Planning for a Distributed Disruption · 2020-01-06 · Planning for a Distributed Disruption: Innovative Practices for Incorporating Distributed Solar into Utility Planning Andrew

Planning for a Distributed Disruption: Innovative Practices for Incorporating Distributed Solar into Utility Planning

AndrewMills1,*,GalenBarbose1,JoachimSeel1,ChangguiDong2,TrieuMai2,BenSigrin2,Jare?Zuboy3

1LawrenceBerkeleyNaFonalLaboratory2NaFonalRenewableEnergyLaboratory

3IndependentConsultant

August2016

ThisprojectwasfundedbytheOfficeofEnergyEfficiencyandRenewableEnergy(SolarEnergyTechnologiesProgram)oftheU.S.DepartmentofEnergy

*[email protected]

Project Overview

•  Analystsprojectthatdistributedsolarphotovoltaics(DPV)willconFnuegrowingrapidlyacrosstheUnitedStates.

•  GrowthinDPVhascriFcalimplicaFonsforuFlityplanningprocesses,potenFallyaffecFngfutureinfrastructureneeds.

•  AppropriatetechniquestoincorporateDPVintouFlityplanningareessenFaltoensuringreliableoperaFonoftheelectricsystemandrealizingthefullvalueofDPV.

Context

•  ComparaFveanalysisandevaluaFonofroughly30recentplanningstudies,idenFfyinginnovaFvepracFces,lessonslearned,andstate-of-the-arttools.

Approach

•  Electricinfrastructureplanning(IRPs,transmission,distribuFon).•  FocusonthetreatmentofDPV,withemphasisonhowDPVgrowthisaccountedforwithinplanningstudies.

Scope

2

Organization of Report

3

Forecas)ngquanFtyofDPV•  SecFon3

RobustnessofdecisionstouncertaintyinDPVquanFty•  SecFon4

CharacterizeDPVasaresourceop)on•  SecFon5

Non-dispatchabilityofDPV•  SecFon6

Loca)on-specificfactorsofDPV•  SecFon7

ImpactofDPVonT&Dinvestments•  SecFon8

AvoidedlossesassociatedwithDPV•  SecFon9

ChangeswithDPVpenetra)on•  SecFon10

IntegraFonofDPVacrossplanningforums•  SecFon11


4










High End of 3rd Party Forecasts Suggests More DPV Than Is Considered By Utilities

5

0%

10%

20%

30%

HECO PG&E LADWP APS NVP TEP PNM ELA TVA FP&L GPC DOM ISO-NE NYISO DEI NSP PAC NWPCC PSE IPC WECC

Hawaii,California,DesertSouthwest South Northeast Midwest Northwest West

DPVpenetraKon(%ofretailsales)

Near-term(~2020)planneresKmate

Long-term(~2030)planneresKmate

Near-term(~2020)3rdpartyforecasts

2020Long-term(~2030)3rdpartyforecasts

Customer-adoption Modeling Brings Customer Decisions Into DPV Forecasting

6

Method DescripFon

Predic)veFactorsUsed

RecentinstallaFonrates

IncenFveprogramtargets

TechnicalpotenFal

PVeconomics

End-userbehaviors

S)pulatedForecast

Assumesend-pointDPVdeployment

HistoricalTrend

Extrapolatesfuturedeploymentfromhistoricaldata

X

Program-BasedApproach

Assumesprogramdeploymenttargetsreached

X

Customer-Adop)onModeling

UsesadopFonmodelsthatrepresentend-userdecisionmaking

X X X X

Shading

TiltAzimuth

Some Planners Use Customer-adoption Models for DPV Forecasting

TechnicalPotenFal

Willingness-to-adopt

Diffusion7

0Payback Period(Years)

2010Ult

imate

Share

(%

)

100%

50%

0%

0Years Since Introduction

3015

Realiz

ed M

ark

et

Pen.

100%

50%

0%

Adaptedfrom:Gagnonetal.2016

*illustra:ve

*illustra:ve


8










Multiple Per-scenario Plans Help Identify Resources that Depend on DPV Forecast

9

Method DescripFon

FactorsAddressed

Netloadchanges

GeneraFonporgoliochanges

Resource-acquisiFonstrategychanges

SingleForecast OneDPV-adopFonforecastused

SubjecttoSensi)vity

CostandperformanceofporgoliosevaluatedunderdifferentsensiFviFes

X

Per-ScenarioPlan

Capacityexpansionmodelsusedtodevelopleast-costplansforvariousscenarios

X X

Acquisi)onPathAnalysis

MulFpleper-scenarioplanscombinedwithtriggereventstoshaperesource-acquisiFonstrategy

X X X

Robustness of Decisions to Uncertainty in DPV Quantity

10

10,000

8,000

6,000

4,000

2,000

0

MW

2015

2020

2025 ...

...

Existing Resources

A

C

B

D

E

10,000

8,000

6,000

4,000

2,000

0

MW

2015

2020

2025 ...

...

Existing Resources

A

C

B

D

E

F

Candidateportfolio

10,000

8,000

6,000

4,000

2,000

0

MW

20

15

20

20

20

25 ...

...

Existing Resources

Load Net Load w/ low DPV

Net Load w/ high DPV

PlanwithLowDPV

PlanwithHighDPV

ForecastsofDPVAdopFonareUncertainàDevelopScenario-SpecificPlans

*illustra:ve

Use Difference In Plans to Identify Trigger Events and Resulting Changes to Plan

11

Source:PacifiCorp(2015)

TriggerEvent PlanningScenario ResourceAcquisi)onStrategy

NearTerm(2015-24) Long-Term(2025-34)

HighersustainedDGpenetraFonlevels

MoreaggressivetechnologycostreducFons,improvedtechnologyperformance,andhigherelectricityretailrates

•  ReduceforwardcontractacquisiFon

•  ConFnuetopursueEE

•  Reduceacquisi:onofgas-firedresources

•  BalanceFmingofthermalacquisiFonwithforwardcontractsandEE

LowersustainedDGpenetraFonlevels

LessaggressivetechnologycostreducFons,reducedtechnologyperformance,andlowerelectricityretailrates

•  IncreaseforwardcontractacquisiFon(primarilybeginning2024)

•  ConFnuetopursueEE

•  Increaseacquisi:onofgas-firedresources

•  BalanceFmingofthermalacquisiFonwithforwardcontractsandEE


12










Characterize DPV as a Resource Option

13

OnlyabouthalfoftheuFlityIRPsincludedDPVasaresource.TheprimarychallengeisdetermininghowtorepresentthecostsandbenefitsofDPVanddisFnguishitfromotherresourceslikeuFlity-scalePV(UPV).

*illustra:ve

Plan Characteris)cCapitalCostDiffersfromUPV?

CapacityFactorDiffersfromUPV?

CapacityCreditDiffersfromUPV?

AvoidedLosses

AvoidedTransmission

AvoidedDistribuFon

DEI(2015) X

GPC(2016) X X

HECO(2013) X X

IPC(2015) X

LADWP(2014)

X X

NSP(2015) X X X X

PG&E(2014) X X X X X X

PSE(2015) X X

TVA(2015) X


14










Evolving Approaches to Capture Non-dispatchability of DPV in Planning

•  Hourly DPV generation profiles capture some potential integration issues, including multi-hour ramping impacts and overgeneration.

•  Sub-hourly variability and uncertainty can be addressed through detailed integration studies.

•  DPV’s resource adequacy contribution is estimated by the capacity credit. 15

0

2

4

6

8

10

12

DEC IPC TEP APS NWPCC

SolarP

VIntegra;

onCost

($/M

Wh)


16










Propensity to Adopt Accounts for Factors Like Customer Demographics

17

Method DescripFon

Predic)veFactorsUsed

LocaFonofexisFngloadorpopulaFon

LocaFonofexisFngDPV

DetailedcustomercharacterisFcs

Propor)onaltoLoad

AssumesDPVisdistributedinproporFontoloadorpopulaFon

X

Propor)onaltoExis)ngDPV

AssumesDPVgrowsinproporFontoexisFngDPV X

PropensitytoAdopt

PredictscustomeradopFonbasedonfactorslikecustomerdemographicsorcustomerload

X X X

Predicting the Location of DPV Adoption Using Propensity to Adopt

18

C - 39

FIGURE 3-8 PG&E SERVICE AREA – SCENARIO 1 - ESTIMATED PV INSTALLED IN 2020 AND 2025

Source:PG&E2015DRP


19










Impact of DPV on T&D Investments: Potential Deferral Value

20 Source:AdaptedfromCohenetal.2016

1.02

1.03

1.04

1.05

1.06

1.07

1.08

0 500 1000 1500 2000

Maxim

umFeede

r

Volta

ge(p

u)

FeederPVPenetra@on(kW)

Allpenetra@onsacceptable,regardlessofloca@on

Somepenetra@onsacceptable,dependingontheloca@on

Nopenetra@onsacceptable,regardlessoftheloca@on

ImpactThreshold

MinimumHos@ngCapacity

MaximumHos@ngCapacity

Impact of DPV on T&D Investments: Hosting Capacity Analysis

21 Source:AdaptedfromEPRI2015

Impact of DPV on T&D Investments: Proactive Planning for DPV

22

$-

$0.1

$0.2

$0.3

$0.4

$0.5

$0.6

0 20 40 60 80 100

Upgrade

Cost($/W

pv)

PVPenetra@on(%ofFeederThermalRa@ng)

Cluster11AllOtherClusters

$-

$0.1

$0.2

$0.3

$0.4

0 20 40 60 80 100

Upgrade

Cost($/W

pv)

PVPenetra?on(%ofFeederThermalRa?ng)

Tradi?onalAdv.InverterAdv.Inverter+Storage

CoststoIncreasetheHosFngCapacityof

FourteenRepresentaFveFeederswith

TradiFonalGridUpgrades

CoststoIncreasetheHosFngCapacityofCluster11Comparing

TradiFonalGridUpgradesto

EmergingOpFons

Source:AdaptedfromNavigant2016a


23










Non-linear Relationship of Losses with Load Is Not Fully Captured in Average Loss Rate

24

Method DescripFon

LossCharacteris)csAddressed

VariedlossesoverFme

Marginallossratediffersfromaverage

Circuit-levellosses

AverageLossRate Appliessingleaveragelossrateacrossallhoursoftheyear

Time-Differen)atedAverageLossRate

DifferenFatesaveragelossratebasedonFmingofavoidedlosses X

MarginalAverageLossRate

Appliesandaveragemarginallossratebasedonlineloading X

Time-Differen)atedMarginalLossRate

DifferenFatesmarginallossratebyhour,month,or“peak”vs.“energy” X X

DetailedAnalysisofLosses

Useacircuit-levelmodeltoanalyzelosses X X X

MostuFliFesappeartouseanaveragelossrate,someuseaFme-differenFatedaveragelossrate,noneappeartousemarginallosses.


25










Studies Contain Few Detailed Discussions of Changes in Value at High Penetration

•  Georgia Power estimates value for different tranches of solar penetration, though details are often redacted.

•  LADWP identifies challenges with overgeneration on some low-load days with high solar.

•  LADWP also notes that electric car charging during the day may mitigate some of the overgeneration challenges.

•  Planners may also want to consider bundling DPV with other enabling technologies to prevent changes in value with high penetration.

26

Source:AdaptedfromGPC2016


27










Integrating DPV Into Planning Requires Coordination Across Generation,

Transmission, and Distribution Planning

•  One approach is to ensure that consistent scenarios, assumptions, and data are passed between different planning entities.

•  Some utilities are developing iterative planning practices that directly link the different planning forums.

•  Emerging tools can help to more fully evaluate the impacts and benefits of DPV and other DER from the distribution system up.

28

Highlights for Innovative Practices •  Forecasting: Customer-adoption modeling for DPV forecasts.

•  Robustness: Develop scenario-specific plans. Use differences in plans to identify “trigger events” that will result in changes to plan.

•  DPV as a Resource: Fully characterize DPV as a resource option. Consider in resource, transmission, and distribution planning.

•  Location of DPV: Forecast location of DPV to improve estimates of the T&D impact and location-specific value of DPV. Use propensity to adopt based on household/customer characteristics.

•  Impacts to T&D: Include DPV in forecast of peak load for transmission and distribution planning. Use hosting capacity analysis to identify needs for proactive distribution investments.

•  Non-dispatchability: Use hourly DPV profiles from SAM, CPR, or PVSyst. Calculate contribution to adequacy with ELCC. Use detailed integration studies to investigate sub-hourly challenges and costs.

•  Avoided Losses: Account for time-varying loss rates.

•  Changes with penetration: Identify costs and benefits for different tranches of DPV. Consider charging EVs when solar output is high. 29

Contact Information

For more information: Andrew D. Mills | [email protected] | 510.486.4059 http://emp.lbl.gov/reports/re

30

Sample of Integrated Resource Plans En)ty TitleandYearArizonaPublicService(APS) 2014IntegratedResourcePlanDominion(DOM) 2015IntegratedResourcePlanDukeEnergyCarolinas/Progress(DEC/DEP) 2014IntegratedResourcePlanDukeEnergyIndiana(DEI) 2015IntegratedResourcePlanEntergyLouisiana(ELA) 2015IntegratedResourcePlanFloridaPower&Light(FPL) TenYearPowerPlantSitePlan:2015-2024GeorgiaPower(GPC) 2016IntegratedResourcePlanHawaiianElectricCompanies(HECO) 2013IntegratedResourcePlanningReportIdahoPower(IPC) 2015IntegratedResourcePlanLosAngelesDepartmentofWaterandPower(LADWP) 2014IntegratedResourcePlanNevadaPower(NVP) 2015IntegratedResourcePlanNorthernStatesPower(NSP) 2015ResourcePlanNorthwestPowerandConservaFonCouncil(NWPCC) SeventhConservaFonandElectricPowerPlan(2016)PacificGas&Electric(PG&E) 2014BundledProcurementPlanPacifiCorp(PAC) 2015IntegratedResourcePlanPublicServiceNewMexico(PNM) 2014IntegratedResourcePlanPugetSoundEnergy(PSE) 2015IntegratedResourcePlanTennesseeValleyAuthority(TVA) IntegratedResourcePlan-2015FinalReportTri-StateG&T 2015IntegratedResourcePlan/ElectricResourcePlanTucsonElectricPower(TEP) 2014IntegratedResourcePlan

31

Sample of Transmission Planning Studies

32

En)ty TitleandYearCaliforniaISO(CAISO) 2015-2016TransmissionPlanningProcessUnifiedPlan.AssumpFonsandStudyPlanISONewEngland(ISO-NE) 2015RegionalSystemPlanNewYorkISO(NYISO) 2015LoadandCapacityDataReport:“GoldBook”PJM 2015RegionalTransmissionExpansionPlanWesternElectricityCoordinaFngCouncil(WECC)

IntegratedTransmissionandResourceAssessment:Summaryof2015PlanningAnalyses

Sample of Distribution Planning Studies En)ty TitleandYearCalifornia:PG&E 2015DistribuFonResourcesPlanCA:SouthernCaliforniaEdison(SCE) 2015DistribuFonResourcesPlanCA:SanDiegoGasandElectric(SDG&E) 2015DistribuFonResourcesPlanHawaii:HECO 2014DistributedGeneraFonInterconnecFonPlanHI:HECO 2015Circuit-LevelHosFngCapacityAnalysisMassachuse?s:NaFonalGrid 2015GridModernizaFonPlanNewYork:NYDepartmentofPublicService 2015DistributedSystemImplementaFonPlanGuidance

33

Planning for a Distributed Disruption · 2020-01-06 · Planning for a Distributed Disruption: Innovative Practices for Incorporating Distributed Solar into Utility Planning Andrew

Documents