows
how to size a butterfly valve.
Example 2.2 Flow coefficient values for the 70 open and fully open positions of
butterfly valves are shown in Table 2.3 based on the inlet and outlet pipe sizes from
4 to 10 .
The crude oil at 60 F with a specific gravity of 0.8 and flow rate of 3500 US gpm
is passing through the butterfly valve at the fully open position. The butterfly
valve’s inlet and outlet pressure values are 10 psi and 7 psi, respectively. What
should be the size of the valve, inlet, and outlet pipe?
Answer
The advantage of the flow coefficient values shown in Table 2.3 is that they have
been corrected to account for the effects of piping geometry factors. The butterfly
valve Cv values in the table have been adjusted to account for pressure losses
caused by pipe fittings, such as reducers and expanders, which might be connected
directly to the butterfly valve’s inlet or outlet connections. The second point is that
in this example, the butterfly valve is used for flow isolation. Therefore, the second
column from the right, which shows the flow coefficient at 70 open, is for flow
control or throttling, and it would not be applicable here. The next
section calculates the required valve flow coefficient for a fully open position.
Table 2.3 Flow coefficient values for butterfly valves based on the inlet and outlet
pipe sizes from 4 to 10 .
Inlet Butterfly Outlet Flow coefficient in 70 Flow coefficient in fully

pipe size valve size pipe size open position open position
4 4 4 496 841
6 4 4 348 442
6 4 6 405 539
6 6 6 1025 1850
8 6 6 784 1016
8 6 8 913 1359
8 8 8 1862 3316
10 8 8 1462 1922
10 8 10 1711 2633
10 10 10 2948 5430
I Pipe Ø
Flow
6″ Ø Pipe 6″ Ø Pipe
6″ Valve
Figure 2.1 A 6 butterfly valve connected to a 6 pipe.
SG 08
Cv = Q Cv = 3500 = 1807 gpm psi
ΔP 3
According to Table 1.3, a 4 butterfly valve can provide a maximum of 841

gpm/psi; that is not sufficient. Furthermore, a 6 butterfly valve with an 8 inlet
pipe and 6 or 8 outlet pipe sizes only can give 1016 or 1359 gpm/psi flow coeffi-
cient values that are less than the required flow coefficient. An 8 butterfly valve
with both 8 inlet and outlet pipe sizes is counted as a valve and piping oversizing
scenario since it gives a very large flow coefficient value of 3316 gpm/psi. The best
solution, as shown in Figure 2.1, is to select a 6 butterfly valve with both 6 inlet
and outlet pipe sizes to provide a Cv value of 1850 gpm/psi.
26 2 Valve Sizing
2.3 Nonreturn (Check) Valve Sizing
Check valves are automatic valves that open with the forward flow and close with
the reverse flow. Among the different types of check valves are swing check valves,
lift check valves, dual plate check valves, etc. Check valves are sometimes installed
after mechanical equipment, such as pumps and compressors, to prevent fluid
reversal. Check valves are also designed to prevent massive backflow when a pipe
breaks and to prevent backflow toward lower pressures. In contrast to a ball, but-
terfly, or gate valve, a check valve does not require an operator, and it is operated
automatically by the fluid inside the pipe. Figure 2.2 illustrates a swing check valve
at the top and a piston check valve at the bottom. In this case, the closure member
swings around the hinge, which is mounted in the seat. As illustrated in Figure 2.2,
top right, a swing check valve is illustrated with the disk moving away from the
seat by the force of the fluid. However, the reverse flow on the top left is stopped
by the disk in the closed position. As shown in the figure, the piston check valves at
the bottom exhibit the same concept of flow movement and stoppage.
Swing check valves have been used widely in water and wastewater piping sys-
tems. These valves are available on the market at a low cost and provide low head
loss when they are open. The disk of the valve swings between 60 and 90
Figure 2.2 A swing check valve at the top and a piston check valve at the bottom.
Source: Sergey Merkulov/Shutterstock.
(fully open) during the passage of fluid. The main problem with this valve during
operation is that the disk slams against the seat during the closing of the valve. The
long stroke of the disk and the sudden closure of the valve because of the weight of
the disk exacerbate the slamming effect in swing check valves. Thus, it can con-
clude that a swing check valve is an economical option for a check valve with a
high slamming effect. The alternative solution to minimize the slamming impact
is to use a dual plate (see Figure 2.3) or an axial flow nozzle check valve
(see Figure 2.4). There are double disks instead of one for the dual plate check
valve. So each half-disk applies less weight force and slamming effect in its closing
action. In addition, the disks are closed by spring force rather than just relying on
their weight. The advantages of a dual plate check valve compared to a swing
check valve are not limited to space-saving and lower slamming rate. The total cost
of a dual plate check valve including the sum of initial, maintenance, and energy
costs are less than a swing check valve.
Plates
Open position
Closed position
Figure 2.3 Dual plate check valve highlighting double plates.
Bore reduction and venturi effect
Figure 2.4 Axial flow nozzle check valve.

28 2 Valve Sizing
There are different pieces of literature in which nozzle check valves are proposed
to be selected after rotating equipment such as pumps and compressors to protect
this expensive equipment from backflow damage. These highly engineered check
valves (see Figure 2.4) are specially designed to eliminate rotating equipment fail-
ure from backflow. The unique design of nozzle check valves minimizes the effect
of water hammering and chattering. In addition, this type of valve provides very
low-pressure drop as well as fast cycling (opening and closing).
It is a common practice to size a check valve in accordance with the pipe
diameter, as discussed for sizing isolation valves. In order to prevent high-
pressure drops and an increase in energy consumption inside the valve, it is
necessary to ensure that the minimum flow capacity maintains the valve disk
or closure member in the fully open position. During periods of sufficient flow,
the valve’s disk is continuously pushed upward and can be completely opened.
As the flow capacity decreases to a minimum and the valve cannot remain fully
open, the disk begins to vibrate. This causes repeated slams on the body and
wears out the valve. Therefore, unscheduled maintenance work may need to
be performed, or the valve may need to be replaced. The occurrence of chatter
or flutter occurs when the flow is insufficient to fully open the valve disk (e.g.
flow through the valve is less than the critical velocity). A valve that chatters or
flutters will eventually fail prematurely. Another disadvantage of a low flow
rate is that the flow cannot achieve the flow capacity and Cv value that the valve
can deliver because of high pressure. In the event that a check valve is too
small, it results in excessive pressure drop, reduced flow, and increased pump
energy requirement. On the other hand, a check valve sized too large may have
an unstable disk, resulting in valve wear, increased maintenance, or even fail-
ure of the valve.
In order to determine the check valve size, there are two key terms: the first is
the cracking pressure of the valve, and the second is the critical velocity. During
the initial opening of a check valve, for example, during the start-up of the pip-
ing system, the pressure applied by the flow upstream of the valve to the valve’s
disk or closure member shall overcome the force exerted by the spring and any
downstream backpressure acting on the back part of the disk. Cracking pres-
sure is the pressure differential that occurs when the upstream pressure over-
comes the spring force and other applicable forces and the valve disk cracks
open. A second significant factor is critical velocity, which is defined as the
required fluid velocity to keep the plates or disks of the valves fully open. In
addition, all check valves should be in a fully open position, and for spring-
loaded check valves such as dual-plate or axial-flow check valves, the force pro-
vided by the flowing fluid must be greater than the spring force. Equation (2.1)
is used to calculate the minimum flow necessary to fully open a swing
check valve.
Minimum Flow Velocity to Fully Open a Swing Check Valve

1
V min = 55 × √ 21
ρ
where:
Vmin: Minimum flow velocity (ft/s);
ρ : Fluid density (lb/ft3).
Example 2.3 A dual-plate check valve in 12 and Class 300 has been selected for
the gas service with a design temperature of 100 C and density of 5.87 kg/m3 in a
piping system with the same size as the valve. The pipe minimum and maximum
flow capacities are 954.4 m3/h and 7818.45 m3/h, respectively. Is the valve size
correct?
Answer
Calculation of the minimum and maximum velocities in the pipe is the first step.
The fluid velocity in the pipe is not uniform throughout the pipe. As a result, a
mean velocity is used, calculated using Eq. (2.2). The velocity of the fluid depends
on the cross-section area of the pipe and the liquid flow rate, as shown in Eq. (2.2)
and Figure 2.5.
Relationship Between the Flow Rate, Flow Velocity, and Cross-Section
Area of the Pipe
Q=A×V 22
where:
Q = liquid flow rate in pipe (m3/s);
A = Cross-section area of the pipe or channel (m2);
V = Mean fluid velocity in pipe (m/s).
The area of the pipe (A) is calculated using Eq. (2.3).
ΔV = AΔx
A
Continuous v
Q = Av Q = Av
pipe
Δx
Figure 2.5 Relationship between the flow rate, flow velocity, and cross-section area of the
pipe. Source: ScientificStock/Adobe Stock.
30 2 Valve Sizing
Pipe Cross-Section Area Calculation
2
ID
A=π 23
2
where:
ID = internal diameter of the pipe (mm).
The internal diameter of the pipe is calculated using Eq. (2.4).
Pipe Internal Diameter Calculation
ID = OD − 2 × t 24
where:
OD = outside diameter of the pipe (mm);
t = piping wall thickness (mm).
The outside diameter of a pipe depends on the NPS; in this case, the outside
diameter of a 12 pipe can be determined by referring to the American Society
of Mechanical Engineers (ASME) B36.10M or ASME B36.19M standard. Gener-
ally, the outside diameter of a pipe equals the NPS if the pipe is 14 and over.
According to ASME B36.19M standards, the outside diameter of a 12 pipe is
12.750 . As a general rule, pipe thickness is calculated based on the relevant ASME
code, such as ASME B31.3, process piping code. In the example, the pipe thickness
is 6.35 mm, so the pipe ID will be calculated as follows:
25 4 mm
ID = 12 750 in × − 2 × 6 35 = 323 85 – 12 7 = 311 15 mm
1 in
The next step is to calculate the pipe cross-section area as follows:

2 2
ID 311 15
A=π =π = 76,035 5 mm2 = 0 07603 m2
2 2
The next step is to calculate the minimum and maximum flow velocities in the
pipe using Eq. (2.2).
Qmax = A × V max 7818 45 m3 h = 0 07603 m2 × V max V max

= 1,02,834 m h = 28 56 m s
Qmin = A × V min 954 4 m3 h = 0 07603 m2 × V min V min

= 12,553 m h = 3 49 m s
The next step is to calculate the flow velocity in the valve using Eq. (2.5).
Flow Velocity Calculation Inside the Nozzle Check Valves

Apipe
V valve = V pipe × 25
Avalve
where:
Vvalve: Fluid velocity inside the valve;
Vpipe : Fluid velocity inside the pipe;
Apipe: Fluid area inside the pipe; and
Avalve: Fluid area inside the valve.
Assuming that the valve flow area is 70% of the pipe cross-section area, the min-
imum and maximum flow velocity values in the valve are calculated as follows:
Apipe 1
V valve min = V pipe min × = × 3 49 × = 4 99 m s
Avalve 07
Apipe 1
V valve max = V pipe max × = 28 56 × = 40 8 m s
Avalve 07
A critical velocity is equal to the minimum flow velocity required to completely
open the valve. The critical velocity of spring-loaded check valves, such as dual
plate and axial flow check valves, is dependent on the spring torque. A valve man-
ufacturer, for example, uses four categories of spring torque: Mini-torque, Low
Torque, Standard Torque, and Super Torque. Table 2.4 shows the critical velocity
associated with each spring torque. The critical velocity values are affected by var-
ious factors, such as the valve design, the fluid type, the spring torque value, or the
valve installation direction. Therefore, they differ from one manufacturer to
another.
In this example, the valve manufacturer has selected a low torque spring giving
the critical water flow velocity of 2 m/s. If a high or super torque spring is used, the
gas flows inside the piping cannot fully open the valve. The valve engineer eval-
uates the critical velocity values given in the table for a valve on the horizontal
Table 2.4 Minimum flow or critical velocity for the dual

plate check valve based on the spring torque in this case.
Spring torque Critical velocity (m/s)
Mini-torque 1.5
Low torque 2
High torque (standard) 3
Super torque 4.4
32 2 Valve Sizing
pipe. In general, valve manufacturers do not recommend installing the check

valves on the vertical lines with a downward flow because the combination of
valve disk weights plus the downward fluid can keep the valve always in the fully
open position. However, for the valve installed in a vertical line with the upward
flow, the critical velocity should be adjusted to be sufficient to overcome the weight
of the valve’s disks and the spring torque. The other important consideration is
that the values provided in Table 2.4 are relevant to the water flow. Equation (2.6)
shows the relationship between the critical water velocity from Table 2.4 and the
valve medium critical velocity.
Critical Velocity Conversion Between Water and the Medium

ρwater
V medium = V water 26
ρmedium
where:
Vmedium: Critical medium velocity or minimum medium velocity to fully open the
dual plate check valve (m/s);
Vwater: Critical water velocity or minimum water velocity to fully open the dual
plate check valve (m/s);
ρwater: Water density equal to 1000 kg/m3;
ρmedium: Medium density in kg/m3.
ρwater 1000
V medium = V water =2× = 26 10 m s
ρmedium 5 87
The minimum velocity to keep the valve in the fully open position is 26.10 m/s,
while the fluid velocity in the valve varies between 4.99 and 40.8 m/s. Thus, the
fluid inside the valve cannot always keep the valve in the fully open position.
The valve is probably oversized. So the solution is to reduce the valve size. This
solution can be coupled with selecting a lower spring torque to reduce the valve
flow restriction. The next example reviews the minimum flow requirement for
an axial flow check valve.
Example 2.4 An axial flow or nozzle check valve in 10 and Class 150 has been
selected for a 10 pipe containing the gas service with a design temperature of
97.5 C and density of 5.87 kg/m3. The valve’s flow rate is constant and equal to
31,263 kg/h, and the valve has a low torque spring (see Table 2.5). Assuming that
the flow area inside the valve is 75% of the pipe cross-section area, is the valve sized
correctly? (Note: Pipe ID is 264 mm.)
Table 2.5 Minimum flow or critical velocity for the axial

flow check valve in this case.
Spring torque Critical velocity (m/s)
Mini-torque 1.5
Low torque 2
High torque (standard) 2.5
Super torque 3
Answer
The critical water velocity for the axial valve considering the low torque spring is
equal to 2 m/s, according to Table 2.5. However, the fluid service, in this case, is a
gas. So the critical water velocity shall be converted to the critical gas velocity by
Eq. (2.6) as follows:
ρwater 1000
V medium = V water =2× = 26 10 m s
ρmedium 5 87
The minimum flow velocity to keep the check valve open is 26.10 m/s. The next
step is to calculate the minimum flow rate in kg/h, and compare it with the valve
normal flow rate, which is 25,263 kg/h.
The cross-section area of the pipe is calculated by Eq. (2.3) as follows:
2 2
ID 0 264
A=π = 3 14 = 0 055 m2
2 2
The area of the valve is 75% of the pipe cross-section area equal to
75% × 0.055% = 0.041 m2.
Now it is possible to calculate the minimum flow rate to keep the check valve
fully open by Eq. (2.2) as follows:
Q m3 s = V m s × A m2 Q = 26 10 m2 s × 0 055 m2
= 1 4355 m3 s × 3600 s h = 5168 m3 h × 5 87 kg m3
= 30,335 kg h
Any flow rate less than 30,335 kg/h cannot keep the axial flow check valve fully
open. The normal flow rate in the valve is constant and equal to 31,263 kg/h. So the
valve can always remain in the fully open position, and the valve is sized correctly
considering the minimum flow requirement.
34 2 Valve Sizing
2.4 Control Valve Sizing
Modern processing plants use many control loops (see Figure 2.6) to produce and
deliver a control product to the market. These control loops, including control
valves, are designed to keep process variables such as pressure, temperature, level,
flow, etc., within the required operating range and ensure that a quality product
has been produced. To avoid the effect of process variations on the desired set
point, sensors collect information from the process variable, and transmitters
transform the information to the controller (e.g. control room). A controller pro-
cesses the information and decides on the action to return the process variable to
the desired point. The most common final control element in process control is the
control valve.
A control valve (see Figure 2.7) is a type of instrument valve selected to control
the fluid flow by changing the fluid passage size and regulating the process vari-
able as close as possible to the desired set point. This section intends to provide a
new control valve size selection method coupled with real industrial examples.
Control valve size selection is divided into two categories for liquids and gases.
The next section explains the sizing of valves for liquids.
2.4.1 Control Valve Sizing for Liquids

There are six steps for a control valve size selection in liquids:
1) Specify the variables required to size the valve;

2) Determine the equation constant (N);
3) Determine the piping geometry factor (FP);
4) Determine the maximum flow rate upstream of the valve (qmax) and the allow-
able pressure drop (ΔPmax);
5) Solve for required flow coefficient (Cv) value using an appropriate equation;
6) Select the appropriate flow coefficient table and the calculated flow coefficient
valve to select the suitable valve size.
Manipulated Controlled Figure 2.6 A control loop.

variable variable
Process
Control Sensor
valve
Controller Transmitter
Figure 2.7 A control valve in the

plant. Source: Dale/Adobe Stock.
2.4.1.1 Specify the Variables Required to Size the Valve

The first step is to specify the variables required to size the valves, such as valve
pressure class, desired valve size, and design type. The pressure classes, which
ASME B16.34 covers, are 150 (PN20), 300 (PN50), 600 (PN100), 900 (PN150),
1500 (PN250), 2500 (PN420), and 4500 (PN720). “PN” stands for pressure nominal.
In addition to size and pressure class, the following process conditions shall be
determined by the process or valve engineer:
• Process fluid (e.g. oil, water);

• Service conditions such as volume rate of flow (q), mass rate of flow (w),
upstream and downstream pressure values (P1 and P2), pressure drop across
the valve (ΔP = P1 − P2), upstream temperature (T1), liquid specific gravity that
is a dimensionless number defined as the ratio of the liquid density at service
temperature to the density of water at 60 F (SGl), vapor pressure absolute of
liquid at inlet temperature (Pv), absolute thermodynamic critical pressure
(Pc), and fluid or gas velocity (v). All parameters for control valve size selection
are summarized in Table 2.6.
36 2 Valve Sizing
Table 2.6 Control valve sizing parameters and definitions.
Symbol Definition Symbol Definition
Cv Valve sizing coefficient P1 Upstream or inlet pressure

d Normal valve size P2 Downstream or outlet pressure
D Inlet diameter of the piping Pc Absolute thermodynamic critical
pressure
FD Valve style modifier, dimensionless Pv Absolute vapor pressure of the
liquid at inlet temperature
FF Liquid critical pressure ratio factor, ΔP Pressure drop across the valve
dimensionless (ΔP = P1 − P2)
Fk Ratio of specific heat factor, ΔPmax(L) Maximum allowable liquid sizing
dimensionless pressure drop
FL Rated liquid pressure recovery factor, ΔPmax(LP) Maximum allowable sizing
dimensionless pressure drop with attached fittings
FLP Combined liquid pressure recovery q Volume flow rate
factor and piping geometry factor of a
valve with attached fittings,
dimensionless (Note: if no fitting
attached, FLP = FL)
FP Piping geometry factor, dimensionless qmax Maximum flow rate at given valve
upstream condition
SGl Liquid-specific gravity that is a T1 Absolute upstream temperature
dimensionless number defined as the
ratio of the liquid density at service
temperature to the density of water at
60 F
SGg Gas-specific gravity that is defined as a W Mass rate of flow
dimensionless number equal to the
ratio of the flowing gas density to the
density of air with both at standard
conditions
k Ratio of specific heats, dimensionless X Ratio of pressure drop to the
upstream or inlet pressure (ΔP/P1)
K Head loss coefficient of a valve due to XT Rated pressure drop ratio factor,
installed fittings attached to the valve, dimensionless
dimensionless
M Molecular weight, dimensionless Y Expansion factor (ratio of flow
coefficient for a gas to that for a
liquid at the same Reynolds
number), dimensionless
N Numerical constant Z Gas compressibility factor,
dimensionless
γ Specific weight at the valve inlet μ Viscosity
condition (Note: The standard
condition is defined as 60 F and
14.7 psia)
2.4.1.2 Determine the Equation Constant (N)

In fact, many equations used in control valve sizing contain a numerical constant
known as parameter N. N is a numerical constant in each flow equation to provide
numeric values depending on the unit system used for calculations. Values for the
various constants and the applicable units are given in Table 2.7. For example, use
N1 if sizing the valve for a flow rate in volumetric units such as gallon per minute
(gpm) or NM3/h. Alternatively, use N6 if the valve is sized based on mass flow rate
units such as lb/h or kg/h.
2.4.1.3 Determine Piping Geometry Factor (FP)

FP is a correction factor that accounts for the pressure losses due to piping fittings
that are used to change the direction or size of the piping or take branches from the
headers such as tees, elbows, and reducers. If fittings are used before or after the
valve, the effect of piping geometry must be considered for the valve sizing. How-
ever, if no fitting is attached to the valve, FP is equal to one and does not impact the
valve-sizing process. Equation (2.7) shows how to calculate the piping geometry.
Piping Geometry Calculation

−1 2
K Cv 2
FP = 1 + 27
N2 d2
where:
N2: Numerical constant found in Table 2.7;

d: Assumed nominal valve size;
Cv: Valve-sizing coefficient.
In Eq. (2.7), the K is the sum of the velocity head loss coefficients of all the
fittings connected to the control valve from both upstream and downstream. The
value of K is calculated from Eq. (2.8).
Valve Head Loss Coefficient Calculation Due to Attached Fittings
K = K 1 + K 2 + K B1 − K B2 28
where:
K1: Head loss or resistance coefficient due to upstream fittings;

K2: Head loss or resistance coefficient due to downstream fittings;
KB1: Inlet Bernoulli coefficient;
KB2: Outlet Bernoulli coefficient.
38 2 Valve Sizing
Table 2.7 Equation constants.
N w q p γ T d, D
N1 0.0865 — NM3/h kPa — — —

0.865 — NM3/h bar — — —
1.00 — gpm psia — — —
N2 0.00214 — — — — — mm
890 — — — — — in.
N5 0.00241 — — — — — mm
1000 — — — — — in.
N6 2.73 kg/h — kPa kg/m3 — —
27.3 kg/h — bar kg/m3 — —
63.3 lb/h — psia lb/ft3 — —
N7 Normal condition 3.94 — NM3/h kPa — K —
(T = 0 C) 394 — NM3/h bar — K —
Standard condition 4.17 — NM3/h Kpa — K —
(T = 16 C) 417 — NM3/h bar — K —
Standard condition 1360 — SCFH psia — R —
(T = 60 F)
N8 0.948 kg/h — kPa — K —
94.8 kg/h — bar — K —
19.3 lb/h — psia — R —
N9 Normal condition 21.2 — NM3/h Kpa — K —
(T = 0 C) 2120 — NM3/h bar — K —
Standard condition 22.4 — NM3/h Kpa — K —
(T = 16 C) 2240 — NM3/h bar — K —
Standard condition 7320 — SCFH psia — R —
(T = 60 F)
The Bernoulli coefficients are used only when the piping diameter in the inlet
and outlet of the valve are not identical. In fact, if the inlet and outlet piping from
the valve are equal sizes, then the Bernoulli coefficients are equal (KB1 = KB2), and
therefore they are dropped from the equation. Inlet and outlet Bernoulli coeffi-
cients are calculated by using Eq. (2.9).
Bernoulli Coefficients Calculation

4
d
K B1 = K B2 = 1 − 29
D
where:
d: Nominal valve size;
D: Internal piping diameter.
The most commonly used fitting before and after a control valve is a reducer
(expander). The head loss or resistance coefficient due to reducers/expanders
are calculated by using Eqs. (2.10) and (2.11).
Head Loss Coefficient for an Inlet Reducer

2
d2
K1 = 0 5 × 1− 2 10
D2
Head Loss Coefficient for an Outlet Reducer

2
d2
K2 = 1 0 × 1− 2 11
D2
Then for a control valve installed between identical reducers/expanders, the

value of the head loss coefficient is calculated as follows:
2 2 2
d2 d2 d2
K = K1 + K2 = 0 5 × 1− +10× 1− =15× 1−
D2 D2 D2
2.4.1.4 Determine the Maximum Flow Rate (qmax) and Maximum Pressure
Drop (ΔPmax)
As its name implies, the maximum flow rate, also called limiting flow rate (qmax) or
choked flow calculated from Eq. (2.12), cannot be increased even by increasing the
pressure drop across the valve. For example, flashing the vapor from liquids when
the static valve pressure drops below the liquid vapor pressure can cause a max-
imum flow rate. It is important to calculate the maximum allowable pressure drop
(ΔPmax) for selecting the correct size of the valve if it is possible for the choked flow
to develop. The calculated ΔPmax value is compared with the actual pressure drop
across the valve specified by the process engineer, and the lesser of these two is
applied to the valve sizing process. If ΔPmax > ΔP, the flow will not be choked
and there is no need to determine the maximum flow rate as per step 4. However,
if a choked flow condition exists (ΔPmax > ΔP), then step 5 for sizing valves for
liquids shall be changed by replacing the actual service pressure differential
(ΔP = P1 − P2) with ΔPmax.
Maximum Flow Rate (qmax)
P1 − F F P v
qmax = N 1 F L Cv 2 12
SGl
40 2 Valve Sizing
Absolute vapor pressure-bar

34 69 103 138 172 207 241
1.0
Liquid critical pressure ratio
0.9
factor–FF
0.8
0.7
0.6
0.5
0 500 1000 1500 2000 2500 3000 3500
Absolute vapor pressure-psia
Figure 2.8 Water critical pressure ratio factor (FF) determination chart.
The liquid critical pressure ratio factor (FF) can be calculated by using Eq. (2.13),
or Figure 2.8 for water. When a liquid pressure drops below its vapor pressure, the
fluid state alters from a liquid to gas, that is called choked flow or vena contracta
pressure condition. FF is the ratio of the vena contracta pressure at choked flow
condition to the vapor pressure of the liquid at inlet temperature.
Liquid Critical Pressure Ratio Factor (qmax)
Pv
F F = 0 96 − 0 28 2 13
Pc
The values of FL for the globe, v-notch ball, and butterfly valves can be extracted
from Tables 2.8, 2.9, and 2.10. Ball valves with standard design are not suitable for
flow control; however, a v-notch ball valve with a unique v-shaped hole in the ball
allows precise flow control. Furthermore, butterfly valves can be used for fluid
control or throttling in addition to fluid isolation.
If the valve is installed with fittings such as reducers attached to it, the combined
liquid pressure recovery factor and piping geometry factor of a valve with attached
fittings known as parameter FLP is applied instead of FL in Eq. (2.12) for maximum
flow rate (qmax) calculation. FLP is calculated as per Eq. (2.14).
Combined Liquid Pressure Recovery and Piping Geometry Factor

Calculation
2 −1 2
K 1 Cv 1
F LP = + 2 14
N 2 d2 F 2L
Table 2.8 Representative sizing coefficients for globe control valves.
Linear cagea
Line and valve size Line size is double the

are equal valve size
Flow coefficient (Cv) Flow coefficient (Cv)

Body size,
in. (DN) Regulating Full open Regulating Full open XT FD FL
1 (25) 16.8 17.7 17.2 18.1 0.806 0.43 0.84

2 (50) 63.3 66.7 59.6 62.8 0.820 0.35
3 (80) 132 139 128 135 0.779 0.30
4 (100) 202 213 198 209 0.829 0.28
6 (150) 397 418 381 404 0.668 0.28
WhisperTM trim cageb
Line and valve size Line size is double the

are equal valve size
Flow coefficient (Cv) Flow coefficient (Cv)

Body size,
in. (DN) Regulating Regulating Regulating Regulating XT FD FL
1 (25) 16.7 17.6 15.6 16.4 0.753 0.10 0.89

2 (50) 54 57 52 55 0.820 0.07
3 (80) 107 113 106 110 0.775 0.05
4 (100) 180 190 171 180 0.766 0.04
6 (150) 295 310 291 306 0.684 0.03
a
Cage improves the control ability of a globe valve and adjusts the amount of flow inside the valve.
A linear cage gives a linear flow characteristic to the valve, as explained in Chapter 1. Linear flow
means that the change in the flow is equal to the opening percentage. For example, 100% of
opening a valve passes 100% of the fluid through the valve.
b
Noise is one of the control valve’s major operational problems caused by the turbulent flow of
vapor, gas, or steam as the fluid passes through the valve. A WhisperTM can significantly reduce
the noise level in control valves if its usage is coupled with the correct valve sizing technique. An
in-depth discussion of valve noise, particularly in control valves, appears in Chapter 6.
42 2 Valve Sizing
Table 2.9 Representative sizing coefficients for single port, globe-style control valve
with cage guiding and balanced plug.
Valve Valve Flow Port diameter Rated travel

size plug style characteristics in. (mm) inches (mm) Cv FL XT FD
1/2 Post Equal 0.38 (9.7) 0.50 (12.7) 2.41 0.90 0.54 0.61
guided percentage
3/4 Post Equal 0.56 (14.2) 0.50 (12.7) 5.92 0.84 0.61 0.61
guided percentage
1 Micro- Equal 3/8 (9.5) 3.07 0.89 0.66 0.72
form percentage 1/2 (12.7) 4.91 0.93 0.80 0.67
3/4 (19.1) 3/4 (19.1) 8.84 0.97 0.92 0.62
Cage Linear 15/16 (33.3) 20.6 0.84 0.64 0.34
guided Equal 15/16 (33.3) 17.2 0.88 0.67 0.38
percentage
1 1/2 Micro- Equal 3/8 (9.5) 3.20 0.84 0.65 0.72
form percentage 1/2 (12.7) 5.18 0.91 0.71 0.67
3/4 (19.1) 3/4 (19.1) 10.2 0.92 0.80 0.62
Cage Linear 1 7/8 (47.6) 39.2 0.82 0.66 0.34
guided Equal 1 5/16 (47.6) 35.8 0.84 0.68 0.38
percentage
2 Cage Linear 2 5/16 (58.7) 1 1/8 (28.6) 72.9 0.77 0.64 0.33
guided Equal 2 5/16 (58.7) 59.7 0.85 0.69 0.31
percentage
3 Cage Linear 3 7/16 (87.3) 1 1/2 (38.1) 148 0.82 0.62 0.30
guided Equal — — 136 0.82 0.68 0.32
percentage
4 Cage Linear 4 3/8 (111) 2 (50.8) 236 0.82 0.69 0.28
guided Equal — — 224 0.82 0.72 0.28
percentage
6 Cage Linear 7 (178) 2 (50.8) 433 0.84 0.74 0.28
guided Equal — — 394 0.85 0.78 0.26
percentage
8 Cage Linear 8 (203) 3 (76.2) 846 0.87 0.81 0.31
guided Equal — — 818 0.86 0.81 0.26
percentage
Table 2.10 Representative sizing coefficients for the v-notch ball and
high-performance butterfly valves used to control the flow.
Flow
Valve Degree of valve coefficient
size (in.) Valve type opening (Cv) FL XT FD
1 V-notch ball valve 60 15.6 0.86 0.53 —

90 34.0 0.86 0.42 —
1 1/2 V-notch ball valve 60 28.5 0.85 0.50 —
90 77.3 0.74 0.27 —
2 V-notch ball valve 60 59.2 0.81 0.53 —
90 132 0.77 0.41 —
High-performance 60 58.9 0.76 0.50 0.49
butterfly valve 90 80.2 0.71 0.44 0.70
3 V-notch ball valve 60 120 0.80 0.50 0.92
90 321 0.74 0.30 0.99
High-performance 60 115 0.81 0.46 0.49
butterfly valve 90 237 0.64 0.28 0.70
90 596 0.62 0.22 0.99
butterfly valve 90 499 0.53 0.19 0.70
90 1110 0.58 0.20 0.99
butterfly valve 90 1260 0.55 0.20 0.70
90 1820 0.54 0.18 0.99
butterfly valve 90 2180 0.48 0.19 0.70
90 3000 0.56 0.19 0.99
butterfly valve 90 3600 0.48 0.17 0.70
90 3980 0.63 0.25 0.99
High-performance 60 2500 — — 0.49
butterfly valve 90 5400 — — 0.70
90 8270 0.37 0.13 1.00
High-performance 60 3870 0.69 0.40 —
butterfly valve 90 8600 0.52 0.23 —
44 2 Valve Sizing
and
K 1 = K 1 + K B1
K1: Resistance coefficient of upstream fittings;

KB1: Inlet Bernoulli coefficient.
The maximum pressure drop across the valve (ΔPmax) is determined from
Eq. (2.15) or (2.16).
Maximum Pressure Drop Calculation for Valves Without Any

Attached Fittings
ΔPmax = F 2L P1 − F F × Pv 2 15
Maximum Pressure Drop Calculation for Valves With Attached Fittings
2
F LP
ΔPmax = P1 − F F × Pv 2 16
FP
2.4.1.5 Solve for Flow Coefficient

Equation (2.17) is used to calculate the modified flow efficient (Cv) value if
ΔPmax > ΔP meaning that the flow will not be choked. Equation (2.18) applies
to adjusted flow coefficient calculation if ΔPmax < ΔP when the flow will
be choked.
Adjusted Flow Coefficient Calculation Without a Choked Flow

(ΔPmax > ΔP)
q
Cv = 2 17
N 1FP P1 − P2 SGl
Adjusted Flow Coefficient Calculation with a Choked Flow (ΔPmax > ΔP)
q
Cv = 2 18
N 1FP ΔPmax SGl
2.4.1.6 Select the Correct Valve Size

As a general rule, the smallest valve with a flow coefficient larger than that calcu-
lated based on either Eq. (2.17) or Eq.(2.18), according to the information provided
in Tables 2.8, 2.9, and 2.10, shall be selected.
Example 2.5 A control globe valve in CL300 (PN of 50 bar) with an equal per-
centage flow characteristic and the guided cage is installed on an 8 line.
The standard concentric reducers are used to install the valve on the line. The fluid
service is liquid propane with a flow rate of 800 gpm and specific gravity of 0.5. The
valve’s inlet pressure, outlet pressure, and inlet temperature values are 300 psig,
275 psig, and 70 F, respectively. The values of absolute thermodynamic
critical pressure and absolute vapor pressure of the liquid at inlet temperature
are 616.3 psia and 124.3 psia, respectively. Select the correct size of the con-
trol valve.
Answer
The flow coefficient of the valve is calculated based on the basic Cv calculation for
liquids as per Eq. (1.1) as follows:
SG 05
Cv = Q = 800 = 113 13
ΔP 300 − 275
Refer to Table 2.9. The smallest globe control valve size with a flow coefficient
larger than 113.13 and an equal percentage flow characteristic is 3 with a Cv value
of 136. Thus, a 3 control valve is assumed first and checked to see whether it can
provide the required flow rate by considering various parameters such as the
installation of reducers and the possibility of choke flow. The next step is to specify
the necessary variables required to size the valve as follows:
• Desired valve design is a 3 CL300 globe valve with an equal percentage flow
characteristic
• Service condition – q = 800 gpm
• P1 = 300 psig (20.7 bar) = 314.7 psia(21.7 bara)
• P2 = 275 psig (19.0 bar) = 289.7 psia(20.0 bara)
• ΔP = 25 psi (1.7 bar)
• T1 = 70 F = 21 C
• SGl = 0.5
• Pv = 124.3 psia(8.6 bara)
• Pc = 616.3 psia(42.5 bara)
The next step is to select an equation constant (N) from Table 2.7. By considering
the pressure values (p) in psia unit and flow rate (q) value in gpm, N1 = 1.
A 3 valve is installed on an 8 line meaning that reducers/expanders are
installed before and after the valve. Thus, it is essential to determine the piping
geometry factor (FP) from Eq. (2.7) to correct the flow coefficient by flow losses
caused by fittings connected to the valve.
46 2 Valve Sizing
−1 2
K Cv 2
FP = 1 +
N2 d2
where:
N2 = 890 refer to Table 2.7, by considering the diameter values in inch units;
d: Nominal valve size = 3 in. (76.2 mm);
Cv: flow coefficient = 136 for a 3 CL300 globe control valve with an equal percent-
age flow characteristic as per Table 2.9.
For a control valve installed between identical reducers/expanders, the value of

the head loss coefficient is calculated as follows:
2 2 2
d2 d2 d2
K = K1 + K2 = 0 5 × 1− +10× 1− =15× 1−
D2 D2 D2
where:
D: Nominal pipe size = 8 ;
d: Nominal valve size = 3 .
2
K = 1 5 × 1 − 32 82 = 1 11
−1 2 −1 2
K Cv
2 1 11 136 2
FP = 1 + N2 d2
= 1+
890 32
−1 2
= 1 28 = 0 88
The value of the rated liquid pressure recovery factor, FL, for a 3 control globe
valve with a guided cage and an equal percentage flow characteristic as per
Table 2.9 is 0.82. Because the given valve in this example is attached to a couple
of reducers (fittings) during the installation, FLP, the combined liquid pressure
recovery factor and piping geometry factor of the valve with attached fittings must
be calculated according to Eq. (2.14).
2 −1 2 2 −1 2
K 1 + K B1 Cv 1 1 35 136 1
F LP = + 2 = +
N2 d2 FL 890 9 0 822
−1 2
= 0 35 + 1 49 = 0 74
4 4
d 3
K B1 = 1 − = 1− = 0 98
D 8
K 1 + K B1 = 0 98 + 0 37 = 1 35
The liquid critical pressure ratio factor (FF) can be calculated by using Eq. (2.13)
as follows:
Pv 124 3
F F = 0 96 − 0 28 = 0 96 − 0 28 = 0 834
Pc 616 3
Now it is possible to calculate the maximum pressure drop (ΔPmax)
from Eq. (2.16):
F LP 2 0 74 2
ΔPmax = P1− F F × Pv = 300 − 0 834 × 124 3
FP 0 88
= 0 71 × 196 33 = 139 40 psi
ΔPmax = 139.40 psi > 25 psi. So choked flow will not happen. The modified flow
coefficient is calculated from Eq. (2.17):
q 800
Cv = = = 128 58
N 1 F P P1 − P2 SGl 1 × 0 88 25 0 5
The required flow coefficient of 128.58 is less than the assumed valve capacity,
which has Cv of 136. So 3 is the correct size. However, if the value of the modified
flow coefficient was higher than 136, then the next larger size (4 in.) would be the
correct valve size most likely.
2.4.2 Control Valve Sizing for Gas and Steam

There are six steps for a control valve size selection in liquids:
1) Specify the variables required to size the valve;

2) Determine the equation constant (N);
3) Determine the piping geometry factor (FP);
4) Determine expansion factor (Y);
5) Solve for required flow coefficient (Cv) value using an appropriate equation;
6) Select the appropriate flow coefficient table and the calculated flow coefficient
valve to select the suitable valve size.
2.4.2.1 Specify the Variables Required to Size the Valve

The first step is to specify the variables required to size the valves, such as valve
pressure class, desired valve design (e.g. balanced globe with a linear cage), and
valve size.
• Process fluid (e.g. oil and water);

• Service conditions such as volume rate of flow (q), mass rate of flow (w),
Upstream and downstream pressure values (P1 and P2), pressure drop across
the valve (ΔP = P1 − P2), upstream temperature (T1), gas-specific gravity that
is defined as a dimensionless number equal to the ratio of the flowing gas density
to the density of air with both at standard conditions (SGg), molecular weight (M),
48 2 Valve Sizing
the ratio of specific heats (k), gas compressibility factor (Z), and specific weight at
the valve inlet condition (γ). If any of the aforementioned parameters are new or
unfamiliar, refer to Table 2.6, where all parameters for control valve size selection
are summarized.
2.4.2.2 Determine the Equation Constant (N)

N, as explained before, is a numerical constant in each of the flow or sizing pro-
cedures to adjust the relevant equation for using different sizing units. Values of
equation constant as well as their applicable units are given in Table 2.7 in the liq-
uid valve sizing section. If sizing the valve, use either N7 or N9 for a flow rate in
volumetric units such as SCF per hour (SCFH) or cubic meter per hour m3/h.
N7 can be used if the gas-specific gravity, SGg, is specified, and N9 can be used only
if the gas’s molecular weight, M, is known. Use either N6 or N8 if sizing the valve for
a mass flow rate such as lb/h or kg/h. N6 can be used if the specific weight at the
inlet valve condition, γ, is known, whereas N8 can be selected if parameter M,
molecular weight has been specified.
2.4.2.3 Determine Piping Geometry Factor (FP)

FP is a correction factor that accounts for the pressure losses due to piping fittings
that are used to change the direction or size of the piping or take branches from the
headers such as tees, elbows, and reducers. If fittings are used before or after the
valve, the effect of piping geometry must be considered for the valve sizing. How-
ever, if no fitting is attached to the valve, FP is equal to one and does not impact the
valve sizing process. Equation (2.7) shows how to calculate the piping geometry.
2.4.2.4 Determine the Expansion Factor (Y)

When sizing components such as valves in the piping system handling compress-
ible gas, it is essential to account for all factors, including the expansion factor that
would affect the flow rate and pressure drop in the piping system. As compressible
gas flows through each component in the piping system, the head loss causes
expansion as well as change in density, temperature, and velocity. The gas expan-
sion process adds resistance to the flow rate, resulting in a flow rate reduction for a
given pressure drop. Thus, the gas expansion factor, Y, shall be taken into account
by using Eq. (2.19) as follows.
Gas Expansion Factor Calculation

X
Y = 1− 2 19
3 FkX T
where,
Fk: The ratio of specific heat factor = k/1.4;

k: Ratio of specific heats;
X: The pressure drop ratio = ΔP/P1; If X < 0.2, the change in gas density and expan-
sion is small, the fluid can be assumed incompressible, and the parameter Y can
be neglected. For X > 0.4, the expansion factor shall be considered as the fluid
can be assumed compressible;
XT: The pressure drop ratio factor for valves installed without attached fittings can
be extracted from Tables 2.8, 2.9, and 2.10. More precisely, XT is the pressure
drop ratio required to produce critical or maximum flow through the valve
when Fk = 1. It should be noted that the value of Y cannot be less than
0.667. When the calculated Y value is less than 0.0667, the Y value shall be
assumed to be 0.667.
Suppose the control valve is installed with fittings such as reducers or elbows
attached to it. In that case, the effect of fittings shall be taken into account by repla-
cing XT by a factor of XTP that can be calculated from Eq. (2.20) as follows.
Pressure Drop Ratio Factor Calculation with Fittings Attached to

the Control Valve (XTP)
2 −1
XT X TKi Cv
X TP = 2 1 + 2 20
FP N5 d2
where:
N5: Numerical constant is taken from Table 2.7;

d: Assumed nominal valve size;
Cv: Valve flow coefficient for a fully open valve taken from either Table 2.8 or 2.9
or 2.10;
FP: Piping geometry factor;
XT: Pressure drop ratio for the valve without any fitting attached can be taken from
Tables 2.8, 2.9, and 2.10;
Ki: Inlet head loss coefficient, which is calculated as follows:
2 4
d2 d
K i = K 1 + K B1 = 0 5 × 1− + 1−
D2 D
where:
K1: Resistance coefficient of upstream fittings is calculated from Eq. (2.10);

KB1: Inlet Bernoulli coefficient is calculated from Eq. (2.9).
50 2 Valve Sizing
2.4.2.5 Solve for the Required Flow Coefficient (Cv)

Flow coefficient value is different from volumetric flow rate than mass flow rate
units. When the gas-specific gravity, SGg, is known for volumetric flow rate units,
the flow coefficient is calculated as per Eq. (2.21).
Flow Coefficient (Cv) Calculation for Volumetric Flow Rate and

Knowing the Gas-Specific Gravity:
q
Cv = 2 21
N 7 F P P1 Y X SGg T 1 Z
If the molecular weight, M, is known for volumetric flow rate units, the flow
coefficient is calculated as per Eq. (2.22).
Flow Coefficient (Cv) Calculation for Volumetric Flow Rate and

Knowing the Molecular Weight
q
Cv = 2 22
N 7 F P P1 Y X M T 1 Z
If the specific weight at the valve inlet, γ, is known for mass flow rate units, the
flow coefficient is calculated as per Eq. (2.23).
Flow Coefficient (Cv) Calculation for Mass Flow Rate and Knowing
the Specific Weight at the Valve Inlet
W
Cv = 2 23
N 6 F P Y XP1 γ
If the molecular weight, M, is known for mass flow rate units, the flow coeffi-
cient is calculated as per Eq. (2.24).
Flow Coefficient (Cv) Calculation for Mass Flow Rate and Knowing the
Molecular Weight
W
Cv = 2 24
N 8 F P P1 Y XM T 1 Z
Example 2.6 A control globe valve in CL300 (PN of 50 bar) with a linear cage is
installed on a 6 line with standard concentric reducers on both sides. The fluid
service is superheated steam with a flow rate of 125,000 lb/h. The valve’s inlet
pressure, outlet pressure, and inlet temperature values are 500 psig, 250 psig,
and 500 F, respectively. The values of specific weight at the valve inlet
condition (γ) and ratio of specific heats (k) are 1.0434 lb/ft3 and 1.28 lb/ft3, respec-
tively. Select the correct size of the control valve.
Answer
The first step is to summarize the data associated with the valve and process
conditions.
1) Desired valve design: CL300 control globe valve with a linear cage;
2) Process fluid: superheated steam;
3) Service conditions:
W = q = 125,000 lb h
P1 = 500 psig 34 5 bar = 514 7 psia 35 5 bar

P2 = 250 psig 17 bar = 264 7 psia 18 3 bar
ΔP = 500 − 250 = 250 psig 17 bar
ΔP 250
X The pressure drop ratio = = = 0 49
P1 514 7
T 1 = 500 F 260 C
γ = 1 0434 lb ft3
k = 1 28
The second step is to calculate the valve’s flow coefficient for superheated steam.
There are different formulas to calculate the valve flow coefficient handling the
steam. First, the valve flow coefficient for saturated steam shall be determined,
which depends on critical or noncritical pressure drop. In this case, the outlet pres-
sure is less than 50% of the inlet pressure, which is considered a critical pressure
drop. A noncritical pressure drop is if the outlet pressure exceeds 58% of the inlet
pressure. The flow coefficient for saturated steam and critical pressure drop is cal-
culated from Eq. (2.25).
Flow Coefficient (Cv) Calculation for Saturated Steam and Critical

Pressure Drop
W
Cv = 2 25
1 61 × P1
where:
W: Steam flow rate (lb/h);

P1: Inlet steam absolute pressure (psia).
52 2 Valve Sizing
w 125, 000
Cv = = = 150 84
1 61 × P1 1 61 × 514 7
The flow coefficient for saturated steam and noncritical pressure drop is calcu-
lated from Eq. (2.26).
Flow Coefficient (Cv) Calculation for Saturated Steam and Noncritical

Pressure Drop
m
Cv = 2 26
21 P1 + P2 ΔP
The temperature of saturated steam, which is pure steam, is 298 F. The flow
coefficient of superheated steam is calculated according to Eq. (2.27).
Flow Coefficient (Cv) Calculation for Superheated Steam
C v Superheated steam = C v Saturated steam 1+ 0 00065dt 2 27
where:
dt: Steam temperature above saturation temperature at the actual pressure ( F).
Cv Superheated steam = Cv Saturated steam 1 + 0 00065dt
= 150 84 1 + 0 00065 500 − 298 = 150 84 × 1 1313 = 170 65
Table 2.11 extracted from Table 2.10 includes flow coefficient values for control
globe valves with a guided cage in size ranges from 3 to 6 .
Table 2.11 Flow coefficient values for cage-guided control globe valves in size range
from 3 to 6 .
Valve Valve Flow Port diameter Rated travel

size plug style characteristics inches (mm) inches (mm) Cv
3 Cage Linear 3 7/16 (87.3) 1 1/2 (38.1) 148

guided Equal — — 136
percentage
4 Cage Linear 4 3/8 (111) 2 (50.8) 236
percentage
6 Cage Linear 7 (178) 2 (50.8) 433
percentage
A 3 valve is undersized because it can produce a maximum flow coefficient of

148, which is lower than 170.65 and insufficient. Conversely, a 6 valve is over-
sized. So a 4 valve with a linear flow characteristic and a flow coefficient value
of 236 can be a suitable choice based on the preliminary flow coefficient calcula-
tion without considering the corrective parameters.
The next step is to find an appropriate equation constant, N, from Table 2.7.
Because the specified flow rate is in the mass unit, pound per hour, and the specific
weight of steam, γ, are given, N6 constant is applied, and N6 = 63.3.
The piping geometry factor is calculated as follows:
−1 2
K Cv 2
FP = 1 +
N2 d2
where:
N2 = 890 refer to Table 2.7, by considering the diameter values in inches;

d: Nominal valve size = 4 in. (101.6 mm);
Cv: Flow coefficient = 236 for a 4 CL300 globe control valve with a linear flow
characteristic as per Tables 2.9 and 2.11.
For a control valve installed between identical reducers/expanders, the value of

the head loss coefficient is calculated as follows:
2 2 2
d2 d2 d2
K = K1 + K2 = 0 5 × 1− +10× 1− =15× 1−
D2 D2 D2
where:
D: Nominal pipe size = 6 ;

d: Nominal valve size = 4 .
2
42
K=15× 1− = 0 463
62
−1 2 −1 2
K Cv 2
0 463 236 2
FP = 1 + = 1+ = 0 95
N2 d2 890 42
The expansion factor, Y, is calculated as follows:

X
Y = 1−
3F k X T
k 1 28
Fk = = = 0 91 and X = 0 49 as calculated in step 1
14 14
54 2 Valve Sizing
Since the control valve size is smaller than the connected piping, fittings (redu-
cers) must be used upstream and downstream of the valve. In that case, the effect
of fittings shall be taken into account by replacing XT by a factor of XTP that can be
calculated from Eq. (2.20) as follows:
2 −1
XT X TKi Cv
X TP = 1+
F 2P N5 d2
where:
XT = 0.69 extracted from Table 2.9 for a 4 globe control valve with a guided cage
and linear flow characteristic;
FP = 0.95 as calculated in the previous step;
N5 = 1000, from equation constants (Table 2.7);
Cv = 236;
d =4 .
2
d2 d 4
K i = K 1 + K B1 = 0 5 × 1 − 2 + 1 −
D D
2 4
42 4
=05× 1− + 1− = 0 96
62 6
2 −1
0 69 0 69 × 0 96 236
X TP = 2 1 + = 0 67
0 95 1000 16
X 0 49
Y = 1− = 1− = 0 73
3F k X T 3 × 0 91 × 0 67
Because the specific weight at the valve inlet, γ, is known for mass flow rate
units, the flow coefficient is calculated as per Eq. (2.23).
W 12,500
Cv = =
N 6 F P Y XP1 γ 63 3 × 0 95 × 0 73 0 49 × 514 7 × γ × 1 0434
= 176
The conclusion is that the 4 globe control valve initially selected was correct
because the assumed valve has a flow coefficient of 236, which is higher than
the corrected Cv value of 176. A 3 control globe valve with the guided cage
and linear flow characteristic has a Cv value of only 148 which is insufficient.
Example 2.7 On this occasion, it has been decided to install an 8 V-notch ball
valve in CL150 (PN of 20 bar) on an 8 pipe. Natural gas is the fluid service with a
flow rate of 6 × 106 standard cubic feet (SCF). Inlet pressure, outlet pressure, and
inlet temperature for the valve are 214.7 psia, 64.7 psia, and 60 F, respectively.
Accordingly, the values of molecular weight (M), gas specific gravity (SGg), and the
ratio of specific heats (k) are 17.38, 0.60, and 1.31, respectively. Is the valve size
correctly selected? (Note: assume an equal percentage flow characteristic for
the valve.)
Answer
First, it is necessary to summarize the data pertaining to the valve and process
conditions.
1) Desired valve design: CL150 V-notch ball valve;

2) Process fluid: natural gas;
3) Service conditions:
W = q = 6 × 106 standard cubic feet SCF
ΔP = 200 − 50 = 150 psig 10 3 bar
ΔP 150
X The pressure drop ratio = = = 0 70
P1 214 7
T 1 = 60 F 16 C = 520 R
M = 17 38
SGg = 0 6
k = 1 31
The second step is to determine the correct value of equation constant (N) from
Table 2.7. As in the example, the flow rate value in SCF is provided, it is possible to
use either N7 or N9 at the standard condition. Consider the use of N7, which is
equal to 1360.
The third step involves finding the value of the piping geometry factor, param-
eter FP. Because the valve and the connected piping are the same size, there are no
reducers required. Therefore, FP = 1.
The fourth step is to find the value of Y, which is the expansion factor. It is nec-
essary to find Y, which is the expansion factor from Eq. (2.19), in the fourth step.
k 1 31
F k The ratio of specific heats factor = = = 0 94
14 14
According to Table 2.10, the value of XT for an 8 V-notch ball valve at the full
opening percentage is equal to 0.18. In the first step, parameter X, the pressure
drop ratio, is calculated, which equals 0.70.
X 07
Y = 1− 1− = − 0 38
3F k X T 3 × 0 94 × 0 18
56 2 Valve Sizing
Y cannot be less than 0.667. Thus, in this case, where the value of Y is calculated
to be −0.038, the minimum Y value of 0.667 should be considered.
Using Eq. (2.21), the fifth step involves calculating the adjusted flow coefficient
(Cv) as follows. It should be noted that the gas Z-factor is assumed to be one.
q 6 × 106
Cv = =
N 7 F P P1 Y X SGg T 1 Z 1360 × 1 × 214 7 × 0 667 0 7 0 6 × 520 × 1
= 655
In accordance with Table 2.10, an 8-in. V-notch ball valve at the fully open posi-
tion will have a Cv value of 3000. In the following step, we shall determine the
opening percentage of the valve that results in a flow efficiency of 655 when we
consider an equal percentage flow characteristic. Based on Eq. (1.13) taken from
Chapter 1, the valve flow coefficient is calculated in terms of valve opening per-
centage for an equal percentage flow characteristic relevant in this case.
C vvp = C vmax × αvp − 1
where:
vp: Valve opening percentage, which is unknown in this case;
Cvvp: The valve flow coefficient of 655 is linked to an unknown valve opening
percentage;
Cvmax: Maximum flow coefficient of the valve at the fully open position = 3000;
α = 50.
655 = 3000 × 50vp − 1 50vp − 1 = 0 2183 vp − 1 × 1 7 = − 0 66 vp = 61 2
So the 8 V-notch ball valve is capable of providing a flow coefficient value of

655 at 61.2% opening. Therefore, the 8 valve size is appropriate.
2.5 Safety Relief Valve Sizing
It is impossible for a chemical process facility to be immune to overpressure, dic-

tating the need for overpressure protection. It is the purpose of the pressure relief
valve (PRV) installed on pressurized equipment (labeled Protected System in
Figure 2.9) to release the overpressure gas or fluid from the equipment into the
flare system. In API standard 520 part 1, the sizing of PRVs is clearly described.
According to API 520, it is critical to understand the similarities and differences
between a relief valve and a safety valve. Both types of safety valves are spring-
loaded and used to protect pressure piping and facilities from overpressure. This
relief valve is typically opened in proportion to the increase in pressure over the
opening pressure, and it is used for noncompressible fluids. Contrary to this, a
safety valve is opened rapidly and is normally used for compressible services.
To flare,
Discharge
header system
BPs
PRV (closed)
Protected
system
Figure 2.9 PRV installation on pressurized equipment connected to a flare system.
There is the possibility of using a safety relief valve suitable for both applications.
Chapter 7 discusses PSVs or PRVs and provides relevant equations and
calculations.
A basic safety valve sizing procedure involves three steps. In the first step, the set
pressure of the PSV is established, which is the gauge pressure at which the pres-
sure relief device will open under normal operating conditions. In the second step,
the relief capacity must be determined. Last, but certainly not least, the pressure
safety device must be sized so that it can provide a capacity equivalent to the relief
pressure.
For sizing pressure safety or relief valves, it is necessary to pay attention to the
valve size, which must correspond to both the inlet and the discharge pipe sizes.
The inlet and outlet ports of a pressure relief or safety valve are typically different
sizes, resulting in the valves being recognized in two sizes: one inlet and the other
outlet. The API 526 standard, which deals with flanged steel PRVs, defines the
sizes of PSVs alphabetically. ASME’s Boiler and Pressure Vessel Code (BPVC)
Sec. VIII refers to boiler construction rules and contains the same letters for
PSV sizing corresponding to different orifice sizes to provide effective discharge
area. In Table 2.12, the API 526 orifice sizes are compared with those of ASME
Sec. VIII based on the alphabetic letters from “D” through “T”. Safety valves
may have several different sizes of inlet and outlet connections, even when the
orifice letter is the same. A 2 × J × 3 and 3 × J × 4 safety valve have the same
size orifice, illustrated by the letter “J,” but have different sizes of inlet and outlet as
shown before and after the orifice letter, respectively.
58 2 Valve Sizing
Table 2.12 ASME and API standard orifice sizes.
Serial Orifice API effective ASME effective PSV inlet × outlet

number designation area (in.2) area (in.2) sizes (in.)
1 D 0.110 0.124 1 ×2
1.5 × 2
1.5 × 2.5
2 E 0.196 0.221 1 ×2
1.5 × 2
1.5 × 2.5
3 F 0.307 0.347 1 ×2
1.5 × 2
1.5 × 2.5
4 G 0.503 0.567 1.5 × 2.5
1.5 × 3
2 ×3
5 H 0.785 0.887 1.5 × 3
2 ×3
6 J 1.287 1.453 2 ×3
2.5 × 4
3 ×4
7 K 1.838 2.076 3 ×4
8 L 2.853 3.221 3 ×4
4 ×6
9 M 3.600 4.065 4 ×6
10 N 4.340 4.900 4 ×6
11 P 6.380 7.205 4 ×6
12 Q 11.05 12.47 6 ×8
13 R 16.00 18.06 6 ×8
6 × 10
14 T 26.00 29.35 8 × 10
What is the reason behind the larger orifice size or effective discharge area in
ASME than API? The required orifice area that is calculated using the equations
discussed later in this chapter is inversely proportional to the coefficient of dis-
charge designated with parameter K in API 520, and with parameter KD in ASME
BVPC Sec. VIII. The ASME Sec. VIII code was revised in 1962 to require that “K”
be used in sizing calculations with 10% safety factor instead of KD. (K = KD × 0.9).
In some texts, the modified coefficient of discharge based on the code or regula-
tion is referred to as the “rated coefficient of discharge.” In the following subsec-
tion, we discuss the sizing of safety valves for gas or vapor in light of critical flow
conditions.
2.5.1 Sizing for Gas or Vapor Relief

2.5.1.1 Critical Flow
When a compressible gas expands across a nozzle or orifice, its velocity, mass flow
rate, and volume increase with decreasing downstream pressure. The ratio of pres-
sure at the valve outlet to pressure at the inlet is known as the critical pressure ratio.
As a general rule of thumb, a critical flow occurs when the outlet pressure is
40–60% of the outlet pressure. Critical flow occurs when the outlet pressure or
downstream pressure reaches 6–4 bars when the inlet pressure is 10 bar. An alter-
native condition for critical flow occurrence is if the critical flow pressure calcu-
lated from Eq. (2.40) is higher than the backpressure. Backpressure is defined later
in this chapter. When operating under critical flow conditions, pressure relief
devices in gas or vapor service are sized according to API 520 using Eqs. (2.28)
and (2.29) for mass flow rate, and Eqs. (2.30), (2.31), (2.32), and (2.33) for volumet-
ric flow rate. In order to select a PRV, it must have a discharge area as large as or
wider than the calculated value of “A” as shown in the equations later.
Sizing Relief Devices for Gas or Vapor in a Critical Flow Condition

According to Mass Flow Rate
W TZ
A= US Customary units 2 28a
C K d P1 K b K c M
13, 160 W TZ
A= SI units 2 29a
C K d P1 K b K c M
Sizing Relief Devices for Gas or Vapor in a Critical Flow Condition

According to Volumetric Flow Rate
V TZM
A= US Customary units 2 30
6 32CK d P1 K b K c
V TZG
1 175 CK d P1 K b K c
32,250 W TZM
A= SI units 2 32
6 32 C K d P1 K b K c
189,750 V TZG
A= SI units 2 33
C K d P1 K b K c
where:
A: Required effective discharge area of the safety relief device, in.2 or mm2;
W: Mass flow rate through the safety device, lb/h or kg/h;
60 2 Valve Sizing
400
380
360
C
340
320
1.0 1.2 1.4 1.6 1.8 2.0

Ratio of specific heats, k
Figure 2.10 Curve for assessing the C factor from the specific heat ratio in the assumption
of ideal gas behavior.
C: An ideal gas constant or coefficient is derived from the ratio of specific heats (k)
as indicated in Eq. (2.34), Figure 2.10, and Table 2.13, respectively, assuming
ideal gas behavior. (If not known, it can be assumed C = 315.) Table 2.14 pre-
sents the C factor and “k” ratio of specific heats for different gases.
Calculation of the Gas Constant from the Specific Heat Ratio for
an Ideal Gas
k+1 k−1
2
C = 520 k 2 34
k+1
Kd: Effective coefficient of discharge and the following values are required to be
considered according to API 520 for primary sizing; the coefficient is equal to
0.975 when the PRV that handles gas, vapors, or steams is installed with a rup-
ture disk or without a rupture disk. When a PRV is not installed and the sizing is
performed for a rupture disk, 0.62 is used. A rupture disk, also called a pressure
safety disk or a burst disk, is a one-time-use device to protect pressure equip-
ment from overpressure situations.
P1: Inlet flowing pressure or relieving pressure, psig or KPa as per Eq. (2.35). If this
pressure is expressed as an absolute pressure. Based on Eq. (2.36), it is equal to
the set pressure plus the allowable overpressure plus atmospheric pressure min-
utes loss pressure. Allowable overpressure, which is typically 10% percent of the
set pressure or design pressure, is used to overcome the spring force acting
against lifting the disk in order to open the valve. A safety relief valve is
Table 2.13 Values of the gas constant (C) based on the ratio of specific heat (k)
according to API 520.
k C k C k C k C
1.00 315 1.31 347 1.60 372 1.90 394

1.01 317 1.31 348 1.61 373 1.91 395
1.02 318 1.32 349 1.62 374 1.92 395
1.03 319 1.33 350 1.63 375 1.93 396
1.04 320 1.34 351 1.64 376 1.94 397
1.05 321 1.35 352 1.65 376 1.95 397
1.06 322 1.36 353 1.66 377 1.96 398
1.07 323 1.37 353 1.67 378 1.97 398
1.08 325 1.38 354 1.68 379 1.98 399
1.09 326 1.39 355 1.69 379 1.99 400
1.10 327 1.41 356 1.70 380 2.00 400
1.11 328 1.41 357 1.71 381 — —
1.12 329 1.42 358 1.72 382 — —
1.13 330 1.43 359 1.73 382 — —
1.14 331 1.44 360 1.74 383 — —
1.15 332 1.45 360 1.75 384 — —
1.16 333 1.46 361 1.76 384 — —
1.17 334 1.47 362 1.77 385 — —
1.18 335 1.48 363 1.78 386 — —
1.19 336 1.49 364 1.79 386 — —
1.20 337 1.50 365 1.80 387 — —
1.21 338 1.51 365 1.81 388 — —
1.22 339 1.52 366 1.82 389 — —
1.23 340 1.53 367 1.83 389 — —
1.24 341 1.54 368 1.84 390 — —
1.25 342 1.55 369 1.85 391 — —
1.26 343 1.56 369 1.86 391 — —
1.27 344 1.57 370 1.87 392 — —
1.28 345 1.58 371 1.88 393 — —
1.29 346 1.59 372 1.89 393 — —
1.30 347 1.60 373 1.90 394 — —
62 2 Valve Sizing
Table 2.14 C factor and “k” ratio of specific heats for different gases.
Ratio of specific
Gas Molecular weight “C ” factor heat (k)
Air 29.96 356 1.40

Ammonia 17.03 356 1.40
Argon 40 348 1.31
Carbon dioxide 44.01 356 1.40
Carbon monoxide 28 356 1.40
Ethane 30.07 336 1.19
Ethylene 28.03 341 1.24
Helium 4 377 1.66
Hexane 86.18 322 1.06
Hydrogen 2.02 357 1.41
Hydrogen sulfide 34 349 1.32
Methane 16.4 348 1.31
N-Butane 58.12 326 1.09
Natural gas (specific gravity = 0.60) 18.9 344 1.27
Nitrogen 28 356 1.40
Oxygen 32 356 1.40
Pentane 72.15 323 1.07
Propane 44.09 330 1.13
Propylene 47.08 332 1.15
Steam 18.01 348 1.31
Sulfur dioxide 64.04 346 1.29
illustrated in Figure 2.11, including its inlet and outlet, as well as some of its
main components, such as the spring and disk.
Calculation of Safety Device Inlet, Upstream, or Relieving Pressure
P1 = Pset + Pover − Ploss 2 35
Calculation of Safety Device Inlet, Upstream, or Absolute Relieving

Pressure
P1 = Pset + Pover + Patmosphere − Ploss 2 36

Figure 2.11 Safety relief valve.

Adjusting
screw
Spring
Disc
Outlet
Base
Inlet
Kb: Capacity correction factor also called backpressure correction factor due to back-
pressure that applies only to conventional or balanced bellows safety valves.
Capacity correction factor is the ratio of the safety valve capacity with backpres-
sure included C1 to the rated valve capacity without backpressure C2 that is typ-
ically applicable for valves handling vapors and gases, based on Eq. (2.37).
Calculation of Backpressure Correction Factor

C1 Safety valve capacity with backpressure included
Kb = = 2 37
C2 Safety valve capacity without backpressure
64 2 Valve Sizing
1.00
0.95
0.90 16% overpressure

Backpressure correction factor, Kb
0.85
10% overpressure
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0 5 10 15 20 25 30 35 40 45 50
Percent of gauge pressure = (PB/PS) × 100
PB = back pressure, in psig,
PS = set pressure, in psig,
Figure 2.12 Backpressure correction factor (Kb) for balanced bellows PSVs in vapors
and gases.
The balanced bellows neutralize the effect of backpressure on the safety valve’s
operation. The backpressure correction factor for conventional and pilot-operated
safety valves is one. Backpressure is the pressure at the outlet of the pressure relief
device as a result of the pressure in the discharge system. The value of Kb for bal-
anced bellows safety valves can be obtained from the valve manufacturer or from
Figure 2.12.
Calculation of the backpressure correction factor for balanced bellows safety
valves is based on the percentage of gauge backpressure calculated according
to Eq. (2.38).
Calculation of Gauge Backpressure for Bellows Safety Valves

Pb Backpressure
Percentage of gauge backpressure = × 100 = × 100
Ps Set pressure
2 38
It is possible to determine the value of Kb using the diagram in Figure 2.13 for
conventional safety valves. In Figure 2.13, it is shown that the percentage of
1.1
1.0
k = 1.0
k = 1.2
0.9
0.87 k = 1.4
k = 1.6
k = 1.8
Kb
0.8
0.7
0.6
See
example
problem
0.5 below
76
40 60 80 100
Percent of back pressure = PB/(PS + Po) × 100 = r × 100
Kb = back pressure correction factor,
PB = back pressure, in psia,
PS = set pressure, in psia,
PO = overpressure, in psi,
Figure 2.13 Backpressure correction factor for gas and vapors based on API 520 for
conventional relief valves.
backpressure and the ratio of specific heat k are used to calculate the value of Kb for
conventional relief valves handling gases or vapors. Equation (2.39) can be used to
calculate the percentage of backpressure for conventional safety valves. A ratio of
specific heat is defined as the ratio of heat capacity at constant pressure to heat
capacity at constant volume. An object’s thermal capacity, or heat, is defined as
the amount of heat required to change its temperature by 1 . The ratio of specific
heat k for a variety of gases can be determined from Table 2.14.
Calculation of Backpressure Percentage for Conventional Safety Valves

Pb
Backpressure percentage = × 100
Ps + Po
Absolute backpressure
= × 100
Absolute set pressure + Overpressure
2 39
Kc: The combination correction factor for installation with a rupture disk upstream
of the PRV. This value is one if there is no rupture disk installed and is 0.9 if a
rupture disk is installed in conjunction with a PRV.
66 2 Valve Sizing
T: Relieving temperature of the inlet gas or vapor, Rankine or Fahrenheit;

Z: Gas compressibility factor or z-factor that is equal to one for an ideal gas;
M: Molecular weight of the gas or vapor at inlet relieving conditions. The molec-
ular weight for some gases can be found in Table 2.14.
V: The flow velocity through the device in SCF per minute (SCFM) at 14.7 psia and
60 F, or normal cubic meter per minute at 0 C and 101.325 KPa.
G: Gas-specific gravity at standard conditions referred to air at standard condi-
tions. Accordingly, the specific gravity of air is one at 14.7 psia and 60 F or
at 0 C and 101.325 KPa.
Example 2.8 When the valve handles air service, it discharges the pressure at
6 bars into the atmosphere. Can it be determined whether the flow is critical
or not?
Answer
It should be noted that according to Table 2.14, the specific heat k of air is 1.40. The
discharge pressure P1 is 6 bars, and the backpressure is equal to the atmospheric
pressure, which is 1.01325 bars. The critical flow pressure is calculated according
to Eq. (2.40) as follows:
Critical Flow Pressure Calculation

k k−1
2
Pcf = P1 2 40a
k+1
where:
Pcf: Critical flow nozzle pressure in psia;

P1: Upstream or relieving pressure in psia;
k: Ratio of specific heats for the gas.
1 4 1 4−1
2
Pcf = 6 = 6 × 0 52828 = 3 17 bar
14+1
As the critical pressure of 3.17 bar is larger than the backpressure of 1.1325 bar,
the flow is critical.
Gas Z-factor or compressibility factor indicates how much the real gas devi-
ates from an ideal gas at the given pressure and temperature. It should be noted
that gases are composed of complex compounds and structures. They are com-
posed of billions of energetic gas molecules that interact with one another. Due
to this reason, it is very difficult and complex to model a real gas. As a result, the
concept of an ideal gas was developed in order to simplify the behavior of gases.
Scientists and engineers use the theory of ideal gases in order to model and pre-
dict the behavior of gases. However, the behavior of real gases matches that of
ideal gases only in some cases, such as when the pressure is atmospheric and the
temperature is in the range of room temperature. Generally, the temperature in
a room is the temperature at which the majority of people would feel comfort-
able. The room temperature is typically in the range of 20 – 22 C. At lower and
higher temperatures, the characteristics of real gases diverge significantly from
those of ideal gases. There are two main rules associated with ideal gases: the
first rule states that all ideal gases have molecules that are in constant motion,
rather than atoms that attract or repel each other. It is believed that intermolec-
ular forces are negligible in ideal gases. There is no type of interaction between
the molecules in an ideal gas model, except for elastic collisions, which are pro-
duced when molecules come into contact with each other and exchange kinetic
energy. However, the total molecular kinetic energy inside the gas is constant,
and there is no change in kinetic energy before and after a collision. The second
important rule is that the molecules occupy negligible volume. Helium is known
as a type of gas that behaves in a very similar way to ideal gases. Unlike an ideal
gas, real gas does not behave according to the laws of kinetic molecular theory.
Therefore, real gases are also called nonideal gases. As a matter of fact, gases
such as carbon dioxide, carbon monoxide, helium, oxygen, and nitrogen are
all real gases. Generally, a real gas has a smaller volume than an ideal gas at
the same pressure and temperature. Thus, real gas can be considered supercom-
pressible. The ratio of real volume to ideal volume is called the Z-factor or com-
pressibility factor, represented by parameter Z and calculated according
to Eq. (2.41).
Gas Z-Factor or Compressibility Factor Calculation
Actual or real volume of the gas at specified pressure and temperature

Z=
Ideal volume of the gas at specified pressure and temperature
2 41
Different methods can be used to calculate the Z-factor or compressibility factor

of a gas. The following section presents the theorem of corresponding states for the
calculation of a Z-factor. The theorem of corresponding states that the Z-factor of
any gas mixture is determined solely by the reduced temperature and pressure, as
indicated by Eq. (2.42).
Relationship Between Z-Factor and Reduced Pressure and Temperature

Z = f pr T r 2 42
68 2 Valve Sizing
where:
Z: Z-factor of the gas;
pr: Reduced pressure (MPa, psi);
Tr: Reduced temperature ( K, R).
The reduced temperature, pressure, and volume are calculated according to

Eqs. (2.43) and (2.44) by dividing the actual temperature, pressure, and
volume by the critical temperature, critical pressure, and critical volume,
respectively.
Reduced Temperature Calculation

T
Tr = 2 43
TC
where:
T: Actual temperature ( K, R);
TC: Critical temperature ( K, R);
Tr: Reduced temperature (dimensionless).
Critical temperature is defined as the maximum temperature of a gas, above

which the gas cannot be liquefied through the application of pressure alone,
regardless of the amount of pressure.
Reduced Pressure Calculation

P
Pr = 2 44
PC
where:
P: Actual pressure (psi, MPa);

PC: Critical pressure (psi, MPa);
Pr: Reduced pressure (dimensionless).
Critical pressure is defined as the pressure a gas exerts when it is in equilibrium

with a liquid phase and at the critical temperature. In other words, the required
pressure to liquefy the gas at a critical temperature.
Table 2.15 shows critical temperature and pressure for some common sub-
stances. In addition, the value of the compressibility factor can be determined from
Figure 2.14 as a function of reduced temperature and reduced pressure.
Table 2.15 The critical pressure and temperature for some common substances.
Critical temperature Critical temperature
Substance F C psig barg
Air −220.94 −140.52 549.08 37.858

Ammonia (NH3) 270 132.4 1636 112.8
Argon −188 −122 705.6 48.7
Butane 305.6 152 550.4 38
Carbon dioxide (CO2) 87.8 31.2 1071.6 73.8
Carbon monoxide (CO) −220.5 −140.3 507.5 35
Chlorine 291 144 118.7 77.1
Decane 653 345 301.7 20.8
Ethane 90.0 32.2 708 48.9
Ethanol (alcohol) 467 242 914 63
Ethylether 381 194 522 36
Ethylene 48.9 9.4 735 50.7
Fluorine −200 −129 808.5 55.8
Helium −456 −271 33.2 2.3
Hydrogen −400 −240 188.2 13.0
Hydrogen chloride 125 51.6 1198 82.7
Isobutane 275 135 592.2 36.5
Isobutylene 293 145 580 40
Isononane 590 310 335.1 23.1
Methane −117 −82.6 673.3 46.5
Nitrogen (N) −232.6 −147 492.4 34
Nitrous oxide (N2O) 97.4 36.4 1047.6 72.3
Oxygen (O2) −181.5 −118.6 732 50.5
Propane 206.1 96.7 617.4 42.6
Propylbenzene 689 365 464.2 32
Propylene 198 92.4 670.3 46.3
Undecane 691 366 287.2 19.8
Water 705 374 3206.2 220.5
70 2 Valve Sizing
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
1.0
0.9
0.8
Compressibility factor
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.5 4.0 4.5 5.0 5.5 6.0 6.5
Reduced pressure
Figure 2.14 Compressibility factor determination chart.
Example 2.9 If the relieving condition is 55 C and 62 bara, what is the com-
pressibility factor of ethylene?
Answer
According to Table 2.15, the critical temperature and pressure for ethylene are
9.4 C and 50.7 barg (51.7 bara), respectively. The next step is to find the reduced
temperature and pressure according to Eqs. (2.43) and (2.44), respectively. The
temperature values in Eq. (2.43) must be expressed in Kelvin (0 C + 273.15 =
273.15 K)
T 55 + 273 15 328 15
Tr = = = = 1 16
TC 9 04 + 273 15 282 19
P 62
Pr = = = 1 20
PC 51 7
The compressibility factor of the gas is approximately equal to Z = 0.71 accord-
ing to Figure 2.14.
Example 2.10 A conventional PSV is designed without its combination with a

rupture disk to handle hydrocarbon gas, which is a mixture of butane and pentane
with a molecular weight of 65, Z-factor = 0.84, and k = 1.09. The required hydro-
carbon flow rate is 24,260 kg/h. The operating temperature of the valve is 167 F
(627 R or 348 K). Meanwhile, the set pressure and the backpressure of the safety
valve are 75 psig and 14.7 psia, respectively. If an overpressure of 10% is tolerable,
select the nozzle size for the PSV based on API 526.
Answer
By using Eq. (2.36) as a starting point, the inlet pressure can be calculated as
follows:
P1 = Pset + Pover + Patmosphere − Ploss = 75 + 7 5 + 14 7 − 0 = 97 2 psia
The second step is to determine whether the flow rate is critical or subcritical. An
easy way to understand this is to compare the outlet pressure or backpressure with
the critical flow pressure. It can be concluded that this is a critical flow condition if
the backpressure is less than the critical flow pressure. The critical flow pressure
can be obtained from Eq. (2.40).
k k−1 1 09 1 09 − 1
2 2
Pcf = P1 Pcf = 97 2 ×
k+1 1 09 + 1
= 97 2 × 0 95712 11 = 97 2 × 0 587 = 57 08 psia
Due to the fact that the critical flow pressure (57.08 psia) is greater than the back-
pressure (14.7 psia), the critical flow condition is met.
In this case, the effective coefficient of discharge, Kd, should be considered equal
to 0.975. From Table 2.13, C = 326. Due to the use of a conventional safety valve,
the Kb factor, or capacity correction factor due to backpressure, is equal to one.
Furthermore, the combination correction factor for installation Kc is also equal
to one, since no rupture disk is used in conjunction with the safety valve. The other
parameters are in US Customary units except for the mass flow rate. So by convert-
ing the mass flow rate from kilograms per hour to pounds per hour, Eq. (2.28) can
be applied in this case for the sizing of a safety valve.
W = 24,260 kg h × 2 20462 lb kg = 53,484 lb h
W TZ 53, 484 627 × 0 84

A= =
C K d P1 K b K c M 326 × 0 975 × 97 2 × 1 × 1 65
= 1 7311 × 2 8465 = 4 93 in 2 = 3179 mm2

2 28b
In accordance with API 526, the size of the standard nozzle must be determined
using Table 2.12. In this example, nozzle designation N is 4.34 in.2 which is small in
size. Therefore, the smallest nozzle with a flow area greater than 4.93 in.2, nozzle
designation “P,” should be selected with the effective discharge area of 6.38 in.2
(4116 mm2).
72 2 Valve Sizing
Example 2.11 Currently, a PRV with a set pressure of 30 psig (207 kPag) is hand-
ling the air service with an airflow capacity of 227 kg/h discharge to the atmosphere.
The maximum overpressure that can be allowed is 3 psi and the temperature at the
inlet is 21 C. What type of API orifice should be used and what would be the nozzle
discharge area? (Note: the compressibility factor of air in this example is equal
to one.)
Answer
To begin with, it is necessary to identify the type of flow. According to Table 2.15, the
critical pressure for air is 37.858 barg (3785.8 kPag). According to Table 2.14, the spe-
cific heat of air is 1.40. The discharge pressure P1 is calculated as follows:
P1 = Set pressure + Overpressure + Atmospheric pressure

= 207 kPag + 20 7 kPag + 101 325 kPa = 329 025 kPag
The critical flow pressure is calculated according to Eq. (2.40) as follows:
Critical Flow Pressure Calculation

k k−1 14 04
2 2
Pcf = P1 = 329 025 × = 208 58 kPag
k+1 14+1
2 40b
The backpressure is equal to atmospheric pressure, which is 1.01325 bars

(101.325 kPag). The flow is considered critical since the backpressure is less than
the critical flow pressure.
The capacity correction factor due to backpressure is one since backpressure
equals atmospheric pressure. The value of gas constant C associated with the ratio
of specific heat k = 1.4 is 356 refer to Table 2.14. The same table indicates that the
molecular weight of air M is 29.96. Because the safety valve is designed for use with
air, the flow is critical, and SI units are in use, Eq. (2.29) is applicable in this case.
W = 227 kg h
T = 21 + 273 = 294 K
13,160 W TZ
A= SI units
C K d P1 K b K c M
13,160 × 227 294

A= = 26 828 × 3 1326
356 × 0 975 × 329 025 × 1 × 0 975 × 1 29 96
= 84 04 mm2 = 0 130,26 in 2
As a result, orifice E with an area of 0.196 in.2 should be selected according to

Table 2.12.
2.5.1.2 Subcritical Flow

Subcritical flows are defined as those in which the backpressure exceeds the crit-
ical pressure. As explained earlier, the critical pressure is calculated according to
Eq. (2.40). In the event the system operates under subcritical flow conditions, the
pressure relief device should be sized according to API 520 using Eqs. (2.45) and
(2.46) for mass flow rates, and Eqs. (2.47)–(2.50) for volumetric flow rates. It is nec-
essary to select a PRV that has a discharge area that is as large as, or larger than, the
calculated value of “A,” as shown in the following equations.
Sizing Relief Devices for Gas or Vapor in a Subcritical Flow Condition

According to Mass Flow Rate
W ZT
735 F 2 K d K c MP1 P1 − P2
17 9 × W ZT
A= SI units 2 46
F2K dK c MP1 P1 − P2
Sizing Relief Devices for Gas or Vapor in a Subcritical Flow Condition

According to Volumetric Flow Rate
V ZTM
4645 F 2 K d K c P 1 P 1 − P2
V ZTG
864 F 2 K d K c MP1 P1 − P2
47 95 × V ZTM
A= SI units 2 49
F2K dK c P 1 P1 − P 2
258 × V ZTG
A= SI units 2 50
F2K dK c P1 P1 − P2
where:
G: The specific gravity of a gas is compared to the specific gravity of air at standard
or normal conditions. It is crucial to know that G = 1 for air at 14.7 psia (101.325
KPa and 0 C);
F2: The coefficient of subcritical flow can either be derived from Figure 2.15 or
calculated from Eq. (2.51).
74 2 Valve Sizing
0.9
F2
0.8
0.7 k = 1.8
1.6 1.4 1.2
1.0
0.6
0,40 0,50 0,60 0,70 0,80 0,90 1,00
Back pressure to relieving pressure ratio (psi/psi)
Figure 2.15 Values of subcritical flow coefficient (F2).
Calculation of Coefficient of Subcritical Flow (F2)
k 2 k 1 − r k−1 k
F2 = r 2 51
k−1 1−r
where:
k: Ratio of specific heats;

r: Ratio of backpressure to upstream pressure, P2/P1.
Other parameters and their values are the same as those discussed previously for
the critical condition.
Example 2.12 A PSV that is installed on the top of a pressure equipment must
have a flow capacity of 53,500 lb/h. In the pressure equipment, the hydraulic fluid
is a mixture of butane and pentane containing a molecular weight of 65, a Z-factor
of 0.84, and a specific heat ratio of 1.09. The set pressure of the safety valve is equal
to the design pressure of the pressure equipment, which is equivalent to 75 psig
with 10% overpressure allowed. The relieving temperature is 627 R, correspond-
ing to 167 F. In this conventional safety valve, a constant backpressure of 55 psi
gauge as well as 7.5 psig buildup backpressure are applied. According to API stan-
dards, what size orifice is required for this valve?
Answer
The first step is to determine whether the fluid is critical or subcritical.

= 75 + 7 5 + 14 7 = 97 2 psia
k k−1 1 09 0 09
2 2
Pcf = P1 = 97 2 × = 57 03 psia
k+1 1 09 + 1
Total back pressure = 55 psig + 7 5 psig = 62 5 + 14 7 = 77 2 psia
As the backpressure (77.2 psia) is higher than the critical pressure (57.03 psia),
the flow is subcritical. Because the flow in this case is subcritical and given as mass
flow rate, and since the units of measurement are US Customary, Eq. (2.45)
applies. Obtaining the coefficient of subcritical flow F2 is the next step. To do
so, first the ratio of backpressure to relieving pressure must be calculated as
follows:
P2 77 2 psia
Ratio of backpressure = = = 0 794
P1 97 2 psia
It should also be noted that the coefficient of subcritical flow F2 which is equal
to 0.86 in Figure 2.15 is a consequence of combining the ratio of specific heat
(k = 1.09) with the ratio of backpressure (0.794). The value of coefficient of
discharge Kd = 0.975. In addition, Kc is the combination correction factor for
installation with a rupture disk upstream of the PRV. This value is one in this case
because there is no rupture disk installed. Inlet relieving temperature should be in
Rankine unit, which is 627 .
W ZT
A=
735F 2 K d K c MP1 P1 − P2
53,500 0 84 × 627
=
735 × 0 86 × 0 975 × 1 65 × 97 2 × 97 2 − 77 2
= 86 81 × 0 6456 = 5 60 in 2 = 3613 mm2
In accordance with Table 2.12, the orifice size of “P” with a discharge area of
6.38 in.2 = 4116 mm2 should be selected.
2.5.2 Sizing for Steam Relief

Equations (2.52) and (2.53) are used to size safety valves in the steam service in US
Customary and SI units, respectively.
76 2 Valve Sizing
Sizing Relief Devices for Steam
W
51 5 × P1 K d K b K c K N K SH
190 4 × W
A= SI units 2 53
P1 K d K b K c K N K SH
where:
A: Minimum required effective discharge area, in.2(mm2);
W: Required flow rate or capacity lb/h (kg/h);
P1: The upstream relieving pressure, psia (kPaa). Upstream or inlet relieving pres-
sure equals the sum of the set pressure, allowable overpressure and atmospheric
pressure;
Kd: Effective coefficient of discharge that is equal to 0.975 when a pressure safety or
relief valve is installed with or without a rupture disk in combination. Alterna-
tively, this value is equal to 0.62 when the PRV is not installed and the sizing is
done for a rupture disk;
Kb: The backpressure correction factor was discussed earlier in the chapter as to
how it can be calculated;
KC: Combination correction factor due to the installation of a rupture disk
upstream of a PSV. The value is one if no rupture disk is installed, and 0.9 if
a rupture disk is installed in conjunction with a safety valve;
KN: Correction factor for the Napier equation. The dimensionless correction factor
is only applicable to the steam flow as per API 520 and is calculated according to
Eqs. (2.54) and (2.55). The factor is only used if the steam flow inlet relieving
pressure P1 is in the range of 1500–3200 psia. The correction factor is
equal to one if the inlet relieving pressure is less than or equal to 1500 psia
(10,339 kPa).
Correction Factor for the Napier Equation Applicable to Steam Flow

K N = 1 if P1 ≤ 1500 psia 10,339 kPa
2 54
If 1500 psia 10, 339 kPa < P1 ≤ 3200 psia 22, 057 kPa
0 1906 × P1 − 1000
KN = US Customary units
0 2292 × P1 − 1061
2 55
0 02764 × P1 − 1000
KN = SI units
0 03324 × P1 − 1061
KSH: Superheat correction factor that can be obtained from Table 2.16. For satu-
rated steam at any pressure, KSH = 1. Steam that is saturated occurs when
the liquid and gaseous phases of water are present simultaneously at a given
Table 2.16 Superheat steam correction factor, KSH, according to API 520 standard.
Set Temperature ( F)
pressure
(psig) 300 400 500 600 700 800 900 1000 1100 1200
15 1.00 0.98 0.93 0.88 0.84 0.80 0.77 0.74 0.72 0.70
20 1.00 0.98 0.93 0.88 0.84 0.80 0.77 0.74 0.72 0.70
40 1.00 0.99 0.93 0.88 0.84 0.81 0.77 0.74 0.72 0.70
60 1.00 0.99 0.93 0.88 0.84 0.81 0.77 0.75 0.72 0.70
80 1.00 0.99 0.93 0.88 0.84 0.81 0.77 0.75 0.72 0.70
100 1.00 0.99 0.94 0.89 0.84 0.81 0.77 0.75 0.72 0.70
120 1.00 0.99 0.94 0.89 0.84 0.81 0.78 0.75 0.72 0.70
140 1.00 0.99 0.94 0.89 0.85 0.81 0.78 0.75 0.72 0.70
160 1.00 0.99 0.94 0.89 0.85 0.81 0.78 0.75 0.72 0.70
180 1.00 0.99 0.94 0.89 0.85 0.81 0.78 0.75 0.72 0.70
200 1.00 0.99 0.95 0.89 0.85 0.81 0.78 0.75 0.72 0.70
220 1.00 0.99 0.95 0.89 0.85 0.81 0.78 0.75 0.72 0.70
240 1.00 1.00 0.95 0.90 0.85 0.81 0.78 0.75 0.72 0.70
260 1.00 1.00 0.95 0.90 0.85 0.81 0.78 0.75 0.72 0.70
280 1.00 1.00 0.96 0.90 0.85 0.81 0.78 0.75 0.72 0.70
300 1.00 1.00 0.96 0.90 0.85 0.82 0.78 0.75 0.72 0.70
350 — 1.00 0.96 0.90 0.86 0.82 0.78 0.75 0.72 0.70
400 — 1.00 0.96 0.91 0.86 0.82 0.78 0.75 0.72 0.70
500 — 1.00 0.96 0.92 0.86 0.82 0.78 0.75 0.73 0.70
600 — 1.00 0.97 0.92 0.87 0.82 0.79 0.75 0.73 0.70
800 — — 1.00 0.95 0.88 0.83 0.79 0.76 0.73 0.70
1000 — — 1.00 0.96 0.89 0.84 0.78 0.76 0.73 0.71
1250 — — 1.00 0.97 0.91 0.85 0.80 0.77 0.74 0.71
1500 — — — 1.00 0.93 0.86 0.81 0.77 0.74 0.71
1750 — — — 1.00 0.94 0.86 0.81 0.77 0.73 0.70
2000 — — — 1.00 0.95 0.86 0.80 0.76 0.72 0.69
2500 — — — 1.00 0.95 0.85 0.78 0.73 0.69 0.66
3000 — — — — 1.00 0.82 0.74 0.69 0.65 0.62
78 2 Valve Sizing
temperature and pressure. Simplified, steam is at equilibrium with the heated

water. In the saturated condition, the rate at which water is vaporized is equal to
the rate at which it is condensed.
Example 2.13 Choose the correct size of PSV that can handle saturated steam
with a relief capacity of 153,000 lb/h. As the set pressure of the valve is 1600 psig,
there is a 10% allowance for overpressure. The valve is designed to discharge into
the atmosphere. A note should be made that a rupture disk is not used in the safety
system in order to prevent catastrophic failure.
Answer
Since the PSV is used for steam and the data given are in US Customary units,
Eq. (2.52) is applicable. Obtaining the values of the parameters in the equation
is an important step before using the equation.
W = 153,000 lb h;
P1 = Set pressure + Overpressure + 14 7
= 1600 + 0 1 × 1600 + 14 7 = 1774 7 psia
Kd: An effective coefficient of discharge of 0.975 is achieved when a pressure safety

or relief valve is installed without a rupture disk combined with a pressure
safety or relief valve.
Kb: The backpressure correction factor is equal to one in this case because the valve
discharge is into the atmosphere.
KC: The combination correction factor is due to the installation of a rupture disk
upstream of a PSV. It is equal to one since there is no rupture disk installed.
KN: According to Eq. (2.54), the correction factor for the Napier equation can be
calculated as follows.
0 1906 × P1 − 1000 0 1906 × 1774 7 − 1000 − 661 74

KN = = = = 1 011
0 2292 × P1 − 1061 0 2292 × 1774 7 − 1061 − 654 24
Superheat correction factor KSH = 1 because saturated steam is handled by

the valve.
W 153,000
A= =
51 5 × P1 K d K b K c K N K SH 51 5 × 1774 7 × 0 975 × 1 × 1 011 × 1
= 1 698 in 2
In accordance with Table 2.12, the selected orifice is K with an effective dis-
charge area of 1.838 in.2.
2.5.3 Sizing for Liquid Relief

In order to properly size a PSV for liquid relief, the following classifications are
made; PRVs that require capacity certification and PRVs that do not require capac-
ity certification.
2.5.3.1 Sizing for Liquid Relief with Capacity Certification

According to ASME Sec. VIII Div. 01, it is mandatory for safety valves designed for
liquid services to obtain capacity certification. An application for capacity certifi-
cation involves testing the valve to determine the rated coefficient of discharge for
liquid valves at 10% overpressure in order to obtain certification. Thus, safety
valves in liquid services that need to comply with ASME code requirements con-
cerning the capacity certificate need to be sized according to Eqs. (2.56) and (2.57).
Sizing Relief Devices for Liquid According to ASME Code
Q G
38 K d K w K c K v P1 − P 2
11 78 × Q G
A= SI units 2 57
K dK w K cK v P1 − P2
where:
A: Required discharge area, in.2 (mm2);

Q: Flow rate US gallons per minute (l/min);
Kd: This is the coefficient of discharge that should be obtained from the valve man-
ufacturer. The value of the discharge coefficient, which is used for a preliminary
valve sizing calculation, can be determined as follows: It reaches a value of 0.65
when a pressure safety or relief valve is installed either with or without a rupture
disk. It is also worth noting that this value is equal to 0.62 when the PRV is not
installed, and the sizing is done for a rupture disk;
Kw: There is a term for this effect called backpressure correction and it is equal to
one if the backpressure is discharged directly into the atmosphere. For bellows
balanced type safety valves, however, the backpressure correction factor must
be obtained from Figure 2.16;
KC: This is a combination correction factor due to the installation of a rupture disk
upstream of a PSV. There is a value of one if there is no rupture disk installed;
however, the value becomes 0.9 if the rupture disk in conjunction with the
safety valve is installed;
Kv: This is the correction factor due to viscosity that can be obtained from Eq. (2.58)
or Figure 2.17. The value of the correction factor as a result of viscosity derived
from the equation or figure depends on the Reynolds number, which will be
80 2 Valve Sizing
1.00
0.95
0.90
0.85
0.80
Kw
0.75
0.70
0.65
0.60
0.55
0.50
0 10 20 30 40 50
Percent of gauge backpressure = (PB/PS) × 100
Kw = correction factor due to back pressure.
PB = back pressure, in psig.
PS = set pressure, in psig.
Figure 2.16 Backpressure correction factor for balanced bellows PRVs in liquid service
(parameter Kw).
1.0
0.9
Kv viscosity correction factor
0.8
0.7
0.6
0.5
0.4
0.3
101 102 103 104 105
R = Reynold’s number
Figure 2.17 Viscosity correction factor for balanced bellows PRVs in liquid service
(parameter Kv).
explained in more detail further in this chapter. The value of Kv = 1 for a non-
viscous fluid.
Calculation of Correction Factor Due to Viscosity (Kv)

−1 0
2 878 342 75
Kv = 0 9935 + + 2 58a
R0 5 R1 5
G: The specific gravity of the liquid corresponds to the density of the liquid relative
to that of water at the temperature of the liquid.
P1: The upstream relieving pressure, psia (kPaa). Upstream or inlet relieving pres-
sure equals the sum of the set pressure, allowable overpressure and atmospheric
pressure.
P2: Backpressure psig (kPag)
First, a safety valve is sized for a nonviscous fluid where Kv = 1. After this, the
calculated orifice discharge area is adjusted based on the Reynolds number and the
viscosity correction factor. It is important to note that after the Reynolds number is
determined from either Figure 2.17 or any other equation explained in the next
section, the viscosity correction factor will be calculated and then applied to the
calculated preliminary discharge area.
Using the Reynolds number (R), it is possible to predict flow patterns in various
fluid flow situations. At low Reynolds numbers, the flow pattern tends to be lam-
inar (sheet-like), whereas at high Reynolds numbers, the flow pattern tends to be
turbulent. The Reynolds number can be calculated by taking the absolute viscosity
at the flowing temperature in centipoises, shown with parameter μ in Eqs. (2.58)
and (2.59), or the viscosity at the flowing temperature in saybolt universal seconds
(SSUs), shown with parameter U as per Eqs. (2.60) and (2.61). Viscosity is a meas-
ure of the resistance that a fluid exhibits when it is being deformed by either a
shear or a tensile force. The term viscosity is derived from the Latin word viscum,
which is a reference to anything that sticks. A fluid with a low viscosity is referred
to as thin, while a fluid with a high viscosity is referred to as thick. Viscosity
describes the internal friction of moving fluids, and a fluid with a higher amount
of viscosity is resistant to movement since the molecules of the fluid create a lot of
internal friction. Low-viscosity fluids, on the other hand, flow better. In
Figure 2.18, there is a comparison between the viscosity of water and honey as
a function of their gravitational flow rates.
Calculation of Reynolds Number Based on Parameter μ

Q × 2800 × G
R= US Customary units 2 58b
μ× A
82 2 Valve Sizing
Figure 2.18 Viscosity comparison between water and honey by studios guy.
Source: https://imgur.com/gallery/O9ahk.
Q × 18,800 × G
R= SI units 2 59
μ× A
Calculation of Reynolds Number Based On Parameter μ
12,700 × Q
R= US Customary units 2 60
U× A
85,220 × Q
R= SI units 2 61
U× A
where:
R: Reynolds number;
Q: Flow rate at the given temperature, US gallons per minute (gpm) (l/min);
G: A liquid’s specific gravity is determined by comparing its density to water at the
liquid’s temperature;
μ: Absolute viscosity at the flowing temperature, centipoise;
A: Effective discharge area of the safety valve in.2 (mm2);
U: Viscosity at the flowing temperature in saybolt universal seconds (SSUs).
Example 2.14 It is essential to size the safety relief valve properly so that it can
handle crude oil with a specific gravity of 0.85 and a viscosity of 2000 saybolt uni-
versal seconds (SSUs) at the flow temperature. The valve has a flow capacity of
2000 gpm, a set pressure of 200 psig, and a backpressure of 40 psig. For valve sizing,
what are the discharge flow area and orifice letter with a 10% allowable overpres-
sure scenario?
Answer
The following data can be obtained from this example:
Q = 2000 gpm, G = 0 85, and U = 2000 SSUs;
P1 = Set pressure + Overpressure + 14 7 = 200 + 0 1 × 200 + 14 7
= 234 7 psia = 220 psig
P2 = Backpressure = 40 psig
Kd: When a pressure safety or relief valve used for liquid services is installed with-
out a rupture disk together with a pressure safety or relief valve, an effective
coefficient of discharge of 0.65 can be achieved.
Kw: The term “backpressure correction” is commonly used for this effect and the
backpressure correction factor can be obtained from Figure 2.16 in the case of
bellows balanced type safety valves. The value of Kw depends on the ratio of
backpressure to set pressure Pb/PS.
Pb 40
= × 100 = 20 K w = 0 97
PS 200
KC: The combination correction factor is due to the installation of a rupture disk
upstream of a PSV. It is equal to one since there is no rupture disk installed.
Due to the fact that the values given are in US Customary units, Eq. (2.56) is used
for sizing the safety valves. But one should be remembered that the valve should be
sized first for a viscid fluid with Kv = 1.
Q G 2000 0 85
AR = =
38 K d K w K c K v P1 − P2 38 × 0 65 × 0 97 × 1 × 1 220 − 40
= 83 4759 × 0 0687 = 5 73 in 2
Using the formula above, the orifice area calculated above equals 5.73 in.2 with-
out considering viscosity correction and is shown as AR. The next step is to deter-
mine the appropriate orifice discharge area that corresponds to the designated
letter as per Table 2.12. The orifice size that is suitable at this stage to calculate
Reynolds number would be “P” with a discharge area of 6.38 in.2. It is then
necessary to calculate the Reynolds number by applying Eq. (2.60) because all
units are in US Customary and the given viscosity is expressed as a parameter U.
84 2 Valve Sizing
12,700 × Q 12,700 × 2000 12,700

R= = = = 5020
U× A 2000 × 6 38 2 53
We can now calculate the correction factor due to viscosity by using Eq. (2.58).
This calculation can be done by the following formula:
−1 0 −1 0
2 878 342 75 2 878 342 75
Kv = 0 9935 + + = 0 9935 + +
R0 5 R1 5 50200 5 50201 5
−1 0
2 878 342 75 −1 0
= 0 9935 + + = 0 9935 + 0 0406 + 0 00096
70 85 3,55,677
−1 0
= 1 03506 = 0 966
As a result, the final orifice discharge area is calculated in the following manner:
AR 5 73
A= = = 5 93 in 2
Kv 0 966
Upon determining the appropriate discharge area for the orifice in accordance
with the designated letter, as shown in Table 2.12, the next step would be to select
the appropriate orifice. For this purpose, the optimum orifice size that would be
suitable at this point would be “P” with a discharge area of 6.38 in.2.
2.5.3.2 Sizing for Liquid Relief Without Capacity Certification

According to ASME Code requirements pertaining to the capacity certificate,
safety valves in liquid services that do need to comply with the ASME code require-
ments need to be sized according to Eqs. (2.62) and (2.63).
Sizing Relief Devices for Liquid Without Capacity Certification
Q G
38K d K w K c K v K p 1 25 P − Pb
11 78 × Q G
A= SI units 2 63a
K dK w K cK v K p 1 25 P − Pb
Kd: The coefficient of discharge should be obtained from the manufacturer of the
valve in question. To determine the preliminary size in this case, 0.62 should be
taken into account.
Kp: It is the correction factor due to overpressure. If the overpressure is considered
25%, the value of the overpressure correction factor is equal to one. Otherwise,
the value of this parameter must be obtained from Figure 2.19. A glance at the
curve in Figure 2.19 indicates that capacity is affected by a change in lift, orifice
discharge coefficient, and change in overpressure up to and including 25%
1.1
1.0
0.9
0.8
Correction factor, Kp
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
10 20 30 40 50
Percent overpressure
Figure 2.19 Capacity correction factor due to overpressure for noncertified pressure safety
valve (PSV) in liquid services.
overpressure. As capacity rises above 25%, the change in overpressure is the only
factor that affects it. The use of noncertified valves under low overpressure can
cause chattering and, therefore, scenarios with more than 10% overpressure
should be avoided.
P: Set pressure psig (kPag);

Pb: Total backpressure psig (kPag);
The other parameters and their values are the same for both certified and non-
certified safety valves in liquid services.
2.5.4 Sizing for Two-Phase Liquid/Vapor Relief

There are various reasons why two-phase flows occur in safety valves. Some of
them are listed here:
• During the safety valve’s operation, a liquid vaporizes;

• From the piping or equipment connected to the valve, a two-phase mixture of
liquid and vapor enters the valve;
86 2 Valve Sizing
Table 2.17 Two-phase flow scenarios.
Two-phase liquid/vapor relief scenario Section
The valve is filled with saturated liquid and saturated vapor and 2.5.4.1
the liquid flashes. There is no noncondensable gas present
(flashing flow).
Condensation of supercritical fluid in the safety valve. 2.5.4.1
High subcooled liquids and either noncondensable gas or 2.5.4.2
condensable vapors, or both, enter the valve, but the liquid does
not flash (frozen flow).
The subcooled liquid enters the valve and flashes. At the inlet, 2.5.4.2
there is no vapor or gas present.
Generic two-phase flow involving a subcooled or saturated liquid Not covered in this
and a noncondensable gas with or without condensable vapor. chapter
• Vapor condenses within the safety valve;

• A supercritical fluid is introduced into the safety valve and condenses.
The two-phase flow mixture is likely to discharge from the safety valve in all four
of the cases mentioned earlier. If a relief valve handles a liquid when the liquid is at
equilibrium between vapor and liquid, this may result in the production of mixed
flow as a result of vapor generation. As a result of the vapor generation, the mass
flow capacity of the valve decreases. According to Table 2.17, there are five possible
scenarios that could occur regarding the two-phase liquid/vapor flow through the
safety valves. We will cover four out of five scenarios in the following sections.
The sizing procedure for two-phase flow is given in Appendix D of API 520 based
on a method called the “Omega method of Leung.” For this method of sizing, the
Omega parameter ω is used, which is a measure of the compressibility of the two-
phase mixture. There are the following steps that must be followed to successfully
complete this method:
• Calculation of the Omega parameter;

• The determination of whether the flow is critical or subcritical;
• Calculation of the mass flux, which refers to the mass flow per unit of area;
• Calculation of the required orifice discharge area and selection of the final ori-
fice size.
The parameters used to determine the size of two-phase safety valves and their
units are presented as follows:
Cp: Specific heat at constant pressure of the liquid at the inlet of the safety valve
(Btu/lb R);
G: Mass flux (lb/s ft2);

hvl0: Latent heat of vaporization at the safety valve inlet. In multicomponent sys-
tems, it is the difference between the vapor and liquid-specific enthalpies at the
safety valve inlet (Btu/lb);
hvls: Latent heat of vaporization at PS; For multicomponent systems, it is the dif-
ference between the vapor and liquid specific enthalpies at PS (Btu/lb);
k: Ratio of specific heats of the vapor. If its value is unknown, it should be assumed
to be equal to one (dimensionless);
P1 or P0: Pressure at safety valve inlet (psia) that is equal to set pressure plus the
allowable overpressure plus atmospheric pressure;
Pa: Downstream backpressure (psia);
Pc: Critical pressure (psia);
Pr: Reduced pressure (psia);
Ps: Saturation pressure (single-component flows) or bubble point pressure (multi-
component flows) at the relieving temperature T0 (psi);
Q: Volumetric flow rate (gpm);
T0: Temperature at safety valve inlet ( R);
Tr: Reduced temperature ( R);
vv0: Specific volume of the vapor at safety valve inlet (ft3/lb);
v0: Specific volume of the two-phase mixture at safety valve inlet (ft3/lb);
vg0: Specific volume of the vapor, gas or combined vapor and gas at the safety valve
inlet (ft3/lb);
vvl0: Difference between the vapor and the liquid specific volumes at the safety
valve inlet (ft3/lb);
vvls: Difference between the vapor and the liquid specific volumes at Ps (ft3/lb);
v9: Specific volume evaluated at 90% of the safety valve inlet pressure (= relieving
pressure), assuming isentropic flashing (ft3/lb);
x0: Vapor (or gas or combined vapor and gas) mass fraction (quality) at safety valve
inlet (dimensionless);
ηa: Ratio between ambient pressure and relieving pressure = Pa/P1
(dimensionless);
ηc: Ratio between critical pressure and relieving pressure (dimensionless);
ηs: Ratio between saturation pressure at relieving temperature and relieving pres-
sure (dimensionless);
ρ10: Density of the liquid at the inlet of the safety valve (lb/ft3);
ρ9: Density evaluated at 90% of the saturation pressure (single-component flows)
or bubble point pressure (multicomponent flows) Ps at T0. The flash calcula-
tion shall be done isentropically (lb/ft3);
ω: Omega parameter (dimensionless);
ωs: Omega parameter for subcooled liquid flows at safety valve inlet
(dimensionless).
88 2 Valve Sizing
2.5.4.1 Sizing for Saturated Liquid and Saturated Vapor, Liquid Flashes
Step 1 Calculate the Omega Parameter (ω)
The definitions of the Omega-Parameter in Eqs. (2.64)–(2.66) can be employed for
multicomponent systems, whose nominal boiling range, that is the difference in
the atmospheric boiling points of the heaviest and the lightest components, is less
than 150 F. For single-component systems with reduced temperature Tr ≤ 0.9
(see Eq. 2.43) and pressure (see Eq. 2.44) Pr ≤ 0.5, either Eq. (2.64) or Eq.(2.65) can
be used.
Calculation of Omega Parameter (ω)
2
x 0 vvo P1 × vvl0 Cp T 0 P1 vvl0
ω= × 1 − 0 37 + 0 185 × 2 64
v0 hvl0 v0 hvl0
2
x 0 vvo Cp T 0 P1 vvl0
ω= + 0 185 × 2 65
v0 k v0 hvl0
Whenever a multicomponent system, whose nominal boiling range is greater

than 150 F, or whenever a single component system is close to its thermodynamic
critical point or supercritical fluids are involved in condensing two-phase flows,
Eq. (2.66) must be used.
v9
ω=9 −1 2 66
v0
Step 2 Determine If the Fluid is Critical or Subcritical
The two-phase flow is critical if the critical pressure is larger than the
backpressure.
If Pc > Pa Critical flow
If Pc < Pa Subcritical flow
The critical pressure can be calculated using Eq. (2.67) as follows:
P c = ηc × P 1 2 67a
In order to determine the ratio of critical pressure ηc either from Eq. (2.68) or
from Figure 2.20 can be used.
ηc 2 + ω2 − 2ω 1 − ηc 2 + 2ω2 ln ηc + 2ω2 1− ηc = 0 2 68a
Step 3 Calculate the Mass Flux

Calculation of the mass flux G for critical and subcritical flow is based on
Eqs. (2.68) and (2.69), respectively.
Calculation of Mass Flux for Critical Flow

P1
G = 68 09 × ηc × 2 68b
ωv0
1.4
1.2
Non-flashing flow Flashing flow

1.0
Critical pressure ratio, ηc
0.8
0.6
0.4
ηc2 + (ω2 – 2ω)(1 – ηc)2 + 2ω2/nηc + 2ω2 (1 – ηc) = 0

0.2
0.0
0.01 0.1 1 10 100
Omega parameter, ω
Figure 2.20 Critical flow ratio determination based on Omega parameter.
Calculation of Mass Flux for Subcritical Flow

P1 − 2 × ω ln Pa P1 + ω − 1 1 − Pa P1
G = 68 09 × 2 69
v0 ω P1 Pa − 1 + 1
Step 4 Calculate the Required Orifice Area of the Safety Valve

Finally, the required area of the safety valve can be computed from Eq. (2.70).
Calculation of the Required Orifice Area of the Safety Valve in Sizing

for Saturated Liquids and Saturated Vapors, Liquid Flashes
1 W
A = 0 04 × × 2 70
K bK cK d G
where:
A: Required discharge area, in.2;
W: Mass flow rate (lb/h);
Kd: This is the coefficient of discharge that should be obtained from the valve man-
ufacturer. The value of the discharge coefficient, which is used for a preliminary
valve sizing calculation, can be considered 0.85 for this case;
Kb: A backpressure correction factor may also be referred to as a capacity correc-
tion factor. It should be noted that backpressure applies only to balanced bel-
lows valves, and it can be obtained from Figure 2.12;
90 2 Valve Sizing
KC: This is a combination correction factor due to the installation of a rupture disk
upstream of a PSV. There is a value of one if there is no rupture disk installed;
however, the value becomes 0.9 if the rupture disk in conjunction with the
safety valve is installed.
Example 2.15 A safety valve must be sized for a two-phase flow of water and its
saturated steam. The set pressure of the safety valve is 145 psig (10% overpressure)
and the valve is expected to deliver a flow rate of 275,600 lb/h. In addition, the dis-
charge of the safety valve is to the atmosphere. The saturated temperature is 830
R, which is the temperature at the inlet of the safety valve. At that temperature, the
physical properties of saturated water and steam are provided as indicated in
Table 2.18. What orifice area and designation must be chosen for this case? (Note:
Vapor or gas or combined vapor and gas mass fraction (quality) at safety valve inlet
is equal to zero.)
Answer
The relieving pressure P is calculated as follows:

= 145 + 14 5 + 14 7 = 174 2 psia
Since the mass fraction of vapor or gas as well as the combined mass fraction of
gas and vapor at the safety valve inlet is equal to zero, it can be concluded that
x0 = 0. The temperature at the valve inlet T0 is 830 R.
The next step is to calculate parameter Omega by using Eq. (2.65) in the follow-
ing manner:
2
x 0 vvo Cp T 0 P1 vvl0
ω= + 0 185 × = 0 + 0 185
v0 k v0 hvl0
2
1 06116 × 830 × 174 2 2 6142 − 0 01824
×
0 01824 853 667
= 1,556,159 14 × 9 25 × 10 − 6 = 14 39
Table 2.18 Water and stem properties at 830 R – the

saturated temperature for water and stem.
vv0 2.6142 ft3/lb
vl0 and v0 ft 3
0.01824 ft3/lb
lb
hvl0 853.667 Btu/lb
Cp 1.06116 Btu/lbR
Following this step, it is necessary to determine whether the fluid is critical or

subcritical. In order to do so, the ratio between critical pressure and relieving pres-
sure ηc = 0.877 is obtained from Figure 2.20.
Pc = ηc × P1 = 0 877 × 174 2 = 152 77 psia 2 67b
Therefore, since the safety valve discharges into the atmosphere, the backpres-
sure value is 14.7 psia, which is less than the critical pressure. So the flow in this
example is considered critical. As a result, in the following step, the mass flux is
calculated in accordance with Eq. (2.68).
P1 174 2
G = 68 09 × ηc × = 68 09 × 0 877 × = 1538 06 lb sft2
ωv0 14 39 × 0 01824
Backpressure correction factor is equal to one because the fluid in the safety valve
is discharged to the atmosphere. The combination correction factor is also equal to
one because to rupture disk is installed in combination with the safety valve.
1 W 1 275, 600
A = 0 04 × × = 0 04 × × = 8 43 in 2
K bK cK d G 1 × 1 × 0 85 1538 06
These calculations result in the selection of orifice Q with a discharge area of

11.05 in.2.
2.5.4.2 Sizing for Subcooled at the Pressure Relief Valve Inlet

As a result of the method presented in this section, a PRV can be sized to handle
a subcooled fluid with a saturated liquid at the inlet using the method described
in this section. Neither noncondensable vapor nor noncondensable gases should
be present at the inlet. If the flow is subcooled, the subcooled liquid flashes
upstream or downstream of the valve, depending on which subcooling region
it falls into.
Step 1 Calculate the Omega Parameter (ω)

The sizing starts with obtaining Omega parameter same as the previous section.
For subcooled liquid flows the Omega-Parameter is generally referred with ωs.
For multicomponent systems with nominal boiling range less than 150 F ωs
can be calculated either from Eq. (2.71) or from Eq. (2.72). For single component
systems with a relative temperature and pressure within the limits Tr ≤ 0.9 and
Pr ≤ 0.5, ωs is given by Eq. (2.71). For multicomponent systems, whose nominal
boiling range is greater than 150 F or for single-component systems close to
the thermodynamic critical point, ωs is given by Eq. (2.72).
92 2 Valve Sizing
Calculation of Omega Parameter (ωs)

vvls 2
ωs = 0 185 ρl0 C P T 0 Ps 2 71
hvls
ρ10
ωs = 9 −1 2 72
ρ9
Step 2 Determine the Subcooling Region

In the event that a liquid enters the safety valve in a subcooled state, it is necessary
to determine whether or not it saturates and the extent of the subcooling region on
the basis of the following equations:
2ωs
If Ps > P0 × 2 73
1 + 2ωs
Low subcooling region where flashing occurs before the valve throat or ori-
fice area
2ωs
If Ps < P0 × 2 74
1 + 2ωs
High subcooling region where flashing occurs at the valve throat or ori-
fice area
Step 3 Determine If the Flow is Critical or Subcritical

The condition for the existence of critical and subcritical flow is provided in
Table 2.19. In particular, it should be noted that the value of critical pressure is
calculated from Eq. (2.67).
Step 4 Calculation of Mass Flux (G)

Equations (2.75) and (2.76) are used to calculate the mass flux under both low and
high subcooling conditions.
Calculation of Mass Flux for Sizing the PSV in Low-Subcooling Systems

05
2 1 − ηs + 2 ωS ηs ln ηs η − ωS − 1 ηs − η
G = 68 09 × P1 ρl0
ωS ηs η − 1 + 1
2 75
η = ηc for critical flow and η = ηa for subcritical flow
Table 2.19 Choosing critical or subcritical flow for PSV sizing in a

subcooling system.
Critical flow Subcritical flow
Low subcooling region Pc > Pa Pc < Pa

High subcooling region Ps > Pa Ps < Pa
Calculation of Mass Flux for Sizing the PSV in High-Subcooling Systems

05
G = 96 3 ρl0 × P1 − P 2 76
For critical flow condition, P = Ps.

For subcritical flow condition, P = Pa.
Step 5 Calculate the Required Orifice Area of the Safety Valve

Calculation of the Required Orifice Area for Sizing a Subcooled

Two-Phase Flow Safety Valve
1 Q ρl0
A = 0 3208 × × 2 77
K bK cK d G
Kd: It is important to obtain the discharge coefficient from the valve manufacturer.
In order to calculate a preliminary valve sizing calculation, the discharge coef-
ficient can be considered to be 0.65 for subcooled liquids and 0.85 for saturated
liquids.
Kb: The backpressure correction factor is also known as the capacity correction
factor. Backpressure only applies to balanced bellows valves, and it can be cal-
culated from Figure 2.12.
KC: This is a combination correction factor resulting from the installation of a rup-
ture disk upstream of a PSV. When there is no rupture disk installed, the value is
one. However, if the rupture disk in conjunction with the safety valve is
installed, the value becomes 0.9.
2.5.5 Sizing for Fire Case and Hydraulic Expansion

The cause of a fire may be due to the leakage of flammable material from equip-
ment and pipelines, or an accident may occur during the operation. When acciden-
tally ignited, this material will immediately pose a danger to adjacent vessels and
equipment. In liquids and gases, burning materials can become an open, free-
burning fire very quickly and can be carried some distance from the source of
the leak by the slope of the ground and by air currents. In operating plants that
handle or process flammable liquids or gases, fire hazards must be taken into
account when selecting the size of safety relief valves. Under normal operating
conditions, any pressure vessel protected by a PRV should be fire-sized in the event
that the equipment may be exposed to fire (even if the contents are not flammable).
When an open fire occurs surrounding equipment or vessels, heat will naturally be
absorbed by anything that comes in contact with the flames and/or hot gases. In
the event of prolonged heat absorption, the contents of the vessel will be heated
and the pressure will rise until the safety relief valve opens.
94 2 Valve Sizing
API 521 and ISO 23251 are two standards dealing with safety requirements for
systems that relieve and depressurize pressure. As shown in this section, we pro-
vide engineering and required equations for determining the size of a safety
valve in two cases of hydraulic expansion and external fire. Thermal expansion
or hydraulic expansion refers to the increase in the volume of liquid caused by
an increase in temperature. A typical instance of this occurs for liquids trapped
in vessels, pipes, heat exchangers and exposed to heat, for example, from elec-
trical coils, ambient heat, fire, etc. The formulas in both standards are identical,
except for the units. The API 521 formulas should be applied using US units,
whereas the ISO 23251 formulas should be applied using SI units. Table 2.20
presents the required parameters and units based on these two standards. With
respect to sizing external fire cases, API 521 distinguishes between wetted and
Table 2.20 Symbols for sizing safety valves in fire cases.
Symbol Description Units (US) Units (SI)
a
A Effective discharge area of valve in.2
a
A Exposed surface area of the vessel ft2
AWS Total wetted surface ft2
m2
αv Cubical expansion coefficient of the liquid at the 1/ F 1/ C
expected temperature
C Specific heat capacity of the trapped liquid Btu/(lb F) J/(kg k)
F Environment factor — —
d Relative density referred to water at 60 F (15.6 C) — —
hvl0 Latent heat of vaporization Btu/lb J/kg
KD Coefficient of discharge — —
Total heat transfer rate Btu/h W
M Molecular mass of the gas lb/lbmol kg/kmol
a
P1 Upstream relieving absolute pressure psi
Q Total absorbed (input) heat to the wetted surface Btu/h J
q Volume flow rate at the flowing temperature gpm m3/s
a
qm or Q Relief load/mass flow rate lb/h
a
T1 Gas temperature at upstream relieving pressure R
a
Tw Recommended maximum vessel wall temperature R
a
ISO 23251 specifies sizing formulas that are identical to those in API 521, which are expressed in
US units. The conversion factors to specified SI units have not yet been provided. It is therefore
recommended that the formula be applied using US units.
unwetted vessels based on the following definitions, and presents a sizing

method for each.
Wetted vessels contain liquids in equilibrium with their vapors or gases. These
vessels contain temperate systems. Due to the heat transfer from the external fire,
partial evaporation of the liquid occurs. Only the portion of the vessel that is in
contact with the liquid within a distance of 25 ft (in ISO 23251, 7.6 m) above
the fire source should be considered for sizing, as shown in Table 2.22. It may
be necessary to use alternative sizing methods if fire exposure results in vapor gen-
eration due to thermal cracking.
A vessel that is unwetted is one in which the walls are either thermally insulated
or filled with gases, vapors, or supercritical fluids. Unwetted vessels contain gas-
eous systems. In addition, vessels with separate liquid and vapor under normal
conditions which become single-phase under relieving conditions may also be
included here. It should be noted, however, that vessels whose walls have become
thermally insulated due to the deposition of coke or materials from the fluids
inside, are still considered wetted for fire sizing purposes, but additional fire pro-
tection is required. Due to the large thermal resistance in unwetted vessels, the
thermal flow from the walls to the interior of the vessel is reduced in comparison
to wetted vessels. When the outside surface of a vessel is exposed to a fire source for
an extended period of time, the temperature within the walls of the vessel can rise
to such an extreme level that a thermal rupture can occur.
2.5.5.1 Hydraulic Expansion (Thermal Expansion)

In the case of a liquid vessel exposed to a heat source, the mass flow rate for sizing
the safety valve can be obtained from Eqs. (2.78) (API 521) and (2.79) (ISO 23251),
provided that the trapped liquid does not evaporate. Nevertheless, according to
API 521, the mass flow rates are usually so small that a safety valve with a size
of NPS 3/4 × NPS 1 (DN 20 × DN 25) should be sufficient.
Hydraulic or Thermal Expansion of a Safety Valve Based on API 521

and ISO 23251 Standards
1 αv Ø
q= API 521 2 78
500 d c
1 αv Ø
q= ISO 23251 2 79
500 d c
The cubical expansion coefficient of a liquid should be determined from the

process data; however, some reference values are given in Table 2.21 for liquids
at 60 F (15.6 C). It should be noted that process design data will provide more
precise values.
96 2 Valve Sizing
Table 2.21 Value of cubical expansion coefficient for hydrocarbon

liquids at 60 F in API 521.
Gravity of liquid (°API) αv (1/ ° F) αv (1/ ° C)
3 – 34.9 0.0004 0.00072

35 – 50.9 0.0005 0.0009
51 – 63.9 0.0006 0.00108
64 – 78.9 0.0007 0.00126
79 – 88.9 0.0008 0.00144
89 – 93.9 0.00085 0.00153
94 – 100 and lighter 0.0009 0.00162
Water 0.0001 0.00018
2.5.5.2 Sizing Safety Valve for the Fire Case

There are two types of fire scenarios discussed in this section: an external fire on a
wetted vessel and an external fire on an unwetted vessel.
2.5.5.2.1 External Fire – Wetted VesselIn the fire case, Table 2.22 details the por-
tions of wetted surfaces that should be taken into account when sizing the
safety valve.
Figure 2.21, for instance, shows two vessels on the left and a sphere on the right.
For those two vessels on the left, the portion of a wetted surface that must be taken
Table 2.22 Wetted surfaces that should be taken into account when sizing a safety
valve in a fire situation.
Class of vessels Portions of liquid inventory
Liquid full (e.g. treaters) All up to the height of 25 ft (7.6 m)

Surge or knockout drums, Normal operating level up to a height of 25 ft (7.6 m)
process vessels
Fractionating columns The normal level in the bottom of the column is increased
by the liquid hold-up from all trays dumped into the
normal level in the column’s bottom; the total wetted area
of the column reaches a height of 25 ft (7.6 m).
Storage tanks Normally, the maximum inventory level is 25 ft (7.6 m),
excluding the portions of the wetted area which are in
contact with the foundations and the ground
Spheres and spheroids As high as 25 ft or as large as the largest horizontal
diameter, whichever is greater
Vertical Sphere
vessel
Horizontal
vessel
Max.
dia.
25 ft
Ground
Figure 2.21 Wetted vessels.
into account is up to the level that is 25 ft above the ground. As for the sphere on
the right, the maximum diameter of the sphere is greater than 25 ft. So the propor-
tion of liquid that must be taken into consideration for the wetted area (surface) is
determined by the diameter of the sphere.
The amount of heat absorbed by a noninsulated vessel filled with a liquid is
dependent on the following considerations:
• The source of fuel for the fire;

• The degree to which a vessel is engulfed with flames, which depends on its size
and shape; and
• The immediacy of firefighting measures and the possibility of draining flamma-
ble materials from the vessel.
The total heat absorption Q for the wetted surface can be estimated by Eq. (2.80)
in case of adequate drainage and prompt firefighting measures and by Eq. (2.81) in
case of insufficient drainage and/or firefighting measures. It is possible to imple-
ment an adequate drainage system for fuels that are flammable by utilizing sewers
and trenches as well as the natural slope of the land.
Total Heat Absorption by the Wetted Surface by Adequate Drainage

and Prompt Fire Fighting
Q = 21,000FA0ws82 US units
2 80
Q = 43,200FA0ws82 SI units
98 2 Valve Sizing
Total Heat Absorption by the Wetted Surface During Insufficient

Drainage and/or Prompt Fire Fighting
Q = 34,500FA0ws82 US units
2 81
Q = 70,900FA0ws82 SI units
where:
Q: Total heat absorption (input) to the wetted surface, expressed in British Ther-
mal Units (BTU) per hour in US units or in Joules in SI units;
F: Environmental factor as per Table 2.23;
AWS: Total wetted surface area ft2, m2 calculated by using the equations given in
Table 2.24. In Figure 2.22, different dimensional designations for wetted vessels
that are partially filled with liquid are shown.
The angle β in Table 2.24 is calculated from Eq. (2.82). Feff is the effective liquid
level up to a maximum distance of 25 ft from the flame source as determined by
Eq. (2.83) or (2.84).
Table 2.23 Environmental factor.
Environmental
Type of equipment factor (F)
Bare or uninsulated vessel 1

Insulated vessel with insulation conductance values in BTUs per
hour for fire exposure conditions
4 0.3
2 0.15
1 0.075
0.67 0.05
0.5 0.0376
0.4 0.03
0.33 0.026
Water application facilities, on a bare vessel 1.0
Depressurizing and emptying facilities 1.0
Earth covered storage 0.03
Below grade storage 0.00
Table 2.24 Calculation of wetted area (surface) for different vessels.
Type of vessel Wetted area (surface) calculation
Sphere AWS = πDFeff

Horizontal cylindrical vessel with flat ends D D
AWS = βD L+ − D sin β − F eff
2 2
Horizontal cylindrical vessel with spherical β
AWS = πD L− D + F eff
ends π
Vertical cylinder with flat ends D
Partially filled (F < L) AWS = πD + F eff
4
Totally filled (F = L) D
AWS = πD + F eff
2
Vertical cylinder with spherical ends AWS = πDFeff
D L
Spherical tank D Cylindrical tank

F F
H H
L Vertical
L
D Horizontal
F
Figure 2.22 Different dimensional designations for wetted vessels that are partially filled
with liquid.
100 2 Valve Sizing
Calculation of Angle β
F
β = cos − 1 1 − 2 82
D
Calculation of the Effective Liquid Level
F eff = min 25 ft; F − H US Customary 2 83
F eff = min 7 6 m; F − H SI unit 2 84
Equation (2.85) is used to determine the mass flow rate of the safety valve. Based
on the mass flow rate or relieving capacity, it is possible to size safety valves and
calculate the required orifice area for vapor or gas services based on one of the
methods discussed previously.
Calculation of the Mass Flow Rate for a Safety Valve in the Fire
Case (Wetted Vessel)
Q
W= 2 85
hvl0
Example 2.16 At a set pressure of 200 psig and an atmospheric backpressure,

benzene is contained in an uninsulated vertical vessel with spherical ends at a tem-
perature of 100 F (559.7 R). The vessel has a diameter of 15 ft, a length of 40 ft,
and an elevation of 15 ft. A maximum fluid level of 32 ft from the bottom of the
vessel is achieved. In the event of a fire, assume that fire-fighting measures will
be implemented promptly and that adequate drainage will be provided. In this
case, what is the recommended type of orifice and area for the safety valve consid-
ering a critical flow? (Note: The latent heat of vaporization for benzene is consid-
ered 114.9 Btu/lbm, K = 1.23, T = 875.5 R, and M = 875.5.)
Answer
The first step is to calculate the wetted surface area by using one of the equations given
in Table 2.24. The vessel is vertical with spherical ends. So the equation in the last row
of the table applies. Feff must be determined before the equation can be used.
D = 15 ft Diameter , L = 40 ft Length , H = 15 ft Elevation , F = 32 ft
Liquid level from the bottom
F eff = min 25 ft; F − H = min 25 ft; 32 − 15 = 25 − 15 = 10 ft
AWS = πDF eff = 3 14 × 15 × 10 = 471 ft2
Using Eq. (2.80), it is now possible to calculate the thermal heat flow.
Q = 21,000FA0ws82 = 21,000 × 1 × 4710 82 = 3,266,603 6 Btu h

Q 3,266,603 6
W= = = 28,430 Btu h
hvl0 114 9
K = 1 23 C = 340 refer to Table 2 13
P1 = Set pressure + Overpressure + Atmospheric pressure = 200 +
21 × 200 + 14 7 = 256 7 psia
Note that overpressure is considered 21% due to the fire case.

Since there is no information to estimate the z factor, Z = 1.
Therefore, the safety valve can be sized based on the vapor or gas pressure and
critical flow using Eq. (2.28) as follows:
W TZ 28,430 875 5 × 1
A= =
C K d P1 K b K c M 340 × 0 975 × 256 7 × 1 × 1 78 11
= 0 3340 × 3 35 = 1 118 in 2
For this application, Orifice J with an area of 1.287 in.2 should be selected.
External Fire – Unwetted Vessel A wetted vessel is one whose walls are either ther-
mally insulated or filled with gases, vapors, or supercritical fluids. Unwetted ves-
sels contain gaseous systems rather than liquids. In this case, the discharge area of
the safety valve is determined by using Eq. (2.86).
Discharge Area of the Safety Valve in the Event of an External Fire and
an Unwetted Vessel
FA
A= 2 86
P1
F refers to the bare metal temperature of the vessel at relief, which can be
calculated based on Eq. (2.87). The value of 0.045 is used if F is not known. In
case the calculated value for F is less than 0.01, a value equal to 0.01 must be
taken, which is the recommended and minimum value. Figure 2.23 can be used
as an alternative to estimate parameter F .
Calculation of F
0 1406 T W − T 1 1 25
F = 2 87
C × Kd T 01 6506
102 2 Valve Sizing
700
0.1406(Twall – T1)1.25 Twall, °R
600 F′ = T1, °R
k = 1.001 CKT10.6506
500
400
°F gas
k = 1.4
300
200 Conservative
Minimum
100
0
0.005 0.015 0.025 0.035 0.045 0.055
Operating factor, Fʹ
Figure 2.23 Estimation of parameter F .
For carbon steel plate materials, the maximum vessel wall temperature Tw is
recommended to be 1100 F (593 C). In the case of plates made of alloys, the
recommended maximum wall temperature must be adjusted appropriately.
The relieving temperature T1 is determined by Eq. (2.88) in relation to the nor-
mal operating temperature and pressure, respectively, Tn and Pn, and the relieving
pressure.
Relieving Temperature Calculation

P1
T1 = Tn 2 88
Pn
Carbon steel vessel (Tw = 1560 R) is filled with air at a set pressure of 100 psig.
This vessel has an exposed surface area of 250 ft2. A normal temperature and pres-
sure are 125 F (584.70 C) and 80 psig (94.7 psia). Which size of the orifice should
be used in this situation?
Answer
Due to the fact that the vessel is filled with air rather than liquid, it must be
unwetted.

= 100 + 21 × 100 + 14 7 = 135 7 psia
P1 135 7
Relieving temperature calculation = T 1 = T n = 584 7 × = 837 84 R
Pn 94 7
According to Table 2.14, the specific heat of air is 1.40. The value of gas constant
C associated with the ratio of specific heat k = 1.4 is 356 (refer to Table 2.14).
We can now calculate F using Eq. (2.87) as follows:
0 1406 T W − T 1 1 25 0 1406 1560 − 837 84 1 25

F = = = 0 019
C × Kd T 01 6506 356 × 0 975 837 840 6506
In addition, the minimum discharge can be calculated from Eq. (2.86) as follows:
FA 0 019 × 250
A= = = 0 40 in 2
P1 135 7
For this application, Orifice G with an area of 0.503 in.2 should be selected.

2.1 Select the correct answer about valve sizing.
A Before a PSV, an 8 × 6 ball valve is the correct size for a subflare line.
B The water service is 60 F with a flow rate of 1000 gpm that passes
through a 3 reduced bore ball valve with a one psi pressure drop. The
valve is correctly sized in this case. (Note: the flow coefficient value
should be taken from Table 2.1.)
C The flow rate for fully opening an axial check valve is 21,048 kg/h. The
minimum flow rate in the valve is 22,234 kg/h showing that the valve is
sized correctly.
D All answers are wrong.
Answer
Option A is incorrect because the valve engineer shall select a full-bore ball
valve for the subflare line allowing sudden fluid or gas service release. Option
B is not correct either because a 3 reduced bore ball valve provides a flow
coefficient equal to a 2 ball valve that is 480 gpm. In contrast, the calculated
flow coefficient based on process data (Temperature = 60 F, Flow
rate = 1000 gpm, Pressure drop = 1 psi, SG = 1 for water) should be
1000 gpm. It means that the valve is undersized, and it should be a 3 full
bore with a flow coefficient value equal to 1300 gpm as per Table 2.1. Option
C shows that the check valve will not be at risk of chattering, and it will be
fully open during the operation. However, this information does not indicate
that the valve is sized correctly. Option D is the correct answer since all other
options are incorrect.
104 2 Valve Sizing
2.2 Which answer is correct about sizing check valves?

A The critical velocity in a swing check valve depends on the spring tor-
que value.
B The critical velocity to fully open a check valve is higher for water
than gas.
C The critical velocity is increased for spring-loaded valves by selecting a
higher torque spring.
D There is no difference between sizing a check valve installed on a hor-
izontal or vertical line.
Answer
Option A is incorrect because a swing check valve does not have a spring.
Option B is not correct either since a lower velocity of water than gas is
needed to fully open a check valve due to the higher density of the water.
Option C is the correct answer as per data provided in Tables 2.4 and 2.5.
Option D is wrong because, for a check valve installed on a vertical line with
an upward flow, the weight of disk(s) for sizing of the valve shall be consid-
ered by the valve engineer. In contrast, disk(s) weight can be neglected for
sizing a check valve installed on a horizontal line.
2.3 Which statements are incorrect regarding the valve sizing in liquid services?
A The possibility of choked flow is evaluated during the valve size
selection.
B Installation of fittings attached to the valve does not impact the
valve size.
C Type of control valve design in terms of plug and cage type as well as flow
characteristics are not taken into account during the valve sizing.
D All three options are incorrect.
Answer
Option B and D are wrong statements.
2.4 A high-performance butterfly valve in pressure class of 300 (PN of 50 bar) is

installed on an 8 line. The fluid service is water with a flow rate of 1600 gpm.
The valve’s inlet pressure, outlet pressure, and inlet temperature values are
300 psig, 275 psig, and 70 F, respectively. It is assumed that the piping geom-
etry factor (Fp) is equal to 0.90, and no choked flow can happen in the valve;
which size for the butterfly valve is correct?
A 6
B 4
C 8
D 10
Answer
Required flow coefficient (Cv) by considering the effect of piping geometry
factor and no choke flow occurrence is calculated by using Eq. (2.17).
q 1600
Cv = = = 355 55
N 1FP P1 − P2 SGl 1 × 0 90 25 1
Referring to Table 2.10, a 3 high performance butterfly valve at a fully open

position can provide a flow coefficient value of 237, which is less than the
required Cv = 355.55. Hence, a 3 butterfly valve is undersized. However,
a 4 high performance butterfly valve can give a flow coefficient value of
499, which is sufficient. Selecting a valve with a larger size than 4 is consid-
ered an oversize valve. Thus, the 4 valve size and option B are correct.
2.5 The required hydrocarbon gas mass flow rate resulting from the operation of
the safety valve is 53,500 lb/h. Fluid service consists of a mixture of butane
and pentane with a molecular weight (M) of 65 and a relief temperature of
627 R, or 348 K. The relief valve set pressure is 75 psig, which is the design
pressure of the component that may be increased by 10% to accommodate
the spring force during the opening. There is a constant backpressure of
55 psig, and there is no pressure loss in the valve. The compressibility factor
of the gas service is 00.84, while the specific heat is 1.09. There is a total
backpressure of 62.5 psig at the valve, whereas the critical flow pressure is
42.6 psi. Which of the following statements is incorrect?
A The safety valve should be sized based on the subcritical flow condition.
B The coefficient of subcritical flow is 0.86 in this case.
C The effective orifice area in this case is 5.6 in.2.
D The valve should be fitted with a “P” sized orifice.
Answer
The total backpressure of 62.5 psig is higher than the critical flow pressure of
42.6. Therefore, the safety valve should be sized according to the subcritical
flow condition, and option A is correct. The subcritical flow coefficient can
be calculated using Eq. (2.51) by using the parameters r and k in the follow-
ing manner.
2.6 A PSV is designed for handling natural gas with a molecular weight of 19 and
a flow capacity of 2675 kg/h. The set pressure for the valve is 1450 kPag, and
the allowable overpressure is 10%. In the presence of atmospheric backpres-
sure and an inlet temperature of 50 C, and assuming a critical flow, what is
the area of the orifice and its type based on API standard?
A Orifice designation F with API effective area of 0.307 in.2
B Orifice designation G with API effective area of 0.503 in.2
106 2 Valve Sizing
C Orifice designation H with API effective area of 0.785 in.2

D Orifice designation J with API effective area of 1.287 in.2
Answer
Equation (2.29), used for the sizing of safety valves for critical flow gases, is
applicable here. Data that are used for safety valve sizing are summarized as
follows:
W = 2675 kg h
T = 50 + 273 = 323 K
T = 50 + 273 = 323 K
P1 = Absolute relieving pressure = Set pressure + Overpressure
+ Atmospheric pressure
= 1450 + 10 × 1450 + 101 = 1696 KPa
Table 2.14 indicates that the constant C for natural gas is 344. The coeffi-
cient of discharge Kd = 0.975. Since the backpressure is equal to the atmos-
pheric pressure, it cannot affect the flow capacity of the valve. So the
capacity correction factor due to backpressure can be assumed as equal
to 1, Kb = 1. Kc is the combination correction factor for installation with
a rupture disk upstream of the PRV. This value is one in this example
because there is no rupture disk installed. It is assumed that the Z-factor,
or gas compressibility factor, is equal to one in this case since its real value
is unknown.
13,160 W TZ
A= SI units
C K d P1 K b K c M
13,160 × 2675 323 × 1
A= = 61 885765 × 4 1231
344 × 0 975 × 1696 × 1 × 1 19
= 255 16 mm2 = 0 395498 in 2
2 29b
As a result, orifice G with an area of 0.503 in should be selected and option
2
B is the correct answer.
2.7 In the previous example, the backpressure is 725 kPag. If all other para-
meters are the same, which API orifice should be selected?
A Orifice designation F with API effective area of 0.307 in.2
B Orifice designation G with API effective area of 0.503 in.2
C Orifice designation H with API effective area of 0.785 in.2
D Orifice designation J with API effective area of 1.287 in.2
Answer
The main difference between this case and the previous case is that since the
backpressure is not discharged into the atmosphere, the value of the back-
pressure correction factor is not equal to one. The ratio of backpressure to
set pressure is equal to 725/1450 = 0.50 and refer to Figure 2.12, the value
of Kb = 0.69. Because all other parameters are equal in these two examples,
the orifice area is calculated as follows:
0 395498
A= = 0 57 in 2
0 69
As a result, orifice H with an area of 0.785 in.2 should be selected and option
C is the correct answer.
2.8 In this instance, a bellows balanced type PSV is handling sodium trisulfide, a
nonviscous liquid with a specific gravity of 1.23. As a result, the valve is capa-
ble of alleviating 475 l of pressure per minute and the set pressure of the valve
is 690 kPa. It should be noted that there is a 10% overpressure allowed and
the backpressure is 207 kPag. If this is the case, then what type of orifice and
effective area should be selected considering that the safety valve will not be
certified?
A Orifice designation G with API effective area of 0.503 in.2
B Orifice designation H with API effective area of 0.785 in.2
C Orifice designation J with API effective area of 1.287 in.2
D Orifice designation K with API effective area of 1.838 in.2
Answer
Since the given values are in SI unit and the safety valve is used in liquid
service, Eq. (2.63) applies. The first step is to collect the given parameters
given above as follows:
Q = 475 l/min (Flow rate)
Kd = 0.62 (Correction factor due to discharge)
Pb = 207 kPag (Backpressure)
P = Ps = 690 kPag (Set pressure)
Pb 207
= =03
Ps 690
Kw = 0.87 (The backpressure correction factor relates to the ratio of the back-
pressure to the set pressure, which is equal to 0.3 and it is derived from
Figure 2.16).
KC = 1 (The combination correction factor is equal to one in this case because
the rupture disk does not need to be installed.).
Kv = 1 (The liquid is nonviscose)
108 2 Valve Sizing
Kp = 0.6 (Correction factor due to overpressure related to 10% overpressure and

taken from Figure 2.19)
G = 1.23 (Specific gravity of the flowing fluid)
11 78 × Q G 11 78 × 475 1 23
A= =
K dK w K cK vK p 1 25 P − Pb 0 62 × 0 87 × 1 × 1 × 0 6 1 25 × 690 − 207
= 17,289 27 × 0 043 = 748 93 mm2 = 1 160 in 2
2 63b
In conclusion, the orifice letter “J” with an effective area of 1.287 in.2 corre-
sponding to Table 2.12 must be selected, and option C is the correct response.
2.9 A safety valve controls the flow of subcooled propane at a volumetric rate of
100 gpm. The set pressure of the relief valve is 260 psig with an overpressure
allowance of 10%. This valve has a total backpressure of 10 psig, including
only build-up backpressure. It has a density of 31.92 lb/ft3 and a specific heat
of 0.6365 btu/lb R at constant pressure at the PRV inlet. In addition, the sat-
uration pressure of propane at 60 F is 107.6 psia, and the specific volume of
propane liquid at the saturation pressure is 0.03160 ft3/lb. The specific vol-
ume of propane vapor and latent heat of vaporization at saturation pressure
are 1.001 ft3/lb and 152.3 Btu/lb, respectively. The temperature at safety
valve inlet is 519.67 R. Which statement is incorrect?
A Omega is 8.515 for subcooled liquid flows at the safety valve inlet.
B In this example, the propane liquid corresponds to the high subcooling
region.
C The mass flux value is 5670 lb/sft2
D “F ” is the correct size of the orifice.
Answer
Due to the fact that propane is flowing at a subcooled temperature, the cal-
culation methods and formulas provided in Section 2.5.4.2 must be applied.
A summary of the provided parameters is the first step.
Q = 100 gal min;
P1 = P0 = Set pressure + Overpressure + Atmospheric pressure
= 260 + 26 + 14 7 = 300 7 psiaPb = Pa = 10 psig
Pa 10
Percent of gauge backpressure = = = 0 038 = 3 8
Set pressure 260
As the downstream backpressure is less than 10% of the set pressure, a con-
ventional relief valve may be used and the back-pressure correction factor is
one. Kb = 1, Kd = 0.65, Kc = 1.
C p = 0 6365 btu lb R
ρ10 = 31 92 lb ft3
Ps = 107 6 psia
vvls: Difference between the vapor and the liquid specific volumes at
Ps (ft3/lb) = Specific volume of propane liquid at the saturation
pressure – Specific volume of propane vapor at the saturation
pressure = 1.001 − 0.03160 = 0.9694 ft3/lb
hvls = 152 3 Btu lb
T 0 = 519 67 R
The next step is to calculate the saturated Omega parameter ωs by using
2
vvls
ωs = 0 185ρl0 Cp T 0 Ps = 0 185 × 31 92 × 0 6365 × 519 67
hvls
2
0 9694
× 107 6 × = 8 515
152 3
Thus, option A is correct. The next step is to determine the subcooling region.
2ωs 2 × 8 515
P0 × = 300 7 × = 284 > Ps = 107 6
1 + 2ωs 1 + 2 × 8 515
Therefore, the liquid falls into the subcooling region, and option B is also
correct. The next step is determining whether the fluid is critical or subcrit-
ical. Since Ps = 107.6 psia > Pa = 10 psig the flow of fluid is critical. The next
task is to calculate mass flux according to Eq. (2.76). (For critical flow con-
dition, P = Ps)
05 05
G = 96 3 ρl0 × P1 − P = 96 3 31 92 × 300 7 − 107 06 = 7571 lb sft2
Therefore, option C is incorrect. It is now possible to calculate the orifice area

using Eq. (2.77).
1 Q ρl0 1
A = 0 3208 × × = 0 3208 ×
K bK cK d G 1 × 1 × 0 65
100 × 31 92
× = 0 208 in 2
7571
Accordingly, the correct orifice is F with an effective discharge area of
0.307 in.2. Therefore, option D is appropriate.
110 2 Valve Sizing
Further Reading
Ahmed, T. (2007). Equations of State and PVT Analysis. Gulf Publishing Company.
American Petroleum Institute (API) 520 (2020). Sizing, Selection, and Installation of
Pressure-Relieving Devices Part 1 – Sizing and Selection. Washington, DC: American
Petroleum Institute (API).
American Petroleum Institute (API) 521 (2007). Pressure-Relieving and Depressuring
Systems, 5e. Washington, DC: American Petroleum Institute (API).
American Petroleum Institute (API) 526 (2017). Flanged Steel Pressure-Relief Valves, 7e.
Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) B16.34 (2017). Process Piping.
New York: American Society of Mechanical Engineers (ASME).
American Society of Mechanical Engineers (ASME) B36.10M (2015). Welded and
Seamless Wrought Steel Pipe. New York: American Society of Mechanical
Engineers (ASME).
American Society of Mechanical Engineers (ASME) B36.19M (2018). Stainless Steel
Pipe. New York: American Society of Mechanical Engineers (ASME).
Ballun, J.V. (2007). A Methodology for Predicting Check Valve Slam. American Water
Works Association (AWWA).
Crosby Valve Inc (1997). Crosby Pressure Relief Valve Engineering Handbook. Crosby
Valve Inc. Technical document number. TP-V300.
Emerson (2019). Control Valve Sizing, 5e. Emerson. [online]. https://www.emerson.
com/documents/automation/control-valve-handbook-en-3661206.pdf (accessed
11 February 2022).
Emerson (2022). Valve Sizing Calculations. Emerson. [online]. https://www.emerson.
com/documents/automation/manual-valve-sizing-standardized-method-fisher-en-
140724.pdf (accessed 11 February 2022).
Ford, R. (2014). Power industry applications: a valve selection overview. Valve World
Magazine 19 (8): 96–103.
Goodwin (2016). Check Valves Technical Catalogue. Goodwin. [online]. https://
dokumen.tips/documents/goodwin-check-valve-technical-cataloguepdf.html
(accessed 8 February 2022).
Guo, B. and Ghalambor, A. (2005). Natural Gas Engineering Handbook, 2e. Gulf
Publishing Company.
International Organization of Standardization (ISO) 23251 (2019). Petroleum,
Petrochemical and Natural Gas Industries – Pressure-Reliving and Depressurizing
Systems, 2e. Geneva: International Organization of Standardization (ISO).
International Society of Automation (ISA) 75.1 (2007). Control Valve Sizing Equations.
Research Triangle Park, NC: International Society of Automation (ISA).
Ludtke, P.R. (1986). Natural Gas Handbook. National Bureau of Standards, US
Department of Commerce.
Further Reading 111
Nesbitt, B. (2007). Handbook of Valves and Actuators: Valves Manual International, 1e.
Oxford: Elsevier.
Oxler, G. (2009). Non-return valve and/or check valve for pump system – a new
approach. Valve World Magazine 14 (4): 75–77.
Rahmeyer, W.J. (1993). Sizing swing check valves for stability and minimum velocity
limits. Transition to the ASME 115: 406–410.
Sotoodeh, K. (2015). Axial flow nozzle check valves for pumps and compressors
protection. Valve World Magazine 20 (1): 84–87.
Sotoodeh, K. (2018). Comparing dual plate and swing check valves and the importance
of minimum flow for dual plate check valves. American Journal of Industrial
Engineering 5 (1): 31–45. https://doi.org/10.12691/ajie-5-1-5.
Sotoodeh, K. (2018). Why are butterfly valves a good alternative to ball valves for utility
services in the offshore industry? American Journal of Industrial Engineering 5 (1):
36–40. https://doi.org/10.12691/ajie-5-1-6.
Sotoodeh, K. (2020). Challenges associated with the bypass valve of control valve in a
sea water service. Journal of Marine Science and Application https://doi.org/10.1007/
s11804-020-00132-8.
Sotoodeh, K. (2021). Analysis and failure prevention of nozzle check valves used for
protection of rotating equipment due to wear and tear in the oil and gas industry.
Journal of Failure Analysis and Prevention https://doi.org/10.1007/s11668-021-
01162-2.
Sotoodeh, K. (2021). A Practical Guide to Piping and Valves for the Oil and Gas Industry,
Sotoodeh, K. (2022). Cryogenic Valves for Liquified Natural Gas Plants, 1e. Austin, TX:
Elsevier (Gulf Professional Publishing).
VALMETALIC (2018). Design and Selection of Check Valves. VALMETALIC. https://
www.valmatic.com/Portals/0/pdfs/DesignSelectionCheckValves.pdf (accessed 7th
February 2022).
112
Cavitation and Flashing
3.1 Introduction
Flow control, also known as regulation or throttling of fluids, is one application for
some valves. Flow control valves adjust the amount of flow moving through the
valve as well as the flow rate inside the valve; this impacts other parameters of
the fluid process such as temperature, pressure, and level. Various sectors of
the oil and gas industry have used globe valves extensively. Generally speaking,
globe valves are covered by the American Petroleum Standard (API 602), which
is titled “Gate, globe, and check valves for sizes DN 100 (4 ) for the petroleum
and natural gas industries.” Valves used for fluid control are susceptible to cavi-
tation, which is a significant problem that is discussed in the following section.
3.2 Cavitation
3.2.1 What is Cavitation?

For oil and gas facilities and components such as pumps, valves, and piping,
erosion or corrosion can be costly. Cavitation can be described as a type of erosion
or erosion-corrosion and is the most severe operational problem for controlling
globe valves. In the oil and gas industry, T-pattern (or tee-pattern) globe valves
(shown in Figure 3.1) are very common. A globe valve with an actuator is referred
to as a control valve. The liquid flows into the valve shown in the figure from the
left side and reaches the middle portion of the valve where the plug (disk) and seats
are located. As the disk is off the seat, the fluid makes two 90 turns and leaves the
valve from the outlet port highlighted by pressure. These two 90 rotations of the
fluid result in a substantial pressure drop inside the valve. There is a narrow area
3.2 Cavitation 113
Actuator force
Seals
Bonnet
Body
Valve plug
Fluid flow – pressure P1 Pressure P2
Valve seat
Differential pressure (ΔP)
Figure 3.1 A T-pattern globe valve.
below the plug where the fluid velocity is very high and the pressure is extremely
low. It is possible for the pressure to drop below the liquid’s vapor pressure in this
narrow area, causing some bubbles to form. The bubbles of gas that are separated
from the liquid by a high-pressure drop can recover their pressure and collapse
firmly, causing pressure waves. Known as cavitation, this can damage parts of
the valve, such as its seats and plugs. Cavitation is a form of corrosion or erosion
that causes metal loss or the creation of pits on the valve trims (internals). Cavi-
tation also causes noise and vibration in addition to accelerated corrosion. A vapor
pressure is a pressure at which a fluid is at thermodynamic equilibrium with its
condensed state. Temperature affects vapor pressure. So, when the pressure of a
fluid drops below its vapor pressure, it changes from liquid to gas. In
Figure 3.2, irregular pitting and erosion-corrosion can be seen on a plug of a con-
trol valve due to cavitation.
3.2.2 Cavitation Essential Parameters

The following are a few of the essential parameters that affect the severity of cav-
itation and its consequences:
• Pressure drop: Cavitation is intensified by higher pressure drops in valves with

higher pressure classes. As an example, a globe valve in pressure class 150 has a
114 3 Cavitation and Flashing
Figure 3.2 Corrosion caused by cavitation on a control valve plug.
maximum pressure drop of 20 bar, while a valve in pressure class 300 has a max-
imum pressure drop of 50 bar.
• Leakage: A leak is the movement of fluid from an area of high pressure to an
area of low pressure. Cavitation is more likely to occur if the valve leaks in a
closed position.
• Material: Harder materials are less susceptible to cavitation. The cavitation
resistance of hard trim materials such as Stellite 6 (UNS R30006) or Stellite
21 in solid or overlay form, as well as 13Chromium martensitic stainless steels
like UNS S41000 and UNS 415000, is higher.
• Valve size and design: In smaller valves, cavitation may be more severe. As a
result of causing less pressure drop, Y-pattern globe and axial valves have less
cavitation risk than T-pattern globe valves. API released the first edition of the
new globe valve design standard (API 623) to control and avoid operation pro-
blems in globe valves such as cavitation, vibration, and leakage. The API 623
specifies applying a hardfacing alloy such as stellite to the valve internals such
as the plug and seat, a larger stem diameter, and ensuring a more durable con-
nection between the plug and the seat.
• Trim design: Valve trim refers to the internal components of the valve that are
in contact with the fluid service. The multistep or anticavitation trim design (see
Figure 3.3) minimizes cavitation risk. By creating a pressure drop in two or more
stages, multistep trimming prevents high-pressure drops inside the valve and
cavitation.
• Flow regime: Turbulent and high-velocity flows increase erosion and cavita-
tion risks.
3.2 Cavitation 115
Multi-step pressure
drop to avoid
excessive cavitation
and noise
Figure 3.3 Trim design with multiple stages.
3.2.3 Cavitation Analysis

The cavitation index or parameter sigma (δ) is the most widely accepted method for
estimating and measuring cavitation in control valves. The cavitation index is
calculated by using Eq. (3.1) as follows:
Cavitation Index (δ) Calculation
P u − Pv
δ= 31
Pu − Pd
where:
δ: Cavity index (dimensionless);

Pd: Downstream (outlet) pressure (psig);
Pv: Adjustment of the vapour pressure of the fluid flowing inside the valve to
account for temperature and atmospheric pressure (psig);
Pu: Upstream (inlet) pressure (psig).
The severity and extension of cavitation for valves including globe valves based
on cavity index values are given in Table 3.1.
Example 3.1 If the inlet and outlet pressure values are 25 bar and 20 bar, respec-
tively, calculate the cavity index for a 6 globe valve CL300 in seawater service
at 25 C. Water has a vapor pressure of 0.36 psig. In this case, is cavitation a
possibility?
Table 3.1 Cavity severity based on cavity index values.
Criteria Cavitation consequence
δ ≥ 2.0 No risk of cavitation

1.7 < δ < 2.0 There is no need for cavitation control
The trim is hardened for protection
1.5 < δ < 1.7 There is some cavitation control required
1.0 < δ < 1.5 Potential for severe cavitation
δ ≤ 1.0 The flashing is occurring
Answer
Inlet and outlet pressure are calculated based on psig as follows:
Pu = 25 × 14 5 = 362 5 psi
Pd = 20 × 14 5 = 290 psi
Pu − Pv 362 5 − 0 36 362 14
δ= = = = 4 995
Pu − Pd 362 5 − 290 72 5
Cavitation is not a concern due to the cavitation index exceeding 2, as shown in
Table 3.1.
3.3 Flashing
Similarly, flashing is another term that is highly associated with cavitation, and it
occurs when the fluid pressure falls below the vapor pressure. At this point, fluid
begins to transform from a liquid to a vapor. There is a flashing in a valve (e.g.
globe or control) if the valve outlet pressure is less than the vapor pressure. In this
case, the pressure inside the valve will not recover, and the vapor will travel down-
stream (after) the valve (see Figure 3.4). Vapor pressure will eventually be restored
to the pipe, and the collapsing vapor will result in mechanical damage, such as
erosion and noise, similar to that which occurs during cavitation. The primary dif-
ference between cavitation and flashing is in the pressure value at the valve outlet;
the cavity flow has a higher pressure than the vapor pressure at the outlet, whereas
the flashing flow has a lower pressure than the vapor pressure (see Figure 3.5).
There are two main conditions that must be met for cavitation and flashing to occur:
• High-pressure drop across the valve;

• The liquid pressure drops below the vapor pressure.
3.3 Flashing 117
P1
Pressure
Vapor P2
pressure
Inlet Valve Outlet
Figure 3.4 A flash occurs when the liquid pressure drops below the vapor pressure.
Inlet pressure
Normal flow
Pressure
Cavitating flow
Outlet pressure
Vapor pressure
Flashing flow
Distance through valve
Figure 3.5 A control valve’s normal, cavitation, and flashing pressure profiles.
It is pertinent to note that cavitation and flashing are the results of choked flow.
A choked flow is a fluid dynamic phenomenon caused by the venturi effect.
Figure 3.6 illustrates how fluid flow is compressed and choked when it passes
through a restricted area. Pressure drops as velocity increases when the fluid
passes through a restricted area.
P1 v1
P2 v2
A2 < A1
v2 > v1
P2 < P1
A2
Increase in fluid speed results in

A1 decrease in internal pressure
Figure 3.6 Choked flow.
Many methods can be used to prevent flashing. For example, one can install a
large control valve and outlet piping. A control valve’s body should not be smaller
than half the diameter of the pipe it is connected to. In addition, there are other
ways to minimize the risk of flashing, such as using hard materials for the trim
(valve internals) and anticavitation trim, as well as incorporating a sacrificial spool
into the piping system. Flashing is usually less damaging than cavitation. It is
important to note, however, that a small piping outlet configuration can cause
cavitation and cause severe piping damage at the valve’s outlet. The flow of fluid
is rotated 90 when using angle globe valves. Because 90 elbows are no longer
required with angle globe valves, less piping space is required. The valve manufac-
turer should be consulted regarding the possibility of cavitation and flashing
within valves and their outlet piping.
3.1 Which of the following statements is correct regarding cavitation?

A Cavitation only occurs in globe valves.
B Cavitation causes vibration, noise, erosion, and corrosion.
C Noise and vibration are the results of pressure drop below the vapor
pressure.
D The valve’s outlet pressure is less than the vapor pressure.
Answer
In fact, option A is incorrect since cavitation can occur in almost any type of
valve used for flow regulation, including globe, v-notch ball, and butterfly
valves. It is option B that is correct. Option C is incorrect since noise and
vibration are caused by vapor pressure recovery and the collapse of the vapor.
Option D is also incorrect since cavitation can only occur when the outlet
pressure of a valve exceeds the vapor pressure.
3.2 What condition increases the chance and severity of cavitation?

A Applying stellite 6 or 21 to the internals of the valve
B Using an axial valve instead of a T-pattern globe valve
C Selecting a larger size and higher-pressure class valve
D Designing multistep trim
Answer
Option C is the correct answer. All other options reduce the chance and
severity of cavitation.
3.3 Which cavitation remedy is illustrated on the right side of Figure 3.7?
A Selecting a smaller size and lower pressure class valve
B Designing a double-stage trim
C Applying stellite to the trim of the valve
D All these answers are incorrect.
Answer
Option B is correct since the pressure drop or reduction on the right side of
the figure occurs in two stages, and the minimum pressure is still above the
vapor pressure. So, cavitation will not occur.
P1 P1
∆P1 ∆PSize = P1–P2
∆PSize = P1–P2 P12
∆P2
Cavitation
P2 P2
begins
–PVC1
–PVC2
PVP PVP
∆PSize = P1–P2 = ∆P1+∆P2
–PVC
Figure 3.7 Cavitation remedy.

3.4 An 8 butterfly valve in pressure class 150 (pressure nominal of 20 bar) is

used to control flow in water service. During operation, the butterfly valve
is open 50–100%. As the differential pressure across the valve varies from
a minimum of 2 bar to a maximum of 4 bar, the ratio of the differential pres-
sure to the inlet operating pressure changes from 0.4 to 0.8. The fluid tem-
perature is 80 F, and the water vapor pressure at this temperature is 0.5069.
What is the correct option?
A The chance of cavitation is higher when the valve is half-open.
B In general, there is no risk of cavitation in this valve.
C The cavity index of the valve is higher when the valve is in a half-open
position.
D All answers are wrong.
Answer
Considering that the cavitation index is greater than two at the fully open
valve position, there is no risk of cavitation, as shown in Table 3.1. At
50% open, there is a risk of severe cavitation. Therefore, option A is the
correct answer, and option B is wrong. When the valve is half-opened, the
cavity index is lower. Therefore, option C is incorrect. Option D is incorrect
because option A is correct.
ΔP 2
100 valve open ΔP = 2 bar =04 = 0 4 P1 = 5 bar P2
P1 P1
Pu − P v 5 − 0 5069
= 3 bar δ= = = 2 25
P u − Pd 5−3
ΔP 4
50 valve open ΔP = 4 bar =08 =08 P1 = 5 bar P2
P1 P1
Pu − Pv 5 − 0 5069
= 1 bar δ= = = 1 123
Pu − Pd 5−1
As the cavitation index is greater than two at the fully open valve position,
therefore, there is no risk of cavitation, as shown in Table 3.1. However,
severe cavitation may result at 50% open position. Therefore, option A is
the correct answer, and option B is false.
3.5 Find the correct statement regarding cavitation.

A Cavitation happens in gas services.
B The first stage of cavitation is the collapse of the vapor bubbles.
C Cavitation energy can cause physical damage to the valve’s internal parts.
D The larger the cavitation index, the higher the cavitation risk.
Answer
Option A is incorrect because cavitation happens in liquid services. Option B
is not correct either since the second stage of cavitation is the collapse of
the vapor bubbles. Option C is the right answer. Option D is wrong, as a larger
cavitation index implies a lower cavitation risk.
3.6 For flow throttling in the water piping system, an 8 plug valve is used. The
upstream pressure of the valve equals 11 psi, and the pressure drop is 4 psi
during throttling. The flow coefficient of the valve at fully open is 1800.
The valve has a flow of 600 gpm during throttling. Figure 3.8 illustrates
the flow curve of the plug valve in this case. Regarding the plug valve’s cav-
itation risk and opening percentage, which statement is true? (Note: In this
case, the vapor pressure of water is −14.4 psi.)
A Cavitation risk is severe, and the valve is less than 20% open.
B Flashing may occur, and the valve is only at a 10% opening position.
C There is no risk of cavitation, and the valve is approximately 40% open.
D All answers are incorrect.
100
90
Plug
80
Butterfly
70
Percent of full open Cv
60
50
Ball
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90
Valve position (degrees from closed position)
Figure 3.8 Valves flow characteristics including the plug valve.

Answer
Pu − Pv 11 − − 14 4
δ= = = 6 35
Pu − Pd 11 − 7
Cavitation cannot occur since the cavity index is greater than two. It is now
possible to calculate the valve flow coefficient during throttling as follows:
SG 1
Cv = Q = 600 = 300
ΔP 4
300
Flow coefficient percentage during flow throttling = = 16 67
1800
According to Figure 3.8, the plug valve opening percentage corresponds to
16.67% of the maximum flow coefficient, or approximately 41%. Answer
C is therefore the correct answer.
3.7 Figure 3.9 shows the cavitation characteristics of three types of valves – plug,
butterfly, and ball. Select the correct statement regarding the valve’s cavita-
tion characteristics chart.
Valve constant cavitation data

16
14
12
Butterfly valve
Cavitation coefficient
10 Safe operating
zone
8
Plug valve
6
Ball valve
4
2 Cavitation
zone
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Valve opening, degrees open
Figure 3.9 Valves cavitation characteristics.

Further Reading 123
A Increases in valve opening percentage increase the risk of cavitation.

B Compared to ball or butterfly valves, the plug valve is less prone to
cavitation.
C Due to the risk of cavitation, it is not recommended to keep the flow
throttling valves open less than 10%.
D A ball valve is always safer than a butterfly valve against cavitation.
Answer
Option A is incorrect because the cavitation risk is reduced by increasing the
valve opening percentage. Option B is not correct either since the plug valve
has the highest risk of cavitation as its curve is closer to the cavitation zone
highlighted at the bottom of the chart in red. Option C is the correct answer.
Option D is wrong as the ball valve is safer than the butterfly valve against
cavitation for an opening percentage of less than 70%.
3.8 Which of the following statements is incorrect regarding flashing and its
conditions?
A When a flow encounters a restriction, such as a reduced bore valve, its
velocity increases, and flashing could occur.
B Flashing occurs when the outlet pressure is greater than the vapor pres-
sure (P2 > Pv).
C Mechanical damages such as erosion and noise are adverse consequences
of flashing.
D Flashing usually causes less damage to the valve than cavitation.
Answer
Option B is incorrect since flashing occurs when the outlet pressure is less
than the vapor pressure (P2 < Pv).
Further Reading
American Petroleum Institute (API) 602 (2016). Gate, Globe and Check Valves for Sizes
DN100 (NPS 4) and Smaller for the Petroleum and Natural Gas Industries, 10e.
Iranian Petroleum Standard (IPS) (2015). Engineering Standard for Control Valves. IPS-
E-IN-160. Tehran: Iranian Petroleum Standard (IPS).
Oxford: Elsevier.
Sotoodeh, K. (2016). Cavitation in globe valves and solutions. Valve World Magazine
21 (3): 32–36.
Sotoodeh, K. (2018). Selecting a butterfly valve instead of globe valve for fluid control in
a utility service in the offshore industry (based on industrial experience). American
Journal of Mechanical Engineering 6 (1): 27–31. https://doi.org/10.12691/ajme-6-1-4.
Sotoodeh, K. (2021). Subsea Valves and Actuators for the Oil and Gas Industry, 1e.
Austin, TX: Elsevier (Gulf Professional Publishing).
Valve-Metallic Valve and Manufacturing (2018). Cavitation in Valves. Valve-Metallic
Valve and Manufacturing. https://www.valmatic.com/Portals/0/pdfs/
CavitationinValves_6-18.pdf (accessed 25 February 2022).
125
Wall Thickness
4.1 Introduction
The bodies and bonnets of industrial valves are pressure-containing parts that lead
to leakage of valve internal fluid into the environment if they fail to function. It is
essential to calculate and select the correct thickness for the body and bonnet to
prevent mechanical failures. The American Society of Mechanical Engineers
(ASME) B16.34, the standard for flanged, threaded, and welding end, covers
various aspects of valves, including pressure-temperature rating, dimensions,
tolerances, materials, nondestructive examination requirements, testing, and
marking. The valve body and bonnet thickness should meet the criteria set in
ASME B16.34.
4.2 ASME B16.34 Minimum Wall Thickness Calculation
4.2.1 Conservation Approach (Mandatory Appendix A)

The minimum valve body thickness according to ASME B16.34, parameter tm, is
provided in both millimeters and inches. The minimum valve body thickness in
ASME B16.34 depends on two valve parameters: internal diameter (parameter d)
and the pressure class of the valve. Generally, the minimum valve wall thickness
increases with the internal diameter and pressure class of the valve. The pressure
classes, which are covered by ASME B16.34, are 150 (PN20), 300 (PN50), 600
(PN100), 900 (PN150), 1500 (PN250), 2500 (PN420), and 4500 (PN720). “PN” stands
for pressure nominal. The internal diameter of a valve is the minimum diameter of
fluid passage through the valve bore. Using Table 4.1 extracted from ASME B16.34,
the inside diameter of the valves can be determined in millimeters and inches. In
general, ASME B16.34 specifies that the inner diameter of the valve shall not be less
than 90% of the diameter of the valve end.
126 4 Wall Thickness
Table 4.1 Valve inside diameter in millimeters and inches, according to valve size
(NPS) and pressure class, as per ASME B16.34.
CL150 CL300 CL600 CL900
NPS mm in. mm in. mm in. mm in. DN
1/2 12.7 0.5 12.7 0.5 12.7 0.5 12.7 0.50 15

3/4 19.1 0.75 19.1 0.75 19.1 0.75 17.5 0.69 20
1 25.4 1.00 25.4 1.00 25.4 1.00 22.1 0.87 25
1 1/4 31.8 1.25 31.8 1.25 31.8 1.25 28.4 1.12 32
1 1/2 38.1 1.50 38.1 1.50 38.1 1.50 34.8 1.37 40
2 50.8 2.00 50.8 2.00 50.8 2.00 47.5 1.87 50
2 1/2 63.5 2.50 63.5 2.50 63.5 2.50 57.2 2.25 65
3 76.2 3.00 76.2 3.00 76.2 3.00 72.9 2.87 80
4 101.6 4.00 101.6 4.00 101.6 4.00 98.3 3.87 100
6 152.4 6.00 152.4 6.00 152.4 6.00 146.1 5.75 150
8 203.2 8.00 203.2 8.00 199.9 7.87 190.5 7.50 200
10 254.0 10.00 254.0 10.00 247.7 9.75 238.0 9.37 250
12 304.8 12.00 304.8 12.00 298.5 11.75 282.4 11.12 300
14 336.6 13.25 336.6 13.25 326.9 12.87 311.2 12.25 350
16 387.4 15.25 387.4 15.25 374.7 14.75 355.6 14 400
18 438.2 17.25 431.8 17.00 419.1 16.50 400.1 15.75 450
20 489.0 19.25 482.6 19.00 463.6 18.25 444.5 17.50 500
22 539.8 21.25 533.4 21.00 511.0 20.12 489.0 19.25 550
24 590.6 23.25 584.2 23.00 558.8 22.00 533.4 21.00 600
26 641.4 25.25 635.0 25.00 603.3 23.75 577.9 22.75 650
28 692.2 27.25 685.8 27.00 647.7 25.50 622.3 24.50 700
30 743.0 29.25 736.6 29.00 695.2 27.37 666.8 26.25 750
32 793.7 31.25 787.4 31.00 736.6 29.00 711.2 28.00 —
34 844.5 33.25 838.2 33.00 781.0 30.75 755.6 29.75 —
36 895.3 35.25 889.0 35.00 825.5 32.62 800.1 31.50 —
38 946.1 37.25 939.8 37.00 872.9 34.37 844.5 33.25 —
40 996.9 39.25 990.6 39.00 920.7 36.25 889.0 35.00 —
42 1047.7 41.25 1041.4 41.00 965.2 38.00 933.4 36.75 —
44 1098.5 43.25 1092.2 43.00 1012.6 39.87 977.9 38.50 —
46 1149.3 45.25 1143.0 45.00 1057.1 41.62 1022.3 40.25 —
48 1200.1 47.25 1193.8 47.00 1104.9 43.50 1066.8 42.00 —
CL150 CL300 CL600 CL900
NPS mm in. mm in. mm in. mm in. DN
50 1250.9 49.25 1244.6 49.00 1149.3 45.25 1111.2 43.75 —

52 1301.7 51.25 1295.4 51.00 1193.8 47.00 — — —
54 1352.5 53.25 1346.2 53.00 1241.2 48.87 — — —
56 1403.3 55.25 1397.0 55.00 1285.7 50.62 — — —
58 1454.1 57.25 1447.8 57.00 1330.1 52.37 — — —
60 1504.9 59.25 1498.6 59.00 1374.6 54.12 — — —
CL1500 CL2500
mm in. mm in. DN
1/2 12.7 0.50 11.2 0.44 15

3/4 17.5 0.69 14.2 0.56 20
1 22.1 0.87 19.1 0.75 25
1 1/4 28.4 1.12 25.4 1.00 32
1 1/2 34.8 1.37 28.4 1.12 40
2 47.5 1.87 38.1 1.50 50
2 1/2 57.2 2.25 47.5 1.87 65
3 69.9 2.75 57.2 2.25 80
4 91.9 3.62 72.9 2.87 100
6 136.4 5.37 111.0 4.37 150
8 177.8 7.00 146.1 5.75 200
10 222.3 8.75 184.2 7.25 250
12 263.4 10.37 218.9 8.62 300
14 288.8 11.37 241.3 9.50 350
16 330.2 13.00 276.1 10.87 400
18 371.3 14.62 311.2 12.25 450
20 415.8 16.37 342.9 13.50 500
22 457.2 18.00 377.7 14.87 550
24 498.3 19.62 412.8 16.25 600
26 539.8 21.25 447.5 17.62 650
28 584.2 23.00 482.6 19.00 700
30 625.3 24.62 517.4 20.37 750
(Continued)
CL1500 CL2500
mm in. mm in. DN
32 — — — — —
34 — — — — —
36 — — — — —
38 — — — — —
40 — — — — —
42 — — — — —
44 — — — — —
46 — — — — —
48 — — — — —
50 — — — — —
52 — — — — —
54 — — — — —
56 — — — — —
58 — — — — —
60 — — — — —
The ASME B16.34 standard covers pressure-temperature ratings for forgings,

castings, bars, plates, and tubular products. Pressure-temperature ratings in ASME
B16.34 indicate the maximum allowable pressure for a specific type of material
(e.g. carbon steel) based on the product pressure class rating (e.g. CL300) and
the process temperature. Bodies of valves are typically made of forging for sizes
<2 or casting for sizes ≥2 . Carbon steel body valves are in material group number
1.1 as ASME B16.34 and have forging and casting designations of A105 and A216
WCB, respectively. American Society for Testing and Materials (ASTM) covers the
provided material designations consisting of a letter (A for ferrous materials) fol-
lowed by a sequential number. Table 4.2 extracted from ASME B16.34 includes the
maximum working pressure in bar unit for carbon steel designations mentioned
earlier established on the pressure class and temperature values. As a general rule,
the working pressure is reduced by increasing the temperature.
Socket-welding end or threaded-end valves may be designed for pressure classes
not designated in Tables 4.1 and 4.2. In this case, it is necessary to perform inter-
polation to find the relevant valve inner diameter. ASME B16.34 mandatory
Table 4.2 Pressure-temperature rating values for material group 1.1 as per ASME B16.34.
Working pressure (bar)
Temperature C CL150 CL300 CL600 CL900 CL1500 CL2500 CL4500
−29 to 38 19.6 51.1 102.1 153.2 255.3 425.5 765.9

50 19.2 50.1 100.2 150.4 250.6 417.7 751.9
100 17.7 46.6 93.2 139.8 233.0 388.3 699.0
150 15.8 45.1 90.2 135.2 225.4 375.6 676.1
200 13.8 43.8 87.6 131.4 219.0 365.0 657.0
250 12.1 41.9 83.9 125.8 209.7 349.5 629.1
300 10.2 39.8 79.6 119.5 199.1 331.8 597.3
325 9.3 38.7 77.4 116.1 193.6 322.6 580.7
350 8.4 37.6 75.1 112.7 187.8 313.0 563.5
375 7.4 36.4 72.7 109.1 181.8 303.1 545.5
400 6.5 34.7 69.4 104.2 173.6 289.3 520.8
425 5.5 28.8 57.5 86.3 143.8 239.7 431.5
450 4.6 23.0 46.0 69.0 115.0 191.7 345.1
475 3.7 17.4 34.9 52.3 87.2 145.3 261.5
500 2.8 11.8 23.5 35.3 58.8 97.9 176.3
538 1.4 5.9 11.8 17.7 29.5 49.2 88.6
Source: ASME B16.34/American Society of Mechanical Engineers.
appendix VI provides a method based on basic equations for obtaining the mini-
mum wall thickness of the valves according to the internal diameter and pressure
class of the valve. Table 4.3 summarizes the basic equations for minimum wall
thickness calculation as per ASME B16.34, mandatory appendix VI.
4.2.2 Nonconservation Method

The wall thickness values, calculated according to mandatory appendix VI, are
very conservative. It means that the provided values for wall thickness are rela-
tively high, which leads to heavier and bulkier valves. Therefore, ASME B16.34
provides another way to calculate valve wall thickness using Eq. (4.1). The mini-
mum wall thickness values, tm, calculated based on the equations in mandatory
appendix VI are all greater than the thickness values determined by Eq. (4.1).
Table 4.3 Basic equations for minimum valve wall thickness calculation, as per ASME
B16.34, mandatory appendix VI.
Valve pressure Inside diameter Metric equation

class, Pc (d ) (mm) tm (mm) Round
150 3 ≤ d < 50 tm(150) = 0.064d + 2.34 Off, one decimal

150 50 ≤ d ≤ 100 tm (150) = 0.020d + 4.5 Off, one decimal
150 100 < d ≤ 1300 tm(150) = 0.0163d + 4.70 Off, one decimal
300 50 ≤ d ≤ 100 tm(300) = 0.030d + 4.83 Off, one decimal
Valve pressure Inside diameter Inch equation

class, Pc (d ) (in.) tm (in.) Round
150 0.12 ≤ d < 2 tm(150) = 0.064d + 0.092 Off, two decimals

150 2≤d≤4 tm(150) = 0.020d + 0.18 Off, two decimals
150 4 < d < 50 tm(150) = 0.0163d + 0.185 Off, two decimals
1500 0.12 ≤ d ≤ 50 tm(1500) = 0.18443d + 0.10 Off, two decimals
Minimum Wall Thickness Calculation Based on ASME B16.34

(An Alternative Method to ASME B16.34 Mandatory Appendix VI)
Pc d
tm = 1 5 × 41
2SF − 1 2Pc
where:
tm: Minimum calculated thickness (mm/in.);

Pc: Pressure class designation number (e.g. for class 150, Pc = 150 and for class
300, Pc = 300);
d: Inside diameter of the valve (mm/in.);
SF: Stress-based constant equal to 7000.
Example 4.1 Calculate the wall thickness of a 12 × 10 reduced bore ball valve
with a pressure class of 150 according to both methods provided in ASME B16.34
and this chapter and compare the results.
Answer
A reduced bore, port, or opening valve has a bore or opening inside the valve smal-
ler than the end connections. The opening or bore of the ball valve is 12 at both
ends of the valve and 10 inside the valve. Therefore, the internal diameter or bore
of the valve is equal to 10 in. Using Table 4.1 to obtain the internal diameter of a
valve with a 10 bore and pressure class of 150 gives an inner diameter of 254 mm.
The next step is to use the basic calculations provided in mandatory appendix VI,
as seen in Table 4.3, as follows:
Pc = Pressure class 150 and 100 < d ≤ 1300 t m 150

= 0 0163d + 4 70 tm
= 0 0163 × 254 + 4 70 = 8 84 mm
The next step is to calculate the valve wall thickness as per the less conservative
approach provided by Eq. (4.1) as follows:
Pc d 150 × 254 57,150

tm = 1 5 × =1 5× = = 4 14 mm
2SF −1 2Pc 2 × 7000 −1 2 × 150 13,820
The valve wall thickness is thinner by 4.70 mm using Eq. (4.1) compared to the
basic equations provided in Table 4.3. As a general rule, the wall thickness values
calculated according to basic equations in mandatory appendix VI in Table 4.3
are typically 3–5 mm thicker than the wall thickness values calculated
using Eq. (4.1).
It is important to remember that the wall thickness values provided by ASME

B16.34 and the methods mentioned earlier are minimums. Other parameters, such
as corrosion allowance (CA), abbreviated as CA, could be factored into the
minimum thickness calculated for noncorrosion-resistant alloys, such as carbon
and low-alloy steels. The amount of CA could be standardized to 3 mm for
carbon and low-alloy steels based on the NORSOK M-001 material selection stand-
ard and the L-001 piping and valves standard. It is essential to avoid adding any CA
to exotic materials such as stainless steel and nickel alloys. CA calculations are
covered in detail in Chapter 5. Equation (4.2) shows the effect of CA on the valve
wall thickness.
Effect of Corrosion Allowance on the Valve Wall Thickness
t m,ca = t m + CA 42
where:
tm,ca: Final valve wall thickness after adding a corrosion allowance (mm/in.);
tm: Minimum valve wall thickness (mm/in.);
CA: Corrosion allowance (mm/in.).
Example 4.2
Find the wall thickness of a threaded-end 2 ball valve in a pressure class of 3000 at
50 C that is made of A216 WCB material. What is the valve’s maximum pressure?
Answer
The pressure classes in Tables 4.1, 4.2, and 4.3 are applied to flanged end valves.
However, welding-end and threaded-end valves are typically designed based on
intermediate pressure class designations, including 3000 psi (206.9 bar), 6000 psi
(413.8 bar), and 9000 psi (620.7 bar) (1 bar = 14.5 psi). It is necessary to apply mul-
tiple linear interpolations in this case. Table 4.4 shows pressure-temperature rating
values for all standard pressure classes as well as 3000 psi for the maximum tem-
perature of 100 C. As shown in the table, 3000 psi falls between CL900 and
CL1500. It is important to remember that 3000 psi pressure class indicates the max-
imum pressure of 206.9 bar at a temperature range from −29 to 38 C.
But the pressure value is decreased by increasing the temperature. So the next
step is to find the associated pressure value of 50 C by using interpolation.
Pressure reduction for CL900 valve by increasing the temperature from ambient
to 50 C = 153.2 − 150.4 = 2.8 bar
Pressure reduction for CL1500 valve by increasing the temperature from ambi-
ent to 50 C = 255.3 − 250.6 = 4.7 bar
Table 4.4 Pressure-temperature rating values for all standard pressure classes as
well as 3000 psi for the maximum temperature of 100 C.
Temperature 3000
C CL150 CL300 CL600 CL900 psi CL1500 CL2500 CL4500
−29 to 38 19.6 51.1 102.1 153.2 206.9 255.3 425.5 765.9

50 19.2 50.1 100.2 150.4 X 250.6 417.7 751.9
100 17.7 46.6 93.2 139.8 X 233.0 388.3 699.0
X = 206 9 − Y
255 3 − 206 9 × 4 7 − 2 8 48 4 × 1 9
4 7−Y = = =09 Y =38
255 3 − 153 2 102 1
X = 206 9 − 3 8 = 203 1
Table 4.5 does not provide the internal diameters of valves in a pressure class of
3000 psi that are required to calculate the valve wall thickness. Suppose the inter-
nal diameter of the 2 class 3000 psi valve is calculated. In that case, there is no
formula in Table 4.3 to determine the wall thickness values for valves in interme-
diate pressure classes such as 3000 psi. Thus, the best way to calculate the wall
thickness of the valve with intermediate pressure class in this example is to use
linear interpolation between the thicknesses of 2 valves in pressure classes of
900 and 1500, considering that there is a direct relationship between the pressure
and thickness values of valves.
Table 4.5 Valve inside diameter for size ranges from 1/2 to 2 and pressure classes
of 900 and 1500, as per ASME B16.34.
CL900 3000 psi CL1500
Size Mm in. mm in. mm in.
1/2 12.7 0.50 — — 12.7 0.50

3/4 17.5 0.69 — — 17.5 0.69
1 22.1 0.87 — — 22.1 0.87
1 1/4 28.4 1.12 — — 28.4 1.12
1 1/2 34.8 1.37 — — 34.8 1.37
2 47.5 1.87 X Y 47.5 1.87
P = 150 4 bar @ T = 50 C, Valve size = 2 Pc = 900 and d = 47 5 mm

20 ≤ d ≤ 50 t m 900 = 0 060d + 4 83 = 7 68 mm
P = 250 6 bar @ T = 50 C, Valve size = 2 Pc = 1500 and d = 47 5 mm
20 ≤ d ≤ 50 t m 1500 = 0 18443d + 0 10 = 8 86 mm
P = 203 1 bar @ T = 50 C, Valve size = 2 tm
8 86 − 7 68 203 1 − 150 4
t m − 7 68 = = 0 62
250 6 − 150 4
t m = 7 68 + 0 62 = 8 3 mm
4.2.3 ASME Sec. VIII Div. 02 Wall Thickness Calculation

ASME BPVC (Boiler and Pressure Vessel Code) Sec. VIII Div. 02 is titled “Rules for
Construction of Pressure Vessels.” It covers rules for the designing, manufactur-
ing, testing, inspection, and certification of pressure vessels and boilers. Piping
and valves are also categorized as pressure vessels. The main concept of ASME
BPVC Sec. VIII Div. 02 is “Design by Analysis,” i.e. using stress analysis, finite ele-
ment analysis (FEA), or finite element method (FEM) to demonstrate the accuracy
and/or correctness of the valve design. Besides, the presented thickness calculation
method by ASME Sec.VIII code is less conservative than that in ASME B16.34.
Pipeline valves installed on oil and gas export pipelines are typically large and
high-pressure class, making them very heavy. Even nonconservative wall thick-
ness calculation in ASME B16.34 adds large tonnes of weight to these valves.
Therefore, the main aim of this section is to propose using ASME Sec.VIII Div.
02 instead of ASME B16.34 for pipeline valves’ wall thickness calculation reduces
the valves’ weight. According to ASME Sec. VIII Div.02, Eq. (4.3) is proposed for
the valve wall thickness calculation, assuming the cylindrical valve shell.
Valve Wall Thickness Calculation as Per ASME Sec. VIII Div. 02
D ps
t= e −1 43
2
t = valve thickness (in.)

D = valve diameter (in.)
p = valve design pressure (psi)
s = Allowable stress (psi) taken from table A.1 of ASME B31.3
e = 2.7182
Example 4.3 Calculate the wall thickness of a 30 CL1500 valve based on meth-
ods in ASME B16.34 and ASME Sec. VIII Div. 02. The valve’s body is made in A216
WCB carbon steel material, and the allowable stress of carbon steel at ambient
temperature is equal to 20,000 psi as per ASME B31.3, process piping code. In addi-
tion, assume that the carbon steel material grade A216 WCB can withstand a
maximum of 250 bar as per ASME B16.34.
Answer
Using Table 4.1 to obtain the internal diameter of a valve with a 30 size and
pressure class of 1500 gives an inner diameter (d) of 625.3 mm.
The next step is to use the basic calculations provided in mandatory appendix VI
of ASME B16.34, as seen in Table 4.3, as follows:
Pc = 1500 pressure class, 150 and 3 < d ≤ 1300 tm 1500 = 0 18443d + 2 54
= 0 18443 × 625 3 + 2 54 = 117 86 mm
The next step is to calculate the valve wall thickness as per the less conservative
approach in ASME B16.34 provided by Eq. (4.1) as follows:
Pc d 1500 × 625 3
tm = 1 5 × =15×
2SF − 1 2Pc 2 × 7000 − 1 2 × 1500
1, 406,925
= = 115 32 mm
12,200
The valve wall thickness is thinner by 2.54 mm using Eq. (4.1) compared to the
basic equations provided in Table 4.3.
The next step is to calculate the valve wall thickness based on Eq. (4.3) from
ASME Sec. VIII Div. 02:
t = valve thickness (in.)

D = valve diameter (in.) = 30
p = valve design pressure (psi) = 250 barg = 250 × 14.5 = 3625 psi
s = Allowable stress (psi) for ASTM A216 WCB carbon steel body material of the
valve = 20,000 psi at ambient temperature as per ASME B31.3
e = 2.7182
D ps 30 3625
t= e −1 = e 20,000
−1 = 15 × 1 1987−1 = 2 98in = 75 71mm
2 2
The valve wall thickness saving between ASME Sec. VIII Div. 02 and noncon-
servative ASME B16.34 method is 39.61 mm.
4.3 Wafer Design Thickness Validation
Flange connection is a prevalent method used to attach a valve to a pipe. A flange is

preferred over welding because flanges can be easily installed and dismantled for
maintenance, inspection, or replacement. Valves can be designed with a flangeless
body to save weight, space, and cost. In fact, a wafer design is defined as a flange-
less design with a facing that permits installation between ASME and manufac-
turer standard (MSS SP) flanges. Figure 4.1 illustrates a drawing of the wafer
body butterfly valve.
Wafer valves that are bolted between flanges (e.g. butterfly or ball valve) should
be designed according to the requirements in ASME B16.34 from Sections
(a) through (f ):
a) The design should provide for bolting using all of the bolt holes and the bolt
circle of the specified flange.
b) Bolt holes parallel to the body run, maybe either be threaded or unthreaded.
Threaded holes may be blind holes suitable for use with bolt studs. When
threaded, the full thread engagement, excluding champers, should be provided
to a depth not less than one nominal bolt diameter.
f′
g′
d′2 n
Axial blind holes d′1 m

L
except flange bolt holes
e t
j
f
t d
g
Axial blind hole
c
Valve rod hole
Figure 4.1 A drawing of the wafer butterfly valve.

c) The required minimum valve body wall thickness, tm, should be measured from
the valve body inside the circumference out to the lesser of the valve body out-
side the circumference, or the circumference of the circle inscribed through the
inner tangent points to the flange bolts.
d) The inner ligament (e in Figure 4.1) of either a through hole or a blinded
threaded hole in the vicinity of a stem penetration should not be less than
25% of the required wall thickness of the body neck but in no case less than
2.5 mm (0.1 in.).
e ≥ 0 25 t and e ≥ 2 5 mm
e) The inner ligament ( f and g in Figure 4.1) for holes parallel to the body run
should not be less than 0.25tm, but in no case less than 2.5 mm (0.1 in.). The
sum of the inner and outer ligaments should not be less than tm.
f ≥ 0 25t m and f ≥ 2 5 mm, g ≥ 0 25t m and g ≥ 2 5 mm, and f + g ≥ t m
f) A ligament within the minimum body wall between two adjacent holes within
the minimum body valve ( j in Figure 4.1) should be 0.25tm or greater, but not
less than 2.5 mm (1 in.).
j ≥ 0 25t m and j ≥ 2 5 mm
a) Wall thickness of the valve body neck (t ) is referred to in section (d). The neck
is the top part of the body in Figure 4.1, where parameters such as d , m, n, g ,
and f are shown. Within the area measured from the outer side of the valve
body, along the direction of the valve neck within the distance of 1.1
d × t m , as illustrated in Figure 4.2, the minimum diameter is d and tm is
the minimum wall thickness calculated by methods explained before according
to ASME B16.34. Beyond the aforementioned 1.1 d × t m regions, straight,
circular sections of valve body necks with inside diameter d shall be provided
with local wall thickness at least equal to t , where t is taken from the appro-
priate class rating equation given in Table 4.3 using the relevant diameter value
shown as d using either Eq. (4.4) or Eq. (4.5).
Calculation of d , a Diameter Used to Determine the Minimum

Body Wall Thickness at the Valve Neck Beyond 1.1 d × t m Regions
2d
150 ≤ Pressure class ≤ 2500 d = 44
3
d Pc
1500 ≤ Pressure class ≤ 4500 d = 27+ 45
48 5000
Figure 4.2 Valve neck section.
d′ t′ 1.1 d · tm
d
t
where:
Pc: Pressure class designation number as defined in Eq. (4.1);

d : Body neck inside diameter (mm/in.);
d : The diameter used to determine body neck valve thickness value beyond 1.1
d × t m regions.
b) For the special case where d > 1.5d, it is necessary that the wall thickness be
equal to or greater than t for the entire body neck length having diameter d
including the 1.1 d × t m body region.
c) If the inner diameter of the valve neck is a lot smaller than the inner diameter of
the valve flow passage, i.e. d/d ≥ 4, the minimum body neck wall thickness
over a distance of L = t m 1+ 1 1 d t m , measured from the intersection of
the valve’s body inside diameter and the axis of the body neck outside diameter
shall not be less than t . t is obtained from Table 4.3 according to the internal
diameter (d ) of the corresponding valve body neck. The neck thickness beyond
the distance of L should be obtained from Table 4.3 according to the value of d .
d) In cases where there is a drilling or tapping on the wall of the valve neck parallel
to the direction of the valve neck axis, the sum thickness of the inner and outer
sides be equal or greater than t or tm. Referring to Figure 4.1,
f ≥ 0 25t 2 , g ≥ 0 25t 2 and f + g ≥ t 2 . (t 2 is the valve body neck thickness
associated with d2 , also shown with parameter “n” in Figure 4.1).
Example 4.4 A ball valve in 3/4 and CL300 made of carbon steel is supplied in
the wafer design (see Figure 4.3). The valve sits between two mating flanges, and
connecting bolts pass through the valve’s body. It is not common to use wafer
design for ball valves, unlike butterfly valves. So the concern of the valve purchaser
is that supplied valves do not have sufficient wall thickness to meet the require-
ments in ASME B16.34. Figure 4.4 shows the machining drawing of the ball valve.
Are the thickness values in the machining drawing sufficient? (Note: 3 mm CA is
added to the valve wall thickness.)
Figure 4.3 Compact 3/4 CL300 wafer

ball valve.
10.55 (C2)
7.6
13 (D2)
522
(E
)
M12
24.05 (C1)
7 (D1)
19
M12
24 (A)
27 (B)
Figure 4.4 Compact wafer ball valve body machining drawing.

Answer
The internal diameter, (d), of a 3/4 CL300 valve is 19.1 mm (refer to Table 4.1).
3 ≤ d < 50 t m 300 = 0 080d + 2 29 = 0 080 × 19 1 + 2 29 = 3 82 mm
The carbon steel valve has a CA equal to 3 mm in this case to mitigate the risk of
corrosion and metal loss. The 3 mm CA should be added to the minimum valve
thickness, as per Eq. (4.2).
t m,ca = 3 82 + 3 = 6 82 mm
tm in sections (c), (d), (e), and (f ) will be considered the minimum wall thicknesses,
including the CA, which is equal to 6.82 mm. It is essential to check and ensure
that each section of the valve body highlighted in Figure 4.4 has a higher thickness
than the minimum thickness requirement and complies with ASME
B16.34 sections (a)–(f ).
The internal bolt holes are based on ASME B.1.1, which matches the flange bolt-
ing dimensional to satisfy the requirement in Section (a). Besides, the bolting
depth in Figure 4.4 is 24 mm (equal to parameter A), and the size of the M12 bolt
is equal to 12 mm, and they are threaded in the body. The bolting length is double
the bolting size, which satisfies the requirement in Section (b). Section (b) states
that “when threaded, the full thread engagement, excluding champers, should be
provided to a depth not less than one nominal bolt diameter.”
Referring to the minimum thickness in Section (c) of ASME B16.34, the mini-
mum thickness values of the valve in Figure 4.4 are A = 24 mm, B = 27 mm,
C1 = 24.05 mm, C2 = 10.55 mm, D1 = 7 mm, and D2 = 13 mm. All six parameters
are more than the minimum allowable thickness equal to 6.82 mm, which quali-
fies the design regarding the minimum thickness and Section (c) of ASME B16.34.
The provisions in paragraph (d) do not apply to this example because the picture
does not include the valve body neck.
Referring to Section (e), the inner ligament (C1 and D1 in Figure 4.4) for holes
parallel to the body run should not be less than 0.25tm, but in no case, less than
2.5 mm (0.1 in.). The sum of the inner and outer ligaments should not be less than
tm. The calculations here demonstrate that section (e) requirements are fulfilled.
Inside diameter (d) = 19 mm, CA = 3 mm
C 1 = 24 05, C 2 = 10 55
t m 300 = 0 08 × 19 + 2 29 = 3 82 mm
t m,ca = 3 82 + 3 = 6 82 mm, 0 25 × 6 82 = 1 705 mm
C 1 = 24 05 > 0 25t m,ca = 1 705 and C1 > 2 5 mm Verified
C 2 = 10 55 > 0 25t m,ca = 1 705 and C2 > 2 5 mm Verified
C 1 + C2 = 34 6 > 6 9 mm Verified
D1 = 7 mm, D2 = 13 mm
D1 = 7 > 0 25t m,ca = 1 705 and D1 > 2 5 mm Verified
D1 + D2 = 20 mm > 6 9 mm Verified
Section (f ) addresses ligaments within the minimum body wall between two
adjacent holes within the minimum body valve. The ligament between two holes
(E in Figure 4.4) should not be less than 0.25tm,ca and greater than 0.25 mm.
t m,ca = 6 82 mm 0 25t m,ca = 0 25 × 6 82 = 1 705 mm
E = 7 6522 mm > 1 705 mm and E = 7 6522 mm > 0 25 mm
Thus, the ligament thickness between the two holes is sufficient and complies
with the requirements of Section (f ) of ASME B16.34. Furthermore, all ASME
B16.34 requirements regarding the minimum wall thickness are fulfilled in
this case.
Example 4.5 A 4 CL150 wafer butterfly valve in 25Cr super duplex has a draw-
ing the same as the one illustrated in Figure 4.1. Calculate the minimum wall
thickness values of “m” and “n.” Consider d1 = 19 mm and d2 = 30 mm.
Answer
The first step is to obtain the internal diameter of the valve (d) based on the valve
size and pressure class as per Table 4.1, which is 101.6 mm. Now, it is possible to
obtain the minimum valve wall thickness (tm) according to Table 4.3 as follows:
100 < d ≤ 1300 t m 150 = 0 0163d + 4 70 = 0 0163 × 101 6 + 4 70

= 6 36 mm
Since the valve body is in a super duplex, no CA is required. Refer to Section (c)
of the valve body neck calculation d d1 = 101 6 19 ≥ 4.
d 101 6
L = t m 1+ 1 1 = 6 36 1+ 1 1 = 34 32 mm
tm 6 36
The minimum body neck thickness (m) associated with d1 within the distance of
L shall not be less than t . t is obtained from Table 4.3 according to the internal
diameter d1 of the corresponding valve body neck.
d1 = 19mm,Pc = 150,3 ≤ d < 50 t m 150 = 0 064d1 + 2 34 = 0 064 × 19 + 2 34

= 3 556mm Minimum m = 3 556mm
The minimum body neck thickness (n) associated with d2 beyond the distance of
L should be obtained from Table 4.3 according to the value of d .
2d2 2 × 30
150 ≤ Pressure class ≤ 2500 d = = = 20
3 3
3 ≤ d < 50 t m 150 = 0 064d + 2 34 = 0 064 × 20 + 2 34 = 3 62

Minimum n = 3 62 mm
4.1 Which statement is correct regarding the pressure-temperature rating tables

for valves?
A Pressure-temperature rating tables for valves are provided according to
ASME B16.5.
B Pressure-temperature rating tables are identical for different materials
such as carbon and stainless steel.
C The working pressure values are decreased by increasing the pres-
sure class.
D The valve pressure class values are reduced by raising the temperature.
Answer
Option A is wrong because ASME B16.5 covers industrial flanges, and ASME
B16.34 provides the pressure-temperature rating tables for valves. Option B is
not correct either since the pressure-temperature rating tables are not iden-
tical for different materials. Option C is wrong as the working pressure
values are increased by increasing the pressure class. Option D is the correct
answer.
4.2 What is the thickness value for a 10 ball valve in pressure class 600 calcu-
lated through a more conservative method?
A 17.04
B 19.32
C 20.34
D 21.38
Answer
The internal diameter of the valve as per Table 4.1 is 247.7 mm.
Pc = 600 and d = 247 7 mm 50 < d ≤ 1300

t m 600 = 0 06777d + 2 54 = 19 32 mm
Option B is the correct answer.
4.3 Find the correct statement about wall thickness calculations for the valves.
A The wall thickness of threaded-end and socket weld-ended valves can be
calculated directly from equations in Table 4.3 according to
ASME B16.34.
B The final and minimum thickness values of a valve made of carbon steel
are equal.
C The thickness of a valve in the 600-pressure class is always more than a
valve in the 300-pressure class.
D Inconel 625 weld overlay applied on the valve body in three-millimeter
thickness to prevent corrosion shall not be counted in the valve wall
thickness calculated through the method and equations provided earlier
in this chapter.
Answer
Option A is wrong because threaded-end and socket weld-ended valves do
not have standard pressure classes as per ASME B16.34. So it is impossible
to calculate the valves’ wall thickness with these two ending connections
directly from the given equations provided in ASME B16.34. Linear interpo-
lation as used in Example 4.2 to obtain the wall thickness of a socket-welded
valve in an intermediate pressure class of 3000 is the correct method to cal-
culate the wall thickness of threaded and socket-welded valves. Option B is
incorrect because, typically, a 3 mm CA is required for carbon steel piping
and valves based on the NORSOK L-001 standard. So the final thickness
of a carbon steel valve is 3 mm thicker than the minimum wall thickness
considering Eq. (4.2). Option C is not correct either since the thickness of
a valve depends on the size in addition to the pressure class. For example,
a 10 CL300 valve is thicker than a 2 CL600 valve based on the calculations
as follows:
10 CL300 ID or d = 254 00 mm from Table 4 1 100 < d ≤ 1300 t m 300

= 0 0334d + 4 32 = 12 80 mm
2 CL600 ID or d = 50 8 mm from Table 4 1 50 < d ≤ 1300 t m 600

= 0 06777d + 2 54 = 5 98 mm
The correct answer is option D. For corrosion mitigation, three millimeters

of Inconel 625 are overlaid on the valve’s body made of carbon or low-alloy
steel. However, the calculated wall thickness prevents mechanical failure of
the valve body, so adding a 3 mm Inconel 625 overlay cannot be considered
mechanical failure prevention, and should be included as part of the calcu-
lated wall thickness.
4.4 Which statements are not correct regarding the minimum valve wall thick-
ness calculation according to the ASME B16.34 standard?
A The wall thickness of a ball and gate valve with an inside diameter of
370 mm and a pressure class of 600 is 27.6 mm.
B Ball valves with the same internal diameter and pressure class but are
made of different materials have different body wall thicknesses accord-
ing to the ASME B16.34 standard.
C A weakness of Table 4.3 extracted from ASME B16.34, which is used for
wall thickness calculation, is that it does not cover all of the possible
internal diameters of valves. Thus, it is impossible to have wall thickness
values for internal diameters that are not covered.
D The methods and calculations provided in ASME B16.34 are always used
for minimum valve wall thickness calculations.
Answer
Option A is correct. First, ASME B16.34 provides the same wall thicknesses
for all valves, including ball and gate valves, as long as they have the same
size and pressure class. Let’s try calculating wall thickness according to the
basic equations provided in mandatory appendix VI in ASME B16.34:
Pc = Pressure class 600 and 50 < d ≤ 1300

t m 600 = 0 06777d + 2 54 = 27 6 mm
Option B is not correct because the type of material and its mechanical
strength do not change the wall thickness of the valve as per the ASME
B16.34 standard. In fact, valves made of different materials will have the
same wall thickness as long as they have the same internal diameter and
pressure class. Option C is not completely correct: although not all of the
internal diameters are not covered by ASME B16.34, it is possible to interpo-
late the valve wall thickness values associated with the missing internal dia-
meters. Option D is incorrect because minimum wall thickness calculations
for some valves like those installed on pipelines are typically calculated based
on ASME Sec. VIII Div. 02 to save thickness and weight compared to ASME
B16.34. Thus, except for option A, all other choices are wrong.
4.5 What is the calculated body thickness for a 30 CL1500 valve installed on the
oil export pipeline calculated according to the nonconservative method in
ASME B 16.34? (Note: the minimum diameter of this valve is 671.26 mm.)
A 75.1 mm
B 90.62 mm
C 123.80 mm
D 141.9 mm
Answer
The valve has a specific (special) bore of 671.26, which does not follow the
value given in Table 4.1. The wall thickness of the valve in this example is
calculated as per Eq. (4.1):
Pc d
t=15×
2SF − 1 2Pc
where:
t: Calculated thickness (mm);
Pc: Pressure class designation number (Class 1500, Pc = 1500);
d: Inside diameter of the valve = 671.26;
SF: Stress-based constant equal to 7000.
1500 × 671 26 1,510,335

t=15× = = 123 80 mm
2 × 7000 − 1 2 × 1500 12,200
Therefore, option C is the correct answer.
4.6 Figure 4.5 illustrates the machining drawing of a ball valve body closure that
is the end section of the valve’s body connected to the piping system. The ball
valve is 1 CL300 in carbon steel with three millimeters of CA. Which
statements are correct?
A The minimum thickness of the valve is 4.32 mm.
B The illustrated body closure meets the minimum wall thickness require-
ments of ASME B16.34.
C The minimum thickness of the valve is 7.32 mm.
D The illustrated body closure does not meet the minimum wall thickness
requirements of ASME B16.34.
Answer
The internal diameter, (d), of a 1 CL300 valve is 25.4 mm (refer to Table 4.1).
3 ≤ d < 50 t m 300 = 0 080d + 2 29 = 0 080 × 25 4 + 2 29 = 4 32 mm
Figure 4.5 Ball valve body closure

10.55 (C2) machining drawing.
9.5029 (A)
24.05 (C1)
∅ 19
22.5 (D1)
21.4 (B)
9 (D2)
The carbon steel valve has a CA equal to 3 mm in this case to mitigate the risk
of corrosion and metal loss. The 3 mm CA should be added to the minimum
valve thickness, as per Eq. (4.2).
t m,ca = 4 32 + 3 = 7 32 mm
Therefore, option A is incorrect and choice C is correct.
Referring to the minimum thickness in Section (c) of ASME B16.34,
the minimum thickness values of the valve in Figure 4.5 are A = 9.5 mm,
B = 21.4 mm, C1 = 24.05 mm, C2 = 10.55 mm, D1 = 22.5 mm, and D2 = 9
mm. All six parameters are more than the minimum allowable thickness equal
to 7.32 mm, which qualifies the design regarding the minimum thickness and
Section (c) of ASME B16.34. The provisions in paragraph (d) do not apply to
this example because the figure does not include the valve body neck.
Referring to section (e), the inner ligament (C1 and D1 in Figure 4.5) for
holes parallel to the body run should not be less than 0.25tm, but in no case
Further Reading 147
less than 2.5 mm (0.1 in.). The sum of the inner and outer ligaments should
not be less than tm. The calculations here demonstrate that section (e)
requirements are fulfilled.
Inside diameter (d) = 19 mm, CA = 3 mm
C 1 = 24 05, C 2 = 10 55
t m 300 = 4 32 mm
t m,ca = 4 32 + 3 = 7 32 mm, 0 25 × 7 32 = 1 83 mm
C 1 = 24 05 > 0 25t m,ca = 1 83 and C 1 > 2 5 mm Verified
C 2 = 10 55 > 0 25t m,ca = 1 83 and C 2 > 2 5 mm Verified
C 1 + C2 = 34 6 > 7 32 mm Verified
D1 = 22 5 mm, D2 = 9 mm
D1 = 22 5 > 0 25t m,ca = 1 83 and D1 > 2 5 mm Verified
D1 + D2 = 31 5 mm > 7 32 mm Verified
Section (f ) addresses ligaments within the minimum body wall between two
adjacent holes that do not exist in Figure 4.5. Therefore, the thickness values
provided in the machining drawing are sufficient and fulfil ASME B16.34
requirements. To sum up, options B and C are correct.
Further Reading
American Petroleum Institute 609 (2004). Butterfly Valves: Double Flanged Lug and
Wafer, 6e. Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) (2012). Design and Fabrication of
Pressure Vessels. Boiler and Pressure Vessel Code. ASME Section VIII Div.02.
New York: American Society of Mechanical Engineers (ASME).
American Society of Mechanical Engineers (ASME) B16.34 (2017). Valves–Flanged,
Threaded, and Welding End. New York: American Society of Mechanical
Engineers (ASME).
American Society of Mechanical Engineers B1.1 (2019). Unified Inch Screw Threads
(UN, UNR, and UNJ Thread Forms). New York: American Society of Mechanical
Engineers (ASME).
American Society of Mechanical Engineers B31.3 (2020). Process Piping. New York:
American Society of Mechanical Engineers (ASME).
Nayyar, M.L. (2000). Piping Handbook, 7e. New York: McGraw-Hill Education.
Oxford: Elsevier.
NORSOK L-001 (2017). Piping and Valves, 4e. Lysaker: NORSOK.
NORSOK M-001 (2004). Materials Selection, 4e. Lysaker: NORSOK.
Parisher, R.A. and Rhea, R.A. (2002). Pipe Drafting and Design, 2e. Austin, TX: Gulf
Professional Publishing.
Sotoodeh, K. (2019). Wafer design valves verification based on ASME B16.34. Springer
Nature Applied Science 1: 1476. https://doi.org/10.1007/s42452-019-1344-0.
Sotoodeh, K. (2022). Cryogenic Valves for Liquified Natural Gas Plants, 1e. Austin, TX:
149
Material and Corrosion
5.1 Introduction
Valves are essential components of piping systems in the oil and gas industry. An
oil and gas plant’s efficiency, safety, and reliability largely depend on fluid hand-
ling and transportation through the piping system, including industrial valves.
Material failure because of corrosion is known as one of the significant causes
of valve failure. Some of the negative impacts of valve failure in the oil and gas
industry, especially the offshore sector, can be summarized as loss of assets and
production and Safety and Environmental issues (HSE), including problems like
hydrocarbon (oil and gas) spillage and environmental pollution, loss of human life
in some cases, jeopardizing safety and reliability, etc. Many valves fail every year
due to poor material selection and corrosion. So, proper material selection and cor-
rosion prevention are important aspects of the valve design.
Corrosion occurs when a material deteriorates due to its interaction with its sur-
rounding environment. Corrosive oil and gas in upstream units such as wellhead
and separation as well as downstream plants such as refineries contain a high
number of undesirable corrosive byproducts such as carbon dioxide (CO2) and
hydrogen sulfide (H2S). The most critical and complicated material calculation
for industrial valves is corrosion allowance. It is important to know that corrosion
allowance is selected to mitigate the corrosion produced by carbon dioxide (CO2).
Carbon dioxide is an odorless, nonflammable, and nontoxic substance, unlike
hydrogen sulfide, and colorless like hydrogen sulfide. The second material equa-
tion covered in this chapter addresses the pitting corrosion resistance of materials.
As a result of the presence of chloride in fluids such as seawater, pitting corrosion
occurs. The third aspect to discuss is the carbon equivalent, which is relevant since
it influences the weldability of carbon steels. The last section of this chapter dis-
cusses hydrogen-induced stress cracking (HISC) corrosion and related equations.
150 5 Material and Corrosion
5.2 Carbon Dioxide Corrosion
5.2.1 Corrosion Mechanism

Sweet corrosion due to carbon dioxide is a serious problem in the oil and gas indus-
try as it causes metal loss and localized corrosion. Therefore, accurate prediction
and modeling of carbon dioxide corrosion is an essential task that should be made
during the basic design phase of oil and gas projects. Thus, one of the main aims of
this chapter is to provide a practical model to predict and calculate the carbon diox-
ide corrosion rate and corrosion allowance selection. To calculate the corrosion
allowance, process parameters such as operating pressure and temperature, fluid
pH, the design life of the pipe, and information about glycol, corrosion inhibitor,
etc., should be considered. The first important point is to understand the mechan-
ism of carbon dioxide corrosion. The first assumption in the proposed model and
other models developed to address carbon dioxide is that CO2 is considered a cause
of uniform corrosion. Uniform corrosion is defined as a type of corrosion that pro-
ceeds at the same rate over the whole metal surface. The second important point
about CO2 corrosion is that it is corrosive in the presence of water. Carbon dioxide
can form carbonic acid in a chemical reaction with water as follows:
CO2 + H2 O H2 CO3
Therefore, the main conclusion is that there will be no corrosion by carbon diox-
ide if there is no water. The third key consideration about this type of corrosion is
that it causes metal loss from the internal surface of the piping and valves.
Figure 5.1 shows carbon dioxide corrosion on a carbon steel pipe. The following
paragraph explains the strategies used to prevent this type of corrosion.
Figure 5.1 Carbon dioxide (sweet) corrosion.

5.2.2 Corrosion Mitigation

Sweet corrosion occurrence has been widely experienced in the oil and gas indus-
try. Three main strategies are commonly used in the oil and gas industry to com-
bat and prevent the carbon dioxide or sweet corrosion. The first strategy is to
increase the wall thickness of the piping and valves as a form of corrosion
allowance. CO2 corrosion, also known as sweet corrosion, causes metal loss in
noncorrosion-resistant materials such as carbon steel. So, adding an extra thick-
ness to the piping and valves – a practice called “corrosion allowance” – is pro-
posed for carbon and low alloy steels. It is typical and logical to add a 1, 1.5, or 3
mm corrosion allowance to the piping and valves made in noncorrosion-resistant
alloys like carbon and low alloy steels. However, adding a 6 mm corrosion allow-
ance to the pipe will make it very thick, heavy, and expensive. A standard
amount of 3 mm corrosion allowance is proposed to be added to piping and
valves in carbon steel material according to Norsok, the Norwegian petroleum
standard. Adding more corrosion allowance such as 6 mm is not proposed since
it makes the piping and valves thicker and heavier. Many different models and
software may be used to calculate the CO2 corrosion rate, all of which are based
on the “Dewaard and Milliams” model. Additionally, the Iranian Petroleum
Standard (IPS) as well as Norsok standard M-506 provide a calculative model
for CO2 corrosion prediction.
The alternative solution and the second strategy to combat sweet corrosion is
using corrosion-resistant alloys (CRAs) such as stainless steel or nickel alloys like
Inconel 625 instead of carbon steel. This practice is common in the Norwegian off-
shore industry. Nickel alloys such as Inconel 625 can be used as a form of solid pipe
or cladding (weld overlay). A cladding carbon steel pipeline with Inconel 625 is
proposed for both hydrogen sulfide and carbon dioxide-containing oil and gas ser-
vices to mitigate both types of corrosion. The third sweet corrosion combat strategy
is injecting the corrosion inhibitor or anticorrosive, a chemical compound inside
the oil and gas services. Different corrosion inhibitors may be injected into the
fluid with different chemistry, physical properties, and solubility. Corrosion inhib-
itor protection is usually from one or more of the following mechanisms: first, the
inhibitor molecule is absorbed by the metal surface; second, the inhibitor makes a
protective film; and the last is that the inhibitor reacts with corrosive substances in
the piping system. The selection of the best and most effective corrosion inhibitor
is performed as per laboratory results, considering specific parameters such as type
of protected material, application method, efficiency, cost, HSE considerations,
and product availability. Like corrosion inhibitors, glycol or methanol can be
injected into the piping system to absorb the water and reduce the corrosion rate.
The following section is about carbon dioxide corrosion rate as well as corrosion
allowance calculations.
5.2.3 Corrosion Rate Calculation

Researchers and oil companies have developed several corrosion estimation mod-
els. Some of the known and commonly used models in the oil and gas industry are
listed as follows:
• DeWaard and Milliams Model (DWM);

• Predict Model (Intercorr);
• Norsok (M-506) (Statoil, Saga Petroleum, Hydro);
• LIPUCOR (Total);
• Hydrocorr (Shell);
• Casandra (British Petroleum);
• CorPos (Corr Ocean);
• Cormed (Elf Aquitaine).
There is no totally valid model for carbon dioxide corrosion estimation due to the
complexity of this type of corrosion. The suggested model in this book uses various
techniques listed here to deliver a practical way of predicting carbon dioxide cor-
rosion in pipelines and piping systems, including industrial valves, where carbon
steel or low-alloy steel material is used. The essential consideration is that the pro-
posed model can only be used for hydrocarbon services such as oil and gas. There-
fore, this section provides a model for corrosion rate calculation that is not
applicable for piping and valves in non-hydrocarbon fluids such as seawater,
drinking water, air, hydraulic oil, and caustic services. After calculating the
CO2 corrosion rate, it is possible to select the most suitable material and the cor-
rosion allowance.
5.2.3.1 Basic CO2 Corrosion Rate

The evaluation of the basic corrosion rate takes into account the effects of operat-
ing temperature and pressure, as well as the mole fraction of CO2. Basic carbon
dioxide corrosion rate prediction is calculated through Eq. (5.1).
Basic CO2 Corrosion Rate Calculation

1710
Log CR = 5 8− + 0 67 log PCO2 51
base
T
where:
CRbase: Base corrosion rate based on DWM (mm/year);

T: Operating temperature (K);
PCO2 : Partial pressure of CO2 (bar).
5.2.3.1.1 Effect of Carbon Dioxide Dry carbon dioxide is noncorrosive, whereas it

is corrosive when dissolved in water. Thus, water presence is necessary for sweet
corrosion to happen, as carbonic acid forms and decreases the pH of the water. The
final pH of the fluid service depends on the temperature and carbon dioxide partial
pressure.
5.2.3.1.2 Effect of Temperature According to Figure 5.1, some laboratory studies

show that the CO2 corrosion rate increases by increasing the temperature up to
almost 70 C, probably due to an increase in mass transfer rate. Above that tem-
perature, the corrosion rate starts to decrease, associated with a corrosion-
protective layer that is explained more in detail later in this chapter (Figure 5.2).
5.2.3.1.3 Effect of Pressure The partial pressure of CO2 depends on the mole
fraction of CO2 and is calculated according to Eq. (5.2):
CO2 Partial Pressure Calculation

PCO2 = Poperation × x CO2 52
mm/year
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0 20 40 60 80 100 120 140 160
Temperature (°C)
Figure 5.2 Effect of temperature on carbon dioxide corrosion rate based on DWM.
where:
PCO2 : Partial pressure of CO2 (bar);

Poperation: Operation pressure (bar);
X CO2 : Mole fraction of CO2.
The mole fraction represents the ratio of the number of molecules of a particular
component to the total number of molecules.
Example 5.1 The operating pressure of a 10 ball valve used for the stop/start of
the gas is 25 bar, and the carbon dioxide mole fraction is 0.008. The valve-operating
temperature is 120 C. Calculate the basic carbon dioxide corrosion rate in milli-
meters per year.
Answer
The partial pressure of the carbon dioxide is calculated as follows:
PCO2 = Poperation × x CO2 = 25 × 0 008 = 0 2 bar
120 C is equal to 393 K.

1710 1710
Log CR = 5 8− + 0 67 log PCO2 = 5 8 − + 0 67 log 0 2
base
T 393
= 5 8 − 4 35 − 0 468 = 0 98 CR base = 9 55 mm year
DeWaard and Milliams generated a nomogram (see Figure 5.3) in 1975 (DWM)
that can be used to estimate the basic carbon dioxide corrosion rate as an alterna-
tive method to Eq. (5.1). Connecting the operating temperature of 120 C on the
left side to the CO2 partial pressure on the right equal to 0.2 passes through the
corrosion rate of 10 mm/year, which is very close to the result provided by the cal-
culations above. The basic corrosion rate of 9.55 mm/year is a high number that
should be reduced to a realistic corrosion rate by corrective factors, as explained in
the next section.
5.2.3.2 Corrective CO2 Corrosion Rate

Additional corrective process parameters shall modify the initially calculated cor-
rosion rate to deliver more accurate estimations of the corrosion rate. This
section suggests effective process parameters to adjust the basic corrosion rate
as follows:
• F (System): This parameter is related to the nonideality of hydrocarbon gas on

the calculated corrosion rate that is designated as the “fugacity coefficient.”
Nonideality of hydrocarbon gas is explained more in detail further in this
chapter.
Figure 5.3 Carbon dioxide CO2 pressure

basic corrosion rate bar 10
Temperature
prediction nomogram as per
°C Scale
the DWM model. 140 factor
130 0.1
120 Corrosion rate
110 mm/year
100 20
1
90 1 10
80
70
60
50 1
40 0.1
Example:
30 0.2 bar CO2 at 120 °C
gives 10 × 0.7 = 7 mm/year
20 0.1
10
0 0.02 0.01
• F (Glycol): This parameter considers the effect of glycol injection (typically

monoethylene glycol or MEG) on the corrosion rate calculations. Glycol will
reduce the corrosivity of the hydrocarbon by absorbing water. Besides, glycol
may be injected into the piping system to prevent hydrate formation. When nat-
ural gas is transported with water in low-temperature and high-pressure condi-
tions, ice-like crystals mainly formed from water can cause flow assurance
problems called hydrate.
• F (Scaling): This parameter evaluates the chance of protective anticorrosion film
of FeCO3 production, which is very likely at high temperatures (above 70 C)
when the pH is high.
• F (Water cut): The presence of free water is absolutely essential for corrosion to
occur. Higher corrosion rates occur when the water amount is high.
• F (Inhibitor): This parameter demonstrates the mitigation impact of any corro-
sion inhibitor injections on the initially calculated corrosion rate. For many
years, it has been common practice in the oil and gas industry to inject inhibitors
into carbon dioxide-containing fluids carried in the carbon steel piping and
valves.
• F (Condensate): It demonstrates the effect of water condensation formation for
nominally dry gas lines in which the corrosion rate is reduced substantially.
• F (pH): It is generally accepted that pH has a critical impact on the corrosion rate
in such a way that low pH values of less than five are considered acidic and pro-
duce an acidic environment that largely extends the corrosion rate.
Thus, the corrective corrosion rate (mm/year) is calculated according

to Eq. (5.3).
CO2 Corrective Corrosion Calculation

CR correct = CR base × F System × F Glycol × F Scale
× F Water cut × F Inhibitor × F Condensate × F PH
53
5.2.3.2.1 Effect of System F (System) is the outcome of the wet gas nonideality
due to both pressure and temperature conditions that can affect the CO2 activity in
piping and valves in a negative manner. The partial pressure of CO2 is assumed to
be 100% efficient (completely ideal) in the base corrosion rate equation of DWM,
which is not realistic in practice. Thus, it is a common practice to replace the CO2
partial pressure by the CO2 fugacity or activity of F(CO2) which is calculated
in Eq. (5.4):
CO2 Fugacity or Activity Calculation

F CO2 = a × PCO2 54
where:
F CO2 : CO2 fugacity or activity;

α: CO2 activity coefficient, known as F (system); calculated based on Eq. (5.5);
PCO2 : Partial pressure of CO2 (bar).
CO2 Activity Coefficient or F (System) Calculation

14
Log a = 0 0031 − ×P 55
T
where:
a: CO2 activity coefficient or F (system);

T: Operating temperature (K);
P: Operating temperature (bar).
It should be noted that if the calculated value of F (system) or α is more than one,
the CO2 is assumed to be 100% active and ideal, and, therefore, F (system) = 1.
Example 5.2 Calculate the activity of carbon dioxide in the piping system in the
previous example where the basic corrosion rate was calculated to be equal to
9.55 mm/year.
Answer
14
Log a = 0 0031 − ×P
T
14
Log a = 0 0031− × 25 log a = −0 01156 a = F System
120 + 273
= 10 −0 01156 = 0 97
F CO2 = a × PCO2 = 0 97 × 0 2 bar = 0 194 bar
5.2.3.2.2 Effect of Glycol (Methanol) Glycol or methanol has a very effective inhib-
itive effect; it functions by absorbing and reducing the water, which results in a
lower corrosion rate and reduced hydrate formation. The reduction of the corro-
sion rate due to the presence of glycol, F (glycol), is calculated based on Eq. (5.6), as
follows:
Effect of Glycol Injection on Corrosion Rate

Log F g = 1 6 × log W g − 3 2 56
where:
F(g): Glycol reduction effect on the corrosion rate;

W(g): Water concentration percentage in the water/glycol mixture.
Note: F (glycol) = 1 for lines without any glycol injection.
Example 5.3 Glycol is injected into two piping systems carrying corrosive CO2
containing gas mixed with water. The water concentration in the water and glycol
mixture is 40% for the first case and 80% for the second case. Calculate the glycol
reduction effect on each case’s carbon dioxide corrosion rate.
Answer
W g 1 = 40 log F g 1 = 1 6 × log W g 1 − 3 2 = 1 6 × log 0 4 − 3 2

−4
= −3 84 F g 1 = 1 445e
W g 2 = 80 log F g 2 = 1 6 × log W g 2 − 3 2 = 1 6 × log 0 8 − 3 2

−4
= −3 36 F g 2 = 4 365e
F g2 4 365e − 4
= = 3 02
F g1 1 445e − 4
Thus, the CO2 corrosion reduction effect is approximately three times higher in
the second case, where the percentage of water in the glycol and water mixture is
80%. It is noticeable that the concentration of water absorbed by the glycol in the
second case is double that in the first case.
5.2.3.2.3 Effect of Scaling The scaling effect is connected to producing a protect-

ing layer on the metal surface at temperatures higher than 70 C. High tempera-
tures will activate the CO2 and lead to its reaction to the iron, which will produce a
protective layer of ferrous carbonate (FeCO3 ) on the metal surface. Scaling forma-
tion reduces the corrosion rate seriously by the corrective factor of F (scaling)
based on Eq. (5.7), as follows:
Effect of Scaling Calculation
2400
Log F S = − 0 6 × log F CO2 − 6 7 57
T
where:
F(S): Scaling effect coefficient;

F CO2 : CO2 fugacity/activity;
T: Operating temperature (K).
It should be noted that scaling effect exists if (F(S)) is less than 1; otherwise, there
is no possibility of scale formation and F(S) = 1 is utilized for corrosion rate
calculations.
Example 5.4 An 8 valve handles the corrosive crude oil service in 100 bar oper-
ating pressure and 120 C operating temperature. Considering the carbon dioxide
mole fraction is 0.002, can scaling be produced in the valve?
Answer
The partial pressure of the carbon dioxide is calculated as follows:
14
Log a = 0 0031 − ×P
T
14
Log a = 0 0031 − × 100 log a = − 0 046
120 + 273
a = 10 − 0 046 = 0 8994
F System = 0 9
F CO2 = a × PCO2 = 0 9 × 0 2 bar = 0 18 bar
2400 2400
Log F Scale = −0 6 × log F CO2 − 6 7 = −0 6 × log 0 18 − 6 7
T 393
= 6 1 + 0 447 −6 7 = − 0 153 F Scale = 10 − 0 153 = 0 70
Since the effect of scaling is equal to 0.7, which is less than one, the scaling can be
produced in the pipes and valves given in Example 5.4.
5.2.3.2.4 Effect of Water Cut Dry hydrocarbons without water are noncorrosive
because the presence of water in oil and gas is vital for the occurrence of corro-
sion. The presence of oil is typically deemed beneficial in corrosion prevention
due to the formation of a protective oil film. But the stability of the protective
layer depends on the water quantity (water cut). The oil-forming film on a metal
surface can be stable at up to a 20–40% water cut, but higher water quantities
eliminate its protective effect. However, it is not actually possible to calculate
and define the minimum volume of water required for corrosion. Since the cal-
culation of the water cut is not possible, the effect of the water cut should be
assumed to be equal to 1 in. the calculations unless, otherwise, the water cut
value is known.
5.2.3.2.5 Effect of Inhibitor For many years, it has been commonplace in the oil
and gas industry to inject various chemicals as corrosion inhibitors in pipes and
valves handling corrosive services to lower the possibility of corrosion by creating
a strong film on the metal surface. It is generally accepted to define the inhibitor’s
capability to reduce the corrosion attack by a parameter known as “inhibitor
efficiency,” calculated as per Eq. (5.8):
Corrosion Inhibitor Efficiency
CR with inhibitor
Inhibitor efficiency = 1 − 58
CR without inhibitor
Inhibitor efficiency is usually described as a percentage. For example, if the cor-

rosion rate is 3 mm/year without corrosion inhibitor injection and 0.3 mm regard-
ing the inhibition outcome, the calculated corrosion inhibitor efficiency is 90%.
Inhibition efficiency above 85% is generally considered ideal; it means an 85% cor-
rosion rate reduction compared to the corrosion rate obtained from Dewaard
and Milliams nomogram based on Figure 5.3. F (inhibitor), which is equal to the
corrosion reduction factor due to inhibitor injection according to the DWM, can be
obtained by Eq. (5.9) as follows:
Effect of Inhibitor Injection
F i = 1 − Inhibitor efficiency 59
It is known to consider 90% efficiency or availability, F (inhibitor) = 0.1 for the

corrosion inhibitor during the engineering design phase. However, corrosion
availability shall be regarded as 85% for condensate, 90% for gas lines, and 95%
for oil streams.
Example 5.5 The valve and material engineers have decided to inject a corrosion
inhibitor inside the piping systems with a corrosion rate of 1 mm/year to
decrease the corrosion rate. The corrosion rate is reduced to 0.05 mm/year due
to inhibitor injection. Calculate the inhibitor efficiency and the effect of inhibitor
injection.
Answer
CR with inhibitor 0 05 mm year
Inhibitor efficiency = 1 − = 1− = 95
CR without inhibitor 1 mm year
Effect of inhibitor injection = 1 – 0 95 = 0 05
5.2.3.2.6 Effect of Condensate Condensate development is taken into account for

gas services with humidity, where cooling below the gas dew point results in con-
densate droplets formation. This situation will likely produce a protective film on
the pipe and substantially lower the corrosion rate. If condensate is made, its effect
will be 0.1; otherwise, F (condensate) equals 1.
5.2.3.2.7 Effect of pH The pH has an essential effect on the corrosion rate

because an acidic environment with a low pH below five essentially boosts the cor-
rosion rate. The pH can be calculated as a function of different parameters like the
carbon dioxide partial pressure, operation temperature, bicarbonate, and other
dissolved gases and chemicals. To calculate the effect of pH on the corrosion rate,
first, pHsat is calculated using Eqs. (5.10) and (5.11).
pHsat Calculation
1307
pHSat1 = 1 36 + − 0 17 × log F CO2 5 10
T
pHSat2 = 5 4 − 0 66 × log F CO2 5 11
The lowest value obtained from Eqs. (5.10) and (5.11) is selected for the follow-
ing step, in which the effect of pH is calculated using Eq. (5.12) or (5.13) depending
on the relationship between pHsat and the effect of pH:
pH Effect Calculation
If pHSat > pHactual
Log F pH = 0 32 pHSat − pHactual 5 12
If pHSat ≤ pHactual
16
Log F pH = 0 13 pHSat − pHactual 5 13
The effect of pH shall be considered equal to one in two cases as follows:
• If the pH of the fluid is unknown;

• Scaling is produced in piping and valves when the scaling effect is less than one.
5.2.3.3 Final CO2 Corrosion Rate

Equation (5.14) is used to calculate the final CO2 corrosion rate by multiplying the
corrected corrosion rate in millimeters per year by the design life of the plant. The
design life of a valve or plant could be 10, 15, 20, 25, or even 30 years.
Final Corrosion Rate Calculation

CRFinal = Design life year × CRcorrective mm year 5 14
The final corrosion rate should be rounded up to 1, 3, or 6 mm to obtain the cor-

rosion allowance as per Table 5.1:
Table 5.1 Final corrosion rates.
Corrosion Average of corrosion Final selected

severity level rate (mm/year) corrosion allowance
Mild <0.04 1.0

Medium 0.04–0.12 3.0
Severe 0.12–0.24 6.0
Note 1
The design life of the plant and valve is considered 25 years in Table 5.1.
The average corrosion rates differ for a design life other than 25 years.
Note 2
For offshore piping and valves on the platforms and ships, the NORSOK stand-
ard proposes standardizing the corrosion allowance for carbon steel piping and
valves to 3 mm. A corrosion allowance of 6 mm on carbon steel should be
avoided as it would increase the piping and valve wall thickness and make
them heavy. However, it is practical to use a 6 mm corrosion allowance for pip-
ing and valves in onshore units such as refineries and petrochemical plants.
Note 3
For onshore piping and valves in the refineries and chemical plants, if the
calculated final corrosion rate exceeds 6 mm, various solutions are recom-
mended; upgrading the material to a CRA such as stainless steel, applying Inco-
nel 625 on the internal surface of carbon steel in contact with the fluid service,
or injection of corrosion inhibitor.
Note 4
No corrosion allowance is required for CRA materials such as austenitic stain-
less steels (e.g. 304, 316), duplex stainless steels, and nickel alloys.
Note 5
It is possible for a piping or valve specification to call for minimum corrosion
allowances of 1.5 or 1 mm in onshore carbon steel applications such as steam
and water that do not contain hydrocarbons.
5.3 Pitting Corrosion
Figure 5.3 illustrates a type of local corrosion known as pitting corrosion that cre-
ates holes inside exposed facilities and susceptible materials. In valves and piping
systems, pitting corrosion may occur internally as a result of fluid service and
5.3 Pitting Corrosion 163
Figure 5.4 Pitting corrosion in

carbon steel. Source: MDPI.
externally as a result of exposure to the corrosive offshore environment (marine).

Internal pitting corrosion is mainly caused by the presence of chloride in corrosive
seawater as well as oil and gas services. Chloride can impair the passive layer on
sensitive materials, such as austenitic stainless steel, where pitting corrosion initi-
ates (Figure 5.4).
Two essential parameters contribute to the severity of pitting corrosion:
the metallurgy of the metallic valve material and the severity of operating con-
ditions. It is common to use various approaches in the industry to prevent
pitting corrosion, including selecting materials with high pitting resistance –
typically with a pitting resistance equivalent number (PREN) higher than
40 – pitting corrosion test, control of pitting-effective process parameters, such
as pH, temperature, and chloride; and lastly, cathodic protection. PREN
depends on the three main elements in a material and is calculated according
to Eq. (5.15).
Pitting Resistance Equivalent Number (PREN) Calculation

PREN = Cr + 3 3 Mo + 16N 5 15
where:
Cr: Chromium;
Mo: Molybdenum;
N: Nitrogen.
Example 5.6 Calculate the maximum PREN values for the materials listed in
Table 5.2 and identify those with high PREN values.
Table 5.2 Chemical compositions of some of the most commonly used stainless steels and nickel alloys in the oil and gas industry.
C Mn P S Cr Ni Mo N
Super austenitic
6MO (UNS S31254) Max. 0.02 Max.1 0.03 0.01 20 18 6 0.21
Nickel alloys
Inconel 600 (UNS N06600) 0.15 1 — 0.015 15.5 72 — —
Inconel 625 (UNS N06625) 0.1 0.5 0.015 0.01 21.5 58 8.5 —
Incoloy 800 (UNS N08800) 0.10 1.5 — 0.015 21 32.5 — —
Incoloy 825 (UNS N08825) 0.05 1 — 0.03 21 42 — —
Duplex
22Cr duplex (UNS S31803) Max. 0.03 Max.2 Max. 0.03 Max. 0.020 25 5.5 3 0.08–0.20
25Cr super duplex (UNS S32750) Max. 0.03 Max. 1.20 Max. 0.035 Max. 0.020 25 7 4 0.24–0.32
25Cr super duplex (UNS S32760) Max. 0.03 Max. 1 Max. 0.03 Max. 0.010 25 7 3.5 0.24–0.32
Austenitic
304 (UNS S30400) 0.08 0–2 Max. 0.045 Max. 0.03 18–20 8–11 Max. 0.1
304L (UNS S30403) Max. 0.03 0–2 Max. 0.045 Max. 0.03 18–20 8–13 Max. 0.1
316 (UNS S31600) 0.08 0–2 Max. 0.045 Max. 0.03 16–18 10–14 2–3 Max. 0.1
316L (UNS S31603) Max. 0.03 0–2 Max. 0.045 Max. 0.03 16–18 10–14 2–3 Max. 0.1
321 (UNS S32100) 0.08 0–2 Max. 0.045 Max. 0.03 17–19 9–12 Max. 0.1
347 (UNS S34700) 0.08 0–2 Max. 0.045 Max. 0.03 17–20 9–13 Max. 0.1
5.4 Carbon Equivalent 165
Answer
PREN6MO = Cr + 3 3 Mo + 16N = 20 + 3 3 × 6 + 16 × 0 21 = 43 16 > 40

6MO is a pitting resistant material
PRENInconel 600 = 15 5 < 40 Inconel 600 is not a pitting resistant material
PRENInconel 625 = 21 5 + 3 3 × 8 5 = 49 55 > 40
Inconel 625 is a pitting resistant material
PRENIncoloy 800 = 21 < 40 Incoloy 800 does not resist pitting
PRENIncoloy 825 = 21 < 40 Incoloy 825 is not a pitting resistant material
PREN22Cr duplex = 25 + 3 3 × 3 + 16 × 0 20 = 38 1 < 40
22Cr duplex is not a pitting resistant material
PREN25Cr super duplex UNS S32750 = 25 + 3 3 × 4 + 16 × 0 32 = 43 32 > 40
25Cr super duplexUNS S32750 is a pitting resistant material
PREN25Cr super duplex UNS S32760 = 25 + 3 3 × 3 5 + 16 × 0 32 = 41 67 > 40
25Cr super duplex UNS S32760 is a pitting resistant material
PRENSS304 = 20 + 16 × 0 1 = 21 6 < 40
SS304 is not a pitting resistant material
PRENSS304L = 20 + 16 × 0 1 = 21 6 < 40
PRENSS316 = 18 + 16 × 0 1 = 19 6 < 40
PRENSS316 = 18 + 16 × 0 1 = 19 6 < 40
PRENSS321 = 19 + 16 × 0 1 = 20 6 < 40
PRENSS347 = 20 + 16 × 0 1 = 21 6 < 40
5.4 Carbon Equivalent
One of the most significant considerations when using carbon steel piping and
welded valves is their ability to be welded. Among other things, carbon and certain
other elements in carbon steel increase its hardness and can make welding difficult
or impossible. Additionally, a high hardness at the weld joints makes it more likely
that corrosion will occur, such as a hydrogen sulfide attack. In fact, reducing the
hardness of materials is a key factor in reducing the risk of sour corrosion. For
these reasons, it is recommended that the carbon content and equivalent carbon
content of carbon steel piping and welded valves be limited. The carbon equivalent
equation is a method of converting alloy elements’ percentages to an equivalent
carbon percentage for carbon steel. Equations (5.16) and (5.17) demonstrate
how the carbon equivalent (CE) can be calculated as described earlier. Pipes
and welded valves made from carbon steel should ideally have a carbon equivalent
value of less than 0.40 in order to ensure acceptable weldability and corrosion
resistance.
Carbon Equivalent Calculation
Mn Cr + Mo + V Ni + Cu
CE = C + + + < 0 40 5 16
6 5 15
Mn
CE = C + 5 17
6
where:
C: Carbon;
Mn: Manganese;
Cr: Chromium;
Mo: Molybdenum;
Ni: Nickel;
Cu: Copper.
If values of other alloys (other than carbon and manganese) are not known,
Eq. (5.17) can be used to calculate the carbon equivalent. It should be noted that
the accepted value of carbon equivalent is usually stated in the project specifica-
tions according to the requirements of the end user. In this book, although the pro-
posed and ideal value for carbon steel CE is less than 0.40, other maximum values
of carbon equivalent have been observed by the author in different projects, such
as 0.39 (stricter than 0.40) or 0.42 (less strict than 0.4), to meet the clients’ or end
users’ requirements.
Example 5.7 The chemical composition percentages for carbon steel welded
valves in three grades, ASTM A106 Gr. A, B, and C are provided in Table 5.3. What
grades offer good weldability?
Table 5.3 Chemical properties percentages of carbon steel pipe grades.
C Mn Cr Mo V Ni Cu
A 106 Gr. A 0.25 0.5 0.3 0.1 0.05 0.4 0.3

A 106 Gr. B 0.30 0.5 0.3 0.1 0.05 0.4 0.3
A 106 Gr. C 0.35 0.5 0.3 0.1 0.05 0.4 0.3
Answer
CE A106Gr A = C + + +
6 5 15
05 0 3 + 0 1 + 0 05 04+03
= 0 25 + + +
6 5 15
= 0 25 + 0 0833 + 0 09 + 0 0466 = 0 47
CE A106Gr B = C + + +
6 5 15
05 0 3 + 0 1 + 0 05 04+03
= 0 30 + + +
6 5 15
= 0 30 + 0 0833 + 0 09 + 0 0466 = 0 52
CE A106Gr C = C + + +
6 5 15
05 0 3 + 0 1 + 0 05 04+03
= 0 35 + + +
6 5 15
= 0 35 + 0 0833 + 0 09 + 0 0466 = 0 57
In accordance with Eq. (5.16), the carbon equivalent of these three grades is 0.47,
0.52, and 0.57 for grades A, B, and C of A106, respectively. CE values exceed 0.4 for
all three grades, which means none of them can be considered acceptable for
welding.
5.5 Hydrogen-Induced Stress Cracking (HISC) Corrosion
Over the last 15–20 years, the subject of HISC has received considerable attention
in the subsea oil and gas industry because of a number of very costly failures of
subsea components. In order to understand the nature of this form of corrosion,
it is imperative to understand that HISC is categorized as the result of an environ-
ment-aided cracking mechanism caused by either applied or residual stress. For
HISC to occur, three factors are required: susceptible material, stress, and hydro-
gen formation due to the use of cathodic protection.
5.5.1 HISC and Vulnerable Materials

The risk of HISC applies to duplex, super duplex, hard nickel alloys, and high-
strength low-alloy steel materials. Ferritic–austenitic stainless steels, such as
duplex and super duplex, offer high mechanical resistance, corrosion resistance,
and low cost, making them suitable for applications in the subsea oil and gas
industry. They have been used in subsea piping and industrial valves for over
20 years. Nonetheless, these steels are also known to be vulnerable to HISC in sub-
sea environments. High-strength carbon and low-alloy steels, which are widely
used for the bodies and bolting of subsea valves, are thought to be susceptible
to HISC due to their large mechanical strengths. Material microstructure plays
a significant role in determining HISC severity. Hydrogen atoms are generally
believed to diffuse between the metallic matrix grain boundaries due to their smal-
ler size.
5.5.2 HISC and Stress

Materials subjected to tensile stress exhibit three distinct reactions in relation to
the occurrence of HISC. In the first step, the material is deformed, and additional
space is provided for hydrogen to enter into the microstructure. Furthermore, the
stress increase in the material and crack initiation provide additional space for
hydrogen to enter. The material stresses and cracks become more intense as a
result. In HISC analysis, it is critical to take into account a variety of loads, such
as the external loads applied by the connected piping, such as axial, bending, and
compression loads, internal loads primarily caused by the internal pressure of the
piping system, accidental or shock loads caused by accidents such as dropped
objects or natural disasters, such as hurricanes, and finally residual stresses due
to fabrication, manufacturing, and welding processes.
5.5.3 HISC and Cathodic Protection

In this form of corrosion, hydrogen atoms produced by cathodic protection move
into the material to cause it to crack. In a subsea environment, all subsea
components such as pipes, valves, and structures are subject to external corrosion.
Cathodic protection is one method of protecting steel against external corrosion. It
is through the transfer and loss of electrons from the anode that metals corrode or
rust. By passing a direct current through a pipe, valve, or structure that serves as a
cathode, corrosion can be mitigated. By reason of the potential difference between
the anode and the cathode, electrons leave the anode and are transferred to the
cathode; this process is known as cathodic protection. The cathodic protection
method also produces atomic hydrogen on the metallic surface, which may result
in HISC corrosion in materials which are vulnerable to corrosion. It is a common
practice to use zinc, magnesium, or aluminum as sacrificial anodes that will easily
corrode and release electrons for the protection of the cathode, which may consist
of pipes, valves, structures, or other components. There is another critical factor,
which is the electrolyte, or seawater because it is necessary in order for electrons to
be able to flow from the anode to the cathode. Cathodic protection is not the only
method for mitigating external corrosion in subsea environments; it can also be
combined with another approach, such as coating.
5.5.4 HISC and DNV Standard

A great deal of effort has been put into identifying the failure mechanisms asso-
ciated with HISC recently. The result is that Det Norske Veritas (DNV) has devel-
oped guidelines for the design of duplex components based on the DNV-RP-F112
standard to prevent HISCs. The gap in the DNV standard is to consider only duplex
and super duplex as susceptible materials to HISC. To mitigate HISC failures, there
are two general criteria for stress and strain limits in the DNV standard. The first
step is creating stress limits based on finite element analysis, also known as linear
elastic analysis. The strain limit, also referred to as elastic-plastic analysis, is
another result obtained from hand calculations. DNV proposes a linear elastic
stress-strain model to evaluate HISC. As a result of linear elastic analysis, or load
analysis, three main principles are observed: stress is always proportional to strain,
the material is assumed to be continually deformed by an increasing load, and the
deformation stops and the material returns to its original shape upon releasing or
removing the load. For HISC analysis, stress is linearized over the thickness of pip-
ing and valves in two ways: membrane stress and bending stress. Peak stress,
another type of stress, can be ignored in HISC analysis. Defining membrane
stress δm as the average, uniform distribution of stress across its thickness, it
could be either compressive or tensile in nature. The bending stress δb, unlike
the membrane stress, varies across the thickness of a component; it is applied
along its longitudinal axis and is either tensile or compression-type. As shown
in Figure 5.5, the distribution of membrane, bending, and membrane plus bend-
ing across the thickness of a component (e.g. pipe or valve) can be seen. Accord-
ing to Eqs. (5.18) and (5.19), DNV guidelines set two limitations for membrane
Figure 5.5 The distribution of

bending and membrane stresses
across the thickness of a component
under stress.
Midsurface
Total Bending Membrane

δm and membrane plus bending stress δm+b for duplex and super duplex materi-
als. According to these equations, the stress in the component subjected to risk
must be limited to a percentage of the specified minimum yield strength (SMYS).
In the oil and gas industry, SMYS is a material mechanical term or characteristic
commonly used to specify the minimum stress that can be supported without
permanent deformation. It is critical to understand that the value of SMYS
decreases with increased temperature, and so the value of SMYS must be
adjusted based on the temperature.
Membrane Stress δ m Limitation as Per DNV Guideline for Duplex and

Super Duplex Steels
δm < αm γ HISC SMYS 5 18
Membrane Plus Bending Stress δ m + b Limitation as Per DNV Guideline

for Duplex and Super Duplex Steels
δm + b < δm + b γ HISC SMYS 5 19
where,
δm: Membrane stress (psi, Pa);

δm + b: Membrane plus bending stress (psi, Pa);
γ HISC: In HISC material quality factor formula, coarse grain materials are rated as
85% and fine grain materials as 100%. A microstructure with fine spacing is
more resistant to HISC, as discussed previously. As a result, DNV RP allows
a higher HISC material factor of 100% for fine spacing compared to 85% for
coarse spacing in the stress criteria. With DNV’s model for mitigating HISC,
the material quality factor is taken into account in such a way that fine-grained
materials with fewer spaces are more resistant to HISC than coarse-grained
materials. More space between material grains actually facilitates the entry of
hydrogen atoms into the grains. There is an important point to remember that
materials with excessive grain spacing and grain flow perpendicular to the main
loading are nonconservative and cannot be analyzed with respect to the HISC
issue by the conservative model and method provided by DNV;
αm: Allowable SMYS factor for membrane stress (= 0.8; dimensionless);
αm + b: Allowable SMYS factor for membrane plus bending stress (= 0.8 or 0.9 or 1
depending on location). The allowable SMYS factor for membrane plus bending
stress with increased stress and weld toes within LRes is 0.8. As shown in
Figure 5.6, the weld toe is the point where the weld face joins the base metal.
The allowable SMYS factor for membrane and bending stress for high-stress
areas outside LRes is 0.9. It is equal to 0.9 and 1 for smooth areas within LRes
and outside LRes, respectively. Figure 5.7 display different values of the
Parent Weld
Weld metal zone
face
Parent
metal
Toe
HAZ
Weld
metal Fusion
Root Excess
line
weld metal
Figure 5.6 A buttweld joint that includes the toe.
δm+b<100% .γHISC.SMYS
Smooth sections without stress raisers
or welds outside of Lres
Stress raisers and welds within Lres
Lres
Stress raisers outside of Lres
Smooth sections within Lres
Figure 5.7 Values of SMYS factor for membrane plus bending stress from DNV standard.
allowable SMYS factor for membrane plus bending stress in accordance with
DNV standards.
The distance Lres from the weld centreline is the distance at which residual strain
and stress must be taken into account. In Eq. (5.20), Lres is the distance from the
weld where HISC poses a higher risk of attack.
Calculation of Distance from Weld Joint Where the Risk of HISC

Attack is Higher as Per DNV-RP-F112
LRes = 2 5 Rt 5 20
where:
R = Nominal pipe radius (mm or in.);

t = Pipe wall thickness (mm or in.).
Example 5.8 A 10 side entry subsea ball valve with a pressure rating of 7500 psi
will be installed at a depth of 2300 m. Through its pup piece, the valve is attached to
the connected subsea piping. The pup piece is, as its name implies, a piece of pipe
that is either integrated with the valve body or welded to it. In essence, the pup
piece is designed to keep the soft (nonmetallic) materials inside the valve away
from the heat generated during the welding process between the valve and the pip-
ing, which will reduce the possibility of the soft materials being melted. A pipe
connected to a valve has the same internal diameter and thickness as the pup piece
at the location where they are welded together. Figure 5.8 shows the dimensional
drawing of the pup piece. In accordance with DNV-RP-F112, calculate the distance
from the weld joint between the piping and the valve pup piece where the risk of
HISC attack is most likely.
Answer
The pipe size is 10 (254 mm), which is also called the nominal pipe size. The nom-
inal pipe radios are half of the pipe diameter, which is equal to 127 mm in this
example. According to the drawing, the pup piece at the connection point to
the pipe is 18.3 mm thick. The thickness value is taken from the figure on the right
where the weld termination details are described in accordance with ASME
B16.25. ASME B16.25 is a standard for the preparation of buttweld joints. Simi-
larly, the thickness of the pup piece and the attached piping is the same. Based
on the measurement of the radios of piping equal to 127 mm and its thickness
equal to 18.3 mm, the value of LRes indicates the distance from welding where
HISC is likely to occur based on Eq. (5.20) as follows:
LRes = 2 5 Rt = 2 5 127 × 18 3 = 2 5 × 2324 1 = 2 5 × 48 21 = 120 52 mm
120.52 mm is the distance from the weld at which HISC is more risky and less
stress is allowed in HISC analysis.
Detail A
- Acc. to ASME B 16.25 -
A 2.5°
°±
37.5
(18.3)
Ø236.4
Ø290
Ø273
1.6 ± 0.8
(Ø236.4)
(Ø273)
Valve side
R 50
14°
60
200
Figure 5.8 Drawing of the 10 subsea ball valve pup piece.
Example 5.9 Consider the 10 subsea ball valve in the previous example to be
made of forged duplex ASTM A182 F51 and operated at ambient temperature. In
terms of metallurgy, the ball valve body and the pup piece are made of
coarse grains. During the FEA, the maximum membrane stress is found to be
226.14 MPa on the transition piece within 120.52 mm of the weld joint to the pipe.
Similarly, the maximum bending stress is associated with the maximum mem-
brane stress of 46.2 MPa. In this case, what is the maximum allowable stress for
the valve material? Will the allowable stress be considered in the HISC analysis?
With the ultimate tensile stress and yield strength of duplex equal to 620 MPa and
450 MPa, respectively, can the valve fail due to HISC by considering the applied
loads in this case?
Answer
The first step is to determine the allowable stress of the forged duplex material in
ASTM A182 F51 based upon its tensile and yield strengths. Allowable stress is
defined as the maximum stress (tensile, compression, or bending) that is allowed
to apply to a component. As a general rule of thumb, as per different ASME codes
and standards, such as ASME B31.3, for process piping, the allowable stress can be
selected from 1/3 tensile stress or 2/3 yield stress, whichever is lower.
2 2
yield stress = × 450 = 300 MPa
3 3
1 1
tensile stress = × 620 = 206 7 MPa
3 3
Thus, the maximum allowable stress for duplex valve materials is 206.7 MPa. It
should be noted that the allowable stresses in HISC analysis based on the DNV
guideline standard is greater than the value of allowable stress equal to 206.7
MPa. For HISC analysis, the allowable stress values (membrane and membrane
plus bending) are calculated from Eqs. (5.18) and (5.19) as follows:
Allowable membrane stress for HISC according to DNV standard

= αm γ HISC SMYS = 0 8 × 0 85 × 450 = 306 MPa
Due to the coarse grain in the duplex material, the material quality factor γ HISC is
0.8. From the results of FEA, it is possible to determine the maximum membrane
stress to be 226.14 MPa < 306 MPa, which means the valve is safe against HISC
while the membrane stress is produced within it. Next, the valve must be checked
for its resistance to HISC during membrane plus bending stress production.
Allowable membrane plus bending stress for HISC according to DNV standard
= αm + b γ HISC SMYS = 0 8 × 0 85 × 450 = 306 MPa
Because the maximum membrane plus bending stress occurs around

LRes = 120.52 mm , where HISC attack is more likely, the allowable SMYS factor
for membrane plus bending stress αm + b is 0.8. Based on FEA calculation, the max-
imum membrane plus bending stress is calculated as 226.14 MPa (membrane
stress) + 46.22 MPa (bending stress) = 272.34 Mpa < 306 MPa, which means the
valve is safe against HISC while the membrane plus bending stresses are produced
within it.
The result of these calculations indicates that there is no risk of HISC occurrence
for the subsea valve with respect to the applied stress. However, we cannot con-
clude that the valve will operate safely under water since the valve must also be
checked for strain during HISC analysis according to the DNV guideline.
Figure 5.9 A component’s length Force Force

changes due to tensile and
compressive stress.
∆L ∆L
L
L
Force Force
Tensile strain Compressive strain
The second parameter that must be checked for HISC after stress is strain, which
relates to the mechanics of materials. Strain calculation is, in essence, the method
of determining the deformation of an object as a result of applied (tensile or com-
pression) stress, as depicted in Figure 5.9, or residual stress.
A component’s strain is defined as the ratio of the change in length to its original
form. Strain is calculated from Eq. (5.21). In fact, strain is a term used to describe
the amount of deformation per unit of dimension. It is a dimensionless quantity
without a unit of measurement. The amount of deformation and strain depends on
many factors, including the magnitude and duration of the stress as well as mate-
rial characteristics such as hardness and mechanical strength.
Strain Calculation
ΔL Change in length or elongation

= = 5 21
L Original length
FEA can be divided into two types: linear and nonlinear. There is a linear rela-
tionship between applied force and component displacement in linear analysis.
Structures and components that are subjected to this type of analysis are normally
those whose stresses remain within the linear elastic range that is maximum equal
to their SMYS. Figure 5.10 shows the stress–strain curve for ductile steel. Ductile
steel is able to go under a large amount of plastic deformation before fracturing.
A yield point, also known as SMYS, is a point of stress at which a small increase in
stress results in a large increase in strain such that the material continues to
deform despite no increase in load. At the yield point, the material will have per-
manent deformation.
Stress, σ
Ultimate strength
Elastic
deformation Fracture
Yield strength
Rise Plastic deformation
Run
Rise
Young’s modulus = Slope =
Run
Strain, ε
Figure 5.10 The stress–strain curve.
A large displacement is evaluated and considered by DNV through nonlinear

FEA using nonlinear strain criteria. Nonlinear strain criteria take into account
both elastic and plastic strains. Based on the DNV HISC guideline, residual strain
values are 0.15% for weld toes and 0.25% for areas except for weld toes. An
illustration of a buttweld joint with different weld areas is shown in Figure 5.7.
According to Table 5.4, the allowable maximum strain from all allowable loads
is calculated after allocating the residual strain values. Accordingly, the maximum
strain obtained through FEA should be lower than the maximum allowable stain
limit shown in Table 5.4.
Table 5.4 Maximum allowable strain limit according to DNV guideline.
Maximum allowable strain limit
Location in component Within Lres from weld Outside Lres from weld (%)
Outside 5% WT from any surface Min. (0.3%; 0.5% − res]) 0.3

Within 5% WT from surface 1.00% − res 1.00
0.6% − res 0.6
Example 5.10 The maximum measured principal strain for the 10 subsea ball
valve in the previous example was 0.0024980 at a distance of 100 mm from the weld
joint, and outside of 5% of the wall thickness from the surface. Does high strain
concentration cause the valve to fail?
Answer
The maximum strain happens at a distance of 100 mm from the weld that is within
the Lres = 120.52 calculated in Example 5.8. Moreover, the location of the maxi-
mum strain is outside 5% of the wall thickness from any surface and, hence,
the maximum allowable strain limit is selected from a minimum of 0.3% and a
maximum of 0.5% – res. As the highest amount of strain is 100 mm away from
the weld joint, it is not on the weld toe, and res = 0.25%. The maximum allowable
strain limit is calculated using Table 5.4:
Maximum allowable strain limit = Min 0 3 ; 0 5 − res
= Min 0 3 ; 0 5 − 0 25 = 0 25 = 0 0025
Due to the maximum strain value of 0.0024980 < 0.0025, the subsea valve will
not fail due to HISC and strain concentration.

5.1 How is carbon dioxide corrosion prevented in the oil and gas industry?
A By adding thickness to the facilities such as piping and valves as a cor-
rosion allowance
B Using corrosion-resistant alloys such as stainless steel as a form of clad-
ding or solid
C Injection of corrosion inhibitor into the corrosive fluid
D All answers are correct.
Answer
All answers are correct, so option D is the right choice.
5.2 In which case must corrosion allowance be considered for the indus-
trial valve?
A A carbon steel body valve is cladded with three millimeters of Inco-
nel 625.
B Martensitic stainless steel is used for the valve internals.
C Either carbon steel or low-alloy steel is used for the valve body.
D The valve body is constructed of austenitic stainless steel.
Answer
It is incorrect to choose option A because the carbon steel body has an Inco-
nel 625 layer that protects it from corrosion. Option B is also incorrect since
corrosion allowance is typically applied to the valve’s body rather than its
internals. The correct answer is Option C. Option D is incorrect because
there is no corrosion allowance needed on CRA materials such as austenitic
stainless steel.
5.3 On the outlet piping of a pressure vessel in a refinery that handles carbon
dioxide-containing corrosive gas at 57 C and 119 bar, a 10 valve is installed.
Keeping in mind that carbon dioxide is the only corrosive compound in the
service with a mole fraction of 0.018 and pH equal to 4 what is the correct
statement?
Assumptions:
• No corrosion inhibitor injection is performed;
• Facilities and components in the refinery, including valves, shall be
designed for 20 years of design life;
• The amount of water in the fluid service is low, less than 1% and the effect
of water cut and condensate are both 0.1.
A Carbon steel is a suitable material for the valve.
B A three-millimeter corrosion allowance is required.
C There is no chance of scale formation in this case.
D The carbon dioxide is deemed 100% active.
Answer
The first step is to calculate the basic CO2 corrosion rate in millimeters
per year.
The partial pressure of the CO2 is calculated as follows:
57 C is equivalent to 330 K
1710 1710
Log CR = 5 8− + 0 67 log PCO2 = 5 8 − + 0 67 log 2 14
base
T 330
= 5 8 − 5 182 + 0 22 = 0 838 CR base = 6 89 mm year
By considering the corrective parameters, it is possible to reduce the basic

corrosion rate, which is very high. The first step is to calculate the activity
of CO2 in the industrial valve.
14 14
Log a = 0 0031 − ×P= 0 0031 − × 119 = − 0 136
T 330
a = F System = 0 73
Option D is incorrect because the activity of carbon dioxide is 0.73 or 73%, so

it is not 100% active.
F CO2 = a × PCO2 = 0 73 × 2 14 bar = 1 56 bar
The glycol and corrosion inhibitor effects are equal to one because no glycol
or corrosion inhibitor is injected into the valve. The effect of water cut and
condensate are both assumed as 0.1.
Two other effects need to be calculated; one is the effect of scaling and the
other is the effect of pH.
2400 2400
Log F Scale = −0 6 × log F CO2 − 6 7 = −0 6 × log 1 56 − 6 7
T 330
= 7 27 + 0 193 − 6 7 = 0 763 F Scale = 100 763 = 5 79
The scaling effect exists if (F(S)) is less than one, which is not happening in
this case; therefore, there is no possibility of scale formation, and option C is
correct, and F(S) = 1 is utilized for corrosion rate calculations.
The next step is to calculate the effect of pH. To calculate the effect of pH
on the corrosion rate, the first step is to calculate the “pHsat” from Eqs. (5.10)
and (5.11).
1307
pHSat1 = 1 36 + − 0 17 × log F CO2
T
1307
= 1 36 + − 0 17 × log 1 56
330
= 1 36 + 3 96 − 0 033 = 5 29
pHSat2 = 5 4 − 0 66 × log F CO2 = 5 4 − 0 66 × log 1 56
= 5 4 − 0 127 = 5 27
The lowest pHsat value is selected for calculating the effect of pH; in this case,
it is 5.27. The pHsat value is higher than the actual pH; therefore, the effect of
pH is calculated from Eq. (5.12) as follows:
If pHSat > pHactual
Log F pH = 0 32 pHSat − pHactual log F pH

= 0 32 × 5 27 − 4
= 0 4064 F PH = 100 4064 = 2 55
The next step is to calculate the corrective corrosion rate as follows:
CR correct = CR base × F System × F Glycol × F Scale
× F Watercut × F Inhibitor × F Condensate × F PH
= 6 89 mm year × 0 73 × 1 × 1 × 0 1 × 1 × 0 1 × 2 55
= 0 128 mm year
Considering the design life of the valve equals 20 years, the final corrosion
rate is 0.128 × 20 = 2.56 mm Carbon steel with a corrosion allowance of
3 mm can be considered for this case. Thus, both options A and B, in addition
to option C, are correct. To sum up, option D is the only wrong statement.
5.4 What is one effect that could increase the basic carbon dioxide corrosion rate
calculated using Eq. (5.1)?
A Scaling
B Fugacity
C pH
D Corrosion inhibitor
Answer
Option C is correct. All other options either reduce the corrosion rate or do
not affect it at all.
5.5 Identify the correct statement about pitting corrosion mechanisms and
prevention.
A Pitting corrosion is a form of uniform corrosion.
B Pitting corrosion is only a type of internal corrosion caused by corrosive
fluids, such as seawater.
C The only way to prevent pitting corrosion is to select materials with a
higher PREN.
D The pitting corrosion process is not confined to austenitic stainless steel.
Answer
Solution A is incorrect since pitting corrosion occurs locally and is not uni-
form. Option B is wrong since pitting corrosion caused by corrosive marine
environments can occur externally. In addition, option C is not correct
because other approaches, such as testing for pitting corrosion as well as con-
trolling effective process parameters on pitting corrosion such as pH, temper-
ature, and chloride content, can prevent pitting corrosion. Accordingly,
Option D is the most appropriate response because carbon steels, nickel
alloys, and duplex stainless steels are all vulnerable to pitting corrosion,
as well as austenitic stainless steels.
5.6 Regarding the calculations in Example 5.6, select the pitting-resistant

material.
A SS347
B 25Cr super duplex
C Incoloy 800
D 22Cr duplex
Answer
The correct answer is Option B, 25Cr super duplex since its PREN is greater
than 40.
5.7 A welded valve is constructed of forged low-temperature carbon steel

(LTCS), material grade ASTM A350 LF2. The chemical composition percent-
age of this material grade can be found in Table 5.5. Which of the following
statement is true regarding the material’s weldability and carbon equivalent?
A As far as the chemical composition and calculation of carbon equivalent
are concerned, ASTM A350 LF2 is weldable.
B It is proposed that ASTM A350 LF2 be welded in this example in a sour
service containing hydrogen sulfide.
C Carbon equivalent is equal to 0.45.
D To achieve an acceptable level of weldability, the carbon content must be
reduced.
Answer
Carbon equivalent of A350LF2 is calculated as follows:
CE A350LF2 = C + + +
6 5 15
08 0 3 + 0 12 + 0 08 04+04
= 0 30 + + +
6 5 15
= 0 30 + 0 1333 + 0 1 + 0 0533 = 0 59
ASTM A350LF2 cannot be welded because its carbon equivalent exceeds

0.40, so option A is incorrect. As ASTM A350LF2 in this example should
not be welded in a sour service due to the high risk of cracking due to hydro-
gen sulfide present, option B is incorrect as well. Carbon equivalent equals
0.59, so Option C is incorrect. Choice D is the correct answer.
Table 5.5 Chemical compositions of the welded valve in A350LF2 material.
Carbon 0.30 Molybdenum 0.12

Manganese 0.8 Phosphorous 0.035
Silicon 0.30 Sulfur 0.040
Copper 0.40 Vanadium 0.08
Chromium 0.30 Columbium 0.02
Nickel 0.40
5.8 According to the DNV guideline for HISC evaluation, which statement is
correct?
A HISC analysis is covered in the guideline for components in the subsea
sector, as well as refineries and chemical plants.
B The stress analysis for HISC evaluation considers plastic deformation.
C Material types susceptible to risk, such as low-alloy steels, hard nickel
alloys, and duplex stainless steels, are all covered by the DNV criteria.
D The strain analysis considers both elastic and plastic conditions.
Answer
Option A is incorrect since HISC only applies to subsea oil and gas
components and is not relevant to refineries and petrochemical plants.
Additionally, option B is not correct since the DNV guideline for stress
analysis in HISC considers only linear and elastic deformation. Option
C is incorrect since DNV criteria only apply to duplex stainless steel.
The correct answer is D.
5.9 In this application, a subsea check valve in 2 size will be constructed from
duplex stainless steel and operated at a pressure of 10,000 psi at a depth of
2300 m. For this valve, two load scenarios are considered: operation and test-
ing. Operating and testing membrane stress values are respectively 172.92
MPa and 190 MPa. Membrane plus bending values for these two operating
and testing conditions are 231.69 MPa and 254.86 MPa, respectively. Based
on the stress analysis, which of the following statements is true regarding
the occurrence of HISC for the valve? (Note: Duplex has an ultimate tensile
strength of 620 MPa as well as a yield strength of 450 MPa.)
A The possibility of HISC occurrence in this case cannot be determined
based on the stress analysis because the material quality factor and the
exact location of the maximum load values are unknown.
B Membrane and bending loads are not large enough to cause HISC in
the valve.
C HISC can tolerate maximum stress of 306 MPa in a worst-case scenario.
D In the worst-case scenario, the maximum allowable stress for HISC
differs between membrane and membrane plus bending stress
conditions.
Answer
As the worst-case scenario for HISC occurrence is when the material con-
tains coarse grains, the value of the material quality factor should be 0.85,
the minimum value. The probability of HISC occurring is highest at the
worst-case scenario when the SMYS factor for both the membrane and
the membrane plus the bending are equal to 0.8. Based upon the worst-case
scenario, the allowable membrane stress, as well as membrane plus bend-

ing stress, are calculated using Eqs. (5.18) and (5.19) as follows:
Allowable membrane stress to prevent HISC at the worst − case scenario
= αm γ HISC SMYS = 0 8 × 0 85 × 450 = 306 MPa
Allowable membrane plus bending stress to prevent HISC at the worst
− case scenario = αm + b γ HISC SMYS = 0 8 × 0 85 × 450 = 306 MPa
Membrane stress during operation = 172 92 < 306 MPa No HISC risk
Membrane stress plus bending during operation = 231 69 < 306 MPa
No HISC risk
Membrane stress during testing = 190 < 306 MPa No HISC risk
Membrane stress plus bending during operation = 254 86 < 306 MPa
No HISC risk
As a result, even without all the data, by assuming the worst-case scenario,
HISC cannot happen in this case, so option B is correct, and option A is
incorrect. Option C is right since in the worst-case scenario, 306 MPa is
the maximum stress that is tolerated before HISC occurs. Option D is incor-
rect because the maximum allowable stress in a worst-case scenario is equal
to 306 MPa in both membrane and membrane plus bending stress condi-
tions. Thus, both options B and C are correct.
5.10 The maximum strain is 0.00178 for the subsea check valve in the previous
example. Based on strain value, which of the following statements is correct
regarding the risk of HISC?
A Due to the strain’s unknown location, the possibility of HISC occurrence
cannot be determined.
B There is no risk of HISC in this case.
C In the worst-case scenario, the maximum allowable strain is 0.0025.
D All options are incorrect.
Answer
According to Table 5.4, the maximum allowable strain in the worst-case
scenario is 0.0025, so option C is appropriate. This happens when the crit-
ical strain occurs in areas other than weld toes, which indicates the residual
strain εres value is 0.25%. Further, the principle strain must be located
within the Lres from weld and outside 5% of the wall thickness from the
surface. Considering that the maximum strain is 0.00178 and less than
the maximum allowable strain in the worst-case scenario, HISC cannot
occur and option B is also correct. Lastly, both options A and D are
incorrect.
Further Reading
American Society of Mechanical Engineers (ASME) (2017). Buttwelding ends. ASME
B16.25, New York, USA.
De Waard, C., Lotz, U., and Milliams, D.E. (1991). Predictive model for CO2 corrosion
engineering in wet natural gas pipelines. Corrosion 47 (12): 976–985.
Det Norske Veritas (DNV) RP-F112 (2008). Design of duplex stainless-steel subsea
equipment exposed to cathodic protection, Hovik, Norway.
Iranian Petroleum Standard (IPS) (2008). Engineering standard for corrosion
consideration in material selection. IPS-E-TP-740, Tehran, Iran.
Oxford: Elsevier.
Norsok (2005). CO2 corrosion rate calculation model. Lysaker: Norsok M-506, Rev.02.
Nustad, G. (2015). Tackling corrosive challenges in the oil and gas industry. Valve
World 20 (11): 56–60.
Sotoodeh, K. (2020). Requirement and calculation of corrosion allowance for piping
and valves in oil and gas industry. Journal of Bio and Tribo Corrosion 6 (21):
https://doi.org/10.1007/s40735-019-0319-4.
Sotoodeh, K. (2020). A review and analysis of industrial valve material failures due to
corrosion and proposals for prevention measures based on industrial experiences in
the offshore sector of oil and gas industry. Journal of Failure Analysis and Prevention
https://doi.org/10.1007/s11668-020-01064-9.
Sotoodeh, K. (2021). A Practical Guide to Piping and Valves for the Oil and Gas Industry,
Sotoodeh, K. (2021). Safety and Reliability Improvement of Valves and Actuators for the
Offshore Oil and Gas Industry Through Optimized Design. University of Stavanger.
Doctor Philosophiae (Dr. Philos) UiS no.573. Faculty of science and technology.
ISBN:978-82-7644-987-7.
Sotoodeh, K. (2021). HISC analysis for valves in the subsea oil and gas industry. Safety
in Extreme Environments 3: 1–8. https://doi.org/10.1007/s42797-021-00030-4.
Sotoodeh, K. (2022). Piping Engineering: Preventing Fugitive Emission in the Oil and Gas
Industry, 1e. New York: Wiley.
Sotoodeh, K. (2022). Case Studies of Material Corrosion Prevention for Oil and Gas
Valves, 1e. Austin, TX: Elsevier (Gulf Professional Publishing).
185
Noise
6.1 Introduction to Sound
In physics, sound is the result of vibrations and is accompanied by acoustic waves

in liquids, gases, and solids. It is true that when an object vibrates, it imparts
energy to the particles around it, causing them to move and this movement is
known as sound waves. The waves caused by vibrations of molecules follow sine
functions characterized by their amplitude and wavelength or frequency. The
two most significant characteristics of a sound are its frequency and intensity
(amplitude or volume or loudness). A sound can be described by its low or high
frequency. The frequency is determined by the speed at which particles vibrate in
the medium, which is defined as the number of waves emitted in a given amount
of time. Frequency is the number of cycles per second measured in Hertz (Hz)
units. For example, a frequency of 1000 Hz means that the waveform is 1000
cycles/s. The human hearing range is between 20 and 20,000 Hz. Each cycle
represents one sequence of waves where the intensity changes from zero to
the highest value, then to the lowest value, and then back to zero, as shown
in Figure 6.1. The length of a wave or wavelength can be measured from one
point on the waveform to the exact point at which the next cycle begins (see
Figure 6.1). Frequency and wavelength are inversely proportional to each other;
therefore, a wave with the greatest frequency will have the shortest wavelength,
and twice the frequency will have half the wavelength. Figure 6.2 depicts sound
waves of equal amplitude (intensity) and increasing frequency from top to bot-
tom. The amplitude of a sound wave is a measure of its height, which is deter-
mined by the distance between its peaks and troughs. In general, the larger the
amplitude of the wave, the louder the sound, indicating that it contains more
energy and may travel further at a lower speed. Pitch or frequency is shown
by the spacing of displayed waves in such a way that the closer the waves are
to one another, the higher the frequency.
186 6 Noise
Peaks
Wavelength
Amplitude One complete

Troughs Cycle
Figure 6.1 Illustration of amplitude, wavelength, and cycle in a sound wave.
Figure 6.2 From top to bottom, sound waves are equal in amplitude and increasing
in frequency.
The temperature, elasticity, and density of the medium determine the speed of
sound. Higher temperatures and moisture cause sound to travel faster, while lower
temperatures and dry air cause sound to travel slower. In any medium, sound can
travel, although the greatest speed occurs in solids and the slowest in gases.
6.2 Introduction to Noise
Generally, the term “noise” is used to refer to a loud, unpleasant, and disruptive
sound that produces undesirable psychological effects on an individual such as
depression and irritation by interfering with their activities such as work, rest,
sleep, etc., as illustrated in Figure 6.3. From a physics point of view, there is no
difference between noise and desired sound. This is because both pertain to vibra-
tions that can move through a medium such as air or water. The difference occurs
when the brain receives and interprets the sound. Noise pollution is projected to
6.2 Introduction to Noise 187
Figure 6.3 Illustrations of the generated noise and its effects on the human brain.
become the third most significant threat to the human environment after air pol-
lution and water pollution.
Sound power level and sound pressure level are two terms which are commonly
confused with one another when discussing acoustic energy. The sound power
level (SWL, LW) is the level of energy emitted by the sound source. The value is
an absolute value that is unaffected by the environment, and it is independent
of distance. The sound pressure level (SPL, Lp) is what ears hear and what sound
meters measure. This parameter is influenced not only by the strength of the
source but also by the surroundings and the distance from the source to the
receiver. As a general rule of thumb, doubling the distance is equivalent to a
6-decibel (dB) reduction in noise levels. While the two parameters differ, they
share the same unit, which is the dB.
Example 6.1 When a worker drills the ground, the sound power level produced
is 100 dB. There are two individuals standing far away from the worker; one is 100 m
away and the other is 50 m away. What is the value of the sound power level at the
locations of these two individuals? If the sound pressure level at 50 m from the
worker is 58 dB, what is the sound pressure level at 100 m where the other person
is standing?
Answer
At the location of these two individuals, the sound power level (SWL) is equal to
100 dB. As doubling the distance reduces the sound pressure level (SPL) by 6 dB,
the SPL at 100 m, where the other person stands, is equal to 58 − 6 = 52 dB.
188 6 Noise
Sound pressure Sound pressure level

Jet take-off
(25 m distance μPa 140 dB
100,000,000
130
Firecrackers
120
10,000,000 Pneumatic
Rock chipper
group
110
100
1,000,000 Noisy workplace
90
Average street traffic 80

1,00,000
70
Business office
60
10,000
Conversational speech 50
40 Living room
Library 1000
30
Bedroom
20
Wood
100
10
20 0
Figure 6.4 A plot of sound pressure and sound pressure level values for different locations.
A plot of sound pressure and sound pressure values for different locations is
shown in Figure 6.4. Furthermore, Table 6.1 provides a summary of the relative
noise levels.
Equation 6.1 can be used to calculate the relationship between sound pressure
level (SPL) and pressure level. As a general rule of thumb, an increase of 6 dB dou-
bles the level of pressure, whereas an increase of 20 dB tens it.
Table 6.1 Relative noise levels.
130 dB Jet aircraft on take-off

120 dB Threshold of feeling
110 dB Train
100 dB Loud highway
90 dB Loud truck
80 dB Plant site
70 dB Vacuum cleaner
60 dB Conversation
50 dB Offices
Relationship Between SPL and Pressure Level
2
Pressure level Pa
Sound pressure level SPL = 20 log10 61
20 − 6
A dB is a logarithmic unit of measurement of sound intensity. This unit is also

used for communications, electronics, and signals. Let us consider two sound-
generating devices. The first produces a sound with a power of P1 and the other
produces a louder sound with a power of P2. Assuming that all other parameters
are equal, the dB difference between these two is equal to 10 log(P1/P2). In the case
where the second device produces twice as much power as the first, the difference
is 10 log 2 = 3 dB. Decibels may be used to determine the sound level of a single
source rather than a ratio. As a result, a reference must be chosen, usually an air
sample with a very low sound power level of 20 μPa.
6.3 Noise in Industrial Valves
The safety and instrument engineers analyze the noise generated by various types
of valves to minimize its impact in diverse sectors of the oil and gas industry, such
as offshore, refineries, and petrochemical plants. The high noise and vibration
levels caused by pressure-reducing devices such as pressure safety valves (PSVs)
and control valves during operation can increase stress levels in the piping system
and cause damage. As a result, it is essential to design piping systems to withstand
noise, vibration, and acoustic fatigue from a safety perspective. In addition, as a
result of the noise and vibration generated at the same time, wear, pits, and
190 6 Noise
fractures of valve connections occur which significantly impair the performance

of the valve. The Occupational Safety and Health Administration (OSHA)
has focused its attention on the noise produced by valves in the field of control
equipment. Noise in industrial valves can be attributed to two main sources:
mechanical work, movement, and vibration within the valves, as well as fluids
moving through the valves.
There are two negative effects of noise: acoustic fatigue and hearing damage.
The noise, however, may be beneficial as a sign of an abnormal condition within
the valve. Due to the negative impacts of noise, measuring the noise level gener-
ated by industrial valves is very important and valuable. The experience in gas pro-
duction, petrochemical and other industries has shown that acoustic energy in
high-capacity gas pressure reduction systems, such as pressure safety and control
valves, can lead to piping fatigue failure after a few hours of operation. Noise expo-
sure must be limited in accordance with safety standards, such as occupational
health standards. It is very important to control the noise level in control valves
as well as PSVs. Although control valves and PSVs are known as the main
noise-producing valves, check valves such as dual plate and swing check valves
are also capable of producing noise due to disk slamming and chattering.
6.3.1 Mechanical Noise and Vibration

Mechanical noises and vibrations are those caused by the movement and vibration
of valve components. There are two possible reasons for mechanical noise and
vibration. Most valve internals react to turbulent flow flowing through them,
resulting in mechanical noise. Valves that undergo rapid changes in flow or
high-pressure drops can also generate noise and vibration in their components.
The second vibration condition is caused by mechanical loads imposed by the con-
nected piping or by the operation of the valve itself, which is a result of the internal
movement of the valve. The noise generated by vibration and mechanical work of
valves is typically unpredictable, which is considered the second concern of proc-
ess engineers. Turbulent flow can cause vibrations in some valve components in
contact with nearby surfaces, leading to an intensification of noise and vibration as
a result. The mitigation of this type of noise involves improving the design of the
valve, such as changing the valve internals, increasing the stem diameter, and bal-
ancing the applied forces. The next subsection describes the noise caused by the
fluid inside the piping system.
6.3.2 Fluid Noise

Fluid noise is primarily associated with control valves. A possibility of noise
generation exists in situations where mass flow rate and pressure drop are high.
Control valve noises are typically associated with cavitation and high-pressure
drop gas applications. Generally speaking, there are two types of fluid service
noise: aerodynamic caused by compressible fluids and hydrodynamic caused by
non-compressible fluids. It is undeniable that liquids are less compressible than
gases. However, the compressibility effect of liquids cannot be ignored if the liquid
pressure is high and its pressure drop is large. Correct trim selection and/or
appropriate insulation can reduce the effect of fluid noise in control valves. The
calculation of noise in control valves caused by fluids is done in accordance with
the International Electrotechnical Commission (IEC) standard.
6.3.2.1 Aerodynamic Noise

Aerodynamic noise is the result of the fluid converting its mechanical energy into
acoustic energy as it passes through the restriction in the valve. It represents the
most severe form of noise resulting from a high-speed or unsteady (turbulent) flow
of fluid. As an example, this type of noise can be generated downstream of vena
contracta, which is the point in a fluid stream where the flow diameter is the smal-
lest, and fluid velocity is at its greatest. By utilizing low-noise trim valves and silen-
cers as well as increasing the thickness of the pipe walls, this type of noise can be
mitigated.
6.3.2.2 Hydrodynamic Noise

Among hydrodynamic noises are those generated by liquid flow, cavitation, and
flashing. In Chapter 3, cavitation and flashing are discussed in more detail. The
most common cause of hydrodynamic noise is cavitation, which occurs when part
of a liquid undergoes two changes of state. It is important to recognize that when
the liquid pressure falls below the vapor pressure in cavitation, the fluid velocity
increases considerably. This increases noise and vibration, as discussed later in this
chapter.
6.3.3 Noise Control Strategies

In general, there are six kinds of measures used to mitigate or prevent noise in pip-
ing systems, including industrial valves, which can be summed up as follows:
• Selection and installation of noise-reducing valves: This method is prima-

rily used for noise control. For instance, several types of valves can be used to
control fluid flow, including globe, axial, butterfly, V-notch ball, and angle globe
valves. Among all these valves, axial and globe valves can offer excellent noise
control levels. It is therefore recommended to choose a globe or axial valve over a
butterfly, V-notch, or other option when noise is a concern.
• Process optimization: Noise levels are strongly influenced by operating con-
ditions. As described below, noise levels can be reduced by reducing the system
192 6 Noise
pressure, differential pressure, or increasing the size or reducing fluid velocity at

the trim exit.
• Modification of valve design: Many different types of low-noise valve trims
(closure members) are available that create multistep pressure reduction in
valves.
• Piping insulation: The noise emissions from the piping system, including
valves, are kept within the noise limit requirements by using acoustic piping
insulation.
• Using thicker pipe: When choosing a thicker pipe with a higher schedule
number, noise and vibration can be reduced.
• Improving the piping connection integrity: The stress on the pipe is
reduced by minimizing the number of pipe fittings and attachments that pro-
duce high stress concentrations at the connection. It can be accomplished by
reducing the number of small bore branch connections (e.g. vents, low point
drains), making branch connections with fittings that ensure a smooth transi-
tion from one branch to the next (e.g. by using tees rather than olets), ensuring
that the header seam does not cross the connection weld line, avoiding pipe-to-
pipe connections (also referred to as stub-in connections), etc.
• Using noise reduction components: When installed after or downstream of
control valves, silencers can be very effective in reducing noise generated by
these valves. A silencer is typically installed in a horizontal piping system
and has a preferred flow direction. Turbulent and high velocity fluid exiting
the control valve expands in the silencer, and the silencer absorbs a significant
proportion of the noise generated.
6.4 Noise Calculations for Pipes and Valves
The noise analysis covered in this section comprises four types; the first is acoustic
fatigue analysis, which includes the calculation of SWL and Mach number for all
types of valves, including control and pressure relief valves. It is the second part
that deals with control valves and the types of fluid noises produced by these
valves, namely hydrodynamic and aerodynamic noises. The final set of calcula-
tions is focused on the noise associated with PSVs according to the method out-
lined in API 521.
6.4.1 Acoustic Fatigue Analysis

The method outlined in this chapter to measure acoustic fatigue is applicable to all
types of valves, both control valves and pressure relief valves.
Figure 6.5 Schematic of valve and

p1 Direction of flow p2
piping system for acoustic fatigue
T1 T2
calculations.
PWL
PWL(x)
6.4.1.1 Sound Power Level Calculations

The American Society of Mechanical Engineers (ASME) publication PVP-Vol. 328,
flow-induced vibration, ASME 1996, by Eisinger may be used to evaluate acoustic
fatigue. The sound power level (SWL or PWL) from the source of noise and
vibration, which is typically a valve or orifice, is calculated according to
Eq. (6.2) (see Figure 6.5).
Calculation of Acoustic Fatigue at the Source of Noise
36 2
P1 − P 2 W T 1 + 273
PWL = 10 log × × 1 2 + 126 1 + SFF
P1 3600 m
62
where:
PWL: the sound power level in dB;

P1: the upstream pressure of the valve (pressure let down source) in bar;
P2: the downstream pressure of the valve (pressure let down source) in bar;
T1: the upstream temperature of the valve in C;
W: Gas or liquid flow rate in kg/h;
M: Molecular weight;
SFF: Sonic flow correction factor. In the presence of sonic conditions, Mach num-
ber is greater than one, so SFF = 6. In the absence of sonic conditions, SFF = 0.
Mach number is discussed further in this chapter.
For valves equipped with low-noise trim, as explained further in this chapter,
Eq. (6.2) is not applicable. As an alternative, the manufacturer of low-noise trim
valves can provide the value of PWL.
The next step is to calculate the attenuation of acoustic fatigue or vibration at a
distance “x” from the valve that is producing the noise using Eq. (6.3). The design
calculations suggest that a distance of 50 times the inside diameter of the valve or
194 6 Noise
the pipe connected to the valve can reduce the PWL by 3 dB in comparison to
the noise produced within the valve.
Calculation of Acoustic Fatigue at a Distance of “x” from the Source of

the Noise
0 06x
PWL X = PWL − 63
D
where:
PWL(X): Level of sound power in dB at a distance of “x” from the source of the
noise (valve);
x: Distance from the valve;
D: Internal diameter of the valve or pipe connected to the valve.
Whenever more than one source generates noise, the noise should be added at
the pipe junctions where the piping from the sources meets as per Eq. (6.4).
Calculation of Acoustic Fatigue at a Distance of “x” from the Source of

the Noise by Considering Multiple Sources of Noise
PWLX = 10 log 10PWL1 10

+ 10PWL2 10
+ 64
where:
PWL1,2,etc.: The power level for each source at a distance of “x” in the piping.
Equation (6.5) is used to calculate PWL’s acceptance level.
Allowable Level of Acoustic Fatigue
Di
PWLA = 173 6 − 0 125 × 65
t
where:
PWLA: Accepted Sound level (dBA);

Di: Inside diameter of the pipe (mm);
t: Thickness of the pipe (mm).
Note 1
Some major oil and gas end users, such as British Petroleum (BP), may set stric-
ter limits on the allowable noise levels than those determined by Eq. (6.5),
which is equal to 155 dB. So even if the calculated value of PWL is less than
x but greater than 155 dB, the design of the valve and connected pipe may
not be reliable regarding noise generation. In this case, it is critical to know
LOF, or the likelihood of failure. The design of piping and valves is safe in such
a case as long as the LOF does not exceed 0.5, according to the BP noise
guidelines.
Note 2
The fact that valves are not the only components that can cause noise is impor-
tant to keep in mind. The noise intensity can also be increased by piping com-
ponents such as bends, tees, and reducers. A discussion of methods for
calculating noise levels produced by components other than valves does not
fit the scope of this chapter.
Note 3
The noise level associated with a PSV can be increased by 5 dB under condi-
tions where an improvement in design is applied to the downstream of the
PSV as follows:
• In piping systems, olets larger than 2 should be replaced with tees made
according to ASME B16.9. Both olets and tees are branching fittings.
• Pipe-to-pipe connections should be avoided when taking branches from the
header. In the event that it is not possible to avoid pipe-to-pipe connections,
an attachment plate or reinforcement plate should be positioned around the
circumference of the pipe.
The design improvement is commonly required for a distance corresponding
to the calculated sound power in excess of the accepted limit, as calculated
according to Eq. (6.5). In the event that the accepted sound level still cannot
196 6 Noise
be attained despite the above-mentioned design improvements, additional

options such as selecting a valve with low-noise trim (applicable to control
valves), evaluating the use of multiple valves in parallel, or reducing the pres-
sure steps can be considered. Other approaches to mitigate noise impacts
include increasing the thickness of the piping walls, reducing the flow capacity,
and applying insulation around pipes.
Example 6.2 The outlet of a pressure relief valve is connected to an 8 pipe with
a standard (STD) wall thickness. Upstream and downstream pressures at the safety
valve are 30 bara and 1 bara, respectively, while the temperature at the valve inlet
and outlet is 39.85 C. The mass flow rate of the gas through the valve is 20, 000 kg/h,
and the molecular weight of the gas is 25. Assuming there is no risk of sonic flow,
is there a risk of acoustic fatigue in this instance?
Answer
The provided process data for the valve and outlet piping are summarized as
follows:
P1: the upstream pressure of the relief valve = 30 bara

P2: the downstream pressure of the relief valve = 1 bara
T1: the upstream temperature of the valve = 39.85 C
W: Gas or liquid flow rate = 20,000 kg/h
M: Molecular weight = 25
SFF = 0
Using Eq. (6.2):

36 2
P1 − P 2 W T 1 + 273
PWL = 10 log × × 1 2 + 126 1 + SFF
P1 3600 m
36 2
30 − 1 20,000 39 85 + 273
= 10 log × × 1 2 + 126 1
30 3600 25
= 10 log 0 8851 × 30 86 × 20 74 + 126 1 = 153 63 dB

Di
PWLA = 173 6 − 0 125 ×
t
Figure 6.6 A schematic of a pipe showing the Outside diameter

internal and external diameters, as well as the wall
thickness.
Inside diameter
Wire diameter (thickness)
In order to determine the PWL’s acceptance level, we should know the internal
diameter and thickness of the outlet piping. As per ASME B36.10/19, standards
for carbon, alloy, and stainless steel pipes, an 8 pipe with a standard (STD) thick-
ness has an outside diameter of 8.625 and a thickness of 0.322 . Equation (6.6)
gives the relationship between the pipe’s internal and external diameters as well
as its wall thickness, as shown in Figure 6.6.
Relationship Between the Internal and External Diameters of a Pipe,

As Well As Its Wall Thickness
ID = OD − 2 × t 66
where:
ID: pipe internal diameter (in., mm);

OD: pipe outside diameter (in., mm);
t: Pipe wall thickness (in., mm);
t = 0.322 and OD = 8.625 ID = 8.625 − 2 × 0.322 = 7.981 .
7 981
PWLA = 173 6 − 0 125 × = 170 50 dB
0 322
The noise level produced by the safety relief valve is 153.63 dB, which is lower
than both 155 and 170.50 dB. Thus, there is no risk of acoustic fatigue in the
relief valve.
198 6 Noise
6.4.1.2 Mach Number

In order to avoid vibration and noise in control and PSVs, fluid velocity must be
considered an extremely crucial parameter. By reducing the valve trim exit veloc-
ity, it is logical to achieve low noise and vibration. It has been a practice in the valve
industry to use a Mach number at the valve outlet to measure the velocity limita-
tion of the fluid. In fluid mechanics, the Mach number is defined as the ratio of
fluid velocity to sound velocity at a given temperature. The Mach number was
named after Austrian physicist Ernst Mach (1838–1916). The Mach Number is
an extremely crucial number in valves for reducing not only noise and vibration
but also erosion damage. Most valve manufacturers recommend a maximum
Mach number of 0.3 to avoid noise and vibration issues. Reducing the Mach num-
ber can be achieved by increasing the valve size or using a silencer. In the case of
PSVs which have a high velocity at their outlet, an expander should be installed
close to the PSVs in order to increase the diameter of the line and limit the velocity
and Mach number. Generally, the Mach number is taken into account for the out-
let of valves according to Eq. (6.7).
Calculation of the Mach Number at the Valve’s Outlet
V outlet
M outlet = 67
C
where:
Moutlet: Outlet Mach number;

Voutlet: Fluid velocity at the trim exit or valve outlet velocity (m/s);
C: Sonic or sound velocity (m/s) calculated using Eq. (6.8) for gas services.
Calculation of Sonic Velocity in a Gas
γRT
C= 68
M
where:
C: Sonic velocity (m/s);

γ: Gas-specific heat ratio (J/(kg K));
R: Universal gas constant = 8315 J/(kg K);
M: Gas molecular weight (g/mol);
T: Gas temperature (K).
Example 6.3 Calculate the speed of sound in air at 0 C and 1 bar absolute
pressure. The specific heat ratio of air is 1.4 J/(kg K) and the air molecular
weight is 28.97 g/mol. Calculate the sound velocity if the air temperature exceeds
to 20 C.
Answer
Speed of sound in air at
0 C 273 15 K = C1 = γ R T1 M = 1 4 × 8315 × 273 15 28 97 = 331 3 m s
Speed of sound in air at
20 C 293 15 K = C 2 = γ R T2 M = 1 4 × 8315 × 293 15 28 97 = 343 23m s
It is possible to calculate the speed of sound in gases, liquids, and solids

using Eq. (6.9).
Calculation of Sonic Velocity in Gases, Fluids, and Solids
K
C= 69
ρ
where:
C: Sonic velocity (m/s);

K: Bulk modulus of elasticity (Pa);
ρ: Density of gas, fluid or solid (kg/m3).
The bulk modulus of elasticity is a measure of an object’s resistance to being

deformed elastically within its volume. Table 6.2 provides density and bulk
modulus of elasticity values for four substances: water, oil, ethylene alcohol and
mercury at 1 bar and 0 C.
Table 6.2 The density and bulk modulus of elasticity of four substances:
water, oil, ethylene alcohol, and mercury at 0 C and 1 bar.
Bulk modulus of elasticity (K) Density

Substance 109 Pa ρ (kg/m3)
Water 2.15 999.8

Oil 1.35 920
Ethylene alcohol 1.06 810
Mercury 28.5 13,595
200 6 Noise
Example 6.4 Calculate the speed of sound in water at 0 C.

Answer
K 2 15 × 109
C= = = 1466 m s
ρ 999 8
The fluid velocity at the trim exit or valve outlet velocity is calculated according
to Eq. (6.10):
Fluid Velocity at the Trim Exist or Valve Outlet Velocity
w w
V outlet = = 2 6 10
Aρ πr ρ
where:
Voutlet: Fluid velocity at the trim exit or valve outlet velocity (m/s);
w: Mass flow rate (kg/m3);
ρ: Density of gas or liquid (kg/m3);
A: A valve outlet area equal to the diameter of the outlet pipe connected to the
valve (m2);
r: A valve outlet radius equal to the internal radius of the outlet pipe connected to
the valve (m).
The flow velocity head at the outlet of the valve is also calculated by
using Eq. (6.11):
Flow Velocity Head at the Valve’s Outlet
P = 1 2 ρV 2outlet 6 11
where:
P: Valve exit velocity head or pressure (Pa);

ρ: Density of gas or liquid (kg/m3);
Voutlet: Fluid velocity at the trim exit or valve outlet velocity (m/s).
The density of a gas is defined as the mass of the gas divided by its volume, as
per Eq. (6.12).
Basic Gas Density Calculation

mg
ρg = 6 12
Vg
where:
ρg : Density of gas (kg/m3);

mg : Mass of the gas (kg);
Vg : Volume of the gas (m3).
A1-pound mole gas has a weight of mg, which is equal to the molecular weight of
the gas, mg. For a 3-lb mole gas, the gas weight, mg, would be triple the molecular
weight, mg, of the gas. As a general rule, the gas weight is equal to the number of
moles multiplied by the molecular weight of the gas, as per Eq. (6.13). The next
step is to convert the volume of the gas to other properties, such as pressure
and temperature according to the gas law, using Eq. (6.14).
Calculation of Number of Moles in a Gas

m
n= 6 13
M
where:
n: Number of lb-mol;
m: Mass of the gas (lb or g);
M: Molecular weight of the gas (lb mol or g mol).
Gas Law Calculation

m
PV = ZnRT = Z RT
M
6 14
V ZRT m MP MP
= = ρ=
m MP V ZRT ZRT
ρ: Density of gas (kg/m3)

V: Volume of the gas (m3);
Z: Gas compressibility factor (dimensionless);
R: Gas universal constant;
P: Gas pressure (Pa);
T: Gas temperature (K)
Based on Eq. (6.7), since the Mach number is measured at the outlet of the valve,
this equation for calculating gas density can be modified in the following manner:
202 6 Noise
Calculation of Gas Density Based on Pressure and Temperature
MP2
ρ= 6 15
ZRT2
Using Eqs. (6.7), (6.8), (6.10), and (6.15) together, the calculation of the Mach
number at the valve outlet can be modified as follows:
V outlet
M outlet = 67
C
γRT
C= 68
M
w w
V outlet = = 2 6 10
Aρ πr ρ
MP2
ρ= 6 15
ZRT2
w ZRT2 M 4Z R w T2
M= × × = × × 6 16
πr 2 MP2 γRT π P2 D2i γ m
In the case of an ideal gas with Z = 1, and R = 8315, the calculation of the Mach
number is simplified as shown by Eq. (6.17).
Calculation of Mach Number for Ideal Gases Based on the Process

Parameters
W T2
M = 116 × × 6 17
P2 D2i γ m
where:
W: Flow rate of gas and liquid (kg/s);
P2: Downstream pressure (Pa);
Di: Outlet pipe inside diameter (m);
T2: Downstream temperature (K);
γ: Ratio of specific heat (dimensionless);
m: Gas molecular weight (g/mol).
Example 6.5 Determine the Mach number for the gas service passing
through the safety valve in Example 6.2. The ratio of specific heat is 1.21 and
the gas is ideal.
Answer
P2: the downstream pressure of the relief value = 1 bara = 105 Pa

T1: the upstream temperature of the valve = 39.85 C = 39.85 + 273.15
= 313 K
W: Gas or liquid flow rate = 20,000 kg/h = 5.56 kg/s
γ: Ratio of specific heat = 1.21
Diameter = Di = ID = 8 625 − 2 × 0 322 = 7 981 = 0 2027 m
z factor = 1
By using Eq. (6.17) M = 116 × W P2 D2i × T 2 γ m = 116 × 5 56
5 2
10 × 0 2027 × 313 1 2 × 25 = 0 1570 × 3 22 = 0 50
6.4.2 Noise in Control Valves

As discussed in Chapter 3, noise may be generated as a result of cavitation in con-
trol valves. Reducing pressure with valves, in particular control valves, can cause
unwanted and bothersome noises. As a matter of fact, control valves often generate
a lot of noise and contribute significantly to the overall noise level at a plant. Most
control system and valve designers specify a maximum noise level of 90 dB in
order to comply with noise standards set by OSHA, the United States Department
of Labour’s Occupational Safety and Health Administration. When calculating
control valve noise, fluid noise is the most critical factor taken into account,
including aerodynamic noise and hydrodynamic noise.
6.4.2.1 Aerodynamic Noise in Control Valves

According to the earlier discussion, aerodynamic noise is caused by the conversion
of mechanical energy associated with the turbulent flow into acoustic energy
through valve restrictions. Many oil and gas companies use IEC 60534-8:3: Indus-
trial Process Control Valves – Part 8 – Noise Considerations – Section 3 – Control
Valve Aerodynamic Noise Prediction Method. It is worth mentioning that the para-
meters that are listed in Table 6.3 are those that are used in the aerodynamic noise
prediction methodology based on the aforementioned standard. A sequential
method is described in this section for determining the aerodynamic noise from
control valves using a sequential model. It should be noted that, although this
method is mainly associated with regime I, it also gives an overview of how to cal-
culate this type of noise in general. In the next paragraph, a variety of flow regimes
based on IEC 60534-8:3 will be briefly described.
204 6 Noise
Table 6.3 Nomenclature in IEC 60534-8:3.
A: Flow area P: Pressure

Cv: Valve flow coefficient pa: Actual pressure outside pipe
C: Gas sound speed ps: Standard pressure outside pipe
(1 Atmosphere)
Cp: Sound speed in piping R: Universal gas constant
dp: Outlet pipe inner diameter T: Gas temperature
dv: Outlet valve inner diameter TL: Pipe transition loss
dvc: Trim jet vena contracta diameter TLfr: Pipe transition loss at ring frequency
Fd: Valve style modifier tp: Pipe wall thickness
FL: Pressure recovery coefficient U: Velocity
f: Sound frequency Wa: Sound power
fo: Acoustic-structural coincident Wm: Stream power of mass flow
frequency
fp: Flow peak frequency γ: Ratio of specific heat
fr: Pipe ring frequency η: Acoustical efficiency factor
Le: Expander and pipe flow noise sound ρ: Gas density
pressure level, A-weighted and 1 m
from the pipe wall
Lg: Pipe Mach number correction factor 1: Upstream of valve or vena contracta
LpAe: A weighted sound pressure level 2: Downstream of valve or vena contracta
m Mass flow rate e: Expander
LpAe: A-weighted sound-pressure level, v: Valve outlet
1 m from the pipe wall on the outlet side
Lpi: Pipe internal sound-pressure level vc: Vena contracta
M: Mach number I: Regime I
Mvc: Mach number at vena contracta
Mw: Molecular weight of gas
In the IEC standard, there are five flow regimes that are defined based on process
parameters such as inlet pressure, downstream pressure, fluid physical data, and
valve pressure recovery factor. These flow regimes are as follows:
• Regime I: Subsonic
• Regime II: Sonic with turbulent flow mixing
• Regime III: No compression but with flow shear mechanism
• Regime IV: Sock cell turbulent flow interaction
• Regime V: Constant acoustical efficiency that produces maximum noise
According to the standard, the methodology consists of seven basic steps as

follows:
1) Calculation of total mechanical stream power Wm at the vena contracta or fluid

restriction point. In fact, the aerodynamic noise is caused by valves inside and
downstream of the valve.
2) If the total fluid power at the vena contracta is calculated, then the fraction that
accounts for noise power may be determined according to the acoustic effi-
ciency factor η.
3) In this step, the acoustic power ratio rw, also known as the ratio between the
sound power in the downstream pipe and the sound power inside the valve, is
determined.
4) Calculation of the flow peak frequency fp for the internal noise inside the pipe.
5) Calculation of the internal sound pressure level Lpi.
6) Calculation of the transmission loss, TL.
7) Calculation of the sound pressure level LpAe.1 m 1 m away from the pipe on the
outlet side.
For the purpose of implementing these steps, the first step is to calculate the
noise source magnitude Wm according to Eq. (6.18) as follows:
Noise Source Magnitude
U 2vc × m
Wm = 6 18
2
The acoustic efficiency factor η can be calculated as per Eq. (6.19) in the follow-
ing manner:
Calculation of Acoustic Efficiency Factor
η = 0 0001 × M 3vc6 6 19
It is important to note that the peak noise frequency depends on the velocity at
vena contracta and the jet vena contracta diameter. Jet vena contracta diameter is a
function of jet pressure recovery and valve style modifier. The next step is to cal-
culate noise peak frequency fp according to Eq. (6.20) as follows:
Calculation of Noise Peak Frequency
0 2U vc
fp = 6 20
dvc
206 6 Noise
Table 6.4 Valve noise propagation values.
Valve type and style rw
Globe valve 0.25

Rotary globe (varimax) valve 0.25
Eccentric rotary plug valve 0.25
Ball valve 0.5
Butterfly valve 0.5
In the next step, we need to calculate the noise propagation through the valve
downstream of the valve that is designated with parameter rw. The ratio of noise
propagation through the valve varies depending on the type and style of valve as
indicated in Table 6.4.
Using Eq. (6.21), it is possible to determine the sound pressure level in the
downstream piping.
Calculation of Sound Pressure Level in the Downstream Piping
3 2 × 109 W a ρ2 C 2
Lpi = 10 log 6 21
d2p
Earlier, it was explained that increased noise levels can result from increasing
Mach numbers M2. In fact, a higher Mach number in the downstream piping
can increase the noise level by the value of Lg as shown in Eq. (6.22) as follows:
Calculation of the Noise Reduction Obtained by Reducing the Mach

Number
1
Lg = 16 log 6 22
1 − M2 10
By using Eq. (6.23), we are able to determine the loss of sound transmission due to
the downstream piping. In addition to this, it is important to pay attention to the
fact that increasing a pipe’s wall thickness leads to an increase in noise loss.
Calculation of the Basic Sound Transmission Loss Through Piping Wall

2
−7 C2 GX pa
TL = 10 log 7 6 × 10 × × × 6 23
tp f p ρ2 C 2 415GY + 1 ps
Transmission loss is determined by frequencies that are calculated according

to Eqs. (6.24) and (6.25). There are two types of frequencies that are calculated
in order to determine the value for transmission loss; one is pipe ring frequency
fr and the other is pipe coincidence frequency (Minimum Transmission
Losses) f0.
Calculation of the Pipe Ring Frequency
Pipe material sound speed Cp

fr = = 6 24
Pipe circumference πdp
Calculation of the Pipe Coincidence Frequency
Pipe circumferencial bending wave speed f C2

f0 = = r × 6 25
Pipe acoustic wave speed 4 Cair
It should be noted that f0 is less than fr (f0 < fr) as illustrated in Figure 6.7.
There is a possibility that three conditions could arise regarding the relationship
between frequency and transmission loss as follows:
1) If fp < f0, there will be a greater amount of transmission loss. This condition can
happen in the standard valves.
2) If fp = f0, there will be the smallest amount of transmission loss. This condition
can happen in the standard valves.
3) If fp > f0, there is a significant increase in transmission loss. This condition can
happen in low-noise trims.
In order to calculate the slope of the transmission loss in each of the three
regimes, we can use the following relationships, which are given in Eq. (6.26).
Figure 6.7 Relationship between frequency

and transmission loss.
TL
fr
fo
Peak frequency
208 6 Noise
Calculation of Transmission Losses in Three Regimes
f0 f0
If f p < f 0 ΔTL = 20 log + 13 log
fp fr
f0 6 26
If f 0 < f p < f r ΔTL = 13 log
fp
fp
If f r < f p ΔTL = 20 log
fr
There is an important reason for knowing that a higher fp (smaller dvc) can
increase the damping of the piping and reduce the noise associated with the
control valve throttling.
The net aerodynamic sound level at the pipe wall is converted to dB as per
Calculation of the Net Aerodynamic Sound Level at the Pipe Wall
LpAe = 5 + Lpi + TL + Lg 6 27
The aerodynamic noise level at 1 m from the pipe wall is calculated

with Eq. (6.28).
Calculation of the Net Aerodynamic Sound Level at the 1 m from the

Pipe Wall
L dp 2 × t p + 2
6 28
pAe,1 m = LpAe − 10 log dp + 2 × t p
6.4.2.2 Hydrodynamic Noise in Control Valves

It is common to use the IEC 60534-8:4: Industrial Process Control Valves – Part 8 –
Noise Considerations – Section 4 – Prediction of Noise Generated by Hydrodynamic
Flow, for Control Valves. Restriction of the valve converts fluid pressure energy
into turbulence, heat, and mechanical waves that produce noise and vibration
inside the valve and the pipes that connect to it. Even though cavitation and
flashing are the main causes of hydrodynamic noise generation, as discussed ear-
lier in this chapter, a turbulent flow can also result in hydrodynamic noise. Note
that the hydrodynamic noise generated by flashing, unstable, and unpredictable
flow conditions, as well as mechanical vibrations, is not included in the afore-
mentioned standard, and as such, this standard only takes into account noise
caused by cavitation and hydraulic turbulence. The standard provides a method-

ology for noise prediction based on fundamental principles of acoustics,
fluid mechanics, and mechanics. Moreover, the method described in this
section for noise prediction can be applied to all types of valves used for flow con-
trol or throttling such as globe, butterfly, and V-notch ball valves. In this standard,
the prediction noise method has been validated by using a water test for nearly all
kinds of control valves, with an inlet pressure of up to 15 bar. Test results indicate
that this method is accurate in the majority of cases within the range of plus or
minus dB. The hydrodynamic noise calculations are divided into two categories:
preliminary calculations and noise calculations. There are nine preliminary
calculations:
6.4.2.2.1 Preliminary Calculations

Pressure and Pressure Ratios In order to predict and measure noise, different pres-
sures and pressure ratios are needed. Differential pressure ratios xF for liquids
depend on the differential pressure across the valve P1 − P2 and the difference
between the inlet and vapor pressure P1 − Pv, as calculated in Eq. (6.29). As dis-
cussed in Chapter 3, the pressure ratio is calculated in much the same way as
the cavity index is calculated in order to evaluate cavitation.
Differential Pressure Ratio Calculation
P1 − P2
xF = 6 29
P1 − Pv
The differential pressure for starting the choked flow is approximately equal to
F 2L P1− Pv where FL is the liquid pressure recovery factor of the valve without
attached fitting. Choked flow is a fluid dynamic phenomenon associated with
the venturi effect. A liquid pressure recovery factor predicts how much pressure
recovery will happen after the vena contracta at the valve outlet. This experimen-
tally determined coefficient is a dimensionless number that accounts for the influ-
ence of the internal geometry of the valve on the maximum capacity of the valve.
From Eq. (6.30), it can be calculated what the differential pressure is at the vena
contracta ΔPc.
Differential Pressure Calculation at the Vena Contracta
ΔPc = Lower than x F P1− Pv or F 2L P1− Pv 6 30

210 6 Noise
In low-pressure differentials, noise is typically generated by turbulent flow.

Whenever xF exceeds the characteristic pressure ratio xFz, cavitation noise over-
whelms the turbulent noise. The following subsection discusses the characteristic
pressure ratio.
Characteristic Pressure Ratio The valve characteristic pressure ratio xFz identifies
the pressure ratio at which cavitation is detected. xFz is dependent on a number of
factors, including the type of valve, its trim shape and type, and the valve flow
capacity. The value of xFz can be estimated using either Eq. (6.31) or Eq. (6.32),
depending on the type of valve trim (internal).
Characteristic Pressure Ratio Calculation
0 90
x FZ = For valve types except multi-hole trims
1 + 3F d C N 34 × F L
6 31
1
x FZ = For multi-hole trims 6 32
4 5 + 1650 × N 0 d2H FL
where:
Fd: Based on the type of valve and closure member as well as the flow coefficient, a
valve style modifier is determined. This dimensionless value must be provided
by the valve manufacturer or calculated according to Eq. (6.33);
C: Valve flow coefficient (Cv or Kv);
N0: Number of independent and identical flow passages in the valve trim or num-
ber of stages of throttling in the valve trim;
dH: Multihole trim hole diameter (m);
N34: It is a numeric constant whose value depends on the type of flow coefficient
(Cv or Kv) used in the calculation of the characteristic pressure ratio. N34 = 1 for
Kv, and N34 = 1.17 for Cv.
Calculation of Valve Style Modifier
1
Fd = 6 33
N0
If xFz is obtained from testing at an inlet pressure of 6 × 105 Pa, then Eq. (6.34)
must be applied to correct for the actual inlet pressure.
Calculation of Differential Pressure Ratio Corrected for Inlet Pressure

0 125
6 × 105
x Fzp1 = x Fz 6 34
P1
Fluid Jet Diameter and Velocity Jets are streams of fluid that are projected into
nozzles or orifices in piping systems and can travel a considerable distance due
to their high momentum. In reality, the velocity of the fluid is increased within
the orifice or nozzle while its pressure is reduced. A great deal of energy is con-
tained within the fluid which is converted into kinetic energy and high-speed fluid
jets. The diameter of the jet fluid can be predicted from Eq. (6.35).
Calculation of Fluid Jet Diameter
DJ = N 14 F d CF L 6 35
where:
N14 = 4.9 × 10−3 When Kv is used in the equation. Otherwise, N14 = 4.6 × 10−3 if Cv
is used.
In Eq. (6.36), the vena contracta velocity Uvc, which is an important parameter
for determining the mechanical power, is calculated.
Calculation of Fluid Jet Velocity
1 2ΔPc
U vc = 6 36
FL ρL
where:
Uvc: Vena contracta velocity (m/s);

FL: Liquid pressure recovery factor of the valve without attached fitting
(dimensionless);
ΔPc: Differential pressure at the vena contracta calculated from Eq. (6.30) (Pa);
ρL: Density of liquid or fluid (kg/m3).
By using Eq. (6.37), the mechanical energy or power Wm dissipated in the orifice
or valve vena contracta can be determined.
Calculation of Mechanical Fluid Stream Power
m × U 2vc F 2L
Wm = 6 37
2
212 6 Noise
where:
Wm: Mechanical power (W);

m: Mass flow rate (kg/s).
Example 6.6 When a single-seat 4 globe valve without multihole trim is

installed on a 4 water pipe, it has a flow coefficient Cv of 90. In addition, the valve
style modifier is 0.42 while the pressure recovery factor is 0.92. The valve has inlet
and outlet pressures of 10 bar and 8 bar, respectively. However, the liquid passing
through the valve has a vapor pressure of 2.32 × 103. Knowing that the density of
the water is 997 kg/m3 and its mass flow rate is 30 kg/s, calculate the fluid jet diam-
eter, the velocity of the water at the vena contracta as well as the mechanical power
produced by the fluid stream. What is the cause of the noise produced in the valve?
Is it mainly due to turbulent flow or cavitation?
Answer
P1 = 10 bar = 106 Pa, P2 = 8 bar = 8 × 105 , Pv = 2 32 × 103

Differential pressure ratio: xF = (P1 − P2)/(P1 − Pv) = (0.2 × 106)/9,997,680 =
0.0200
Pressure differential or choked pressure across the vena contracta: ΔPc =
Lower than x F P1− Pv or F 2L P1− Pv
x F P1− Pv = 0 020 × 9,997,680 = 2 × 105 Pa

2
F 2L P1− Pv = 0 92 × 9,997,680 = 8 46 × 105 Pa
ΔPc = 2 × 105 Pa
Characteristic pressure ratio: x FZ = 0 90 1 + 3F d C N 34 × F L
(For valve types except for multihole trims) x FZ = 0 90
1 + 3 × 0 42 90 1 17 × 0 92 = 0 9 3 54 = 0 254 > x F = 0 0200
Since the characteristic pressure ratio is more than the differential pressure
ratio, turbulence flow is the primary cause of noise in the valve.
Characteristic pressure ratio corrected for the inlet pressure: xFzp1 =
xFz((6 × 105)/P1)0.125 = 0.254((6 × 105)/106)0.125 = 0.238
Water jet diameter: DJ = N 14 F d CF L = 4 6 × 10 − 3 × 0 42 90 × 0 92 =
0 01758 m
Vena contracta velocity: U vc = 1 F L 2ΔPc ρL = 1 0 92 × 4 × 105 997 =
21 77 m s
2 2
Mechanical water stream power: W m = m × U 2vc F 2L 2 = 30 × 21 77 0 92
2 = 6017W
6.4.2.2.2 Based on the function of acous-

Secondary Calculations (Noise Prediction)
tic efficiency, an amount of the mechanical power Wm computed from Eq. (6.37) is
converted to the valve internal noise Wa. Acoustic conversion efficiency, also
called acoustic efficiency factor η, describes the percentage of fluid power that
is converted to the noise inside the valve. The acoustic power ratio rw represents
the fracture of sound power radiated into the pipe. There are two different sets of
equations and conditions that are used to calculate the sound noise; the first is dur-
ing flow turbulent conditions (Eq. 6.38), while the second is during cavita-
tion (Eq. 6.39).
Noise Calculation for Turbulent Flow

If ΔP < x Fzp1 P1− Pv W a = ηturbulant × W m × r w 6 38
Noise Calculation for Cavitation Flow Condition
If ΔP > x Fzp1 P1 − Pv and x Fzp1 > x F > 1 Wa
6 39
= ηturbulant + ηcavitation × W m × r w
where:
Wm: Mechanical power (W);

Wa: Sound power (W);
ηturbulant: Acoustic efficiency factor at fluid turbulent condition (dimensionless)
calculated from Eq. (6.40);
ηcavitation: Acoustic efficiency factor at fluid cavitation condition (dimensionless)
calculated from Eq. (6.41);
rw: Acoustic power ratio (dimensionless).
rw is determined by the type of valve, and it can be found in Table 6.5.
Table 6.5 Acoustic power ratio rw.
Valve or fitting rw
Globe 0.25
Butterfly 0.5
Eccentric rotary plug 0.25
Ball 0.25
Expanders 1
214 6 Noise
Acoustic Efficiency Factor for Turbulent Flow
U vc
ηturbulant = 10 − 4 6 40
CL
CL: Speed of sound in liquid (m/s)
Cavitation is the second part of the process after the liquid is converted to vapor
bubbles. This occurs at the vena contracta or at a point with a high-pressure drop
and increased fluid velocity. The second part of this process is the collapse of the
vapor bubbles due to the fluid pressure exceeding the vapor pressure. As the bub-
bles collapse, a large amount of jet energy is created which is a source of noise and
serious damage to the valve. Water is known to be one of the most destructive
fluids during cavitation. In the cavitation region, xFzp1 > xF > 1ηcavitation is calcu-
lated as follows:
Acoustic Efficiency Factor for Cavitation Flow
05
P 1 − P2 1 1 − x Fzp1
ηcavitation = 0 32ηturbulant × e5xFzp1 ×
ΔPc x Fzp1 1 − xF
6 41
05
xF 15
× × x F − x Fzp1
x Fzp1
Example 6.7 In the previous example, it was necessary to determine whether

the fluid was turbulent or cavitating. In addition, what are the values of the acous-
tic efficiency factor and the sound power?
Answer
P1 = 10 bar = 106 Pa, P2 = 8 bar = 8 × 105

ΔP = 0 2 × 106 Pa = 2 × 105 Pa
x Fzp1 P1 − Pv = 0 238 106 − 2 32 × 103 = 0 238 × 9,997,680
= 2 38 × 105 Pa
ΔP < x Fzp1 P1 − Pv The flow is turbulent
Using Eq. (6.40), we can calculate the acoustic efficiency factor for the turbulent
flow as follows:
U vc 21 77
ηturbulant = 10 − 4 = 10 − 4 = 1 47 × 10 − 6
CL 1482
It is important to know that parameter CL is the speed of sound in the water that
is equal to 1482 m/s in this case.
In terms of acoustic power ratio, rw.The globe valve has a value of 0.25. It is now
possible to calculate the noise power for a turbulent flow as per Eq. (6.38) accord-
ing to what follows:
Wa = ηturbulant × Wm × rw = 1.47 × 10−6 × 6017 × 0.25 = 0.00234 W
6.4.3 Noise in Pressure Safety or Relief Valves

However, safety valves are not the primary issue when considering noise emission,
safety valves are evaluated more and more with regard to noise emission, espe-
cially if they discharge to open air, which can cause significant noise pollution
within a short space of time. The high noise levels produced by pressure-reducing
devices, such as PSVs, contribute to high-stress levels in the piping system and lead
to damage as illustrated in Figure 6.8.
The noise calculations are generally based on the expansion of the steam or gas
at the end of the pipe. Specific characteristics of safety valves, such as the geometry
of the valve outlet, are not taken into account in the calculations. Noise emission
tests are not typically conducted on safety valves. Additionally, the frequency of
Figure 6.8 Failure of the relief piping system connected to a PSV due to noise and
acoustic fatigue.
216 6 Noise
the noise is not determined by the noise calculations for the safety valve. Also,
worth mentioning is the fact that, unlike control valves, there is no low-noise
trim available for safety valves. For some specifications, such as NORSOK L-
002, piping system layout, design, and structure analysis, there is a maximum
permissible noise limit, as shown in Eq. (6.5). In cases where the calculated noise
level exceeds the maximum permissible level, different approaches, such as the
use of a silencer, are recommended. Another method of reducing noise is to
decrease the mass flow of air passing through the valve by limiting the opening
of the valve. There are three ways to calculate the noise emission level. They are
as follows:
• ISO 4126-9
• API 521
• VDI 2713
It should be noted that noise calculations based on all three standards are inde-
pendent of the safety valve designs provided by manufacturers. Due to this, the
design of the various safety valves does not affect the noise level as long as they
have the same capacity. In general, two physical values that were discussed earlier
in this chapter are relevant to the noise evaluation:
• Sound power level (PWL) is a measure of the amount of energy being generated
and emitted by the noise source which is the safety valve. The sound power level
is independent of the distance from the source of the noise, as previously
explained.
• The sound pressure level describes the pressure variation caused by the noise
source depending on the distance from it. This is the type of noise that affects
human hearing.
6.4.3.1 Calculation of Noise Emission According to ISO 4126-9

According to ISO 4126-9, safety valve noise calculations are based on symbols,
designations, and units defined in Table 6.6.
The sound power level of the safety valve, PWL, expressed in dB, can be
estimated by Eq. (6.42) as follows:
Calculation of the Sound Power Level of a Safety Valve
PWL = 20 log 10 − 3 dA – 10 log v + 80 log u – 53 6 42
It is possible to calculate the sound pressure level SPLr, expressed in dB at a dis-

tance r from the point of discharge to the atmosphere by using Eq. (6.43) as follows:
Table 6.6 Symbols, designations, and units used in calculating noise for safety valves
as defined in ISO 4126-9.
Used
symbols Designations Units
dA Internal diameter of outlet pipe mm

v Specific volume of the stream at relieving pressure and temperature m3/kg
u Velocity of fluid in outlet pipe m/s
r Distance from noise source M
Calculation of the Sound Pressure Level of a Safety Valve
SPLr = PWL − 10 log 2πr 2 6 43
Example 6.8 It has a 6 -diameter pipe with a standard thickness or schedule and
it is located on the outlet of a PSV. As the safety valve is dealing with natural gas
whose specific volume is 1.2 m3/kg in this instance, the gas volume will be handled
safely. How much sound power level will there be at 2 and 10 m away from the
safety valve if the velocity of the gas in the outlet pipe is 15 m/s? What is the sound
pressure level near the safety valve at a distance of 2 and 0 m?
Answer
The outside diameter of the 6 pipe, according to ASME B36.10, is 6.625
(168.3 mm), which is equal to 168.3 mm. Furthermore, the wall thickness for a
6 pipe with a standard wall thickness is 7.11 mm. So the internal diameter of
the pipe is calculated as follows:
dA = Pipe inside diameter = Pipe outside diameter
− 2 × Thickness = 168 3 − 2 × 7 11 = 154 08 mm
The sound power level (PWL) that is calculated according to Eq. (6.42) is inde-
pendent of the distance, meaning it is the same in both a 2-m and a 10-m distance.
PWL = 20 log 10 − 3 dA − 10 log V + 80 log u − 53

= 20 log 10 − 3 × 154 08 − 10 log 1 2 + 80 × log 15 − 53
= − 16 245 − 0 79 + 94 09 − 53 = 24 dB
Therefore, the sound power level (PWL) is 24 dB both at a distance of 2 and 10 m

from the safety valve. As a next step, the noise pressure level (SPL) needs to be
calculated at a distance of two and ten meters from the safety valve.
218 6 Noise
2
SPL2 = PWL − 10 log 2πr 2 = 24 − 10 log 2 × π × 2
= 24 − 10 × 1 4 = 10 dB
2
SPL10 = PWL − 10 log 2πr 2 = 24 − 10 log 2 × π × 10
= 24 − 10 × 2 8 = − 4 dB
Accordingly, the PWL at 2 m and 10 m is 10 dB and −4 dB, respectively.
6.4.3.2 Calculation of Noise Emission According to API 521

Noise levels are normally calculated in accordance with API 521 at a distance of
30 m or 100 ft from the point of discharge. However, it is possible to calculate the
noise level at other distances from the discharge of the safety valve using an equa-
tion which will be discussed further in this chapter. In Table 6.7, you are given a
list of the symbols, designations, and units used in API 521 for the sizing of
safety valves.
Table 6.7 Symbols, designations, and units used in calculating noise for safety valves as
defined in API 521.
Used symbols Designations Units
L30 (100) Noise or sound pressure level at 30 m (100 ft) from the dB
point of discharge
L Noise or sound pressure level at the valve discharge as dB
extracted from Figure 6.9
Lp Sound pressure level at distance r dB
r Distance from the source of the sound m ft
qm Mass flow through the valve kg/s lb/s
C Speed of sound in the gas inside the valve that can be m/s ft/s
calculated from either Eq. (6.45) or Eq. (6.46)
k Ratio of the specific heats in the gas —
M Relative molecular mass of the gas —
T Gas temperature k R
PR, X Pressure ratio across a valve that is calculated as the ratio —
of relieving pressure to back pressure
Y Sound pressure level, L30 (100) that can be determined from dB
Figure 6.9 by considering the value of the pressure ratio or
by using Eq. (6.44)
70
60
50
40
30
20
1.5 2 3 4 5 6 7 8 9 10
Pressure ratio, PR
Absolute relieving pressure
Absolute back pressure
Figure 6.9 Noise intensity at the valve outlet L = (L30 − 10 log10(0.5qm × C2)) based on
the pressure ratio according to API 521 standard.
In order to calculate the noise level at 30 m from the discharge point of the PSV
to the atmosphere, Eq. (6.44) is applied:
Calculation of the Noise Level at 30 m from the PSV’s Discharge Point

to the Atmosphere
L30 = L + 10 log 10 0 5qm × C2 6 44
Calculation of the Speed of Sound in the Gas Inside the Valve
C = 91 2 kT M 0 5 m s SI units 6 45
C = 223 kT M 0 5 ft s USC 6 46
By applying Eqs. (6.47) and (6.48), the noise level Lp can be adjusted for distances
that differ from 30 m (100 ft).
r
Lp = L30 − 20log SI units 6 47
30
r
Lp = L30 − 20log USC units 6 48
30
220 6 Noise
Example 6.9 Figure 6.10 illustrates a pressure vessel protection system in

12 × 10 for the protection of pressure vessels. The set pressure of the valve is
42 barg. So the inlet of the valve is Class 300, which is equal to 50 barg design pres-
sure. In addition, the outlet line of the valve is Class 150. There is a mass flow rate of
59,507.55 kg/h passing through the PSV and the temperature of the fluid is 110 C.
In addition, the specific heat of the gas is 1.18 and the molecular weight of the gas is
21.88. There is a 7.50 bar back pressure on the valve. What are the sound pressure
levels at both one meter and thirty meters away from the safety valve? In this case,
it is important to note that there is a 10% overpressure on the safety valve.
Answer
The first step is to get parameter C by using Eq. (6.45). To do so, it is important to
convert the temperature from degrees Celsius to Kelvin using the following
formula.
Gas temperature is 110 C,
TemperatureKelvin = TemperatureCentigrade + 273 15 TemperatureKelvin
= 110 + 273 15 = 383 15
05
1 18 × 383 15
C = 91 2 = 414 57 m s
21 88
Mass flow rate qm = 59,507 55 kg h = 16 53 kg s
Figure 6.10 12 × 10 PSV for

Pressure
protection of the vessel.
relief valve
12 in Valve
10 in.
Pipe CSO
entrance valve
Protected
vessel
Valve relieving pressure bara = Set pressure +

Overpressure + Atmospheric pressure
PR =
Valve back pressure bara = Valve backpressure barg
+ Atmospheric pressure
42 + 0 1 × 42 + 1 47 2
= = = 5 55
7 50 + 1 85
Thus, based on Figure 6.9, the noise intensity level or sound pressure level at the
discharge of the valve corresponds to the pressure ratio of 5.55. This level is approx-
imately 55.7 dB. Now it is possible to calculate and obtain the value of the noise
intensity level at a distance of 30 m from the safety valve at a distance of 30 m
by using Eq. (6.44).
L30 = L + 10 log 10 0 5qm × C2

= 55 7 + 10 log 10 0 5 × 16 53 × 414 572
= 55 7 + 10 log 1,420,491
= 55 7 + 61 52 = 117 22 dB
Using Eq. (6.47), it is possible to measure the noise level at a distance of 1 m

from the valve outlet.
r 1
Lp = L30 − 20log L1 = 117 22 − 20log
30 30
= 117 22 − 20log 0 03333
= 117 22 + 29 54 = 146 76 dB
6.4.3.3 Calculation of Noise Emission According to VDI 2713

As a rule, noise levels for steam services are usually calculated according to VDI
2713. In Table 6.8, you will find a list of the symbols, designations, and units used
in VDI 2713 to determine the size of safety valves.
The following equation is used in order to calculate the noise level of steam.
qm
Lw = 17 log + 51 log T − 15 6 49
20
The distance-dependent noise level can be calculated in the following way:

LA = Lw × 10 × log A 6 50
222 6 Noise
Table 6.8 Symbols, designations, and units used in calculating noise for safety valves
as defined in VDI 2713.
Used symbols Designation Units
Lw Noise level dB
LA Noise level at the distance of r meter(s) dB
qm Mass flow rate kg/h
P Set pressure Bar
αd Coefficient of discharge —
T Temperature Kelvin
R Radios of the imaginary hemisphere are used to measure the m
distance from the source of the noise
A Surface of the “imaginary hemisphere” having a radius of r m2
A = 2πr2

6.1 The control valve in question is 4 in size and connected to 8 inlet and outlet
piping with a schedule 40 wall thickness. The flow passing through the valve
is air with a capacity of 30,000 kg/h at a temperature of 100 C. Valve inlet
and outlet pressures are 30 bar and 12 bar, respectively. Which of the follow-
ing statements is true regarding the noise generated by the valve? The air
density and molecular weight are 1.225 kg/m3 and 28.97 g/mol, respectively.
A The calculated Mach number of the valve is less than 0.3.
B The sonic speed in this case is 290 m/s.
C The valve has an outlet velocity of less than 150 m/s.
D The noise produced by the valve is approximately 150 dB, which does
not exceed the allowable level.
Answer
Sonic velocity is calculated by using Eq. (6.8) as follows:
γRT 1 4 × 8315 × 100 + 273 15

C= =
M 28 97
= 387 22 meter per second m s
Thus, option B is incorrect. The next step is to calculate the valve outlet veloc-
ity as per Eq. (6.10). Parameter r is the internal radius of an 8 pipe STD
thickness connected to the valve. Using Example 6.2, a pipe with an outside
diameter of 8 and a thickness of 0.322 has an outside diameter of 8.625

and a thickness of 0.322 . The internal diameter and radius of the pipe
are calculated as follows:
t = 0 322 and OD = 8 625 ID = 8 625 – 20 322
= 7 981 Internal radius
= 3 995 in = 0 10135
w w 30, 000 3600
V outlet = = 2 = = 210 91 m s
Aρ πr ρ π × 0 10135 2 × 1 225
The outlet velocity of the valve is greater than 150 m/s. Therefore, option C is
incorrect.
Equation (6.7) can be used to determine the Mach number by knowing the
sonic velocity and the valve outlet velocity:
V outlet 210 91
M outlet = = = 0 544
C 387 22
Mach number is greater than 0.3. So option A is incorrect. The last step is to
determine the noise generated in the valve by Using Eq. (6.2):
36 2 12
P1 − P2 W T 1 + 273
PWL = 10 log × × + 126 1 + SFF
P1 3600 m
36 2 12
30 − 12 30, 000 100 + 273
= 10 log × × + 126 1
30 3600 28 97
= 10 log 0 1590 × 69 44 × 21 464 + 126 1

= 23 747 + 126 1 = 149 85 dB
Equation (6.5) provides the following formula for calculating the allowable
noise level:
Di 7 981
PWLA = 173 6 − 0 125 × = 173 6 − 0 125 ×
t 0 322
= 173 6 − 3 0982 = 170 50 dB
There is a stricter noise limit of 155 dB. Since the produced noise level is less
than both limits, it is acceptable and option D is the correct response.
6.2 A relief valve, which is connected to a pipe of 8 in. in diameter, has a max-
imum flow capacity of 1,40,000 kg/h. At the downstream of the valve, there is
a developed back pressure of 8 bar and the discharge temperature of the valve
is 40 C. Inlet pressure and temperature of the valve are 30 bar and 40 C,
224 6 Noise
respectively. The ideal gas passing through the valve has a molecular weight
of 25 and a heat ratio of 1.21. Assuming that the outlet pipe has a standard
thickness and no sonic condition exists (SFF = 0), which of the following
statements is correct regarding noise and acoustic fatigue analysis for
the valve?
A Mach number at the valve outlet is calculated to be 0.40.
B The power of the produced sound is 160 dB.
C The produced sound exceeds the acceptable sound power limit.
Answer
Consider Example 6.2, which shows that an 8 pipe with a standard (STD)
wall thickness has an outside diameter of 8.625 (0.2190 m), an inside diam-
eter of 7.981 (0.2027 m), and a thickness of 0.322 (0.0082 m). Given here is a
summary of all provided piping and process data:
P1: The upstream pressure of the relief valve = 30 bar;
P2: The downstream pressure of the relief valve = 8 bar = 8 × 105 Pa
T1: The upstream temperature of the valve = 39.85 C = 39.85 +
273.15 = 313 K
W: Gas or liquid flow rate = 140,000 kg/h = 38.89 kg/s
γ: Ratio of specific heat = 1.21
Outlet pipe inside diameter = Di = ID = 8.625 – 20.322 = 7.981 = 0.2027 m
Outlet pipe wall thickness = t = 0.0082 m
z factor = 1
In order to calculate the Mach number, we need to use Eq. (6.16), where Z
is equal to 1 and R is equal to 8315.
W T2 38 89
M = 116 × × = 116 ×
P2 D2i γm 8 × 105 × 0 20272
313
× = 0 1372 × 3 22 = 0 44
1 21 × 1 25
According to Eq. (6.2), the sound power is calculated as follows:
36 2 12
P1 − P2 W T 1 + 273
PWL = 10 log × × + 126 1 + SFF
P1 3600 m
36 2 12
30 − 1 1,40,000 39 85 + 273
= 10 log × × + 126 1
30 3600 25
= 10 log 0 885 × 1512 × 20 76 + 126 1
= 44 43 + 126 1 = 170 54 dB

Di 0 2027
PWLA = 173 6 − 0 125 × = 173 6 − 0 125 × = 170 51
t 0 0082
Option A is incorrect as the Mach number is equal to 0.44. Option B is incor-
rect as well since the generated sound power exceeds 160 dB. Since the pro-
duced sound is 170.54 dB, it exceeds the acceptance level of 170.51, as well as
155 dB. Therefore, option C is the correct answer. Option D is incorrect since
option C is the correct answer.
6.3 In Figure 6.11, three noise sources are illustrated, namely PSVs C, A, and D.
Each of these PSVs has an 8 inlet pipe with a standard (STD) wall thickness.
The sound power levels at point “B” located 10 m from each PSV source are
as follows:
• Sound power level at point B due to pipe section A − B = 160 dB
• Sound power level at point B due to pipe section C − B = 163 dB
• Sound power level at point B due to pipe section D − B = 165 dB
Which of the following statements is true?
A Due to multiple sources of noise, the sound level produced at point B is
greater than 170 dB.
B Since the calculated noise level at point B does not exceed the established
limit, the piping at point B can handle the noise safely.
C Depending on the likelihood of piping failure resulting from the noise,
the amount of noise produced at point B may be acceptable.
D By reducing the wall thickness and increasing the flow rate, the pipe
system’s noise problem can be mitigated.
Figure 6.11 The three PSVs are located

10 m from point B.
D
226 6 Noise
Answer
From Eq. (6.4), the sound power level generated at point B due to the dis-
charge of multiple sources of noise that are PSVs C, A, and D at the same
time is calculated as follows:
PWLX = 10 log 10 PWL1 10

+ 10 PWL2 10
+ 10 PWL3 10
= 10 log 10160 10
+ 10163 10
+ 10165 10
= 10 log 1016 + 1016 3 + 1016 5 = 10 × 16 79 = 167 9 dB
Using Eq. (6.5), the sound power at point B, which is located at a distance of
10 m from each PSV source, can be calculated as follows. Referring to Exam-
ple 6.2, an 8 pipe with standard (STD) wall thickness has an outside diam-
eter of 8.625 (0.2190 m), an inside diameter of 7.981 (0.2027 m), and a
thickness of 0.322 (0.0082 m).
Di 0 2027
PWLA = 173 6 − 0 125 × = 173 6 − 0 125 × = 170 51
t 0 0082
It is incorrect to choose option A since the sound level produced at point B as

a result of the simultaneous operation of three PSVs is 167.9 dB, which is
below 170 dB. Furthermore, option B is incorrect as well, as even if the cal-
culated noise level is below the allowable noise level, the calculated noise
levels are still higher than 155 dB. So we cannot conclude that the pipes
at point B are safe from noise. The correct answer is C. In this case, even
if the predicted noise is above 155 dB, the piping design at point B can be
considered safe if the likelihood of failure is less than 0.5. Option D is incor-
rect because increasing the pipe wall thickness and reducing the flow rate are
two methods for mitigating the noise risk in the piping system. Increasing
the number of PSVs in the system can reduce the flow rate at the PSVs’
outlet piping.
6.4 Water is carried by a 4 globe control valve with a flow rate of 40 kg/s. The
valve has an inlet pressure and an outlet pressure of 10 bars and 6.5 bars,
respectively. In this case, water density is assumed to be 997 kg/m3 and its
vapor pressure is 2320 Pa. Considering the speed of sound in water to be
1400 m/s, which of the following statements is wrong regarding the hydro-
dynamic noise in the control valve? (Note: Liquid pressure recovery fac-
tor = 0.92, valve style modifier = 0.42, and valve flow coefficient Cv is 90.)
A Acoustic noise is caused primarily by cavitation in the valve.
B The water jet velocity in vena contracta is 28.8 m/s.
C The power of the sound produced is more than 1 W.
D Currently, the Acoustic power ratio stands at 0.25.
Answer
In order to determine whether the hydrodynamic noise in the control valve is
primarily caused by cavitation or turbulent flow, the following two condi-
tions should be examined:
1) Whenever the differential pressure ratio xF exceeds the characteristic
pressure ratio xFz, cavitation noise domains the turbulent noise. Accord-
ing to the following calculations, cavitation is a dominant cause of noise
in the control valve.
P1 − P2 106 − 6 5 × 105 3,50,000

xF = = = = 0 3508
P1 − Pv 106 − 2320 9,97,680
0 90
x FZ = For valve types exceptmulti-hole trims
1 + 3F d C N 34 × F L
0 90 0 90
x FZ = =
1 + 3F d C N 34 × F L 1 + 3 × 0 42 90 1 17 × 0 92
0 90
= = 0 2543
3 5385579
Another condition that indicates cavitation is the predominant cause of the
noise is the following: ΔP > xFzp1(P1 − Pv). By examining the calculations
later, we can see that the second condition for noise generation in the valve
due to cavitation is met. Thus, option A is correct.
ΔP = 106 − 6 5 × 105 = 350,000

0 125 0 125
6 × 105 6 × 105
x Fzp1 = x Fz = 0 2543 = 0 2386
P1 106
x Fzp1 P1− Pv = 0 2386 106− 2320 = 238,046
Next, it is necessary to calculate the fluid velocity at the vena contracta. In

order to accomplish this, the pressure differential ΔPc at the vena contracta is
first calculated as follows:
ΔPc = Lower than x F P1− Pv or F 2L P1− Pv
x F P1− Pv = 0 3508 106− 2320 = 349,986 144 Pa
2
F 2L P1− Pv = 0 92 106− 2320 = 844,436 352 Pa
ΔPc = 349,986 144 Pa
We can now calculate the velocity at vena contracta as follows:
1 2ΔPc 1 2 × 349,986 144

U vc = = = 28 80 m s
FL ρL 0 92 997
228 6 Noise
Therefore, option B is also valid. Next, it is necessary to calculate the

mechanical power generated by noise.
m × U 2vc F 2L 40 × 28 802 × 0 922

Wm = = = 14,040 76 W
2 2
The following part discusses how to calculate the portion of mechanical
power that is converted to sound power under conditions of cavitation. At
this stage, three parameters must be determined: the acoustic efficiency fac-
tor at turbulent condition ηturbulant and acoustic efficiency factor at cavitation
conditions ηcavitation, as well as the acoustic power ratio rw. Referring to
Table 6.5, the acoustic power ratio should be 0.25. Thus, option D should
be selected.
U vc 28 80
ηturbulant = 10 − 4 = 10 − 4 = 2 057 × 10 − 6
CL 1400
05
P1 − P2 1 1 − x Fzp1
ηcavitation = 0 32ηturbulant × e5xFzp1 ×
ΔPc x Fzp1 1 − xF
05
xF
= 1 243 × 10 − 6
15
× × x F − x Fzp1
x Fzp1
Sound power Wa for cavitation dominant flow is calculated as follows:
W a = ηcavitation + ηturbulant W m × r w = 3 3 × 10 − 6 × 14,040 76

× 0 25 = 0 01158 W
Since the sound power is less than one watt, option C is wrong.
6.5 A globe valve is used to control the flow of water in a piping system. The flow
through the valve corresponds to a pressure drop across the valve of 1 psi,
which is equivalent to 100 gpm. Based on a liquid pressure recovery factor
of 0.89 and a valve style modifier of 0.45, what is the diameter of the liquid
jet inside the valve?
A 19.53 mm
B 20.20 mm
C 18.19 mm
D 17.50 mm
Answer
The first step is to calculate the flow coefficient Cv value of the valve as
follows:
SG SG = 1 for water
Cv = Q = 100 = 100
ΔP 1
Now it is possible to calculate the water jet diameter inside the valve by
using Eq. (6.35).
DJ = N 14 F d CF L = 0 0046 × 0 45 100 × 0 89
= 0 01953 m = 19 53 mm
6.6 In a globe valve, the acoustic efficiency factors are 2 × 10−6 and 1.9 × 10−6 for
turbulent and cavity conditions, respectively. What would be the hydrody-
namic sound power as a function of the cavitation flow conditions and
the mechanical stream power of 14,000 W?
A 0.01365 W
B 0.01546 W
C 0.01687 W
D 0.01837 W
Answer
Cavitation is the type of fluid, and the sound power is calculated using
W a = ηturbulant + ηcavitation × W m × r w = 2 × 10 − 6 + 1 9 × 10 − 6
× 14,000 × 0 25 = 0 01365 W
6.7 In a liquid service, throttling is achieved via a globe valve. Due to turbulent
flow, the valve produces a considerable amount of noise in the piping system.
The liquid pressure recovery factor is 0.9, and the pressure loss at vena con-
tracta is 2,00,000 Pa. The mass flow rate of water through the valve is assumed
to be 35 kg/s, and the density of water is assumed to be 1000 kg/m3. What
percentage of the fluid’s mechanical energy is converted into noise? (Note:
Assume that the sound velocity in water is 1400 m/s.)
A 70%
B 50%
C Less than 1%
D 5%
Answer
In order to determine the fluid mechanical power, the first step is to deter-
mine the fluid velocity at the vena contracta area according to Eq. (6.36).
1 2ΔPc 1 2 × 200,000
U vc = = = 22 22 m s
FL ρL 09 1000
230 6 Noise
2
m × U 2vc F 2L 35 × 22 222 × 0 9
Wm = = = 6998 60 W
2 2
Considering that noise is generated by turbulent flow, the acoustic efficiency
factor for turbulent flow should be calculated according to Eq. (6.40) as
follows:
U vc 22 22
ηturbulant = 10 − 4 = 10 − 4 = 1 587 × 10 − 6
CL 1400
The value of rw is 0.25 for a globe valve. As a result, the sound power gen-
erated by turbulent flow can now be calculated using Eq. (6.38) as follows:
W a = ηturbulant × W m × r w = 1 587 × 10 − 6 × 6998 60 × 0 25

= 0 002776 W
Wa 0 002776
= = 3 9675 × 10 − 7
Wm 6998 60
Considering that less than 1% of fluid mechanical power is converted to
noise, option C is the correct answer.
6.8 A safety valve is used for handling natural gas with a molecular weight of 19
g and a relieving capacity of 7502 ft/h. Inlet relieving temperature is 120 F,
set pressure is 210 psi, overpressure is 10%, and backpressure is 50 psi. What
is the noise level at a distance of 500 ft away from the valve discharge when
the valve is opened based on API 521 standard? The specific ratio of heat for
natural gas is 1.27.
A 89.79 dB
B 103.8 dB
C 54.5 dB
D 28 dB
Answer
The first step is to calculate the absolute relieving pressure as follows:
P1 = Absolute relieving temperature = Set pressure + Overpressure
+ Atmospheric pressure = 210 + 210 × 10 + 14 7 = 245 7 psia
Backpressure absolute pressure = 50 psig + 14 7 = 64 7 psia
Valve relieving pressure 245 7
PR = = =38
Valve back pressure bara 64 7
It has been determined that the noise intensity at the discharge point is
54.5 dB (L = 54.5 dB) (refer to Figure 6.9). It’s important to understand that
Further Reading 231
the temperature of the gas in SI unit is measured in Kelvin, which is calculated

as follows:
Gas temperature is 120 F, TemperatureKelvin = 322.039 K
By using Eq. (6.45), the next step is to calculate the speed of sound in the
gas inside the valve using the formula as follows:
05 05
C = 91 2 kT M = 91 2 × 1 27 × 322 039 19 = 423 13 m s
The mass flow rate is 7502 lb/h = 2.084 lb/s = 0.945 kg/s
Now it is possible to calculate and obtain the value of the noise intensity
level at a distance of 30 m (100 ft) from the safety valve at a distance of 30 m
by using Eq. (6.44).
L30 100 ft = L + 10 log 10 0 5qm × C 2

= 54 5 + 10 log 10 0 5 × 0 945 × 423 132
= 54 5 + 10 log 84,596 = 54 5 + 49 27 = 103 77 dB
Using Eq. (6.47), it is possible to measure the noise level at a distance of 500 ft
equal to 150 m from the valve outlet.
r
L150 500 ft = L30 100 ft − 20log L150 500 ft
30
150
= 103 77 − 20log = 103 77 − 20 log 5
30
= 103 77 − 13 98 = 89 79 dB
Thus, the correct choice is A.
Further Reading
Ahmed, D. (2011). Acoustic fatigue assessment of piping components by finite element
analysis. Proceeding of ASME Pressure Vessel and Piping Conference, Baltimore, MA,
USA (1–21 July 2011).
American Petroleum Institute (API) (2014). API Standard 521, Pressure-Relieving and
Depressuring Systems, 6e. Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) B16.9 (2018). Factory-Made
Wrought Buttwelding Fittings. New York: American Society of Mechanical
Engineers (ASME).
American Society of Mechanical Engineers (ASME) B36.10/19 (2004). Carbon, Alloy
and Stainless-Steel Pipes. New York: American Society of Mechanical
Engineers (ASME).
Baumann, H.D. (1997). Solve valve noise and cavitation problems. Hydrocarbon
Processing 76 (3): 45–50.
232 6 Noise
Dresser Flow Control (2020). Noise control manual. Bulletin OZ3000 01/02.
Fagerlund, A.C. (1986). Recommended maximum valve noise level. Proceedings of the
ISA/86 International Conference and Exhibition, Houston, TX, USA (November).
Fisher Controls International (2001). Control Valve Handbook, 3e. Fisher.
Glaunach (2010). The Silencer Handbook. A General Introduction Into Noise and Its
Prevention. Glaunach.
International Electrotechnical Commission (IEC) 60534-8-4 (2015). Industrial Process
Control Valves – Part 8-4: Noise Considerations-Prediction of Noise Generated by
Hydrodynamic Flow. Geneva: International Electrotechnical Commission (IEC).
International Organization for Standardization (ISO) 4126-9 (2008). Safety Devices for
Protection Against Excessive Pressure. Application and Installation of Safety Devices
Excluding Stand-Alone Bursting Disc Safety Devices. Geneva: International
Organization for Standardization (ISO).
Momber, A.W. (2005). Hydrodemolition of Concrete Surfaces and Reinforced Concrete,
67–103. Elsevier https://doi.org/10.1016/B978-185617460-2/50003-2.
Oxford: Elsevier.
NORSOK L-002 (2009). Piping System Layout, Design and Structure Analysis, 3e.
Lysaker: NORSOK.
Norwegian Oil Industry Association (2013). Valve Technology. Norsk Olje & Gass.
Pettigrew, M.J., Paidoussis, M.P., Weaver, D.S., and Au-Yang, M.K. (1996). Flow-
induced vibration. Proceeding Paper. Conference: American Society of Mechanical
Engineers (ASME) Pressure Vessels and Piping Conference, Montreal, Canada (21–26
July 1996). Other Information: PBD: 1996.
Randall, E.I.M. (2001). PSV noise – criteria, limits and prediction. Valve World. https://
docplayer.net/39961976-Psv-noise-criteria-limits-and-prediction.html (accessed
12 March 2022).
Samson (2008). Improvement of IEC 60534-8-3 for noise prediction in control valves.
https://www.samsongroup.com/document/w01950en.pdf (accessed 2 June 2022).
Shahda, J. (2010). Predicting control valve noise in gas and steam applications: valve
trim exit velocity head vs. valve outlet Mach number. Dresser Masoneilan. https://
www.plantservices.com/assets/Media/1003/WP_Valve.pdf (accessed
13 March 2022).
Sotoodeh, K. (2019). Noise and acoustic fatigue analysis in valves (case study of noise
analysis and reduction for a 12 × 10 pressure safety valve). Journal of Failure
Analysis and Prevention 19: 838–843. 10.1007/s11668-019-00665-3.
Statoil (2014). Noise Control in Projects. Guideline, GL0563. 1st revision. Oslo: Statoil.
The Engineering Toolbox (2022). Speed of sound equations. https://www.
engineeringtoolbox.com/speed-sound-d_82.html (accessed 13 March 2022).
VDI 2713 (1974). Noise Reduction in Thermal Power Stations. Berlin: Deutsches Institut
fur Normung E.V. (DIN).
233
Water Hammering
7.1 Introduction
There is a risk of water hammering when fluid parameters, including flow, velocity,
and pressure, change suddenly in a piping system due to various circumstances such
as pump stopping or starting as well as valve opening or closing. High pressure is
generated in liquid services when the velocity of a mass of liquid or large velocity
liquid is suddenly reduced. Thus, the mobile energy of the liquid is converted into
pressure energy. Water hammering is not a kinetic energy problem but rather an
acoustic problem. Waves produced in the pipe as a result of water hammering are
much faster than liquid velocity. It has undesirable and devastating consequences
such as noise, wear, tear, and eventually the collapse and rupture of the piping system,
as well as the associated valves. Consequently, it is extremely important to understand
this problem in order to prevent it as well as to calculate and analyze the pressure
change in the piping caused by water hammering. More specifically, the primary
cause of water hammering is a rapid change in flow rate in the piping system caused
by the shutoff or startup of a pump or the opening or closing of a valve. The effects of a
water hammer can range from small changes in pressure and velocity to relatively
high pressure that can result in the bursting and failure of pipes, pipe fittings, valves,
and, in some cases, damage to pumps. For instance, a pump stoppage can cause a
severe form of water hammering, and a sudden power failure that causes all the
pumps to fail at once can cause an extremely severe case of water hammering. The
following section describes in detail what causes water hammering in check valves.
7.2 Water Hammering and Pressure Loss in Check Valves
Nonreturn valves, or check valves, open with the forward flow and close with the
reverse flow. Essentially, the check valve’s main function is to prevent equipment
and facilities, such as pumps, from reverse flow or backflow. In the oil and gas
234 7 Water Hammering
industry, there are several types of check valves, including swing, dual plate, axial,
piston, and ball lift. Several factors influence the selection of check valves, includ-
ing initial and maintenance costs, heat loss and energy costs, nonslam character-
istics, location of the valve (e.g. after pumps or compressors), fluid compatibility,
sealing ability, and flow characteristics. It is common to install nonslam check
valves, such as axial types, following pumps, and compressors. In various indus-
tries such as oil and gas, pumps are used to pressurize and move fluid (liquid)
through the piping system. In order to pressurize and move gas services, compres-
sors are used. It is important to know that all check valves are operated automat-
ically by the fluid flow inside the piping, without any need for an operator.
The most essential requirement for check valves, particularly those that are
installed after equipment such as pumps and compressors, is that they provide less
resistance to the flow in the normal direction that opens the valve. In contrast, a
check valve should exhibit unlimited resistance to backflow, also referred to as
reverse flow. Figure 7.1 illustrates a swing check valve, which has a disk that
moves upward as fluid moves through the piping system and the valve. Swing
check valves are available, low-cost, and offer relatively low-pressure loss when
fully open. The swing check valve is a popular choice of valve in water piping sys-
tems and, compared to other check valves, swing check valves are relatively inex-
pensive. Moreover, the valves provide very low-pressure drops, or head losses,
when fully opened. Referring to Eq. (7.1), head loss refers to the loss of fluid pres-
sure inside the valve compared to the pressure inside the piping. Valve pressure
drop is typically greater than pipe pressure drop. It means that the pressure drop
in the valve is 4 bars if the valve has an inlet pressure of 20 bars and an outlet pres-
sure of 16 bars.
Figure 7.1 A swing check valve

in the open position. Source:
American Journal of Industrial
Engineering/Science and
Education Publishing Co. Ltd.
Valve Head Loss

ΔPv = Pp – Pv 71
where:
ΔPv: Head loss or fluid pressure drop inside a valve;

Pp: Fluid pressure in the pipe;
Pv: Fluid pressure in the valve.
The operator company’s experience shows that check valves are not entirely
open when fluid passes through the valve. For example, the disk of a valve swings
only about 60 from the closed position and is approximately 30 from 100% or a
fully open position. Swing check valves are closed when the disk closes due to the
weight of the disk. So, if the fluid flow is interrupted, the disk quickly closes and
slams against the seat during the complete closure of the valve. The long stroke of
the disk coupled with the sudden closing of the valve caused by the weight of the
disk exacerbates the slamming effect of swing check valves. The sudden slamming
of the valve disks causes water hammering, which is a form of hydraulic shock, as
shown in Figure 7.2. The valve closes when the fluid cannot move forward. So, its
kinetic energy turns into waves that apply pressure and load to the pipe wall. These
shocks are actually pressure surges or waves that cause noise and damage to piping
systems. The constant opening and closing of the disk make slamming, which
causes water hammering and a higher rate of wear on the disk and valve seat.
Valve open – moving water
3 Valve closes – water hammer
Figure 7.2 Water hammering in the piping system due to rapid valve closure.
Example 7.1 The pressure of the water in the pipes is 12 bars. If 25% of the pip-
ing fluid pressure loss occurs in the valve, what is the valve outlet pressure?
Answer
Pp: Fluid pressure in the pipe = 12 bar

ΔPv: Fluid pressure drop or head loss in the valve = 25 % × 12 = 3 bar
ΔPv = Pp − Pv Pv = Pp – ΔPv = 12 − 3 = 9 bar
The valve has an outlet pressure of 9 bar.

Alternatively, the head loss can be calculated by using Eq. (7.2) as follows:
An Alternative Method of Calculating Valve Head Loss
KvV 2
ΔH = 72
2g
where:
ΔH: The head loss or pressure loss as measured by the water column (m, ft);
Kv: Valve flow resistance coefficient (dimensionless);
V: Fluid velocity (m/s, ft/s);
g: Acceleration due to gravity (9.81 m/s2, 32.2 ft/s2).
Equation (7.3) can be used to calculate the valve flow resistance coefficient using
the valve internal diameter and valve flow coefficient.
Calculation of the Valve Flow Resistance Coefficient
d4
K v = 980 73
C 2v
where:
Kv: Valve flow resistance coefficient (dimensionless);

d: Valve diameter (in.);
Cv: Valve flow coefficient.
Alternative solutions to prevent water hammering can be used when selecting

check valves. Many valve manufacturers equip swing check valves with dashpots,
shock absorbers, or cushioning/damping systems, which are practical accessories
for dampening swing check valves. In Figure 7.3, the check valve is equipped with
a dashpot which counterbalances the effect of slamming so that the disk does not
close too quickly. Although it should be noted that adding a dashpot as a slamming
Figure 7.3 A swing check valve

fitted with a dashpot. Source: DeZurik.
and water hammering prevention solution creates four new problems, the first is
that the valve closes slower, which causes a higher rate of backflow and reduces
the resistance of the valve. As a result, the pump installed upstream or before the
swing check valve must be able to handle some backflow. In addition, the swing
check valve with dashpot has a higher pressure drop than a standard check valve.
The fluid pressure drop is more critical and problematic when the piping system is
vertically installed with upward fluid movement. Thirdly, the swing check valve
with the dashpot is more expensive. The dashpot contains high-pressure oil, often
exceeding 2000 psi, and is very expensive. Furthermore, since dashpots apply high
loads to the hinge pin that connects the disk to the valve body, a check valve with a
large hinge pin diameter is required. The last challenge is that swing check valves
with dashpots require additional maintenance because they contain more moving
parts. Using a dashpot for swing check valves to prevent water hammering is pro-
hibited in many projects.
Studies have demonstrated that in order to prevent a check valve from slamming
and creating water hammering problems, the check valve must be closed rapidly
before reverse flow begins or the valve must be closed very slowly once reverse flow
has developed and reached the valve. As shown in Figure 7.4, a swing check valve
disk slams firmly onto the seat by the force of reverse flow, resulting in a strong
shock wave and water hammering inside the valve and connected piping. In this
case, the valve disk was unable to be closed quickly before reverse flow occurred.
Furthermore, it could not be shut slowly as the reverse flow hit the disk with a solid
force on the seat. According to the same studies, at least two conditions must be
met to achieve a nonslamming feature for check valves. First, the disk or closure
k
oc
Sh ve
wa
Close
Reverse
Reverse Flow
Flow Stopped
Figure 7.4 Swing check valve slamming.
Plates
Figure 7.5 A dual plate check valve.
member of the valve should have low inertia and low friction. Second, the travel
distance of the disk must be short, or the spring must assist the motion of the disk.
As shown in Figure 7.5, one alternative solution to minimizing the slamming
effect is to switch from swing check valves to dual plate check valves. The dual
plate check valve has two disks instead of one. So, each half-disk is less weighty
and less likely to slam when closing. The closing of the valve disk is accomplished
by the spring force rather than the weight of the disks, in order to reduce the slam-
ming effect and water hammering. Another advantage of dual plate check valves is
that their initial cost, total cost, maintenance, and energy costs are lower than
swing check valves. The dual check valve has two disadvantages. The first is
the fact that there is still a low-to-medium slamming effect present for this type of
valve, and the second is that it has a reduced bore, which results in a relatively
high-pressure drop and energy loss. Dual plate check valves are manufactured
according to the American Petroleum Institute (API) 594 standard, which has
the ability to have a bore or port that is approximately 80% of the area connected
to the piping.
Example 7.2 After a pump, a 12 dual plate check valve is installed in pressure
class 150 (pressure nominal of 20 bars) with a flow coefficient of 4000. According to
the American Society of Mechanical Engineers (ASME B16.34), the internal diam-
eter of the valve is standard for valves to be 304.8 mm in diameter. When the fluid
velocity inside the valve is 22 m/s, what is the pressure drop inside the valve based
on that information? What would be the pressure drop of a 100-ft pipe run if the
pressure drop was equal to 1.5? Then the pressure drop in the valve would be equal
to what length of the connected pipe?
Answer
The internal diameter of the valve is 304.8 mm or 12 . In the first step, we must
calculate the valve flow coefficient resistance by using Eq. (7.3).
d4 124
K v = 890 2 = 890 × = 1 15344
Cv 4000 2
According to Eq. (7.2), now it is possible to calculate the head loss in terms of the
length of the water column.
KvV 2 1 15 × 222
ΔH = = = 28 4539 m = 93 35 ft
2g 2 × 9 81
The water at a 10-m depth has a pressure of 1 bar. So, the loss of pressure in the
valve is 2.84 bar.
1 15344
Equivalent pressure loss in the piping = × 100 ft = 76 90 ft
15
The pressure drop produced by a 12 dual plate check valve in pressure class 150
is equal to 76.90 ft or 23.44 m.
In addition to the issues mentioned earlier, there are several additional issues
associated with dual plate check valves, the most significant ones being pressure
drop and medium slamming behavior. Consequently, it has become common to
use axial flow check valves or nozzle check valves after pumps and compressors
in Norwegian offshore projects. There are several characteristics of axial flow

check valves that make them nonslamming, with a very low-pressure drop com-
pared to dual plate check valves and a fast closing characteristic. Even though axial
check valves are more expensive than swing and dual plate check valves initially,
they can reduce pressure drops, protect rotating equipment from damage, and save
a lot of energy in the long run. These valves are widely used in offshore platforms,
subsea, refineries, pipelines, LNG, and petrochemical plants in the oil and gas
industry. Its short axial disk travel to the seat, spring-assisted design, and low-mass
disk contribute to the quick closing of these valves. Fast closing reduces the risk of
equipment damage caused by backflow and protects expensive mechanical equip-
ment. Figure 7.6 illustrates a nozzle check valve with a low-mass disk that is ener-
gized with a spring that can be depressed easily by the hand of the operator to keep
the valve closed. Another advantage of axial check valves is their robust design,
which can withstand perfectly the loads and vibration generated by upstream
equipment and facilities such as pumps and compressors. Typically, axial check
valves are fabricated from a single piece without any flange connections, maximiz-
ing the valve’s resistance to loads and minimizing leakage risk.
While the initial or purchasing cost of an axial check valve is higher than both
dual plate and swing check valves, it is the cheapest of the three when considering
the total cost of valves. The total cost of valves is not only determined by the initial
cost but also by the energy and maintenance costs. Compared to the dual plate and
swing check valves, axial check valves require very little maintenance and provide
less pressure loss. The cost estimation for 12 check valves over 40 years is pre-
sented in Table 7.1.
Figure 7.6 A nozzle check valve.

Table 7.1 Comparison of the costs of 12 check valves over 40 years of operation
in a plant.
Type of Purchase Maintenance

check valve (initial) cost Energy cost cost Total cost
Nozzle or axial 8000 USD 16,000 USD 1,000 USD 25,000 USD
Dual plate 3000 USD 22,000 USD 6,000 USD 31,000 USD
Swing 5000 USD 31,000 USD 20,000 USD 56,000 USD
Example 7.3 The dual plate check valve is replaced by the nozzle check valve in
the previous example. The flow coefficient resistance of the nozzle check valve is
0.83. Based on its flow coefficient resistance, how many meters of the pipe length
would be lost due to the nozzle check valve?
Answer
KvV 2 0 83 × 222
ΔH = = = 20 475 m = 67 175 ft
2g 2 × 9 81
Each 10 m of water has a pressure of 1 bar. So, the valve loses 2.04 bars of
pressure.
0 83
Equivalent pressure loss in the piping = × 100 ft = 55 33 ft
15
A 12 nozzle check valve in pressure class 150 produces a pressure drop equal to
55.33 ft or 16.86 m of pipe.
Based on Eq. (7.4), the head loss from valves, including check valves, is con-
verted to the electrical power and cost of a pump required to overcome the pressure
or head loss from the valve. Two main components typically determine the cost of
electrical energy; a demand charge and an energy charge. The energy charge repre-
sents the consumption of kilowatt-hours of electricity with a fee of approximately
0.05 USD/kWh. In contrast, the demand charge can be charged at a higher rate and
represents the cost of electricity generation capacity, with a cost of approximately
10 USD/kW. Demand charges can be affected by the time of day.
Energy Cost Due to the Valve Head Loss

1 65 × Q × ΔH × Sg × C × U
A= 74
E
where:
A: Annual energy cost, dollars per year;

Q: Flow rate, gallon per minute (gpm);
ΔH: Head loss, feet of water;
Sg: The specific gravity (SG) of fluids in the piping system, including the valves
(dimensionless), and the SG for water is equal to one;
C: Cost of electricity in USD per kilowatt per hour;
U: An electricity usage percentage that is equal to 1 if the pump is running 24 hours
a day;
E: Pump efficiency which is typically assumed equal to 0.8.
Example 7.4 Calculate the head loss for swing check valves, dual plate check
valves, and nozzle check valves in 12 sizes. Flow resistance coefficients k are
respectively 1.58, 1.15, and 0.83 for swing, dual plate, and nozzle check valves.
The fluid in the piping is the oil with a density of 800 kg/m3, a velocity of 12.76
ft/s, and a flow rate of 4500 gpm. Calculate the annual energy costs in three cases
of using each type of valve, assuming that electricity is used 50% of the time at the
cost of 0.05 USD/kWh?
Answer
According to Eq. (7.2), the first step is to calculate the head loss in terms of the
water column length.
KvV 2
Loss of pressure in the swing check valve = ΔH =
2g
2
1 58 × 12 76
= = 3 9946 ft water column
2 × 32 2
As each foot of water can generate 0.43352 lb of pressure, the pressure generated
by a column of 3.9946 ft is 1.73 psi.
K v V 2 1 15 × 12 762
Loss of pressure in the dual plate check valve = ΔH = =
2g 2 × 32 2
= 2 9074 ft water column = 1 26 psi
KvV 2
Loss of pressure in the nozzle check valve = ΔH =
2g
0 83 × 12 762
= = 2 0984 ft water column = 0 9097 psi
2 × 32 2
The annual energy cost of using a swing check valve A
1 65 × Q × ΔH × Sg × C × U 1 65 × 4500 × 3 9946 × 0 8 × 0 05 × 0 5
= =
E 08
= 741 50 USD
The annual energy cost of using a dual plate check valve A

1 65 × Q × ΔH × Sg × C × U 1 65 × 4500 × 2 9074 × 0 8 × 0 05 × 0 5
= =
E 08
= 539 69 USD
The annual energy cost of using a nozzle check valve A
1 65 × Q × ΔH × Sg × C × U 1 65 × 4500 × 2 0984 × 0 8 × 0 05 × 0 5
= =
E 08
= 389 52 USD
In this example, the density of oil is 800 kg/m2. So, the SG of oil is equal to 0.8
when calculating energy costs.
7.3 Water Hammering Calculations
Joukowsky’s formula is used to measure water hammering to calculate the pres-

sure change that results from a rapid velocity change, as shown in Eq. (7.5). Based
on this equation, it is evident that the greater the magnitude of the velocity change,
the greater the magnitude of the pressure change, as well as the wave load
and speed.
Water Hammering Evaluation as a Column of Fluid Length Pressure

According to Joukowsky’s Formula
ΔQ × C
ΔH = 75
gA
where:
ΔH: Fluid (water) pressure change expressed as water column (m, ft);
ΔQ: Flow rate change (m3/s, ft3/s);
C: Velocity of the pressure wave (m/s, ft/s) calculated from Eq. (7.8);
g: Acceleration due to gravity (9.81 m/s2, 32.2 ft/s2);
A: Pipe area (m2, ft2).
Example 7.5 Due to the closing of a swing check valve, the velocity of the water
fluid in the piping system is suddenly changed from 3 to 0, causing a wave speed of
1100 m/s. What is the increase in pressure due to the change in velocity?
Answer
In piping, the change in flow rate is dependent on the change in fluid velocity and
the area of the pipe. Here is a summary of the calculation of water hammer pres-
sure changes as a function of water column length using Eq. (7.6):
ΔQ × C ΔV × A × C ΔV × A × C
ΔQ = ΔV × A ΔH = = =
gA gA gA
ΔV × C 3 × 1100
= = = 336 39 m
g 9 81
One bar of pressure is produced by every 10 m of water. So, 33.64 bars of pressure
are produced by 336.39 m of water. Due to an increase in fluid velocity of 3 m/s, the
water pressure is increased by 33.64 bar.
Calculation of Water Hammering Through the Use of a Column of Fluid

Pressure by Considering Velocity Changes or Reverse Velocity
V ×C
ΔH = 76
g
where:
ΔH: Fluid (water) pressure change in the form of a water column (m, ft);
V: Reverse velocity or change in velocity (m/s, ft/s);
C: Pressure wave velocity (m/s, ft/s) calculated from Eq. (7.8);
g: Acceleration due to gravity (9.81 m/s2, 32.2 ft/s2).
Note: Based on field experience, a water hammer that occurs in the range of
15–30 m of water column equals approximately 50–100 ft representing a mild slam
that is tolerable. In contrast, water hammers over 30 m of the water column, or
over 100 ft, are significantly loaded and dangerous and therefore require the selec-
tion of a different check valve or the modification of the existing valve design to
incorporate a spring or dashpot.
Equation (7.7) measures the water hammering effect for various fluids in the
form of maximum pressure variation. A pressure increase in a pipe due to water
hammering is determined by a number of factors, including the fluid properties of
density and velocity, as well as the generated wave velocity in the pipe.
Calculation of Water Hammering Based on Maximum Pressure Variation

by Incorporating Velocity Change or Reverse Velocity
ΔPmax = ρ × C × V 77
where:
ΔPmax: The maximum pressure increase in the pipe as a result of water hammer-
ing (Pa);
ρ: Maximum density of the fluid (kg/m3);
C: Wave velocity (m/s);

V: A reverse velocity or a change in velocity in a pipe (m/s).
A wave speed or acoustic velocity produced in the pipe due to water hammering
can be calculated using Eq. (7.8). As was mentioned earlier, the wave speed is
much faster than that of the liquid.
Wave Speed or Acoustic Velocity in the Pipe
Eb
C= if the pipe is rigid 78
ρ
Ec
C= if the pipe is elastic
ρ
where:
1 1 Dk
= +
Ec Eb Ep t
where:
C: Wave velocity due to water hammering (m/s or ft/s);

ρ: Water density (1000 kg/m3 or 0.01347 slug/in.3);
Eb: Bulk modulus of water (2.1 × 109 N/m2 or 3.0 × 105 psi);
Ec: Effective bulk modulus of water in the elastic pipe (N/m2, psi);
Ep: Modulus of elasticity of pipe material (N/m2, psi);
t: pipe wall thickness (m/ft);
D: pipe diameter or NPS (m/ft);
k: This factor is influenced by the pipe anchorage method and Poisson’s ratio ε of
the pipe material. Depending on the piping material, Poisson’s ratio can vary
from 0.25 to 0.35. However, it is common to standardize Poisson’s ratio at 0.25.
5
k= − for pipes free to move longitudinally;
4
k = 1 – 2 for pipes anchored at both sides;
k = 1 – 0.5 for pipes with expansion joints
From Table 7.2, the modulus of elasticity Ep can be determined based on the
piping material.
In general, when a piece of material is stretched in one direction, it tends to
become thinner in the lateral direction – and if a piece is compressed in one direc-
tion, it tends to become thicker in the lateral direction. In Physics, Poisson’s ratio is
defined as the ratio between the relative contraction strain (transverse, lateral, or
Table 7.2 Modulus elasticity Ep of piping material.
Pipe material Ep (N/m2) Ep (psi)
Gray cast iron 1.1 × 1011 16 × 106

11
Malleable cast iron 1.6 × 10 23 × 106
Reinforced concrete 1.6 × 1011 25 × 106
10
Copper 9.7 × 10 14 × 106
Steel 1.9 × 1011 28 × 106
radial strain) to the applied load and the relative extension strain (or axial strain) in
the direction of the applied load. The term strain refers to the deformation of a
material as a result of stress. It is simply a ratio of the changed length to the original
length.
Typically, the sound waves generated in the piping exacerbate the slamming
effect in the valve. Before the pressure wave is reflected from the end of the pipe
and returned to the valve, it cannot affect the pressure process within the valve and
the elevated slamming action. In order to measure the time it takes the sound wave
to return to the valve, Eq. (7.9) is used. To prevent water hammering, it is imper-
ative to select and design the valve in a way that it can be closed once the sound
wave returns to the valve.
The Maximum Amount of Time It Takes for a Sound Wave to Cause

Valves to Slam and Water Hammering
L
t=2× 79
c
where:
t: The time required for the sound wave to return (s), and the valve must be closed
after that period;
L: Pipe length (m);
C: Wave speed or velocity (m/s).
Example 7.6 Figure 7.7 shows the dynamic characteristic of a dual plate check
valve installed after a pump in the water containing piping. In the figure, the hor-
izontal axis represents the deceleration of fluid flow after the pump is stopped,
expressed in feet per second squared. A vertical axis represents the maximum
reverse velocity or velocity change, expressed in feet per second. Following the
pump stoppage, fluid deceleration is approximately 40 ft/s2. What is the pressure
rise in the steel pipe caused by the pump stoppage as well as the slamming of the
2.0
1.8
1.6
Reverse velocity, ft/s
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0 10 20 30 40 50 60
Deceleration, ft/s2
Figure 7.7 Test-based dynamic characteristic of a dual plate check valve.
dual plate check valve if the wave velocity in the steel pipe is 3200 ft/s? Is the
amount of calculated slamming acceptable? In this instance, what is the water
hammering strategy?
Answer
The reverse flow velocity associated with the fluid deceleration of 40 ft/s2 is 1.2 ft/s
according to Figure 7.7. The water hammering can now be calculated in terms of
water column length according to Eq. (7.6) as follows:
V ×C 1 2 × 3200
ΔH = = = 119 25 ft
g 32 2
There is more than 100 ft of water hammer, which indicates dangerous slam-
ming and hammering that should be avoided by changing the valve type or mod-
ifying its design (e.g. adding a dashpot). A foot of water has a pressure of 0.43352
psi. So, the amount of water pressure caused by 119.25 ft of water is 51.70 psi.
Equation (7.7) may be used as an alternative method to calculate the increase
in pressure in the pipe caused by water hammering. It is important to note that
the water density is 1000 kg/m3.
ΔPmax = ρ × C × V
= 1000 kg m3 × 3200 ft s × 0 3048 m ft × 1 2 ft s × 0 3048 m ft
= 356, 747 6736 Pa = 51 74 psi
Example 7.7 The swing check valve is installed in a piping system handling oil
with an SG of 0.8. Upon closure of the valve, a reverse flow velocity of 1 m/s and a
wave velocity of 1100 m/s are produced. Calculate the increase in oil pressure. If
the wave sound must travel through an 1110-m pipe and return to the valve, what
is the minimum time for the valve to be closed in order to prevent water
hammering?
Answer
According to Eq. (7.7), oil pressure increases due to reverse flow velocity as follows:
ΔPmax = ρoil × C × V = SG × ρwater × C × V

= 0 8 × 1000 × 1100 × 1 = 880, 000 Pa
= 8 8 bar = 127 63 psi
The next step is to calculate the length of time necessary for the pressure wave to
traverse the pipe and return to the valve based on Eq. (7.9).
L 1100
t=2× 2× = 2 seconds
c 1100
In order to prevent water hammering, the valve must be closed after the arrival
of the sound wave. Therefore, the valve must be closed after two seconds.
Using Eq. (7.10), it is possible to calculate the water hammer pressure in a piping
system due to sudden valve closure:
Produced Water Hammer Pressure by Knowing the Valve Closing Time
ΔV × L
ΔP = 0 070 7 10
Δt
where:
ΔP: An increase in pressure or a spike in pressure (psi);

ΔV: Flow velocity change (ft/s);
Δt: The closing time of the valve (second);
L: Length of the upstream pipe (ft).
1 ft = 0 3048 m;
1 ft s = 0 3048 m s;
1 psi or lb in2 = 6894 8 Pa or N m2
Example 7.8 Calculate the pressure increase caused by the closure of a valve in a
pipe in a time of two seconds, resulting in a decrease in velocity from 20 to 15 m/s.
The pipe upstream of the valve is 20 m in length.
Answer
Except for valve closure time, all the given parameters must be converted from SI
units to imperial units first.
Initial velocity = 20 m s = 6 096 ft s
Secondary velocity = 15 m s = 4 572 ft s
Velocity change = 6 096 ft s – 4 572 ft s = 1 524 ft s
The length of the pipe upstream of the valve = 20 m = 6.096 ft
ΔV × L 1 524 × 6 096
ΔP = 0 070 = 0 070 = 0 32 psi
Δt 2
7.1 Which of the following statements about water hammering is true?

A A very common problem with ball and butterfly valves is water hammering.
B The devastating effects of water hammering are restricted to piping
systems, including valves.
C Water hammering is directly related to the pipe area.
D Water hammering is directly proportional to changes in pressure and
wave velocity.
Answer
Option A is incorrect since water hammering occurs most frequently in
check valves and not in ball and butterfly valves. Option B is not acceptable
either, as the undesirable effects of water hammering are not limited to pip-
ing systems, as it can also harm pumps. According to Joukowsky’s formula,
option C does not hold true since increasing the pipe area through different
methods, such as increasing the wall thickness, reduces the severity of water
hammering. The correct option is D.
7.2 A swing check valve is installed after a pump on a water-handling piping

system. The speed of water inside the valve is 20 m/s. The port size of the
valve is 80% of the connected pipe, creating a flow resistance coefficient of
1.58. To reduce this flow resistance coefficient and energy loss in the swing
check valve, a nozzle check valve is installed with a flow resistance
coefficient of 0.83. If the fluid velocity is the same in both types of valves, how
much pressure drop is reduced by switching from swing to nozzle check
valves?
A 1.5 bar
B 2 bar
C 0.5 bar
D 3 bar
Answer
Equation (7.2) can be used to calculate the head loss of the valves as follows:
Drop in pressure at the swing check valve

KvV 2 1 58 × 202
ΔH = = = 32 21 m
2g 2 × 9 81
Every 10 m of water produces 1 bar of pressure. So, 32.21 m of water causes a

3.21 bar loss in the swing check valve.
Drop in nozzle check valve pressure

KvV 2 0 83 × 202
ΔH = = = 16 92 m
2g 2 × 9 81
A 10-m length of water generates 1 bar of pressure, and thus 16.92 m of water
causes 1.692 bar of pressure loss in the nozzle check valve.
The difference in pressure loss generated by the swing and axial nozzle
check valves = 3.21 − 1.692 = 1.518 bar. Therefore, the most appropriate
option is A.
7.3 According to the following assumptions, how much USD is saved annually
by using a nozzle check valve instead of a swing check valve after a pump:
1) The fluid velocity is 20 m/s and the flow rate is 8000 gpm.
2) The fluid service in the pipes and valves is water.
3) Electricity is used at a rate of 70% and the cost of electricity is 0.06
USD/kWh.
4) The pump located upstream of the check valve has an efficiency of 0.8.
5) For swing and nozzle check valves, respectively, the flow coefficient resis-
tances are 1.58 and 0.83.
A 400 USD
B 3200 USD
C 30,000 USD
D 28,000 USD
Answer
As a first step, calculate the head loss difference between the swing and axial
nozzle check valves by applying Eq. (7.2) as follows:
KvV 2 1 58 − 0 83 × 202
ΔH = = = 15 29 m = 50 16 ft water
2g 2 × 9 81
Using a nozzle check valve rather than a swing check valve will
result in areduction of energy costs per year A
1 65 × Q × ΔH × Sg × C × U
=
E
1 65 × 8000 × 50 16 × 0 8 × 0 06 × 0 7
= = 27, 808 70 USD
08
The best answer is choice D.
7.4 A series of experiments were conducted on an 8 CL1500 dual plate and noz-
zle check valve installed after a pump in the water system in order to meas-
ure water hammering generated as a result of the valve closure. The reverse
flow velocity generated by stopping the pump is 0.75 m/s for the dual plate
valves and 0.3 m/s for the nozzle check valves. What is the correct answer?
(Note: assume a wave velocity of 3200 ft/s.)
A In this case, installing a nozzle check valve after the pump is recommended.
B The pressure increase caused by water hammering in the dual plate
check valve is 90 psi.
C In comparison with the axial check valve, the severity and value of water
hammering in dual plate check valves are more than three times higher.
D Water hammering causes more than 50 psi pressure increase in the axial
check valve.
Answer
Equation (7.6) can be used to calculate the water hammering in terms of the
water column length as follows:
Case 1
V ×C
Water hammering in the dual plate check valve ΔH =
g
0 75 m s × 3 28 ft m × 3200 ft s
=
32 2 ft s2
= 244 47 ft Pressure increase = 244 47 × 0 43352 = 106 psi
Case 2
V ×C
Water hammering in the axial check valve ΔH =
g
0 30 m s × 3 28 ft m × 3200 ft s
=
32 2 ft s2
= 97 79 ft Pressure increase = 97 79 × 0 43352 = 42 39 psi
Based on field experience, a water hammer within a range of 50–100 ft is

acceptable. Water hammering in the dual plate check valve is equivalent
to a water column of 97.79 ft; therefore, it is acceptable. However, water
hammers that exceed 100 ft are not tolerated and are considered dangerous.
Therefore, the water hammer created by the dual plate check valve is not
acceptable, and the dual plate check valve is not a suitable valve choice in
this application. There is an error in option B since the water hammering
in the dual plate check valve causes a pressure increase of 106 psi, not 90
psi. The intensity of water hammering in a pipe as a result of a dual plate
check valve closing compared to an axial check valve closing = 106/
42.39 = 2.5 that is less than three. So, option C is incorrect. In addition,
option D is incorrect since water hammering increases the pressure in the
axial check valve by 42.39 psi, which is less than 50 psi. Therefore, option
A is the correct answer.
7.5 Figure 7.8 illustrates the dynamic characteristics of various types of check
valves as determined by the results of experimental testing. Based on the
given chart, which is the correct statement regarding the risk of valve slam-
ming and water hammering?
A In the case of a reverse flow velocity of 0.1 m/s, the risk of water ham-
mering is exceptionally high for swing and dual plate check valves.
B Tilted check valves are more susceptible to slamming and water ham-
mering than swing check valves.
C The deceleration of 5 m/s squared gives mild slamming for swing check
valves but no slamming for nozzle check valves.
D A dual plate check valve will slam mildly with a reverse fluid velocity of
0.45 m/s.
Answer
Based on the chart, option A is incorrect since 0.1 m/s reverse flow velocity
does not cause slamming or water hammering for both dual and nozzle
check valves. Option B is also not correct since the slamming and hammer-
ing associated with a tilted check valve is less severe than that of a swing
check valve. The correct choice is option C. Option D is incorrect as the
reverse fluid velocity of 0.45 m/s would result in severe slamming of a dual
plate check valve.
m/s2
0 5 10 15
2.0 Tilted disc 0.6
Swing
check check valve
1.8 Ball
Severe
check 0.5
slam
1.6 Swing-flex Dual disc
Reverse velocity, ft/s
check valve check valve

1.4
0.4
1.2
Surgebuster
m/s
1.0 0.3
slam
Mild
0.8 Silent
check valve 0.2
0.6
0.4
0.1
slam
Nozzle
No
check
0.2
0.0
0 10 20 30 40 50 60
Deceleration, ft/s2
Figure 7.8 A comparison of the dynamic characteristics of various types of check valves.
7.6 A check valve closure produces a wave velocity of 1100 m/s in the piping sys-
tem. What is the minimum duration of time for a valve to be closed in order
to prevent a slamming effect in the piping system if the length of pipe the
produced wave must travel and return is 2200 m?
A One second
B Two seconds
C Three seconds
D Four seconds
Answer
L 2200
t=2× 2× = 4 seconds
c 1100
To prevent water hammering, the valve must be closed after the arrival of the
sound wave, which means that it shall be closed after four seconds. There-
fore, option D is the correct answer.
7.7 A check valve closes in one second, reducing the velocity of water in a piping
system from 10 ft/s to zero. Based on the length of the downstream pipe being
50 ft, how much pressure is generated by the water hammering effect in the
piping system?
A 30 psi
B 35 psi
C 40 psi
D 45 psi
Answer
For this particular case, Eq. (7.10) is applied to calculate the water hammer-
ing as follows:
ΔV × L 10 ft s × 50 ft
ΔP = 0 070 = 0 070 × = 35 psi
Δt 1 second
7.8 A steel pipe has a diameter of 18 in. and a wall thickness of 2 in., and it is
5000 ft long on a uniform slope. A pipe transports water from a reservoir
and discharges it to the atmosphere at an elevation of 150 ft below the reser-
voir’s free surface. There is a valve installed at the downstream end of the
pipe, which allows for a flow rate of 25 ft3/s. Figure 7.9 illustrates a schematic
of the valve and the sloped pipe that discharges the water into the environ-
ment. What is the maximum water hammer pressure generated at the valve
if the valve closes in 1.4 seconds? Is the valve suddenly slammed and closed?
(Note: there is no longitudinal stress in the pipe, and the pipe is free to move
longitudinally throughout.)
A 857 psi due to sudden closure of the valve
B 450 psi due to slow closure of the valve
C 790 psi due to sudden closure of the valve
D 400 psi due to slow closure of the valve
Answer
The following is a summary of the given data in the question:
Steel pipe length L = 5000 ft
Reservoir
150 ft
Valve
Figure 7.9 An illustration of the valve and sloped valve that discharges water into the
environment.
Pipe diameter D = 18 in = 1 5 ft
Pipe thickness t = 2 in = 0 166667 ft
Flow rate Q = 25 cfs
Valve closing time t = 1 4
Firstly, the acoustic wave velocity, parameter C, must be calculated. Due to
the fact that the pipe is steel and not plastic, the following equations apply:
Ec
C=
ρ
where:
1 1 Dk
= +
Ec Eb Ep t
Eb represents the bulk modulus of water, which is 3.0 × 105 psi. According to
Table 7.2, the modulus elasticity Ep of pipe material in steel equals 28 × 106.
Assuming a Poisson’s ratio of 0.25, the following equation can be used to
determine the value of k:
5
k= − = 1 25 – 0 25 = 1
4
1 1 18 in × 1
= 5 + E c = 2 74 × 105 psi
Ec 3 0 × 10 28 × 106 × 2 in
Ec 2 74 × 105 psi
C= = = 4510 ft s
ρ 0 01347
Equation (7.9) is used to calculate the maximum amount of time it takes for a
sound wave to cause valves to slam and water hammer
L 5000
t=2× =2× = 2 2 second
c 4510
The valve closes in 1.4 seconds, resulting in the sudden closure of the valve
and the hammering of water in the pipe due to the pressure wave. In order to
calculate the fluid velocity, the next step is to calculate the area of the pipe.
d2 1 52
A=π =π = 1 7670 ft2
4 4
The following equation expresses the relationship between the volumetric

flow rate, area, and velocity of fluid inside the pipe:
Q 25
Q=V ×A V= = = 14 1 ft s
A 1 7670
By using Eq. (7.7), it is possible to calculate the pressure increase in the pipe
as a result of water hammering:
ΔPmax = ρ × C × V = 1000 kg m3 × 4510 ft s × 0 3048 m ft × 14 1 ft s

× 0 3048 m ft = 5, 907, 797 Pa = 857 psi
Further Reading
American Petroleum Institute (API) 594 (2004). Check Valves: Flanged, Lug, Wafer and
Butt-welding, 6e. Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) B16.34 (2017). Valves – Flanged,
Engineers (ASME).
Ballun, J.V. (2007). A Methodology for Predicting Check Valve Slam. American Water
Works Association (AWWA).
Ford, R. (2014). Power industry applications: A valve selection overview. Valve World
Magazine 19 (8): 96–103.
Gustorf, H. and Root, P. (2016). Developments in axial valve design. Valve World
Magazine 21 (10): 49–53.
ITT Water & Wastewater AB (2019). Hydraulic Transient Analysis: Preventing Water
Hammer. ITT Water & Wastewater AB: Sundbyberg.
Kruisbrink, A. (2010). The need for dynamic characteristics of check valves. Valve
World Magazine 15 (9): 65–66.
Oxford: Elsevier.
NORSOK L-001 (2017). Piping and Valves. Revision 4. Lysaker: NORSOK.
Norwegian Oil Industry Association (2013). Valve Technology. 2nd revision. Norsk
Olje & Gass.
Oxler, G. (2009). Non-return valve and/or check valve for pump system – a new
approach. Valve World Magazine 14 (4): 75–77.
Provoost, G.A. (1982). The dynamic characteristic of non-return valves. Conference
Paper Submitted to 11th Symposium of the Section of Hydraulic Machinery, Equipment
and Cavitation, Netherlands (13–17 September 1982).
Further Reading 257
Schmitz, H. (2018). What is water hammer and how do I fix it? Livintanor. https://
livinator.com/whats-water-hammer-and-how-do-i-fix-it/ (accessed 23 March 2022).
Sotoodeh, K. (2015). Axial flow nozzle check valves for pumps and compressors
protection. Valve World Magazine 20 (1): 84–87.
Sotoodeh, K. (2018). Comparing dual plate and swing check valves and the importance
of minimum flow for dual plate check valves. American Journal of Industrial
Engineering. 5 (1): 31–35. https://doi.org/10.12691/ajie-5-1-5.
Sotoodeh, K. (2021). Analysis and failure prevention of nozzle check valves used for
protection of rotating equipment due to wear and tear in the oil and gas industry.
Journal of Failure Analysis and Prevention https://doi.org/10.1007/s11668-021-
01162-2.
VALMETALIC (2018). Design and selection of check valves. https://www.valmatic.
com/Portals/0/pdfs/DesignSelectionCheckValves.pdf (accessed 22 March 2022).
258
Safety Valves
8.1 Introduction
Pressure safety valves (PSVs) and pressure relief valves (PRVs) are collectively
referred to as safety valves, which are used in all pressurized equipment and facil-
ities such as piping, pumps, compressors, turbines, and boilers when overpressur-
izing can occur. The principles of protection from overpressurizing are briefly
presented in Chapter 2. In addition, the sizing of safety valves is covered in full
detail in Chapter 2. The objective of this chapter is to provide an overview of essen-
tial design considerations and calculations for PSVs and PRVs, including the relief
of pressure and capacity as well as generating reaction forces. There is a noticeable
sound made by these valves during operation, which can be measured using the
methods and calculations described in Chapter 6. For whatever reason, if some-
thing goes wrong with the pressure piping or equipment which leads to overpres-
sure scenarios, the excessive pressure can damage or burst the expensive
equipment and piping. PRVs and PSVs are installed over pipes and equipment
to prevent undesirable events and enhance safety and reliability. If the pressure
inside the equipment or piping exceeds the allowable limit, the safety valve auto-
matically opens and releases the extra pressure. The result is a reduction in pres-
sure across the protected equipment or piping. Both PSVs and PRVs are sometimes
referred to as spring valves. There are two main differences between a PSV and a
PRV; the first is that PSVs are typically used on gas services as well as steam ser-
vices. However, PRVs are primarily used to provide overpressure protection for
liquid handling systems. Another main difference between these two types of
valves is that a PRV opens proportionally to the increase in fluid pressure, while
a PSV opens suddenly. As shown in Figure 8.1, three PSVs are installed on the pip-
ing system to protect it in case of an overpressure situation. Three PSVs are used in
lieu of one for safety and reliability reasons. In fact, if one or two PSVs malfunction
and require maintenance, the third one may be utilized. Three ball valves are
PSVs
Ball valves
Piping
Figure 8.1 Three PSVs installed on piping for overpressure protection.

Source: Pichitstocker/Adobe Stock.
located upstream of PSVs to isolate the piping from the downstream PSV when it is
removed for maintenance.
8.2 Safety Valve Parts
Figure 8.2 depicts a conventional safety valve, which features the following essen-
tial parts: The main pressure-containment element is a body made of carbon steel
or another corrosion-resistant alloy such as stainless steel. In valves, pressure-
containing components are those components whose malfunction could cause
leakage from the valve into the environment. It is important to note that the fluid
enters the valve through a port at the bottom, and it exits through the left port fol-
lowing a 90 rotation.
8.3 Safety Valve Design and Operation
8.3.1 Design and Operation Parameters

Three primary standards are relevant to the design and operation of safety valves.
The first is American Petroleum Institute (API) 520, Sizing, Selection, and Instal-
lation of Pressure-Relieving Devices Part 1 – Sizing and Selection, the second is
260 8 Safety Valves
Bonnet
Adjusting nut
Adjusting screw
Seal
Sealing
Spring washer
Spring
Pin
Jacket
Valve guide
Stop
Disc valve
To exhaust
Lift ring
system
Adjusting ring
Seat
Body
Figure 8.2 An essential component of a conventional safety valve.
API 526, Flanged Steel Pressure-Relief Valves, and the third is API 521,
Pressure-Relieving and Depressuring Systems. In determining how safety valves
should be designed, it is critical to determine what factors can increase pres-
sure inside protected piping and facilities. It is also critical to determine how
much each factor can increase pressure inside protected components and
equipment. Table 8.1 summarizes the most relevant parameters for the design
of a safety valve.
The following are some terminologies associated with safety valve design
parameters:
Relieving pressure: It is also referred to as the inlet flow pressure as indicated
by parameter P1 and can be calculated using Eq. (8.1). Using Eq. (8.2), this pressure
can be expressed as absolute relieving pressure. It is equal to the set pressure plus
the allowable overpressure plus atmospheric pressure minutes loss pressure. The
relieving pressure is equal to the test pressure.
Table 8.1 Factors that should be considered in the design of safety valves.
1) Fluid properties 3) Relieving conditions

A) Type of fluid service and its state A) Required relieving
B) Molecular weight capacity
C) Viscosity B) Set pressure
D) Specific gravity C) Allowable overpressure
E) Ratio of specific heats D) Superimposed backpressure
F) Compressibility factor E) Buildup backpressure
F) Relieving temperature
2) Operating conditions
A) Operating pressure
B) Operating temperature
C) Maximum allowable
working pressure (MAWP)
Calculation of Safety Device Inlet, Upstream, or Relieving Pressure
P1 = Pset + Pover − Ploss 81
Calculation of Safety Device Inlet, Upstream, or Absolute Relieving

Pressure
P1 = Pset + Pover + Patmosphere − Ploss 82
The maximum allowable working pressure (MAWP) is the highest pressure at

which a device or equipment can be safely operated without any change in over-
pressure scenarios. The MAWP can be considered equal to the design pressure of
the equipment, according to some valve literature.
A set pressure refers to the level of increasing static inlet pressure at which a PRV
opens or leaks. In general, the set pressure is equal to the valve opening pressure,
the popping pressure, or the start of leakage pressure. A bench test is conducted on
a PSV to determine set pressure and seat tightness at ambient temperature by
adjusting the valve spring. The first step in designing a safety valve is determining
a suitable set pressure, which is typically expressed as a percentage of MAWP. The
set pressure could be less, equal, or greater than the MAWP.
Overpressure or allowable overpressure is the pressure increase over the
set pressure of a safety valve, which is usually expressed as a percentage.
262 8 Safety Valves
Table 8.2 Set pressure and maximum accumulated pressure limits for safety valves.
Single device installation Multiple device installation
Maximum
Maximum Maximum set Maximum
set accumulated pressure accumulated
pressure (%) pressure (%) (%) pressure (%)
Nonfire case
First relief device 100 110 100 116
Additional device — — 105 116
Fire case
First relief device 100 121 100 121
Additional device — — 105 121
Supplement device — — 110 121
All values are expressed as a percentage of the MAWP.
The overpressure scenario may occur in both the protected pressure equipment
and the safety valve.
The maximum accumulated pressure is calculated using Eq. (8.3) by adding the
set pressure and the allowable overpressure.
Calculation of Maximum Accumulated Pressure
Maximum accumulated pressure = Set pressure + Allowable overpressure

83
In accordance with ASME codes, Table 8.2 summarizes the maximum accumu-
lation and set pressure for PRVs. Based on Table 8.2, two conditions are considered
for overpressure protection of pressurized equipment and piping: fire conditions
and nonfire conditions. Furthermore, one or more safety devices may be installed
on the pressurized equipment.
Example 8.1 For a nonfire case, a single safety valve is installed on a piping
system with a MAWP of 100 psi. Given the maximum values of set pressure
and maximum accumulated pressure, what would be the values of allowable
overpressure and absolute relieving pressure assuming that there is no pressure
loss in the valve?
Answer
The maximum set pressure for a single safety valve in nonfire conditions is
MAWP = 100 psi (see Table 8.2). According to the same table, the maximum
accumulated pressure for this case is approximately 1.1 times the MAWP of
110 psi. Knowing the set pressure and maximum accumulated pressure, it is
possible to calculate overpressure as per Eq. (8.3) as follows:
Maximum accumulated pressure = Set pressure + Allowable overpressure
Allowable overpressure = Maximum accumulated pressure − Set pressure =
110−100 = 10 psi
Thus, the maximum allowable overpressure is 10% of the set pressure. The next
step is to calculate the absolute relieving pressure using Eq. (8.2) as follows:
P1 = Pset + Pover + Patmosphere − Pover + Patmosphere − Ploss

= 100 + 10 + 14 7 = 124 7 psi
Example 8.2 The top of a pressure vessel is equipped with two safety valves in
the event of a fire. If MAWP = 100 psi, and the set and accumulated pressure
values are at their maximum, calculate the absolute relieving pressure for both
valves. (Note: no pressure loss inside both valves.)
Answer
The maximum set pressure for the first valve in the fire case = MAWP = 100 psi
The maximum set pressure for the second valve in the fire case = 105%
MAWP = 105 psi
The maximum accumulated pressure for the first and second valve in the fire
case = 121% MAWP = 121 psi
Allowable overpressure (first valve) = Maximum accumulated pressure −
Set pressure = 121 − 100 = 21 psi
Allowable overpressure (second valve) = Maximum accumulated pressure −
Set pressure = 121 − 105 = 16 psi
Absolute relieving pressure first valve = Pset + Pover + Patmosphere − Ploss

= 100 + 21 + 14 7 = 135 7 psi
Absolute relieving pressure second valve = Pset + Pover + Patmosphere − Ploss
= 105 + 16 + 14 7 = 135 7 psi
Backpressure refers to the statistical outlet pressure of the safety valve, which can
either be fixed or variable. In plain English, backpressure is defined as a pressure
that exists at the outlet of a safety valve that affects the opening pressure and the
valve capacity. Furthermore, for a conventional safety valve subjected to constant
264 8 Safety Valves
backpressure, the set pressure is effectively reduced by an amount proportional to

the backpressure. Due to this condition, a pressure increase equal to the backpres-
sure should be applied to the required set pressure in order to compensate. In this
case, the initial set pressure plus the constant backpressure is considered the
adjusted set pressure. In the event of variable backpressure, the effective set pres-
sure will be dependent on the fluctuations in the backpressure. The effects of back-
pressure on a safety valve depend on many factors such as the type of backpressure
(fixed or variable), the valve position (open or closed), the fluid phase (gases,
vapors, or liquids), and the type of valve construction (balanced bellows or conven-
tional). Backpressure is equal to the sum of superimposed and accumulated
backpressures.
Superimposed backpressure is a static pressure that exists at the outlet of a pres-
sure relief device before and during the PSV operation, which is caused by other
pressures in the discharge system. As an example, if a PSV outlet is attached to a
closed system with a pressure of 2 psi, the superimposed backpressure is 2 psi. If
the discharge outlet of the PSV is connected to the atmosphere, the superimposed
backpressure would be 0 psi.
Buildup backpressure refers to the existing pressure at the outlet of a pressure
relief device resulting from the flow passing through the safety device into a dis-
charge system. Buildup backpressure is affected by the flow rate and the flow
phase and varies from one safety valve to another as well as in different situations.
Later in this chapter, more information about methods for calculating buildup
pressure will be discussed.
Backpressure correction factor Kb or KB is the ratio of the safety valve capacity
with backpressure included to the rated valve capacity without backpressure that
is normally applicable for valves that handle vapors and gases calculated in accord-
ance with Eq. (8.4).
Backpressure correction factor Kb or KB is the ratio of the safety valve capacity
with backpressure included C1 to the rated valve capacity without backpressure
C2 that is typically applicable for valves handling vapors and gases, based on
Eq. (8.4). It is not possible to provide a standard value for the backpressure correc-
tion factor since this depends on the design of the valve. The responsibility lies
with the manufacturers to provide their own data about this parameter, which
is backed up by testing.
Calculation of Backpressure Correction Factor
C1 Safety valve capacity with backpressure included

Kb = = 84
C2 Safety valve capacity without backpressure
Calculation of the backpressure correction factor for balanced bellows safety

valves is based on the percentage of gauge backpressure calculated according
to Eq. (8.5).
Calculation of Gauge Backpressure for Bellows Safety Valves
Pb Backpressure
Percentage of gauge backpressure = × 100 = × 100
Ps Set pressure
85
It is possible to determine the value of Kb using the diagrams in Figures 8.3 and
8.4 for bellows and conventional safety valves, respectively. Bellows or balanced
bellows safety valves are those which are designed to minimize the effect of back-
pressure on the operational characteristics of the valve. As a result, the bellows
surround an area equal to the inlet of the orifice to protect it from backpressure
resulting from the discharge of the safety valve. However, it should be noted that
1.00
0.95
0.90
Backpressure correction factor, Kb
16% Overpressure
0.85
10% Overpressure
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0 5 10 15 20 25 30 35 40 45 50
Percent of gauge pressure = (PB/PS ) × 100
PB = Back pressure, in psig,
PS = Set pressure, in psig,
balanced bellows valves.
266 8 Safety Valves
1.1
1.0
k = 1.0
k = 1.2
0.9
0.87 k = 1.4
k = 1.6
k = 1.8
Kb
0.8
0.7
0.6
See
example
problem
below
0.5
40 60 76 80 100
Percent of back pressure = PB/(PS + Po) × 100 = r × 100
Kb = Back pressure correction factor,
PB = Back pressure, in psia,
PS = Set pressure, in psia,
PO = Overpressure, in psi,
conventional relief valves.
the main benefit of using a bellows safety valve is that there is no effect of back-
pressure on the relieving pressure.
In Figure 8.4, it is shown that the percentage of backpressure and the ratio of
specific heat k are used to calculate the value of Kb for conventional relief valves
handling gases or vapors. Equation (8.6) can be used to calculate the percentage of
backpressure for conventional safety valves. The ratio of specific heat is defined as
the ratio of heat capacity at constant pressure to heat capacity at constant volume.
Heat or thermal capacity is defined as the amount of heat required to change the
temperature of an object by one degree. Ratio of specific heat k for different gases
can be extracted from Table 8.3.
Calculation of Backpressure Percentage for Conventional Safety Valves

Pb
Backpressure percentage = × 100
Ps + Po
Absolute backpressure
= × 100
Absolute set pressure + Overpressure
86
Table 8.3 Ratio of specific heats “k” for different gases.
Gas Molecular weight Ratio of specific heat (k)
Air 29.96 1.40

Ammonia 17.03 1.40
Argon 40 1.31
Carbon dioxide 44.01 1.40
Carbon monoxide 28 1.40
Ethane 30.07 1.19
Ethylene 28.03 1.24
Helium 4 1.66
Hexane 86.18 1.06
Hydrogen 2.02 1.41
Hydrogen sulfide 34 1.32
Methane 16.4 1.31
N-Butane 58.12 1.09
Natural gas (specific gravity = 0.60) 18.9 1.27
Nitrogen 28 1.40
Oxygen 32 1.40
Pentane 72.15 1.07
Propane 44.09 1.13
Propylene 47.08 1.15
Steam 18.01 1.31
Sulfur dioxide 64.04 1.29
There is also a backpressure correction factor Kw used for balanced bellows pres-
sure relief or safety valves in liquid services. This value can be derived directly from
Figure 8.5 based on the percentage of gauge backpressure referred to as API 520.
Example 8.3 The set pressure in a balanced bellows safety valve is 100 psig and
the backpressure is 35 psig. What is the value of backpressure correction factor if
the maximum allowable overpressure is 10% of the set pressure?
Answer
In this case, the percentage of gauge pressure is 35, and a 10% overpressure curve
should be applied. According to Figure 8.3, the backpressure correction factor is
approximately 0.94, which indicates that the backpressure reduces the valve
capacity by 6%.
268 8 Safety Valves
1.00
0.95
0.90
0.85
0.80
Kw
0.75
0.70
0.65
0.60
0.55
0.50
0 10 20 30 40 50
Percent of gauge backpressure = (PB/PS ) × 100
Kw = Correction factor due to back pressure.
PB = Back pressure, in psig.
PS = Set pressure, in psig.
Figure 8.5 The capacity correction factor Kw of balanced bellows PRVs in liquid services
caused by backpressure.
Example 8.4 The set pressure or overpressure of a conventional safety valve are
100 psi and 10 psi, respectively. It is estimated that the superimposed and buildup
backpressures are 70 psig and 10 psi, respectively. In light of the fact that the valve
deals with carbon dioxide, what is the percentage reduction in valve capacity
caused by backpressure?
Answer
Backpressure = Superimposed backpressure + Buildup backpressure = 70 + 10 =
80 psig
Absolute backpressure = Backpressure + 14.7 = 80 + 14.7 = 94.7 psia
Pb
Backpressure percentage for the conventional safety valve = × 100
Ps + Po
Absolute backpressure 94 7
= × 100 = × 100 = 76
Absolute set pressure + Overpressure 100 + 14 7 + 10
According to Table 8.3, carbon dioxide has a specific heat ratio k of 1.4. As shown
in Figure 8.4, 0.87 represents the backpressure correction factor Kb. Due to the
backpressure, this valve’s capacity has been reduced by 13%.
Conventional relief valves do not perform satisfactorily when excessive back-

pressure develops after the safety valve opens to discharge gas or liquid. In this
instance, the backpressure tends to reduce the valve disk lifting forces, resulting
in the valve remaining open. In addition, excessive backpressure may cause the
valve to operate in an unstable manner, resulting in the valve vibrating and chat-
tering. In this case, the disk of the valve can move up and down and cause contact
with the seat of the valve, which damages its internals. In addition to changing the
set pressure, backpressure can introduce corrosive conditions into the inner cham-
ber of the safety valve. In the case of a conventional safety valve, backpressure
must not exceed 10% of the set pressure when the valve’s allowable overpressure
is 10%. When the backpressure is expected to exceed the provided limit, bellows
balanced safety valves must be used instead. A balanced bellows safety valve is
required in the second circumstance when the superimposed backpressure varies
significantly from the set pressure. When using a balanced bellows safety valve, the
total backpressure can reach a maximum of 50% of the set pressure.
Coefficient of discharge Kd is a unitless valve which is typically supplied by the
valve manufacturer and it is typically used for calculation of flow passing through
the safety valve. Based on Eq. (8.7), this factor is simply calculated by dividing the
actual flow capacity of the safety valve by the theoretical flow capacity of the valve,
and it is assumed to be equal to 0.9.
Calculation of Coefficient of Discharge

Actual flow capacity of the safety valve
Kd = 87
Theoritical flow capacity of the safety valve
There is a very rough estimation of buildup backpressure (see Eq. 8.8) that does
not fully consider important factors such as fluid properties and pressure losses
due to the discharge piping arrangement.
An Estimation of the Buildup of Backpressure

A1
Pbb = P1 × 88
A2
where:
Pbb: Buildup backpressure (psia);

P1: Pressure at the inlet or the relieved pressure (psia) calculated from Eq. (8.2);
π × Valve inlet size2
A1: Cross-sectional area of the valve inlet = in 2 ;
4
π × Valve outlet size2
A2: Cross-sectional area of the valve outlet = in 2
4
270 8 Safety Valves
Example 8.5 What is a rough estimate of backpressure buildup at a PSV 2 × 3

when the set pressure is 150 psig and 10% overpressure is permitted? (Note that
there is no pressure loss within the valve.)
Answer
According to Eq. (8.2), the inlet or relieving pressure is calculated as follows:
P1 = Pset + Pover + Patmosphere − Ploss = 150 + 15 + 14 7 = 179 7 psia
A1 = The cross-sectional area of the valve inlet =
4
π × 22
= = 3 14 in 2
4
A2 The cross-sectional area of the valve outlet =
4
π × 32
= = 7 07 in 2
4
3 14
Pbb = 179 7 × = 79 81 psia = 65 1 psig
7 07
There is a more accurate way to calculate the buildup backpressure at any ter-
minal point (t) located on the safety valve outlet piping or at the discharge point
shown in Figure 8.6 with number 2.
Lh Figure 8.6 Schematic of a safety valve for

calculating backpressure (flow safe).
t
Vent LV
stack
A2 2
n
1
A1
F7000/8000 series
typical installation
Equation (8.9) is employed to calculate the buildup of backpressure at terminal

pressure, which is point “t.”
Calculation of the Terminal Pressure at Point “t”

K d × K b × A1
Pt = k k−1 1 2
89
At × P 1 × 2 k + 1 1 Z
where:
Pt: Terminal pressure (psi);

Kd: Coefficient of discharge calculated from Eq. (8.7) (dimensionless);
Kb: Backpressure correction factor (dimensionless);
A1: Cross-sectional area of the valve inlet in 2 ;
4
At: Internal area of the outlet pipe at the point “t” (in.2);
Z: Gas compressibility factor calculated according to Eq. (8.10) (dimensionless);
k: Ratio of specific heats obtained from Table 8.3 for different gases
(dimensionless);
P1: Inlet pressure (psi) calculated according to Eq. (8.10).
Note 1
In thermal physics and thermodynamics, the ratio of specific heat is the ratio of
heat capacity at constant pressure to heat capacity at constant volume. Heat
capacity can be defined as the amount of heat necessary to change an object’s
temperature in one unit.
Note 2
In thermodynamics, the gas compressibility factor indicates how much the real
gas differs from the ideal gas at given pressure and temperature. The Z-factor of
gas is close to 1 at low pressures and high temperatures, which means that the
behavior of real gases is similar to that of ideal gases. A gaseous mixture typi-
cally consists of billions of energetic gas molecules that interact with one
another. As a result, it is extremely complex and difficult to model a real gas.
Thus, the concept of an ideal gas was developed in order to simplify the behavior
of gases. Scientists and engineers can model and predict the behavior of gases
using the theory of ideal gases. There are two main rules associated with ideal
272 8 Safety Valves
gases: the first rule states that the molecules of ideal gases do not attract or repel
one another, but are constantly in motion. In ideal gases, intermolecular forces
are negligible. The only type of interaction between molecules in an ideal gas
model is an elastic collision, which occurs when molecules strike against each
other and kinetic energy is exchanged. It should be noted, however, that the total
molecular kinetic energy inside the gas remains constant, and the kinetic energy
does not change before or after the collision. Furthermore, the molecules occupy
very little volume inside the gas. Helium is known as a type of gas that shares
many characteristics with ideal gases. In contrast to ideal gases, real gases do not
follow the laws of kinetic molecular theory. Nonideal gases are also known as
real gases. Carbon dioxide, carbon monoxide, helium, oxygen, and nitrogen
are all examples of real gases. When the pressure and temperature are the same,
the volume of a real gas is typically less than the volume of an ideal gas. A real
gas can therefore be considered supercompressible. The ratio of the real volume
to the ideal volume is known as the Z-factor or compressibility factor, which is
determined by the parameter Z and measured using Eq. (8.10). Gases possess a
Z-factor that is close to 1 at low pressures and high temperatures, meaning that
they behave very similarly to ideal gases. The volume of most gases at a low
pressure and high temperature is so great that the volume of the molecules
can be ignored, and the distance between the molecules is so large that any
attractive or repulsive forces do not affect the behavior of the gas. However,
as pressure increases, the volume of a gas decreases. So, the volume of mole-
cules should also be taken into account. Furthermore, the attractive or repellent
forces between molecules become relevant.
Gas Z-Factor or Compressibility Factor Calculation
Actual or real volume of the gas at specified pressure and temperature

Z=
Ideal volume of the gas at specifiedf pressure and temperature
8 10
Example 8.6 There is a set pressure of 150 psig for a PSV in a size of 2 × 3 with
an allowable overpressure of 10%. In this valve, air is received and connected to a
line coming from the outlet, which has a size of 3 and a schedule of 40. There is a
1.5-ft horizontal run of the outlet pipe, which is followed by a 6-ft vertical run of
the vent stack. How much pressure is present at the outlet pipe taking the follow-
ing assumptions into consideration?
A The actual flow capacity of the valve is equal to 0.915 times that of the
theoretical value.
B The backpressure has no effect on the flow capacity of the PSV.
Answer
Due to the fact that the ratio of the valve’s actual capacity to its theoretical capacity
is 0.915, then Kd = 0.915. Moreover, since the backpressure has no effect on the
PSV flow capacity, Kb = 1. As air is an ideal gas, Z = 1, and according to
Table 8.3, the specific heat ratio of air is equal to 1.4. The cross-sectional area
of the valve inlet and outlet pipes at point “t” is calculated as follows:

A1 = Cross-sectional area of the valve inlet = = 3 14 in 2
4
P1: Inlet pressure (psi) calculated according to Eq. (8.10):
P1 = Pset + Pover + Patmosphere − Ploss = 150 + 15 + 14 7 = 179 7
Outlet pipe internal diameter = Outlet pipe outside diameter − 2 × Thickness

= 3 5 in − 2 × 0 216 in = 3 068 in
The values for the pipe’s external diameter and thickness are taken from ASME
B36.10 standard.
At Internal area of the outlet pipe at the point“t”
π × Outlet pipe inside diamter2 π × 3 0682
= = = 7 39 in 2
4 4
K d × K b × A1
Pt = k k−1 1 2
At × P 1 × 2 k + 1 1 Z
0 915 × 1 × 3 14
= 1 4 1 4−1 1 2
7 39 × 179 7 × 2 1 4 + 1 1 1
= 0 004 psia
Equation (8.11) is used to calculate the backpressure at valve discharge, which is

point “2.”
Calculation of the Backpressure at Point “2”
2 2k k −1 1 2
2ln P2 P1 + r +f L K d × K b × An × P 1 k 2
P2 = P2t + × × ×
ID At Z k +1
8 11
274 8 Safety Valves
where:
P2: Outlet or discharge pressure at point “2” (psi) calculated according

to Eq. (8.11);
Pt: Terminal pressure (psi);
P1: Inlet pressure (psi) calculated according to Eq. (8.1) or (8.2);
Kd: Coefficient of discharge calculated from Eq. (8.7) (dimensionless);
Kb: Backpressure correction factor (dimensionless);
An: Cross-sectional area of the valve outlet = in 2 ;
4
At: Internal area of the outlet pipe at the point “t” (in.2);
r: Sum of the outlet pipe resistance coefficient (dimensionless);
f: Outlet pipe friction factor (dimensionless);
ID: Outlet pipe internal diameter (in.);
L: Output pipe length equal to the sum of the horizontal and vertical lengths (in.);
Z: Gas compressibility factor (dimensionless);
k: Ratio of specific heats obtained from Table 8.3 for different gases
(dimensionless).
Opening or cracking pressure refers to the increasing inlet static pressure of a

safety valve at which there is a measurable lift or at which the discharge becomes
continuous, as determined by sight, feel, or hearing. Accordingly, the opening pres-
sure, sometimes known as cracking pressure, is a pressure value that is lower than
the set pressure at which leakage begins from the safety valve.
Closing or reseat pressure refers to the pressure required for a valve to reseat after
it has been opened. Actually, the main point is that the safety valve is usually
reseated at a pressure that is slightly lower than the set pressure, as shown in
Figures 8.7 and 8.8. The difference between these two pressure values calculated
using Eq. (8.12) is known as blowdown.
Calculation of Blowdown
Blowdown = Set pressure − Closing pressure reseat pressure 8 12
The reseat or closing pressure is calculated by knowing the set pressure and the
inlet line pressure loss to the safety valve as per Eq. (8.13). According to ASME
Section VIII Div. 01 and API 520 part II, the nonrecoverable losses in the inlet pipe
to the safety valve shall not exceed 3% of the set pressure. Nonrecoverable losses
are primarily due to friction rather than static losses. As an example, if the design
pressure for a pressure vessel is 10 bars, and the PSV is installed 10 m above the
vessel, the static pressure loss for the water supply is approximately 1 bar, and
the valve must not be set at more than 9 bars (10 − 1 = 9 bar).
Pressure %
Where PRVs normally sized
(except fire case)
110
Set pressure
tolerance Allowable
overpressure
Typical set pressure 100 MAWP
Blowdown % of set pressure
95
Reseat pressure of PRV
Typical operating
Leak test pressure 90 pressure
Figure 8.7 A chart illustrating the relationship between reseat pressure, set pressure,
and blowdown.
Calculation of the Closing or Reseat Pressure

Closing or reseat pressure = Set pressure − Non-recoverable losses at the
valve inlet line = Set pressure − 3 × Maximum set pressure
8 13
It should be noted that most safety valve manufacturers and codes and standards
recommend a difference between reseat and operating pressure of 3–5% in order to
ensure proper reseating of the valve and the restoration of proper tightness.
Example 8.7 Approximately, 10 m above a water storage tank handling water

with a density of 1000 kg/m3 is a PSV. If the design pressure of the storage tank
is 10 bars and the allowable overpressure is 10%, what are the values of the max-
imum set pressure, fixed set pressure, relieving pressure, and reseat pressure?
Moreover, the fixed set pressure must be a round number and as close to the max-
imum set pressure as possible, and there should not be any head loss in the valve.
Answer
The static pressure loss is approximately 1 bar due to the location of the safety
valve, which is 10 m above the storage tank. Consequently, the maximum and
set pressure are almost equal to 9 bar.
Static pressure loss = 1 bar
Maximum set pressure of the PSV = Fixed set pressure = 9 bar
276 8 Safety Valves
Vessel Typical characteristics of

Pressure-vessel requirements pressure safety relief valves
Maximum allowable
accumulated pressure 121 Maximum relieving pressure
(fire exposure only) for fire sizing
120
Maximum allowable
accumulated pressure Maximum relieving pressure
116 for process sizing
for multiple-valve installation
115
(other than fire exposure) Multiple valves
Percent of maximum allowable working pressure (gauge) Margin of safety
Single valves
due to orifice
selection
(varies)
Maximum allowable Maximum allowable set pressure
accumulated pressure 110 for supplemental valves
for single-valve installation (fire exposure)
(other than fire exposure)
Overpressure (maximum)
105 Maximum allowable set pressure

for supplemental valves (process)
Overpressure (typical)
Maximum allowable
working pressure 100 Maximum allowable set pressure
or design pressure Simmer for single valves (average)
(hydrotest at 150) (typical)
Start to open
Blowdown (typical)
95 Seat clamping force
Reset pressure (typical)

for single valve
Usual Maximum normal 90 Standard-leak-test pressure

operating pressure Setting ±3%
Not specified by
Tolerances Blowdown ASME code,
simmer Section VIII
Tightness: API Standard 527
85
Notes:
1. The operating pressure may be any lower pressure required.
2. The set pressure and all other values related to it may be moved downward if the operating pressure permits.
3. This figure conforms with the requirements of Section VIII. Division I. of the ASME Code.
4. The pressure conditions shown are for safety relief valves instafled on a pressure vessel (vapor phase).
Figure 8.8 Allowable operating, working, relief, set, and blow-down pressures in a
safety valve.
Thus, the PSV opens at a pressure less than the fixed set pressure of 9 bar when
the pressure in the storage tank is 10 bar.
PSV relieving pressure = 9 + 10 × 9 + 1 = 10 9 bara = 9 9 barg
Maximum safety valve inlet line loss = 3 × 9 bar = 0 27 bar
Equation 8 13 Reseat pressure = Set pressure − 3 × Maximum set
pressure = 9 − 0 27 = 8 73 bar
Simmer refers to a situation where the inlet pressure approaches the set pressure
without still opening or popping the safety relief valve. In fact, simmer makes a
small opening between the disk and the seat, which is called the startup opening.
As a general rule, PRVs begin simmering or opening at a pressure of 90% of the set
pressure. According to Eq. (8.14), simmer may be calculated as follows:
Simmer Calculation
Simmer = Maximum allowable set pressure − Start to open pressure

8 14
Figure 8.8 illustrates allowable operating, working, relief, set, and blow-down
pressures in a safety valve.
The following paragraph summarizes the pressure changes during the operation
of a safety valve. A pressure less than the set pressure causes the PRV to open ini-
tially. Gradually, the pressure reaches the set pressure and then exceeds the set
pressure based on allowable overpressure. Once the pressure reaches the relieving
pressure, which is considered a maximum pressure scenario, the pressure gradu-
ally decreases until the valve is closed.
8.3.1.1 Overpressure Criteria

It is imperative that all piping and equipment be protected when internal or exter-
nal pressure exceeds the design condition of the system due to a variety of reasons,
such as an emergency, operational error, instrument malfunction, or fire. Pres-
sure-relieving devices were designed to ensure that a system or any of its compo-
nents would not be subject to or at risk of a pressure that exceeded the allowable
pressure specified in the relevant design code. According to ASME B31.3, process
piping code, overpressure is referred to as occasional variations in pressure that
may occur in piping systems, including safety valves. There is no doubt that the
allowances for pressure increases are correlated with temperature. Actually, the
most severe coincident pressure and temperature condition should be taken into
consideration as the design pressure can be equal to the pressure rating. There is,
278 8 Safety Valves
however, the possibility of exceeding the design pressure in some cases as an over-
pressure scenario if all the following conditions are met:
a) Pipes and valves shall not contain pressure-containing components made from
materials such as cast iron or other nonductile or brittle materials.
b) The stress generated from overpressure cannot exceed the yield strength of the
piping and valve material at temperature.
c) In the case of an overpressure scenario, the longitudinal stress generated in pip-
ing and valves must be contained within limits stipulated by the relevant
ASME code.
d) During the life of the piping system, the total number of pressure-temperature
variations above the design conditions is not to exceed 1000 times.
e) The increased pressure should, in no case, exceed the test pressure used for the
piping system, which is equal to 1.5 times the design pressure.
f) There may be occasional variations above design conditions, provided that they
do not exceed one of the following limits:
Under certain circumstances and with the approval of the client or end user, it is
permitted to exceed the pressure rating or the allowable stress for pressure design
at the temperature of the increased condition not more than:
a) 33% for more than 10 hours at any one time and no more than 100 h/year, or
b) 20% for no more than 50 hours at any one time and no more than 500 h/year.
8.3.2 Principle of Operation

As illustrated in Figure 8.9, a safety valve is acti-
vated as a result of two dynamic forces caused by
the velocity and pressure of gas or liquid. In the
closing or idle position, the internal pressure of
the protected equipment, such as a pressure ves-
sel, is less than the set pressure of the safety valve;
P2· therefore, the spring retains the disk of the valve
against the seat and keeps the valve in the closed
D· position. When Eq. (8.15) is satisfied, the valve is
closed.
Safety Valve Closing Condition

P1·
Figure 8.9 A safety valve πD2 πD2

P1 × < P2 × + Kx 8 15
schematic. 4 4
where:
P1: Inlet pressure (Pa or N/m2);

P2: Outlet pressure (Pa or N/m2);
D: Seat diameter refers to the smallest diameter of contact between the fixed and
moving portions of the valve’s pressure-containing elements (m);
Kx: Action of spring (N).
Conversely, the valve is open, or in the working position, when the internal pres-
sure of the protected equipment exceeds the valve’s set pressure, causing the spring
to be compressed. As the fluid is released, the internal pressure is reduced. As soon
as the internal pressure falls below the valve’s set pressure, the spring closes the
valve against the seat. If the condition in Eq. (8.16) is satisfied, the valve is open.
Safety Valve Opening Condition
πD2 πD2
P1 × > P2 × + Kx 8 16
4 4
Example 8.8 Flanges are used to connect a 6 pipe with a standard thickness to
an inlet of a PSV attached to a pressure vessel. Accordingly, the internal diameter
of the PSV flange end is equal to the pipe’s internal diameter, but the diameter of
the seat where the disk sits on the seat is 80% of the flange diameter. If the inlet
pressure is 10 bar and the outlet pressure is 2 bar, what is the minimum spring
force necessary to maintain the closed position of the valve?
Answer
Following ASME B36.19, the standard for steel pipes, the outside diameter of a
6 pipe is 6.625 equal to 168.3 mm, and the standard wall thickness is 0.280 equal
to 7.11 mm. Equation (8.17) is used to calculate the internal diameter of the pipe
based on the given outside diameter and the wall thickness:
Pipe Internal Diameter Calculation

ID = OD − 2 × T 8 17
where:
ID: Internal diameter of the pipe (in., mm);

OD: Outside diameter of the pipe (in., mm);
T: Pipe thickness (in., mm);
ID = OD − 2 × T = 168.3 − 2 × 7.11 = 154.08 mm = 0.154 m;
PSV seat diameter = 80 % × 0.154 = 0.1232 m;
280 8 Safety Valves
Figure 8.10 Schematic

Spring representation of a safety valve
FS with the bonnet vented to the
Spring bonnet valve discharge.
Disc area (AO)
PB PB Vent
Disk
guide PB
Disk
PB PB
PV
Nozzle
area (AN)
P1 : Inlet pressure = 10 bar = 106;

P2 : Inlet pressure = 2 bar = 2 × 105.
The condition in Eq. (8.15) must be met in order for the valve to remain in the
closed position:
πD2 πD2
P1 × < P2 × + Kx Minimum“Kx” or spring force to keep the valve
4 4
πD2 πD2 πD2
in closed position = P1 × −P2 × = P1 −P2
4 4 4
π × 0 12322
= 106 −2 × 105 = 9536 49 N
4
As illustrated in Figure 8.10, a safety valve schematic diagram is illustrated in

which the pressure inside the valve bonnet is released to the valve discharge or
outlet line. The force acting on the disk with the area of AD is equal to the valve
inlet pressure multiplied by the area of the nozzle AN, which is equal to the sum of
the spring force FS and the force caused by backpressure PB acting on the top of the
disk AN. In this case, the minimum amount of force required to open the valve is
determined by Eq. (8.18).
The Minimum Force Necessary to Open the Safety Valve

Pv AN = F S + PB AN 8 18
where:
Pv: Fluid inlet pressure;

AN: Nozzle area;
FS: Spring force;
PB: Backpressure.
In order to select a PRV, three factors must be taken into consideration: mini-
mum inlet diameter D, the flow rate through the PSV, also known as the relieving
flow, and time for the valve to open. Equations (8.19), (8.20), and (8.21) are used to
calculate these parameters.
Calculation of the Minimum Inlet Diameter of a Safety Valve
Qmax
D = 28 6 × 8 19
H 0max
23
where:
D: Minimum diameter of a safety valve’s inlet (mm);

Qmax: Discharged flow rate at the maximum pressure and opening (l/s);
Hmax: Pressure head at maximum valve opening (m).
Example 8.9 Calculate the minimum diameter of the inlet of a safety valve
that discharges a flow of 400 l/s at the maximum opening with a head pressure
of 234 m.
Answer
Qmax 400 20
D = 28 6 × 23 = 28 6 × = 28 6 × = 163 mm
H 0max 2340 23 35
Calculation of the Relieving Flow Through a Safety Valve
Head
Qrelv = Qmax × 8 20
Head at the maximum openning
Example 8.10 Calculate the relieving flow for the safety valve in the previous
example if the head is 160 m.
Answer
Head 160
Qrelv = Qmax × = 400 × = 330 76 l s
Head at the maximum openning 234
282 8 Safety Valves
Calculation of the Time Required for Valve to Open
D 100
t=K× × 8 21
50 1 25 × H open
where:
t: Time required for the valve to open (second);

K: Constant that equals 0.4 for fast opening and 0.9 for slow opening valves
(dimensionless);
D: Minimum diameter of a safety valve’s inlet (mm);
Hopen: The head at which the valve opens (m).
Example 8.11 Estimate the time required to open a safety valve if the minimum
diameter of the safety valve’s inlet is 200 mm and the head at which it opens
is 100 m.
Answer
We will consider two cases of fast opening and slow opening for the valve in this
example.
D 100
Fast opening t=K× ×
50 1 25 × H open
200 100
=04× × = 1 43 seconds
50 1 25 × 100
D 100
Slow opening t=K× ×
50 1 25 × H open
200 100
=09× × = 3 2 seconds
50 1 25 × 100
8.3.3 Safety Valve Reaction Forces

Reaction forces can be generated by PSVs, slug flow forces, and water hammering.
Water hammering due to changes in flow velocity and sudden valve closure is dis-
cussed in Chapter 7. Slug flow forces are not covered since they are not related to
industrial valves. When the pressure inside the apparatus exceeds the set pressure
or allowable pressure level, the PSV opens, the excess pressure is released to the
flare lines, and the extremely high pressure returns to its normal, safe level. As the
safety valve is open and blowing, reaction forces are generated by a combination of
backpressure and sudden impulses. While reaction forces at the inlet are relatively
small, the main reaction force is generated at the outlet, especially in the case of gas
fluid service. This is because high velocity and an increase in outlet pressure are
expected. The impact of the PSV load should be taken into account so that the pipe-
work is adequately supported.
The PSV load can be calculated in two ways. The first is based on ISO 4126-9,
Safety devices for protection against excessive pressure – Application and instal-
lation of safety devices other than standalone bursting disc safety devices, based
on Eq. (8.22). A steady-state flow should be considered for the calculations of
the PSV reaction forces according to ISO 4126-9. Flow in a steady state refers to
a process in which the fluid properties at any point in the system do not change
over time. These fluid properties include temperature, pressure, velocity, and mass
flow rate.
The Reaction Forces for Pressure Safety Valves or Relief Valves as

Defined by ISO 4126-9 in Steady-State Flow
Qm V out Aout
F= + Pb− Pu 8 22
3600 10
where:
F: Reaction force (N);

Qm: Mass flow (kg/h);
Vout: Velocity of the fluid in the outlet pipe (m/s);
Pb: Backpressure (MPa);
Pu: Superimposed backpressure (MPa);
Aout: Flow area of outlet pipe (mm2).
Example 8.12 With an inlet of 4 and an outlet of 6 , a PSV has a capacity of

53,484 lb/h. A gas with a density of 0.68 kg/m3 passes through the safety valve.
The static pressure at the outlet of the PSV during the opening process is 2 bars,
whereas the buildup backpressure is 3 bars. Calculate the PSV reaction forces
according to ISO 4126-9, taking into account that the wall thickness of the outlet
pipe is 0.280 in. (7.112).
Answer
The first step is to calculate the mass flow rate in kilograms per hour in SI units as
follows:
Qm = 53,484 lb h = 53,484 lb h × 0 453592 kg lb = 24,260 kg h

= 6 739 kg s
In order to calculate the outlet piping flow area, the following formula
must be used:
284 8 Safety Valves
According to ASME B36.10, the piping standard, a 6 pipe has an outside diam-
eter of 6.625 in. (168.275). Based on the given outside diameter and the wall
thickness, the following Eq. (8.6) is used to calculate the internal diameter of
the pipe:
ID = OD − 2 × T = 168 275 − 2 × 7 112 = 154 051 mm = 0 154051 m
From the pipe’s internal diameter, the flow area of the pipe can be calculated as
follows:
ID2 154 0512
Aout = π =π = 18,638 29 mm2
4 4
The following equation gives the relationship between mass flow rate, flow area,
the density of fluid at the outlet pipe, and fluid density. Using the mass flow rate,
pipe flow area, and density of the fluid inside the valve and the outlet pipe, one can
obtain the fluid velocity as follows:
Qm 6 739
Qm = V out × Aout × ρ V out = = = 66 33 m s
Aout ρ 0 154051 × 0 68
While the valve is opening, the static backpressure at the outlet of the PSV is
2 bar or 0.2 MPa. In addition, the buildup backpressure is 3 bars or 0.3 MPa.
Calculating backpressure is as follows:
Backpressure = superimposed back − pressure + buildup backpressure
= 0.2 + 0.3 = 0.5 MPa
Q V out Aout 24,260 × 66 33
F= m + Pb− Pu =
3600 10 3600
18,638 29
+ 0 5−0 2 = 447 + 559 15 = 1006 15 N
10
According to API 520 Part II contains sizing, selection, and installation of
pressure-relief devices in refineries. According to Eq. (8.23) extracted from API
520 for calculation of reaction forces from PSVs, the standard assumes that the
discharge from the PSV is discharged into the atmosphere (open discharge),
and the flow condition is steady-state. In a closed system under steady-state con-
ditions, the PSV reaction forces are not so large that they can be ignored.
Pressure Safety or Relief Valve Reaction Forces as Per API 520 Part II
in Open Discharge Condition for Gas Service
In imperial units, the following equation should be used:
W KT
F= × + AP 8 23a
366 K+1 M
Table 8.4 Definitions of parameters used for PRV load calculation.
Used
symbol Designation Unit (metric or imperial)
F Reaction force at the point of discharge to N (newton) Pound

the atmosphere
W Flow of any gas or vapor kg/s Pound mass per
hour (lbm/hour)
K Ratio of specific heats at the outlet — —
conditions calculated from Cp/Cv equation
Cp Specific heat at constant pressure — —
Cv Specific heat at constant volume — —
T Temperature at the outlet K R (Rankine same as
Fahrenheit scale)
M Molecular weight of the process fluid —
A Area of the outlet at the point of discharge mm2 in.2
P Static pressure within the outlet at the Barg psig
point of discharge
In metric units, the following equation should be used:
KT
F = 129 W × + 0 1 AP 8 23b
K+1 M
A description of each parameter and its unit of measurement is provided in

Table 8.4.
Example 8.13 A PSV is installed for a gas service with a flow rate of 81, 804 kg/h.
The ratio of specific heat at the outlet of PSV is 1.001 and the temperature at the
outlet of the PSV is 515.6 C. The molecular weight of the process fluid is 18.02 and
the area of the outlet at the point of discharge is 345.0424 cm2. The static pressure
within the outlet at the point of discharge is 1.05 kg/cm2. Calculate the PSV reac-
tion force in Newton.
Answer
All the given data can be converted to metric units as per Table 8.5, as follows:
Using all the given data in Eq. (8.23), the PSV reaction load is calculated as follows:
KT
F = 129 W + 0 1 AP
K+1 M
1 001 × 788 75
= 129 × 22 72 kg s + 0 1 × 34504 2378 × 1 03
2 001 × 18 02
= 2930 88 × 4 68 + 3554 = 17,270 N
286 8 Safety Valves
Table 8.5 Provided data for PSV load calculation in gas service and the conversion to
metric units.
Given Given data value in metric

Provided data data value system
Flow of gas or vapor (W) 81,804 kg/h 22.72 kg/s

Ratio of specific heat (K) 1.001 1.001
Temperature at inlet (T) 515.6 C 788.75 K
Modular weight of process fluid (M) 18.02 18.02
Area of outlet at the point of 345.0424 cm2 34,504.2378 mm2
discharge (A)
Static pressure within the outlet at the 1.05 kg/cm2 1.03 bar
point of discharge (P)
Note: Some of the literature recommends applying a dynamic load factor of 2 to the
PSV reaction forces. Thus, in this case, the reaction force that is calculated through
this equation should be doubled to 34,540 N.
Using Eqs. (8.24) and (8.25), there is an alternative method to calculate reaction
forces for a PRV discharging gas, vapor, or steam directly to the atmosphere with-
out any discharge piping, using API 520.
PSV Reaction Force for Open Discharge of Gas, Vapor, or Steam to

the Atmosphere (Open Discharge) in Accordance with API 520
Gas/Vapor
CKAP1 k k + 1
F= + Fg
332 7 8 24
KAP1 k r
Fg = − Pa × Ao
1 383A0
If Fg is less than or equal to zero, use Fg = 0
Steam
AP1
F= + Fs
1 335
8 25a
AP1
Fs = − Pa × Ao K n
1 886A0
If Fs is less than or equal to zero, use Fs = 0

where:
F: The total reaction force at the point of discharge to the atmosphere in

pounds-force;
Fg: The reaction force due to static pressure at the valve outlet for gas/vapor appli-
cations, expressed in pounds;
Fs: The reaction force due to static pressure at the valve outlet for steam applica-
tions in pounds;
C: The coefficient is obtained from the ratio of the specific heats of the gas or vapor
at standard conditions in Table 8.7 (dimensionless);
K: Effective coefficient of discharge equal to 0.975 for safety valves handling gas,
vapor, or steam according to API 520. According to API 520, the effective coef-
ficient of discharge for safety valves in liquid service is 0.65 and for two-phase
flows, it is 0.85;
A: An effective discharge area in square inches is selected from the ASME or API
standards according to Table 8.6;
Table 8.6 ASME and API standard orifice sizes.
Serial Orifice API effective ASME effective PSV inlet × outlet

number designation area (in.2) area (in.2) sizes (in.)
1 D 0.110 0.124 1 ×2
1.5 × 2
1.5 × 2.5
2 E 0.196 0.221 1 ×2
1.5 × 2
1.5 × 2.5
3 F 0.307 0.347 1 ×2
1.5 × 2
1.5 × 2.5
4 G 0.503 0.567 1.5 × 2.5
1.5 × 3
2 ×3
5 H 0.785 0.887 1.5 × 3
2 ×3
6 J 1.287 1.453 2 ×3
2.5 × 4
3 ×4
7 K 1.838 2.076 3 ×4
8 L 2.853 3.221 3 ×4
4 ×6
9 M 3.600 4.065 4 ×6
10 N 4.340 4.900 4 ×6
11 P 6.380 7.205 4 ×6
12 Q 11.05 12.47 6 ×8
13 R 16.00 18.06 6 ×8
6 × 10
14 T 26.00 29.35 8 × 10
288 8 Safety Valves
Table 8.7 Values of the gas constant (C) based on the ratio of specific heat (k)
according to API 520.
K C k C k C k C
1.00 315 1.31 347 1.60 372 1.90 394

1.01 317 1.31 348 1.61 373 1.91 395
1.02 318 1.32 349 1.62 374 1.92 395
1.03 319 1.33 350 1.63 375 1.93 396
1.04 320 1.34 351 1.64 376 1.94 397
1.05 321 1.35 352 1.65 376 1.95 397
1.06 322 1.36 353 1.66 377 1.96 398
1.07 323 1.37 353 1.67 378 1.97 398
1.08 325 1.38 354 1.68 379 1.98 399
1.09 326 1.39 355 1.69 379 1.99 400
1.10 327 1.41 356 1.70 380 2.00 400
1.11 328 1.41 357 1.71 381 — —
1.12 329 1.42 358 1.72 382 — —
1.13 330 1.43 359 1.73 382 — —
1.14 331 1.44 360 1.74 383 — —
1.15 332 1.45 360 1.75 384 — —
1.16 333 1.46 361 1.76 384 — —
1.17 334 1.47 362 1.77 385 — —
1.18 335 1.48 363 1.78 386 — —
1.19 336 1.49 364 1.79 386 — —
1.20 337 1.50 365 1.80 387 — —
1.21 338 1.51 365 1.81 388 — —
1.22 339 1.52 366 1.82 389 — —
1.23 340 1.53 367 1.83 389 — —
1.24 341 1.54 368 1.84 390 — —
1.25 342 1.55 369 1.85 391 — —
1.26 343 1.56 369 1.86 391 — —
1.27 344 1.57 370 1.87 392 — —
1.28 345 1.58 371 1.88 393 — —
1.29 346 1.59 372 1.89 393 — —
1.30 347 1.60 373 1.90 394 — —
psia mpaa Bara

3200 22.060 220.6
3100 21.370 213.7
3000 20.690 206.9
2900 20.000 200.0
2800 19.310 193.1
2700 18.620 186.2
2600 17.930 179.3
2500 17.240 172.4
2400 16.550 165.5
2300 15.860 158.6
2200 15.170 151.7
2100 14.480 144.8
2000 13.790 137.9
1900 13.100 131.0
1800 12.410 124.1
1700 11.720 117.2
1600 11.030 110.3
1500 10.340 103.4
0.95 1.00 1.05 1.10 1.15 1.20
Correction factor, Kn
Figure 8.11 Correction factor for high pressure steam, Kn.
Ao: Valve outlet cross-sectional area (in.2) calculated from Eq. (8.25);
P1: The absolute relieving pressure is calculated using Eq. (8.2) (psia);
Pa: Atmospheric pressure equal to 14.7 psi;
k: Ratio of specific heats of the fluid as per Table 8.3 (dimensionless);
Kn: According to Figure 8.11, the high-pressure steam correction factor is used
when the steam absolute relieving pressure is greater than 1400 psia. As a result
of the ASME standard, this factor is used to compensate for the deviation
between the steam flow as determined by Napier’s equation and the actual sat-
urated steam flow at high pressures. It can be calculated from Eq. (8.26) rather
than from the figure to obtain the high-pressure steam correction factor;
Kr: The correction factor for ratios of specific heat of other than 1.4 according to
Table 8.8.
Calculation of the Valve Outlet Cross-Sectional Area
Valve outlet size2

Ao = π × 8 25b
4
290 8 Safety Valves
Calculation of the Correction Factor for High-Pressure Stream Kn

0 1906P1 − 1000
Kn = 8 26
0 2292P1 − 1061
Table 8.8 Correction factors for

the specific heat.
K Kr
1.01 1.15
1.05 1.13
1.10 1.11
1.15 1.09
1.20 1.07
1.25 1.05
1.30 1.03
1.35 1.02
1.40 1.00
1.45 0.98
1.50 0.97
1.55 0.95
1.60 0.94
1.65 0.93
1.70 0.91
1.75 0.90
1.80 0.89
1.85 0.87
1.90 0.86
1.95 0.85
2.00 0.84
Example 8.14 In a subflare system, a safety valve with dimensions of

1 1/2 × 2 1/2 is managing natural gas with a specific gravity of 0.6. The valve
has a capacity of 5900 lb/h, and the valve is designed to have an orifice area of
0.397 in.2. If the valve set pressure is set at 210 psig with only 10% overpressure,
calculate the reaction forces generated by the safety valve during operation if there
is no pressure loss within the valve. (Note: API standard is used for the design of
this valve.)
Answer
According to Eq. (8.2), the relieving pressure is calculated as follows:
Table 8.3 states that the ratio of specific heat (k) for natural gas is 1.27. As a
result, the correction factor for ratios of specific heat Kr according to Table 8.8
is 1.04. The next step is to calculate the cross-sectional area of the valve outlet
based on the valve outlet size as determined by Eq. (8.25) as follows:
Valve outlet size2 2 52

Ao = π × =π× = 4 909 in 2
4 4
There is a calculated orifice area of 0.397 in.2, which is not a standard orifice size.
According to the API standard, the standard orifice size that meets the estimated
orifice area in Table 8.6 is orifice designation G, equal to 0.503 in.2. According to
Table 8.7, the gas constant is obtained as 344 based on the value of specific heat
ratio. All given and calculated data are summarized as follows:
K = 0 975, A = 0 503 in 2 , P1 = 245 7 psia, K r = 1 04, Pa = 14 7 psia,

Ao = 4 909 in 2 , k = 1 27, C = 344
In the next step, the reaction force due to the static pressure at the safety valve
outlet will be calculated based on Eq. (8.24) as follows:
KAP1 k r 0 975 × 0 503 × 245 7 × 1 04
Fg = − Pa × Ao = − 14 7
1 383A0 1 383 × 4 909
× 4 909 = 18 45 lb
The total force generated by the safety valve can now be calculated as follows:
CKAP1 k k + 1 344 × 0 975 × 0 503 × 245 7 1 27 1 27 + 1

F= + Fg =
332 7 332 7
+ 18 45 = 111 64 lb
Equation (8.27) is used for the calculation of reaction forces for a PRV dischar-
ging liquid into the atmosphere.
PSV Reaction Force for Open Discharge of Liquid Service to the

Atmosphere (Open Discharge)
3 44 × 10 − 7 × W 2
F= 8 27
ρA0
292 8 Safety Valves
where:
F: Reaction force at the point of discharge to the atmosphere (lb);

W: Actual relieving capacity (lb/h);
A0: Area at discharge (in.2);
ρ: Density of the fluid (lb/ft3).
Equation (8.28) is used for the calculation of reaction forces for a PRV dischar-
ging a two-phase gas and liquid flow into the atmosphere, according to API 520-
part II recommended practice.
PSV Reaction Force for Open Discharge of Gas–Liquid Service to the

Atmosphere (Two-Phase Discharge)
W2 X 1−X
F= + + Ao P E − P A 8 28
2 898 × 106 A0 ρg ρL
where:
F: Reaction force at the point of discharge to the atmosphere (lb);

W: Actual relieving capacity (lb/h);
Ao: Area at discharge outlet to the atmosphere (in.2);
x: Mass fraction of gas and vapor portion calculated as WG/W;
WG: Actual relieving capacity of the gas (lb/h);
ρg: Vapor density at exit (lb/ft3);
ρl: Liquid density at exit (lb/ft3);
PE: Pressure at pipe exit (psia);
PA: Ambient pressure (psia).
Example 8.15 A PSV has a 1 CL150 raised face flange inlet and a 2 CL150
raised face flange outlet. The design pressure and temperature are 285 psi gauge
and 86 F and the material of the valve and connected pipe are in carbon steel.
The piping material grade is ASTM A106 Gr. B. The overpressure crude oil in
the liquid phase is released from the valve to the atmosphere. The density
of the crude oil in this case is 800 kg/m3. The actual capacity of the valve is
26,896 lb/h. The area of discharge from the PSV is 0.196 in.2. What is the reaction
load of the PSV?
Answer
Equation (8.27) should be used for the calculation of the reaction forces. All the
required data are in imperial units, except for the density of the fluid.
ρ = 800 kg m3 = 49 94 lb ft3
3 44 × 10 −7 × W 2 3 44 × 10 −7 × 26,896 2 248 85
F= = = = 25 41lb = 113N
ρA0 49 94 × 0 196 9 79
Applying the design load factor of 2 provides a 51-lb equal to 226 N reaction force
for the PSV.
The main question is, what is the limit of reaction forces generated by the PSV? It
should be noted that the PSV reaction force is an occasional load. The total of the
sustained loads SL plus occasional loads So should be at maximum equal to 1.33 of
the allowable stress at hot temperature. (SL + So ≤ 1.33 × Sh). The term “sustained
loads” refers to a combination of internal pressure and weight loading that are con-
stant during the lifetime of the pipe. In this context, weight refers to both the pipe
itself and related components such as valves, flanges, strainers, etc. It is also impor-
tant to consider the weight of the fluid during the operation or during the hydrot-
est. The weight of some pipes is increased by the use of thermal insulation or
internal cladding. In addition, in cold regions, snow can accumulate on the piping.
So, the weight of the snow is considered in calculating the sustained load on the
piping. Sustained load analysis takes into account only longitudinal or axial prin-
cipal stress (δa), which is produced by the fluid pressure. The sustained loads may
result in the collapse of the piping if the load is not appropriately managed through
adequate pipe supports at the correct location. The ASME B31.3 code requires that
the sustained load or stress always be less than, or at a maximum equal to, the hot
allowable stress, parameter Sh. Sustained load is equal to weight plus longitudinal
stress. Equation (8.29) shows the limitation and calculation of sustained loads:
Limitation and Calculation of Sustained Loads

S L ≤ Sh S a + Sw ≤ S h 8 29
where:
SL: Sustained load equal to the weight loads plus longitudinal (axial) loads (psi);
Sh: Allowable stress at temperature or hot temperature (can be extracted from
ASME B31.3; psi);
Sa: Axial or longitudinal stress as a result of piping inside pressure, which can also
be shown as δa (psi);
Sw: Stress due to weight (psi).
Example 8.16 A PSV is estimated to produce a load of 51 lb. It is assumed that

none of the occasional loads occurs at the same time. The total sustained load is at
its maximum and equal to the allowable stress at hot temperatures. The PSV and
its connecting pipe are constructed from carbon steel, and the pipe is constructed
294 8 Safety Valves
of A106 Gr.B. Assuming the operating temperature of the PSV is ambient, is the
produced load by the PSV acceptable?
Answer
If the sustained load is equal to the allowable stress at hot temperature (SL = Sh),
then the maximum allowable occasional stress would be 0.33 × Sh.
SL + So ≤ 1 33 × Sh Sh + So ≤ 1 33 × Sh So ≤ 0 33 × Sh
The piping material connected to the PSV is in ASME A106 Gr. B and the tem-
perature is 86 F. Accordingly, the maximum allowable stress on a pipe at a given
temperature is 20,000 psi. There is an allowable PSV load of 0.33 × 20, 000 psi,
which equals 6600 psi. The pressure produced by the PSV, 51 lb, is much less than
6600 psi. So, the PSV load is within acceptable limits.
8.3.4 Safety Valve Capacity Conversion

It is possible to determine the capacity of a safety or relief valve in terms of a gas or
vapor other than the medium for which the valve is rated and designed using
ASME Section VIII Div. 01 and appendix 11. Equations (8.30), (8.31), and (8.32)
are used to calculate the capacity of a relief valve for steam, air, gas or vapor,
respectively. This formula can also be used when the required flow rate of any
gas or vapor is known and it is required to determine the rated capacity of a valve
for steam or air service.
Safety Valve Rated Capacity for Steam

W s = 51 5KAP 8 30
Safety Valve Rated Capacity for Air
M
W a = CKAP 8 31
T
C = 356, M = 28.97
Safety Valve Rated Capacity for Gas or Vapor
M
W = CKAP 8 32
T
In addition, if the compressibility factor Z should be included in Eq. (8.32) for the
light hydrocarbons, then the formula would be adjusted as follows:
M
W = CKAP
ZT
where:
Ws: Rated capacity lb/h of steam;

Wa: Rated capacity, converted to lb/h of air at 60 F inlet temperature;
W: Flow of any gas or vapor in lb/h;
C: The constant for gas or vapor that depends on the ratio of specific heats from
Table 8.7;
K: Coefficient of discharge calculated from Eq. (8.7);
A: Actual discharge area of the safety valve (in.2);
P: Discharge pressure (psia) calculated according to Eq. (8.2);
M: Molecular weight obtained from Table 8.3;
T: Absolute temperature at the valve inlet ( F + 460).
Example 8.17 The capacity of a safety valve is 3500 lb/h under a set pressure of
190 psi. What is the capacity of the valve to relieve air at 100 F for the same set
pressure?
Answer
Using Eq. (8.30) to calculate the safety valve’s capacity to handle steam:
3500
W s = 51 5KAP 3500 = 51 5KAP KAP = = 67 96
51 5
Using Eq. (8.31) to calculate the safety valve’s capacity to handle air:
M 28 97
W a = CKAP = 356 × 67 96 × = 5503 lb h
T 460 + 100
Example 8.18 A safety valve installed on a pressure vessel must release 5200 lb
of propane per hour considering the valve inlet temperature is 125 F. What is the
overall capacity of this valve in pounds of steam per hour if it has been designed for
steam service?
Answer
In accordance with Table 8.3, the ratio of the specific heat k for propane is 1.13.
The molecular weight of propane as taken from the same table of data is found
to be 44.09. In accordance with Table 8.7, the gas constant value C is found to
be 330. Using Eq. (8.32), the safety valve rated capacity for a gas or vapor can
be calculated as follows:
M 44 09
W = CKAP 5200 = 330KAP KAP = 57 395
T 460 + 125
296 8 Safety Valves
It is now necessary to apply Eq. (8.30) in order to determine the safety valve rat-
ing based on the steam service as follows:
W s = 51 5KAP = 51 5 × 57 395 = 2955 lb h
Example 8.19 It is necessary to discharge 1000 lb of ammonia per hour through

a PSV at a temperature of 150 F. Is the total capacity of this valve the same as the
set pressure that is used for steam?
Answer
It is known that ammonia has a molecular weight of 17.03 and its specific heat
ratio is 1.4. Based on Table 8.7, we can determine that the value of constant C based
on Table 8.7 should be 355.5. Equation (8.32) is used to calculate the safety valve
rated capacity for ammonia by the following formula:
M 17 03
W = CKAP 1000 = 355 5KAP
T 460 + 150
1000
KAP = = 15 394
64 96
In order to determine the safety valve rating for steam service, the following
Eq. (8.30) must be applied:
W s = 51 5KAP = 51 5 × 15 394 = 792 80 lb h
8.1 In the context of nonfire conditions, two PSVs are installed above pressure
equipment. For the second safety valve, the MAWP is 100 psi and the values
of set pressure and accumulated pressure are at their maximum. Identify the
incorrect statements from the following list:
A The set pressure for the valve is 105 psi.
B The allowable overpressure is 10 psi.
C The maximum accumulated pressure is 116 psi.
D Based on no head loss in the valve, the absolute relieving pressure is
130.7 psi.
Answer
Based on Table 8.2, the maximum set pressure and maximum accumulated
pressure values for the second safety valve in the nonfire case are 1.05 times
and 1.16 times the MAWP, respectively.
Set pressure = 1 05 × 100 = 105 psi

Maximum accumulated pressure = 1 16 × 100 = 116 psi
Equation (8.2) Allowable overpressure = Maximum accumulated
pressure − Set pressure = 116 − 105 = 11 psi
Equation (8.2) Absolute relieving pressure = Set pressure + overpres-
sure + atmospheric pressure − pressure loss = 105 + 11 + 14.7 = 130.7 psi
Therefore, the incorrect answer is B since the allowable overpressure
is 11 psi.
8.2 The MAWP for a PSV made of carbon steel material is 20 bar, which is equal
to the design pressure. During normal operation, the safety valve experiences
overpressure conditions. Under which conditions is it acceptable to experi-
ence an overpressure condition? Carbon steel material has a tensile strength
of 70 ksi and a yield strength of 36 ksi.
A The total number of pressure-temperature variations above the design
conditions over the design life of the piping system is 2000 times.
B As a result of the overpressure scenario, the safety valve is subjected to a
stress of more than 40 ksi.
C Overpressure reached 35 bars only once in the overpressure scenario.
D In all three cases described here, an overpressure scenario is not
acceptable.
Answer
The condition described in option A is not acceptable because the total number
of overpressure scenarios shall not exceed 1000 times during the design life of
the piping system. It is also incorrect to choose Option B since the stress gen-
erated during the overpressure scenario must not exceed the yield strength of
the carbon steel material, which is 36 ksi. Option C is not appropriate because
it is not acceptable to have an overpressure scenario of more than 1.5 times the
design pressure equal to 30 bar. Option D is the right answer.
8.3 Which sentence is correct regarding the PRV installed on the pressure vessel
illustrated in Figure 8.12?
A The amount of force produced by a PRV during opening is negligible.
B The PRV is used to drain liquid.
C Overpressurized fluid in the pressure vessel is released into the atmos-
phere through the PRV.
D None of these.
Answer
Option A is incorrect as the generated relief valve load to the atmosphere,
also known as open discharge, is not negligible. In regard to option B, it is
298 8 Safety Valves
F
Ao
(Cross-sectional area)
Long-radius elbow
Vent pipe
Pressure
relief valve
Support to resist
weight and
reaction forces
Vessel
Figure 8.12 PSV on a pressure vessel.
not possible to determine whether gas or liquid is released into the environ-
ment from the figure. Therefore, option B is not entirely accurate. Accord-
ingly, option C is correct, as the overpressure fluid in the vessel is
released into the environment through the PRV. Therefore, option C is
the correct answer, and option D is incorrect.
8.4 Which statement regarding the maximum relieving pressure and set pres-
sure relationship to MAWP is accurate, assuming that the valve’s atmos-
pheric pressure and pressure loss are both zero?
A The set pressure of a pressure vessel should not exceed 10% of the MAWP
of the vessel when a single pressure relief device protects the vessel.
B The relief pressure shall not exceed 16% of the maximum operating
pressure of a pressure vessel protected by multiple pressure relief
devices.
C The pressure relief shall not exceed 21% of the MAWP of the vessel when
the vessel is protected from fire.
D All three options are correct.
Answer
As shown in Table 8.2, the maximum set pressure of a safety valve when a
single device is installed on a pressure vessel is 10% of MAWP. Therefore,
option A is correct. As both atmospheric and pressure losses occur within
the valve, the relieving pressure and accumulated pressure are equal to
the set pressure plus overpressure. Accordingly, options B and C are correct.
Therefore, option D is the correct answer.
8.5 There is a PSV installed 10 m above pressure equipment with a design pres-
sure of 20 bar. When the maximum set pressure and the fixed set pressure are
equal, the PSV opens at 18 bar, and the allowable overpressure is 10%, which
of the following statements is incorrect?
A There is a relieving pressure of 20.9 bar.
B A simmer is equal to 1 bar.
C The maximum head loss in the inlet line to the PSV is 0.7 bar.
D Closing or reseat pressure is 18.43 barg.
Answer
The first step is to determine the fixed set pressure of the PSV. The static pres-
sure loss associated with PSV’s location at a distance of 10 m is 1 bar. There-
fore, it is possible to calculate the maximum and fixed set pressure as follows:
Maximum set pressure = Fixed set pressure = Design pressure − Static
pressure loss = 20 − 1 = 19 bar
Relieving pressure = Set pressure + Overpressure

= 19 + 19 × 10 = 20 9 barg
Referring to Eq. (8.14), simmer is the difference between the set pressure
and the start to open pressure as calculated as follows:
Simmer = 19 − 18 = 1 bar.
The maximum head loss according to both API 521 and ASME
Sec. VIII Div.01 is 3 % × 19 bar = 0.57 bars.
In relation to Eq. (8.13), the closing pressure or reseat pressure is the
difference between the set pressure and the maximum head loss,
19 − 0.57 = 18.43 barg.
Therefore, all options except C are correct.
8.6 According to the manufacturer of a pressure vessel, the maximum allowable

working pressure is 6 bar, and the maximum allowable accumulated
pressure is 6.3 barg. In the case of a PSV used to protect a pressure vessel from
overpressure scenarios, what is the set pressure of the PSV if the overpressure
amounts to 10%? In this case, a single PSV is used to protect the equipment.
A 5.7 bar
B 6 bar
C 6.3 bar
D 6.5 bar
300 8 Safety Valves
Answer
Options C and D are not valid, since if a single PSV is used, its set pressure
cannot exceed MAWP, which is 6 bar, as shown in Table 8.2. Using
Eq. (8.2), the maximum allowable accumulated pressure can be calculated
as follows:
Maximum accumulated pressure = Set pressure
+ Allowable overpressure 10 of the set pressure

Maximum accumulated pressure
= 1 1 × Set pressure Set pressure =
11
63
= = 5 7 bar Option A is the correct answer
11
8.7 In determining the set pressure of a safety valve, which of the following
statements is correct?
A The set pressure of the PSV should never exceed the maximum allowable
working pressure (MAWP).
B The set pressure of the PSV should never exceed the maximum allowable
accumulation pressure (MAAP).
C The set pressure of the PSV should never exceed the normal working
pressure (NWP) of the protected equipment.
Answer
Option B is the correct answer since MAAP equals the sum of set pressure
and allowable overpressure.
8.8 According to the standards for conventional safety valves, the required set
pressure is 8.5 barg, unless, of course, the valve experiences a contact back-
pressure of 1.0 barg. In light of the effects of backpressure on the set pressure,
how much should the set pressure be adjusted?
A 7.5 bar
B 8.5 bar
C 9.5 bar
D 10.5 bar
Answer
The set pressure, also known as cold set pressure is reduced due to the effect
of backpressure. This has to be compensated for by adding the backpressure
value to the cold set pressure so as to get the adjusted set value, which is equal
to 8.5 + 1 = 9.5 bar. Therefore, option C is the correct answer.
8.9 There is a PSV installed on the pressure equipment that handles the natural gas.
The valve has an inlet and an outlet size of 1.5 and 2.5 , respectively, and the
orifice size selected is G. Considering the fact that 10% overpressure is allowed,
the required capacity of the valve is 6500 lb/h and the set pressure is 150 psi. Dur-
ing the operation of a valve, what is the total reaction forces that are produced?
A 60.25 lb
B 64.67 lb
C 68.16 lb
D 72.55 lb
Answer
According to Eq. (8.2), the relieving pressure is calculated as follows:
Table 8.3 states that the ratio of specific heat (k) for natural gas is 1.27. As a
result, the correction factor for ratios of specific heat Kr according to Table 8.8
is 1.04. The next step is to calculate the cross-sectional area of the valve outlet
based on the valve outlet size as determined by Eq. (8.25) as follows:
Valve outlet size2 2 52

Ao = π × =π× = 4 909 in 2
4 4
According to the API standard, the standard orifice size that meets the esti-
mated orifice area in Table 8.6 is orifice designation G, equal to 0.503 in.2.
According to Table 8.7, the gas constant is obtained as 344 based on the value
of specific heat ratio. All given and calculated data are summarized as
follows:
K = 0 975, A = 0 503 in 2 , P1 = 179 7 psia, K r = 1 04, Pa = 14 7 psia,

Ao = 4 909 in 2 , k = 1 27, C = 344
In the next step, the reaction force due to the static pressure at the safety valve
outlet will be calculated based on Eq. (8.24) as follows:
KAP1 k r 0 975 × 0 503 × 179 7 × 1 04
Fg = − Pa × Ao = − 14 7
1 383A0 1 383 × 4 909
× 4 909 = − 5 89 lb
Since Fg is less than zero, it should be considered equal to zero.

The total force generated by the safety valve can now be calculated as follows:
CKAP1 k k+1
F= + Fg
332 7
344 × 0 975 × 0 503 × 179 7 1 27 1 27 + 1
= + 0 = 68 16 lb
332 7
Thus, option C is the correct choice.
302 8 Safety Valves
8.10 In order for a safety valve to function properly, it should be able to release
10,000 ft3/min of air at 60 F, and its atmospheric pressure should be equal
to 14.7 psia. If the set pressure of the safety valve for both fluid services is
equal, what is the flow capacity of this safety valve in pounds per hour for
saturated steam?
A 25,460 lb/h
B 28,168 lb/h
C 26,750 lb/h
D 23,233 lb/h
Answer
The first step is to convert the air flow capacity from cubic feet per minute
to pounds per hour as follows:
W a = 10,000 × 0 0766 × 60 = 45,960 lb h
Now, Eq. (8.31) is used to calculate the rated capacity of the valve for air
as follows:
M
W a = CKAP C = 356, M = 28 97 45, 960
T
28 97 45,960
= 356 × KAP × KAP = = 546 96
460 + 60 84 028
Now, Eq. (8.30) is used to calculate the rated capacity of the valve for
steam as follows:
W s = 51 5KAP = 51 5 × 546 96 = 28,168 lb h
Further Reading
American Petroleum Institute (API) 520 (2020). Sizing, Selection, and Installation of
Pressure-Relieving Devices Part 1 – Sizing and Selection. Washington, DC: American
Petroleum Institute (API).
American Petroleum Institute (API) 521 (2007). Pressure-Relieving and Depressuring
Systems, 5e. Washington DC: American Petroleum Institute (API).
American Petroleum Institute (API) 526 (2017). Flanged Steel Pressure-Relief Valves, 7e.
American Society of Mechanical Engineers (ASME) (2004). Carbon, Alloy and
Stainless-Steel Pipes. ASME B36.10/19. New York: American Society of Mechanical
Engineers (ASME).
Further Reading 303
American Society of Mechanical Engineers (ASME) (2012). Design and Fabrication of

Pressure Vessels. Boiler and Pressure Vessel Code. ASME Section VIII Div.02. New
York: American Society of Mechanical Engineers (ASME).
Crosby Valve Inc (1997). Crosby Pressure Relief Valve Engineering Handbook. Technical
document number. TP-V300. Crosby Valve Inc.
Hellemans, M. (2009). The Safety Relief Valve Handbook: Design and Use of Process
Safety Valves to ASME and International Codes and Standards, 1e. Oxford:
Butterworth-Heinemann, and imprint of Elsevier.
Nesbitt, B. (2007). Handbook of Valves and Actuators: Valves Manual International,
1e. Oxford: Elsevier.
Sotoodeh, K. (2022). Piping Engineering: Preventing Fugitive Emission in the Oil and Gas
Industry, 1e. New York: Wiley.
Sotoodeh, K. (2022). Cryogenic Valves for Liquified Natural Gas Plants, 1e. Austin,
TX: Elsevier (Gulf Professional Publishing).
Spirax Sarco (2009). Module 9.3, safety valve selection.
304
Safety and Reliability
9.1 Introduction
In different industries, including oil and gas, valves are essential components of
piping systems that regulate and control fluid flow, prevent backflow, and main-
tain safety. In the offshore industry, various types of valves are used, including ball
valves, gate valves, butterfly valves, globe valves, and check valves. Various rea-
sons such as corrosion and mechanical malfunction can result in industrial valves
failing to function. This can lead to adverse consequences such as environmental
pollution, asset loss, loss of production, or even death. It is true that valves are con-
nected to pipes in order to produce or transport goods. These valves are an essential
part of the process, as their failure could halt the entire process. Safety and relia-
bility assurance of industrial valves is such an important topic that has been
addressed in past studies. In addition, a subsurface safety valve (SSV) is also
installed in the well. In the process of oil and gas production, they monitor the well
for abnormal conditions such as an overpressure scenario. This is to prevent blow-
outs and make sure the well is closed in case of any abnormal conditions. The reli-
ability analysis of SSVs has been a popular topic among researchers. In a separate
study, pressure relief valves that are installed on piping and mechanical equipment
to relieve the overpressure fluid or gas from the piping and equipment have been
subjected to a reliability analysis. In the offshore oil and gas industry, there have
recently been a couple of studies that focus on online monitoring and strict factory
acceptance tests (FATs) to enhance the safety and reliability of valves. For exam-
ple, the effect of partial stroke testing on the reliability of safety-critical valves is
analyzed in previous research. The safety-critical valves are those that are con-
nected to either a process shutdown (PSD) system or an emergency shutdown sys-
tem (ESD). PSDs and ESDs are two types of shutdown systems for process systems.
As part of the plant and facility safeguarding systems, PSD systems minimize or
Emergency Safety layer

shutdown
Prevent Shutdown Trip alarm
systems Process Process control layer
shutdown
Process alarm
Process
control Process Process control layer
system value Normal
Figure 9.1 Plant safety implementation and protection with PSD and ESD layers.
prevent the occurrence and consequences of process parameters like pressure,

temperature, and level exceeding operating limits. PSD systems are capable of
stopping a part or the entire operation or of depressurizing or blowing down parts
of the process. Process shutdown valves are actuated valves connected to the basic
process control systems that ensure that the process is managed and controlled
within acceptable limits. The use of an ESD system minimizes or prevents emer-
gency conditions such as extreme or abnormally high values of process parameters
(e.g. high fluid flow in piping) or leakage or emission of hydrocarbons to the envi-
ronment. In this sense, ESD differs from PSD in that it is a safety layer rather than a
process control layer, as illustrated in Figure 9.1. In general, PSD and ESD valves
in oil and gas projects are listed in safety analysis reports (SARs) and must meet
safety integrity level (SIL) 2 or 3. SIL, an important parameter in evaluating
the safety and reliability of industrial valves, is discussed in more detail further
in this chapter.
9.2 Safety Standards
A safety-related system protects equipment and industrial processes where danger

or risk may arise as a result of a failure or malfunction. It is important to remember
that in the old days, industrial safety was focused on areas such as safe work prac-
tices, the control of hazardous materials, and the protection of personnel and
equipment. The safety issues and strategies that are addressed today, however,
are far deeper and more complex than those that were addressed in the past. There-
fore, safety systems of today reduce risk while operating in a manner that contin-
uously improves productivity and profitability by reducing risk. The safety
management process was largely self-regulated until the 1980s. Following the
advent of electronic control devices, more complex issues arose concerning
306 9 Safety and Reliability
manufacturing systems, environmental protection, and the assets and facilities in

a plant. As a result, new internal safety standards have emerged and continue to
.
develop, including the International Electrotechnical Commission (IEC) 61508

and 61511. The number of manufacturers integrating their products with these
standards is increasing. With the advent of these standards, many essential ben-
efits were gained by the final user, which are outlined as follows:
• The risk and reliability requirements, specifications, and design processes are
formulated using a more technical and scientific approach;
• Definition of risk that is more accurate;
• The ability to demonstrate the effectiveness of safety-related systems more easily
and widely;
• Implementing safety-related issues in a more cost-effective manner;
• Maintenance operations can be evaluated more effectively.
IEC 61508 is an international standard published by the International Electro-

technical Commission addressing the functional safety of electrical, electronic,
and programmable electronic safety-related systems. The standard is divided into
seven parts which can be summed up as follows:
• Part 1: General requirements (Normative);

• Part 2: Requirements for electrical/electronic/programmable electronic safety-
related systems (Normative);
• Part 3: Software requirements (Normative);
• Part 4: Definitions and abbreviations (Informative);
• Part 5: Examples of methods for determination of SILs (Informative);
• Part 6: Guidelines on the application of parts 2 and 3 (Informative);
• Part 7: Overview of techniques and measures (Informative).
Part 6 of IEC 61508 provides some essential definitions that will be discussed in
the following section.
Hazard: The term refers to a possible cause of harm, recognizing that the term
includes both threats that occur within a short period of time such as fires
and explosions and those that have a long-term effect on a person’s health, such
as the release of toxic substances such as hydrogen sulfide.
Hazardous situation: The term refers to situations in which a person is exposed
to certain hazards.
Hazardous event: Generally, it refers to a hazardous situation that leads to a neg-
ative result.
Risk: It is a combination of the probability of occurrence of harm and the severity
of that harm.
Tolerable risk: According to the current values of society, it is an acceptable risk

based on the given context.
Residual risk: The residual risk is the risk that remains even after protective mea-
sures have been taken.
IEC 61511 is an international standard titled “Safety Instrumented Systems for

Process Industry Sector: Functional Safety” that was developed in order to inte-
grate safety instruments into safety systems in the process industry sector as an
extension of IEC61508 for the process industry. Standard IEC 61508 governs
the safety instrumented system (SIS) for process industries, including sensors, logic
solvers, and final elements in particular. Figure 9.2 depicts a SIS architecture that
consists of three pressure transmitters, a logic solver, and finally two actuated
valves as the last two components.
The pipeline pressure transmitters monitor the pipeline pressure against the pre-
defined limit and send signals to the logic solver where it is determined what
action is appropriate based on the logic solver output in the form of an electrical
signal. This enables the final element to perform an action, such as shutting down
the processing system through a valve, after receiving the electrical signal from the
logic solver. As a result of this configuration, overpressure incidents are avoided by
removing the source of overpressure or reducing the likelihood of an overpressure
event to such a low level that it is no longer considered likely. SISs are widely used
Figure 9.2 A safety instrumented system (SIS) architecture. Source: Courtesy of Emerson
Electric Co.
in oil and gas industries, for example, with a variety of design and application con-
cepts. The main benefits of SIS can be summed up as the protection of equipment,
personnel, and the environment, as well as the reduction of capital expenditures
and operations costs. In addition to this, IEC 61511 deals with the interface
between SIS and other safety-related systems relating to process hazards and risks.
This standard includes three parts that are listed as follows:
• Part 1: Framework, definitions, and requirements for systems, hardware, and

software;
• Part 2: Guidelines for the application of the IEC 61511 standard;
• Part 3: An overview of the examples and methods for determining safety integ-
rity in the application of hazard and risk analysis.
9.3 Risk Analysis
It is the purpose of all safety standards to minimize the risk that is inherent to the
manufacturing and processing processes. To achieve this goal, it is necessary to
either eliminate the risk or reduce it to a level that is acceptable. In a real-world
scenario, the risk could be negligible, tolerable, or unacceptable. Thus, more spe-
cifically, the purpose of safety systems is to reduce the risk to a level that is accept-
able or tolerable. Hence, safety can be defined as a state of being free from
unacceptable risks. According to Eq. (9.1), the risk is given a score or calculated.
Calculation of Risk Score

Risk Score or risk priority number RPN = Hazard frequency or probability
× Hazard severity
91
It is possible to determine the probability and severity of a hazard based on

Tables 9.1 and 9.2, respectively. Accordingly, the lowest and highest scores for
the hazard frequency (probability) and severity are 1 and 10, respectively. This
makes the risk score range from a minimum of one to a maximum of one hundred.
There must be a reduction or elimination of those risks that have higher scores
(e.g. RPN = 95). However, if the risk score is as low as five, the risk factor is most
likely tolerable.
Failure mode and effect analysis (FMEA) is an assessment method used to iden-
tify and eliminate known and potential problems and errors from a system before
it is delivered to the customer. This tool is widely used for determining the relia-
bility of systems. Among the major advantages of this system is that it enhances
Table 9.1 Determination of the risk probability score.
Rating Occurrence probability Minimum probability percentage (%)
10 Extremely high 50
9 Very high 33
8 Very high 10 – 15
7 High 5
6 Marginal 1
5 Marginal 0.25
4 Unlikely 0.05
3 Low 0.007
2 Very low 0.0007
1 Remote 0.000007
Table 9.2 Determination of the severity of the risk.
Rating Effect Severity effect
10 Hazardous without Very high severity without warning

warning
9 Hazardous with warning Very high severity with warning
8 Very high Destructive and unsafe failure
7 High System is inoperable due to major equipment
damage
6 Moderate System is inoperable due to minor damage
5 Low System is inoperable without damage
4 Very low High degradation of performance
3 Minor Less performance degradation
2 Very minor Minimal system operability
1 None No effect
safety and reliability, and reduces the possibility of failure, which, in turn,
increases customer satisfaction. Injury to humans and adverse environmental
effects can be minimized through the use of failure reduction opportunities to
improve health, safety, and environment (HSE). At the moment, this is the most
common risk analysis approach being used. In accordance with historical records,
Initiate FMEA of an item
Select a component to analyze
Identify failure modes of the selected component
Are there any

Select the failure mode of analyze failure mode to
analyze?
Identify the effect of the failure mode
Determine severity of final effect Are there other

components for
analysis?
Identify potential causes of the failure mode
Estimate the frequency or probability of the

occurrence for the failure Complete the
FMEA
Figure 9.3 FMEA procedure chart.
FMEA was first developed by the American military in the 1940s and implemented
in the 1950s for the design of flight control systems. Figure 9.3 is an illustration of
steps and a procedure for implementing FMEA in practice.
Example 9.1 In the offshore oil and gas industry, pipeline valves are known for
being the most expensive and critical valves. An example of a pipeline valve is dis-
cussed in this example. The valve is connected to the emergency shutdown system
and it is considered to be a part of SIS. This valve was found to have four specific
types of risk that are listed in Table 9.3. The table also includes estimates of the
probability and severity of the listed risks. Which of the risks has the highest pri-
ority that needs to be reduced or eliminated? In addition, which risk item is accept-
able as it is?
Answer
In order to calculate the RPN, each risk item is assigned a severity and occurrence
score as well as a risk severity-occurrence score. These values are entered into
Table 9.4. On the right side of the table, in the last column, we can see RPNs
Table 9.3 Identified risks for the pipeline valve.
Risk item Potential failure mode or risk item Severity (S) Occurrence (%)
A Damage to the valve body as a result of Hazardous 1

pipeline loads with
warning
B Ingress of dense oil in the seat Very high 5
arrangement and malfunction of the valve
seat
C Damage to the seat and ball during the Hazardous 0.007
commissioning and hydrotest as a result without
of dirt and welding debris warning
D Failure in shutting down the emergency Hazardous 0.000007
shutdown valve during the subsea with
pipeline failure warning
Table 9.4 Scores of risks associated with pipeline valves.
Risk
priority
Risk Potential failure Severity Occurrence Occurrence number
item mode or risk item Severity (S) score (%) score (RPN)
A Damage to the Hazardous 9 1 6 54

valve body as a with
result of pipeline warning
loads
B Ingress of dense Very high 8 5 7 56
oil in the seat
arrangement and
malfunction of the
valve seat
C Damage to the Hazardous 10 0.007 3 30
seat and ball without
during the warning
commissioning
and hydrotest as a
result of dirt and
welding debris
D Failure in shutting Hazardous 9 0.000007 1 9
down the with
emergency warning
shutdown valve
during the subsea
pipeline failure
are calculated and risk item B has the highest score equal to 56. This makes it the
top priority to reduce or eliminate this risk. Risk item A has been assigned a score
of 54 and has been given the second highest priority for risk reduction measures,
having a score of 54. In terms of risk number, item D has the lowest risk score. So it
can be accepted as is.
9.4 Basic Safety and Reliability Concepts
In this section, some fundamental concepts that are used in the IEC 61508 stand-
ard are explained in order to provide a deeper understanding of safety issues.
9.4.1 System Incidents and Failures

It is very important to understand the concept of system incidents and failures
before we go into the details of basic safety and reliability concepts as well as
how these concepts are calculated. Figure 9.4 shows the most relevant system inci-
dents and failures, those incidents and failures that are all connected to the sys-
tem’s safety and reliability.
MTBF
MTTF MTTD MTTR MTTF
Correct Correct
Diagnose Repair
behavior behavior
First failure Begin repair End repair Second failure
Figure 9.4 A system incidents and failures chart.

9.4.1.1 Failure Rate

The failure rate refers to the frequency of failure for a component per unit of time,
which is typically shown as a Greek letter λ. Several safety parameters are calcu-
lated based on failure rates, such as reliability and incidents that are explained in
more detail later in this chapter. There is a general rule of thumb that the failure
rate of a component or system is related to time. This means that the failure rate
and its effect will vary over the system’s or the component’s lifespan. In the case of
a valve, a failure occurring five years after its installation might have a greater
impact on the cost and risk of the operation than if the valve failed after only
one year of operation.
There is a method to estimate the failure rate of a valve by analyzing its per-
formance history over the years of operation. In order to calculate the failure
rate of a valve, it is simply necessary to divide the number of failures by the
total time the valve has been in operation (see Eq. (9.2)). It is usually the valve
manufacturer or the end user who provides a history of the valves’ failures dur-
ing operation.
Calculation of the Failure Rate Based on the Valve’s Performance

During Operation
Number of failures n
λ= = 92
Total time of the valve in operation τ
Example 9.2 For the past seven years, 300 industrial valves have been operating
in a refinery. Since the valves have been in operation for seven years, only five
valves have failed. In this example, what is the average failure rate for the group
of valves?
Answer
The average failure rate per hour for the group of valves is calculated as follows:
Number of failures n 5
λ= = =
Total time of the valve in operation τ 300 × 7 × 365 × 24
= 2 7 × 10 − 7 failures h
It is helpful to know that some people and engineers prefer to use the failure rate
in the unit of the number of failures per year. This is because the number of failures
per hour in this instance is so low. As a result, the average failure rate per year is
calculated as follows:
Number of failures n 5
λ= = =
Total time of the valve in operation τ 300 × 7
= 0 00238 failures year
Specifically, IEC 61511 states that the estimated rates of failure of a subsystem
can be determined through a quantified failure mode analysis of the design using
failure data from a recognized industry source or from previous experience of
using the system in the same environment as it is intended to be used for its
intended application. This needs to be done in a way in which the experience is
sufficient to prove the claimed mean time to failure (MTTF) on a statistical basis
up to a single-sided confidence interval of at least 70%. The calculation of the fail-
ure rate at the upper, (single-sided) confidence interval of 70% can be performed by
applying Eq. (9.3).
Calculation of Failure Rate Representing a 70% Single-Side

Upper Confidence Interval
1
λ= Z0 3 2 n + 1 93
2τ
where:
τ: Accumulated exposure time;

n: Number of reported failures; and
Z0.3 2(n + 1): 30% percentage of the chi-square distribution with 2 (n+1) degrees of
freedom.
In this analysis, the failure rate is calculated as per Eq. (9.2) in order to deter-
mine the final result. In the study, we took into account the total percentage of
failures, the distribution of failures into operation modes, such as dangerous,
safe, detected, and undetected, and finally, the 70% single-sided upper confi-
dence interval.
Final Failure Rate Calculation
Final failure rate h − 1 = Failure rate × Percentage of failure

× Distribution percentage into operational modes
open or closed or both positions
× Failure distribution percentage into dangerous,
safe, detected, and undetected modes
94
Example 9.3 As can be seen in Table 9.5, we have forecasted failure rates for
butterfly valves during the period 2014–2020. All valve failures reported by an
end user to the valve manufacturer over the first three years of valve operation
when the valve was covered by the manufacturer’s warranty. How reliable are but-
terfly valves according to your estimation?
Table 9.5 The number of butterfly valve failures during operations reported by end
users to valve manufacturers from 2014 to 2020.
Number of Number of reported Total hours in the Estimated failure

Year valves failures service rate (h−1)
2014 1,673 — — —
2015 1,961 — — —
2016 3,169 — — —
2017 4,648 5 6,803 8.4 × 10−8
2018 3,243 12 9,778 1.4 × 10−7
2019 3,351 13 11,060 1.3 × 10−7
2020 3,243 1 11,242 1.0 × 10−8
Total 21,288 31 38,883 9.1 × 10−8
Answer
τ: accumulated exposure time = 38,883 hours;
n: number of reported failures = 31;
Z0.3 2(n + 1): 30% of chi-square distribution with 2 (n+1) degrees of freedom.
Using the 70% single-sided upper confidence interval for the calculation of
failure rate as per Eq. (9.3) is provided as follows:
1
λ= Z 0 3 2 n + 1 = 10 − 7 h − 1
2τ
Among the most important things to keep in mind is that there is a relationship
between probability of failure rate and failure rate that can be expressed
in Eq. (9.5).
Relationship Between Probability of Failure and Failure Rate
Probability of failure P = λ × t 95
Example 9.4 There is a failure rate of 500 FIT for safety valves. What is the prob-
ability of the valve failing during one year of operation?
Answer
A FIT of 500 indicates 500 failures in one billion hours of operation
Number of failures 500
λ= = 9 = 5 × 10 − 7 failures h
Total time of the valve in operation 10
Failure
Dangerous (D) Safe (S)
Dangerous Dangerous Safe undetected Safe detected

undetected (DU) detected (DD) (SU) (SD)
Figure 9.5 Failure mode classification.
The duration of operation is one year which is equal to 365 × 24 = 8760 hours.
Using Eq. (9.5) Probability of failure (P) = λ × t = 5 × 10−7 × 8760 =
−3
4.38 × 10
Failure rates (probabilities) are categorized into two types: safe failure (λS) or
dangerous failure (λD) rates. A dangerous failure rate relates to failures that will
prevent achieving the required SIL. Safe failures put the safety function in its safe
state, such as the emergency shutdown of a valve in case of failure. It is noticeable
that SIL addresses only dangerous failures (λD) The dangerous failures can be
either detected or undetected. The possibility of detected dangerous failures and
undetected dangerous failures are shown as λDD and λDU. Equation (9.6) shows
the relationship between λD, λDD and λDU. Failure modes’ classification is shown
in Figure 9.5.
Relationship Between Dangerous Failure Rate and Dangerous

Detected and Dangerous Undetected Failure Rates According
to IEC 61508
λD = λDU + λDD 96
Safe failure rates are divided into safe detected and safe undetected failure rates.
The safe failure rates can be divided into two categories: safe detected failure rates
and safe undetected failure rates. Furthermore, it is important to note that the safe
failure rate is equal to the sum of the safe detected and undetected failure rates, as
indicated in Eq. (9.7).
Relationship Between Safe Failure Rates

λS = λSU + λSD 97
where:
λS: Safe failure rate;

λSU: Safe undetected failure rate;
λSD: Safe detected failure rate;
SD: Estimated percentage of failures belonging to the “safe detected” category;
SU: Estimated percentage of failures belonging to the “safe undetected” category;
DD: Estimated percentage of failures belonging to the “dangerous detected”
category;
DU: Estimated percentage of failures belonging to the “dangerous undetected”
category.
9.4.1.2 Repair Rate

Repair rate is an evaluation of the frequency of successful repairs performed per
unit of time on a failed component. In general, it is signified by the Greek letter
μ (Mu) and is used to calculate the metrics that are specified in the following
paragraphs. In mathematic terms, the repair rate is defined as follows:
Calculation of Repair Rate
1
μ= 98
MTTR
Number of repairs
μ= 99
Total maintenance time
where:
The mean time to recovery (MTTR) is discussed later in this chapter.
9.4.1.3 Mean Time to Failure (MTTF)

During the operation, an industrial valve is installed, and it will function until the
first failure occurs. The average time it takes from the time an operation begins
until the first failure occurs is known as the mean time to failure (MTTF). In order
to determine MTTF, Eqs. (9.10) to (9.11) can be used.
Calculation of MTTF
Total hours of opeartion
MTTF = 9 10
Total number of units
MTTF = The time of the first failure – The time when the system
or component starts operating
9 11
Calculation of MTTF Based on the Failure Rate

∞
MTTF = E t = tf t dt 9 12
0
∞
MTTF = R t dt 9 13
0
1
MTTF = limit lim R s = 9 14
s 0 λ
where:
MTTF: Mean time to failure;

E(t): Expected value of MTTF;
R(t): An indicator of system or component reliability that indicates the probability
that the system or component will not fail in the time period from 0 to “t.” In
other words, it indicates the probability that the given item will survive at least
until time “t”;
s: In math, the Laplace transform variable can be defined as a linear transforma-
tion which takes a parameter like “x” and transforms it into a new and more
complex variable called “s.” As a result of this factor, we may also be able to
convert some of the differential equations into pure algebraic equations;
R(s): It refers to the Laplace transform for the reliability function, R(t);
λ: Failure rate.
9.4.1.4 Mean Time Between Failure (MTBF)

Mean time between failure (MTBF) is defined as the average time duration
between two repairable failures that is calculated in accordance with
Calculation of Mean Time Between Failure (MTBF)
Total hours of operation between two failures

MTBF = 9 15
Total number of incidents
Refer to Figure 9.4, MTBF can be calculated as per Eq. (9.16) alternatively.
An Alternative Way to Calculate Mean Between Failure (MTBF)
MTBF = Mean time to detection MTTD + Mean time to recovery MTTR

+ Mean time tofailure MTTF
9 16
9.4.1.5 Mean Time to Repair and Recovery (MTTR)

An MTTR is the average time required to repair a failed component and restore
it to operational status. This metric includes the time spent during the alert and
diagnostic processes, prior to the initiation of repair activities. The average time
spent solely on the repair process is called the mean time to repair and it is also
shown with MTTR and it is calculated according to Eq. (9.17).
Calculation of Mean Time to Repair (MTTR)
Total hours of maintennace

MTTR = 9 17
Total number of repairs
9.4.1.6 Mean Time to Detection (MTTD)

Mean time to detection (MTTD) is the average time elapsed between the occur-
rence of a component failure and its detection, which is calculated according
to Eq. (9.18).
Calculation of Mean Time to Detection (MTTD)
Total hours of incident detection

MTTD = 9 18
Total number of incidents
Example 9.5 In the case of an industrial valve, the MTBF value is 3,000,000,
MTTF = 2,999,992, and the repair time is eight hours. Calculate the mean time
to detection (MTTD).
Answer
Using Eq. (9.16): MTBF = Mean time to detection (MTTD) + Mean time to recov-
ery (MTTR) + Mean time to failure (MTTF) 3,000,000 = MTTD + 2,999,992 +
8 3,000,000 = MTTD + 3,000,000 MTTD = 0
Thus, the mean time to detect the failure is zero.
9.4.2 Reliability and Unreliability

The reliability of a system such as an industrial valve is the possibility that it will
perform correctly during a specific period of time. The valve does not need to be
repaired or maintained during the period of operation. Additionally, the system
adheres to defined performance according to protocols and specifications. The
more comprehensive definition of the term reliability can be expressed as follows:
The probability that a component, equipment, or system will be able to satisfac-
torily perform its intended function when required to do so, under given condi-
tions, such as design limits and environmental conditions, operating time, and
frequency and accuracy of maintenance for a specific period of time. Following

are four important aspects of this definition:
• The component, equipment, or system’s intended function must be understood.

This aspect refers to the explicit definition of what the device is expected to do in
order to accomplish the task.
• It is necessary to specify the time when the component or system must be oper-
ated. Secondly, this aspect relates to the requested operability of a system, such
as an industrial valve.
• A satisfactory performance of the component or system must be established.
Lastly, there is the issue of evaluating what the device is capable of with regard
to a good performance.
• There must be a clear understanding of the specific design limits. Lastly, there is
the issue of the operating conditions under which the device operates, such as
pressure and temperature.
In the mathematical sense, reliability is the probability that a device will work
correctly in the interval starting from time 0 to time “t” in the time interval. An
example of the reliability chart as a function of time for an industrial valve is illus-
trated in Figure 9.6. When the operating time is increased from “t” to TTF (Time to
Failure), where it is estimated that the device will fail with a probability of almost
100%, the reliability is reduced from 100% to zero. It is predicted that at time “t,”
the probability of failure will be 73%. In other words, the probability that the sys-
tem will operate without failure from 0 to “t” is 0.73. It is not possible to calculate
0.73
Operating time
0
t TTF Time
Failure time
Figure 9.6 Reliability chart as a function of time for an industrial valve.

reliability for time periods longer than “t,” as this is outside the normal operating
time period of the system. It is important to note that the reliability of the system
depends on the failure rate, which will be discussed in more detail later in this
chapter.
Example 9.6 In the case of a newly manufactured and tested industrial valve, it
operates correctly when it is put into service at time t = 0. What is the reliability of
the valve when it is put into service? What is the reliability of the valve over an
indefinite period of time? Is it true that the valve has a 95% reliability rate?
Answer
The reliability of the valve may be almost 100% once it is placed into operation.
Since the valve will eventually fail, its likelihood of success and reliability for
an infinite period will be zero. A statement such as a valve has a 95% reliability
rating is not accurate because reliability should be expressed over a period of time.
It is accurate to state, for example, that the valve has a reliability of 98% over a
period of 10,000 hours.
As an alternative method of calculating the reliability, there is an Eq. (9.19)

which can be used to perform the calculation.
General Reliability Function

1
× dR t = − λ t dt 9 19
Rt
where:
R(t): Reliability function;

λ: Failure rate (1/h).
Equation (9.23) is obtained by integrating both sides of Eq. (9.20) over the time
interval [0, t] as follows:
Rt t
1
× dR t = − λ t dt 9 20
1 Rt 0
t
ln R t = − λ t dt 9 21
0
t
− λ t dt
R t =e 0 9 22
R t = e − λt 9 23
Example 9.7 The failure rate for a valve in an industrial setting can be assumed
to be 0.00002 failures/h. In this case, calculate the valve mean time to failure.
Answer
Based on the failure rate λ, MTTF can be calculated by using Eq. (9.14) as follows:
1 1
MTTF = limit lim R s = = = 50,000 hours
s 0 λ 0 00002
This means that the valve is expected to fail after 50,000 hours of operation, a
very long time after it has been installed.
Example 9.8 It can be assumed that a valve that is used for an industrial appli-
cation will fail on average 0.0004 times/h. Calculate the reliability and mean time
to failures for 15 hours of continuous operation.
Answer
Equations (9.22) and (9.23) are used to calculate the reliability of the valve in the
following way:
t
− λ t dt
R t =e 0 = e − λt = e − 0 0004 × 15 = 2 7183 − 0 006 = 0 9940
∞ ∞ ∞
1
MTTF = R t dt = e − λt dt = e − 0 0004 × t dt = = 2500 hours
0 0 0 0 0004
The reliability of the industrial valve has been calculated to be 0.9940 and the
mean time to failure is estimated to be 2500 hours.
It is not unusual to find engineering systems, such as an industrial valve, com-

posed of a variety of components and configurations that are subject to reliability
analysis. In the case of safety-critical valves, for example, these consist of three
components, which are a valve, an actuator, and a control panel, as will be
explained in greater detail later on in this chapter. As a result of this, it is important
to take into account the overall reliability of the system as a whole. As a matter of
fact, the failure of one component in a system can result in the dysfunction of the
system as a whole. In order to determine the reliability of the whole system, the
combinatorial rules of probability are applied based on the type of subsystems
and elements which are connected in order to comprise the whole. For safety crit-
ical valves, it is very common to use a series network, which consists of three
blocks or units, including a valve, actuator, and control panel. This means that
a safety critical valve is typically comprised of three units acting in series. So in
the event of one component failing, the entire system will also fail. In other words,
1 2 3 m
Figure 9.7 A series of network diagram including “m” components.
the entire system must function normally for the system to be successful.
Figure 9.7 illustrates a series of “m” components that are located in a series.
The reliability of the series systems network is expressed by Eq. (9.24).
Calculation of the Reliability Series

Rs = P X 1 X 2 X 3 … X m 9 24
where:
Rs: Series system reliability or probability of success;

Xi: An event donating the success of unit i, for i = 1, 2, 3, …, and m;
P(X1X2X3… Xm): Probability of occurrence of events X1, X2, X3, …, Xm.
For a system that can fail due to the failure of each individual component, the
series system reliability can be calculated using Eqs. (9.25) and (9.26):
Rs = P X 1 × P X 2 × P X 3 × × P Xm 9 25
where:
P(X) is the occurrence probability of event Xi, for i = 1, 2, 3, …, and m;

Considering Ri = P(xi).
Calculation of the Reliability Series (Main Equation)

m
Rs = Ri 9 26
i=1
where:
Ri is the unit i reliability for i = 1, 2, 3, …, and m.
For identical units where Ri = R, it is possible to simplify the calculation of the

reliability series by using Eq. (9.27) as follows:
Calculation of the Reliability Series for Identical Units

Rs = 1 − m 1− R 9 27
Using Eq. (9.28), we are able to calculate the failure rates of a series network as
the sum of the failure rates of each component.
Calculation of the Failure Rate Series for N Components

N
λS = λi 9 28
i=1
where:
λS: The failure rate of the entire system including all N components in series;
λi: The failure rate of component “i” that could be a number from one to N.
Example 9.9 Imagine that four identical valves are installed in a pipeline in such
a way that failure of any one of them as a single unit can cause problems during the
transportation of crude oil from the plant to the customer. Calculate the reliability
of the crude oil transportation system including these four valves. All four valves
have a reliability of 0.97.
Answer
Using Eq. (9.26), it is possible to calculate the reliability of the entire system by
following the steps as follows:
4
Rs = Ri = R1 × R2 × R3 × R4 = 0 97 × 0 97 × 0 97 × 0 97
i=1
= 0 8853
Since all units have the same reliability, it is possible to use Eq. (9.27) in lieu of
Eq. (9.26) to calculate the total reliability of the crude oil transportation system as
follows:
Rs = 1 − m 1− R = 1 − 4 × 1− 0 97 = 0 88
It is important to know about the failure rate allocation method. In this method,
failure rates are allocated to components of a system based on the knowledge of the
required failure rate for a given system. The assumptions that are associated with
this method are as follows:
• System components for a series configuration

• System components fail independently
• Time to component failure is exponentially disturbed
Accordingly, the calculation of the system failure rate is based on Eq. (9.28). If
the system failure rate is λrs, then the allocation component failure rate must meet
the condition in Eq. (9.29).
N
λ∗i ≤ λsr 9 29
i=1
where:
λ∗i is the failure rate allocated to component i = 1, 2, 3, n according to the required

failure rate λrs.
The following steps are required in this method:
1) Estimate the component failure rates λi for i = 1, 2, 3, …, n using the past data.
2) Calculate the relative weight θi of component i using Eq. (9.30).
Calculation the Relative Weight
λi
θi = N
9 30
λi
i=1
It should be borne in mind that θi indicates the relative failure rate of component
i, and
N
θi = 1 9 31
i=1
By using the relationship given in Eq. (9.32), it will be possible to assign the
required failure rate to component i.
Calculation of Required or Allocated Failure Rate
λ∗i = θi × λrs 9 32
Example 9.10 Let us suppose that there are five independent control subsystems
in series and that the failure rate of the whole system is expected to be 0.0006
failures per hour. Using the experience from the past, the failure rate for
these five subsystems are estimated at λ1 = 0.0001 failures/h, λ2 = 0.0002
failures/h, λ3 = 0.0003, λ4 = 0.0004, and λ5 = 0.0005. Assign the required system
failure rate to five subsystems.
Answer
To begin with, the first step would be to calculate the failure rate for the entire
system according to the past experience as per Eq. (9.28).
N
λS = λi = 0 0001 + 0 0002 + 0 0003 + 0 0004 + 0 0005
i=1
= 0 015 failures h
Using Eq. (9.30), the relative weights for five subsystems are calculated as
follows:
0 0001
θ1 = = 0 00667 λ∗1 = θ1 × λrs = 0 00667 0 0006
0 015
= 0 00004 failures h
0 0002
θ2 = = 0 1333 λ∗2 = θ2 × λrs = 0 1333 0 0006
0 015
0 0003
θ3 = =02 λ∗3 = θ3 × λrs = 0 2 0 0006 = 0 00012 failures h
0 015
0 0004
θ4 = = 0 2667 λ∗4 = θ4 × λrs = 0 2667 0 0006
0 015
0 0005
θ5 = = 0 333 λ∗5 = θ5 × λrs = 0 333 0 0006
0 015
Besides series networks, parallel networks are also

widely used and it represents a system with units as
illustrated in Figure 9.8 that work at the same time
as an alternative to series networks. The success of
1
the system depends on the functioning of at least one
of the units.
A block represents a unit of the system. According to
Eq. (9.33), the failure probability of the parallel system
2
or network can be calculated.
Failure Probability of a Parallel System

3 FP = P X 1X 2X 3… X m 9 33
where:
m FP: Failure probability of the parallel system;

Xi: An event donating the failure of unit i, for
i = 1, 2, 3, …, and m;
Figure 9.8 A parallel
network diagram including P(X1X2X3… Xm): Probability of occurrence of events
“m” components. X1, X2, X3, …, Xm.
In addition, the failure probability of a parallel system can also be calculated

according to Eq. (9.34).
An Alternative Method for Calculating the Failure Probability

of a Parallel System
m
FP = Fi 9 34
i=1
where:
Fi: The failure probability of each unit for unit numbers i = 1, 2, 3, …, m.
The reliability of a parallel network can now be calculated right away with the
help of Eq. (9.35).
Calculation of the Reliability for a Parallel System

m
RP = 1 − F P = 1 − Fi 9 35
i=1
where:
RP: Parallel system reliability.
For identical units with similar failure probability, the calculation of reliability
for a parallel system can be simplified according to Eqs. (9.36) and (9.37).
Calculation of the Reliability of a Parallel System Consisting

of Identical Units
RP = 1 − F m 9 36
m
Since R+F=1 RP = 1 − 1 − R 9 37
Figure 9.9 depicts the reliability plot for parallel systems with m’s of 1, 2, 3, and 4.
Based on the picture here, it can be clearly seen that the parallel system reliability
increases in proportion to the unit reliability and number of units.
Example 9.11 Two identical pipeline valves are installed and operated on a pipe-
line; one of them is the replacement of the other. Ideally, at least one of the valves
should function properly to ensure that the oil is successfully transported from the
1 4
m=
2
3 m=
m=
Parallel network
reliability Rp
1
m=
1
Unit reliability R
Figure 9.9 A reliability plot of a parallel system based on m = 1, 2, 3, and 4 units.
platform to the end user. Assuming that each of these two valves has a reliability of
0.96, calculate the reliability of the system with respect to these two valves.
Answer
In order to calculate the reliability of a parallel system consisting of identical units,
we can turn to Eq. (9.37). The answer is as follows:
m 2
RP = 1 − 1 − R = 1 − 1 − 0 96 = 0 9984
Accordingly, the reliability of the pipeline system is equal to 0.9984 with regard
to these two pipeline valves.
Alternatively, a reliability block diagram as illustrated in Figure 9.10 may be

used for a complex system listing all the units or subsystems as well as their failure
rate values. In this scenario, the analytical methods must be revised as shown
above to take into consideration the large-scale and complex nature of the net-
work. It should be noted that in the illustrated RBD, both parallel and series
connections between various components are present, and this is what is known
as the hybrid model. This can be seen in the figure below, where in this example,
the reliability or any other relevant required data can first be retrieved from sub-
system B by using parallel reliability equations. After that, this complex network
will be simplified to one that has four subsystems in series.
Subsystem B
λB
Subsystem A Subsystem B Subsystem C Subsystem D

λA λB λC λD
Subsystem B
λB
Figure 9.10 An example of a network diagram that includes series and parallel blocks
along with their failure rates.
Example 9.12 An illustration of the arrangement of safety units in an entire

system is given in Figure 9.11. Estimate the reliability of the entire system includ-
ing all seven units.
Answer
First, we need to calculate the level of reliability of subsystem B, which includes four
units in parallel so that we have B1, B2, B3, and B4. As a result, it is possible to cal-
culate the reliability of the parallel network B with the help of Eq. (9.35), as follows:
B4
RB = RP = 1 − F P = 1 − Fi = 1 − 1 − 0 2 1 − 0 8 1 − 0 6 1 − 0 7
i = B1
= 1 − 0 0192 = 0 9808
Subsystem B1
RB1 = 0.2
Subsystem A Subsystem B2 Subsystem C Subsystem D

RA = 0.5 RB2 = 0.8 RC = 0.6 RD = 0.7
Subsystem B3
RB3 = 0.6
Subsystem B4
RB4 = 0.7
Figure 9.11 An example of a network diagram that contains seven series and parallel
blocks, along with their failure rates.
Subsystem A Subsystem B Subsystem C Subsystem D

RA = 0.5 RB = 0.9808 RC = 0.6 RD = 0.7
Figure 9.12 An illustration of a simplified network is shown in the previous figure.
As a result, we are now able to simplify this network diagram to four blocks of
A, B, C, and D as illustrated in Figure 9.12.
The reliability of this network including the four blocks in series is calculated
using Eq. (9.26) which is as follows:
D
Rs = Ri = R A × RB × RC × RD
i=A
= 0 5 × 0 9808 × 0 6 × 0 7 = 0 2060
Thus, the overall reliability of the entire system, including seven units, is 20.60%.
We need to know about unreliability, which can be considered a measure of fail-

ure. The term unreliability is defined as the probability of a system or component
failing within a time interval from 0 to “t.” Unreliability is calculated according
to Eq. (9.38).
Calculation of Unreliability
Unreliability U t = 1 − Reliability t 9 38
It is possible to measure the degree of unreliability of an item by considering the

failure function (F(t)) and calculate this according to Eq. (9.39).
Calculation of the Failure Function

t
F t = Pr T ≤ t = f u du for t > 0 9 39
0
Using Eq. (9.40), we can express the relationship between reliability and failure
functions.
The Relationship Between the Reliability and Failure Functions

∞
R t = Pr T > t = 1 − F t = f u du 9 40
t
Example 9.13 An industrial valve has a reliability of 0.98 over a period of

1000 hours of use. What is the exact unreliability value for the valve during the
given time period?
Answer
Unreliability U(t) = 1 − Reliability (t) = 1 − 0.98 = 0.02
9.4.3 Availability and Unavailability

The concept of reliability, as it was mentioned earlier, refers to the period when a
system or component can operate successfully and during the period when it can-
not be repaired. Availability is a term introduced to describe an operational con-
dition of a device, a system, or its component parts, even if it was corrected either
by maintenance or repair. This means that if a system or device works, it is avail-
able for use regardless of the fact that it has already been repaired or if it has been
running successfully without requiring any kind of repair from the start of oper-
ation. Figure 9.13 is a representation of the relationship between reliability, avail-
ability, unreliability, and unavailability of a device or component in order to
provide successful or unsuccessful operation.
Availability is the function of failure rate, repair time, and operating time that is
calculated in accordance with Eq. (9.41). A good system or component is expected
to have an availability close to one. In order to calculate the unavailability,
Eq. (9.41) is used.
Calculation of Availability
Operating time MTTF
Availability = =
Operating time + Repair time MTTF + MTTR
MTBF
=
MTBF + MTTR
9 41
Reliability Unreliability Unsuccessful
Successful
Unavailability Availability
Figure 9.13 Relationship between reliability, availability, unreliability, and unavailability.

Example 9.14 There is an average repair time of 8 hours for a safety device with
an operating life of 50,000 hours without failure. For a period of 20 hours, calculate
the reliability and unreliability of the device, as well as the availability of the
device.
Answer
The first step is to calculate MTTF by using Eq. (9.10):
Total hours of opeartion 50,000
MTTF = = = 50,000 hours
Total number of units 1
Now, it is possible to calculate the failure rate by using Eq. (9.14):
1 1
λ= = = 0 00002 failures h
MTTF 50,000
Equations (9.22) and (9.23) are used to calculate the reliability of the valve in the
following way:
t
− λ t dt
R t =e 0 = e − λt = e − 0 00002 × 20
= 2 7183 − 0 0004 = 0 99960 = 99 960
Thus, the reliability of the safety device is 99.96%.
Using Eq. (9.38): Unreliability U(t) = 1 − Reliability (t) = 1 − 0.99960 = 0.0004
Equation (9.41):
Operating time
Availability = =
Operating time + Repair time
MTTF 50,000
= = 0 99984 = 99 984
MTTF + MTTR 50,000 + 8
Alternatively, a calculation of the availability can be done using the failure rate
and repair rate provided in Eq. (9.42) as follows:
Calculation of the Availability Based on the Alternative Method
Repair rate μ
Availability = = 9 42
Failure rate + Repair rate λ+μ
Example 9.15 A safety-critical system will have a failure rate of 0.0000005 fail-
ures/h with a repair rate of 0.1 failures/h. What is the availability of the system?
Answer
Repair rate 01
Using Eq. (9.42) Availability = =
Failure rate + Repair rate 0 1 + 0 0000005
= 0 999995 = 99 9995
The availability of a series systems network is expressed by Eq. (9.43).

Calculation of the Availability Series
As = P X 1 X 2 X 3 … X m 9 43
where:
As: Series system availability connected to probability of success;

Xi: An event donating the success of unit i, for i = 1, 2, 3, …, and m;
P(X1X2X3… Xm): Probability of occurrence of events X1, X2, X3, …, Xm.
For a system that can fail due to the failure of each individual component, the
series system availability can be calculated using Eqs. (9.44) and (9.45):
As = P X 1 × P X 2 × P X 3 × × P Xm 9 44
where:
P(X) is the occurrence or availability probability of event Xi, for i = 1, 2, 3, …, and m;

Considering Ai = P(xi).
Calculation of the Availability Series (Main Equation)

m
As = Ai 9 45
i=1
where:
Ai is the unit i availability for i = 1, 2, 3, …, and m.
For identical units where Ai = A, it is possible to simplify the calculation of the

availability series by using Eq. (9.46) as follows:
Calculation of the Availability Series for Identical Units
As = 1 − m 1− A 9 46
Example 9.16 As a result of the intertwining of two safety systems, failure of any
one of them results in the failure of the whole system as a whole. In the first sys-
tem, the failure rate is 0.0002 failures/h, and the rate of repair is 0.1 failures/h. The
second system has a failure rate of 0.0003 failures/h and a repair rate of 0.2 fail-
ures/h. What is the reliability of this system as a whole?
Answer
The first thing we need to do is calculate the availability of both systems by apply-
ing Eq. (9.42) as follows:
Repair rate
A1 = Availability of the first system =
Failure rate + Repair rate
01
= = 0 9980
0 0002 + 0 1
Repair rate
A2 = Availability of the second system =
Failure rate + Repair rate
01
= = 0 49925
0 0003 + 0 2
Equation (9.45) is used to calculate the availability of the entire system as
follows:
2
As = Ai = A1 × A2 = 0 9980 × 0 49925 = 0 49825
i=1
In order to calculate the unavailability of a system as a function of the operating

and repair times, we can use Eq. (9.47).
Calculation of Unavailability
Operating time
Unavailability = 1 − 9 47
Example 9.17 In an industrial application, a valve has been operated for

60,000 hours without any failures. After 60,000 hours of operation for eight hours,
the valve has already been repaired once. What is the value of the availability and
unavailability of this valve?
Answer
Operating time 60,000
Availability = =
Operating time + Repair time 60,000 + 8
= 0 99986 = 99 986
Unavailability = 1 − Availability = 1 − 0 99986 = 0 00014 = 0 014
According to Eq. (9.48), the unavailability of the parallel system or network can
be calculated.
Unavailability of a Parallel System

U P = P X 1X 2X 3… X m 9 48
where:
UP: Unavailability of the parallel system;

Xi: An event donating the failure of unit i, for i = 1, 2, 3, …, and m;

P(X1X2X3 … Xm): Probability of occurrence of events X1, X2, X3, …, Xm.
In addition, the unavailability of a parallel system can also be calculated accord-

ing to Eq. (9.49).
An Alternative Method for Calculating the Unavailability of a Parallel

System
m
UP = Ui 9 49
i=1
where:
Ui: The unavailability of each unit for unit numbers i = 1, 2, 3, …, m.
The availability of a parallel network can now be calculated right away with the
help of Eq. (9.50).
Calculation of the Availability for a Parallel System

m
AP = 1 − U P = 1 − Ui 9 50
i=1
where:
AP: Parallel system availability.
Example 9.18 In parallel with each other, four safety systems are working. The
operating and repair times of these four systems can be found in Table 9.6. How
available is the entire system as a whole?
Table 9.6 The operating and repair times for four subsystems that
operate in parallel.
Subsystem number Operating time (hours) Repair time (hours)
1 50,000 10
2 60,000 12
3 65,000 15
4 70,000 18
Answer
As a first step, according to Eq. (9.41), we need to calculate the availability for each
subsystem as follows:
Availability 1 = =
= 0 9980 = 99 80
Unavailability U1 = 1 − A1 = 1 − 0 9980 = 0 002
Availability 2 = =
= 0 9998 = 99 98
Availability 3 = =
= 0 99977 = 99 977
Availability 4 = =
= 0 99979 = 99 979
Equation (9.50) can be used to determine the availability of a parallel network
immediately as follows:
4
AP = 1 − U P = 1 − U i = 1 − 0 002 0 0002 0 00023 0 00021
i=1
= 1 − 0 00000000000001932 = 0 99999999
9.5 Safety Integrity Level (SIL) Calculations
9.5.1 SIL
SIL is a term attached to international standards such as IEC 61508 which provides
suppliers and end users with a common framework to design products and systems
when it comes to safety-related applications. The IEC has published IEC 61508 as
an international standard for electrical, electronic, and programmable electronic
safety-related systems. With the help of SIL, it is possible to specify and design
safety systems in a scientific, numerical manner, in order to allow for a quanti-
tative evaluation of failure risk. A SIL is a scientific and numerical method for
Table 9.7 SIL Levels according to IEC 61508 for low-demand conditions.
SIL PFD Risk reduction factor
SIL 1 0.1–0.01 10–100

SIL 2 0.01–0.001 100–1,000
SIL 3 0.001–0.0001 1,000–10,000
SIL 4 0.0001–0.00001 10,000–100,000
defining and evaluating safety systems, which quantifies the risk of failure. As can
be seen in Table 9.7, SIL levels range from 1 to 4, and it is significant to note that
these values are for low-demand operations. In the following section, we will dis-
cuss these modes of operation in additional detail. A SIL 4 system has the lowest
probability of failure or probability of failure on demand (PFD) and the highest
reliability. PFD is a measure of the effectiveness of a safety function. The likelihood
that the safety function will not work when needed is represented by this value.
Consequently, SIL 1 has the highest probability of failure and the lowest level
of safety. For ESD valves, SIL 2 or 3 is typically required, as described earlier.
In relation to a reliability analysis for a safety system, IEC 61508 describes two or
three types of requirements. The IEC 61508 categorizes the frequency of demands
into three categories, called demand modes of operation.
Low-demand: Safety system requirements or functions are occurring less fre-

quently than once per year. Some characteristics of low-demand systems
include: a safety system’s response is rarely needed. It may also be possible
for a failure to remain hidden for some time, so the safety system is not func-
tioning for this reason.
High-demand: Safety system requirements or functions are occurring more
frequently than once per year.
Continuous mode: The demands or functions of a safety system are always
present.
As a general rule, high-demand and continuous modes of operation share two

main characteristics. First, the safety system provides a response to prevent failure
caused by risks occurring frequently or constantly. Furthermore, the failure of
the safety system may directly result in a hazardous incident if the system
malfunctions.
For a high-demand mode or continuous mode of operation, PFH is calculated as
the average frequency of dangerous failure of the safety function. Table 9.8 pro-
vides the correlation between PFH and SIL. In Table 9.9, SILs are correlated with
the associated average PFDs and PFHs.
Table 9.8 SIL Levels according to IEC 61508 for high-demand and continuous
operation conditions.
SIL PFH Risk reduction factor
SIL 1 0.00001–0.000001 100,000–1,000,000

SIL 2 0.000001–0.0000001 1,000,000–10,000,000
SIL 3 0.0000001–0.00000001 10,000,000–100,000,000
SIL 4 0.00000001–0.000000001 100,000,000–1,000,000,000
Table 9.9 Levels of SIL according to IEC 61508 for low and high demand as well as
continuous operation.
Average PFD (low-demand Average PFH (high-demand or

SIL mode of operation) continuous mode of operation)
1 ≥10−2–<10−1 ≥10−6–<10−5
−3 −2
2 ≥10 –<10 ≥10−7–<10−6
3 ≥10−4–<10−3 ≥10−8–<10−7
−5 −4
4 ≥10 –<10 ≥10−9–<10−8
9.5.2 Probability of Failure on Demand (PFD)

PFD stands for probability of failure on demand. In many cases, safety systems are
designed to work in the background, monitoring a process. However, they don’t
take any action until a safety limit has been exceeded, at which point they must
take action to keep the process safe. As with reliability, PFD is also a probability
value ranging from 0 to 1, inclusive. For example, a PFD of zero means there is no
likelihood of failure. There is no question that a system designed for high reliability
must exhibit a small PFD value (very close to zero). There’s only one way to
determine how low the PFD has to be, and that’s based on how necessary the
component or system is to the fulfilment of our human needs.
IEC 61508 presents a simplified formula for the calculation of PFD that is
given based on Eq. (9.51).
Calculation of PFD Based on IEC 61508
avg = λD,G t GE
PFDG 9 51
where:
λD,G: Group failure frequency of dangerous failures;

tGE: Group equivalent mean downtime.
Note: The next section will explain the concept of mean downtime.
In addition, PFD can be calculated according to Eq. (9.52) as follows:
Frequency of tolerable accidents
PFDavg = 9 52
Frequency of accidents without protection
As a result of Eq. (9.53), there is a relationship between availability and PFD.
Relationship Between Availability and PFD
Availability = 1 − PFDavg 9 53
According to Eq. (9.54), there is a correlation between risk reduction factor (RRF)
and probability of loss (PFD).
Correlations Between PFD and RRF
1
RRF = 9 54
PFDavg
Example 9.19 The safe PFD for a SIS is 0.001/year, while the dangerous prob-
ability is 0.0001/year. Calculate the values of the PFD, availability, and risk reduc-
tion factor.
Answer
PFDavgS = 0 001 year, and PFDavgD = 0 0001 year
PFDavg = PFDavgS + PFDavgD = 0 001 + 0 0001 = 0 0011 year
Using Eq. (9.53) Availability = 1 − PFDavg = 1 − 0.0011 = 0.9989 = 99.89%

1 1
Using Eq. (9.54) RRF = = = 909 09
PFDavg 0 0011
In this regard, we observe that the values of the PFD, the availability and the risk
reduction factor are 0.0011/year, 99.89%, and 90.09%, respectively.
9.5.3 Mean Downtime

The mean downtime of a system due to both dangerous detected and undetected
failures tCE includes the mean downtime of the system due to dangerous
λD
λDU λDD
tc1 = T1 + MRT tc2 = MTTR
2
tCE
IEC 325/2000
Figure 9.14 1oo1 reliability block diagram.
undetected failures tc1 plus the mean downtime of the system due to dangerous
detected failures tc2 as shown in Figure 9.14. tc1 and tc2 are calculated using
Eqs. (9.55) and (9.56):
Calculation of the Mean Downtime of the System Due to

Dangerous Undetected Failures As Well As Dangerous Detected
Failures (Refer to IEC 61508)
T1
t c1 = + MRT 9 55
2
t c2 = MTTR 9 56
where:
T1: Proof test interval (hour);

MTTR: Mean time to restoration (hour);
MRT: Mean repair time (hour).
MTTR = Time to detect the failure + Time spent before starting the repair
+ The effective time to complete the repair
+ The time before the component is put back into operation
9 57
MRT = MTTR − Time to detect the failure 9 58

Figure 9.14 shows the reliability block diagram for architecture 1oo1:
Note 1
There is a distinction to be made between how MTTR is defined and calculated
in Eq. (9.57) and what was explained earlier in this Chapters 9−17.
Note 2
A typical industrial valve that is a component of SIS is composed of three
parts; the valve, the actuator, and the control panel. Generally speaking,
an actuator is a mechanical device that is installed on the top of the valve
to operate (open and close) the valve and control the valve movement. In
order to control the flow of hydraulic fluid through the actuator, a control
panel that consists of valves, a pressure gauge and tubes is used. As a side
note, we should mention that the majority of actuators that are part of SIS
work with hydraulic oil. It is true that these actuators convert hydraulic
power to mechanical motion to operate valves through hydraulic power.
As part of our SIL analysis, we use the 1oo1 architecture since these three
components are considered as a single unit. Based on Figure 9.15, the archi-
tecture of the system is composed of a single channel in which any failure
of any part compromises the safe operation of the system. The reliability
equations are simplified when using the 1oo1 architecture, as opposed to
scenarios where other architectures are employed.
Operation failure
OR
Valve OR
Actuator Control panel
Figure 9.15 Fault tree for safety critical valves connected to the SIS.
Placing values of tc1 and tc2 from Eqs. (9.55) and (9.56) into Eq. (9.59):
λDU T1 λDD
t CE = + MRT + MTTR 9 59
λD 2 λD
9.5.4 Diagnostic Coverage

The number of dangerous failures detected by automatic online diagnostic tests is
called diagnostic coverage (DC). DC is the detected dangerous failure rate divided
by the total rate of the dangerous failures, using Eq. (9.60).
Calculation of Diagnosis Coverage (DC)
λDD
DC = 9 60
λD
Therefore, λDD = λD × DC 9 61
λDU = λD − λDD = λD − λD × DC λDU = λD 1− DC 9 62
9.5.5 Safe Failure Fraction (SFF)

The average PFD (PFDG) is calculated in accordance with Eqs. (9.63) and (9.64):
PFDG = λDU + λDD t CE 9 63
T1
PFDG = λD × t CE = λDU + MRT + λDD × MTTR 9 64
2
Another key safety parameter is called the safe failure fraction (SFF), which veri-
fies the suitability of a device for use in a system for a particular SIL. In fact, SFF is
a term used to describe the proportion of failures that are safe (not hazardous) plus
hazardous failures that are possible to detect by applying a test. On the basis of
Eq. (9.65), the SFF is calculated as the ratio of the average safe failure rate plus
the average dangerous detection rate to the average total failure rate:
Calculation of Diagnosis Safe Failure Fraction (SFF)
λS + λDD λtotal − λDU

SFF = = 9 65
λS + λD λtotal
SFF and SIL can be correlated based on the table for different levels of hardware
fault tolerance. In terms of hardware fault tolerance, it can either be N = 0, 1, or 2
and can either be type A or B. The type B model provides a higher level of safety for
Table 9.10 Correlation between SIL and SFF for hardware fault tolerance – type A (IEC).
Hardware fault tolerance – type A
Safe failure fraction (SFF) 0 1 2
SFF < 60% SIL 1 SIL 2 SIL 3

60% < SFF < 90% SIL 2 SIL 3 SIL 4
90% < SFF < 99% SIL 3 SIL 4 SIL 4
SFF ≥ 99% SIL 3 SIL 4 SIL 4
more complex diagrams. Type A can be used with a single block diagram and sim-
ple 1oo1 architecture, including all safety critical valves. In Table 9.10, the maxi-
mum allowable SIL for a safety function implemented by a type A safety is
illustrated for the different hardware types used. The hardware fault tolerance
of N indicates that the loss of safety functions could be caused by a fault of
N + 1. Because architecture-type 1oo1 has been selected in this chapter, which
means that failure of any component may result in the failure of the entire system,
the hardware fault tolerance value has been set to 0 in order to achieve the highest
level of safety and reliability possible.
Example 9.20 On an offshore platform, a 20 hydraulically actuated ball valve is

connected to an ESD. This valve is used for the transportation of the produced gas
from offshore to the final user. This valve is composed of a carbon steel body, and
it can handle a maximum pressure of 250 bar. There are three failure rates for
the valve, hydraulic actuator, and control panel summarized in Table 9.11. The
following parameters need to be calculated:
1) DC for valves, actuators, control panels, and the entire system.

2) Assuming that the proof test interval is 9760 hours, the mean repair time is
24 hours, and the time needed to detect the problem is zero, calculate the aver-
age time that the whole system is unable to run.
Table 9.11 The failure rate values for the 20 valve, its actuator, and its control panel.
Component λS λD λDU λDD
20 valve 0 1.02 × 10−7 3.3 × 10−8 6.9 × 10−8

Actuator 0 4 × 10−8 1.7 × 10−8 2.3 × 10−8
−9 −7 −7
Control panel 1.3 × 10 1.3 × 10 1.3 × 10 0
3) The PFD for the entire system.

4) Safe failure fracture (SFF) for the entire system.
A minimum SIL 3 is also required by the end user for this valve that is actuated.
Does the actuated valve meet the expectation that is required?
Answer
In the first step, we should calculate the safe failure rate λS, dangerous failure rate
λD, dangerous detected λDD, and undetected failure rate λDU, for the entire system.
λS Total = λS Valve + λS Actuator + λS Control panel
= 0 + 0 + 1 3 × 10 − 9 = 1 3 × 10 − 9
λD Total = λD Valve + λD Actuator + λD Control panel
= 1 02 × 10 − 7 + 4 × 10 − 8 + 1 3 × 10 − 7 = 2 72 × 10 − 7
λDU Total = λDU Valve + λDU Actuator + λDU Control panel
= 3 3 × 10 − 8 + 1 7 × 10 − 8 + 1 3 × 10 − 7 = 1 8 × 10 − 7
λDD Total = λDD valve + λDD Actuator + λDD Control panel
= 6 9 × 10 − 8 + 2 3 × 10 − 8 + 0 = 9 2 × 10 − 8
Now, it is possible to calculate diagnosis coverage for the valve, actuator, control
panel, and the entire system separately by using Eq. (9.60) as follows:
λDD 6 9 × 10 − 8
For valve DC = = = 67 65
λD 1 02 × 10 − 7
λDD 2 3 × 10 − 8
For actuator DC = = = 57 5
λD 4 × 10 − 8
λDD 0
For control panel DC = = =0
λD 1 3 × 10 − 7
λDD 9 2 × 10 − 8
For entire system DC = = = 33 82
λD 2 72 × 10 − 7
The next step would be to conduct a calculation to obtain the mean downtime of
a system due to both dangerous detected and undetected failures tCE for the entire
system using Eq. (9.59).
T1: Proof test interval (hour) is 9760 hours

MTR: Maintenance repair time = 24 hours
Mean repair time (MRT) = MTTR − time to detect the failure
Time to detect the failure is zero and thus, MTTR = MTR = 24 hours
λDU T1 λDD
t CE = + MRT + MTTR
λD 2 λD
1 8 × 10 − 7 9760 9 2 × 10 − 8
t CE = + 24 + × 24
2 72 × 10 − 7 2 2 72 × 10 − 7
= 3245 + 8 12 = 3253 hours
Thus, the average time that the whole system is unable to run is 3253 hours.
Now, it is possible to calculate the PFD for the entire system based on Eq. (9.64)
as follows:
PFDG = λD × t CE = 2 72 × 10 − 7 × 3253 = 8 85 × 10 − 4
The next step is to calculate the SFF for the whole system using Eq. (9.65):
λS + λDD 1 3 × 10 − 9 + 9 2 × 10 − 8
SFF = = = 34 13
λS + λ D 1 3 × 10 − 9 + 2 72 × 10 − 7
According to Table 9.10, the SFF is less than 60%, considering zero fault toler-
ance and type A. This means that the level of safety achieved by the system is not
sufficient to meet the requirements of the project.
Example 9.21 By applying more restrictive tests and accurate sensors to monitor
the failures on the 20 valve considered in the previous example, the failure rate
values are modified to the values provided in Table 9.12.
There are a number of parameters that need to be calculated:
1) DC for valves, actuators, control panels, and the entire system.

2) Based on the assumption that the proof test interval, the mean repair time, and
the time to detect the problem remain unchanged, calculate the average period
of time that the whole system cannot be operated.
3) For a given system, the PFD is calculated.
4) SFF for the entire system.
There is also a requirement from the end user for this valve that is actuated to
have a minimum SIL 3. How well does this valve fulfil the requirements set by the
end user?
Table 9.12 Modified failure rate values for the 20 valve, its actuator, and its control panel.
Component λS λD λDU λDD
20 valve 0 8.65 × 10−8 2.57 × 10−8 5.99 × 10−8

Actuator 0 2.58 × 10−8 2.3 × 10−9 2.35 × 10−8
−6 −7 −7
Control panel 1.05 × 10 4.95 × 10 4.95 × 10 0
Answer
In the first step same as the previous example, we should calculate the safe failure
rate λS, dangerous failure rate λD, dangerous detected λDD, and undetected failure
rate λDU, for the entire system.
λS Total = λS Valve + λS Actuator + λS Control panel
= 0 + 0 + 1 05 × 10 − 6 = 1 05 × 10 − 6
λD Total = λD Valve + λD Actuator + λD Control panel
= 8 55 × 10 − 8 + 2 58 × 10 − 8 + 4 95 × 10 − 7 = 6 06 × 10 − 7
λDU Total = λDU Valve + λDU Actuator + λDU Control panel
= 2 57 × 10 − 8 + 2 3 × 10 − 9 + 4 95 × 10 − 7 = 5 23 × 10 − 7
λDD Total = λDD Valve + λDD Actuator + λDD Control panel
= 5 99 × 10 − 8 + 2 35 × 10 − 8 + 0 = 8 34 × 10 − 8
Now, it is possible to calculate diagnosis coverage for the valve, actuator, control
panel, and the entire system separately by using Eq. (9.60) as follows:
λDD 5 99 × 10 − 8
For valve DC = = = 69 25
λD 8 65 × 10 − 8
λDD 2 35 × 10 − 8
For actuator DC = = = 91 09
λD 2 58 × 10 − 8
λDD 0
For control panel DC = = =0
λD 4 95 × 10 − 7
λDD 8 34 × 10 − 8
For entire system DC = = = 13 76
λD 6 06 × 10 − 7
λDU T1 λDD
t CE = + MRT + MTTR
λD 2 λD
5 23 × 10 − 7 9760 8 34 × 10 − 8
t CE = + 24 + × 24 = 4235 30 hours
6 06 × 10 − 7 2 6 06 × 10 − 7
Thus, the average time that the whole system is unable to run is 4235.30 hours.
Now, it is possible to calculate PFDG using Eq. (9.64) as follows:
PFDG = λD × t CE = 6 06 × 10 − 7 × 4235 30 = 2 56 × 10 − 3
The next step is to calculate the SFF using Eq. (9.65):
λS + λDD 1 05 × 10 − 6 + 8 34 × 10 − 8
SFF = = = 68 44
λS + λD 1 05 × 10 − 6 + 6 06 × 10 − 7
Based on the results shown in Table 9.10, the SFF is between 60 and 90%. As a
result, the system’s safety level is SIL 2, which is not acceptable to the end user.
9.6 Condition Monitoring (ValveWatch) 347
Hence, in order to improve the safety and reliability of the system, various tech-
niques need to be used such as more stringent tests and/or using sensors as part of
condition monitoring or ValveWatch.
9.6 Condition Monitoring (ValveWatch)
Probably, the first literature about the use of online condition monitoring or Valve-
Watch on the Norwegian continental shelf was written in 2002. As shown in
Figure 9.16, a complete set of pressure sensors is installed on a safety-critical valve
in order to illustrate the type of hardware included in the ValveWatch system. The
following is a brief explanation of each of the sensors:
1) Actuator pressure sensor: It is a sensor attached to the pressurized part of the

actuator and measures the air or hydraulic pressure applied to the actuators. In
case of a reduction in actuator pressure compared to normal circumstances (e.g.
160–200 bar hydraulic oil pressure), there could be various reasons for this,
such as external or internal leakage in the actuator, or perhaps a problem with
the actuator control system.
2) Strain gauge: There is a sensor installed on top of the valve which measures
the torque or force that is exerted between the valve and the actuator. In phys-
ics, torque is a measure of the force that can cause an object to rotate around its
axis. A practical way to calculate the torque is to multiply the force by distance
according to Eq. (9.66). It has been observed that when the operating force
detected by this sensor is higher than normal, it is likely caused by problems
with stem packing such as misalignment or overtightening of the stem packing
Position
Actuator transmitter Acoustic
pressure
leak sensor
sensor
Dynamic
Torque/force pressure
strain gage leak sensors
Figure 9.16 ValveWatch sensors by MRC Global. Source: MRC Global Inc.
which causes extra friction with the stem. In order to prevent leakage from the
stem to the external environment, valve packings are used as stem sealings. The
strain gauge is able to detect the bending of the stem, which causes packing
wear and external leakage, to occur. If an operator applies extra torque on
the stem higher than the maximum stem load capacity, also known as the
MAST (Maximum Allowable Stem Torque), the operator may damage and
bend the stem as well as wear it out. The bending and damage to the valve stem
lead to an external leakage from the valve.
3) The valves use dynamic pressure sensors in order to detect leakage inside the
valves. Figure 9.16 shows two dynamic pressure leak sensors, one located on
the pipe and the other located in the valve cavity. Each of these sensors is sen-
sitive to noise levels and can convert them into the value of pressure.
Calculation of the Valve Torque

τ=F×r 9 66
where:
τ: the magnitude of the torque to open or close the valve (Nm);

F: the force applied on the valve stem (N);
r: the distance from the stem to the center of closure member (m).
9.1 In an offshore oil rig, an industrial valve is installed on a piping system that
carries crude oil. It is estimated that the MTTF is 2,000,000 hours and the
mean time to restore (MTTR) is 10 hours. What is the sentence that is not
correct regarding the safety and reliability of the valve?
A Valve availability is more than 99%.
B Statistically, the unavailability of the valve is less than 0.1%.
C Valve reliability of one during its first 2,000,000 hours of use.
D There is a 0.5% reliability of the valve after 2,000,000 hours of operation.
Answer
The availability is calculated according to Eq. (9.41) as follows:
Operating time
Availability =
2,000,000
= = 99 9995
2,000,000 + 10
Therefore, option A is the correct choice. The next step will be to calculate
unavailability as follows:
Unavailability = 1 − Availability = 1 − 0 999995 = 0 000005 = 0 0005
Therefore, option B is correct as well. It should be noted that Option C is the

correct answer because no failures occurred during the first 2,000,000 hours
of the valve’s operation. Option D is incorrect because we have no informa-
tion regarding the operational reliability of the valve after 2,000,000 hours of
use. So we cannot provide any reliability data for the valve.
9.2 There are 2000 industrial valves that have been working in an industrial
plant for 10 years. During this time, only 20 of them have failed. In which
of the following options would you find the correct failure rate and MTTF
values for this group of valves?
A λ = 1.14 × 10−7, MTTF = 0.88 × 107
B λ = 1.14 × 10−6, MTTF = 0.88 × 107
C λ = 1.14 × 10−7, MTTF = 0.88 × 106
D λ = 1.14 × 10−6, MTTF = 0.88 × 106
Answer
The average failure rate for the group of valves is calculated as follows:
Number of failures
λ=
Total time of the valve in operation
n 20
= = = 1 14 × 10 − 7 failures h
τ 2000 × 10 × 365 × 24
1 1
MTTF = = = 0 88 × 107
λ 1 14 × 10 − 7
Therefore, option A is the correct answer.
9.3 Which of the following statements concerning SIL formulas and calculations
is correct?
A The probability of dangerous failure is equal to the sum of the detected
dangerous failure rate and the undetected dangerous failure rate.
B The correlation between the SFF and SIL cannot be established.
C DC is the detected dangerous failure rate divided by the total rate of the
failures.
D If there is a dangerous failure, the safety function goes into its safe state,
such as closing a valve in an emergency situation.
Answer
The correct choice is option A. It would be wrong to choose option B because
there could be a correlation between the SFF and the SIL. Option C is incor-
rect because DC is the ratio between the number of dangerous failures
detected by the number of dangerous failures as a whole. It is incorrect to
choose option D since safe failures put the emergency shutdown valve in
a safe position.
9.4 Consider a pipeline that contains four identical valves and assume that each
valve has a reliability of 0.93. In the event of one of the valves failing, the
whole pipeline system will cease to function. In terms of safety, how reliable
is the pipeline system with regard to these four identical valves?
A 81.2%
B 73.4%
C 90.5%
D 92.8%
Answer
Four identical valves are arranged in series in the system. Equation (9.27) is a
formula we can use to determine the reliability of a series system or network
that consists of identical units. The answer is:
Rs = 1 − m 1− R = 1 − 4 × 1− 0 93 = 0 72
A second method to calculate the reliability of the entire system would be
to multiply all reliability numbers together as shown in the following
calculation:
4
Rs = Ri = R1 × R2 × R3 × R 4
i=1
= 0 93 × 0 93 × 0 93 × 0 93
= 0 7480
There is an average reliability of 73.4% calculated by using two methods. So
option B is the correct answer.
9.5 In a piping system, two valves are installed; the first is located in the main
line, and the second is located in the bypass line. The two identical valves
work independently, and the probability of failure is 0.05 for each. Assuming
that the system will function successfully if only one valve is functioning,
what is the reliability of the entire system?
A 99.75%
B 98.85%
C 97.60%
D 95.50%
Answer
The system consists of two parallel valve units. As a first step, we have to
determine the reliability of each valve by identifying the probability of failure
for each as follows:
R Reliability = 1 − F Failure probability = 1 − 0 05 = 0 95
To determine the reliability of a parallel system consisting of identical units,

we can use Eq. (9.37). The answer is:
m 2
RP = 1 − 1 − R = 1 − 1 − 0 95 = 0 9975
Accordingly, the reliability of the entire system including two valves is equal
to 0.9975 = 99.75%. Thus, option A is the correct answer.
9.6 There are seven independent valves installed in series in a piping system.
A failure rate of 0.009 failures/h is expected from the entire piping system,
which includes seven valves. As a result of past experience, the failure rates
for the first to seventh valve have been determined to be 0.001 failures/h,
0.002 failures/h, 0.003 failures/h, 0.004 failures/h, 0.005 failures/h, 0.006 fail-
ures/h, and 0.007 failures/h, respectively. Which of the following statements
is correct?
A The allocated failure rate to the first valve is 6.4 × 10−5 failures/h.
B The allocated failure rate to the seventh valve is 1.93 × 10−4 failures/h.
C The allocated failure rate to the second valve is 1.29 × 10−4 failures/h.
D The allocated failure rate to the third valve is 9.64 × 10−5 failures/h.
Answer
In order to begin, the first step would be to calculate the failure rate for the
entire system based on past experience as per Eq. (9.28).
N
λS = λi = 0 0001 + 0 0002 + 0 0003 + 0 0004 + 0 0005 + 0 0006
i=1
+ 0 0007 = 0 028 failures h
Using Eqs. (9.30) and (9.32), the relative weights and allocated reliability
values for five subsystems are calculated as follows:
0 0001
θ1 = = 0 00357 λ∗1 = θ1 × λrs
0 028
= 0 00357 0 009 = 3 2 × 10 − 5 failures h
0 0002
θ2 = = 0 00714 λ∗2 = θ2 × λrs
0 028
= 0 00714 0 009 = 6 4 × 10 − 5 failures h
0 0003
θ3 = = 0 010714 λ∗3 = θ3 × λrs
0 028
= 0 010714 0 009 = 9 64 × 10 − 5 failures h
0 0004
θ4 = = 0 0142857 λ∗4 = θ4 × λrs
0 028
= 0 0142857 0 009 = 1 29 × 10 − 4 failures h
0 0005
θ5 = = 0 017857 λ∗5
0 028
= θ5 × λrs = 0 017857 0 009
= 1 61 × 10 − 4 failures h
0 0006
θ6 = = 0 02143 λ∗4
0 028
= θ4 × λrs = 0 02143 0 009
= 1 93 × 10 − 4 failures h
0 0007
θ7 = = 0 025 λ∗5
0 028
= θ5 × λrs = 0 025 0 009
= 2 25 × 10 − 4 failures h
Thus, option D is the correct answer.
9.7 The probability of detecting a total failure for an actuated valve connected to
SIS by a traditional method of inspection is 35%. In order to detect defects in
the actuated valves more reliably, three sensors are added to the valves and
actuators as part of a process called condition monitoring or ValveWatch. As
a result, these sensors, which are labelled A, B, and C, can identify 66.66%,
15.83%, and 20.83% of system defects, respectively. Can you identify which of
the following statements could be the outcome of using the ValveWatch for
the valve in this case?
A Reducing the SIL
B Increasing the DC
C Reducing the SFF
D Reducing the mean time to failure (MTTF)
Answer
ValveWatch results can be applied to failure detection more easily, thereby
increasing the likelihood of dangerous detected failures λDD being detected,
thereby increasing the DC value. As a result, the correct answer is option B.
Both options A and C are incorrect because increasing DC will result in an

increase in reliability, SIL and SFF. Option D is not correct either because the
MTTF value is independent of the improvement in failure detection.
9.8 There are values of 0.1 and 0.01, respectively, for safe failure rates and dan-
gerous failure rates for an actuated valve connected to SIS. What are the
values of the total mean time to failure as well as the total mean time
between failures?
A MTTFtotal = MTBFtotal = 100 years
B MTTFtotal = MTBFtotal = 10 years
C MTTFtotal = MTBFtotal = 9.9 years
D MTTFtotal = MTBFtotal = 9.09 years
Answer
λS = 0 1 year and λD = 0 01 year
λtotal = λD + λS = 0 11 year
1 1
MTTFtotal = MTBFtotal = = = 9 09 years
λtotal 0 11
Thus, option D is the right choice.
9.9 The following values are provided for the safe and dangerous failure rates:
λDD = 0 14 year, λDU = 0 04 year, λSD = 0 22 year, λSU = 0 5 year
For the second condition, all these values remain unchanged, except for the
failure rate of dangerous undetected value, which is changed to 0.4/year.
What is the difference between the calculated safe failure fraction (SFF)
in two conditions?
A 20.2%
B 27.3%
C 32.4%
D 38.6%
Answer
Equation (9.65) is used to calculate the SFF in both conditions as follows:
λS + λDD λSD + λSU + λDD
First condition SFF = =
λS + λD λSD + λSU + λDD + λDU
0 22 + 0 5 + 0 14 0 86
= = = 0 9555 = 95 55
0 22 + 0 5 + 0 14 + 0 04 0 90
λS + λDD λSD + λSU + λDD

Second condition SFF = =
λS + λD λSD + λSU + λDD + λDU
0 22 + 0 5 + 0 14 0 86
= = = 0 6825 = 68 25
0 22 + 0 5 + 0 14 + 0 4 1 26
The difference between SFF in first and second conditions = 95.55% −
68.25% = 27.3%
Therefore, option B is the correct answer.
9.10 The failure rate of a safety component is 200 FIT. What is the likelihood of
this device failing in ten years?
A 1.43 × 10−2
B 1.77 × 10−2
C 1.98 × 10−2
D 1.75 × 10−2
Answer
The failure rate of 200 FIT means that 200 failures have happened over a
billion hours.
Number of failures 200
λ= = 9 = 2 × 10 − 7 failures h
Total time of the valve in operation 10
The duration of operation is ten years which is equal to 365 × 24 × 10 =
87, 600 hours.
Using Eq. (9.5) Probability of failure (P) = λ × t = 2 × 10−7 ×
−2
87, 600 = 1.75 × 10
Thus, option D is the correct answer.
Further Reading
Basilio, A., Landrini, G., Capelle, T.V. et al. (2010). Plant Engineering and Maintenance
According to IEC 61508 and IEC 61511 Standards, 3e. International Technology for
Safety.
Colombo, D., GBA, L., Pereira, D.R. et al. (2020). Regression-based finite element
machines for reliability modeling of downhole safety valves. Reliability
Engineering & System Safety 198: 106894.
Gross, R.E. (2008). Reliability testing of pressure relief valves. Proceeding Paper: ASME
Pressure Vessel and Piping Conference, San Diego, CA, USA, 57–166. Paper number:
PVP2004-2610 (25–29 July 2004).
Further Reading 355
International Electrotechnical Commission (IEC) (2010). IEC 61508, Functional Safety

of Electrical/Electronic/Programmable Electronic Safety-Related Systems, 2e. Geneva:
International Electrotechnical Commission (IEC).
International Electrotechnical Commission (IEC) (2016). IEC 61511, Functional
Safety – Safety Instrumented System for the Process Industry Sector, 2e. Geneva:
International Electrotechnical Commission (IEC).
Jeeves, S. (2016). Valve monitoring across disciplines and operations. MRC Global.
Valve World Conference, Dusseldorf. Presentation # PV16035 (29 November–1
December 2016).
Juvik, T., Carr, R., Hermansen, T., and Hale, S. (2002). Online valve monitoring system
used on off-shore platforms in the North Sea. 21st International Conference on
Offshore Mechanics and Artic Engineering. Oslo: ASME. Paper No. OMAE2002-28403
(23–28 June 2002).
Kim, J.H., Lee, J.H., and Jo, Y. (2018). A study on monitoring method for accident
prevention of industrial valves. Proceeding Paper: International Science and
Technology Conference, Paris, France (18–20 July 2018).
Lemme, A. and Furseth, H.J. (2019). Expansion of the Condition Monitoring Strategy for
ESD and PSD Valves on Johan Sverdrup Phase 1 – A Case Study in Equinor ASA.
University of Stavanger, Faculty of Science and Technology.
Lundteigen, M.A. and Rausand, M. (2007). The Effect of Partial Stroke Testing on the
Reliability of Safety Valves. Department of Production and Quality Engineering,
NTNU, ESREL.
Molnes, E., Sundet, I., and Vatan, J. (1992). Reliability of Surface Controlled Surface
Safety Valve. Phase 4. ETDEWEB. Technical Report. Report Number STF-75A91038.
MRC Global (2020). ValveWatch-sensors. https://www.mrcglobal.com/Global-Region/
Products/ValveWatch/Sensors (accessed 13 April 2022).
Oxford: Elsevier.
Norskolje & Gass (2018). Application of IEC 61508 and IEC 61511 in the Norwegian
Petroleum Industry (Recommended SIL Requirement). Norwegian Oil and Gas
Association. Rev. 03.
Rausand, M. and Vatn, J. (1998). Reliability modeling of surface controlled subsurface
safety valves. Reliability Engineering & System Safety 61 (1–2): 159–166.
Reza, M. (2019). System reliability and availability calculation. https://www.bmc.com/
blogs/system-reliability-availability-calculations/#:~:text=Reliability%20is%
20calculated%20as%20an,be%20used%20to%20calculate%20Reliability (accessed
15 April 2022).
Sangolt, J.K., Bjoroy, M., and Erstad, T. (2019). Application of Industry 4.0 Concepts for
Increasing Reliability of Valves. Department of Mechanical and Marine Engineering,
Western Norway University of Applied Sciences.
SINTEF (2006). Reliability Prediction Method for Safety Instrumented System – The PDS
Method Handbook. SINTEF. ISBN: 82-14-03899-5.
Smith, D.J. (2011). Reliability, Maintainability and Risk: Practical Methods for Engineers.
Butterworth-Heinemann. ISBN: 978-0-08-096902-2.
Sotoodeh, K. (2019). Safety integrity level in valves. Journal of Failure Analysis and
Prevention 19: 832–837. https://doi.org/10.1007/s11668-019-00666-2.
Sotoodeh, K. (2020). Development of a numeric method to validate the reliability
improvement of safety critical valves during operation through online monitoring
implementation. Journal of Sensing and Imaging https://doi.org/10.1007/s11220-
020-00323-1.
Sotoodeh, K. (2021). Safety and Reliability Improvement of Valves and Actuators for the
Offshore Oil and Gas Industry Through Optimized Design. University of Stavanger.
Doctor Philosophiae (Dr. Philos) UiS no.573. Faculty of Science and Technology.
ISBN: 978-82-7644-987-7.
357
10
Valve Operation
10.1 Introduction
A valve must have a means of operation in order to function properly. It may be

possible to divide all valves operating in the same way, except for pressure relief
and check valves, into two groups; those which are controlled by hand and those
which are controlled with actuators. There are valves that are manually controlled
or manual valves that are operated by either a lever or a handwheel and a gearbox.
As seen in Figure 10.1, a manually operated valve with a handwheel and a
gearbox that is being driven by an individual (a human) is shown.
The calculations in this section are focused primarily on the torque and forces
generated by the manual operator or actuator to operate the valve. Additionally,
the generated loads from the operator or actuator are transferred through the valve
stem to the internals of the valve in order to operate (open or close) the valve.
Therefore, the valve internals, especially the stem, must be designed robustly in
order to withstand the forces applied by the operator or actuator. There are also
several valves which are used in cryogenic or extremely cold service environments
like liquefied natural gas (LNG), and nitrogen. In fact, some of these valves are
capable of operating at –196 C. It is no surprise that LNG has since become a very
popular source of energy as a clean and environmentally friendly fuel. So many
investments have been made in LNG plants, and more investments will be made
in the future. Cryogenic valves are designed with an extended bonnet and stem.
The length of the extension should be sufficient to keep the stem packing from
coming into contact with the cryogenic material passing through the valve. When
the valve packing or stem sealing is exposed to a freezing temperature, there is a
high probability that the packing will freeze and leak. By increasing the length of
the stem, parameter L, the risk of stem buckling and misalignment for gate and
globe valves will increase. Buckling in engineering is the deflection or deformation
of a component such as a bar or a pipe under the pressure of a compression load.
358 10 Valve Operation
Gearbox
Handwheel
Valve
Figure 10.1 An operator who operates a manual valve by using a handwheel and gearbox.
Source: Pichitstocker/Adobe Stock.
With a rotating motion, the stems of ball valves and butterfly valves are subject to
torsional deflection. In this chapter, we explain the necessary calculations to
design the valve stems to be able to withstand the loads from the valve operator
or actuator while preventing buckling and torsional deflections.
10.2 Valve Torque
In the technical world, torque is defined as the amount of force that can be exerted
on a valve operator to rotate the valve closure member to be opened or closed. One
Newton-meter of torque has the same meaning as one Newton force applied per-
pendicularly to a 1-m-long moment arm, which is the result of one Newton force
applied perpendicularly to the end of the moment arm. In accordance with the
definition provided, the torque on the valve lever or handwheel is calculated
according to Eqs. (10.1) and (10.2), respectively.
Calculation of the Torque for the Lever-Operated Valves

Torque = Length of the lever × Operator single hand force 10 1
Example 10.1 A single-hand force is required to operate a lever-operated valve

by a human operator, which is 100 N. In the case of a 70 Nm torque produced by
the operator, what is the length of the lever?
Answer
Torque = Length of the lever × Operator single hand force 70
= Length of lever × 100 Length of lever = 0 7 m = 70 cm
Calculation of the Torque for the Gear-Operated Valves

Torque = Diameter of the handwheel × Operator single hand force
10 2
Example 10.2 Figure 10.2 depicts the hands of a human operator using a hand-
wheel to operate a manual valve. The handwheel is 300 mm in diameter, as illus-
trated in the figure. The force produced by the operator’s hand when holding the
valve is 200 N. Calculate the required torque for the valve to be operated.
Answer

= 0 3 × 200 = 60 Nm
In physics, torque can be calculated using Eq. (10.3) as follows:
Calculation of Torque in Physics

τ=F×r 10 3
where:
τ = Magnitude of torque (Nm);

F = Force applied on the bare stem valve (N);
r = Distance of the force from the rotational axis (m).
Figure 10.2 An operator who operates 150 mm 150 mm

a manual valve equipped with a
handwheel.
F
40 teeth
20 teeth
A B
Figure 10.3 Gears in a gearbox for increasing input force.
In order to enhance the force and torque produced by the hand of the operator,
the handwheel-operated valves can be fitted with a gearbox. Typically, gearboxes
are used in conjunction with handwheels in order to facilitate valve operation. It
can be seen as a simple and cheap method of valve operation in which gearing is
used to increase the force and efficiency produced by the operator by moving the
handwheel. There are gears in the gearbox, which are wheels with teeth that slot
together. Assume that there is a gearbox made up of two gears, as illustrated in
Figure 10.3; the first one on the ride, the “driving” gear, is smaller and has 20 teeth.
With 40 teeth, the second gear is larger, referred to as the “driven” gear. Considering
the number of teeth on the driving gear in this case, the gearbox, in this case,
increases the input force applied by the operator through the handwheel. Depending
on the gear ratio, which is calculated by Eq. (10.4), the amount of increase will vary.
Calculation of Gear Ratio

Number of teeth on driven gear
Gear ratio = 10 4
Number of teeth on driver gear
As a result, in this case, the gear ratio is equal to two. In the example given ear-
lier, if the operator places 300 Newtons (N) of force on the handwheel with two
hands, then the gearbox in the example will increase the force to 600 Newtons (N).
Example 10.3 Assuming that a valve with a gear ratio of three is used in the pre-
vious example, how much torque is produced by the operator’s one hand to oper-
ate the valve using the gear ratio of three?
Answer
The hand of the operator will have the ability to increase the produced torque
and force three times by using a gearbox with a gear ratio of three. This is equiv-
alent to an increase in torque from 60 to 180 Nm when using a gearbox with a ratio
of three.
The force applied to the valve should overcome the unbalanced forces resulting
from the pressure drop; impart sufficient force to the seat to prevent leakage;
and overcome the packing friction force on a bare stem valve using Eq. (10.5)
as follows:
Calculation of the Total Force Acting on the Valve Stem
F total = F pressure drop + F seat + F packing 10 5
As a result of the pressure drop across the valve, an unbalanced force is gener-
ated, which is calculated as follows using Eq. (10.6):
Calculation of the Required Force for the Valve Operation Due to

the Pressure Drop
F pressure drop N = Seat area m2 × Pressure drop N m2 10 6
The seat of a ball valve is placed around the ball and its area is determined
by Eq. (10.7).
Calculation of the Ball Valve Seat Area
π
Seatarea = Seat external bore2 − Ball or valve internal bore2 10 7
4
Example 10.4 In an oil service, a 10 ball valve with a pressure class of 300
(50 bar pressure nominal, refer to ASME B16.34 standard) is used. There is a
258 mm external bore on the seat, and a 252 mm minimum valve bore. There is
a 0.5 diameter stem on the valve, which is double reinforced with packing. There
are various packing friction values with a stem which are based on the ASME pres-
sure class and stem diameter which are listed in Table 10.1. Find the force required
for this valve to operate. What would be the required torque to operate the valve if
its operation is controlled by a handwheel with a diameter of 0.4 m?
Table 10.1 Packing friction forces.
Packing type
Stem ASME Single Double Reinforced

diameter (mm) pressure class PTFE PTFE PTFE
12.5 150 170 250 284

300 428
600 563
900 720
1500 855
Answer
In the case of a ball valve, the seating force is provided by the fluid and the spring
force behind the seat, so Fseat = 0 in this case. As a next step, the area of the ball
valve seat needs to be calculated according to Eq. (10.7) as follows:
π
Seatarea = Seat external bore2 − Ball or valve internal bore2
4
π π
= 2582 − 2522 = × 3060 = 2403 25 mm2 = 0 0024032 m2
4 4
The maximum pressure drop across a valve can be assumed equal to the max-
imum class pressure of the valve, which can be equated to the nominal pressure of
50 bar in this case. Each bar is equal to 100,000 (N/m2). Thus, the pressure drop
across the valve is 5,000,000 (N/m2). Now it is possible to calculate the force of
the pressure drop using Eq. (10.6) as follows:
F pressure drop N = Seat area m2 × Pressure drop N m2

= 0 0024032 × 5,000,000 = 12,016 N
The next step is to obtain the required force to overcome the packing friction
from Table 10.1. Fpacking = 428 and now it is possible to calculate the total force
on the stem valve using Eq. (10.5) as follows:
F total = 12,016 + 428 = 12,444 N
Given that the valve is operated manually by a handwheel, it is possible to cal-
culate the torque required for operation using Eq. (10.2) as follows:

= 0 4 × 12,444 = 4977 6 Nm
10.3 Stem Design
10.3.1 MAST Calculations

A valve manufacturer/supplier defines MAST as the maximum torque/thrust that
can be applied to a valve train without causing damage. Engineering units for
MAST are Nm or lb-in. The definition of valve drive train according to API 6D
is “All parts of a valve drive between the operator and obturator (closure member,
such as a disc or a ball), including the obturator but excluding the operator.” In
order to ensure torque delivered by the actuator is always less than the valve MAST
during the engineering stage, it is the valve manufacturer’s responsibility to calcu-
late and provide valve MAST during the engineering stage. Previously, it was
explained that the torque applied to the valve train comes from the valve operator,
which can either be manual (e.g. a hand wheel with a gearbox) or actuated. The
drive train in a ball valve includes a solid round stem, a stem key, and a ball, as well
as a ball and stem joint. In order to make sure that the actuator’s maximum output
torque/thrust does not exceed the valve MAST at any point of travel, it is crucial to
keep in mind that the actuator’s maximum output torque/thrust is based upon the
maximum supply voltage or operating pressure. In the case of manual valves, there
is generally not enough torque generated by the gearbox to exceed MAST for the
valve itself. In addition, the other reason for the high torque values for actuators is
to have a safety factor (e.g. 2), calculated using Eq. (10.8), when designing and siz-
ing the actuators.
Calculation of the Actuator Safety Factor

Actuator torque value
Safety factor = 10 8
Valve torque value
Example 10.5 A 30-in. Class 1500-actuated top entry ball valve is installed on an
oil export pipeline constructed of low-temperature carbon steel grade A352 LCC.
The top entry design reduces the possibility of leakage from the valve because it is
welded to the pipeline from both sides. This valve also includes an emergency
shutoff function with a hydraulic actuator for quick action. High-pressure oil sup-
plied to the actuator (e.g. 180 barg) along with high torque and a large spring in the
actuator enable fast opening and closing of the valve. The safety factor of the actu-
ator is two, and the torque values required from the valve manufacturer to open
and close the valve are shown in Table 10.2. The meaning of each torque abbre-
viation is provided later. What is the minimum torque that the actuator should be
able to generate?
Table 10.2 30 class 1500 ball valve torque values.
Valve to open torques (Nm) Valve to close torques (Nm)
30 class 1500 top entry ball valve, delta pressure = 250 barg
BTO Running ETO BTC Running ETC
110,016 16,215 36,852 110,016 16,215 36,852
There are six torque values associated with valves including ball valves:
1) Break to open (BTO): The torque is measured when the valve is closed and the
ball is in contact with just one seat under pressure. The torque, also known as
breakaway torque, is the strongest torque in this case.
2) Running torque (RT): It refers to the torque of the valve when the ball opens
at approximately 35 –45 .
3) End to open (ETO): The torque is applied when the ball is in an 80 position
and closed to fully open the valve.
4) Break to close (BTC): In the fully open position, the torque is required to
break the open position of the valve and close it.
5) End to close (ETC): When the valve is about to close, the torque is required to
fully close the valve.
Answer
Considering the safety factor of two, the selected actuator must provide at least
twice the torque of the valve at each position. The following formula determines
the minimum actuator torque at different positions:
Minimum actuator BTO torque = 2 × Valve BTO = 2 × 110,016 = 220,032 Nm
Minimum actuator running torque = 2 × Valve running = 2 × 16,215 = 32,430 Nm
Minimum actuator ETO torque = 2 × Valve ETO = 2 × 36,852 = 73,704 Nm
Minimum actuator BTC torque = 2 × Valve BTC = 2 × 110,016 = 220,032 Nm
Minimum actuator ETC torque = 2 × Valve ETC = 2 × 36,852 = 73,704 Nm
As such, the minimum torque an actuator must produce is 32,430 Nm, which is
the running torque.
As illustrated in Figure 10.4, which includes four sections outlined later, the
most complex shape of the stem is that of a ball valve.
Section 1
Section 4
D1
Section 2
Section 3
Figure 10.4 The stem of a ball valve consists of four sections.
r
1) The top section contains two keys
2) Circular part in the middle a
3) The rectangular section at the bottom
4) Stem keys M
It is important to note that only one stem key

can be used on each valve stem in some cases as
illustrated in Figure 10.5. In order to calculate
the MAST for the stem section with one or b
two stem keys, we utilize Eq. (10.9). Figure 10.5 The portion of the
stem containing one key.
Calculation of MAST for the Stem Section with One or Two

Stem Keys
0 536 × Y S × r 3
MAST in a stem sectionwith one or two keys = MC1 =
B × 1000
10 9
where:
MAST is Nm;
Y S = Yield strength N mm2
Reduction of diameter at the top of the stem where the keys
are located on the actuator side
r= mm ;
2
B is calculated based on Roark’s equation in accordance with Eq. (10.10) as
follows:
r Calculation of Parameter B Based

on Roark’s Equation for the Stem
Section with a Single Key
M a
2 3
b b b
B = K1 + K2 + K3 + K4
r r r
10 10
b
where for 0.5 ≤ (a/b) ≤ 1.5
Figure 10.6 A section of the
stem that contains two keys.
a a 2
K 1 = 1 1690 − 0 3168 + 0 0490
b b
a a 2
K 2 = 0 43490 − 1 5096 + 0 8687
b b
a a 2
K 3 = − 1 1830 + 4 2764 − 1 7024
b b
a a 2
K 4 = + 0 8812 − 0 2627 − 0 1897
b b
In the case of two stem keys on the top of the stem (see Figure 10.6), the MAST
calculations are made in accordance with Eq. (10.9), but parameter B is calculated
differently in accordance with Eq. (10.11).
Calculation of Parameter B Using Roark’s Equation for

the Stem Section with Two Keys
2 3
b b b
B = K1 + K2 + K3 + K4 10 11
r r r
where for 0.5 ≤ (a/b) ≤ 1
a a 2
K 1 = 1 2512 − 0 5406 + 0 0387
b b
a a 2
K 2 = − 0 9385 + 2 3450 + 0 3256
b b
a a 2
K 3 = 7 2650 − 15 338 + 3 1138
b b
a a 2
K 4 = − 11 152 + 33 710 − 10 007
b b
According to Eq. (10.12), the MAST in the circular section is calculated as

follows:
Calculation of MAST for the Rounded Section of the Stem
0 536 × Y S × π × d3
MAST in the rounded area of the stem = MC2 =
16,000
10 12
where:
MAST is Nm;
YS = Yield strength (N/mm2);
d is the diameter of the stem at the rounded end.
Example 10.6 The stem of a 16 ball valve is made of a super duplex material that
has a yield strength of 80 Ksi. If the maximum torque of the valve is 2711 Nm, what is
the diameter of the stem at the rounded end?
Answer
The yield strength of the stem in super duplex material is 80 Ksi, which is equal to
5.516 × 108 N/m2
The stem of the valve is calculated as follows using Eq. (10.12):
0 536 × Y S × π × d3
MAST in the rounded area of the stem = 2711
16,000
0 536 × 5 516 × 108 × π × d3

= 0 43 × 108
16,000
= 9 29 × 108 × d3 d3 = 0 0463 m3 d
= 0 359 m = 359 mm
Thus, the stem diameter in this case to withstand

2b
2711 N.m is equal to 359 mm.
2a
Equation (10.13) can be used to calculate the
MAST at the rectangular section of the stem illus- Figure 10.7 Section of the
trated in Figure 10.7. stem that is rectangular.
b Figure 10.8 Part of the stem with keys.
D a
Calculation of the Rectangular Section of the Stem Using MAST
MAST in the rectangular sectionof the stem = MC3

3 6b 10 13
0 536 × Y S × 2a × 4b2 3+
2a
= Nm
1000
The next part of the stem for MAST is the part where the stem keys are installed,
as illustrated in Figure 10.8. In accordance with Eq. (10.14), the calculation of
MAST at the stem is performed with stem key(s) such as the one illustrated in
the figure here.
Calculation of MAST at the Stem with Stem Key(s)
0 6 × Sm × D × L × a 0 6 × 0 67 × Y S × D × L × a
MC4 = ×n= ×n
2000 2000
0 402 × Y S × D × L × a
= ×n
2000
10 14
where:
a: width of key (mm);

L: length of key (mm);
n: number of keys;
D: stem diameter (mm);
b: key groove (mm).
The final valve MAST is calculated using Eq. (10.15) as follows:
Calculation of the Final MAST

Valve final MAST = Min MC1, MC2, MC3, MC4 10 15
Torque is typically greater as a function of force for valves with a larger size and a
higher-pressure rating. Larger valves with a higher-pressure class usually possess
bigger internals. So a higher force and consequently more torque is needed to oper-
ate the internals of the valve. By increasing the diameter of the valve stem or using
a stronger material, it is possible to increase the torque that the valve stem can
withstand.
10.3.2 Buckling Prevention

A cryogenic valve has an extended stem, meaning that its length is increased. The
length of a stem, parameter L, increases the risk of stem buckling and misalign-
ment. When applied to engineering, buckling refers to the deflection or deforma-
tion of a component, such as a bar or a pipe, that is subjected to a compression load.
In Figure 10.9, a column is illustrated buckling under an axial compressive load (P).
The same principle applies to the valve stem. The ratio of stem length to stem
diameter, parameter L/D, has a direct impact on buckling: the higher the L/D ratio,
the higher the chance of buckling. Thus, the stem extension of cryogenic valves
increases the risk of buckling. So some measures should be taken as per the relevant
cryogenic valve standards.
Figure 10.9 Buckling of a P P

column under an axial
compression load P.
B B
A A
There is a particular emphasis on stem design to prevent buckling for gate and
globe valves with a linear stem motion in the MSS SP134 standard. In ball and but-
terfly valves, a rotary motion exposes the stem to torsional deflection, as explained
further in this chapter. The MSS SP134 standard recommends that the ratio of stem
length to diameter be limited for linear stem motion valves in order to prevent
buckling. MSS SP 134 presents two equations for buckling prevention, depending
on the disk/gate design: the first type of disk/gate is called a stem-guided disk/gate,
and the second type is called a body-guided disk/gate. The stem-guided disk/gate
and globe valves are those in which the movement of the disk/gate is guided and
controlled by the stem.
Based on MSS SP134, gate and globe valves, which have a stem-disk connection
through the stem nut, should have a limited stem length to diameter ratio when in
the closed position as described by Eq. (10.16).
Ratio of Stem Length to Diameter in a Closed Position for

Stem-Guided Disk Globe and Gate Valves
L π E
≤ 10 16
d 2 2Spl N
where:
L: Unsupported length of the stem between the upper stem guide and the interface
between the stem and the disk (m./in.);
d: Stem diameter (m/in.);
E: Modulus of elasticity of stem material (pound per square inch (psi)/Pa);
Spl: Material proportional limit of the material, which is generally less than the
material yield strength (psi/Pa). It is advisable to use the material yield strength
instead of the material proportional limit in case of a worst-case scenario. On a
stress-strain curve, Figure 10.10 illustrates the proportional limit for a material;
N = 2.
An alternative design, called a body-guided disk/gate, is more robust than the

stem-guided disk/gate. It is an advantage of a body-guided disk design that the
movement of the disk takes place entirely within the body enclosure and is guided
by the valve body.
Example 10.7 According to the API 600 standard for gate valves, a 12-in. CL600
wedge-gate valve has a stem diameter of 2 in. or 50.80 mm. Due to the application
of this valve in cryogenic environments, the bonnet and stem should be extended.
The length of the stem without support is 45 in. or 1150 mm in the closed position,
Stress
Ultimate strength
Yield point Break

Elastic limit
Proportional limit
Strain
Figure 10.10 Proportional limit region on a stress-strain curve.
according to the MSS SP 134 standard. This valve is designed to operate at a min-
imum operating temperature of –129 C or –200 F. The stem of the valve contains
stainless steel grade 304 austenitic stainless steel with a modulus of elasticity of
29.5 × 106 psi at the minimum operating temperature and a modulus of elasticity
of 28 × 106 psi at ambient temperature, which is the maximum operating temper-
ature. There is no change in the yield strength of SS304 at the different tempera-
tures given earlier and it is equal to 30,000 psi. The gate of the valve has a stem
guide. Is it possible that the stem of the valve can withstand a buckling axial force
when the valve is in the closed position?
Answer
Since the gate valve is a stem-guided gate, the stem length-to-diameter conditions
provided in Eq. (10.16) should be met.
L π E
≤
d 2 2Spl N
L 45 in π E
= = 22 5
d 2 in 2 2Spl N
3 1415 29 5 × 106
= = 1 57 245 83
2 2 × 3000 × 2
= 24 61 at minimum operating temperature
Therefore, L d ≤ π 2 E 2Spl N at minimum operating temperature, since

22.5 < 24.61. So there is no risk of stem buckling for the wedge-gate valve at min-
imum operating temperature.
Now the equation should be checked for the maximum (normal) operating con-
dition as follows:
π 2 E 2Spl N = 3 1415 2 28 × 106 2 × 3000 × 2 = 1 57 × 233 33 =
23 98 at a maximum operating temperature equal to room temperature.
Therefore, L d ≤ π 2 E 2Spl N at a minimum operating temperature since
22.5 < 23.98. So there is no risk of stem buckling for the wedge-gate valve at min-
imum operating temperature.
Due to the fact that the ratio of the stem length to a diameter at maximum and
minimum operating temperatures is within the limits prescribed, the stem is not at
risk of buckling during operation.
The modulus of elasticity, or elastic modulus, is a quantity that measures the
resistance of an object to elastic deformation. According to Eq. (10.17), the modulus
of elasticity is calculated as the ratio of stress to strain. In physics, strain is the
amount of deformation or displacement of an object compared to its reference
or initial length (refer to Eq. (10.18)). In addition to the effect of stress and strain
on the modulus of elasticity, temperature also has an influence on the modulus of
elasticity. As a matter of fact, the modulus of elasticity is reduced by increasing the
temperature. In the ASME B31.3 process piping code, the modulus of elasticity
values can be found in a table C-6, which contains the values for different materials
at various temperature ranges. Steels such as 304 and 316, which are austenitic
stainless steels, are commonly used for cryogenic valve parts, including the stem.
According to ASME B31.3 table C-6, the modulus of elasticity of austenitic stain-
less steels is 28.3 × 106 psi at ambient temperature. There is a type of elastic defor-
mation that results from stress, which is a nonpermanent deformation, which
means that once the stress is released, the stressed object returns to its origi-
nal form.
Modulus of Elasticity Calculation

δ
E= 10 17
ε
where:
E: Modulus of elasticity (psi/Pa);

δ: Stress (psi/Pa);
ε: Strain (dimensionless).
Strain Calculation
ΔL
ε= = α × ΔT 10 18
L
where:
ε: Strain (dimensionless);
ΔL: Change in length of pipe (m/in./ft);
L: Initial length of pipe (m/in./ft).
Example 10.8 Under tensile stress, a steel bar has a cross-sectional area of
100 mm2and a length of 50 mm. Calculate the value of the modulus of elasticity
when the ratio of force to change in length for the bar is 410 × 103 N/mm.
Answer
The following parameters are given in this example:
As a result of the combination of Eqs. (10.17) and (10.18), the modulus of elas-
ticity is given as follows:
δ F A F L 50
E= = = × = 410 × 103 × = 205 × 103 N mm2
ε ΔL L ΔL A 100
Based on MSS SP134, gate and globe valves with a body-guided disk/gate con-
nection, in the closed position, should be designed so that the stem length to the
diameter ratio is determined by Eq. (10.19).
Ratio of Stem Length to Diameter in the Closed Position for

Body-Guided Disk/Gate Globe and Gate Valves
L π E
≤ 10 19
d 2 Spl N
where:
L: Unsupported length of the stem between the upper stem guide and the interface
between the stem and the disk (m/in.).
d: Stem diameter (m/in.).
E: Modulus of elasticity of stem material (pound per square inch (psi)/Pa).
Spl: Material proportional limit of the material, which is generally less than the
material yield strength (psi/Pa). It is recommended that the material yield
strength should be used in place of the material proportional limit in the
worst-case scenario. As shown in Figure 10.10, the proportional limits for a

material are illustrated by a stress-strain curve.
N = 2.
In fact, a body-guided disk/gate connection is more robust than a stem-guided

disk/gate connection. So a lower risk of buckling is expected for the stem with a
more robust design.
Example 10.9 The size, pressure class, and minimum and maximum operating
temperatures of the two gate valves are the same. In addition, the stems for these
valves are of the same material and diameter. In this case, there are three assump-
tions: The first is that one valve has a body-guided disk and the other has a stem-
guided disk. The second point is that the ratio of length to diameter should be as
high as possible without buckling. According to the third assumption, the unsup-
ported length of the stem of the body-guided disk is 500 mm. What would be the
stem’s unsupported length for the other design?
Answer
Because the material of the stem is identical in both valves, the values of modulus
of elasticity and material proportional limit will also be identical. Given that both
valve stems have the same diameter, the only difference between the stem of the
first and second valves is their length. The stem length to diameter ratio should be
the maximum without buckling, meaning that:
L π E
= Applicable to the stem-guided disk
d 2 2Spl N
500 π E
= Applicable to the body-guided disk
d 2 Spl N
The following result is obtained by dividing both sides of the above equations:
L 1
= L = 353 56 mm
500 2
Consequently, the maximum unsupported length of the stem for stem-guided
disks is 353.56 mm.
10.3.3 Torsional Deflection Prevention

Since quarter-turn valves do not experience axial (linear) loads, their stems are not
at risk of buckling. This is because the stems of quarter-turn valves, i.e. ball,
Torque
Diameter D
Angle θ
Figure 10.11 Twisting a piece of the valve stem.
butterfly, and plug types are not at risk of buckling. The length of the stem to the
diameter of the valve must be as close to one as possible in order to prevent exces-
sive torsional deflection of the stem. The term torsional deflection refers to the
angular displacement that occurs when a twisting force is applied to a component.
The torsion of a component is defined as the twisting of its one end, in this case, the
stem of a quarter-turn valve, as shown in Figure 10.11. Torque is a measure of the
force used to twist an object, in this case, the stem of a valve. A torque is applied to
one end of the stem, which is called the rotating end. The angle of twist, parameter
θ, of the shaft or stem of a valve that is under a torsional load is defined as the angle
formed between the fixed and rotating ends as a result of torque.
According to an MSS SP 134 standard requirement, the stem length and diam-
eter combination of a quarter-turn valve should limit the stem torsional deflection
or angle of twist to π/90 or 0.0349 radian or 2 . The angle of the twist is determined
by Eq. (10.20).
Angle of Twist
TL π
θ= ≤ 10 20
GJ 90
where:
θ: Angle of twist (radian);

T: Maximum stem design torque (Nm);
L: Length of the stem from the point where torque is applied to the connection to
the closure member (m);
Table 10.3 The values of the modulus of rigidity or modulus

of shear for some important materials.
Modulus of rigidity or shear modulus

Material (parameter G) in GPa (109 Pa)
Carbon steel 77
Inconel 625 79
Monel 66
Stainless steel 77
Titanium 41
G: Modulus of rigidity or shear modulus (Pa or N/m2). In general terms, the mod-
ulus of rigidity is defined as the ratio of shear stress to the displacement of the
material or shear strain. Shear stress is the result of opposing forces applied by
two separate objects to a cross-section that is on the same plane as the cross-
section of the material. Shear forces tend to cause material deformation through
slippage along the plane parallel to the imposed stresses. The values of the mod-
ulus of rigidity for some essential materials are provided in Table 10.3;
J: Polar moment of inertia of stem indicating the torsion deformation resistance of
the shaft m4.
The polar moment of inertia for a circular object such as a valve stem or shaft
depends on the stem diameter and is calculated using Eq. (10.21):
Polar Moment of Inertia for a Valve Stem in a Circular Shape
πd4
J= 10 21
64
where:
J: Polar moment of inertia of stem indicating the torsion deformation resistance of

the shaft (m4);
d: Valve stem diameter (m).
10.3.4 MAST Limitation for Quarter-Turn Cryogenic Valves

In accordance with MSS SP 134, the MAST for a quarter-turn extended stem valve
in cryogenic service shall not exceed the limit established by Eq. (10.22).
Limit of MAST in Quarter-Turn Valves
πd3 τpl
MAST < 10 22
12n
where:
MAST: Maximum allowable stem torque (Nm).

d: Stem diameter (m).
τpl: The shear stress proportional limit for the stem material. It should be noted
that this value is very close to the shear strength point of the material. Standard
values for shear strength have not been published, but can be estimated to be
60% of the tensile strength. The unit of proportional shear stress is Pa or N/m2.
n: Safety factor of 2.
Example 10.10 In a cryogenic application with a low temperature of –196 C, a

20-in. CL300 ball valve requiring a maximum torque value of 6500 Nm is used.
This valve can handle a maximum stem torque of 20,500 Nm. The valve stem is
made of stainless steel 316 with a tensile strength of 480 MPa. To prevent stem fail-
ure caused by the loads applied by the valve operator, what is the minimum stem
diameter for this valve?
Answer
There is a maximum operating torque for the valve of 6500 Nm, which is less than
the MAST value, which is 20,500 Nm in this case. This is a crucial point to keep in
mind because if the maximum operating torque is higher than the MAST, then the
stem of the valve has a high risk of being damaged or failing during operation.
According to Eq. (10.22), the minimum stem diameter leads to the minimum
value of πd3τpl/12n on the right-hand side. πd3τpl/12n must be at least slightly over
the MAST to be considered a minimum value. In this case, we can assume that
πd3τpl/12n is only one unit above the MAST. Therefore,
πd3 τpl
= 20,501 Nm, and
12n
τpl = Shear strength = 0 6 × Tensile strength
3 1415 × d3 × 288 × 106

= 0 6 × 480 MPa = 288 × 106 Pa or N m2
12 × 2
= 20,501 d3 = 543 7 × 10 − 6 d
= 8 16 × 10 − 2 m = 0 0816 m = 81 6 mm
10.1 The maximum torque required for the operation of a 4-in. cryogenic ball
valve in pressure class 300 is 500 Nm. The handwheel’s diameter should
not exceed 700 mm. In accordance with the design standard of the valve,
the maximum operator hand force is 360 N. Can an operator apply 360 N
of force to provide the required torque? If not, what would be required
to generate the torque necessary for valve operation?
A Yes. It is possible for the operator to apply the required torque in order
to operate the valve.
B No. The operator is only able to produce 300 Nm torque in this case,
which is insufficient.
C In this instance, a gearbox with a gear ratio of two is required.
D All of the answers are incorrect.
Answer
Maximum torque is produced when the diameter of the handwheel and the
force applied by the operator are both at their maximum. The largest diam-
eter of the handwheel is 700 mm, and the maximum torque produced by
rotating the handwheel is calculated as follows:
Torque = 360 N × 0 7 m = 252 Nm
Due to the maximum amount of torque produced by the operator,

option A is incorrect; option B states that the operator is only able to pro-
duce 300 Nm, whereas the actual torque produced is 252 Nm. As a result,
option B is wrong as well. Option C is correct, as using a gearbox with a gear
ratio of two could result in a torque of 504 Nm, which is more than the
required torque. Hence, option D is incorrect since option C is correct.
10.2 For the stem of a ball valve with a rounded geometry, there are two mate-
rial options: the first is super duplex which has a yield strength of 80 Ksi,
and the second is Inconel 718 which has a yield strength of 150 Ksi.
The required torque for the operation of the valve is 5000 Nm. In what
number of millimeters is the diameter of the stem larger when a stem con-
structed of super duplex material is selected instead of an Inconel 718
material?
A 83.5 mm
B 85 mm
C 90 mm
D 95 mm
Answer
YS for a duplex material = 80 Ksi = 5.516 × 108 N/m2
YS for an Inconel 718 material = 150 Ksi = 1.034 × 109 N/m2
According to Eq. (10.12), the MAST of a rounded stem is calculated as
follows:
MAST in the rounded area of the stem = (0.536 × YS × π × d3)/16, 000
The super duplex stem material has been chosen 5000 = (0.536 ×
5.516 × 108 × π × d3)/16, 000 d3 = 0.086 d = 0.4417 m = 441.7 mm
The Inconel 718 material has been chosen 5000 = (0.536 × 1.034 ×
109 × π × d3)/16,000 d3 = 0.046 d = 0.3582 m = 358.2 mm
The diameter difference between a super duplex and Inconel 718
stems = 441.7 − 358.2 = 83.5 mm. Thus, option A is the correct answer.
10.3 The stem of a large size ball valve that is installed on a large oil export pipe-
line is composed of four segments; two keys, a circular and rectangular sec-
tion, and the area where the stem keys are installed. For these four sections,
the calculated MAST values are 270,000 Nm, 1,450,000 Nm, 190,000 Nm,
and 930,000 Nm, respectively. What is the final MAST value for the stem
as a whole assembly of these sections?
A 270,000 Nm
B 1,450,000 Nm
C 190,000 Nm
D 930,000 Nm
Answer
According to Eq. (10.15), the final MAST must be selected from the lowest
torque value given earlier, which is equal to 190,000 Nm. Therefore, option
C is the correct response.
10.4 The below table lists the torque values for a valve and its actuator. What
would be the correct number for the actuator’s safety factor?
BTC (N) BTO (N) ETO (N) ETC (N)
8 slab gate valve 8,145 27,450 8,322 23,241

Hydraulic fail-safe closed actuator 70,737 62,553 45,735 55,452
Safety factor 8.68 2.28 5.50 2.39
A 1
B 2
C 3
D 4
Answer
The most appropriate answer is option B with a safety factor of two. The
choice of option A is incorrect since the safety factor of one implies that
the actuator torque values should be close to the valve torque values, which
is not common in the industry, including in this case. Neither option C nor
D is correct, as safety factors 3 and 4 are not met for torque values of BTO
and ETC.
10.5 The stem of a valve is made of a steel round bar that has been put under a
stress of 280 MPa. The stem has a diameter of 80 mm and a length of
240 mm. A modulus of elasticity of 205 GPa is calculated for the bar. Which
of the following statements is true?
A The strain value is 0.0010.
B The strain value is 0.002.
C A force of 1.40 MN is applied to the stem.
Answer
E: Modulus of elasticity = 205 GPa = 205 × 109
Using Eq. (10.17) E = (δ/ε) ε = (δ/E) = ((280 × 106)/(205 ×
109)) = 0.00137
As a result, both options A and B are incorrect. Therefore, the next step
would be to calculate the applied force on the valve stem. By multiplying
the stress by the cross-sectional area of the stem, it is easy to estimate
the force applied to the stem.
The cross-sectional area of the stem (A) = π(d2/4)(d = Diameter of the
stem) = π(0.0802/4) = 0.00503 m2
F = δ × A = 280 × 106 × 0 00503 = 1 4084 × 106
The calculated force is 1.40 MN. So option C is the correct choice. Option
D is incorrect since option C is the correct choice.
10.6 A quarter-turn valve requires 200 N of force to overcome pressure drop, 160 N
of force to overcome seating, and 140 N of force to overcome packing and
stem friction. Suppose that the valve is to be manually operated by a gearbox
and handwheel, that the size of the handwheel is 0.4 m, and that the torque
to be produced by the handwheel operator is equal to 100 Nm, which of the
following statements is correct?
A The force produced by the operator is sufficient to move the internal
mechanisms of the valve without the aid of a gearbox.
B It is necessary to apply a force of 500 N to the valve stem in order to

operate it.
C A force of 350 N is applied by the operator using the handwheel.
D A gearbox is required, and its gear ratio should be two.
Answer
In this situation, the force generated by the operator is equal to the ratio of
torque (100 Nm) to handwheel diameter (0.4 m) or 400 N. Therefore, option
C is incorrect because it states that the force produced by the operator is 350 N.
In order to determine the force required for valve operation, which is applied
to the stem, we must add together the forces required to overcome friction,
including a pressure drop (200 N), seating (160 N), and packing (140 N),
which is 500 N. Therefore, option B is the correct choice. The first option is
incorrect due to the fact that the force produced by the operator is 400 N,
but a force of 500 N is necessary to operate the valve. Thus, in order to increase
the force applied by the operator from 400 to 500 N, a gearbox is required.
By this, we mean that the gear ratio should be less than two. As a result,
option D is incorrect. The correct choice is, therefore, option B.
10.7 Stainless steel 304 is the material used for the construction of a body-guided
wedge gate valve stem, which is 10 mm in diameter. The modulus of elas-
ticity of the stem in this case should be considered to be 28 × 106 psi during
the operation of the valve. It is noted that the yield strength of SS304 at the
different operating temperature ranges of valves remains the same and is
equal to 30,000 psi. How long can the stem of the valve be before it becomes
bent or buckled?
A 0.16 m
B 0.22 m
C 0.34 m
D 0.47 m
Answer
According to MSS SP134, gate and globe valves with a body-guided disk/
gate connection should have a stem length to diameter ratio in the closed
position determined by Eq. (10.19) as follows:
L π E
≤
d 2 Spl N
where:
d = 100 mm
E = 28 × 106 psi
Spl = Proportional limit = Yield strength = 30,000 psi

N = 2 (Safety factor)
The length of the stem is maximum when both sides of the equation
become identical.
π E 3 14 28 × 106
= × = 1 57 × 21 60 = 33 92
2 Spl N 2 30,000 × 2
L
= 33 92 L = 33 92 × 10 = 339 2 mm = 0 34 m
d
Thus, option C is the correct answer.
10.8 A 20-in. CL300 ball valve operates at a maximum torque of 4500 Nm in a

cryogenic application at a low temperature of –196 C. This valve is designed
to handle a maximum stem torque of 18,500 Nm. The stem of the valve is
made from stainless steel 316 with a tensile strength of 480 MPa. Considering
that the diameter of the stem is 79 mm in this case, what would be the max-
imum length of the stem in order to prevent the stem from failing due to tor-
sional deflection?
A 350 mm
B 276 mm
C 250 mm
D 220 mm
Answer
According to Eq. (10.20), increasing the stem’s length increases the stem’s
angle of twist, and the maximum allowable angle of twist is 2 , π/90, or
0.0349 radian. This means that the maximum stem length should result
in a maximum of 0.0349 radian angle of twist. For the purpose of calculat-
ing the maximum stem length using Eq. (10.20), it should be noted that
parameter T, the maximum design stem torque value, is equal to 18,500
Nm. In addition to the modulus of rigidity, parameter G, is equal to
77 × 109 Pa or N/m2 for stainless steel, as per Table 10.3. The polar moment
of inertia, parameter J, is equal to πd4/64(d = 0.079 m) or 1.9 × 10−6 m4.
TL TL 18,500 × L
θ= 0 0349 = 0 0349 =
GJ GJ 77 × 109 × 1 9 × 10 − 6
L = 0 276 m = 276 mm
As a result, the maximum length of a stem that can prevent twisting by

more than 2 degrees is 276 mm. So option B is correct.
10.9 There are two globe valves that have the same dimensions, pressure class, and
minimum and maximum operating temperatures. The stems are made of the
same material and have the same diameter. For this case, we have to make
three assumptions: the first assumption is that the ratio of length to

diameter should be the maximum value without the risk of buckling; sec-
ond, one valve has a disk that is guided by the body, while the other has a
disk that is guided by the stem; and as a third assumption, it is assumed that
the unsupported length of the stem of the stem-guided disk is 1000 mm. If
the stem of the other design were unsupported, what would be its length?
A 1414 mm
B 1560 mm
C 1710 mm
D 1800 mm
Answer
There is no difference in the parameters of Eqs. (10.16) and (10.19)
between the two valves, with the exception of the length of the stem.
There should be a maximum ratio of stem length to diameter in order
to avoid buckling, meaning that:
1000 π E
= This applies to the stem-guided disk
d 2 2Spl N
L π E
= This applies to the body-guided disk
d 2 Spl N
By dividing both sides of the equations earlier 1000 L = 1 2

L = 1414 mm
Therefore, option A is the correct answer.
10.10 A cryogenic ball valve has a stem diameter of 200 mm and is constructed
from stainless steel 316 material. What is the maximum amount of MAST
that the valve can have without risk of stem failure?
A 106 Nm
B 0.2 × 106 Nm
C 0.3 × 106 Nm
D 0.4 × 106 Nm
Answer
In accordance with MSS SP 134, the measure of the MAST for a quarter-
turn extended stem valve in cryogenic service should not exceed the mea-
surements of Eq. (10.22).
πd3 τpl
MAST <
12n
The stem of the valve is made of stainless steel 316 and has a tensile
strength of 480 MPa. The stem material shear stress proportional limit
is parameter τpl. This value is very close to the shear strength point of a
material. This value can be estimated as follows:
τpl = Shear strength = 0 6 × Tensile strength = 0 6 × 480 Mpa

= 288 × 106 Pa or N m2
The value of πd3τpl/12n on the right side of the equation is calculated as
follows:
πd3 τpl π 0 2 3 × 288 × 106
= = 0 3 × 106
12n 12 × 2
We can say that MAST should be almost as high as 0.3 × 106. Thus,
option C is the correct answer.
Further Reading
American Petroleum Institute (API) (2012). Standard for Actuator Sizing and Mounting
Kits for Pipeline Valves. API 6DX, 1e. Washington, DC: American Petroleum
Institute (API).
American Petroleum Institute (API) (2014). Specification for Pipeline and Piping Valves.
API 6D, 24e. Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) (2004). Valves – Flanged, Threaded,
and Welding End. New York: American Society of Mechanical Engineers (ASME).
Gokilakrishnan, G., Divya, S., Rajesh, R., and Selvakumar, V. (2014). Operating torque
in ball valves: a review. International Journal for Technological Research in
Engineering 2 (4): 2347–4718.
International Organization for Organization (ISO) (2007). Pipeline Transportation
Systems: Pipeline Valves. ISO 14313, 2e. Geneva: International Organization for
Organization (ISO).
International Organization for Organization (ISO) (2011). Petroleum and Natural Gas
Industries: Mechanical Integrity and Sizing of Actuators and Mounting Kits for
Pipeline Valves. ISO 12490, 1e. Geneva: International Organization for
Organization (ISO).
Oxford: Elsevier.
Sotoodeh K. (2019). Actuator sizing and selection. Springer Nature Applied Science, 1,
1207. https://doi.org/10.1007/s42452-019-1248-z
Sotoodeh K. (2019). The importance of maximum allowable stem torque in valves.
Springer Nature Applied Science, 1, 433. https://doi.org/10.1007/s42452-019-0445-0
Sotoodeh, K. (2022). Cryogenic Valves for Liquified Natural Gas Plants, 1e. Austin:
Young, C. and Budynas, R. (2002). Roark’s Formulas for Stress and Strain, 7e.
New York: McGraw-Hill.
385
11
Miscellaneous
11.1 Introduction
Equations and calculations related to joint efficiency, as well as valve sealing, are
discussed in this chapter. Joint efficiency is important because it refers to the
valve’s ability to provide a secure seal between metallic components, such as
the body and bonnet, to prevent leakage. Welded bodies and bonnets provide
the tightest connections with the highest joint efficiency. However, the majority
of valves are body pieces or body and bonnets that are bolted together. The second
part of this chapter deals with the design formulas and calculations that are asso-
ciated with valve stem sealing. The seals on valves, especially the stem seals, play
an imperative role because they reduce the possibility of leaks and fugitive emis-
sions from valves. In addition, they increase the life of the valves, improving per-
formance, as well as reducing the amount of maintenance required. In view of this,
it is essential to dedicate a separate chapter to valve seals. This is because a key
aspect of valve seals is their ability to minimize the risk of emissions of gases
and other environmental pollutants through the seal. It is very critical to select,
design, and arrange valve seals in the most efficient way possible to prevent leak-
age and increase the life and performance of the valve. There are several parts of a
valve that require sealing, such as the stem, the body and the bonnet, and the body
of the valve pieces. Valve stem seals are known as the most important leakage
points when it comes to fugitive emissions for two reasons. First, valve stem seals
are classified as dynamic seals due to their movement. The seals are divided into
two categories: static seals and dynamic seals. It should be noted that static seals
are used between two nonmoving parts of the valve, while dynamic seals are used
between a moving or rotating valve part and another part that may be in motion or
not. Seals that are dynamic in nature are more critical since they are at a higher
risk of wearing and tearing due to friction between one or two moving parts. Fur-
thermore, unlike the valve seat or seat insert, leakage from the valve stem seals is
released to the environment and is one of the major sources of fugitive emissions.
386 11 Miscellaneous
Therefore, it is well worth paying more attention to valve stem seals in more detail.
The term fugitive emission refers to the unintentional and undesirable release,
leakage, or discharge of gases or vapors from equipment or facilities containing
pressure, and from components inside industrial plants, such as valves, piping
flanges, pumps, storage tanks, and compressors.
11.2 Joint Efficiency
In this section, we will focus on three joint efficiency conditions; welded

joints, bolted joints, and threaded joints. Next, we will discuss the efficiency of
welded joints.
11.2.1 Weld Joint Efficiency

It is crucial to understand that welded joints are defined by their efficiency, and
therefore, in some literature, the term “welding joint efficiency” is used. American
Petroleum Institute (API) standards and American Society of Mechanical Engi-
neers (ASME) codes include references to joint efficiency. This is a numerical
value expressed as the ratio between the strength of the welded part and the
strength of the base material. This parameter is calculated as follows
using Eq. (11.1).
Sw
E= 11 1
Sb
where:
E: weld joint efficiency;

Sw: weld strength (Ksi);
Sb: base metal strength (Ksi).
According to paragraph 2.1.6 of ASME B16.34, the minimum acceptable welding

joint efficiency is 0.8 or 80%. The ideal condition is when the weld joint strength is
equal to the base material strength, in which case the joint efficiency would be 1.00
or 100%. Joint efficiency is also applicable between two welded components, such
as two pieces of pipe or a pipe and a valve. Joint efficiency is determined by the type
of welding and the type of post welding inspection. Nondestructive testing (NDT)
refers to an inspection technique that is applied to materials, especially after weld-
ing or machining, in order to find possible defects and discontinuities in the mate-
rial or the welding without destroying it or damaging it. On welding, NDTs are
typically applied after a post weld heat treatment (PWHT). The primary purpose
of PWHT is to reduce residual stress during welding. The residual stress refers to
the stress that remains in the material after welding, leading to cracks, brittle frac-
tures, and ultimately failure of the weld. NDT methods include visual inspection,
liquid penetration test, magnetic particle test, ultrasonic test (UT), and radiogra-
phy test (RT). In contrast to the first three methods, ultrasonic tests and RTs are
volumetric NDTs, which are designed to detect defects below the surface. An RT is
a type of NDT that increases the efficiency of the weld joint. It is possible to
increase weld joint efficiency to 100% by applying an RT test. In Table 11.1,
Table 11.1 Longitudinal weld joint quality factor for piping according to
ASME B31.3 code, table 302.3.4.
Weld joint
efficiency
No. Type of joint Type of seam Examination (E )
1 Furnace Straight As required by 0.6

buttweld the specification
or code
2 Electrical Straight or As required by 0.8
resistance weld spiral the specification
or code
Additional spot 0.9
radiography
Additional 100% 1.00
radiography
3 Electrical fusion Straight or As required by 0.8
weld (single butt spiral the specification
weld) or code
Additional spot 0.9
radiography
radiography
4 Electrical fusion Straight or As required by 0.8
weld (double spiral the specification
buttweld) or code
Additional spot 0.9
radiography
radiography
5 Submerged arc API 5L Straight with As required by 0.95
weld, gas metal one or two the specification
arc weld seams spiral
we found that the longitudinal weld joint factors for different welding mechanisms
were summarized based on the data in table 302.3.4 in ASME B31.3. It should be
noted that electrical resistance welding and electrical fusion welding provide a
higher level of weld joint efficiency than furnace buttwelding.
Furthermore, the joint efficiency of the fillet welds could be 0.45, 0.45, or 0.50
depending on whether or not the fillet weld is single or double. Welding in which
two pieces of metal meet at an angle or perpendicularly is known as “fillet
welding.”
11.2.2 Bolted Joint Efficiency

This section discusses two types of bolted joints; the first is the bolted joint of the
bonnet or the cover, and the second is the bolted joint of the body.
11.2.2.1 Bolted Bonnet or Cover Joints

With a bolted body and bonnet valve design, the body and bonnet are like flanges
that are joined with bolts and nuts. A gasket is placed between the body and the
bonnet to prevent leakage. The body and bonnet are connected by bolts, making it
possible to maintain the valve. The bolting of bonnets and covers to valve bodies
shall be threaded according to ASME B1.1 for inch-dimensioned bolting and, as a
minimum, shall satisfy the following bolt cross-sectional area requirements
as Eq. (11.2).
Bolted Bonnet or Cover Joint Requirement According to ASME B16.34

Ag
Pc × ≤ K 1 × Sa ≤ 9000 11 2
Ab
where:
Ab: Total effective bolt tensile stress area that is calculated either from Eq. (11.6)
or (11.7);
Ag: Area enclosed by the seal effective outside periphery of a gasket, O-ring, or
another seal, with the exception that in the case of a ring-joint gasket, the
bounded area is determined by the pitch diameter of the ring-type joint gasket;
K1: It is 65.26/MPa when the allowable bolt stress is expressed in MPa units and
0.45/psi when the allowable stress is expressed in psi unit;
Pc: Pressure rating class designation. In accordance with the mandatory appendix
B of ASME B16.36, the value of the designation pressure rating is the same as the
value of the pressure rating. There are several pressure ratings that are used for
industrial valves, including 150, 300, 600, 900, 1500, and 2500. The pressure
class designation number for a valve with a pressure class of 150 is equal to
150, for example. In the same way, if the valve pressure class is 300, the pressure
class designation number is 300;
Sa: Allowable bolt stress at 38 C or 100 F. If the value is greater than 137.9 MPa or
20,000 psi, use 137.9 MPa (20,000 psi). The allowable bolt stress values shall be
taken from the ASME boiler and pressure vessel code (BPVC), section II, part D.
When you select calculated tensile stress area in a bolt calculation program such
as connector-bolt property manager in a bolt calculation program, the program
calculates the tensile stress area (At), which is the minimum area of the threaded
portion of the bolt according to Eq. (11.3) as follows:
Calculation of the Tensile Stress Area According to the

Bolt Calculation Program
2
d3 + d2 n
At = × 11 3
2 4
where:
P: Pitch, axial distance between two adjacent threads;

n: 1/P, number of bolt threads per inch or millimeter;
d: nominal bolt shank diameter.
The values of d2 and d3 are shown in Figure 11.1, and they are calculated as
follows:
d3 = d − 1 2268 × P 11 4
d2 = d − 0 6495 × P 11 5
h/8
h/2 Nut
h1 60° h3
h
h/2
Screw
h/4
h/6
R
d d2 d 1 P d3 d2 d
Figure 11.1 A schematic showing the contact between bolts and nuts. Source: Courtesy:
Dassault systems.
For bolt dimensions in millimeters, the formula for the tensile stress area would
be Eq. (11.6) and for bolt dimensions in inches, Eq. (11.7).
Calculation of the Bolt Tensile Stress Area Based on the

Bolt Diameter and Thread Count
2
1
At = 0 7854 × d − 0 9382 × The units are mm 11 6
n
2
1
At = 0 7854 × d − 0 9743 × The units are in 11 7
n
In Table 11.2, the nominal bolt diameter, pitch, and number of bolt threads per
millimeter for ISO M large bolts are provided.
Table 11.2 Nominal bolt diameter, pitch, and number of bolt threads per millimeter
for ISO M large bolts.
Nominal bolt Number of bolt threads

diameter d (mm) Pitch P (mm) per millimeter (n)
M 1.60 0.35 2.86

M 2.00 0.40 2.50
M 2.20 0.45 2.22
M 3.00 0.50 2.00
M 4.00 0.70 1.43
M 4.50 0.75 1.33
M 5.00 0.80 1.25
M 6.00 1.00 1.00
M 8.00 1.25 0.80
M 10.00 1.50 0.67
M 12.00 1.75 0.57
M 16.00 2.00 0.50
M 20.00 2.50 0.40
M 24.00 3.00 0.33
M 30.00 3.50 0.29
M 36.00 4.00 0.25
M 42.00 4.50 0.22
M 48.00 5.00 0.20
M 56.00 5.50 0.18
M 64.00 6.00 0.17
Example 11.1 Calculate the bolt tensile stress area for a large ISO M bolt with a
diameter of 2 mm. The area of tensile stress of the bolt is increased by how many
square millimeters if the bolt diameter doubles?
Answer
Table 11.2 indicates a value of 2.5 for the number of threads per millimeter. Given
the diameter of the bolt and the number of threads per millimeter, using Eq. (11.6),
the area of tensile stress of the bolt can be calculated as follows:
2 2
1 1
At = 0 7854 × d − 0 9382 × = 0 7854 × 2 − 0 9382 ×
n 25
= 0 7854 × 2 63 = 2 07 mm2
According to Table 11.2, the number of threads per millimeter becomes 1.43
when the bolt diameter is doubled and increased to 4 mm. It is now possible to
calculate the area of tensile stress for the bolt with a diameter of 4 mm as follows:
2 2
1 1
At = 0 7854 × d − 0 9382 × = 0 7854 × 4 − 0 9382 ×
n 1 43
= 0 7854 × 11 18 = 8 78 mm2
8 78 − 2 07 = 6 71 mm2
As a result of doubling the bolt diameter from 2 to 4 mm in this example,
6.71 mm2 of bolt tensile stress surface is created.
Example 11.2 There are eight bolts that are used to connect the body and bonnet
of a valve, which is made of carbon steel material and has a pressure rating of 1500.
The valve is operating at ambient temperature. As far as bolts and nuts are con-
cerned, the bolts and nuts used are the American Society for Testing and Materials
(ASTM) A193 B7 and A194 Gr.2H, respectively. In terms of the bolt’s major diam-
eter, it is 0.875 in., and there are 14 threads/in. on the bolt. If we take into consid-
eration the gasket diameter between the body and the bonnet to have a sealing
diameter of 4.5 in., what would be the total effective area of tensile stress? Can
the bolts provide adequate force to secure the area between the body and the bon-
net tightly without the possibility of leaks – as required by ASME B16.34?
Answer
In order to calculate the bolt stress area, Eq. (11.7) must be taken into account as
follows:
2 2
1 1
At = 0 7854 × d − 0 9743 × = 0 7854 × 0 875 − 0 9743 ×
n 14
= 0 7854 × 0 649 = 0 51 in 2
Due to the eight bolts connecting the body and bonnet of the valve, the total bolt
stress area can be calculated as follows:
Ab = 8 × 0 51 in 2 = 4 076 in 2
Since the diameter of the seal is equal to 4.5 in., the area bounded by the effective
outside of the gasket can be calculated as follows:
Ag Area enclosed by the seal − effective outside periphery of a gasket,

d2 4 52
O-ring,or other seal = π d Gasket diameter = π = 15 904 in 2
4 4
The valve pressure class is 1500 meaning that Pc = 1500.

The maximum allowable stress for A193 B7 bolts refers to two ASME codes
including ASME BPVC Sec. II and ASME B31.3, the process piping code, is equal
to 20,000 psi at room temperature. It is critical to keep in mind that when the tem-
perature increases, the value of the allowable stress is typically reduced. The value
of K1 is 0.45/psi when the allowable stress is expressed in psi unit. In order to
ensure that the bolting (bolts and nuts) is able to provide enough force to ensure
the body and bonnet joint remains tightly joined together with no chances for
leaks, the following requirements must be met.
Ag
Pc × ≤ K 1 × Sa ≤ 9000
Ab
The left side of the equation is calculated as follows:
Ag 15 904
Pc × = 1500 × = 5852 7
Ab 4 076
The middle section of the equation is calculated as follows:
K 1 × Sa = 0 45 × 20,000 = 9000
Therefore, 5852.7 ≤ 9000 ≤ 9000 is met and the bolts used for the body and bon-
net joint of the valve provide sufficient forces for the tight sealing of the valve.
11.2.2.2 Bolted Body Joints

The bolted body joints shall be threaded in accordance with ASME B1.1 for inch
dimensional bolts and ASME B1.13M for metric bolts and, at a minimum, shall
meet the following bolt cross-sectional area requirements given in Eq. (11.8).
Bolted body joint requirement according to ASME B16.34.
Ag
Pc × ≤ K 2 × Sa ≤ 7000 11 8
Ab
K2: It is 50.76/MPa when the allowable bolt stress is expressed in MPa units and
0.35/psi when the allowable stress is expressed in psi unit.
Example 11.3 A ball valve is constructed from low-temperature carbon steel

material and has a pressure rating of 300. There are eight bolts connecting its
two body pieces. The valve operates at a temperature of 50 C. In terms of bolts
and nuts, the bolts and nuts used are ASTM A320 L7 and A194 Gr.4, respectively.
There are 14 threads/in. on the bolt, and the bolt has a diameter of 0.875 in. Given
that the gasket diameter between the body and the bonnet is 3.5 in., what is the
total area of tensile stress applied to the bolts? Do the bolts provide adequate force
to secure the area between two body pieces tightly without the possibility of leaks,
as required by ASME B16.34?
Answer
The bolt stress area can be calculated using Eq. (11.7) as follows:
2 2
1 1
At = 0 7854 × d − 0 9743 × = 0 7854 × 0 875 − 0 9743 ×
n 14
= 0 7854 × 0 649 = 0 51 in 2
Due to the eight bolts connecting the body and bonnet of the valve, the total bolt
stress area can be calculated as follows:
Ab = 8 × 0 51 in 2 = 4 076 in 2
Given that the seal has a diameter of 3.5 in., the area bounded by the effective
outside of the gasket can be calculated as follows:
Ag Area enclosed by the seal − effective outside periphery of a gasket,

d2 3 52
O-ring, or other seal = π d Gasket diameter = π = 9 621 in 2
4 4
The maximum allowable stress for A320 L7 bolts refers to two ASME codes
including ASME BPVC Sec. II and ASME B31.3, the process piping code, is equal
to 25,000 psi at 50 C. It is, however, required to use the allowable stress equal to
20,000 psi when the allowable stress exceeds 20,000, as in this case. The following
requirements must be met in order to ensure that the bolting (bolts and nuts) is
able to provide adequate forces to ensure that the body pieces remain tightly
attached with no possibility of leaks.
Ag
Pc × ≤ K 2 × Sa ≤ 9000
Ab

Ag 9 621
Pc × = 300 × = 708 11
Ab 4 076
K 2 × Sa = 0 35 × 20,000 = 7000
Therefore, 708.11 ≤ 7000 ≤ 7000 is met and the bolts used for the body pieces’
connection of the valve provide sufficient force for tight sealing of the valve.
11.2.3 Threaded Joint Efficiency

This section discusses two types of threaded joints; the first is the threaded joint of
the bonnet or the cover and the second is the threaded joint of the body. The fol-
lowing section discusses threaded bonnets or cover joints.
11.2.3.1 Threaded Bonnet or Cover Joints

In cases where bonnets or covers are attached to valve bodies by threads, the
thread shear area must satisfy the following criteria as per Eq. (11.9):
Threaded Bonnet or Cover Joint Requirement According to

ASME B16.34
Ag
Pc × ≤ 4200 11 9
As
where:
As: Total effective thread shear area.
Example 11.4 Imagine a valve that is in the pressure class 300 and has a sealing
diameter of 4.5 in. There is a threaded contact between the body of the valve and
the bonnet. In light of that, what would be the minimum effective threaded shear
area necessary for a secure connection between the body and the bonnet?
Answer
Ag Ag
Pc × ≤ 4200 Pc × ≤ As
As 4200
Ag 45
Minimum As = Pc × = 300 × = 0 32 in 2
4200 4200
The minimum effective threaded shear area is 0.32 in.2.
11.2.3.2 Threaded Body Joints

As a minimum requirement, threaded body joints must satisfy the following
thread shear area requirements according to Eq. (11.10):
Threaded Body Joint Requirement According to ASME B16.34

Ag
Pc × ≤ 3300 11 10
As
where:
As: Total effective thread shear area.
Example 11.5 The pressure class 150 valve has a minimum effective threaded
shear area of 0.40 in.2, which is in accordance with the minimum pressure class
150. Two pieces of the valve’s body are attached together and there is a threaded
contact between them. In light of this, what would be the maximum gasket diam-
eter necessary for a secure connection between the threaded joint of the body and
the valve?
Answer
Equation (11.10) is valid in this case because it describes the threaded joint
between two body pieces of a valve.
Ag Ag
Pc × ≤ 3300 150 × ≤ 3300 Ag ≤ 8 8
As 0 40
Therefore, the maximum gasket area in this case is 8.8 in.2. It is now possible to
determine the diameter of the gasket as follows:
Ag: Area enclosed by the seal-effective outside periphery of a gasket, O-ring, or
other seal = 8.8 = π(d2/4) (d : Gasket diameter) π(d2/4) = 8.8 in.2 d = 3.35 in.
11.3 Stem Sealing
In order to prevent leakage from the valve stem area to the environment, stem seal-
ing is critical. Furthermore, accuracy and tight sealing of valve stem seals are
essential in fugitive emission applications where hazardous fluids, such as gases
or vapors, can escape from valves that are not properly sealed. Hazardous fluid
is defined as a fluid service in which the potential for personnel exposure is judged
to be significant and in which a single exposure to a very small quantity of toxic
Yoke
Flat washer Gland bolt
Nut
Gland follower
Stem
Packing
Figure 11.2 Stem packing arrangement including gland follower.
fluid can produce serious irreversible harm to someone upon breathing or bodily
contact, even when prompt restorative measures are taken. The valve stem sealing
can be performed with soft thermoplastics and elastomers, such as Teflon (PTFE),
lip seals, V-packs, and Vitons, as well as graphite or graphoil. As illustrated in
Figure 11.2, valve stem seals may consist of a compression packing that consists
of soft materials placed in an area around the valve stem known as the “stuffing
box” and compressed by a valve part known as the “gland or gland follower” to
create stem seals.
The sealing capability of compression packing is defined as its ability to expand
laterally against the stem and stuffing box wall when the gland is tightened. Based
on Eq. (11.11), the relationship between stem seal lateral stress and stem seal axial
stress can be expressed as follows:
Relationship Between Lateral and Axial Stresses in Stem Seals

1−μ
δl = δa 11 11
μ
where:
δl: Lateral stress;

δa: Axial stress;
μ: Poisson’s ratio (ratio of lateral expansion to axial compression of an elastic solid
compressed between two faces calculated according to Eq. (11.12)).
Calculation of Poisson’s Ratio

Lateral strain
μ= 11 12
Longitudinal strain
Strain is defined as deformation per unit of the original length, which is calcu-
lated using Eq. (11.13).
Calculation of Strain
ΔL
= 11 13
L
where:
ε: Strain;
ΔL: Length difference;
L: Length.
Example 11.6 If soft rubber with a Poisson’s ratio of 0.5 is used for stem sealing,
what would be the ratio of lateral to axial stress?
Answer
In order to express the relationship between the lateral stress on the stem seal and
the axial stress on the stem seal, Eq. (11.11) can be used as follows:
1−μ 1−0 5
δl = δa and μ = 0 5 δl = δa δl = δa
μ 05
As the lateral and axial forces are equal, the ratio of lateral to axial stresses is one.
Torque is a measure of the force required to operate a valve between its open and
closed positions. There are several factors that determine the torque required to
operate a valve, such as the packing or stem sealing material, the number of stem
sealing rings, the load of the gland flange, the stem surface finish, the temperature,
and the cycle frequency. Equation (11.14) can be used to calculate the force of
packing on the valve stem.
Calculation of the Packing Force on the Stem

F = π × d × H × GS × μ × Y 11 14
where:
F: Force required to overcome the packing friction (N);

d: Stem diameter (m);
H: Packing height (m);
GS: Compressive stress on the packing (axial gland stress on the packing) (N/m2)
that is calculated according to Eq. (11.15);
μ: Packing coefficient of friction (dimensionless);
Y: Ratio of lateral to axial load transference (for example, 0.5);
GS: Known as the axial gland stress on the packing, is calculated using Eq. (11.15).
Calculation of Axial Gland Stress on the Packing
Tn 02b
GS = 11 15
π D 2 − d2 4
where:
T: the torque applied to the gland flange bolts calculated as per Eq. (11.16);
n: the number of bolts;
b: the nominal bolt diameter;
D: the packing box diameter;
d: the stem diameter.
Calculation of Axial Gland Stress on the Packing
kFBb
T= 11 16
12n
where:
T: Required bolt torque (lb-ft);

k: Nut factor that is typically considered 0.2;
FB: bolt force (lb);
b: Nominal bolt diameter (in.);
n: Number of gland bolt.
Example 11.7 In the gland flange, there are two bolts with a diameter of 3/8
that provide a torque value of 20 Nm. How much torque should be applied to
the bolts to tighten the gland and the packing?
Answer
Equation (11.16) is applicable in this case. The value of torque equal to 20 Nm cor-
responds to 14.75 lb of force foot.
kFBb 12T n 12 × 14 75 × 2
T= FB = = = 4720 lb
12n kb 02× 3 8
Example 11.8 In the previous example, if the stem diameter of the gland pack-
ing is 1/2 in. and the packing box diameter is 1 1/2 in., calculate the compressive
stress on the gland packing.
Answer
For this case, Eq. (11.15) must be used as follows:
Tn 02b 14 75 × 2 0 2 × 3 8
GS = =
π D −d
2 2
4 π 1 52 − 0 52 4
14 75 × 2 0 2 × 3 8 393 33
= = = 250 41 psi
π 1 52 − 0 52 4 1 57075
Example 11.9 How much force is required to overcome the packing friction on
the valve stem if the diameter of the valve stem is 50 mm and the packing height is
20 mm? Axial gland pressure on the packing is 1.4 × 106 newtons per square meter
and the packing coefficient of friction is 0.5. In addition, the lateral to axial force
ratio is 0.5. If the valve stem diameter is 50 mm and the packing height is 20 mm,
how much force is required to overcome the packing friction on the valve stem?
Answer
Equation (11.14) can be used to calculate the force of packing on the valve stem.
F = π × d × H × GS × μ × Y = π × 0 05 × 0 02 × 1 4 × 106 × 0 5 × 0 5 = 1099 5 N
11.1 A ball valve is constructed of a body and a bonnet that are butt-welded
together. It is estimated that the weld strength to base metal is 0.8. What
is the correct answer?
A It is more accurate to refer to the welding process used to join the valve
body and bonnet as furnace buttweld.
B Through the application of a 100% radiography test, the joint efficiency
can be increased to 100%.
C The valve-welded joint between the body and bonnet has a lower strength
compared with the fillet-welded joint.
D Weld joint efficiency can be increased by up to 85% by using spot
radiography.
Answer
Option A is incorrect because the furnace buttweld is the weakest type of
welding, based on Table 11.1, which can only provide a joint efficiency of
60% or 0.6, while the joint efficiency, in this case, would be 80% or 0.8. The
correct answer is B. The choice of option C is incorrect because the joint
efficiency through fillet welding is generally lower than through buttweld-
ing, and it can range from 40 to 50%. Option D is incorrect since spot radi-
ography, as shown in Table 11.1, can increase the joint’s efficiency to 90%.
Accordingly, option B is the correct answer.
11.2 Which of the following statements is true regarding the area of tensile stress
value for various bolting diameters?
A For an M6 bolt, the area of tensile stress is 20.1 mm2.
B The tensile stress area value for a bolt with a diameter of M12 is double
that of a bolt with a diameter of M6.
C For an M12 bolt, the area of tensile stress is 70 mm2.
D All of the options are incorrect.
Answer
There is only one thread per millimeter on an M6 bolt as per Table 11.2,
which has a diameter of 6 mm. According to Eq. (11.6), the area of tensile
stress for an M6 bolt is calculated as follows:
2 2
1 1
At = 0 7854 × d − 0 9382 × = 0 7854 × 6 − 0 9382 ×
n 1
= 0 7854 × 25 62 = 20 12 mm2
Table 11.2 indicates that the number of threads per millimeter for an M12
bolt is 0.57. In accordance with Eq. (11.6), the area of tensile stress for an
M12 bolt is calculated as follows:
2
1
At = 0 7854 × 12 − 0 9382 ×
0 57
2
1
= 0 7854 × 12 − 0 9382 ×
0 57
= 0 7854 × 107 20 = 84 20 mm2

There is no doubt that option A is the correct answer. M6 bolts have a

tensile stress area almost four times greater than M12 bolts, which is the
reason why option B is wrong. Option C is incorrect for the reason that
an M20 bolt’s area of tensile stress is 84.20 square millimeter. Option
D is incorrect as well since option A is the correct answer.
11.3 A top entry ball valve is installed on an oil export pipeline constructed from
low-temperature carbon steel material and has a pressure rating of 1500 psi.
The body and bonnet of the valve are connected by 24 bolts. At all times, the
valve operates at a temperature that is less than 70 C. Bolts and nuts used
are ASTM A320 L7 and A194 Gr.4, respectively. The bolt has 8 threads/in.,
and its diameter is 1 in. When the gasket diameter between the body and
the bonnet is 10 in., what is the total area of tensile stress applied to the
bolts? Do the bolts provide sufficient force to secure the area between
two body pieces tightly without the possibility of leaks, as required by
ASME B16.34?
A The bolt stress area is 10 in.2, and the bolts provide adequate force to
seal the joint.
B The bolt stress area is 11.5 in.2, and the bolts do not provide adequate
force to seal the joint.
C The bolt stress area is 14.5 in.2, and the bolts provide adequate force to
seal the joint.
D The bolt stress area is 17 in.2, and the bolts do not provide adequate force
to seal the joint.
Answer
2
1
At = 0 7854 × d − 0 9743 ×
n
2
1
= 0 7854 × 1 − 0 9743 ×
8
= 0 7854 × 0 7712 = 0 6057 in 2

Due to the 24 bolts connecting the body and bonnet of the valve, the total
bolt stress area can be calculated as follows:
Ab = 24 × 0 6057 in 2 = 14 538 in 2
According to the following formula, since the seal has a diameter of 10 in.,
the area bounded by the actual outside of the gasket can be calculated.
Ag: Area enclosed by the seal-effective outside periphery of a gasket, O-ring,
or other seal = π(d2/4) (d : Gasket diameter) = π(102/4) = 78.54 in.2
According to two ASME codes including ASME BPVC Section II and

ASME B31.3, the process piping code, the maximum stress for A320 L7
bolts is equal to 25,000 psi at a maximum temperature of 70 C. However,
it is required to use the allowable stress equal to 20,000 psi when the allow-
able stress exceeds 20,000, which is the case here. It is critical to meet the
following requirements in order to ensure that the bolting (bolts and nuts)
is strong enough to provide adequate forces to ensure that the body pieces
remain tightly attached with no leakage potential.
Ag
Pc × ≤ K 1 × Sa ≤ 9000
Ab
Ag 78 54
Pc × = 1500 × = 8103 6
Ab 14 538
K 1 × Sa = 0 45 × 20,000 = 9000
Therefore, 8103.6 ≤ 9000 ≤ 9000 is met and the bolts used for the body
and bonnet joint of the oil export pipeline valve provide sufficient forces
for tight sealing of the valve. Therefore, option C is the correct answer.
11.4 In order to connect two body pieces of a ball valve in a pressure class of 600,
twelve 1/2 -in. bolts are installed with 13 coarse threads per inch. As far as
bolts and nuts are concerned, the bolts and nuts used are the ASTM A193
B7 and A194 Gr.2H, respectively. The operating temperatures of the valve
are ambient. What is the largest gasket diameter that can be used in order to
keep body joints in place during operation?
A 14.10
B 15.30
C 15.90
D 16.20
Answer
2
1
At = 0 7854 × d − 0 9743 ×
n
2
1
= 0 7854 × 0 5 − 0 9743 ×
13
= 0 7854 × 0 1807 = 0 1419 in 2

Due to the twelve bolts connecting the body pieces of the valve, the total
bolt stress area can be calculated as follows:
Ab = 12 × 0 1419 in 2 = 1 703 in 2
The allowable stress of the bolts in A193 B7 is 20,000 psi. Thus, the middle
section of the equation is calculated as follows:
K 2 × Sa = 0 35 × 20,000 = 7000
The maximum gasket diameter results in a maximum value of the left

side of the equation Pc × (Ag/Ab) and the maximum value of Pc × (Ag/Ab)
is when it is equal to K2 × Sa = 7000.
Ag Ag
Pc × = 7000 600 × = 7000 Ag = 198 68 in 2
Ab 1 703
d2
Ag = π 198 68 in 2
4
d2
=π d2 = 252 d = 15 90 in
4
11.5 There is a threaded connection between the body pieces of a ball valve in
pressure class 150. The gasket area is at its maximum. What is the ratio of
the gasket area to the effective thread shear area of a ball valve?
A 20
B 22
C 24
D 26
Answer
In this case, Eq. (11.10) is valid since it describes the threaded joint between
two body pieces of a valve.
Ag
Pc × ≤ 3300
As
Gaskets should not result in an area exceeding the left side of the equation
by more than 3300.
Ag Ag Ag
Pc × ≤ 3300 150 × = 3300 = 22
As As As

11.6 The lateral force caused by a stem seal is double the axial force created by
the gland on the sealing. What is Poisson’s ratio of the stem seal?
A 0.13
B 0.23
C 0.33
D 0.43
Answer
Equation (11.11) can be used to express the relationship between the lateral
stress on the stem seal and the axial stress on the stem seal as follows:
1−μ 1−μ
δl = δa 2= 2μ
μ μ
= 1−μ 3μ = 1 μ = 0 33
11.7 If the valve stem diameter is 2 and the packing height is 1 , how much
force is required to overcome the packing friction on the valve stem?
The axial gland pressure on the packing is 200 psi and the packing coeffi-
cient of friction is 0.5. Additionally, the ratio of lateral to axial force is one.
A 570 lb
B 628 lb
C 750 lb
D 820 lb
Answer
Equation (11.14) can be used to calculate the force of packing on the
valve stem.
F = π × d × H × GS × m × Y
= 3 1415 × 2 × 1 × 200 × 0 5 × 1
= 628 3 − lb force
11.8 When two bolts are used to fasten a gland flange with a diameter of 1/2
and torque values of 12 lb-ft are applied, what is the value of compressive
stress on a gland packing? It should be noted that the diameter of the gland
packing is 1/2 in. and the diameter of the packing box is 1 1/2 in.
A 152.87
B 155.90
C 164.78
D 171.90
Further Reading 405
Answer
For this case, Eq. (11.15) must be used as follows:
Tn 02b 12 × 2 0 2 × 0 5 240
GS = = = = 152 87 psi
π D 2 − d2 4 π 1 52 − 0 5 2 4 1 57
Further Reading
Aksenov, A., Iliine, K., Luniewski, T. et al. (2006). Oil leakage through a valve stem seal.
ABAQUS Users’ Conference 2005, Stockholm, Sweden (18–20 May 2006).
American Petroleum Institute (API) 6D (2014). Specification for Pipeline and Piping
Valves, 24e. Washington, DC: American Petroleum Institute (API).
American Society of Mechanical Engineers (ASME) (2016). Process Piping. ASME
B31.3. New York: American Society of Mechanical Engineers (ASME).
American Society of Mechanical Engineers (ASME) (2020). Boiler and Pressure Vessel
Code (BPVC) VIII Div.02. Rules for Construction of Pressure Vessels Division 2 –
Alternative Rules. New York: American Society of Mechanical Engineers (ASME).
American Society of Mechanical Engineers (ASME) (2020). Boiler and Pressure Vessel
Code (BPVC) Section II. Materials. New York: American Society of Mechanical
Engineers (ASME).
American Society of Mechanical Engineers (ASME) B1.1 (2019). Unified Inch Screw
Threads (UN, UNR, and UNJ Threaded Forms). New York: American Society of
Mechanical Engineers (ASME).
American Society of Mechanical Engineers (ASME) B16.34 (2020). Valves – Flanged,
Engineers (ASME).
Cox, J.C. (2010). Thwarting fluid-system leaks: saving energy, reducing leaks. Digital
Refining, Processing, Operations and Maintenance 15 (9): 75–79.
Dassault Systems (2022). Tensile stress area of a bolt. https://help.solidworks.com/
2021/english/SolidWorks/cworks/r_Tensile_Stress_Area_Bolt.htm (accessed 5
May 2022).
Drago, J. and Evans, W. (2017). Gauging the force effects of stem packing on valve stem
actuation. Proceedings ASME/NC Pump and Valve Symposium, PVS2017-3518, Silver
Spring, Maryland, USA (17–18 July 2017).
Engineering Edge (2019). Welded joint efficiency table recommendations.
https://www.engineersedge.com/weld/welded_joint_efficiency_14419.htm
(accessed 4 May 2022).
Fluid Sealing Association (2018). Compression Packing Technical Manual, 4e. Wayne,
PA: Fluid Sealing Association.
Gobind, K. (2016). Valve fugitive emissions vs. current industry practices – part 2. Valve
World Magazine. 21 (10): 121–122.
Oxford: Elsevier.
Picard, D. (1999). Fugitive emission from oil and natural gas activities. Journal of Good
Practice Guideline and Uncertainty Management in National Greenhouse Gas
Inventories. 103–127.
Sotoodeh, K. (2020). A review of valve stem sealing to prevent leakage from the valve
and its effect on valve operation. Journal of Failure Analysis and Prevention
https://doi.org/10.1007/s11668-020-01050-1.
Zulaika, L.C. (2016). Torque increase due to heavy packing tightening in valves. AMPO.
Valve World Conference 2016, Dusseldorf, Germany (29 November–1
December 2016).
407
Index
Note: Page number followed by ‘f’ and ‘t’ refer to figures and tables respectively.
a and API standard orifice

absolute pressure measurement 5 sizes 58t, 287t
absolute relieving pressure 230 Boiler and Pressure Vessel Code
acoustic conversion efficiency 213 (BPVC) 57
acoustic efficiency factor 205, 213 liquid services 79
for cavitation flow 214 American Society of Mechanical
for turbulent flow 214, 230 Engineers (ASME) B16.34
acoustic fatigue 190, 193f, 215 standard 125
analysis 192–203, 224 ASME BPVC Sec. VIII Div. 02,
noise and 215f 134–135
acoustic power ratio 213t, 228 conservation approach 125–129
actuator pressure sensor 347 minimum valve body
actuators 13, 345t, 379t thickness in 125
aerodynamic noise 191 nonconservation method 129–134
in control valves 203–208 NPS and pressure class 126–128
air: pressure-temperature ratings
flow coefficient for 3–6 in 128, 129
flow factor for 6–8 amplitude, of sound wave 185
safety valve rated capacity for 294 angle of twist 375
allowable overpressure 261, 262 anticavitation trim 117
allowable stress 173 anticorrosive 151
American Society for Testing and API 520, 267
Materials (ASTM) 128 ratio of specific heat 288
A350 LF2, 181 relief valve reaction forces as 284–286
American Society of Mechanical superheat steam correction
Engineers (ASME) 30 factor 77t
408 Index
API 623, 114 body closure machining drawing 146f

API 526 standard 57 flow curves for 11f, 19f
API standard orifice sizes 58t, 287t pup piece 173f
API 521 standards 94, 218–221 stem of 365f
ASME B16.25 172 torque values 364t
ASME B31.3 code 293, 393 v-notch 40, 43t, 56, 191
longitudinal weld joint quality bellows safety valves 265
factor 387 bench test 261
ASME BPVC Sec. VIII Div. 02 134–135 bending stress 169, 169f, 171
ASME B16.34 standard 144 Bernoulli coefficients 38
atmospheric pressure values 5 blowdown pressures 274, 276
austenitic stainless steel 163, 178 body-guided disk/gate 370, 373, 381
availability 331–336, 348 Boiler and Pressure Vessel Code
and probability of failure on demand (BPVC) 57
(PFD) 339 bolted body joints 392–394
axial check valves 240 bolted bonnet/cover joints 388–392
axial flow check valve 27–28, 27f, bolted joint efficiency 388–394
31, 33t bolt stress 393, 402
axial gland stress 398 bolts vs. nuts 389f
breakaway torque 364
b break to close (BTC) 364
backpressure 59, 64, 71, 72, 107, 263, break to open (BTO) 364
264, 269 buckling 369f
safety valve for 270f direct impact on 369
at valve discharge 273 buildup backpressure 264
backpressure correction factor 63, 76, bulk modulus, of elasticity 199, 199t
78, 79, 83, 89, 91, 93, 108, burst disk 60
264–265, 267 butterfly valves 24, 25f, 25t, 40, 43, 104,
for balanced bellows PRVs 80f 120, 315t, 358
for balanced bellows PSVs 64f buttweld joint 171f, 176
for gas and vapors 64–65, 265, 265f
in liquid service 80 c
balanced bellows cage-guided control valves 52t
capacity correction factor 268f capacity correction factor 63, 71, 72,
PRVs in liquid service 80, 268 85f, 89, 93, 268f
PSVs in vapors and gases 64f, 65f capacity index 1
viscosity correction factor 80f carbon dioxide (CO2) 149
balanced bellows safety valve 63, activity coefficient 156
265, 269 corrosion. See sweet corrosion
ball valves 22, 358, 361, 393, 399 effect of 153
Index 409
fugacity or activity 156 cladding carbon steel 151

partial pressure 153–154, 156, 158 closing pressure 274
temperature 153 CL300 wafer ball valve 139
carbon equivalent 165–167 coefficient of discharge 58, 269
of A350LF2 181 cold set pressure 300. See also set
carbonic acid 150 pressure
carbon steel: combination correction factor 91, 93
body valves 128 compact wafer ball valve 139f
pipe grades 167 compressibility factor 66–68, 272–273
vessel 102 determination chart 70f
welded valves 166 compressible fluids 191
carbon steel material 297 compressive stress 175f
carbon steel pipe grades 167t condensate development 160
carbon steel valve 146 condition monitoring
cathodic protection 168–169 (ValveWatch) 347–348, 352
cavitation 112–116, 208, 214 conservative method 142
analysis 115–116 constant backpressure 264
characteristics 122f continuous modes of operation 337
on control valve plug 114f IEC 61508 for 338
essential parameters 113–115 control loop 34
and flashing 116 control valves 9, 13, 112, 222
remedy 119f aerodynamic noise in 203–208
cavitation index 115–116, 116t, 120 authority 14
characteristic pressure ratio 210, 212 hydrodynamic noise in 208–215, 227
check valves: noise 190–191, 203–215
axial flow nozzle 27f in plants 35f
dual plate 27f, 238–240 plug, cavitation on 114
dynamic characteristics and pressure relief valves 192
of 253, 253f control valve sizing 34–56
flow coefficient 12t Bernoulli coefficients 38
nonslamming feature for 237 coefficients for 41
nozzle 27, 28, 31, 32, 239, 240 equation constant 37, 48
sizing 26–33, 104 expansion factor 48
swing. See swing check valve for gas and steam 47–56
water hammering and pressure loss for liquids 34–47
in 233–243 parameters and definitions 36t
chloride 163 piping geometry 37, 48
in fluids 149 in plant 35
choked flow 2, 39, 117, 118f, 209 pressure drop ratio factor 49
flow coefficient for 44 selection method 34
410 Index
conventional relief valves 269 critical pressure 68, 88

conventional safety valves 64, 74, and temperature 69t
260, 263 critical pressure drop 51
backpressure percentage for 65–66, flow coefficient 3–5, 51–52
266–269 flow factor 7
component of 260f critical pressure ratio 59
overpressure of 268 critical temperature 68, 69
standards for 300 critical velocity values 31, 32
correction factor axial flow check valve 33t
for high pressure steam 289f dual plate check valve 31t
for specific heat 290t crude oil transportation system 324
corrective CO2 corrosion rate 154–161 cryogenic ball valve 383
condensate development 160 cryogenic valves 357, 369
glycol (methanol) effect 157–158 quarter-turn 376–377
inhibitor efficiency 159–160 cubic meter per hour 48
pH effect 160–161
scaling effect 158–159 d
system effect 156–157 decibels 189
water cut effect 159 demand charges 241
corrosion 112, 149 demand modes of operation 337
accurate prediction and depression 186
modeling 150 Det Norske Veritas (DNV)
carbon dioxide. See sweet corrosion standard 169–177
carbon equivalent 165–167 HISC guideline 176, 182
by cavitation 114f “Dewaard and Milliams” (DWM)
final rate 161t model 151, 153–155, 159–160
hydrogen-induced stress cracking diagnostic coverage (DC) 342
(HISC) 149, 167–177 differential pressure 1
inhibitor efficiency 159–160 differential pressure ratio 209
inhibitor injections 155, 160 dimensionless correction factor 76
pitting 162–165 direct flow rate control 13
corrosion allowance 150, 151, 162, 177 discharge coefficient 89
corrosion inhibitors protection 151, 179 discharge pressure 72
corrosion-resistant alloys (CRAs) 151, disk/gate design 370
162, 259 DN 12
cracking pressure 274 DNV-RP-F112, 172
of valve 28 downstream plants 149
critical flow pressure 66, 71, 72 “driven” gear 360
to mass flow rate units 59 dry carbon dioxide 153
mass flux for 88–89 dual plate check valve 27, 27f, 28, 29, 31,
volumetric flow rate 59–60 190, 238, 238f, 239, 240
Index 411
minimum flow/critical velocity linear strain criteria 176

for 31t nonlinear strain criteria 176
test-based dynamic characteristic types 175
of 247f finite element method (FEM) 134
ductile steel 175 fire case 93–103
dynamic seals 385 mass flow rate for 100–101
safety valve for 96–103
e symbols for 94
elastic deformation 372 fire conditions 262
elasticity, bulk modulus of 199 flanges 136, 279
elastic-plastic analysis 169 flashing 116–118, 123, 208
elastomers 396 cavitation and 116
electronic control devices 305 flow 2
emergency shutdown system pressure profiles 117f
(ESD) 304, 305, 305f flow capacity:
end to close (ETC) 364 flow coefficient and. See flow
end to open (ETO) 364 coefficient
energy charge 241 rangeability 12–14
environmental factor 98, 98t turndown 12–14
equal percentage 9 valve authority 14–15
equation constant 37, 38t valve gain 15–16
erosion 112, 116 flow coefficient 1, 56
advantage 24
f for air and gases 3–6
factory acceptance tests (FATs) 304 for butterfly valves 24, 25t
failure function 330–331 cage-guided control valves 52t
failure mode and effect analysis chart and flow curve 8–12
(FMEA) 308, 310f for choked flow 44
failure mode classification 316f critical pressure drop 51–52
failure probability, of parallel and factor 8
system 326–327 and flow factor 18
failure rate allocation method 324 for full-bore ball valves 23t
failure rates 313–317 gas-specific gravity 50
series and parallel blocks with 329 for gate, globe, and check valves 12t
of series network 324 for liquids 1
values 343t, 345, 349 mass flow rate units 50
failures chart 312f for noncritical pressure drop 52
ferritic–austenitic stainless steels 168 for reduced bore ball valves 23t
ferrous carbonate (FeCO3) 158 for saturated steam 51–52
fillet welding 388 solve for 44
finite element analysis (FEA) 134 for superheated steam 52–56
412 Index
flow coefficient (cont’d) sizing relief devices for 73

on valve opening percentage 9–10 sonic velocity in 198–200
flow control 112 specific gravity of 3t–5t, 73
flow curve 8–12 Z-factor 56, 66, 271–272
flow factor 2, 18 gas–liquid service 292–293
for air and gases 6–8 gas-specific gravity 50
flow coefficient and 8, 18 gate valves 370, 373
for liquids 2–3 flow coefficient 12t
flow rate 1, 29 gauge backpressure 64–65, 265
maximum 39–40 gearbox 358f, 360f
valve opening percentage 10 gearboxes 360
flow velocity values 29, 29f, 31 gear-operated valves 359
fluid noise 190–191 gear ratio 360
fluid pressure drop 237 general arrangement (GA) valve 11
fluids, sonic velocity in 199–200 gland or gland follower 396
fluid velocity 29 gland packing 399
frequency 185 globe-style control valve 42
and transmission loss 207f globe valves 6f, 112, 118, 357, 370, 382
fugacity coefficient 154 flow coefficient 12t
fugitive emission 385, 386 sizing coefficients for 41t–42t
full-bore ball valves 22–23, 23t, 103 T-pattern 113f
glycol 155, 157–158, 179
g
gain values 15–16 h
gallon per minute (gpm) 37 handwheel 358f, 359f
gas compressibility factor 106, 271 handwheel-operated valves 360
gas/gases: hardness 175
backpressure correction factor hardware fault tolerance 343
for 265f, 266f hazard 306
C factor and “k” ratio of 62 hazardous event 306
constant 61t, 103, 288 hazardous fluids 395
control valve sizing for 47–56 hazardous situation 306
density calculation 201 head loss coefficient 39, 53
expansion factor 48–49 health, safety, and environment
flow coefficient for 3–6 (HSE) 309
flow factor for 6–8 heat/thermal capacity 266, 271
law calculation 201 high-demand systems 337
number of moles in 201 IEC 61508 for 338t
on pressure and temperature 202 high-pressure steam correction
safety relief valve sizing for 59–75 factor 289–290
safety valve rated capacity high-subcooling systems 93
for 294–295 hybrid model 328
Index 413
hydraulic expansion (thermal function of time for 320

expansion) 93–103 noise in 189–192
on API 521 standards 95–96 reliability. See reliability
on ISO 23251 standards 95–96 safety. See safety
hydraulic pressure 347 safety valves. See safety valves
hydraulic shock 235 inhibitor efficiency 159–160
hydraulic turbulence 209 inhibitor injections 155, 160
hydrocarbon liquids, cubical expansion inlet flow pressure 260
coefficient for 96t intensity 185
hydrocarbon services 152 internal pitting corrosion 163
hydrodynamic noise 191 International Electrotechnical
in control valves 208–215, 227 Commission (IEC) 306
preliminary calculations 209–213 Iranian Petroleum Standard (IPS) 151
secondary calculations 213–215 ISO 4126-9:
hydrogen atoms 168 safety valve noise 216–218, 283
hydrogen-induced stress cracking in steady-state flow 283–284
(HISC) 149, 167–177 isolation valve sizing 22–25
attack, risk 172 ISO 23251 standards 94
and cathodic protection 168–169
DNV guideline for 176, 182 j
and DNV standard 169–177 jet fluid, velocity 211
and stress 168 joint efficiency 385, 400
and vulnerable materials 168 bolted 388–394
hydrogen sulfide (H2S) 149, 306 threaded 394–395
attack 166 weld 386–388
Joukowsky’s formula 243
i
ideal gas 272 l
concept of 67 Laplace transform variable 318
constant/coefficient 60 leakage 114
IEC 60534-8:3 203, 204 lever-operated valves 358
IEC 61508 306–307, 312 likelihood of failure (LOF) 195
continuous operation conditions 338 limiting flow rate 39
for high-demand conditions 338t linear elastic analysis 169
for low-demand conditions 307, linear elastic stress-strain model 169
337t, 338 linear flow 9
PFD based on 338–339 characteristics 9–10
IEC 61511 standard 306–308 linear motion valves 10
Inconel 625, 151 linear stem motion valves 370
industrial safety 305 liquefied natural gas (LNG) 357
industrial valves: liquid critical pressure ratio
cavitation. See cavitation factor 40, 47
414 Index
liquid flashes 88–91 mean downtime 339–342

liquid pressure recovery factor 209 mean time between failure (MTBF) 318
liquid relief, PSV for 79–85 mean time to detection (MTTD) 319
with capacity certification 79–84 mean time to failure
without capacity certification 84–85 (MTTF) 317–318, 322
liquids: mean time to repair and recovery
control valve sizing for 34–47 (MTTR) 319, 340, 348
cubical expansion coefficient mechanical noise 190
of 95, 96 mechanical vibrations 208
flow coefficient for 1 membrane plus bending stress 171f,
flow factor 2–3 174, 182, 183
pressure drops 117 membrane stress 169, 169f, 173
pressure recovery 40–44 metallic components 385
longitudinal weld joint quality metal loss 150
factor 387t, 388 methanol 157
low-demand systems 337t, 338 miscellaneous
low-noise trim 193, 196 joint efficiency. See joint efficiency
low-subcooling systems 92 stem sealing 357, 385, 395–399
low-viscosity fluids 81 modulus of elasticity 246t, 372, 373
modulus of rigidity 376t
m monoethylene glycol (MEG) 155
Mach number 198–203, 206, 223 MSS SP134 standard 370–371, 383
manual valve 358f, 359f multicomponent system 88, 91
mass flow rate units 50
critical flow conditions 59 n
subcritical flow condition 73 Napier’s equation 76, 78, 289
wetted vessel 100 natural disasters 168
mass flux 92–93 natural gas 54, 155
for critical flow 88 network diagram 323f
for subcritical flow 89 nickel alloys 151, 164
material failure 149 noise 116
material yield strength 373 and acoustic fatigue 215f
maximum accumulated pressure 262, acoustic fatigue analysis 192–203, 215
263, 300 aerodynamic. See aerodynamic noise
maximum allowable stem torque API 521, safety valves as 218–221
(MAST) 348, 363–369 control strategies 191–192
in quarter-turn valves 377 in control valves 203–215
maximum allowable strain and effects on human brain 187
limit 176t, 177 emission tests 215
maximum allowable working pressure exposure 190
(MAWP) 261, 296, 299 fluid 190–191
Index 415
hydrodynamic 191 NORSOK L-002, 216

in industrial valves 189–192 NORSOK L-001 standard 143
intensity 195, 219f nozzle check valves 27, 28, 31, 32,
ISO 4126-9, safety valves as 216–218 239–240, 240f, 249–251
levels, relative 189 numerical constant 37
Mach number 198–203
mechanical 190 o
negative effects of 190 occasional variations 277
peak frequency 205, 206 Occupational Safety and Health
for pipes and valves 192–222 Administration (OSHA) 190
pollution 186 oil and gas industry:
prediction method 209 carbon dioxide corrosion in 177
in pressure safety or relief diverse sectors of 189
valves 215–222 models in 152
propagation values 206 offshore 304
for safety valves 217t, 218t, 222t piping systems in 149
sound 185–186 stainless steels and nickel alloys
sound power level 193–197 in 164t
source magnitude 205 strategies 151
VDI 2713, 221–222 valve failure in 149
and vibration 190 oil-forming film 159
noise-producing valves 190 “Omega method of Leung” 86
nomenclature, in IEC 60534-8:3 204t Omega parameter 86–88, 91–92
nominal bolt 390t critical flow ratio on 89f
nominal pipe size (NPS) 12, 22, 172 saturated 109
non-compressible fluids 191 open discharge 297
nonconservation method 129–134 opening pressure 274
noncorrosion-resistant materials 151 operating time 320
noncritical pressure drop 51 overpressure 259f, 261, 277
flow coefficient 5–6
flow coefficient for 52 p
flow factor 7–8 packing friction forces 362t, 399
nondestructive testing (NDT) 386 parallel network diagram 326f
nonfire conditions 262 parallel system:
non-hydrocarbon fluids 152 availability 335–336
nonideal gases 67, 272 reliability 327–328
nonideality, of hydrocarbon gas 154 unavailability 334–335
nonlinear strain criteria 176 parameter
nonrecoverable losses 274 estimation 102f
nonreturn valves. See check valves sigma 115
nonslam check valves 234 used for PRV load calculation 285t
416 Index
peak stress 169 pressure class 35

percentage of backpressure 266–269 pressure drops:
percentage of gauge ratio factor 49
backpressure 64, 265 for valves 1, 44, 117
pH effect 160–161 pressure level 188
pipe anchorage method 245 pressure rating 277
pipe coincidence frequency 207 pressure ratios 209
pipe cross-section area 29–30, 33 pressure-reducing devices 189
pipe fittings 24 pressure relief valves (PRVs) 56, 73, 91,
pipeline system 350 192, 196, 258
pipeline valves 134 factors 281
risk for 311t installation on pressurized
risks associated with 311t equipment 57f
pipe/piping system 149, 150, 156, 350 noise in 215–222
acoustic fatigue analysis 192–197 parameters used for 285t
flow rate, flow velocity, and cross- reaction forces for 283–284
section area 29f pressure-relieving devices 277
handling 48 pressure safety disk 60
internal and external diameters 197 pressure safety valves (PSVs) 22, 57, 78,
internal diameter 279–280 105, 189, 226, 258, 281, 286t
longitudinal weld joint quality factor control and 198
for 387t critical or subcritical flow for 92t
Mach number 198–203 discharge point of 219
noise for 192–222 flange 279
for overpressure protection 259f in gas service 286
valves and 162 in high-subcooling systems 93
water hammering in 235 impact of 283
pipe ring frequency 207 for liquid relief 79
piping geometry factors 24, 37, 53 in low-subcooling systems 92
liquid pressure recovery and 40–44 MAWP for 297
piping injected gadget (PIG) 22 noise in 215–222
pitting corrosion 162–165 noncertified 85f
in carbon steel 163f on pressure vessel 298f
mechanisms and prevention 180 reaction forces for 283–287,
pitting resistance equivalent number 291–293
(PREN) 163, 165 vessel protection system 220
plant, control valve in 35 pressure-temperature rating values 128,
plug valve 121f 129t, 133t, 142
Poisson’s ratio 245, 255, 396, 397 pressure vessel, PSV on 298f
of stem sealing 404 principle of operation 278–282
post weld heat treatment (PWHT) 386 probability of failure on demand
post welding inspection 386 (PFD) 338–339
Index 417
availability and 339 and unavailability 334–336

risk reduction factor and 339 and unreliability 319–332
process optimization 191–192 relief piping system 215f
process shutdown (PSD) relief valves 57
system 304, 305f relieving pressure 260, 270, 301
process variables 34 repair rate 317
quarter-turn cryogenic valves 376–377 reseat pressure 274
quarter-turn valves 10–11, 24, 374, 380 residual risk 307
MAST in 377 residual stress 175
quick opening 9 resistance coefficient 39
Reynolds number 79, 81–83
r risk:
radiography test (RT) 387 analysis 308–312
rangeability 12–14, 18 for pipeline valve 311
“rated coefficient of discharge” 58 probability score 309t
ratio of specific heat 65, 266, 267t, 271 severity 309t
API 520 standard 288 risk priority number (RPN) 308
C factor and “k” 62t risk reduction factor (RRF) 339
gas constant on 61t Roark’s equation 366–367
reaction forces: rotating end 375
for PSVs 283–287 running torque (RT) 364
safety valves 282–294 rupture disk 60, 65, 90
real gases 67
reduced bore ball valves 22, 23t s
reference point 5 safe failure fraction (SFF) 342–347,
regulation. See flow control 343t, 353
relative noise levels 189t safety:
reliability: diagnostic coverage 342
availability 331–334 ESD systems 304–305
block diagram 340f management process 305
chart 320f mean downtime 339–342
concept of 331 probability of failure on demand
diagnostic coverage 342 (PFD) 338–339
and failure function 330–331 PSD systems 304–305
function 321 risk analysis 308–312
mean downtime 339–342 SIL. See safety integrity level (SIL)
parallel system 327–328 standards 305–308
safe failure fraction (SFF) 342–347 system incidents and
safety-critical valves 304 failures 312–319
series systems network 323 safety analysis reports (SARs) 305
system incidents and Safety and Environmental issues
failures 312–319 (HSE) 149
418 Index
safety-critical valves 304, 322, 341 minimum inlet diameter of 281

safety critical valves, fault tree for 341f noise for 217t, 218t, 222t
safety instrumented system opening condition 279
(SIS) 307–308, 307f operation 259–278
safety integrity level (SIL) 305, 336–338 parts 259
condition monitoring principle of operation 278–282
(ValveWatch) 347–348 reaction forces 282–294
diagnostic coverage 342 relieving flow for 281
formulas and calculations 349 schematic 278f
mean downtime 339–342 set pressure 262t, 300
probability of failure on demand saturated steam 51, 302
(PFD) 338–339 flow coefficient for 51–52
safe failure fraction (SFF) saturation pressure 108
and 342–347, 343t saybolt universal seconds (SSUs) 81
safety relief valve 63f, 277 scaling effect 158, 179
safety relief valve sizing 56–103 SCF per hour (SCFH) 48
absolute relieving pressure 62–63 seat diameter 279
backpressure correction factor 63–64 sensors 347
critical flow pressure 66, 71, 72 sequential model 203
for fire case 93–103, 94f series systems network 323
for gas or vapor relief 59–75 set pressure 105, 261, 263, 264
hydraulic expansion (thermal in balanced bellows safety valve 267
expansion) 93–103 of safety valves 300
for liquid relief 79–85 shear modulus 376
procedure 57 shear stress 376, 383
relieving pressure 62 shutdown systems 304
for steam relief 75–78 silencers 192
subcritical flows 73–75 simmer 277
two-phase liquid/vapor relief 85–93 simplified network 330f
types 56 single component system 88, 91
safety valves 230, 258. See also pressure single port control valve 42
relief valves (PRVs); pressure safety slamming 252
valves (PSVs) swing check valves 238f
for backpressure 270f sloped valve 254f
blow-down pressures in 276f slug flow forces 282
capacity conversion 294–296 socket-welding 128
closing condition 278–279 solids, sonic velocity in 199–200
conventional 260 sonic flow correction factor (SFF) 193
design 259–278, 261t sonic velocity 222
maximum accumulated pressure limits in fluids and solids 199–200
for 262t in gas 198–200
Index 419
sound 185–186, 186f. See also noise strain gauge 347

transmission loss 206–207 strain limit 169
sound power level (SWL) 187, 193–197, stress 168
216, 217, 224 stress limits 169
safety valve 216–217 stress–strain curve 175, 176f, 370, 371f
sound pressure level (SPL) 187, 188, stuffing box 396
188f, 193, 216 subcritical flow coefficient 73–75, 74f
in downstream piping 206 subcritical pressure drop 88
and pressure level 188–189 flow coefficient 5–6
safety valve 217–218 subsurface safety valve (SSV) 304
sound pressure and 188 superheated steam:
sound waves 185–186 correction factor 77t
specific gravity 2 valve’s flow coefficient for 51–56
common gases 3t–5t superimposed backpressure 264, 269
of gas 73 sustained loads 293–294
specified minimum yield strength sweet corrosion 150f
(SMYS) 170, 175 corrective process. See corrective CO2
for membrane plus bending corrosion rate
stress 171 corrosion rate 153f, 155f
spring-loaded check valves 31 on DeWaard and Milliams 153–155
spring torque, categories of 31 final corrosion rates 161–162
spring valves 258 inhibitor efficiency 159–160
stainless steel 151, 164, 259 mechanism 150
startup opening 277 mitigation 151
static pressure loss 275 mole fraction 154
static seals 385 oil and gas industry 177
steady-state flow 283 prediction 152
ISO 4126-9 in 283–284 rate calculation 152–162
steam, control valve sizing for 47–56 reduction effect 158
steam relief 75–78 temperature on 153
Stellite 6 (UNS R30006) 114 swing check valves 26–27, 26f, 27, 29,
stem design: 190, 234–237, 237f, 248, 249
buckling prevention 369–374 in open position 234f
MAST 363–369 slamming 238f
torsional deflection system incidents 312f
prevention 374–376
stem-guided disk/gate 370, 383 t
stem packing arrangement 396f Teflon (PTFE) 396
stem sealing 357, 385, 395–399 tensile stress 168, 175, 175f, 373,
Poisson’s ratio of 404 400, 401
strain 175, 183, 246, 372, 373, 397 test pressure 260
420 Index
thermal expansion 93–103 flow rates and opening

thermodynamic equilibrium 113 percentages 16t
thickness value 172 flow resistance coefficient 236
threaded body joints 395 head loss 236
threaded bonnet/cover joints 394 inside diameter in
threaded-end valves 128 millimeters 126t–128t
threaded holes 136 Mach number 198–203
threaded joint efficiency 394–395 neck section 138f
throttling of fluids. See flow control noise for 192–222
tolerable risk 307 noise propagation values 206t
torque 375, 378, 397 outlet velocity 200
torsional deflection 375 and piping system 162
prevention 374–376 size of 16, 103
T-pattern (tee-pattern) globe valve 112, and sloped valve 254
113, 113f stem 375f
transmission loss 206–208, 207f style modifier 210
trim design 114 types 304
with multiple stages 115 wall thickness. See wall thickness
turbulent flow 190 valve authority 14–15, 20
acoustic efficiency factor for 214, 230 valve discharge 280
turndown 12–14 valve flow capacity 1. See also flow
two-phase flow scenarios 86t capacity; flow coefficient
two-phase liquid/vapor relief 85–93 valve gain 15–16
valve opening percentage
u flow coefficient on 9–10
ultrasonic test (UT) 387 flow rate 10
unavailability 331, 334–336 valve operation:
uniform corrosion 150 stem design. See stem design
unreliability 319–331 valve torque 358–362
unwetted vessels 95, 101–103 valve pup piece 172
external fire and 101 valve sizing 22
control. See control valve sizing
v isolation 22–25
valve(s): in liquid services 104
for acoustic fatigue 193 nonreturn (check) 26–33
acoustic fatigue analysis 192–197 safety relief. See safety relief valve
cavitation characteristics 122f sizing
characteristic pressure ratio 210 valve stem seals 385
flow characteristics of 8–9, 8f, 17, 121 valve torque 358–362
flow coefficient. See flow coefficient and actuator 379
flow factor 18 ball valve 364t
Index 421
gear-operated valves 359 equations for minimum

lever-operated valves 358 valve 130t
in physics 359–360 net aerodynamic sound level 208
valve trim 114 pipe’s internal and external
ValveWatch system 347–348, 347f, 352 diameters 197
vapor/gas pressure 101, 113 wafer design thickness
vapor relief 85–93 validation 136–142
mass fraction of 90 water critical pressure ratio factor 40f
safety relief valve sizing for 59–75 water cut effect 159, 179
sizing relief devices for 73 water hammering 233, 235f, 282
vapor, safety valve rated capacity calculations 243–249
for 294–295 in piping system 235
VDI 2713, safety valves noise and pressure loss in check
as 221–222 valves 233–243
vena contracta 209–212 slamming and 252
venturi effect 117 wave speed or acoustic
vibrations 190, 193 velocity 245–246
viscosity 81, 82f welded valve 181t
viscosity correction factor 80f welding process 172
v-notch ball valve 40, 43t, 56, 191 weld joint efficiency 386–388
volumetric flow rate 50 wetted vessels 95, 96–101, 97f
critical flow conditions 59–60 for different surface 99t
vulnerable materials 168 dimensional designations for 99f
in fire situation 96t
w total heat absorption 97
wafer butterfly valves 136f
wafer design thickness y
validation 136–142 yield point 175
ball valve body machining 139
wall thickness 143 z
ASME B16.34. See American Society of Z-factor, of gas 56, 66, 67, 74, 106,
Mechanical Engineers 271–272
(ASME) B16.34
WILEY END USER LICENSE AGREEMENT
Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.

Industrial Valves Calculations_Wiley

Uploaded by

Copyright:

Available Formats

Industrial Valves Calculations_Wiley

Uploaded by

Document Information

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

Industrial Valves Calculations_Wiley

Uploaded by

Copyright:

Available Formats

Industrial Valves

Calculations for Design, Manufacturing,

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Limit of Liability/Disclaimer of Warranty

Library of Congress Cataloging-in-Publication Data

Cover Image(s): Rawf8/Adobe Stock Photos; Who is Danny/Adobe Stock Photos

Set in 9.5/12.5pt STIXTwoText by Straive, Pondicherry, India

2.5.1 Sizing for Gas or Vapor Relief 59

3 Cavitation and Flashing 112

4 Wall Thickness 125

5 Material and Corrosion 149

5.2.3.1 Basic CO2 Corrosion Rate 152

7 Water Hammering 233

8 Safety Valves 258

9 Safety and Reliability 304

10 Valve Operation 357

10.3.2 Buckling Prevention 369

Flow Coefficient (Cv) Calculation for Liquids

Cv: Flow coefficient (US gal per minute [gpm]/psi);

Note: Water-specific gravity is equal to one.

Specific Gravity (SG) Calculation

ρfluid: Density of the fluid passing through the valve;

To summarize, the flow coefficient calculated by Eq. (1.1) is a parameter

Flow Factor (Kv) Calculation for Liquids

Kv: Flow coefficient (m3/h bar);

Cv: Flow coefficient (Standard cubic feet per minute [SCFM]/psi);

Table 1.1 Specific gravity of some common gases.

Gas Specific gravity

Acetylene (ethyne) – C2H2 0.90

Table 1.1 (Continued)

Gas Specific gravity

Butane – C4H10 2.0061

Table 1.1 (Continued)

Gas Specific gravity

Flow Coefficient Calculation for Air and Gases (Noncritical

PO: Outlet gas absolute pressure (psia);

Air Gas inlet Gas outlet

Q SG T + 460 100 0 565 60 + 460

= 0 0735 × 0 2984 = 0 022 SCFM psi

Kv: Flow factor (m3/h bar);

PO: Outlet gas absolute pressure (bara);

Example 1.4 Isobutane with temperature and pressure values of 20 C and

Relationship Between Flow Coefficient and Factor

1.2 Flow Coefficient Chart and Flow Curve

Inherent control valve characteristics

Figure 1.2 Flow characteristics of valves.

Valve Flow Rate Calculation Based on Valve Opening Percentage for a

qvp: Valve flow rate linked to the valve opening percentage;

Valve Flow Coefficient Calculation Based on Valve Opening Percentage

Cvvp: Valve flow coefficient linked to the valve opening percentage;

C vvp = Cvmax × vp C v5 = 700 × 0 05 = 35 and

Valve Flow Rate Calculation Based on Valve Opening Percentage for an

qvp = qmax × αvp−1 1 12

Valve Flow Coefficient Calculation Based on Valve Opening Percentage

C vvp = C vmax × αvp − 1 1 13

Cvvp = Cvmax × αvp−1 Cv10 = 640 × 500 1 − 1 = 640 × 0 02957 = 18 93

Tables of estimated Cv values

Figure 1.3 Flow curves for a 38 CL1500 ball valve.

Reduced Full Piston Ball Swing

1/4 8 1.7 1.7 1 0.9 0.8