Fuzzy methodology applied to Probabilistic Safety Assessment for digital system in nuclear power plants

  • Published on
    11-Sep-2016

  • View
    213

  • Download
    1

Transcript

<ul><li><p>Nuclear Engineering and Design 241 (2011) 3967 3976</p><p>Contents lists available at ScienceDirect</p><p>Nuclear Engineering and Design</p><p>j ourna l ho me page: www.elsev ier .com/</p><p>Review</p><p>Fuzzy min nucl</p><p>Antonio a Instituto de E75 Zip Code 2b Programa de c Instituto Naci</p><p>a r t i c l</p><p>Article history:Received 2 SepReceived in reAccepted 25 Ju</p><p>Contents</p><p>1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39672. Description of digital feedwater control system (DFWCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39683. Descr4. Applic</p><p>4.1. 4.2. </p><p>5. Result6. Concl</p><p>AcknoRefer</p><p>1. Introdu</p><p>Traditionmonitoringexisting plaplant desigNuclear Reresearch plasupporting </p><p> CorresponE-mail add</p><p>lapa@ien.gov.b</p><p>0029-5493/$ doi:10.1016/j.iption of fuzzy inference system approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3968ation of the proposed approach to DFWCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3970Fuzzy membership function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3971Fuzzy rule base application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3972s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3974usion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3976wledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3976</p><p>ences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3976</p><p>ction</p><p>ally, nuclear power plants (NPP) have functions for, protecting and control that use analog systems. Soments have replaced current analog systems while newns have fully incorporated digital systems. The U.S.gulatory Commission (NRC) dened a digital systemn that establishes a coherent set of research programsregulatory needs. A deterministic engineering criterion</p><p>ding author. Tel.: +55 21 21733899; fax: +55 21 21733909.resses: tony@ien.gov.br, antonio.cesar@pq.cnpq.br (A.C.F. Guimares),r (C.M.F. Lapa), malu@ien.gov.br (M.d.L. Moreira).</p><p>is used to the current licensing process for digital systems. How-ever, at present, there are no consensus methods for quantifying thereliability of digital systems. The objective of the NRC digital systemrisk research is to identify and develop methods, analytical tools,and regulatory guidance to support (1) NPP regulatory decisionsusing information on the risks of digital systems, and (2) includingmodels of digital systems into NPP PRAs (Probabilistic Risk Assess-ments). An example of this type of method is the traditional EventTree/Fault Tree (ET/FT) approach that does not explicitly modelthe interactions between the plant system that is being modeledand the plant physical processes, nor the exact timing of theseinteractions. In the past few years, methods and tools for probabilis-tic modeling of digital systems were investigated. The reviewingwork in Chu et al. (2008) has analyzed operating experience,</p><p> see front matter 2011 Elsevier B.V. All rights reserved.nucengdes.2011.06.044ethodology applied to Probabilistic Safety Assessment for digital systemear power plants</p><p>Csar Ferreira Guimaresa,, Celso Marcelo Franklin Lapab,c, Maria de Lourdes Moreirab,c</p><p>ngenharia Nuclear (IEN), Diviso de Reatores, Via Cinco, s/n , Cidade Universitria, Rua Hlio de Almeida, Postal Box 68550,1941-906 Rio de Janeiro, BrazilPs-Graduac o em Cincia e Tecnologia Nucleares do IEN, Brazilonal de C&amp;T de Reatores Nucleares Inovadores, Brazil</p><p> e i n f o</p><p>tember 2010vised form 8 June 2011ne 2011</p><p>a b s t r a c t</p><p>A fuzzy inference system (FIS) modeling technique to treat a nuclear reliability engineering problem ispresented. Recently, many nuclear power plants (NPPs) have performed a shift in technology to digitalsystems due to analog obsolescence and digital advantages. The fuzzy inference engine uses these fuzzyIF-THEN rules to determine a mapping of the input universe of discourse over the output universe ofdiscourse based on fuzzy logic principles. The risk priority number (RPN) (typical of a traditional failuremode and effects analysis FMEA) is calculated and compared to fuzzy risk priority number (FRPN),obtained by the use of the scores from expert opinions. It was adopted the digital feedwater controlsystem as a practical example in the case study. The results demonstrated the potential of the inferencesystem to this class of problem.</p><p> 2011 Elsevier B.V. All rights reserved.locate /nucengdes</p></li><li><p>3968 A.C.F. Guimares et al. / Nuclear Engineering and Design 241 (2011) 3967 3976</p><p>developing failure rate estimates using Hierarchical Bayesian anal-ysis, and performing failure modes and effects analyses (FMEAs)of digital systems. The experience has been acquired with theworks presented in the review. The NRC together with the BNL(Brookhaveuse of tradimentation this paper.</p><p>Failure mnique (Stamor potentiasystems anmanagemeure mode a(KBS) to esttem (Zadehwith fuzzy fuzzy logic.can be classand Sugenoand defuzziwas the fuzin most invand Guimarit was deveto scale anyinference syquanticati</p><p>The knowitative reprand incorpintuitive th</p><p>In this arital feedwawater reactet al., 2008trollers, depto gain a furelevant comnents and tdeterminedtic Safety Adata are abscal initiatinthe experts</p><p>The threied for thisTop-Level (DFWCS systhe DFWCSMain Feedw(BFV) contrferential Indrelated to tlowest levedata were a</p><p>In this pcontroller, aing the propof Module) the Main CP</p><p>The newnamed Fuzzadherence oalso endorsmodel the u</p><p>2. Description of digital feedwater control system (DFWCS)</p><p>The test case for applying the approach proposed involves adigital feedwater control system (DFWCS) of a two-loop pressur-</p><p>ater rr coosyste</p><p> (FW, maregusociaary in Chn of ing fionalusesdwa</p><p>t thethe plly c</p><p> the p purmat</p><p>ed tos. Se</p><p> feed plandes aing oand preselves</p><p> showDFWS areFWPs</p><p> DFWontroWCS) anvel, fcontr</p><p> onlyS. Syption000 es of</p><p>crip</p><p> purnsisce eng fry seogic q. (1</p><p>IF x1</p><p> Flia</p><p>nd oe hasrameF-THn National Laboratory) has conducted research on thetional reliability modeling methods for digital instru-and control (I&amp;C) systems, which is one the subject of</p><p>ode and effects analysis (FMEA) is an important tech-atis, 1995) that is used to identify and eliminate knownl failures to enhance reliability and safety of complexd is intended to provide information for making risknt decisions. It is being proposed here a modied fail-nd effects analysis (FMEA) and knowledge base systemimate the risk using scores from experts. Fuzzy logic sys-, 1987) is a name for the systems that have relationshipconcepts (like fuzzy sets and linguistic variables) and</p><p> The most popular fuzzy logic systems in the literatureied into three types: pure fuzzy logic systems, Takagis fuzzy system, and fuzzy logic systems with fuzzierer (Wang, 1993). The methodology used in this paperzy logic systems with fuzzier and defuzzier, as usedestigations, e.g. Pillay and Wang (2003), Xu et al. (2002)es and Lapa (2004a,b). In Guimares and Lapa (2006),loped a methodology which uses risk priority number</p><p> parameter characteristics of the system and a fuzzystem for estimating risk from expert opinion about theon of the variables.ledge-based fuzzy systems allow a descriptive or qual-</p><p>esentation of expressions such as remote or high,orate symbolic statements that are more natural andan mathematical equation.ticle, it is applied the new approach and concept to dig-ter control system (DFWCS) of a two-loop pressurizedor (PWR). The DFWCS was analyzed in detail in (Chu), including its function, components, associated con-endencies and interfaces, and digital features, in orderll understanding of the way the DFWCS and each of its</p><p>ponents operate. The failure modes of DFWCS compo-he impact of each of them on the system function were</p><p> by performing an FMEA. In the traditional Probabilis-ssessment (PSA) study, operational, historic and failureolutely necessities, however using the FMEA, the criti-g events and its consequences can be determined from opinions.e different levels of detail of the FMEA that can be stud-</p><p> system are dened as: The rst level FMEA, namedfor the system level), includes the analysis of the wholetem. The second level of the FMEA includes modules of, with the major ones being the Main CPU, Backup CPU,ater Valve (MFV) Controller, Bypass Feedwater Valve</p><p>oller, Feedwater Pump (FWP) controller, Pressure Dif-ication (PDI) controller, and the optical isolator that ishe watchdog timer (WDT) signal. The third level (thel or components) is the one in which more probabilisticvailable from publicly available sources.aper, the rst level (named Top-Level) as well as FWP</p><p> main module of the DFWCS, was considered for apply-osed approach. In a future paper the second level (Levelcombined with the third level (Level of Components) ofU will be considered.</p><p> approach proposed combining FMEA and Fuzzy Logic,y FMEA, and the set of results demonstrated the greatf the Fuzzy FMEA approach to this kind of problems. Ites the advantages of using a fuzzy inference system toncertainty parameters levels in risk analysis.</p><p>ized wreactowater Pumpssystemwater and asseconddetail functioprovidoperattion cathe feeso thaWhen maticaunlessFor theof autoassumimpactdigitalpowerure moconsistpump Fig. 1 ow vawhichto the the FWof the by theually cthe DF(215%(S/G lelevel) showsDFWCDescri53MC5Featur</p><p>3. Des</p><p>Thebase coinferenmappito fuzzfuzzy lform E</p><p>R(l) : </p><p>whereinput aPracticnient ffuzzy Ieactor. Each of the two reactor-coolant loops contains alant pump and a steam generator (S/G). The main feed-m (FWS) consists of steam-turbine-driven FeedwaterPs), minimum ow control valves, a pump-seal waterin feedwater regulating valves (MFRVs), bypass feed-lating valves (BFRVs), high-pressure feedwater heaters,ted piping and instrumentation. The feedwater of eachloop is controlled by a DFWCS, which is described inu et al. (2008) in Section 4. During plant operation, thethe FWS is to remove heat from the primary system byeedwater to the S/Gs. Degradation that exceeds certain</p><p> parameters or total loss of the FWS during this opera- a reactor trip. The plant contains two secondary loops;ter in each loop is controlled by an identical DFWCS,</p><p> analysis of a single DFWCS is applicable to the other.lant is in the power operation mode, a DFWCS auto-</p><p>ontrols the feedwater in its associated secondary loop,lant operators have set the DFWCS in the manual mode.pose of this study, failure of a DFWCS is dened as lossic and manual control of its related loop. This loss is</p><p> cause a reactor trip because it can result in undesiredction 4 provides a functional and physical overview of awater control system (DFWCS) of an operating nucleart (NPP). The information is the basis for performing fail-nd effects analysis (FMEA). The NPP has two units, eachf two reactor coolant loops. There are a reactor coolanta steam generator (S/G) for each reactor coolant loop.nts a simplied diagram of the FWS (without mini-</p><p>, seal water system, and high pressure feedwater heater)s the location of some of the sensors that provide inputCSs of the two FWS trains. Note that the two trains of</p><p> aligned together at the discharge as well as the suction. The sensors from the reactor coolant loops are sharedCSs of the two FWS trains. Typically, the FWS is man-lled below 2% power and automatically controlled by</p><p> above 2%. It has two automatic modes of operation, lowd high (above 15%) power, operating in three-elementeedwater ow, and steam ow) and single-element (S/Gols, respectively. In Fig. 2 is a simplied diagram that</p><p> one of the reactor coolant loops with its associatedstem level description, Control Modes and algorithms,</p><p> of Azonix MAC 7000 Controllers and Fischer &amp; PorterControllers, Dependencies and Interfaces, and Digital</p><p> the DFWCS, can be found in detail in Chu et al. (2008).</p><p>tion of fuzzy inference system approach</p><p>e fuzzy logic system is the system where the fuzzy rulets of a collection of fuzzy IF-THEN rules, and the fuzzyngine uses these fuzzy IF-THEN rules to determine aom fuzzy sets in the input universe of discourse U Rnts in the output universe of discourse V R based onprinciples. The fuzzy IF-THEN rules are of the following):</p><p>is Fl1 and . . . xn is Fln, THEN y is G</p><p>l (1)</p><p>nd Gl are fuzzy sets, x = (x1, . . ., xn)T U and y V areutput linguistic variables, respectively, and l = 1, 2, . . ., M.</p><p> shown that these fuzzy IF-THEN rules provide a conve-work to incorporate human experts knowledge. EachEN rule (Eq. (1)) denes fuzzy set Fl1x . . . xF</p><p>ln Gl in the</p></li><li><p>A.C.F. Guimares et al. / Nuclear Engineering and Design 241 (2011) 3967 3976 3969</p><p>Fig. 1. A simplied diagram...</p></li></ul>

Recommended

View more >