Integrate fault tree analysis and fuzzy sets in quantitative risk assessment

1. International INTERNATIONAL Journal of Advanced JOURNAL Research OF in Engineering ADVANCED and Technology RESEARCH (IJARET), IN ISSN ENGINEERING
0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
AND TECHNOLOGY (IJARET)
ISSN 0976 - 6480 (Print)
ISSN 0976 - 6499 (Online)
Volume 5, Issue 10, October (2014), pp. 12-20
© IAEME: www.iaeme.com/ IJARET.asp
Journal Impact Factor (2014): 7.8273 (Calculated by GISI)
www.jifactor.com
12
IJARET
© I A E M E
INTEGRATE FAULT TREE ANALYSIS AND FUZZY SETS IN
QUANTITATIVE RISK ASSESSMENT
Rachid Ouache1, Ali A.J Adham2, Noor AzlinnaBinti Azizan3
1, 2, 3Faculty of Technology, University Malaysia Pahang, 26300, Malaysia
ABSTRACT
Quantitative risk assessment is the most important step to judge the results of risk estimation
in the process of decision making to improve level of safety. Quantitative risk assessment hasfaced
big problemwith complexity of engineering systemsin term of reliability. In this study, reliability of
risk Assessment proposed to solve problem of uncertainty based on fuzzysets and fault tree analysis
to precise values of top event. The results demonstrated that the model proposed is the best to solve
problem of uncertainty in quantitative risk assessment.
Keywords: Quantitative risk assessment, Fault tree analysis, Fuzzyset, Reliability.
1. INTRODUCTION
In recent decades the world has witnessed an increase in the number of major accidents and
disasters, where considerable efforts are made to control the safety of industrial systems. Risk
assessment approach are designed primarily to reduce the existing risk inherent in engineering
system to a level considered tolerable and maintain it over time.The method of fault tree analysis
used to study the reliability and dependability of complex systems (Ding et al, 2010;Gupta et al,
2007).A study of system reliability, the probabilities are often considered accurate and fully
determinable. It is also assumed that all the information on the performance of the reliability of the
system and its components is provided (Wang et al, 2009). The Improve of reliability for prolonging
the life of the item based on two steps essential, on the one hand, study reliability issues and on the
other hand, estimate and reduce the failure rate. Two approaches for risk analysis, which can be
qualitatively and quantitatively. The qualitative approach used when there is a source of danger, and
when there are no safeguards against exposure of the hazard, and then there is a possibility of loss or
injury. In complex engineering systems, there are often safeguarded against exposure of hazards for
maximizing the level of safeguards, and minimize the level of risk. The quantitative risk assessment
is the approach concerned with this study. Since quantitative risk analysis involves estimation of the

2. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
degree or probability of loss, risk analysis is fundamentally intertwined with the concept of
probability of occurrence of hazards.This studyproposes FTAuses fuzzy setstoprecisethe values of
top events which help us to precisethe value of risk.
13
2. FUZZY FAULT TREE ANALYSIS FFTA
Probabilistic safety assessment by FTA was considered as an important tool to evaluate the
systems engineering. Boolean algebras are used to mathematically represent the tree diagramand
calculate the output of every logic gate (Haimes et al, 2004; Epstein Rauzy 2005; Ericson
2005;Huang, Tonga Zuo 2004, N. Limnios, 2007). The occurrence probability of the undesired top
event is afunction of the reliability data of primary events, which are also known as basic
events(IAEA 2007; Yang 2007).Fault tree analysis can give two types of results, i.e. qualitative
andquantitative results.
2.1 FAULT TREE ANALYSIS
The FTA is composed of a number of symbols to describe events,Boolean gates, and page
transfers. Transfer event symbols are pointers toindicate sub-tree branches that are used elsewhere in
the tree (Ericson 2005; Vesely et al. 1981).
2.1.1 BOOLEAN ALGEBRA
Boolean algebra is the algebra of fault events used in a fault tree tomathematically represent
the relationship between input and output faultevent of a Boolean gate in the tree. This relationship
describes a situation where anoutput of the gate either fails or not (Haimes 2004;Vesely et al. 1981).
Table.1: Booleanalgebras
Rules Engineering symbolism Mathematical Symbolism
Idempotent law X.X=X
X+X=X
X= X
X= X
Distributive law X. (Y+Z)= X.Y+X.Z X= (XY)(XZ)
Commutative law X.Y=Y.X XY=YX
Absorption law X.(X+Y)=X
X+(X.Y)=X
X(XY)=X
X(XY)=X
2.1.2 FAILURE PROBABILITY CALCULATION
The failure probability of an output event for two or more independent inputevents combined
by a Boolean OR gate calculated using Eq.(1). And by a Boolean AND gate calculated using Eq. (2).

3. Where is the failure probability of the input event , and n is the number of inputevents
to the Boolean gate.

4. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
14
2.2 FUZZY SET THEORY
2.2.1 Fuzzy Sets
The utility of fuzzy sets lies in their ability to model uncertain or ambiguous data, Fuzzy Sets
FS is important to observe that there is an intimate connection between Fuzziness and Complexity.
Fuzzy sets provide a means to model the uncertainty associated with vagueness, imprecision, and
lack of information regarding a problem or a plant, etc (Dubois, 1980, Zadeh, 1978).
The soft computingis useful tool for solving problems in many fields.It is a high-performance
language for technical computing (Dubois, 1998). Typical usage includes: Development of
algorithms, Data analysis, exploration and visualization, Mathematics and computing, Scientific and
technical graphics, Modeling, simulation and prototyping (Cavallo et al, 1996; Canos,2008;Bouchon
et al, 1995).
Fuzzy set theory deal with mathematically model information uncertainties and the theory has
been developed and applied in a number of real world applications (FuzzyTECK,1995; Chen, 2001;
Gebhardt, 1995; Zadeh,1965)
μ

5. μ

6. The value of the membership function μA(X) represents the membership grade of x in X. The
closer value to 1 is, the stronger degree of membership of x in A is. (Bector Chandra 2005; Lu et
al.2007).
μA!B(x)= max(μA(x), μB(x)) = μA(x) !μB(x) (4)
μAB(x)= min(μA(x), μB(x)) = μA(x) μB(x) (5)
μ Ac(x)=1- μA(x) (6)
2.2.2 FUZZY NUMBERS
A fuzzy number is one type of fuzzy sets with normalized membership function. A fuzzy
number is subset of real line R who membership function μ#$(x) Can be a continuous mapping
from R into a closed interval [0,1]. The membership functionμ#$(x) has the following characteristics
(Dubois Prade 1978;Wang et al. 2006).
The membership function of the number can be expressed as follows.
μ#$(x) =
μ#%' ( ) ' ) *

7. * ) ' ) +
μ,#%'+ ) ' ) -
(7)
In a special case when b=c, the trapezoidal fuzzy number into a trianglar fuzzy number
(Abbasbandy Hajjari 2009; Wang et al.2006).
.#%'= /01
201(8)
.#,%' 30/
304(9)

8. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
15
Figure.1: Trapezoidal and triangular fuzzy numbers
2.2.3 Fuzzy inference system FIS
The output level z of each rule is weighted by firing strength w of the rule (Guh et al, 2008;
Castillo, 1999a).The final output of the system is weighted average of all the rule output which is
given as:Final output= 5 6787 97
:;
5 67 97
:;
(10)
The Sugeno proposed a fuzzy inference method that guarantees the continuity of the
output.Tomohiro Takagi and Michio Sugeno introduced a mathematical tool to build a fuzzy model
of a system where fuzzy implication and reasoning are used in their paper in the year of 1985
(Sugeno, 1985).A FISwith five functional blocks described in Figure.2.
Figure.2: Fuzzy inference system
3. CASE STUDY
The storage tank is designed to hold a flammable liquid under slight nitrogen positive
pressure under controls pressure (PICA-I)(CCPS,2000).The incident isthe top event that will be
developed in the fault tree, Construction of the fault tree FTbased on the knowledge of the system
andinitiating events in the hazard operability HAZOP study, the fault tree is constructed manually.
Equipment and valves Instruments
FV = Flow control Valve
T = Tank
P = pump
PV =Pressure control Valve
RV = Relief Valve
1 = 1 inch size
P = Pressure
T = Temperature
L = Level, F = Flow
I = Indicator
C = Controller
A = Alarm, H = High,
L = Low
#$

9. .#%' .#,%'
#$

10. .#%' .#,%'
!
#
$
#
$

11. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
,()
*
0()
*
0
% %(
,!$7$
,'586
2 2
' 1
16
-
,
%'
*
%(#)
%
+

13. 1
/
(#)
Figure.3: Flammable liquid storage tank
, $
.
!
!
/
,$
0
Figure. 4 shows fault tree constructed manually. Every event is labeled sequentially with a T
for the top event, M for intermediate event, andB for a basic or undeveloped event. The procedure
starts at the top event, Tank rupture due to overpressure, and determines the possible events that
could lead to this incident (CCPS, 2000).
,!7$
95'6
*
$
!
9//516
2
)
3
$
!
91/586
)
%
9''516
3
$
4'51
6
.
4151
6
%(#)(6
%45
'
%(#)4/5'6
, $
4.516
+#
4+516
/
*
$
*4:516
%
48516
Figure.4: Fault tree analysis for tank rupture due to overpressure using boolean algebra
Tables.2 Shows calculate frequency using Boolean algebras and fuzzy inference methods, the
results by fuzzy sets are more precise than classical method , and in tables we can see two kinds of
results. The first illustrate using same imputs in classical with results more precise, the second is the

14. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
results by change the value of Input which influence directy in output. The results show the methodto
calculate top event as follow:
Table.2: Tank rupture due to overpressure using Boolean algebras and fuzzy inference methods
Calculate frequency using Boolean algebras method
Figure.5.a Rules inferencess process of Tank
17
If
Tank overpressured
=1.51*10-2/yr
And
Pressure relief system
failure = 2.5*10-3
Then
Tank rupture due to
overpressure=2*10-5/yr
-Level intervals of frequency usingfuzzy inference method
1
If
0.0151
And
0.0025
Then
2.53*10-5
2 [0.0101 0.0133] 0.0025 2.52*10-5
3 [0.0134 0.0201] 0.0025 2.53*10-5
4 0.0151 [0.002 0.003] 2.53*10-5
5 0.0101 [0.002 0.003] 2.52*10-5
rupture due to overpressure
Figure.5.bThree dimensional diagram, Tank
rupture due to overpressure , Pressure relief
system failure and Tank overpressured
Figure.5.c Two dimensional diagram, Tank
rupture due to overpressure with Tank
overpressured
Figure.5.dTwo dimensional diagram, Tank
rupture due to overpressure with Pressure
relief system failure
Fault tree analysis after calculating the frequency using fuzzy inference method, the results
show that fuzzy sets is very appropriate for precise the value of initiating events for risk assessment.

15. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
,!$7$
,'581;86
)
2 2
' 1
3
$
!
/
*
$
*4:516
18
,!7$
958;'6
8
*
$
!
9//516
2
91/58/;86
)
%
9''58/1
3
$
4'51
.
4151
%(#)(6
%45
'6
%(#)4/5'6
, $
4.516
+#
4+516
%
48516
Figure.6: Fault tree analysis for tank rupture due to overpressure using fuzzy inference
4. DISCUSSION
From our results above we can say that:
The results of fault tree analysis using fuzzy sets is more powerfull than classical methods,
which help us to precise the value of top event and to deal with uncertainty. The Surface Viewer
shows a three-dimensional curve that represents the mapping from inputs and outputs, the diagram
drawn using three equations essential (7)-(9).From the simulation results, it may be observed that the
fuzzy model can be successfully employed to instantaneously determine the exactitude of fault event
with only two input specifications.
The two-dimensional diagrams show the relationship between one input and one output,
whereas the three dimensional diagrams show the relationship between two inputs and one output
and the tables show also the change of outputs by changing the inputs.The two dimension diagram
for OR gate happened one steps as line continue, this show the extent of one input directly on output,
however the AND gate, the two dimensional diagramshow three steps which means that one input
not influence directly on output.The same analysis for three dimensional diagram for both OR and
AND gates.Results calculated of the rule output using equation (10).
5. CONCLUSION
In this paper, we have proposed a fuzzy fault tree analysis FFTA for reliability quantitative
risk assessmentas new model to solve problem of imprecise and uncertainty of results.The
application of this model gavebest results for the decision making.FFTA asnewmodelfor reliability
quantitative risk assessment based onFuzzy inference system FIS which considered as the best tool to
precise the value of top events of fault tree analysis.The fault tree analysis is constructed using two
different calculation approaches, boolean algebrasclassic and fuzzy inference using sugeno method,

16. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
whereas the fuzzy inference system gave results more precise than boolean algebras method, which
consider as comlementry to solve problems of uncertainty.Fault tree analysis with fuzzy sets is good
solution for reliability quantitative risk assessment.
19
REFERENCES
[1] Abbasbandy, S. Hajjari, T. 2009, 'A new approach for ranking of trapezoidal fuzzy
numbers',Comput. Math. Appl., vol. 57, no. 3, pp. 413-419.
[2] Bector, C.R. Chandra, S. 2005, 'Fuzzy numbers and fuzzy arithmetic', in J. Kacprzyk
(ed.),Fuzzy Mathematical Programming and Fuzzy Matrix Games, Springer, Berlin,
pp. 39-56.
[3] Bouchon-Meunier B, Yager R, Zadeh L (1995) Fuzzy logic and soft computing.World
Scientific, Singapore.
[4] Canos, L. Liern, V. 2008, 'Soft computing-based aggregation methods for human resource
management', Eur. J. Oper. Res., vol. 189, no. 3, pp. 669-681.
[5] Castillo O, Melin P (1994) Developing a new method for the identification of
microorganisms for the food industry using the fractal dimension. J Fract 2(3):457–460.
[6] Castillo O, Melin P (1999a) A new fuzzy inference system for reasoning with multiple
differential equations for modelling complex dynamical systems. In: Proceedings of
CIMCA’99, IOS Press, Vienna, Austria, pp. 224–229.
[7] CCPS, 2000, Guidelines for Chemicals Process Quantitative Risk Analysis, 2nd edition,
American institute of chemical engineers (AICHE).
[8] Chen G, Pham TT (2001) Introduction to fuzzy sets, fuzzy logic, and fuzzy control systems.
CRC, Boca Raton, FL.
[9] Ding, Y., Zuo, M.J., Lisnianski, A. Li, W. 2010, 'A framework for reliability
approximation of multi-state weighted k-out-of-n systems', IEEE Trans. Reliab., vol. 59, no.
2, pp. 297- 308.
[10] Dubois D, Prade H (1980) Fuzzy sets and systems: theory and applications.Academic, New
York.
[11] Dubois D,Prade H (1998) Soft computing, fuzzy logic, and artificial intelligence, soft
computing a fusion of foundations, methodologies and applications, vol 2(1). Springer, Berlin
Heidelberg New York, pp. 7–11.
[12] Dubois, D. Prade, H. 1978, 'Operations on fuzzy numbers', Int. J. Syst. Sci., vol. 9,
pp. 613- 626.
[13] Epstein, S. Rauzy, A. 2005, 'Can we trust PRA?',Reliab. Eng. Syst. Saf., vol. 88, no. 3, pp.
195-205.
[14] Ericson, C.A. 2005, 'Fault tree analysis', in Ericson (ed.), Hazard Analysis Techniques for
System Safety, John Wiley Sons, Virginia, pp. 183-221.
[15] Gebhardt J (1995) New industrial applications of the fuzzy-PLC proceedings of the 3.
In: European congress on fuzzy and intelligent technologies (EUFIT 95), Aachen, 08/95.
[16] Guh, Y.Y., Po, R.W. Lee, E.S. 2008, 'The fuzzy weighted average within a generalized
means function', Comput. Math. Appl., vol. 55, no. 12, pp. 2699-2706.
[17] Gupta, S. Bhattacharya, J. 2007, 'Reliability analysis of a conveyor system using hybrid
data', Qual. Reliab. Eng. Int., vol. 23, no. 7, pp. 867-882.
[18] Haimes, Y.Y. 2004, 'Fault trees', in Risk Modeling, Assessment, and Management, 2nd edn,
John Wiley Sons, New Jersey, pp. 525-569.
[19] Huang, D., Chen, T. Wang, M.J.J. 2001b, 'A fuzzy set approach for event tree analysis',
Fuzzy Sets Syst., vol. 118, no. 1, pp. 153-165.

17. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 – 6480(Print),
ISSN 0976 – 6499(Online) Volume 5, Issue 10, October (2014), pp. 12-20 © IAEME
20
[20] Huang, H.Z., Tonga, X. Zuo, M.J. 2004, 'Posbist fault tree analysis of coherent systems',
Reliab. Eng. Syst. Saf., vol. 84, no. 1, pp. 141-148.
[21] IAEA 2007, IAEA safety glossary, terminology used in nuclear safety and radiation
protection, IAEA, Vienna, Austria.
[22] Limnios .N.,Fault Tree, UK. ISTE, 2007.
[23] Lu, J., Zhang, G., Ruan, D. Wu, F. 2007, Multi-Objective Group Decision Making:
Methods, Software and Applications with Fuzzy Set Techniques, Imperial College Press,
London.
[24] Sugeno, M.,1985, Industrial applications of fuzzy control, Elsevier Science Pub. Co.
[25] Vesely, W.E., Goldberg, F.F., Roberts, N.N. Haasl, D.F. 1981, Fault Tree Handbook,
Systems and Reliability Research, in U.S.N.R. Commission (ed.) vol. Nureg-0492,
Washington, D.C.
[26] Wang, Y.M., Chin, K.S., Poon, G.K.K. Yang, J.B. 2009, 'Risk evaluation in failure mode
and effects analysis using fuzzy weighted geometric mean', Expert Syst. Appl., vol. 36, no. 2,
pp. 1195-1207.
[27] Wang, Y.M., Yang, J.B., Xu, D.L. Chin, K.S. 2006, 'On the centroids of fuzzy numbers',
Fuzzy Sets Syst., vol. 157, no. 7, pp. 919-926.
[28] Yang, G. 2007, 'Potential failure mode avoidance', in Life Cycle Reliability Engineering,
John Wiley Sons, Hoboken, NJ, pp. 194-235.
[29] Zadeh, L.A. 1965, 'Fuzzy sets', Inform. Control, vol. 8, pp. 338-353.
[30] Zadeh, L.A. 1978, 'Fuzzy sets as a basis for a theory of possibility', Fuzzy Sets Syst., vol. 1,
no. 1, pp. 3-28.
[31] Saleh Alawi Ahmad, Usama H. Issa, Moataz Awad Farag and Laila M. Abdelhafez,
“Evaluation of Risk Factors Affecting Time and Cost of Construction Projects in Yemen”,
International Journal of Management (IJM), Volume 4, Issue 5, 2013, pp. 168 - 178,
ISSN Print: 0976-6502, ISSN Online: 0976-6510.
[32] S.G.Umashankar and Dr.G.Kalivarathan, “Prediction of Transportation Specialized Views
of Median Safety by using Fuzzy Logic Approach”, International Journal of Civil
Engineering Technology (IJCIET), Volume 4, Issue 1, 2013, pp. 38 - 44, ISSN Print:
0976 – 6308, ISSN Online: 0976 – 6316.
[33] Er. Amit Bijon Dutta and Dr. M. J. Kolhatkar, “Study of Risk Management in Construction
Projects”, International Journal of Management (IJM), Volume 5, Issue 6, 2014, pp. 32 - 39,
ISSN Print: 0976-6502, ISSN Online: 0976-6510.