Pressure Relief Valve Inspection Interval

Thiago Trotta, Charles Kashou, and Nancy Faulk

Siemens Energy, Inc, 4615 Southwest Fwy, Suite 900, Houston, TX 77027; nancy.faulk@siemens.com (for correspondence)
Published online 20 April 2017 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/prs.11892

Determining the correct interval for pressure relief valve inspections or shop inspection and overhaul. How often
inspection, testing, and maintenance remains a major chal- should these inspections take place?
lenge for facilities covered by the U.S. Occupational Safety Determining the correct interval for a pressure relief valve
and Health Administration Process Safety Management Stan- (PRV, which is a type of PRD) inspection and testing, as well
dard. To this end, guidance is provided by API Standard as any indicated maintenance, remains a major challenge for
510, Pressure Vessel Inspection Code: In-Service Inspection, facilities covered by the U.S. Occupational Safety and Health
Rating, Repair, and Alteration, by API Recommended Prac- Administration (OSHA) Process Safety Management (PSM)
tice 576, Inspection of Pressure-relieving Devices, by API Standard [2] or by similar regulatory requirements. To this
Standard 520, Sizing, Selection, and Installation of Pressure- end, guidance is provided by API Standards 510 and 520 Part
Relieving Devices, Part II-Installation, and by NB-23, Nation- II, as well as API RP 576 and the National Board Inspection
al Board Inspection Code—Part 2 Inspection. Furthermore, Code—Part 2 [3]–[6]. Furthermore, ASME BPVC, Sections VIII
ASME BPVC, Sections I and VIII, provide general guidelines and I, provide general guidelines for the repair of pressure
for the repair of pressure relief valves. relief valves [7,8]. Note that a definite time interval between
However, the testing and inspection interval listed, up to inspections or tests should be established for every PRV on
ten years, is the maximum time span between shop inspec- operating equipment.
tions and overhaul. Further direction is often requested for Guidance is often requested for determining the proper
determining the proper interval for valves in typical process interval for valves in typical process services, especially in
services, especially in cases of PRV chattering. Recent API cases of pressure relief valve instability. This paper first
STD 520 Part II guidance on performing engineering analy- describes the instabilities which may affect PRV performance
ses for PRV installations, based on service and specific instal- and their causes, then describes how the inspection interval
lation, is included here. may be determined based on risk-based inspections as well
In this paper, a decision-making approach to determining as additional engineering analysis.
these intervals based on a combined understanding of
risk-based inspection, quality assurance, engineering analyses, INSTABILITY PHENOMENA AND CAUSES
and facility experience is presented. The approach provides pro- The principal reason for inspecting and maintaining PRVs
cess operators and managers with additional assistance in is to ensure that they can provide overpressure protection
making this determination. V C 2017 American Institute of Chem-
when needed. Inspections determine the condition of the
ical Engineers Process Saf Prog 37: 37–41, 2018 valve and look for signs of damage or other concerns. In
Keywords: PRV inspection interval; PRV maintenance; addition to corrosion, deposits, and plugging, there are insta-
PRV risk-based inspection bility phenomena which can cause damage to the seating
surface of a PRV and prevent it from performing properly.
INTRODUCTION These are described in the following sections.
Pressure-relieving devices (PRD) are installed on pressure
vessels and boilers in order to relieve excess pressure which Cycling, Fluttering, and Chattering
may result from the overpressure causes described in the The total non-recoverable pressure loss between the pro-
American Petroleum Institute Standard 521 (API STD 521) [1]. tected equipment and the PRV should not exceed 3% of the
These include process upsets, operator error, external fires, PRV set pressure, to prevent valve instability; with the fol-
and other scenarios. Failure to function properly on demand lowing exceptions to the aforementioned criteria [9]:
could result in vessel overpressure, and possibly loss of con-
tainment with subsequent explosions, fires, or toxicity  Thermal relief valves: PRVs designed solely to protect
impacts. There are also consequences associated with leakage against the overpressure caused by liquid hydraulic
of PRDs. To ensure personnel safety as well as protection of expansion, due to ambient or process heating. They are
equipment, facilities and the environment, it is essential that largely oversized for the relief requirement, which could
the PRD be properly designed, installed, inspected regularly, lead to cycling. Moreover, it may result in exceeding the
and maintained in good operating condition. 3% criteria. However, this would only be a concern for
Inspections are used to determine the general physical isolatable equipment; should it happen, the amount of
and operating conditions of the PRD, and assess whether it trapped liquid would be limited. In that case, as the relief
meets the requirements for a given installation and service. device opens and discharges a volume of liquid, it subse-
These inspections are of two types: visual on-stream quently recloses due to loss of built-up pressure on the
inlet of the valve. It would not immediately reopen
because it takes some time for the pressure to build up
C 2017 American Institute of Chemical Engineers
V again. Chattering is not expected for this case.

Process Safety Progress (Vol.37, No.1) March 2018 37

  Remotely sensed pilot operated relief valves: this arrange- Excessive Built-up Backpressure
ment for pilot operated relief valves permits the relief Built up backpressure, due to flow in the discharge pip-
device to sense the pressure directly from the protected ing of the PRV, results in a force on the valve disk, which in
equipment. This prevents the relief device from reclosing turn may reclose the valve. If the backpressure is excessively
due to high, non-recoverable pressure losses. large, the valve will close, only to reopen again after the
flow stops and the force ceases. This is also a cause for chat-
The concern in exceeding the 3% rule is that it may result tering on the relief device [14].
in one of the following instability phenomena which may be
observed on pressure relief valves [10]. Acoustic Interaction
The rapid opening of a relief valve results in an also-
Cycling rapid drop on the pressure upstream of the valve disk. As a
Cycling refers to low frequency opening and reclosing of result, a pressure reduction wave travels through the
a relief device, tending to occur when the relief requirement upstream fluid; when it comes in contact with an upstream
is small compared to its capacity. The pressure in the system reservoir it reflects and travels back, as a compression wave
decreases, then builds back up again periodically. This low towards the valve disk. If this wave reaches the disk before
frequency movement does not usually result in damage to it closes, the valve may discharge in a stable manner or it
the PRV; however, it does impact the ability of the valve to may flutter. If the disk is closed, however, it may lead to a
reseat and may lead to wear over time. cycling or a chattering effect [15].

Retrograde Condensation
Fluttering If the fluid during the relief scenario is supercritical (e.g.,
Flutter occurs while the device is open. Moving parts of pressure relief valves protecting high pressure system on
the valve are rapidly reciprocating, however, the disk does hydrocracker units), the drop in pressure as the valve opens
not contact the seat; instead, it reciprocates around a point, may lead to partial condensation of the fluid. When this con-
creating pressure pulsations. Over time this may lead to the densation occurs, there will be an increased pressure drop
valve becoming stuck open. This phenomenon wears out effect due to the contraction of the fluid from supercritical to
the components of the valve. liquid [16].

Chattering Improper Valve Selection

Chattering is a very high frequency and high amplitude Vapor certified valves will open faster than liquid certified
oscillation; in the worst case, the disk may move between its ones. For services with only liquid relief, a vapor certified
two extremes of completely closed and maximum lift. The valve would open much faster than what is required, as the
frequency of the oscillations may exceed several hundred pressure buildup on a liquid service is also slower, which in
hertz [11]. Different from the cycling, the main consequences turn may lead to reclosing the valve right after it starts reliev-
of this event are possible loss of containment, damage to ing [17].
moving parts of the device, and damage to the equipment to
Oversized Relief Valves
which the valve is connected. Chattering can be very
An oversized valve is one in which the capacity greatly
destructive to the valve seat.
exceeds the relief requirement. This commonly occurs when
A recent method has been proposed in API STD 520 Part
specifying a PRV for equipment which has credible scenari-
II to investigate if a relief device would chatter upon its relief
o(s) with large relief requirements; however, other scenarios
[12]. Experience has shown that a relief device may start or minor upsets may only need to relieve a fraction of that
chattering when the unrecoverable pressure losses are below amount. Conservatism in sizing (oversizing) can lead to on–
the 3% recommended limit; conversely, even if the losses are off cycling [18].
above 3%, the PRV may not chatter or result in failures due
to relieving events. Because the relationship between inlet Body Bowl Choking
pressure drop and chattering is not well understood, detailed An additional cause for PRV instability is given by the
requirements for an engineering analysis are the responsibili- Center for Chemical Process Safety (CCPS) [19]. The velocity
ty of the user. The standard states that the user’s engineering at the outlet flange of a conventional PRV reaches sonic
analysis may be qualitative or quantitative and shall be docu- velocity in cases of larger PRV sizes and higher set pressures.
mented. Readers are cautioned that an engineering analysis This phenomenon is known as body bowl choking. When
shall not be applied if the relief device has a history of this occurs, the pressure in the valve body rises regardless of
chattering. the back pressure at the PRV outlet. This pressure rise in the
PRV body may in turn cause reduced lift and/or unstable
Causes of PRV Instability motion of the valve.
Several sources discuss the main causes for PRV instability
and are listed below. PRV instability is a complex occurrence Inherently Unstable Operation
and often cannot be attributed to a single cause. In addition to the factors outlined above, the stability of
the PRV is influenced by dynamic forces acting to open and
close the valve, and the various factors which affect those
Excessive Inlet Pressure Losses forces. Darby et al. derived and tested a model for the open-
As the relief valve opens, the pressure on the inlet nozzle ing lift dynamic response of a pressure relief valve in gas/
decreases due to the pressure losses resulting from friction in vapor service [20–22]. This response depends on physical
the inlet piping. If the pressure losses are significant, the characteristics of the valve and the valve flow characteristics,
inlet pressure may fall below the reseating value for the PRV, operating conditions, inlet line and discharge piping, and
which will result in the device reclosing. As the pressure on capacitance of the protected vessel. They concluded the
the protected equipment builds up again, the valve will model was capable of reasonable replication of the dynamic
reopen. This effect may lead to chattering on the relief response of valves tested, including instabilities, and further
device [13]. concluded that it is difficult to generalize the influence of

38 March 2018 Published on behalf of the AIChE DOI 10.1002/prs Process Safety Progress (Vol.37, No.1)
any one parameter due to the highly nonlinear nature of the The resulting risk, recommended inspection interval, and
system. next inspection date can then be determined.

DETERMINING INSPECTION INTERVAL

Both the National Board Inspection Code (NBIC) Part 2 PRV RBI Decision-Making
and API RP 576 provide guidance on inspection and test fre- The following decision-making occurred in two steps: (1)
quencies for pressure relief valves, including manual checks, Reviewing applicable sections from API RP 581 [27]; and (2)
pressure tests (pop tests), and service intervals (preventive Reviewing equipment and piping applicable data to deter-
maintenance) [23,24]. Both standards state that the normal mine what items that addressed COF and POF could actually
interval between shop inspection/overhaul is determined by be analyzed.
operating experience and the environment the relief device
is subjected to; therefore, it is expected that a valve with one Risk Matrix
or more of the following would require a shorter interval
The case study was performed using API RP 581 5x5 risk
than a valve in clean, non-fouling or non-corrosive service:
matrix [28].
 Corrosive or fouling service;
 Common discharge header;
Consequence of Failure (COF)
 System critical to plant operation;
Consequence is set up as A – E (X-axis of the Risk
 Discharge particularly detrimental (fire hazard, environ-
Matrix), equivalent to Very Low, Low, Moderate, High, and
mental damage, toxicity);
 Subject to vibration; Very High (respectively). The COF is evaluated by consider-
 Low differential between set and operating pressures; ing each PRV and noting what equipment (and/or piping) or
 Frequent operational upsets; or pieces of equipment are protected by that particular valve.
 Leakage problems. Observe the RBI analysis for the protected equipment and
start with the COF for that equipment. If there is more than
Establishing shop inspection/test history for a valve is one piece of equipment protected by the valve, use the high-
therefore vital to the process of establishing its test interval. est COF to be conservative. If a pump is the piece of equip-
Where several sequential tests of the “as received” valve ment being protected then defer to the piping associated
reflect test results consistent with the cold differential test with the pump and use the COF for the piping.
pressure (CDTP), and no change in service is expected for After determining the starting COF for each PRV, deter-
the valve, an increase in the test interval may be considered, mine the types of overpressure scenarios, the discharge loca-
if allowed by local regulations. On the other hand, if the tion of the relief valve, and if there are multiple relief valves
tests show a history of erratic results, evidence of PRV insta- in parallel with the valve that is being reviewed, for
bility, or significant deviation from CDTP, the test interval example.
should decrease and/or the valve installation should be mod-
ified to improve performance.
Note that both standards suggest inspection frequencies Overpressure Scenarios
where test records and/or inspection histories are not avail- Identify the overpressure scenarios. Utilize API RP 581
able; the NBIC provides these for various equipment types, Table 7.2 (Default Initiating Event Frequencies) and consider
for example, annually for PRV’s in steam service or five years the overpressure scenarios that have an Event Frequency of
for PRV’s in propane or refrigerant service. Both standards 1 per 10 years or less, per the case study. Therefore the fol-
state that if the effects of corrosion, system fluid, or service lowing are considered.
conditions are unknown (for new processes, e.g.), a relative- The following scenarios have an Event Frequency of 1
ly short inspection interval, not to exceed one year, should per year.
be established.
 Runaway Chemical Reaction
RISK-BASED INSPECTION OF PRV’S The following scenarios have an Event Frequency of 1
API RP 576 states, “In API 510, the subsection on per 5 years.
pressure-relieving devices establishes a maximum interval  Tower Pump Around Failure or Reflux Pump Failure
between device inspection or tests of 10 years, unless quali- The following scenarios have an Event frequency of 1
fied by a risk-based inspection (RBI) assessment.” [25]. This per 10 years.
approach is accomplished through RBI method of utilizing  Blocked discharge, without administrative controls in
“Risk” to manage and prioritize an inspection program based place
on API RP 581 [26].  Loss of cooling water utility
The following steps are based on a recent RBI case study,  Thermal/hydraulic expansion relief, without administra-
in which it is assumed that all PRV’s were inspected in accor- tive controls in place
dance with API RP 576, and includes the assumption that all  Control Valve (CV) failure, initiating event is same direc-
PRV’s have been sized, selected and installed per API STD tion as CV normal fail position (i.e., Fail safe)
520 Part II.  Liquid overfilling without administrative controls
The details of Consequence of Failure (COF) and Proba-
bility of Failure (POF) regarding PRV’s will be demonstrated
below. Discharge Location
Adjustments can be made to the PRV consequence based
 COF was analyzed by defining the types of overpressure on the discharge location for each PRV, For example:
scenarios, the discharge location of the relief valve, and if
there were multiple relief valves in parallel.  Atmosphere—Increase the (COF level)
 POF was analyzed by defining the number of overpres-  Flare system—Decrease the (COF level)
sure scenarios, fluid service severity, relief valve type, if a  Closed Process—Increase the (COF level)
rupture disk existed upstream of the relief valve, and  Closed Drain (underground piping, or liquid portion of
inspection history. the flare system)—Decrease the (COF level)

Process Safety Progress (Vol.37, No.1) Published on behalf of the AIChE DOI 10.1002/prs March 2018 39
Probability of Failure (POF) based on the operating company’s risk-ranking acceptance
Probability (Likelihood) is set up as 1 – 5 (Y-axis of the criteria.
Risk Matrix) equating to Highly Unlikely, Unlikely, Possible, With these defined Consequence and Probability Num-
Somewhat likely, and Very Likely (respectively). All PRVs bers and Letters, there are resulting risks of High Risk, Medi-
started with a likelihood category of “3” or “Possible.” um High Risk, Medium Risk and Low Risk.
Observe the number of overpressure scenarios, fluid service Note: The distribution of the number of PRV’s are placed
severity, relief valve type, and if a rupture disk existed in the Risk Matrix proper designations.
upstream of the relief valve, and the inspection history. As a result, the summary of the consequent risk distribu-
tion shall be as follows:
Number of Overpressure Scenarios  High Risk 5 (Total numbers of PRV’s assigned in the high
Observe the Estimated Demand Rate (from applicable sce- risk matrix section)
narios in Table 7.2—Default Initiating Event Frequencies)  Medium High Risk 5 (Total numbers of PRV’s assigned in
total number of scenarios (Total for all equipment protected). the medium high risk matrix section)
The POF should be adjusted based on the number of over-  Medium Risk 5 (Total numbers of PRV’s assigned in the
pressure scenarios. medium risk matrix section)
 Low Risk 5 (Total numbers of PRV’s assigned in the low
Fluid Service Severity risk matrix section)
There are two different types of PRV Service Severity as
described in API RP 581 [29]. One is based on “Fail to Open” Based on the resulting risk information, the next inspec-
cases and the other is based on “Leak” only cases. Determine tion intervals should be focusing on high risk PRV’s while
the severity level for each PRV based on the descriptions in the inspection frequency regarding the low risk PRV’s can be
API RP 581 Categories of PRD Service Severity (Fail Open reduced or even greatly reduced compared to the usual
Case) Table 7.4 and (Leak Case) Table 7.11. The probability inspection program. This results in considerable inspection
will be adjusted relative to fluid service based on the deter- and general cost reduction. Therefore, implementing RBI
mined severity level for the fluid service. If the device pro- provides a dependable methodology of determining the best
tected multiple pieces of equipment, note the service for combination of “inspection frequencies and inspection
each piece of equipment and note the severity for each one. scopes.”
Use the worst severity to make the adjustments.
ROLE OF PRESSURE RELIEF ANALYSIS
Valve Type Pressure relief analysis (PRA) studies and periodic revali-
Because the Probability of Failure on Demand vs. Time in dations provide a comprehensive database of potential sour-
Service is based on Weibull parameters which were deter- ces of overpressure and can be used to develop a concern
mined using industry failure rate data, using the valve type list summarizing any deviations from industry and operating
as a variable for probability of failure was considered unrea- company standards. During a pressure relief analysis on a
sonable in this case study (See API RP 581 Fail to Open unit, it is possible to identify and quantify:
Table 7.5/Figure 7.2 and Leakage Table 7.12/Figure 7.3).
 Non-recoverable pressure losses on the inlet piping to the
relief valves;
Rupture Disk Upstream  Built-up backpressures on the outlet piping;
If there was a rupture disk upstream of the PRV then the  Undersized overpressure scenarios;
probability should be slightly lowered.  Oversized overpressure scenarios;
 Acoustic interaction on the piping;
Historical Inspection Data  Improper valve selection;
Review the case study inspection data and determined  Improperly installed relief valves;
which types of relief device inspections have been per-  Liquid static head on piping that could prevent relief
formed. Determine the actual inspection interval for the last valves from opening prior to the protected equipment
two inspections by comparing the date for the last shop exceeding its MAWP plus allowable accumulation;
inspection/overhaul and the second to last shop inspection/
overhaul. If there was no second to last shop inspection/ Obtaining all the above mentioned data, which are out-
overhaul, then the other option is to use the date three years puts of PRA, is beneficial to identify relief valves that could
prior to the last shop inspection/overhaul. present instability (even if they never presented before) and
Check if a visual On-stream inspection had been per- to judge in a more assertive approach which actions can be
formed. Evaluate if the last shop pre-test “Passed” or taken to prevent instability for that devices.
“Failed.”
Grade the last inspection data based on API RP 581 Table CONCLUSION
7.7—Inspection and Testing Effectiveness. A 5 Highly Effec- As previously mentioned, the 10-year interval should be
tive to D 5 Ineffective. Record the next inspection due date. considered as a maximum time between inspections. A
Determine if the next inspection was overdue relative to risk-based approach, as illustrated on this paper, is able to
the date of the current analysis. If the next inspection is improve on the inspection management of PRVs by provid-
overdue, then the probability of failure should be increased. ing a dependable methodology of determining the best com-
bination of “inspection frequencies and inspection scopes.”
Resulting Risk
By now the PRV has a defined COF (shown as Conse- DISCLAIMER
quence on API RP 581 5 x 5 Risk Matrix) A – E, Very Low, The information contained in this document represents
Low, Moderate, High and Very High (respectively). It also the current view of the authors at the time of publication.
has a defined POF (shown as Probability on API RP 581 5 x Process safety management is complex and this document
5 Risk Matrix) 1 – 5, Highly Unlikely, Unlikely, Possible, cannot embody all possible scenarios or solutions related to
Somewhat likely, and Very Likely (respectively). These are compliance. This document is for informational purposes

40 March 2018 Published on behalf of the AIChE DOI 10.1002/prs Process Safety Progress (Vol.37, No.1)
only. Siemens makes no warranties, express or implied, in 15. American Petroleum Institute, API Standard 520, Sizing,
this paper. Selection, and Installation of Pressure-Relieving Devices,
Part II-Installation, 6th Edition, 2015, §7.2.4.
16. American Petroleum Institute, API Standard 520, Sizing,
Literature Cited Selection, and Installation of Pressure-Relieving Devices,
1. American Petroleum Institute, API Standard 521, Part II-Installation, 6th Edition, 2015, §7.2.5.
Pressure-relieving and Depressuring Systems, 6th Edition, 17. American Petroleum Institute, API Standard 520, Sizing,
2014 Table 1. Selection, and Installation of Pressure-Relieving Devices,
2. U.S. Department of Labor Occupational Safety and Health Part II-Installation, 6th Edition, 2015, §7.2.6.
Administration, 29 CFR 1910.119 – Process Safety Man- 18. Design Institute for Emergency Relief Systems (DIERS),
agement of Highly Hazardous Chemicals, 1992. Emergency Relief System Design Using DIERS Technolo-
3. American Petroleum Institute, API Standard 510, Pressure gy, Wiley, New York, NY, 1992, pp. 91–93
Vessel Inspection Code: In-Service Inspection, Rating, 19. Center for Chemical Process Safety (CCPS), Effect of Body
Repair, and Alteration, 10th Edition, 2014. Bowl Choking on Pressure Relief Valve Stability, AIChE Acad-
4. American Petroleum Institute, API Standard 520, Sizing, emy Webinar, Available at http://www.aiche.org/acade-
Selection, and Installation of Pressure-Relieving Devices, my/videos/conference-presentations/effect-body-bowl-
Part II-Installation, 6th Edition, 2015. choking-on-pressure-relief-valve-stability, Accessed 25
5. American Petroleum Institute, API Recommended Practice January 2017
576, Inspection of Pressure-relieving Devices, 3rd Edition, 20. R. Darby, The dynamic response of pressure relief valves
2009. in vapor or gas service part I: Mathematical model, J Loss
6. National Board of Boiler and Pressure Vessel Inspectors Prevent Process Ind 26 (2013), 1262–1268.
(NBBI) NB 23-2015, National Board Inspection Code 21. A. Aldeeb, R. Darby, and S. Arndt, The dynamic response
(NBIC) – Part 2 Inspection, 2015 Edition. of pressure relief valves in vapor or gas service part II:
7. American Society of Mechanical Engineers, ASME Boiler Experimental investigation, J Loss Prevent Process Ind 31
and Pressure Vessel Code, Section VIII – Division 1, 2015 (2014), 127–132.
Edition. 22. R. Darby, and A. Aldeeb, The dynamic response of pres-
8. American Society of Mechanical Engineers, ASME Boiler sure relief valves in vapor or gas service part III: Model
and Pressure Vessel Code, Section I, 2015 Edition. validation, J Loss Prevent Process Ind 31 (2014), 133–141.
9. American Petroleum Institute, API Standard 520, Sizing, 23. National Board of Boiler and Pressure Vessel Inspectors
Selection, and Installation of Pressure-Relieving Devices, (NBBI) NB 23-2015, National Board Inspection Code
Part II-Installation, 6th Edition, 2015, §7.3.4. (NBIC) Part 2 Inspection, 2015 Edition, §2.5.8.
10. American Petroleum Institute, API Standard 520, Sizing, 24. American Petroleum Institute, API Recommended Practice
Selection, and Installation of Pressure-Relieving Devices, 576, Inspection of Pressure-relieving Devices, 3rd Edition,
Part II-Installation, 6th Edition, 2015, §7.1–7.2. 2009, §6.4.2.
11. A. Borg and S. Jakobsson, On the Stability of Pressure 25. American Petroleum Institute, API Recommended Practice
Relief Valves: A numerical study using CFD, Master’s The- 576, Inspection of Pressure-relieving Devices, 3rd Edition,
sis, Chalmers University of Technology, 2014, Available at 2009, §6.4.1.3.
http://publications.lib.chalmers.se/records/fulltext/ 26. American Petroleum Institute, API Recommended Practice
211055/211055.pdf, Accessed 25 January 2017. 581, Risk-Based Inspection Technology, 2nd Edition, 2008.
12. American Petroleum Institute, API Standard 520, Sizing, 27. American Petroleum Institute, API Recommended Practice
Selection, and Installation of Pressure-Relieving Devices, 581, Risk-Based Inspection Technology, 2nd Edition,
Part II-Installation, 6th Edition, 2015, §7.3.6. 2008, Part I §7.
13. American Petroleum Institute, API Standard 520, Sizing, 28. American Petroleum Institute, API Recommended Practice
Selection, and Installation of Pressure-Relieving Devices, 581, Risk-Based Inspection Technology, 2nd Edition,
Part II-Installation, 6th Edition, 2015, §7.2.2. 2008, Part I Figure 4.2.
14. American Petroleum Institute, API Standard 520, Sizing, 29. American Petroleum Institute, API Recommended Practice
Selection, and Installation of Pressure-Relieving Devices, 581, Risk-Based Inspection Technology, 2nd Edition,
Part II-Installation, 6th Edition, 2015, §7.2.3. 2008, Part I §7.6.

Process Safety Progress (Vol.37, No.1) Published on behalf of the AIChE DOI 10.1002/prs March 2018 41