GlobalSolarReport:

2025Edition

TheStateofPVPerformance

?2025RaptorMaps.Allrightsreserved.

1

?2025RaptorMaps.Allrightsreserved.

2

Contents

Introduction

ExecutiveSummary

Chapter1:ScalingPower,SharpeningPerformance

Chapter2:AnatomyofPowerLoss

Chapter3:BenchmarkingDCHealth

Conclusion

AboutRaptorMaps

Introduction

RaptorMapsstudiestheperformanceandriskprofilesofsolarassetsaroundtheworldinordertodistillindustrytrendsandtohighlightopportunitiesfortherapidlygrowingsolarindustry.

ThisreportisasummaryofRaptorMaps’research.

The2025GlobalSolarReportdrawsupon193GWdcofutility-

scaleandcommercialandindustrial(C&I)solarPVanalysis,with67GWdcofanalysisin2024alone.Inthisreport,youwillfind:

·Assetperformancemetricsoverthelast5yearsandfindingsonmacro-trendsdrivingunderperformance

·Commonproblemsfoundandtheirimpactonpowerproduction

·Trendsinsiteworkautomationandresourcedeployment·Awiderangeofperformancebenchmarksby:

Typeofpaneltechnologyutilized

Capacityofinstallation

U.S.electricpowermarketsandU.S.states

NoteonUnderperformanceData:

Weassignunderperformance(powerloss)byeachgranularissueobservedtocalculatepowerloss.Thisreportcorrectsforsitesinspectedmultipletimesinasinglecalendaryear(anincreasingtrend)andsignificantoutliers.

3

ExecutiveSummary

In2024,thesolarindustrydemonstratedremarkablegrowthandresilience,surpassing2

terawatts(TW)inglobalcapacityjusttwoyearsafterbroaching1terawatt[1].The

photovoltaic(PV)markethascementeditspositionasarobustandcompetitive

alternativetotraditionalenergygeneration,withtheInternationalEnergyAgency(IEA)

projectingsolartoaccountfor80%ofrenewablesgrowthby2030withstronginstallationfromsomeoftheworld’slargesteconomies:[2][3]

·Chinaisexpectedtoaccountfor60%ofglobalcapacityexpansion.

·TheEuropeanUnionisforecastedto2xtheirinstalledcapacityby2030.

·Indiaisexpectedtoemergeasthefourth-largestmarket,growing3.5xincapacity.

·TheUnitedStatesinstalled40.5GWin2024,bringingthetotalto219GW.Thesectorhassolidifieditsroleasaneconomicdriver,with$60billioninprivateinvestmentand

280,000jobsnationwide.

However,expansivegrowthinthesolarindustryisnotwithoutitschallenges,especiallyasamaturingfleetrevealsnewandattimesunexpectedriskcategories:

·Equipment-drivenunderperformancehastripledoverthelast5years,highlightingacontinuedneedforefficientlydeployingresourcesforreportingissuesandcorrective

maintenance.

·Increasingriskprofiles,emphasizedbysolar’sexposuretoclimatechange-driven

risks,suchasstorms,floodsandfires.Industrymodelscontinuetounderestimatethe

effectsofweather-relatedeventsby300+%[4].Assetownershaveturnedtocreative

methodstomanagerisk,suchaspreventativeinvestmentindesignandmaintenance

toreducecosts.

·Ongoinglaborconstraints,especiallyasoperatorscontendwithbothincreasingfleetsizesandaverageinstalledcapacity.IntheU.S.specifically,“nearly70%ofthetruecostofsolarO&Mcomesfromlabor-basedinflationarycosts”[5].

[1][2]“SolarPVpowercapacityintheNetZeroScenario,2015-2030.”IEA,

/data-and-statistics/charts/solar-pv-power-capacity-in-the-net-zero-

scenario-2015-2030

.

[3]“SolarMarketInsightReportQ42024.”SEIA/WoodMackenzie,10Jan.2025,

/research-resources/solar-market-insight-report-q4-2024

.

[4]“SolarRiskAssessment.”kwHAnalytics,2024,

/solar-risk-assessment

.

i[

EQ\*jc3\*hps16\o\al(\s\up1(5),n)

i“

EQ\*jc3\*hps16\o\al(\s\up1(T),g)

EQ\*jc3\*hps16\o\al(\s\up1(h),h)

t

EQ\*jc3\*hps16\o\al(\s\up1(e),s)

EQ\*jc3\*hps16\o\al(\s\up1(True),/)

CostofSolarO&MIsIncreasing.”O(jiān)rigis/WoodMackenzie,2022.

/insights/the-true-cost-of-solar-om-is-increasing-

RaptorMapsGlobalSolarReport:2025Edition

RaptorMapsGlobalSolarReport:2025Edition5

Politicalheadwinds,suchasPresidentTrump’stariffsandtheU.S.leavingtheParis

ClimateAccords,couldimpactthegrowthofU.S.solarasinfrastructureinvestorsmightgravitatetowardslenientpoliciesfordomesticnaturalgas.Themomentumofsolar

growthremainsuncertainintheU.S.,evenasDecemberprojectionsstillshowed43GWdcaddedannuallyoverthenextfiveyears[6].

Thesethreeindustry-widetrends-rapidgrowth,changingenergy

policylandscape,andincreasingfinancialpressure-underscoretheimportanceofensuringthatalreadyinstalledassetsperformas

modeledtocontinuedeliveringreliable,low-costcleanenergy.

Tothatend,RaptorMapshasuncovered4maintrendsinthisyear’sReport:

·Whilethegrowthinaverageunderperformanceyear-over-yearhassloweddown,thereisstillsignificantopportunityforownerstoclosethegapbetweenmodeledand

actualperformance.In2024,theaveragesolarfarminourdatasetlostupto

$5,720/MWdcperyearfromunresolvedequipmentissues.Extrapolatedtothewholeindustry,thatindicatesanannualizedpotentialrevenuelossof$10.0billion[7]fromsolarunderperformance.

·OwnersandoperatorsarebothincreasinglyadoptingAIandautomationtoreducetruckrolls,deployingrobotics(suchasremotelyoperateddrones)toinvestigate

specializedusecases.Amongtheseinspections,erosionandcrackinginspectionshavebecomemostpopular,comprising23%and18%ofnon-aerialthermographyinspections,respectively.

·Highpriorityissuesaccountedfor42%oftotalidentifiedissues,butwereresponsiblefor90%ofobservedrevenueloss.

·Overthepast5years,underperformancehasincreasedacrosssitesofallsizes,butin2024,weobservedthewidestrangeinperformanceacrossC&Iassets.

[6]“SolarMarketInsightReportQ42024.”SEIA/WoodMackenzie,10Jan.2025,

/research-resources/solar-market-insight-report-q4-2024

.

[7]Excludesresidentialcapacityandusesglobaltotalaccordingto

IEA

.

RaptorMapsGlobalSolarReport:2025Edition6

Aboutthedata

TheRaptorMapsplatformidentifiesandcategorizesissuesonsolarassetsthatcan

causesafety,compliance,orperformanceissues.Ournormalizedandaggregated

datasetof193GWcontainsdataonissuesrangingfromdamagederosionBMPsto

modulecrackingtothermographicanomaliesthatindicatepowerloss.Thisallowsustoanalyzetheperformanceandconditionsacrosssolarassetsandidentifyindustry-widetrends.

MostofthisreportwillfocusonDChealth,modulecracking,andequipment-drivenunderperformance–apotential$10Bproblemfortheglobalsolarindustry.

In2024alone,theRaptorMapsplatformanalyzeddatafrom67GWofPVsystems,of

which52GWwereforDChealthanalysis.Notably,lostcapacityfromequipment

anomalieshastripled(3.1x)overthelastfiveyears,underscoringthegrowingchallengeforsolarownersandoperators.ThefollowingsectionhighlightskeytrendsinDChealthperformance,usingdatatovalidatepatternsthatmanyintheindustryhavealready

sensed.

Powerlossiscalculatedbyapplyingalossfactortoeachimpactedmodulebasedontheseverityofthedetectedissue.

Thisyear’sreportappliesarefinedmethodologytoaccountforassetswithmultiple

inspectionswithinthesameyear,whichisarisingtrendthatRaptorMapshasobserved.

RaptorMapsDataset

CumulativeGWovertime

200.0

150.0

100.0

50.0

0.0

21.3GW-Non-AerialThermographyDataset

172.0GW-AerialThermographyDataset

201920202021202220232024

Chapter1

SCALINGCAPACITY,

SHARPENINGPERFORMANCE

In2024,theaveragesolarassetexperiencedapotentialrevenuelossof$5,720perMWdc,peryearfromequipment-drivenunderperformance.

Thatrepresentsincreasesof15%from2023and214%overthepast5years.

7

RaptorMapsGlobalSolarReport:2025Edition8

UnderperformanceContinuestoRise

AverageSystemPowerLoss

%ofTotalCapacity

5.77%

6.00%

5.00%

4.00%

3.00%

2.00%

1.84%

1.00%

0.00%

20202021202220232024

In2024,powerlossfromDChealthandotherequipmentissues(“powerloss”)continueditssteadyclimb,withsolarsitesexperiencing5.77%underperformance,upfrom5.00%in2023[1].Theconsistentincreaseinpowerlossisexpectedasownersandoperatorsnavigatethecomplexitiesofmanagingincreasinglylargerutility-scalesolarsitesandportfolios.

[1]PowerlosscalculationsassumeDCpowercapacityforannualfigures.

$6000

$5000

$4000

$3000

$2000

$1000

$0

AverageSystemRevenueLossPotential,

Annualized$perMWdc

$5,720

$977

20202021202220232024

Withtheindustryadjustingtorisingdemand,itiscriticalforownerstotakeproactivestepsinrecoveringlostrevenue—theaverageassetexperiencedannualizedlossesof$5,720/MWdcin2024,upfrom$4,617MWdcin2023.Forcontext,a100MWsiteexperiencing5.77%

underperformancefromyeartwoofoperationsuntildecommissioningwouldseeitsprojectIRRdropby249basispoints(Modelsource:

NREL

,RaptorMapsanalysis)

RaptorMapsGlobalSolarReport:2025Edition9

IndustryTrendsUnderpinningRisingUnderperformance

RaptorMaps’datasetprovidesanindicatorintotheoverallDChealthofalreadyinstalledsolarcapacity,providingfurthercontextintothefollowingtrendsinlaborand

manufacturing.

First,intheUnitedStates,thegrowthofoperationallaborisfailingtokeeppacewiththerapidexpansionofsolarcapacity.AccordingtotheUSIRECJobsCensusand

SEIA/WoodMackenzie,solarcapacityhasgrownby182%overthepastfiveyears

(2019-2023),whileoperationallaborhasincreasedbyonly91%duringthesameperiod[1][2].With27%ofoperations&maintenancehiringduetochurn,thecostsofhiring,onboarding,andretentionexertadditionalpressuresonbudgetsandtime[3].

Second,majorstridesinmanufacturinghavebeenmadeoverthepastdecade,as

trackedbythePVELScorecard[4].Infact,morethan20newmanufacturerswereratedTopPerformerswithinthescorecard.However,PVELhasalsoobservedacontinued

riseinBOM-levelfailureswith41%ofBOMsexperiencingatleastonefailure-upfrom32%in2023.Asthesolarindustrycontinuestomatureandscale,efficientlymonitoringandmanagingpotentialequipmentfailureswillbecritical.

Finally,boththemagnitudeandfrequencyofnaturalcatastropheeventscontinueto

grow.Inthe2010s,therewere13.1NOAAbillion-dollarweathereventsperyear.Theannualcountmorethandoubledin2024with27billion-dollarweatherevents[5].Fromhailstormsthatcausecrackingtolightningstormsthatcanstrikeequipment,ownersandoperatorscontinuetoprioritizetimelyextremeweatherresponse,increasingly

relyingontechnologyandinnovativeoperatingprocedurestomitigaterisk.

[1][3]“NationalSolarJobsCensus”IREC,2024,

/census-solar-job-trends/

.

[2]“SolarMarketInsightReportQ42024.”SEIA/WoodMackenzie,10Jan.2025,

/research-resources/solar-market-

insight-report-q4-2024

.

[4]“The2024PVModuleReliabilityScorecard”KiwaPVEL2024,

/

.

[5]“U.S.Billion-DollarWeatherandClimateDisasters”NOAANationalCentersforEnvironmentalInformation(NCEI),2025.

/access/billions/

.

CumulativeU.S.InstalledSolarCapacity,GWdc

+182%

200

150

100

50

0

2018

2019

2020

2021

2022

2023

2018

2019

2020

2021

2022

2023

+91%

000

000

000

000

000

0

25

20

15

10

5

,

,

,

,

,

U.S.EmploymentinSolarOperations&Maintenance,allfunctions(source:IREC)

RaptorMapsGlobalSolarReport:2025Edition10

Spotlight:OnlyaFractionofSolarJobsforMaintenance

U.S.SolarJobsbySector

300

250

200

150

100

5

,

,

,

,

,

,

000

000

000

000

000

000

0

Operations&Maintenance

AllOthers

WholesaleTrade&Distribution

Manufacturing

Installation&

ProjectDevelopment

20192020202120222023

[1]“NationalSolarJobsCensus”IREC,2024,

/census-solar-job-trends/

.

In2023,26%ofOperations&Maintenancejobswereinthe“Installation&Repair”category

Othercategories:

Management/Professional(29%)Administrative(20%)

Sales(12%)

ProductionandManufacturing(8%)AllOther(6%)

6%

12%

8%

29%

20%

26%

27%ofNewHiresinOperations&Maintenance

werebackfillsduetoChurn

27%

Othercategories:

43%

ExistingEmployeeAddsSolarResponsibilities(43%)

NewPositions(29%)

29%

RaptorMapsGlobalSolarReport:2025Edition11

In2024,Performance

IntelligenceIsBridgingPowerLossGaps

Lastyear,RaptorMapsidentifiedmorethan10.8millionDChealthissuesacross44uniquegigawattsofsolarassetsanalyzed,or249issuesperMWdc.Asmoresolarisinstalledandasubstantialportionofinstalledassetsage,assetmanagersandO&Mteamscontinuetofacethedifficultactofbalancingbetweencorrectiveand

preventativemaintenanceofDChealthissuesandbeyond.AccordingtoRaptorMaps,dataset,32%ofidentifiedissueswerecategorizedas“HighPriority”,accountingfor90%oftheobservedpowerloss.56%and12%ofidentifiedissueswere“LowPriority”and“MediumPriority”,respectively.Aerialinspectionsallowforrapiddiagnosingandgeolocatingofthose“HighPriority”issues,especiallythosecausingSCADAalerts,

providingthenecessaryintelligenceforeffectiveandefficientremediation.

Thedataemphasizesthecriticalityofmaintainingdetailedrecordsofsolarasset

performanceandsiteinfrastructureconditions.Whileaddressingimmediaterepairs

naturallydemandsattentionwhenresourcesarelimited,theanalysisrevealsthatnearly70%ofidentifiedissuesfallintomediumorlowprioritycategories,makingitvery

difficulttotrack,muchlessremediate,non-highpriorityissuesthatareatriskof

escalatingtolargerproblems-andwhichissuesarenot.Thisrealityhasledmany

organizationstoinvestincentralizedsystemsofrecordandmoreadvancedmethodsofcapturingdatatomaintaincleardocumentationofthesedevelopingissues,includingmorefrequentandtargetedinspections.

PrioritizingDCHealthIssues:Frequencyvs.PowerLossImpact

100%

80%

60%

40%

20%

0%

IssueFrequencyPowerLoss

Priority

Low

Medium

High

RaptorMapsGlobalSolarReport:2025Edition12

AShiftTowardsTargeted,RepeatedInspections

Thekeytoemployingperformanceintelligenceisprioritizingthehighest-impactissuesandleveragingdatatomakemoreinformedresourcingdecisions.Inlarge-scalesites

between50-200MW,low-prioritydefectscontributejust0.2kWofpowerlossperdefect,whereashigh-prioritydefectscause6-8kWofpowerlossperdefect.However,itis

importanttonoteforC&Iandsmallerutility-scalefarms,module-leveldefaultscancause15-20%ofpowerlossandarenotinsignificantwhenaddressingsiteperformance.GivencurrentlaborconstraintsandtighterO&Mbudgets,fixingeverydefectisneitherpracticalnorcost-effective.Savvyoperationsteamshavebeguntoshifttowardsdata-informed,targetedinterventionstorecoverpowerfasterwhileminimizingwindshieldtime.

Furthermore,industrydatareflectsashiftawayfromtreatingaerialinspectionsasisolatedassessmentsandratherasascalableapproachtowardspreventativemonitoring.Notably,someownersandoperatorshaveinternalizedthistrend,inspectingthesamesites70%

morein2024thanin2023.

Withagrowingvolumeofdefectstoremediateormonitor—estimatedatnearly249

defectsperMWperyearin2024—operationsteamsarefacinganoverwhelmingbacklogofissuestomanage,andincreasinginspectionfrequencyisnotviable.Thesetrends

highlighttheurgentneedfortechnologysolutionstoefficientlyidentify,prioritize,and

monitorissuesonsolarfarmswithoutmobilizingpersonnel.Asaresult,manyoperatorsareturningtonewtoolstosupportbalancingbetweenmonitoring,preventative,and

correctivemaintenance,ensuringthemostcriticalissuesareaddressedattherighttime.Software-enabledsensorsandon-siteroboticsprovidefrequent,autonomous,and

granulardiagnosesintositehealth.Thisallowsteamstobedeployedefficiently,ina

timelymanner,andonlywhenhumaninterventionisneeded—reducingunnecessarytruckrolls,automatingreports,andensuringtechniciansarrivepreparedwithclearguidanceandnecessaryequipment,ensuringtechniciansafetyandreducingtimespenton-sitepertask.

PowerLoss,bySeverity,bySiteSize

<5MW

5-10MW 10-20MW 20-50MW 50-100MW 100-200MW 200+MWDatasettotal

0%20%40%60%80%100%

IssueFrequency,bySeverity,bySiteSize

<5MW

5-10MW 10-20MW 20-50MW 50-100MW 100-200MW 200+MWDatasettotal

0%20%40%60%80%100%

Legend

Low

Medium

High

IssueFrequencyperMWdc

2024AveragebySiteSize

<5MW

5-10MW

10-20MW

20-50MW

50-100MW

100-200MW

200+MW

Datasetavg

0100200300400500

RaptorMapsGlobalSolarReport:2025Edition13

DifferentNeedsforDifferentFleets

Elevatedpowerlossfromequipmentissuescreatesdifferentchallengesandopportunitiesforeachsolarindustrysegment.Forexample,commercialandindustrial(C&I)siteslessthan5MWdcexperiencedanaverageof264identifiedDChealthissuesperMWdc,whichpresentssignificantlostrevenueriskwhenaggregatedatthefleet-level.WithC&Iand

communitysolarfleetstypicallyspreadacrossmanydifferentlocations,balancing

preventativeandcorrectiveeffortscanbecomedifficult,butwithoutthoseefforts,sitesriskhighermagnitudeeventssuchasfiresoroutages.

Inresponse,someassetownersandoperatorshaveunlockedeconomiesofscopeby

expandingthecapabilitiesoftheirteams.Forexample,severalofRaptorMaps’customershavetrainedtechnicianstoflydronesandself-performaerialdatacollection,enablinghighfrequencyperformanceinsightsontheirassetsandthereallocationoftechniciantime

previouslyspentonmanualinspectionstomoreimpactfulactivities.

Assitesgetlarger,thevolumeofdataandconditionsthatassetmanagersandO&M

teamsmustcareaboutincreasesexponentially,withmanyrequiringsomeformofmanualintervention.However,increasingportfolios,sitesizes,anddistancesfrompopulation

centershavemadeitdifficulttoconductalltheactivitiesundercontract.Asaresult,

RaptorMapshasobservedariseinautomationofmanyvisualinspectionactivitieswith

remotelyoperatedrobotics.Forexample,RaptorMapscustomersleverageremotely

operateddronesinstalledon-sitetoconductinvestigationsofalertsfromSCADAsystems,whichallowfortechnicianstobedeployedforremediationwiththerighttools,supplies,

anddata.Furthermore,usingremotelyoperatedroboticsasafirstmethodofresponse

reducestechnicianexposuretopotentiallyhazardousorhigh-voltagesituations,drasticallyincreasingsafetyandsupportingworkforceretentionefforts.

AutomatingGround

Inspections&SiteWalks

Remoteoperationsarenottheonlysolutiontopreservingtechniciantime--customersarealsowidelydeployingaerialinspectionsfornon-DChealthusecases.In2024,civilworksinspectionswerethemostpopulartypeofnon-DChealthinspectionanalyzedintheRaptorMapsplatform.Erosion/SWPPPandPerimeterInspectionsmakeupmorethanathirdofallnon-DChealthorders.Inaddition,SpecialtyCrackingInspections(e.g.forthedetectionofhairlinecracking)havealsogrownpopularasawaytomitigatetheriskoffireandgroundfaultsonutility-scalesites,makingup18%ofnon-aerial

thermographyinspections.

TypicalgroundinspectionsarebeingreplacedacrosstheentireO&Mecosystemandsteadilyincreasingincadenceaswell.Ownersandoperatorshavebeenincreasingtherateofaerialinspectionsontheirassetsovertime,withinspectionspersiterising70%from2023to2024.ThisincludesalltypesofinspectiontypesintheRaptorMaps

dataset.Historically,from2019to2023,ownersandoperatorstypicallyconductedanannualaerialinspectionwithanoccasionalmid-yearcheck-inafterabigstormor

performanceissues.

Morerecently,theevolvingneedsoftheassetownerhavedriveninspectioncadencesupdrastically.Technologyhasenabledtheon-demandcollectionandanalysisofdataat-scale,allowingformoreproactivemanagementofriskwhilereducingwhentruckrollsareneeded.Ownerscannowuseaerialinspectionsasapreventativemeasure,rather

thandeployingreactionaryinspections-especiallyonutility-scalesiteswheresiteriskscanexacerbateandbecomeverycostly.

/yrInspectionsperSite

70%Increasein

Breakdownof2024Non-Aerial

ThermographyInspections,bytype

Wiring

5%

Substation

9%

Vegetation

1%

Erosion/SWPPP

23%

P.O.I.*

12%

Cracking

18%

Perimeter

15%

Construction

17%

14

*P.O.I.-PointofInterconnection

RaptorMapsGlobalSolarReport:2025Edition15

ExampleofRiskManagementAnalyticsfromRaptorMaps

SubstationP0ThermalSignature

VegetationGrowthonCABLine

Erosion&WaterManagementIssues

Chapter2

ANATOMYOFPOWERLOSS

RaptorMaps’PowerLossdatasetmeasureslostcapacityduetoDChealthandotherequipment

issues.Whilesystem-levelfaultscontinuetocausethemostpowerloss,2024alsosawanincreaseinphysicaldamageandmodule-levelissues,whichpresentincreasedfireandpowerlossrisksthat

ownersneedtomanageacrosstheirfleet.

16

RaptorMapsGlobalSolarReport:2025Edition17

RaptorMaps,

PowerLossDataset

TheRaptorMapsdatasetcontains172GWofDChealth-drivenpowerlossanalyticsand21GWofrisk

managementanalytics(e.g.vegetationmonitoringorequipmentdefectdetection),taggedwithprecise

geolocationsofthedetectedissueandanactionablecategorizationofwhatthatissueis.Thefollowing

sectionprovidestrendsinwhatRaptorMapshas

detectedtobetheleadingdriversofpowerloss,

providingownersandoperatorswithbenchmarksfortheirownassets.

ForthepurposesoftheGlobalReport,similaranomalytypesweregrouped