World Applied Programming, Vol (3), Issue (9), September 2013.

417-426
ISSN: 2222-2510
2013 WAP journal. www.tijournals.com

Fuzzy Risk Assessment and Categorization, based on Event

Tree Analysis (ETA) and Layer of Protection Analysis
(LOPA): Case Study in Gas Transport System
Sohrab Khaleghi *

Saeed Givehchi

Saeed Karimi

Health, Safety and Environment (HSE)

Department, University of Tehran, Iran.
sohrabkhaleghi@ut.ac.ir

Health, Safety and Environment (HSE)

Department, University of Tehran, Iran.

Health, Safety and Environment (HSE)

Department, University of Tehran, Iran.

Abstract: Process safety and risk assessment are vital demands for any industry to characterize hazards and
their risks for personnel, environment and loss of money. Risk matrix is a very useful tool to estimate risk of
process or equipment that helps decision-making processes. Thus fuzzy logic method for risk assessment is
selected as a new and efficient way to industry resource management. This study generally includes quantitative
reviews of possible accidents, based on previous accident experiences that may occur in a typical natural gas
transport system. For current risk assessment study the possibility exists to limit failure in case definition and
risk modeling to only accidents that may include fire, explosion and toxic effect risks. Consequently a fuzzy
risk matrix is extracted based on Layer of Protection Analysis (LOPA) and Event Tree Analysis (ETA)
procedures for analyzing 3 leakage scenarios based on the size of leakage, then classical risk indexes are
compared with those obtained from the fuzzy approach. Results of case study showed that risk indexes of fuzzy
risk assessment in small and medium leakage differ 30% and 8% to classical risk indexes respectively, and
shown that fuzzy risk assessment overcomes uncertainties and imprecision of classical risk assessment.
Keywords: Fuzzy logic, Risk assessment, Risk categorization, Decision-Making

I. INTRODUCTION
Nowadays, artificial intelligence (AI) computational methods, such as knowledge-based systems, neural networks,
genetic algorithms, and fuzzy logic, have been increasingly applied to several industrial researches [1-6].
Establishment and development of soft set theory and its applications are used in recent years and are extended in
combined forecasting, decision-making, information science and so on [7-9]. Soft set theory is usually associated with
other mathematical theories by many researchers for its characteristics, such as fuzzy set theory [10]. Fuzzy Logic is an
approach for measuring imprecision in estimation and assessment, and may be preferred to Probability theories when
capturing activity during uncertainty in situations where data base are either imprecise or inaccurate [11].
Recently information is defined by imprecision and uncertainty, therefore probabilistic risk analysis has been developed
for states in which measured data about the precision and reliability of a system are restricted and expert knowledge is
the best and only source of information available [12]. To overcome these uncertainties, increasing efforts have been
taken to introduce expert systems and sets for effective decision making in any process [13]. So fuzzy systems have the
capacity of developing the functionality of engineering systems and sets with linguistic terms in data analyzing,
processing or decision-making [14]. The fuzzy logic basis is easy to understand and expressed in a natural language,
flexible and can deal with imprecise data without necessarily needing the mathematical model of the system to be
controlled [15].
Chemical and process units contain huge amounts of dangerous chemical products and substances that may be exposed
to any kinds of hazards, like natural and process hazards [16]. Risk assessment must be an analytical process for

417

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

identifying the potential hazards, and consequently harmful effects of these hazards [17]. In recent years of oil and gas
industry, quantitative risk analyses have provided valuable information for decision process in the planning phase [18].
Successful safety management of such processes requires knowledge in safety management for judgment about types of
industry hazards. So risk assessment is a solution for assessing and analyzing industry hazards and is an appropriate tool
for managers to use resources to prevent loss of financial, time and reputation in industry. So the original purpose of risk
analysis is helping to better decision-making process about risks.
Hence tools and concepts of optimization are widespread in decision-making, design, and planning [19], and must be
used in risk management process in chemical and process industries by different types of risk assessment methods.
Risk analysis techniques based on particular characteristics are divided into four categories: deterministic, probabilistic,
qualitative and quantitative. Some methods of risk analysis are: HAZOP, Event Tree Analysis (ETA), Fault Tree
Analysis (FTA), Quantitative Risk Assessment (QRA), Layer Of Protection Analysis (LOPA), and Probabilistic Risk
Assessment (PRA) that is used to identify potential accident scenarios, estimate their likelihoods and consequences, and
improve system safety and operation, etc [20,21]. But in a scientific approach, there are three main routes for risk
assessment:
1) Qualitative route
2) Semi-quantitative route
3) Quantitative route [22].
First method is based on compliance assessment; in this method a reference state is available, then general risk
assessment is used to compare compliance or noncompliance with standards.
Second method is based on components categorization, final risk indexes or numbers are obtained though different
methods. And the third method is based on risk measurement assessment which is called Quantitative Risk Assessment
(QRA) [22].
However, knowledge is developing rapidly, but there is still lacking and uncertain process information, implicit in the
variable, models and in the large accident hazards consequently [23], and risk and reliability analyses involve estimating
the probabilities of hazardous events such as hardware or system failures and consequently their severity [24].
Classical risk assessment usually does not precise and its results do not sensitive to accidents severity and frequency. As
clearly shown in Figure 1, fuzzy risk assessment can be a solution to decision makers to have detailed and precise
information about risks and risk management.

Figure 1. Classical set and fuzzy set for safe and unsafe state [23].

418

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

This study generally includes quantitative reviews of possible accidents that may occur, based on previous accident
experience or judgment. In this paper, data of a natural gas pipeline manipulated by two conventional risk analysis
methods, ETA and LOPA and after that by using fuzzy logic uncertainty and imprecision of risk parameters reduced),
finally extracted risk numbers by classical matrix risk assessment compared to fuzzy risk assessment method to illustrate
of benefits of fuzzy risk assessment.
II. FUZZY AND ITS APPLICATIONS IN RISK ASSESSMENT
Fuzzy modeling is one of the most powerful tools to estimate relation between input-output of nonlinear systems. In this
method, fuzzy numbers are referred to variables that express uncertainties [25]. A fuzzy number expresses relation
between an imprecise or uncertain number, X and a membership function, that is between (0, 1) [26].
In other words, the fuzzy logic tool provides a technique to deal with imprecision and information, the fuzzy theory
provides a mechanism for representing linguistic constructs such as many, low, medium, few, etc.
In the new view of analysts and decision makers, risk analysis is being done with great uncertainties because risk
assessment requires detail information about damage frequency rate of particular processes and equipment components.
This data usually are imprecise and uncertain. One of the advantages of fuzzy logic in the field of risk assessment is
dealing with these problems [23].
Risk analysis helps us to take both certain and uncertain quantities and calculate them to what extent a specific events or
scenarios can be occur in future? But an usual and useful idea exists in facing with any type of uncertainties that includes
lack of knowledge and vagueness, that believes the best way to solve this obstacles is utilizing fuzzy logic [27]. Based on
fuzzy logic theory or fuzzy set, this method can be used in systems that include imprecision or uncertainty, and solved
systems that there is not a specified or precise boundary in it [25], such as Failure Mode and Effect Analysis (FMEA)
method that developed a risk priority number that is qualitative and is accompanied with uncertainties [28].
Fuzzy logic is based on series of linguistic human rules. Fuzzy systems are transforming these rules to their mathematic
equivalent and simplifying system designers work In this case, risk decision makers. In addition we will have more
accurate expressions about system behaviors in the real world. Other advantages of fuzzy logic are facility and
flexibility, because fuzzy logic can manipulates imprecise problems that have had incomplete information, and it can
model any arbitrary and complicated function [29,30].
There are two types of fuzzy inference systems:
1) Takagi-Sugeno models;
2) Mamdani models;
Output membership function of Takagi-Sugeno models contains constant values or linear functions of input variables,
while output membership functions of Mamdani models are fuzzy sets, that could convert to linguistic information for
modeling [31]. These models are described in details in Takagi & Sugeno and Mamdani & Assilian papers [32,33].
Generally, the fuzzy logic provides an inference structure that enables appropriate human reasoning capabilities [34].
Concept of fuzzy set in risk assessment is mathematical formulation of imprecise parameters in a risk analysis method.
In this particular case, all of parameters and risk components (frequency and severity of a consequence) are defined
based on terms of fuzzy set, and risk assessment matrix formed finally. These concepts are presented in Figure 2
Consequently results of risk matrix are used for process hazard analysis (PHA). The basis definition of risk matrix
compromises of two parameter, first an occurred event scenario in a particular accident scenario and (second,)its
frequency. In other word for making a risk matrix, the following steps should be doing:
1)
2)
3)
4)

Categorization and scaling of event severity and frequency;

Categorization and scaling of output risk index;
Constructing risk-based rules knowledge;
Graphical edition of the risk matrix [35].

419

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

Figure 2. Typical procedure for Fuzzy risk matrix definition.

The use of fuzzy logic in the field of safety, risk and reliability analysis has been presented in several books and papers
that show the importance of this method in industries [36-41].
III. FUZZIFICATION PROCESS AND FUZZY RISK MATRIX DEVELOPMENT
III.1 Fuzzification process
In summary,Fuzzy Logic System (FLS) includes following components:
1) The fuzzifier, which task is converting frequency variables (F), consequence severity (S), and risk
categorizations (R), with crisp number and mapping them into fuzzy sets ( F , S , R ).
2) The inference engine, which task is mapping of input fuzzy sets into fuzzy output sets by using knowledge base
rules. This procedure is followed by IF-THEN rules, established on the basis of human knowledge or
mathematical calculation.
3) The defuzzifier task is vice versa of fuzzifier and reaches to a precise and deffuzified output for any variable
[34]. These components are showed in Figure 4 in general form.

Figure 4. General form of a typical fuzzy set.

420

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

III.2 Fuzzy risk matrix development

Contents of subheading 2 goes here. In order to form fuzzy sets, 7 level of probabilities, 5 level of consequence
severities, and 4 level of risks are defined that conformity with some papers [23-25]. These fuzzy sets are shown in Table
1 Acceptability of the different approaches in a risk assessment has been varied and related to safety manager of
organization [42].
Because of defining of Gaussian distribution and overlaying of numbers, open and close intervals are used as shown in
Table1.
Table 1. Fuzzy sets for fuzzy risk matrix.
Linguistic variables of severity
Frequency(F)

Fuzzy set
Very High

F [10-2,1) [1/year]

High

F [10-1,10-3 )

Moderate

F [10-2,10-4 )

Low

F [10-3,10-5 )

Very Low

F [10-4,10-6 )

Unlikely

F [10-5,10-7 )
F<10-6

Remote
Severity of consequence(S)

Negligible

S [1,2]

Low

X
XF (10-8,100)

XS [1,5)

S (2,3]

Moderate

S (3,4]

High

S (4,5]
S>5

Catastrophic
Risk category(R)

Description range

A:Acceptable

R [0,2]

TA: Tolerable Acceptable

TNA: Tolerable Unacceptable
NA: Unacceptable

XR (0,5)

R [1,3]
R [2,4]
R [3,5]

The Relations between fuzzy variables are defined by Mamdani rules that are engineering knowledge-based rules, by
using membership functions (MFs) based on Table 1 Then 35 rules for risk assessment are defined. For example, for
estimating the risk number: IF consequence severity is Catastrophic and event frequency is Remote, THEN the risk is
TA. Figure 5 shows the fuzzy surface of risk matrix and the classical risk matrix for comparison.
It should be mentioned that in this paper for establishment of severity fuzzy logic systems, consequences intensity and
size of release used to estimate more accurate severity in fuzzy risk surface. For this purpose, four categories of violence
are defined based on experimental information [43,44]: Toxic effect (0-2 min), Pool fire (2-5 min), Flash fire (5-10
min), and jet fire (10-15 min), these information are summarized in Figure 6. Three categories are defined for size of
release: 0-50 (mm), 50-100 (mm), and greater than 150 mm. This categorization is shown in Table 2. So 12 Rules have
been set up for estimating the severity of consequence, for example: IF release is Rupture and time is TX, THEN the
severity is Moderate. In this paper triangular methods are used for membership function of release time in fuzzy severity
consequence, and Gaussian method is used for other fuzzy sets. These data are illustrated in Figure 7 clearly.

421

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

Figure 5. Fuzzy surface and classical risk matrix

Figure 6. Discharge flow rate profile from leakage [43].

Table 2. Fuzzy sets for fuzzy severity matrix

Linguistic variables of severity
Time of release(T)

Fuzzy set

Description range

Toxic effects(TX)

,2) (min)

Flash fire(FF)

,10)

Jet fire(JF)

,15)

Small

,50) (mm)

Medium

X L (0,150)

,150)

Full Bore Rupture

Severity(S)

X T (0,15)

,5)

Pool fire(PF)

Leak size(L)

L >150

Negligible

,2]

Low
Moderate
High

2,3]

3,4]

S
4,5]
S>5

Catastrophic

422

X S (1,5)

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

Figure 7. Fuzzy severity of consequence matrix for different scenarios.

IV. CASE STUDY AND SCENARIO DEVELOPMENT

IV.1. Case study description
This paper is done in one of the oil and gas exporter islands in Persian Gulf. The selected area is a gas recovery site; the
gas recovery facility is a grass roots development and will consist of the simultaneous development of two separate
plants (gas treatment and recovery) that share common utilities and liquid export systems. This site incorporates
thousands of process equipment items and piping, but not all of them are capable of producing severe fire, explosion and
toxic impacts onto their nearby buildings and/or personnel, because risks are function of the type of material that is going
to be released, inventory, operating pressure, operating temperature, leak size, leak duration, etc. For current risk
assessment study the possibility exists to limit failure case definition and risk modeling to only those accidents capable
of fire, explosions and toxic risks in plant.
IV.2.Scenario development
In this paper three different accident initiator are defined, which is usually represented by a variety of leakages or full
bore ruptures with different sizes in pressurized equipment and fittings due to process upset, incorrect materials of
construction, design error, over pressurization, corrosion, equipment malfunction, operational error, and abnormal
operation. Table 3 shows leakage scenarios discussed in this paper.
Table 3. Leakage scenarios of piping in gas recovery plant.
Initiating Event
Small leakage

Range of Leak Hole Size(mm)

Smaller than 50

Representative Leak Hole Size (mm)

25

Medium leakage

50 mm -150

100

Full Bore Rupture leakage

Larger than 150

Equivalent to size of largest connection

Series of small, medium, and full bore rupture scenarios for a segment of the plant was selected to compare classical and
fuzzy risk assessment methods. For this purpose ETAs of gas compression stations scenarios are illustrated in Fig. 8 and
any leakage frequencies of them are determined. The scenario frequencies (F) are usually obtained by analysis of
previous accident experiences represented as general process data that is supported with a large database of historical
failure frequencies [45]. It should be mentioned that the means of Ignition in Fig. 8 is equivalent to failure of second IPL.
This type of definition is expressed to other factors such failure of emergency shut-down systems and fire extinguisher
systems.

423

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

Figure 8. ETAs of scenarios for leakage in a typical compression gas station of gas recovery plant [45].

V. RESULTS AND DISCUSSION

Table 4 shows results of three initiating Events and their typical severity of consequence, it should be mentioned that
these numbers are just for comparison of classical and fuzzy sets. But Table 5 presents outcomes of fuzzy logic system
(FLS) that explicitly demonstrate the difference between classical and fuzzy view of risk indexes that are based on ETAs
presented in Figure 8.. In two sample leakage, difference between fuzzy risk indexes and classic risk indexes in a small
leakage scenario is about 30% and 8% in a medium leakage scenario. These results are in accordance to results of
Markowski and Mannan research [23], that have 20-30% differences between classical and fuzzy risk assessment. It
should be noted that determining and discussing any risk assessment methods and risk boundaries are related to the
request, planning, and ability of organization management for risk mitigation. This detailed information in risk index
enables us to obtain more precise and better decision making processes for improvement of process safety. In addition
this helps us to select more reliable and effective IPLs for the same processes in the future or make better decisions about
safety resource management. It is obvious that the final risk index strongly depends on the structure of risk assessment
matrix.

Table 4. Results of classical and fuzzy severity of consequence and their comparison
Initiating Event
Small leakage(25mm)
Medium leakage(75mm)
Full Bore Rupture(150mm)

Time of release(min)

Classical severity(S)

0-2
2-5
10-15

1
3
5

424

Fuzzy severity( )
1.63
3.4
4.48

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

Table 5. Results of classical and fuzzy risk indexes and their comparison for Toxic effect
Initiating Event

Frequency(F)

Classical Severity of
consequence(S)

Small leakage
Medium leakage
Full Bore Rupture

6.19E-03
4.53E-05
7.07E-05

1
2
3

Fuzzy Severity of
consequence( )
1.63
2.11
3.07

Classical risk
index(R)

Classical risk
category

2
2
2

TA
TA
TA

Fuzzy Risk
index( )
2.61
2.16
2.11

VI. CONCLUSION
Process safety is an important component to any industry that must be measured by risk assessment methods, but risk
assessment accompanies to uncertainties and imprecision. Fuzzy logic system enables to overcome this uncertainties and
imprecision in risk assessments and helps us to have a better view about risk indexes and predictions of process hazards.
LOPA facilitates the determination of more precise, causeconsequence pairs and therefore, improves scenario
identification. In addition to combination of this with fuzzy logic resulted in improvement of decision making process
about risk indexes and facing hazards. In this case study, classical risk category was TA (Tolerable acceptable), but by
using the fuzzy logic differences of three scenarios was shown, that helps safety managers to select better decisions to
mitigate of risk and prediction more efficient IPLs.
The use of fuzzy logic in the area of risk assessment has proven notable advantages in calculating the risk categorization
and its rating for any process, equipment, etc., and is an easy, user friendly and time saving method. Furthermore, fuzzy
based models are reliable and could produce the most accurate and detailed results in future studies by improving and
innovating in fuzzy logic and other related soft sets to overcome imprecision and uncertainties of any common risk
assessment methods such as What-IF, FMEA, HAZOP, PRA, etc. or even any special risk assessment method for a special
process.
REFERENCE
[1]
[2]
[3]

[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]

Peche R, and Rodrguez E (2012). Development of environmental quality indexes based on fuzzy logic A case study,
Ecological Indicators, vol(23), 555565.
Gharibi H, Mahvi A H, Nabizadeh R, Arabalibeik H, Yunesiana M, Sowlat M H (2012). A novel approach in water quality
assessment based on fuzzy logic, Journal of Environmental Management, vol(112), 87-95.
Akko S (2012). An empirical comparison of conventional techniques, neural networks and the three stage hybrid adaptive
Neuro Fuzzy Inference System (ANFIS) model for credit scoring analysis: The case of Turkish credit card data, European
Journal of Operational Research, vol(222), 168178.
Li D F (2012). A fast approach to compute fuzzy values of matrix games with payoffs of triangular fuzzy numbers, European
Journal of Operational Research, vol(223) 421429.
Zarandi A (2010). On fuzzy information theory, Indian Journal Of Science and Technology, vol(3), No9.
Valizadeh Z, Ezzati R, Khezerloo S (2012). Approximate symmetric solution of dual fuzzy systems regarding two different
fuzzy multiplications, Indian Journal Of Science and Technology, vol(5), No(2).
Jiang Y C, Liu H, Tang Y, Chen Q M (2011). Semantic decision making using ontology-based soft sets, Mathematical and
Computer Modeling, vol(53), 11401149.
Roy A R, and Maji P K (2007). A fuzzy soft set theoretic approach to decision making problems, Journal of Computational
and Applied Mathematics, vol(223), 540542.
Bobillo F, and Straccia U (2012). Generalized fuzzy rough description logics. Information Sciences, vol(189), 4362.
Xiao Z, Xia S, Gong K, Li D (2012). The trapezoidal fuzzy soft set and its application in MCDM, Applied Mathematical
Modeling, vol(36), 58445855.
Haji Yakhchali S (2012). A path enumeration approach for the analysis of critical activities in fuzzy networks, Information
Sciences, vol(204), 2335.
Guikema S D (2009). Natural disaster risk analysis for critical infrastructure systems: An approach based on statistical learning
theory, Reliability Engineering & System Safety, vol(94), 855860.
Lee S, Yang J, Han J (2012). Development of a decision making system for selection of dental implant abutments based on the
fuzzy cognitive map, Expert Systems with Applications, vol(39), 1156411575.

425

Sohrab Khaleghi et al. World Applied Programming, Vol (3), No (9), September 2013.

[14]

[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]

[45]

Costa Silva G, Palhares R M, Caminhas W M (2012). Immune inspired Fault Detection and Diagnosis: A fuzzy-based
approach of the negative selection algorithm and participatory clustering, Expert Systems with Applications, vol(39), 12474
12486.
Li Q, Chen W, Li Y, Liu S, Huang J (2012). Energy management strategy for fuel cell/battery/ultracapacitor hybrid vehicle
based on fuzzy logic, Electrical Power and Energy Systems, vol(43), 514525.
Markowski A S, Mannan M S., Bigoszewska A (2009). Fuzzy logic for process safety analysis, Journal of loss prevention in
the process industries, vol(22), 695-702.
Milne I, Ritchie R, Karihaloo B L (2003). Comprehensive structural integrity, Vol(1), Elsevier publishers.
Reda W, Mosleh A, Vinnem J E, Aven T (2009). On the use of the hybrid causal logic method in offshore risk analysis,
Reliability Engineering and System Safety, vol(94) 445455.
Ben-Haim Y(2012). Doing our best: Optimization and the management of risk, Risk Analysis, vol(32) 1326-1332.
Marengo C R, Flores J D L., Molina A L, Roman R V, Vazquez V C, Mannan M S(2013). A formulation to optimize the risk
reduction process based on LOPA, Journal of loss prevention in the process industries, vol(26), 488-494.
Mohaghegh Z, and Mosleh A(2009). Incorporating organizational factors into probabilistic risk assessment of complex sociotechnical systems: Principles and theoretical foundations, Safety Science, vol(47), 11391158.
Mather J, and Lines I G (1999). Assessing the risk from gasoline pipelines in the UK based on a review of historical
experience, WS Atkins safety and reliability. Birchwood. Warrington. Cheshire WA3 7WA.
Markowski A S, and Mannan M S(2009). Fuzzy logic for piping risk assessment (pfLOPA), Journal of loss prevention in the
process industries, vol(22), 921-927.
Guikema S D, and Goffelt J P(2008). A Flexible Count Data Regression Model for Risk Analysis, Risk Analysis, vol(28), 213223.
Zadeh L A(1965). Fuzzy sets, Information and control, vol(8), 338-353.
Sadiq R, Al-zahrani M A, Sheikh A K, Husain T, Farooq S(2004). Performance evaluation of slow sand filters using fuzzy
rule-based modeling, Environmental Modeling & Software, vol(19), 507-515.
Nielsen T, and Aven T(2003). Models and model uncertainty in the context of risk analysis, Reliability engineering and system
safety, vol(79), 309-317.
Liu H C, Liu L, Liu N, Mao L X(2012). Risk evaluation in failure mode and effects analysis with extended VIKOR method
under fuzzy environment, Expert Systems with Applications, vol(39) 1292612934.
Zadeh L A(1984). Making compute think like people, IEEE spectrum, vol(8), 26-32.
Haack S(1979). Do we need fuzzy logic?, International Journal of Man-Machine Studies, vol(11) 437-445.
Huang Z, and Hahn J(2009). Fuzzy modeling of signal transduction networks, Chemical Engineering Science, vol(64), 20442056.
Takagi T, and Sugeno(1985). M. Fuzzy identification of systems and its application to modeling and control, IEEE
Transactions on Systems, Man and Cybernetics, vol(15),116-132.
Takagi T, and Sugeno(1985). M. Fuzzy identification of systems and its application to modeling and control, IEEE
Transactions on Systems, Man and Cybernetics, vol(15),116-132.
Sivanandam S N, Sumathi S, Deepa S N(2007). Introduction to fuzzy logic using MATLAB, Springer-verlag Berlin
Heidelberg, New York.
Markowski A S, Mannan M S(2008). Fuzzy risk matrix, Journal of hazardous materials, vol(159) 152-157.
Solzano E, and Gozzani V(2006). A fuzzy set analysis to estimate loss intensity following blast move interaction with process
equipment, Journal of loss prevention in the process industries, vol(19), 343-352.
Geymar J A B, and Ebecken N F F(1995). Fault tree: a knowledge-engineering approach, IEEE transactions on reliability,
vol(44:1), 37-45.
Ding Y, and Lisnianski A(2008). Fuzzy universal generating functions for multi-state system reliability assessment, Fuzzy sets
and systems, vol(159), 307-324.
Guimarees A C F, and Ebecken N F F(1995). Fuzzy FTA: a fuzzy fault tree system for uncertainty analysis, Annals of nuclear
energy, vol(26), 523-532.
Markowski A S, Mannan M S, Kotynia A, Pawlak H(2012). Application of fuzzy logic to explosion risk assessment, Journal
of loss prevention in the process industries, vol(24), 780-790.
Klir G J, and Yuan B(1995). Fuzzy sets and fuzzy logic theory and applications. Prentice Hall PTR publication. Upper Sanddle
River. New Jersey.
Aven T(2001). A risk concept applicable for both probabilistic and non-probabilistic perspectives, Safety Science, vol(49),
10801086.
GB 50183(2004). Fire Protection Design of Petroleum and Natural Gas Engineering, Chinese Standards.
Van den Bosch C J H, and Weterings R A P M(2005). Methods for the calculation of physical effects due to releases of
hazardous materials, Due to releases of hazardous materials (liquids and gases) Yellow book,3th. Ed., Rev.2, Committee for
the Prevention of Disasters, Netherland.
Spouge J R, and Fannemark E(2006). Det Norske Veritas (DNV) Technical Note 14, Process equipment failure frequencies
Rev.3, Issued at risk net.

426