Академический Документы
Профессиональный Документы
Культура Документы
RichardPerez
BenjaminL.Norris
ThomasE.Hoff
November2012
Preparedfor:
MidAtlanticSolarEnergyIndustriesAssociation
&
PennsylvaniaSolarEnergyIndustriesAssociation
Preparedby:
CleanPowerResearch
1700SoscolAve.,Suite22
Napa,CA94559
Acknowledgements
Thisreportwasfundedbythefollowingorganizations:
TheReinvestmentFundsSustainableDevelopmentFund
MidAtlanticSolarEnergyIndustriesAssociation
AdvancedSolarProducts
SMAAmericas
VoteSolar
RenewablePower
GeoscapeSolar
TheauthorswishtoexpresstheirgratitudetoRachelHoffforcollectingandanalyzingFERCfilingsfrom
thesixutilities,producingthePVfleetsimulations,andconductingthepeakloaddayanalysis;alsoto
PhilGruenhagenforresearchingandpreparingthePJMloadandpricingdata.
1
ExecutiveSummary
Thisreportpresentsananalysisofvalueprovidedbygridconnected,distributedPVinPennsylvaniaand
NewJersey.Theanalysisdoesnotprovidepolicyrecommendationsexcepttosuggestthateachbenefit
mustbeunderstoodfromtheperspectiveofthebeneficiary(utility,ratepayer,ortaxpayer).
Thestudyquantifiedtenvaluecomponentsandonecostcomponent,summarizedinTableES1.These
componentsrepresentthebenefits(andcosts)thataccruetotheutilities,ratepayers,andtaxpayersin
acceptingsolarontothegrid.Themethodologiesforquantifyingthesevaluesaredescribedfurtherin
Appendix2.
TableES1.Valuecomponentdefinitions.
ValueComponent Basis
FuelCostSavings Costofnaturalgasfuelthatwouldhavetobepurchased
foragasturbine(CCGT)plantoperatingonthemarginto
meetelectricloadsandT&Dlosses.
O&MCostSavings OperationsandmaintenancecostsfortheCCGTplant.
SecurityEnhancementValue Avoidedeconomicimpactsofoutagesassociateddueto
gridreliabilityofdistributedgeneration.
LongTermSocietalValue Potentialvalue(definedbyallothercomponents)ifthe
lifeofPVis40yearsinsteadoftheassumed30years.
FuelPriceHedgeValue Costtoeliminatenaturalgasfuelpriceuncertainty.
GenerationCapacityValue CosttobuildCCGTgenerationcapacity.
T&DCapacityValue FinancialsavingsresultingfromdeferringT&Dcapacity
additions.
MarketPriceReduction Wholesalemarketcostsincurredbyallratepayers
associatedwithashiftindemand.
EnvironmentalValue Futurecostofmitigatingenvironmentalimpactsofcoal,
naturalgas,nuclear,andothergeneration.
EconomicDevelopmentValue Enhancedtaxrevenuesassociatedwithnetjobcreation
forsolarversusconventionalpowergeneration.
(SolarPenetrationCost) Additionalcostincurredtoacceptvariablesolar
generationontothegrid.
TheanalysisrepresentsthevalueofPVforafleetofPVsystems(thatis,alargesetofsystems
generatingintothegrid).Fourdifferentfleetconfigurations(e.g.,fixed,southfacing,30degreetilt
angle)wereevaluatedateachofsevenlocations(Pittsburgh,PA;Harrisburg,PA;Scranton,PA;
2
Philadelphia,PA;Jamesburg,NJ;Newark,NJ;andAtlanticCity,NJ).Theselocationsrepresentadiversity
ofgeographicandeconomicassumptionsacrosssixutilityserviceterritories(DuquesneLightCo.,PPL
UtilitiesCorp,PECOUtilitiesCorp,JerseyCentralP&L,PSE&G,andAtlanticElectric).
Theanalysisrepresentedamoderateassumptionofpenetration:PVwastoprovide15%ofpeakelectric
loadforeachstudylocation(higherpenetrationlevelsresultinlowervalueperMWh).PVwasmodeled
usingSolarAnywhere,asolarresourcedatasetthatprovidestimeandlocationcorrelatedPVoutput
withloads.LoaddataandmarketpricingwastakenfromPJMforthesixzones,andutilityeconomic
inputswerederivedfromFERCsubmittals.AdditionalinputdatawastakenfromtheEIAandtheBureau
ofLaborStatistics(producerpriceindices).
LevelizedvalueresultsforthesevenlocationsareshowninFigureES1andTableES2.Detailedresults
forallscenariosareincludedinAppendix3.
FigureES1.Levelizedvalue($/MWh),bylocation(South30).
Thefollowingobservationsandconclusionsmaybemade:
TotalValue.Thetotalvaluerangesfrom$256perMWhto$318perMWh.Ofthis,thehighest
valuecomponentsaretheMarketPriceReduction(averaging$55perMWh)andtheEconomic
DevelopmentValue(averaging$44perMWh).
MarketPriceReduction.Thetwolocationsofhighesttotalvalue(HarrisburgandScranton)are
notedfortheirhighMarketPriceReductionvalue.Thismaybetheresultofagoodmatch
betweenLMPandPVoutput.Byreducingdemandduringthehighpricedhours,acostsavingsis
realizedbyallconsumers.Furtherinvestigationofthemethodsmaybewarrantedinlightoftwo
argumentsputforthbyFelder[32]:thatthemethodologydoesaddressinducedincreasein
demandduetopricereductions,andthatitonlyaddressesshortruneffects(ignoringthe
impactoncapacitymarkets).
EnvironmentalValue.Thestateenergymixisadifferentiatorofenvironmentalvalue.
Pennsylvania(withalargecomponentofcoalfiredgenerationinitsmix)leadstohigher
environmentalvalueinlocationsinthatstaterelativetoNewJersey.
T&DCapacityValue.T&Dcapacityvalueislowforallscenarios,withtheaveragevalueofonly
$3perMWh.Thismaybeexplainedbytheconservativemethodtakenforcalculatingthe
effectiveT&Dcapacity.
FuelPriceHedge.Thecostofeliminatingfuturefuelpurchasesthroughtheuseoffinancial
hedginginstrumentsisdirectlyrelatedtotheutilityscostofcapital.Thismaybeseenby
comparingthehedgevalueinJamesburgandAtlanticCity.Atautilitydiscountrateof5.68%,
JerseyCentralPower&Light(theutilityservingJamesburg)hasthelowestcalculatedcostof
capitalamongthesixutilitiesincludedinthestudy.Incontrast,PSE&G(theutilityserving
Newark)hasacalculateddiscountrateof8.46%,thehighestamongtheutilities.Thisisreflected
intherelativehedgevaluesof$24perMWhforJamesburgand$44perMWhforNewark,
nearlytwicethevalue.
GenerationCapacityValue.ThereisamoderatematchbetweenPVoutputandutilitysystem
load.Theeffectivecapacityrangesfrom28%to45%ofratedoutput,andthisisinlinewiththe
assignedPJMvalueof38%forsolarresources.
TableES2.LevelizedValueofSolar($/MWh),byLocation.
Pittsburgh Harrisburg Scranton Philadelphia Jamesburg Newark AtlanticCity
Energy
FuelCostSavings $41 $41 $41 $38 $42 $39 $41
O&MCostSavings $20 $20 $20 $18 $21 $19 $20
TotalEnergyValue $61 $60 $60 $56 $63 $58 $61
Strategic
SecurityEnhancementValue $23 $23 $23 $22 $23 $22 $22
LongTermSocietalValue $28 $29 $29 $27 $28 $28 $28
TotalStrategicValue $51 $52 $52 $49 $51 $50 $50
Other
FuelPriceHedgeValue $31 $42 $42 $47 $24 $44 $25
GenerationCapacityValue $22 $16 $17 $22 $19 $26 $18
T&DCapacityValue $6 $1 $1 $3 $1 $8 $2
MarketPriceReductionValue $35 $67 $69 $54 $52 $51 $54
EnvironmentalValue $54 $55 $55 $52 $23 $22 $23
EconomicDevelopmentValue $44 $45 $45 $42 $45 $44 $45
(SolarPenetrationCost) ($23) ($23) ($23) ($22) ($23) ($22) ($22)
TotalOtherValue $170 $203 $206 $199 $143 $173 $144
TotalValue $282 $315 $318 $304 $257 $280 $256
5
Contents
Acknowledgements.......................................................................................................................................ii
ExecutiveSummary.......................................................................................................................................1
Introduction:TheValueofPV.......................................................................................................................7
FuelCostSavings.......................................................................................................................................7
O&MCostSavings.....................................................................................................................................7
SecurityEnhancementValue....................................................................................................................8
LongTermSocietalValue..........................................................................................................................8
FuelPriceHedgeValue.............................................................................................................................8
GenerationCapacityValue.......................................................................................................................8
T&DCapacityValue...................................................................................................................................8
MarketPriceReduction............................................................................................................................9
EnvironmentalValue.................................................................................................................................9
EconomicDevelopmentValue..................................................................................................................9
SolarPenetrationCost..............................................................................................................................9
ValuePerspective......................................................................................................................................9
Approach.....................................................................................................................................................10
Locations.................................................................................................................................................10
PenetrationLevel....................................................................................................................................11
FleetConfigurations................................................................................................................................11
ScenariosandFleetModeling.................................................................................................................12
Results.........................................................................................................................................................13
UtilityAnalysis.........................................................................................................................................13
PVTechnicalAnalysis..............................................................................................................................13
ValueAnalysis.........................................................................................................................................15
6
FutureWork................................................................................................................................................20
Appendix1:DetailedAssumptions.............................................................................................................21
Appendix2:Methodologies........................................................................................................................23
Overview.................................................................................................................................................23
UnitsofResults.......................................................................................................................................23
PVProductionandLossSavings..............................................................................................................24
LossSavings.............................................................................................................................................26
FuelCostSavingsandO&MCostSavings...............................................................................................28
SecurityEnhancementValue..................................................................................................................29
LongTermSocietalValue........................................................................................................................30
FuelPriceHedgeValue...........................................................................................................................31
GenerationCapacityValue.....................................................................................................................32
T&DCapacityValue.................................................................................................................................33
MarketPriceReductionValue................................................................................................................33
EnvironmentalValue...............................................................................................................................43
EconomicDevelopmentValue................................................................................................................45
SolarPenetrationCost............................................................................................................................47
MethodologyReferences........................................................................................................................48
Appendix3:DetailedResults......................................................................................................................51
Pittsburgh................................................................................................................................................52
Harrisburg...............................................................................................................................................54
Scranton..................................................................................................................................................57
Philadelphia.............................................................................................................................................59
Jamesburg...............................................................................................................................................61
Newark....................................................................................................................................................63
AtlanticCity.............................................................................................................................................65
7
Introduction:TheValueofPV
ThisreportattemptstoquantifythevalueofdistributedsolarelectricityinPennsylvaniaandNew
Jersey.Itusesmethodologiesandanalyticaltoolsthathavebeendevelopedoverseveralyears.The
frameworksupposesthatPVislocatedinthedistributionsystem.PVthatislocatedclosetotheloads
providesthehighestvalueperunitofenergytotheutilitybecauselinelossesareavoided,thereby
increasingthevalueofsolarrelativetocentrallylocatedresources.
ThevalueofPVmaybeconsideredtheaggregateofseveralcomponents,eachestimatedseparately,
describedbelow.ThemethodsusedtocalculatevaluearedescribedinmoredetailintheAppendices.
FuelCostSavings
DistributedPVgenerationoffsetsthecostofpowergeneration.EachkWhgeneratedbyPVresultsin
onelessunitofenergythattheutilityneedstopurchaseorgenerate.Inaddition,distributedPVreduces
systemlossessothatthecostofthewholesalegenerationthatwouldhavebeenlostmustalsobe
considered.
Underthisstudy,thevalueisdefinedasthecostofnaturalgasfuelthatwouldotherwisehavetobe
purchasedtooperateagasturbine(CCGT)plantandmeetelectricloadsandT&Dlosses.Thestudy
presumesthattheenergydeliveredbyPVdisplacesenergyatthisplant.
Whethertheutilityreceivesthefuelcostsavingsdirectlybyavoidingfuelpurchases,orindirectlyby
reducingwholesalepowerpurchases,themethodofcalculatingthevalueisthesame.
O&MCostSavings
UnderthesamemechanismdescribedforFuelCostSavings,theutilityrealizesasavingsinO&Mcosts
duetodecreaseduseoftheCCGTplant.Thecostsavingsareassumedtobeproportionaltotheenergy
avoided,includinglosssavings.
8
SecurityEnhancementValue
ThedeliveryofdistributedPVenergycorrelatedwithloadresultsinanimprovementinoverallsystem
reliability.Byreducingtheriskofpoweroutagesandrollingblackouts,economiclossesarereduced.
LongTermSocietalValue
Thestudyperiodistakenas30years(thenominallifeofPVsystems),andthecalculationofvalue
componentsincludesthebenefitsprovidedoverthisstudyperiod.However,itispossiblethatthelife
canbelongerthan30years,inwhichcasethefullvaluewouldnotbeaccountedfor.Thislongterm
societalvalueisthepotentialextendedbenefitofallvaluecomponentsovera10yearperiodbeyond
thestudyperiod.Inotherwords,iftheassumedlifewere40yearsinsteadof30,theincreaseintotal
valueisthelongtermsocietalvalue.
FuelPriceHedgeValue
PVgenerationisinsensitivetothevolatilityofnaturalgasorotherfuelprices,andthereforeprovidesa
hedgeagainstpricefluctuation.Thisisquantifiedbycalculatingthecostofariskmitigationinvestment
thatwouldprovidepricecertaintyforfuturefuelpurchases.
GenerationCapacityValue
InadditiontothefuelandO&Mcostsavings,thetotalcostofpowergenerationincludescapitalcost.To
theextentthatPVdisplacestheneedforgenerationcapacity,itwouldbevaluedasthecapitalcostof
displacedgeneration.Thekeytovaluingthiscomponentistodeterminetheeffectiveloadcarrying
capability(ELCC)ofthePVfleet,andthisisaccomplishedthroughananalysisofhourlyPVoutput
relativetooverallutilityload.
T&DCapacityValue
Inadditiontocapitalcostsavingsforgeneration,PVpotentiallyprovidesutilitieswithcapitalcost
savingsonT&Dinfrastructure.Inthiscase,PVisnotassumedtodisplacecapitalcostsbutratherdefer
theneed.ThisisbecauselocalloadscontinuetogrowandeventuallynecessitatetheT&Dcapital
investment.Therefore,thecostsavingsrealizedbydistributedPVismerelythecostofcapitalsavedin
theinterveningperiodbetweenPVinstallationandthetimeatwhichloadsagainreachthelevelof
effectivePVcapacity.
9
MarketPriceReduction
PVgenerationreducestheamountofloadontheutilitysystems,andthereforreducestheamountof
energypurchasedonthewholesalemarket.Thedemandcurveshiftstotheleft,andthemarketclearing
priceisreduced.Thus,thepresenceofPVnotonlydisplacestheneedforenergy,butalsoreducesthe
costofwholesaleenergytoallconsumers.Thisvalueisquantifiedthroughananalysisofthesupply
curveandthereductionindemand.
EnvironmentalValue
OneoftheprimarymotivesforPVandotherrenewableenergysourcesistoreducetheenvironmental
impactofpowergeneration.Environmentalbenefitscoveredinthisanalysisrepresentfuturesavingsfor
mitigatingenvironmentaldamage(sulfurdioxideemissions,watercontamination,soilerosion,etc.).
EconomicDevelopmentValue
DistributedPVprovideslocaljobs(e.g.,installers)athigherratesthanconventionalgeneration.These
jobs,inturn,translatetotaxrevenuebenefitstoalltaxpayers.
SolarPenetrationCost
InadditiontothevalueprovidedbyPV,therearecoststhatmustbefactoredinasnecessarytoaccept
variablesolargenerationontothegrid.Infrastructuralandoperationalexpenseswillbeincurredto
managetheflowofnondispatchablePVresources.Thesecostsareincludedasanegativevalue.
ValuePerspective
Thevalueofsolaraccrueseithertotheelectricutilityortosociety(ratepayersandtaxpayers),
dependinguponcomponent.Forexample,PVreducestheamountofwholesaleenergyneededtoserve
load,resultinginsavingstotheutility.Ontheotherhand,environmentalmitigationcostsaccrueto
society.
10
Approach
Locations
Sevenlocationswereselectedtoprovidebroadgeographicalandutilitycoverageinthetwostatesof
interest(seeTable1).FourlocationswereselectedinPennsylvaniarepresentingthreeutilities
1
and
threelocationswereselectedinNewJersey,eachservedbyaseparateutility.
Table1.Studylocationsummary.
Location Utility
2011Utility
PeakLoad
(MW)
PVFleet
Capacity
(MW)
PA
1 Pittsburgh DuquesneLightCo. 3,164 475
2 Scranton PPLUtilitiesCorp. 7,527 1,129
3 Harrisburg PPLUtilitiesCorp. 7,527 1,129
4 Philadelphia PECOEnergyCo. 8,984 1,348
NJ
5 Jamesburg JerseyCentralP&L 6,604 991
6 Newark PSE&G 10,933 1,640
7 AtlanticCity AtlanticCityElectric 2,956 443
Theselocationsrepresentadiversityofinputassumptions:
Thelocationsspantwostates:PAandNJ.Thesestatesdifferingenerationmix(percentageof
coal,gas,nuclear,etc.),andthisisreflectedindifferentenvironmentalcostassumptions(see
Appendix2).
Thelocationsdifferinsolarresource.
1
ScrantonandHarrisburgarebothservedbyPPLUtilities.
11
Thelocationsrepresentsixdifferentutilityserviceterritories.Eachoftheseutilitiesdifferby
costofcapital,hourlyloads,T&Dlossfactors,distributionexpansioncosts,andgrowthrate.
PenetrationLevel
Fleetcapacitywassetto15%oftheutilitypeakload.Thisassumptionwasintendedtorepresenta
moderatelongtermpenetrationlevel.
ThevalueofsolarperMWhdecreaseswithincreasingpenetrationforseveralreasons:
ThematchbetweenPVoutputandloadsisreduced.AsmorePVisaddedtotheresourcemix,
thepeakshiftstononsolarhours,therebylimitingtheabilityofPVtosupportthepeak.
Linelossesarerelatedtothesquareoftheload.Consequently,thegreatestmarginalsavings
providedbyPVisachievedwithsmallamountsofPV.Byaddinglargerandlargerquantitiesof
PV,thelosssavingscontinuetobegained,butatdecreasingrates.
Similarly,themarketpricesarenonlinear,andPVismosteffectiveincausingmarketprice
reductionwithsmallPVcapacity.
Basedontheaboveconsiderations,thisstudyisintendedtorepresentamoderateleveloflongtermPV
penetration.Withpenetrationlevelslessthan15%,thevalueofsolarwouldbeexpectedtobehigher
thantheresultsobtainedinthisstudy.
PeakloadsforeachutilitywereobtainedfromhourlyloaddatacorrespondingtoPJMloadzones,and
thesewereusedtosetthefleetcapacityasshowninthetable.
FleetConfigurations
FourPVsystemconfigurationswereincludedinthestudy:
South30(southfacing,30degreetilt,fixed)
Horizontal(fixed)
West30(westfacing,30degreetilt,fixed)
1Axis(trackingat30degreetilt)
Thesewereselectedinordertocapturepossiblevariationsinvalueduetothedifferentproduction
profiles.Forexample,Westfacingsystemsaresometimesfoundtobethebestmatchwithutilityloads
12
andhavethepotentialtoprovidemorecapacitybenefits.Ontheotherhand,trackingsystemsdeliver
moreenergyperunitofratedoutput,sotheyhavethepotentialtooffermoreenergybenefits(e.g.,fuel
costsavings).
ScenariosandFleetModeling
Valuewasdeterminedforeachof28scenarios(fourfleetconfigurationsateachofsevenlocations).For
modelingpurposes,fleetsweredescribedbylatitudeandlongitudecoordinates,ACrating,amodule
deratefactor(90%),inverterefficiency(95%)andotherlossfactor(90%).Thesefactorswereconsistent
acrossallscenarios.
Fleetsweremodeledforallhoursof2011usingSolarAnywheresatellitederivedirradiancedataand
simulationmodelwitha10kmx10kmpixelresolution.
2
Underthisprocedure,thefleetoutputforeach
scenarioislocationandtimecorrelatedwithhourlyPJMzonalloads.
2
http://www.solaranywhere.com.
13
Results
UtilityAnalysis
UtilityanalysisresultsareshowninTable2,obtainedfromananalysisofFERCfilingsandPJMhourly
datausingmethodsdevelopedpreviouslyforNYSERDA.
3
Theseinclude:
Utilitydiscountrate
Utilitysystemlossdata
Distributionexpansioncosts(presentvalue)
Distributionloadgrowthrate
Distributionlossdata
Notethatactualutilitycostsareusedinthisanalysisbecausetheyarethebasisofvalue.Forthisreason,
theutilitycostofcapitalisrequired(e.g.,anassumedorcommonvaluecannotbeused).Theresults
maythereforediffer,inpart,duetodifferencesinutilitydiscountrate.
PVTechnicalAnalysis
AsummaryoffleettechnicalperformanceresultsispresentedinTable3.Annualenergyproductionis
themodeledoutputfor2011.Capacityfactoristheannualenergyproductionrelativetoabaseload
plantoperatingat100%availabilitywiththesameratedoutput.GenerationcapacityisEffectiveLoad
CarryingCapability(ELCC)expressedasapercentageofratedcapacity.T&DCapacityisameasureofthe
directannualpeakloadreductionprovidedbythePVsystemexpressedasapercentageofrated
capacity.
3
NorrisandHoff,PVValuationTool,FinalReport(DRAFT),NYSERDA,May2012.
14
Table2.Utilityanalysisresults.
Table3.Technicalresults,bylocation(South30).
ValueAnalysis
Figure1showsthevalueresultsinlevelizeddollarsperMWhgenerated.Figure2showsthedatain
dollarsperkWinstalled.ThisdataisalsopresentedintabularforminTable4andTable5.Detailed
resultsforindividuallocationsareshowninAppendix3.
Thetotalvaluerangesfrom$256perMWhto$318perMWh.Ofthis,thehighestvaluecomponentsare
theMarketPriceReduction(averaging$55perMWh)andtheEconomicDevelopmentValue(averaging
$44perMWh).
ThedifferencesbetweenTable4andTable5areduetodifferencesinthecostofcapitalbetweenthe
utilities.Forexample,AtlanticCityhasthehighestvalueperinstalledkW,butAtlanticCityElectrichas
oneofthelowestcalculateddiscountrates(Table2).Therefore,whenthisvalueislevelizedoverthe30
yearstudyperiod,itrepresentsarelativelylowvalue.
Otherobservations:
MarketPriceReduction.Thetwolocationsofhighesttotalvalue(HarrisburgandScranton)are
notedfortheirhighMarketPriceReductionvalue.Thismaybetheresultofagoodmatch
betweenLMPandPVoutput.Byreducingdemandduringthehighpricedhours,acostsavingsis
realizedbyallconsumers.Furtherinvestigationofthemethodsmaybewarrantedinlightoftwo
argumentsputforthbyFelder[32]:thatthemethodologydoesaddressinducedincreasein
demandduetopricereductions,andthatitonlyaddressesshortruneffects(ignoringthe
impactoncapacitymarkets).
EnvironmentalValue.Thestateenergymixisadifferentiatorofenvironmentalvalue.
Pennsylvania(withalargecomponentofcoalfiredgenerationinitsmix)leadstohigher
environmentalvalueinlocationsinthatstaterelativetoNewJersey.AsdescribedinAppendix2,
thePAgenerationmixisdominatedbycoal(48%)comparedtoNJ(10%).
T&DCapacityValue.T&Dcapacityvalueislowforallscenarios,withtheaveragevalueofonly
$3perMWh.Thismaybeexplainedbytheconservativemethodtakenforcalculatingthe
effectiveT&Dcapacity.
FuelPriceHedge.Thecostofeliminatingfuturefuelpurchasesthroughtheuseoffinancial
hedginginstrumentsisdirectlyrelatedtotheutilityscostofcapital.Thismaybeseenby
comparingthehedgevalueinJamesburgandAtlanticCity.Atarateof5.68%,JerseyCentral
Power&Light(theutilityservingJamesburg)hasthelowestcalculatedcostofcapitalamongthe
16
sixutilitiesincludedinthestudy.Incontrast,PSE&G(theutilityservingNewark)hasacalculated
discountrateof8.46%,thehighestamongtheutilities.Thisisreflectedintherelativehedge
valuesof$24perMWhforJamesburgand$44perMWhforNewark,nearlytwicethevalue.
Figure1.Levelizedvalue($/MWh),bylocation(South30).
17
Figure2.Value($/kW),bylocation(South30).
18
Table4.Value(levelized$/MWh),bylocation(South30).
Table5.Value($/kW),bylocation(South30).
FutureWork
Inthecourseofconductingthisstudy,severalobservationsweremadethatsuggestfurtherrefinement
totheseresultsshouldbeconsidered:
Themarketpricereductionestimatedaspartofthepresentstudywillhavetobeascertainedas
PVdevelopsandpenetratestheNJandPAgrids.Inparticular,theimpactofPVinducedprice
reductiononloadgrowth,hencefeedbacksecondaryloadgrowthinducedmarketprice
increaseassuggestedbyFelder[32]shouldbequantified.Inaddition,thefeedbackofmarket
pricereductiononcapacitymarketswillhavetobeinvestigated.
Inthisstudy15%PVcapacitypenetrationwasassumedamountingtoatotalPVcapacityof
7GWacrossthesevenconsideredutilityhubs.Sincebothintegrationcostincreasesandcapacity
valuediminisheswithpenetration,itwillbeworthwhiletoinvestigateotherpenetration
scenarios.ThismaybeparticularlyusefulforPAwherethepenetrationissmallerthanNJ.In
addition,itmaybeusefultoseethescenarioswithpenetrationabove15%.Forthesecases,it
wouldbepertinenttoestablishthecostofdisplacing(nuclear)baseloadgenerationwithsolar
generation
4
sincethisquestionisoftenbroughttotheforefrontbyenvironmentallyconcerned
constituentsindenselypopulatedareasofNJandPA.
Othersensitivitiesmaybeimportanttoassessaswell.Sensitivitiestofuelpriceassumptions,
discountrates,andotherfactorscouldbeinvestigatedfurther.Inparticularthechoicemade
heretousedocumentedutilityspecificdiscountratesanditsimpactontheperMWhlevelized
results
5
couldbequantifiedandcomparedtoanassumptionusingacommondiscountrate
representativeofaverageregionalbusinesspractice.
TheT&Dvaluesderivedforthepresentanalysisarebasedonutilitywideaverageloads.
Becausethisvalueisdependentupontheconsidereddistributionsystemscharacteristicsin
particularloadgrowth,customermixandequipmentagetheT&Dvaluemayvaryconsiderably
fromonedistributionfeedertoanother.Itwouldthereforebeadvisabletotakethisstudyone
stepfurtherandsystematicallyidentifythehighestvalueareas.Thiswillrequirethe
collaborationoftheservicingutilitiestoproviderelevantsubsystemdata.
4
Consideringintegrationsolutionsincludingstorage,wind/PVsynergyandgasgenerationbackup.
5
NotethattheperkWvalueresultsaremuchlessdependentuponthediscountrate
21
Appendix1:DetailedAssumptions
InputassumptionsthatarecommonacrossallofthescenariosareshowninTable6.
Table6.Inputassumptionsandunitscommontoallscenarios.
PVdegradationisassumedtobe0.50%peryearindicatingthattheoutputofthesystemwilldegrade
overtime.Thisisaconservativeassumption(PVdegradationislikelytobelessthan0.5%peryear).
Studiesoftenignoredegradationaltogetherbecausetheeffectissmall,butitisincludedherefor
completeness.
Thestudyperiodistakenas30years,correspondingtotypicalPVlifetimeassumptions.
PVisassumedtodisplacepowergeneratedfrompeakingplantsfueledbynaturalgas.Gasturbine
capital,O&M,heatrate,andescalationvaluesaretakenfromtheEIA.
6
Plantdegradationisassumedto
bezero.
6
UpdatedCapitalCostEstimatesforElectricityGenerationPlants,U.S.EnergyInformationAdministration,
November2010,availableathttp://www.eia.gov/oiaf/beck_plantcosts/pdf/updatedplantcosts.pdf.Takenfrom
Table1,page7.Costsareescalatedto2012dollars.
INPUTASSUMPTIONS
PVCharacteristics
PVDegradation 0.50% peryear
PVSystemLife 30 years
GenerationFactors
GenCapacityCost $1,045 perkW
GenHeatRate(FirstYear) 7050 BTU/kWh
GenPlantDegradation 0.00% peryear
GenO&MCost(FirstYear) $12.44 perMWh
GenO&MCostEscalation 3.38% peryear
GarverPercentage 5.00% PctofAnnPeak
NGWholesaleMarketFactors
EndofTermNGFuturesPriceEscalation 2.33% peryear
22
CostsforgenerationO&Mareassumedtoescalateat3.38%,calculatedfromthechangeinProducer
PriceIndex(PPI)fortheTurbineandpowertransmissionequipmentmanufacturingindustry
7
overthe
period2004to2011.
NaturalgaspricesusedinthefuelpricesavingsvaluecalculationareobtainedfromtheNYMEXfutures
prices.Theseprices,however,areonlyavailableforthefirst12years.Ideally,onewouldhave30years
offuturesprices.Asaproxyforthisvalue,itisassumedthatescalationafteryear12isconstantbased
onhistoricallylongtermpricestocovertheentire30yearsofthePVservicelife(years13to30).The
EIApublishednaturalgaswellheadpricesfrom1922tothepresent.
8
Itisassumedthatthepriceofthe
NGfuturesescalatesatthesamerateasthewellheadprices.
9
A30yeartimehorizonisselectedwith
1981gaspricesat$1.98perthousandcubicfeetand2011pricesat$3.95.Thisresultsinanaturalgas
escalationrateof2.33%.
7
PPIdataisdownloadablefromtheBureauindustryindexselectedwastakenasthemostrepresentativeofpower
generationO&M.BLSdoespublishanindexforElectricpowergenerationbutthisisassumed.
8
USNaturalGasPrices(Annual),EIA,releasedate2/29/2012,availableat
http://www.eia.gov/dnav/ng/ng_pri_sum_dcu_nus_m.htm.
9
TheexactnumbercouldbedeterminedbyobtainingoverthecounterNGforwardprices.
23
Appendix2:Methodologies
Overview
ThemethodologiesusedinthepresentprojectdrewuponstudiesperformedbyCPRforotherstates
andutilities.Inthesestudies,thekeyvaluecomponentsprovidedbyPVweredeterminedbyCPR,using
utilityprovideddataandothereconomicdata.
Theabilitytodeterminevalueonasitespecificbasisisessentialtothesestudies.Forexample,theT&D
CapacityValuecomponentdependsupontheabilityofPVtoreducepeakloadsonthecircuits.An
analysisofthisvalue,then,requires:
Hourbyhourloadsondistributioncircuitsofinterest.
HourlyexpectedPVoutputscorrespondingtothelocationofthesecircuitsandexpectedPV
systemdesigns.
Localdistributionexpansionplancostsandloadgrowthprojections.
UnitsofResults
Thediscountingconventionassumedthroughoutthereportisthatenergyrelatedvaluesoccuratthe
endofeachyearandthatcapacityrelatedvaluesoccurimmediately(i.e.,nodiscountingisrequired).
10
ThePresentValueresultsareconvertedtoperunitvalue(PresentValue$/kW)bydividingbythesizeof
thePVsystem(kW).AnexampleofthisconversionisillustratedinFigure3forresultsfromaprevious
study.TheyaxispresentstheperunitvalueandthexaxispresentssevendifferentPVsystem
configurations.ThefigureillustrateshowvaluecomponentscanbesignificantlyaffectedbyPVsystem
configuration.Forexample,thetrackingsystems,byvirtueoftheirenhancedenergyproduction
capability,providegreatergenerationbenefits.
10
Theeffectofthiswillbemostapparentinthatthesummationsofcashflowsstartwiththeyearequalto1
ratherthan0.
24
Figure3.Sampleresults.
Thepresentvalueresultsperunitofcapacity($/kW)areconvertedtolevelizedvalueresultsperunitof
energy($/MWh)bydividingpresentvalueresultsbythetotalannualenergyproducedbythePVsystem
andthenmultiplyingbyaneconomicfactor.
PVProductionandLossSavings
PVSystemOutput
AnaccuratePVvalueanalysisbeginswithadetailedestimateofPVsystemoutput.Someoftheenergy
basedvaluecomponentsmayonlyrequirethetotalamountofenergyproducedperyear.Othervalue
components,however,suchastheenergylosssavingsandthecapacitybasedvaluecomponents,
requirehourlyPVsystemoutputinordertodeterminethetechnicalmatchbetweenPVsystemoutput
andtheload.Asaresult,thePVvalueanalysisrequirestime,location,andconfigurationspecificPV
systemoutputdata.
Forexample,supposethatautilitywantstodeterminethevalueofa1MWfixedPVsystemorientedat
a30tiltfacinginthesouthwestdirectionlocatedatdistributionfeederA.DetailedPVoutputdata
thatistimeandlocationspecificisrequiredoversomehistoricalperiod,suchasfromJan.1,2001to
Dec.31,2010.
25
Methodology
ItwouldbetemptingtousearepresentativeyeardatasourcesuchasNRELsTypicalMeteorological
Year(TMY)dataforpurposesofperformingaPVvalueanalysis.Whilethesedatamayberepresentative
oflongtermconditions,theyare,bydefinition,nottimecorrelatedwithactualdistributionlineloading
onanhourlybasisandarethereforenotusableinhourlysidebysidecomparisonsofPVandload.Peak
substationloadsmeasured,say,duringamidAugustfivedayheatwavemustbeanalyzedalongsidePV
datathatreflectthesamefivedayconditions.Consequently,atechnicalanalysisbasedonanything
otherthantimeandlocationcorrelatedsolardatamaygiveincorrectresults.
CPRsSolarAnywhereandPVSimulatorsoftwareserviceswillbeemployedunderthisprojectto
createtimecorrelatedPVoutputdata.SolarAnywhereisasolarresourcedatabasecontainingalmost14
yearsoftimeandlocationspecific,hourlyinsolationdatathroughoutthecontinentalU.S.andHawaii.
PVSimulator,availableintheSolarAnywhereToolkit,isaPVsystemmodelingservicethatusesthis
hourlyresourcedataanduserdefinedphysicalsystemattributesinordertosimulateconfiguration
specificPVsystemoutput.
TheSolarAnywheredatagridwebinterfaceisavailableatwww.SolarAnywhere.com(Figure4).The
structureofthedataallowstheusertoperformadetailedtechnicalassessmentofthematchbetween
PVsystemoutputandloaddata(evendowntoaspecificfeeder).Together,thesetwotoolsenablethe
evaluationofthetechnicalmatchbetweenPVsystemoutputandloadsforanyPVsystemsizeand
orientation.
PreviousPVvalueanalysesweregenerallylimitedtoasmallnumberofpossiblePVsystem
configurationsduetothedifficultyinobtainingtimeandlocationspecificsolarresourcedata.Thisnew
valueanalysissoftwareservice,however,willintegrateseamlesslywithSolarAnywhereand
PVSimulator.ThiswillallowuserstoreadilyselectanyPVsystemconfiguration.Thiswillallowforthe
evaluationofacomprehensivesetofscenarioswithessentiallynoadditionalstudycost.
26
Figure4.SolarAnywheredataselectionmap.
LossSavings
Introduction
Distributedresourcesreducesystemlossesbecausetheyproducepowerinthesamelocationthatthe
powerisconsumed,bypassingtheT&Dsystemandavoidingtheassociatedlosses.
Losssavingsarenottreatedasastandalonebenefitundertheconventionusedinthismethodology.
Rather,theeffectoflosssavingsisincludedseparatelyforeachvaluecomponent.Forexample,inthe
sectionthatcoversthecalculationofEnergyValue,thequantityofenergysavedbytheutilityincludes
boththeenergyproducedbyPVandtheamountthatwouldhavebeenlostduetoheatinginthewires
iftheloadwereservedfromaremotesource.Thetotalenergythatwouldhavebeenprocuredbythe
utilityequalsthePVenergyplusavoidedlinelosses.Losssavingscanbeconsideredasortofadderfor
eachbenefitcomponent.
27
Thissectiondescribesthemethodologyforcalculatinglosssavingsforeachhour.Theresultsofthese
calculationsarethenusedinsubsequentsections.AsillustratedinFigure5,itwillbeimportanttonote
that,whilethemethodologydescribesthecalculationofanhourlylossresult,thereareactuallytwo
differentlosscalculationsthatmustbeperformed:systemlosses,representingthelossesincurredon
boththetransmissionanddistributionsystems(betweengenerationload,L,andendusedemand,D),
anddistributionlosses,representinglossesspecifictodistributionsystemalone.
Figure5.Systemlossesversusdistributionlosses.
Thetwolossesarecalculatedusingthesameequation,buttheyareeachapplicableindifferent
situations.Forexample,EnergyValuerepresentsabenefitoriginatingatthepointofcentral
generation,sothatthetotalsystemlossesshouldbeincluded.Ontheotherhand,T&DCapacityValue
representsabenefitasmeasuredatadistributionsubstation.Therefore,onlythelossessavedonthe
distributionsystemshouldbeconsidered.
Theselectionofsystemversusdistributionlossesisdiscussedseparatelyforeachsubsequent
benefitsection.
Methodology
Oneapproachanalystshaveusedtoincorporatelossesistoadjustenergyandcapacityrelatedbenefits
basedontheaveragesystemlosses.Thisapproachhasbeenshowntobedeficientbecauseitfailsto
capturethetruereductioninlossesonamarginalbasis.Inparticular,theapproachunderestimatesthe
L
D
Distributionlosses
Systemlosses
28
reductioninlossesduetoapeakingresourcelikePV.Resultsfromearlierstudiesdemonstratedthatloss
savingscalculationsmaybeoffbymorethanafactoroftwoifnotperformedcorrectly[6].
Forthisreason,thepresentmethodologywillincorporateacalculationoflosssavingsonamarginal
basis,takingintoaccountthestatusoftheutilitygridwhenthelossesoccur.CleanPowerResearchhas
previouslydevelopedmethodologiesbasedontheassumptionthatthedistributedPVresourceissmall
relativetotheload(e.g.,[6],[9]).CPRhasrecentlycompletednewresearchthatexpandsthis
methodologysothatlosssavingscannowbedeterminedforanylevelofPVpenetration.
FuelCostSavingsandO&MCostSavings
Introduction
FuelCostSavingsandO&MCostSavingsarethebenefitsthatutilityparticipantsderivefromusing
distributedPVgenerationtooffsetwholesaleenergypurchasesorreducegenerationcosts.EachkWh
generatedbyPVresultsinonelessunitofenergythattheutilityneedstopurchaseorgenerate.In
addition,distributedPVreducessystemlossessothatthecostofthewholesalegenerationthatwould
havebeenlostmustalsobeconsidered.Thecapacityvalueofgenerationistreatedinaseparate
section.
Methodology
ThesevaluescanbecalculatedbymultiplyingPVsystemoutputtimesthecostofthegenerationonthe
marginforeachhour,summingforallhoursovertheyear,andthendiscountingtheresultsforeach
yearoverthelifeofthePVsystem.
Therearetwoapproachestoobtainingthemarginalcostdata.Oneapproachistoobtainthemarginal
costsbasedonhistoricalorprojectedmarketprices.Thesecondapproachistoobtainthemarginal
costsbasedonthecostofoperatingarepresentativegeneratorthatisonthemargin.
Initially,itmaybeappealingtotaketheapproachofusingmarketprices.Thereare,however,several
difficultieswiththisapproach.Onedifficultyisthatthesetendtobehourlypricesandthusrequire
hourlyPVsystemoutputdatainordertocalculatetheeconomicvalue.Thisdifficultycanbeaddressed
byusinghistoricalpricesandhistoricalPVsystemoutputtoevaluatewhatresultswouldhavebeenin
thepastandthenescalatingtheresultsforfutureprojections.Amoreseriousdifficultyisthat,while
hourlymarketpricescouldbeprojectedforafewyearsintothefuture,theanalysisneedstobe
29
performedoveramuchlongertimeperiod(typically30years).Itisdifficulttoaccuratelyprojecthourly
marketprices30yearsintothefuture.
Amorerobustapproachistoexplicitlyspecifythemarginalgeneratorandthentocalculatethecostof
thegenerationfromthisunit.ThisisoftenaCombinedCycleGasTurbine(CCGT)poweredusingnatural
gas(e.g.,[6]).ThisapproachincludestheassumptionthatPVoutputalwaysdisplacesenergyfromthe
samemarginalunit.Giventheuncertaintiesandcomplicationsinmarketpriceprojections,thesecond
approachistaken.
FuelCostSavingsandO&MCostSavingsequalsthesumofthediscountedfuelcostsavingsandthe
discountedO&Mcostsavings.
SecurityEnhancementValue
BecausesolargenerationiscloselycorrelatedwithloadinmuchoftheUS,includingNewJerseyand
Pennsylvania[26],
theinjectionofsolarenergynearpointofusecandelivereffectivecapacity,and
thereforereducetheriskofthepoweroutagesandrollingblackoutsthatarecausedbyhighdemand
andresultingstressesonthetransmissionanddistributionsystems.
TheeffectivecapacityvalueofPVaccruestotheratepayer(seeabove)bothatthetransmissionand
distributionlevels.Itisthuspossibletoarguethatthereservemarginsrequiredbyregulatorswould
accountforthisnewcapacity,hencethatnoincreasedoutageriskreductioncapabilitywouldoccur
beyondtheprePVconditions.Thisisthereasonthisvalueitemaboveisnotincludedasoneofthe
directlyquantifiableattributesofPV.
Ontheotherhandthereisampleevidencethatduringheatwavedrivenextremeconditions,the
availabilityofPVishigherthansuggestedbytheeffectivecapacity(reflectingofallconditions)e.g.,
see[27],[28],onthesubjectofmajorwesternandeasternoutages,and[29]onthesubjectoflocalized
rollingblackouts.Inaddition,unlikeconventionalcentralizedgenerationinjectingelectricity(capacity)at
specificpointsonthegrid,PVactsasaloadmodulatorthatprovidesimmediatestressreliefthroughout
thegridwherestressexistsduetohighdemandconditions.Itisthereforepossibletoarguethat,all
conditionsremainingthesameintermsofreservemargins,aloadsidedispersedPVresourcewould
mitigateissuesleadingtohighdemanddrivenlocalizedandregionaloutages.
30
Lossesresultingfrompoweroutagesaregenerallynotautilitys(ratepayers)responsibility:societypays
theprice,vialossesofgoodsandbusiness,compoundedimpactsontheeconomyandtaxes,insurance
premiums,etc.ThetotalcostofallpoweroutagesfromallcausestotheUSeconomyhasbeen
estimatedat$100billionperyear(Gellings&Yeager,2004).Makingtheconservativeassumptionthata
smallfractionoftheseoutages,5%,areofthehighdemandstresstypethatcanbeeffectivelymitigated
bydispersedsolargenerationatacapacitypenetrationof15%,
11
itisstraightforwardtocalculate,as
shownbelow,that,nationally,thevalueofeachkWhgeneratedbysuchadispersedsolarbasewouldbe
oftheorderof$20/MWhtothetaxpayer.
TheUSgeneratingcapacityisroughlyequalto1000GW.At15%capacitypenetration,takinganational
averageof1500kWh(slightlyhighernationwidethanPAandNJ)generatedperyearperinstalledkW,
PVwouldgenerate225,000GWh/year.Byreducingtheriskofoutageby5%,thevalueofthisenergy
wouldthusbeworth$5billion,amountingto$20perPVgeneratedMWh.
Thisnationalvalueof$20perMWhwastakenforthepresentstudybecausetheunderlyingestimateof
costwasavailableonanationalbasis.Inreality,therewouldbestateleveldifferencesfromthis
estimate,butthesearenotavailable.
LongTermSocietalValue
Thisitemisanattempttoplaceapresentvalue$/MWhonthegenerallywellacceptedargumentthat
solarenergyisagoodinvestmentforourchildrenandgrandchildrenswellbeing.Considering:
1. Therapidgrowthoflargenewworldeconomiesandthefinitereservesofconventionalfuels
nowpoweringtheworldeconomies,itislikelythatfuelpriceswillcontinuetorise
exponentiallyfastforthelongtermbeyondthe30yearbusinesslifecycleconsideredhere.
2. Theknownveryslowdegradationoftheleading(silicon)PVtechnology,manyPVsystems
installedtodaywillcontinuetogeneratepoweratcostsunaffectedbytheworldfuel
marketsaftertheirguaranteedlifetimesof2530years
Oneapproachtoquantifythistypeoflongviewattributehasbeentoapplyaverylowsocietaldiscount
rate(e.g.,2%orless,see[25])tomitigatethefactthatthepresentdayimportanceoflongterm
expenses/benefitsisessentiallyignoredinbusinessasusualpractice.Thisisbecausediscountratesare
11
Muchlessthanthatwouldhavepreventedthe2003NEblackout.See[30].
31
usedtoquantifythepresentworthoffutureeventsandthat,andtherefore,longtermrisksand
attributesarelargelyirrelevanttocurrentdecisionmaking.
Herealesscontroversialapproachisproposedbyarguingthat,onaverage,PVinstallationwilldeliver,
onaverage,aminimumof10extrayearsofessentiallyfreeenergyproductionbeyondthelifecycle
consideredinthisstudy.
Thepresentvalueoftheseextra10years,allotherassumptionsonfuelcostescalation,inflation,
discountrate,PVoutputdegradation,etc.remainingthesame,amountsto~$25/MWhforallthe
cities/PJMhubsconsideredinthisstudy.
FuelPriceHedgeValue
Introduction
Solarbasedgenerationisinsensitivetothevolatilityoffuelpriceswhilefossilbasedgenerationis
directlytiedtofuelprices.Solargeneration,therefore,offersahedgeagainstfuelpricevolatility.One
waythishasbeenaccountedforistoquantifythevalueofPVshedgeagainstfluctuatingnaturalgas
prices[6].
Methodology
ThekeytocalculatingtheFuelPriceHedgeValueistoeffectivelyconvertthefossilbasedgeneration
investmentfromonethathassubstantialfuelpriceuncertaintytoonethathasnofuelprice
uncertainty.Thiscanbeaccomplishedbyenteringintoabindingcommitmenttopurchasealifetimes
worthoffueltobedeliveredasneeded.Theutilitycouldsetasidetheentirefuelcostobligationup
front,investingitinriskfeesecuritiestobedrawnfromeachyearasrequiredtomeettheobligation.
Theapproachusestwofinancialinstruments:riskfree,zerocouponbonds
12
andasetofnaturalgas
futurescontracts.
Considerhowthismightwork.SupposethattheCCGToperatorwantstolockinafixedpricecontract
forasufficientquantityofnaturalgastooperatetheplantforonemonth,oneyearinthefuture.First,
theoperatorwoulddeterminehowmuchnaturalgaswillbeneeded.IfEunitsofelectricityaretobe
generatedandtheheatrateoftheplantisH,E*HBTUsofnaturalgaswillbeneeded.Second,ifthe
correspondingfuturespriceofthisnaturalgasisP
NGFutures
(in$perBTU),thentheoperatorwillneedE*
12
Azerocouponbonddoesnotmakeanyperiodicinterestpayments.
32
H*P
NGFutures
dollarstopurchasethenaturalgasoneyearfromnow.Third,theoperatorneedstosetthe
moneyasideinariskfreeinvestment,typicallyariskfreebond(rateofreturnofr
riskfree
percent)to
guaranteethatthemoneywillbeavailablewhenitisneededoneyearfromnow.Therefore,the
operatorwouldimmediatelyenterintoafuturescontractandpurchaseE*H*P
NGFutures
/(1+r
riskfree
)
dollarsworthofriskfree,zerocouponbondsinordertoguaranteewithcertaintythatthefinancial
commitment(topurchasethefuelatthecontractpriceatthespecifiedtime)willbesatisfied.
13
ThiscalculationisrepeatedoverthelifeoftheplanttocalculatetheFuelPriceHedgevalue.
GenerationCapacityValue
Introduction
GenerationCapacityValueisthebenefitfromaddedcapacityprovidedtothegenerationsystemby
distributedPV.TwodifferentapproachescanbetakentoevaluatingtheGenerationCapacityValue
component.Oneapproachistoobtainthemarginalcostsbasedonmarketprices.Thesecondapproach
istoestimatethemarginalcostsbasedonthecostofoperatingarepresentativegeneratorthatison
themargin,typicallyaCombinedCycleGasTurbine(CCGT)poweredbynaturalgas.
Methodology
Thesecondapproachistakenhereforpurposesofsimplicity.Futureversionofthesoftwareservicemay
addamarketpriceoption.
OncethecostdataforthefullydispatchableCCGTareobtained,thematchbetweenPVsystemoutput
andutilityloadsneedstobedeterminedinordertodeterminetheeffectivevalueofthenon
dispatchablePVresource.CPRdevelopedamethodologytocalculatetheeffectivecapacityofaPV
systemtotheutilitygenerationsystem(see[10]and[11])andPerezadvancedthismethodandcalledit
theEffectiveLoadCarryingCapability(ELCC)[12].TheELCCmethodhasbeenidentifiedbytheutility
industryasoneofthepreferablemethodstoevaluatePVcapacity[13]andhasbeenappliedtoavariety
ofplaces,includingNewYorkCity[14].
TheELCCisastatisticalmeasureofeffectivecapacity.TheELCCofageneratingunitinautilitygridis
definedastheloadincrease(MW)thatthesystemcancarrywhilemaintainingthedesignatedreliability
13
[E*H*P
NGFutures
/(1+r
riskfree
)]*(1+r
riskfree
)=E*H*P
NGFutures
33
criteria(e.g.,constantlossofloadprobability).TheELCCisobtainedbyanalyzingastatistically
significanttimeseriesoftheunit'soutputandoftheutility'spowerrequirements.
GenerationCapacityValueequalsthecapitalcost($/MW)ofthedisplacedgenerationunittimesthe
effectivecapacityprovidedbythePV.
T&DCapacityValue
Introduction
ThebenefitthatcanbemostaffectedbythePVsystemslocationistheT&DCapacityValue.TheT&D
CapacityValuedependsontheexistenceoflocationspecificprojectedexpansionplancoststoensure
reliabilityoverthecomingyearsastheloadsgrow.Capacityconstrainedareaswhereloadsareexpected
toreachcriticallimitspresentmorefavorablelocationsforPVtotheextentthatPVwillrelievethe
constraints,providingmorevaluetotheutilitythanthoseareaswherecapacityisnotconstrained.
DistributedPVgenerationreducestheburdenonthedistributionsystem.Itappearsasanegativeload
duringthedaylighthoursfromtheperspectiveofthedistributionoperator.DistributedPVmaybe
consideredequivalenttodistributioncapacityfromtheperspectiveofthedistributionplanner,provided
thatPVgenerationoccursatthetimeofthelocaldistributionpeak.
DistributedPVcapacitylocatedinanareaofgrowingloadsallowsautilityplannertodefercapital
investmentsindistributionequipmentsuchassubstationsandlines.Thevalueisdeterminedbythe
avoidedcostofmoneyduetothecapitaldeferral.
Methodology
IthasbeendemonstratedthattheT&DCapacityValuecanbequantifiedinatwostepprocess.Thefirst
stepistoperformaneconomicscreeningofallareastodeterminetheexpansionplancostsandload
growthratesforeachplanningarea.Thesecondstepistoperformatechnicalloadmatchinganalysis
forthemostpromisinglocations[18].
MarketPriceReductionValue
TwocostsavingsoccurwhendistributedPVgenerationisdeployedinamarketthatisstructuredwhere
thelastunitofgenerationsetsthepriceforallgenerationandthepriceisanincreasingfunctionofload.
First,thereisthedirectsavingsthatoccurduetoareductioninload.Thisisthesameasthevalueof
34
energyprovidedatthemarketpriceofpower.Second,thereistheindirectvalueofmarketprice
reduction.Distributedgenerationreducesmarketdemandandthisresultsinlowerpricestoallthose
purchasingpowerfromthemarket.Thissectionoutlineshowtocalculatethemarketsavingsvalue.
CostSavings
AsillustratedinFigure6,thetotalmarketexpendituresatanygivenpointintimearebasedonthe
currentpriceofpower(P)andthecurrentload(L).TherateofexpenditureequalsPL.Totalmarket
expendituresafterPVisdeployedequalsthenewprice(P*)timesthenewload(L*),orP*L*.Cost
savingsequalthedifferencebetweenthetotalbeforeandafterexpenditures.
Cost So:ings = P I - P
(1)
Thefigureillustratesthatthecostsavingsoccurbecausethereisbothachangeinloadandachangein
price.
Figure6.Illustrationofpricechangesthatoccurinmarketasresultofloadchanges.
Equation(1)canbeexpandedbyadding-P
I + P
Iandthenrearrangingtheresult.
Cost So:ings = P I + (-P
I + P
I) - P
(2)
35
= (P - P
)I + P
(I -I
= __
P - P
I - I
] I + P
_ (I - I
LetI = I - I
andP = P - P
andsubstituteintoEquation(2).Theresultisthat
Cost So:ings = _P +
P
I
I -P_ I
(3)
PerunitcostsavingsisobtainedbydividingEquation(3)byI.
Pcr unit Cost So:ings = P
cct Sungs
+
P
I
I - P
(4)
Discussion
Equation(4)suggeststhattherearetwocostsavingscomponents:directsavingsandmarketprice
suppression.Thedirectsavingsequaltheexistingmarketpriceofpower.Themarketpricereduction
valueisthesavingsthattheentiremarketrealizesasaresultoftheloadreduction.Thesesavings
dependsonthechangeinload,changeinprice,andexistingload.Itisimportanttonotethatthechange
inloadandtheexistingloadcanbemeasureddirectlywhilethechangeinpricecannotbemeasured
directly.Thismeansthatthechangeinpricemustbemodeled(ratherthanmeasured).
Itisusefultoprovideaninterpretationofthemarketpricereductioncomponentandillustratethe
potentialmagnitude.ThemarketpricereductioncomponentinEquation(4)hastwoterms.Thefirst
termistheslopeofthepricecurve(i.e.,itisthederivativeasthechangeinloadgoestozero)timesthe
36
existingload.Thisisthepositivebenefitthatthewholemarketobtainsduetopricereductions.The
secondtermisthereducedpriceassociatedwiththedirectsavings.
TheleftsideofFigure7presentsthesameinformationasinFigure6,butzoomsoutontheyaxisscale
ofthechart.Thefirsttermcorrespondstotheyellowarea.Thesecondtermcorrespondstothe
overlappingareasofthechangeinpriceandchangeinloadeffects.
Themarketpricecurvecanbetranslatedtoacostsavingscurve.TherightsideofFigure7presentsthe
perunitcostsavingsbasedontheinformationfromthemarketpricecurve(i.e.,theleftsideofthe
figure).Thelowerblacklineisthepricevs.loadcurve.Theupperlineaddsthemarketpricesuppression
componenttothedirectsavingscomponent.Itassumesthatthereisthesameloadreductionforall
loadsasintheleftsideofthefigure.Thefigureillustratesthatnomarketpricesuppressionexistwhen
theloadislowbutthemarketpricesuppressionexceedthedirectcostsavingswhentheloadishigh.
Thesavingisdependentupontheshapeofthepricecurveandthesizeoftheloadreduction.
Figure7.Direct+marketpricereductionvs.load(assumingconstantloadreduction).
TotalValue
Theprevioussectionscalculatedthecostsavingsataspecificinstantintime.Thetotalcostsavingsis
calculatedbysummingthisresultoverallallperiodsintime.Theperunitcostsavingsiscalculatedby
dividingbythetotalenergy.(Notethatitisassumedthateachunitoftimerepresents1unit).Theresult
isthat:
37
Thisresultcanbeviewedgraphicallyastheprobabilitydistributionoftheloadtimestheassociatecost
savingscurveswhenthereisaconstantloadreduction.Multiplytheloaddistributionbythetotalper
unitsavingstoobtaintheweightedaverageperunitcostsavings.
Figure8.Applyloaddistributiontocalculatetotalsavingsovertime.
Application
Asdiscussedabove,alloftheparametersrequiredtoperformthiscalculationcanbemeasureddirectly
exceptforthechangeinprice.Thus,itiscrucialtodeterminehowtoestimatethechangeinprice.
Thisisimplementedinfoursteps:
1. ObtainLMPpricedataanddevelopamodelthatreflectsthisdata.
2. UsetheLMPpricemodelandEquation(4)tocalculatethepricesuppressionbenefit.Notethat
thisdependsuponthesizeofthechangeinload.
3. ObtaintimecorrelatedPVsystemoutputanddeterminethedistributionofthisoutputrelative
totheload.
4. MultiplythePVoutputdistributiontimesthepricesuppressionbenefittocalculatethe
weightedaveragebenefit.
38
HistoricalLMPandtimeandlocationcorrelatedPVoutputdataarerequiredtoperformtheanalysis.
LMPsareobtainedfromthemarketandthePVoutputdataareobtainedbysimulatingtimeand
locationspecificPVoutputusingSolarAnywhere.
Figure9illustrateshowtoperformthecalculationsusingmeasuredpricesandsimulatedPVoutputfor
PPLinJune2012.TheleftsideofthefigureillustratesthatthehistoricalLMPs(blackcircles)areusedto
developapricemodel(solidblackline).Thecenterofthefigureillustrateshowthepricemodelisused
withEquation(4)isusedtocalculatethepricesuppressionbenefitforeveryloadlevel.Sincethis
benefitdependsuponthesizeofthechangeintheload,thefigurepresentsarange.Thesolidblueline
isthebenefitforaverysmallPVoutput.Thedashedbluelinecorrespondstothebenefitfora1,000
MWPVoutput.Therightsideofthefigure(redline)presentsthedistributionofthePVenergyrelative
totheload(i.e.,theamountofPVenergyproducedateachloadlevel,sohighervaluescorrespondto
morefrequentweighting).Theweightedaveragepricesuppressionbenefitiscalculatedbymultiplythe
PVoutputdistributiontimesthepricesuppressionbenefit.Notethatinpractice,theactualcalculationis
performedforeachhouroftheanalysissincethepricesuppressionbenefitisafunctionofboththeload
andthePVoutput.
Figure9.IllustrationofhowtocalculatebenefitusingmeasureddataforJune2011.
Step1: Step2: Step3:
Developpricemodelbasedon
measuredLMPdata
Calculatepricereductionbenefit
(dependsonoutput)
DevelopPVoutputdistribution
Figure10presentstheresultsforthethreestepsforeachmonthin2011.
39
40
Figure10.MeasuredandmodeledLMPs(blackcirclesandlines),pricesuppressionbenefit
(solidblueforsmalloutputanddashedbluefor1,000MWofoutput)andPVoutput
distribution(PPL2011).
41
Results
AsillustratedinTable7thepricereductionbenefitsaremorethandoublethedirectsavingsfora100
MWofPVandslightlyexceedthedirectsavingfor1,000MWPV,foracombinedvaluerangingfrom
$127/MWhto$180/MWh.
Table7.Marketsavingsillustration.
100MW 1,000MW
DirectSavings $58 $58
MarketPrice
Reduction $122 $69
Total $180 $127
AcomparisonofdirectmarketsavingsandenergysavingsascalculatedinthisstudyisshowninTable8.
FuelcostsavingsandO&Mcostsavingsarecombinedbecausetheyrepresentthesamecoststhatare
includedinmarketprice.DirectsavingswerecalculatedforeachhourasPL,summedfortheyear,and
escalatedatthesamerateeachyearasnaturalgasfuturesbeyondthe12yearlimit.
42
Table8.Directmarketsavingscomparison(Newark,South30).
Theresultsshowthatdirectmarketsavingsare39%abovetheenergysavings.Thisdiscrepancyreflects
thefactthatthetwoquantities,whilerepresentingthesamevaluecomponents,useentirelydifferent
approaches.Fuelcostsavingsarederivedfromnaturalgasfutures,discountedattheutilitydiscount
rate,andappliedagainstanassumedCCGTheatrate.DirectmarketsavingsarebasedonhourlyPJM
zonalpricesfor2011.
Theenergysavingsachievedbytheutilityisbasedonavoidedmarketpurchases.However,historical
marketpricesarenotnecessarilyanindicatoroffutureyears,especiallyfor30yearsintothefuture.For
thisreason,theenergysavingsmethodologyusedinthisanalysisismorecloselytiedtothe
fundamentalsofthecost:fuelandO&Mcoststhatmustberecoveredbythemarketplacefor
generationtobesustainableinthelongrun.
ZonalPriceModel
Tocalculatethemarketpricereductioninequation(4),azonalpricemodelwasdevelopedasfollows.A
functionF()maybedefinedwhosevalueisproportionaltomarketclearingpriceusingtheform:
F(IooJ) = Ac
BxLoud
C
+
wherecoefficientsA,B,C,andDareevaluatedforeachutilityandforeachmonthusinghourlyPJM
zonalmarketpricedata,amountingtoatotalof84individualmodels.
Pisthezonalwholesaleclearingprice,andP*isgivenby:
P
P
=
F(IooJ - FlcctPowcr - IossSo:ings)
F(IooJ)
Value Value
($/kW) ($/MWh)
FuelCostSavings $709 38.8
O&MCostSavings $345 18.9
TotalEnergySavings $1,054 57.7
DirectMarketSavings $1,470 80.4
43
Themarketpricereduction(in$/MWh)iscalculatedusingtherelevantterminEquation(4)and
multiplyingbythechangeinload,includinglosssavings.
EnvironmentalValue
Introduction
ItiswellestablishedthattheenvironmentalimpactofPVisconsiderablysmallerthanthatoffossil
basedgenerationsincePVisabletodisplacepollutionassociatedwithdrilling/mining,andpowerplant
emissions[15].
Methodology
TherearetwogeneralapproachestoquantifyingtheEnvironmentalValueofPV:aregulatorycost
basedapproachandanenvironmental/healthcostbasedapproach.
TheregulatorycostbasedapproachvaluestheEnvironmentalValueofPVbasedonthepriceof
RenewableEnergyCredits(RECs)orSolarRenewableEnergyCredits(SRECs)thatwouldotherwisehave
tobepurchasedtosatisfystateRenewablePortfolioStandards(RPS).Thesecostsareapreliminary
legislativeattempttoquantifyexternalcosts.Theyrepresentactualbusinesscostsfacedbyutilitiesin
certainstates.
Anenvironmental/healthcostbasedapproachquantifiesthesocietalcostsresultingfromfossil
generation.EachsolarkWhdisplacesanotherwisedirtykWhandcommensuratelymitigatesseveralof
thefollowingfactors:greenhousegases,SOx/NOxemissions,miningdegradations,groundwater
contamination,toxicreleasesandwastes,etc.,thatareallpresentorpostponedcoststosociety.Several
exhaustivestudieshaveestimatedtheenvironmental/healthcostofenergygeneratedbyfossilbased
generation[16],[17].Theresultsfromenvironmental/healthcostbasedapproachoftenvarywidelyand
canbecontroversial.
Theenvironmental/healthcostbasedapproachwasusedforthisstudy.
Theenvironmentalfootprintofsolargenerationisconsiderablysmallerthanthatofthefossilfuel
technologiesgeneratingmostofourelectricity(e.g.,[19]).Utilitieshavetoaccountforthis
environmentalimpacttosomedegreetoday,butthisisstillonlylargelyapotentialcosttothem.Rate
basedSolarRenewableEnergyCredits(SRECs)marketsinNewJerseyandPennsylvaniaasameansto
meetRenewablePortfolioStandards(RPS)areapreliminaryembodimentofincludingexternalcosts,
44
buttheyarelargelydrivenmorebypoliticallynegotiatedprocessesthanbyareflectionofinherent
physicalrealities.Theintrinsicphysicalvalueofdisplacingpollutionisrealandquantifiablehowever:
dependingonthecurrentgenerationmix,eachsolarkWhdisplacesanotherwisedirtykWhand
commensuratelymitigatesseveralofthefollowingfactors:greenhousegases,SOx/NOxemissions,
miningdegradations,groundwatercontamination,toxicreleasesandwastes,etc.,whichareallpresent
orpostponedcoststosociety(i.e.,thetaxpayers).
Theenvironmentalvalue,EV,ofeachkWhproducedbyPV(i.e.,notproducedbyanotherconventional
source)isgivenby:
EI = x
EC
n
=0
WhereEC
i
istheenvironmentalcostofthedisplacedconventionalgenerationtechnologyandx
i
isthe
proportionofthistechnologyinthecurrentenergymix.
Severalexhaustivestudiesemanatingfromsuchdiversesourcesasthenuclearindustryorthemedical
community([20],[21])estimatetheenvironmental/healthcostof1MWhgeneratedbycoalat$90250,
whilea[nonshale
14
]naturalgasMWhhasanenvironmentalcostof$3060.
ConsideringNewJerseyandPennsylvaniaselectricalgenerationmixes(Table9)andassumingthat(1)
nuclearenergyisnotdisplacedbyPVattheassumedpenetrationlevel
15
and(2)thatallnaturalgasis
conventional,theenvironmentalvalueofeachMWhdisplacedbyPV,hencethetaxpayerbenefit,is
estimatedat$48to$129inPennsylvaniaand$20to$48inNewJersey.
Weretainedavaluenearthelowerrangeoftheseestimatesforthepresentanalysis.
14
ShalegasenvironmentalfootprintislikelyhigherbothintermsofenvironmentdegradationandGHGemissions.
15
Thestudythereforeascribesnoenvironmentalvaluerelatedtonucleargeneration.Scenarioscancertainlybe
designedinwhichnucleargenerationwouldbedisplaced,inwhichcasetheenvironmentalcostofnuclear
generationwouldhavetobeconsidered.Thisisacomplexandcontroversialsubjectthatreflectstheprobabilityof
catastrophicaccidentsandtheenvironmentalfootprintoftheexistinguraniumcycle.Thefactthatthe
environmentalliabilityisassumedtobezerounderthepresentstudymaythereforebeconsideredaconservative
case.
45
Table9.Environmentalinputcalculation.
EconomicDevelopmentValue
TheGermanandOntarioexperiencesaswellastheexperienceinNewJersey,wherefastPVgrowthis
occurring,showthatsolarenergysustainsmorejobsperunitofenergygeneratedthanconventional
energy([21],[22]).Jobcreationimpliesvaluetosocietyinmanyways,includingincreasedtaxrevenues,
reducedunemployment,andanincreaseingeneralconfidenceconducivetobusinessdevelopment.
Inthisreport,onlytaxrevenueenhancementfromthejobscreatedasameasureofPVinduced
economicdevelopmentvalueisconsidered.Thismetricprovidesatangiblelowestimateofsolar
energyslikelylargermultifacetedeconomicdevelopmentvalue.InPennsylvaniaandNewJersey,this
lowestimateamountstorespectively$39and$40perMWh,evenundertheveryconservative,butthus
farrealistic,assumptionthat80%ofthePVmanufacturingjobswouldbeeitheroutofstateorforeign
(seemethodologysection,below).
Methodology
Inaprevious(NewYork)study[24],netPVrelatedjobcreationnumberswereuseddirectlybasedupon
OntarioandGermanyshistoricalnumbers.Howeverthisassumptiondoesnotreflectstherapidchanges
ofthePVindustrytowardslowerprices.Inthisstudyafirstprincipleapproachisappliedbasedupon
GenerationMix
48% Coal 43.2 to 120.0
15% NaturalGas 4.5 to 9.0
34% Nuclear 0.0 to 0.0
3% Other 0.0 to 0.0
EnvironmentalValueforPA 47.7 to 129.0
10% Coal 9.0 to 25.0
38% NaturalGas 11.4 to 22.8
50% Nuclear 0.0 to 0.0
2% Other 0.0 to 0.0
EnvironmentalValueforNJ 20.4 to 47.8
Pennsylvania
NewJersey
ProratedEnvironmentalCost
($/MWh)
46
thedifferencebetweentheinstalledcostofPVandconventionalgeneration:inessencethisapproach
quantifiesthefactthatpartofthepricepremiumpaidforPVvs.conventionalgenerationreturnstothe
localeconomyintheformofjobshencetax.
Therefore,assumingthat:
TurnkeyPVcosts$3,000perkWvs.$1,000perkWforcombinecyclegasturbines(CCGT)
TurnkeyPVcostiscomposedof1/3technology(modules&inverter/controls)and2/3structure
andinstallationandsoftcosts.
20%oftheturnkeyPVtechnologycostand90%oftheothercostsaretraceabletolocaljobs,
while50%oftheCCGTareassumedtobelocaljobs,thus:
o ThelocaljobstraceableamountspentonPVisequalto: [
0.2
3
+
0.92
3
Suuu =
$1,99ukw
o AndthelocaljobstraceableamountspentonCCGTisequalto:u.S 1uuu = $Suukw
PVsystemsinNJandPAhaveacapacityfactorof~16%,producing~1,400kWhperyearper
kW
AC
andCCGThaveanassumedcapacityfactorof50%,producing4,380kWhperyear,
therefore
o ThelocaljobstraceableamountspentperPVkWhinyearoneis:1,900/1,400=$1.42
o ThelocaljobstraceableamountspentperCCGTkWhinyearoneis:500/4,380=$0.114
ThenetlocaljobstraceablebetweenPVandCCGTisthusequalto1.420.11=$1.30
AssumingthatthelifespanofbothPVandCCGTis30years,andusingalevelizingfactorof8%,
thenetlocaljobstraceableamountpergeneratedPVkWhoveritslifetimeamountsto:
1.Su
0.081.08
30
1.08
29
= $0.116/kWh
AssumingthatlocallytraceableO&McostsperkWhforPVareequaltothelocallytraceable
O&McostsforCCGT,
16
butalsoassumingthatbecausePVrelatedT&Dbenefitsdisplacea
commensurateamountofutilityjobsassumedtobeequaltothisbenefit(~0.5centsperkWh),
thenetlifetimelocallytraceablePVCCGTdifferenceisequalto0.1160.005=$0.111/kWh
FinallyassumingthateachPVjobisworth$75K/yearafterstandarddeductionshencehasa
combinedStateandFederalincometaxrateof22.29%inPAand22.67%inNJ
17
andthateach
16
Thisincludesonlyafractionofthefuelcoststheotherfractionbeingimportedfromoutofstate.
17
Fortheconsideredsolarjobincomelevel,theeffectivestaterate=3.07%inPAand3.54%inNJandthe
effectivefederalrate=19.83%.TheincreasedfederaltaxcollectioniscountedasanincreaseforNewJerseys
47
newjobhasanindirectjobmultiplierof1.6,
18
itcanbearguedthateachPVMWhrepresentsa
netnewjobrelatedtaxcollectionincreaseforNJequaltoalevelizedvalueof$111NWh
u.2267 1.6 = $4uNWh,andataxcollectionincreaseforPAequalto$111NWh u.2229
1.6 = $S9NWh.
SolarPenetrationCost
Itisimportanttorecognizethatthereisalsoacostassociatedwiththedeploymentofsolargeneration
onthepowergridwhichaccruestotheutilityandtoitsratepayers.Thiscostrepresentsthe
infrastructuralandoperationalexpensethatwillbenecessarytomanagetheflowofnoncontrollable
solarenergygenerationwhilecontinuingtoreliablymeetdemand.ArecentstudybyPerezetal.[31]
showedthatinmuchoftheUS,thiscostisnegligibleatlowpenetrationandremainsmanageablefora
solarcapacitypenetrationof30%.Forutilitiesrepresentativeofthedemandpatternandsolarload
synergiesfoundinPennsylvania,thispenetrationcosthasbeenfoundtorangefrom0to5centsper
kWhwhenPVpenetrationrangesfrom0%to30%incapacity.Uptothislevelofpenetration,the
infrastructuralandoperationalexpensewouldconsistoflocalizedloadmanagement,[usersited]
storageand/orbackup.
19
Atthe15%levelofpenetrationconsideredinthisstudy,thecostof
penetrationcanbeestimatedfromthePerezetal.study
18
at$1020/MWh.
taxpayer,becauseitcanbereasonablyarguedthatfederaltaxesare(1)redistributedfairlytothestatesand(2)
thatfederalexpensebenefitallstatesequally.
18
indirectbasemultipliersareusedtoestimatethelocaljobsnotrelatedtotheconsideredjobsource(heresolar
energy)butcreatedindirectlybythenewrevenuesemanatingfromthenew[solar]jobs
19
Atthehigherpenetrationlevelsthetwoapproachestoconsiderwouldberegional(orcontinental)
interconnectionupgradeandsmartcouplingwithnaturalgasgenerationandwindpowergenerationthecostof
theseapproacheshasnotbeenquantifiedaspartofthisstudy.
48
MethodologyReferences
[1].Hoff,T.E.,Wenger,H.J.,andFarmer,B.K.DistributedGeneration:AnAlternativetoElectricUtility
InvestmentsinSystemCapacity,EnergyPolicy24(2):137147,1996.
[2].Hoff,T.E.FinalResultsReportwithaDeterminationofStackedBenefitsofBothUtilityOwnedand
CustomerOwnedPVSystems:Deliverable1.3.5.2,SMUDReport,2002.
[3].Hoff,T.E.,andMargolis,R.DistributedPhotovoltaicsinNewJersey,NRELReport,2003.
www.cleanpower.com/research/distributedgeneration/DistributedPVInNewJersey.pdf.
[4].Hoff,T.E.,Norris,B.L.,andWayne,G.PotentialEconomicBenefitsofDistributedPhotovoltaicsto
theNevadaPowerCompany.Report,2003.
www.cleanpower.com/Content/Documents/research/distributedgeneration/NevadaPower2003.pdf
[5].Hoff,T.E.andMargolis,R.M.MovingTowardsaMoreComprehensiveFrameworktoEvaluate
DistributedPhotovoltaics.Report,2005.
www.cleanpower.com/Content/Documents/research/customerPV/EvaluationFramework.pdf
[6].Hoff,T.E.,Perez,R.,Braun,G.,Kuhn,M.,andNorris,B.TheValueofDistributedPhotovoltaicsto
AustinEnergyandtheCityofAustin,2006.
www.cleanpower.com/Content/Documents/research/distributedgeneration/AE_PV_ValueReport.pdf.
[7].Norris,B.L.,Hoff,T.E.,Perez,R.PVValueAnalysisforWeEnergies,2009.
[8].Perez,R.Zweibel,K.,andHoff,T.E.SolarPowerGenerationintheUS:Tooexpensive,ora
bargain?SubmittedtoEnergyPolicy.2011.
[9].Hoff,T.andShugar,D.S.,TheValueofGridSupportPhotovoltaicsInReducingDistributionSystem
Losses,IEEETransactionsonEnergyConversion,10(9):569576.,1995.
[10].Garver,L.L.EffectiveLoadCarryingCapabilityofGeneratingUnits.IEEETransactions,Power
ApparatusandSystems.Vol.Pas85,no.8,1966.
[11].Hoff,T.CalculatingPhotovoltaics'Value:AUtilityPerspective,IEEETransactionsonEnergy
Conversion3:491495,1988.
49
[12].Perez,R.,Seals,R.,andStewart,R.AssessingtheLoadMatchingCapabilityofPhotovoltaicsforUS
UtilitiesBasedUponSatelliteDerivedInsolationData,IEEETransactions,pp.11461149(23d.PV
Specialists,Louisville,KY),1993.
[13].Hoff,T.,Perez,R.Ross,JP,andTaylor,M.PhotovoltaicCapacityValuationMethods,SolarElectric
PowerAssociationReport#0208,2008.
[14].Hoff,T.E.andPerez,R.ShiningontheBigApple:SatisfyingNewYorkCitysPeakElectricalNeeds
withPV.Report,2008.
www.cleanpower.com/Content/Documents/research/distributedgeneration/PVForNewYorkCity.pdf.
[15].Fthenakis,V.,Kim,H.C.,Alsema,E.,EmissionsfromPhotovoltaicLifeCycles.Environmental
ScienceandTechnology,42(6):p.21682174,2008.
[16].DevezeauxJ.G.EnvironmentalImpactsofElectricityGeneration.25
th
UraniumInstituteAnnual
Symposium.London,UK,2000.
[17].Epstein,P.Fullcostaccountingforthelifecycleofcoal.AnnalsoftheNewYorkAcademyof
Sciences,2011.
[18].Hoff,T.E.IdentifyingDistributedGenerationandDemandSideManagementInvestment
Opportunities,TheEnergyJournal:17(4),1996.
[19].Fthenakis,V.,Kim,H.C.,Alsema,E.,2008.EmissionsfromPhotovoltaicLifeCycles.Environmental
ScienceandTechnology,42(6):p.21682174
[20].DevezeauxJ.G.,2000.EnvironmentalImpactsofElectricityGeneration.25thUraniumInstitute
AnnualSymposium.London,UK(September,2000).
[21].Epstein,P.2011.Fullcostaccountingforthelifecycleofcoal.AnnalsoftheNewYorkAcademyof
Sciences.February,2011.
[22].Louw,B.,J.E.WorrenandT.Wohlgemut,2010.EconomicImpactsofSolarEnergyinOntario.
CLearSkyAdvisorsReport(www.clearskyadvisors.com).
[23].BanWeissG.etal.,2010SolarEnergyJobCreationinCalifornia,UniversityofCaliforniaat
Berkeley.
50
[24].Perez,R.,K.ZweibelandT.E.Hoff,(2011):SolarPowerGenerationintheUS:TooExpensive,ora
Bargain?JournalofEnergyPolicy,39(2011),72907297.
[25].TolR.S.J.,J.Guo,C.J.Hepburn,andD.Anthoff2006.DiscountingandtheSocialCostofCarbon:a
CloserLookatUncertainty,EnvironmentalScience&Policy,9,205216,207.
[26].Perez,R.,R.Margolis,M.Kmiecik,M.SchwabandM.Perez,(2006):Update:EffectiveLoad
CarryingCapabilityofPhotovoltaicsintheUnitedStates.Proc.ASESAnnualConference,Denver,CO
[27].Perez,R.,R.Seals,H.Wenger,T.HoffandC.Herig,1997.PVasaLongTermSolutiontoPower
Outages.CaseStudy:TheGreat1996WSCCPowerOutage.Proc.ASESAnnualConference,Washington,
DC.
[28].PerezR.,B.Collins,R.Margolis,T.Hoff,C.HerigJ.WilliamsandS.Letendre,2005.Solutiontothe
SummerBlackoutsHowdispersedsolarpowergeneratingsystemscanhelppreventthenextmajor
outage.SolarToday19,4,July/August2005Issue,pp.3235.
[29].LetendreS.andR.Perez,2006.UnderstandingtheBenefitsofDispersedGridConnected
Photovoltaics:FromAvoidingtheNextMajorOutagetoTamingWholesalePowerMarkets.The
ElectricityJournal,19,6,6472.
[30].PerezR.,B.Collins,R.Margolis,T.Hoff,C.HerigJ.WilliamsandS.Letendre,(2005)Solutiontothe
SummerBlackoutsHowdispersedsolarpowergeneratingsystemscanhelppreventthenextmajor
outage.SolarToday19,4,July/August2005Issue,pp.3235.
[31].Perez,R.,T.HoffandM.Perez,(2010):QuantifyingtheCostofHighPVPenetration.Proc.ofASESNational
Conference,Phoenix,AZ.
[32].Felder,F.A.,ExaminingElectricityPriceSuppressionDuetoRenewableResourcesandOtherGrid
Investments,TheElectricityJournal,Vol.24,Issue4,May2011.
51
Appendix3:DetailedResults
52
Pittsburgh
TableA41.Technicalresults,Pittsburgh.
TableA42.Value($/kW),Pittsburgh.
TableA43.LevelizedValue($/MWh),Pittsburgh.
FigureA41.Value($/kW),Pittsburgh.
FigureA42.LevelizedValue($/MWh),Pittsburgh.
54
Harrisburg
20
TableA44.Technicalresults,Harrisburg.
TableA45.Valueresults($/kW),Harrisburg.
20
ScrantonandHarrisburgconstitutetwoexamplesofa15%penetrationwithinPPLterritory.Strictlyspeakingthis
doesnotamounttoa30%penetration,buttwoexamplesof15%gridpenetrationwhereresourcewouldbe
deployedineitherlocation,illustratinghowresultsareinfluencedbythelocationchoice,everythingelse(utility
andeconomicassumptions)beingequal.
South30 Horiz West30 1Axis
FleetCapacity(MWac) 1129 1129 1129 1129
AnnualEnergyProduction(MWh) 1,809,443 1,565,940 1,461,448 2,274,554
CapacityFactor(%) 18% 16% 15% 23%
GenerationCapacity(%ofFleetCapacity) 28% 27% 26% 32%
T&DCapacity(%ofFleetCapaccity) 14% 14% 14% 14%
South30 Horiz West30 1Axis
Energy
FuelCostSavings $751 $652 $608 $942
O&MCostSavings $366 $318 $296 $459
TotalEnergyValue $1,117 $969 $904 $1,401
Strategic
SecurityEnhancementValue $424 $368 $343 $532
LongTermSocietalValue $530 $460 $429 $665
TotalStrategicValue $954 $827 $772 $1,196
Other
FuelPriceHedgeValue $786 $682 $636 $985
GenerationCapacityValue $297 $287 $274 $336
T&DCapacityValue $24 $24 $24 $24
MarketPriceReductionValue $1,241 $1,224 $1,171 $1,335
EnvironmentalValue $1,011 $877 $819 $1,268
EconomicDevelopmentValue $827 $717 $669 $1,037
(SolarPenetrationCost) ($424) ($368) ($343) ($532)
TotalOtherValue $3,761 $3,444 $3,249 $4,454
TotalValue $5,832 $5,240 $4,925 $7,051
55
TableA46.LevelizedValueresults($/MWh),Harrisburg.
FigureA43.Value($/kW),Harrisburg.
FigureA44.LevelizedValue($/MWh),Harrisburg.
57
Scranton
TableA47.Technicalresults,Scranton.
TableA48.Value($/kW),Scranton.
TableA49.LevelizedValue($/MWh),Scranton.
FigureA45.Value($/kW),Scranton.
FigureA46.LevelizedValue($/MWh),Scranton.
59
Philadelphia
TableA410.Technicalresults,Philadelphia.
TableA411.Valueresults($/kW),Philadelphia.
TableA412.LevelizedValueresults($/MWh),Philadelphia.
FigureA47.Value($/kW),Philadelphia.
FigureA48.LevelizedValue($/MWh),Philadelphia.
61
Jamesburg
TableA413.Technicalresults,Jamesburg.
TableA414.Valueresults($/kW),Jamesburg.
TableA415.LevelizedValueresults($/MWh),Jamesburg.
FigureA49.Value($/kW),Jamesburg.
FigureA410.LevelizedValue($/MWh),Jamesburg.
63
Newark
TableA416.Technicalresults,Newark.
TableA417.Valueresults($/kW),Newark.
TableA418.LevelizedValueresults($/MWh),Newark.
FigureA411.Value($/kW),Newark.
FigureA412.LevelizedValue($/MWh),Newark.
65
AtlanticCity
TableA419.Technicalresults,AtlanticCity.
TableA420.Valueresults($/kW),AtlanticCity.
TableA421.LevelizedValueresults($/MWh),AtlanticCity.
FigureA413.Value($/kW),AtlanticCity.
FigureA414.LevelizedValue($/MWh),AtlanticCity.