RAGAGEP Considerations for Relief and

Flare Systems
Georges A. Melhem and Casey Houston
ioMosaic Corporation, Salem, NH 03079; melhem@ioMosaic.com (for correspondence)
Published online 11 August 2016 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/prs.11839

Recent emphasis on Recognized and Generally Accepted National Fire Protection Agency, Center for Chemical Process
Good Engineering Practices (RAGAGEP) increased the scope Safety (CCPS), the Chlorine Institute (CI) guidelines, and
of relief systems risk factors that require evaluation to develop applicable manufacturer’s recommendations.
complete and compliant Pressure Relief and Flare Systems OSHA further clarifies “Appropriate Internal Standards”
documentation. Failure to comply with RAGAGEP ((d)(3)(ii)) and the use of “Shall” and “Should” in RAGAGEP. The appli-
is the most cited element of the Process Safety Management cation of internal standards and the employer’s decision to
requirements. follow (or to not follow) RAGAGEP is expected to be made
This paper discusses how RAGAGEP considerations now to meet or exceed the protective requirements of published
require evaluation and proper documentation of risk factors RAGAGEP.
that are often overlooked including but not limited to: disper- Failure to comply with RAGAGEP ((d)(3)(ii)) is the most
sion analysis, thermal radiation, noise, vibration risk, reac- cited element of the PSM requirements [1].
tion forces and structural supports, metal cold temperatures
due to expansion cooling and two phase flow, hot tempera-
tures due to fire exposure and/or runaway reactions, PRV RISK FACTORS
stability, chemical reaction systems, and loss of high pres- When complying with the “relief system design and design
sure/low pressure interface. basis” Process Safety Information (PSI) element of the PSM
Important RAGAGEP considerations for these additional standard (1910.119(d)(3)(i)(D)), the traditional approach
risk factors are highlighted and discussed. Recommendations included application of ASME and American Petroleum Insti-
are provided on how to best address these factors in the eval- tute (API) standards. Relief devices were sized and installed
uation and documentation of design basis. V C 2016 American
per ASME and API 520. Overpressure scenarios were consid-
Institute of Chemical Engineers Process Saf Prog 36: 18–23, ered and relief requirements quantified in accordance with API
2017 521. The primary goals of the analysis and documentation was
Keywords: recognized and generally accepted good engi- to ensure that relief devices provided adequate capacity, that
neering practices; relief systems; flare systems the inlet pressure drop and built-up backpressure were within
recommended limits, and that the physical installation met the
RAGAGEP
requirements (eliminate restrictions on the inlet line and
Recent updates to Recognized and Generally Accepted
absence of pockets in the discharge line, for example) [2,3].
Good Engineering Practices (RAGAGEP) have expanded the
There are several additional categories that require special
scope of risk factors that require evaluation to develop com-
attentions and are often overlooked when reviewing or
plete Pressure Relief and Flare System (PRFS) documentation.
establishing the design basis for relief and flare systems [1,4].
Additionally, the United States Occupational Safety & Health
These categories are essential for compliance with RAGA-
Administration (OSHA) issued Standard Interpretation
GEP. They include but may not necessarily be limited to:
“RAGAGEP in Process Safety Management Enforcement” on
June 8, 2015 with respect to the Process Safety Management 1. Dispersion analysis and safe discharge location
(PSM) standard 29 CFR 1910.119. a. Thermal radiation
In the Standard Interpretation, OSHA identifies three pri- 2. PRV stability
mary sources of RAGAGEP: 3. Emergency relief systems piping
a. Vibration risk
1. Published and widely adopted codes
2. Published consensus documents 1. Noise
b. Reaction forces and structural support,
3. Published non-consensus documents
c. Temperature excursions
Explicitly included in the three primary sources of RAGA- 1. Cold metal temperatures due to expansion cooling
GEP are industry standards and recommended practices such and two phase flow
as American Society of Mechanical Engineers (ASME), 2. Hot temperatures due to fire exposure and/or run-
away reactions,
Originally presented at the Global Congress of Process Safety,
4. Other factors
AICHE Spring 2016 Meeting, Houston, Texas, April 11–13, 2016. a. Reaction systems
b. Loss of high pressure/low pressure interface and liquid
C 2016 American Institute of Chemical Engineers
V displacement.

18 March 2017 Process Safety Progress (Vol.36, No.1)

Dispersion Analysis and Safe Discharge Location
In addition to the Environmental Protection Agency (EPA)
regulations, such as 40 CFR 68 and 40 CFR 60, Subparts J &
JA in particular, RAGAGEP include evaluating relief system
dispersion characteristics.
API STD 521 recognizes that disposal to atmosphere can
be safe and has been demonstrated. However, in accordance
with §5.8.1 careful attention is required to ensure that atmo-
spheric disposal of relief streams is safe. §5.8.2 and §5.8.3
provide further recommendations on the evaluation of the
flammable and toxic hazards, respectively. Furthermore, API
STD 520 Part II suggests that dispersion analysis may be
used to evaluate the hazards associated with bellows valve
bonnet venting.
Perhaps more importantly, the OSHA National Emphasis
Programs (NEP; Directives CPL 03-00-010 and CPL 03-00-014)
provide guidance for Compliance Safety and Health Officers Figure 1. Dispersion analysis for PSM compliant documenta-
(CSHO) when inspecting refineries and chemical facilities. tion. [Color figure can be viewed in the online issue, which
While it is recognized that the Process Hazard Analysis is available at wileyonlinelibrary.com.]
(PHA) team is expected to identify any hazard associated
with an atmospheric release, the actual dispersion assess-
ment is often a recommendation from the finding.
The process data required for performing dispersion ignition occurs and can present hazardous conditions for
analyses is a readily available result of typical PRFS evalua- personnel within the confines of nearby structures [6]. Stack
tion protocols. With the guidance and methodology provid- downwash is more pronounced when flows out of the relief
ed in API STD 521, evaluating flammable and toxic system discharge piping are low with low velocities. Low
hazards associated with an atmospheric relief should be flow velocities also lead to poor dispersion and more persis-
included in any relief system design and design basis doc- tent dispersing clouds.
umentation [5]. Environmental impact should be considered, especially in
Flame outs during flaring can lead to the dispersion of the case of two-phase discharges where rain-out of liquid
flammable, toxic, or environmental pollutants. During flame droplets is possible.
outs, flares will behave like dispersion stacks. Dispersion
analysis should be conducted to ensure that any delayed Thermal Radiation
ignition or non-ignition do not create hazardous conditions Thermal radiation from flare stacks as well as vent stacks
from thermal radiation (flame jets, vapor cloud fires), explo- and the atmospheric discharge of pressure relief devices
sions, and/or toxicity or environmental impact. should be considered for both personnel and equipment
In installations where relief devices are designed to dis- safety. Where flammable materials are discharged and igni-
charge two-phase materials special care is required to assess tion occurs (immediate or delayed), vent stacks will behave
delayed ignition explosion potential as well as toxicity like flare stacks. See Figure 2.
impact. Many two-phase discharges will include liquid aero- There have been instances where flammable vapors being
sols and these aerosols make the dispersion mixture heavier discharged to the atmosphere ignited due to static or light-
than air and increase the mass of the cloud. Note that two- ning or the presence of nearby ignition sources from other
phase discharges can also occur from oversized relief devices equipment and hot surfaces or open flames. Immediate igni-
in vapor service where two-phase are present in the source tion can cause flame jets from momentum dominated dis-
vessel. charges from pressure relief devices. Delayed ignition can
Liquid rainout represents a challenging problem for two- cause vapor cloud fires and explosions followed by flame
phase discharge directly to atmosphere. Rainout can occur if jets.
the droplets produced during thermal or mechanical break
of the liquid/two-phase discharge from a relief device are PRV Stability
large enough to settle out. If ignited, they can cause wide- Pressure Relief Valve (PRV) stability and the 3% “rule” has
spread fires and behave like flaming rain. been under the limelight over the past several years as the
Hot hydrocarbon releases from relief devices discharging subject of litigation, research, and modeling [7]. PRV instabili-
to atmosphere can condense and create mists. However, ty is very rarely the cause of incidents leading to serious
these mists will almost always remain airborne because they accidents. However, it is important to ensure that when the
will be in the range of 10 microns. The presence of flamma- last line of defense is asked to perform, any relief devices
ble mists or small droplets can substantially increase the operate in a safe and stable manner.
explosion severity of the dispersing material if delayed igni- The 3% “rule” remains a recommendation and not a
tion occurs. requirement in RAGAGEP. The “rule” appears in both ASME
The exit velocity, angle, and density of the discharged Boiler and Pressure Vessel Code Section VIII Division I
material from a relief device along with wind speed and (BPVC-VIII-I) Non-Mandatory Appendix M and as a “should”
direction will determine if the dispersing jet or cloud will in API STD 520 Part II. However, based on OSHA’s RAGA-
touchdown at distances further than the location of the dis- GEP interpretation, any deviation from a “should” item
charge. If the limiting concentration of interest is for odor requires that the measure is “at least as protective, or that the
threshold or toxicity limits, these touchdown downwind dis- published RAGAGEP is not applicable.”
tances can be significant. See Figure 1. API practice formed the foundation of the ASME guid-
Stack downwash can also be an issue with discharges ance. In the past API RP 520 Part II has allowed an
from relief devices that are close or below roof lines. Stack “Engineering Analysis” to demonstrate that non-recoverable
downwash can exacerbate explosion potential if delayed inlet pressure drop (IPD) greater than 3% of the set pressure

Process Safety Progress (Vol.36, No.1) Published on behalf of the AIChE DOI 10.1002/prs March 2017 19
Figure 2. Thermal radiation due to ignition of atmospheric discharge. [Color figure can be viewed in the online issue, which
is available at wileyonlinelibrary.com.]

is safe, but has been silent on a method. Based on significant

research and experience, the 6th edition of API 520, now a
Standard, includes an engineering analysis (§7.3.6) and pro-
vides valuable guidance to the reader [8,9].
Make no mistake, CSHOs are instructed to review PSI
documentation with respect to the 3% “rule” (see CPL 03-00-
010 §X.A.3.1). A prudent relief system design will incorporate
the RAGAGEP available in API STD 520 Part II.
Excessive inlet line pressure loss or backpressure can
cause PRV instability. Stable PRV installations are required by
ASME and API. The governing case for PRV instability may
be very different from the governing case for PRV flow
capacity sizing or reaction forces for structural support.
The 3% inlet pressure loss rule is experience based and is
not sufficient to guarantee PRV stability. The 3% inlet pres-
sure loss rule does not directly address acoustic interaction
between the valve disk and the fluid/piping system. See
Figure 3. The 3% inlet pressure loss rule cannot be met for
some simple installations even with a very short inlet pipe. It Figure 3. Excessive inlet loss or backpressure may cause
has been shown that installations with more than 3% inlet PRV instability. [Color figure can be viewed in the online
pressure loss can operate in a stable manner while installa- issue, which is available at wileyonlinelibrary.com.]
tions with less than 3% inlet pressure loss can operate in an
unstable manner. Stability may need to be assessed and con-
firmed for all credible scenarios and not necessarily just the
one with the highest flow rate or the largest steady state
reaction force. An engineering Analysis as defined by API Emergency Relief Systems Piping
520 II should be considered even when the 3% inlet pres- The emergency relief system (ERS) consists not only of
sure loss rule is met. the relief device (PRV, rupture disk, etc.), but also the
PRV instability can occur due a variety of reasons and attached inlet and discharge piping. In the same vein as PRV
instability may occur with other than the dominant design stability, it is important to ensure that the entire ERS, ade-
scenario. This includes but is not limited to intermittent flow, quately and safely handles the effluent from an overpressure
excessive inlet pressure loss, excessive backpressure, body scenario.
bowl choking, oversized pressure relief device, and condens- In addition to meeting the aforementioned inlet pressure
ing flow [10,11]. drop and built-up backpressure limits, additional relief

20 March 2017 Published on behalf of the AIChE DOI 10.1002/prs Process Safety Progress (Vol.36, No.1)
 viscous liquid or two-phase relief [13]. These simplified tech-
niques, including isothermal assumptions, may not capture
choke points and fluid expansion or contraction that result
in excessive reaction forces, acoustic vibration risk, and low
temperature concerns, which are required to be evaluated by
RAGAGEP [14].

Vibration Risk
The assessment of fatigue failure due to vibration risk is
now required by API-521. Fluid flow in pipes generates tur-
bulent energy (pressure fluctuations). Dominant sources of
turbulence are associated with flow discontinuities in the
piping systems (e.g., partially closed valves, short radius,
mitered bends, tees or expanders). The level of turbulence
intensity is a function of pipe size, fluid density, viscosity,
Figure 4. Vibration risk required per API 521. [Color figure
velocity, and structural support. High noise levels are gener-
can be viewed in the online issue, which is available at
ated by high velocity fluid impingement on the pipe wall,
wileyonlinelibrary.com.]
turbulent mixing, and if the flow is choked, shock waves
downstream of flow restriction, which leads to high frequen-
cy excitation/vibration [15].
system design basis documentation criteria, in accordance Fatigue failure of relief and/or process piping caused by
with RAGAGEP, include: vibration can develop due to the conversion of flow
mechanical energy to noise. Factors that have led to an
1. Vibration Risk, increasing incidence of noise vibration related fatigue failures
2. Reaction Forces, and in piping systems include but are not limited to (a) increas-
3. Temperature Excursions (hot and cold) ing flow rates as a result of debottlenecking which contrib-
utes to higher flow velocities with a correspondingly greater
ASME BPVC-VIII-I Non-Mandatory Appendix M (M-7)
level of turbulent energy, (b) frequent use of thin-walled
requires consideration that discharge line vibration is consid-
ered minimized. Furthermore, API STD 521 §5.5.12 discusses piping which results in higher stress concentrations, particu-
acoustic fatigue and recommends a method to calculate the larly at small bore and branch connections, (c) design of
sound power level (SPL). For SPL greater than 155 dB further process piping systems on the basis of a static analysis with
modeling is recommended to evaluate the risk. Experience little attention paid to vibration induced fatigue, (e) and lack
has shown that the high velocities and pressure drop of emphasis of the issue of vibration in piping design codes.
encountered in many relief system piping designs may fail to Piping vibration is often considered on an ad-hoc or reactive
meet the 155 dB criteria; for these situations, mitigation basis. According to the UK Health and Safety Executive
options are provided in §5.5.12.3. See Figure 4. (HSE), 21% of all piping failures offshore are caused by
Large high-pressure relief devices may generate the thrust fatigue/vibration. Typical systems at risk include large com-
of a jet engine when they open! RAGAGEP requires that the pressor recycle systems and high capacity pressure relief
relief system piping is adequately designed, braced, and sup- depressuring systems. For relief and flare piping, flow
ported to accommodate the static and dynamic loads. Reac- induced turbulence and high frequency acoustic excitations
tion forces in relief system piping are evaluated in are key concerns [16].
accordance with ASME B31.3 and API STD 520 Part II §5.8.2. Excessive AIV/FIV can cause piping and equipment fail-
ASME BPVC-VIII-I Non-Mandatory Appendix M (M-12) rec- ures. Even when complete failure of the piping does not
ommends evaluating the bending moments and stresses for occur, excessive AIV/FIV can cause cracks allowing air inges-
both steady-state flow and transient dynamic loads upon tion/ingress into flare and vent systems creating an explosion
relief device opening. API STD 521 provides guidance on hazard. Use of equivalent pipe length methods for fitting
reaction forces in §5.5.14 with respect to disposal systems losses can lead to inaccurate estimates of sound power levels
and includes reaction forces in the list of recommended min- that are required to assess vibration risk. High superimposed
imum relief system design documentation in §4.7.3.9 item e. constant backpressure can reduce the value of sound power
Another risk factor in ERS piping design is low tempera- level from AIV/FIV sources including those that are from
ture. Most often associated with two-phase or condensing pressure relief systems.
flow in relief device discharge lines, but also experienced
due to all-vapor relief, low temperature may result in embrit- Reaction Forces
tlement. Auto-refrigeration, low-temperatures, and selecting Reaction forces are not always properly calculated and
appropriate discharge piping mechanical design limits are documented. Dynamic loads and dynamic load durations vs.
discussed in API STD 521 §5.5.13.2 with cautions related to structural response time for piping need to be considered.
fluid properties found in §5.2 and §5.2.2.4. OSHA Directive See Figure 5. How long does it take to fill the discharge pipe
CPL 03-00-010 instructs CSHOs to evaluate piping circuits for a relief system for example? Reaction forces can be differ-
and ensure the PSI includes the materials of construction. ent for different pressure relief device types, rupture disks
Although often associated with the Mechanical Integrity PSM vs. modulating or pop action pressure relief valves. The
element, ERS piping materials of construction are easily eval- design case for reaction forces may be different than the
uated by a prudent PRFS designer [12]. governing case for relief device flow capacity. The use of
When evaluating ERS piping, it is critically important to equivalent pipe length methods for fitting losses can lead to
consider the solution methodology. Traditional approaches inaccurate estimates of reaction forces especially for gas and
such as Crane Technical Paper No. 410 and “L/D” hydraulic multiphase flow. Safe upper limits for reaction forces can be
calculations were developed based on Newtonian liquids established by neglecting the attached relief piping to the
and typical steady-state flow rates, which are not necessarily relief device, especially where rupture disks are the primary
appropriate for high-pressure, high-velocity vapor, highly relief device.

Process Safety Progress (Vol.36, No.1) Published on behalf of the AIChE DOI 10.1002/prs March 2017 21
 Figure 7. Temperature excursions can cause metal failures.
Figure 5. Properly calculate and document reaction forces. [Color figure can be viewed in the online issue, which is
[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
available at wileyonlinelibrary.com.]

Figure 6. Low temperature downstream of an expansion. Figure 8. Proper selection of vessel hydrodynamics and
[Color figure can be viewed in the online issue, which is relief phase. [Color figure can be viewed in the online issue,
available at wileyonlinelibrary.com.] which is available at wileyonlinelibrary.com.]

Special considerations must be given to dynamic reaction For gas flow the inside pipe wall metal surface tempera-
forces from pressure relief devices in liquid service or high ture can be higher than the fluid temperature especially
pressure systems. These systems can lead to substantial pres- when the flow velocity is high. See Figure 6. Use of equiva-
sure spikes and reaction forces during fluttering and/or chat- lent pipe length methods for fitting losses can lead to inaccu-
tering than loosen bolts and damage piping with potential rate estimates of pipe metal temperatures for gas flow and
loss of containment. multiphase flow because of how single and multiple choke
points can be miscalculated [12].
Temperature Excursions
Excursions in temperatures, hot or cold, can lead to metal
failure. Cold temperatures caused by expansion cooling can Other Factors
lead to embrittlement. Piping and equipment downstream of
the expansion point, typically the flow limiting area, can be Reactive Chemicals
at fluid temperatures if condensation occurs. Piping and Reaction systems require a higher level of care and design
equipment downstream of the expansion point can collect basis documentation. Relief systems design and evaluation
cold liquids from condensation when it occurs [12,14]. where reactive chemicals are involved is more about chemis-
Hot temperatures caused by fire or flame jet impingement try (desired and undesired) than it is about fluid mechanics.
or runaway reactions can lead to weakening of the metal A key element that is often missed is the correct reaction sys-
strength and subsequent metal failure. See Figure 7. Hot tem- tem characterization under runaway conditions. Scenarios
peratures due to flame jet or pool fire impingement or flame leading to runaway reactions need to be considered and the
radiation can cause vessel failure at the reseat point of the potential for two-phase flow from pressure relief devices
PRV if the fire duration is long enough [17–19]. must be considered as well. See Figure 8.

22 March 2017 Published on behalf of the AIChE DOI 10.1002/prs Process Safety Progress (Vol.36, No.1)
 Two phase flow will almost always occur during runaway 6. G.A. Melhem, An Overview of Explosion Modeling for
reactions. The dispersion and/or effluent handling of reac- Facility Siting, Fall DIERS Users Group Meeting, Reno,
tion products which may include toxics and/or flammable Nevada, 2010.
should also be considered. We note that runaway reactions 7. C. Cunio and G.A. Melhem, A guide to the legal frame-
will often continue in the discharge piping and effluent han- work of the PSM standard for engineers, Process Saf Prog
dling systems [20–24]. 33 (2014).
8. G.A. Melhem, A Simplified Model for PRV Stability Analy-
Loss of High Pressure/Low Pressure Interface and Liquid sis with Examples, AIChE/DIERS Fall Meeting, Houston,
Displacement TX, 2015.
The loss of high pressure (HP)/low pressure (LP) inter- 9. G.A. Melhem, An Overview of the API Simple Force Bal-
face is often poorly evaluated and in some cases missed all ance Method for PRV Stability, Fall DIERS Users Group
together. The loss of HP/LP interface should be evaluated to Meeting, Mobile, Alabama, 2014.
(a) ensure that the downstream equipment can handle the 10. G.A. Melhem, Effective Modeling of Body Bowl Choking,
energy and/or mass accumulation and (b) to also ensure that Fall DIERS Users Group Meeting, Mobile, Alabama, 2014.
the upstream equipment can handle the rapid depressuriza- 11. G.A. Melhem, Modeling the Dynamics for PRV Stability
tion. The loss of the HP/LP interface can occur as a result of for Liquid Systems, 2nd Joint European and US DIERS
automatic controls failure/malfunction, and/or inadvertent Group Meeting, Dusseldorf, Germany, June 10-12, 2015.
valve opening/human error [25,26]. Potential outcomes of 12. G.A. Melhem and D. Gaydos, Properly calculate vessel
the loss of the high pressure/low pressure interface include and piping wall temperatures during depressuring and
but may not necessarily be limited to: relief, Process Saf Prog 34 (2015).
13. G.A. Melhem, Sizing relief systems for high viscosity two-
 Overfilling of the downstream equipment, i.e. as the liq-
phase flow, Chem Eng Prog (2004), 29–34.
uid is displaced from the upstream equipment into the
14. G.A. Melhem, A General Purpose Method for The Estima-
downstream equipment and the resulting high pressure
tion of Multiple Chokes in Gas/Vapor and Two-Phase
associated with liquid overfilling of the downstream
Flow, Safety Relief Valve Benchmarks, 21st DIERS Users
equipment
Group Meeting, Pittsburgh, PA, 1997.
 High temperatures associated with the rapid compression
15. Estimate Vibration Risk for Relief and Process Piping.
of the vapor space of the downstream equipment associ-
Melhem, Georges A. s.l.: AIChE/CCPS, 2013. Proceedings
ated with liquid displacement,
of the 9th Global Congress on Process Safety.
 Multiphase flow associated with increasing liquid level,
16. G.A. Melhem, An Overview of ioVIPERTM for the Assess-
liquid entrainment, and/or high superficial vapor veloci-
ment of Piping Vibration Risk, Spring DIERS Users Group
ties caused by pressure relief device actuation in the
Meeting, Houston, Texas, 2010.
downstream equipment,
17. J. Dunjo and G.A. Melhem, Catastrophic Vessel Failure
 Low temperatures caused by rapid depressurization in the
upstream equipment and ultimately in the downstream due to Fire Exposure: Wall Dynamics & Consequence
equipment as pressure is further reduced when the gas Analysis, Proceedings of the AIChE 11th Global Congress
breaks through to the downstream equipment, on Process Safety, Austin, TX, 2015.
 Possible hydrate formation in the upstream and/or down- 18. G.A. Melhem, R.P. Stickles, and D.R. Eckhardt, Improve
stream equipment the design of fire emergency relief systems, Chem Eng
Prog (1995), 50–57.
19. G.A. Melhem, R.P. Stickles, and D.R. Eckhardt, Improve
CONCLUSION the design of fire emergency relief systems, Plant Saf
Recent industry experience and research has resulted in (1996), 201.
additional guidance in standards, recommended practices, 20. G.A. Melhem, Reactivity screening made easy, Process
and other RAGAGEP. OSHA’s timely interpretation of RAGA- Saf Prog 23 (2004), 99–107.
GEP warrants that PRFS design and design basis calculations 21. G.A. Melhem, “Systematic Evaluation of Chemical Reac-
and documentation expand to meet what were formerly con- tion Hazards,” 2nd International Symposium on Runaway
sidered unessential requirements. The recent emphasis on Reactions, Pressure Relief Design and Effluent Handling,
RAGAGEP raises the bar for relief systems design and evalua- American Institute of Chemical Engineers (AIChE), 1998,
tion studies. pp. 399–443.
22. G.A. Melhem and H.G. Fisher, Reactive Storage Tanks: To
Insulate or Not to Insulate? 30th DIERS Users Group
LITERATURE CITED Meeting, Houston, Texas, 2002.
1. G.A. Melhem, RAGAGEP Requirements for Relief Systems 23. G.A. Melhem and P. Howell, Designing emergency relief
Design Basis and Documentation, AIChE/DIERS Fall systems for runaway reactions, Chem Eng Prog (2005),
Meeting, Houston, TX, 2015. 23–28.
2. G.A. Melhem, A systematic approach to relief and flare 24. G.A. Melhem, H.G. Fisher, and D.A. Shaw, An advanced
systems evaluation, Process Saf Prog 32 (2013). method for the estimation of reaction kinetics, scale-up
3. G.A. Melhem, Relief system’s last line of defense, only and pressure relief design, Process Saf Prog 14 (1995).
line of defense? Process Saf Prog 25 (2006). 25. G.A. Melhem, Model the Dynamics of Heat Exchanger
4. G.A. Melhem, Deal with Controversial Topics in Pressure Tube Failure, Fall DIERS Users Group Meeting, Reno,
Relief Systems, Spring DIERS Users Group Meeting, Nevada, 2010.
Orlando, Florida, 2009. 26. G.A. Melhem, The Anatomy of Liquid Displacement and
5. G.A. Melhem, Dispersion Analysis - A Tutorial, DIERS Vapor Breakthrough, Spring 2014 DIERS Users Group
Users Group Meeting, Philadelphia, PA, 2013. Meeting, Houston, TX, 2014.

Process Safety Progress (Vol.36, No.1) Published on behalf of the AIChE DOI 10.1002/prs March 2017 23