Professional Documents
Culture Documents
com
ScienceDirect
Procedia Engineering 97 (2014) 2323 2332
Abstract
The aim of this research work is to study the behaviour of the CHU (compressor house unit) of a medium size coal fired thermal
power plant using fuzzy - approach (quantitative approach). The system has been modelled using PN and various reliability
indices viz. failure rate, repair time, MTBF, ENOF, Availability and reliability are computed at different spread/uncertainty level.
The results are helpful to the system analyst to analyse the behaviour of the system and to plan suitable maintenance policy for
improving the system availability.
2014
2014TheTheAuthors. Published
Authors. by Elsevier
Published Ltd. This
by Elsevier is an open access article under the CC BY-NC-ND license
Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014.
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014
Keywords: Fuzzy methodology (FM); fuzzy - approach; Petri net (PN); Mean time between failure (MTBF)
1. Introduction
The goal of maximum generation and long run availability of the system may be obtained by making the system
failure free which is an inescapable phenomenon in an industrial system. Thus to achieve high availability of the
system it is necessary for the system analyst to emphasis on purging of sudden failure which results in heavy
generation loss .CHU (compressor house unit) of the coal fired thermal power plant is a complex repairable
industrial system where it is necessary to analyse the system behaviour for maintaining its high availability. From
the survey it has been noted that lot of research has been done by different researcher in the direction of behaviour
analysis of various units of different process industries (running 247 hrs) using probabilistic approach. Arora and
Kumar [1] dealt with the availability and MTBF and discuss the availability analysis of steam and power generating
system in the thermal power plant using Markov birth death approach. Arora and Kumar [2] again presented the use
of the Markov approach for the stochastic behaviour analysis and maintenance planning of the ash handling system
_________________
1877-7058 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014
doi:10.1016/j.proeng.2014.12.330
2324 Dilbagh Panchal and Dinesh Kumar / Procedia Engineering 97 (2014) 2323 2332
of a thermal power plant. The probabilistic approach so applied is inadequate to address the uncertain and vague
failure/repair data. Thus to handle such uncertainty in the failure/repair data FM has been used by the various
researchers in different field. Knezevic and Odoom [3] dealt with the reliability of the repairable system in which
fuzzy methodology and fuzzy lambda tau technique is used for calculating the various reliability indices. Sharma et
al. [4] presented the use of FM for the behaviour prediction of forming unit of the paper mill. Presently FM has been
applied successfully in following field like Structural reliability [5-6]; Safety and risk engineering [7]; Fault
diagnosis [8-9]; Software reliability [10]; Human reliability [11]. Thus from literature survey it is observed that FM
has not been applied in the CHU unit of the thermal power plant industry. Thus an attempt has been made by the
author to apply the FM for the reliability analysis of the considered system.
In the present study static part of the PN has been used to analyze the system behaviour (with assumptions that the
transitions are not timed). The PN model corresponding to logical OR / AND gate is shown in fig1.
T1 T1
T2 T3 T4 T2 T3 T4
(a) (b)
!
(5)
The basic arithmetic operation like addition, subtraction, multiplication and division etc. used for calculating the
fuzzy expression are extended through the use of extension principle [16-17]
4. Methodology
Fuzzy - methodology is used to compute the various reliability parameters of the considered system. The
procedural steps involved in this approach are shown in fig 3.
2326 Dilbagh Panchal and Dinesh Kumar / Procedia Engineering 97 (2014) 2323 2332
Expert knowledge
Historical records and reliability database
Log book record
System /reliability analyst
I R
N E
F L
O I
R A
M B
A Fuzzy input Defuzzified values I
T L
I Fault tree and I
O Fuzzifier Petrinet model Defuzzifier T
N Y
Crisp input I
E
X N
Fuzzy output D
T
R I
A Construct fuzzy
C
Obtain reliability I
C indices function in reliability indices MF
T using arithmetic E
the form of and
I operation S
O
N
Intercooler
Air dryer
Moisture remover system
Cleaner Heater
(a) (b)
Fig 5: (a) Fault tree model (b) Petrinet model
2328 Dilbagh Panchal and Dinesh Kumar / Procedia Engineering 97 (2014) 2323 2332
Step1.Information collection phase: Failure/repair data related to each component of the system is collected from
the CHU maintenance log book and verified with the maintenance field/ maintenance manager expert given in
table1.
Table 1.Failure/repair data for CHU
Component Failure rate (i) Repair time (i)
(Failures/hr) (hrs)
High pressure side (HPS), Low pressure side 40
(LPS)
Intercooler 15
Air tank 10
Air dryer, Heater 12
Cleaner 12
Step2.Computation phase: Failure/repair data collected from the various sources with ambiguity /vagueness are
translated into triangular fuzzy numbers using TFN membership, with 15% spread on the crisp input value (fig. 6
for HPS of the considered system).The - expression for conventional AND/OR expression is given in table
2 and OR /AND gate transition expression are represented by equation 6, 7 and 8.
! !
40
n-Input gate
expression
Dilbagh Panchal and Dinesh Kumar / Procedia Engineering 97 (2014) 2323 2332 2329
(6)
(7)
(8)
Step3. Estimation of reliability parameters: Using expression as given in table 3 various reliability parameters
viz. Failure rate, repair time, MTBF, MTTR, ENOF, availability and reliability are computed with right and left
spread at different alpha cut/membership grades (the range lies within 0-1, with an increment of 0.1) and shown by
graphs in fig 7(a,d) for 15% uncertainty.
Table 3: Various reliability parameter
Reliability indices Expression
Mean time to failure
Mean time to repair
Mean time between failure
Availability
Reliability
Expected number of failures
1 1
0.9 L.S 0.9 L.S
Degree of Membership
Degree of Membership
1 1
0.9 0.9 L.S
0.8 0.8
Degree of Membership
L.S R.S
Degree of Membership
Degree of Membership
0.8 0.8 R.S
0.7 R.S
0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0.8000 0.8200 0.8400 0.8600 0.13000 0.15000 0.17000 0.19000
Reliability Unreliability
(c)
1 1
0.9 L.S 0.9 L.S
Degree of Membership
Degree of membership
Step4.Defuzzification: In order to make the maintenance decision with maintenance action defuzzifiaction is done
to convert the fuzzy output into crisp value using centre of area (COA) method. Further to quantify the system
behaviour more pragmatically the Defuzzified value for various reliability parameter are calculated at different
spreads( 15%, 25% and 60%) and is given in table 4. The % change in reliability parameters is given in table 5.
Table 4: Crisp and defuzzified value
7. Behavioural analysis
Table 5 shows the % change and increase/decrease trend for the various reliability parameters of the considered
system. From table 5 it is observed that the defuzzified value at different spread (15%, 25% and 60%) for failure
rate, repair time, reliability, MTBF and unavailability increase with the increase in spread where as availability,
ENOF and unreliability decrease with increase in spread. Based upon the observations the system analyst/
maintenance manager will choose a feasible defuzzified value rather than crisp value and recommend for the
revision of targeted goal.
8. Conclusion
With the study a quantitative approach for the CHU has been used for predicting the behaviour of the system. Effort
have been made to model the uncertain/vague information for the considered system and various reliability
parameters viz. failure rate, repair rate, MTBF, ENOF and availability are computed and the trend of reliability
parameters has been analysed for studying the failure behaviour of the considered system.
2332 Dilbagh Panchal and Dinesh Kumar / Procedia Engineering 97 (2014) 2323 2332
Acknowledgements
The authors thanks to Mr. Kharti Lal (maintenance helper-1, CHU) and Mr. D. Prasad (maintenance helper-2, CHU)
of coal fired thermal power plant dist. Panipat, Haryana, India for providing every possible help for the work.
References
[1]. N. Arora, D. Kumar, Availability analysis of steam and power generation systems in thermal power plant,
International Journal of Microelectronics Reliability, 37(1995) 795-799
[2]. N. Arora, D. Kumar, Stochastic and maintenance planning of ash handling system in thermal power plant,
International Journal of Microelectronics Reliability,37(1996) 5 819-834
[3]. J. Knezevic, E.R. Odoom, Reliability modelling of repairable systems using Petri nets and fuzzy Lambda-
Tau methodology, Reliability Engineering & System Safety, 73 (2001) 1, 1-17.
[4]. R.K. Sharma, D. Kumar, P. Kumar, Predicting uncertain behaviour of industrial system using FM: a
practical case, Appl. Soft Comp. 8 (1) (2008) 96109.
[5]. M. Savoia, Structural reliability analysis through fuzzy number approach, with application to stability,
Computers & Structures, 80 (2002) 12, 1087-102.
[6]. F. Biondini, F. Bontempi, P.G. Malerba, Fuzzy reliability analysis of concrete structure, Comput.
Struct.82 (2004)1033-1052.
[7]. H.S. Sii, S.H. Ruxton, J.T. Wang, A fuzzy logic based approach to qualitative safety modeling for marine
systems, Reliab. Eng. Syst. Saf.73 (2001) 1934.
[8]. T.S. Liu, S.B. Chiou, The application of Petri nets to failure analysis, Reliab. Eng. & Syst. Saf. 57(1997)
129-1428.
[9]. F. Mustapha, S.M. Sapun, N. Ismail, A.S. Mokhtar, A computer based intelligent system for fault diagnosis
of an aircraft engine, Eng. Comput.21 (1) (2004) 7890.
[10]. K. Popstojanova, K.S. Trivedi, Architecture-based approach to reliability assessment of software systems,
Perform. Eval. 45 (23), (2001) 179204.
[11]. M. Konstandinidou, Z. Nivolianitou, C. Kiranoudis, and N. Markatos, A fuzzy modeling application of
CREAM methodology for reliability, Reliability Engineering & System Safety, Vol. 91 No. 6, (2006) pp.
706-16.
[12]. C.A. Petri, Communication with automata. PhD thesis, University of Bonn, Technical Report (English)
RADC-TR-65-377, Giriffis (NY): Rome Air Development Centre; 1962.
[13]. T.S. Liu, S.B. Chiou, The application of Petri nets to failure analysis, Reliab. Eng. & Syst. Saf. 57(1997)
129-1428.
[14]. J.L. Peterson, Petri Net Theory and the Modelling of Systems, Englewood Cliffs, NJ: Prentice-Hall, 1981.
[15]. L.A. Zadeh, Fuzzy Sets, Fuzzy Logic, Fuzzy Systems: Selected Papers, World Scientific, Singapore,
1996.
[16]. H. Zimmermann, Fuzzy Set Theory and its Applications, third ed., Kluwer Academic Publishers, 2000.
[17]. B. Kokso, Fuzzy Engineering, Prentice Hall, Englewood Cliffs, NJ, 1999.
[18]. T.J. Ross, Fuzzy Logic with Engineering Applications, McGraw-Hill, NY, 1995.
[19]. K. Tanaka, An introduction to Fuzzy Logic for Practical Applications, Springer, New York, 2001.
[20]. B.S. Dhillion, C. Singh, Engineering Reliability: New Techniques and Applications, Wiley, New York,
1991.