integrated reliability index whitepaper draft
TRANSCRIPT
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
1/11
1
Integrated Reliability Index Concepts
Thedevelopmentofanintegratedreliabilityindexaimstoinform,
increasetransparency,andquantifytheeffectivenessofriskreduction
and/ormitigationactions. Thegoalistoprovidetheindustrymeaningful
trendsofthebulksystemperformanceandguidanceonhowtheycanimprove
reliabilityandsupportriskinformeddecisionmaking.
Underthe
direction
of
the
Operating
Committee
(OC)
and
Planning
Committee
(PC),
the
Reliability Metrics Working Group (RMWG) has developed a portfolio of eighteen (18)
Adequate Level of Reliability (ALR) metrics.1 At its March 2011 meeting, the Operating
Committee (OC) and Planning Committee (PC) approved the Severity Risk Index (SRI)2
calculationandsupportedthedevelopmentofanIntegratedReliabilityIndex(IRI),whichcanbe
constructedbasedontheriskmodelillustratedinFigure1. Thismodelattemptstocapturethe
universeofriskandlinksthatriskcalculationtothereliabilityofthebulkpowersystem.
Figure1RiskModelforBulkPowerSystem
1 http://www.nerc.com/docs/pc/rmwg/RMWG_AnnualReport6.1.pdf2http://www.nerc.com/docs/pc/rmwg/Integrated_Bulk_Power_System_Risk_Assessment_Concepts_Final.pdf
EventDrivenIndex(EDI)IndicatesRiskfrom
MajorSystemEvents
Standards/StatuteDriven
Index(SDI)Indicates
Risks
fromSevereImpactStandardViolations
ConditionDrivenIndex(CDI)
IndicatesRisk
from
KeyReliabilityIndicators
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
2/11
2
EventDriven Index (EDI)TheEventDriven Index (EDI)providesabasis forprioritizationofevents based on bulk power system integrity, equipment performance, and/or engineering
judgment. The event severity indicators can serve as a high value risk assessment tool.
Stakeholderscanusethetooltomeasuretheseverityoftheseeventsandevaluatedisturbance
history.The
relative
severity
ranking
of
events
considers
both
the
occurrence
of
an
event
and
its impact, inordertoquantifytheeventrisktoreliability. The indexEDI increases iftherisk
valueassociatedwithsignificanteventsisreducedoveratrendingperiod. TheEDIcalculation
andmetricaggregationaredescribedinAppendixA.
ConditionDriven Index (CDI)ConditionDriven Index (CDI) focusesona setofmeasurablesystemconditionstoassessbulkpowersystemreliability. These reliability indicators identify
factors that positively or negatively impact reliability and are earlypredictors of the risk to
reliabilityfromconditionbasedALRmetrics. Acollectionoftheseindicatorsmeasureshowfar
reliabilityperformanceisfromdesiredoutcome,andiftheperformanceisheadingortrending
in the preferred direction. The index CDI increases if the risk value associated with key
reliabilitymetrics
covering
major
risk
factors
to
reliability
is
reduced
over
atrending
period.
TheCDIcalculationandmetricaggregationaredescribedinAppendixB.
Standards/Statute Driven Index (SDI) The Standards/Statute Driven Index (SDI) measuresimprovementincompliancewithReliabilityStandards.
3 TheviolationsincludedinSDIallhave
high Violation Risk Factors (VRFs) and were rated as potentially severe reliability impacts.
Basedon these twodimensions,knownunmitigatedviolationsareweightedandnormalized
overthenumberofapplicableregisteredentitiessubjecttoaparticularstandardrequirement.
Theindexincreasesifthecomplianceimprovementisachievedoveratrendingperiod. TheCDI
calculationandmetricaggregationaredescribedinAppendixC.
As
this
time,
the
IRI
is
intended
to
be
used
as
a
historical
measure,
neither
a
real
time,
nor
forwardlookingperformance score. As such, the index iscalculatedusingactual fielddata.4
The value of the IRI can be calculated based on the assessed risk of the above three
componentsandtheirrelativeweightings,asshowninEquation1below:
(1)
where:
weighting of event component;
normalized Event Driven Index;
weighting of metric component;
normalized Condition Driven Index;
weighting of standard compliance component; normalized Standards Driven Index;
Integrated Reliability Index for a speciic period.
3Detailedstandards/statutedrivenindicatorscanbeviewedathttp://www.nerc.com/filez/pmtf.html.4Fromarealtimeorforwardlookingpointofview,onecanalsocreateotherindiceslinkingrealtimeorforecast
variablesandfactoringinprobabilities. Astheindustrygainsexperiencewiththisbackwardlookingmetricitwill
continuetoconsiderwhethercertainrealtimeorforecastmetricsarevaluableindicatorsofthereliabilityofthe
bulkpowersystem.
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
3/11
3
Thevalueof IRIwill range from0 to100,and canbeaggregatedatvarious levelsofdetail,
includingattheNERC,InterconnectionorRegionallevels. Thethreeweightscanbeadjustedas
we learn more and gain experience after one to two years of trending. At this time it is
suggestedthataweightingof
(2)
isusedinitiallyastheEDIreflectsactualexperiencesofthepowersystemwhileCDIandSDIare
morereflectiveofriskstothepowersystem.
ThefollowingfactorsaretobeconsideredwhendevelopingeachoftheIRIcomponents:
Covermajorityofrisks
Showaconsistentlinktoreliability
Considerrandomnessandindependence
Shouldunderstandthedegreetowhichanyofthethreeindicesmayrelatetootherindices
Conceptually,thecomponentsmightbeconsideredassuccessratesforeventsexperienced
(asmeasuredbyEDI),conditionperformance(asmeasuredbyCDI),andcompliancehistory(as
measuredby SDI). Components shouldbemeasurable for specific timeperiodsof interest,
suchasquarterlyorannually.
Another important concept is the extent to which the IRImeasures the intersection or the
union of riskswithin the bulk powers system. At this time, there appears to be benefit in
shrinking the size of the populations (or weightings) for CDI and SDI. Also, currently the
equationreflectstheadditionofrisks,ratherthanthemeasurementofanyoftheunions(and
increasedweightings)ofeachofthecomponentsoftheIRI. Thisconceptisshowngraphically
inFigure
2.
Depending
upon
the
model
preferred,
the
equation
and
relevant
indices
should
be
modifiedappropriately.
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
4/11
4
Figure2UnionIRIorIntersectingIRI
IRIIndexComponentsAsdiscussedpreviouslyIRIisintendedtobeacompositemetricwhichintegratesseveralforms
of individual risks to the bulk power system. Upon their factoring andweighting a lagging
indicator will be developed on a quarterly or annual basis. At this time three component
indicesarebeingconsidered. Eachofthesecomponentsisdiscussedingreaterdetailbelow. It
is important to caution that since they range across many stakeholder organizations, these
concepts are developed as starting points for continued study and evaluation. Additional
supportingmaterialsarefoundintheAppendicesasmuchoftheindividualindicescalculations
andsupportingtrendinformationisinaformativestate.
SampleIntegratedReliabilityIndex(IRI)CalculationAt this time, the RMWG believes some form of blended weighting may serve to start to
populate IRI characteristic curves at a high and generic level. Based upon feedback from
stakeholders,thisapproachmaychange,orasdiscussedfurtherbelow,weightingfactorsmay
varybasedonperiodicreviewandriskmodelupdate.TheRMWGwillcontinuetherefinement
of the IRI calculation and consider other significant factors that impact reliability (e.g.,
intentionaland controlled loadshedding); further, itwillexploredevelopingmechanisms for
enablingongoingrefinementwhichshouldbeinfluencedbyawidesetofstakeholders.
RMWG
recommends
the
Event
Driven
Index
(EDI)
be
weighted
the
highest
(50%)
since
events
actuallyoccurred,indicatinghowsystemwasperforming.TheConditionDrivenIndex(CDI)and
Standards/Statute Driven Index (SDI) are weighted 25% each since they are indicators of
potentialriskstoBPSreliability. UsingEquation3,IRIcanbecalculatedasfollows:
0.50 0.25 0.25 (3)
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
5/11
5
AAppppeennddiixx AA EEvveenntt DDrriivveenn IInnddeexx ((EEDDII)) CCaallccuullaattiioonn
EventDrivenIndex(EDI)The
event
severity
risk
index
(SRI)
was
developed
to
measure
relative
severity
ranking
of
events
basedoneventoccurrence rateand their impact to thebulkpower system. Impact canbe
amongmultipledimensionssuchasload(asaproxyforcustomers)orlossoffacilities(suchas
generators, transmission lines, substations or communications facilities). These measures
provideaquantitativeapproachtodeterminewhicheventshavemore impactonbulkpower
systemreliability. Inotherwords,themetricsarean integratedmeasurementsystem,which
classifies an events impact on each of the components that are critical to the holistic
performanceofthebulkpowersystem.
Figure3plotsthe20082010NERCSRIperformance,includinghistoricbenchmarkeventsusing
theapprovedSRIformula.Theimpactconsidersloadlossandduration,ACtransmissioncircuit
andgeneration
losses
occurred
from
recorded
events,
including
weather
and
other
natural
events. The load lossanddurationdataweregathered fromEOP0045andOE417
6 reports.
ThedailygenerationandACtransmissioncircuitoutageswereobtainedfromGADS(Generating
AvailabilityDataSystem)andTADS (TransmissionAvailabilityDataSystem),respectively. The
eventcategory7anditsassociatedseverityriskrangearealsoshowninFigure3.
Figure3NERC20082010SeverityRiskIndexversusHistoricBenchmarkDays
5http://www.nerc.com/files/EOP0041.pdf6http://www.eia.doe.gov/cneaf/electricity/forms/instfor417.pdf7http://www.nerc.com/docs/eawg/Event_Analysis_Process_Field_test_DRAFT_102510Clean.pdf
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
6/11
6
The EDI component is derived from RMWGs Severity Risk Index (SRI) values8 and can be
computedasfollows:
(4)
wheretheSRIisadailySeverityRiskIndexandthedurationisaspecifictimeperiodofinterest,
suchasquarterlyorannually. ThevalueofEDIwillrangefrom0and100
BasedontheSRIvalues inFigure3andEquation4,quarterlyEDItrending forthepastthree
yearsispresentedinFigure4.
NosignificantEDItrendchangesareobservedfrom2008to2010.
Figure4NERC20082010EventDrivenIndexbyQuarter
8http://www.nerc.com/docs/pc/rmwg/Integrated_Bulk_Power_System_Risk_Assessment_Concepts_Final.pdf
99.70
99.75
99.80
99.85
99.90
99.95
100.00
EventDriven
Index(EDI)
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
7/11
7
AAppppeennddiixx BB CCoonnddiittiioonn DDrriivveenn IInnddeexx ((CCDDII)) CCaallccuullaattiioonnIn 2009, the RMWG developed the S.M.A.R.T criteria
9 (Specific, Measurable, Attainable,
Relevant and Tangible), shown in Table 1, as the ranking process providing a consistent
approachto
identify
ahigh
ranking
subset
of
metrics.
Table1SMARTMethodandRating(*SMARTScore=SumofRatingsforS,M,A,R,andT)
Table1SMARTMethodandRating*
Rating
S - Specific/Simple
Be easily understoodand not driven bycommercial factors (i.e.tariff)Identify factors thatpositively or negativelyimpact reliability
Address reliabilityproblems and solutions
M - Measurable
Be easily measuredwith regularly collectedinformation
Measure past andcurrent reliability
Measure progress in
ensuring reliability
Measure effectivenessof reliability standardsand enforcementprograms
A - Att ainable
The industry canprovide the rightresources (i.e. funding,time and ability) toimprove reliability
Reliability will bemeasurably improved
R - Relevant
Linked to reliabilitygoals
P rovide meaningfulinformation
P rovide feedback forimproving the Reliability
Standards
T - Tangible/Timely
Reflect current toppriority issues andpossess a sense ofurgency
Identify reliability gapsand point to existingstandards that need to
be modified or newstandards that need tobe developed
3
Defined in a NERCStandard
Not driven bycommercial factors
Addresses specificreliability issues
Easily measured andreported regularly
Historical data existsat REs and/or NERCand is currentlyrequired in NERCStandard
Directly measure
effectiveness ofstandard andenforcement programs
Compelling businesscase suggests goodchance of beingapproved throughbusiness planning andtariff approvalprocesses
Reliabilityimprovements will be
easily seen
Direct link to reliabilitygoalsFocus on failures andpossible solutions orimprovements
Provide directfeedback for improvingthe ReliabilityStandards
Directly links tocurrent top priorityreliability issues andpossess a sense ofurgency
Clearly identifiesreliability gaps andpoints to standardimprovement needs
2
Defined within theindustry
Not driven bycommercial factors
Addresses reliabilityissues
Easily measured andreported on occasionsSome historical dataexists at REs and/orNERC and is currentlyrequired in NERCstandard
Measure effectivenessof standard andenforcement programsin a long run
Additional resourceswill be required andhave a reasonablechance of beingapproved throughbusiness planning andtariff approvalprocesses
Reliabilityimprovements will beapparent within areasonable period oftime (months)
Some link to reliabilitygoals
Not directly focus onfailures and possiblesolutions orimprovements
May provide feedbackfor improving theReliability Standards ina long run
Some link to currenttop priority reliabilityissues
May reveal reliabilitygaps in a long run
1
Defined somewhere
May have somecommercial factors
May relate to reliabilityissues
Easily measured andnot reported
Some historical dataexists at REs and/orNERC
No link toeffectiveness ofstandard andenforcement programs
Significant resourceswill be required wellbeyond normalbusiness planning andtariff approval levels
Reliabilityimprovements will onlymarginal, or evidentover an extendedperiod of time (years)
No link to reliabilitygoals
Not Focus on failuresand possible solutionsor improvements
Not tied to a standardimprovement
No link to current toppriority reliability issuesand does not possess asense of urgency
Does not identifyreliability gaps.
9 http://www.nerc.com/docs/pc/rmwg/RMWG_Metric_Report090809.pdf
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
8/11
8
Figure 1: Metric Development Process
RMWG recommendsusingacombinationofRelevantandTangible ratingsofeachmetricas
weighting factors. Table 2 lists eachmetricsRelevant and Tangible scores10
and combined
metricweightingfactors.
Table2CDIMetricWeightingFactor(WF)andTrendRating(TR)
ALR Description Relevant Tangible WF2010
PerformanceTrend TrendRating
13 PlanningReserveMargin 2 3 5Slight
Improvement4
14BPSTransmissionRelatedEventsResultinginLoss
ofLoad3 3 6
Slight
Improvement4
15 SystemVoltagePerformance 3 2 5 NoData 0
112 FrequencyResponse 3 2 5 NoData 0
23 ActivationofUnderFrequencyLoadShedding 2 2 4 Inconclusive 3
24 AveragePercentNonRecoveryDCS 3 3 6 Inconclusive 3
25DisturbanceControlEventsGreaterThanMost
SevereSingleContingency2 2 4
Slight
Improvement4
35InterconnectionReliabilityOperatingLimit/
SystemOperatingLimit(IROL/SOL)Exceedance3 3 6 NewData 0
41AutomaticTransmissionOutagesCausedby
ProtectionSystemMisoperations3 3 6 Inconclusive 3
61 TransmissionConstraintMitigation 3 2 5 NewData 0
62 EnergyEmergencyAlert3(EEA3) 3 3 6Slight
Improvement4
63 EnergyEmergencyAlert2(EEA2) 3 3 6Slight
Improvement4
611AutomaticTransmissionOutagesCausedby
ProtectionSystemMisoperations3 2 5 Inconclusive 3
612AutomaticTransmissionOutagesCausedby
HumanError3 2 5 Inconclusive 3
613AutomaticTransmissionOutagesCausedby
FailedACSubstationEquipment3 2 5 Inconclusive 3
614AutomaticTransmissionOutagesCausedby
FailedACCircuitEquipment3 2 5 Inconclusive 3
615 ElementAvailabilityPercentage(APC) 2 2 4 NewData 0
616TransmissionSystemUnavailabilityon
Operational,PlannedandAutoSustained
Outages
2 2 4 NewData 0
10http://www.nerc.com/filez/Approved_Metrics.html
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
9/11
9
To integrate individualmetricswithdifferingunitsofmeasure,thefivecommontrendratings
(TR)were identifiedtoquantifyeachmetricperformance level,shown inTable2. Thesefive
ratingsrangefrom1to5. Ifnometricdatatrendisyetavailable,thetrendratingisassigned0.
Thesignificantimprovementtrendhasthehighestratingof5,andothertrendratings,aslisted
below:
TR 5 for Significant Improvement,
4 for Slight Improvement,
3 for Neutral or Inconclusive,
2 for Slight Deterioration, and
1 for Significant Deterioration.
0 for New Data or No Data
Theindividualmetriccontributioncanbecomputedas
(5)
Thevalue
of
metric
contribution
will
range
from
0to
100.
The
CDI
aggregates
each
metric
contributionasaweightedaveragebyrelativeimportance. Forexample,SMARTratingsofthe
abovefivemetricsservesastheirimportancevaluesintheCDIequationbelow:
(6)
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
10/11
10
AAppppeennddiixx CC SSttaannddaarrddss//SSttaattuuttee DDrriivveenn IInnddeexx ((SSDDII))
TheRMWGrecommendsusingtheReliabilityImpactStatement(RIS)andViolationRiskFactor11
(VRF)asselectioncriteriatoidentifythesubsetofstandardrequirementstobeincludedinthe
SDI. RISistheassessmentofrisktothebulkpowersystemwhenarequirementisviolated,as
determinedbytheRegionalEntity. ThethreeRISlevelsareminimalimpact,moderateimpact
andsevereimpact. ThefactorsincludedintheRISare
TimeHorizon
Relativesizeoftheentity
Relationshiptootherentities
Possiblesharingofresponsibilities
Voltagelevelsinvolved
Sizeofgeneratororequipmentinvolved
Abilityto
project
adverse
impacts
beyond
the
entitys
own
system
Failuretomaintainvegetationclearancecouldcauseorexacerbateacascadingoutageisan
example of severe impact; lack of regular maintenance and testing could result in
misoperations isanexampleofmoderate impact; theentity is conducting training,but its
training program did not include a plan for the initial operator training is an example of
minimalimpact.
Basedon theabovecriteria,thetwentysixstandard requirements inTable3were identified
fromNERCs3yearcompliancedatabaseashavingasevereRISandhighVRF.
Table3StandardRequirements*IncludedinSDI
*RequirementsareidentifiedfromtheNERCcompliancedatabasewheretheirviolationshave
severeRIS.
11http://www.nerc.com/page.php?cid=2|20|285
Standard Req. Standard Req. Standard Req. Standard Req. Standard Req.
EOP-001-0 R1. FAC-009-1 R1. PER-002-0 R3. PRC-005-1 R2. TOP-004-2 R1.
EOP-003-1 R7. IRO-005-2 R17. PER-002-0 R4. TOP-001-1 R3. TOP-004-2 R2.
EOP-005-1 R6. PER-001-0 R1. PRC-004-1 R1. TOP-001-1 R6. TOP-006-1 R6.
EOP-008-0 R1. PER-002-0 R1. PRC-004-1 R2. TOP-001-1 R7. TOP-008-1 R2.
FAC-003-1 R1. PER-002-0 R2. PRC-005-1 R1. TOP-002-2 R17. VAR-001-1 R1.
FAC-003-1 R2.
-
7/30/2019 Integrated Reliability Index WhitePaper DRAFT
11/11
11
ByapplyingasimilaraggregationasEDI,theSDIcanbecalculatedas
100
(7)
where:
integrated standard compliance index for a speciic period;
weighting of a particular requirement violation; number of known unmitigated violations for the selected requirement;
number of registered entities who are required to comply
with the selected requirement.
Assuming an equal weighting (wV) for all requirements, Figure 5 provides the quarterly SDI
trendsforpastthreeyears.
Figure5NERC20082010SDITrendbyQuarter
Figure5providesthequarterlySDItrendsforpastthreeyears,indicatingtheriskduetoknown
severe impactviolationshasdecreased,andthe industryhasachievedahigherreliability level
throughstandardscompliancesincethirdquarter2009.
99.70
99.75
99.80
99.85
99.90
99.95
100.00
Stan
dardsDrivenIndex(%)