Effects Protection SystemHidden Failures On Bulk .
Effects of Protection System Hidden Failures onBulk Power System ReliabilityFang Yang, Student Member, IEEE, A. P. Sakis Meliopoulos, Fellow, IEEE, George J. Cokkinides,Member, IEEE, and Q. Binh Dam, Student Member, IEEEAbstract - Protection System hidden failures have beenrecognized as a contributing factor to power system cascadingoutages. However, in the current bulk power system reliabilityassessment practice, protection systems are generally assumed tobe perfect, and the effects of hidden failures in protection systemsare not taken into account. In this paper, the impact of protectionsystem hidden failures on bulk power system reliability isinvestigated. A breaker-oriented bulk power system networkmodel is developed to include detailed system substationconfigurations and associated protection system schemes.Protection system constituents, such as transducers, relays,circuit breakers, may suffer from hidden failures. Hiddenfailures existing in transducers and relays can be detected by theadvanced system real time monitoring and analysis technologies.Thus, the major concern of this work focuses on the analysis ofhidden failures in circuit breakers. The hidden failure effectsanalysis shows that some initial system disturbances can result inthe unnecessary outages of intact power system equipmentbecause of hidden failures in circuit breaker trip mechanisms.Contingencies resulting from hidden failure outages are furtherevaluated by a security-constrained adequacy evaluationmethodology to obtain their influence on system reliability. Theproposed analysis procedure is demonstrated with a breakeroriented 24-substation reliability test system, which is developedbased on the IEEE 24-bus reliability test system and integratesexplicit substation and protection system models in the networkmodel. Evaluation results show that protection system hiddenfailures downgrade the system reliability level because they leadto the outages of undamaged equipment following initial systemdisturbances.Index Terms - Hidden failure, protection system, circuit breakeroriented substation model, bulk power system reliability assessment,advanced real time power system monitoring and analysis, securityconstrained adequacy evaluation.vvXI. INTRODUCTIONJN bulk power system plannig and operaThon, an importantiprocedure is reliability assessment [1]. The current practicein reliability assessment mainly includes N-i contingencyanalysis for independent and common mode equipmentoutages, and N-2 security analysis is performed only for someThis work was supported in part by the Power Systems EngineeringResearch Center (PSERC).Fang Yang, A. P. Sakis Meliopoulos, George J. Cokkinides, and Q. DamBinh are with the School of Electrical and Computer Engineering, GeorgiaInstitute of Technology, Atlanta, GA 30332-0250 USA ( e-mail:gtg65 1 jamail.gatech.edu,sakis.me1iopou1os a ece.gatech.edukiie. om as1nt ibama .gcheu.1-4244-0228-X/06/ 20.OO 2006 IEEEcredible cases. Most of contingencies that involve multiplecomponent outages are considered as a result of severalindependent succeeding events in the system and usually not amajor concern in the reliability assessment procedure.However, recent research [2-4] shows that protection systemhidden failures may cause multiple component outages thatare dependent upon each other, i.e., an initial componenttr i ofu oteoutagenca lead o the cascadin systemom entsrbecasof the prtenmalfuncTherefore, protection system hidden failures have beencrecognized as a contributing factor in spreadingpower systemdisturbances and even causing system blackouts. Sinceprotection systems are generally assumed to be perfect whenconsidering bulk power system reliability, the effects ofprotection system hidden failures are not considered in thecurrent practice of bulk power system reliability assessment.Hidden failures in protection systems are defined [5] as "apermanent defect that will cause a relay or a relay system toincorrectly and inappropriately remove a circuit element(s) asa direct consequence of another switching event." In otherwords, hidden failures remain hidden during normal systemconditions, and when system disturbances occur, such asfaults or overloads, hidden failures are exposed and causeunnecessary outages. The existence of hidden failures inprotection systems makes the stressed system situation evenworse and reduces the system reliability level.Protection systems consist of many components, such astransducers (current and voltage transformers), relays, circuitbreakers, and so on, which contribute to the detection andremoval of faults [6]. Hidden failures may exist in any ofthese constituents. Most of present research focuses onstudying hidden failures in relays. For example, themechanism and consequence of some possible hidden failuremodes in various relays used for transmission systemprotections are analyzed in [7, 8]. On the other hand, theanalysis of hidden failures in other protection systemcomponents, such as transducers and circuit breakers, has notreceived much attention.The application of various intelligent electronic devices(IEDs) in power system substations, such as phasormeasurement units, digital protection relays, and so on, hasmade it possible to implement advanced real time systemmonitoring and analysis technologes [9-1 1], based on whIchthe transducer outputs and relay settings can be validated andverified on line. However, such technologies can not examinethe circuit breaker's ability to trip, even though the trip coil. .517circuit may be monitored. Hiddcen failures in th1e circuitbreaker trip mechanism can cause circuit breakers fail to open.Therefore, this work concentrates on analyzing the impact of
hidden failures in circuit breaker trip mechanisms on bulk impact of protection system hidden failures on bulk powerpower system reliability.system reliability is investigated.In this paper, a framework of bulk power system reliabilityBusAassessment, i.e., the security-constrained adequacy evaluation(SCAE) methodology, is extended to include the effects ofprotection system hidden failures on bulk power systemProtection System1CB1I --4-.---------------reliability. The detailed SCAE methodology can be found incB6TOIreferences [12-14]. It incorporates operational practices andLL4simulates contingencies efficiently in a realistic manner. In the-CT1extended SCAE framework, the hidden failure effects analysisA.x.0-- 1E , CT2 \/T,is performed on each power system substation to obtainpossible contingencies resulting from hidden failure outagesCB25as well as their conditional probabilities given the occurrenceof initial system disturbances. Hidden failure contingenciesL3L2and other system contingencies are then subject to the threeSystem2steps of SCAE methodology:(a) contingencyXPtedioncomputation. The effects of protection system hidden failureson bulk power system reliability are demonstrated with abreaker-oriented 24-substation reliability test system CB34 -T-3-.CT3T4CB4; I TC4C---------------------------------11. METHODOLOGYBus BThis section describes the proposed methodology that canconsider the effects of protection system hidden failures onbulk power system reliability.Figure 1. A breaker-and-a-half bus arrangement substation modelB. Protection System Hidden FailuresA. Breaker-Oriented SubstationEach alrsdof protectionnisystemnisihrnsuffer fromcomponents mayehns.SmSince protection systems are assumed to be perfect in the hdeigcurrent bulk power system reliability assessment procedure, pible hiddenhof major components in protection'oos1beh1denfailuressystem substations are generally simplified to buses, and Xdifferent transmission lines simply converge at buses to systems are briefly analyzed below.connect generators or to serve loads. To consider hiddenCurrent Transformer (CT)failures in protection systems, we develop the breakeroriented substation model [15] in this study. Such modelAfter a fault occurs, fault currents may lead to currentprovides the substation configuration by converting each bus transformer saturation,. in which the secondary current of theof the power system into a substation with specific bus'arrangements (breaker and a half, ring, and so on). The currenttselection of bus arrangements follows usual design proceduresand practices. The substation models are then an integral partVoltage Transformer (VT)of the bulk power system network model and reflect the reallife existence of substation configurations. Figure 1 shows anFor some voltage transformers, such as coupling capacitorexample breaker-oriented substation model with a breakerageand-a-half bus arrangement, which consists of six circuittransformers (CCaVTs), theoutput of the voltagetransformermaybe(C1orinoigotonsignificantly different from the actualbreakers (CBI1 to CB6) and four incoming/outgoing piayvlaeatrafuttransmission lines (LI to L4).The breaker-oriented substation model adds a new level ofdetail in the network model, based on which the protectionsystem schemes for various power system components can beWhen the system operating condition changes, such asintroduced into the networkmodel. For instance, twosysroducedintotheneteorkmodesign.aoverload occurs a relayy may fail to detect the systemexamples of the protection system design are shown in Figure systemy'statuscorrectlyand trip the intact yssystem components asaline.1. Protection systems 1 and 2 are to protect transmission l rneLI and Bus B, respectively. Major constituents in these twoprotection systems include current transformers (CT1 to CT4)a voltage transformer (VT), relays (Ri and R2), and circuitbreakers (CB 1 to CB4) associated with trip coils (TC1 toAnfalrsithtipmcnsmoacrutbekr,uhTC4). The detailed substation and protection system modelsmake it possible to study the impact of protection systems on a h pncruto rpciso h ici rae ltsfipowe sysemprforance Spcifially in his ork,the to separate because of welding, will lead to circuit breaker fai' to trip.sbreakers toC6.n518
C. Impact ofAdvanced System Monitoring andAnalysisTechnologiesNowadays, besides conventional RTUs, a variety ofintelligent electronic devices (IEDs) are available in powersystem substations, including phasor measurement units,digital protection relays, and so on. Compared to thelimitation, inaccuracy, and delay in traditional SCADA data,more redundant, accurate, and real time system data can beobtained from these IEDs. Based on IED measurementssubstation level system information extraction functions, suchas substation level state estimation and alarm processing, cansignificantly improve the capability of system real timemonitoring and analysis.The advanced system monitoring and analysis functionsmake it possible to perform real time validation andverification for transducer outputs and relay settings. If anyhidden failures exist in CTs or VTs that cause their outputsfail to reflect actual primary system statuses, the substationlevel state estimation based on IED measurements can identifybad data in a fast and reliable way. In addition, based on thereal time synchronized measurements of system states, the realtime system model can be built and relay settings can beverified to avoid misoperation caused by relay outdatedsettings. Therefore, the advanced system monitoring andanalysis capability brought by the application of IEDs candetect hidden failures existing in transducers and relays inprotection systems considerably. However, such techniquesdo not detect the circuit breaker's ability to trip. Hiddenfailures in the circuit breaker trip mechanisms will remainuncovered until circuit breakers fail to open during systemdisturbances. Hence, in this work, the consideration ofprotection system hidden failures concentrates on hiddenfailures in the circuit breaker trip mechanism.D. Probabilistic Modeling of Hidden Failures in the circuitbreaker trip Mechanism (CBTM)This section presents the probabilistic modeling of hiddenfailures in the CBTM. For the example substation modelshown in Figure 1, which has six circuit breakers, theindependent and common mode hidden failure models of thecorresponding CBTMs are described below.Independent Hidden Failures of CBTMsEach CBTM can cycle between normal and hidden failurestatuses. This process can be modeled as a two-state Markovprocess with constant transition rates. We assume that theoccurrences of such hidden failures are independent with theirown failure and repair rates. Two-state Markov models for thesix CBTMs (CBTM1 to CBTM6) are shown in Figure 2, inwhich Ax and ,px represent failure and repair rates of eachCBTM, ,HFHF(,ECBTM4t 3CBTM5CBTM6242526P4V,HFP5V,JVCHFP6HFigure 2. Two-state Markov models of CBTMsThe differential equations that govern transitions for eachCBTM between the normal and hidden failure statuses are:dp (t)(t)A(1)dtwhere px (t) is the row vector that contains normal and hiddenfailure status probabilities (i.e., px(t) and qx(t) ) of eachCBTMx,Px (t) [Px (t),qx (t)](2)Also, the probabilities of normal and hidden failure statusesfor each CBTM satisfy the following condition:Px (t) q() 1,(3)where o px (t) 1 and 0 qx (t) 1In addition, A is the transition intensity matrix for CBTMX,xi.e.,KAx Ax]A(4)L/x -1'xiThe initial condition represents a CBTM with the probabilityof normal status set to one and the probability of hiddenfailure status set to zero.Px (O) [1 01.The solution to the above differential equations gives theprobabilities of normal and hidden failure statuses of eachCBTMx: ix exp(-(Ax px)t),x AxXexp(-(Ax px)t).qx(t) 1-p(t) A¾X Ax ¾ XPx (t)Ax X(5)If only long-term status probabilities are of interest, thenormal and hidden failure status probabilities of each CBTMare expressed as following:PxX(oo519Xx( (6)
For the substation shown in Figure 1, each combination ofthe operational statuses of six CBTMs constitutes a substationstate. The different combinations of CBTM statuses generate atotal of 64 (26) substation states in its state space. Part of thisstate space (states 1 to 16) is shown in Table 1.Table 1 Partial State Enumeration for the Example Substation (States 1-16)1234567811121314 15 16of failure and repair rates of all transitions from the currentstate to an adjacent state (transition rates found in the samerow). The sum of all elements in each row of the transitionmatrix is zero. The initial state is assumed to be at the state 1in Table 1, in which every component works in the normalstatus. The solution to the differential equations gives theprobability of each substation states at any instant of time.E. Hidden Failure Effects AnalysisHidden failures in CBTMs can cause the trip of intactCBTM3XXXxxxCBTM4equipment following system disturbances, which reduces thexxxCBTM5system reliability level. In this section, an approach of hiddenxxx1CBTM6Ix failure effects analysis for each system substation is proposedIX indicates a hidden failure status ofthe CBTMto obtain possible hidden failure outages following any initialfaults. The proposed hidden failure effects analysis procedureBecause we assumepthat hide faiof CtMsare is illustrated by the example substation shown in Figure 1.independent, the probability of each substation state can be We assume that the substation is under state 3 as enumeratedobtained by multiplying the probability of each CBTM status. in Table 1. In state 3, the trip mechanism of circuit breaker 2For example, for substation state 3 in Table l, in which (CB2) has hidden failure that can cause CB2 fail to open. If anCBTM2 is in the hidden failure status, all others are in the initial fault Ft occurs on transmission line LI as shown innormal status, the probability for this substation state is Figure 1, circuit breakers 1 and 2 should open to isolate thecalculated as follows:faulty circuit LI accordingly. Since CB2 fail to open due to itshidden failure, circuit breaker 3 (CB3) that is adjacent to CB2(7)will open and then cause the outage of intact componentPS3 ptPCBTm 1tqC 2 (t)}7J PCB (t)1 3transmission line L2 following the initial fault on transmissionline LI. The conditional probability of the hidden failureThe sum of the substation state probabilities is forced to 1 outage of L2, given the incidence of the initial fault on LI, issimply because the total 64 states are mutually exclusive and the occurrence probability of substation state 3 as calculatedin Equation (7). Such effects analysis procedure can betheir union forms the certain event, i.e.,repeated for all other possible initial faults. The results,64including initial faulty circuits, associated hidden failure8Z Ps() 1(8)outages, and corresponding conditional probabilities are listedk lin Table 2. We can see that the initial faults on LI, L2, BusI,or Bus2 will cause outages of intact equipment in thesubstation under state 3, while the initial faults on otherdo not cause any hidden failure outages. For thecomponentsThe independent hidden failure model of CBTMs preventsthere are total of 64 states, a cut-offexamplesubstation,considering common mode failures that involve simultaneouscanbeto reduce states in the stateprobabilitypredefinedhidden failures of two or more CBTMs as the result of a singlethataresubecttohiddenfailure effects analysis.Thespaceyoutage event. For example, a loss of the power obabilityto twoormoretripcstateswhich supplies power to twowhichor suppliesmore trip coils,can causedue to their small incidence. Furthermore, themultiple circuit breakers the enter hidden failure status be considered effectshiddenfailureanalysis procedure can be performed forsimultaneously. When considering common mode hiddenstateand all substations in the system.failures of CBTMs in the substation shown in Figure 1, every substationdifferential equations that govern the transitions amongsubstation states, initial conditions, and sum of probabilitiesInitial FaultHidden Failure Conditionalare given below:OutageProbabilityCBTMICBTM2XXx x x x xXX X X Xpowerdps(t) Ps(t).AsdtPs ( ) [I 0 064Zk iPsk (t) o.0](Where pS(t) is the row vector of substation state probability,and A is the substationtransition intensity matrix, Offsdiagonal terms (i, j) of matrix A have the failure/repair ratefrom state i to state]j, and diagonal terms are the negative sum520Fault on Bus AFault on LI or Bus 1Fault on L2 or Bus 2Fault on Bus BFault on L3 or Bus 4N/AL2 and Bus 2L2 and Bus 1N/AN/AN/AFault on L4 or Bus 5N/AN/APS3Ps3N/AN/AF. Security-Constrained Adequacy Evaluationfor BulkPoe ytm eibltorder to assess bulk power system reliability, the Inseuiycntaeddqayevlton(CBmethodology is applied to evaluate the system contingencies
resulting from independent outages, common mode outages, system has been replaced with a substation with specific busand hidden failure outages. This section briefly describes the arrangement (ring, breaker and a half, and so on). The busSCAE methodology, the detailed information about this arrangement at each node and the location of each circuitbreaker become the explicit part of the network model.methodology can be found in [12-14].As an example, bus 180 of the original IEEE 24-bus RTS,The SCAE methodology, which is implemented based onanalytical techniques, encompasses three main steps: (a) which connects to one unit and four circuits, is replaced withcritical contingency selection, (b) effects analysis, and (c) the substation 180 of a mixed breaker-and-a-half and doublereliability index computation. To improve the accuracy and breaker scheme as illustrated in Figure 4. The overallefficiency in contingency selection and effects analysis, an conversion procedure from the original bus-oriented system toadvanced power flow model, i.e., the single phase quadratized the breaker-or
advanced system real time monitoring and analysis technologies. considering bulk power system reliability, the effects of Thus, the major concern ofthis workfocuses on the analysis of protection system hidden failures are not considered in the hidden failures in circuit breakers.
Demand-side mark et failures happen when de-mand curves do not reflect consumers' full willing-ness to pay for a good or service. Suppl y-side market failures occur when supply curves do not reflect the full cost of producing a good or service. Demand-Side Market Failures Demand-side market failures arise because it is impossible
13.29 % failures were Class IV, 12.65 % failures were identified as class II, 12.02 % failures as class V and 5.69 % failures were categorized in class I failure. Conclusion: Though earlier literature reported caries as the most
a) The two aspects of market failures are demand-side market failures and supply side market failures. b) Demand-side market failures: They occur when the demand curves do not take into account the full willingness of consumers to pay for a product. Eg.: None of us will be willing to pay to view a wayside fountain because we can view it without .
Coiled Tubing Bias Welds Recent Failures Trend October 29 th, 2014 ICoTA Roundtable, Calgary, Canada Tomas Padron, Coiled Tubing Research and Engineering (CTRE) Bias Weld Failures - Outline . Great Yarmouth P2216 3 2014 ICoTA Roundtable - Bias Weld Failures 4.
3. Current Transformer Testing . Testing current transformers helps to detect: 3.1. Installation related failures . Transportation damages, wiring errors and manufacturing defects. 3.2. In-service related failures . Degradation of accuracy class, shorted turns, magnetized core, burden failures in secondary circuit and insulation material .
Failure Modes Potential Causes of Failures Potential Effects Recommendedof Failures Current Controls S e v D t R P N . For each potential failure – identify the effects of that failure mode . – Applying failure mode effects and criticality analysis in .
Fundamentals of Protection Protection System – A complete arrangement of equipment that fulfills the protection requirements Protection Equipment – A collection of devices excluding CT, CB etc Protection Scheme – A collection of protection equipment providing a defined function. 34! Zones of Protection
patience and understanding during the long and comprehensive revision process. We believe you will find it was well worth the wait. Deborah E. Wilson, DrPH, CBSP L. Casey Chosewood, M.D. Director Director Division of Occupational Office of Health and Safety Health and Safety Centers for Disease Control National Institutes of Health and Prevention Bethesda, Maryland Atlanta, Georgia September .
