Shariff, Zaini - 2013 - Inherent Risk Assessment Methodology in Preliminary Design Stage A Case Study For Toxic Release
Shariff, Zaini - 2013 - Inherent Risk Assessment Methodology in Preliminary Design Stage A Case Study For Toxic Release
Shariff, Zaini - 2013 - Inherent Risk Assessment Methodology in Preliminary Design Stage A Case Study For Toxic Release
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 October 2011
Received in revised form
2 November 2012
Accepted 17 December 2012
At preliminary design stage, process designers normally lack of information on the risk level from
process plant. An inherently safer process plant could be designed if the information of risk levels could
be known earlier at the preliminary design stage. If the risk level could be determined, there is a possibility to eliminate or reduce the risk by applying the well-known concept: inherent safety principle.
This paper presents a technique to determine the risk levels at preliminary process design stage using
a 2-region risk matrix concept. A model to calculate the severity and likelihood of a toxic release accident
was developed in Microsoft Excel spreadsheet. This model is integrated with process design simulator,
iCON to allow for data transfer during preliminary design stage. 2-region risk matrix is proposed and
used to evaluate the acceptability of the inherent risk based on the severity and likelihood rating. If the
inherent risk level is unacceptable, modication for improvement can be done using the inherent safety
principles. A case study has been carried out to illustrate the benet of applying this newly developed
technique. It was successfully shown that an inherently safer plant could easily be designed by applying
this technique.
2012 Elsevier Ltd. All rights reserved.
Keywords:
Inherent risk
Inherent safety
Risk matrix
Risk assessment
Inherently safer design
1. Introduction
Process safety must be measured and addressed in the whole
life cycle of a process system or a facility (Greenberg & Cramer,
1991). Many guidelines and procedures have been developed,
especially over the last three decades, with respect to risk assessment and safety of chemical process plants, which often include gas
plants, petrochemical plants, reneries, etc., by the industries
themselves, for instance, HAZOP by ICI, DOW Fire and Explosion
Index by DOW Chemicals (Leong & Shariff, 2008). Thus, risk
assessment and safety aspects of process plants should be given
a high priority, and have been further intensied after the Flixborough and the Bhopal incidents.
There are many established methodologies to identify, analyze,
prioritize and manage hazard arising from different stages of
a plant. For example, the CCPS (1996) of the American Institute of
Chemical Engineers (AIChE) has identied a number of hazards
analyses techniques and methodologies, which are deemed suitable for respective plant design stages. The recommendations are
shown in Fig. 1. Taylor (1994) also illustrated the safety program in
another perspective, which is shown in Fig. 2. The program includes
606
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
Phase 1
Phase 2
Phase 3
Conceptual
Engineering
Basic
Engineering
Detailed
Engineering
1. Preliminary
Hazard
Analysis of
Project
1. Relative
Ranking e.g.
DOW Fire &
Explosion &
Chemical
Exposure
Index
Project
Stages
Hazard
Evaluation
1. Hazard and
Operability
(HazOp) Study
2. Failure Mode
and Effect
Analysis
(FMEA)
Phase 4
Phase 5
Equipment
Procurement
and
Construction
Commissioning
1. Check List
Review
2. What-if review
1. Pre-Startup
Safety Review
(PSSR)
Feasibility
Studies
Conceptual
Design
Detail
Design
Safety
Concept
Construction
Commissioning
Construction Safety
Analyses
Operations
As Built
Risk Analyses
Quantitative Risk
Studies
HAZOP
Structural
Reliability Analysis
Economic Risk
Assessment
Follow up
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
Fig. 3. The design impossibility and inherently safer design (Hurme & Rahman, 2005).
607
Denition
608
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
Input from
User
Modification based
on Inherent Safety
(IS) principle
Evaluation of inherent
Yes risk
level using 2-region risk matrix
Risk Acceptable?
NO
YES
User proceed with design
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
609
Table 2
Comparison between QRA and TRIRA.
Criteria
QRA
TRIRA
Stage to be applied
Purpose
Table 4
Likelihood ratings (US Department of Defense, 2000).
Likelihood of occurrence
Very high
High
Moderate
Low
Very low
Unlikely
100
101
102
103
104
105
P
P
P
P
P
P
>
>
>
>
>
>
101
102
103
104
105
106
Table 3
Severity ratings (EPA, 2001; NRC, 2001).
Severity level Observed effect
AEGL-1
AEGL-2
AEGL-3
610
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
Name
Description
Upstream Op
Downstream Op
VapFrac
T [C]
P [kPa]
MoleFlow/Composition
HYDROGEN
AMMONIA
NITROGEN
Total
MassFlow [kg/h]
VolumeFlow [m3/hr]
StdLiqVolumeFlow
[m3/hr]
StdGasVolumeFlow
[SCMD]
Energy [W]
H [kJ/kmol]
S [kJ/kmol-K]
MolecularWeight
MassDensity [kg/m3]
Cp [kJ/kmol-K]
ThermalConductivity
[W/m-K]
Viscosity [Pa-s]
molarV [m3/kmol]
ZFactor
a)
Stream1
Purification_Column.In
1.00
400.0
800.00
Fraction
kgmole/h
240.00
0.2000
840.00
0.7000
120.00
0.1000
1.00
1200.00
18151.00
8371.666
48.787
6.8228E+5
7.896E+6
23689.2
199.923
15.13
2.1682
42.237
0.1118
2.4859E-5
6.976
0.9978
b)
Streams Summary
1
3
SI
C
kPa
kgmole/h
kg/h
m3/hr
W
kJ/kmol
kJ/kmol-K
Stream1
1.0000
400.00
800.00
1200.00
18151.00
8371.67
7896393.17
23689.18
199.92
kg/m3
kJ/kmol-K
kJ/kmol-K
W/m-K
Pa-s
m3/kmol
2.17
42.24
33.80
0.11
0.00
6.98
stage, a slightly modied case study from Shariff and Zaini (2010)
was used. This case study was based on the worst-case accident
scenario on pipeline rupture for purication column in a typical
ammonia production plant, and the schematic diagram is given in
Fig. 7. It was assumed that a maximum concentration will occur at
the center of the puff cloud from the release of pipeline rupture,
and it occurs in 10-min durations (Crowl & Louvar, 2002). The
design intention for this purication column was to have an
acceptable inherent risk condition for the residence at 1200 m from
the plant in the case of pipeline rupture. If the inherent risk value is
UNACCEPTABLE, the inherent safety principles will be used to
reduce the inherent risk to the ACCEPTABLE level.
TRIRA was used to extract data from process design simulator
iCON to MS-Excel in order to determine the severity value of
ammonia release at 1200 m. This is the distance for the 10-min
duration of the accidental released in the case of the worst condition. The design criterion of the plant was decided to be in
ACCEPTABLE region in the case of accidental released by referring
to 2-region risk matrix table as given in Fig. 6. The extracted data
Fig. 9. Data from severity estimation component before lowering the inlet pressure P.
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
611
Fig. 10. Typical fault tree analysis for pipeline rupture from incident likelihood estimation component (Khan & Abbasi, 2000; Khan & Amyotte, 2002; Khan et al., 2001).
overpressure; therefore, the cost and safety concerns during operation of the plant can be minimized. At this pressure, the calculated
amount of ammonia released was 119 ppm as given in Fig. 12 with
the severity rating of AEGL-1 as referring to Table 3. Since the
overpressure is no longer an issue, thus a modied fault tree is
developed as shown in Fig. 13. The incident likelihood was reduced
to 4.5 105/year, and the likelihood rating is VERY LOW as shown
in Table 4. The potential risk after the mitigation is now at the
ACCEPTABLE region as shown in Fig. 14 with the risk factor of A5.
From this case study, it can be concluded that the implementation of inherent safety principle in TRIRA has improved the
process plant safety at preliminary design stage. Besides, other alternatives of inherent safety principle such as minimization, substitution and simplication technique can also be used to assure the
overall design objective is met. However, it is worth noting that any
Fig. 11. Risk matrix before reducing the inlet pressure P and incident likelihood.
612
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
Fig. 12. Data from severity estimation component after lowering the inlet pressure P.
Fig. 13. Modied fault tree analysis for pipeline rupture from incident likelihood estimation component.
Fig. 14. Risk matrix after reducing the inlet pressure P and incident likelihood.
A.M. Shariff, D. Zaini / Journal of Loss Prevention in the Process Industries 26 (2013) 605e613
613