PaulDenholm,WesleyCole,A.WillFrazier,KaraPodkaminer,andNateBlairStorageFuturesStudyTheChallengeofDefiningLong-DurationEnergyStorageSuggestedCitation:Denholm,Paul,WesleyCole,A.WillFrazier,KaraPodkaminer,andNateBlair.2021.TheChallengeofDefiningLong-DurationEnergyStorage.Golden,CO:NationalRenewableEnergyLaboratory.NREL/TP-6A40-80583.https://www.nrel.gov/docs/fy22osti/80583.pdf.StorageFuturesStudyTheChallengeofDefiningLong-DurationEnergyStoragePaulDenholm,WesleyCole,A.WillFrazier,KaraPodkaminer,andNateBlairiiiThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.AcknowledgmentsTheauthorswouldliketothankthefollowingindividualsfortheircontributions.EditingandotherresearchsupportwasprovidedbyClaireBolyard,MichaelDeneen,MadelineGeocaris,andMikeMeshek.HelpfulreviewandcommentswereprovidedbySamBaldwin,JaquelinCochran,ChrisNamovicz,KeithParks,GianPorro,andPaulSpitsen.ThisworkwasauthoredbytheNationalRenewableEnergyLaboratory,operatedbyAllianceforSustainableEnergy,LLC,fortheU.S.DepartmentofEnergy(DOE)underContractNo.DE-AC36-08GO28308.FundingprovidedbyU.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergySolarEnergyTechnologiesOffice,U.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergyWindEnergyTechnologiesOffice,U.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergyWaterPowerTechnologiesOfficeandU.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergyOfficeofStrategicAnalysis.TheviewsexpressedhereindonotnecessarilyrepresenttheviewsoftheDOEortheU.S.Government.ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratory(NREL)atwww.nrel.gov/publications.U.S.DepartmentofEnergy(DOE)reportsproducedafter1991andagrowingnumberofpre-1991documentsareavailablefreeviawww.OSTI.gov.ivThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.PrefaceThisreportisoneinaseriesoftheNationalRenewableEnergyLaboratory’sStorageFuturesStudy(SFS)publications.TheSFSisamultiyearresearchprojectthatexplorestheroleandimpactofenergystorageintheevolutionandoperationoftheU.S.powersector.TheSFSisdesignedtoexaminethepotentialimpactofenergystoragetechnologyadvancementonthedeploymentofutility-scalestorageandtheadoptionofdistributedstorage,andtheimplicationsforfuturepowersysteminfrastructureinvestmentandoperations.Theresearchfindingsandsupportingdatawillbepublishedasaseriesofpublications.ThetableonthenextpageliststheplannedpublicationsandspecificresearchtopicstheywillexamineundertheSFS.Thisdocumentexploresthedefinitionof“longduration”asappliedtoenergystorage.Giventhegrowinguseofthisterm,auniformdefinitioncouldaidincommunicationandconsistencyamongvariousstakeholders.ThereislargeandgrowinguseoftheAdvancedResearchProjectsAgency–Energy(ARPA-E)definitionofgreaterthan10hours.However,theterm“long-durationenergystorage”isoftenusedasshorthandforstoragewithsufficientdurationtoprovidefirmcapacityandsupportgridresourceadequacy.Theactualdurationneededforthisapplicationvariessignificantlyfromaslittleasafewhourstopotentiallymultipledays.Thisdualuseofthetermmeansthattherecannotbeasimple,uniform,andstaticdefinitionoflong-durationstoragethatcapturesitsabilitytoprovidefirmcapacityandalsoaidsconsistentcommunication.Toaddressthisissue,theNationalRenewableEnergyLaboratoryrecommendsthatqualitativedescriptionsoflong-durationenergystoragealwaysbeaccompaniedbyquantitativedescriptions,andthatpowersectorstakeholdersbedeliberateinhowtheychoosetodefinelong-durationenergystoragetechnologies.TheSFSseriesprovidesdataandanalysisinsupportoftheU.S.DepartmentofEnergy’sEnergyStorageGrandChallenge,acomprehensiveprogramtoacceleratethedevelopment,commercialization,andutilizationofnext-generationenergystoragetechnologiesandsustainAmericangloballeadershipinenergystorage.TheEnergyStorageGrandChallengeemploysausecaseframeworktoensurestoragetechnologiescancost-effectivelymeetspecificneeds,anditincorporatesabroadrangeoftechnologiesinseveralcategories:electrochemical,electromechanical,thermal,flexiblegeneration,flexiblebuildings,andpowerelectronics.Moreinformation,anysupportingdataassociatedwiththisreport,linkstootherreportsintheseries,andotherinformationaboutthebroaderstudyareavailableathttps://www.nrel.gov/analysis/storage-futures.html.vThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.TitleDescriptionRelationtothisReportTheFourPhasesofStorageDeployment:AFrameworkfortheExpandingRoleofStorageintheU.S.PowerSystemExplorestherolesandopportunitiesfornew,cost-competitivestationaryenergystoragewithaconceptualframeworkbasedonfourphasesofcurrentandpotentialfuturestoragedeployment,andpresentsavaluepropositionforenergystoragethatcouldresultincost-effectivedeploymentsreachinghundredsofgigawatts(GW)ofinstalledcapacity.Providesbroadercontextontheimplicationsofthecostandperformancecharacteristicsdiscussedinthisreport,includingthespecificgridservicestheymayenableinvariousphasesofstoragedeployment.Thisframeworkissupportedbytheresultsofscenariosinthisproject.EnergyStorageTechnologyModelingInputDataReportReviewsthecurrentcharacteristicsofabroadrangeofmechanical,thermal,andelectrochemicalstoragetechnologieswithapplicationtothepowersector.Providescurrentandfutureprojectionsofcost,performancecharacteristics,andlocationalavailabilityofspecificcommercialtechnologiesalreadydeployed,includinglithium-ionbatterysystemsandpumpedstoragehydropower.Providesdetailedbackgroundaroundthebatteryandpumpedstoragehydropowercostandperformancevaluesusedasinputstothemodelingperformedinthisproject.EconomicPotentialofDiurnalStorageintheU.S.PowerSectorAssessestheeconomicpotentialforutility-scalediurnalstorageandtheeffectsthatstoragecapacityadditionscouldhaveonpowersystemevolutionandoperations.Featuresaseriesofcost-drivengrid-scalecapacityexpansionscenariosfortheU.S.gridthrough2050andexaminesthedriversforstoragedeployment.DistributedStorageCustomerAdoptionScenariosAssessesthecustomeradoptionofdistributeddiurnalstorageforseveralfuturescenariosandtheimplicationsforthedeploymentofdistributedgenerationandpowersystemevolution.Analyzesdistributedstorageadoptionscenariostotestthevariouscosttrajectoriesandassumptionsinparalleltothegridstoragedeployments.TheChallengeofDefiningLong-DurationEnergyStorageDescribesthechallengeofasingleuniformdefinitionforlong-durationenergystoragetoreflectbothdurationandapplicationofthestoredenergy.Thisreport.GridOperationalImplicationsofWidespreadStorageDeploymentAssessestheoperationandassociatedvaluestreamsofenergystorageforseveralpowersystemevolutionscenariosandexplorestheimplicationsofseasonalstorageongridoperations.Considerstheoperationalimplicationsofstoragedeploymentandgridevolutionscenariostoexamineandexpandonthegrid-scalescenarioresultsfoundwiththeRegionalEnergyDeploymentSystem(ReEDS).StorageFuturesStudy:ExecutiveSummaryandSynthesisofFindingsSynthesizesandsummarizesfindingsfromtheentireseriesandrelatedanalysesandreports,andidentifiestopicsforfurtherresearch.Includesadiscussionofallotheraspectsofthestudyandprovidescontextfordiscussioninthisreport.viThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.TableofContentsPreface........................................................................................................................................................iv1Introduction...........................................................................................................................................12FirstThingsFirst:Defining“Duration”ofEnergyStorage..............................................................23DefiningLongDurationToCommunicateConsistently...................................................................34DefiningLongDurationToEstablishItsAbilityToProvideResourceAdequacy........................45AFurtherComplication:TheImpactofEconomicandTechnologyCapabilities.........................96Conclusions........................................................................................................................................10References.................................................................................................................................................121ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.1IntroductionAstheshareofU.S.powergenerationfromvariablerenewableenergy(VRE)grows,anewvisionistakingshapeforlong-durationenergystorage(LDES)toensureaffordableandreliableelectricity.Inthisvision,LDESisdeployedatlargescaletoprovideresourceadequacy1tothegridandsupportdecarbonizationoftheelectricitysystem.However,thelackofauniformdefinitionofLDESinhibitsclearcommunicationabouttheneedsofthecurrentandfuturegrid,includingscenariosapproaching100%decarbonizationrelyingprimarilyonrenewableenergy.Energystoragedurationistypicallyexpressedintermsofthenumberofhoursastoragedevicecanprovidecontinuousoutputatitsratedcapacity.DefinitionsofLDESintheliteraturerangefromaslittleas2hourstoasmuchasmultipledaysorevenmonths.Therearetwomainreasonstoestablishaconsistentdefinition:1.Createacommonlanguagetoaidcommunicationtoensurestakeholdersareworkingunderconsistentassumptionsandunderstanding.2.Establishcharacteristicsneededtoprovidefirmcapacityandsupportresourceadequacy,particularlyforestablishingregulatoryormarketrulesorotherstandards.ItisrelativelystraightforwardtodefineLDESforthefirstreason(acommoncommunicationframework)andareviewoftheliteraturesuggestsdurationsofatleast10hourscouldapproachaconsensus-baseddefinition,givenitscurrentusebyanumberofindustryandgovernmentorganizationsandgrowinguseintheacademicandgeneralliterature.Wesuggestcautioningeneraluseofthisdefinition,however,asitinherentlyconflictswiththesecondmotivationforadefinitionofLDES(basedonitsabilitytoprovidefirmcapacity).Thisapplication-baseddefinitionhasimportantimplicationsformaintainingareliablegrid,establishingmarketrules,andoptimalplanningfordecarbonizationofthepowersystem.Itisdifficult—ifnotimpossible—toreconcilethetwodifferentapproachestodefiningLDESandarriveatasinglenumericalvalueforduration(orevenrangeofvalues)thatdefinesLDESforbotheaseofcommunicationandusingthistermasashorthanddescriptionforstoragethatprovidesfirmcapacity.Theabilityofstoragetoprovidefirmcapacity(measuredintermsofcapacitycredit)rangessignificantlybasedonregionaldemandpatternsandgridmix,includingtheamountofrenewableenergyandstoragealreadyinplace.Therefore,thedurationofstorageneededtoprovidehighcapacitycreditcanspananenormousrange,fromasfewasabout2–4hoursforsomelocationsintoday’sgridtomultipledaysinfuturegridswithverylargerenewableenergyandstoragedeployment.Asaresult,LDEScannotsimultaneouslyhaveasimpleuniformnumericalvalueandbeusedasathresholdvalueformeasuringcapacitycredit.1Resourceadequacy(orsimply“adequacy”)isdefinedbytheNorthAmericanElectricReliabilityCorporation(NERC)as“Theabilityoftheelectricsystemtosupplytheaggregateelectricaldemandandenergyrequirementsoftheend-usecustomersatalltimes,takingintoaccountscheduledandreasonablyexpectedunscheduledoutagesofsystemelements”(1).Thisincludesmeetingpeakdemandduringperiodsofhotorcoldweather,duringperiodsoflowVREoutput,duringscheduledorunscheduledplantoutages,orduringextremeweather.2ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Thisdiscrepancyincreasesthechallengeofcommunicatingthepotentialroleandopportunitiesforstorageofvariousdurations,especiallywhenconsideringtheeconomicsofdifferenttechnologiesthatmayprovidedifferenceservices.Regulatoryandmarketframeworkswilllikelyneedtoevolvetoaccommodatethisreality.However,thelackofasimpleuniformdefinitionprovidesanopportunitytoeducatekeystakeholdersaboutthecriticalimportanceofevaluatingresourceadequacywithincreaseddeploymentofrenewablesandmultiplestoragetechnologies.2FirstThingsFirst:Defining“Duration”ofEnergyStorageFirst,itisimportanttoestablishthedefinitionofstorage“duration.”Thisdocumenttakestheperspectiveoftheenduserofastationarystoragedevice,includinggridplanners,operators,andutilities.Fromthisperspective,durationhasafairlystraightforwarddefinitionsummarizedbytheU.S.EnergyInformationAdministration(2):Thedurationofabatteryisthelengthoftimethatastoragesystemcansustainpoweroutputatitsmaximumdischargerate,typicallyexpressedinhours.Theenergycapacityofthebatterystoragesystemisdefinedasthetotalamountofenergythatcanbestoredordischargedbythebatterystoragesystem.Itisimportanttoemphasizethatweinterpretbothenergyanddurationasmeasuringtheusableenergyanddurationavailabletotheplantorsystemoperator,netofenergyheldbacktomaintainminimumandmaximumstateofchargeorotherfactors.Thismeansthattheamountofusableenergystoredisequaltothenetpowerratingmultipliedbytheduration.Forexample,a1-MW(ACrating)batterywith4hoursofdurationhas4MWhofusablestoredenergythatcanbedeliveredtothegrid.The“gross”storagecapacityneededtoachievethenetcapacityisaseparatefactordeterminedbythemanufacturerordevelopertoensurethatthenetdurationisavailabletotheenduser.Theuseofnetvs.grosscapacityfordefiningdurationhassignificantprecedent.Pumpedstoragehydropowerplants,whichrepresentthevastmajorityofenergystoragedeployedtodate,aretraditionallymeasuredbytheamountofstoredwaterthatcanactuallybeused,accountingfortheminimumandmaximumlevelsofboththelowerandupperreservoir,asopposedtothetotalamountofwaterinthereservoir.2Notethatthisdefinitiondescribesonlytheamountofenergystored,nothowlongitwillbestoredbeforeuse.3However,thetwoquantitiesarepotentiallyrelatedasdurationsincrease,asdeviceswiththecapacitytostoremultipledaysofenergywilllikelyneedtostorethisenergyformultipledays(orlonger)beforedischarging.Finally,thisdefinitiondoesn’tconsiderthetime2Thisisalsoanalogoustounusablecushiongasinhydrocarbonstorage.Inaddition,netpowerratingsofpowerplantshavelongbeenthestandarddefinitionofplantpoweroutput,accountingforplantoperatingrequirementssuchasparasitics(e.g.,crushers,fans).3Onearticleuses“longterm”tomeasuretheamountoftimestorageisheldbeforedischarging,butwefoundlittleindicationthisuseiscommon,andfoundmultipledocumentswhere“term”and“duration”arebothusedtorepresentstoragecapacity(3,4).3ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.neededtochargethedevice,whichcanimpacttheabilityofadevicetoachieveitsrateddurationforcertainapplications,asdiscussedlater.3DefiningLongDurationToCommunicateConsistentlyItcanbeusefultoassignanumericalvaluetoaqualitativeadjectiveusedinthepowerindustryforbotheaseandconsistencyofcommunication.Forexample,wecommonlyusetheterms“distribution”and“transmission”todescribevoltagelevelswithinthepowergridinsteadofstatinganumericalvalue.Itiseasiertouse“distributionvoltage”insteadof“voltagelevelsat34.5kVandbelow.”Inthesecases,theactualvalueassignedtothisadjectivecanbesomewhatarbitrary,butovertime,orthroughactualstandards,thevaluecanbecomeinstitutionalizedorcodified.Giventhattheterm“long-durationenergystorage”isalreadypartofthepowersystemvernacularwithoutacleardefinition,havingaconsensus-baseddurationvaluewouldaidincommunication.Ifthisisourprimarymotivationfordefininglong-duration,itisprobablyeasiesttouseexistingliteraturetoderivesomethingclosetoaconsensusvalue.WecanfindnoevidencethatLDEShasbeendefinedinadefinitivemannerbyastandardsorganizationsuchastheInternationalOrganizationforStandardizationorInstituteofElectricalandElectronicsEngineers.Table1providesasummaryofaliteraturereviewof39documentsthatdefinelongdurationwithanumericalvalue,typicallyexpressedinhours.Weidentifythetypeofdocument,dividingjournalarticlesbetweenthosefocusedonactualgridapplicationsforstorageandthosethatfocusontechnologydevelopment.Weexcludedefinitionspublishedorpromotedbyindividualtechnologydevelopers.WedonotincludealargenumberofdocumentsthatdiscussLDESinsignificantdetailbutdonotexplicitlydefineaduration.Table1.SourcesDefiningLong-DurationStorageDuration(Hours)CitationCount(numberofcitesfollowedbyreferences)U.S.Dept.ofEnergyJournal(technologyfocus)NationalLabReportJournal(gridfocus)MediaUtility/Trade/ConsultantOtherTotal>21(2)1≥43(5–7)2(8,9)2(10,11)3(12–14)2(15,16)1(17)13≥61(18)1≥81(19)1(20)2≥102(21,22)2(23,24)7(4,25–30)1(31)3(32–34)15Beyonddiurnala1(35)1(36)5(3,37–40)7Total4631443539aTypicallymultidaytoseasonal4ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Whilenotanexhaustivereview,Table1demonstratesalargerangeofdefinitions,withasignificantnumberofdefinitionswiththreethresholdvalues:≥4hours,≥10hours,andwhatwecall“beyonddiurnal,”whichintheliteraturegenerallycorrespondstomultidaytoseasonalstorage.Thisrangereflectshowdifferentstudiesofenergystorageoftenconsiderdifferentaspects,includingdifferenttechnologies(e.g.,abatterywith4hoursofcapacity,whichhaslongerdurationthanmostcurrentlydeployed)ordifferentgridscenarios(e.g.,astudyofafuturegridwithverydifferentrequiredattributesthantoday’s).Wefoundseveralarticlesthatdiscusstheissueswesummarizelaterinthisdocument,includingthechallengeofprovidingadefinitionthatbothusesaconsistentnomenclatureandisbasedonapplicationsthatLDEScanserve(41,42).Despitethelargerangeindefinitions,thereappearstobeatleastsomejustificationforconsidering≥10hoursasaconsensusduration,basedontwofactors.First,ithasthelargestnumberofcitationsinoursurvey.Second,thereappearstobegrowinguseofthisvaluefollowingitsusebytheAdvancedResearchProjectsAgency–Energy(ARPA-E),whichdefinesLDESas10–100hours(21).Thisprogramandcorrespondingvalueiscitedspecificallyinseveralarticles,andmorerecentlythisvaluewasusedintheU.S.DepartmentofEnergy’sLongDurationStorageShottarget(22).WhiletheARPA-Edefinitionalsoestablishesanupperbound,thereislittlediscussionofanupperboundintheliterature.However,itmaybevaluabletoestablishnomenclaturetodistinguishbetweentechnologiesgenerallythoughtofashavingtechnicaloreconomiclimitstodurationsmuchbeyond12hourstoafewdays(e.g.,pumpedstorage,manybatteries,pumpedthermalstorage)andthosewithmuchlongercapacities(multipledaysandbeyond)thatmightbebettercharacterizedas“seasonal,”suchaspowertogas.Althoughitmaybepossibletoapplya10-plus-hourdefinitionforthesakeofconvenience,thebroadrangeindefinitionsfrom4hourstomultipledayspointstotheneedtounderstandthesedifferences.VariationindefinitionsofLDEScanhaveimportantconsequencesifthesedefinitionsareusedtocommunicateneedsinthecurrentorfuturegrid,ortoestablishpolicyormarketrules.ThisleadsustooursecondmotivationandapproachtodefiningLDES,basedonservicesthatLDEScanprovide.Thissecondapproachwilldemonstratetheinherentchallenge(orperhapsimpossibility)ofachievingauniformdefinitionofLDES.4DefiningLongDurationToEstablishItsAbilityToProvideResourceAdequacyInadditiontoeaseofcommunication,itisalsocommontoassignanumericalvaluetoaqualitativeadjectiveforavarietyofregulatoryandmarketreasons.Forexample,a“majorsource”forcertainemissionsisdefinedbytheU.S.EnvironmentalProtectionAgencyasemittingatleast10tons/year(43).Evenwithoutanunderlyingregulatoryneed,itmaybeusefultogenerateanapplication-basedthresholdforadefinitionlike“longduration.”ForLDES,thisapproachtoadefinitionultimatelylinksathresholddurationvaluetoaspecificapplication—itsabilitytoprovidefirmcapacity.Energystorageisincreasinglybeingdeployedforthepurposesofprovidingfirmcapacityandsupportingresourceadequacy.Theprovisionoffirmcapacityisalsoasignificantaspectofthe5ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.discussionofLDESthatappearsintheliterature.4Thisalsoreflectstheoften-claimed“need”forLDESnoworinthenearfuture,whereretirementsofexistinggenerationassetsorgrowthinelectricitydemandincreasetheneedforresourcesthatcanmeetthisdemand.DefiningLDESintermsoftheminimumdurationneededtoprovidefirmcapacityresultsinalargerangeofdurationsthatvaryovertimeandbylocation.Asaresult,thisapproachtoadefinitionultimatelyconflictswithourfirst(consistencyandeaseofcommunication).Toexplorethepossibleinconsistencyindefinitions,wemustfirstexaminethecapacitycredit,oreffectiveload-carryingcapability(ELCC)ofstorageasafunctionofduration.Simplystated,if1MWofstorageistocompletelyreplace1MWofconventionalgenerationcapacity,itneedstohavesufficientdurationtoprovideequalorgreatercapacitycredit,orELCC.ELCCreflectstheabilityofageneratortobeavailableduringtheperiodofhighestriskofanoutage,whichtypicallycorrespondstoperiodsofpeakdemand—or,increasingly,peaknetdemand,wherenetdemandisthenormaldemandminusthecontributionofVRE.Figure1showsanexampleofasimpleapproximationapproachforcalculatingthedurationofstorageneededtoreducethenetpeakloadbyacertainstoragepowercapacity(essentiallyrepresenting100%capacitycredit).Inthisexamplewearesimulatingreplacing1,700MWofconventionalpeakinggenerationcapacityinFlorida.Wemeasuretheamountofenergyneededtoreducethenetloadbytheratedcapacityofthestoragedevice,whichinthiscaseisabout7,000MWh,correspondingtoabouta4-hourduration.Figure1.MeetingthepeakwithenergystorageSimpleapproachessuchasthoseshowninFigure1canbeusedtoestimatethecapacitycreditofstorageasafunctionofduration,essentiallyusingalinearderate.Inourpreviousexample,a1,700-MWdevicewith4-hourdurationcanprovideapproximately1,700MWofELCC(100%4ManyofthedocumentsinTable1describetheroleofLDESwithavarietyofterms—includingcontributingtoreliability,addressingshortfallsinrenewableenergysupply(particularlywhenevaluatingscenarioswithhighrenewableenergydeployment),ormeetingpeakdemandduringperiodsofveryhotorcoldweather.Theseallarefundamentallydescribingresourceadequacy(1).22,00024,00026,00028,00030,00032,00034,00036,00038,00040,00042,00012AM4AM8AM12PM4PM8PMLoad(MW)HourofDayStorageDischargeChargingLoadNetLoadReductioninnetpeakdemandresultingfromstorage6ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.capacitycredit),andastoragedevicewiththesamepowerratingwith2-hourdurationcouldbederatedto850MW(meaningitwouldoutput850MWfor4hours).Thisexampleforaparticularlocationandgridmixfindsthatstoragewithabout4-hourdurationcancontributeabout100%ofitspowerratingtosystemresourceadequacy.Whilethisrepresentsanapproximation,moredetailedprobabilisticapproacheshavefoundthat4-hourdurationstoragedevicescanprovidehighcapacitycreditinmanypartsoftheUnitedStatesthataresummer-peaking(44–47).ThisvaluealsoconformstothethresholdvaluesetforfullcapacitycreditestablishedbymanymarketregionsintheUnitedStates(48).5Theimplicationofthisresultisthatatleastinsomelocations,thereiscurrentlynoinherentneedforstoragewithatleast10hoursofdurationtoprovidesystem-levelresourceadequacy.Asaresult,anapplication-baseddefinitionofLDES(intermsofminimumdurationneededtoprovidefirmcapacity)doesnotmatchour“easeofcommunication”definitionsuggestedbytheliterature(10+hours).Furthermore,wecannotgenerateauniformlyconsistentapplication-baseddefinitionbecausethethresholddurationvaluevariesgreatlyduetofourmainreasons,asdiscussedinthefollowingsections.4.1Reason#1:Thecapacitycreditofstoragevariesbasedonregionalloadpatterns.Theabilityofstoragetoservepeakdemandperiodsdependsontheshapeanddurationofthosepeaks.Figure2takesthesameapproachasinFigure1,butinthiscasecalculatesthedurationofstorageneededforNewYork,accountingforthedifferentsizeofthesystem.6BecauseNewYorkexperienceslongerloadpeaks(illustratedherebythepurpleareabeingwider),thatsystemwouldrequireabout5.5hourstoachievethesamenetloadreductioninproportiontotheFloridacaseabove.5Establishingthecapacitycreditofstorageasafunctionofdurationisimportantformanyregulatoryandmarketreasons,includingtheabilityofautilityorload-servingentitytomeetresourceadequacystandardssetatthelocal,state,orregionallevel.Italsoestablishestheabilityofindividualplantstoreceivecapacitypaymentsinwholesalemarketsoraspartofpowerpurchaseagreements.6ThepeakloadinFloridainthisyearwasabout1.6timesthatinNewYork,sowearesimulatingaproportionallysmallerstoragepowercapacity(1,065MWinNewYorkvs.1,740MWinFlorida).7ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Figure2.Longernetloadpeaksdecreasethecapacitycreditofstorage(NewYorkexample)Ingeneral,longerpeakstypicallyoccurinlocationsthatareeitherlessstronglysummer-peakingorwinter-peaking.Thisresultsinasignificantregionalvariationinthecapacitycreditofstorage,andresultsindifferentthresholdsfordurationstoachieve100%capacitycredit.4.2Reason#2:Thecapacitycreditofstoragevariesbasedonrenewableenergydeployment.ThelengthofthenetloadpeakisalsoimpactedbyVREdeployment,particularlysolarphotovoltaics(PV).Figure3providesanexampleillustratingacasewheresubstantiallyreducingthesystempeakwithstoragemayrequire8ormorehoursofduration(bluearrow)withnoPVdeployedinthesystem.However,thesamesystemwhenderiving20%ofannualenergyfromPVwouldrequireonly4hoursofdurationtoachievethesamelevelofnetloadreductionwithstorage(grayarrow).Figure3.IncreasedPVdeploymentnarrowsthenetloadpeakandincreasesthecapacitycreditofstorage14,00016,00018,00020,00022,00024,00026,00012AM4AM8AM12PM4PM8PMLoad(MW)HourofDayStorageDischargeChargingLoadNetLoadLongerpeaksrequiringmorestoredenergy010,00020,00030,00040,00050,00060,00006121824NetDemand(MW)HourofDay0%PV5%PV10%PV15%PV20%PV8ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Particularlyinstronglysummer-peakingsystems,PVdeploymentcansubstantiallyalterthecapacitycreditofstorageandthereforereducethedurationofstoragerequiredtoachieve100%capacitycredit.4.3Reason#3:Thecapacitycreditofstoragevariesbasedonstoragedeployment.Akeyelementistheimpactofstoragedeploymentitself;storageinherentlyshavesthepeakandcreateslongernetloadpeaks.Figure4providesanexampleshowinghowthesequentialadditionofstorageresultsinwiderpeaks,usingourFloridaexample.Figure4.ImpactofstoragedeploymentondurationneededInthisexample,maintaininghighcapacitycreditrequiresatransitionfrom4-hour-durationtoeventually10-hour-durationsystemsasstoragedeploymentincreases.4.4Reason#4:Thecapacitycreditofeven10+hourstoragetechnologiesmaybeverylowindecarbonizedenergysystems.TheexampleinFigure4showshowtheadditionofstorageincreasesthedurationneededforthenextunitofstoragetomaintainhighcapacitycredit.WhilethiscanbeoffsettosomeextentwithadditionalPV(Figure3),atsomepointpeaknetloadscanbeshiftedtoperiodsofrelativelylowPV(andwind)output.WithenoughVREandstoragedeployment,peaknetloadperiodscanlastseveraldays,whichwouldrequirefurtherstoragedeploymentstohavecorrespondingdurationstomaintainhighcapacitycredit.Anumberofstudies,includingseveralinTable1,examinescenariosthatapproachorachieve100%renewableenergysupplyandidentifythepotentialneedforstoragetochargeweeksorevenmonthsbeforeperiodsofhighnetdemand,andthendischargeformultipledays(48).Inthesecases,10hoursofstoragecanhaveverylowcapacitycredit,andeven100hoursmaybeinsufficientforsomeapplicationssuchasaddressingextendedoutagesoftransmissionincongestedloadpockets.22,00024,00026,00028,00030,00032,00034,00036,00038,00040,00042,00012AM4AM8AM12PM4PM8PMLoad(MW)HourofDay4HrStorage8HrStorage10HrStorageChargingLoad(allstorage)NetLoad(w/ocharging)Asmorestoragepowercapacityisdeployed,netloadpeaksbecomelongerIncreasingstoragepowercapacity9ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Overall,thesefourreasonslikelyexplainmuchofthediscrepancyindefinitionsseeninTable1.DefinitionsofLDESthatemphasizeitsabilitytoprovidefirmcapacityspanawiderangeofusecasesandgridconditions.Definitionsthatuse4-hourdurationmaybeconsideringthecurrentorhistoricalgrid,where4hoursmaybesufficienttomeetpeaksummerdemand.Definitionsthatuseadurationofmorethan4hoursmayreflectcurrentsystemsthatarewinter-peaking,ornear-futuresystemswhereshiftingloadpatternsandincreasedstoragedeploymentsresultinnetloadpeaksthatarelongerthan4-hours.Inaddition,LDESdefinitionsofdaysandbeyondareoftenbasedonstudiesofgridsthatrelymostlyonVREandmayrequiresufficientstoredenergytoaddressmultidayperiodsofbelow-averagewindorsolarenergysupply.5AFurtherComplication:TheImpactofEconomicandTechnologyCapabilitiesThepreviousexamplesshowthatthereisawiderangeinthedurationthresholdneededforhighcapacitycredit.However,communicatingthisintermsofa“need”forlong-durationstorage(oranyspecificminimumduration)overlooksthemorefundamentalissuethatthisneedisultimatelydrivenbytheeconomicsofcompetingstorageoptions,includingthepotentialroleofshorter-durationstorageappropriatelyderated.Forexample,Figure5showsascenarioinwhichthecombinationofVREandstorageproducesnetloadpeaksofabout10hours.7Asdiscussedpreviously,thiscouldbeprovidedbya10-hourdurationdevice,ora6-hourdurationdevicederatedto60%powercapacity.Inthesecases,thederatemeansthatthepowercomponentoftheshorter-durationstoragesystemispotentiallyoversized(morecostly)relativetothelonger-durationsystem.However,thisoversizedpowercapacityprovidestheplantadditionalopportunitytochargeduringhigh-powercurtailmentevents,particularlyinscenariosofsignificantPVdeployment.ThisisillustratedinFigure5bythesurplusgenerationwindow,whereaderated6-hourdurationdevicecantakeadvantageofitsabilitytochargeforshorterdurationsathighpower,providingenergytime-shiftingopportunitiesthatmaypartiallyorcompletelyoffsettheincreaseinpower-relatedcosts.7ThisimageshowsasimulationoffourdaysinJanuaryin2050inaregionoftheeasternUnitedStatesfromastudywhererenewableenergyprovidesabout80%ofthenation’selectricity(50).10ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.Figure5.Opportunitiesforderatedstoragedevicestotakeadvantageofhigh-powercurtailmenteventsTherelativepowercapacitiesofcharginganddischargingalsoplaceslimitstowhatlong-durationstoragecanachieveintermsofusableduration.IfweassumeourstoragedeviceinFigure5hasthesamechargeanddischargerate,thena10-hourdurationdevicewithan80%round-tripefficiency(RTE)wouldrequire14hourstocharge,representing24hoursofcontinuousoperation(eitherchargingordischargingat100%ofratedpower),whichistheupperboundoffeasibledailyoperation.Thisoperationwouldalsorequireenergyfromthegridtobeavailableduringtheentire14-hourperiodwhenthestoragedeviceisnotdischarging.Inreality,theremaybeshorterperiodsoflow-cost,off-peakgenerationavailableforcharging,whichcouldrequireahigherpowerratingtochargetomeetthedurationrequirement.Thisallpointstothefactthathighcapacitycreditcannotbebasedsolelyonthedurationofthestoredenergy,butalsoitsabilitytorechargeinatimelyandeconomicmanner,andanydefinitionofLDESmustaccountforthislimitation.Ultimately,thechoicebetweentechnologiesisdrivenbyeconomics,whichaddsyetanotherdimensiontodefininglong-durationstorage.Insomecasesthe“need”forstoragewithlongdurationcouldbemetwithderatedshorter-durationstorage,whoselossincapacityvaluecanbeoffsetbyincreasedenergyvaluefrommoreflexibleoperation.6ConclusionsThegrowingroleofvariablegenerationresourcesinthepowergridhasledtotheperception,withsignificantanalyticbasis,thattherewillultimatelybeaneedtomovebeyondstoragedeploymentswith4hoursofduration,currentlydominatedbylithium-ionbatteries.Thisperceptionhasresultedincallsfortheuseoflong-durationenergystorage,recognizingthepotentialfornetloadpeaksthatmayextendto8ormorehoursundervariousscenariosofstorageandrenewableenergydeployment.Yetthelackofauniformdefinitionoflongdurationinhibitsaclearcommunicationabouttheneedsofthecurrentandfuturegrid,includingscenariosapproaching100%decarbonizationrelyingprimarilyonrenewableenergy.Becauseofthedifferentmotivationsandpracticesforhowlong-durationstorageisdiscussedandanalyzed,wedonotrecommendthatasingledefinitionforlong-durationstoragebeused.AlthoughasinglequalitativedurationthresholdforLDESwouldbeusefulforcommunications,Surplusgenerationwindow(6hours)NetLoadPeak(10hours)12pm12pm12pm12pmPVOtherREStoragePeakerNormalloadLoad+Charging11ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.itisultimatelytooarbitraryanddoesnotidentifythedurationrequiredforstoragetoprovideresourceadequacytothegridasitevolves.However,wedorecommendthatqualitativedescriptionsforstoragedurationshouldalwaysbeaccompaniedbyaquantitativedefinition(e.g.,“inthisworkweconsiderlong-durationstoragesystemstohavedurationof4ormorehours”).Further,werecommendthatanalystsandotherswhoareactiveinthestoragespaceconsiderwhytheyhavechosentheirdefinitions.Forthosewhoaresimplylookingtoadoptalong-durationstoragedefinitionforconvenience,werecommendusingARPA-E’sdefinitionof10–100hoursgiventhat(1)thistimeframefulfillsthe“highcapacitycredit”requirementuntilsystemsareapproaching80%ormoreVREcontributions(withcommensuratedeploymentsofstoragetechnologies),(2)italignswiththelargestshareofliteraturecitations,and(3)itprovidesafairlycleardistinctionbetweenincumbenttechnologiessuchaslithium-ionbatteriesandtrulyseasonalstoragetechnologiessuchashydrogenorpowertogas.However,webelievethatcautionisneeded,particularlyiftheuseofthetermisusedtoimplyasystemwideneedtosupportresourceadequacy.Claimsthat10-hourdurationofstorageisneededtoprovideresourceadequacyinthecurrentgriddoesnotalignwithanalysisthatshows4hourscanbelargelysufficient,atleastintheneartermandinsomeregions.Furthermore,theneedfordurationsofmorethan4hoursislessenedbytheincreaseddeploymentofsolarPVandtheabilitytoderateshorter-durationstorage(ifsufficientlycost-effective),makingtheneedfortechnologieswithspecificdurationsasmuchofaneconomicissueasatechnicalone.Therefore,theneedforstoragewithdurationsof10ormorehourslargelyhingesonafuturegridwithaspecificsetofconditionsincludingregionalloadpatterns,renewableenergydeployment,previousstoragedeployments,andtheeconomicsofcompetingstorageoptions.Finally,asdiurnalstoragedeploymentincreases,thereisapointatwhichmultidaytoseasonalstoragemaybenecessarytosupportresourceadequacyandtoallowfurthercost-effectivedecarbonizationofrenewableresources.Asaresult,therecannotbeauniformandbroadlyapplicabledefinitionofLDESthathasasanunderlyingbasisitsabilitytosupportresourceadequacy.Thisoutcomemaybesomewhatinconclusivebutreflectsthegrowingcomplexityofresourceadequacyassessmentingeneral.Manyresources,includingwind,solarPV,anddemandresponsehavetime-,region-,anddeployment-basedvariationsincapacitycredit.Regulatoryandmarketframeworkswillneedtoevolvetoaccommodatethisreality.Theroleofstorageofvaryingdurationswillultimatelybedeterminedbytheireconomiccostsandbenefitsforprovidingresourceadequacyandtheotherservicesthatstoragecanprovide.Webelievethatasresearchers,analysts,andotherscarefullyconsiderthemotivationoftheirchosenstoragedefinitions,communicationamongstakeholderswillimproveandwewillbeabletocollectivelyadvancetheunderstandingoftheroleofstorageinpowersystems.12ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.References1.NorthAmericanElectricReliabilityCorporation(NERC).2021.GlossaryofTermsUsedinNERCReliabilityStandards.Washington,D.C.:NERC.https://www.nerc.com/files/glossary_of_terms.pdf2.U.S.EnergyInformationAdministration(EIA).2020.BatteryStorageintheUnitedStates:AnUpdateonMarketTrends.Washington,D.C.:EIA.https://www.eia.gov/analysis/studies/electricity/batterystorage/pdf/battery_storage.pdf3.S.Gonzato,K.Bruninx,andE.Delarue.2021.“Longtermstorageingenerationexpansionplanningmodelswithareducedtemporalscope.”AppliedEnergy298:117168.https://doi.org/10.1016/j.apenergy.2021.1171684.J.J.HargreavesandR.A.Jones.2020.“LongTermEnergyStorageinHighlyRenewableSystems.”FrontiersinEnergyResearch8:219.https://doi.org/10.3389/fenrg.2020.002195.R.E.CiezandD.Steingart.2020.“AsymptoticCostAnalysisofIntercalationLithium-IonSystemsforMulti-hourDurationEnergyStorage.”Joule4:597–614.https://doi.org/10.1016/j.joule.2020.01.0076.G.F.Frate,L.Ferrari,andU.Desideri.2020.“Multi-CriteriaEconomicAnalysisofaPumpedThermalElectricityStorage(PTES)WithThermalIntegration.”FrontiersinEnergyResearch8.https://doi.org/10.3389/fenrg.2020.000537.P.ByrneandP.Lalanne.2021.“ParametricStudyofaLong-DurationEnergyStorageUsingPumped-HydroandCarbonDioxideTranscriticalCycles.”Energies14.https://doi.org/10.3390/en141544018.M.Kintner-Meyer,P.Balducci,W.Colella,M.Elizondo,C.Jin,T.Nguyen,V.Viswanathan,andY.Zhang.2012.NationalAssessmentofEnergyStorageforGridBalancingandArbitrage:Phase1,WECC.Richland,WA:PacificNorthwestNationalLaboratory.https://www.pnnl.gov/main/publications/external/technical_reports/PNNL-21388.pdf9.SandiaNationalLaboratories.2021.ISSUEBRIEFLong-DurationEnergyStorage.Albuquerque,NM:SandiaNationalLaboratories.https://www.sandia.gov/ess-ssl/wp-content/uploads/2021/01/Issue-Brief-Long-Duration-Energy-Storage-January-2021_v2.pdf10.J.G.Simpson,G.Hanrahan,E.Loth,G.M.Koenig,andD.R.Sadoway.2021.“Liquidmetalbatterystorageinanoffshorewindturbine:Conceptandeconomicanalysis.”RenewableandSustainableEnergyReviews149:111387.https://doi.org/10.1016/j.rser.2021.11138711.U.Helman,B.Kaun,andJ.Stekli.2020.“DevelopmentofLong-DurationEnergyStorageProjectsinElectricPowerSystemsintheUnitedStates:ASurveyofFactorsWhichAreShapingtheMarket.”FrontiersinEnergyResearch8:275.https://doi.org/10.3389/fenrg.2020.53975213ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.12.L.Collins.2021.“Long-durationenergystoragesetfor‘verysteep’growthassectorenterscommercialphase.”RechargeNews.https://www.rechargenews.com/technology/long-duration-energy-storage-set-for-very-steep-growth-as-sector-enters-commercial-phase/2-1-96632613.A.GrundyandA.Colthorpe.2020.“Contenders:Longdurationenergystoragetechnologies,andwho’sbehindthem.”EnergyStorageNews.https://www.energy-storage.news/contenders-long-duration-energy-storage-technologies-and-whos-behind-them/14.F.MayrandF.Oldenburg.2021.“‘Longer-durationstorage’anditsroleinthefutureofenergy.”EnergyStorageNews.https://www.energy-storage.news/longer-duration-storage-and-its-role-in-the-future-of-energy/15.E.Childs,M.Roumpani,S.Dueñas,P.Sanchez,J.Gorman,M.Davidson,andL.Backer.2020.LongDurationEnergyStorageforCalifornia’sClean,ReliableGrid.Berkeley,CA:StrategenConsulting,LLC.16.WoodMackenziePower&Renewables.2020.U.S.EnergyStorageMonitor.17.N.Gallardo.2021.“Long-durationenergystorage:atechnoeconomiccomparativeanalysiswithcasestudiesinMexico.”Master’sthesis,KTHSchoolofIndustrialEngineeringandManagement.18.SandiaNationalLaboratories.2021.“‘BIG’EnergyStorage:PrioritiesandPathwaystoLong-DurationEnergyStorage.”DOELong-DurationEnergyStorageWorkshop,9–10March2021.19.LongDurationEnergyStorageAssociationofCalifornia.2021.“LongDurationEnergyStorageinCalifornia.”https://www.storeenergyca.org/background20.J.Spector.2020.“TheFirstMajorLong-DurationStorageProcurementHasArrived.”GreentechMedia.https://www.greentechmedia.com/articles/read/the-first-long-duration-storage-procurement-has-arrived21.M.TuttmanandS.Litzelman.2020.“WhyLong-DurationEnergyStorageMatters.”ARPA-EBlogPost.https://arpa-e.energy.gov/news-and-media/blog-posts/why-long-duration-energy-storage-matters22.U.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergy.2021.“LongDurationStorageShot.”https://www.energy.gov/eere/long-duration-storage-shot23.C.Amy,M.Pishahang,C.C.Kelsall,A.LaPotin,andA.Henry.2021.“High-temperaturePumpingofSiliconforThermalEnergyGridStorage.”Energy233:121105.https://doi.org/10.1016/j.energy.2021.12110524.Z.Ma,X.Wang,P.Davenport,andJ.Martinek.2021.“EconomicAnalysisofanElectricThermalEnergyStorageSystemUsingSolidParticlesforGridElectricityStorage.”14ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.ProceedingsoftheASME202115thInternationalConferenceonEnergySustainability,16–18June2021.https://doi.org/10.1115/ES2021-6172925.P.Albertus,J.Manser,andS.Litzelman.2020.“Long-durationelectricitystorageapplications,economics,andtechnologies.”Joule4:21–32.https://doi.org/10.1016/j.joule.2019.11.00926.J.A.Dowling,K.Z.Rinaldi,T.H.Ruggles,S.J.Davis,M.Yuan,F.Tong,N.S.Lewis,andK.Caldeira.2020.“RoleofLong-DurationEnergyStorageinVariableRenewableElectricitySystems.”Joule4:1907–1928.https://doi.org/10.1016/j.joule.2020.07.00727.M.Jafari,M.Korpås,andA.Botterud.2020.“Powersystemdecarbonization:Impactsofenergystoragedurationandinterannualrenewablesvariability.”RenewableEnergy156:1171–1185.https://doi.org/10.1016/j.renene.2020.04.14428.R.H.SchulteandF.C.Fletcher.2020.“OntheValueofTime-DiversifiedRenewableEnergyUsingInterregionalHVDCTransmission.”ElectricityJournal33(10):106861.https://doi.org/10.1016/j.tej.2020.10686129.J.Zhang,O.J.Guerra,J.Eichman,andM.A.Pellow.2020.“BenefitAnalysisofLong-DurationEnergyStorageinPowerSystemswithHighRenewableEnergyShares.”FrontiersinEnergyResearch8:313.https://doi.org/10.3389/fenrg.2020.52791030.F.J.deSisternes,J.D.Jenkins,andA.Botterud.2016.“Thevalueofenergystorageindecarbonizingtheelectricitysector.”AppliedEnergy175:368–379.https://doi.org/10.1016/j.apenergy.2016.05.01431.H.B.Ratz,R.Robichaud,L.Bird,andN.Hutchinson.2020.TheRoleofLong-DurationEnergyStorageinDeepDecarbonization:PolicyConsiderations.Washington,D.C.:WorldResourcesInstitute.http://www.indiaenvironmentportal.org.in/files/file/role-long-duration-energy-storage-deep-decarbonization-policy.pdf32.Z.Ma,J.D.McTigue,P.Li,R.Yang,Y.Ding,andC.N.Markides(Eds.).2020.“Long-DurationandLong-TermEnergyStorageforRenewableIntegration.”FrontiersResearchTopic.https://www.frontiersin.org/research-topics/10251/long-duration-and-long-term-energy-storage-for-renewable-integration33.J.D.Ogland-Hand,J.Bielicki,B.Adams,T.A.Buschek,andM.O.Saar.2021.“UsingSedimentaryBasinGeothermalResourcestoProvideLong-DurationEnergyStorage.”PresentedattheWorldGeothermalCongress,24–27October2021,Reykjavik,Iceland.https://doi.org/10.3929/ethz-b-00046759534.A.N.C.Edington.2019.“TheRoleofLongDurationEnergyStorageinDecarbonizingPowerSystems.”Master’sthesis,MassachusettsInstituteofTechnology.35.J.Gifford,Z.Ma,andP.Davenport.2020.“ThermalAnalysisofInsulationDesignforaThermalEnergyStorageSiloContainmentforLong-DurationElectricityStorage.”FrontiersinEnergyResearch8.https://doi.org/10.3389/fenrg.2020.0009915ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.36.C.K.HoandA.Ambrosini.2020.“Chapter12:ThermalEnergyStorageTechnologies.”In2020U.S.DOEEnergyStorageHandbook.Albuquerque,NM:SandiaNationalLaboratories.37.J.Bistline,G.Blanford,T.Mai,andJ.Merrick.2021.“Modelingvariablerenewableenergyandstorageinthepowersector.”EnergyPolicy156.https://doi.org/10.1016/j.enpol.2021.11242438.J.E.T.BistlineandG.J.Blanford.2021.“Impactofcarbondioxideremovaltechnologiesondeepdecarbonizationoftheelectricpowersector.”NatureCommunications12.https://doi.org/10.1038/s41467-021-23554-639.E.Du,H.Jiang,J.Xiao,J.Hou,N.Zhang,andC.Kang.2021.“Preliminaryanalysisoflong-termstoragerequirementinenablinghighrenewableenergypenetration:AcaseofEastAsia.”IETRenewablePowerGeneration15(6):1255–1269.https://doi.org/10.1049/rpg2.1210440.N.A.Sepulveda,J.D.Jenkins,A.Edington,D.S.Mallapragada,andR.K.Lester.2021.“Thedesignspaceforlong-durationenergystorageindecarbonizedpowersystems.”NatureEnergy6:506–516.https://doi.org/10.1038/s41560-021-00796-841.J.Spector.2020.“So,WhatExactlyIsLong-DurationEnergyStorage?”GreentechMedia.https://www.greentechmedia.com/articles/read/so-what-exactly-is-long-duration-storage-explained42.RenewableEnergyWorld.2017.“ALongerLookatLong-durationEnergyStorage.”https://www.renewableenergyworld.com/storage/a-longer-look-at-long-duration-energy-storage/#gref43.U.S.EnvironmentalProtectionAgency.2021.“SummaryoftheCleanAirAct.”https://www.epa.gov/laws-regulations/summary-clean-air-act44.A.W.Frazier,W.Cole,P.Denholm,D.Greer,andP.Gagnon.2020.“AssessingthepotentialofbatterystorageasapeakingcapacityresourceintheUnitedStates.”AppliedEnergy275:115385.https://doi.org/10.1016/j.apenergy.2020.11538545.K.CardenandN.Wintermantel.2019.EnergyStorageCapacityValueontheCAISOSystem.Hoover,AL:AstrapéConsulting.46.K.Carden,N.Wintermantel,andA.Krasny.2019.CapacityValueofEnergyStorageinPJM.Hoover,AL:AstrapéConsulting.47.R.Sioshansi,S.H.Madaeni,andP.Denholm.2014.“ADynamicProgrammingApproachtoEstimatetheCapacityValueofEnergyStorage.”IEEETransactionsonPowerSystems29:395–403.https://doi.org/10.1109/TPWRS.2013.227983948.P.Denholm,W.Cole,A.W.Frazier,K.Podkaminer,andN.Blair.2021.TheFourPhasesofStorageDeployment:AFrameworkfortheExpandingRoleofStorageintheU.S.Power16ThisreportisavailableatnocostfromtheNationalRenewableEnergyLaboratoryatwww.nrel.gov/publications.System.Golden,CO:NationalRenewableEnergyLaboratory.https://www.nrel.gov/docs/fy21osti/77480.pdf49.J.CochranandP.Denholm(Eds.).2021.TheLosAngeles100%RenewableEnergyStudy(LA100).Golden,CO:NationalRenewableEnergyLaboratory.NREL/TP-6A20-79444.https://maps.nrel.gov/la100/50.U.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergy.2021.SolarFuturesStudy.Washington,D.C.:EERE.https://www.energy.gov/sites/default/files/2021-09/Solar%20Futures%20Study.pdfNationalRenewableEnergyLaboratory15013DenverWestParkway,Golden,CO80401303-275-3000•www.nrel.govNRELprintsonpaperthatcontainsrecycledcontent.NRELisanationallaboratoryoftheU.S.DepartmentofEnergyOfficeofEnergyEfficiencyandRenewableEnergyOperatedbytheAllianceforSustainableEnergy,LLCNREL/TP-6A40-80583•November2021