undp derisking renewable energy investment

7/27/2019 UNDP Derisking Renewable Energy Investment

1/156

Contents

Copy goes here

Derisking Renewable Energy Investment

Book Title

c

DEFINITIONS

Nam ipsum mauris

Dapibus et tristique ac,consectetur ac nunc.Fusce pulvinar eros inlibero eleif end sodalesfringilla risus lobor ryultis. Duis ullamcorperlaoreet sapien faucibustincidunt.

Aliquam sit amet pelitulentesque nunc. Sed eutmauris tellus, et eleifendurna. Fusce ultricies treyvehicula mauris pulvinarullamcorper.

Derisking RenewableEnergy Investment

A Framework to Support Policymakers in SelectingPublic Instruments to Promote Renewable EnergyInvestment in Developing Countries


2/156


3/156

DEFINITIONS

Nam ipsum mauris

Dapibus et tristique ac,consectetur ac nunc.Fusce pulvinar eros inlibero eleif end sodalesfringilla risus lobor ryultis. Duis ullamcorperlaoreet sapien faucibustincidunt.

Aliquam sit amet pelitulentesque nunc. Sed eutmauris tellus, et eleifendurna. Fusce ultricies treyvehicula mauris pulvinarullamcorper.

1

Figures, Tables and Boxes 2

Acronyms 6

Foreword 8

Executive Summary 11

Introduction 28

1. The Role o Public Instruments in Reducing Financing 31Costs or Renewable Energy in Developing Countries1.1 High Financing Costs or Renewable Energy 31

1.2 The Role o Public Instruments in Reducing Financing Costs 35

1.3 Challenges to Identi ying an Appropriate Public Instrument Mix 38

2. A Framework to Select Public Instruments to Promote 43Renewable Energy Investment2.1 Stage 1: Risk Environment 462.2 Stage 2: Public Instruments 55

2.3 Stage 3: Levelised Cost 67 2.4 Stage 4: Evaluation 75

3. Illustrative Country Case Studies 79 3.1 Approach to the Modelling Exercise 79

3.2 Country Results or South A rica 883.3 Country Results or Panama 96

3.4 Country Results or Mongolia 103 3.5 Country Results or Kenya 112

4. Implications or Public Finance o Scaling Up Renewable Energy 1214.1 Public Finance E ectiveness to Trans orm Renewable Energy Markets 1214.2 Public Finance E ciency to Trans orm Renewable Energy Markets 1234.3 The Distributional Impact o Renewable Energy Policies 1254.4 Scaled-up Climate Change Mitigation Outcomes 127

Conclusion 131

Annexes 135 A. Methodology and Data or the Illustrative Modelling Exercise 135

B. Re erences 147

Table o Contents



4/156

Contents

Copy goes hereNulla ut nisl etlyneque lobortis islu

egestas sit amet quisconsectetur ac nunc.

FVivamu nisi mi elis,tincidun non blandit

vestibu luma.

2

FIGURESExecutive SummaryFigure 1: Impact o nancing costs on wind and gas power generation costs in developed and

developing countriesFigure 2: Shi ting the risk-reward pro le o renewable energy projectsFigure 3: Public instrument selection or large-scale renewable energyFigure 4: Public derisking instruments can reduce nancing costs o renewable energy investmentsFigure 5: Overview o the ramework to select public instruments to promote renewable

energy investmentFigure 6: The our country case studies and their illustrative combinations o public instrumentsFigure 7: Illustrative modelling exercise or Kenya (Wind, 1GW): selected resultsFigure 8: Scaled-up mitigation actions blending derisking instruments and per ormance-based

payments

IntroductionFigure 9: Investments in clean energy by type o countries (USD billions)

Chapter 1Figure 10: The core drivers o the LCOEFigure 11: The di erent capital intensity o electricity production rom wind energy and combined

cycle gasFigure 12: Impact o nancing costs on wind and gas power generation costs in developed and

developing countriesFigure 13: Shi ting the risk-reward pro le o renewable energy investmentFigure 14: Public instrument selection or large-scale renewable energy

Chapter 2Figure 15: Public derisking instruments can reduce nancing costs o renewable energy investmentsFigure 16: Overview o the ramework to support policymakers in selecting public instruments to

promote renewable energy investmentFigure 17: Overview o Stage 1: Risk EnvironmentFigure 18: Drivers and components o investor risk or renewable energy investmentFigure 19: Illustrative nancing cost water all quanti ying the impact o risks on increasing

nancing costsFigure 20: Interview questions to quanti y the impact o risk categories on the cost o equity and debtFigure 21: Illustrative simpli ed application o the methodology to determine the impact o risk

categories on increasing nancing costsFigure 22: Overview o Stage 2: Public InstrumentsFigure 23: The e ects o policy and nancial derisking on investor risk Figure 24: Illustrative post-derisking cost o equity water all, identi ying the impact o public instruments

in reducing the incremental nancing costs attributable to investor risk categoriesFigure 25: Overview o Stage 3: Levelised Cost

Figures, Tables and Boxes




5/156

3

Figure 26: Illustrative baseline energy mixFigure 27: Illustrative comparison o the LCOE o pre- and post-derisking renewable energy

investments in comparison to the baseline energy mixFigure 28: Overview o Stage 4: EvaluationFigure 29: Visualisation o the investment leverage ratioFigure 30: Visualisation o the savings leverage ratioFigure 31: Visualisation o end-user a ordabilityFigure 32: Visualisation o carbon abatement potential and costFigure 33: Key drivers or sensitivity analyses

Chapter 3Figure 34: The our country case studies and their illustrative combinations o public instruments

South A ricaFigure 35: Energy generation mix in South A rica (1971 to 2009)Figure 36: Wind map o South A ricaFigure 37: Impact o risk categories on nancing costs or wind energy investment in South A rica,

business-as-usual scenarioFigure 38: Impact o policy derisking instruments on reducing nancing costs or wind energy

in South A ricaFigure 39: LCOE or the marginal baseline and wind investment in South A ricaFigure 40: Per ormance metrics or the selected package o policy derisking instruments in

promoting 8.4 GW o wind energy investment in South A rica

PanamaFigure 41: Energy generation mix in Panama (1971 to 2009)Figure 42: Wind map o PanamaFigure 43: Impact o risk categories on nancing costs or wind energy investment in Panama,


in PanamaFigure 45: LCOE or the marginal baseline and wind investment in PanamaFigure 46: Per ormance metrics or the selected package o policy derisking instruments in

promoting 1 GW o wind energy investment in Panama

MongoliaFigure 47: Energy generation mix in Mongolia (1985 to 2009)Figure 48: Wind map o MongoliaFigure 49: Impact o risk categories on nancing costs or wind energy investment in Mongolia,


in MongoliaFigure 51: LCOE or the marginal baseline and wind investment in Mongolia




6/156

Contents




vestibu luma.

4

Figure 52: Per ormance metrics or the selected package o policy derisking instruments inpromoting 1 GW o wind energy investment in Mongolia

Ken aFigure 53: Energy generation mix in Kenya (1971 to 2009)Figure 54: Wind map o KenyaFigure 55: Impact o risk categories on nancing costs or wind energy investment in Kenya,


in Kenya

Figure 57: LCOE or the marginal baseline and wind investment in KenyaFigure 58: Per ormance metrics or the selected package o policy derisking instruments

in promoting 1 GW o wind energy investment in Kenya

Chapter 4Figure 59: Overview o the modelling exercises results or investment leverage ratiosFigure 60: Overview o the modelling exercises results or savings leverage ratiosFigure 61: Overview o the modelling exercises results or end-user a ordabilityFigure 62: Overview o the modelling exercises results or carbon abatement

ConclusionFigure 63: Scaled-up mitigation actions blending derisking instruments and per ormance-based

payments

AnnexesFigure 64: The modelling exercises uel price assumptions

TABLESChapter 1Table 1: Examples o the evolution o public instruments in the short-, medium- and long-term

Chapter 2Table 2: Typical stakeholders or large-scale renewable energy projectsTable 3: A generic multi-stakeholder barrier and risk table or large-scale, on-grid

renewable energy deployment in developing countriesTable 4: A generic public instrument table or large-scale, on-grid renewable energy

deployment in developing countries

Chapter 3Table 5: The modelling exercises public instrument table

South A ricaTable 6: Investor eedback on risk categories or wind energy investment in South A rica




7/156

5

Table 7: Example sensitivity analyses on the South A rica case studys per ormance metricswhen varying key inputs by +/- 10%

Table 8: Summary assumptions or the South A rica case study

PanamaTable 9: Investor eedback on risk categories or wind energy investment in PanamaTable 10: Example sensitivity analyses on the Panama case studys per ormance metrics

when varying key inputs by +/- 10%Table 11: Summary assumptions or the Panama case study

MongoliaTable 12: Investor eedback on risk categories or wind energy investment in MongoliaTable 13: Example sensitivity analyses on the Mongolia case studys per ormance metrics when

varying key inputs by +/- 10%Table 14: Summary assumptions or the Mongolia case study

Ken aTable 15: Investor eedback on risk categories or wind energy investment in KenyaTable 16: Example sensitivity analyses on the Kenya case studys per ormance metrics when

varying key inputs by +/- 10%Table 17: Summary assumptions or the Kenya case study

Annexes

Table 18: Interview sample size in each o the modelling exercises countriesTable 19: The modeling exercises assumptions or policy derisking instruments e ectivenessTable 20: The modelling exercises assumptions on costing o nancial derisking instrumentsTable 21: The modelling exercises assumptions or the baseline energy mixTable 22: The modelling exercises assumptions or marginal baseline emission actorsTable 23: The modelling exercises assumptions or wind energy ull load hoursTable 24: The modelling exercises assumptions on technology speci cations or wind energy

BOXESChapter 1Box 1: How di erent investment types a ect the pricing o risk into nancing costs

Chapter 2Box 2: Methodology or quanti ying the impact o risk categories on increasing nancing costsBox 3: The di erent e ects o policy and nancial derisking instruments

AnnexesBox 4: The eight investment assumptions or wind energy in the our countriessBox 5: The modelling exercises LCOE ormula




8/156

Contents




vestibu luma.

6

General AcronymsBAU Business as usualBDP UNDP Bureau or Development PolicyBM Build marginBNEF Bloomberg New Energy FinanceBOO Build-own-operateBOP Balance-o -plantCCGT Combined cycle gas turbineCDM Clean Development MechanismCM Combined marginECN Energy Research Centre o the NetherlandsEIA Environmental impact assessmentEITT UNDP Energy, In rastructure, Transport and Technology teamEPC Engineering, procurement and constructionESCO Energy service companyFDI Foreign direct investmentFiT Feed-in tari GDP Gross domestic productGCF Green Climate Fund

GEF Global Environment FacilityGHG Greenhouse gasGW GigawattHDI Human Development IndexIEA International Energy AgencyIIASA International Institute or Applied Systems AnalysisIPP Independent power producerIRENA International Renewable Energy AgencyISO Organisation or International StandardisationkW KilowattkWh Kilowatt-hourLCOE Levelised cost o electricityLDC Least developed countryMIGA Multilateral Investment Guarantee Agency (World Bank)MW MegawattNAMA Nationally Appropriate Mitigation ActionNIMBY Not in my back yardNMM New market mechanismNREL National Renewable Energy Laboratory


Acronyms

Acronyms


9/156

7Derisking Renewable Energy Investment

Acronyms

O&M Operations and maintenanceOECD Organisation or Economic Cooperation and DevelopmentOM Operating marginPDD CDM project design documentPoA CDM programme o activitiesPPA Power purchase agreementPPP Purchasing power parityPRI Political risk insurancePV PhotovoltaicREFIT Renewable energy eed-in tari REN21 Renewable Energy Policy Network or the 21st CenturyRET Renewable energy technologyRFP Request or proposalUN DESA United Nations Department o Economic and Social A airsUNDP United Nations Development ProgrammeUNEP United Nations Environment ProgrammeUNFCCC United Nations Framework Convention on Climate ChangeUN REDD United Nations Collaborative Programme on Reducing Emissions rom De orestation and

Forest DegradationUSD United States dollarVAT Value-added tax

South A rica Case Study AcronymsNERSA National Energy Regulator o South A ricaSAWEA South A rican Wind Energy AssociationZAR South A rican rand

Panama Cast Study AcronymsANAM Autoridad Nacional del AmbienteASEP Autoridad Nacional de Servicios PublicosENEL Ente Nazionale per l'Energia Elettrica (Italy)

ETESA Empresa de Transmisin Elctrica S.ASNE Secretara Nacional de Energa

Mongolia Case Study AcronymsERA Energy Regulatory Authority

Kenya Case Study AcronymsKETRACO Kenya Electricity Transmission CompanyKPLC Kenya Power and Lighting Company


10/156


11/156

Derisking Renewable Energy Investment 9

Foreword

The challenge o addressing these barriers has inspired the development o a wide array o public measuresto promote renewable energy. However, public instruments to catalyse clean energy nance come at acost. Irrespective o the instrument port olio that is selected, there will be a cost to industry, consumersor the tax-payer. This publication, together with its accompanying nancial tool, introduces an innovativeramework developed by UNDP to help decision-makers quantitatively compare di erent public instrumentsand their environmental and cost e ectiveness. To illustrate how the ramework can support policydecision making in practice, the publication provides the results rom case studies in our countries. It drawson these results to discuss possible directions or enhancing the e ectiveness and e ciency o publicnance to catalyse renewable energy investment and provide universal access to clean, secure and a ordableenergy services.

I hope that policymakers, development practitioners and the renewable energy community at large willnd this rst version o the ramework help ul. UNDP looks orward to collaborating with our partners tourther develop and re ne it thereby hope ully making a contribution to addressing global warming and toassisting developing countries in meeting their sustainable energy objectives.

Rebeca Gr nspanAssociate Administrator,United Nations Development Programme


12/156

Executive Summary


13/156


Executive Summary

Executive Summary

THE ROLE OF PUBLIC INSTRUMENTS IN REDUCING FINANCINGCOST OF RENEWABLE ENERGYAround the world, developing countries are seeking to rapidly scale-up renewable energy investment. Thisshi t to renewable energy is driven by a number o considerations. Many developing countries are strugglingto meet ast-growing energy demand. About 1.3 billion people still lack access to electricity and 2.7 billionto modern energy services, with their human development held back through energy poverty (UN, 2011).Meanwhile, rising global uel prices and resource scarcities are making developing countries increasinglyvulnerable to oil prices. Over one-third o low-income countries already pay more than 10 percent o theirgross domestic product (GDP) to secure their oil supply (Economy Watch, 2011; Seth, 2012).

At the same time, the technology cost o renewable energy has been experiencing remarkably steady allsover the past decades (nearly 98 percent or solar photovoltaic (PV) modules since 1979, or instance (IRENA,2012a)). It has been suggested that a sustained technology push by a ew pioneer countries could urtherreduce technology costs, enabling renewable energy to out-compete ossil uels by the end o this decade.However, barriers towards a ull-scale transition to renewable energy in developing countries lie not justin technology costs but in the challenges o securing long-term a ordable nance. Financing cost is theprimary determinant o generation cost or renewable sources, as renewable energy (other than biomass andbio uel) has no uel cost but does have high up ront investment costs.

The nancial sums involved in a rapid shi t to low-emission energy systems are enormous. According tothe Global Energy Assessment (GEA, 2012), global investment in energy e ciency and low-carbon energy

generation will need to increase to between USD 1.72.2 trillion per year compared with present levelso about USD 1.3 trillion over the coming decades to meet the combined challenges o energy access,energy security and climate change. In order to success ully scale-up renewable energy in developingcountries, it is clear that private sector investment must be at the ore ront. In principle, with enablingpolicies and investment practice aligned, global capital markets, amounting to some USD 212 trillion innancial assets (McKinsey, 2011), have the size and depth to step up to the investment challenge. However,project developers in developing countries o ten struggle to access the large quantities o nancing theyneed. When available, the inancing cost o this up ront investment is substantially higher than indeveloped countries, translating into higher power generation costs or renewable energy technologies.

The di erence in nancing costs (debt and equity) can dramatically a ect the competitiveness o renewableenergy versus ossil uel technologies in developing countries. Figure 1 compares the 2012 levelised cost o electricity (LCOE)1 o a generic onshore wind energy plant and a combined-cycle gas plant in a developedcountry with those o the same plants in a developing country. In a developed country bene ting rom lownancing costs, wind power can be almost cost-competitive with gas, despite the present a ordability o natural gas. All other assumptions kept constant, in a developing country with higher nancing costs, windpower generation cost becomes 40 percent more expensive than that o gas because o the up ront capitalintensity o wind technologies.

1 The levelised cost o electricity (LCOE) is a popular metric to compare di erent types o systems rom renewable energy projects, where theup ront capital cost is high and the uel cost is near-zero, to a natural gas plant, where the capital cost is lower but the uel cost is higher. LCOEallocates the costs o an energy plant across its use ul li e to give an e ective price per each unit o energy ( or example, USD/kWh).

In order tosuccess ully scale-urenewable energyin developingcountries, it is cleathat private sectorinvestment mustbe at the ore ront.


14/156

Derisking Renewable Energy Investment12

Executive Summary

The higher nancing costs in developing countries refect a number o perceived or actual in ormational,technical, regulatory, nancial and administrative barriers and their associated investment risks. A countryneeds to provide potentially very high return rates to investors to succeed in attracting private investmentsor wind power development i independent power producers (IPPs) ace barriers in access to grids, lengthyand uncertain processes to issue permits, limited local supply o expertise or a lack o long-term price

guarantees.

Rather than a problem o capital generation, the key challenge o unding the transition towards a low-carbon energy system is to address existing investor risks that a ect the nancing costs and competitivenesso renewable energy in developing countries. The task o addressing these investor risks has inspired thedevelopment o a wide array o public instruments over recent years. Public derisking measures can broadlybe divided into two groups:

Developed country(Wind vs. Gas)

Developed CountryCost of Equity = 10%Cost of Debt = 5%

Developing CountryCost of Equity = 18%Cost of Debt = 10%

Developing country(Wind vs. Gas)

P R E

T A X L C O E U S D

C E N T S / k W h

Gas(CCGT)

Wind(onshore)

Gas(CCGT)

Wind(onshore)

Financing Cost(Equity)

Financing Cost(Debt)

Operating Cost(incl. fuel cost)

Investment Cost/Depreciation

2.9

1.1

0.9

1.8

6.7

+ 40%

0.7

4.5

0.30.6

6.1

2.9

0.8

2.4

3.3

9.4

0.7

3.9

0.9

1.0

6.5+ 6%

Figure 1: Impact o nancing costs on wind and gas power generation costs

All assumptions besides the nancing costs are kept constant between the developed and developing country.For technology assumptions, see inputs or wind energy and gas (CCGT) in Section A.3 (Annex A); a 70%/30% debt/equity capitalstructure is assumed; nancing costs are based on data in the our country case study (Chapter 3), assuming a non-investment gradedeveloping country.Operating costs appear as a lower contribution to LCOE in developing countries due to discounting e ects rom higher nancing costs.

The highernancing costsin developing

countries refecta number o perceived or

actual investmentrisks.


15/156


Executive Summary

Polic derisking instruments seek to remove the underlying barriers that are the root causes o risks. Theseinstruments include, or example, support or renewable energy policy design, institutional capacitybuilding, resource assessments, grid connection and management, and skills development or localoperations and maintenance (O&M).

Financial derisking instruments do not seek to directly address the underlying barriers but, instead,trans er the risks that investors ace to public actors, such as development banks. These instruments caninclude, or example, loan guarantees, political risk insurance (PRI) and public equity co-investments.

Recognising that not all risks can be eliminated through policy derisking or trans erred through nancial

derisking, e orts to reduce risks can be supplemented by direct nancial incentives (price premiums, taxbreaks, carbon o sets, etc.) to compensate or residual incremental costs and to thereby increase returns. Theoverall aim is to achieve a risk/return pro le that can attract private sector investment.

Figure 2 provides a conceptual illustration o the approach. The igure illustrates a shi t rom a commerciallyunattractive investment opportunity (right) to a commercially attractive one (top). This is achieved throughtwo actions: irst, by reducing the risk o the ac tivity, or example, through a regulatory policy such asguaranteed access to the grid or IPPs; and, second, by increasing the return on investment by, or example,creating nancial incentives, such as a premium price or renewable energy.

F I N A N C I A L R E T U R N

RISK OF INVESTMENT

Infeasiblerenewable

energy project

Example:price premium

Feasiblerenewable

energy project

Example: guaranteed access to the grid

Figure 2: Shi ting the risk reward pro ile o renewable energ projects

Source: Glemarec (2011), adapted.

Public deriskingcan be supplementeby direct nancialincentives tocompensate orresidual incrementalcosts.


16/156


17/156


Executive Summary

A key constraint or policymakers is that they currently lack a way to quantitatively compare di erent setso public instruments. In order to better understand and clearly communicate the impact o di erentcombinations o public derisking mechanisms in a given context, UNDP has developed a ramework thatenables planners and decision-makers to quanti y assumptions.

Figure 3: Public instrument selection or large scale renewable energ


+

Direct Financial Incentives(I positive incremental cost)

Examples:

FiT/PPA price premium

Select Cornerstone Instrument

Examples:

Feed-in tari

PPA-based bidding process

Select PolicDerisking Instruments

Examples:

Long-term RE targets

Streamlined permits process

Improved O&M skills

Select FinancialDerisking Instruments

Examples:

Public loans

Partial loan guarantees

Political risk insurance

Tax credits

Carbon o sets


18/156


Executive Summary

A FRAMEWORK TO SELECT PUBLIC INSTRUMENTS TOPROMOTE RENEWABLE ENERGY INVESTMENT

The theory o change 3 underlying the ramework is that one o the main challenges or scaling-uprenewable energy technologies in developing countries is to lower the nancing costs that a ect theircompetitiveness against ossil uels. As these higher nancing costs refect barriers and associated risks inthe investment environment, the key entry point or policymakers to oster renewable energy technologies isto address these risks and thereby lower overall li e-cycle costs. This theory o change draws rom UNDPsexperience in renewable energy market trans ormation in over 80 developing economies (Glemarec et al.,2012) as well as rom the ndings o a recent UNDP research partnership with Deutsche Bank on eed-in tari s(DB Climate Change Advisors, 2011). Figure 4 illustrates this theory o change and how public instruments,through addressing barriers to investment, can reduce the nancing costs o renewable energy investmentsand attract capital at scale. 4

3 Theory o change is an increasingly common concept used in international development (Vogel, 2012). While there is no single de nition o theterm, it is here used to articulate UNDPs underlying assumptions o how and why change might happen as a result o a public programmes actions.

4 In this gure, operational and investment costs are shown as remaining constant. There are two main reasons or this. First, it is recognised thatbarriers to investment can lead to higher operational and investment costs. For example, an investor may incur additional costs in a prolongedattempt to obtain permits i the permits process is poorly designed. Similarly, an investor may incur additional costs in fying technicians romabroad or project commissioning and O&M in the absence o a local supply o expertise. The ramework is based on the assumption thatthe possibility o higher costs brought about by investor barriers are actored into the up ront investment decision and result in t he investordemanding a higher return on investment, which translates into a higher cost o capital (McKinsey 2012, DB Climate Change Advisors, 2011).Second, the gure only addresses the role o public derisking instruments. The gure there ore does not include direct nancial incentives,such as production tax credits or accelerated depreciation, which do reduce operational and investment costs.

USD/kWh

USD/kWh

Pre-DeriskingLife-cycle Costs (LCOE)

Post-DeriskingLife-cycle Costs (LCOE)

Renewable EnergyRenewable Energy

Cost of Equity

Financing Costs

Technology Costs

Operational Costs

Investment Costs/Depreciation

Cost of Debt

Figure 4: Public derisking instruments can reduce nancing costs o renewable energ investments

One o the mainchallenges or

scaling-up renewableenergy in developing

countries is to lowerthe nancingcosts that a ect their

competitivenessagainst ossil uels.


19/156


Executive Summary

The ramework aims to support policy decis ion making by quant itat ively comparing di erent publicinstrument port olios and their impacts. The intent o the ramework is not to provide one predominantnumeric result, but instead is to acilitate a structured, transparent process whereby key inputs andassumptions are made explicit, so that they can be checked, debated and enriched to strengthen thedesign o market trans ormation initiatives or renewable energy.

As illustrated in Figure 5, the ramework is organised into our stages, each o which is, in turn, dividedinto two steps. UNDP is also releasing an LCOE-based nancial tool in Microso t Excel, available at UNDP'swebsite (www.undp.org), to accompany the ramework. Chapter 2 provides a detailed description o therameworks our stages.

Stage 1: Risk Environment identi es the set o investment barriers and associated risks relevant to therenewable energy technology, and analyses how the existence o investment risks can increase nancingcosts. 5

Stage 2: Public Instruments selects a mix o public derisking instruments to address the investor risks andquanti es how they in turn can reduce nancing costs. This stage also determines the cost o the selectedpublic derisking instruments.

Stage 3: Levelised Cost determines the degree to which the reduced inancing costs impact therenewable energys li e-cycle cost (LCOE). This is then compared against the current baseline generationcosts in the country.

Stage 4: Evaluation assesses the selected public derisking instrument mix using our per ormancemetrics, as well as through the use o sensitivity analyses. The our metrics are: (i) investment leverage ratio,(ii) savings leverage ratio, (iii) end-user a ordability and (iv) carbon abatement.

5 A key step in Stage 1 is determining a multi-stakeholder barrier and risk table or the particular renewable energy. This table identi ies a seto independent risk categories in the investment universe which can subsequently be submitted to numeric treatment under the ramework.Independent (i.e., non-overlapping) risk categories are important as strongly correlated risk categories would undermine the rameworksquanti ication process.

The ramework acilitates astructured,transparentprocess wherebykey inputs andassumptionsare madeexplicit.


20/156


Executive Summary

Figure 5: Overview o the ramework to select public instruments to promote renewable energ investment

Step 1

Step 1

Step 2

Step 2

Determine a multi-stakeholder barrierand risk table or the renewable energyinvestment

Select one or more public deriskinginstrument(s) to mitigate the identi edrisk categories

Quanti y the impact o risk categories onincreased nancing costs

Quanti y the impact o the public deriskinginstrument(s) to reduce nancing costs

Quanti y the public costs o the publicderisking instrument(s)

Main Output:Multi-stakeholder Barrier and Risk Table

Main Output:Public Instrument Table

STAKEHOLDERS BARRIER RISK CAT EGORY

End-usersBarrier #1

Risk #1Barrier #2

Supply chainBarrier #3

Risk #2Barrier #4

BARRIERRISK

CATEGORY

POLICYDERISKING

INSTRUMENT

FINANCIALDERISKING

INSTRUMENT

Barrier #1Risk #1

Instrument #1

Barrier #2 Instrument #2

Barrier #3 Risk #2 Instrument #3 Instrument #1

Main Output:Financing Costs Water all

Main Output:Post-Derisking Water all

%%

Best-in-Class(Developed Country)

Cost o Equity/Debt

Risk #2

Risk #3

Risk #1

Pre-Derisking(Developed Country)Cost o Equity/Debt

%%

Pre-DeriskingCost o Equity/Debt

DeriskingInstrument

#2

DeriskingInstrument

#1

Post-DeriskingCost o Equity/Debt

Stage 1:Risk Environment

Stage 2:PublicInstruments


21/156


Executive Summary

Step 1

Step 1

Step 2

Step 2

Calculate the levelised cost o electricity(LCOE) or the baseline energy generationmix in the particular country

Analyse the selected public instrument(s)in terms o our key per ormance metrics

Quanti y the LCOE or renewable energyinvestment in the (i) pre-derisking and(ii) post-derisking scenarios

Calculate the incremental cost (or savings) bycomparing these scenarios vs. the baseline

Per orm sensitivity analyses on key inputsand assumptions

Main Output:Baseline LCOE

Main Output:4 Per ormance Metrics

Coal 52%

Diesel Oil 21%

Hydro 27%

InvestmentLeverage Ratio

SavingsLeverage Ratio

End-userA ordability

CarbonAbatement

Main Output:Incremental Cost (via LCOE)

Main Output:Sensitivity Analyses

USD/kWh

USD/kWh

USD/kWh

BaselineActivity

Renewable EnergyInvestment

Pre-Derisking

Renewable EnergyInvestment

Post-Derisking

USD/kWh

BaselineActivity

Varying key inputs

Stage 3:LevelisedCost

Stage 4:Evaluation


22/156


Executive Summary

ILLUSTRATIVE COUNTRY CASE STUDIES

In order to demonstrate how the ramework can be used in practice, the report describes a simpli iedmodelling exercise to promote large-scale, onshore wind energy in our selected countries: Kenya,Mongolia, Panama and South A rica.

As set out in Figure 6, the our countries represent a range o renewable energy market conditions, refectingdi erent investment environments and baseline electricity generation costs. For example, South A ricahas a high sovereign rating investment environment, combined with relatively low-cost electricity (where

the baseline energy mix is dominated by inexpensive coal). In contrast, Kenya has a low sovereign ratinginvestment environment, combined with relatively high-cost electricity (where the baseline energy mixhas a high share o expensive uel-oil-based generation).

Onshore wind energy is chosen as it represents a mature renewable energy technology with a strongtrack record and good data availability. All our countries have strong, untapped wind resources andalready have guaranteed price and market-access cornerstone instruments or wind energy in place.Kenya and Mongolia have implemented FiTs, while Panama and South A rica have deployed PPA-basedbidding.

The modell ing exercise assumes a long-term, 20-year national target or wind investment in each o theour countries: 8.4 GW in South A rica, and 1 GW each in Kenya, Mongolia and Panama. In South A rica,the Governments announced 2030 target has been used. In the other three countries, the long-term20-year targets are the exercises own assumptions. The objective was to create an ambitious vision orwind energy in each country but, at the same time, to cap wind energys share o the anticipated uturegeneration mix at a level whereby intermittency issues could be managed.

The two-by-two instrument matrix illustrated in Figure 6 above provides an organising basis with which toselect a plausible set o policy and nancial derisking instruments to complement the existing cornerstoneinstrument in each country.

Financial derisking or wind energy, a relatively mature renewable energy technology, is assumed notto be required in countries with high sovereign ratings (South A rica, Panama). Financial deriskinginstruments are assumed to be a requirement in countries with low sovereign ratings (Mongolia,Kenya).

A direct nancial incentive, in the orm o a price premium in the tari , is modelled when the LCOE o wind energy is higher than the baseline electricity generation costs (in South A rica, Mongolia). No pricepremium is assumed necessary in cases where wind power is less expensive than the baseline generationcosts (in Panama, Kenya).

The use o the ramework requires the collection o a large amount o data and many assumptions. Over30 investors and other wind energy stakeholders in the our countries were interviewed or the modellingexercise. However, in order to keep the overall exercise manageable, several modelling simpli cations havebeen adopted. Many input parameters, such as wind technology costs, have been standardised across all ourcountries. Actual costs might di er considerably rom country to country and project to project. A number

The modellingexercise's our

countries representa range o marketconditions, refectingdi erent investment

environments andbaseline electricitygeneration costs.


23/156


Executive Summary

o key assumptions on the scope o the modelling exercise have also been made regarding: intermittency,where balancing costs are not actored into the study; the cost o the transmission grid, which is e ectivelyexcluded rom the analysis; and, with regard to the cost o ossil uels, unsubsidised uel costs have been usedin order to remove the distortive e ects o subsidies and to allow or comparison between the our countries.

None o the above simpli ying assumptions undermines the integrity o the modelling exercise. However,should policymakers wish to use the ramework or detailed policy analysis, additional in-depth countryconsultations would be required to collect empirical data to ne-tune the input parameters and modellingassumptions. The ramework allows or the degree o complexity used to be tailored on a case-by-case basis.

A presentation o the results o the our country case studies is given in Chapter 3. The ull data-sets andassumptions or the modelling exercise are set out in Annex A. As an illustration, Figure 7 shows some o the key ramework outputs or Kenya and the case studys 1 GW 20-year target or wind energy investment. The gure shows outputs or the modelling exercises business-as-usual (BAU) scenario, where Kenyas FiT iscomplemented only by nancial derisking instruments, and or a post-derisking scenario, where Kenyas FiTis complemented by both policy and nancial derisking instruments.

* For the modelling exercise, the investment environment is classi ed using sovereign ratings rom credit rating agencies as a generalindicator. High refects a sovereign rating o BBB- or above (commonly re erred to as investment-grade); low refects a sovereignrating below BBB- (non-investment grade)

Figure 6: The our countr case studies and their illustrative combinations o public instruments

General investment environment *

B a s e

l i n e e n e r g y g e n e r a

t i o n

c o s

t

L O W

C O S T

B A S E L I N E

H I G H C O S T

B A S E L I N E

HIGH SOVEREIGN RATING LOW SOVEREIGN RATING

South A rica 8.4 GW

Panama 1 GW

Mongolia 1 GW

Ken a 1 GW

Cornerstone instrument: PPA bidding

Cornerstone instrument: PPA bidding

Cornerstone instrument: FiT

Cornerstone instrument: FiT

Policy derisking instruments


Direct nancial incentive: premium

Direct nancial incentive: premium

Financial derisking instruments


+

+

+

+

++




24/156


25/156


Executive Summary

Stage 4: Evaluation4 Per ormance Metrics

x 8.0*

M I L L I O N U S D

Wind EnergyInvestments

1,980

Cost of Post-Derisking

Instruments

131

112

Cost of BAUInstruments

248

x 15.1

19

Policy derisking instrumentsPresent value of costs over 20 years


Marginal baseline LCOE (unsubsidised)

-6%

U S D c e n t / k W h

17.11618

141210

86420

-53%

LCOEPost-Derisking

LCOE BAU

8.7 8.1

M I L L I O N U S D

x 14.2

19

( P r e s e n t v a

l u e o v e r

2 0

y r s

)

4,516

275

4,241

IncrementalCosts

Post-Derisking

SavingsIncrementalCosts-BAU

Cost of Policy

Derisking

-6%

U S D / t C O

2 e

0

-20

-40

-60

-80

-100

-120

-140

-120.9-113.5

BAU Post-Derisking

37.4 Mt CO 2e (20 yrs)

INVESTMENT LEVERAGE R ATIO(Metric 1)

END-USER AFFORDABILITY(Metric 3)

SAVINGS LEVERAGE RATIO(Metric 2)

CARBON ABATEMENT(Metric 4)

Source: interviews with wind investors and developers; modelling exercise; see Table 5, Table 17 (Chapter 3) and Annex A or details onassumptions and methodology.For Stage 1: the cost o debt and equity assume supporting nancial derisking instruments are in place. The cost o debt shown is thecommercial rate assuming nancial derisking is in place.For Stage 2: the impacts shown are average impacts over the 20-year modelling period, assuming linear timing e ects.

* In the BAU scenario the ull investment target may not be met.


26/156


Executive Summary

IMPLICATIONS FOR PUBLIC FINANCE TO PROMOTERENEWABLE ENERGY AT SCALE

A number o practical ndings emerge rom a comparative analysis o the illustrative results across the ourcase study countries. While a more detailed modelling exercise may substantially re ne the gures obtained,it is likely that the overall implications will stay the same.

In analysing the results, each o the rameworks our per ormance metrics provides a di erent perspectiveor the policymaker seeking to promote renewable energy at scale.

Public Finance E ectiveness (Metric 1: Investment Leverage Ratio)A undamental goal o the policymaker is to catalyse concrete private sector investment. A inding o this illustrative modelling exercise is that the presence o a cornerstone instrument, such as a FiT orPPA bidding process, by itsel does not guarantee this investment. Instead, the results show that thereis a role or complementary policy and inancial derisking measures to target the residual risks that acornerstone instrument alone cannot address and that can otherwise suppress investment.

This point is particularly well illustrated by the case study o Panama. Despite the country having a PPAbidding process in place, an attractive investment climate and low wind power generation costs whencompared to an existing high-cost baseline, nancial closure with banks or the rst wind licences awardedhas yet to occur. The nancing cost water all or Panama clearly shows that a number o non-price barriersremain and that additional derisking e orts are required to complement the existing PPA bidding process. The modelling exercise shows that the impact o such additional derisking e orts could be dramatic. WithPanamas unique combination o avourable actors, a relatively small amount o policy derisking couldcatalyse 100 times its cost in private investment.

More broadly, these ndings illustrate that renewable energy market trans ormation takes time. Despite theact that a FiT or PPA bidding process has been in place in the our case study countries or several years, itmay not be immediately e ective. Barriers to renewable energy investment are o ten deeply embedded,refecting long-held practices centred on ossil- uels and monopolistic market structures. A cornerstoneinstrument, such as a FiT, complemented by policy and nancial derisking, can there ore be seen as thestarting point on a longer path to trans orming a market or renewable energy investment.

Policy andnancial derisking

instruments targetthe residual risks

that a FiT alonecannot addressand which can

otherwise suppressinvestment.


27/156


Executive Summary

Public Finance Efciency (Metric 2: Savings Leverage Ratio)A second key nding rom the modelling exercise is that derisking measures can generate signi cant publicsavings in all our countries. In low-cost baseline countries, such as South A rica and Mongolia, deriskinginstruments reduce the price premium required to make renewable energy competitive with conventionalenergy. In South A rica, with a large 8.4 GW wind target, the modelling nds that an estimated USD 40 millionin policy derisking instruments can result in a USD 2.3 bi llion reduction in the price premium over the 20-yeartarget, a savings leverage ratio o over 50.

Less intuitive but just as critical, derisking instruments can unlock the savings associated with the lower cost

o renewable energy in high-cost baseline countries, such as Panama and Kenya. For example, a modestinvestment in additional policy derisking instruments in Kenya, estimated at about USD 20 million in thissimpli ed modelling exercise, could unlock USD 4.5 billion in negative incremental costs over the next 20years as compared to an unsubsidised baseline.

For the two low-cost baseline countries (South A rica and Mongolia), wind energy remains more expensivethan the baseline even a ter derisking, and this can result in a net cost to taxpayers or electricity consumers.In such cases, an implication o the modelling exercise is that the ambition o a countrys long-term visionor wind energy can be an important actor. Although local content requirements have proven controversial,in South A rica, or example, the ambitious 8.4 GW target or wind energy could provide a solid oundationor the local manu acturing sector. The experience o countries, such as China and India shows that localmanu acturing can greatly reduce the total installed cost o wind energy (IRENA, 2012b) and generate oreigndirect investment (FDI) and green jobs. In Mongolia, with the modelling assumptions (1 GW, domesticlow cost baseline) employed, the economic case in avour o public nancial incentives or wind energy isquestionable. However, a more ambitious, export-oriented vision or wind in Mongolia, partneringwith neighbouring countries with higher baseline costs, could dramatically alter the cost equation andcompetitiveness o Mongolian wind energy.

Distributional Impact o Public Interventions (Metric 3: End-user A ordability)Ultimately the generation costs o renewable energy, as well as those o any associated public measures,will be met by the end-user (industry, households) and/or the taxpayer. The results o the modelling exerciseshow that, i passed on to the consumer, the use o derisking instruments to complement a FiT or PPA biddingprocess has the potential to increase a ordability o the renewable energy technologies in all our countries.

E orts to promote renewable energy are commonly blamed or causing high energy costs in countries thathave adopted ambitious clean energy targets. However, contrary to this widespread belie , the modellingexercise indicates that well-designed and implemented public measures can o er tangible bene ts in theorm o reduced household energy bills in countries with high baseline power costs. In Kenya, the LCOE o wind energy a ter derisking (USD 8.1 cents per kWh) is a ull 53 percent lower than the unsubsidised baselinecost (USD 17.1 cents per kWh), creating potentially very large bene ts or low-income ratepayers.

Derisking measurreduce the pricepremium in low-cobaseline countries.and unlock savingsin the high-cost

baseline countries.

Well-designedand implementedpublic measurescan o er tangiblebene ts in theorm o reducedhouseholdenergy bills.


28/156


Executive Summary

The result s o the case studies, where ossi l uel subsidies have expressly been excluded, show thatrenewable energy is competitive against unsubsidised ossil uel technologies in many developing countries.Globally, subsidies to ossil uels are at least ve times larger than nancial incentives to renewable energytechnologies (IEA, 2012) and distort the true competiveness o renewable energy (Schmidt et al ., 2012).In low-baseline countries, the most cost-e ective means o reducing direct nancial incentives or renewableenergy is to phase-out or phase-down ossil uel subsidies. In high-baseline countries, ossil uel subsidiesthat are intended to help the consumer may have the perverse e ect o preventing investment in ar morea ordable renewable energy alternatives.

Scaling-Up Climate Change Mitigation Outcomes (Metric 4: Carbon Abatement) The modelling exercise presented in this publication shows that derisking renewable energy investment canlower the abatement costs o CO 2 emissions in the our countries investigated. For example, in South A rica,meeting the 8.4 GW wind energy target over 20 years could result in emission reductions amounting to 604million tonnes o CO 2, with derisking measures reducing the cost o abatement rom USD 12 to USD 8.20 pertonne o CO 2.

The importance o derisking in reducing abatement costs is applicable to every developing country that haslisted climate mitigation pledges under the Copenhagen Accord. It also has signi cant implications or thedesign o modalities and mechanisms to scale-up climate mitigation e orts, such as Nationally AppropriateMitigation Actions (NAMAs), a re ormed Clean Development Mechanism (CDM), New Market Mechanisms(NMMs), and public payments rom vertical unds, such as the Global Environment Facility (GEF) and GreenClimate Fund (GCF). A number o these international mechanisms envisage per ormance-based paymentsor emission reductions. The results o the case studies suggest the desirability o incorporating up rontgrant-based derisking activities to complement per ormance-based payments in such mechanisms, therebyreducing the overall carbon abatement cost. Figure 8 below summarises this per ormance-based paymentapproach.

Figure 8: Scaled up mitigation actions blending derisking instruments and per ormance basedpa ments

Polic & Financial

Derisking

Public Support orScaled up Mitigation

Actions

Per ormance based

Pa ments

+=

Carbon Marketsor Public Finance

Public Finance

Complementingper ormance-

based paymentmechanisms with

derisking activitiescan reduce theoverall carbon

abatement cost.


29/156


Executive Summary

CONCLUSION There are many di erent ways to create markets or renewable energy, and the path each country takes willdepend on its speci c national context, goals and resource endowments.

A central conclusion o the report is that it is important or policymakers to address the risks to renewable energyinvestment in a systemic and integrated manner. In all our case study countries, the rameworks nancingcost water alls clearly demonstrate that a range o risks exist in the investment environment. Any isolated,short-term e ort ocusing on a subset o risks and relying on a subset o inst ruments is unlikely to sustainab lytrans orm renewable energy markets.

A complementary conclusion is that investing in derisking measures, bringing down the nancing costs o renewable energy, appears to be cost-e ective when measured against paying direct nancial incentives tocompensate investors or higher risks. Instead o using scarce public unds to pay higher electricity tari s,it can be advantageous to irst reduce and manage typical renewable energy risks ( or example, thoseassociated with power markets, permits, and transmission), and thereby change the undamental risk rewardtrade-o that energy investors ace in a given country.

The ramework introduced in this report can help to estimate the costs o derisking instruments and theamount o up ront grant required. It can also help to assess the direct nancial incentives required to meetthe derisked incremental costs o renewable energy and calibrate a per ormance-based payment schemeaccordingly.

However, it is important to be realistic about the di culties associated with modelling derisked incrementalcosts in the absence o what is o ten scarce historical empirical data and when con ronting long-rununcertainties, such as those relating to technological evolution. The sensitivity analysis conducted or theour country case studies shows that relatively small changes in key model input parameters can result inmajor variations. The ramework can support, but not substitute or, in-depth policy decision-making andconsultation processes involving all key stakeholders.

Any isolated,short-term e ortocusing on asubset o risks

and relyingon a subset o instruments isunlikely tosustainablytrans ormrenewableenergy markets.


30/156


Introduction

Around the world, developing countries are seeking to rapidly scale-up renewable energy investment. Thisshi t to renewable energy is driven by a number o considerations. Many developing countries are acingast-growing energy demand coupled with increasing exposure to rising ossil uel imports and prices. About1.3 billion people still lack access to electricity and 2.7 billion to modern energy services (UN, 2011), withtheir human development held back through energy poverty. At the same time, rising global uel pricesand resource scarcities are making developing countries increasingly vulnerable to oil prices. Nine o the 25low-income countries or which data are available already pay more than 10 percent o their gross domesticproduct (GDP) as an average over 2006-2010, to simply secure their oil supply (Economy Watch, 2011; Seth,2012). This number is expected to increase in the uture, given the transition rom traditional energy sources( or example, uel, wood and waste) to modern energy systems. On the other hand, costs o energy e cientand renewable energy technologies are alling and renewable energy is o ering increasingly attractiveinvestment opportunities.

The nancial sums involved in a rapid shi t to low-emission energy systems are enormous. UN-DESA, orexample, has estimated that it would cost up to USD 250270 billion per year to shi t developing countriesto 20 percent renewable energy by 2025 (DeMartino & Le Blanc, 2010). According to the Global EnergyAssessment (GEA, 2012), global investment in energy e ciency and low-carbon energy generation will needto increase to between USD 1.72.2 trillion per year compared to present levels o about USD 1.3 trillionper year over the coming decades to meet the combined challenges o energy access, energy security andclimate change. I developing countries are going to success ully scale-up their use o renewable energy, it isclear that private sector investment must be at the ore ront.

Global capital markets, representing some USD 212 trillion in nancial assets (McKinsey, 2011), includingUSD 71 trillion managed by institutional investors (CPI, 2013), in principle have the size and depth to stepup to the investment challenge. 6 The existence o signi cant potential or low-carbon investments, withmany options already available and cost-e ective, should make a compelling case or businesses, privateinvestors and households to independently adopt mitigation and adaptation technologies. Nonetheless,investment in seemingly straight orward renewable energy technologies aces a range o in ormational,technical, institutional and nancial barriers. As a result, global investment in renewable energy su ersrom severe regional imbalances. Figure 9 compares investments in Organisation or Economic Cooperationand Development (OECD) countries, the BASIC countries (Brazil, South A rica, India and China) and others.Outside the BASIC countries, developing countries have consistently accounted or less than 10 percent o investment in clean energy over each o the last nine years (BNEF, 2013). In order to sustain and acceleraterenewable market growth across all developing economies, signi cant public nancial resources romboth national and international actors will be required to establish an enabling investment environmentto attract capital at scale.

The challenge o addressing these barriers to meet the increasing demand or clean energy has inspired thedevelopment o a wide array o policy and nancing instruments to shi t investments rom ossil uels to moreclimate- riendly alternatives. However, public policies to catalyse clean energy nance come at a cost or

Introduction

6 Many renewable energy investment opportunities should appeal to institutional investors (pension unds, insurance companies, etc.) seekingattractive, low-risk, long-term investment per ormance. Yet general in rastructure investment only accounts or around 1 percent o the assetallocation o the average pension und, and speci ically green in rastructure accounts or around 3 percent o that (BNEF, 2013). Institutionalinvestors ace a series o constraints in investing in renewable energy projects, including limitations on investing in illiquid assets, transaction coststo maintain a direct investment capability and sector diversi ication requirements. However, a recent study rom CPI (2013) estimates that withpolicy and investment practices properly aligned, pension unds and insurance companies could supply about a quarter o the equity and hal o the debt renewable energy projects and all o the corporate equity and debt that would eed into renewable energy to oster a low-carbon society.

Outside theBASIC countries,developing countries

have consistentlyaccounted or less

than 10 percento investment in

clean energy.


31/156


2004 2005 2006 2007 2008 2009 2010 2011 2012

$54bn

$80bn

$114bn

$164bn

$191bn

$187bn

$251bn

$302bn

$269bn

9%

35%

56%

6%

27%

67%

6%

26%

68%

7%

27%

66%

7%

25%

68%

6%

22%

73%

5%

7%

6%

20%

74%

17%

76%

14%

84%

OECD Countries BASIC Countries Other Developing Countries

Introduction

industry, consumers or taxpayers. As such, and with an array o other development priorities competing orscarce public resources, the challenge or policymakers is to identi y the most cost-e ective policy port oliotailored to the particular national energy context. A key constraint or policymakers to meet this challengeis that they currently lack means o quantitatively comparing di erent public instruments and their impacts.As a contribution to address this gap, this publication presents an innovative ramework developed by UNDPto support decision-making or renewable energy investment.

The paper is structured in our chapters. Chapter 1 spells out the theory o change underpinning UNDPsapproach to derisking renewable energy investment. In Chapter 2, the overall structure and the individualsteps o the ramework are described. To illustrate its potential and limits, Chapter 3 applies the ramework tolarge-scale, onshore wind energy development in our illustrative countries: Kenya, Mongolia, Panama andSouth A rica. Based on the ndings o the our-country study, Chapter 4 discusses the implications o theramework or public interventions to promote renewable energy market trans ormation.

Figure 9: Investments in clean energ b t pe o countries USD billions

Source: Bloomberg New Energy Finance (2013)


32/156

1.1 High Financing Costs or Renewable Energy

1.2 The Role o Public Instruments in Reducing Financing Costs

1.3 Challenges to Identi ying an Appropriate Public Instrument Mix

The Role o Public Instruments in Reducing Financing Costs or Renewable Energy inDeveloping CountriesChapter 1


33/156


1Historically, some o the most signi cant barriers to renewable energy development have been technical.

Today, the lack o long-term operational experience, technical standards and quality control that plaguedrenewable energy technologies in the past has mostly been addressed. Renewable energy technologieshave also shown considerable potential or price reduction.

However, the large up ront investment costs inherent to renewable energy remain a major impediment tothe development o the sector in developing countries. This chapter describes the impact o high nancingcosts in developing countries on the nancial viability o renewable energy investment. It then discusses howpublic instruments can improve the risk-reward pro le o renewable energy investments, either by reducingrisks or by providing a nancial incentive, and thereby attracting private sector capital. It concludes with adiscussion on the challenges in identi ying an appropriate public instrument mix to cost-e ectively addressinvestment barriers and associated risks.

1.1 HIGH FINANCING COSTS FOR RENEWABLE ENERGYINVESTMENT

When analysing how to promote investment in renewable energy, a use ul metric is the levelised cost o electricity (LCOE). While by no means a per ect metric, 7 the use o the LCOE does allow or a like- or-likecomparison o the li e-cycle generation costs o di erent technologies. In this way, it provides a measureo a renewable energy technologys competitiveness and can assist in determining the need or publicly-unded nancial incentives. The LCOE is commonly used by policymakers, energy planners and researchers tosupport decision-making and ormulate energy policy.

Figure 10 sets out the core drivers o the LCOE calculation. More detail on the LCOE ormula, as usedin the rameworks inancial tool, can be ound in Annex A. In simpli ied terms, the LCOE takes

the li e-cycle costs o an energy project and divides these costs by the projects electricity generation over itsli etime, to give a cost per unit o electricity generated, or example, in USD cents per kWh. An LCOE calculationor any given energy technology takes into account both technology costs, made up o investment costsand O&M (including uel costs), and nancing costs, made up o the cost o equity and debt.

The Role o Public Instruments

in Reducing Financing Costs orRenewable Energy in DevelopingCountries

Chapter 1: The Role o Public Instruments in Reducing Financing Costs or Renewable Energy in Developing Countries

7 Comparisons based on LCOE do not, or example, actor in the costs o intermittency balancing and the di erent value o peak/o peak generationcosts (Joskow, 2011) or port olio and merit-order e ects with renewable energy. However, or the purposes o eed-in tari design and in the lessmature power markets o ten ound in developing countries, LCOE is well suited (Schmidt et al ., 2012). In addition, it should be noted that LCOEocuses on generation costs, and does not capture the macro-economic bene ts o uel price certainty, greenhouse gas abatement, greenemployment, energy security and other aspects (Kammen and Pacca, 2004).


34/156


Technology costs or new renewable energy technologies have experienced a remarkably steady decreasein investment costs over the past decades (IRENA, 2012a, IPCC, 2011; Peters et al., 2011). In the case o solar PV energy, module costs have allen rom about USD 40 per Watt in 1979 to USD 0.90 in 2012 (IRENA,2012a), realising a near 98 percent cost reduction over the period.

Given this rapid all in renewable technology costs, it has been suggested that a sustained technologypush by a ew initial leader countries could prove enough to urther reduce the generation cost o renewableenergy and enable renewable energy to out-compete ossil uels by the end o this decade (Lilliestamet al ., 2012). However, this is unlikely to be the case in most developing countries in the near uture. Thebarriers towards a ull-scale transition to renewable energy lie not just in technology costs but also in the

challenges associated with securing long-term a ordable nance.

While the technology costs o many renewable energy technologies are rapidly declining, they continueto have high up ront investment costs compared with ossil uel energy projects. Figure 11 illustrates thedi erent cost pro les o electricity generation rom wind energy and combined-cycle gas plants. Investmentcosts account or approximately 80% o the total li etime technology costs or wind energy but only accountor around 15% in the case o gas. Annual operating costs are relatively low or wind energy, but, due to theimpact o uel costs, predominate in the case o gas. In essence, renewable energy investments exchangelong-term uel costs or up ront investment costs.


USD/kWh

Levelised Cost of Electricity (LCOE)

Cost of Equity

Financing Costs

Technology Costs

Operational Costs

Investment Costs/Depreciation

Cost of Debt

Figure 10: The core drivers o the LCOE


35/156



As a consequence, project developers in developing countries o ten struggle to access the large quantitieso up ront nancing needed or renewable energy investments. When available, the cost o nancing is o tensubstantially higher than in developed countries. Given their up ront capital intensity, renewable energyprojects are particularly sensitive to high costs o debt and equity. Over the long li etimes o in rastructureinvestments, the impact o higher nancing costs on renewable energys competitiveness can be quitesubstantial.

Furthermore, banks in some developing countries provide credit only on a short-term basis in order tomanage lending risks, posing an additional challenge to renewable energy projects that require long-termnancing to be cost-competitive. So, or example, or onshore wind in developed countries, wind projects

seeking nancing have been able to obtain commercial loans with tenors o up to 18 years (Tan, 2012), closeto the anticipated li etime o the underlying equipment. However, commercial loans or renewable energyin developing countries, where available, typically have a loan tenor equating to around only hal or less o the projects anticipated li etime. Moreover, lenders tend to demand a higher contribution o equity romproject developers when there is higher perceived investment risk. As the cost o debt is lower than the costo equity, this increased share o equity can have a signi cant impact on overall nancing costs. 8

$2.0 m

$1.5 m

$1.0 m

$0.5 m

$0.0 m

Wind (onshore)

$2.0 m

$1.5 m

$1.0 m

$0.5 m

$0.0 m

YEAR

YEAR

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Gas (combined cycle)

Investment Costs Operating Cost (including Fuel)

C O S T S I N U S D / M W

I N S T A L L E D

Figure 11: The di erent capital intensit o electricit production rom wind energ and combined c cle gas

For technology assumptions, see inputs or wind energy and gas (CCGT) in Section A.3 (Annex A).

8 Upcoming Basel III regulations in the nancial sector could unintentionally urther limit the ability o banks to provide long-term, non-recourseproject nance (BNEF, 2013). Under Basel III, renewable energy projects cannot be counted as liquid and will negatively a ect the liquiditycoverage ratio o banks.

Given theirup ront capitalintensity, renewablenergy projectsare particularlysensitive to highnancing costs.


36/156


As can be seen on the le t-hand side o Figure 12, the developed country LCOE or wind energy, atUSD 6.7 cents per kWh, is close to being competitive with the gas LCOE, at USD 6.1 cents per kWh. This gridparity to gas is ound in many developed countries today (Ernst & Young, 2012). However, on the right-handside, the developing country LCOE or wind jumps to USD 9.4 cents per kWh, a 40 percent increase over thedeveloped country. The LCOE or gas also increases to USD 6.5 cents per kWh, but in this case represents onlya 6 percent increase over the developed country cost given its lower up ront capital requirements. The overallnet e ect is a clear loss in competitiveness or wind energy (and renewable energy in general) in developingcountries, which can be ully attributed to its capital intensity combined with high nancing costs.

Given their high capital intensity, the di erences in nancing costs and terms can dramatically a ect thecompetitiveness o renewable energy investments versus ossil uel-based technologies in developingcountries. Figure 12 compares the 2012 LCOE o identical onshore wind energy and combined cycle gasinvestments in illustrative developed and developing countries. The LCOEs are calculated using the sameinputs or technology costs (investment costs, O&M costs, uel costs) as well as electricity generated ( ull-loadhours), and only varying nancing costs between the two types o countries. In the illustrative developedcountry, the cost o debt is modelled at 5 percent and the cost o equity at 10 percent; in the illustrativedeveloping country, the cost o debt increases to 10 percent and the cost o equity to 18 percent.


Developed country(Wind vs. Gas)

Developed CountryCost of Equity = 10%Cost of Debt = 5%

Developing CountryCost of Equity = 18%Cost of Debt = 10%

Developing country(Wind vs. Gas)

P R E T

A X L C O E

U S D C E N T S / k W h

Gas(CCGT)

Wind(onshore)

Gas(CCGT)

Wind(onshore)

Financing Cost(Equity)

Financing Cost(Debt)

Operating Cost(incl. fuel cost)

Investment Cost/Depreciation

2.9

1.1

0.9

1.8

6.7

+ 40%

0.7

4.5

0.30.6

6.1

2.9

0.8

2.4

3.3

9.4

0.7

3.9

0.9

1.0

6.5+ 6%

Figure 12: Impact o nancing costs on wind and gas power generation costs in developed anddeveloping countries

For technology assumptions, see inputs or wind energy and gas (CCGT) in Section A.3 (Annex A); a 70%/30% debt/equity capital

structure is assumed; nancing costs are based on data in the our country case study (Chapter 3), assuming a non-investment gradedeveloping country.Operating costs appear as a lower contribution to LCOE in developing countries due to discounting e ects rom higher nancing costs.


37/156



1.2 THE ROLE OF PUBLIC INSTRUMENTS IN REDUCINGFINANCING COSTS

The higher nancing costs or renewable energy in developing countries refect a number o perceived oractual risks to investment (Glemarec, 2011; Glemarec et al ., 2012). Some risks may be speci c to renewableenergy. For example, there may be shortcomings in new grid codes or renewable energy, or concernsregarding the ability o the grid to manage intermittent renewable energy sources. Other risks mayrefect broader concerns regarding the investment environment in the particular country, such as poormacro-economic per ormance or political instability. Taken together, these higher nancing costs can beunderstood as the extra reward required by investors to compensate them or the extra risks that comewith capital-intensive, long-term renewable energy investments in developing economies.

Investors adjust their required risk/return pro les to take into account risks in the investment environment.As risks can result in negative nancial impacts or investors, investors require a higher return to compensateor the possibility o this impact (Glemarec et al ., 2012; McKinsey, 2012; DB Climate Change Advisors, 2011).

The degree to which investors accurately price barriers and risks into their nancial return requirementsdepends in practice on the particular type o investment being made (Box 1).

Box 1: How di erent investment t pes a ect the pricing o risk into nancing costs

A number o di erent actors can a ect whether risks are priced into investors nancing costs. These include: Corporate nance vs. project structures. I a corporate nance structure is taken, a bank will typically lend on the strengtho the business balance sheet, and the cost o debt is less likely to price in the speci c risks aced by the renewable energyinvestment. I a project nance structure is taken, the only collateral a bank can have recourse to is the underlying assets o the investment. In this case, a bank is likely to per orm detailed due diligence on the investment itsel , and in turn will pricein the various risks aced by the investment.

Core vs. non core investments. I the investment is a non-core activity, the risks associated with the investment are lesslikely to a ect nancing costs. A common example o non-core activities in this context are energy e cient investments. Asan illustration, an upgrade o an ine cient industrial boiler in a textile actory may simply use a cost o nancing associatedwith the actorys core activity (textile manu acturing). Here, the particular risks o the energy e ciency upgrade will haveminimal a ect on the nancing costs used. On the other hand, i an investment is a core activity, the cost o nancing used

is more likely to incorporate the associated risks. An example is an energy e cient boiler upgrade in a textile actory wherethe investment is now ully outsourced to an energy service company (ESCO). The energy e ciency investment undertakenby the ESCO is now in line with the core activities o the ESCO, and as such the nancing costs used are more likely toincorporate the particular risks o the energy e cient investment.

Unsophisticated vs. sophisticated investors. Generally, the more experienced and sophisticated the investor, the morecapable the investor is o pricing in the particular risks. So, or example, an inexperienced IPP investing its own equitymay underestimate the uncertainty associated with certain risks, resulting in a low cost o equity and overly aggressivebidding in a PPA-based bidding process. Similarly, commercial banks in an immature domestic nancial sector may notbe ully com ortable with investments related to renewable energy technologies, resulting in a conservatively pricedhigh cost o debt.


38/156


Investment by the private sector in large-scale, renewable energy is typically per ormed with non-recourse project nance structures. Commercial bank lending and equity investing tends to accurately priceinvestment risks into the nancing costs. As a result, a country needs to provide very high return rates toattract private investments or wind power development i IPPs ace barriers, such as limited access to grids,lengthy and uncertain processes to issue permits, limited local supply o expertise or a lack o long-termprice guarantees. 9 The key challenge to scaling-up renewable energy technologies in developing countriesis to lower the nancing costs that a ect these technologies competitiveness with ossil uel energy. Ashigher nancing costs refect risks in the investment environment, the entry point or policymakers to osterrenewable energy technologies in developing countries is to address these investment risks.

In order to meet this challenge, policymakers in developing countries have been exploring a broad spectrumo public instruments (Glemarec, 2011). The common objective o these instruments is to create conditionsor attractive investment risk/reward pro les, adapted to di erent types o investors, either through reducingrisks (and hence lowering the weighted average cost o capital demanded or these investments) or increasingrewards (through premium prices, tax credits, etc.).

Figure 13 provides a conceptual illustration o the approach. The gure illustrates a shi t rom a commerciallyunattractive investment opportunity (right) to a commercially attractive one (top). This is achieved throughtwo actions: rst, by reducing the risk o the activity (derisking), or example through a regulatory policy,such as guaranteed access to the grid or IPPs; and, second, by increasing the return on investment throughnancial incentives, such as a price premium or renewable energy.


Example:price premium

Example: guaranteed access to the grid

F I N A N C I A L R E T U R N

RISK OF INVESTMENT

Infeasible

renewableenergy project

Feasiblerenewable

energy project

Figure 13: Shi ting the risk reward pro ile o renewable e nerg investment

9 These risks and their impact on renewable energy nancing are discussed in greater detail in Chapter 2 o this report.



39/156



Public derisking measures can broadly be divided into two groups: policy derisking instruments and nancialderisking instruments.

Polic derisking instruments address and attempt to remove the underlying barriers that are the rootcauses o risks. These instruments utilise policy and programmatic interventions to mitigate risk. Forexample, renewable energy projects typically involve obtaining a number o permits and approvals,including generation licences, environmental impact assessments (EIAs) and land rights. Unclearand overlapping institutional responsibilities related to renewable energy permitting, or lack o sta experience with renewable energy, can increase transaction costs, delay revenues and discourageinvestment. A policy derisking approach might involve streamlining the permitting process, clari ying and

standardizing institutional responsibilities, reducing the number o process steps and providing capacitybuilding to programme administrators.

Financial derisking instruments do not seek to directly address the underlying barrier but, instead,unction by trans erring the risks that investors ace to public actors, such as development banks. Theseinstruments can include development banks loans and guarantees, political risk insurance and publicequity co-investments. Financial derisking instruments can also indirectly address certain underlying barriersthrough learning-by-doing and track-record e ects. For example, in countries with immature and under-capitalised nancial sectors, local banks may be concerned about lending their limited capital to borrowers inan unproven sector such as renewable energy. Partial loan guarantees rom a development bank can providethese local banks with the security they need to issue loans, whereby a portion o the risk o de ault is trans erredto a public actor. In this way, nancial derisking instruments can kick-start the local nancial sectors

involvement in renewable energy.

Recognising that all risks cannot be eliminated through policy derisking or trans erred through nancialderisking, e orts to reduce risks might need to be complemented by a third group o public instruments,direct inancial incentives , to compensate or any residual risks and costs. These incentives can take anumber o di erent orms including price premiums, tax breaks, such as production tax credits, and proceedsrom carbon o sets.


40/156


1.3 CHALLENGES TO IDENTIFYING AN APPROPRIATE PUBLINSTRUMENT MIX

To promote renewable energy at scale, policymakers typically need to proceed in two steps:

Select an appropriate combination o public instruments centred around a cornerstone instrument.

Assess the cost e ectiveness o these di erent public instruments throughout the market trans ormationprocess.

i. Selecting an appropriate combination o public instruments

Identi ying an appropriate combination o public instruments can prove very challenging in practice. The severity o investment barriers to renewable energy varies across locat ions and technologies.Selecting rom the many hundreds o di erent public instruments (Glemarec, 2011) can be overwhelming.A public instrument can prove highly e ective in one country context and ace major compliance issuesin another. Di erent resource endowments, market conditions and national goals mean that there is noone-size- its-all best public instrument mix.

Decision-makers tasked with selecting an optimal mix o public instruments will irst need to surveyall the investor risks to renewable energy development in the country. They will have to identi y howunderlying barriers a ect di erent stakeholders and translate into investment risks. They will then haveto identi y a mix o public instruments that e ectively target all major investment risks.

While policymakers can use a range o di erent instruments to address renewable energy investmentrisks and their underlying barriers, certain types o instruments have achieved greater prominencethan others and are o ten re erred to as cornerstone instruments. A cornerstone instrument targets keyinvestment risks and is the oundation upon which all complementary policy and inancial deriskinginstruments are built.

Mechanisms that provide renewable energy generators with a PPA, ensuring a xed long-term price orpower and guaranteed access to the electricity grid, are o ten the cornerstone instrument or renewableenergy market trans ormation e orts. Such cornerstone instruments o ten re erred to as eed-intari s (FiTs), but can also be designed around auctions or bidding processes. 10 When necessary,

FiTs can also include an above-market price, in the orm o a premium, in order to increase thereturn on investment. Thus, FiTs are both a policy derisking instrument (market access to the grid andmust-take requirements) and a nancial derisking instrument (guaranteed price over a period o 15-25 years) that can also ac t, when needed, as a nancial incentive instrument (through a price premium),shi ting the entire risk-reward pro le o a renewable energy investment. FiTs have become the most widelyused mandated price and market instrument in the renewable electricity sector, having been adopted byat least 65 countries and 27 states/provinces as o 2012 (REN21, 2012).


10 Recognising that there are ew clear dividing lines between FiTs and PPA-based auctions/bidding processes both result in project developersentering into long-term PPAs at a xed price this report uses the term FiTs at times to cover both types o cornerstone instrument. For acomparative discussion on FiT and auctions/bidding in non-OECD countries see Becker and Fischer (2013).

A cornerstoneinstrument targets

key investmentrisks and is the

oundation uponwhich all

complementarypolicy and nancial

deriskinginstruments

are built.


41/156



FiTs are popular with developers and investors because they can mitigate the speci c risks associated withthe nancial pro le o renewable energy projects (von Flotow & Friebe, 2011; Brer & Wstenhagen, 2009).As discussed above, approximately 80 percent o the li etime total technology cost o wind energy is relatedto up ront costs o the wind turbine, oundations and grid connection. By establishing a secure uturerevenue stream, FiTs minimise the risk associated with long-term, xed-cost investments. As renewableenergy generation is not exposed to variations in uture ossil uel prices, a FiT can thus dramaticallyimprove the relative nancial attractiveness o a renewable energy investment versus its conventionalenergy alternative.

Figure 14 below sets out the key components o a public instrument port oli

undp derisking renewable energy investment

Documents