- Email: [email protected]
A strategy for the generation of robust accident scenarios in quantitative risk assessment using multi-component analysis
European Symposium on Computer Aided Process Engineering - 10 S. Pierucci (Editor) 9 2000 Elsevier Science B.V. All rights reserved.
733
A S T R A T E G Y F O R THE G E N E R A T I O N OF R O B U S T A C C I D E N T S C E N A R I O S IN QUANTITATIVE RISK A S S E S S M E N T USING M U L T I - C O M P O N E N T ANALYSIS Ku Hwoi Kim, Ji Ho Song, Dongil Shin and En Sup Yoon School of Chemical Engineering, Seoul National University Shillim-Dong, Kwanak-Gu, Seoul 151-742, Korea A strategy for producing robust accident scenarios in quantitative risk assessment, which is performed in the process design or operation steps, is proposed. Most governments over the world require industrial companies to submit proper emergency plans through the off-site risk assessment. However, there have been no systematic approaches or criteria for generating reasonable virtual accident scenarios, and it is very difficult to get the unified or coherent assessment result. To get over these shortcomings, this study proposes a strategy for analyzing process elements and then selecting and generating robust accident scenarios that simulate the worst accidents most likely to happen and should be foremost considered. The result through the proposed analysis enhances the reliability of the produced risk scenario and prevents the risks from being overestimated. The obtained result should be more helpful in the proper process design and emergency planning.
1. INTRODUCTION Chemical industries are composed of complicated processes with many recycle streams of energy and materials, regulated by environmental and safety considerations. As concerns about protection from accidents and environmental problems increase, we need better process technology and safety management systems that can deal with process safety more efficiently in real time [1]. Worldwide chemical processes are in need of off-site risk assessment as well as the on-site one. Most governments over the world require industrial companies to submit proper emergency plans through the off-site risk assessment; Korea is also preparing for executing IRMS (Integrated Risk Management System) along with PSM (Process Safety Management) and SMS (Safety Management System). These kinds of analyses are helpful in determining appropriate safety devices, capacity of safety facilities, and the minimum distance from residential areas. Therefore, more and more petroleum and oil/gas companies are adopting these technologies to improve the safety and productivity. However, there have been no systematic approaches or criteria for generating reasonable virtual accident scenarios, and it is very difficult to get the unified or coherent assessment result. Robust accident scenario selection is considered essential for the success of offsite consequence analysis because analysis results may vary significantly according to the selection of scenarios.
2. RISK MANAGEMENT PROGRAM The off-site consequence analysis technology is a world-widely accepted method, which aims to establish an appropriate emergency planning for the off-site area. But due to the limitations of conventional hazard analysis techniques, consequence analysis has the same drawback; analysis results are different according to the individual analyst's view. In the United States, EPA (Environmental Protection Agency) has announced 'RM (Risk Management) Program' in 1996,
734 and every industrial company is supposed to submit reports about the off-site consequence analysis of the release of specific substances [2, 3]. The most notable feature of RM Program is that it asks to carry out consequence analysis for a WCS (Worst Case Scenario) and an ACS (Alternative Case Scenario) for each hazardous material. The worst case scenario is defined as the release of the largest quantity of a regulated substance from a vessel or process line failure that results in the greatest distance to an endpoint. The endpoint is the concentration, explosion overpressure or radiant heat at which serious human health effects or environmental damage could occur from exposure to a release of that substance. Usually, for regulated substances, the release distance that impacts off-site area is fairly long. Parameters required in modeling the scenarios for WCS and ACS are similar. 2.1. Drawbacks in the establishment of accident scenarios The most important part in a consequence analysis program is to determine accident scenarios that are likely to occur in a process. Generally, there are three kinds of methods in deciding accident scenarios: qualitative methods, quantitative methods, and methods using past accident data. HAZOP study and What-If analysis are examples of qualitative methods. Event Tree Analysis (ETA) is an example of the quantitative method [4-8]. Accident data of five years in a similar process are analyzed and used as an imaginary scenario in past-accident data based methods. Each method has its own fortes and drawbacks, and it's difficult to apply these methods in real consequence analysis, because there is no systematic selection criteria for the scenarios among the above methodologies. In qualitative methods, only kinds of accident results are presented and they cannot be applied in ranking or selecting accident scenarios. In quantitative methods like ETA, results change according to the selection of the initial event. In the RM Program, WCS is calculated only using the maximum capacity (i.e. not using the state information of the process or operational condition), and the result tends to be overestimated than the real one. Therefore, to overcome these drawbacks, a method based on the qualitative result, considering the process condition, material property and equipment behavior, etc., and can apply the result in a quantitative manner is required. The result of off-site consequence estimation is presented as toxicity level, heat radiation or overpressure, and used as the basis of emergency planning [9]. When an accident occurs, we can analyze the unit or equipment that affects the surrounding area. Existing methods for calculating the risk depend heavily on the individual analyst's view in generating accident scenarios; the calculation represents so variety of results. Sometimes, heavier risk in a process is overlooked because it would not consider the status of the process. Therefore, we should consider chemical property, meteorological condition and the equipment behavior to achieve calculating the accurate effect to the surrounding area in the off-site consequence estimation.
3. REASONING A L G O R I T H M FOR THE ACCIDENT SCENARIO SELECTION In this study, we propose a new reasoning algorithm through process partition and process component analysis to improve the reliability of accident scenario selection. Process elements are analyzed and then the proposed strategy selects and generates the robust accident scenario of a worst case that is most likely to happen and should be foremost considered. (This concept is being extended using statistical methods, but that part is not included in this paper.) The scenario reasoning algorithm consists of the following steps: macro decomposition, micro decomposition, equipment analysis, scenario selection, and the consequence analysis (see Figs. 1 and 2). In the macro decomposition, process units are selected according to their functions and the meteorological condition around the area. For the decomposition, the chemical plant is classified into the feed system, reaction system, separation system, storage system, and utility system.
735
STEP I : MACRO DECOMPOSITION [ I STEP II : MICRO DECOMPOSITION I
+
] STEP III : EQUIPMENT ANALYSIS I
Topography~1 Equipment1I Componentb RELIABILITY ANALYSIS 4,
RISK
InferenceEnginej.[ Library I
"~"~
No
Yes
[ STEP IV : SCENARIO SELECTION I
+
ScenadoSelec~on
I
Fig. 1. Structure of the scenario selection mechanism
STEPv : EFFECTANALYSIS
I
Fig. 2. Flow diagram of the algorithm
Meteorological characteristics and the surrounding condition are also considered: the main unit is defined, and meteorological characteristics and the topography of the selected unit are considered. The procedure for selecting units through functional decomposition is represented in Fig. 3. We can conclude that unit 2, unit 4, and unit 6 give potential hazards resulting to damages in the surrounding area. Considering unit 2 and unit 4 are utility units with minor risk compared with the main unit, we can determine that unit 2 and unit 4 from the macro decomposition may cause major damage to the surrounding area. For the second step, we propose ESM (Equipment Screening Method) analyzing the process condition and selecting the process equipment with higher priority risk ranking. Equipment characteristics such as material property, operating condition, flow-rate, capacity, safety devices, age and accident history are analyzed using ESM, which is a sequential reasoning method. In case of material property, we use NFPA (National Fire Protection Association) code to confirm the flammability and toxicity; the criterion of this property is more than 3 NFPA rating. In the next stage, we consider whether the equipment is operated in high pressure or temperature, and equipments with high flow-rate or capacity are determined. In the fourth stage, we decide whether the selected equipments have safety devices. In the final stage, we consider the age and accident history for individual equipment using the sequential screening method.
w,o
r Operating~ Condition ~
t c...-~ ]
?, Utility
v
Fig. 3. Unit selection using process partition
Priority 1
Priority 2
Fig. 4. Reasoning scheme for ES Method
736 Table ! Example of failure modes for equipment behavior Equipment Valve Pump Fail on Failure mode Open Transfer off Close Seal leak/rupture Rupture Pump casing leak/rupture Leak
Heat exchanger Leak/rupture (tube to shell) Leak/rupture (shell to tube) Plugged Fouling
The analyzed process elements are ranked and risk grades are determined. According to the grades, risk assessment is performed. Fig. 4 shows all five stages of the reasoning algorithm. In the equipment analysis, the effect estimation for the selected equipment in the step II (micro decomposition) is accomplished: equipment with high severity is researched to find a detailed accident scenario. We use effect analysis method for the failure mode of the selected equipment to identify single equipment failure modes and each failure mode's potential effect on the system and the plant. This mode describes how equipment fails and is determined by the system's response to the equipment failure (see Table 1). In the scenario selection, we infer possible effects depending on the failure mode of the equipment. Possible scenarios for each failure mode are so variable that risk rankings are assigned according to the potential hazard of material and the magnitude of abnormal situation. Table 2 shows the failure mode and scenario selection procedures.
4. CASE STUDY: LPG STORAGE FACILITY
This system is one of typical LPG transportation and storage facilities, including propane underground cavern, propane coalescer, propane dryer, and a propane storage tank, illustrated in Fig.5.
Table 2 Example of scenario selection for a failure mode .
.
.
.
Identi.fication
Mode
Effect
Valve A on the chlorine line
Fail open
Excess flow of chlorine to the heater May cause a high level in the cleaning bed
Fail closed
No flow of chlorine to the cleaning bed Excess water flow to the cleaning bed
Material Chlorine
Risk Ranking C
Excess chlorine and water Water
D
Minor
Water
Minor
737
: ...... {~..i
P
t
minor hazard
[~'orZ
............................................
.
~t
:
}, Propane tans
r
Priority 2 Priority 1 Fig. 6. ESM application for LPG facility
Fig. 5. PFD for LPG storage facility
4.1. Accident scenario selection using the proposed algorithm In step I (macro decomposition), the entire process is decomposed into unit processes and process units are selected according to their functions and the meteorological condition around the area. Step II is the micro decomposition step. Through this step, ESM is applied to the 5 valves, 3 pumps and 1 heat exchanger, which have been selected as the most influential process components to the off-site area when an accident occurs. The result is shown in Figure 6; an accident due to pump A and valve C is the most hazardous one. In step III (equipment analysis), the analysis is performed to the process elements chosen in the step II, and the elements are ranked and risk grades are determined. Table 3 shows a part of the analysis. In step IV (scenario selection), according to the result of step III, propane releases due to the rupture of valve E or the open of valve E caused by the failure and rupture of pump sealing are selected as the most suitable accident scenarios. Step V is effect analysis: a consequence analysis is performed using the scenarios chosen in the step IV. An analysis program is being developed as part of this research, and that program may be used in this step (See Fig. 7 and Fig. 8 for the screenshot of implemented system). An analysis result can also be obtained using one of commercial software packages available today. Table 3 Scenario selection for LPG facility (in case of Pump A)
Identtfication Pump A on the liquefied
Mode Fail open
Effect Excess flow of propane to
Material
Risk Ranking
Propane
Minor
the propane underground
propane line
cavern Fail transfer
No flow of propane to the
off
propane underground
Minor
cavem Seal rupture
Large release of propane to the surrounding area
Propane
A
738
Fig. 7. Screenshot of main menu and input data
Fig. 8. Result of equipment analysis
5. CONCLUSION A strategy for producing robust accident scenarios in the quantitative risk analysis, which is performed in the process design or operation steps, has been proposed and tested to one of the LPG facilities and the DAP process (not shown in this paper). The obtained result of the systematic analysis enhances the reliability of the generated risk scenarios and prevents the risks from being overestimated; the result should be more helpful in the proper process design and emergency planning. The proposed strategy is being implemented as a part of the government-supported, quantitative process hazard analysis system, and expected to be successfully applied to most of mandated off-site consequence analyses in Korea. REFERENCES
1. Greenberg HR and Cramer JJ, Risk Assessment and Risk Management for the Chemical Process Industry, Van Nostrand Reinhold, 1991. 2. EPA, RMP Offsite Consequence Analysis Guidance, EPA, 1996. 3. Murphy JF and Zimmermann KA, "Making RMP hazard Assessment Meaningful", Process Safety Progress, 17(4), 238-242, 1998. 4. Hoist S, Hjertager BH, Solberg T and Malo KA, "Properties of Simulated Gas Explosion of Interest to the Design Process", Process Safety Progress, 17(4), 278-287, 1998. 5. Khan FI and Abbasi SA, "Techniques and Methodologies for Risk Analysis in Chemical Process Industries", Journal of Loss Prevention in the Process Industries, 11(4), 261-277, 1998. 6. CCPS, Vapor Cloud Source Dispersion Models, CCPS of the AIChE, 1989. 7. CCPS, Guidelinesfor Evaluating the Characteristics of Vapor Cloud Explosions, Flash Fires, and BLEVEs, CCPS of the AIChE, 1994. 8. CCPS, Guidelinesfor Hazard Evaluation Procedures, 2 nd Ed., CCPS of the AIChE, 1992. 9. Amaldos J, Casal J, Montiel H, Sanchez-Carricondo M and Vilchez JA, "Design of a Computer Tool for the Consequences of Accidental Natural Gas Releases in Distribution Pipes", Journal of Loss Prevention in the Process Industries, 11(2) ,135-148, 1998.
733
A S T R A T E G Y F O R THE G E N E R A T I O N OF R O B U S T A C C I D E N T S C E N A R I O S IN QUANTITATIVE RISK A S S E S S M E N T USING M U L T I - C O M P O N E N T ANALYSIS Ku Hwoi Kim, Ji Ho Song, Dongil Shin and En Sup Yoon School of Chemical Engineering, Seoul National University Shillim-Dong, Kwanak-Gu, Seoul 151-742, Korea A strategy for producing robust accident scenarios in quantitative risk assessment, which is performed in the process design or operation steps, is proposed. Most governments over the world require industrial companies to submit proper emergency plans through the off-site risk assessment. However, there have been no systematic approaches or criteria for generating reasonable virtual accident scenarios, and it is very difficult to get the unified or coherent assessment result. To get over these shortcomings, this study proposes a strategy for analyzing process elements and then selecting and generating robust accident scenarios that simulate the worst accidents most likely to happen and should be foremost considered. The result through the proposed analysis enhances the reliability of the produced risk scenario and prevents the risks from being overestimated. The obtained result should be more helpful in the proper process design and emergency planning.
1. INTRODUCTION Chemical industries are composed of complicated processes with many recycle streams of energy and materials, regulated by environmental and safety considerations. As concerns about protection from accidents and environmental problems increase, we need better process technology and safety management systems that can deal with process safety more efficiently in real time [1]. Worldwide chemical processes are in need of off-site risk assessment as well as the on-site one. Most governments over the world require industrial companies to submit proper emergency plans through the off-site risk assessment; Korea is also preparing for executing IRMS (Integrated Risk Management System) along with PSM (Process Safety Management) and SMS (Safety Management System). These kinds of analyses are helpful in determining appropriate safety devices, capacity of safety facilities, and the minimum distance from residential areas. Therefore, more and more petroleum and oil/gas companies are adopting these technologies to improve the safety and productivity. However, there have been no systematic approaches or criteria for generating reasonable virtual accident scenarios, and it is very difficult to get the unified or coherent assessment result. Robust accident scenario selection is considered essential for the success of offsite consequence analysis because analysis results may vary significantly according to the selection of scenarios.
2. RISK MANAGEMENT PROGRAM The off-site consequence analysis technology is a world-widely accepted method, which aims to establish an appropriate emergency planning for the off-site area. But due to the limitations of conventional hazard analysis techniques, consequence analysis has the same drawback; analysis results are different according to the individual analyst's view. In the United States, EPA (Environmental Protection Agency) has announced 'RM (Risk Management) Program' in 1996,
734 and every industrial company is supposed to submit reports about the off-site consequence analysis of the release of specific substances [2, 3]. The most notable feature of RM Program is that it asks to carry out consequence analysis for a WCS (Worst Case Scenario) and an ACS (Alternative Case Scenario) for each hazardous material. The worst case scenario is defined as the release of the largest quantity of a regulated substance from a vessel or process line failure that results in the greatest distance to an endpoint. The endpoint is the concentration, explosion overpressure or radiant heat at which serious human health effects or environmental damage could occur from exposure to a release of that substance. Usually, for regulated substances, the release distance that impacts off-site area is fairly long. Parameters required in modeling the scenarios for WCS and ACS are similar. 2.1. Drawbacks in the establishment of accident scenarios The most important part in a consequence analysis program is to determine accident scenarios that are likely to occur in a process. Generally, there are three kinds of methods in deciding accident scenarios: qualitative methods, quantitative methods, and methods using past accident data. HAZOP study and What-If analysis are examples of qualitative methods. Event Tree Analysis (ETA) is an example of the quantitative method [4-8]. Accident data of five years in a similar process are analyzed and used as an imaginary scenario in past-accident data based methods. Each method has its own fortes and drawbacks, and it's difficult to apply these methods in real consequence analysis, because there is no systematic selection criteria for the scenarios among the above methodologies. In qualitative methods, only kinds of accident results are presented and they cannot be applied in ranking or selecting accident scenarios. In quantitative methods like ETA, results change according to the selection of the initial event. In the RM Program, WCS is calculated only using the maximum capacity (i.e. not using the state information of the process or operational condition), and the result tends to be overestimated than the real one. Therefore, to overcome these drawbacks, a method based on the qualitative result, considering the process condition, material property and equipment behavior, etc., and can apply the result in a quantitative manner is required. The result of off-site consequence estimation is presented as toxicity level, heat radiation or overpressure, and used as the basis of emergency planning [9]. When an accident occurs, we can analyze the unit or equipment that affects the surrounding area. Existing methods for calculating the risk depend heavily on the individual analyst's view in generating accident scenarios; the calculation represents so variety of results. Sometimes, heavier risk in a process is overlooked because it would not consider the status of the process. Therefore, we should consider chemical property, meteorological condition and the equipment behavior to achieve calculating the accurate effect to the surrounding area in the off-site consequence estimation.
3. REASONING A L G O R I T H M FOR THE ACCIDENT SCENARIO SELECTION In this study, we propose a new reasoning algorithm through process partition and process component analysis to improve the reliability of accident scenario selection. Process elements are analyzed and then the proposed strategy selects and generates the robust accident scenario of a worst case that is most likely to happen and should be foremost considered. (This concept is being extended using statistical methods, but that part is not included in this paper.) The scenario reasoning algorithm consists of the following steps: macro decomposition, micro decomposition, equipment analysis, scenario selection, and the consequence analysis (see Figs. 1 and 2). In the macro decomposition, process units are selected according to their functions and the meteorological condition around the area. For the decomposition, the chemical plant is classified into the feed system, reaction system, separation system, storage system, and utility system.
735
STEP I : MACRO DECOMPOSITION [ I STEP II : MICRO DECOMPOSITION I
+
] STEP III : EQUIPMENT ANALYSIS I
Topography~1 Equipment1I Componentb RELIABILITY ANALYSIS 4,
RISK
InferenceEnginej.[ Library I
"~"~
No
Yes
[ STEP IV : SCENARIO SELECTION I
+
ScenadoSelec~on
I
Fig. 1. Structure of the scenario selection mechanism
STEPv : EFFECTANALYSIS
I
Fig. 2. Flow diagram of the algorithm
Meteorological characteristics and the surrounding condition are also considered: the main unit is defined, and meteorological characteristics and the topography of the selected unit are considered. The procedure for selecting units through functional decomposition is represented in Fig. 3. We can conclude that unit 2, unit 4, and unit 6 give potential hazards resulting to damages in the surrounding area. Considering unit 2 and unit 4 are utility units with minor risk compared with the main unit, we can determine that unit 2 and unit 4 from the macro decomposition may cause major damage to the surrounding area. For the second step, we propose ESM (Equipment Screening Method) analyzing the process condition and selecting the process equipment with higher priority risk ranking. Equipment characteristics such as material property, operating condition, flow-rate, capacity, safety devices, age and accident history are analyzed using ESM, which is a sequential reasoning method. In case of material property, we use NFPA (National Fire Protection Association) code to confirm the flammability and toxicity; the criterion of this property is more than 3 NFPA rating. In the next stage, we consider whether the equipment is operated in high pressure or temperature, and equipments with high flow-rate or capacity are determined. In the fourth stage, we decide whether the selected equipments have safety devices. In the final stage, we consider the age and accident history for individual equipment using the sequential screening method.
w,o
r Operating~ Condition ~
t c...-~ ]
?, Utility
v
Fig. 3. Unit selection using process partition
Priority 1
Priority 2
Fig. 4. Reasoning scheme for ES Method
736 Table ! Example of failure modes for equipment behavior Equipment Valve Pump Fail on Failure mode Open Transfer off Close Seal leak/rupture Rupture Pump casing leak/rupture Leak
Heat exchanger Leak/rupture (tube to shell) Leak/rupture (shell to tube) Plugged Fouling
The analyzed process elements are ranked and risk grades are determined. According to the grades, risk assessment is performed. Fig. 4 shows all five stages of the reasoning algorithm. In the equipment analysis, the effect estimation for the selected equipment in the step II (micro decomposition) is accomplished: equipment with high severity is researched to find a detailed accident scenario. We use effect analysis method for the failure mode of the selected equipment to identify single equipment failure modes and each failure mode's potential effect on the system and the plant. This mode describes how equipment fails and is determined by the system's response to the equipment failure (see Table 1). In the scenario selection, we infer possible effects depending on the failure mode of the equipment. Possible scenarios for each failure mode are so variable that risk rankings are assigned according to the potential hazard of material and the magnitude of abnormal situation. Table 2 shows the failure mode and scenario selection procedures.
4. CASE STUDY: LPG STORAGE FACILITY
This system is one of typical LPG transportation and storage facilities, including propane underground cavern, propane coalescer, propane dryer, and a propane storage tank, illustrated in Fig.5.
Table 2 Example of scenario selection for a failure mode .
.
.
.
Identi.fication
Mode
Effect
Valve A on the chlorine line
Fail open
Excess flow of chlorine to the heater May cause a high level in the cleaning bed
Fail closed
No flow of chlorine to the cleaning bed Excess water flow to the cleaning bed
Material Chlorine
Risk Ranking C
Excess chlorine and water Water
D
Minor
Water
Minor
737
: ...... {~..i
P
t
minor hazard
[~'orZ
............................................
.
~t
:
}, Propane tans
r
Priority 2 Priority 1 Fig. 6. ESM application for LPG facility
Fig. 5. PFD for LPG storage facility
4.1. Accident scenario selection using the proposed algorithm In step I (macro decomposition), the entire process is decomposed into unit processes and process units are selected according to their functions and the meteorological condition around the area. Step II is the micro decomposition step. Through this step, ESM is applied to the 5 valves, 3 pumps and 1 heat exchanger, which have been selected as the most influential process components to the off-site area when an accident occurs. The result is shown in Figure 6; an accident due to pump A and valve C is the most hazardous one. In step III (equipment analysis), the analysis is performed to the process elements chosen in the step II, and the elements are ranked and risk grades are determined. Table 3 shows a part of the analysis. In step IV (scenario selection), according to the result of step III, propane releases due to the rupture of valve E or the open of valve E caused by the failure and rupture of pump sealing are selected as the most suitable accident scenarios. Step V is effect analysis: a consequence analysis is performed using the scenarios chosen in the step IV. An analysis program is being developed as part of this research, and that program may be used in this step (See Fig. 7 and Fig. 8 for the screenshot of implemented system). An analysis result can also be obtained using one of commercial software packages available today. Table 3 Scenario selection for LPG facility (in case of Pump A)
Identtfication Pump A on the liquefied
Mode Fail open
Effect Excess flow of propane to
Material
Risk Ranking
Propane
Minor
the propane underground
propane line
cavern Fail transfer
No flow of propane to the
off
propane underground
Minor
cavem Seal rupture
Large release of propane to the surrounding area
Propane
A
738
Fig. 7. Screenshot of main menu and input data
Fig. 8. Result of equipment analysis
5. CONCLUSION A strategy for producing robust accident scenarios in the quantitative risk analysis, which is performed in the process design or operation steps, has been proposed and tested to one of the LPG facilities and the DAP process (not shown in this paper). The obtained result of the systematic analysis enhances the reliability of the generated risk scenarios and prevents the risks from being overestimated; the result should be more helpful in the proper process design and emergency planning. The proposed strategy is being implemented as a part of the government-supported, quantitative process hazard analysis system, and expected to be successfully applied to most of mandated off-site consequence analyses in Korea. REFERENCES
1. Greenberg HR and Cramer JJ, Risk Assessment and Risk Management for the Chemical Process Industry, Van Nostrand Reinhold, 1991. 2. EPA, RMP Offsite Consequence Analysis Guidance, EPA, 1996. 3. Murphy JF and Zimmermann KA, "Making RMP hazard Assessment Meaningful", Process Safety Progress, 17(4), 238-242, 1998. 4. Hoist S, Hjertager BH, Solberg T and Malo KA, "Properties of Simulated Gas Explosion of Interest to the Design Process", Process Safety Progress, 17(4), 278-287, 1998. 5. Khan FI and Abbasi SA, "Techniques and Methodologies for Risk Analysis in Chemical Process Industries", Journal of Loss Prevention in the Process Industries, 11(4), 261-277, 1998. 6. CCPS, Vapor Cloud Source Dispersion Models, CCPS of the AIChE, 1989. 7. CCPS, Guidelinesfor Evaluating the Characteristics of Vapor Cloud Explosions, Flash Fires, and BLEVEs, CCPS of the AIChE, 1994. 8. CCPS, Guidelinesfor Hazard Evaluation Procedures, 2 nd Ed., CCPS of the AIChE, 1992. 9. Amaldos J, Casal J, Montiel H, Sanchez-Carricondo M and Vilchez JA, "Design of a Computer Tool for the Consequences of Accidental Natural Gas Releases in Distribution Pipes", Journal of Loss Prevention in the Process Industries, 11(2) ,135-148, 1998.