Sandia National Laboratories is a multi-mission laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000. SAND NO. 2011-XXXXP

AssessingPowerSystemResiliencetoAdverseWeatherEvents

AndreaStaidINFORMSComputingSocietyConference

January17,2017

WhatisInfrastructureResilience?

Definitions:§ Theabilitytoprepareforandadapttochangingconditionsand

withstandandrecoverrapidlyfromdisruptions.– PresidentialPolicyDirective21§ Theabilitytoreducethemagnitudeand/ordurationofdisruptive

events.– NationalInfrastructureAdvisoryCouncil§ Anticipate– Identifyandplanforadverseevents(deliberateattacks,accidents,

ornaturaldisasters)§ Absorb– Continueoperatingaftershockstosystem§ Adapt– Adjustsysteminrealtimetominimizeadverseimpacts§ Recover– Returnsystemoperationstonormalstateasfastaspossible

Howisitdifferentfromreliability?§ Resilience focusesonlow-probability,high-consequenceevents§ Reliability focusesonhigh-probability,low-consequenceevents

§ Day-to-Dayoperations2

Howtoimproveresilience?

§ Mustbeabletomeasureit!§ Usearisk-basedapproach

§ Identifythreatsofconcern§ Resilienttowhat?Improvementsmustbetargetedataspecifictypeofthreat.

§ Increasedresilienceagainsthurricanesmaynothelpwithresiliencetoterrorattacks.Approachisgenerallythreat-specific.

§ Assesslikelihoodofsystemdisruptiongivenathreat§ Evaluateconsequencesofdisruption

§ Moreresilientsystemswillminimizelikelihoodofdisruption,severityofconsequences,orboth

§ Needaquantifiablemetricgivenaspecificthreat

3

ResiliencyAnalysisProcess

§ Frameworkforquantificationofpowersystemresilience§ Thisframeworkenablesdecisionmakingtoobtain

demonstrableresilienceimprovements

4

Prob

abilit

y of

Con

sequ

ence

s [$

] G

iven

Thr

eat X

Consequences [$]

Reduced Expected Financial Consequence

Reduced Risk

Baseline System Resilience

Resilience of System after Improvements Improvements must

cost significantly less than E-E’

E’(C) E(C)

§ Resultingresiliencemetricsareprobabilistic

§ Theframeworkisflexible:§ Canhandledifferenttypes

ofthreats§ Providesinformationfor

differenttypesofdecisionmakers

ResilienceFramework

5

Define Resilience

Goals

Define System & Resilience

Metrics

CharacterizeThreats

Determine Level of Disruption

Define & Apply System Models

Calculate Consequence

Evaluate Resilience

Improvements

Populate

Define Resilience

Goals

Define System & Resilience

Metrics

CharacterizeThreats

Determine Level

of Disruption

Define & Apply System Models

Calculate Consequence

Evaluate Resilience

Improvements

Create

ScenarioAnalysis:IdentifyThreats

6

§ Whatisthepossiblethreatspace?

§ Consequences(andprotectivemeasures)canvarydrasticallyamongthreats.

§ Focusonimprovingsystemresilienceagainstanindividualthreat.

ScenarioAnalysis:CharacterizeIndividualThreat

§ Givenhigh-levelthreatcharacterization,thenextstepistofurtherrefinethedescriptionofthespecificthreats

… …

Historicalinformationandforecastmodelsusedtoguidespecificationofpossibleeventsandtheirrelativelikelihoods

p1 p2 pn

Category4,north-of-peninsulastormtrack

Category5,eyetracksovermetropolitanarea

Category2,landfallathightide

…

ScenarioAnalysis:DisruptingtheSystem

… …p1 p2 pn

Category4,north-of-peninsulastormtrack

Category5,eyetracksovermetropolitanarea

Category2,landfallathightide

……

Givenaspecificmanifestationofadisruptionevent,wethenspecifyadistribution ofinfrastructureimpacts

DamageRealizationN

DamageRealizationK

Assumeuniformprobabilities

Forexample:1. Normaldistributionofgeneratorfailures,

withu=20,s=52. Normaldistributionoflinefailures,with

u=40,s=7

§ Thefinalstepistotranslatedisruptioneventsintosystemimpacts

CaseStudy– AEPAdverseWeather

9

§ AmericanElectricPower(AEP)isalargeelectricutilityintheU.S.§ Serving5.4millioncustomersin11states§ OwnslargesttransmissionnetworkintheU.S.– Morethan40,000

miles– and31GWofgeneratingcapacity

§ Interestedinimprovingresiliencytoadverseweathereventsintheir‘East’territory

ImprovementOptions

§ Transmissionsystemresiliencycanbeimprovedby:§ Hardeninglines

§ Reducinglikelihoodoffailureforindividuallines§ Long-termplanningdecision

§ Generatorre-dispatchand/ortransmissionswitching§ Inadvanceofastorm,re-dispatchsystemtomaintainpowertocustomers

§ Real-timedecisionswhenadverseweatherinforecast

§ Demonstratebenefitofpossibleactionusingscenariosofweatherevents§ Compareproactivedecisionsto‘businessasusual’case

10

AnalyzingSystemResponse

§ Wouldtherebeabenefittore-dispatchingthesysteminadvanceofastorm?Canthisincreaseresilience?

§ Usehistoricalstormstobuildscenariosofrealisticfuturestorms§ Futurestormsunlikelytoexactlymirrorpaststorms,butsystem

weaknessescanbecaptured§ Historicalprobabilityofoutagevariesdrasticallyacrosslines

§ Scenariogenerationforthreeapplicationareas:§ Identifycandidatesforlinehardening§ Demonstratevalueofreal-timegeneratorre-dispatch§ Developtoolforreal-timeuseinadvanceofoncomingweather

§ Inwork

11

Number of Outages per Circuit

Density

0 20 40 60 80 100

0.00

0.02

0.04

0.06

0.08

0.10

Number of Outages per Event (Log Scale)

Density

0 1 2 3 4 5

0.0

0.2

0.4

0.6

0.8

1.0

AvailableData§ Transmissioncircuitoutagedatafrom~1990– 2015§ Mostcircuits seeveryfewoutages,butsomehavemany§ Moststorms causeveryfewoutages,butsomestormsresultin

hugenumbers(e.g.,June2012Derecho,SuperstormSandy)

12

ScenarioGeneration

§ Needtorepresentspectrumofstormdamage:§ Randomsamplingfor‘standard’scenarios§ Augmentedsamplingfor‘worstcase’scenarios

§ Forcelargernumberofoutages(uptoselectedmaximum),stillusingbaselineprobabilitiestosample

§ Complications:§ Probabilitiesoflineoutagescanbeconditionalontypeofweather

§ But,realdataismessy,mostoftheoutagecause-codescannotbetrusted§ Somestormeventshavespatiallydistributedoutages,othersvery

concentrated§ Needto‘force’scenariostorepresentbothtypes.Don’tyethaveenoughdatatodothiswell

§ Outagesmustrepresentcascadingfailuresinsystem§ Linkscenariosto‘contingency’dataonpropagatingfailures§ Givesrealisticrepresentationofhowsystemwouldhandleanoutage 13

AnticipativeOperations§ StochasticOptimizationallowingforanticipativeoperations

§ Basedonuncertainscenariosofadverseweatherimpact,howbesttore-dispatchgeneratorstominimizelossofload?

14

BASELINE RE-DISPATCH

NextSteps– Real-TimeTool

§ Createscenariosforuseinreal-timedecisionsasstormapproaches

§ Linkweatherdatatoprobabilityofoutage;Circuitoutagewillbeafunctionof:§ Windspeed,Precipitation,Temperature,Lightningforecast,etc.

§ Generatescenariosbasedonuncertaintyinweatherforecast,stormtrack,andprobabilityoffailure

§ Simulatesuggesteddecisionsforre-dispatchonAEPsystemforactualweatherevents

§ Quantifychangeinconsequences(lossofload)acrosssystem

15

SomeFinalThoughts

§ Demonstratedbenefittosystemoperatorsbybeingproactivewhenitcomestoweather

§ Inordertoquantifythechangesinsystemresiliencetoadverseweather,wechosetodevelopscenariosof‘stormevents’representedbytransmissionlinefailures

§ Workingtowardsareal-timetool;movesfromtheoreticallyimprovingresiliencetoofferingsuggestedactionstominimizelossofload§ Highlydata-dependent,shouldbeexciting!

16

Questions?

Contact:AndreaStaidastaid@sandia.gov

Acknowledgements:AmericanElectricPowerforprovidingthedataDHSandDOE/EPSAforfunding

17