Upload File

poster title poster title continued - nrel · pdf fileposter title poster title continued ......

  • Home
  • Documents
  • Poster Title Poster Title continued - NREL · PDF filePoster Title Poster Title continued ... Energy was simulated for each of the studied resource files using the PVSyst Software
1
Poster Title Poster Title continued Researchers’/Presenters’ Names Institution/Organization/Company Methods Methods Figure 3 Project Energy Yield for all three case studies Why Resource Matters: Impacts of Preconstruction Resource Data on LongͲTerm Production Estimates and Project Finance Paul Thienpont – Meteorologist, Marie Schnitzer – VP Consulting Services, Rebecca Tilbrook – Solar Services Team Lead • Determine how various sources of solar and meteorological data impact production estimates • Determine how uncertainty of data sets inhibit the ability to leverage the project • Determine how plant overproduction can lead to lost revenue • Determine how plant overproduction can result in lower rate of returns for debt, tax equity, and cash equity finance partners Accurate resource and energy production assessments are needed to establish project revenues with confidence thereby sizing debt appropriately. However, when plant performance exceeds the preͲ construction energy estimate, what are the financial implications? This presentation will review the impacts of overproduction on tax equity and debt financing models, while evaluating the major variables that contribute to plants outperforming the preͲconstruction energy estimates, such as: solar resource input data, plant loss assumptions, and uncertainty. Some developers are still relying on higher uncertainty or nonͲsite specific solar resource data for inputs into energy simulation models. These datasets typically have a larger uncertainty band and are less accurate, either underͲestimating or overͲestimating the resource at the project location. When used in an energy simulation there are direct impacts on the forecasted energy production. Additionally accurate plant loss considerations are crucial to best represent the longͲterm production of the solar plant. When poor plant loss assumptions are made in conjunction with inaccurate and high uncertainty solar resource datasets, actual plant production may vary considerably from the preͲconstruction estimate. Plant production and the associated revenue are key inputs into financial models used to size debt or tax equity contributions. Therefore, relying on a poor quality resource and energy estimate can lead to less than optimal debt sizing which can result in lower returns on investment for equity partners. Abstract Objectives Methodology Resource Inputs: To test the impacts of potential overͲproduction AWST evaluated three geographically diverse sites across the United States , studying the impacts of how various resource inputs predict longͲterm resource versus a high quality ground reference. The sources of data included Typical Meteorological Year (TMY) from: • ' Satellite derived dataset • ' National Solar Radiation Database (NSRDB) TMY3 [1] • ' Ground measured data from the United States Climate Reference Network (USCRN) [2] The ground measurements from the USCRN sites were used as the baseline for all results and assumed to be representative of the actual irradiance on site. The three project sites selected were in: • ' Merced, California • ' Tuscon, Arizona • ' Millbrook, New York Uncertainty for each dataset were assessed utilizing AWST’s standard approach. Energy Analysis: Energy was simulated for each of the studied resource files using the PVSyst Software. All three sites were simulated using AWS Truepower standard loss assumptions. Basic plant designs are as follows: • ' 12.5 MW DC / 10.0 MW AC (DC/AC Ratio: 1.25) • ' Generic 300 W polycrystalline module • ' Generic 500 kW inverter • ' Row tilt optimized for each project location using PVsyst • ' Modeled without near shading • ' Plant loss assumptions were applied consistently for each project location and resource analysis Uncertainty Analysis: Uncertainty around resource data, as a percentage of the resource data, is highly variable and dependent on the source of data, type of campaign, and accuracy of the sensors utilized. • ' Measurements: 1% Ͳ 5%, depending on the measurement campaign Investor Interests: • ' Cash Equity Investors: LongͲTerm energy production estimates (Reliant on P50 analysis) • ' Tax Equity Investors: Interests are greatest in the beginning of the project lifeͲcycle (Reliant on P50 analysis) • ' Debt Lenders: Interested in only the minimum ability to pay back loan (Reliant on P90/P99 analysis) Financial Model Key Performance Metrics: • ' Debt Ratio: Ratio of acquired debt to cash equity • ' Tax Equity Contribution: Amount financed through tax equity investor • ' Projected Internal Rate of Return (IRR): Estimated IRR for the cash equity investor when financing with each resource dataset • ' Actual IRR: Realized IRR for the cash equity investor when running the actual ground measured resource data file through the financial model for the Satellite TMY and NSRDB TMY3 Figure 3 presents the gross energy yield and longͲterm net energy yield for the NSRDB TMY3 and Satellite modeled TMY for each case study location. • ' From this graph the most notable trend is that the NSRDB TMY3 underͲpredicts the energy yield when compared to the ground measured data at all three locations. • ' The Satellite Modeled TMY is more variable across each region. Figure 4 presents the projected revenue for the Satellite Model and the NSRDB TMY3 at the Merced, CA location where both estimates underͲpredicted the resource and energy • ' In both cases the projected revenue is lower than what the plant would actually produce. • ' For the NSRDB TMY3 the estimate is low enough that over the longͲterm the PPA overproduction threshold would be met, resulting in lost revenue of approximately 0.5%. Figure 4 Project Revenue for the Merced, CA case study • ' Satellite Modeled: 5% Ͳ 10%, depending on the resolution of the model and the project location • ' NSRDB TMY3: 10% or greater, depending on the proximity of the dataset to the project location Projected Revenue: Power Purchase Agreement (PPA) pricing for each project location was developed using Locational Marginal Pricing (LMP)[3]: • ' The Base Price was assumed to be 1.5 x the average LMP • ' Time of Day and Seasonal (TOD) multipliers were developed from the raw LMP prices Potential project revenue was estimated from the hourly net energy and the TOD PPA price Production thresholds were assumed within the PPA structure. • ' Guaranteed Energy was assumed to be the annual P50 estimate • ' Overproduction limitations began at 110% of the Guaranteed Energy and paid 75% of the TOD PPA price. • ' Default was assumed to occur at 70% of the annual P50 Financial Model: Traditional Debt financing and Tax Equity Financing structures were evaluated: • ' CAPEX: $2.2/W AC Installed Capacity • ' OPEX: $25/kW AC /year Installed Capacity Debt Sizing: sized using a DSCR of 1.0 and P99 production estimates Tax Equity Investment sized assuming an 8.0% IRR Results Figure 5 Example of low uncertainty (left) and high uncertainty (right) Figure 5 illustrates how energy estimates with larger uncertainty bands can impact the P90/P99. • ' When financing with debt, use of higher uncertainty resource datasets will limit the risk lenders are willing to take on. Figure 6 presents the Debt Ratio, Projected IRR, and Actual IRR for a traditional debt financing structure. • ' The results of financing with the Satellite Modeled TMY for the Merced, CA case study shows that the cash equity investor would have yielded an Actual IRR of 0.5% lower than if a higher quality resource dataset was used for financing. If we assume a $10M cash equity investment over a 25Ͳyear project lifetime, the 0.5% IRR differential would have a cash equivalent of roughly $1.3M of unrealized revenue potential. • ' The results from the NSRDB TMY3 analysis yield actual IRR losses of nearly double those of the satellite model, which is due to the larger uncertainty band around the resource data. This point Illustrates the importance of high quality resource input data to reduce the spread between P50 to P90/P99 estimates. Figure 7 presents the Tax Equity Contribution, Projected IRR, and Actual IRR for a tax equity financing structure. • ' The results of financing with the Satellite Modeled TMY for the Merced, CA case study shows that the cash equity investor would have yielded an Actual IRR of 0.5% lower than if a higher quality resource dataset was used for financing. As in the scenario above, for a $10M cash equity investment over a 25Ͳyear project lifetime the cash equivalent is roughly $1.3M left on the table. Conclusions • ' Choosing an accurate and reliable source of solar resource data is critical for project financing • ' UnderͲestimating solar resource can lead to plant overproduction and lost revenue to PPA caps • ' Plant overͲproduction can result in lower returns for lenders, tax equity, and cash equity finance partners Solar Resource • GHI, DNI, DHI • POA Energy • Losses • Uncertainty Revenue • PPA • Over/Under Financial Model • CAPEX • OPEX • Debt Sizing Figure 1. Flow Chart of Case Study Methodology Annual Degradation (0.5 – 1 %) Transposition To Plane of Array (0.5 – 2%) Energy Simulation, Plant Losses (3 – 5 %) Solar Resource Uncertainty (5 – 17%) Figure 2. Sources of Energy Uncertainty Figure 6 Debt Financing Structure Project Yields Figure 7 Tax Equity Debt Structure Project Yields References [1] National Solar Radiation Data Base “1991 – 2010 Update” http://rredc.nrel.gov/solar/old_data/nsrdb/1991Ͳ2010/ [2] United States Climate Reference Network “Hourly Data” http://www1.ncdc.noaa.gov/pub/data/uscrn/products/hourly02/ [3] California ISO “Open Access SameͲtime Information System” http://oasis.caiso.com/ [4] Paul Thienpont, AWS Truepower “Is Solar Overproduction Costing You?” 18 November 2014 Webinar, https://www.awstruepower.com/knowledgeͲcenter/webinars/ AWS Truepower, LLC | 463 New Karner Road | Albany, New York 12205 | +1Ͳ518Ͳ213Ͳ0044 | awstruepower.com Paul Thienpont, Meteorologist – AWS Truepower, LLC, [email protected] © Copyright 2015

Upload: lekien

Post on 14-Mar-2018

218 views

Category:

Documents


3 download

Report

TRANSCRIPT

Page 1: Poster Title Poster Title continued - NREL · PDF filePoster Title Poster Title continued ... Energy was simulated for each of the studied resource files using the PVSyst Software

Poster TitlePoster Title continued

ResearchersrsquoPresentersrsquo NamesInstitutionOrganizationCompany

Methods

Methods

Figure3 ProjectEnergyYieldforallthreecasestudies

WhyResourceMatters ImpactsofPreconstructionResourceDataonLongͲTermProductionEstimatesandProjectFinance

PaulThienpontndash MeteorologistMarieSchnitzerndash VPConsultingServicesRebeccaTilbrookndash SolarServicesTeamLead

bull Determinehowvarioussourcesofsolarandmeteorologicaldataimpactproductionestimates

bull Determinehowuncertaintyofdatasetsinhibittheabilitytoleveragetheproject

bull Determinehowplantoverproductioncanleadtolostrevenue

bull Determinehowplantoverproductioncanresultinlowerrateofreturnsfordebttaxequityandcash equityfinancepartners

Accurateresourceandenergyproductionassessmentsareneededtoestablishprojectrevenueswith confidencetherebysizingdebtappropriatelyHoweverwhenplantperformanceexceedsthepreͲ constructionenergyestimatewhatarethefinancialimplications

Thispresentationwillreviewtheimpactsofoverproductionontaxequityanddebtfinancingmodels whileevaluatingthemajorvariablesthatcontributetoplantsoutperformingthepreͲconstruction energyestimatessuchassolarresourceinputdataplantlossassumptionsanduncertainty

SomedevelopersarestillrelyingonhigheruncertaintyornonͲsitespecificsolarresourcedataforinputs intoenergysimulationmodelsThesedatasetstypicallyhavealargeruncertaintybandandareless accurateeitherunderͲestimatingoroverͲestimatingtheresourceattheprojectlocationWhenusedin anenergysimulationtherearedirectimpactsontheforecastedenergyproductionAdditionally accurateplantlossconsiderationsarecrucialtobestrepresentthelongͲtermproductionofthesolar plantWhenpoorplantlossassumptionsaremadeinconjunctionwithinaccurateandhighuncertainty solarresourcedatasetsactualplantproductionmayvaryconsiderablyfromthepreͲconstruction estimate

Plantproductionandtheassociatedrevenuearekeyinputsintofinancialmodelsusedtosizedebtortax equitycontributionsThereforerelyingonapoorqualityresourceandenergyestimatecanleadtoless thanoptimaldebtsizingwhichcanresultinlowerreturnsoninvestmentforequitypartners

Abstract

Objectives

Methodology Resource Inputs TotesttheimpactsofpotentialoverͲproductionAWSTevaluatedthreegeographicallydiversesitesacrossthe

UnitedStatesstudyingtheimpactsofhowvariousresourceinputspredictlongͲtermresourceversusahigh qualitygroundreferenceThesourcesofdataincludedTypicalMeteorologicalYear(TMY)from bull Satellitederiveddataset bull NationalSolarRadiationDatabase(NSRDB)TMY3[1] bull GroundmeasureddatafromtheUnitedStatesClimateReferenceNetwork(USCRN)[2]

ThegroundmeasurementsfromtheUSCRNsiteswereusedasthebaselineforallresultsandassumedtobe representativeoftheactualirradianceonsite

Thethreeprojectsitesselectedwerein bull MercedCalifornia bull TusconArizona bull MillbrookNewYork

UncertaintyforeachdatasetwereassessedutilizingAWSTrsquosstandardapproach Energy Analysis EnergywassimulatedforeachofthestudiedresourcefilesusingthePVSyst Software AllthreesitesweresimulatedusingAWSTruepower standardlossassumptions Basicplantdesignsareasfollows

bull 125MWDC 100MWAC (DCACRatio125) bull Generic300Wpolycrystallinemodule bull Generic500kWinverter bull RowtiltoptimizedforeachprojectlocationusingPVsyst bull Modeledwithoutnearshading bull Plantlossassumptionswereappliedconsistentlyforeachprojectlocationandresourceanalysis

Uncertainty Analysis Uncertaintyaroundresourcedataasapercentageoftheresourcedataishighlyvariableanddependenton

thesourceofdatatypeofcampaignandaccuracyofthesensorsutilized bull Measurements1Ͳ 5dependingonthemeasurementcampaign

Investor Interests

bull CashEquityInvestorsLongͲTermenergyproductionestimates(ReliantonP50analysis) bull TaxEquityInvestorsInterestsaregreatestinthebeginningoftheprojectlifeͲcycle(ReliantonP50

analysis) bull DebtLendersInterestedinonlytheminimumabilitytopaybackloan(ReliantonP90P99analysis)

Financial Model Key Performance Metrics bull DebtRatioRatioofacquireddebttocashequity

bull TaxEquityContributionAmountfinancedthroughtaxequityinvestor bull ProjectedInternalRateofReturn(IRR)EstimatedIRRforthecashequityinvestorwhenfinancing

witheachresourcedataset bull ActualIRRRealizedIRRforthecashequityinvestorwhenrunningtheactualgroundmeasured

resourcedatafilethroughthefinancialmodelfortheSatelliteTMYandNSRDBTMY3

Figure3presentsthegrossenergyyieldandlongͲtermnetenergyyieldfortheNSRDBTMY3and SatellitemodeledTMYforeachcasestudylocation

bull FromthisgraphthemostnotabletrendisthattheNSRDBTMY3underͲpredictstheenergy yieldwhencomparedtothegroundmeasureddataatallthreelocations

bull TheSatelliteModeledTMYismorevariableacrosseachregion

Figure4presentstheprojectedrevenuefortheSatelliteModelandtheNSRDBTMY3atthe MercedCAlocationwherebothestimatesunderͲpredictedtheresourceandenergy

bull Inbothcasestheprojectedrevenueislowerthanwhattheplantwouldactuallyproduce bull FortheNSRDBTMY3theestimateislowenoughthatoverthelongͲtermthePPA

overproductionthresholdwouldbemetresultinginlostrevenueofapproximately05

Figure 4 ProjectRevenuefortheMercedCAcasestudy

bull SatelliteModeled5Ͳ 10dependingontheresolutionofthemodelandtheprojectlocation

bull NSRDBTMY310orgreaterdependingontheproximityofthedatasettotheprojectlocation

Projected Revenue PowerPurchaseAgreement(PPA)pricingforeachprojectlocationwasdevelopedusingLocationalMarginal

Pricing (LMP)[3] bull TheBasePricewasassumedtobe15xtheaverageLMP bull TimeofDayandSeasonal(TOD)multipliersweredevelopedfromtherawLMPprices

PotentialprojectrevenuewasestimatedfromthehourlynetenergyandtheTODPPAprice

ProductionthresholdswereassumedwithinthePPAstructure bull GuaranteedEnergywasassumedtobetheannualP50estimate

bull Overproductionlimitationsbeganat110oftheGuaranteedEnergyandpaid75oftheTODPPAprice bull Defaultwasassumedtooccurat70oftheannualP50

Financial Model TraditionalDebtfinancingandTaxEquityFinancingstructureswereevaluated

bull CAPEX$22WAC InstalledCapacity bull OPEX$25kWACyearInstalledCapacity

DebtSizingsizedusingaDSCRof10andP99productionestimates TaxEquityInvestmentsizedassumingan80IRR

Results

Figure5 Exampleoflowuncertainty(left)andhighuncertainty(right)

Figure5illustrateshowenergyestimateswithlargeruncertaintybandscanimpactthe P90P99

bull Whenfinancingwithdebtuseofhigheruncertaintyresourcedatasetswilllimittherisk lendersarewillingtotakeon

Figure6presentstheDebtRatioProjectedIRRandActualIRRforatraditionaldebtfinancing structure

bull TheresultsoffinancingwiththeSatelliteModeledTMYfortheMercedCAcasestudy showsthatthecashequityinvestorwouldhaveyieldedanActualIRRof05lowerthanifa higherqualityresourcedatasetwasusedforfinancingIfweassumea$10Mcashequity investmentovera25Ͳyearprojectlifetimethe05IRRdifferentialwouldhaveacash equivalentofroughly$13Mofunrealizedrevenuepotential

bull TheresultsfromtheNSRDBTMY3analysisyieldactualIRRlossesofnearlydoublethoseof thesatellitemodelwhichisduetothelargeruncertaintybandaroundtheresourcedata ThispointIllustratestheimportanceofhighqualityresourceinputdatatoreducethe spreadbetweenP50toP90P99estimates

Figure7 presentstheTaxEquityContributionProjectedIRRandActualIRRforataxequity financingstructure

bull TheresultsoffinancingwiththeSatelliteModeledTMYfortheMercedCAcasestudy showsthatthecashequityinvestorwouldhaveyieldedanActualIRRof05 lowerthanif ahigherqualityresourcedatasetwasusedforfinancingAsinthescenarioabovefora $10Mcashequityinvestmentovera25Ͳyearprojectlifetimethecashequivalentisroughly $13Mleftonthetable

Conclusions bull Choosinganaccurateandreliablesourceofsolarresourcedataiscriticalforprojectfinancing

bull UnderͲestimatingsolarresourcecanleadtoplantoverproductionandlostrevenuetoPPAcaps

bull PlantoverͲproductioncanresultinlowerreturnsforlenderstaxequityandcashequityfinance partners

Solar Resource

bull GHI DNI DHI bull POA

Energy bull Losses bull Uncertainty

Revenue bull PPA bull OverUnder

Financial Model

bull CAPEX bull OPEX bull Debt

Sizing

Figure1FlowChartofCaseStudyMethodology

Annual Degradation (05 ndash 1 )

Transposition To Plane of Array

(05 ndash 2)

Energy Simulation Plant Losses

(3 ndash 5 )

Solar Resource Uncertainty (5 ndash 17)

Figure2SourcesofEnergyUncertainty

Figure6DebtFinancingStructureProjectYields

Figure7 TaxEquityDebtStructureProjectYields

References [1]NationalSolarRadiationDataBaseldquo1991ndash 2010Updaterdquo httprredcnrelgovsolarold_datansrdb1991Ͳ2010 [2]UnitedStatesClimateReferenceNetworkldquoHourlyDatardquo httpwww1ncdcnoaagovpubdatauscrnproductshourly02 [3]CaliforniaISOldquoOpenAccessSameͲtimeInformationSystemrdquohttpoasiscaisocom [4]PaulThienpontAWSTruepower ldquoIsSolarOverproductionCostingYourdquo18November2014 WebinarhttpswwwawstruepowercomknowledgeͲcenterwebinars

AWSTruepowerLLC|463NewKarner Road|AlbanyNewYork12205|+1Ͳ518Ͳ213Ͳ0044|awstruepowercom

PaulThienpontMeteorologistndash AWSTruepowerLLCpthienpontawstruepowercom

copyCopyright2015

Title of the Poster Authors of the poster Affiliation - Email

POSTER PRESENTER TITLE OF POSTER NUMBER 1 2 3 4epsassets.manchester.ac.uk/medialand/eee/pdf/pgr_conference... · PRESENTER TITLE OF POSTER 1 Abdulaziz ... The inclined plane test

Poster Title Primary Author Organization

POSTER LAYOUT H01 modelo poster i3S...Title POSTER_LAYOUT_H01 Created Date 20151201194539Z

Poster Name & Discipline Designation Abstract Title

NOTES 1.Poster Title

Parallel and Scalable Combinatorial String and Graph ... · Title: 48x36 Poster Template Subject: Free PowerPoint poster templates Keywords: poster presentation, poster design, poster

Poster Title Goes Here

Movie Poster Project Rubric - rcboe.org · Movie Poster Project Rubric Names:_____ Points 0 5 10 15 Score Book Title (10 Points) Title is omitted from poster Title is used incorrectly

Title of your Poster Presentation

S.No Poster Title Author's Name Poster Number 2016/List of Posters.pdf ·  · 2016-06-07S.No Poster Title Author's Name Poster Number ... Characterisation and Applications of Tin

Title of your poster presentation: Subtitle of the poster A ......Title of your poster presentation: Subtitle of the poster The ChEESE project has received funding from the European

Title And Poster

poster · Title: poster Created Date: 10/1/2019 2:21:06 PM

Infrastructure Poster V1 - ACT Acorn · Title: Infrastructure Poster V1 Created Date: 20190205144812Z

Townend Bradley - A2 poster · Title: Townend Bradley - A2 poster Created Date: 20150112105952Z

kaizan poster - qcfimc.com poster.pdf · Title: kaizan poster Author: Administrator Created Date: 20140425061431Z

Poster Session (continued) - Duke University...Poster Session (continued) Nickolich Outcomes for Percutaneous Coronary Intervention vs DUKE INTERNAL MEDICINE RESIDENCY PROGRAM Resident

Full Title of your Poster Here

Open 5 poster A3...Title Open 5 poster A3 Created Date 20200225091938Z

Poster Title XXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXX€¦ · Poster Title XXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXX Abstract Your poster will be printed on an HP designjet

ABSTRACT 4’x4’ POSTER TITLE 4’x4’ POSTER TITLE, PRINTED @100% 4’x4’ POSTER TITLE AUTHORS AFFILIATES CHARTS/FIGURESMATERIALS INTRODUCTION CONCLUSIONS METHODS

Title of poster Authors - Hulsen

Poster A2 Congresso€¦ · Title: Poster A2 Congresso Created Date: 20191031151816Z

POSTER TITLE AND NUMBER INDEX Poster Zonesplantbiology2015.aspb.org/wp-content/uploads/2015/07/FINAL-Poster... · POSTER TITLE AND NUMBER INDEX . Poster Zones . ... Axillary Bud versus

Party Poster Small 2 · Title: Party Poster Small 2 Created Date: 20170816193016Z

C:UsersAlbertDesktopPPFORMATO POSTER Presentación (1)deim.urv.cat/~blas.herrera/poster congreso_Linea 1... · Title: C:UsersAlbertDesktopPPFORMATO POSTER Presentación (1) Author:

KVPY Poster curvedkvpy.iisc.ernet.in/main/resources/poster-hindi.pdf · Title: KVPY Poster curved Author: a Created Date: 20100707100021Z

! 2009 WSCG Title page POSTER 1

poster~2 - tomdoughty.com€¦ · Title: poster~2 Author: Richard Created Date: 20080214182549Z

poster · 2012. 10. 28. · Title: poster Author: GR Created Date: 20120831094103Z

POSTER TITLE - Northwestern University

Poster Title - CPSA USA · CPSA 2013 Poster Assignments: Poster # Title 1 Fine Isotopic Pattern Recognition for Small Molecule Identification 2 The increasingly recognized potential

Continued Afm Exploration Real Poster

English Title and poster "Gladiator"