92224 DIREC TIONS IN DE VELOPMENT Energy and Mining The Design and Sustainability of Renewable Energy Incentives An Economic Analysis Peter Meier, Maria Vagliasindi, and Mudassar Imran contributions by Anton Eberhard and Tilak Siyambalapitiya with ­ The Design and Sustainability of Renewable Energy Incentives Direc tions in De velopment Energy and Mining The Design and Sustainability of Renewable Energy Incentives An Economic Analysis Peter Meier, Maria Vagliasindi, and Mudassar Imran with contributions by Anton Eberhard and Tilak Siyambalapitiya © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 17 16 15 14 This work is a product of the staff of The World Bank with external contributions. The findings, interpreta- tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Meier, Peter, Maria Vagliasindi, and Mudassar Imran. 2015. The Design and Sustainability of Renewable Energy Incentives: An Economic Analysis. Directions in Development. Washington, DC: World Bank. doi: 10.1596/978-1-4648-0314-7. License: Creative Commons Attribution CC 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Responsibility for the views and opinions expressed in the adaptation rests solely with the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any ­ third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@ worldbank.org. ISBN (paper): 978-1-4648-0314-7 ISBN (electronic): 978-1-4648-0315-4 DOI: 10.1596/978-1-4648-0314-7 Cover photo: © Thinkstock. Used with the permission of Thinkstock. Further permission required for reuse. Cover design: Naylor Design. Library of Congress Cataloging-in-Publication Data has been requested. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Contents Acknowledgments xv Executive Summary xvii Abbreviations xxi Chapter 1 Introduction 1 Background 1 Key Issues 4 Objectives 4 Why Is Renewable Energy Important for Poor Countries? 7 Taxonomy of Financial Incentive Mechanisms 8 Economic vs. Financial Incentives 12 Organization of the Rest of the Report 13 Notes 13 Bibliography 14 Chapter 2 The Economic Rationale for Renewable Energy 17 Analytical Framework 17 Local Environmental Damage Costs 19 Discount Rate 26 The Social Cost of Carbon 29 Fossil-Fuel Price Subsidies 31 Renewable Energy and Employment 34 Specific Questions for the Case Studies 38 Methodology 39 Notes 40 Bibliography 41 Chapter 3 Case Study: Vietnam 43 Sector Background 43 Power Sector Development 43 Renewable Energy Development 45 Renewable Energy Resource Endowment: The Supply Curve 52 Production Costs 56 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   v   vi Contents The Avoided Social Cost of Thermal Generation 66 Carbon Accounting and the Clean Development Mechanism (CDM) 66 Renewable Energy Targets 69 Design of Incentive Schemes 70 Incremental Costs and Their Recovery 74 Impact of Renewable Energy Tariffs on the Consumer 79 Decreasing the Consumer Cost with International Assistance 81 The Cost of Fossil-Fuel Subsidies 81 Conclusions 82 Notes 85 Bibliography 89 Chapter 4 Case Study: Sri Lanka 91 Sector Background 91 Renewable Energy Development 94 Renewable Energy Resource Endowment and the Renewable Energy Supply Curve 95 Capital Costs 100 The Avoided Social Cost of Thermal Generation 100 Carbon Accounting and CDM 102 Renewable Energy Targets 104 Design of Incentive Schemes 105 Incremental Costs and Their Recovery 111 Impact of Renewable Energy Tariffs on the Consumer 114 The Cost of Fossil-Fuel Subsidies 117 Financing New and Renewable Energy 118 Conclusions 119 Notes 120 Bibliography 122 Chapter 5 Case Study: Indonesia 125 Sector Background 125 Renewable Energy Development and the Resource Endowment 128 Renewable Energy Targets 129 Production Costs 129 Geothermal Development Policy Issues 129 The Renewable Energy Supply Curve 132 Carbon Accounting and CDM 133 Design of Incentive Schemes 134 Detailed Design of the Geothermal Feed-In Tariff 138 Incremental Costs and Their Recovery 141 Potential Impact of Incremental Costs on the Consumer 143 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Contents vii Buying Down the Price of Renewable Energy with International Assistance 144 The Environmental Costs of the Electricity Subsidy 145 Conclusions 148 Notes 151 Bibliography 153 Chapter 6 Case Study: South Africa 155 Sector Background 155 Renewable Energy Development 158 Renewable Energy Targets 160 Design of Incentive Schemes 160 Impact of Renewable Energy Tariffs on the Consumer 165 Conclusions 165 Note 168 Bibliography 168 Chapter 7 Case Study: Tanzania 171 Sector Background 171 Renewable Energy Development 174 Renewable Energy Targets 175 Design of Incentive Schemes 175 Conclusions 183 Notes 184 Bibliography 184 Chapter 8 Case Study: The Arab Republic of Egypt 187 Sector Background 187 Renewable Energy Development 192 Renewable Energy Targets 196 Production Costs 199 Design of Incentive Schemes 201 Carbon Accounting 205 Incremental Costs and Their Recovery 207 Buying Down the Incremental Costs with Concessionary Finance 210 Conclusions 213 Notes 216 Bibliography 216 Chapter 9 Case Study: Brazil 219 Sector Background 219 Renewable Energy Development 220 Renewable Energy Targets 221 Design of Incentive Schemes 222 Financing of Incremental Costs 230 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 viii Contents Conclusions 231 Note 233 Bibliography 233 Chapter 10 Case Study: Turkey 235 Sector Background 235 Renewable Energy Resource Endowment 237 Renewable Energy Development 237 Renewable Energy Targets 237 Design of Incentive Schemes 238 Conclusions 248 Notes 250 Bibliography 250 Chapter 11 Summary and Conclusions 251 The Financial Health of Power Utilities 251 Widespread Consumer and Political Support 252 Setting Renewable Energy Targets 255 The Difficulty of Predicting Unexpected Consequences 256 Risk and Reward 257 Institutional Barriers 257 The Avoided Cost of Carbon 258 Definitions of Renewable Energy 259 The Transparency of Tariffs 260 Auctions 261 Note 262 Bibliography 263 Appendix A Dealing with Uncertainty in Setting Renewable Energy Targets: Croatia 265 Appendix B Multi-Attribute Decision Analysis and Trade-Off Plots 269 Appendix C Estimating Incremental Costs from Renewable Energy Supply Curves 277 Boxes 1.1 A Paradox in the Design of the Erneuerbare-Energien-Gesetz (EEG) Surcharge, German Renewable Energy Sources Act 2000 9 2.1 The Renewable Energy Supply Curve in India 21 2.2 The Welfare Impacts of Fuel Subsidies 32 2.3 Lessons Learned from the German Energiewende (Energy Transition) 35 3.1 Development of Small Hydro in Vietnam 45 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Contents ix 3.2 Counterpoint: Small Hydro Development in the Lao People’s Democratic Republic 46 3.3 The Capacity Value of Renewables in China 60 4.1 The ESDP On-Lending Program for Renewable Energy Finance 95 4.2 Sri Lanka’s Feed-In Tariff, 2009 112 5.1 Counterpoint: Geothermal Development in the Philippines and Kenya 137 5.2 Local Environmental Damage Costs of Thermal Generation in Indonesia 140 5.3 Deadweight Losses of Electricity Tariff Subsidies 147 11.1 Lessons from Germany: Coping with Higher Shares of Renewables 252 Figures 2.1 Economic Rationale for Renewable Energy: Optimal Quantity (QFIN) at Financial Cost of Thermal Energy (PFIN) 18 2.2 Optimal Quantity (QECON) at the Economic Cost of Thermal Energy (PECON) 18 2.3 Optimal Quantity of Renewable Energy, Taking into Account the Environmental Damage Cost 20 2.4 The Economic Rationale for Renewable Energy: China 20 B2.1.1 Renewable Energy Supply Curve in India, by States and Energy Source 22 2.5 Variation in Damage Cost Estimates 23 2.6 Damage Costs of NOX Emissions vs. Per Capita GDP in Selected European Countries 25 2.7 The Impact of a Discount Rate on an Optimal Capacity Expansion Plan: Sri Lanka 28 2.8 Impact of Coal Price Subsidies 31 2.9 Energy Subsidies, by Fuel, in Non-OECD Countries 33 B2.3.1 Development of Renewables-Based Electricity Generation and Investment 35 B2.3.2 Development of Renewables-Based Jobs and Ownership, 2012 36 3.1 Vietnam’s Capacity Expansion Plan 44 3.2 Distribution of Tariffs, and Individually Negotiated Tariffs 48 3.3 2020 RE Targets for Electricity Generation: Vietnam and the European Union 51 3.4 The 2025 RE Supply Curve for Vietnam: Installed Capacity 52 3.5 The 2025 RE Supply Curve for Vietnam: Generation 55 3.6 Impact of the REMP Development Scenario: Economic + Wind Demonstration Program, 2009–25 57 3.7 Required FIT to Maintain 15 Percent FIRR vs. Capital Costs 59 B3.3.1 Capacity Displacements in the North China Grid 60 B3.3.2 Capacity Displacements in Zhejiang 61 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 x Contents 3.8 Operation of the Nam Mu, Daily Peaking, Small Hydro Project 62 3.9 Operation of the Proposed Ly Son Island Wind Project 63 3.10 Dispatch of a Portfolio of Small Hydro Projects 63 3.11 Average Contribution of Small Hydro during the Five Peak Hours of the Day, 2009 64 3.12 Vietnam’s Expected Generation Mix (2008–25) 67 3.13 Wet Season Generation, Typical July Week 68 3.14 Averaging Intervals in the Avoided Cost Tariff 78 4.1 Generation Mix: The Vision, 2009–27 93 4.2 Wind Characteristics in Sri Lanka 97 4.3 Wind Generation and the Daily Load Curve 98 4.4 Distribution of Capital Costs for Small Hydro Projects 101 4.5 The Least Cost Expansion Plan, 2009–27 103 4.6 Greenhouse Gas Emissions vs. System Cost Trade-Offs 106 4.7 Monthly Production from Small Hydro Projects, 2000–02 109 4.8 Tariff Expectations: Baseline, 2009–27 115 4.9 Tariff Impact of the 10 Percent RE Target: 2009 Forecast 115 4.10 Forecast of CEB Fuel Use, 2009–27 118 5.1 The Forecasted Generation Mix 126 5.2 Electricity Price vs. Installed Capacity 130 5.3 Supply Curve for Java-Bali 133 5.4 PLN CO2 Emissions 134 B5.3.1 Impact of Tariff Subsidy 147 6.1 Average Nominal and Real Eskom Electricity Prices 160 7.1 Announced FIT and Surplus Power from Existing Power Plants, 2007–12 179 7.2 A Hypothetical Case, Illustrating Negative Returns in the Initial Years 180 7.3 Cash Flow Profiles for a Mini-Hydro SPP 180 7.4 Cash Flow Profiles for a Mini-Hydro SPP with a 5 Percent Growth in FIT 181 7.5 Optional Approaches to Improve the Cash Flow of Mini-Hydro SPPs 182 8.1 Structure of the Egyptian Power Sector 188 8.2 Evolution of Oil Production and Consumption and Reserves 190 8.3 Evolution of Natural Gas Production and Consumption and Reserves 191 8.4 Evolution of Fuel Mix Production and Installed Capacity, FY2002/03 to FY2011/12 193 8.5 Evolution of Hydropower in the Arab Republic of Egypt, 1955–2025 194 8.6 Evolution of Hydropower in the Arab Republic of Egypt, FY2007/08 to FY2010/11 195 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Contents xi 8.7 Wind-Installed Capacity and Production, FY2002/03 to FY2011/12 196 8.8 Fuel Savings Due to the Implementation of Wind Energy Projects, FY2002/03 to FY2011/12 196 8.9 Integrated Gas-Steam-Solar Combined Cycle 197 8.10 CO2 Emissions Reduction Due to the Implementation of Wind Energy Projects, FY2002/03 to FY2011/12 198 8.11 The Egyptian Solar Plan Approved by the Cabinet in July 2012 198 8.12 Impact of CDM on Incremental Costs of Wind and CSP 206 8.13 Stakeholder Impacts, No Foreign Assistance 210 8.14 Impact of IBRD Financing 211 8.15 Revised Finance Package 215 9.1 Price Evolution through Wind Auctions, 2009–12 227 9.2 Price Evolution through Hydroelectric Auctions, 2007–11 227 9.3 Price Evolution through Biomass Auctions, 2007–11 228 10.1 Evolution of the Power Market in Turkey: Key Phases 237 10.2 Evolution of the Power Market in Turkey: Generation and Installed Capacity 238 10.3 The First TEIAS Wind Capacity Auction: Capacity Allocations 241 10.4 Development of Installed Wind Capacity, 2000–12 242 10.5 Licensing Process 244 B11.1.1 The German Feed-In Premium 253 B11.1.2 Enhancing System Flexibility in Germany 253 B11.1.3 Another View of System Flexibility in Germany 254 B.1 Emissions vs. Stack Height and Population Weighted Index, Sri Lanka 271 B.2 Illustrative Trade-Off Plot 272 B.3 Power Sector Options in Vietnam 273 C.1 Castlerock Supply Curve, Java and Sumatra (with the Old Ceiling Price) 278 C.2 Impact of a 12.5 Cents/kWh Ceiling Price 278 Tables 1.1 Taxonomy of Financial Incentive Mechanisms for Renewable Energy 5 2.1 Externality Costs of Coal Generation 19 2.2 Local Externality Damage Costs in Selected Countries 21 2.3 Damage Cost of NOX Emissions from Combined-Cycle Gas Turbines in the Arab Republic of Egypt 24 2.4 Damage Cost Estimates ($/ton Emissions per Million People per $1,000 of Per Capita GDP Income) 25 2.5 Damage Costs of NOX Emissions from Combined-Cycle Gas Turbines in the Arab Republic of Egypt 26 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 xii Contents 2.6 Discount Rates in World Bank Renewable Energy Projects 27 2.7 Carbon Valuations in World Bank Studies and Project Appraisals 30 2.8 The Avoided Cost of Carbon for Concentrated Solar Power 30 The Welfare Impact of Subsidy Removal B2.2.1 33 Standardized Tariffs B3.2.1 47 Project Characteristics B3.2.2 47 3.1 The 2009 Avoided Cost Tariff, D/kWh 49 3.2 Realization of the Avoided Cost Tariff, Run-of-the-River Project, North Vietnam 49 3.3 Tariff Realizations for Different Types of Projects, 2009–12 50 3.4 CO2 Emissions per Kilowatt-Hour Generated in Selected Countries, 2008 52 3.5 Status of Wind Power Development in Vietnam 53 3.6 Economically Optimal Renewable Energy: Installed Capacity 55 3.7 Economically Optimal Renewable Energy in Vietnam Generation (GWh): Projected to 2025 56 3.8 Wind Support Tariff Comparisons 59 3.9 Capacity Penalty, Wind Project with Annual Load Factor of 26.9 Percent 61 3.10 Tariff Required for 15 Percent FIRR (Post Tax) 65 3.11 Tariff Required (in Cents): 3 Percent Annual Real Price Escalation for Rice Husk 65 3.12 Avoided Social Cost of Gas Generation 68 3.13 Potential CDM Revenue 69 3.14 Capacity Factors and Wind Auction Prices in Brazil 72 3.15 Wind Capacity Factors in Selected Countries 73 3.16 Design of Existing RE Incentive Schemes in Vietnam 73 3.17 CPC’s Incremental Financial Cost of Wind Energy, 2011 74 3.18 EVN’s Incremental Financial Cost of Wind Energy, 2011 74 3.19 Small Hydro Project: Typical Purchase Costs vs. Wholesale Price, 2010–11 75 3.20 Summary of Incremental Network Costs 76 3.21 Connection Costs at Large Hydro and Thermal Projects 76 3.22 Merit Order Stack, 2010 77 3.23 Impact of a 12 Cents/kWh Feed-In Tariff on CPC Cash Flows 79 3.24 Impact on Consumers: 1 Percent Additional Wind Power by 2020 80 3.25 Buying Down the Cost 81 3.26 Impact of Price Increases to Reduce Subsidy 82 4.1 Electricity Sales 92 4.2 Wind Resources of Sri Lanka 96 4.3 Wind Projects in Sri Lanka 97 4.4 Small Hydro Projects in Sri Lanka, 2002–12 99 4.5 Potential Small Hydro Projects in the Estate Sector 99 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Contents xiii 4.6 Status of Grid-Connected RE Projects, March 31, 2013 100 4.7 2009 Feed-In Tariff Cost Assumptions 101 4.8 Grid Emission Factors, 2008–11 104 4.9 Carbon Point Forecasts, 2008–20 104 4.10 Avoided Costs of Carbon 105 4.11 Implementation Scenario for NCRE: 2010 Projection 107 4.12 Avoided Cost Tariff, 1996–2011 108 4.13 Original Recommendation for the Small Power Purchase Tariff 109 4.14 Flat-Rate Feed-In Tariffs 113 4.15 Summary Evaluation of Tariff Designs 114 4.16 Impact of a 1 Percent Increase in Renewable Energy 116 4.17 Impact of RE on CEB Revenue Requirements 117 5.1 Regional Imbalances of Electricity Supply 127 5.2 Renewable Energy Targets: The 2006 National Energy Plan 129 5.3 Prices for Geothermal Projects 130 5.4 Changes in Capacity 132 5.5 The New Geothermal Feed-In Tariff (Established in 2012) 136 B5.1.1 Global Installed Geothermal Capacity, December 2010 137 5.6 Avoided Costs of Thermal Generation 139 5.7 Summary Evaluation of Tariff Designs 142 5.8 Impact of Subsidy on Assumptions (Java and Sumatra) 142 5.9 Impact of Incremental Costs 143 5.10 Buying Down the Incremental Costs 145 5.11 Impact of Electricity Subsidies, 2011 and 2021 146 5.12 Impact of the Price Elasticity Assumption, 2011 146 6.1 Eskom’s Power Stations 156 6.2 South African Integrated Resource Plan, 2010–30 158 6.3 Economic Development Threshold and Target Levels for Wind Energy 163 6.4 Capacity of Renewable Energy Made Available for Bids and Finally Allocated to Preferred Bidders 164 6.5 Prices for Renewable Energy: REFITs, REBID Caps, and Average Bids 164 7.1 Hydro Candidates 173 7.2 National Demand Forecast 174 7.3 SPP Tariffs for the Main Grid, 2008–12 176 7.4 Mini-Grid Feed-In Tariffs, 2009–12 177 7.5 Required Additional Incentives to FIT for Project Viability 183 8.1 Major Dams on the Nile River 194 8.2 Wind Projects 195 8.3 Technical Specifications of the Kuraymat Solar Field 197 8.4 Future PV Projects 199 8.5 Production Costs: Generation Alternatives 200 8.6 Impact of Capital Cost Reductions 201 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 xiv Contents 8.7 Incentive Mechanisms 201 8.8 Wind Projects Currently under Development by the Egyptian Government 203 8.9 Wind Projects That Will Be Built and Operated by the Private Sector on a BOO Basis to Supply the National Electricity Network 204 8.10 Wind Projects to Be Built and Operated by the Private Sector for Self-Consumption or to Directly Sell to Consumers 204 8.11 Avoided Cost of Carbon: Concentrated Solar Power 205 8.12 Avoided Cost of Carbon: Wind 206 8.13 Incremental Financial Cost to the Arab Republic of Egypt, $ million (as NPV) 207 8.14 Reconciliation of Economic and Financial Flows, $ Million, as NPV at 10 Percent Discount Rate (Domestic Finance Only) 209 8.15 Financing Options 211 8.16 Proposed Application of Funds 212 8.17 Comparison of Effectiveness in Buying Down the Incremental Costs 212 8.18 Reconciliation of Economic and Financial Flows ($ Million, as NPV at 10 Percent Discount Rate): Revised Finance Package 214 8.19 Incremental Financial Flows for Tariff Support, Revised Financial Package 215 8.20 Tariff Support, 50 Percent Grant 216 9.1 Targets under the PROINFA 222 9.2 Average FIT Levels under the PROINFA 224 9.3 Renewable Auctions 225 9.4 Financing Conditions Offered by BNDES under the PROINFA 229 9.5 BNDES Financing Conditions Offered to Generation Projects 229 9.6 A Comparison of the PROINFA and RE-Specific Auctions 232 10.1 The Results of the First TEIAS Wind Capacity Auction 242 10.2 Feed-In Tariff 246 10.3 Proposed Premium for Use of Equipment Manufactured Locally 247 11.1 Renewable Energy Targets in the Case Study Countries 256 11.2 Avoided Cost of Carbon 258 11.3 Size Thresholds for “Small Hydro” Projects 259 11.4 Tariff Assumptions in the Philippines (for Biomass) 261 A.1 Economically Optimal Quantity of Renewables 266 A.2 Payoff Matrix (Net Benefits in 2010, in € Million) 267 A.3 Revised Payoff Matrix: Future Coal Plant Unlikely 267 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Acknowledgments This report was developed by a team led by Maria Vagliasindi, lead economist, Energy Anchor. The team is grateful for the input provided by Mohab Hallouda, Margherita Fabbri, JangHo Park, Yun Wu, Sebastian Lopez-Azumendi, and Joeri de Wit (all from the World Bank) and Hatem Elrefaei (an external expert). The report would not have been possible without the invaluable advice, com- ments, and suggestions of expert advisory and peer reviewers, including John Besant-Jones, lead energy advisor (World Bank); Luiz Maurer, principal specialist (International Finance Corporation, IFC); Katharina Gassner, senior economist (IFC); Silvana Tordo, lead energy economist (Sustainable Energy Department); Fan Zhang, energy economist (Europe and Central Asia Energy Unit); and Joshua Gallo, infrastructure specialist (Public-Private Infrastructure Advisory Facility). Excellent editorial support was provided by Fayre Makeig. Financial support from the Norwegian Trust Fund for Private Sector and Infrastructure (NTF-PSI) is gratefully acknowledged. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   xv   Executive Summary Several factors—including the growing demand for energy to fuel economic development, the need to diversify into environmentally sustainable supply sources and ensure energy security, and climate change considerations—have contributed to a need for accelerating public and private investment in r ­ enewable energy (RE). Many countries have designed incentive structures to induce pri- vate investment in renewables, especially those that involve higher or incremen- tal costs. There remains much debate, however, on the cost-effectiveness—from an economic and financial perspective—of various incentives to promote renew- ables around the world, and on how to best address issues related to regulatory design and affordability. In an attempt to contribute to the lively debate, this study provides a global taxonomy of the economic and financial incentives provided by RE support schemes. It summarizes economic models of the sustainability and affordability of such support schemes, alongside operational advice on how the regulatory design may need to be modified to minimize the impact on the budget and be affordable to the poor, as well as how to identify—and fill—the financing gap. In line with its objectives, the study examines a range of issues associated with RE development that fall under the following broad categories: the effectiveness of incentive mechanisms, the details of tariff design, the integration of climate finance considerations into existing regulatory processes, and financing and affordability issues. Under each category the report represents the first systematic attempt to respond to questions such as: What types of incentive schemes have proven to be the most successful in attracting private investment in renew- able generated electricity? How do feed-in tariffs (FITs) compare with renewable portfolio standards (RPSs), quota systems, avoided-cost-based tariffs, funds, and auctions? How important are the details of FIT system design, which may include capping (government-established limits on installation)? From a broader policy perspective, how cost-effective are RE solutions in reducing or controlling greenhouse gas (GHG) emissions? What are the incremental costs of renewables, who pays for them, and what is the impact of RE support mechanisms on consumers? The novelty of this work is the fact that it introduces a rigorous and objective economic perspective on current RE support mechanisms and an empirical analysis of the strengths and weaknesses of these mechanisms—both of which The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   xvii   xviii Executive Summary are much needed in a debate often dominated by widespread misconceptions. The economic rationale for RE is straightforward: the optimum amount of RE for grid-connected generation is given by the intersection of the RE supply curve with the avoided cost of thermal electricity generation. The proposed analytical framework (a) differentiates and illustrates trade- offs—among local, regional, and national impacts, in the short and long run; (b) captures distributional impacts (since subsidies to cover the incremental costs of RE may have very different beneficiaries); and (c) captures externalities and compares (where possible) alternative projects based on equivalent output and cost (comparing, for example, RE and energy efficiency projects against those using fossil fuels). Accordingly, the study advocates for the need to get the eco- nomic, financial, and institutional basics right for the deployment of RE. The study’s integration of RE subsidies with fossil-fuel subsidies is another novel and important contribution. This allows important comparisons. For example, to reduce carbon intensity in the economies of developing countries, is it more efficient to deploy RE or implement alternative options, such as eliminat- ing subsidies on fossil fuels? It is easily shown that both social and global welfare increases as a result of eliminating subsidies; any reduction in fossil-fuel subsidies is a win-win situation. But the political economy of such reforms represents a major challenge. The work is based on case studies of Vietnam, Indonesia, Sri Lanka, South Africa, Tanzania, the Arab Republic of Egypt, Brazil, and Turkey, selected to provide a representative sample of varied energy endowments (coal, natural gas, ­ and hydro-based systems) and policy incentives (from FITs to auctions). The case studies compare the incremental cost of RE (from wind to mini-hydro, solar, and biomass) with the average cost of generation (again, highly dependent on the energy resource endowment) and determine the impact of alternative support mechanisms on the government budget and residential consumers. An analytical framework provides the underpinnings of the case studies, and provides the background for the principal research hypothesis of this report: more attention to the principles of economic analysis and market efficiency leads to more sus- tainable and effective policies. The main premise is that the economic rationale for RE lies at the heart of effective incentive mechanism design. The main lessons emerging from the case studies are clear and inescapable; successful RE policies: • Will only be effective once the state-owned utilities who are the buyers of grid-connected RE are themselves in good financial health (in all of the case study countries, the power utilities are under financial duress). • Need to be grounded in economic analysis and accompanied by the applica- tion of market principles to ensure economic efficiency. • Require a sustainable, equitable, and transparent recovery of incremental costs. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Executive Summary xix These points are elaborated as follows: 1. The first and arguably most important conclusion about sustainable incentives for RE relates to the financial health of the power utilities. Until such time as these utilities are in good financial health, and operate under a transparent regulatory system that sets electricity tariffs on a sustainable basis—and allows for the incremental costs of RE to be passed to the consumer—they will con- tinue to oppose what they see as unnecessary costs that will worsen their already poor financial situation. The idea that one can achieve sustainable recovery of incremental costs for RE where utilities are in financial distress is unrealistic. 2. Although numerous studies have advocated using economic principles as a basis for RE targets, few countries have in fact done so. The lack of intellec- tual rigor in setting RE targets lies at the heart of the slow uptake of RE generation in most of the case study countries. Targets that bear no relation- ship to the economic realities of the incremental costs of RE are rarely achieved; even worse are those targets (and associated support tariffs) issued in the complete absence of knowledge about the magnitude of the incremen- tal costs implied (the most notable recent example of which is the 2012 Indonesian geothermal FIT). 3. Incentives can be successful in enabling significant private sector investment in RE. But this is merely a necessary condition, not a sufficient condition. To be successful, such a tariff needs to be transparent in its methodology, be accom- panied by a nonnegotiable power purchase agreement; propose clear arrange- ments for transmission costs, and be clear about the magnitude of expected incremental costs and how these will be recovered. Transparency is important because private developers and their lenders require assurance about the evo- lution of the tariff in the future, and need to understand the methodology of its derivation so that they can themselves make an assessment of future cash flows. Transparency in setting and adjusting a support tariff will necessarily support its acceptance. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Abbreviations ACT avoided cost tariff ADB Asian Development Bank AfD French Development Assistance AfDB African Development Bank ANEEL Brazilian Electricity Regulatory Agency ASEAN Association of Southeast Asian Nations ASTAE Asia Sustainable and Alternative Energy Programme (World Bank) AWDR average weighted deposit rate (Sri Lanka) BBBEE broad-based black economic empowerment bbl barrel bcm billion cubic meters BEP best efficiency point (hydro turbine) BM build margin (CDM methodology) BNDES Brazilian Development Bank BNE best new entrant BO build, operate BoI Board of Investment BOO build, own, and operate BOOT build, own, operate, transfer BOT build, operate, transfer bp basis point bpd barrels per day BTU British thermal unit BWEA British Wind Energy Association CAPM capital asset pricing model CCCT combined-cycle combustion turbine CCGT combined-cycle gas turbine CDCF Community Development Carbon Fund The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   xxi   xxii Abbreviations CDM clean development mechanism CEB Ceylon Electricity Board (Sri Lanka) CEPEL Centro de Pesquisas de Energia Elétrica CER certified emission reduction CF capacity factor CFB circulating fluidized bed cif cost insurance freight CNPE National Energy Policy Council CO2 carbon dioxide COD closure of development COP Copenhagen Conference of Parties CPC Central Power Company (Vietnam) CPC Ceylon Petroleum Corporation (Sri Lanka) CPI consumer price index CRESP China Renewable Energy Scale-up Program CS consumer surplus CSP concentrated solar power CTF Clean Technology Fund cumec cubic meter per second CV compensating variation DC direct current DFCC Development Finance Corporation of Ceylon DSCR debt service cover ratio DSI Directorate of State Hydraulic Works DSM demand-side management dwt dead weight ton ECX European Carbon Exchange EdL Electricite de Laos EEA European Environment Agency EEG Erneuerbare-Energien-Gesetz (Renewable Energy Sources Act) EEHC Egyptian Electricity Holding Company EETC Egyptian Electricity Transmission Company EgyptERA Egyptian Electric Utility and Consumer Protection Regulatory Agency EIA Environmental Impact Assessment EIPS Environmental Issues in the Power Sector (Sri Lanka, World Bank study) EMA Energy Market Authority (of Singapore) EME Exempt Micro Enterprise The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Abbreviations xxiii EML Electricity Market Law EMRA Energy Market Regulatory Authority EMRRP Estate Micro Hydro Rehabilitation and Re-Powering Project EPC engineering, procurement, and construction (contract) ERAV Electricity Regulatory Authority of Vietnam ERPA Emissions Reduction Purchase Agreement ERR economic rate of return ESDP Energy Services Delivery Project (Sri Lanka) Eskom Electricity Supply Commission of South Africa ESMAP Energy Sector Management Assistance Program (World Bank) ESP electrostatic precipitator EU European Union EUA EU Allowance Unit of one ton of CO2 EU-ETS European Union Emission Trading Scheme EV equivalent variation EVN Electricity of Vietnam EWURA Energy and Water Utilities Regulatory Authority FERC Federal Energy Regulatory Commission (U.S.) FGD flue gas desulphurization FIDIC International Federation of Consulting Engineers FIRR financial internal rate of return FIT feed-in tariff fob free on board FOREX foreign exchange FS feasibility study FTP2 second fast-track program GDP gross domestic product GEF Global Environment Facility GHG greenhouse gas GoV Government of Vietnam GTZ Gesellschaft für Technische Zusammenarbeit (Germany) GW gigawatt = 1,000 MW GWh gigawatt-hour ha hectare HCMC Ho Chi Minh City HFO heavy fuel oil HHV higher heating value HRSG heat recovery steam generator The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 xxiv Abbreviations HSFO high sulphur fuel oil HX heat exchanger IBRD International Bank for Reconstruction and Development IDA International Development Association IDC interest during construction IEA International Energy Agency IEC International Electrotechnical Commission IFC International Finance Corporation IFI international financial institution IMF International Monetary Fund IoE Institute of Energy (Vietnam) IPCC Intergovernmental Panel on Climate Change IPP independent power producer IRP integrated resource plan IRR internal rate of return ISCC Integrated Solar Combined Cycle ISO International Standards Organisation IUP Izin Usaha Panas Bumi (geothermal business license) JCC Japan Crude Cocktail JICA Japan International Cooperation Agency JSC Joint Stock Company KCal kilocalories KfW German Development Bank kg kilogram kW kilowatt kWh kilowatt-hour = 3,412 BTU LCOE levelized cost of energy LDU local distribution utility LECO Lanka Electricity Company (Private) Limited (Sri Lanka ­distribution company) LF load factor LFG landfill gas LHV lower heating value LIBOR London inter-bank offer rate LNG liquefied natural gas LoI letter of intent LR licensing regulation LRMC long-run marginal cost LV low voltage The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Abbreviations xxv m/sec meter per second m2 square meter m2/yr square meter per year MADA multi-attribute decision analysis MAE Wholesale Electric Energy Market MARD Ministry of Agriculture and Rural Development (Vietnam) MEM Ministry of Energy and Minerals MEMR Ministry of Energy and Mineral Resources MENA Middle East and North Africa MENR Ministry of Energy and Natural Resource mmBTU million British thermal units MME Ministry of Mines and Energy MMS mandated market share MNES Ministry of Non-conventional Energy Sources (India) MoE Ministry of Energy MOEE Ministry of Electricity and Energy MoF Ministry of Finance MoIT Ministry of Industry and Trade (Vietnam) MoNRE Ministry of Natural Resources and Environment (Vietnam) MoU memorandum of understanding MSWI municipal solid waste incineration Mt metric ton MT million tons mtpa million tons per annum MUV manufacture unit value (index) MV medium voltage MW megawatt = 1,000 kW MWh megawatt-hour MWL minimum water level NCRE nonconventional and renewable energy (Sri Lanka) NERSA National Energy Regulator of South Africa NIF Neighbourhood Investment Facility (European Union) NLDC National Load Dispatch Centre (Vietnam) NOX nitrogen oxide NPC National Power Corporation NPV net present value NREA New and Renewable Energy Authority NREL National Renewable Energy Laboratory NTF-PSI Norwegian Trust Fund for Private Sector and Infrastructure The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 xxvi Abbreviations O&M operation and maintenance OCCT open-cycle combustion turbine OCGT open-cycle gas turbine OECD Organisation for Economic Co-operation and Development OLS ordinary least squares OM operating margin (CDM methodology) OPEC Organisation of Petroleum Exporting Countries OPEX operating expenses ORB OPEC Reference Basket (crude oils) PA Privatization Administration PAD Project Appraisal Document (of the World Bank) PC pulverized coal PDA project development agreement PDD project design document (of the UNFCCC) PDP Power Development Plan (Vietnam) PDP7 Seventh Power Development Plan (Vietnam) PGE Pertamina Geothermal Energy PLN Perusahaan Listrik Negara (Indonesian State Electric Utility Company) PM-10 particulate matter (no greater than 10 microns in diameter) PPA power purchase agreement PPIAF public-private infrastructure advisory facility PPP purchasing power parity PPPs public-private partnerships PROINFA program for the promotion of renewable energy PS pumped storage PSO public service mechanism PTC production tax credit PUCSL Public Utility Commission of Sri Lanka PURPA Public Utilities Regulatory Policy Act (United States) PV photovoltaic QF qualifying facility QSE qualifying small enterprise R&D research and development RE renewable energy REAP Renewable Energy Action Plan (Vietnam) REDP Renewable Energy Development Project (World Bank, Vietnam) REIPPP Renewable Energy Independent Power Producer Procurement The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Abbreviations xxvii REL REPA Renewable Energy Law Wind Energy Potential Map of Turkey REMP Renewable Energy Master Plan (Vietnam) RER Certificate Renewable Energy Resource Certificate RERED Renewable Energy for Rural Economic Development RESPP renewable energy small power producer RfP request for proposals ROE return on equity RoR run-of-river RPS renewable portfolio standard RSA Republic of South Africa SBV State Bank of Vietnam SC supercritical SCADA supervisory control and data acquisition SCC social cost of carbon SCF standard conversion factor SCF statement of cash flow SEFI Sustainable Energy Finance Initiative SEIERP System Efficiency Improvement, Equitization and Renewables Project (World Bank) SGD Singapore dollars SHP small hydro project SIDA Swedish International Development and Cooperation Agency SLF system load factor SLPUC Sri Lanka Public Utilities Commission SLSEA Sri Lanka Sustainable Energy Authority SMO system market operator SMS Turkish State Meteorological Service SO system operator SO2 sulphur dioxide SPDF special purpose debt facility SPP small power producer SPPA standardized power purchase agreement SV switching value T&D transmission and distribution TA technical assistance TANESCO Tanzania Electric Supply Company TCM thousand cubic meters TEAS Turkish Electricity Generating and Transmission Corporation TEDAS Turkish Electricity Distribution Company The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 xxviii Abbreviations TEK Turkish Electricity Authority TGC tradable green certificate TJLP long-term interest rate TKB Turkish Development Bank TOE tons of oil equivalent TOOR transfer of operating rights ToR terms of reference TRY Turkish Lira TSKB Turkish Industrial Development Bank TSO transmission system operator TSP total suspended particulates TWh terawatt-hour U.S. United States (of America) UAE United Arab Emirates UK United Kingdom UNEP United Nations Environment Programme UNFCCC United Nations Framework Convention on Climate Change UREA Uttranchal Renewable Energy Agency (India) USC ultra super critical VAT value added tax VCGM Vietnam competitive generation market VEPF Vietnam Environmental Protection Fund VM volatile matter VND Vietnamese dong VSL value of statistical life VSPP very small power producer WACC weighted average cost of capital WASP Wien Automatic System Planning WKP Wilayah Kerja Pertambangan Panas Bumi (geothermal work areas as known in Bahasa, Indonesia) WTI West Texas Intermediate (crude oil) YEGM Yenilenebilir Enerji Genel Müdürlüg˘ ü (General Directorate of Renewable Energy) YOLL years of life lost All monetary amounts are in U.S. dollars (US$) unless otherwise indicated. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 1 Introduction Background Rapid urbanization and economic growth, new demographic trends, and climate change are key challenges that developing countries must face as they strive to meet growing energy demand. These and other challenges call for an acceleration of public and private investment in renewable energy (RE). The adoption of the Kyoto Protocol in early 2005 spurred exponential growth in mainstream RE investment around the world. In 2008, for the first time, RE— including large hydropower projects—attracted more power sector investment globally than fossil-fuel-based technologies (UNEP, SEFI, and Bloomberg New Energy Finance 2012). Contributing to this exponential growth was an alignment of global factors: rapid growth in energy demand in emerging economies such as those of China and India, increased competition for energy resources, geopolitical tension and energy security concerns, rising oil and gas prices, as well as the entry into force of the Kyoto Protocol, and the rise of climate change in the political agenda more generally. The traditional functions of energy policy and regulation are to ensure access to adequate and reliable supply, protect consumers from high prices, and ensure that private sector entities will be able to recoup their investment. A fourth goal—decreasing environmental impact—is often added. These goals sometimes conflict with one another. Improved access to reliable, secure, affordable, climate- friendly, and sustainable energy can represent a formidable challenge. Changes can be particularly costly if a move toward a low-carbon solution is implemented through an increasing proportion of RE. Most renewable sources of energy are more expensive than conventional ones; in most cases this is because of high capital costs, spurring changes in the level and composition of investment. In addition, most forms of renewable generation—though good substitutes for conventional sources of energy—are poor in providing capacity at peak time. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   1   2 Introduction A recent review of private sector investment in RE draws several important conclusions about the effectiveness of incentive mechanisms:1 • Developing countries that have introduced feed-in tariffs (FITs) are almost four times more likely to attract private investment in RE—resulting in about seven times more total investment—than countries where such support mechanisms have not been introduced. ­ • The introduction of FITs (and, more broadly, of other support mechanisms) is positively and significantly associated with the introduction of public-private partnerships (PPPs)2 in renewable electricity generation, controlling for several variables (including supply and demand factors), economy-wide governance indicators, and sectoral controls. FITs affect both the entry and the level of investment in renewable-based energy, though they became much less signifi- cant when it comes to determining the amount of investment. This second point suggests the need to revisit the implied allocation of risks between the public and private sectors over time to ensure that FITs produce the desired volume of investment. • In contrast, broader economy-wide governance factors, including the degree of corruption and political competition, are most often considered by private investors as they decide whether to invest in renewable-based generation. This reinforces the hypothesis that private investors seem to be adequately pro- tected against risk: once they have entered the market, they can accommodate the governance environment. • Countries that have enhanced transmission investment have also paved the way for attracting more investment in renewables. This confirms that attracting more private investment depends on the broad policy environment, and not just financial support or incentive mechanisms. Avoiding costly construction delays due to regulatory uncertainties, and lack of transmission and ­infrastructure access pose significant obstacles to timely, successful project development. The effectiveness of FITs and renewable portfolio standards (RPSs) in deploy- ing RE can be framed within the pioneering debate between the use of quantity versus price instruments. In the absence of market imperfections, both policies have the same welfare outcome (Weitzman 1974, 477−91). In the presence of market failures, however, each policy has its relative merits. The key advantage of a FIT is that it reduces investor risk by offering a guaranteed price. On the other hand, a FIT that is too generous can stifle innovation and unnecessarily increase procurement costs. The advantage of an RPS is that it typically stimulates cost- effective procurement by inducing competition between suppliers. On the other hand, such competition may deter the entry of risk-averse RE suppliers and limit the ability to foster technologies that require time to become more competitive. Functional form choices for the independent variable range from binary The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 3 indicators (a policy exists, or it does not) to nominal measures (level of FIT or RPS target) to more nuanced measures that aim to better capture the incentives that the policies provide. Other dimensions, along which these regression approaches differ in scope (in terms of countries and years), are relevant to policy design and range-of-control variables. Overall, there is still a lack of consensus among studies about whether, which, and in what way renewable policies have been successful in stimulating RE. Case studies and cross-sectional regressions typically find that FITs and RPSs are most effective in RE deployment (del Rio Gonzalez 2007, 994−1012; 2008, 2917−29; Haas and others 2011, 2186−93; Lesser and Su 2008, 981−90; Lipp 2007, 5481−95; for RPSs, see Allegappan, Orans, and Woo 2011, 5099−104; Menz and Vachon 2006, 1786−96). But econometric studies using panel data do not con- firm these results. For example, in the case of RPS studies, the presence of an RPS has been found to increase, not affect or even reduce, RE penetration (Carley 2009, 3071−81; Shrimali and Kniefel 2011, 4726−41; Yin and Powers 2010, 1140−49). Some discrepancy between findings is due to the different methodolo- gies that have been used. For example, between cross-sectional and panel data, the ability of panel data to control for time-invariant unobservables encourages greater confidence in the robustness of results. Another reason for discrepancy between findings depends on the different extents to which different studies take account of policy design and policy contexts. The many dimensions of policy design and context are difficult to capture by simple quantitative indicators. Major progress may be observed in the econometric literature, however. Zhang (2013), for example, models several FIT design elements and finds that high feed-in rates do not necessarily lead to an increased uptake of wind power in European countries, but guaranteed grid access and length of feed-in contracts are crucial policy characteristics for RE deployment (Delmas and Montes- Sancho 2011). Several studies, including those of Delmas and Montes-Sancho (2011) and Zhang (2013), attempt to account for the existence of a lag between the enactment of policy and measured policy output. This lag arises because it takes time for investors to respond to incentives and is particularly relevant for technologies with high up-front capital costs, such as RE. Existing studies focus almost exclusively on the United States and the European Union (EU). This focus undoubtedly reflects not only the prevalence and experience of RE policies in developed countries, but also the difficulty of assembling data for quantitative analysis in developing countries. There is limited understanding of renewable policy design considerations that are specific and important to developing countries. Studies also focus on effectiveness as a measure of policy success, rather than cost-effectiveness or efficiency. For example, Zhang (2013) suggests that high subsidies in Europe’s FIT program may have driven up investment costs by allowing installation at low-wind-speed sites. Similarly, Menz and Vachon (2006) suggest that an effective RPS can facilitate the adoption of renewable capacity in states with low resource potential. These results are critical to ensure sound spending of public funds in support of RE generation. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 4 Introduction Key Issues From these general observations follow the main questions to be considered by the case studies included in this report: • Effectiveness and efficiency of incentive mechanisms. What types of incentive schemes prove to be the most successful in attracting private investment in renewable-generated electricity? How do FITs compare with RPSs, quota sys- tems, and auctions in terms of effectiveness and efficiency? How can the com- bination of different incentive schemes (see the taxonomy outlined in table 1.1) (Mitchell, Bauknecht, and Connor 2006; Rickerson and Grace 2007) be used to maximize effectiveness, while reducing the cost burden on the budget and on vulnerable consumers (Cory, Couture, and Kreycik 2009)? • Details of tariff design. How important are the details of the FIT system design, which may include capping (government-established limits on installation); tariffs differentiated by technology; tariff inflation-indexation; duration; and the methodology used to determine tariff levels and to revise tariffs, purchase obligations, the introduction of tariff degressions, and the specific “burden- sharing” system? Which of these design factors will make the business environ- ment for renewables more or less attractive to private investors? • The broader energy policy environment. How effective is the deployment of RE in reducing the carbon intensity of developing country economies, relative to alter- native options, such as eliminating subsidies on fossil fuels? What is the evidence from several of the case study countries that provide large subsidies to coal and gas generators (the Arab Republic of Egypt, South Africa, Indonesia, Vietnam)? • Financing and affordability issues. What is the incremental cost of RE relative to that of fossil fuels? Who pays for it? Are donor grants, concessionary loans, and carbon finance provided by the global community, consumers, or taxpayers? What is the impact of RE support mechanisms on consumers? Is it equitable and affordable for poor consumers in developing countries to contribute? Can the costs be passed on to just large customers (rather than poorer residential customers)? Yet in some countries (including Germany), it is the large custom- ers who are exempt from consumer levies to recover the incremental costs.3 Objectives The main objectives of this study are to offer (a) a global taxonomy of the eco- nomic and financial incentives provided by renewable support schemes and (b) an economic modeling of the sustainability and affordability of such support schemes. Also included is operational advice on how the regulatory design may need to be modified to minimize budgetary impact and be affordable to the poor, with an aim to identify—and fill—the financing gap. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 5 Table 1.1 Taxonomy of Financial Incentive Mechanisms for Renewable Energy Category Type of instrument Who pays Examples Price Production cost- Design decision Algeria (since 2002) based feed-in Austria (since 2002) tariffs (FITs) Belgium Brazil (since 2002 until 2010) Bulgaria (since 2007) Canada (Prince Edward Island, since 2004; Ontario, since 2006) China (since 2005) Cyprus (since 2003) Czech Republic (since 2002) Estonia (since 2003) France (since 2001) Germany (since 1990) Greece (since 1994) Hungary (since 2003) Ireland India (since 1993) Israel (since 2004) Italy (since 1992) Kenya (since 2008) Korea, Rep. (since 2003) Latvia (since 2001) Lithuania (since 2002) Luxembourg (since 1994) Malaysia (since 2010) Malta (since 2010) Netherlands (since 2011) Nicaragua (since 2004) Norway (since 1999) Pakistan (since 2006) Philippines (since 2008) Portugal (since 1999) Slovak Republic (since 2003) Slovenia (since 1999) Spain (since 1994) Sri Lanka (since 2011) South Africa (since 2009—not implemented) Switzerland (since 1991) Tanzania (since 2008) Thailand (since 2006) Turkey (since 2005) Uganda (since 2008) United Kingdom (since 2010) United States (California since 1978; Hawaii since 2008; Oregon and Vermont since 2009) Avoided cost tariffs Design decision Indonesia (2012 geothermal tariff) (ACTs) Sri Lanka (1998−2010) Vietnam table continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 6 Introduction Table 1.1  Taxonomy of Financial Incentive Mechanisms for Renewable Energy (continued) Category Type of instrument Who pays Examples Premiums over Design decision Czech Republic generation market Denmark (premium only) price (“adders”) Estonia Italy Netherlands (premium only) Slovenia Spain Thailand (“adders”) Premiums over retail Consumers, China-Shanghai (“jade” tariff) price (“green voluntarily tariffs”) Quantity Direct auctions for Design decision Brazil (since 2007) price China (wind concessions until 2009) Egypt, Arab Rep. Indonesia (price-based tenders for geothermal work areas) Morocco Peru (since 2009) South Africa (since 2009) Turkey (since 2008) Auctions for subsidy Design decision Thailand (funded from the tax on petroleum products) Renewable portfolio Australia (since 2001) standards (RPSs) Belgium (Flanders since 2002; Walloon since 2003; Brussels since 2004) Canada (Nova Scotia, Ontario, and Prince Edward Island, since 2004) Chile (since 2008) China (since 2007) Italy (since 2001) India (at a state level, Maharashtra since 2003; at a national level, since 2008) Japan (since 2003) Korea, Rep. (since 2012) Latvia Lithuania Poland (since 2005) Philippines (since 2008) Romania (since 2008) Sweden (since 2003) United Kingdom (England, Wales, and Scotland, since 2002; Northern Ireland, since 2005) United States (30 states and the District of Columbia, with Iowa the first, since 1983) Direct Grants and capital Government Belgium subsidies Cyprus Czech Republic Egypt, Arab Rep. Finland Greece Hungary Jordan table continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 7 Table 1.1  Taxonomy of Financial Incentive Mechanisms for Renewable Energy (continued) Category Type of instrument Who pays Examples India Latvia Lithuania Malta Morocco Philippines (grants to consumers for photovoltaic systems) Tunisia Sale of carbon credits Global Most developing countries (for example, CDM) community Grants Global Egypt, Arab Rep. (EU Neighbourhood Investment community facility, grant for CSP) Many countries (grant component of IDA loans, for example, Nepal hydro rehabilitation project) Indirect Preferential taxes Taxpayers Belgium Finland Greece India (accelerated depreciation on wind farms) Spain Tunisia United States (PTC) Most developing countries (import duty and VAT concessions) Preferential domestic Government Bulgaria financing Brazil (low-cost loans to RE producers by BNDES) Estonia Germany Malta Netherlands Poland Slovenia Thailand Preferential foreign Global Indonesia (Carbon Trust Fund support to financing and loan community geothermal projects) guarantees Most developing countries have access to the Carbon Trust Fund Source: Authors’ elaboration. Note: BNDES = Brazilian Development Bank; CDM = clean development mechanism; CSP = concentrated solar power; EU = European Union; IDA = International Development Association; PTC = production tax credit; PV = photovoltaic; VAT = value added tax. The taxonomy is not meant to be exhaustive, but to provide a few representative examples in each category. Why Is Renewable Energy Important for Poor Countries? To date, few World Bank discussions about the need for increasing RE in develop- ing countries have directly confronted one of the fundamental realities of the global climate change debate: governments in poor, developing countries believe that they should not bear the incremental costs of RE in the same way as the governments of, say, Germany, Switzerland, and Sweden. Such beliefs are fundamental to the question of who pays. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 8 Introduction Everyone prefers to be seen as “green,” so there are countless examples of RE targets promulgated as political statements that have little realistic chance of being achieved, or of governments going through the motions of introducing RE incentive schemes that at first glance appear to be based on wishful thinking, but in fact are based on a great reluctance to do anything that results in increases of the electricity tariff. An example from our case studies underscores this point. Vietnam has at best a modest wind resource, and what it does have is highly seasonal (and more sea- sonal than that of Europe or Latin America).4 Given that wind is very high up on the RE supply curve, and that Vietnam has significant small hydro and bio- mass resources that can be exploited at a much lower cost, there is no economi- cally rational reason for Vietnam to pursue wind power. Only years of relentless donor advocacy have persuaded the government to introduce a wind FIT—but one set at such a low level (7.8 cents/kilowatt-hour, kWh) as to have no realistic chance of enabling any wind farms. Understandably, the Government of Vietnam is reluctant to introduce a wind FIT at a level comparable to other Asian countries (16 cents/kWh in the Philippines, 19 cents/kWh in Sri Lanka). With inflation and already sharply increasing electricity tariffs being a real problem in Vietnam, the idea of imposing a consumer levy to recover incremental costs of wind power has been politically unattainable, with the result that the draft Renewable Energy Master Plan, which was submitted in 2009 and which proposed such a consumer levy, has little chance of eventual approval. Yet at the same time, Vietnam has implemented a highly successful small hydro program, with 800 megawatts (MW) enabled since 2009 through its avoided cost tariff (ACT) and standardized power purchase agreement (SPPA). The program has been successful precisely because one could demonstrate that small hydro was economically efficient, with costs at or below the avoided social cost of the thermal alternative. This highlights one of the main themes of this report: namely, that economic rationality lies at the heart of any successful RE program, and that the single most important issue is the transparent recovery of incremental costs. We know of no successful RE program based on attempts to bury incremental costs in nontrans- parent subsidies. The most expensive RE program in the world—in Germany—has been achieved by a transparent consumer levy. In 2012 residential customers paid 25 cents/kWh for electricity, of which the surcharge for the FIT levy accounted for 3.59 cents/kWh, or 13.9 percent of the average bill (see box 1.1). This surcharge will rise to 5.28 cents/kWh in 2013 (excluding value added tax, VAT).5 Taxonomy of Financial Incentive Mechanisms The economic rationale for RE lies at the heart of the design of incentive mecha- nisms. Our proposed taxonomy of incentive mechanisms recognizes four general categories: • Price incentives, as when the government intervenes to provide RE generators with preferential output prices, with the result that the market determines the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 9 Box 1.1 A Paradox in the Design of the Erneuerbare-Energien-Gesetz (EEG) Surcharge, German Renewable Energy Sources Act 2000 Renewable energy (RE) has had a price-curbing impact on wholesale prices in recent years, as additional supply has shifted the demand curve, particularly in the case of wind (Sensfuß, Ragwitz, and Genoese 2007). Because the surcharge is calculated as the difference between the feed-in tariff (FIT) and spot market prices, lower prices mean an increased surcharge (see the figure B1.1.1, panel a). Exempted industrial consumers are net beneficiaries: because of RE they pay lower electricity prices, and almost no surcharge. Households and other small consumers do not benefit from lower prices (due to the merit order effect), as these are not passed on to them (for lack of effective competition among distributors). By contrast, their surcharge payments are increased since they also pay for the “extra cost” share of the exempted industry (see figure B1.1.1, panel b). a. The development of the EEG surcharge, 2000–13 6 5 EURO cents/kWh 4 3 2 1 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Note: kWh = kilowatt hour. b. The breakdown of the EEG surcharge Market Arrears, premium, 3 Liquidity 13 reserve, 3 Reduction of electricity price, 13 Cost of renewable Industry energy privilege, support, 25 43 box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 10 Introduction Box 1.1  A Paradox in the Design of the Erneuerbare-Energien-Gesetz (EEG) Surcharge, German Renewable Energy Sources Act 2000 (continued) Is it true that all costs are passed through to users? The answer is no, as there are other costs attributable to RE sources. These include, for example, the additional cost for basic and balanc- ing energy that is needed because of the fluctuating input of electricity from photovoltaic (PV) and especially wind energy systems. Other factors are grid expansion due to the integration of power from renewables, and administrative costs incurred by grid operators for implementa- tion of the EEG. These additional cost factors are difficult to quantify. They have been esti- mated to total between €300 million and €600 million, the dominant share of which is due to basic and balancing energy. On the other hand, the expansion of renewables also involves a number of beneficial effects that are not reflected in the operating cost factors so far considered. Apart from the reduction in wholesale electricity prices effected by the EEG, the external costs of electricity generation from fossil fuels that are avoided by using RE sources are particu- larly important from a macroeconomic point of view. If these costs were allocated in strict accordance with the “polluter pays” principle, the price of electricity from non-RE sources would be much higher. In this connection a study for the Federal Ministry for the Environment, Nature Conservation, and Nuclear Safety of Germany (Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit, BMU) came to the conclusion that the external costs saved by EEG electricity, between €5.84 billion and €20.44 billion in 2012, were more or less equal to the additional procurement costs for the EEG. Electricity generation from RE sources also results in a significant reduction in imports of coal and natural gas into Germany. In 2012 this reduced Germany’s bill for fuel imports by about €25 billion. One must also remember the positive effects of RE on growth and employ- ment. The basis for this positive trend is the rise in domestic sales of RE that has been in prog- ress for years and—to an increasing extent—the export success of the German renewables sector. The latter is profiting considerably from the fact that the EEG has set in motion a tech- nological development that has given Germany a leading position on the world market in vari- ous fields in the renewables sector. The fact that the EEG itself is increasingly proving to be an export hit, reflects this trend and is one of its main driving forces. Source: Lauber 2013. quantity of RE provided at the stipulated price (though in some countries a cap is placed on the quantity). • Quantity incentives, as when the government sets a target for the quantity of RE to be provided, with the result that the marketplace determines the price (for example, through an auction for a given quantity of megawatt-hours [MWh] to be delivered some years hence). • Direct support. Cash support is provided directly to RE generation projects, either as direct cash subsidies from governments, or as cash from the sale of carbon credits (clean development mechanism, CDM). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 11 • Indirect support. Support is provided to developers through tax rebates and incentives, low-cost loans from government-owned development banks, or concessionary carbon financing. Within each category there are many different specific mechanisms, as listed in table 1.1. Moreover, most countries have in place more than one such mecha- nism, which makes policy interaction and compatibility important issues, since the combined impact may result in inefficient outcomes.6 In some cases, who pays is defined by the nature of the incentive mechanisms. For example, a preferential rate of income tax is necessarily carried by taxpayers, and green tariffs are necessarily carried by consumers. But for those incentives identified in table 1.1 as a “design decision,” who pays must be decided by the government as a matter of policy design. The incremental costs of a FIT can be paid by consumers or from several different sources (as in the case of Vietnam’s wind FIT, by the utilities and the Vietnam Environmental Protection Fund, VEPF). Table 1.1 lists the policies directly aimed at increasing RE generation. But this list excludes the policies that are not expressly directed at promoting RE, but which may in fact have a much greater impact on RE by removing the distortions that lead to the need for RE incentives in the first place. The two main policies in this regard are: • Subsidies on fossil fuels, which make RE appear more expensive than it really is (as in Vietnam and Indonesia). • Subsidies on the retail tariff, whose elimination would (other things being equal) reduce all electricity generation and greenhouse gas (GHG) emissions far more than is achieved by the policies listed in table 1.1 (illustrated by Indonesia, where the Ministry of Finance subsidy to the Indonesian utility Perusahaan Listrik Negara [PLN, or Indonesian State Electric Utility Company] runs to several billions per year). These are important questions for countries whose subsidies are large: from both the Vietnam and Indonesia case studies it can be concluded that removing fuel subsidies would have a far greater impact on GHG emissions—and the amount of RE that would become competitive without subsidy—than the addi- tional RE likely to be enabled by the proposed FITs. Meanwhile, the removal of institutional barriers often unlocks much more RE than attempts to introduce price incentives. This is well illustrated by the Indonesian example: in fact the main barrier to achieving the geothermal targets is not inadequate tariffs, but the barriers faced by private developers in dealing with an often-dysfunctional permitting system in the provinces, and tender com- mittees that lack technical capacity and have awarded tenders at unrealistic prices by developers who lack technical and financial capacity.7 On the other hand, in Vietnam, it is relatively easy to build a small hydro project (SHP), but provincial authorities’ capacity for reviewing SHPs is weak, resulting in wide- spread allegations of environmental damage and the perception that too many SHPs are being built. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 12 Introduction Economic vs. Financial Incentives The different incentives listed in table 1.1 all relate to the financial engineering of projects in an attempt to achieve bankable projects by reducing the financial costs (or increasing the financial benefits through preferential tariffs). Together with reducing subsidies on fossil fuels and on power tariffs, these measures can be advocated on economic efficiency grounds by bringing financial costs closer to economic costs and thereby improving the allocation of resources in the economy.8 But none of these incentives does anything to change the realities of the underlying economic costs and benefits. A quite different set of policies is required to provide incentives to reduce economic costs, or improve the technical effi- ciency of RE. Examples of such incentives include: • Domestic manufacture of RE equipment. The foremost example of this is China, whose low-cost equipment has done far more to promote bankable RE proj- ects in Asia than all of the financial incentives listed in table 1.1. In Vietnam hydro- and wind-turbine generators manufactured in China cost 60 percent of the equipment manufactured in Europe.9 Many countries have attempted to promote domestic manufacture through domestic content provisions (for example, eligibility for low-cost loans from government-owned development banks as in Brazil, or the bonus in the Malaysian FIT for biomass equipment manufactured locally and the bonus in the Turkish FIT and the South Africa local content provision for RE auctions). • Operational optimization. Many hydro projects are not based on a clear under- standing of how reservoir-operating rules and flow-discharge decisions affect generation, resulting in operation at points quite distant from the so-called best efficiency point (BEP), with significant generation penalties (amounting to as much as 5 percent in total annual net generation).10 • Institutional transaction costs. The original Global Environment Facility (GEF) that funded many RE projects between 1995 and 2005 expressly recognized the importance of reducing institutional barriers. Experience over the past decade has shown that the costs of delay attributable to institutional dysfunc- tion have a major impact on economic returns: excellent examples are the mini hydro projects funded by the Philippines Rural Power Project (if the 2.5 MW Sevilla project implemented by the Boheco Rural Electricity coopera- ­ conomic tive had been built over two years rather than the actual four years, its e rate of return [ERR] would have increased from 21 percent to 26 percent) (World Bank 2013a). • Support for transmission integration. Twenty percent of China’s installed wind capacity is reported to be idle for lack of transmission connection or transmis- sion system bottlenecks. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 13 Organization of the Rest of the Report Chapter 2 presents the analytical framework that underpins the case studies, and provides the background for the principal research hypothesis of this report, which is better attention to the principles of economic analysis and market effi- ciency leads to more sustainable and effective policies. Chapters 3–10 present country case studies for Vietnam, Indonesia, Sri Lanka, South Africa, Tanzania, Egypt, Brazil, and Turkey. The conclusions of the study are presented in chapter 11. Each of the main issues presented above is discussed, and appropriate conclusions are drawn. The main lessons for those who design RE support mechanisms are clear and inescap- able. Successful RE policies: • Need to be grounded in economic analysis and the application of market principles to ensure economic efficiency. ­ • Will only be effective once the state-owned utilities that are the buyers of grid- connected RE are themselves in good financial health (in all of the case study countries, the power utilities are under financial duress). • Require a sustainable, equitable, and transparent recovery of incremental costs. Finally, some appendixes illustrate the application of useful techniques in eco- nomic analysis (taken from World Bank practice) that have been found effective in communicating analytical ideas to policy makers and stakeholder consultation meetings. Appendix A (setting RE targets in Croatia) illustrates basic tools from decision analysis; appendix B (multi-attribute decision analysis in Vietnam) shows how trade-off plots are useful in comparing RE generation with other options for reducing GHG emissions; and appendix C (estimating incremental costs in Indonesia) shows how RE supply curves can be used for estimating subsidies. Notes 1. See Vagliasindi (2013) for the overall report and Vagliasindi (2012) for the statistical analysis, which is based on panel data analysis. 2. The public-private partnership (PPP) requires careful definition. In some countries (for example, Indonesia) PPPs are simply independent power producers (IPPs) with sovereign guarantees. In others, PPPs imply equity contributions from a government or international financial institution (IFI) (for example, the International Finance Corporation, IFC). 3. It is often supposed that the incremental costs of the German feed-in tariff (FIT) are spread to all consumers in Germany through a levy, but power-intensive industrial consumers (and the railways) benefit from various degrees of exemption. See, for example, Neuhoff and others (2013). 4. For a detailed discussion of this point, see chapter 3 on Vietnam, section “Renewable Energy Resource Endowment: The Supply Curve.” 5. See Neuhoff and others (2013, 42). The transmission system operators publish the level of the FIT surcharge every October for the following year. But it may well be The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 14 Introduction noted that given the high level of the surcharge, public support is beginning to wane, and calls for a more competitive system are growing. 6. See, for example, conclusions of the Bank’s 2011 review of the design of policy instru- ments (Azuela and Barroso 2011). 7. As noted below, another reason for failure to reach targets is that the targets them- selves were established as political statements rather than being grounded in eco- nomic analysis, with the result that targets were in any event unachievable at a reasonable cost. 8. As we see in chapter 2, the financial supply curve for renewable energy (RE) lies above the economic supply curve, so all other things being equal, the point of intersec- tion with the avoided costs of thermal generation will be at a lower quantity of RE. 9. See chapter 3 for further discussion of this question: it is true that low-cost Chinese hydro turbines for small hydro projects often have lower efficiencies and higher out- age rates than their European counterparts, but these are far outweighed by the lower up-front costs (provided that equipment is sourced from reputable manufacturers). 10. See, for example, Kali Gandaki Hydro Rehabilitation Project Appraisal Document (World Bank 2013b). Bibliography Allegappan, L. R. Orans, and C.Woo. 2011.“What Drives Renewable Energy Development?” Energy Policy 39 (9): 5099–104. Azuela, G., and L. Barroso. 2011. “Design and Performance of Policy Instruments to Promote the Development of Renewable Energy: Emerging Experience in Selected Developing Countries.” Energy and Mining Board Discussion Paper 22, World Bank, Washington, DC. Carley, S. 2009. “State Renewable Energy Electricity Policies: An Empirical Evaluation of Effectiveness.” Energy Policy 37 (8): 3071–81. Cory, K., T. Couture, and C. Kreycik. 2009. Feed-in Tariff Policy: Design, Implementation and RPS Policy Interactions. National Renewable Energy Laboratory (NREL). Technical Report NREL/TP-6A2-45549. http://www.nrel​ .Gov/docs /fy09osti/45549.pdf. del Rio Gonzalez, P. G. 2007. “An Integrated Assessment of the Feed-In Tariff System in Spain.” Energy Policy 35 (2): 994–1012. ———. 2008. “Ten Years of Renewable Electricity Policies in Spain: An Analysis of Successive Feed-In Tariff Reforms.” Energy Policy 36 (8): 2917–29. Delmas, M., and M. Montes-Sancho. 2011. “U.S. State Policies for Renewable Energy: Context and Effectiveness.” Energy Policy 39 (5): 2273–88. Haas, R., M. Ragwitz, G. Resch, C. Panzer, S. Busch, and A. Held. 2011. “Efficiency and Effectiveness of Promotion Systems for Electricity Generation from Renewable Resources—Lessons from EU Countries.” Energy Policy 36 (4): 2186–93. Lauber, V. 2013. “Can Germany’s Energiewende Still Be Slowed Down?” Presentation made at 8th Internationale Energiewirtschaftstagung (IEWT), Technical University of Vienna, Austria, February 2013. Lesser, J., and X. Su. 2008. “Design of an Economically Efficient Feed-In Tariff Structure for Renewable Energy Development.” Energy Policy 36 (3): 981–90. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Introduction 15 Lipp, J. 2007. “Lessons for Effective Renewable Electricity Policy from Denmark, Germany and the United Kingdom.” Energy Policy 35 (11): 5481–95. Menz, F., and S. Vachon. 2006. “The Effectiveness of Different Policy Regimes for Promoting Wind Power: Experiences from the States.” Energy Policy 34 (14): 1786–96. Mitchell, C., D. Bauknecht, and P. M. Connor. 2006. “Effectiveness through Risk Reduction: A Comparison of the Renewable Obligation in England and Wales and the Feed-In System in Germany.” Energy Policy 34 (3): 297–305. Neuhoff, K., S. Bach, J. Diekmann, M. Beznoska, and T. El-Laboudy. 2013. “Distributional Effects of Energy Transition: Impacts of Renewable Electricity Support in Germany.” Economics of Energy and Environmental Policy 2 (1): 41–54. Rickerson, W., and R. C. Grace. 2007. The Debate over Fixed Price Incentives for Renewable Electricity in Europe and the United States: Fallout and Future Directions. Washington, DC: The Heinrich Boll Foundation. Sensfuß, F., M. Ragwitz, and M. Genoese. 2007. “The Merit-Order-Effect: A Detailed Analysis of the Price Effect of Renewable Electricity Generation on Spot Market Prices in Germany.” Fraunhofer Institute Systems and Innovation Research, Fraunhofer Working Paper 07/2007, Karlsruhe, Germany. Shrimali, G., and J. Kniefel. 2011. “Are Government Policies Effective in Promoting Deployment of Renewable Electricity Resources?” Energy Policy 39 (9): 4726–41. UNEP (United Nations Environment Programme), SEFI (Sustainable Energy Finance Initiative), and Bloomberg New Energy Finance. 2012. Global Trends in Sustainable Energy Investment 2012. Geneva, Switzerland: UNEP. Vagliasindi, M. 2012. “The Role of Policy Driven Incentives to Attract PPPs in Renewable- Based Energy in Developing Countries: A Cross-Country Analysis.” Policy Research Working Paper 6120, World Bank, Washington, DC. ———. 2013. Revisiting Public-Private Partnership in the Power Sector. Washington, DC: World Bank. Weitzman, M. 1974. “Prices vs. Quantities.” Review of Economic Studies 41 (4): 477–91. World Bank. 2013a. Implementation Completion Report, Philippines Rural Power Project. World Bank, Washington, DC. ———. 2013b. “Kali Gandaki Hydro Rehabilitation Project Appraisal Document.” Report No. 74861-NP, World Bank, Washington, DC. Yin, H., and N. Powers. 2010. “Do State Renewable Portfolio Standards Promote In-State Renewable Generation?” Energy Policy 38 (2): 1140–49. Zhang, F. 2013. How Fit Are Feed-in Tariff Policies? Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 2 The Economic Rationale for Renewable Energy Analytical Framework The economic rationale for renewable energy (RE) is straightforward: the ­ optimum amount of RE for grid-connected generation is given by the intersection of the RE supply curve with the avoided cost of thermal electricity generation figure 2.1). Very little RE will be competitive with the avoided thermal cost if (­ that cost is based on financial prices: in almost all Asian countries that have their own fossil-fuel resources, subsidized prices to power utilities are widespread. Only where the marginal thermal resource is imported (unsubsidized) oil is RE competitive (as was the case in Sri Lanka in the early 2000s); where the thermal generation price is based on coal, little if any RE is competitive. If thermal energy is correctly valued at the border price PECON (which equals = PFIN + a, the subsidy), then the optimal quantity of RE increases, as depicted in figure 2.2. These principles constitute the basis for the original avoided cost tariffs (ACTs) for RE in Sri Lanka, Indonesia, and Vietnam. In Sri Lanka, which has no domestic fossil resources, the marginal thermal production cost was set by imported diesel fuel, so the acceptance of an RE tariff set at this avoided cost was easily achieved in 1998. In Vietnam this was more difficult, since at the time of its introduction in 2009, the avoided financial cost of thermal generation to the state-owned utility (Electricity of Vietnam, EVN) was based on extensive subsi- dies to coal and domestic gas used for power generation. But as additional gas- fired combined-cycle-gas-turbine (CCGT) plants came online, with prices linked to international prices,1 EVN accepted a tariff based on the cost of the marginal thermal project. This is discussed further in chapter 3. But even if the cost of fossil energy is correctly valued at the border price, this needs to be further adjusted to reflect the local environmental damage costs of fossil energy—that is, the damage caused by local air pollutants (PM10,2 SOX,3 NOX4), or the environmental damage costs associated with coal mining (to the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   17   18 The Economic Rationale for Renewable Energy Figure 2.1 Economic Rationale for Renewable Energy: Optimal Quantity (QFIN) at Financial Cost of Thermal Energy (PFIN) Price Renewable energy supply curve Thermal energy cost (oil) PFIN Thermal energy cost (subsidized coal) Quantity QFIN Figure 2.2 Optimal Quantity (QECON) at the Economic Cost of Thermal Energy (PECON) Price Renewable energy supply curve Thermal energy cost, border price PECON Fuel subsidy, α PFIN Thermal energy cost, financial price Quantity QFIN QECON The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 19 Table 2.1 Externality Costs of Coal Generation Rand cents/kWh US cents/kWh Positive externalities 18.00 2.40 Negative externalities Combustion air pollution −1.35 −0.18 Biodiversity loss −0.70 −0.09 Acid mine drainage −2.10 −0.28 Fuel production health impacts (coal mining) −0.36 −0.05 Total negative externalities −4.51 −0.60 Net benefit 13.49 1.80 Source: Edkins and others 2010. Note: kWh = kilowatt-hours. extent these are not already reflected in the economic cost of coal supplied to a coal-burning project). As shown in the example of South Africa, in the case of coal these externality costs may be substantial (table 2.1). Nevertheless, there are also positive externalities to be included, which as shown in this table exceed the negative ­ externalities—these benefits derive mainly from the avoidance of the health effects from indoor air pollution associated with kerosene lighting and diesel self-­ generation. However, while these net benefits are relevant for evaluation of the no project alternative, when comparing coal with RE alternatives these same benefits also accrue to RE, so it is only the comparison of the negative externali- ties that matter. Such environmental damage costs represent real economic costs to the national economy, and their avoidance should be reflected as a benefit in the economic analysis of RE. In effect, the real social cost of thermal generation is its economic price (that is, without subsidy) plus the per kilowatt-hour (kWh) local environmental damage cost. As shown in figure 2.3, at this cost (PENV) = PECON + E, the economic quantity of RE increases further, to QENV . Just this framework was used to underpin the case for RE in China, as is sum- marized in figure 2.4. The quantity of additional RE increases from 79 terawatt- hours (TWh) to 89 TWh when the environmental damage cost of coal, estimated at 0.4 yuan/kWh (0.48 cents/kWh), is added to the economic cost of coal-fired generation.5 Appendix C shows how such supply curves can be used in practice to illustrate and estimate incremental costs. Local Environmental Damage Costs Table 2.2 summarizes estimates of the environmental damage costs of thermal projects in several developing countries. The difficulty with such aggregate damage cost estimates is that they are not transparent with respect to a whole range of important assumptions: the popula- tion affected, per capita income, the quality of the fuel (and the efficiency with The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 20 The Economic Rationale for Renewable Energy Figure 2.3 Optimal Quantity of Renewable Energy, Taking into Account the Environmental Damage Cost Price Renewable Thermal energy cost energy Including local env. supply curve damage cost PENV Local env. damage cost, ε PECON Fuel subsidy, α PFIN Thermal energy financial price Quantity QFIN QECON QENV Note: env. = environmental. Figure 2.4 The Economic Rationale for Renewable Energy: China 60 Supply curve, S 50 Q(BAU) = 36 TWh Q(ECON) = 79 TWh Q(ENV) = 89 TWh P[coal + env] = Price (fen/kWh) 40 36 f/kWh P[coal] = 32 f/kWh 30 20 10 0 20 40 60 80 100 120 Quantity (TWh) Source: Spencer, Meier, and Berrah 2007. Note: kWh = kilowatt-hours; TWh = terawatt-hours. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 21 which it is burnt), the height of the stack at which the pollutant is emitted, and the pollution control technology in place. Therefore, application of such aggre- gate per kilowatt-hour emission factors to any specific project comparison, or policy evaluation, can be very misleading. Compounding the difficulty, a signifi- cant part of the damage cost from the air pollutant is related to the cost of mortality—how to value the cost of human life is the key question. This is ­ recognized in the latest European Union (EU) studies, which show damage costs ­ based on two main methodologies: the value of statistical life (VSL), and years of life lost (YOLL).6 Thus, for example, the damage cost estimate per kilogram (kg) of PM-10 (particulate matter no greater than 10 microns in diam- eter) emissions in Germany varies from €28.9/kg using YOLL, to €81/kg using VSL (EEA 2011). Perhaps it is not surprising that even using the same methodology across all countries (or across provinces in the large countries), the damage cost estimates for specific pollutants vary widely. In both Europe and China (figure 2.5) regional variations in damage costs span an order of magnitude. EU and U.S. estimates of health damages are often scaled by per capita gross domestic product (GDP) figures, adjusted by purchase-power parity when transferred to developing countries (the so-called benefit-transfer method). Table 2.2 Local Externality Damage Costs in Selected Countries Cent/kWh Date of estimate Source India Coal 1.21 2010 See box 2.1 South Africa Coal 0.60 2010 See table 2.1 China Coal 0.1–1.0 2006 World Bank (2005) Indonesia Coal 0.32 2010 (1) Gas 0.087 2010 (1) Heavy fuel oil 2.2 2010 (1) Egypt, Arab Rep. Gas CCGT 0.03 2013 NOX only Note: (1) see box 5.2 (in chapter 5 of this report) for details. CCGT = combined-cycle gas turbine; CRESP = China Renewable Energy Scale-Up Program; NOX = nitrogen oxide. Box 2.1 The Renewable Energy Supply Curve in India A good example of an renewable energy (RE) supply curve is that prepared by a recent World Bank study for India (Sargsyan and others 2011). The production cost of coal is 5.65 cents/kWh (3.08 rupees [Rs]/kWh), to which is added the estimated local environmental damage cost of 1.21 cents/kWh, which intersects the RE supply curve at about 38 gigawatts (GW). The ­ additional global environmental premium is 2.24 cents/kWh (based on a carbon valuation of $32/carbon dioxide, CO2), which enables an additional 13 GW—to bring the total to 51 GW (see figure B2.1.1). This would constitute a rational basis for setting an all-India target for RE. box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 22 The Economic Rationale for Renewable Energy Box 2.1  The Renewable Energy Supply Curve in India (continued) Figure B2.1.1 Renewable Energy Supply Curve in India, by States and Energy Source 6 5 With global environmental Economic cost of generation, Rs/kWh premium 4.96 Rs/kWh Wind Rajesthan and West Bengal and 4 Wind Kerala and Maharashtra With local environmental Biomass Various states* premium 3.74 Rs/kWh Haryana Biomass Wind 3 Coal avoided cost Gujarat Biomass AP and MP 3.08 Rs/kWh Wind Wind and SHP TN and AP MP 2 All Karnataka states 1 0 0.11 37.8 51.0 67.9 Cumulative capacity (in GW) Note: SHP = small hydro project; TN = Tamil Nadu; MP = Madhya Pradesh; AP = Andhra Pradesh. $1 = Rs. 54.5. In the case studies presented in this report, some of the issues associated with such supply curves will be discussed in more detail. For example, India in particular suffers from low (and declining) average load factors in its wind projects, so gigawatt-hours rather than megawatts is the preferred unit of comparison. And different RE technologies also have very different capacity values, which require some adjustment to the RE cost if expressed simply as Rs($)/kWh. But whatever the difficulties, such an analysis is always a better basis for setting an RE target than mere political statement of aspirational goals. Source: Sargsyan and others 2011. Table 2.3 shows such an exercise for NOX emissions in the Arab Republic of Egypt, estimated at about 0.1 cent/kWh using the U.K. damage costs. Had the calculation been based on German damage costs, the estimate would be three times higher.7 The rationale for such adjustment is therefore doubtful. Figure 2.6 shows the relationship between damage cost estimates for NOX (as a/kg) versus per capita GDP for European countries. There is little evidence of correlation. The practice of scaling by per capita GDP would certainly not work within Europe, so there is little reason to suppose it would work across developing countries. These problems were recognized in a 2000 World Bank study that estimated health damage costs from air pollution across six major cities in developing countries. As shown in table 2.4, damage cost estimates varied by two orders of ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 23 magnitude across (a) ground-level emissions, typical of self-generation, and (b) large-scale utility projects, which have high stacks and are typically located in areas remote from densely populated cities. The damage cost estimates of table 2.3 are recalculated in table 2.5, using the average values for medium-stack-height emission factors (CCGTs rarely have the sort of high stacks used at coal projects). The damage cost per kilowatt-hour is one-tenth of the benefit transfer estimate listed in table 2.2. Figure 2.5 Variation in Damage Cost Estimates a. Europe (based on YOLL) Switzerland Germany Austria Hungary France Slovak Republic Slovenia Romania Croatia Czech Republic Italy Belgium Netherlands Moldova Poland Bosnia and Herzegovina Bulgaria Ukraine Belarus United Kingdom Lithuania Ireland Denmark Macedonia, FYR Albania Spain Latvia Sweden Turkey Estonia Greece Finland Portugal Cyprus Malta 0 2 4 6 8 10 12 14 16 18 20 2005 damage cost, 1000 Euro/Kg NOX Source: EEA 2011. Note: YOLL = years of life lost; kg = kilogram; NOX = nitrogen oxide. figure continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 24 The Economic Rationale for Renewable Energy Figure 2.5  Variation in Damage Cost Estimates (continued) b. China Xinjiang Xizang Qinghai Yunnan Heilongjiang Gansu Ningxia Jilin Guangdong Fujian Neimongu Hainan Liaoning Shaanxi Sichuan Hunan Shanxi Tianjin Hubei Henan Hebei Guizhou Jiangxi Shanghai Beijing Guangxi Chongqing Zhejiang Anhui Shandong Jiangsu 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 Yuan/kWh TSP SO2 NOX Source: World Bank 2005. Note: (in 2005, $1 = 8.25 Yuan). The total damage cost in Shandong of around 0.08 Yuan would be (in 2005) 0.97 cent/kWh. In Yunnan the damage cost is one-tenth of this, about 0.1 cent/kWh. NOX = nitrogen oxide; SO2 = sulfur oxide; TSP = total suspended particulates. Table 2.3  Damage Cost of NOX Emissions from Combined-Cycle Gas Turbines in the Arab Republic of Egypt Unit CCGT 1. NOX damage cost, utility emissions 2005 €/ton 5,181 2. NOX damage cost, utility emissions $/ton 6,735 3. Adjusted to 2013 prices $/ton 8,206 4. Emission factor gms/kWh 0.71 5. EU damage cost cents/kWh 0.6 6. PPP Euro zone, per capita GDP $ 35,657 7. Country PPP $ 7,057 8. Local damage cost $/ton 1,333 9. Egypt, Arab Rep., damage cost cents/kWh 0.095 Source: World Bank 2013. Note: CCGT = combined-cycle gas turbine; EU = European Union; GDP = gross domestic product; gms = grams; kWh = kilowatt-hours; NOX = nitrogen oxide; PPP = purchasing power parity. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 25 Figure 2.6  Damage Costs of NOX Emissions vs. Per Capita GDP in Selected European Countries 15 Germany Austria 2005 damage cost, 1,000 Euro/Kg NOX Hungary Slovak Republic France 10 Slovenia Romania Croatia Czech Republic Italy Belgium Netherlands Moldova Poland Bosnia and Herzegovina Bulgaria 5 Ukraine Belarus United Kingdom Lithuania Ireland Macedonia, FYR Albania Spain Denmark Latvia Sweden Turkey Estonia Greece Finland Portugal Malta Cyprus 0 0 20 40 60 Per capita GDP, $1,000 Source: Data from EEA 2011. Note: GDP = gross domestic product; kg = kilogram; NOX = nitrogen oxide. Three outliers—Switzerland, Norway, and Luxembourg—have been removed, as their economic conditions are unique in Europe. Table 2.4  Damage Cost Estimates ($/ton Emissions per Million People per $1,000 of Per Capita GDP Income) High stack (modern Medium stack (large Low stack (small boilers power plants) industry) and vehicles) PM-10  Range 20−54 63−348 736−6,435  Average 42 214 3,114 SO2  Range 3−8 10−56 121−1,037  Average 6 33 487 NOX  Range 1−3 3−13 29−236  Average 2 9 123 Source: Lvovsky and others 2000. Note: GDP = gross domestic product; PM-10 = particulate matter (no greater than 10 microns in diameter); NOX = nitrogen oxide; SO2 = sulphur dioxide. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 26 The Economic Rationale for Renewable Energy Table 2.5  Damage Costs of NOX Emissions from Combined-Cycle Gas Turbines in the Arab Republic of Egypt NOX Damage cost $/ton/million population/$1,000 GDP 9 GDP (PPP) $1,000/capita 7.1 Population Million 2 Cost per ton $/ton 127.8 Emission factor, CCGT gm/kWh 0.71 Damage cost Cent/kWh 0.009 Source: World Bank 2013. Note: CCGT = combined-cycle gas turbine; GDP = gross domestic product; gm = grams; kWh = ­kilowatt-hours; NOX = nitrogen oxide; PPP = purchasing power parity. The point is simply that there is high uncertainty in the cost estimates for local environmental externalities. This means that, in turn, targets for RE set on the basis of such estimates are also associated with similar uncertainties—though the impact in practice will also depend on the slope of the RE supply curve. Discount Rate Supply curves are based on a ranking of potential projects according to their levelized cost of energy, defined as: ∑ (1 + r ) Ci i i =1,..n LCOE = ∑ (1 + r ) E i i i =1,..n where r = Discount rate LCOE = Levelized cost of energy Ei = Net energy generation in year i Ci = Economic cost incurred in year i n = Economic life The levelized cost is thus critically dependent upon the choice of the discount rate. RE is generally more capital intensive than fossil energy, for which a greater part of the cost (of fuel) lies in the future. Consequently, the lower the discount rate, the more favorable RE appears by comparison—which is quoted by some as a reason for using lower discount rates when evaluating RE alternatives.8 Discount rates across countries vary: as shown in table 2.6, discount rates in the Bank’s RE project portfolio have varied from 8 percent to 15 percent. For example, in the Philippines the rationale for the high 15 percent discount rate (as used in the solar PV program) is that public sector projects ought not to crowd out private sector investment, and that therefore public sector hurdle rates (at least in the energy sector) should be higher than the typical weighted average cost The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 27 Table 2.6  Discount Rates in World Bank Renewable Energy Projects Country Rate (%) Renewable energy technologies evaluated Philippines 15 Solar homes (PV) Peru 14 Small hydro, solar homes (PV) (Peru Rural Electrification Project) India 12 Solar homes (PV), small hydro China 12 Small hydro, wind, bagasse, landfill gas Vietnam 10 Large and small hydro South Africa 10 Landfill gas, small hydro, pulp and paper cogeneration: Renewable Energy Market Transformation Project (carbon finance for renewables) Sri Lanka 10 Small hydro, wind, village (micro) hydro, solar homes Cape Verde 10 Wind Croatia 8 Biomass (combined heat and power), wind, small hydro Note: PV = photovoltaic. of capital (WACC) for private sector companies. If one argues that the optimal quantity of RE is given by the intersection of the RE supply curve with the avoided social cost of thermal generation (that is, including the cost of externali- ties), the same discount rate should be used for both sets of calculations. If the discount rate used in the least-cost expansion plan is 10 percent, one cannot justify a comparison with an RE option whose levelized costs are calculated on the basis of a 6 percent discount rate. Low discount rates should be used with caution, and should not be used merely as a substitute for attempting quantification of environmental impacts. At the same time, they do need country-by-country scrutiny. For example, almost every country that uses formal capacity expansion planning models (such as WASP or EGEAS) use 10–12 percent as the discount rate. At least in theory, the discount rate used for power sector planning should reflect the Government’s actual opportunity cost of capital (OCC)—which may or may not be 10–12 ­ percent as is often assumed. That such a rate is not always appropriate is illustrated by the recent example of the economic analysis for the proposed Noor II & III concentrated solar power (CSP) projects in Morocco (World Bank 2014), where the state-owned Morocco power utility ONEE has long used 10 percent (real) for its discount rate in its least-cost planning studies. One measure of the Government’s actual OCC is the cost of recent bond issues in foreign currency,9 for which a nominal rate of percent would be reasonable.10 Given an inflation assumption of 2 percent (for 6 ­ both Morocco inflation and trade-weighted FOREX), the corresponding real rate would be 4 percent. Now it might be argued that additional $2 billion bond issue earmarked expressly for CSP would require a somewhat higher coupon rate, a reasonable assumption for the real discount rate used for economic analysis would be 5 percent. That is significantly lower than the standard 10 percent assumption, and has a correspondingly large impact on the results. The main lesson here is simply that an economic analysis needs to examine a range of discount rates. Figure 2.7 illustrates the impact of the discount rate on Sri Lanka’s capacity expansion plan. Much as expected, at the lower discount rate of 8.5 percent, the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 28 The Economic Rationale for Renewable Energy Figure 2.7 The Impact of a Discount Rate on an Optimal Capacity Expansion Plan: Sri Lanka a. 10% discount rate 700 600 500 400 Net capacity additions, MW 300 200 100 0 –100 –200 –300 –400 09 10 11 12 13 14 15 16 21 23 24 25 26 27 28 17 18 19 20 22 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 b. 8.5% discount rate 1,000 900 800 700 600 Net capacity additions, MW 500 400 300 200 100 0 –100 –200 –300 –400 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Coal Hydro Renew Oil LNG Source: World Bank 2010a. Note: Negative numbers indicate retirements. LNG = liquefied natural gas; MW = megawatts. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 29 planning model chooses to build additional medium-scale hydro projects in 2014 and 2015, reducing the need for additional combustion turbine capacity. In the later years of the planning horizon, the model builds additional coal capacity, percent rather than the combustion turbines built in the reference case using a 10 ­ discount rate. The Social Cost of Carbon In economic analysis, the relevant global environmental premium is not the financial revenue that may be obtained from the sale of carbon credit on global carbon markets, but the global social cost of carbon (SCC) that reflects the actual damage costs of increasing atmospheric concentrations of greenhouse gases (GHGs). The literature on the SCC is growing, with estimates ranging from a small net benefit to costs of several hundred dollars a ton. Thus almost any estimate would find some support. Tol’s (2008) meta-analysis of peer-reviewed literature— which updated an earlier 2005 meta-analysis (Tol 2005)—cites 211 studies, and finds an average estimate of $120/ton carbon ($33/ton CO2) for studies pub- lished in 1996−2001, and $88/ton carbon ($24/ton CO2) for studies published since 2001. Tol concludes in the 2005 study that “it is unlikely that the marginal damage costs of emissions exceeds $50/ton carbon ($14/ ton CO2) and are likely to be substantially lower than that.” Much of the economics literature on the subject is highly technical, ­ particularly with respect to the choice of discount rate and assumptions about future global economic growth and income inequalities: in general one can say that the lower the discount rate, the higher the SCC (a value that may also change over time). The high valuation given in a report by Stern (2007) (“the current SCC might be around $85/ton CO2”) is largely a ­ consequence of the use of a very low discount rate.11 A 2007 Intergovernmental Panel on Climate Change (IPCC) report highlighted the wide range of SCC estimates, given in the literature as $4−$95/ton CO2. In the United States, regulatory impact analysis requires consideration of the SCC12 using a range of discount rates (from 2.5 percent to 5 percent), with carbon values that increase over time. For example, at a 5 percent discount rate the valuation is $12/ton in 2015, rising to $27/ton by 2050; at a 2.5 percent discount rate the valuation rises from $58/ton to 98$/ton by 2050. In the United Kingdom the Department of Environment recommended, in 2007, a value of £25/ton ($37/ton) CO213; this was subsequently updated to a time-dependent s ­ ystem ranging from £23/ ton CO2 in 2015 rising to £48/ton by 2025 ($36–$76/ton CO2). The World Bank—like other international financial institutions (IFIs), such as the Asian Development Bank (ADB) and African Development Bank AfDB)— does not publish an official estimate of the value of the SCC to be used in eco- nomic analysis. In the typical economic analysis of RE projects, recent practice has been to calculate the economic rate of return (ERR) with and without The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 30 The Economic Rationale for Renewable Energy consideration of GHG emissions. The choice of valuation is left to the economist assigned the task of estimating the economic returns (table 2.7). The approach taken in this study is not to choose any particular value for the SCC, but to calculate the avoided cost of carbon associated with a particular RE option. This is the value of CO2 that makes the cost of an RE project exactly equal to the least-cost thermal alternative. What is particularly important about such calculations is that they only have meaning relative to the option against which the RE is being compared. For example, in South Africa, where concentrated solar power (CSP) would be compared to coal (which has a high GHG emission factor), the avoided cost of carbon for CSP is much lower than in Egypt, where ­ the comparison is with natural gas, whose emission factor per net k ­ ilowatt-hour (in a highly efficient CCGT) is just one-third that of coal (table 2.8). In this report the calculations presented for the avoided cost of GHG emis- sions are based on discounted GHG emissions (using the same rate as costs and benefits are discounted): in this definition, the avoided cost, in $/ton, is defined as the value that must be given to a ton of avoided GHG emissions (i.e., a ­benefit) Table 2.7 Carbon Valuations in World Bank Studies and Project Appraisals Country $/ton CO2 Reference India 32 Sargsyan and others 2011 Indonesia 30 Geothermal project appraisal PAD Vietnam 30 Trung Son hydro project PAD Egypt, Arab Rep. 5−50 Wind Power Development Project PAD South Africa 29 Medupi coal project PAD Morocco 30 Ourzazate I CSP PAD Central Asia 13–43 CASA-1,000 transmission project PAD EEA 44 IPCC 4−95 Note: EEA = European Environment Agency; ton CO2 = ton of carbon dioxide; PAD = Project Appraisal Document (of the World Bank); CASA-1000 = HVDC transmission project to export summer hydro surplus from the Kyrgyz Republic and Tajikistan to Afghanistan and Pakistan. Table 2.8 The Avoided Cost of Carbon for Concentrated Solar Power Production cost, Carbon shadow Country Technology cents/kWh price, $/ton CO2 South Africa Medupi coal = least cost 5.8 0 CSP no storage, 25% LF 14.8 115 CSP storage, 40% LF 17.0 143 CSP storage Eskom estimate 17.9 155 Egypt, Arab Rep. Kom Ombo (1) (against Gas CCGT) — 267 Source: South Africa: World Bank 2010b, 2013. Note: CSP = concentrated solar power; CCGT = combined-cycle gas turbine; CO2 = carbon dioxide; Eskom = Electricity Supply Commission of South Africa; kWh = kilowatt-hours; LF = load factor; ton CO2 = ton of carbon dioxide; — = not available. (1) see chapter 8 for a detailed discussion of the Kom Ombo project. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 31 to bring the incremental costs of a RE option to the hurdle rate. It could also be termed the switching value as used in sensitivity analysis and risk assessment. Note that this differs to the term “marginal abatement cost” of GHG as used by the CTF and GEF, in which the net present value of costs and benefits (at the dis- count rate used) is divided by the undiscounted lifetime GHG emissions (i.e., assuming a zero discount rate). In general one should avoid arithmetic operations on quantities based on different discount rates. The calculation based on undiscounted emissions will typically be 50–60 percent lower than in the switching value definition.14 Fossil-Fuel Price Subsidies The impact of fuel subsidies is readily illustrated. Consider figure 2.8, which shows the demand for electricity, the RE supply curve, and the price of thermal energy in a competitive generation market, PCGM, assuming that the coal price is subsidized in the amount a. The quantity consumed at this price, Q, is given by the intersection of the demand curve with PCGM. The quantity of renewables will be R (namely that quantity whose production cost is less than PCGM), and the balance will be fossil generation, T (R + T = Q). Now suppose that the subsidy on domestic coal is removed, which increases the price to P*. At this higher price, the demand curve intersects at Q*. More RE will be economic at the higher price P*, and the quantity of fossil energy reduces to T* (R* + T* = Q*). Figure 2.8 Impact of Coal Price Subsidies Price Renewable energy supply curve P* α PCGM Demand Quantity R T Q R* T* Q* The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 32 The Economic Rationale for Renewable Energy Thus there are three important consequences of reducing the subsidy on coal: • Less electricity is consumed. • The amount of fossil energy, and hence GHG emissions, is reduced. • The amount of RE increases. It is easily shown (in box 2.2) that both social and global welfare increases as a result of the elimination of the subsidy: the reduction in fossil-fuel subsidies is a win-win. The International Energy Agency (IEA 2012) Energy Outlook15 estimates subsidies on energy consumption in the largest countries outside the Organisation for Economic Co-operation and Development (OECD) at $523 billion in 2011—almost $110 billion higher than in 2010, based on the IEA’s price-gap methodology (figure 2.9). This applies to several of this report’s case study countries (Egypt, Vietnam, Indonesia, South Africa, and Brazil). Most countries ­ Box 2.2 The Welfare Impacts of Fuel Subsidies The cost of a fuel subsidy to a government is Tα, equal to the area E + F + I + K + M. At the sub- sidized level of consumption Q, consumers enjoy a net benefit equal to the area under the supply curve less their cost, the so-called consumer surplus, equal to the area A + B + E + F + I + H + K. RE producers enjoy the producer surplus C. And GHG emissions are Tα where α is the relevant emission factor. Price Renewable energy supply curve A P* H α B E F I K M PCGM C Demand D G J L Quantity R T Q R* T* Q* box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 33 Box 2.2  The Welfare Impacts of Fuel Subsidies (continued) Once the subsidy is eliminated, the government benefits by the amount of that subsidy. The consumer surplus shrinks to A + H, but RE producers increase their surplus to C + B + E. So the balance of costs and benefits can be shown as in table B2.2.1. In other words, society gains (because the cost of the subsidy exceeds the increase in consumer surplus enjoyed under the subsidy), and the global environment gains (because ­ there is less fossil generation). Table B2.2.1 The Welfare Impact of Subsidy Removal With subsidy No subsidy Net impact Government (subsidy cost) −E − F − I − K − M 0 +E + F + I + K + M Consumers +A + B+ E + F + H + I + K A+H −B − E − F − I − K RE producers +C C+B+E +B + E Society A+B+C+H−M A+B+C+B+E+H +E + M Global environment Tα T*α α(T – T) Note: RE = renewable energy. Figure 2.9 Energy Subsidies, by Fuel, in Non-OECD Countries Iran, Islamic Rep. Saudi Arabia Russian Federation India China Venezuela, RB Egypt, Arab Rep. Iraq United Arab Emirates Indonesia Mexico Algeria Uzbekistan Kuwait Pakistan Thailand Argentina Ukraine Malaysia Qatar Kazakhstan Turkmenistan Bangladesh Ecuador Nigeria 0 10 20 30 40 50 60 70 80 90 US$ Billion Oil Natural Gas Coal Electricity Source: IEA 2012. Note: OECD = Organisation for Economic Co-operation and Development. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 34 The Economic Rationale for Renewable Energy have declared policies to eliminate these subsidies, but implementation is almost always slower than the announced schedule, for sudden removal of subsidies often has significant political repercussions. Vietnam is a case in point: notwith- standing the declared intention of removing subsidies for fuels used in power generation, and the commitment to market principles declared in the Electricity Law, both domestic coal and natural gas prices for power generation remain subsidized.16 Renewable Energy and Employment A widely cited benefit of RE is employment creation.17 But assessment of the benefits of increased employment requires caution. Economic analysis normally treats the cost of labor as an input, not as an output. A highly labor-intensive biomass technology may create local employment, but if the economic costs of the biomass project are above the avoided social costs, then employment in the economy as a whole may fall (because if households spend more on electricity, they will spend less on other goods and services). The argument that RE development will create “green jobs” is frequently heard in the United States, European countries, and some developing countries (such as China). Generalizations from limited country experience, mainly in RE equipment manufacturing in the OECD countries, are no substitute for careful country-specific analysis; more research is needed to better understand the issue of green jobs. The large employment benefits noted in such countries are a consequence of RE technology manufacture, particularly in countries that manufacture and export equipment, such as Spain and Denmark in the case of wind power (see box 2.3). So the question is: what is the extent to which these job gains apply to countries that do not have domestic manufacturing capacity for renewable generating equipment or reasonable prospects for ­ doing so? Another question that needs to be answered is whether such studies on the job creation benefits of RE also include the loss of jobs in those energy indus- tries that are displaced by the RE. In countries where there are large benefits from RE replacing coal, more RE could mean fewer coal miners, and lower employment in factories that manufacture gas turbines and coal-fired steam ­ generators. Within this context it is important to build a methodology to (a) contribute to a better understanding of the main effects/mechanisms to depict employ- ment impacts and (b) provide a clear definition of gross impact studies (­ sectoral) and net impact studies (economy wide). Even if some comparisons indicate that RE and energy efficiency projects generated more employment than fossil fuels, such comparisons fail to consider both the costs of delivery of equal outputs using different fuel mixes and the cost of public funds. Such simplifying assumptions may lead to misleading estimates. First, the lack of evidence on the cost of using alternative energy sources to generate the same The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 35 Box 2.3 Lessons Learned from the German Energiewende (Energy Transition) The ideas presented in a book titled Energiewende (Energy Transition) by Öko-Institut Freiburg found fertile soil in 1980 in Germany, in the context of anti-nuclear protests, two oil crises, fears of acid rain, and the emerging climate-change problem. It made the case for a change—a tran- sition from fossil fuels over to renewables and energy efficiency. This thinking inspired the Erneuerbare-Energien-Gesetz (EEG or the Renewable Energy Sources Act) of 2000 and its 2004 amendment. Both were adopted by Social Democratic-Green parliamentary groups against intense Conservative-Liberal opposition that came back to power in 2009. In the meantime, the EEG had a strong impact: (a) it was highly effective in stimulating RE with growth and invest- ment, increasing new renewable energy (RE) output (other than hydro) by a factor of five from 2000 to 2011 (see figure B2.3.1); (b) it created a strong wind power, biomass, and photovoltaic (PV) industry, which generated new employment for about 365,000 persons by 2012 (see figure percent of new RE generation capacity is owned by private per- 2.3.2, panel b); and (c) over 50 ­ sons and farmers (see figure B2.3.2, panel b). All utilities together own from 2 percent to 7 per- cent (PV, wind)—only for hydro does this go up to 90 ­ percent (see figure B2.3.2, panel a). There were three Conservative-Liberal attempts to slow down Energiewende in 2010, by scrapping the nuclear phase-out (a bridge technology for renewables), planning caps and steeper degressions for RE, and introducing a flexible cap for PV energy to limit new PV installations to 3 GW per year. In 2012 a plan for new, more drastic caps on PV and other technologies was implemented, and in 2013 there was a proposal to cap the EEG surcharge. From the Conservative-Liberal perspective several emerging issues needed to be addressed. First, the extraordinary deployment of PV since 2009, which surpassed all expectations and cost estimates; the dramatic increase in the EEG surcharge from €1.3 to Figure B2.3.1  Development of Renewables-Based Electricity Generation and Investment a. Trends in renewable-based generation (1990–2011) 1,40,000 EEG: 1,20,000 January 2009 1,00,000 EEG: EEG: April 2000 August 2004 80,000 Amendment to BauGB: GWh 60,000 November 1997 StromEinspG: 40,000 January 1991–March 2000 20,000 0 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 11 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 19 Geothermal Photovoltaics Biomass* Wind energy Hydropower box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 36 The Economic Rationale for Renewable Energy Box 2.3  Lessons Learned from the German Energiewende (Energy Transition) (continued) Figure B2.3.1  Development of Renewables-Based Electricity Generation and Investment (continued) b. Trends in renewable investment (2004–12) 25 Investment (€ billion) 20 15 10 5 0 04 05 06 07 08 09 10 11 12 20 20 20 20 20 20 20 20 20 Note: EEG = Renewable Energy Sources Act; GWh = gigawatt-hour. Figure B2.3.2  Development of Renewables-Based Jobs and Ownership, 2012 140,000 120,000 100,000 Job (number) 80,000 60,000 40,000 20,000 0 Wind Biomass Solar Hydropower Geothermal 2004 2007 2010 2011 2012 box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 37 Box 2.3  Lessons Learned from the German Energiewende (Energy Transition) (continued) Figure B2.3.2  Development of Renewables-Based Jobs and Ownership, 2012 (continued) Other electric utilities, 2.7 International Other, electric utilities, 1.2 2.7 Four largest utilities, 6.5 Private persons, 39.7 Industry, 9.3 Farmers, 10.8 Funds/Banks, 11.0 Project developers, 14.4 €5.28 cents from 2009 to 2013; an “imminent” grid congestion from PV and wind ­ technologies; damage to profitability of needed fossil-fuel generation due to the priority dispatch for RE, which implies that hard coal and gas plants lose lucrative operating hours (the noon peak demand is now increasingly covered by PV); and gas generation being affected by cheap coal due to low U.S. shale gas and Emission Trading System (ETS) prices. Source: Lauber 2013. output may lead to overestimating the net benefits of job creation by RE and energy efficiency projects (relative to fossil-fuel projects), by not including their cost. A rigorous methodology should first differentiate and illustrate trade-offs among (a) local, regional, and national impacts and (b) short- and long-run impacts. Second, it should illustrate the extent to which classic economic proj- ect analysis does not adequately reflect the employment-creation objectives of the government. Third, it must capture distributional impacts (since subsidies to cover incremental costs of RE may have very different beneficiaries) and employment-related externalities. Fourth, it might compare, where possible, alternative projects based on equivalent output and cost between The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 38 The Economic Rationale for Renewable Energy (a) renewable-energy and energy-efficiency projects and (b) fossil-fuel projects. To our best knowledge, no study reviewed to date compares projects costing the same amount (or producing the same output) along both employment and cost metrics. Of course, it is perfectly valid for governments to stimulate employment in disadvantaged areas of a country, even at the cost of lower employment in richer urban areas, if this is viewed in the context of social equity. But an impor- tant distinction might be noted: promoting employment in specific regions reflects the equity objective of the government, and not the economic-­ efficiency objective. The correct approach for economic analysis is to shadow-price labor costs.18 For example, the economic cost (or the opportunity cost) of employing otherwise unemployed rural workers is zero—and different to the actual wage rate that would be used in the financial analysis. But few RE projects will have much need for unskilled labor, whether during construction, or during operation (when unskilled labor, at best, would extend to the security staff at a wind farm or a CSP project). Specific Questions for the Case Studies The analytical framework requires that the effectiveness of incentive mecha- nisms be assessed and compared by a set of rational criteria, as follows: • Economic efficiency. How close is an RE support tariff to the avoided social cost of thermal energy (which for developing countries means economic cost + avoided local externalities)? How close is the target quantity to the economic optimum (intersection of the economic supply curve with the avoided social cost of thermal energy)? • Market principles. Does the design require the application of market princi- ples? An auction meets this criterion perfectly (provided there are safeguards against collusion and abuses). Access to a subsidy on the basis of first come, first served, or “all come” (as in Germany’s feed-in tariff, FIT) does not meet market ­ principles (and constitutes the worst possible way of providing access to support). • Transparency. Is the methodology of preferential pricing published? Can devel- opers and their lenders come to their own conclusions about the future evolu- tion of the tariff level? Does the mechanism provide for adjustment to changes in the law? • Sustainable recovery of incremental costs. Are the incremental costs known? (In a surprising number of cases they are not!) Is the mechanism for recovery of these costs sustainable (that is, is the mechanism for raising the necessary funds, and for disbursing them, seen as credible by developers and their lenders)? • Adaptability. Is there a predictable mechanism for updating the tariff and adjusting it for external changes (changes in technology costs, changes in tax rates, changes in fossil energy prices in the case of ACTs)? The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 39 The evaluation of the policy framework should similarly follow a set of ratio- nal criteria: • Targets. Are the targets set as political statements, or on the basis of a rational economic analysis (supply curve methodology, or affordability)? Are targets reasonably achievable, and are they in harmony with the support measures necessary to achieve them? In the case of renewable portfolio standards (RPSs) and mandatory renewable shares, are the penalties for not meeting them reasonable? • Energy subsidies. How does the additional quantity of RE—made economic by reducing fossil-fuel subsidies—compare with the quantity of RE to be sup- ported by RE incentives? How does the quantity of thermal energy that would not be required if retail tariff subsidies were eliminated compare with the quantity of RE supported by targeted RE incentives? Methodology For each of the case study countries, the following set of calculations will be presented: • From the current least-cost power sector development plan, the expected gen- eration mix for 2020, the generation shares and gigawatt-hours of each major fuel and technology, and the gigawatt-hours of retail sales. • Estimate of the consequences if 1 percent of generation were replaced by RE. What would be replaced is the most expensive of the thermal generation, by some RE whose tariff could be calculated to provide the developer with a target financial internal rate of return (FIRR) based on typical commercial lending rates. This allows a calculation of the total financial incremental costs—and the impact on consumers were this amount to be recovered from them. • An estimate of the tariff (and incremental cost) decrease prompted by the various incentives listed in table 1.1 (taxonomy of incentives): a clean develop- ment mechanism (CDM), carbon finance, subsidized loans from government- owned development banks, tax incentives, and so on. • A comparison of the impact on the consumer from reducing any fossil-fuel subsidies. Reducing fossil-fuel subsidies would increase the generation cost passed to the consumer, for which there is also a GHG emission reduction benefit. How does this compare with the cost to the consumer if the consumer is charged with a levy to recover RE generation costs? • Finally, a comparison of the residual incremental cost, as may need to be cov- ered from the direct government budget, with government spending on edu- cation and health (or some other appropriate indicator of spending for poverty alleviation). This question is core, as developing country governments are reluctant to incur the incremental cost of RE in the face of the overriding objectives of poverty alleviation and economic growth. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 40 The Economic Rationale for Renewable Energy Notes 1. Gas delivered to the Ca Mau CCGT project in Vietnam is indexed to the Singapore fuel oil price. 2. Particulate matter (no greater than 10 microns in diameter). 3. Sulphur oxides. 4. Nitrogen oxides. 5. The additional quantity QBAU is the much lower quantity of renewable energy likely to be implemented in the absence of explicit RE policy, due mainly to institutional and regulatory constraints (such as problems in negotiating PPAs, obtaining permits, and obstacles imposed by utility buyers who have traditionally opposed the emer- gence of IPPs for fear of losing market share). 6. VSL (value of statistical life) is used in most U.S. and European studies as a basis for mortality and is based on contingent valuation methods typical in American accident liability lawsuits. Most development economists argue that valuations based on YOLL (years of life lost) are more appropriate for the premature mortality typically associ- ated with pollution-aggravated respiratory diseases. 7. The damage cost of $1,333/ton NOX is consistent with the $473/ton (at 2002 prices) cited in the Bank’s 2003 Energy/Environmental Review (though the derivation of that estimate is unclear) (World Bank/EEAA 2003). 8. For example, a study on wind energy in Vietnam (Global Green Energy 2004) argues percent that “OECD uses a discount rate of 6 percent as standard, thereby justifying a 7 ­ rate” (rather than the 10 percent actually used by the Government of Vietnam). 9. In the case of an open economy, capital can be considered a tradable good, and the EOCC will be the world supply price of capital (U.S. treasuries, or long term LIBOR plus some country specific risk premium). Many developing countries now have domestic bond markets which can provide further information. 10. Morocco issued $500 million, 30-year 144a/Regulation S bonds in December 2012 at a coupon of 5.5 percent. The issue was reopened in May 2013 to increase the issue to $750 million for a tap of 237.5 basis points over U.S. Treasuries, and is currently trad- ing at a discount. As such, a nominal discount rate of 6 percent for modeling purposes seems reasonable. 11. For a good discussion of these issues, and a review of the assumptions in the Stern Review, see, for example, Hope and Newbery (2007). See also Grubb, Jamasb, and Pollitt (2008). 12. Interagency Working Group, Technical Update of the Social Cost of Carbon (SCC) for Regulatory Impact Analysis under Executive Order 12866, May 2013. 13. DEFRA, The SCC and the Shadow Price of Carbon, December 2007; Department of Energy and Climate Change, Carbon Appraisal in UK Policy Appraisal: A Revised Approach: A Brief Guide to the New Carbon Values and their Use in Economic Appraisal. 14. The rationale for not discounting GHG emissions is that it is the cumulative stock of GHG emissions in the atmosphere that matters, not the time at which it is emitted. However, there is an emerging consensus that the economic benefit of a ton of avoided GHG emissions increases over time as the concentration of atmospheric GHGs reaches the tipping point (recall the discussion of SCC, above, and the valua- tions being used by the U.S. Interagency Working Group on the SCC, and others). 15. The subject first received detailed analysis by the IEA in 1999 (IEA 1999). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 The Economic Rationale for Renewable Energy 41 16. Whether the use of subsidized fuel prices also distorts power sector investment deci- sions is unclear. An assessment of the use of financial prices rather than economic prices in Vietnam’s Sixth Power Development Plan found little impact on the optimal capacity mix as proposed by the plan (Economic Consulting Associates 2006). 17. An argument made, for example, by Kammen, Nozafari, and Prull (2012). 18. A related problem is the extent to which the cost of accidents and deaths to coal min- ers should be separately considered as an additional externality and added to the social cost of coal generation (as are included, for example, in the South African damage cost estimates of table 2.1). In economic theory higher occupational health hazards should be reflected in higher wage rates for miners, compared to other potential occupations that experience lower rates of occupational mortality, and hence do not classify as an externality. But this would be true only in a perfectly competitive and mobile labor market. For example, whether miners in the mining areas of northern Vietnam have real alternative employment opportunities (whether in the mining areas or elsewhere in Vietnam) may be debated. Bibliography Economic Consulting Associates. 2006. Economic Chapter of Power Sector Masterplan No.6. Report to the World Bank, Washington, DC, January. Edkins, M., H. Winkler, A. Marquard, and R. Spalding-Fecher. 2010. External Cost of Electricity Generation, Contribution to the Integrated Resource Plan 2 for Electricity. Report to the Department of Environment and Water Affairs, Energy Research Centre, University of Cape Town, South Africa. EEA (European Environment Agency). 2011. Revealing the Costs of Air Pollution from Industrial Facilities in Europe. Report EEA Technical Report 12/2011, Copenhagen, Denmark. Global Green Energy. 2004. Wind Farm Development in Vietnam, Prefeasibility Study: Developing Commercial Wind Energy Projects in Vietnam. New York, USA: Global Green Energy. Grubb, M., T. Jamasb, and M. G. Pollitt, eds. 2008. Delivering a Low Carbon Electricity System: Technologies, Economics and Policy. Cambridge, U.K.: Cambridge University Press. Hope, C., and D. Newbery. 2007. Calculating the Social Cost of Carbon. Cambridge University Electricity Policy Research Group, Cambridge, U.K. IEA (International Energy Agency). 1999. Insights, World Energy Outlook: Looking at Energy Subsidies: Getting the Price Right. Paris: IEA. ———. 2012. World Energy Outlook 2012. Paris: IEA. Kammen, D., M. Nozafari, and D. Prull. 2012. Sustainable Energy Options for Kosovo. Renewable & Appropriate Energy Laboratory Energy & Resources Group, University of California, Berkeley. Lauber, V. 2013. “Can Germany’s Energiewende Still Be Slowed Down?” Presentation made at 8th IEWT, Tech. Univ. Vienna, Vienna, Austria, February. Lvovsky, K., G. Hughes, D. Maddison, B. Ostrop, and D. Pearce. 2000. “Environmental Costs of Fossil Fuels: A Rapid Assessment Method with Application to Six Cities.” Environment Department Paper 78, World Bank, Washington, DC. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 42 The Economic Rationale for Renewable Energy Sargsyan, G., M. Bhatia, G. Banerjee, K. S. Raghunathan, and R. Soni. 2011. Unleashing the Potential of Renewable Energy in India. Washington, DC: World Bank. Spencer, R., P. Meier, and N. Berrah. 2007. Scaling Up Renewable Energy in China: Economic Modelling Method and Application. Energy Sector Management Assistance Program (World Bank) (ESMAP) Knowledge Exchange Series #11, June, Washington, DC. Stern, N. 2007. The Economics of Climate Change: The Stern Review. Cambridge, U.K.: Cambridge University Press. Tol, R. 2005. “The Marginal Damage Costs of Carbon Dioxide Emissions: An Assessment of the Uncertainties.” Energy Policy 33: 2064–74. ———. 2008. “The Social Cost of Carbon: Trends, Outliers and Catastrophes.” Economics e-Journal 2: 2008–25. World Bank. 2005. Economic Analysis for the China Renewable Energy Scale-Up Programme (CRESP). Washington, DC: World Bank. ———. 2010a. Sri Lanka, Environmental Issues in the Power Sector. Washington, DC: World Bank. ———. 2010b. Medupi Economic Analysis Background Report. Washington, DC: World Bank. ———. 2013. Kom Ombo CSP: Economic and Financial Analysis. Washington, DC. ———. 2014. Noor II and III Concentrated Solar Power Projects. Project Appraisal Document. Washington, DC. World Bank/EEAA. 2003. Egypt: Energy/Environmental Review. Report prepared by ERM, April, Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 3 Case Study: Vietnam Sector Background Vietnam has seen major economic growth over the past three decades, and significant progress in reducing poverty. It is a densely populated country that, in ­ the past 30 years, has had to recover from the ravages of war, the loss of financial support from the old Soviet Bloc, and the rigidities of a centrally planned econ- omy. The seeds of this expansion were planted more than two decades ago, with the 1986 launch of the renovation process known as Doi Moi.1 Vietnam has since witnessed a rapid transition to a globalized, market-based economy. The progress is reflected in the growth of electricity use and the electrification rate. In 1976, the first year after unification, just 2.5 percent of rural households had access to electricity; per capita electricity consumption was 45 kilowatt- hours (kWh)/year, and aggregate consumption was 2,300 gigawatt-hours (GWh). This rose to just 65 kWh/capita in 1985, before the start of Doi Moi. But since 1986, average growth has been a dramatic 9 percent: by 2009 consumption reached 72 terawatt-hours (TWh) and by 2012, 105 TWh—an increase of 11.25 percent over 2011 (Gencer and others 2011). Per capita annual consump- tion has grown from less than 50kWh/capita in 1976 to over 1000 kWh/year in 2013. While there is much uncertainty about future growth rates, especially in light of expected tariff increases, even at a modest 7 percent growth rate, 2020 sales should reach 180 TWh. The electrification rate is now close to 98 percent. Power Sector Development Two large hydro projects, Hoa Binh (1,920 megawatts, MW) and Yali (720 MW), both built with Russian assistance, provided the impetus for large-scale electrifi- cation of the country. Most of the domestic hydro resources are in the North (and in the Central highlands); all of the domestic coal (anthracite) resources are also in the North, while natural gas is exclusively in the South. These regional disparities were accommodated by the construction of the 500 kilovolt (kV) grid, which connected the major generating projects to the major load centers in Hanoi and Ho Chi Minh City (HCMC). Several large gas-fired combined-cycle The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   43   44 Case Study: Vietnam combustion turbines (CCCT) have also been built in the South over the past 15 years, some as independent power producers (IPPs). Figure 3.1 illustrates the least-cost capacity expansion plan by technology. Net capacity retirements are shown as negative entries. This shows that the projected increases in demand will be met by a combination of hydro, gas, and coal—with coal, in particular, playing an increasingly important role, albeit with increasingly sophisticated technologies new to Vietnam (such as supercritical pulverized coal). The domestic anthracite resource is nearing its end, and so the first imported coal-fired generating station is expected by 2017. Most of the large hydro projects, however, will be completed by 2017—the 2,400 MW Son La and the 1,200 MW Lai Chau projects are the last major hydro projects expected in Vietnam itself—and Vietnam is looking to several medium-sized hydro projects in the Lao People’s Democratic Republic to supply additional peaking power in the 2015−20 period. Beyond 2020 the main uncertainties include whether additional gas can be found to fuel combined-cycle gas turbines (CCGTs), whether the energy to gross Figure 3.1 Vietnam’s Capacity Expansion Plan 6 5 4 Net capacity additions, 1,000 MW 3 2 1 0 –1 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Oil, CT, other SHP/RE CFB CCGT Imports CCCT_LNG Hydro Imported coal Domestic coal Hydro PS Source: Meier 2011. Note: Negative values indicate plant retirements. CCCT = combined-cycle combustion turbine; CCGT = combined-cycle gas turbine; CFB = circulating fluidized bed; CT = combustion turbine; LNG = liquefied natural gas; PS = pumped storage; RE = renewable energy; SHP = small hydro project. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 45 domestic product (GDP) elasticity can be reduced, and whether (and when) Vietnam should commit to nuclear power. Renewable Energy Development Vietnam (like China) was one of the leaders in small hydro development long before the linkage of renewable energy (RE) to climate change, the principal motivation being remote rural electrification. But with the expansion of the national grid to even the most-remote provinces, most of these older off-grid small hydros were abandoned during the 1990s (see box 3.1). The seminal work on RE is the 2001 Renewable Energy Action Plan (REAP), a joint study by the World Bank, the Ministry of Industry and Trade (MoIT), and the electric utility (Electricity of Vietnam, EVN) (Bogach and others 2001). It called for new grid-connected small hydro projects (SHPs), community isolated hydro grids (to electrify up to 90,000 off-grid households), resource data devel- opment (especially for wind), 25,000−50,000 household scale systems (photo- voltaic [PV] and improved pico-hydro units), and extensive technical assistance (for example, in the development of a standardized power purchase agreement, SPPA) for grid-connected small power producers. The effort to promote off-grid electrification using RE must be judged a ­ failure—not least because of the much faster expansion of the national grid: by Box 3.1  Development of Small Hydro in Vietnam Small hydro development in Vietnam falls into five main phases: • 1960−75 and 1981−85 saw extensive construction of small hydro projects (SHPs) in remote areas to service mini-grids. Most were built with funds from the state budget for construc- tion of civil works, with equipment imported from China and Eastern Europe. • 1985−90 saw a diversification of the forms of investment: some projects were funded by the central budget, and many others were built by military units, cooperatives, and local com- munities (most with provincial assistance). • 1990−95: SHP development slowed down due to lack of investment capital for construction of new stations and lack of equipment and spare parts for replacement and repair. At the same time, the national power grid began to expand rapidly into rural areas. The use of pico- hydro units expanded greatly during this era—by some estimates as many as 150,000 such units, each less than 500 watts (W), had come into use. • 1995−2002: SHP development slowed down further, and as the grid expanded into remote areas previously served by SHPs, these were abandoned. Some 200 stations of 5−50 kilowatt (kW) capacity stopped operation, many at multipurpose facilities (power and irrigation). • 2002−present: There was significant expansion of larger grid-connected SHPs in the 2−30  MW size range, most developed by private developers, particularly provincial con- struction companies. Pico units have now all but disappeared. Source: MoIT 2011. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 46 Case Study: Vietnam 2008 just one district (Muong Te) in the remote province of Lai Chau remained unelectrified. Off-grid projects assisted by the World Bank, the Japan International Cooperation Agency (JICA), and the Swedish International Development and Cooperation Agency (SIDA) made little progress in the face of rapid escalation of civil costs and the difficulties of construction in remote areas. But the prospects for grid-connected small hydro were much better, and by the end of 2011, the Electricity Regulatory Authority of Vietnam (ERAV) esti- mated some 590 MW of small hydro had been connected to the grid, greater than the 175−251 MW target established by the Action Plan (see as a counter- factual the small hydro development in Lao PDR summarized in box 3.2).2 This compares to just a few thousand households electrified by SHPs in off-grid areas, compared to the 10-year target of 90,000−150,000 households. In 2009, with the assistance of the World Bank, the MoIT prepared a draft Renewable Energy Master Plan (REMP). This plan established targets and pro- posed to recover the incremental costs of RE by a consumer tariff levy funded Box 3.2 Counterpoint: Small Hydro Development in the Lao People’s Democratic Republic Small hydro development in the Lao People’s Democratic Republic serves as the perfect counterfactual for Vietnam. Notwithstanding the significant potential for small hydro, very little has been accomplished: as of the date of writing, the few grid-connected small hydro projects (SHPs) operating in Lao PDR have a total generation of just 11 megawatts (MW). To date, government  policy has been to promote the role of independent power producers (IPPs) through an incentive-based system. Potential developers carry out feasibility studies under a memorandum of understanding (MOU) with the Government of Lao PDR. Yet, while there are currently over 30 active MOUs across the country for small hydropower, few developers have been able to raise adequate finance. Some MOUs are held by speculators, holding attractive sites dormant. Concession agreements are power purchase agreements (PPAs) negotiated on a case-by-case basis. In 2012 efforts were made to introduce a new approach, following Vietnam’s example, with a standardized power purchase agreement (SPPA) and a published tariff, to be issued every five years, and inflation adjusted every year: the tariff proposed a time-of-day structure differentiated by season (compare the tariff in tables B3.2.1 and B3.2.2 with table 3.2 for Vietnam). The tariffs are linked to the utility’s (Electricity of Lao) 22 kV tariff, ranging from 95 percent of the tariff for 1 MW projects, to 80 percent for 10−15 MW projects. The SPPA and new tariff would be available to all SHPs no smaller than 15 MW in size. The approach calls for the government to prepare batch projects (with Asian Development Bank [ADB] assistance), and award to developers by competitive tendering (with awards based on highest royalty). The projects are designed for high heads and minimal storage and based on dry-season flows. The first four projects have the following design parameters: high plant fac- tors and high heads. Costs are in the range of $1,600–$2,100/kW (considerably higher than for Vietnam). box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 47 Box 3.2  Counterpoint: Small Hydro Development in the Lao People’s Democratic Republic (continued) Table B3.2.1 Standardized Tariffs Connection Jan–Jun Weighted capacity (MW) Jan–Jun peak off-peak Jul–Dec Peak Jul–Dec off-peak average 0–1 0.1324 0.0543 0.0659 0.0420 0.07631 1–5 0.1254 0.0518 0.0624 0.0398 0.07229 5–10 0.1067 0.0503 0.0611 0.0458 0.06827 10–15 0.1005 0.0474 0.0575 0.0431 0.06426 Note: Exchange rate: $1 = KN 8,055. KN = kip; MW = megawatts. Table B3.2.2 Project Characteristics Project Nam Hong Nam Hao Nam Long Nam Pe Province Units Borikhamxai Houaphan Houaphan Phongsaly 3 Average flow m /s 12 7 3 4 Design flow m3/s 6.1 4.3 1.8 1.6 Gross head m 237 210 285 380 Net head m 224 197 267 343 Canal length m 10,800 11,000 7,279 21,433 Penstock length m 530 520 966 938 Permanent road km 8.6 2.4 3.4 3.5 Transmission lines km 28 0.3 8.7 7.6 Generator output kW 11,660 7,100 3,800 4,500 Plant factor % 0.79 0.71 0.73 0.85 Production/year GWh/year 80 44.2 24.1 34 Overnight cost $ 18.8 12.4 7.3 9.2 Investment per kW $ 1,630 1,740 1,950 2,070 Levelized cost/kWh $ 0.037 0.044 0.048 0.043 Tariff per kWh $ 0.064 0.068 0.072 0.072 Note: GWh = gigawatt-hour; km = kilometer; kW = kilowatt; kWh = kilowatt-hour; m = meter; m3/s = cubic meters per second. As of the time of writing, the proposal for an SPPA and a published tariff has not been acted on, and they face opposition both from the developers and the utility. Yet without these essential reforms, it is hard to see substantive progress. Source: Anderson 2012. by a Renewable Energy Fund.3 But both the fund and the consumer levy did not find favor with the government, and the REMP has yet to be approved. Prior to 2009 small hydro tariffs were negotiated between developers and EVN on an ad hoc, project-by-project basis, with tariffs in the range of D575−D625/kWh (2.8–4.4 cents/kWh) (figure 3.2). The process did not work well, as developers gamed the system in the expectation of being negotiated down to a 12 percent rate of return on equity. January 2009 saw the start of a new system, with the introduction of an SPPA for qualified RE facilities not greater than 30 MW, which provided for an avoided cost tariff (ACT) to be published by the MoIT every year. In the recognition that such a tariff would be mainly of interest to small hydro, the tariff was expressly The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 48 Case Study: Vietnam Figure 3.2  Distribution of Tariffs, and Individually Negotiated Tariffs 8 Number of projects 7 6 5 4 3 2 1 0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 720 740 760 Tariff, D/kWh Source: MoIT Survey of Small Hydro developers 2007. Note: At the 2007 average exchange rate $1 = D15,740. kWh = kilowatt-hour. designed to reward daily peaking projects rather than pure run-of-river (RoR) projects, by providing a capacity charge for peak-hour dry-season production (table 3.1).4 This is calculated as the avoided capacity cost of a gas-fired CCCT. Other important characteristics include: • Regional differentiation (across the three main regions of Vietnam). The small regional variations in tariff arise because of differences in avoided transmission losses. For example, the output of an SHP in the North results in less gas-fired generation in the South (which is otherwise imported into the North across the 500 kV network), so the amount of generation reduction in the South (the benefit of RE) is larger than the injection of RE in the North when transmis- sion losses are taken into consideration. • Surplus energy charge (which applies to wet season energy produced in excess of monthly load factors greater than 90 percent). This was introduced as a concession to the distribution companies, which (under must-take provisions of the SPPA) are obliged to accept small hydropower in wet years even though they are already in surplus from large hydro. • Transmission connection. The SPPA requires that developers be responsible for the costs of the connection to the nearest substation, or to the nearest passing transmission line. With this tariff, a typical RoR SHP could in 2009 achieve an average tariff of just D563/kWh (3.2 cents/kWh)—not much of an increase compared to the old tar- iff regime (table 3.2). By 2010 this had increased to D686/kWh (3.3 cents/kWh). Just 20 percent of the total remuneration (row [20]) is from the capacity charge (a reflection of the low value of such projects to the EVN system). Table 3.3 summarizes the calculations for other types of RE projects. In 2012 a typical daily peaking hydro could achieve a tariff of 5 cents/kWh, as compared to just 3.3 cents/kWh for a RoR project. Realization from wind farms is just 4 cents/kWh—obviously not a level at which wind farms would be profitable. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 49 Table 3.1 The 2009 Avoided Cost Tariff, D/kWh Dry season (November−June) Wet season (July−October) Normal Peak Normal Surplus Peak hours hours Off-peak hours hours Off-peak energy [1] [2] [3] [4] [5] [6] [7] North 435 419 415 483 472 470 235 Center 403 411 418 418 427 439 220 South 428 427 426 453 451 447 223 Capacity charge 1,674 As cents/kWh North 2.5 2.4 2.4 2.8 2.7 2.7 1.3 Center 2.3 2.3 2.4 2.4 2.4 2.5 1.3 South 2.4 2.4 2.4 2.6 2.6 2.6 1.3 Capacity charge 9.6 Source: Decision No. 74/QD-DTDL, dated December 24, 2008. Note: At the 2009 exchange rate $1 = D17,490. kWh = kilowatt-hour. Table 3.2 Realization of the Avoided Cost Tariff, Run-of-the-River Project, North Vietnam 2009 tariff, 2009 revenue, 2012 tariff, 2012 revenue, GWh [%] D/kWh D (million) [%] D/kWh D (million) [%] Dry season Peak [capacity] 3,928 n.a. 1,674 6,575 21.9 1,805 7,090 19.4 Peak [energy] 3,928 7 435 1,709 5.7 619 2,431 6.7 Normal 10,214 19 419 4,280 14.3 596 6,088 16.7 Off-peak 4,714 9 415 1,956 6.5 554 2,612 7.1 Total 18,856 35 n.a. 14,520 48.4 n.a. 18,221 49.8 Wet season Peak 7,165 13 483 3,461 11.5 596 4,270 11.7 Normal 18,628 35 472 8,792 29.3 557 10,376 28.4 Off-peak 5,113 10 470 2,403 8.0 538 2,751 7.5 Surplus 3,485 7 235 819 2.7 269 937 2.6 Total 34,391 65 n.a. 15,475 51.6 n.a. 18,334 50.2 Total 53,247 100 n.a. 29,995 100 n.a. 36,555 100 Average load factor 43.4% Average, D/kWh 563.3 686.5 Exchange rate 17,490 20,690 Cents/kWh 3.2 3.3 Capacity charge, D m 6,575 7,090 Average capacity remuneration, D/kWh 123 133 (as % of total) 21.9 19.4 Average energy remuneration, D/kWh 440 553 Source: ERAV 2012. Note: n.a. = not applicable. kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 50 Case Study: Vietnam Table 3.3 Tariff Realizations for Different Types of Projects, 2009–12 2009 2010 2011 2012 Run-of-the-river small hydro D/kWh 563 565 646 687 Cents/kWh 3.2 3.0 3.2 3.3 Daily peaking small hydro D/kWh 877 870 979 1,029 Cents/kWh 5.0 4.6 4.8 5.0 Wind farm D/kWh 655 683 802 828 Cents/kWh 3.7 3.6 3.9 4.0 Rice husk gasifier D/kWh 668 677 794 862 Cents/kWh 3.8 3.6 3.9 4.2 Source: Meier 2013. Note: kWh = kilowatt-hour. By the end of 2011, 88 new projects for 572 MW had signed SPPAs under the new tariff, of which 30 projects (256 MW) were already in operation. The tariff itself was set on the basis of EVN’s avoided cost of the marginal thermal genera- tor (which is CCGT, as discussed further). This success in enabling small hydro at such low tariffs was made possible only by the extensive use of Chinese equip- ment, with costs about one-third less than that of European equipment. Despite the success of the ACT in enabling small hydro, as expected by the tariff’s designers, no wind projects have been enabled by the tariff; and only in late 2012 did two bagasse projects (at existing sugar mills) join the tariff. There is presently just one wind farm in operation in Vietnam, the 20 MW project in Binh Thuan.5 This has led to calls for feed-in tariffs (FITs) to support biomass and wind generation; several proposed biomass projects based on rice husk combustion have languished, awaiting a new biomass FIT. But much of the rice husk is already productively used, mainly as a heat source for brick making and ceramics. Following years of unrelenting pressure from donor advocacy (including count- less field trips to developed countries and several major studies rehearsing the arguments for a wind FIT), in June 2011 the MoIT issued a wind FIT of 7.8 cents/kWh.6 One cent/kWh was to be provided by the Vietnam Environmental Protection Fund (VEPF), the balance of 6.8 cents would be paid by the buyer (most wind resources are in the southern part of central Vietnam, and would be connected to the Central Power Company). This tariff has been widely criticized by developers as providing inadequate remuneration, and indeed no new wind projects have been enabled by the tariff.7 Nor is it clear whether the VEPF has a sustainable source of funding for this purpose. At the time of writing, the government is considering the issuance of a FIT for biomass, and is assessing the desirability of a FIT for solar PV. The government is also preparing a Renewable Energy Decree (or possibly a law) to codify the gov- ernment’s approach to supporting RE development. The absence of wind and biomass generation notwithstanding, it is worth noting that in 2009 Vietnam generated 39 percent of its electricity from RE. ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 51 While this includes large hydropower, members of the European Union (EU) also include large hydro in their RE targets (figure 3.3). In Vietnam this percent- age is set to decline with the significant increase in coal generation expected in the next 15 years, and by 2020 the fraction of electricity generated by RE will fall to 33 percent. But as evident from figure 3.4, even at this lower level, Vietnam’s renewables share will be greater than most EU countries, including Germany. Vietnam’s average power sector emissions per kilowatt-hour generated also compares well by international standards, as shown in table 3.4. Emissions are low because of the high proportion of hydro and gas in Vietnam’s power genera- tion mix. Figure 3.3  2020 RE Targets for Electricity Generation: Vietnam and the European Union Sweden Latvia Finland Austria Portugal Denmark Romania Estonia Slovenia Vietnam Lithuania France EU-27 Spain Germany Italy Greece Bulgaria Slovak Republic Poland Ireland Hungary Czech Republic Netherlands United Kingdom Belgium 0 10 20 30 40 50 60 70 80 90 2007 Actual 2020 Target Vietnam, 2020 target Source: Meier 2013. Note: EU = European Union; RE = renewable energy. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 52 Case Study: Vietnam Figure 3.4 The 2025 RE Supply Curve for Vietnam: Installed Capacity GW 2,500 2,000 1,500 D/kWh 1,000 500 0 1,000 2,000 3,000 4,000 5,000 Cumulative installed capacity, MW SHP Wind Wind Bagasse Pg Pfin LFG Ricehusk MSWI geothermal Pecon Source: MoIT 2011. Note: GW = gigawatt; kWh = kilowatt-hour; LFG = landfill gas; MSWI = municipal solid waste incineration; SHP = small hydro project. Table 3.4 CO2 Emissions per Kilowatt-Hour Generated in Selected Countries, 2008 gms CO2 /kWh South Africa 895 China 745 Indonesia 726 Malaysia 656 United States 535 Thailand 529 Global average 502 United Kingdom 487 Philippines 487 Germany 441 Sri Lanka 430 Vietnam 413 Source: IEA 2010. Note: Includes industrial process heat. gms CO2 = grams of carbon dioxide; kWh = kilowatt-hour. Renewable Energy Resource Endowment: The Supply Curve Other than for small hydro, for which a detailed master plan is available (PECC1 2001), and for agricultural wastes (biomass, which can be reliably inferred from official data on agricultural production), the other RE resources suitable for grid- connected projects are either largely unknown (if not quite speculative, as in the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 53 case of geothermal), too small to make any significant contribution (such as landfill gas), or vastly overestimated in light of existing evidence (as in the case of wind, where estimates of “physical potential” have little practical meaning). The small-hydro master plan identified some 2,925 MW in 408 potential SHPs in a size range of 5−30 MW, with average costs of $1,283/MW. This is somewhat less than the results of the 2007 MoIT survey that identified 3,443 MW in 319 projects that were at some stage in the project pipeline. How much of this capacity is actually economic is debatable, but according to the REMP, some 2,000 MW of small hydro was considered economic at EVN’s avoided financial cost. The size of the wind resource is subject to the usual meaningless estimates of gross physical potential. According to one early World Bank−supported study, the potential in areas identified as having average annual wind speeds of 7−8 meters per second (m/sec) is 102 GW (plus another 9 GW in areas of wind speeds > 8 m/sec) (TrueWind Solutions 2001). But the first official study in 2007 estimated the actual technical potential (at sites with wind speeds >6 m/sec at 60 meters above ground) at just 1,785 MW (EVN 2007). More recent assess- ments show as much as 5,200 MW has been proposed at various stages of the project pipeline, concentrated in just two provinces (Ninh Thuan and Binh Thuan) (table 3.5). Since the planning process in Vietnam encourages early Table 3.5 Status of Wind Power Development in Vietnam Status Province Number of projects Installed capacity IR IP TD UC IO 1 Lang Son 1 200 1 2 Quang Binh 3 n.a. 3 3 Quang Tri 1 30 1 4 Binh Dinh 3 251 1 2 5 Phu Yen 1 45 1 6 Dak Lak 2 n.a. 2 7 Gia Lai 1 41 1 8 Lam Ding 2 330 2 9 Ninh Thuan 16 1,105 9 6 1 10 Binh Thuan 20 1,541 17 1 1 1 11 Ba Ria 2 112 1 1 12 Tien Giang 2 150 1 1 13 Ben Tre 2 280 2 14 Tra Vinh 2 123 2 15 Soc trang 6 690 6 16 Bac Lieu 1 99 1 17 Ca Mau 2 300 2 Total 67 5,297 49 11 4 2 1 Source: Meier 2013. Note: Status: IR = investment report; IP = investment project; TD = technical design; C = under construction; IO = in operation. n.a. = not applicable. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 54 Case Study: Vietnam inclusion of projects in the officially approved list, this list is not necessarily an indication of actually bankable projects.8 Vietnam has a plentiful endowment of potential biomass, rice husk, rice straw, bagasse, coffee shells, wood fuel, and wood waste. An estimated 60 ­percent of the rural population uses biomass to meet the needs of heat for cooking, agricultural processing, and small-scale food production. In addition, biomass is used as fuel for producing heat and electricity in rural industrial production, such as rice husk for firing bricks, firewood or woody biomass for firing ceramics (in the South), and electricity and steam production in sugar mills from bagasse. Much biomass use is still based on old, outdated, and inef- ficient technologies. Several rice husk gasification and combustion projects are under development, but progress has been slow in the absence of an adequate tariff. None have been enabled under the existing subsidy scheme offered by the VEPF or the ACT. Vietnam has been successful in developing small-scale household biogas sys- tems, with nearly 200,000 biogas systems installed in households with an aver- age digester capacity of about 7−10 cubic meters (m3) per system, and used for cooking, lighting, and running small electricity generators. But the potential for biogas development in Vietnam remains significant—especially at the medium and large scale—not only to handle animal waste but also waste from food pro- cessing. Another important potential biomass source for fuel in Vietnam is rice straw (one ton of paddy produces 1 ton of straw) estimated at some 40 million tons in 2010, 54 percent of which is produced in the Mekong River Delta region. Bagasse is used for cogeneration of heat (steam) and power in sugar mills: the current installed capacity of cogeneration systems in all sugar mills is around 150 MW. But only three sugar mills—Son La sugar mill (Son La province), La Nga sugar mill (Dong Nai province), and Bourbon Tay Ninh sugar mill (Tay Ninh province)—are selling their surplus electricity to the grid. The Bourbon sugar mill has installed the largest bagasse cogeneration plant in Vietnam, with a capac- ity of 24 MW, of which 9−10 MW is used for the sugar mill, and the rest is sold to the grid. Several new projects are under development but, absent an appropri- ate support tariff, progress has been slow. The most careful and detailed assessment of the RE resource for grid-­ connected energy generation is found in the 2008 Draft Renewable Energy Master Plan, which evaluated all projects in the plausible development pipeline (except for small hydro, whose resource potential was extrapolated from the small hydro master plan). Figure 3.4 shows the 2025 supply curve for installed capacity: about 2,400 MW is enabled by the 2008 estimate of economic cost, of which 220 MW is available at the avoided financial cost (that is, at the ACT). Just a single 30 MW wind farm is economic when the carbon benefit is included (table 3.6). The corresponding supply curve in terms of generation is shown in figure 3.5 and table 3.7. At the 2008 estimate of the avoided economic cost (PECON), some 11 TWh are economic by 2025, likely to be around 5 percent of generation. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 55 Table 3.6 Economically Optimal Renewable Energy: Installed Capacity MW enabled by ACT MW enabled by MW enabled by CDM (PFIN ) support to PECON (PG ) Total SHP 2,030 0 0 2,030 Rice husk 0 0 69 69 Bagasse 185 65 0 250 LFG 52 0 0 52 MSWI 0 0 0 0 Geothermal 0 152 0 152 Wind 0 0 30 30 Total 2,267 217 99 2,583 Source: MoIT 2011. Note: ACT = avoided cost tariff; CDM = clean development mechanism; LFG = landfill gas; MSWI = municipal solid waste incineration; MW = megawatt; SHP = small hydro project. Figure 3.5 The 2025 RE Supply Curve for Vietnam: Generation 2,500 2,000 1,500 D/kWh 1,000 500 0 1,000 2,000 3,000 4,000 5,000 Cumulative installed capacity, MW SHP Wind Wind Bagasse Pg Pfin LFG Rice husk MSWI geothermal Pecon Source: MoIT 2011. Note: kWh = kilowatt-hour; LFG = landfill gas; MSWI = municipal solid waste incineration; MW = megawatt; RE = renewable energy; SHP = small hydro project. Based on this assessment of the supply curve, the REMP developed four scenarios: • Scenario 1: economic quantity of grid-connected renewables, household electrification completed in 2025. ­ • Scenario 2: economic quantity of grid-connected renewables, rural electrifica- tion completed in 2020. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 56 Case Study: Vietnam Table 3.7 Economically Optimal Renewable Energy in Vietnam Generation (GWh): Projected to 2025 Generation at PFIN Generation at PECON Generation at PG SHP 8,814 8,814 (81.7%) 8,814 Rice husk 0 0 299 Bagasse 667 896 (8.3%) 896 LFG 214 214 (2.0%) 214 MSWI 0 0 0 Geothermal 0 862 (8.0%) 862 Wind 0 0 91 Total 9,694 10,785 (100%) 11,175 Source: MoIT 2011. Note: GWh = gigawatt-hour; LFG = landfill gas; MSWI = municipal solid waste incineration; SHP = small hydro project. • Scenario 3: scenario 1 + wind development demonstration program (630 MW by 2020). • Scenario 4: scenario 1 + all identified grid-connected renewables (maximum grid-connected potential). The main assumptions for the evaluation of incremental costs included a wind FIT of 10 cents/kWh, no subsidy for SHPs, and biomass and geothermal at the estimated cost of generation (capped at PECON of D1,200/kWh). The incremen- tal costs were assumed to be raised by a consumer levy. The REMP assumed that the electricity levy would also be used to fund the off-grid rural electrification program,9 and to fund subsidies for household-scale biogas and solar water heating. Figure 3.6 shows the results of scenario 3 (economic quantities plus a ­ wind demonstration program of 630 MW by 2020). Subsidies for rural electrifi- cation peak in the period 2015−20, then fall off as the goal is achieved (here in 2025). The subsidy for wind is about $125 million per year, but the impact on the tariff is just D20/kWh (0.1 cent/kWh). One reason for the low tariff levy is the very high demand forecast used in 2008: that is, 128 TWh in 2012 (compared to the actual 2012 consumption of 105 GWh), though this is partially offset by underestimating the consumer tariff (2012 estimate of D1,012/kWh compared to the actual D1,361/kWh). When the levy is recalculated under currently more realistic assumptions, it more or less doubles to around D40/kWh (0.2 cents/kWh). Production Costs Wind For the baseline estimate of wind production costs, we make the following assumptions: • Capital cost $1,750/kW (based on the 2011 Institute of Energy, Vietnam [IoE] assessment of Chinese wind turbine prices; wind farms based on European turbine prices are assessed at $2,250/kW).10 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 57 Figure 3.6 Impact of the REMP Development Scenario: Economic + Wind Demonstration Program, 2009–25 a. Electricity levy, USc/kWh 0.10 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 11 09 10 12 13 19 20 21 22 23 18 24 14 15 16 25 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 b. Support requirements, US$ million 250 200 150 100 50 0 11 18 20 21 22 23 09 10 12 13 19 24 14 25 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Heat O -grid Grid-connected Source: MoIT 2011. Note: USc = U.S. cents. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 58 Case Study: Vietnam • Average annual plant load factor of 26.7 percent (based on the IoE evaluation of wind speed data of 7 meters per second [m/sec] at 85 meters hub height). • Financing 25 percent equity/75 percent debt (the state-owned Vietnamese Development Bank stipulates a minimum of 15 percent equity, but commer- cial banks would want to see much higher equity). • Local commercial financing: end 2012 Vietnamese prime rate (14 percent) + 2 percent = 16 percent; term 7 years including 2 years grace (during construction). • Balance sheet financing (so project debt service cover ratio [DSCR] is not a binding constraint). • Annual operation and maintenance (O&M) costs (2 percent of capital cost). • Two-year construction period. • Domestic inflation 6 percent, dollar inflation 2 percent, and exchange rate depreciation 3.9 percent.11 • Target financial internal rate of return (FIRR) after tax = 15 percent (nominal). • Corporate tax rate is 28 percent, depreciation over 20 years (no tax holidays or accelerated depreciation).12 The model calculates the required tariff in cents/kWh to meet this target FIRR, under the assumption that the U.S.-denominated tariff remains constant (that is, adopting the way in which the present wind FIT is adjusted). The required tariff under the above assumptions is 11.3 cents/kWh. For European wind turbine costs, the required tariff is 14.6 cents/kWh.13 As shown in figure 3.7, it is evident that the current wind FIT of 7.8 cents would require an annual load factor of 38 percent using Chinese equipment, and almost 50 percent using European equipment. Such load factors are simply not achievable given the actual wind regime. Table 3.8 summarizes wind power support tariffs in other Asian countries. Note the wide range in support levels, from Vietnam’s 7.8 cents/kWh to 24 cents/kWh in the Philippines. This wide range cannot be explained by the differences in assumptions about the cost of wind turbines, which is certainly much less than a factor of 3. Chinese wind turbines may cost 70−80 percent as much as machines from European turbine manufacturers, but not as little as 20−25 percent. This illustrates the difficulty of the government setting a produc- tion-cost-based FIT in the face of great information asymmetry: government officials only rarely have access to the relevant current market information, so assumptions about production costs are often little more than guesses or judg- ments as to what might be reasonable. Capacity Value The calculations above simply reflect the tariff required of the developer. But this is not necessarily the same as the cost to the buyer, because wind energy is not dispatchable, and contributes little reliable power during peak hours. In short, it has little capacity value.14 The general rule of thumb is that the capacity credit The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 59 Figure 3.7 Required FIT to Maintain 15 Percent FIRR vs. Capital Costs 25 20 USc/kWh 15 10 5 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 Annual load factor $/kW = 2,250 [European] $/kW = 1,750 [Chinese] 7.8 USc/kWh FIT Source: Meier 2013. Note: FIT = feed-in tariff; FIRR = financial internal rate of return; kW = kilowatt; kWh = kilowatt-hour; USc = U.S. cents. Table 3.8  Wind Support Tariff Comparisons Assumed load factor Cents/kWh Vietnam 26.9% 7.8 Thailand 17.6a China 30–35.6% 8.1–9.7 Philippines Not provided 24 Sri Lanka 36% ~20 (in 2009) 15 (2012 proposal) Note: kWh = kilowatt-hour. a. 11 cent adder + 6.5 off-peak base rate. may be approximated by the ratio of annual average capacity factors. If a wind project of x MW with a load factor of 25 percent displaces coal generation with a load factor of 75 percent, it should be given a capacity credit γ = 0.33x [MW]. This means that the capital cost of the wind project should be burdened with the fixed costs of open-cycle gas turbine (OCGT) capacity (the cheapest form of The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 60 Case Study: Vietnam capacity) in the amount of (1 – γ) [MW]. The British Wind Energy Association (BWEA) estimates that the capacity credit for wind is around 33 percent at 5 percent wind penetration, falling to 20 percent at 15 percent penetration.15 As shown in box 3.3, studies in China showed capacity values for wind and small hydro of 43 percent and 47 percent, respectively. Therefore, a Vietnamese wind power project with a 26.8 percent load factor has a capacity credit of 32 percent. With OCGT at $350/kW, this means that EVN incurs an additional capital cost of $239/kW, or 1.192 cents/kWh Box 3.3 The Capacity Value of Renewables in China Rules of thumb are all very well, but do they have any basis in reliable studies? The only way the capacity impacts can be realistically assessed is in a capacity expansion optimization model, in which the least-cost plan is perturbed by forcing in renewable energy (RE) and evaluating how much thermal capacity is actually avoided (or deferred). There are few such studies; one was part of the economic analysis conducted for the China Renewable Energy Scale-up Program (CRESP) project in China. In an initial modeling study, the impacts of a wind development plan of 2,600 MW of additional wind capacity over 10 years in the North China grid were assessed: as shown in the figure below this resulted in a displacement of 836 MW of coal and 256 MW of oil-fired combined-cycle gas turbine (CCGT)—in effect a capacity credit of 43 percent (see figure B3.3.1). ­ A second modeling study examined 1,000 MW of additional small hydro in the Zhejiang grid: this resulted in a 402 MW decrease in coal capacity, and 60 MW in oil-fired CCCT, a capacity credit of 47 percent (see figure B.3.3.2). ­ Figure B3.3.1 Capacity Displacements in the North China Grid 3 Change in installed capacity, GW 2 1 0 –1 –2 Coal Nuclear Oil CCCT Hydro Small hydro Wind Source: World Bank 2005. Note: CCCT = combined-cycle combustion turbine; GW = gigawatt. box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 61 Box 3.3  The Capacity Value of Renewables in China (continued) Figure B3.3.2 Capacity Displacements in Zhejiang 1,500 Change in installed capacity, GW 1,000 500 0 –500 Coal Nuclear Oil CCCT Hydro Small hydro Wind Source: World Bank 2005. Note: CCCT = combined-cycle combustion turbine; GW = gigawatt. Table 3.9 Capacity Penalty, Wind Project with Annual Load Factor of 26.9 Percent Units 1 OCCT cost $/kW 350 2 Capacity credit [proportion] 0.32 3 Capacity penalty $/kW 239 4 Capital recovery factor [proportion] 0.12 5 Annual cost $/year/kW 28.1 6 Annual generation kWh/year 2,356 7 Cost per kWh Cents/kWh 1.192 Source: ERAV 2012. Note: kW = kilowatt; kWh = kilowatt-hour; OCCT = open-cycle combustion turbine. (table 3.9). This additional cost needs to be considered in incremental cost calculations. Detailed studies of how RE projects are operated in Vietnam provide addi- tional insights. Figure 3.8 shows the average monthly dispatch for each of the three tariff blocks (peak, normal, and off-peak) for the 12 MW Nam Mu daily peaking SHP. This shows that even during the dry season, the average monthly dispatch during peak hours is around 8 MW; during the system peak in November it is 10 MW. During the wet months of July−August, the plant runs more or less at its full capacity of 12 MW throughout the day. Of course, there is little or no generation in the dry season during off-peak hours—but the economic motiva- tion to build daily peaking capacity rather than pure RoR is clear. And, clearly, Nam Mu is dispatchable and has significant capacity value.16 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 62 Case Study: Vietnam Figure 3.8 Operation of the Nam Mu, Daily Peaking, Small Hydro Project 14 Installed capacity = 12.0 MW 12 10 8 MW 6 4 2 0 ch ne t r r y y ay ly r r ril us be be be be ar ar Ju ar Ju M Ap g nu ru m m em to M Au b ve ce Oc Ja pt Fe De No Se Peak Off-peak Normal Source: ERAV 2012. Note: MW = megawatt. This may be compared to the analogous evaluation of a wind project proposed for Ly Son Island, (poorly) served by old diesels (figure 3.9).17 The output is strongly seasonal: just 2 MW on average for most of the year, and just 7 MW in the peak month (December) and a little less than 5 MW in January (for an annual average load factor of 22 percent). In short, such a project has very little capacity value, especially when compared to daily peaking hydro. For small hydro, then, the capacity value is much more significant. For 2009 we have examined monthly dispatch data for a portfolio of 11 SHPs for which we have individual hourly data, representing an installed capacity of 89.5 MW ­ (figure 3.10). With heavy representation of central region projects, the contribu- tion of the portfolio to the coincident peak month is high:18 the average dispatch in the peak hours of this month is 82 MW, and the average peak-hour contribu- tion in the dry season is 62 MW. Even in the dry season of January−April, the portfolio has a capacity value of around 50 MW for the five peak hours of the day. The same is true for the EVN system as a whole, for which we have aggregate hourly data from the regional load dispatch centers (but not at the individual project level). This shows significant contribution during the peak hours of the dry season throughout the year, and is a clear demonstration of the capacity value of a portfolio of SHPs. Strictly speaking, capacity value can be gauged by the contribution to the November coincident system peak: as is clear from figure 3.11, the SHP portfolio as a whole contributes about 85 percent of its ­ installed capacity to the November system peak. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 63 Figure 3.9 Operation of the Proposed Ly Son Island Wind Project 14 Installed capacity = 12.0 MW 12 10 8 MW 6 4 2 0 ch ne st y ry ay ly r r r r ril be be be be ar gu Ju ua ar Ju M Ap nu em to m m M Au br ve ce Oc Ja Fe pt De No Se Peak O -peak Normal Source: ERAV 2012. Note: MW = megawatt. Figure 3.10  Dispatch of a Portfolio of Small Hydro Projects 100 Installed capacity = 89.5 MW 80 60 MW 40 20 0 ry y ch ril ay ne ly t r r r r be be be be us ar Ju Ap ua M ar Ju g ru m to m m M n Au b te ve ce Oc Ja Fe p De No Se Peak O -peak Normal Source: ERAV 2012. Note: MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 64 Case Study: Vietnam Figure 3.11 Average Contribution of Small Hydro during the Five Peak Hours of the Day, 2009 400 350 300 250 MW 200 150 100 50 0 ne ch ay ly r r y st ry r r ril be be be be ar Ju gu ua ar Ju M Ap nu to m m em M br Au ve ce Oc Ja Fe pt De No Se Source: National Load Dispatch Centre (NLDC) daily reports. Note: MW = megawatt. Biomass A technology-based FIT for biomass has two critical assumptions: capital cost and fuel cost. The MoIT will need to make estimates for both to be able to issue a technology-specific FIT. Table 3.10 shows the tariff required to achieve a 15 percent FIRR (posttax), under the following assumptions: • Rice husk combustion: 20 MW. • No carbon revenue, or ash sales. • Domestic inflation, 6 percent; Organisation for Economic Co-operation and Development (OECD) inflation, 2 percent. • Two-year construction time. • Thirty percent equity, 70 percent debt (domestic loan, 16 percent interest, 7 years including a 2-year grace during construction, interest during construc- tion [IDC] capitalized). • Own use: 9 percent. • Fuel rate: 1.5 kg/kWh. • No real escalation in rice husk prices (that is, increase at assumed rate of domestic inflation). • First year fixed operation and maintenance (O&M) costs: 4 percent of capital costs, increasing to 5 percent in year 6 and 7 percent in year 11. • Annual tariff adjustment to allow for exchange rate depreciation (similar to the wind FIT mechanism). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 65 Table 3.10 Tariff Required for 15 Percent FIRR (Post Tax) Rice husk price, $/ton Capital cost, $/kW 18 20 22 24 26 28 30 32 34 36 1,200 6.36 6.75 7.13 7.52 7.91 8.30 8.69 9.07 9.46 9.85 1,300 6.60 6.98 7.37 7.76 8.15 8.54 8.93 9.31 9.70 10.09 1,400 6.83 7.22 7.61 8.00 8.39 8.78 0.16 9.55 9.94 10.33 1,500 7.07 7.46 7.85 8.24 8.63 9.01 9.40 9.79 10.18 10.57 1,600 7.31 7.70 8.09 8.48 8.86 9.25 9.64 10.03 10.42 10.81 1,700 7.55 7.94 8.33 8.71 9.10 9.49 9.88 10.27 10.66 11.04 1,800 7.79 8.18 8.56 8.95 9.34 9.73 10.12 10.51 10.89 11.28 1,900 8.03 8.41 8.80 9.19 9.58 9.97 10.36 10.74 11.13 11.52 2,000 8.27 8.65 9.04 9.43 9.82 10.21 10.59 10.98 11.37 11.76 2,100 8.50 8.89 9.28 9.67 10.06 10.45 10.83 11.22 11.61 12.00 Feasible at the same tariff as wind (7.8 USc/kWh) Feasible at the avoided social cost of thermal generation (10.8 USc/kWh) Requires tariff higher than the avoided social cost (uneconomic) Source: ERAV 2012. Note: kW = kilowatt; kWh = kilowatt-hour; USc = U.S. cents. Table 3.11 Tariff Required (in Cents): 3 Percent Annual Real Price Escalation for Rice Husk First year rice husk price, $/ton Capital cost, $/kW 18 20 22 24 26 28 30 32 34 36 1,200 7.3 7.8 8.3 8.8 9.3 9.8 10.3 10.8 11.3 11.8 1,300 7.6 8.1 8.6 9.1 9.6 10.1 10.6 11.1 11.5 12.0 1,400 7.8 8.3 8.8 9.3 9.0 10.3 10.8 11.3 11.8 12.3 1,500 8.1 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 1,600 8.3 8.8 9.3 9.8 10.2 10.8 11.3 11.8 12.3 12.8 1,700 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 1,800 8.8 9.3 9.8 10.3 10.8 11.3 11.7 12.2 12.7 13.2 1,900 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 2,000 9.2 9.7 10.2 10.7 11.2 11.7 12.2 12.7 13.2 13.7 2,100 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 Feasible at the same tariff as wind (7.8 USc/kWh) Feasible at the avoided social cost of thermal generation (10.8 USc/kWh) Requires tariff higher than the avoided social cost (uneconomic) Source: ERAV 2012. Note: kW = kilowatt; kWh = kilowatt-hour; USc = U.S. cents. This shows, for example, that at $1,800/kW and a $22/ton rice husk price, a tariff of 8.56 cents/kWh is required to achieve the 15 percent FIRR. But the experience of Thailand shows that rice husk price escalation is a major risk once a support tariff is issued: if the real rate of rice husk price escalation is 3 percent, the baseline tariff increases to 9.8 cents/kWh, as shown in table 3.11. The discussion about biomass targets for electricity generation and the FIT required to enable bankable projects is often ill informed. From the perspective of reducing carbon emissions, it matters little whether it is achieved by electricity The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 66 Case Study: Vietnam generation or by use as a fuel for heat—it is only important that agricultural waste is not burnt or dumped into rivers. A significant proportion of rice husk is used as a fuel for brick-making, at rice mills, and in the rural ceramics industry— in all cases it replaces oil as a fuel. Moreover, if used for power generation, bio- mass displaces gas; if used for process heat, it displaces oil. Since greenhouse gas (GHG) emissions from oil are much greater than those from gas, diverting bio- mass from process heat into electricity generation may well result in an increase in aggregate GHG emissions, not a decrease.19 It is well understood that rice husk prices were bid up in Thailand once rice husk projects received a generous adder. But in Vietnam, rice husk prices have been bid up by the successful development of rice husk pelletizing, with ready markets for pellets in the Republic of Korea and Japan. In the space of the past few years, prices have already reached as much as $35/ton. One may note that from a global GHG perspective, it does not matter whether rice husk displaces oil in Vietnam or oil in Korea and Japan. The Avoided Social Cost of Thermal Generation Examination of Vietnam’s power generation shows that even during the wet season, when the output of the North’s hydro projects peak, CCGTs (in the South) generate power throughout the day. Consequently the avoided cost of thermal generation is given by the CCGT with the highest variable cost ­ (figure 3.12)—which in practice means the Ca Mau project, whose gas price is indexed to the Singapore fuel oil price.20 Production cost simulations prepared for the Seventh Power Development Plan (Vietnam) (PDP7) showed that CCGTs would be run 24 hours a day until at least 2025 (figure 3.13). Consequently, one may take the avoided social cost of thermal generation in Vietnam as a CCGT, with gas priced at international levels. ­ alaysia-Singapore Table 3.12 shows the calculation for Vietnam, based on the M gas power purchase agreement (PPA) as the benchmark for the border price: Malaysia and Vietnam share the Ca Mau gas field. The avoided social cost (PECON) is 11.7 cents/kWh for a $111/barrel (bbl) OPEC Reference Basket21 price (­average 2012). We note this is higher than the 11 cents/kWh required for wind generation, so in the absence of a capacity penalty wind power is economic. There are no reliable health damage studies for nitrogen oxide (NOX) emis- sions from gas generation in Vietnam, so PENV ~ PECON. The calculations for the financial price of gas, at the Ca Mau pricing formula of 0.45 of the Singapore border price, show the financial avoided cost at 5.7 cents/kWh (=PFIN).22 Carbon Accounting and the Clean Development Mechanism (CDM) Vietnam well illustrates the problems of carbon accounting. After a slow start,23 by 2012 there were 29 grid-connected RE projects registered with the United Nations Framework Convention on Climate Change (UNFCCC). All of these The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 67 Figure 3.12 Vietnam’s Expected Generation Mix (2008–25) 120 100 80 Net capacity additions, MW 60 40 20 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Oil, CT, other CCCT_LNG SHP/RE Hydro CCGT CFB Imported coal Domestic coal Imports HydroPS Source: Meier 2011. Note: CFB = circulating fluidized bed; CCCT = combined-cycle combustion turbine; CT = combustion turbine; MW = megawatt; PS = pumped storage; RE = renewable energy; SHP = small hydro project. projects used the standard UNFCCC methodology for calculating emission fac- tors to be used, and those under 15 MW used the simplified methodology that calculates the emission factor as the weighted average of the build and operating margins. The most recent official calculation by Vietnam’s Designated National Authority (DNA) is 0.54 kg carbon dioxide (CO2)/kWh, based on the average of the build margin (BM) 0.4722 and the operating margin (OM) 0.6095. But at the margin, it is the CCGTs that are displaced by additional RE in Vietnam, whose GHG emission factors are much lower than the grid average. In the economic analysis one should use the best estimate of the actual impact, not an accountant’s artifact—even though the UNFCCC methodology is to the advantage of the developing country when calculating the magnitude of carbon credits. Consequently in the calculations that follow, we use the CCGT emission factor in economic analysis, but the DNA estimate of emission factors for calcu- lating the financial contribution potentially made by the CDM in buying down the incremental costs. Table 3.13 illustrates the potential impact CDM revenues may have on the revenue of a daily peaking small hydro with a tariff of 5 cents/kWh. At $15/ton CO2, The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 68 Case Study: Vietnam Figure 3.13  Wet Season Generation, Typical July Week 8 6 1,000 MW 4 2 0 ly ay y ay ay ay ay ay da (Ju nd sd sd id rd nd 4) es Fr tu ne ur o Su Tu M Sa Th ed W Oil Hydro Gas CCGT Coal Source: ERAV 2012. Note: CCGT = combined-cycle gas turbine; MW = megawatt. Table 3.12 Avoided Social Cost of Gas Generation Units Basis 1 World oil price $/bbl 111.5 OPEC Reference Basket 2 Singapore HFO price ratio Number 0.942 3 HFO $/bbl 105 OPEC Reference Basket 4 HFO mmBTU/bbl 6.29 5 Singapore HFO price $/mmBTU 16.71 6 Singapore gas price ratio Number 0.9 7 Singapore border price $/mmBTU 15.0 8 Ca Mau price $/mmBTU 6.8 Ca Mau gas supply agreement, 0.45 of Singapore price 9 Assumed price $/mmBTU 15.0 Border price 10 Transportation $/mmBTU 1.1 Ca Mau gas supply agreement 11 Delivered gas price $/mmBTU 16.1 12 CCCT heat rate BTU/kWh 7,250 ERAV 13 Avoided variable cost $/kWh 0.117 14 Exchange rate D/$ 20,830 15 Avoided variable cost D/kWh 2,436 Source: ERAV 2012. Note: bbl = barrel; BTU = British thermal unit; CCCT = combined-cycle combustion turbine; ERAV = Electricity Regulatory Authority of Vietnam; HFO = heavy fuel oil; kWh = kilowatt-hour; mmBTU = million British thermal units; OPEC = Organisation of Petroleum Exporting Countries. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 69 Table 3.13 Potential CDM Revenue Carbon price $/ton CO2 10 15 20 25 30 Emission factor kg/kWh 0.54 0.54 0.54 0.54 0.54 Carbon price D/kg 207 310 414 517 621 Value D/kWh 112 168 223 279 335 CER fraction Number 0.75 0.75 0.75 0.75 0.75 Value D/kWh 83.8 125.7 167.6 209.5 251.4 Cents/kWh 0.4 0.6 0.8 1.0 1.2 Typical tariff Cents/kWh 5 5 5 5 5 Impact % 8 12 16 20 24 Source: MoIT 2011. Note: Exchange rate for 2012 at $1 = D20,690. CER = certified emission reduction. CDM = clean development mechanism; kg = kilogram; kWh = kilowatt-hour; tonCO2 = ton of carbon dioxide. the incremental revenue—if 75 percent of carbon offsets can be sold—is an addi- tional 16 percent. But with only 26 SHPs registered among the 88 projects (known to ERAV) likely to be eligible, two issues have discouraged developers: the up-front transaction costs, and the increasing difficulty of demonstrating additionality. Renewable Energy Targets Although there are no headline targets for RE (as in the case of Sri Lanka or Indonesia), RE targets do appear in a variety of government documents and plans. For new and RE the National Energy Strategy, approved in 2007, estab- lished a target of 3 percent of commercial primary energy by 2010, increasing to 5 percent by 2020.24 The Seventh Power Development Plan (PDP7), approved by the prime minister in July 2011,25 established the following capacity targets for RE: ­ • Wind energy: 1,000 MW by 2020; 6,200 MW by 2030. • Biomass crop residues: 500 MW in 2020; 2,000 MW in 2030. • Hydropower: from 9,200 MW in 2010 to 17,400 MW by 2020. The expected 2020 generation mix is 19.6 percent hydro, 46.8 percent coal, 24 percent gas, 4.5 percent RE (small hydro, wind, and biomass), 2.1 percent nuclear, and 3 percent imported power. In addition, to achieve 100 percent household electrification by 2020, it was decided to electrify the remaining 600,000 households not likely to be connected to the grid by RE. Attached to the PDP7 is an officially approved list of projects. In the case of RE, the total additions from “wind farm and renewable energy” between 2011 and 2020 is 1,660 MW (slightly more than the 1,500 MW listed above). Individual projects are not identified—only the aggregate annual targets. The best-founded targets, based (at least in part) on a supply curve analysis, are those set out in the MoIT’s draft REMP, as discussed above. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 70 Case Study: Vietnam Neither the PDP7 nor the National Energy Strategy targets have much practi- cal significance, in part because they are unsupported by any detailed analysis and are simply political statements. Also, as noted above, the REMP has yet to be approved by the government, so the REMP targets in any event have no official standing. A draft of a proposed RE Decree currently under consideration by the MoIT, does not mandate numerical values of RE targets, but makes the MoIT responsible for issuing a set of headline targets and empowers the MoIT to adjust preferential tariffs to ensure that such targets are met. But in the absence of agreement on how any incremental costs will be recovered, the prospects for issuing targets in the near future are small. Indeed, this report takes the position that issuing targets in the absence of a process for sustainable incremental cost recovery is pointless. Design of Incentive Schemes VEPF Subsidy Scheme In principle, the Ministry of Finance/Ministry of Natural Resources and Environment (MoNRE) Circular 58 provides a mechanism for providing subsidy to RE projects,26 under which the VEPF would provide a subsidy equal to the difference between actual production cost plus fair return, and the tariff offered by EVN. But to date, no grid-connected RE project has been enabled by this mechanism. The one application that was made to the VEPF to support a bio- mass project was unsuccessful. The scheme provides no incentive for developers to seek CDM funding, nor even to design an efficient project—and precisely what constitutes a “fair return” was not defined. Avoided Cost Tariff The design of an ACT was entrusted to ERAV, which developed the basic ratio- nale of setting the tariff on the basis of the avoided cost of gas generation. There remain some very high-cost diesel and fuel oil generation projects in the EVN system, but these are either being phased out, or serve mainly for system fre- quency support at remote parts of the network, and would not therefore be displaced by RE. We have already noted, above, the success of this tariff in enabling SHPs. One of the design features of the SPPA that proved unnecessary was the ­ cap-and-collar option, under which a developer would be guaranteed receipt of a minimum of 90 percent of the tariff prevailing on the date of PPA signature, in exchange for agreeing to a maximum payment of 110 percent of the tariff.27 Of the 88 SPPAs signed till end 2011, none elected for this option—which is really a vote of confidence in the tariff methodology and the announced procedure for annual adjustment. Wind Feed-In Tariff The basis for the level of the wind FIT has never been made public. In fact the Institute of Energy (IoE) conducted a detailed study of wind farm production The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 71 costs (based on Chinese wind turbines) and proposed a tariff of 10−11 cents kWh. This found no favor with the utility EVN and the government. Notwithstanding that wind power cannot substitute for base-load coal generation, it was proposed that the avoided cost of coal generation might be a suitable yardstick, which was estimated at 6.8 cents/kWh (though the calculations of this estimate were never published). The VEPF was enlisted to provide an additional 1 cent/kWh, though as noted, the plausibility of this contribution has been questioned given that the VEPF itself has no sustainable source of revenue.28 Notwithstanding anecdotal reports that Chinese developers have expressed interest, the plausibility of the level of the tariff may be judged by comparison with the wind FITs in China—which have higher rates of remuneration for much better wind regimes. Fichtner, the German wind power and adviser to the MoIT, has long advo- cated the merits of a “stepped” FIT, exactly following the German model (Fichtner 2009). Under this principle, the FIT is adjusted for the load factor—the higher the load factor, the lower the FIT. The consultant’s report argued that: “A stepped FIT design leads to a homogenous generator profit that is nearly the same across all load factors, and will decrease windfall profits.” Concerns about “windfall profits” are almost always a reflection of poorly designed policies—in this case first-come, first-served access to a preferential technology-specific tariff. The solution is to make access to a guaranteed price and must-run dispatch dependent on a competitive process, rather than build further complexity into an already inefficient mechanism. Such a design feature has no merit for Vietnam (or indeed for any other devel- oping country where economic efficiency must be the priority). Wind developers should be encouraged to develop the best sites, not poor sites. The rationale for its adoption in Germany was regional equity—an effort to promote wind devel- opment in the interior of the country, rather than the Northwest coast, where most of Germany’s wind farms are located (and where the wind regime is best). But such motivation is a luxury for rich countries, and has no basis in economic efficiency. The principle of “degression,” wherein FITs are reduced by some fixed rate over time, was also first introduced in Germany, and has been advocated as an incentive to early investment and for reducing technology costs. But the existing Vietnam wind FIT is so low that discussion of degression is academic. Finally, arguments for technology-dependent FITs or auctions have no merit for relatively poor developing countries. Targets by technology, it is argued, are necessary to encourage all forms of RE. But why? What matters is that GHG emissions are reduced, not by what technology (or policy) this is achieved. The marketplace is a much better mechanism to decide what mix of technologies can most cost-effectively meet given levels of RE generation. It is claimed that in a technology-neutral auction, in Vietnam most of the offered capacity would be hydro; a general lack of wind capacity promises to constrain wind power in the country. But why wind power should be seen as an end in itself remains a mystery. Clearly, large quantities of wind power can be enabled if the FIT is set at high levels, since the certainty of a FIT makes for a The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 72 Case Study: Vietnam more predictable revenue stream, and is therefore favored by both lenders and potential private sector investors. But the question is at what cost, and who pays. Great caution needs to be exercised when making international comparisons of wind energy prices. While the Decision 37 wind power tariff is certainly the lowest in Asia, prices even lower than 7.8 cents/kWh have been reported in Latin America, especially in Brazil, which has pioneered RE auctions (Cunha and others 2012). Support for wind power in Brazil was initially provided through a ­ fixed FIT (the Program for the Promotion of Renewable Energy [PROINFA] scheme), with a tariff of 15.7 cents/kWh (table 3.14). In the first auction, in 2009, some 1,800 MW was offered for an average of 8.5 cents/kWh; in subse- quent auctions, prices fell to the 6–7 cents/kWh range. But such low prices are made possible only by an unusually good wind regime: annual load factors of around 50 percent, and average annual wind speeds of 9 m/sec. Such high- capacity factors are extremely unlikely in Vietnam. Moreover, the gap between developer expectations and actual performance in practice remains wide. In Brazil the National Electricity Operator (ONS) issues a monthly report tracking production of several wind farms. In the Northeast, just two wind farms that have been on line for more than a year were operating above estimated capacity factors over the past 12 months. The largest discrepan- cies include Praia do Morgado, a 28.8 MW wind farm owned by Energimp/ Cemig, which claims an estimated capacity factor of 50 percent but has operated at an average of 31 percent in the 12 months to March; and Praia Formosa, a 104.4 MW project owned by SIIF Énergies, which claims an estimated capacity factor of 39 percent but has been operating at an average of 28.4 percent.29 Indeed, independent experts cast significant doubts on whether the actual performance of auctioned projects will live up to their claimed capacity factors ­ (see, for example, Barroso 2012). Table 3.15 shows a comparison of capacity fac- tors, by country, for wind farms in operation. Summary Evaluation Table 3.16 compares various tariff designs using the criteria noted in chapter 2. Only the ACT can be judged successful. Table 3.14 Capacity Factors and Wind Auction Prices in Brazil Brazilian Power, MW Capacity factor (%) real/kWh Cents/kWh PROINFA, feed-in tariff 1,288 32.5 308.3 15.72 2009 LER reserve energy auction 1,807 43.3 167.38 8.54 2010 LFA alternative sources auction 1,584 43.9 147.19 8.24 2010 LER reserve auction 528 50.5 134.25 7.52 A-3 2011 auction 1,067 45.4 101.35 6.18 2011 LER reserve auction 861 49.8 101.56 6.20 A-5 2011 auction 976 49.0 105.12 6.41 Source: Gornsztejn 2012. Note: See chapter 9 for further details on Brazilian RE auctions. kWh = kilowatt-hour; LER = Reserve Energy Auction; LFA = Alternative Source Auction; MW = megawatt; PROINFA = Program for the Promotion of Renewable Energy. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 73 Table 3.15  Wind Capacity Factors in Selected Countries End-2009 installed capacity 2009 capacity factor % Ireland 1,270 29.0 United Kingdom 4,058 28.7 Greece 1,087 21.9 Portugal 3,535 27.1 Sweden 1,560 22.0 Denmark 3,480 23.0 Netherlands 2,221 23.6 Spain (incl. Canary Islands) 19,149 23.0 France 4,538 22.3 Germany 25,777 17.4 Italy 4,850 16.2 United States — 28.8 India — 12.0 Egypt, Arab Rep. — 38.6 China — 20.0 Brazil (auction bids) — 43.0 Sources: Europe: Renewable UK 2011; India: World Bank; Others: International Renewable Energy Agency (IRENA). Note: — = not available. Table 3.16  Design of Existing RE Incentive Schemes in Vietnam 2011 VEPF subsidy scheme Avoided cost tariff Wind feed-in tariff Introduced 2008 2009 2011 Achievement to date, 0 0 MW GWh 0 0 Economically efficient No Yes, in principle No (no incentives for developer (though tariffs have yet to (level far below avoided social to reduce costs or seek reach actual marginal cost) cost) CDM) Market principles No Yes No (first come, first served) (first come, first served) Sustainable recovery No Not applicable No of incremental (no sustainable source of (unclear that VEPF can cover costs funding) incremental costs) Transparency Yes Yes No (methodology published) (rationale unclear) Adaptability Not applicable Yes Limited (updated annually, reviewed (only adjustment for FOREX to by regulator) maintain $ denominated price) Successful? No Yes No Note: CDM = Clean Development Mechanism; FOREX = foreign exchange; GWh = gigawatt-hours; MW = megawatts; VEPF = Vietnam Environmental Protection Fund. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 74 Case Study: Vietnam Incremental Costs and Their Recovery Wind Power The incremental financial cost to the buyer (the Central Power Company, CPC) is the difference between the cost of wind energy (at the FIT) and the cost of conventional energy from the generation market operator (that is, the average wholesale price as calculated by ERAV), as shown in table 3.17. It is immediately obvious that the incremental cost to the CPC is higher than the 1 cent/kWh that Decision 37 provides as a subsidy to the buyer from the VEPF. But D662/kWh is not the true incremental cost, because when 1 kWh of addi- tional wind power is purchased by the CPC from a wind farm, the g ­ eneration market (EVN) responds by reducing dispatch in its most expensive thermal genera- tor by 1 kWh, and therefore avoids the marginal financial operating cost at that facility, and not the average wholesale price. For Ca Mau—the most expensive CCGT in the EVN system—this marginal fuel cost calculates to D1,296/kWh, so the actual (financial) incremental cost to EVN is D321/kWh (1.55 cents/kWh)30 (table 3.18). Small Hydro In principle, the ACT for RE makes a buyer indifferent to whether a given quan- tity of RE is purchased from an SHP, or whether the same quantity of thermal energy is purchased from the system market operator (SMO). But distribution companies31 have noted that in some areas, large concentrations of SHPs impose significant additional network development costs that need to be recovered, and that therefore potentially affect the retail tariff. In the past, the PCs have also noted that purchases of energy from SHPs are more expensive than purchases at the bulk-supply tariff. Even though this should in principle be equalized across PCs over the long term, in the short run this difference may raise cash-flow issues. ­ Table 3.17 CPC’s Incremental Financial Cost of Wind Energy, 2011 D/kWh Cents/kWh Wind power tariff 1,617 7.8 As stipulated in Decision 37 Wholesale cost 956 4.6 As calculated by ERAV Incremental financial cost 662 3.2 Source: ERAV 2012. Note: CPC = Central Power Company; kWh = kilowatt-hour; ERAV = Electricity Regulatory Authority of Vietnam. Table 3.18 EVN’s Incremental Financial Cost of Wind Energy, 2011 D/kWh Cents/kWh Wind power tariff 1,617 7.80 As stipulated in Decision 37 EVN avoided cost 1,296 6.25 Ca Mau Incremental financial cost 321 1.55 Source: ERAV 2012. Note: kWh = kilowatt-hour; EVN = Electricity of Vietnam. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 75 Table 3.19 Small Hydro Project: Typical Purchase Costs vs. Wholesale Price, 2010–11 2010 2011 RoR Daily peaking RoR Daily peaking Wholesale price 718 718 891 891 SHP 600 740 640 850 Impact of SHP −118 22 −251 −41 Source: ERAV 2012. Note: RoR = run-of-the-river; SHP = small hydro project. In 2010 the wholesale price (before equalization) was D718/kWh, whereas a typical RoR SHP had a tariff of around D600/kWh, so RoR prices (and projects with old PPAs with average tariffs of D625/kWh) were cheaper than the whole- sale price. But purchases from daily peaking projects under the ACT were D22/kWh more expensive (table 3.19). But by 2011 even daily peaking power projects under the ACT, at D850/kWh, were D41/kWh cheaper than the wholesale price. Energy from RoR projects is cheaper still. Indeed, for reasons discussed further below, the actual avoided costs of SHP are greater than the ACT. But the problem of high network developments costs associated with SHPs is significant: these costs arise because local loads in some of the rural provinces where there is much SHP development are smaller than the SHP output, espe- cially in the wet season, which must therefore be evacuated to more distant load centers. In practice, often because of the very long distances involved, this means additional 110 kV development costs. For example, nowhere is this disparity greater than in Muong Te district in Lai Chau province (in Vietnam’s Northwest): by 2020 the local loads are unlikely to exceed 10 MW, but 120 MW of small hydro will feed into the Muong Te 110 kV substation. This power needs evacuation to the national grid. But a clear identi- fication of incremental 110 kV development costs, attributable solely to SHPs, is more difficult than it at first appears. Many 110 kV lines in remote areas would be built anyway as part of the national strategy to extend the grid into rural areas: the presence of SHPs simply accelerates the timing of these lines. Indeed, a review of provincial transmission and distribution (T&D) development plans reveals that by 2020 almost all district towns, even in the most remote areas, would be served by the grid, even where there is no SHP development. In 2011 ERAV prepared a detailed examination of these incremental transmission costs in six provinces with large concentrations of small hydro. The ­ detailed transmission plans were examined in these projects, and 110 kV n ­ etwork costs, classified by whether they were needed solely for SHP power evacuation, or would be needed even in the absence of small hydro. The results are shown in table 3.20: for the six provinces, some $67 million in incremental network devel- opment costs were identified, equivalent to $51/MW in additional investment costs. These costs are not recovered in the present avoided cost generation tariff. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 76 Case Study: Vietnam Table 3.20 Summary of Incremental Network Costs To 2015 2016–20 Total Incremental costs (D billion) Dak Nong 117 84 201 Nghe An 153 0 153 Gia Lai 506 0 506 Lai Chau 236 109 345 Son La 198 0 198 Total 1,210 193 1,403 Dak Nong 63.4 87.5 71.7 Nghe An 48.0 0.0 41.1 Gia Lai 71.1 0.0 65.8 Lai Chau 66.2 28.2 46.5 Son La 33.7 33.7 Total 56.0 32.6 51.0 Total, D billion/MW 1.1 0.7 1.0 Source: ERAV 2012. Note: MW = megawatt. Table 3.21 Connection Costs at Large Hydro and Thermal Projects 220 kV Installed capacity Cost Project Type Circuits x km MW D billion D billion/MW $/MW Non Trach 1 CCGT (2 x 0.7 km) + (4 x 0.7 km) 450 18.3 0.04 2.0 O Mon 1 CCGT 600 66.7 0.11 5.4 Average CCGT 1,050 85.0 0.08 3.9 Nghi Son 1 coal 2 x 6.7 km 600 130.0 0.22 10.6 Son Dong coal 2 x 18 km 220 73.4 0.33 16.3 Average coal 820 203.3 0.25 12.1 Srepok 4 hydro 2 x 6.7 km 70 30.9 0.44 21.5 A luoi hydro 2 x 30 km 150 146.0 0.97 47.5 Dong Nai 3 hydro 2 x 30 km 180 81.7 0.45 22.2 Dong Nai 4 hydro 2 x 11.4 km 340 39.9 0.12 5.7 Huoi Quang hydro 2 x 17.9 km 560 149.6 0.27 13.0 Trung Son hydro 2 x 63 km 260 452.8 1.74 84.9 Ban Chat hydro 2 x 27.4 km 220 163.5 0.74 36.3 Average hydro 1,780 1,064.3 0.60 29.2 Source: ERAV 2012. Note: CCGT = combined-cycle gas turbine; km = kilometer; MW = megawatt. How do these costs compare to the connection costs of thermal generation and of large hydro—which in most cases are at 220 kV and the responsibility of the National Power Transmission Company? As shown in table 3.21, the average connection cost of CCGTs is $3.9/MW, $12.1/MW for coal, and $29.2/MW for large hydro. So even when the avoided thermal connection costs were subtracted The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 77 from the small hydro requirements, there remains a balance of around $40/MW that is apparently unrecovered. This unrecovered cost is, however, offset by a deviation from strict marginal cost evaluation in the calculations used to determine the avoided fuel cost. The original intention of the tariff design was to calculate the average fuel costs of a set of the most expensive thermal plants that corresponded to the inventory of renewables in the portfolio. So, if there were 500 MW of small hydro operating, one would calculate the average variable cost of the most expensive 500 MW of thermal capacity. This was already a deviation from a strict marginal cost evalu- ation, which would simply have applied the variable cost of the single-most expensive project.32 But this last method would have resulted in a tariff that was unacceptable to the PCs, and therefore the final procedure adopted an averaging interval whose width was determined by the regulator. To show the impact of this averaging interval, consider Vietnam’s most expen- sive thermal projects, as shown in table 3.22. Ca Mau has the highest variable (fuel) cost at D884/kWh (4.67 cents/kWh), followed by the Formosa imported coal project at D671/kWh (3.77 cents/kWh), and then the other CCGTs listed in order of decreasing cost. Assume that in the peak hour, all of these plants are operating and are stacked in the order shown in figure 3.14. In the peak hour, if there were an additional 270 MW of SHPs, the most expensive thermal project (Ca Mau) would be backed down, so the avoided cost is D884/kWh. But the ACT calculates the average of thermal costs across a capacity band across all six plants (3,798 MW), which brings the amount to D636/kWh. So the benefit to EVN is the difference between these two values, D248/kWh. In normal, off-peak hours, neither Formosa nor Ca Mau would be ­ dispatched33; 300 MW of SHP would result in 270 MW of Ba Ria (D545/kWh), plus 30 MW of Phu My 4 (D494/kWh) being backed down, for an average of D540/kWh. But the ACT averages costs over all four of the operating plants, namely D499/kWh. In short, the actual avoided costs are higher than the reimbursement (presently) provided to the SHP in the ACT. Table 3.22 Merit Order Stack, 2010 Fuel cost Capacity D/kWh Cents/kWh MW Ca Mau 884 4.67 1,286 Formosa 671 3.55 150 Ba Ria 545 2.88 270 Phu My 1 494 2.61 1,021 Phu My 21 494 2.61 858 Phu My 4 494 2.61 213 Total 3,798 Source: National Load Dispatch Centre (Vietnam) (NLDC) avoided cost tariff (ACT) calculations. Note: Exchange rate at $1 = D18,920. kWh = kilowatt-hour; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 78 Case Study: Vietnam Figure 3.14 Averaging Intervals in the Avoided Cost Tariff 4 Actual 3 displacement 1,000 MW 2 Used in act calculation 1 0 Peak Normal Ca Mau Formosa Ba Ria Phu My 4 Phu My 21 Phu My 1 Source: ERAV 2012. Note: MW = megawatt. The conclusion is that even though the incremental transmission cost was not included in the ACT, this is offset by the wide averaging interval used for the calculation of the avoided fuel cost, with the result that the ACT as issued is a good approximation of EVN’s avoided cost. But the problem is that the incre- mental transmission investments have to be provided up front by the PCs before the SHPs are operating, which in the present situation of stressed cash flows they find difficult to mobilize. Cost Recovery At the time of writing, a sustainable mechanism for incremental cost recovery has yet to be established. As noted, the MoIT is presently drafting a RE Decree that may or may not include a new proposal to establish an RE fund supported by a consumer levy. With the bulk of the RE having been provided by small hydro at costs lower than the EVN’s actual avoided cost, there has been no need for a cost-recovery mechanism. All this would change if a FIT were introduced for wind at around 12 cents/ kWh—the level recommended by the many advocates of wind power in Vietnam. The difficulty is that the bulk of the wind resource falls into the service area of the CPC, the distribution company in the central region, with whom the PPA would be signed. Serving mainly the areas with low-load density, the CPC is one of the weakest of Vietnam’s distribution companies, whose cash flow and margins are causing problems even in the absence of such additional cash obliga- tions. As the signatory to the PPA, it would be required to meet invoices from the wind farms within 30 days of billing in cash. But as shown in table 3.23, The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 79 Table 3.23 Impact of a 12 Cents/kWh Feed-In Tariff on CPC Cash Flows 2011 2012 2015 2020 2022 2023 Without RE purchases Sales growth % 12 12 12 12 12 Total kWh sales TWh 10 11 16 27 34 39 Retail tariff [table 3.10] D/kWh 1,242 1,292 1,370 1,370 1,370 1,370 Consumer bill D billion 12,296 14,322 21,347 37,621 47,191 52,854 Purchased from SO TWh 11 12 17 30 37 42 Average purchase cost to PC D/kWh 956 994 1,055 1,055 1,055 1,055 SO purchases D billion 10,287 11,983 17,860 31,476 39,483 44,221 Distribution margin D billion 2,008 2,339 3,487 6,145 7,708 8,633 With RE purchases Installed capacity MW 26 26 255 850 1,360 1,785 RE energy purchased GWh 64 64 637 2,122 3,395 4,456 Average cost of RE D/kWh 2,520 2,520 2,520 2,520 2,520 2,520 RE purchases D billion 160 160 1,604 5,348 8,556 11,230 Energy from SO TWh 11 12 16 28 34 37 SO purchases D billion 10,227 11,919 17,189 29,237 35,902 39,520 Distribution margin D billion 2,008 2,339 3,487 6,145 7,708 8,633 Incremental capacity costs D billion 0 0 170 609 986 1,301 Total cost D billion 12,395 14,419 22,449 41,340 53,153 60,685 Incremental costs Additional cash required D billion 100 97 1,102 3,719 5,961 7,830 As % of distribution margin % 5.0% 4.2% 31.6% 60.5% 77.3% 90.7% Source: ERAV 2012. Note: CPC = Central Power Company; kWh = kilowatt-hour; PC = pulverized coal; RE = renewable energy; SO = system operator; TWh = terawatt-hour. the incremental cash costs of meeting the 2020 PDP7 target (1,000 MW, of which 850 MW is assumed to be in the CPC’s service area) represent 60.5 ­ percent of the CPC’s distribution margin. This is the main reason for the RE fund. Without the fund, it is very doubtful whether the CPC can meet its current cash obligations for the incremental costs of a significant amount of wind power, and at some point the CPC may simply refuse to sign additional PPAs. There will be difficulties even at the 7.8 cent tariff of Decision 37: at higher tariffs the incremental cash requirements will be impos- sible to meet. But with the guarantee of the fund, which could disburse the incremental costs upon submission of the seller’s invoices within weeks, the CPC’s cash shortfall is quickly made up. Impact of Renewable Energy Tariffs on the Consumer In the case of Vietnam, one can only examine the future impact of RE tariffs on the consumer, either because existing tariffs have no impact (because they are unsuccessful, as in the case of the wind FIT) or because, by definition, the ACT involves no incremental costs. Therefore, we examine the potential impact on The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 80 Case Study: Vietnam Table 3.24 Impact on Consumers: 1 Percent Additional Wind Power by 2020 Units Source 1   2020 baseline generation, PDP7 TWh 211 PDP7 2   Target energy to be replaced [%] 1.0 3   Target energy to be replaced GWh 2,118 4   Load factor of wind [] 0.269 IoE assumption 5   MW of wind required [MW] 899 6   Cost of wind power Cents/kWh 11.3 7   Wind capacity penalty Cents/kWh 1.2 8   Avoided cost of gas-fired CCGT Cents/kWh −5.7 9   Incremental cost Cents/kWh 6.8 10   Total incremental cost [$ million] 145 11 Impact on consumer 12   Retail sales TWh 194 8% growth over 2012 13   Incremental cost per kWh Cents/kWh 0.07 14 Incremental cost per kWh D/kWh 15.5 15   Baseline tariff D/kWh 1,450 Expected 2013 tariff 16   Tariff increase [] 1.1% 18 Avoided GHG emissions 19   Emission factor kg/kWh 0.4 Emission factor for natural gas 20   Avoided GHG emissions Million kg 847 21   Carbon value $/ton 171 Note: CCGT = combined-cycle gas turbine; GHG = greenhouse gases; GWh = gigawatt-hour; IoE = Institute of Energy; kWh = kilowatt-hour; MW = megawatt; PDP7 = Seventh Power Development Plan; TWh = terawatt-hour. the consumer if the wind FIT were raised to the 12.9 cents estimated as the baseline financial cost (including the capacity penalty). Table 3.24 shows the calculations, assuming that 1 percent of the 2020 generation (2,118 GWh) would be replaced by wind power, which would ­ require an additional 899 MW of wind power. The total incremental variable cost is $145 million. When spread over total consumer sales (194 TWh), the incremental cost per retail kilowatt-hour is D15/kWh (0.07 cents/kWh), corresponding to a 1.1 ­percent consumer tariff increase. While this increase may seem small to some, the Government of Vietnam evidently sees this as unacceptably too high. Indeed, replacing gas generation with wind implies a high carbon price to the consumer, equal to $171/ton CO2eq. Another way of assessing the magnitude of this incremental cost is to compare it with the avoided emissions that result from a 5 percent increase in electricity price (the expected increase in the 2013 retail tariff), from a decline in overall energy consumption (as might be the consequence of a small improvement in the income elasticity of electricity demand). A 1 percent demand reduction would again be accommodated by reducing the generation of 2,118 GWh in the most expensive thermal generation, namely, CCGT gas (equivalent to the output of a 280 MW CCGT). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 81 Decreasing the Consumer Cost with International Assistance Can the high consumer cost of carbon reduction be “bought” down by other interventions? The following seven measures may be considered: • Sale of carbon credits: for which we assume $15/ton CO2 in a seven-year Emissions Reduction Purchase Agreement (ERPA), renewed once, with 70 percent of expected CERs sold (this corresponds roughly to the terms of the project development agreement [PDA] currently under consideration). • Government-owned development bank financing: based on a subsidized interest rate of 12 percent (versus 18 percent for normal commercial borrowing), 15 years including a 2-year grace period. • Income tax exemption. • Accelerated depreciation: five years rather than the 20 years in the baseline. • International Bank for Reconstruction and Development (IBRD) finance: 24.5 years, including 9 years’ grace, $ London inter-bank offer rate (LIBOR) swap 2.85 percent + 2 percent spread. • Carbon finance: Clean Technology Fund (CTF), noninterest bearing, 40 years, 10-year grace period, service charge at 0.35 percent.34 The results are shown in table 3.25. Almost 100 percent of the incremental cost can be bought down by carbon finance (under typical CTF terms); the IBRD loan brings the remaining incremental cost to Vietnam (for a 30 MW wind farm) to $50 million. By comparison, the CDM revenue (at $15/ton CO2, achievable in the past in some years, but doubtful in the short term) buys down just percent of the incremental cost. 6 ­ The Cost of Fossil-Fuel Subsidies There are two major sources of fuel subsidy in Vietnam. The first is for coal, sup- plied to the EVN’s coal-burning power stations by the state-owned monopoly coal company (VINCOMIN). The price to EVN is fixed by the government, and Table 3.25  Buying Down the Cost Tariff required, Buy down, Remaining incremental cents/kWh $ million cost, $ million Baseline 11.3 0.0 41.7 CDM 10.8 2.6 6% 39.2 Income tax exemption 9.7 8.9 21% 32.8 Accelerated depreciation 10.8 2.8 7% 39.0 SBV loan 9.8 8.6 21% 33.1 IBRD 5.7 31.7 80% 10.1 CTF 4.0 41.6 99% 0.2 Note: CDM = clean development mechanism; CTF = clean technology fund; IBRD = International Bank for Reconstruction and Development; kWh = kilowatt-hour; SBV = State Bank of Vietnam. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 82 Case Study: Vietnam Table 3.26 Impact of Price Increases to Reduce Subsidy 1 Baseline tariff D/kWh 1,361 2 New tariff D/kWh 1,491 3 Increase [percent] 9.55 4 Inflation [percent] 6.00 5 Real price increase [percent] 3.35 6 Price elasticity (−0.2) [percent] 99.34 7 Demand contraction at consumer [GWh] 1,277 8 At bus bar [GWh] 1,379 9 Avoided costs D billion 2,056 10 Loss of consumer surplus D billion 90 11 Deadweight loss recaptured D billion 1,966 12 Deadweight loss recaptured $ million 94 13 Emission reduction Million kg 552 Note: kg = kilogram; kWh = kilowatt-hour. kept at a low level in the interest of keeping down the electricity price. The theory is that profitable export sales can offset losses in domestic sales to EVN. The government also imposes a 20 percent export tax on coal. Promises that domestic coal prices would be raised to at least the production cost, if not the international border price, have been made since 2006, but to date, price increases have been small. The net result is that VINCOMIN is now making large losses, and the present situation is not sustainable. Significant price increases are expected to be passed to the consumer in 2013 and in the coming years. Gas is also subsidized. As noted, gas from the older onshore fields is priced at production cost, without a depletion premium, while offshore; Ca Mau gas is priced at roughly 45 percent of the border price. But at least one can say that the cost is above the production cost, which is clearly not the case for coal. In table 3.26 we show the impact of the proposed tariff increase in 2013, from D1,361/kWh to D1,491/kWh, which at 6 percent inflation provides for a real price increase of 3.35 percent. At a price elasticity of −0.2, this results in 1,277 GWh less demand at the consumer level, or 1,379 GWh at the bus bar. This corresponds to avoided costs of D2,056 billion, offset by loss of consumer surplus, for a net economic gain of $94 million (which represents the recapture of the deadweight loss of the subsidy). Conclusions We draw the following conclusions from this case study: • Resource endowment. Compared to the wind resource in China (or to the U.S. state of Texas, or Scotland, or the Arab Republic of Egypt), the wind resource in Vietnam is modest. Its most unfavorable characteristic is its high degree of The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 83 seasonal variation. By contrast, Vietnam’s competitive advantage lies with small hydro: at least another 1,000 MW could be exploited at costs below or at the avoided social cost of thermal generation. • Targets. Although RE targets are included in a number of official documents, to date there is no widely publicized headline target. But given the lack of agree- ment on incremental cost recovery, there is indeed no point in setting targets. • Design of incentive schemes. Paradoxically, Vietnam has both the best and the worst of designs. Both the VEPF and the wind FIT are poorly designed, not transparent, and unsuccessful. On the other hand, Vietnam’s ACT, coupled with institutional reforms (such as standardization of the PPA and regulatory devolution to the provinces for small projects) is one of the Asian RE success stories. • Recovery of incremental costs. In large measure because of the success of the small hydro program, creating a mechanism for the recovery of incremental costs for the more expensive renewable technologies (wind and biomass) has been seen as a low priority: the proposal for an RE fund to facilitate disburse- ment to wind developers languishes. But the cash-flow calculations for the CPC, where most of the wind resource is located, shows clearly that without such a fund, the CPC cannot meets its obligations for timely cash payments to wind farms as required by any bankable PPA. Without such a mechanism, the chances of large-scale wind development in Vietnam are in any event small. • Impact on consumers. The impact on the consumer tariff of an additional percent of RE from wind power by 2020 is estimated at 0.07 cents/kWh, or 1 ­ 1.1 percent of the estimated 2020 tariff. That may seem small, but reflects an incremental cost of $145 million, and an avoided cost of carbon of $171/ton CO2. It is clear that such increases in tariffs are not politically acceptable for the time being (the best evidence of which is the continued refusal of the gov- ernment to approve the proposed REMP, which proposes a consumer tariff levy of about this magnitude). • Transmission development. Most of the discussion about the need to develop the transmission infrastructure to enable RE development has been in support of wind power. But Vietnam’s experience shows that successful small hydro development is no less dependent on transmission network development. The problem in Vietnam has been not so much the magnitude of the incremental investment required (at $51/MW a small increase [3 percent] compared to the capital costs of around $1,500/kW), but that the entities responsible for transmission have weak cash flows and difficulty in meeting even the normal investment requirements. International financial institution (IFI) and bilateral assistance to the sector has been directed primarily to generation (in the case The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 84 Case Study: Vietnam of the Bank’s RE Development Project, on lending for developers): future sup- port should also be made available to the PCs for related transmission network development. • Regulatory framework. Although there is a regulator (ERAV), as part of the MoIT it is not sufficiently independent (as is the case, for example, in the Philippines, or the Public Utilities Commission [PUC] in Sri Lanka). Nevertheless, ERAV has high technical competence, and has been at the fore- front of innovative RE tariff development, as evidenced by its 2009 introduc- tion of the ACT. Its annual review of the tariff has been transparent and timely. • Fossil-fuel subsidies. The average wholesale cost of electricity is expected to rise considerably over the next few years, as the subsidies to the coal industry have become unsustainable, forcing price rises in the cost of coal to EVN. Coal proj- ects, even when paying full market price, are not at the margin of the merit order so this will have little impact on the ACT or the avoided social cost of thermal generation. But the average price of electricity to the PCs will increase, pushing average prices higher than the current ACT for RE. This will make it easier for the MoIT to raise the ACT, which will be helpful to the development of further SHPs. • Off-grid RE. Notwithstanding the expectations of the 2001 REAP, powering small grids of less than 1 MW in remote areas with small hydro have mostly proven unsuccessful. Costs have trebled over early estimates, and competent construction supervision in remote areas has proven virtually impossible.35 Vietnam has yet to develop a sustainable institutional model for electrifying these remote areas with RE at reasonable cost—in which it is of course in good company with its Association of Southeast Asian Nations (ASEAN) neighbors.36 • Environmental impact of renewable. Small hydro development is not without its environmental problems. Often the most damaging impacts arise in road con- struction necessary for projects in the remote hilly areas subject to torrential rains in the wet season. But there have also been problems related to dam safety. This is largely a consequence of regulatory devolution from the MoIT (which until 1997 had an important technical review function) at the center to the provincial authorities—authorities whose capacity to evaluate and monitor dam safety and environmental issues of even small projects is often ­ weak—for projects less than 30 MW. • Buying down incremental costs. The analysis shows that carbon finance, the IBRD, and concessionary loads can indeed buy down the incremental costs of wind power. But the question of who pays does little to improve the balance The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 85 sheet or the economic comparisons, and does not change the high avoided cost of carbon, relative to other more cost-effective renewables, notably hydro. • The main problem. In 2011–12 the pace of small hydro development slowed, largely a consequence of a credit squeeze, high interest rates, and increasing civil costs. While the World Bank−supported RE Development Project offers financing support (on-lending though commercial banks) at longer tenors than those available in the commercial banking system, for reasons explained above interest rates are tied to the commercial lending rate, so there have been few takers, and substantial funds remain undisbursed. Calls for making subsidized loans through the Vietnam Development Bank available have been resisted, largely on grounds that social and agricultural development sectors should have priority on the available funds: such funds would indeed ­ buy down the consumer burden. But poverty alleviation and rural develop- ment have a much higher priority for the government, thus challenging a rational economic case for extending subsidized loans for grid-connected RE generation. In short, Vietnam is a success story for RE even though it has refused thus far (and in our view, correctly so) to provide subsidy for wind and biomass. The ACT (and related PPA reforms) has enabled 800 MW of small hydro at no incremental cost to the government or consumers. Provided the government adheres to its announced policy to end coal price subsidies to EVN, and to raise consumer tariffs to better reflect the true cost of supply, the impact on GHG emissions relative to the existing baseline will be much the same as another 800 MW of RE generation. Indeed, reducing subsidies incurs no incremental costs, but rather brings net economic benefits, as deadweight losses are recaptured. Notes 1. The reform program involved a set of measures aimed to gradually move from central planning to market mechanisms and to open up the economy to trade and foreign investment. Key measures included: • Agricultural sector reform. Agricultural collectives were dismantled, land was distrib- uted among farming households, and peasants were given land-use rights for 20 years. These land-use rights could be renewed, and there was also the option of selling or mortgaging the land. • Price reform. Controlled prices for most goods and services were abolished. • Macroeconomic reform. Production and consumption subsidies were eliminated from the budget. Interest rates on loans to state firms were raised above the level of inflation. • Increased integration with the international economy. The opening of Vietnam’s econ- omy to international markets was initiated with the unification of the country’s multiple exchange rates and the devaluation of the dong, followed by gradual struc- tural reforms in foreign trade and investment. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 86 Case Study: Vietnam 2. A 2008 MoIT survey showed the following project pipeline for SHPs: Number of Total installed Average project projects capacity, MW size, MW MoU 178 2,175 12.2 Under construction, no tariff information 21 260 12.4 Under construction, tariff known 67 630 9.4 Under construction, signed power purchase agreement (PPA) 11 101 9.2 In operation 42 278 6.6 Total 319 3,443 10.8 3. The need for such a fund had been identified already in 2001 in the RE Action Plan (REAP), though under the original proposal it was to be funded just by the Government of Vietnam and donors. 4. The classification of hours into peak (4 hours), normal (14 hours), and off-peak (6 hours) follows that of the retail tariff design. 5. This 20 × 1.5 MW wind farm, built by a Joint Stock Company, started operation in 2009, and used German Fuhrlaender Turbines. How the project was financed in the absence of a PPA is unknown. The project operated under an interim agreement with EVN at a reported 4.5 cents/kWh. It is hardly surprising that this project is the first (and only one) to have signed up for the new wind FIT of 7.8 cents/kWh. 6. Decision 37/2011/QD-TTg, June 29, 2011: On the Mechanism Supporting the Development of Wind Power Projects in Vietnam. 7. The GTZ-Fichtner Report on wind power in Vietnam estimated the average levelized cost of wind generation as a function of the quality of the wind regime, assuming an after-tax return on equity of 15 percent, Nordex S70 turbines, weighted average cost of capital (WACC) of 11.5 percent, and capital cost of $1,813−$1,842/kW, depending on hub height. Based on these calculations, Fichtner recommends an initial FIT of 10.5 cents/kWh, which would “allow for developing an average site under good conditions.” Poor Fair Good Average annual wind speed at 60 meters m/sec 5.8 6.7 7.22 Full load hours Hours 1,929 2,712 3,055 Annual load factor % 22.0% 31% 35% Average levelized cost (20-year life) Cents/kWh 16.5 12.0 10.8 D/kWh 3,399 2,472 2,224 Note: Exchange rate: $1 = D20,600; €1 = D28,000. 8. This problem not restricted to RE; the officially approved list of large thermal and hydro projects in the Seventh Power Development Plan (Vietnam) (PDP7) is also a poor indication of what is likely to be realized: the issue is simply that projects that are not included in the official list find subsequent approvals difficult to obtain. 9. Through a mix of s mall hydro (Northwest), PV and wind-diesel hybrids (coastal islands), household PV solar home systems, and some biogas (in the South). 10. By comparison, in the Philippines the capital cost assumption for calculating the FIT is $2,758/kW. 11. Annual foreign exchange (FOREX) depreciation rate = (1 + domestic inflation rate)/ (1 + $ inflation rate). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 87 12. According to the 2012 tax code revisions, losses can now be carried forward to a maximum of five years. 13. The reported cost of the 20 × 1.5 MW Tuy Phong wind farm—the only operating wind farm in Vietnam at present—is $80 million ($2,666/kW) (Tuan 2010). This is for the first phase of the project, but may include road and site development costs for its total development, which is planned at 120 MW. Also, the capital cost reported in the Project Design Document submitted to the United Nations Framework Convention on Climate Change (UNFCCC) for Clean Development Mechanism (CDM) registration shows a capital cost of D798 billion (at the 2006 exchange rate of $1 = D16,000, equal to $1,664/kW). 14. Estimates of the capacity credit in a range of U.S. systems was first reviewed by Grubb and Meyer (1992). The capacity credit was found to generally decrease as the level of wind in the system increases. For example, in the Kansas Gas and Electric system the capacity credit falls from 50 percent at 5 percent wind penetration (wind megawatts as a percentage of system peak) to 30 percent at 20 percent penetration. At low pen- etration levels (5−10 percent), most estimates of capacity credit are between 20 ­percent and 50 percent. The flood of more recent studies on the topic vary little in their conclusions, as acknowledged by leading industry groups such as the British Wind Energy Association (BWEA). 15. http://www.ofgem.gov.uk/Networks/Trans/.../8449-19604_BWEA.pdf. 16. Even though such small plants are not in fact under the control of the regional or national load dispatch centers, the project operator has strong incentives to dispatch into the peak period because of the incentive provided by the ACT—namely a capac- ity payment of D1,805/kWh (see table 3.3) for power delivered during peak dry season hours. 17. In 2008 the ADB proposed a wind-diesel hybrid for Ly Son Island. But the economic analysis showed that even with an off-peak tariff to encourage ice-making during the night (fishermen presently pick up ice from the mainland before heading out to sea), the effective load factor of wind power was just 14 percent. The level of subsidy required for the hybrid was little less than the current level of subsidy to maintain an old diesel unit, and the proposal was abandoned. Now under consideration are small- scale coal units (as in the many Indonesia Islands presently served by old diesel units), but the (Chinese) technology for such units has yet to be successfully demonstrated and its environmental impacts are yet unresolved. 18. There are significant climate differences across the major regions of Vietnam. In the North, the wet season is July−September; in the Central Highlands, it is September− December. The system coincident peak is in November. 19. An excellent example of the law of unexpected consequences. 20. There are a number of older CCGTs in Vietnam whose heat rates are significantly below that of Ca Mau, but whose gas price is subsidized. These are lower (cheaper) in the merit order. 21. Organisation of Petroleum Exporting Countries (OPEC) Reference Basket (crude oils). 22. The financial price from older domestic gas fields is much lower: • $3.2/million British thermal units (mmBTU) for gas delivered from Block 6.1 of the Nam Con Son field, escalating at 2 percent per year. The price includes taxes, gas transmission costs, and PetroVietnam’s fees. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 88 Case Study: Vietnam • $2.2/mmBTU for gas delivered from Block PM3-CAA in the Southwest basin of Ca Mau. The wellhead price also escalates at 2 percent per year, to which is added a transportation cost of about $0.9/mmBTU. 23. Climate Focus 2008. By mid-2008 there were only two registered CDM projects in Vietnam—the Rang Dong Gas Flaring Reduction project and a single SHP (2 MW, Sing Muc). This compared to 54 SHPs approved in China, and 39 in India by mid-2008. 24. Office of the Prime Minister, Approving Vietnam’s National Energy Development Strategy to 2020, with Outlook to 2030, 1855/QD-TTg, December 27, 2007. 25. Office of the Prime Minister, On Approval of the National Power Development Plan between 2010 and 2020, with Outlook to 2030, 1208/DD-TTg, July 21, 2011. 26. Circular 58/2008/TTLT-BTC-BTN&MT, Ministry of Finance and Ministry of Natural Resources and Environment, Financial Mechanisms, Policies of Investment Project under the Clean Development Mechanism. 27. See the Sri Lanka case study, chapter 4, for a similar provision in the Sri Lanka ACT. 28. Its main sources include a 2 percent levy on certified emission reduction (CER) sales, and revenues from environmental permits. Although it has published a subsidy scheme for grid-connected RE projects, the one application for a biomass combustion project was unsuccessful. 29. ABEEólica (the Brazilian wind power association) reports that 12 wind farms across the country performed at an average capacity factor of 32.54 percent in the first three months of 2012, compared with 15.82 percent during the corresponding period in 2011. This documents the large annual variations that may be encountered in operat- ing wind farms, with a variability that is much greater than for small hydro. 30. Note that this is lower than the avoided social cost, which is based on the full inter- national border price of gas, not 0.45 as mentioned in the Ca Mau gas supply agreement. 31. The distribution companies, which have been unbundled from EVN were formerly known as power companies (PCs). Some serve urban areas, but the three most affected by RE development are the Northern, Central, and Southern power compa- nies, with generally low load densities and serving large rural areas. To maintain a national uniform tariff, the PCs are cross-subsidized by the urban PCs (such as the Hanoi and HCMC power companies), a process known as equalization. 32. As noted, fuel oil and diesel plants, in fact the most expensive projects in the EVN system, were excluded from the calculations precisely on the grounds that these would not likely be displaced by RE projects given their role in meeting peak-hour system stability at the extremities of the network. 33. In reality, Formosa is a base-load plant, and would not follow the daily load curve. 34. Terms as per the Indonesian IBRD/CTF loan for the Ulubelu and Lahendong Geothermal Projects. 35. This is true of small-scale electrification of systems supported by the JICA, by the Swedish SIDA, as well as the Bank-supported program in Muong Te. 36. For example, the Philippines has struggled to find a sustainable model for PV solar home systems (SHSs): the model currently under consideration in which rural coops provide PV SHSs as a fee-for-service may be sustainable (the so-called “PV main- streaming” model), but has over twice the cost of providing the same level of service with small diesels. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Vietnam 89 Bibliography Anderson, A. 2012. “Towards the Sustainability of Hydropower in the Mekong Region: Options for Improved Project Design and Technologies.” Presentation at GIZ work- shop, Bangkok, February. Barroso. 2012. “Renewable Energy Auctions: the Brazilian Experience.” Presentation to Workshop on Energy Tariff-based Mechanisms, IRENA, November 2012. https:// www.irena.org/DocumentDownloads/events/2012/November/Tariff/4_Luiz​ _Barroso.pdf. Bogach, S., A. Cabraal, J. Exel, and P. Anh. 2001. “Renewable Energy Action Plan.” ASTAE/ESMAP, World Bank: Washington, DC. Climate Focus. 2008. Renewable Energy Small Power Producers in Viet Nam: Carbon Finance Consultancy. Report to MoIT and the World Bank, August, World Bank, Washington, DC. Cunha, G., L. Barroso, F. Porrua, and B. Bezzeras. 2012. “Fostering Wind Power through Auctions: The Brazilian Experience.” International Association for Energy Economics Newsletter, 2nd Quarter 2012. ERAV (Electricity Regulatory Authority of Vietnam) 2012. Review of the Avoided Cost Tariff for Small Grid-connected Renewable Energy Generation Projects. Hanoi: ERAV. EVN (Electricity of Vietnam). 2007. Wind Resource Assessment for Power Generation. Hanoi, Vietnam: EVN. Fichtner. 2009. A Regulatory Framework for Wind Power in Vietnam. Report to Ministry of Industry and Trade, Hanoi, Vietnam, November. Gencer, D., P. Meier, R. Spencer, and V. Hung Tien. 2011. State and People, Central and Local, Working Together: The Vietnam Rural Electrification Experience. Asia Sustainable and Alternative Energy Program, World Bank, Washington, DC. Gornsztejn, J. 2012. “Financing Wind Power Development in Brazil.” Presentation at Multi- stakeholder workshop on Wind Energy, Copenhagen, Denmark, April 13–15, 2012. Grubb, M., and N. Meyer. 1992. “Wind Energy: Resources, Systems and Regional Strategies.” In Renewable Energy: Sources for Fuels and Electricity, edited by T. Johansson, H. Kelly, A. Reddy, and R. Williams. Washington, DC: Island Press. IEA (International Energy Agency). 2010. CO2 Emissions from Fuel Combustion. Paris: IEA. Meier, P. 2011. Trung Son Hydro Project: Economic Analysis. Washington, DC: World Bank. ———. 2013. Implementation of Renewable Energy Policy in Vietnam. Hanoi: MoIT. MoIT (Ministry of Industry and Trade). 2011. Draft Renewable Energy Masterplan. Hanoi: MoIT. PECC1 (Power Engineering Consulting Joint Stock Company 1). 2001. “Small Hydro Masterplan.” PEEC1, Hanoi, Vietnam. Renewable UK. 2011. “International Comparisons: Turbine Densities and Capacity Factors.” http://www.bwea.com/pdf/publication/RenewableUK_Turbine_Density_Study.pdf. TrueWind Solutions. 2001. “Wind Energy Resource Atlas of Southeast Asia.” http://­ siteresources.worldbank.org/EXTEAPASTAE/Resources/wind_atlas_complete.pdf. Tuan, N. A. 2010. “Opportunities and Challenges to Scaling Up Wind Power in Vietnam.” Presentation at ADB Consultation on Wind Power, Manila, Philippines, June. World Bank. 2005. Economic Analysis for the China Renewable Energy Scale-Up Programme (CRESP). Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 4 Case Study: Sri Lanka Sector Background The Sri Lankan economy has seen robust annual growth at 6.4 percent over the last decade, well above its regional peers. Following the end of the civil conflict in May 2009, gross domestic product (GDP) growth rose initially to 8 percent, largely reflecting a “peace dividend” following the end of the civil conflict, and underpinned by strong private consumption and investment. Growth was around 7 percent in 2013, driven by a rebound in the service sector which now accounts for approximately 60 percent of GDP (World Bank 2014). The Ceylon Electricity Board (CEB) is the main state-owned vertically integrated power utility. It has the monopoly for large hydro and transmission, and for most generation, together with a number of independent power pro- ducers (IPPs). Two entities are involved in distribution: the CEB and the Lanka Electricity Company Pvt. Ltd. (LECO).1 The LECO was established in 1983 to distribute electricity in areas previously served by local authorities (municipal councils and so on), mainly between Galle and Negombo along the ­ Western coastal belt: the LECO purchases its entire supply from the CEB at 33 and 11 kV. In 2011 the CEB had 4.7 million customers, the LECO 490,000 (SLSEA 2011). Load growth over the past two decades has averaged around 7 percent (table 4.1), though with much lower growth in the years of severe power cuts (that have occurred frequently in dry years, largely for reasons of the failure to build additional base load generation projects).2 Even though tariffs have increased over the past few years, in real terms the tariffs have increased little: indeed, as noted below, in real terms the tariffs have barely changed since the mid-1980s.3 Nevertheless, even if the electricity inten- sity of the economy has been permanently reduced, with the resumption of robust economic growth, continued electricity demand growth—and the conse- quent need for additional generation investment—must be expected. But whether the on-going rehabilitation of the war-distributed areas will result in significant additional electricity demand is unclear: the bulk of the consumption in these areas is residential.4 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   91   92 Case Study: Sri Lanka Table 4.1 Electricity Sales GWh 2000 2005 2009 2010 2011 2012 Domestic 2,061 2,866 3,373 3,651 3,928 4,063 Religious 37 49 51 55 59 63 Industrial 2,203 2,732 2,773 3,148 3,379 3,528 Commercial 1,073 1,465 2,059 2,224 2,490 2,614 Street lighting 68 141 133 130 133 139 Total 5,443 7,253 8,389 9,209 9,989 10,407 Growth rate (%) 5.9 3.7 9.8 8.5 4.2 Source: SLSEA 2012. Primary Energy Use The primary energy supply in Sri Lanka is dominated by biomass (43.7 percent in 2011, down from 50.7 percent in 2000), followed by petroleum (43.4 ­percent), hydro (8.5 percent), and renewable energy (RE) (excluding large hydro 1.5 ­percent) (SLSEA 2011). A small amount of coal is imported for industry. Biomass supplies are largely from home gardens and from the replanting program of the rubber industry, while rice husk and other agricultural waste is also increasingly used for energy requirements. Petroleum requirements—crude oil and refined products—are imported. About 40 percent of petroleum product consumption is for power generation (as diesel, fuel oil, and naphtha), with the remainder largely used in transport. Hydropower was the main resource for elec- tricity generation until the mid-1990s, after which the growing demand for electricity has been met mainly by oil. Wind power generation is limited to a small pilot-scale plant. Institutional and Regulatory Framework Sri Lanka’s energy industry is managed by two ministries (power and energy, and the petroleum and petroleum resource development). Although all electricity utilities remain under direct or indirect state ownership, there is significant pri- vate sector participation in power generation and in petroleum distribution. Biomass remains in the informal market, but an estimated 20 percent of biomass is traded. The Public Utility Commission of Sri Lanka (PUCSL) was established in 2002 by an Act of Parliament, with the objective of regulating the utility industries. The PUCSL is expected to regulate the electricity, petroleum, and water sectors, and possibly other utilities at a later date. Individual industry acts of parliament must be amended to enable the PUCSL to perform its regulatory functions. After some delays, the new Electricity Act5 was declared operational as of April 2009. The PUCSL commenced its functions by issuing temporary licenses. A regulatory manual was published in May 2009,6 and the Ministry of Power and Energy issued policy guidelines to the PUCSL. As provided by the act, the longer-term licenses have since been issued (in October 2009) to all the key ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 93 industry players (the CEB, the LECO, and IPPs), while other licenses (small power ­producers, and so on) are being now being regularized to be in compliance with the act. Power Sector Development Until the mid-1990s, the largest share of electricity generation was from hydro- power. With most of the major hydroelectric potential developed by then, non- hydro sources have met most of the additional demand in the past decade. By 2009 only 42 percent of the total electricity demand was provided by the hydro- electric power plants (both large and small), compared with 94 percent in 1995. Sri Lanka’s main problem over the past 20 years has been its reliance on oil-based power generation because of a lack of indigenous fossil fuels: only recently has the first coal project been built. The RE share in electricity generation (that is, mainly large hydro) is expected to decline further as the growing demand is to be met with thermal generation. But the 2006 National Energy Policy declared that Sri Lanka would endeavor to serve 10 percent of electricity generation (in energy terms) with nonconventional and renewable energy (NCRE) sources by 2015,7 with the planned generation mix as shown in figure 4.1. The government has announced a new initiative to grow biomass as a com- mercial fuel, by recognizing grown biomass as the fourth commercial plantation crop.8 An incentive scheme is already in place to grow biomass as an undercrop in coconut plantations. Recent efforts have explored the large-scale development of biomass plantations, with fast-growing, coppicing varieties. Experiments with Gliricidia have met with some success. Initially, one manufacturing industry Figure 4.1  Generation Mix: The Vision, 2009–27 40,000 35,000 Annual generation, GWh 30,000 25,000 20,000 15,000 10,000 5,000 0 20 1 09 10 20 1 12 13 18 19 20 22 23 24 25 26 14 15 16 27 17 2 1 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Renew Oil Coal Hydro Source: World Bank 2010. Note: GWh = gigawatt-hour; NCRE = nonconventional and renewable energy. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 94 Case Study: Sri Lanka commenced purchasing Gliricidia sticks from outgrowers for use in a boiler, and this concept and practice has now spread to several industries. Renewable Energy Development Off-grid hydro development started more than 100 years ago as a source of power for tea plantations, and many such schemes lasted into the 1950s.9 Most were abandoned with the extension of the national grid into the Sri Lankan hill country, and with the advent of low-cost diesels. In the modern era, interest in grid-connected hydro started in the late 1980s, when the CEB Hydro Master Plan identified a number of small projects. But by the mid-1990s just one such project had been developed. In 1995 the govern- ment and the World Bank/Global Environment Facility (GEF) established the Energy Services Delivery Project (ESDP), which was designed to support a range of RE projects, including the first 3 megawatts (MW) wind demonstration proj- ect, support for solar photovoltaic (PV) for household electrification, a village hydro program for off-grid electrification, and support for private sector develop- ment of grid-­ connected small hydro for plants below 10 MW. Key reforms included the ­ introduction of a standardized power purchase agreement (SPPA) and an avoided cost tariff (ACT) for renewable energy (see below for more details). This was so successful that a successor program followed in 2003–07— the Renewable Energy for Rural Economic Development (RERED) project, which received additional financing for 2008–11. Prior to the ESDP, there was no interest in the commercial financing of renewables. The ESDP disbursed $24 million through two development banks and three commercial banks; under the RERED follow-up project one development bank, one commercial bank, and two leasing companies were added, as well as two finance companies and a rural development bank providing independent credit financing outside the World Bank project. The ­ Sri Lanka ESDP has been successful not only in serving as a catalyst to the establishment of a viable, private sector, small hydro industry, it has also been successful in establishing a broader basis for commercial financing for renew- ables (box 4.1). The SPPA offered by the CEB is a standardized, nonnegotiable 15-year con- tract. The contract specifies the conditions, current prices, and pricing policy on which electricity will be purchased by the CEB. The first SPPA was signed in 1996. Investor confidence in Sri Lanka is so far seen mainly in the development of small hydro; investors have shown some interest in developing biomass and wind power plants, but with little success. The new tariff policy announced in 2007 changed this situation (see below), and applications are reported to be flowing in to the Sri Lanka Sustainable Energy Authority (SLSEA) in large num- bers to develop non-hydro projects. In 2007 the SLSEA was created: energy-efficiency programs previously residing in the Energy Conservation Fund were transferred to the SLSEA, with ­ the additional task of formulating strategies to ensure energy security and RE The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 95 Box 4.1 The ESDP On-Lending Program for Renewable Energy Finance The arrangements under Sri Lanka’s World Bank−financed Energy Services Delivery Project (ESDP) were as follows: • Funds were provided to the Government of Sri Lanka as an International Development Association (IDA) credit under typical terms, for which the government carries the exchange risk. • The government in turn nominated the Development Finance Corporation of Ceylon (DFCC) to administer the program, which operated a special account set up in the Central Bank of Sri Lanka. • Developers obtain finance from qualified commercial banks under normal lending terms, with interest at the normal bank rate (“average weighted deposit rate, AWDR” + 5 percent).10 • The commercial banks then refinance, at the AWDR, with the administrator of the program (DFCC), some portion (typically 75−80 percent) of this loan. This was designed to achieve the following objectives: • Banks that had previously been reluctant to lend to developers for small hydro projects (SHPs) (on grounds of unfamiliar risk, unwillingness to lend at long loan tenors) can offload the risk by refinancing from the DFCC. • Developers obtain 10-year loans, significantly longer than the 3–7 years normally obtainable. • Developers deal with normal commercial banks, and establish normal long-term banking relationships that build confidence among all parties over time, so that banks become more familiar with the risks (or lack thereof) entailed in SHPs. development—effectively implementing these on behalf of the Ministry of Power and Energy. The Sri Lanka Public Utilities Commission (SLPUC) really only became effective in 2010 when the first tariffs (based on a new meth- odology) were approved. Jurisdiction over RE tariffs, previously calculated (with- out regulatory oversight) by the CEB, was also transferred to the SLPUC: its first tariff issuance for revisions to the 2009 feed-in tariff (FIT) came into force in mid-2012 (see further discussion below). Renewable Energy Resource Endowment and the Renewable Energy Supply Curve Biomass Sri Lanka has an extensive potential biomass resource for power generation, and for some time so-called dendropower, based on fast-growing species planted in degraded marginal land, has been advocated as a power source. A hectare planted The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 96 Case Study: Sri Lanka with 5,000 Gliricidia, Acacia, or Cassia trees in the dry zone of Sri Lanka would produce about 25–30 tons (dry) per hectare (ha) per year. At a rate of 5,000/ha, an estimated 0.8–1.6 million ha of suitable land (estimates vary greatly!), 12,000−24,000 gigawatt-hours (GWh) could be produced per year. In 2008–09 there were ­proposals to use such biomass as a supplementary fuel at the first coal power project in Puttalam, but these were not pursued because of difficulties in establishing a viable supply chain. A detailed resource assessment is under way at the SLSEA. A 10 MW biomass project in the Trincomalee area to produce power for a cement plant was registered for the clean development mechanism (CDM) in 2009.11 Of the substantial quantities of agricultural waste, much is already being used as a source of heat. Paddy is the main agricultural crop in Sri Lanka, grown in some 0.76 million ha across the country, and tea, rubber, and coconut—which are major export crops—are grown on another 0.8 million ha. The potential power generation capacity from residue generated from these fields is substantial—but again the main problem is its economic collection. ­ Wind As everywhere, generalized assessments of wind resource potential mean little. The 2003 National Renewable Energy Laboratory (NREL) wind energy atlas of Sri Lanka suggests a wind potential of 24,520 MW (Elliot and others 2003) (table 4.2). But according to the SLSEA, due to system absorption limitations, under business-as-usual circumstances its short-term potential is limited to 200 MW. Sri Lanka’s first wind project, a 3 MW pilot supported by the GEF, was commissioned in 1999. This was not a particularly successful project, with ­ production significantly below expectations.12 But the project served its purpose ­ as a pilot, and production at the first recent project built under the new FIT has met the annual capacity factor expectations of the full scale (at 30 percent) (table 4.3). Sri Lanka’s monsoonal climate results in the characteristic wind-speed pattern shown in figure 4.2: between November and April, average wind speeds are less than 5.0 meters per second (m/sec), but during the monsoon are around 9 m/sec. Table 4.2  Wind Resources of Sri Lanka Wind speed Percent Wind Wind power at 50 at 50 meters, Land Lagoon Total windy Total installed class meters, Watts per m2 m/sec area, km2 area, km2 area, km2 land, % capacity, MW Good 4 400–500 7.0–7.5 2,341 664 3,005 3.6 15,000 Excellent 5 500–600 7.5–8.0 788 41 829 1.2 4,150 Excellent 6 600–800 8.0–8.8 517 0 517 0.8 2,600 Excellent 7 >800 >8.8 501 0 501 0.8 2,500 Total 4,147 795 4,852 6.4 24,250 Source: Elliot and others 2003: table 7.1. Note: km2 = kilometers squared; m2 = square meters; m/sec = meters per second; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 97 Table 4.3  Wind Projects in Sri Lanka Installed capacity, MW Start-up Owner Configuration Hambantota 3 1999 CEB Mampuri 10 March 2010 Senok 8 x 1.25 MW Suzlon Norocholai 9.75 — — — Seguwantivu 14.2 May 2013 — — Vidatamunai 9.6 May 2013 — — Source: SLSEA. Note: CEB = Ceylon Electricity Board; MW = megawatt; — = not available. Figure 4.2  Wind Characteristics in Sri Lanka 10 9 Average wind speed, (m/sec) 8 7 6 5 4 3 2 1 0 Jan Feb Mar April May June July Aug Sept Oct Nov Dec Source: http://www.windpower.lk. Note: m/sec = meters per second. In the case of wind, not only seasonal variations but also daily variations deter- mine its value to the grid. Figure 4.3 compares the typical daily pattern of wind speed for sites on the west coast (where the current batch of wind projects is located) that benefit from the daily sea breeze, which, on average, peaks between 16:00 and 18:00 but then declines rapidly—exactly when the load peak increases. But with a difference of just a few hours, the storage in the hydro system should be able to absorb the wind peak production without major impact on the normal pattern of hydro releases. Small Hydro Sri Lanka’s small hydro program has been a success, and was supported by a series of World Bank/GEF projects that provided both financial assistance (through an on-lending program with Sri Lanka’s domestic banks) and technical assistance. As shown in table 4.4, since the first projects were commissioned in 2002, 188 MW of small hydro in 77 projects has been added under these programs. Currently under construction (with the last tranche of lending ­ support) are a further 11 projects (55 MW). The average size of the projects is ­ relatively small, though gradually increasing, reaching just 2.4 MW in 2012.13 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 98 Case Study: Sri Lanka Figure 4.3  Wind Generation and the Daily Load Curve a. The daily load curve 2,500 2,000 1,500 MW 1,000 500 0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 Hour 2010 2007 2009 2008 2006 2005 b. Daily wind speed pattern 9 8 Average wind speed, m/sec 7 6 5 4 3 2 1 0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 Hour Source: Daily load curve from SLSEA (2011) National Energy Balance; daily wind speed pattern from http:// www.windpower.lk. Note: m/sec = meters per second. Estate Sector Hydro Significant potential exists for SHPs in the plantation estates: a survey of 276 sites estimated a potential of around 97 MW (table 4.5) (Fernando 1999). Of the 137 sites in old estates, 49 were found to be in operation, 14 not in operation (but relatively easily rehabilitated), and 74 abandoned. The Asian Development Bank (ADB) is currently funding a project to restore some of these old projects.14 Village Hydro The village hydro program was no less successful. By the end of the RERED in 2011, 174 schemes had been successfully completed, serving some 6,100 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 99 Table 4.4 Small Hydro Projects in Sri Lanka, 2002–12 Installed capacity Average Cumulative Cumulative number Cumulative average Projects added added, MW size, MW capacity, MW of projects size, MW 2002 17 30.9 1.8 30.9 17 1.8 2003 2 4.5 2.2 35.3 19 1.9 2004 12 33.7 2.8 69.1 31 2.2 2005 7 13.7 2.0 82.7 38 2.2 2006 12 20.8 1.7 103.5 50 2.1 2007 2 7.5 3.8 111.0 52 2.1 2008 10 19.0 1.9 130.1 62 2.1 2009 5 23.6 4.7 153.7 67 2.3 2010 3 12.5 4.2 166.2 70 2.4 2011 5 13.9 2.8 180.0 75 2.4 2012 2 7.6 3.8 187.7 77 2.4 Source: http://www.energyservices.lk/statistics/esd_rered.htm. Note: MW = megawatt. Table 4.5 Potential Small Hydro Projects in the Estate Sector Number of Utilized, Potential, Largest Smallest Site classification sites MW MW site, kW site, kW Old estate sites 137 6.1 23.7 1,665 5 New estate sites 71 20.7 1,127 8 Nonestate sites 49 53.0 5,192 44 Total 257 97.4 Source: Fernando 1999. Note: kW = kilowatt; MW = megawatt. rural households in remote areas—with an average size of just under 10 ­kilowatts (kW), serving an average of 35 households each. The success of the program was dependent not just on a good design concept (generally very high heads, and sized according to the availability of dry-season flows),15 and the ESDP/RERED financing facility, but on a significant “sweat equity” com- ponent involving beneficiary households—which was successful because of the relatively high educational levels in Sri Lanka’s rural areas (also reflected in the capability of its village leaders). The availability of good local consulting engineers in Colombo to design and advise these projects was another reason for successful project completion.16 Such a sustainable model has eluded Vietnam, whose off-grid small hydro program in the modern era has been much less successful. Supply Curves Formal RE supply curves, of the type shown in figure 3.6 for Vietnam, do not appear to have been constructed to date. Table 4.6 shows the commissioned capacity as of March 31, 2013, and the projects for which the CEB has signed an SPPA: this can be taken as the current project pipeline potentially realizable in the next few years. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 100 Case Study: Sri Lanka Table 4.6 Status of Grid-Connected RE Projects, March 31, 2013 Number of Installed projects capacity, MW Commissioned   Mini hydropower 111 238.990   Biomass—agricultural and industrial waste power 2 11.000   Biomass—dendro power 1 0.500   Solar power 4 1.378   Wind power 9 73.650   Total—commissioned 128 330.518 SPPA signed   Mini hydropower 72 167.262   Biomass—agricultural and industrial waste power 4 21.300   Biomass—dendro power 2 4.000   Solar power 10 56.770   Wind power 1 10.000 Total—commissioned 89 259.332 Source: CEB. Note: RE = renewable energy; MW = megawatt; SPPA = standardized power purchase agreement. Capital Costs The first set of SHPs built between 1998 and 2003 had average completed finan- cial capital costs of $1,055/kW—though with considerable variation around the average (figure 4.4).17 By 2009 the capital cost assumption for small hydro had risen to $1,620/kW (as shown in table 4.7, together with the capital cost assumptions for the other technologies). The Avoided Social Cost of Thermal Generation Unlike Vietnam, for which the calculation of the avoided social cost of thermal generation is straightforward because it involves just a single fuel in a single technology (combined-cycle gas turbine, CCGT),18 the rapidly changing gen- eration mix makes such calculation more difficult in Sri Lanka. With increasing coal generation, whether renewables will displace oil (as at present) or coal will depend on the extent to which the large storage hydro projects can provide the offsetting load following energy and capacity. There is some evidence that this may be the case (Siyambalapitya 2001); if so, there may be no oil generation for some hours, and the portfolio of renewable energy would permit coal units to be backed down for several months in the wet season (when both wind and SHP production is at its peak). But with some future coal units being planned as IPPs, this may be constrained by take-or-pay clauses (which would argue for two-part IPP tariffs to give the CEB more operational flexibility).19 While a study of the transmission system implications of larger amounts of renewable energy has been completed,20 a comparable study of generation The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 101 Figure 4.4  Distribution of Capital Costs for Small Hydro Projects 4 3 Frequency 2 1 0 300 500 700 900 1,100 1,300 1,500 1,700 1,900 Capital cost, $/kW Source: World Bank 2003. Note: kW = kilowatt. Table 4.7  2009 Feed-In Tariff Cost Assumptions Annual operating Capital cost costs as % of Assumed capacity SL Rs million/MW $/kW capital cost factor (%) Minihydro 190 1,621 3.00 42 Wind 230 1,962 3.00 32 Biomass 217 1,852 4–5.00 80 Agricultural waste 217 1,852 4–5.00 80 Municipal waste 313 2,671 7.00 80 Waste heat recovery 217 1,852 1.33 67 Note: At the exchange rate of $1 = SL Rs 117.2 (August 31, 2009). kW = kilowatt; MW = megawatt. ­ ispatch, which would involve a detailed chronological production simulation d (assessing merit order dispatch over at least hourly intervals over the next 20 years for several scenarios of RE penetration), is not yet available. For the purpose of this report, therefore, we rely on an estimated economic value of new and renewable energy to the CEB system, based on the Wien Automatic System Planning (WASP) model, with and without a portfolio of renewables (that must be forced into the solution), which was estimated at 7.12 cents per kilowatt-hour (kWh). The baseline plan without renewables also does not build any of the other alternatives offered for the express purpose of low carbon development, including wind, small hydro, medium hydro (defined as hydro projects in the 10−100 MW size range),21 liquefied natural gas (LNG), or electricity generated in coal-based supercritical projects in southern India.22 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 102 Case Study: Sri Lanka Figure 4.5 illustrates the two capacity expansion plans: in the base case ­figure 4.5a) the only new hydro project is the Upper Kotmale project, under ( construction at the time this plan was formulated. Figure 4.5b shows the capac- ity expansion plan for the NCRE scenario (which continues beyond 2015 to maintain the same 10 percent share), and figure 4.5c shows the differences between the two. Two 300 MW coal units are displaced entirely, and beyond 2022, some larger units are delayed. Carbon Accounting and CDM With the expected growth in coal generation, the grid emission factors for CDM will increase (table 4.8). The first coal project was commissioned in 2011, in which year the combined margin increased from 0.53 to 0.73; additional increases are expected in the coming years as the share of coal in the generation mix increases further. But the prospects for buying down the incremental costs from CDM have become poor: although the grid emission factor in Sri Lanka is increasing, the certified emission reduction (CER) price is decreasing. Table 4.9 shows the early May 2013 forecasts for European Union (EU) Allowance Unit of one ton of CO2 (EUA) and CER. Because of the oversupply of CERs eligible for Phase III of the European Union Emission Trading Scheme (EU-ETS, 2013–20), the CER price is significantly below the forecast EUA price, and remains below €1/ton throughout the forecast period.23 The table also shows the current ECX (European Carbon Exchange) futures prices for CER for delivery by end- December of each year. The Sri Lanka: Environmental Issues in the Power Sector (EIPS) study (World Bank 2010) allowed calculation of the avoided cost of carbon for the range of low-carbon-emission options—all of which, as noted, needed to be forced into the expansion plan. The options considered were: • LNG (CCGT to replace those coal units not under construction or under active development). • Medium hydro: Uma Oya, (150 MW),24 Broadlands (35 MW), Moragolla (27 MW), and Ging Ganga (49 MW). • NCRE (wind and small hydro, as listed in table 4.11). • “Green scenario” (LNG+NCRE). • Supercritical coal. The Sri Lankan power system is still too small to be able to accommodate 500−600 MW scale supercritical units. They are indeed more efficient than sub- critical, so greenhouse gas (GHG) emissions decline, but being so large, their size exceeds the annual increase in base-load requirement, so there is additional excess capacity, and hence the system PV increases. Table 4.10 shows the result of this analysis. When life-cycle emissions are also considered, most RE options have somewhat lower avoided costs. But in the case The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 103 Figure 4.5 The Least Cost Expansion Plan, 2009–27 a. The baseline 800 Net capacity additions, MW 600 400 200 0 –200 –400 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 b. Renewable energy scenario 1000 Net capacity additions, MW 800 600 400 200 0 –200 –400 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 c. Changes in the expansion plan 1000 Net capacity additions, MW 800 600 400 200 0 –200 –400 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 Coal Oil Hydro Renew LNG Source: World Bank 2010. Note: LNG = liquefied natural gas; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 104 Case Study: Sri Lanka Table 4.8  Grid Emission Factors, 2008–11 Our 2008 2009 2010 2011 Build margin 0.5986 0.6081 0.5684 0.7491 Operating margin 0.6990 0.6975 0.6920 0.7047 Combined margin 0.6487 0.6520 0.6302 0.7269 Source: SLSEA. Table 4.9 Carbon Point Forecasts, 2008–20 Point carbon price forecasts ECX CER futures Open interest, ​ EUA, €/ton CER, €/ton Price, €/ton €/ton 2008 22.4 17.4 2009 13.3 11.8 2010 14.5 12.4 2011 13.3 9.8 2012 7.6 2.9 2013 3.0 0.5 0.42 64,028 2014 4.0 0.6 0.45 36,903 2015 5.0 0.6 0.51 11,241 2016 5.0 0.5 0.56 2,382 2017 6.0 0.4 0.69 4,426 2018 6.0 0.4 0.73 560 2019 8.0 0.3 0.93 697 2020 8.0 0.2 1.04 1,146 Source: ECX. Note: CER = certified emission reduction; ECX = European Carbon Exchange; EUA = European Union Allowance Unit of one ton of carbon dioxide. of LNG, the avoided cost increases, a consequence of high methane emissions associated with liquefaction, transportation, and regasification. These options all lie in quadrant IV of the trade-off between system cost and GHG emissions—that is, higher costs allow lower GHG emissions—and do not include the win-win options. But as shown in figure 4.6, demand-side manage- ment (DSM) lies in the win-win quadrant—so an avoided cost is not defined. But the remaining potential for DSM is quite limited, so the quantity of GHG emis- sion reduction is quite small compared to the supply options. Pumped storage lies in the trade-off quadrant II—GHG ­ emissions increase, while system costs decrease—a simple consequence of the fact that the pumping energy is provided by coal. Appendix B provides further information on trade-off curves and the tools of multi-attribute decision analysis. Renewable Energy Targets The 10 percent target is an aspirational political statement, not one that is based on economic analysis. The SLSEA expects that this will be met largely by a combination of existing small hydro, new small hydro, wind, and The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 105 Table 4.10 Avoided Costs of Carbon Combustion Life-cycle impacts only emissions Composite renewables scenario to meet 10% 87 80 target (NCRE) Medium hydro 37 34 Supercritical coal 7 6 LNG 86 98 Green (LNG+NCRE) 79 81 Source: World Bank 2010. Note: LNG = liquefied natural gas; NCRE = nonconventional and renewable energy. biomass: table 4.11 shows the implementation scenario that would meet such a target, with relative shares of different technologies based on current expec- tations of what might be feasible. At present the bulk of the energy is still small hydro (70 percent in 2012), but the SLSEA expects that wind and biomass will take up increasing shares, with wind accounting for 37 percent of NCRE by 2025. This forecast calls for 100 MW of wind by 2013, but what is achieved is just 40 MW. To meet the target in 2015 requires 220 MW of wind capacity, but this is unlikely to be attained. Design of Incentive Schemes ­ rojects, A range of policy incentives are in place to encourage grid-connected RE p including: • There is no solicitation process; all projects are on a first-come, first-served basis, provided only that they meet the CEB’s technical standards for connection. • The power purchase agreement (PPA) is standardized and nonnegotiable (thus avoiding lengthy negotiations). • The support tariff is published and uniformly applied to all small power ­ producers (SPPs) (until 2006–07 it was based on avoided costs, and then subsequently replaced by a technology-specific FIT). ­ • Projects qualify for Board of Investment (BoI) concessions if they meet the standard criteria laid out by the board. In general, projects with an investment exceeding SL Rs 50 million qualify for BoI incentives, which offer duty-free import of investment equipment and material, and a tax holiday between five and eight years, and a concessionary tax rate thereafter. • Financing support (through competitive interest rates and the ESDP/RERED projects). • A “net” metering facility is available to all LECO customers (also to be extended to the CEB in due course). Consumers are free to use any qualified RE25 source, based on availability and affordability. The capacity limit is the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 106 Figure 4.6  Greenhouse Gas Emissions vs. System Cost Trade-Offs 11,500 IV. Trade-off I. Lose-lose Green 11,000 System NPV US$ million LNG 9,400 10,500 9,350 Hydro No coal System NPV US$ million 10,000 9,300 NCRE 9,250 Force supercritical Baseline 9,500 Hydro 9,200 PS Baseline Force supercritical PS DSM III. Win-win PS+DSM DSM II. Trade-off 9,150 PS+DSM 9,000 60 70 80 90 100 110 9,100 Discounted emissions million tons CO2 90 92 94 96 98 100 Discounted emissions million tons CO2 Source: World Bank 2010. Note: CO2 = carbon dioxide; DSM = demand-side management; LNG = liquefied natural gas; NCRE = nonconventional and renewable energy; NPV = net present value; PS = pumped storage. Case Study: Sri Lanka 107 Table 4.11 Implementation Scenario for NCRE: 2010 Projection 2009 2010 2011 2012 2013 2014 2015 2020 2025 Capacity   Small hydro MW 170 190 200 210 250 270 280 340 390   Wind MW 0 10 40 70 100 160 220 320 470  Biomass MW 12 13 14 14 15 15 36 61 121  Other MW 0 2 4 6 8 10 12 27 52 Total MW 182 215 258 300 373 455 548 748 1,033 Energy   Small hydro GWh 469 566 596 625 745 804 809 983 1,093   Wind GWh 0 25 102 184 272 449 597 869 1,235  Biomass GWh 42 44 47 63 66 72 188 346 740  Other GWh 0 9 18 26 35 44 53 130 273 Total GWh 511 643 762 899 1,117 1,369 1,647 2,328 3,342 Energy shares   Small hydro [%] 92 88 78 70 67 59 49 42 33   Wind [%] 0 4 13 20 24 33 36 37 37  Biomass [%] 8 7 6 7 6 5 11 15 22  Other [%] 0 1 2 3 3 3 3 6 8 Source: World Bank 2010. Note: GWh = gigawatt-hour; MW = megawatt; NCRE = nonconventional and renewable energy. contract demand, subject to a maximum of 42 kVA.26 In any given month, the customer will be billed for the net purchase from the grid. Any surplus exports are credited to the bill, to be used at any time, in any month in the future. Credits can be carried through until the end of the net metering ­ contract (10 years). Avoided Cost Tariff Sri Lanka introduced a standardized PPA for SHPs in 1997 at the start of the World Bank−financed ESDP, based on a published avoided cost based tariff (ACT). The CEB’s actual avoided energy costs (that is, without any capacity credit) are updated annually: table 4.12 shows the tariff for the 16 years of the operation of the system. The ACT system was introduced a number of years before the establish- ment of an independent regulator. The CEB’s calculation of avoided costs was at times controversial, and the disputes were not resolved by the investigation and report of an independent expert (Siyambalapitya 2001), resulting in sev- eral developers instigating arbitration and court actions against the CEB on grounds of alleged inconsistencies and mistakes in the tariff calculation. Nevertheless, despite the pleadings of the developers, the success of the pro- gram speaks for itself. Only short-run avoided variable costs were considered in the calculation. Small hydro plants connect to the 33 kV system, so costs were adjusted for The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 108 Case Study: Sri Lanka Table 4.12 Avoided Cost Tariff, 1996–2011 Dry season Wet season Exchange rate Dry season Wet season (SL Rs/kWh) (SL Rs/kWh) (SL Rs:$) (cents/kWh) (cents/kWh) 1996 2.9 2.9 55.2 5.3 5.3 1997 3.4 2.9 58.9 5.7 4.9 1998 3.5 3.1 64.7 5.4 4.9 1999 3.2 2.7 70.6 4.6 3.9 2000 3.1 2.8 76.6 4.1 3.6 2001 4.2 4.0 89.2 4.7 4.5 2002 5.1 4.9 95.4 5.4 5.1 2003 6.1 5.9 96.3 6.3 6.1 2004 5.7 5.0 101.0 5.6 4.9 2005 6.1 5.3 100.4 6.0 5.3 2006 6.7 5.8 103.6 6.5 5.6 2007 7.6 6.9 110.2 6.9 6.3 2008 9.7 8.9 108.1 8.9 8.3 2009 11.2 10.6 114.7 9.7 9.2 2010 11.9 11.1 112.8 10.6 9.8 2011 11.2 10.2 110.3 10.1 9.3 Note: kWh = kilowatt-hour. average losses to the 33 kV level. The avoided cost was computed separately for the dry season (February to April) and the wet season (May to December and January). The seasonal tariff that was announced by the CEB every year is a three-year moving average of the past three years’ avoided energy costs. If the announced tariff for a particular year fell below 90 percent of the tariff during the year in which the SPPA was signed for a given small power producer (SPP), the tariff applicable would be the tariff of the previous year. A number of methodological issues arise in the Sri Lanka approach, most notably the absence of a capacity credit. While a single 3 MW run-of-river (RoR) hydro may have little impact on capacity deferrals, a portfolio of 100 MW of small hydro, when taken as a whole, is very unlikely to have zero capacity credit. Even in the driest months, the output from the portfolio is not zero (figure 4.7). It should be noted that the original recommendation for the ACT (Vernstrom 1995) did argue for a capacity credit (table 4.13). In any event, the calculation of avoided variable costs in the CEB’s t­ hermal plants tells only part of the story in a system with significant amounts of conventional annual storage hydro. A report prepared by an independent consultant in 2001 (commissioned to help mediate disputes between some developers and the CEB), showed a significant benefit to end-year reservoir storage (equivalent to some 23 GWh, equal to 10 percent of the total SHP contribution of 294 GWh in 2001); in addition, the SPP contributed to a reduction of unserved energy demand of another 11 GWh (Siyambalapitya 2001). These benefits were not included in the CEB’s e ­stimates of avoided costs. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 109 Figure 4.7 Monthly Production from Small Hydro Projects, 2000–02 8 6 Monthly GWh 4 2 0 January April July October January April July October 2001 2002 Glassaugh MHP Mandagal oya M Ellapita ella Rakwana gan Delgoda MHP Carolina MHP Kolonna Dick oya Source: World Bank 2003. Note: GWh = gigawatt-hour; SHP = small hydro project. Table 4.13 Original Recommendation for the Small Power Purchase Tariff Tariff Dry season (SL Wet season (SL Dry season Wet season component Time of day Rs/kWh) Rs/kWh) (cents/kWh) (cents/kWh) HV Energy Peak 4.37 4.11 8.21 7.73 Off-peak 3.43 2.68 6.45 5.04 Capacity Peak 1.67 0.18 3.14 0.34 Off-peak 1.67 0.18 3.14 0.34 MV Energy Peak 4.7 4.48 8.83 8.42 Off-peak 3.64 2.85 6.84 5.36 Capacity Peak 3.48 1.79 6.54 3.36 Off-peak 1.83 0.2 3.44 0.38 Source: Vernstrom1995, exhibit S-2. Note: (1) at the 1995 exchange rate: $1 = SL Rs 53.2; HV = high voltage; kWh = kilowatt-hour; MV = medium voltage. Revisions to the Sri Lankan Tariff Support System Notwithstanding its unique success, in 2007 Sri Lanka changed its ACT system in favor of a production-cost-based FIT. The argument for the change was as follows: • Avoided costs were expected to decline after 2011 (in real terms) once the first coal projects finally came into operation. While existing developers would be protected by the 90 percent floor (thus preserving 90 percent of the expected revenue at the time of the PPA signature, regardless of the actual future ACT), this would be of no value to new projects entering at the lower tariff. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 110 Case Study: Sri Lanka • To ensure that the benefits of renewable energy flow into society (to whom the natural resources belong) in the longer term, after the developers have been given adequate returns on their investments (in other words, in the plain speak of economics, to capture the producer’s surplus). • To encourage renewables other than small hydro (wind, biomass) to come on line, which are not viable on the basis of avoided costs. Only small hydro was considered viable at the present tariff. Low heads, smaller projects (<500 kW), biomass, and wind all require more than avoided costs to be viable. The ­ government had just declared a 10 percent target of grid energy by noncon- ventional renewable energy by 2015, and the view was that small hydro alone could not achieve this target. • Predictable tariffs would bolster the bankability of projects. • The best SHP sites had been developed, and additional projects would require higher tariffs. None of the stated reasons are valid objections to an ACT; rather, they reflect the shortcomings of the way in which the tariff was implemented: • Argument (1) correctly anticipates that average avoided costs may fall when the coal plants come on line—though it is hard to see how auto-diesel-fueled combined-cycle combustion turbines (CCCTs) would not be at the margin even when coal plants are in the system. A better response to this problem would have been to introduce a capacity credit (as originally recommended by Vernstrom (1995). • Argument (2) is the classic preoccupation of government committees that worry about “windfall profits” to developers. It may well be true that the tariff would increase in the next few years, subsequently to decrease again after the coal plants come on stream. But there are better ways to deal with this prob- lem than to introduce FITs, for example, by making risk-sharing symmetrical: if a developer benefits from the 90 percent floor price, the buyer should also have benefited from a corresponding cap.27 Moreover, if there is a concern about developers of good SHP sites capturing site rents (though objection [5] states that no low-cost SHP sites remain), the best way to deal with that prob- lem is through bidding, as introduced in Zhejiang. • Argument (3) correctly notes that wind and biomass are presently uneco- nomic, and would not be developed at the present ACT. • Argument (4) is true. But there was no evidence that the variations in the ACT discouraged bankable projects: the number of projects attested to the bank- ability of the tariff. In reality, the change in policy resulted from the alignment of interests of the established developers (whose industry association has become increasingly vocal as the small hydro industry developed) and of the government, which wished to demonstrate concrete steps had been taken to achieve the 10 percent target. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 111 The 2007 Feed-In Tariff The new system provided a technology-specific tariff offered for the main RE technologies considered to be promising: wind, small hydro, biomass (in two categories, agricultural wastes and biomass plantation crops [“dendro-thermal”]), and municipal solid waste. The Sri Lankan FIT embodied a number of unique and important features: • Recognizing the reality of short loan tenors in the commercial banking system, developers’ costs are frontloaded during the first years of debt service. Therefore, to achieve an acceptable debt service cover ratio (DSCR), the tariff needs to be higher in the early years. A tiered system was therefore introduced, under which the tariff was highest in years 1−6, then lower in years 7−15, and lower still in years 16−18. • A bank guarantee was required to ensure that the SPP operates in years 7−12 of the second tier, in return for the high tariffs paid in the first tier. The guarantee was to be provided to the buyer in years 1−6, and returned from year 7 onwards. • The target financial internal rate of return (FIRR) was set at a high 22 percent return on equity. • In addition to the tiered tariff, a flat-rate tariff was offered that required no bank guarantees (table 4.14). The methodology is to calculate the liveliest tar- iff using a discount rate equal to the weighted average cost of capital (WACC) (that is, 40 percent equity at 22 percent, 60 percent debt at 19.22 percent = 20.33 percent). • In 2009 the SLPUC published a spreadsheet that was used as a basis for the updated tariff (see box 4.2). Few FITs in the world are published with this degree of ­ transparency (although the Philippines regulator publishes a list of proposed assumptions). The hazards of setting production-cost-based FITs are well illustrated by the latest tariff issuance, issued by the PUCSL in October 2012, just following a sharp depreciation in the exchange rate. The commission used an exchange rate of $1 = SL Rs 132.86, but no sooner had the tariff been issued than the Sri Lankan rupee started to appreciate against the dollar: in April 2013, the ­ average exchange rate was $1 = SL Rs 128.9 (table 4.14). Summary Evaluation Table 4.15 summarizes our evaluation of the tariff incentive systems in place in Sri Lanka. Incremental Costs and Their Recovery The 2006 Energy Strategy called for the establishment of a fund, to be funded by a tax. A mechanism to recover the incremental costs is obviously required if another goal of the strategy—namely that the “NCRE shall not cause any addi- tional burden on end use customer tariffs”28—is to be met. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 112 Case Study: Sri Lanka Box 4.2 Sri Lanka’s Feed-In Tariff, 2009 Option 1: Three-tier Tariff The tariffs for the first tier (that is, years 1–6) are as follows (for SPPAs signed in 2007,​ in SLR/kWh): Year of operation 1 2 3 4 5 6 Mini Hydro Nonescalable 7.99 7.99 7.99 7.99 7.99 7.99 escalated O&M 0.48 0.52 0.56 0.60 0.64 0.69 Total 8.47 8.51 8.54 8.59 8.63 8.68 Wind Nonescalable 13.88 13.88 13.88 13.88 13.88 13.88 escalated O&M 1.67 1.80 1.93 2.08 2.23 2.40 Total 15.55 15.68 15.81 15.95 16.11 16.27 Biomass Nonescalable 5.22 5.22 5.22 5.22 5.22 5.22 escalated fuel 5.00 5.25 5.51 5.78 6.07 6.37 escalated O&M 0.84 0.90 0.97 1.04 1.12 1.20 Total 11.06 11.37 11.70 12.04 12.41 12.79 The tariffs for the second tier (years 7–15) are substantially lower: Year of operation 7 8 9 10 11 12 13 14 15 Mini Hydro Nonescalable 2.79 2.79 2.79 2.79 2.79 2.79 2.79 2.79 2.79 escalated 1.48 1.59 1.71 1.84 1.97 2.12 2.28 2.45 2.63 O&M Total 4.28 4.29 4.50 4.63 4.77 4.91 5.07 5.24 5.42 Wind Nonescalable 4.85 4.85 4.85 4.85 4.85 4.85 4.85 4.85 4.85 escalated O&M 2.57 2.76 2.97 3.19 3.43 3.68 3.96 4.25 4.57 Total 7.43 7.62 7.82 8.04 8.28 8.54 8.81 9.10 9.42 Biomass Nonescalable 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 escalated fuel 6.68 7.01 7.36 7.73 8.11 8.51 8.98 9.37 9.84 escalated O&M 1.29 1.39 1.49 1.60 1.72 1.85 1.98 2.13 2.29 Total 9.80 10.22 10.67 11.15 11.65 12.18 12.74 13.33 13.95 And for the third tier (years 16–18) are lower still, as follows: Year of operation 16 17 18 19 20 Mini Hydro Nonescalable 2.06 2.17 2.27 2.39 2.51 escalated O&M 2.82 3.03 3.26 3.5 3.76 Total 4.89 5.2 5.53 5.89 6.27 box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 113 Box 4.2  Sri Lanka’s Feed-In Tariff, 2009 (continued) Year of operation 16 17 18 19 20 Wind Nonescalable 2.06 2.17 2.27 2.39 2.51 escalated O&M 4.9 5.27 5.66 6.08 6.53 Total 6.97 7.44 7.94 8.47 9.04 Biomass Nonescalable 2.06 2.17 2.27 2.39 2.51 escalated fuel 10.32 10.84 11.37 11.94 12.53 escalated O&M 2.46 2.64 2.84 3.05 3.27 Total 14.85 15.64 16.48 17.37 18.31 Option 2: Flat Tariff Technology All inclusive rate (SL Rs/kWh) for years 1–20 Mini-hydro 7.10 (1,070 SL Rs/kWh) Wind 12.83 (1,934 SL Rs/kWh) Biomass 11.87 (1,789 SL Rs/kWh) Source: SLSEA. Note: O&M = operation and maintenance; SPPA = standardized power purchase agreement. Table 4.14  Flat-Rate Feed-In Tariffs At assumed At actual April Issued flat rate SLPUC exchange 2013 exchange tariff rate rate (132.86) rate SL Rs Cents/kWh Cents/kWh Mini-hydro 16.7 12.6 13.3 Mini-hydro, local 17.15 12.9 13.6 Wind 20.62 15.5 16.4 Wind, local 21.22 16.0 16.9 Biomass (dendro) 25.09 18.9 19.9 Biomass (agricultural and industrial waste) 17.71 13.3 14.1 Municipal solid waste 26.1 19.6 20.7 Waste heat 9.19 6.9 7.3 Source: SLPUC. Note: kWh = kilowatt-hour; SLPUC = Sri Lanka Public Utilities Commission. The Sustainable Energy Authority Act established three funds. The first is the Fund of the Authority (which covers the authority’s expenses); second, the Sri Lanka Sustainable Energy Fund (with the main source of revenue tax on imports of fossil-fuel products, and from which subsidies to RE producers were to be funded); and third, the Sustainable Energy Guarantee Fund (which provides loan guarantees for energy-efficiency projects). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 114 Case Study: Sri Lanka Table 4.15 Summary Evaluation of Tariff Designs Avoided cost tariff Feed-in tariff small hydro Feed-in tariff wind Introduced 1997 2007 2007 Achievement to date, MW 187 [?] 40 Economically efficient Yes No No Market principles Yes No No (first come, first served) (first come, first served) Sustainable Recovery of Yes No No incremental costs (by definition) (though not the fault of the (though not the fault of the tariff design itself) tariff design itself) Transparency Yes Yes Yes Adaptability Yes Yes (in principle) (but with Yes (in principle) (but with (updated annually) long delays in issuance of long delays in issuance updated tariff) of updated tariff) Successful? Yes Yes To be seen (extensive project pipeline) Note: MW = megawatt; [?] = not available yet. In the first few years of the scheme, the Fund was unable to pay the CEB’s invoices for incremental costs. The original expectation was that as more and more of the PPAs of existing SHPs expired, and compensation would fall to the low “tier 3” level (and below the CEB’s avoided cost), then even with much higher tariffs for wind and new small hydro, the average tariff of the entire NCRE portfolio would be close to the avoided cost (thereby meeting the requirement that NCRE not increase consumer tariffs). But for whatever reason, these fore- casts proved optimistic. Subsequently, the SLPUC allowed these costs as part of the CEB’s general revenue requirements, but it is unclear whether this is to be a permanent feature of the tariff methodology.29 Impact of Renewable Energy Tariffs on the Consumer Figure 4.8 shows the tariff expectations of the baseline. In the absence of the additional NCRE, only the existing SHPs are assumed, whose tariff declines because, as the old PPAs expire, average compensation will be limited to the “tier 3” level of around 3 cents/kWh by 2025 (when the last 15-year PPA under the old tariff system expires). The CEA’s overall tariff also declines from the pre-coal era of 13 cents/kWh to around 10 cents/kWh by 2016, and to around 8 cents/kWh by 2025. Figure 4.9 shows the impact of the 10 percent RE target on the consumer tariff. These results are based on a financial model of the CEB that forecasts the CEB’s revenue requirements (and which include a return on equity). By 2020 the impact is about 1 cent/kWh (sold). The tariff increase is 8 percent in 2015, 12.5 percent in 2020, and 17.5 percent by 2025. This obviously conflicts with the above-noted stipulation that NCRE should not increase tariffs! The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 115 Figure 4.8 Tariff Expectations: Baseline, 2009–27 20 15 U.S. cents/kWh 10 5 0 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 Tariff Existing small hydro Thermal fuel cost Source: World Bank 2010. Figure 4.9 Tariff Impact of the 10 Percent RE Target: 2009 Forecast 20 15 U.S. cents/kWh 10 5 0 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 +Renewable energy Baseline Tariff increment Source: World Bank 2010. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 116 Case Study: Sri Lanka Nevertheless, it is not an unreasonable question to ask why, with a generally declining tariff that comes with the substitution of coal for oil, is an increasing penetration of more expensive renewable energy such a problem for the CEB: since the tariff is projected to decrease, the additional cost of NCRE simply means the tariff decline is somewhat smaller than it would otherwise have been. But at least in the short term, the main reason for the CEB’s objections is that its overall financial situation remains poor, and until such time as the CEB is in sustainable financial health, any incremental expenditure (and above all any claim on its cash flow) will be opposed. Only once the principle that incremental expenditures for renewable energy are treated no differently than purchases of energy from other IPPs in the retail tariff methodology, and its additional cost of working capital is allowed, will the CEB become indifferent to renewable energy—always assuming that a 10 percent share of non-dispatchable and inter- mittent energy poses no problems for network operation and stability. Table 4.16 shows the calculation of our comparative consumer tariff indicator that estimates the impact on the consumer tariff of a 1 percent increase in the quantity of RE generation. These estimates use the revised FITs of 2012, which particularly in the case of wind have somewhat lower rates than in 2009. According to the baseline demand forecast for 2020, generation is 23,950 GWh, sales are 20,870 GWh, and the average 2020 tariff (for the least-cost baseline) is 8.04 cents/kWh.30 The incremental energy required is 204 GWh, supplied by the mix of renewable energy as shown in the NCRE scenario of table 4.10. Table 4.16 Impact of a 1 Percent Increase in Renewable Energy Units 1 2020 Baseline generation GWh 23,952 2 Target energy to be replaced % 1.0% 3 Target energy to be replaced GWh 240 4 Target energy to be replaced Hydro Wind biomass other 5 Share 0.42 0.37 0.15 0.06 6 Generation GWh 101 89 36 14 7 Financial cost SL Rs/kWh 16.70 20.62 21.00 15.31 8 Financial cost Cents/kWh 14.5 17.9 18.3 13.3 9 Financial cost Cents/kWh 7.1 7.1 7.1 7.1 10 Incremental cost Cents/kWh 7.4 10.8 11.2 6.2 11 Total $ million 7.5 9.6 4.0 0.9 12 Total incremental cost $ million 22.0 13 Impact on consumer 14   Retail sales GWh 20,869 15   Average consumer tariff Cents/kWh 8.04 16   Total cost [RR] $ million 1,678 17   Tariff increase % 1.3 18   Tariff increase Cents/kWh 0.105 Note: GWh = gigawatt-hour; kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 117 The incremental cost is $22 million, which results in a tariff increase of percent (or 0.105 cents/kWh). 1.3 ­ Table 4.17 explains why renewable energy is seen by the CEB as costly. While incremental purchases from IPPs (RE producers) amount to $314 million, fossil- fuel savings are only $68 million. As noted earlier, some coal units would be displaced and deferred by NCRE, but the incremental decline in revenue requirements for debt service, and the CEB’s equity return on new projects, amounts to only another $33 million, for a net increase in revenue requirements of $216 million.31 Sri Lanka is perhaps an outlier, insofar as the fuel displaced by renewable energy may well be coal (since the storage in the many large hydro projects serves as the matching mechanism to the load curve)—but since coal is the cheapest of all of the fossil fuels, the incremental costs of renewable energy are correspondingly high. The Cost of Fossil-Fuel Subsidies Unlike many other case study countries (for example, Indonesia, the Arab Republic of Egypt, and Vietnam), Sri Lanka lacks its own fossil resources,32 and must import all its thermal fuels for power generation. Most of the fuel used in the power sector is auto diesel, of which a substantial fraction is imported, since the refinery at Sapugaskanda, operated by the government-owned Ceylon Petroleum Corporation (CPC), cannot meet the entire domestic demand; this diesel does not benefit from subsidy. Table 4.17 Impact of RE on CEB Revenue Requirements US$ million Fossil fuel costs −68 Fixed O&M 0 Debt service, principal −1 Debt service, interest −24 IPP capacity payments 0 Purchases from RE IPPs 314 Past debt service 0 T&D investment 0 Other expenses 0.2 LNG terminal costs 0 CEB equity return, new projects −8 Total revenue requirement 216 Source: World Bank 2010. Note: CEB = Ceylon Electricity Board; IPP = independent power producers; LNG = liquefied natural gas; O&M = operation and maintenance; RE = renewable energy; T&D = transmission and distribution. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 118 Case Study: Sri Lanka The question of whether the CEB benefits from subsidies relates mainly to heavy fuel and naphtha (as a fuel for some CCGTs). Residual fuel is delivered directly to the CEB’s heavy diesels, and is so viscous it must be heated for transportation. With both the CEB and CPC making large losses,33 it is not surprising that it is alleged (by the CPC) that the CEB benefits from subsidized heavy fuel. The problem is not that significant subsidies on this fuel are in place by explicit government design, but that during the years when international oil prices were rising—such as in 2010–11—the price adjustment mechanism was insufficiently flexible. Since domestic prices are fixed by the government, if these are not adjusted regularly and frequently, then the CPC incurs large losses because it buys at the increasing international price but has to sell at the still-to-be-adjusted domestic price. In this situation it is not surprising that the CPC has made losses in 2010 and 2011. But from the perspective of the RE policy, this is largely moot. The use of heavy furnace oil is expected to be phased out in the next few years (figure 4.10); residual oil will be used for somewhat longer (since there is no other use for it, and it cannot be exported).34 Financing New and Renewable Energy In the case of small hydro, with the ESDP and RERED support, the local banking sector seems comfortable with lending: the track record of the industry has been good to date, so capital mobilization for SHPs should not be a major problem Figure 4.10  Forecast of CEB Fuel Use, 2009–27 14 12 Fuel consumption, 106 tons 10 8 6 4 2 0 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 Coal trinco Coal Naptha Residual Furnace oil Auto diesel Source: World Bank 2010. Note: CEB = Ceylon Electricity Board. “Coal” is that used in the Puttalam (west coast) coal projects; “Coal Trinco” is the coal consumed at the projected projects in Trincomalee (on the east coast).35 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 119 given the end of the ESDP and RERED financing facilities—also shown by the large number of pending projects (see table 4.4). The present level of the small hydro FIT will allow projects to be built at much higher capital costs than the first set of SHPs. In the case of wind projects, note that two of the three wind projects recently completed are 100 percent equity, and the third is 50/50 equity/debt—made possible by the small size of the projects (less than 20 MW). The extent to which this reflects the skepticism of lenders about the economics of wind projects is not known, but such levels of equity do not provide a sustainable model for achiev- ing the NCRE wind energy targets. Conclusions The conclusions of Sri Lanka’s experience with renewable energy can be sum- marized as follows: • Targets. The 10 percent target for renewable energy was a political statement issued at a time of power shortages, and was not supported by a credible analy- sis of its economic impact. By end 2013, only 40 MW of wind power was in place, making the required 220 MW of wind that would be necessary to meet the 2015 target most unlikely. A new target of 20 percent by 2020 has now been proposed. • Design of incentive schemes. Sri Lanka’s ACT must be judged a great success, having enabled some 188 MW of grid-connected small hydro schemes in 77 projects. This was replaced by a FIT, which was the result of the alignment of interests of the government (which wanted to demonstrate practical ­measures in support of achieving its 10 percent 2015 RE target) and of the Small Hydro Developers Association (who faced a declining ACT with the increased pene- tration of coal, and wanted an increase rather than a decrease in tariff!). • Recovery of incremental costs. While the ACT was in operation, recovery of incre- mental costs was not an issue. But with the higher levels of support under the FIT, recovery of incremental costs in the first few years of the tariff have been problematic, because the expectations of offsetting tariff declines in the existing SHPs have not materialized. This has created difficulties in timely payment of the CEB’s invoices for its incremental costs, leading in turn to a halt of further issuance of Letters of Intent (LoIs). Although tariffs are expected to decline significantly in the face of the move from oil to coal, the CEB is still making large financial losses, and therefore will continue to oppose having to absorb the incremental costs of renewable energy until such time as it is in better financial health, and the tariff methodology explicitly recognizes that these incremental costs are part of the CEB’s legitimate revenue requirements. • Impact on consumers. The impact on the consumer tariff of an additional percent of renewable energy in 2020 is estimated at 0.15 cents/kWh, or an 1 ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 120 Case Study: Sri Lanka increase in the average consumer tariff of 1.3 percent (corresponding to an incremental cost at 2012 price levels of $22 million). Since the tariff is ­ projected to fall from the present 13 cents/kWh to around 8 cents/kWh (as a result of coal projects replacing expensive auto-diesel generation), this may be seen as an acceptable increase. But meeting the NCRE target of 10 percent implies a 13 percent tariff increase, which is unlikely to be acceptable. • Regulatory framework. The SLPUC has been slow to assume its tariff responsi- bilities, and there was much delay in the revision of the 2009 tariffs—the new tariff was only issued in mid-2012. • Fossil-fuel subsidies. The extent to which subsidies on the CEB’s purchases of heavy fuel have risen during the past few years of oil price increases is a func- tion not of an explicit intent to subsidize the CEB, but is a consequence of the slow oil price adjustment system. In any event, with the expected transition from oil to coal, this issue is moot—unless difficulties with implementing the Trincomalee coal projects again push Sri Lanka into oil generation. • Off-grid renewable energy. Unlike Vietnam, Sri Lanka’s village hydro program has been successful. It has developed an institutional model that is closely aligned to the capacities of its rural beneficiaries, and that can be replicated on a large scale. Rehabilitation of the estate sector SHPs is also promising. Notes 1. Although still owned by the government, it is mandated to run on commercial lines, and has made major progress in rehabilitating the distribution system: its losses were greater than 50 percent when it took over its new franchise area; its losses today are 8 percent. 2. Finally resolved in March 2011, when the new coal project at Norocholai, North of Colombo, started operation. 3. For a detailed assessment of the influence of tariffs on demand growth and the demand forecast, see Sri Lanka: Environmental Issues in the Power Sector (World Bank 2010) (hereafter cited simply as EIPS). 4. The CEB continued to serve the eastern province throughout the conflict period that ended in 2008, but new investment on grid extensions and service quality improve- ments were not implemented. Network losses (both technical and commercial) are high in the Eastern Province owing to lack of investment and poor supervision. In the Northern Province the CEB redeveloped the Jaffna distribution network after 1995, operated it as a mini-grid served by diesel-engine generators on short-term contracts, and continued to do so even after the end of the conflict until the transmission link was reestablished. Several other towns in the North were provided with a limited power supply by the CEB using diesel engines, intermittently throughout the duration of the conflict. Sri Lanka Electricity Act No 20 of 2009: http://www.pucsl.gov.lk/download​ 5. /Electricity /electricity20act202009.pdf. 6. http://www.pucsl.gov.lk/download/pucsl/regulatory20manual.pdf. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 121 7. NCRE is defined as small hydro (less than 10 megawatts, MW), wind, biomass, and other sources such as energy from agricultural waste, landfill gas, and municipal waste. 8. The others being rice, tea, and rubber. 9. Colonial planters were the first to tap hydropower in small streams to generate electricity and motive power for their plantation industries. It is estimated that around ­ 500 such micro-hydro plants had been in operation in the early part of the twentieth century. 10. But these sub-loan maturities are limited to 10 years including a maximum 2-year grace, and no more than $3 million to any individual project. 11. This would use 75 percent paddy husk and 25 percent fuel wood to produce 72 GWh/year, using a moving grate boiler. A survey around Trincomalee indicated that 240,000 million tons (MT) of paddy husk are available as unmanaged agricultural waste, which would otherwise be left to decay or be burned in the open air. The annual requirement of rice husk for the 10 MW power plant is 81,000 MT, which would account for 34 percent of the available rice husk in the area. To produce annual requirements of fuelwood (27,000 MT), an effective land area of 1,000 ha is required; Gliricidia is proposed as the fuelwood species. 12. At appraisal, the annual load factor was estimated at 27.5 percent. When the wind turbines went to tender, the turbine manufacturers asserted that the wind regime would provide just 22 percent. In the first few years of the project, the actual load factor was around 15 percent. The main problem was the location; because of various factors (including the air force, and the interest of a nearby bird sanctuary) the project had been moved to a less-favorable location. 13. This is much smaller than in Vietnam, where the average size of small hydro projects is 11 MW. 14. The Estate Micro Hydro Rehabilitation and Re-Powering Project (EMRRP), which is funded by the ADB Sustainable Power Sector Support project. 15. This is possible where heads are very high requiring a low volume of water, which means that the design flow is only a small fraction of the average annual flows. Many engineers would describe this as underutilization of the potential—since most of the wet season water remains unused. 16. Some of these villages now receive grid electricity, but they are now eligible to sell into the grid under the SPPA: Athureliya is a 21.8 kW village micro-hydro scheme that was the first to sell electricity. 17. Economic costs were estimated to be slightly lower, at $963/kW. 18. As discussed in chapter 3, throughout the typical short-to-medium term planning horizon, gas CCGT is forecast to run in Vietnam 24 hours/day until at least 2025. 19. In other words, the IPP should recover its capital costs, equity returns, and fixed operating costs in a fixed charge, independent of generation, with a variable charge ­ covering fuel costs. 20. See Siemens Power Technologies UK 2008. This study showed that with a targeted program of network reinforcement, by 2013 some 690 MW of embedded generation from renewable energy could be absorbed, under the assumption that the output of CEB generators would be reduced by this amount. 21. There are four such projects in Sri Lanka: Uma Oya (150 MW), Broadlands (35 MW), Moragolla (27 MW), and Ging Ganga (49 MW)—none of which are in the CEB’s least-cost plan at 10 percent. But as noted in section “Renewable Energy The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 122 Case Study: Sri Lanka Development,” at the lower discount rate of 8.5 percent these hydro projects are in the least cost plan. 22. And brought to northern Sri Lanka by a dedicated transmission line (including an underground cable section and back-to-back direct current [DC]). 23. In early 2012 many observers were still optimistic about Phase III of the EU-ETS, with a survey of forecasts showing a consensus expectation of around €8.5/ton. But with the April 2013 vote of the EU parliament on backfilling (which many observers thought would revive the market), and with the continuing economic malaise in the Eurozone, recovery of the CER price to former levels is seen as very unlikely. 24. The Uma Oya hydro project is under construction (for start-up in 2015). 25. Including waste-heat recovery. 26. Corresponds to a three-phase, 60 Ampere supply, which is the highest rating for a retail supply. 27. Precisely this symmetry of risk sharing was introduced in the Vietnam ACT, 10 years later (see section “Design of Incentive Schemes” in chapter 3). 28. This stipulation that there be no impact on end-use tariffs would appear to preclude an RE levy on electricity similar to the universal charge in the Philippines. 29. In Vietnam, as noted in chapter 3, purchases of renewable energy from small power producers are a pass-through in the retail tariff methodology (just as are purchases from the system operator for conventional power). 30. This is considerably lower than the present tariff of SL Rs 13.42/kWh (12.2 cents/kWh). 31. There is no change in IPP capacity payments, since these are fixed in the PPAs and are unaffected by more or less renewable energy. 32. Indonesia has oil, coal, and gas; Vietnam has coal and gas; and Egypt has gas. 33. In 2012 the CPC lost SL Rs 94.5 billion ($859 million), and the CEB lost SL Rs 65 billion ($591 million). 34. Fuel oil has long been exported, because the product mix at the refinery does not match that of the domestic market. 35. Trincomalee was the originally proposed location for the first coal project, as recom- mended by the original feasibility study in 1986, mainly for it excellent deep-water sheltered harbor, which would allow coal impacts on cape-size vessels. But this area was under the control of the Tamil Tigers, so the site for the coal plant was shifted first to the south, and then to the Puttalam area, north of Colombo (where the first coal project is now operating). But with the end of the conflict, Trincomalee is again the preferred site for coal projects. Bibliography Elliot, D., M. Schwartz, G. Scott, S. Haymes, D. Heimiller, and R. George. 2003. “Wind Energy Resource Atlas of Sri Lanka and the Maldives.” NREL, U.S. Department of Energy. Fernando, S. 1999. An Assessment of the Small Hydro Potential in Sri Lanka. Colombo: Resource Management Associates. http://www. microhydropower.net/download/esd​ _smallhydro.pdf. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Sri Lanka 123 Siemens Power Technologies UK. 2008. Technical Assessment of the Generation Absorption Capacity of the Sri Lanka Power System. Final Report to DFCC Bank, Colombo, Sri Lanka. Siyambalapitya, T. 2001. Study on Grid Connected Small Power Tariff in Sri Lanka. Washington, DC: World Bank. SLSEA (Sri Lanka Sustainable Energy Authority). 2011. Sri Lanka Energy Balance, 2011. Colombo, Sri Lanka. ———. 2012. Energy Statistics 2012. Colombo, Sri Lanka. Vernstrom, R. 1995. Published Small Power Purchase Tariff for Sri Lanka. Report to the World Bank and the Ceylon Electricity Board, Washington, DC, May. World Bank. 2003. ESD Implementation Completion Report. World Bank, Washington, DC. ———. 2010. Sri Lanka: Environmental Issues in the Power Sector. World Bank, Washington, DC. ———. 2014. Sri Lanka Overview. World Bank, Colombo. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 5 Case Study: Indonesia Sector Background Indonesia’s power industry expanded rapidly from the early 1980s to the late 1990s, when the Asian financial crisis seriously disrupted the Indonesian econ- omy. But since then, the power sector has been gradually recovering, especially in the past few years. By June 2012 the total installed generation capacity of the national power system reached 35,167 megawatts (MW),1 making it one of the largest in Southeast Asia. But given the size of its population, Indonesia’s per capita electricity consumption, at 655 kilowatt-hours (kWh) per capita per year, and electrification ratio, at 71 percent,2 are still low compared to other middle- income countries. The state-owned national power company, PT Perusahaan Listrik Negara (PLN), has the constitutionally mandated responsibility for Indonesian electric- ity supply. It is a vertically integrated power company and generates, transmits, and distributes most of the electricity in the country. PLN is solely responsible for Indonesia’s transmission systems. But acting as the single authorized buyer at the wholesale level, PLN buys electricity from an increasing number of indepen- dent power producers (IPPs) and some large captive power plants. In 2011 PLN sold 160 terawatt-hours (TWh) of electricity to some 45.9 million customers nationwide.3 The Geological Agency of Indonesia (2010) estimates that Indonesia holds some 28,500 MW of geothermal resources, a significant proportion of the global potential. As of 2012, however, only some 1,190 MW of geothermal power capacity had been commissioned. Nevertheless, Indonesia ranks third behind the United States (3,093 MW) and the Philippines (1,904 MW) in terms of installed geothermal power generation capacity. The government has announced an ambi- tious target for the development of this resource (9,500 MW by 2025), the bulk of which is to be achieved by the private sector. But renewables development must be put into the context of the overall PLN generation plan, which is overwhelmingly coal. According to PLN’s current investment plan, the share of coal in the generation fuel mix will increase from around 35 percent in 2012 to roughly 70 percent by 2020 (figure 5.1). A range The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   125   126 Case Study: Indonesia Figure 5.1 The Forecasted Generation Mix GWh 450 400 350 300 250 TWh 200 150 100 50 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Geothermal Hydro Gas Imports Coal MFO Biomass LNG HSD Source: PLN 2012b, figure 5.3. Note: GWh = gigawatt-hour; HSD= high-speed diesel; LNG = liquefied natural gas; MFO = marine fuel oil; TWh = terawatt-hour. of major challenges faced by PLN—and that also have a claim to PLN’s limited financial resources—make achievement of the geothermal target very difficult. Large Investment Requirements Significant investments from both the public and private sectors are required to meet fast-growing energy demand, and to increase access to modern and sustain- able energy solutions for all. PLN’s financial condition is critical to the financial viability of the power sector as a whole, and the sector’s ability to attract the large amounts of capital required to keep up with the growing electricity demand. The magnitudes involved are formidable: according to PLN’s latest invest- ment plan (2012–22), between 2012 and 2021 the total investment requirement is $107 billion, of which $77.2 billion is for generation, $16 billion for transmis- sion, and $13.8 billion for distribution. Although PLN’s $2 billion bond issue in 2012 was oversubscribed (PLN 2012a), and there are plans for about one-third of the new generation additions to be provided by the private sector, the chal- lenge for PLN is clear. Low Rate of Electrification Electrification levels remain low, especially outside Java-Bali. The current national electrification rate is 71 percent, leaving 78 million people without elec- tricity access, or access to only very unreliable non-grid supply. Most of those without access to electricity live in the remote areas of Java and Bali, or in islands The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 127 outside the area covered by the Java-Bali system. In 2011 the electrification rate was 76 percent in Java-Bali and 64 percent in the rest of Indonesia. The govern- ment has set an electrification target of 92 percent by 2021.4 High and Highly Subsidized Tariffs Electricity tariffs that are significantly below the cost of supply undermine PLN’s financial viability and lead to large government subsidies. In 2011, the subsidy of Rp 93 trillion accounted for 45 percent of PLN’s total revenue requirements of Rp 206 trillion ($21 billion) a sharp increase from 2005 when the subsidy was just Rp13 trillion (World Bank 2012). Tariffs below cost-recovery levels are the main barrier to improving energy efficiency and achieving greater private sector participation. At the same time, Indonesia’s average cost of power generation is very high (15.5 cents/kWh in 2011), which is in large part attributable to the unusually high share of oil in the thermal generation fuel bill.5 Gas Supply Subsidized domestic gas prices and underdeveloped gas infrastructure have caused delays in the expansion of the domestic gas sector and have contributed to a suboptimal generation mix: the consequences of gas shortages in the short run are either that more oil is used to meet intermediate and peak demand, and/or that coal projects are used as load followers (with significant efficiency penalties). In the longer term, as applies to the investment plan, less domestic gas translates into more (expensive) liquefied natural gas (LNG). Geographical Imbalance In 2012 the Java-Bali system accounted for 77 percent of total PLN sales; Sumatra, 15 percent; and all the rest of Indonesia, just 8 percent (table 5.1). As discussed below, this creates great difficulties for the implementation of geother- mal energy, much of which is in remote provinces far from the institutional center in Jakarta. The island fragmentation also creates significant problems for electric- ity planning: PLN uses the Wien Automatic System Planning (WASP) IV capacity expansion planning model for seven major grids, plus a further 97 ­ systems with peak demand of more than 1 MW for which a simpler supply-demand balance model is used to forecast generation requirements. Much of the planning work for the Eastern Islands is devolved to PLN’s regional offices. Table 5.1 Regional Imbalances of Electricity Supply 2012 2021 TWh TWh Expected annual growth rate Sumatra 26 15% 62 17% 10.5% Java/Bali 132 77% 259 72% 7.9% Rest of Indonesia 14 8% 37 10% 11.4% Total 172 100% 358 100% Source: PLN 2012b. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 128 Case Study: Indonesia Renewable Energy Development and the Resource Endowment Lack of incentives, a complicated and uncertain regulatory environment ­ combined with the relatively weak institutional capacity of central and local institutions, and the weak and low coverage of transmission networks, has hindered substantial development of renewable energy (RE) resources, espe- cially geothermal, hydropower, and biomass. Geothermal Systematic development of geothermal energy began with the enactment of the 2003 Geothermal Law. This opened geothermal development to direct private participation through competitive tendering, and provided an active role for the regional government to conduct these tenders and issue licenses. Prior to this law, geothermal work areas were awarded to developers on a memorandum of under- standing (MoU) basis without competition. But progress since enactment of Law 27/2003 has been slow. The tender ­ process for geothermal working areas (known as Wilayah Kerja Pertambangan Panas Bumi [geothermal work areas as known in Bahasa, Indonesia] WKPs)6 has revealed various impediments to rapid expansion of geothermal power capacity. The pricing framework for geothermal power has been revised several times, but the cost differential between geothermal power and coal-fired generation has only recently been taken up by the government. A Ministry of Finance (MoF) regulation provides that the incremental cost of geothermal energy be funded under the public service mechanism (PSO).7 So-called legacy WKPs are those that were awarded prior to 2003; a number of these are proposed to be expanded in the near future. The current develop- ment framework provides for private sector entities to take on the bulk of the exploration risk. Government involvement in geothermal energy is through a Pertamina subsidiary (Pertamina Geothermal Energy, PGE). Small Hydro As in Sri Lanka, small-scale hydropower started in the tea plantations. In West Java, one of the main tea regions in Indonesia, the first turbine was installed in 1885. At this time turbines were providing shaft power to tea rollers and other machinery in the tea factory, but not directly driving genera- tors. Later, with advancing turbine and generator technology, hydroelectric power plants were built. In 1910, 40 private tea plantations owned hydro- power plants, and by 1925 there were 400 such projects with a total capacity of some 12.5 MW. In the modern era, off-grid hydro has been actively pro- moted, but comprehensive information about the extent of these projects is not available. Nevertheless, as noted below, the target of several hundred megawatts of off-grid hydro should be within reach. Moreover, because many of these systems are on small islands that will never become accessible to the interconnected national grid, these are much less likely to become abandoned (as, for example, in Vietnam). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 129 Renewable Energy Targets The National Energy Plan 2006 sets targets for a range of RE technologies (table 5.2), with the share of renewable energy in the primary energy supply to grow from 4.3 percent in 2006 to 17 percent by 2025. The 9,500 MW target appears to be reasonable in light of the resource estimates published by Indonesia’s geological agency. But as with estimates of ­ wind potential in Vietnam, these are quite misleading since they are divorced from economics. A 2007 study estimates the exploitable potential at 9,000 MW, spread across 50 fields, with a maximum potential of 12,000 MW (WestJEC additions 2007). PLN’s latest investment plan (2012–22) anticipates total capacity ­ by 2021 of 57,300 MW, so the additional 8,200 MW of geothermal capacity represents 14 percent of the required capacity additions. But with geothermal ­ capacity typically requiring investment at $4,000–5,000 per kilowatt (kW), com- pared to coal at $1,500/kW, the geothermal proportion of total capital investment is much greater. In 2010 the Ministry of Energy and Mineral Resources (MEMR) set an addi- tional target of 3,967 MW to be achieved by the end of 2014—a target that is unlikely to be achieved given the current installed capacity of around 1,300 MW. This target lists 43 projects, of which 37 (3,627 MW) are to be developed by the private sector, and 6 (340 MW) by PGE. Most of the larger projects are on Java or Sumatra. Production Costs The price per kWh for operating geothermal projects, and for WKPs under development where the price is available, is shown in table 5.3.8 There is no evidence of scale economies, except that very small projects below 10 MW have costs above 9.5 cents/kWh (figure 5.2).9 Geothermal Development Policy Issues It is widely asserted that a major reason for the slow pace of geothermal develop- ment in Indonesia is that most developers are reluctant to assume exploration risk (Fichtner 2011). It is held that only the largest of companies with access to Table 5.2 Renewable Energy Targets: The 2006 National Energy Plan Installed capacity, MW Geothermal 9,500 Biomass 810 Grid-connected small hydro 500 Off-grid small hydro 330 Wind 255 Solar 80 Source: National Energy Plan 2006. Note: MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 130 Case Study: Indonesia Table 5.3 Prices for Geothermal Projects Project Status MW Cents/kWh Tangkuban perahu 1 Under development Java 110 5.56 Tampomas Under development Java 40 6.24 Silangkitang (Sarula 1a and b) Under development Sumatra 110 6.79 Namora i langit (Sarula 2) Under development Sumatra 330 6.79 Suoh sekincau Under development Sumatra 220 6.90 Darajat 1, 2, and 3 Existing West Java 255 6.95 Kamojang 1−4 Existing West Java 200 7.03 Cisolok-cisukarame Under development Java 50 7.09 Sibayak Existing North Sumatra 12 7.10 Bedugul Existing Bali 10 7.15 Wilis/ngebel Under development Java 165 7.55 Sorik merapi Under development Sumatra 55 7.96 Kaldera dano Under development Java 110 8.35 Wyang Windu 1 and 2 Existing West Java 227 8.39 Salak 4, 5, 6 Existing West Java 377 8.46 Ijen Under development Java 110 8.58 Ungaran Under development Java 44 9.08 Muaralaboh (liki oinangawan) Under development Sumatra 220 9.40 Atadei Under development LEMBATA 5 9.50 Rajabasa Under development Sumatra 220 9.52 Dieng 1 Existing West Java 60 9.81 Sokoria Under development FLORES 5 13.03 Jaboi Under development SABANG 7 17.78 Jailolo Under development TERNATE 10 18.01 Source: For projects under development, the prices are as shown in Castlerock Consulting (2010, exhibit 4.2). Note: kWh = kilowatt-hour. Figure 5.2 Electricity Price vs. Installed Capacity 20 15 USc/kWh 10 5 0 100 200 300 400 Installed capacity, MW Operating Under development Note: kWh = kilowatt-hour; MW = megawatt; USc = U.S. cents. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 131 balance sheet financing (which in Indonesia would include Chevron and PGE) have the necessary equity to adequately fund exploration, and that smaller developers whose projects are ultimately dependent upon nonrecourse project ­ finance simply do not have the necessary equity to carry the exploration risk. Moreover, only the large entities can themselves mitigate this risk by portfolio diversification.10 Therefore, the argument is made that to attract more and better entrants, ways must be found to mitigate this risk. The problem arises because under the Indonesian geothermal development model, bidders for WKPs have had to esti- mate an electricity price prior to exploration (albeit subject to renegotiation when the business license is awarded once a decision to proceed is made by the developer). It is not generally possible for a project company to raise debt finance for exploration work, and therefore the question arises whether enough equity can be raised commensurate with the risks and rewards. But the extent to which smaller companies have in fact been discouraged from bidding for WKPs, or have lost tenders to larger competitors for reasons of an uncompetitive electricity price, is unclear. The Castlerock Consulting study (2010, exhibit 4.2) presents a list of WKPs and their status (as of December 2010), and, where tenders are complete, the winning tender and in some cases the “price per kWh”11—but provides no information on the number and identity of the unsuccessful bidders, or on the prices offered by unsuccessful bidders. Indeed, there are many other reasons why projects are stalled: while Castlerock identifies 13 WKPs with “commercial problems,”12 12 WKPs have permit problems (predominately land and forestry permits). Discussions with developers confirm ­ that permitting issues are a major problem, particularly for land and forestry per- mits, and these represent one of the main obstacles to timely implementation.13 But discussions with developers also suggest that a major problem lies in the tendering process: many successful bidders for smaller projects are unable to deliver projects at the excessively low prices bid. Indeed, the $10 ­million perfor- mance bond requirement for winning bidders is not enforced, and the bid bonds are typically far too low to discourage speculators and unqualified entities. Exploration Risk For all the general discussion about exploration risks, no rigorous quantification of the exploration risk has been undertaken to date, which would inform us to the extent to which higher tariffs would in fact mobilize the additional equity required. But a study that assessed Indonesian drilling success performance suggests that geothermal exploration risk in Indonesia is smaller than in most other countries with prospective geothermal resources (Sanyal and others 2011). Average depths per well are smaller than elsewhere, and megawatts per well are higher. The ­ average megawatt per successful well, which was around 6 MW initially, has gradually increased to 9−10 MW, and the overall success rate—with over 200 wells drilled, is now around 60 percent. Similar learning curve effects are observed at individual fields: in the well-developed Kamojang The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 132 Case Study: Indonesia field, the drilling success rate—initially around 40 percent, is now 70 percent. The study concludes that the current well development cost per megawatt is $300,000−$400,000, so with a median well size of 9 MW the average cost per well is $3 million−$4 million.14 The view among developers is less optimistic. Many note that the “low-­ hanging fruit” has already been picked, and that unit drilling costs are on the increase. Costs for a full-sized development well (seen in recent feasibility stud- ies) are around $6 million,15 but some expect costs to be $8 million and more. Capacity Reductions Another yardstick for gauging the extent of risk during the exploration phase is the change in estimated capacity between expectation at the outset and current capacity. If exploration risks are high, one would expect the final project design to have lower capacity than originally expected. But of 52 WKPs listed in the Castlerock status survey, 47 show no change. One project has increased its esti- mated capacity from 205 MW to 220 MW (Chevron project in Suoh Sekincau, Lampung, Sumatra), and five projects show a decrease (table 5.4). A somewhat higher failure rate is implied by the Castlerock reassessment of overall geothermal potential in the WKPs. That analysis takes into account the probabilistic variations in input parameters, and the revised potentials represent the expected value of commercial potential. Of the 52 WKPs examined, only 10 show no change, 7 show an increase, 20 show a decline, and 14 (or 27 ­ percent) show zero potential. This compares to the current overall drilling failure rate of 38 percent (Sanyal and others 2011). The Renewable Energy Supply Curve Castlerock prepared supply curves for all major geothermal areas in Indonesia, for which the curves for Java-Bali are shown in figure 5.3. These recognize the wide range of uncertainty in the supply curves, having been derived with a probabilistic model of geothermal exploration and exploitation.16 These are compared to the avoided costs of coal (6.1 US cents [USc]/kWh) and with the costs of coal plus local and environmental externalities (8.1 USc/kWh)—­ following exactly the procedure recommended in section “Renewable Energy Table 5.4 Changes in Capacity Project Location Developer Original MW Revised MW Change MW Suoh Sekincau Sumatra Chevron 205 220 15 Tampomas Java Wika Jabar Power 45 40 −5 Salak Java Chevron 40 0 −40 Darajat Java Chevron 110 0 −110 Parutra 1, 2, and 3 Java GDE 180 55 −125 Wayang Windu 3 and 4 Java Star Energy 240 110 −130 Source: Castlerock Consulting 2012. Note: MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 133 Figure 5.3 Supply Curve for Java-Bali 16 14 12 LCOE, USc/kWh 10 8 6 4 2 0 0 200 400 600 800 1000 1200 1400 Cumulative installed capacity, MW Pessimistic Coal + externality (8.1 USc/kWh) Average Coal (6.1 USc/kWh) Optimistic Source: Castlerock Consulting 2010, exhibit 2.6. Note: kWh = kilowatt-hour; LCOE = levelized cost of energy; MW = megawatt; USc = U.S. cents. Development and the Resource Endowment,” the economic quantities are given at the point where the supply curves intersect the avoided costs of coal. These supply curves need to be updated, because drilling costs have increased signifi- cantly since 2009 when these curves were prepared. This supply curve (and similar ones for Sumatra and the Eastern islands) reveal rather lower estimates of potential than the government expects—if 10  cents/kWh were set as the feed-in tariff (FIT) (or as the tariff ceiling), then the potential is just another 1,300 MW on Sumatra and 900 MW on Java-Bali (under the average cost and capacity conditions shown in figure 5.3). Expressed differently, to achieve the target of 9,500 MW is wishful thinking not just because of the state of the actual resource, but the incremental cost required to achieve it. Carbon Accounting and CDM With the large number of new coal projects, the greenhouse gas (GHG) emis- sions of Indonesia’s power sector are expected to increase from 150 million tons of carbon dioxide (CO2) per year to 280 million tons by 2021 (figure 5.4). These estimates are based on the Intergovernmental Panel on Climate Change (IPCC) default emission factors. The 2010 Castlerock report states that sale of certified emission reductions (CERs) could cover a significant portion of the incremental costs of geothermal (up to 38 percent of total incremental cost in 2014, 45 percent in 2016, The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 134 Case Study: Indonesia Figure 5.4 PLN CO2 Emissions Million Tons/Year 300 250 200 Million tons CO2 150 100 50 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Coal LNG Gas MFO HSD Source: PLN 2012b. Note: CO2 = carbon dioxide; LNG = liquefied natural gas. and 31 percent in 2020).17 Indeed, the economic analysis presented in that report assumes CER revenue at $20/ton CO2. But as discussed in section “Carbon Accounting and CDM” in chapter 4 (see table 4.9), the prospects for such carbon prices in the next five to seven years are minimal, given the over- registered CERs. supply of ­ Design of Incentive Schemes Over the past 15 years, Indonesia has issued a number of incentive schemes for renewable energy: • The 1995 avoided cost tariff (ACT) for small RE producers. • The competitive tariff and tendering scheme for geothermal development (in a series of regulations to implement the Geothermal Law of 2003). • The geothermal fund. • Feed-in tariff for geothermal of August 2012. The Avoided Cost Tariff of 1997 Indonesia introduced an ACT in December 1995.18 Indeed, the ACT and the standardized power purchase agreement (SPPA) were the cornerstones of a World Bank/Global Environment Facility (GEF) RE small power project approved in 1997. The basic principle of the tariff was that it was to be based on 100 percent of PLN’s avoided costs, but differentiated by region. The original scheme envisaged nonfirm contracts with a single energy charge, and The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 135 firm contracts with energy and capacity charges (the latter with complicated escalation provisions to hold constant the rupiah-dollar exchange rate for the first five years, with the objective of protecting foreign debt service obligations). But with the disruptions of the 1997–98 financial crisis, few small hydro projects (SHPs) were completed and the Bank’s project—which would have made loans available at longer tenors—was cancelled in late 1998.19 Geothermal Law of 2003 In 2003 the Government of Indonesia issued Geothermal Law No. 27/2003, which required all new geothermal concessions to be competitively tendered for development. To be consistent with the country’s law on decentralization, the authority to carry out most geothermal tenders rested with the local or provincial governments. But most sub-national institutions lacked the capacity and experience to carry out multimillion-dollar international tenders, and many public institutions faced capacity constraints in planning and managing geothermal developments. The result was a number of poorly structured geothermal development oppor- tunities being tendered and none achieving financial closure. With a lack of pre- liminary information regarding the field and the credibility of the information offered being questioned (despite Indonesia having a vast database of mapped geothermal fields and related information), many leading geothermal developers did not participate in the tenders. Those that did participate proceeded to rene- gotiate the terms after the concession was awarded. Since the tenders did not include a standardized power purchase agreement (PPA) with PLN, the financial prospects of the offer were undermined. In practice, the electricity price offered by a developer at the WKP stage could be (slightly) renegotiated at the award of the geothermal business license (known as Izin Usaha Panas Bumi (IUP),20 though in 2010 the government set an upper bound of 9.7 cents/kWh. The renegotiation has been limited to indexation and escalation:21 the base price itself (i.e., the price set in the first year of commercial operation) is not subject to change. The Geothermal Fund In 2012 the government created a $220 million geothermal fund for the pur- poses of geothermal exploration, funded by the state budget, and administered by a unit of the MoF. In principle, this can fund up to $30 million per WKP for geothermal exploration. But the precise workings of this fund have yet to be sorted out, though there are several proposals for how this might work—­ including the possibility of a secured loan to a license holder for a new or legacy WKP to conduct exploration. But the problem yet to be solved is what happens if the exploration program does not lead to a commercial development: if the developer remains at risk (secured through collateral) then there is little benefit to take out the loan. But with 100 percent collateral demanded as security for a loan, there have been no takers thus far. The government is now considering a proposal to use the fund to provide up-front de-risking of WKPs prior to tender, as a public good, with resource data ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 136 Case Study: Indonesia from at least three wells. Because the cost of private equity for funding first-stage exploration is very expensive, de-risking provided as a public good can achieve significant tariff reductions (from 1 to 3 cents/kWh), particularly in smaller proj- ects in the Eastern Islands, that attract little interest from the big developers. The smaller the project, the greater is the impact of such up-front de-risking (Meier, Lawless, and Randle 2014a). The proposal is for the costs of up-front exploration to be recovered from the developer only at the time of financial closure, when most of the risk has been taken out of the project, and the costs of recovery in the tariff can be achieved at a weighted average cost of capital (WACC) far below the returns required by early-stage, high-risk private equity. The 2012 Geothermal Feed-In Tariff A fixed FIT for grid-connected geothermal projects was introduced on August 16, 2012 (table 5.5).22 This replaced the earlier system of competitively bid tariffs that was part of the geothermal tender system. These new tariffs are higher than currently paid by plants in operation in Java and Sumatra (6.95–9.81 cents/kWh),23 and higher than the ceiling price of 9.7 cents/kWh for projects currently under development. The expectation of the government was that the higher tariffs would motivate developers to accelerate geothermal development, given that over the past few years progress in achieving government targets has been slow. The FIT was based on the recommendation of the tariff study by Castlerock Consulting (2010, 2012), which proposed that the tariff be based on the cost of the alternative fossil generation. Geothermal projects operate at high annual plant factors (85–95%), and unlike most other RE forms, which are non-­ dispatchable (such as wind), they serve as an excellent substitute for base-load coal, upon which generating capacity. Indonesia is relying for the bulk of its future base-load-­ But this tariff has been unsuccessful: not a single PPA has been signed under this tariff, which has been widely criticized (for a counterfactual see the geother- mal development in the Philippines and Kenya, summarized in box 5.1). The MEMR has recognized the problems, and is currently considering a new tariff issuance that returns to the previous system of competitively tendered projects subject to a price ceiling based on the benefits of geothermal energy. Table 5.5 The New Geothermal Feed-In Tariff (Established in 2012) Tariff (cent/kWh) No Region High voltage Medium voltage 1 Sumatra 10 11.5 2 Java, Madura, and Bali 11 12.5 3 South Sulawesi 12 13.5 4 North Sulawesi 13 14.5 5 NTB, NTT, Maluku, and Papua New Guinea 15 16.5 6 Maluku and Papua New Guinea 17 18.5 Source: Ministry of Energy and Mineral Resources. Note: kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 137 Box 5.1 Counterpoint: Geothermal Development in the Philippines and Kenya The Philippines’ build, operate, transfer (BOT) model for geothermal development has been more successful than Indonesia’s model. With 1,900 MW of installed capacity, the Philippines is the leading developing country for successful geothermal development (see table B5.1.1). The key difference is that geothermal risk is taken by the government-owned geothermal company. The first application of the BOT-based geothermal public-private partnership (PPP) in the Philippines is the World Bank−supported Leyte-Cebu Geothermal Power Project—a 200 MW geothermal project to be implemented by a private firm through a BOT contract with PNOC EDC, the publicly owned national geothermal development company. PNOC EDC provides the exploration and development of the geothermal field, while the power plant contractor designs, supplies, installs, and commissions the plant for a predetermined cooperation period of 10 years. During the cooperation period, PNOC EDC pays for the plant through an energy conversion tariff (essentially a BOT fee), which covers operating costs and provides for capital recovery and return on capital. Plant ownership is transferred and handed over to PNOC EDC at the end of the cooperation period. Finding commercial funding for the private BOT contractors was not a problem because the exploration (geothermal resource) risk and the off-take risk were carried by the state through PNOC EDC and the National Power Corporation (NPC), the national power utility. Furthermore, payments to the BOT contractor were backed by a government undertak- ing in case of default by PNOC EDC or the NPC. In Kenya the January 2010 FIT included a fixed tariff for geothermal. The stated objectives of the Kenyan FIT system are to facilitate resource mobilization (by providing investment s ­ ecurity and market stability for investors) and reduce transaction and administrative costs and delays (by eliminating the conventional bidding processes). The tariff provides a fixed payment of 8.5 cents/kWh delivered at the interconnection point for 20 years, and is subject to an SPPA. It applies only to the first 500 MW (first come, first served), and only to plants not less than 70 MW. Table B5.1.1  Global Installed Geothermal Capacity, December 2010 Installed capacity, Geothermal generation, Share of geothermal in generation 2010, MW GWh mix, % United States 3,093 17,014 0.4 Philippines 1,904 10,723 17.6 Indonesia 1,197 8,297 5.6 Mexico 958 7,056 2.7 Italy 843 5,520 1.7 New Zealand 628 4,200 9.6 Iceland 575 4,038 24.5 Japan 536 2,752 0.3 El Salvador 204 1,519 25.5 Kenya 167 1,180 16.7 Costa Rica 166 1,131 11.9 Sources: ESMAP 2012; Kenya Energy Regulatory Commission 2012. Note: GWh = gigawatt-hour; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 138 Case Study: Indonesia Detailed Design of the Geothermal Feed-In Tariff The FIT introduced in August 2012 has been a failure, and the reasons for that failure are worth reviewing in some detail. Lack of Transparency Developers were confused about the basis of the tariff, since normally a “feed-in tariff” is understood to be based on the estimated costs of the producer, not the avoided costs of the buyer.24 The ministry (MoF) published no information on the basis, methodology, and assumptions of the tariff. The regulation was silent on the impact of changes in law on the tariff. According to the MoF Regulation 9/2012, there is a 5 percent royalty on steam and 2.5 percent royalty on the gross electricity price. There are also fixed fees payable for exploration $2/hectare (ha), and $4/ha during exploitation. Were these royalties and fees to change in the future, it is a reasonable question for developers to ask whether and how it is intended that the FIT be adjusted. But these concerns are testimony to the misunderstandings surrounding ter- minology. There should indeed be some provision for updating, but if the tariff is based on avoided costs, it is the avoided costs of the buyer, not the costs of the seller, that need review. Transmission Costs The tariff was silent about transmission costs. It was unclear whether the FIT was to include or exclude the costs of connection. Tendering Under the existing system, tenders involved price competition, albeit subject to some renegotiation of the PPA at a later time. But the FIT fixed the tariff in advance: for any particular region, all now have access to the same tariff. But if there is to be no price competition, then on what basis are developers to be selected for new WKPs? A new selection methodology would be required. The MEMR (and Castlerock) have suggested a “quality selection” approach (“beauty contest”)25 but this becomes increasingly difficult if there are several contestants who can all demonstrate financial strength and documented geo- thermal experience, making a qualitative differentiation subjective. Castlerock argues that such a process is the basis for selecting oil and gas developers in the United Kingdom. But a successful licensing program based purely on qualitative factors requires significant institutional and technical expertise that is at least as good as the companies seeking licenses, likely to be true in the case of the United Kingdom and New Zealand (which has a similar licensing round pro- cess) but unlikely to be true for the MEMR, and even less likely to be true of tenders in the hands of regional governments. Indeed, several developers have expressed frustration at some of the prequalification processes conducted in Indonesia, where obviously qualified and experienced developers have not made the shortlist.26 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 139 Economic Basis for the Tariff The economic basis for the FIT was provided by Castlerock. The methodology was to calculate the cost of coal (or diesel) generation in six regions of the ­ country, then add premia for fuel volatility, local environmental externalities, and global externalities. In short, the basis is the avoided social cost of thermal generation—which is coal in the large systems, and diesel in most of the outlying ­ small islands—in some of which the PLN investment plan calls for small coal projects where there are no geothermal resources (3 MW or less) because government policy requires that no new diesels be built.27 The result of these ­ calculations is shown in table 5.6: these are suggested as “minimum values” and are based on levelized costs based on some WACC, and a coal cost of $80/ton. This analysis raises several issues: • The basis of the global environmental cost ($10/ton CO2) is based on a review of the European Union Emission Trading Scheme (EU-ETS) and carbon mar- ket conditions. Even if it were appropriate for a global environmental cost to be reflected in an Indonesian tariff, the value to use should be based not on the current state of carbon markets, but on studies of actual damage costs (such as the Stern Report, or the American Inter-agency Working Group on the Social Cost of Carbon).28 • The “fuel volatility” adjustment is incomprehensible. • The estimate of local environmental damage cost is based on arbitrary adjust- ments (see box 5.2 for details of the damage cost estimates). Whatever may be these objections, Castlerock emphasized several important aspects of the tariff—none of which, sadly, were incorporated into the tariff as issued by the MEMR in August 2012: • The need to clarify the responsibility for the costs of transmission. • The need to regularly update the tariff and to stipulate a mandatory review period based either on time (for example, every two years) or on number of tenders completed (for example, every 5 or 10 tenders). Table 5.6 Avoided Costs of Thermal Generation Cents/kWh Others Java-Bali Sumatra Sulawesi Coal (small islands) Diesel (80% coal/20% diesel) Conventional 7.7 8 9.8 13.6 44.2 19.7 Fuel volatility 0.1 0.1 0.2 0.2 0.8 0.3 Local environmental costs 0.5 0.3 0.3 0.3 0.05 0.22 Global environmental cost 0.8 0.9 0.8 0.8 0.8 0.8 Total cost 9.2 9.3 11.1 14.8 45.9 21.1 Source: Castlerock Consulting 2010, exhibit 3.14. Note: kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 140 Case Study: Indonesia Box 5.2 Local Environmental Damage Costs of Thermal Generation in Indonesia The most recent assessment of the local environmental damage costs from Indonesian ­ thermal generation projects is by Kusumawati, Sugiyono, and Bongaerts (2010), who studied damage costs at the Paiton coal project, the Gresik gas project, and the Muara Karang oil project. These projects have emission factors for the main criteria pollutants as follows: ­ Emission Factors, Grams/kWh Project SO2 NO2 PM-10 Paiton coal 4.34 4.56 0.67 Muara Karang oil 11.7 2.32 0.29 Gresik gas 0 1.79 0 Using the SIMPACT model, damage costs per kWh were estimated as follows: Damage Costs, Cents/kWh (at 2010 Price Levels) Gresik Muara Karang Paiton Gas HFO Coal PM-10 0 1.301 0.207 SO2 0 0.517 0.016 NO2 0.051 0.063 0.008 Sulfates 0 0.148 0.042 Nitrates 0.036 0.173 0.045 Total 0.087 2.202 0.318 These damage cost estimates differ slightly from those estimated by Liun, Kuncoro, and Sartono (2007), who use the same SIMPACT model as Kusumawati, Sugiyono, and Bongaerts (2010): Damage Costs, Cents/kWh (at 2010 Price Levels) Gresik Muara Karang Paiton Suralaya Tanjung Jati Gas HFO Coal Coal Coal Kusumawati, Sugiyono, and Bongaerts (2010 prices) 0.087 1.301 0.207 Liun, Kuncoro, and Sartono 200729 0.074 0.097 0.646 Kusumawati, Sugiyono, and Bongaerts (2010) and Liun, Kuncoro, and Sartono (2007) use U.S. damage cost estimates, adjusted by purchasing power parity adjusted per capita gross domestic product (GDP) (though as noted in chapter 2 of this report, since this procedure is clearly invalid across the EU countries, it is not clear why this should be valid for the even greater ­differences between the Organisation for Economic Co-operation and Development [OECD] and developing countries). None of these estimates can be considered reliable, except for the order of magnitude of damage costs.30 Source: Kusumawati, Sugiyono, and Bongaerts 2010; Liun, Kuncoro, and Sartono 2007. Note: HFO = heavy fuel oil; kWh = kilowatt-hour; PM-10 = particulate matter (no greater than 10 microns in diameter); NO2 = nitrogen dioxide; SO2 = sulphur dioxide. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 141 Level of the Tariff The additional incentive provided by the new tariff can be gauged by a compari- son with existing tariffs. In principle it appears that the range of FITs is above the tariffs negotiated in the past. But under the existing arrangements, the price negotiated at the PPA stage is generally subject to escalation and indexation. Some developers have noted that at least in the case of Sumatra, the FIT of 10 cents/kWh may not constitute an improvement over the present ceiling of 9.7 cents + future escalation. Indeed, the lack of transparency led to a situation where the MoF believed the tariff would lead to higher prices, and the developers that it would lead to lower prices. Stakeholder Consultation In part the confusion about the tariff was a consequence of the complete lack of stakeholder consultation. The MEMR issued a tariff without proper consultation with the two parties most affected: the MoF, who bears the incremental costs, and the developers themselves. Impact of the Tariff Arguably the main deficiency was that the MEMR issued a tariff without under- standing what would be the impact on incremental costs. Under the present Indonesian tariff system, these are carried by not by PLN, but by the MoF—who is under intense pressure to reduce subsidies. Conclusions For developers and their lenders to have confidence in the tariff system, its cal- culation must be according to a known methodology. It is therefore important for the MEMR to state the rationale for the tariff, and present the calculations involved. The tariff as issued, and the apparent reluctance of the MEMR to issue the necessary clarifications, makes this a textbook example of a poor regulation, which in fact has hindered resolution of the problems of geothermal develop- ment rather than being helpful. Table 5.7 shows our summary evaluation of the various stages of the geothermal tariff implementation. Incremental Costs and Their Recovery New estimates have recently been made of the probable level of incremental costs, based on the current estimate of PLN’s avoided cost for a base load genera- tion of 6.7 cents/kWh. The level of subsidy will depend upon the outcome of tender bids, and on the outcome of the many renegotiations of past PPAs that are currently being sought—outcomes that will be subject to tariff ceilings based on the avoided costs of the buyer plus adjustments for externalities (for example a valuation of avoided GHG emissions based on $30/ton CO2) and local regional economic development multipliers. Table 5.8 shows the results of The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 142 Case Study: Indonesia Table 5.7 Summary Evaluation of Tariff Designs 2003 geothermal law Avoided cost tariff (tendering and negotiation) New feed-in tariff Introduced 1997 2003 2012 Achievement to date, MW 0 1,300 0 Economically efficient Yes Yes No Market principles Yes Yes No (competitive tendering) Sustainable recovery of Yes No No incremental costs (by definition) (though not the fault of the (though not the fault of the tariff tariff design itself) design itself) Transparency No Yes No (no published tariff) MEMR published no explanations Clarity of transmission Yes Yes No provisions (included in scope of final (no mention in tariff) negotiation) Adaptability Yes Yes No (in principle) (no provision for updating) Successful? No Yes No (Asian financial crisis) Note: MEMR = Ministry of Energy and Mineral Resources. Table 5.8 Impact of Subsidy on Assumptions (Java and Sumatra) MoF subsidy if tender prices are at: @tariff ceiling Installed Incremental LCOE with @tariff adjusted for Ceiling capacity capacity LCOE de-risking ceiling de-risking [1] [2] [3] [4] [5] [6] [7] Cents/kWh MW MW $ million $ million $ million $ million PLN avoided cost (2014) 6.7 186 186 0 0 0 0 Old ceiling 9.7 1,583 1397 120 104 168 152 Higher ceilings 11.0 1,900 317 197 141 298 242 11.5 1,900 0 197 141 298 242 Proposed ceiling 12.5 1,949 49 214 149 316 251 13.5 2,028 79 248 170 345 267 14.0 2,094 66 277 188 368 279 15.0 2,156 62 305 206 388 290 16.0 2,237 82 348 234 413 299 17.0 2,292 55 381 256 428 303 18.0 2,332 40 407 274 438 305 19.0 2,332 0 407 274 438 305 20.0 2,362 30 430 291 445 306 Source: Meier, Lawless, and Randle 2014a. Note: kWh = kilowatt-hour; LCOE = levelized cost of energy; MoF = Ministry of Finance; MW = megawatt; PLN = Perusahaan Listrik Negara (Indonesian State Electric Utility Company). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 143 these calculations, as a function of the proposed tariff ceiling, and as a function of tender outcomes31: • Tender prices at or near the levelized cost of energy (LCOE), as based on the Castlerock supply curves (column [4]). • LCOE adjusted for lower tariffs as are expected from the up-front de-risking using the Geothermal Fund (column [5]). • Tender prices under the pessimistic assumption that a published tariff ceiling will push up bid prices to the ceilings (column [6]). • At the tariff ceiling adjusted for up-front de-risking (column [7]). While the calculated incremental costs for the currently proposed ceiling of 12.5 cents/kWh are modest compared to the overall magnitude of the MoF tariff subsidies to PLN ($149 million–$316 million per year), that is only for a geothermal total of some 2,000 MW. This should be compared to the $10 ­ billion of overall subsidy provided by MoF to PLN in 2013 to cover the shortfall between revenue requirements and the consumer tariff. However, to achieve the much more ambitious target as envisaged by the so-called second fast-track program (FTP2)—an additional 4,925 MW on top of the existing 1,335 MW by 2020—the incremental costs will be much greater, on the order of $800 ­ million–$1.0 billion. Potential Impact of Incremental Costs on the Consumer Because of the disconnect between the PLN tariff and the consumer tariff, the impact of the incremental costs of geothermal energy is in the first instance felt as an increase in the PSO (that is, an increase in the subsidy provided by the MoF), rather than as an increase in the consumer tariff. Table 5.9 shows the Table 5.9 Impact of Incremental Costs Units 2020 PLN sales GWh 310,000 1% as geothermal energy GWh 3,100 Incremental cost Cents/kWh 0.058 $ million 179.8 Cost-reflective tariff Rp/kWh 1351 Cents/kWh 15.49 Revenue requirement $ million 48,007 Incremental cost to PLN % 0.37 Consumer tariff Rp/kWh 737 Cents/kWh 8.45 Consumer bill $ million 26,189 Incremental cost to consumer % 0.69 Note: GWh = gigawatt-hour; kWh = kilowatt-hour; PLN = Perusahaan Listrik Negara (Indonesian State Electric Utility Company). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 144 Case Study: Indonesia impact of 1 percent additional geothermal energy, assuming a worst-case average incremental cost of 5.8 cents/kWh. The impact on the tariff is just 0.17 percent, significantly less than the corresponding impact in Vietnam (1.1 percent) or Sri Lanka (1.3 percent)—a reflection of the relatively small incremental cost of geothermal power compared to wind. Buying Down the Price of Renewable Energy with International Assistance Buying down the value of renewable energy is well illustrated by the example of the Ulubelu Geothermal Project, one of two PGE geothermal projects sup- ported by the World Bank. The total (economic) cost for the 110 MW project is $359 million ($3,300/kW). Alternative financing approaches can be assessed as follows: • The project is funded entirely by PGE equity, for which the relevant cost of capital is the cost of equity (assumed in the Project Appraisal Document [of the World Bank] [PAD]32 at 14 percent). • The debt is funded by the International Bank for Reconstruction and Development (IBRD) for which typical terms are 30-year London inter-bank offer rate (LIBOR)33 of 3.87 percent (as of November 2010) + 1.15 percent fixed spread = 5.02 percent, and a term of 24.5 years with a 9-year grace period. • The debt is funded by highly concessional carbon finance—in this case by the Clean Technology Fund (CTF) for which typical terms are 0.25 percent ser- vice charge over 40 years, with a 10-year grace period. • The actually proposed financing: 44.3 percent PGE equity, 32.2 percent IBRD, and 23 percent CTF.34 The distinguishing feature of geothermal financing is the very long period of capital investment—here assumed at eight years (four years for exploration plus four years of construction). This is longer even than a large hydro storage project or nuclear project (assuming no litigation delays), and stands in stark contrast to other RE projects such as wind, for which two years would be a comparable preoperational period for a 110 MW project. Thus the presumption is that debt finance will start only in year 6 (at the start of construction), and that all the exploration is funded by equity. This stands in sharp contrast to the standard IPP model, in which equity contributions are required pari passu with debt (excepting up-front development expenses carried by the developer). Table 5.10 shows the results of the financial analysis. The financial model calculates the tariff that would be necessary to achieve a 14 percent financial internal rate of return (FIRR) for PGE (which is the assumption of the PAD). As an equity-only project, the required tariff is 10.3 cents/kWh (which is above the 10 cents/kWh FIT for Sumatra). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 145 Table 5.10  Buying Down the Incremental Costs PGE FIRR nominal Tariff Incremental cost (1) % Cents/kWh $ million All equity 14 10.3 827.7 IBRD only 14 8.3 320.9 Blended, as proposed 14 7.7 185.9 CTF only 14 7.0 0.0 Note: CTF = Clean Technology Fund; FIRR = financial internal rate of return; kWh = kilowatt-hour; IBRD = International Bank for Reconstruction and Development; PGE = Pertamina Geothermal Energy. If the debt portion is funded entirely by the IBRD, the required tariff falls to 8.3 cents/kWh, and the incremental costs fall to $321 million; if funded entirely by the CTF, the tariff required is exactly 7 cents/kWh (the assumed cost of coal generation). For the blended IBRD/CTF financing, the incremental costs are bought down to $185 million. The Environmental Costs of the Electricity Subsidy There are few countries in the world that provide as large a subsidy to elec- tricity as Indonesia.35 In 2011 the PLN’s costs would have required a tariff of Rp 1,351/kWh (15.5 cents/kWh), but the actual average tariff was just Rp 737/kWh (8.45 cents/kWh). Of course, one of the reasons for the high tariff is Indonesia’s unusual dependence on oil for power generation. But while tariff increases will (obviously) reduce the consumer surplus, this is more than offset by a decrease in the PLN’s costs, recapturing the 2011 dead- weight losses of $790 million/year. Moreover, this comes with a reduction of GHG emissions of some 20 million tons/year (assuming the impact is a reduc- tion in coal generation). In rows [17]–[23] of table 5.11 we show the quantity of geothermal power that would be necessary to achieve the same level of GHG emission reduc- tion: under the optimistic assumption of an average incremental cost of just 3 cents/kWh, 3,021 MW of geothermal would be required, at an incremental cost of $675 million/year. With the expected growth of generation, by 2021, the deadweight losses asso- ciated with the subsidy rise to $1.65 billion (column [1]); elimination of the subsidy would reduce GHG emissions by 42 million tons/year. To achieve the same GHG emission reduction by geothermal would require 6,387 MW, at an incremental cost of $2.1 billion. These calculations are obviously a function of the assumed price elasticity (table 5.12), and are notional because elimination of the present level of subsidy would need to be phased in gradually. At lower levels of (long-run) price elastic- ity, the impact of subsidy elimination will be less. Moreover, the PLN’s costs and its tariff requirement should decline over the next few years as oil generation is gradually phased out, so the amount of tariff increase associated with elimination of the subsidy would be smaller. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 146 Case Study: Indonesia Table 5.11 Impact of Electricity Subsidies, 2011 and 2021 2021 2011 Units [1] [2] Areas in box 5.3 1 Baseline tariff Rp/kWh 737 737 P1 2 Cents/kWh 8.4 8.4 3 New tariff Rp/kWh 1,351 1,351 P2 4 Cents/kWh 15.5 15.5 5 Real price increase % 83.31 83.31 6 Price elasticity −0.25 −0.25 7 Price elasticity adjustment % −14.1 −14.1 8 Demand contraction at consumer GWh 46,815 22,494 Q1–Q2 9 Q1 GWh 333,000 160,000 10 Q2 GWh 286,185 137,506 11 PLN avoided costs Billion Rp 204,462 98,240 +C+D+B 12 Loss of consumer surplus Rp billion −190,090 −91,334 B+C 13 Deadweight loss recaptured Rp Billion 14,372 6,906 D 14 $ million 1,647 792 15 Emission factor Kg/kWh 0.90 0.90 16 Avoided GHG emissions Million tons/year 42 20 17 Equivalent geothermal capacity MW 6,387 3,021 18 Average load factor [Proportion] 0.85 0.85 19 Geothermal generation GWh 47,558 22,494 20 Avoided GHG emissions Million tons/year 43 20 21 Incremental cost Cents/kWh 4.5 3 22 $ million 2,140 675 23 Avoided cost $/ton 50.0 33.3 Note: 2011 exchange rate: $1 = Rp 8,724. GHG = greenhouse gas; kWh = kilowatt-hour; GWh = gigawatt-hour; MW = megawatt; PLN = Perusahaan Listrik Negara (Indonesian State Electric Utility Company). Table 5.12 Impact of the Price Elasticity Assumption, 2011 GHG emissions, Required Incremental cost of Demand Deadweight million tons geothermal geothermal Price elasticity reduction, % loss, $ million CO2 /year capacity, MW capacity, $ million −0.05 −3.0 168 4 641 143 −0.10 −5.9 331 8 1,264 282 −0.15 −8.7 489 13 1,867 417 −0.20 −11.4 643 16 2,453 548 −0.25 −14.1 792 20 3,021 675 −0.30 −16.6 936 24 3,572 798 −0.35 −19.1 1,076 28 4,107 917 Note: CO2 = carbon dioxide; GHG = greenhouse gas; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 147 Nevertheless, the conclusion is inescapable: reducing large consumer tariff subsidies is win-win for the environment and the economy; 20 million tons/year GHG emission reduction incurs no incremental cost, but comes together with a macroeconomic benefit of $790 million/year. By 2021 the annual benefit of sub- sidy elimination increases to $1.6 billion, with 42 million tons of CO2 avoided. Box 5.3  Deadweight Losses of Electricity Tariff Subsidies At the subsidized price of P1, the quantity Q1 is demanded (160 TWh) in 2011. At the tariff ­ orresponding to PLN’s costs, P2 (Rp 1,351/kWh), the smaller quantity of Q2 is demanded c (286 TWh assuming a price elasticity of −0.25). At the old price P1, the consumer surplus is the area A + B + C. If the subsidy is removed, then at the higher price P2 the consumer surplus is only the area A—so there is a loss of ­ consumer surplus of (B + C ). But at the old price PLN makes a loss B + C + D, whereas at P2, PLN’s revenue covers its costs. So there is a gain to PLN of B + C + D. Therefore, the net gain is the area D—the so-called deadweight loss associated with the tariff subsidy. Note that the validity of consumer surplus (CS) to assess changes in welfare consequent to price changes is subject to important qualifications (see figure B5.3.1), notably that income and price elasticities and the magnitude of the price shift should all be small: where these conditions do not hold, then the method of equivalent variation (EV) or compensating varia- tion (CV) should be used (though these typically bracket the CS estimate, as shown in the numerical example provided in Bacon [1995]). The estimates presented here using CS may therefore be subject to an error of +10−20 percent. But at this level of uncertainty, the main message remains unchanged: large subsidies incur significant deadweight losses! Figure B5.3.1 Impact of Tariff Subsidy Price Demand curve A P2 Rp 1351/kWh D Tariff subsidy B C P1 Rp 737/kWh E F Q2 Q1 Quantity Note: kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 148 Case Study: Indonesia Conclusions Resource Endowment and Targets There is little question that Indonesia has a substantial geothermal resource endowment. But the magnitude of the target, 9,500 MW, is best described as wishful thinking, announced before an understanding of the incremental costs, and of the consequences for electricity subsidies. Much of the perception that the geothermal program is in trouble is related to apparently little progress in achieving the 9,500 MW target or even the revised 2014 target. But today’s 1,100 MW of geothermal capacity, achieved largely at quite modest incremen- tal costs, is no mean achievement: it is modest only when measured by an unreasonable yardstick. Design of Incentive Schemes In principle, an ACT can be successful in enabling significant private sector investment in renewable energy—as well illustrated by Vietnam’s ACT. But this is merely a necessary condition, not a sufficient condition. To be successful, such a tariff needs to be transparent and—as to methodology—be accompanied by a nonnegotiable SPPA, propose clear arrangements for transmission costs, and be clear about the magnitude of expected incremental costs and how these will be recovered. Unfortunately, the August 2012 FIT meets none of these conditions, and will likely be soon replaced. A tariff that is so misunderstood that some developers believed the published values to be lower bounds for negotiation, when in fact these were fixed tariffs, is obviously in trouble. Tendering Concession systems for natural resources work best where what is to be auc- tioned is well defined: the better the information available at the time of bid, the easier it is for bidders to make realistic assessments of the risk-return trade-off. From the perspective of achieving the government geothermal targets, it is the larger projects (most likely on Java and Sumatra, say those larger than 50 MW) where the efficiency gains of a tender system are likely to justify the transaction costs—where there is sufficient demand to absorb the output without difficulty, and where costs of exploration failure to the PLN system are small because even successful projects would represent just a few percent of the grid requirements. These are the projects where the bulk of the financial impact of the FIT on PLN subsidy requirements will be felt, and where efficiency gains are the most important. Indeed, it seems useful to make a distinction between larger projects whose main rationale is carbon emission reductions, and smaller projects on smaller islands whose principal rationale is a more cost-effective alternative to diesel generation—here geothermal costs are in the 12−20 cents/kWh range, and still provide a cheaper option than diesel generation. But for the smaller projects on smaller islands, the consequences of explora- tion failure for PLN are much greater, because of the difficulties of assuring The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 149 alternative supplies. These prospects will in any event be of less interest to large companies who are in a much better position to assume exploration risks. Whether “quality selection” works is primarily a function of the capacity of the government entities that make the selection to make informed judgments. That such capacity exists in the United Kingdom (U.K.) Ministry of Energy that regulates the U.K. oil and gas sector (held up by Castlerock as the model) seems reasonable, but whether such capacity exists in Indonesia, particularly in the regions, may well be questioned. Regulatory Framework and the Geothermal Fund While in principle the geothermal fund could be used for exploration, there is concern about the consequences of an unsuccessful exploration program—not all exploration programs will lead to a commercial prospect (indeed, if they were all successful, there would be no risk, and no need for the fund in the first place). But because there are in place severe penalties for misuse of government funds, there is a fear that the individual or entity that makes the final decision will be subject to prosecution for having “wasted” government monies. In the Philippines the exploration risk has been successfully passed to a state company established expressly for this purpose, but in Indonesia the 2003 Geothermal Law appears to require the private sector to assume this risk, and so rational application of the resources in the geothermal fund has been subject to inter- minable delays. It is unclear whether the oil and gas sector is the best model for the geother- mal sector, because in oil and gas there is a much closer alignment of interests between the oil company and the government: the primary interest of both parties is simply the maximization of physical production (barrels of oil), and the negotiation is simply about how that revenue is shared equitably between the two. Except in the highly unlikely case of the discovery of a giant superfield, the price of what is extracted is unaffected by the success or failure of the explora- tion program itself. But in the case of geothermal development, there are addi- tional conflicting public interests: on the one hand, the government wants the lowest electricity price possible (PLN, MoF), but also the maximum royalty revenue (local/regional government) and the maximum quantity (the MEMR has to meet its geothermal targets). Tariff Subsidies Few countries have electricity price subsidies as high as Indonesia: in 2011 the average tariff was just 8.5 cents/kWh compared to PLN’s cost of 15.5 cents/kWh. Such subsidies are also badly targeted, with small residential consumers captur- ing no more than 25 percent of the total subsidy. Although is there is some uncertainty about the price elasticity of demand, it is clear that reducing con- sumer subsidies is win-win for electricity costs and GHG emissions: to achieve the same reduction of GHG emissions as achieved with the elimination of sub- sidies, by 2021 some 6,400 MW of geothermal capacity would be needed, with an incremental cost of more than $2 billion per year. But reducing the general The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 150 Case Study: Indonesia electricity subsidy is accompanied by a benefit of $1.6 billion as the deadweight losses are recaptured. Buying Down Incremental Costs Highly concessional carbon finance—as represented by the CTF—is the key to buying down the incremental costs of renewable energy for developing countries. But how much additional carbon finance is actually available to Indonesia remains unclear. The Main Problem Few countries in the world have achieved significant development of their geo- thermal resources without a significant portion of the exploration risk being assumed by the state. Indonesia is the only country that modeled its geothermal development on the oil and gas sector, where it is typical that the private sector assumes the exploration risk—in return for substantial returns when exploration is successful. Most importantly, the price of the output, oil, is set by the interna- tional global oil market, whereas for geothermal, the output price is set by the government—and preferably as low as possible. Fortunately, revisions to the Geothermal Law are currently before parliament, which would declassify geo- thermal development as a “mining” activity, which should make environmental and forestry permits easier to obtain. But short of the repeal of the 2003 Geothermal Law, Indonesia must find some other path to move forward on geothermal development. The conclusions of this case study are straightforward: • Recognize the differences between large projects in Java and Sumatra, whose main objective is the avoidance of GHG emissions, and whose scale is sufficient to warrant the interest of international developers, and the smaller projects in the outlying islands, where the objective is to avoid the high cost of diesel generation. In these small projects, the transaction costs of tendering are high, and PGE should take the lead on behalf of the state. The geothermal fund should be able to support exploration in these areas, and since the beneficiary of successful exploration would be PGE, the issue surrounding possible waste of public funds to the benefit of the private sector (where exploration is unsuccessful) should not arise. In these areas the criterion for proceeding with development is whether the costs exceed the cost of diesel generation. • Clarify the arrangements for transmission. The responsibility for constructing the transmission line should be passed to the developer, and handed over at the time of commissioning. The costs should be recovered by an adder to the bid tariff (which should be limited to that of the generation project). • Update the resource supply curves developed in 2010 by Castlerock to provide a more reliable basis for estimating future incremental costs. Drilling costs in particular have risen dramatically since most of the cost estimates for Indonesia projects were developed in the 2004–07 time frame. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 151 • Replace the 2012 FIT with a new tariff that is based on a consultative process involving the main stakeholders, notably the PLN, the MoF, and the develop- ers. No tariff that is misunderstood by those primarily affected can be success- ful. The methodology of the tariff must be transparent, and published. • To be compliant with the 2003 Geothermal Law, tendering should continue to be price based, but with a focus on improving the tender process. The priority should be to create a new central entity to conduct tenders on behalf of pro- vincial entities (and indeed also serve this function for other forms of renew- able energy such as hydro). Most importantly, by up-front de-risking using geothermal fund resources, the number and quality of bidders is likely to improve. The commercial terms of the PPA should be fixed at the time of tender, and not subject to ad hoc, post-tender negotiations. Notes 1. Total PLN installed capacity as of June 30, 2012, is 35,169 MW. In addition, there were 26 IPPs with a contract capacity of 5,634 MW. 2. PLN Statistics 2011, table 19. 3. PLN Statistics 2011, tables 1 and 4. 4. In Indonesia a household is considered electrified if it has at least a solar lantern. village is considered electrified if at least 10 percent of its households are connected. A ­ 5. The subsidy mechanism is known as PSO (public service mechanism). 6. In Bahasa, Wilayah Kerja Pertambangan Panas Bumi, hence WKP. 7. Ministry of Finance regulation 111/2007. 8. Some of the prices are quoted in rupiah (Rp); these were converted at the exchange rate of Rp 9,590 per $1. 9. The prices shown in table 5.2 are the so-called base prices, applicable to the first year of commercial operation: in many PPAs, some portion of this base price is escalated at the U.S. producer price index. 10. For example, of the 52 WKPs in progress as of December 2011, PGE is the developer for 11 projects (Castlerock Consulting 2010). 11. It is also unclear from the report whether the price refers to that bid at the tender price, or whether the price is the final price negotiated in the power purchase agree- ment (PPA). 12. These include instances where winning bid prices are higher than the cap of 9.7 cents/kWh imposed by Permen 32/2009; where a price has been agreed, but the PPA has not been signed; and where a developer is unwilling to invest in further exploration absent a binding commitment from PLN to purchase power. 13. This is a problem not just for geothermal development, but for all of PLN’s generation projects. Currently there are negotiations under way for a service level agreement between PLN and 11 ministries that stipulate performance obligations among the parties. 14. For reasons that are unclear, the one statistic that is not presented in this report is the per well drilling cost. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 152 Case Study: Indonesia 15. See, for example, the feasibility study for Ullubelu, which presents costs of $6 million for development wells, and $1 million for reinjection wells. 16. The pessimistic assumptions are based on costs and capacity both one standard devia- tion lower than the average, the optimistic assumptions one standard deviation above the average. 17. Castlerock Consulting 2010, Exhibit 6.2. The assumption is that CER revenues are available only up to 2020, and that all CERs produced could be sold at that price. But it would be quite unusual for an Emissions Reduction Purchase Agreement (ERPA) to cover 100 percent of total potential CO2 emissions. 18. Known under the Bahasa Indonesian abbreviation PSKSK. 19. At the time of project appraisal in 1996, the prospects for a successful operation were good since the following conditions were in place: (a) the Indonesian authorities had put in place important policy and regulatory changes, (b) PLN was supportive of the project and willing to purchase power from all eligible private projects, (c) four of the strongest commercial banks in Indonesia had expressed interest in participating in the project, and (d) there was a firm pipeline of projects ready to use funding pro- vided by the Bank and GEF at a reasonably early date. But the value of the rupiah plummeted from Rp 2,341/$1 in September 1996 to Rp 17,000/$1 by January 1998. As a consequence, the capital costs in rupiah terms became too high, and the antici- pated investments were no longer viable commercially. 20. Izin Usaha Panas Bumi. 21. In the typical PPA, some part of the base price is indexed to the U.S. producer price index. 22. Decree of the Minister of Energy and Mineral Resources: Assignment to PT PLN to Purchase Power from Geothermal Plants and Standard Purchasing Price of PT PLN to Geothermal Power Plants, Number 22, year 2012, August 16, 2012. 23. See table 5.2 for details. 24. This confusion was not helped by Castlerock’s own comments. For example, Castlerock states that an FIT should be “dynamic, changing as external conditions change, for instance, if there is a sustained increase in the fossil fuel prices, the FIT should be adjusted accordingly” (Castlerock Consulting 2012, section 3.4, 3−21). This is correct for a FIT based on avoided social cost. But then Castlerock goes on to state that “if the FITs do not yield a level of developer interest (as demonstrated by the number and quality of firms competing for WKP tenders), then the government may consider raising the FIT.” But this is inconsistent with the concept of avoided cost— which should be completely independent of developer costs or developer interest. If one is going to make arbitrary adjustments to a tariff simply to achieve some target, then there is no point to claiming it is based on the avoided costs of PLN. 25. One hesitates to use this term given its generally negative connotations, but the term has been used by the MEMR, and appears in the Revision Matrix for Regulation 59/2007. 26. In any event, the scale and nature of the U.K. North Sea Oil and Gas licensing pro- gram is entirely different from that of the Indonesian geothermal WKP selection. For example, on October 26, 2012, the U.K. Ministry of Energy awarded 167 licenses for exploration rights in the 27th Licensing Round; 224 applications had been received for 330 blocks of the U.K. Continental shelf. This follows the award of 46 licenses in May 2012 for areas off the coast of Scotland. Licenses are indeed selected on the basis The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Indonesia 153 of the U.K. Energy Ministry judgments about the best exploration program rather than by auction; licensees pay a nominal escalating annual rental fee. There are strin- gent requirements about relinquishing acreage, environmental management, and non-­ performance to agreed work programs. 27. There are significant uncertainties about the practicality of a proliferation of small coal units in the small outlying islands, to say nothing of the potential environmental impacts. There is presently just one such project in operation using Chinese equip- ment, but local manufacture is being planned. 28. See section “The Social Cost of Carbon” in chapter 2. 29. The World Bank Ulubelu and Lahendong Geothermal Project economic analysis used a value of 0.546 cents/kWh as the coal damage cost, as stated as the average of the high and low estimates in the Liun study for Suralaya. We were unable to confirm these estimates in the original paper. 30. Indeed, none of these various studies have been published in the peer-reviewed literature. 31. See Appendix C for details of such calculations. 32. Project Appraisal Document (of the World Bank). 33. London Interbank Offered Rate. 34. For the project as a whole, which also includes the Lahendong Geothermal Project: the equity amounts are $125 million, CTF; $175 million, IBRD; and $243 million, PGE. 35. It is also worth noting that the subsidy is poorly targeted: the principal beneficiaries of the subsidy are large industrial and commercial customers; very small residential customers capture just 24 percent of the total subsidy disbursement (World Bank 2012). Bibliography Bacon, R. 1995. “Measurement of Welfare Changes Caused by Large Price Shifts.” World Bank Discussion Paper 273, Washington, DC. Baker & McKenzie. 2012. “Long Awaited Indonesian Geothermal Tariff Issued.” Client Alert, September 2012. Castlerock Consulting. 2010. Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs. Phase 1 Report, Report to Ministry of Energy and Mineral Resources, Castlerock Consulting, December 8, Jakarta, Indonesia. ———. 2012. Geothermal Power Development Project: Development and Implementation of Pricing and Incentive Policy. Phase 2, Final Implementation Report, Report to Ministry of Energy and Mineral Resources, May 21, Jakarta, Indonesia. ESMAP (Energy Sector Management Assistance Program). 2012. Geothermal Handbook: Planning and Financing Power Generation. Technical Report 002/12, ESMAP, Washington, DC. Fichtner. 2011. Geothermal Business Transaction Characteristics and Proposed Improvements. Report to the Ministry of Minerals and Natural Resources, Jakarta, Indonesia. Geological Agency of Indonesia. 2010. Geothermal Resources and Development in Indonesia. Jakarta, Indonesia. Kenya Energy Regulatory Commission. 2012. Feed-In Tariff Policy. 2nd rev. Nairobi, Kenya. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 154 Case Study: Indonesia Kusumawati, W., A. Sugiyono, and J. Bongaerts 2010. “Using the QUERI Model-AirPacts Program to Assess the External Costs of Three Power Plants in Indonesia with Three Different Energy Sources.” IMRE Journal 4 (1). http:// www.wordpress.hrz.tu-freiberg​ .de/wordpress-mu/journal/kusumawati.pdf. Liun, E., A. Kuncoro, and E. Sartono. 2007. “Environmental Impacts Assessment of Java’s Electricity Generation Using SimPacts Model.” Proceeding of International Conference on Advances in Nuclear Science and Engineering, 379−84. http://www.batan.go.id​ /­ppin/lokakarya/ICANSE07/article/D3.5-Edwaren_Environmental.pdf. Meier, P., J. Lawless, and J. Randle. 2014a. Indonesia Geothermal Tariff Reform: Tariff Methodology Report. World Bank and Asian Development Bank, Jakarta, March. ———. 2014b. Unlocking Indonesia’s Geothermal Potential. World Bank and Asian Development Bank, Jakarta, June. PLN (PT Perusahaan Listrik Negara). PLN Statistics 2011. PLN Annual Report, Jakarta, Indonesia. ———. 2012a. “$2million Global Medium Term Note Program.” Offering Memorandum, Jakarta, Indonesia, October 8. ———. 2012b. “Long-Term Investment Plan 2012–2022.” Jakarta, Indonesia. Sanyal, S., J. Morrow, M. Jayawardena, N. Berrah, S. Li, and Suryadarma. 2011. “Geothermal Resource Risk Assessment in Indonesia, a Statistical Inquiry.” Proceedings of the 36th Workshop on Geothermal Reservoir Engineering, Stanford University, January 2011. (This paper is based on the more detailed report by Geothermex, An Assessment of Geothermal Risks in Indonesia, Report to the World Bank, Washington, DC, June 2010.) WestJEC. 2007. Master Plan Study for Geothermal Power Development in the Republic of Indonesia. Jakarta, Indonesia. World Bank. 2012. Indonesia: Fuel to Power Value Chain Study: Preliminary Findings. Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 6 Case Study: South Africa Sector Background South Africa relies heavily on coal, which supplies around 70 percent of its ­ primary energy and more than 90 percent of its electricity. The country also has a highly energy-intensive economy. These combined factors mean that South Africa’s carbon emissions, on a per capita and gross domestic product (GDP) basis, are disproportionately high (although, in total, they amount to little over 1 percent of global emissions). South Africa does not face any formal commit- ments under the United Nations Framework Convention on Climate Change (UNFCCC) to mitigate climate change but, mindful of risks to its future international competitiveness, it has pledged to reduce its carbon emissions below ­ a business-as-usual scenario. It is within this context that South Africa has embarked on an ambitious renewable energy (RE) program to diversify its energy mix. The country has abundant wind and, especially, solar resources, although exploiting these still comes at a higher cost than its cheap coal (ignoring ­ externalities). After first exploring feed-in tariffs (FITs) for grid-connected RE, a competitive tender system has been implemented that has engendered a great deal of interest from private developers and financiers, and has seen prices fall in subsequent bid rounds. In 2012 South Africa ranked among the top 10 countries globally in terms of RE investments: over $9 billion1 was invested in 2,460 mega- watts (MW) of grid-connected wind, photovoltaic (PV), and concentrated solar power (CSP). The country presents an interesting case of having introduced incentive schemes for RE within an environment of heavy dependence on fossil fuels and a relatively low-cost electricity environment. South Africa’s publicly owned national utility, the Electricity Supply Commission of South Africa (Eskom), generates 96 percent of the country’s electricity, which amounts to just over half of the electricity generated in Sub- Saharan Africa. Private generators contribute about 3 percent of national output (mostly for their own consumption) and local municipalities contribute less than 1 percent. Power generation is heavily dependent on coal (92 percent) with nuclear, hydroelectricity, bagasse (from sugarcane), and emergency diesel-fired turbines accounting for the rest. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   155   156 Case Study: South Africa Eskom owns and controls the national integrated high-voltage transmission grid and distributes about 60 percent of electricity directly to customers. The remaining electricity distribution is undertaken by about 179 local authorities, which buy bulk electricity supplies from Eskom. Eskom imports power from Mozambique and, in the past, also has imported from the Democratic Republic of Congo and Zambia. It also sells electricity to neighboring countries (Botswana, Lesotho, Mozambique, Namibia, Swaziland, Zambia, and Zimbabwe). Imports and exports constitute about 5 percent of total electricity on the Eskom system. Eskom’s power stations are listed in table 6.1. Direct electricity sales to mines and industrial customers account for more than 40 percent of Eskom’s electricity sales. Eskom also operates retail distribution services for 4.65 million customers (4.5 million of these are house- ­ holds) and the municipal distributors service slightly more customers. About 85 ­percent of South Africans have access to electricity. Eleven of Eskom’s 13 coal-fired power stations are located in Mpumalanga Province in the northeast; the other two are at Lephalale in Limpopo Province Table 6.1 Eskom’s Power Stations Name Location Fuel Available MW Arnot Middelburg Coal 2,232 Camden Ermelo Coal 1,430 Duvha Witbank Coal 3,450 Grootvlei Balfour Coal 950 Hendrina Hendrina Coal 1,865 Kendal Witbank Coal 3,840 Komati Middelburg Coal 940 Kriel Bethal Coal 2,850 Lethabo Sasolburg Coal 3,558 Majuba Volksrust Coal 3,843 Matimba Lephalale Coal 3,690 Matla Bethal Coal 3,450 Tutuka Standerton Coal 3,510 Acacia Cape Town Gas/petroleum 171 Ankerlig Atlantis Gas/petroleum 1,327 Gourikwa Mossel Bay Gas/petroleum 740 Port Rex East London Gas/petroleum 171 Gariep Orange River Hydro 360 Vanderkloof Orange River Hydro 240 Drakensberg Bergville Pumped storage 1,000 Palmiet Grabouw Pumped storage 400 Koeberg Cape Town Nuclear 1,830 Total 41,847 Source: Eskom Annual Report 2012. Note: The table excludes four small, nonoperating hydro plants in Transkei. The balance of non-Eskom- generating capacity totals about 1,150 MW and is located mainly at Sasol’s synfuels plant (520 MW), Kelvin (128 MW), Rooival (155 MW), Pretoria West (100 MW), Steenbras (180 MW), and mini-hydro (65 MW). Eskom = Electricity Supply Commission of South Africa; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 157 and at Sasolburg. The two major hydro stations are located on the Orange River in the center of the country. Eskom’s Koeberg nuclear power station is located 30 kilometers (km) north of Cape Town. The open-cycle (kerosene/diesel) tur- bines are on the coast and are used for emergency peaking loads. Peak demand is also supplied by pumped storage schemes in the Cape and in the Drakensberg mountains in KwaZulu-Natal. The South African power system is thus character- ized by power stations that are concentrated in the interior near the mines and industries of Gauteng and Johannesburg, and long transmission lines down to coastal areas, which depend on power transfers from the northeast. Eskom embarked on a massive investment program in the 1970s and 1980s. It overestimated demand growth and, in the 1990s, there was significant over- capacity. But after a decade or more of little investment, Eskom is having to play catch-up. It is building two massive new coal-fired plants—Medupi and Kusile— each 4,800 MW, as well as a new pumped storage scheme, Ingula. At the same time, it has commenced procurement of its first RE power: a 100 MW wind farm, Sere, and a 100 MW CSP plant. These last two power projects have been funded mainly by several public lenders: the World Bank and African Development Bank, and the Clean Technology Fund. The engineering, procure- ment, and construction (EPC) contracts have been competitively bid for, but the final power costs will be blended (nontransparently) into Eskom’s average power generation costs. The government has also accepted that independent power producers (IPPs) should be allowed to enter the market. A rough 70:30 spilt between Eskom and the private sector was accepted by the cabinet after the Energy Policy White Paper was published in 1998. Work commenced on the design of a Nordpool-like power exchange. But with looming power shortages, the prospective competitive wholesale market was abandoned in 2004 in favor of a single-buyer model, with Eskom being the offtaker. For many years, however, the policy and regulatory framework for procuring IPPs was not put in place. As described below, this changed with the initiation of the RE IPP program in 2012. South Africa has a fairly rigid energy-planning system. By law, an electricity plan (Integrated Resource Plan, IRP) has to be produced by the Department of Energy (DOE) (although in practice this is delegated to the planners within Eskom). Based on this plan, the minister of energy makes periodic “determina- tions” of what power needs to be built and when. The regulator can only license new capacity within these ministerial determinations. The most recent IRP is for the period 2010−30 and is shown in table 6.2. For the first time it included RE options. These were “forced” into the plan as they were not least cost but were necessary for South Africa to meet its carbon mitiga- tion pledges, described in the following section. South Africa’s electricity once ranked among the cheapest in the world. Eskom’s average electricity sales price in 2007−08 was as low as 2.5 cents per kilowatt-hour (kWh). Effectively it had paid for much of its existing capacity, and prices were close to short-run marginal costs. But with the commencement of a new investment program of more than $50 billion (a large proportion of which The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 158 Case Study: South Africa Table 6.2 South African Integrated Resource Plan, 2010–30 Committed build DOE Cogene- RTS Ingula OCGT ration, capacity Medupi Kusile (pumped IPP own Landfill, Sere Decommis- (coal) (coal) (coal) storage) (­diesel) build Wind CSP hydro (wind) sioning MW MW MW MW MW MW MW MW MW MW MW 2010 380 0 0 0 0 260 0 0 0 0 0 2011 679 0 0 0 0 130 0 0 0 0 0 2012 303 0 0 0 0 0 300 0 100 100 0 2013 101 722 0 333 1,020 0 400 0 25 0 0 2014 0 722 0 999 0 0 0 100 0 0 0 2015 0 1,444 0 0 0 0 0 100 0 0 −180 2016 0 722 0 0 0 0 0 0 0 0 −90 2017 0 722 1,446 0 0 0 0 0 0 0 0 2018 0 0 723 0 0 0 0 0 0 0 0 2019 0 0 1,446 0 0 0 0 0 0 0 0 2020 0 0 723 0 0 0 0 0 0 0 0 2021 0 0 0 0 0 0 0 0 0 0 −75 2022 0 0 0 0 0 0 0 0 0 0 −1,870 2023 0 0 0 0 0 0 0 0 0 0 −2,280 2024 0 0 0 0 0 0 0 0 0 0 −909 2025 0 0 0 0 0 0 0 0 0 0 −1,520 2026 0 0 0 0 0 0 0 0 0 0 0 2027 0 0 0 0 0 0 0 0 0 0 0 2028 0 0 0 0 0 0 0 0 0 0 −2,850 2029 0 0 0 0 0 0 0 0 0 0 −1,128 2030 0 0 0 0 0 0 0 0 0 0 0 Total 1,463 4,332 4,338 1,332 1,020 390 700 200 125 100 −10,902 Source: Eskom’s Annual Report. Note: CCGT = combined-cycle gas turbine; CSP = concentrated solar power; DOE = Department of Energy; FBC = fluidized bed combustion; IPP = independent power producer; OCGT = open-cycle gas turbine; MW = megawatt; PF = pulverized fuel; RTS = rotary triboelectrostatic separator; PV = photovoltaic. was required to finance the two new coal-fired power stations), tariffs had to be increased to sustain Eskom’s financial viability (even though the utility success- fully accesses private capital markets and has secured a substantial sovereign guarantee). Figure 6.1 shows how electricity prices have risen in nominal and real terms. The regulator has agreed on above-inflation increases for the next five years. Renewable Energy Development While the official RE policy has not been very effective in applying practical implementation strategies, policies to mitigate climate change have had a much more profound impact. In many respects this is surprising. As a non−appendix A The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 159 New build options Coal (PF, FBC, Gas CCGT OCGT Import Total new Total system imports) (natural gas) (diesel) hydro Wind Solar PV CSP Nuclear build capacity MW MW MW MW MW MW MW MW MW MW 0 0 0 0 0 0 0 0 640 44,535 0 0 0 0 0 0 0 0 809 45,344 0 0 0 0 0 300 0 0 1,103 46,447 0 0 0 0 0 300 0 0 2,901 49,348 500 0 0 0 400 300 0 0 3,021 52,369 500 0 0 0 400 300 0 0 2,564 54,933 0 0 0 0 400 300 100 0 1,432 56,365 0 0 0 0 400 300 100 0 2,968 59,333 0 0 0 0 400 300 100 0 1,523 60,856 250 237 0 0 400 300 100 0 2,496 63,352 250 237 0 0 400 300 100 0 2,010 65,362 250 237 0 0 400 300 100 0 1,212 66,574 250 0 805 1,143 400 300 100 0 1,365 67,939 250 0 805 1,183 400 300 100 1,600 2,358 70,297 250 0 0 283 800 300 100 1,600 2,424 72,721 250 0 805 0 1,600 1000 100 1,600 3,835 76,556 1000 0 0 0 400 500 0 1,600 3,500 80,056 250 0 0 0 1,600 500 0 0 2,350 82,406 1000 474 690 0 0 500 0 1,600 1,414 83,820 250 237 805 0 0 1000 0 1,600 2,764 86,584 1000 948 0 0 0 1000 0 0 2,948 89,532 6,250 2,370 3,910 2,609 8,400 8,400 1,000 9,600 45,637 country under the Kyoto Protocol, South Africa does not face any commitments to reduce greenhouse gas (GHG) emissions. Nevertheless, research work was commissioned by the Department of Environmental Affairs on long-term mitiga- tion strategies, and these provided the basis for President Zuma to make a pledge at the Copenhagen Conference of Parties (COP) in 2009 that South Africa would reduce its carbon dioxide (CO2) emissions by 34 percent below a business-as-usual scenario by 2020 and by 42 percent by 2025, provided the ­ international community supported South Africa with financial aid and the trans- fer of appropriate technology. This peak, plateau, and decline scenario for carbon emissions subsequently informed the development of the IRP 2010−30. The power sector in South Africa contributes roughly half of South Africa’s carbon The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 160 Case Study: South Africa Figure 6.1 Average Nominal and Real Eskom Electricity Prices Cents/kWh 9 8 7 6 5 4 3 2 1 0 1996 1998 2000 2002 2004 2006 2008 2010 2012 Real 2008 values Nominal tariff Source: NERSA annual reports. Note: Eskom = Electricity Supply Commission of South Africa; kWh = kilowatt-hour. emissions and an effective emissions cap was set of around 275 metric tons (Mt)/year CO2 equivalent. A subsequent National Climate Change Response White Paper, published in 2011, provided a wider band for emission caps but maintained the peak, plateau, and decline trajectory. South Africa’s voluntary Copenhagen pledge to reduce its carbon emissions from a business-as-usual scenario set the stage for new procurement strategies for RE. Renewable Energy Targets In 2003 the government published a RE Policy White Paper that set a target of reaching 10,000 gigawatt-hours (GWh) of RE production by 2013. For years, very little was done to achieve this target and there was a great deal of confusion surrounding what this target actually meant: was it a cumulative or annual tar- get? Did it include RE sources other than electricity? The Department of Energy clarified that the target would be met by a combination of biomass, wind, solar, and small hydroelectricity. Design of Incentive Schemes FITs have been the most widely applied support mechanism internationally to encourage the growth of grid-connected RE. But have RE feed-in tariffs (REFITs) provided desirable or optimal outcomes in terms of affordable and competitive electricity prices? Could competitive tenders or auctions offer lower prices while still providing adequate incentives for RE suppliers to enter the market? South Africa at first explored the option of FITs but then abandoned them in favor of The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 161 competitive bids for RE (REBIDs). The initial outcomes have been encouraging: there has been a great deal of market interest and subsequent bidding rounds have seen prices fall. Could there be lessons in this for other countries? The Birth and Death of REFITs in South Africa South Africa relies on coal for electricity production. But in the face of climate change concerns, it has embarked on a transition to lower carbon-emitting tech- nologies. The electricity plan (IRP 2010) included, for the first time, ambitious targets for RE, namely 18,800 MW of wind and solar, out of a total projected system capacity of around 90,000 MW by 2030. In 2009 the National Energy Regulator of South Africa (NERSA) approved a REFIT policy. Tariffs were designed to cover generation costs—plus a real return on equity of 17 percent—and to be fully inflation indexed (NERSA 2009). The first-published FITs (assuming an exchange rate of R8 per $1)— 15.6 cents/kWh for wind, 26 cents/kWh for CSP (troughs with 6 hours storage), and 49 cents/kWh for PV—were generally regarded as generous by developers. But considerable uncertainty remained, including the legality of FITs within South Africa’s public procurement framework, and delays in finalizing power purchase agreements (PPAs) and interconnection agreements with the national utility, Eskom. In March 2011 the NERSA unexpectedly released a consultation paper with lower FITs, arguing that a number of parameters—such as the cost of debt and exchange rates—had changed. The new wind tariffs were 25 percent lower, CSP was down by 13 percent, and PV down by 41 percent (in nominal rand terms). Furthermore, the capital component of these tariffs could no longer be fully inflation indexed. Importantly, in its revised financial assumptions, the NERSA did not change the required return for equity investors of 17 percent (NERSA 2011). More policy and regulatory uncertainty was to come. After receiving legal advice that FITs were inconsistent with public finance and procurement laws, the DOE announced that a competitive bidding process for RE would be launched, known as the Renewable Energy Independent Power Producer Procurement (REIPPP) program. Subsequently, the regulator abandoned FITs: not a single megawatt of power had been signed in the two years since the launch of the REFITs (although it is probably fair to admit that a practical procurement pro- cess for REFITs was never actually implemented). These developments were met with dismay by many RE project developers that had secured sites and had initiated resource measurements and environmental impact assessments. ­ Subsequently, however, it was these early developers that were ready to benefit from the first round of competitive bids. The Birth of the REIPPP Program in South Africa The DOE, with the assistance of the Public Private Participation Unit in the National Treasury, and a phalanx of international transactional advisors, commenced work on bid documents. A Request for Qualification and Proposals ­ was issued in August 2011. A compulsory bidders’ conference was held in The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 162 Case Study: South Africa September of that year and attracted over a 1,000 participants, many from abroad. A total of 3,625 MW of new power capacity was offered with overall procurement caps for specified technologies—mainly wind and PV but also smaller amounts for concentrated solar, biomass, biogas, landfill gas, and small hydro (see table 6.2). The tender for different technologies was held simultaneously. Bidders could bid for more than one project and also for different technologies. Projects had to be larger than 1 MW and an upper limit was placed for different technologies: for example, 50 MW for CSP and 140 MW for wind projects. A further 100 MW was reserved for small projects below 5 MW. Price caps were specified for each of these technologies at levels not dissimilar to the NERSA’s 2009 REFITs, all of them much higher than the national utility’s average generation tariff of around 5 cents/kWh at the time. Standard 20-year, local-currency-denominated power purchase agreement (PPA) contracts were offered for the different technologies with the offtaker being the national utility, Eskom. Up to five discrete bidding rounds were envisaged, at more or less six-month intervals, with the first round of bids due in November 2011. Qualification Criteria In the first request for proposals (RfPs) the full 3,625 MW was made available. The evaluation process involved a two-step process. In the first, bidders had to satisfy certain minimum threshold requirements in six areas: environment, land, commercial/legal, economic development, financial, and technical. For example, wind developers were required to provide 12 months of wind data for the des- ignated site and an independently verified generation forecast. Project developers were responsible for identifying appropriate sites and for paying for measure- ment and early development costs at their own risk. Wind turbines had to be international standard International Electrotechnical Commission (IEC) 61400-1 certified. These economic development requirements, in particular, were com- plex, incorporating 17 sets of minimum thresholds and targets that needed to be met (table 6.3). For example, for wind projects, at least 12 percent of the share- holding in the project company had to be by black South Africans and a further 3 percent by local communities. At least 1 percent of project revenues had to go to socioeconomic contributions. The minimum threshold for local content was set at 25 percent, while a target of 45 percent was encouraged. Bid bonds or guarantees had to be posted, equivalent to $12,500/MW of nameplate capacity of the proposed facilities, and the amount was doubled once preferred bidder status had been announced. Bidders who satisfied the threshold requirements then entered the second step of evaluation where bid prices counted 70 percent with the remaining 30 percent weighting given to composite scores on job creation, local content, preferential procurement, enterprise development, and socioeconomic develop- ment. Bidders were asked to provide two prices: one fully indexed by inflation, the other partially indexed with the bidder being able to determine the propor- tion that would indexed. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 163 Table 6.3 Economic Development Threshold and Target Levels for Wind Energy Economic development factor Subcriteria Threshold Target Job creation South Africa–based employees who are citizens 50% 80% South Africa–based employees who are black citizens 30% 50% Skilled employees who are skilled black citizens 18% 30% South Africa–based employees that are citizens from local 12% 20% communities Local content Value of local content spend Wind 25% Wind 45% PV 35% PV 50% Ownership Shareholding by black people in the project company 12% 30% Shareholding by black people in the contractor responsible for construction 8% 20% Shareholding by black people in the operations contractor 8% 20% Shareholding by local communities in the project company 3% 5% Management control Black top management n.a. 40% Preferential procurement BBBEE procurement spend n.a. 60% QSEs and EMEs procurement n.a. 10% Women-owned vendors procurement n.a. 5.0% Enterprise development Enterprise development contributions n.a. 0.6% Adjusted enterprise development contributions n.a. 0.6% Socioeconomic development Socioeconomic development contributions 1.0% 1.5% Adjusted socioeconomic development contributions 1.0% 1.5% Source: South Africa’s Department of Energy. Note: BBBEE = broad-based black economic empowerment; EME = exempt micro enterprise; PV = photovoltaic; QSE = qualifying small enterprise; n.a. = not applicable. Round One Outcomes Fifty-three bids were received initially, totaling 2,128 MW. A large legal, techni- cal, financial, and governance evaluation team was assembled in a high-security environment with 24-hour voice and closed-circuit television (CCTV) monitor- ing. The team included local legal firms Bowman Gilfillan, Edward Nathan Sonnenberg, Ledwaba Mazwai, Webber Wentzel, and BKS, as well as interna- tional firms Linklaters for legal, Mott Macdonald for technical, and Ernst & Young and PricewaterhouseCoopers for the financial and governance reviews. The evaluation resulted in 28 qualifying bids, amounting to 1,416 MW of new capacity. For the first round, a deadline of July 2012 was set for financial closure (the date was later extended), and closure of development (COD) had to be reached by the end of 2014. Although bidders could not know for certain the total capacity that would be bid, they probably assumed that the tight deadlines, and challenging threshold qualification criteria, would result in the total capacity bid being less than the total made available in round one. Accordingly, prices bid were mostly uncom- petitive and only marginally below the caps specified in the RfPs. Direct and PPAs were signed in November 2012 between the government, Eskom, and each of the 28 successful bidders, resulting in a total investment of close to $6 billion. Much of the debt component was provided by local South African commercial banks. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 164 Case Study: South Africa Second Round The design of the second bid round incorporated the above lessons, and less capacity (1,284 MW) was offered to stimulate more competition (table 6.4). The second round closed in March 2012. Seventy-nine bids were received total- ing 3,255 MW; 51 of these bids met the qualifying criteria, of which 19 were granted preferred bidder status (next-best bids would have resulted in more than the full window being allocated). Wind and solar PV prices in the second round were much more competitive: on average, 20 percent for wind and 40 percent for solar PV (table 6.5)! The range of prices bid was also wider, with wind prices varying from 10 cents/kWh to 12 cents/kWh and solar PV from 17.5 cents/kWh to 22 cents/kWh. The price of CSP fell by 7 percent, with one preferred bidder taking up the remaining available capacity. There was little competition in small hydro, with only two qualifying bids, both at the capped price. Table 6.4 Capacity of Renewable Energy Made Available for Bids and Finally Allocated to Preferred Bidders Capacity (MW) Available in Allocated in Available in Allocated in Remaining in Technology round 1 round 1 round 2 round 2 round 3 Wind 1,850 634 650 562.5 563.5 Solar PV 1,450 631.5 450 417.1 401.1 Solar CSP 200 150 50 50.0 0 Small hydro 75 0 75 14.3 60.7 Landfill gas 25 0 25 0 25 Biomass 12.5 0 12.5 0 12.5 Biogas 12.5 0 12.5 0 12.5 Total 3,625 1,415.5 1,275.0 1,043.9 1,165.6 Source: South Africa’s Department of Energy. Note: CSP = concentrated solar power; MW = megawatt; PV = photovoltaic. Table 6.5 Prices for Renewable Energy: REFITs, REBID Caps, and Average Bids Price R/kWh Price cents/kWh Round 1 Round 2 Technology REFIT 2009 REFIT 2011 Bid cap average average Round 2 average Wind 1.25 0.94 1.15 1.14 0.90 11.25 Solar PV 3.94 2.31 2.85 2.76 1.65 20.63 Solar CSP 2.10 1.84 n.a. 2.69 2.51 31.38 Small hydro 0.94 0.67 1.03 n.a. 1.03 12.88 Landfill gas 0.65 0.54 n.a. n.a. n.a. n.a. Biomass 1.18 1.06 1.07 n.a. n.a. n.a. Biogas 0.96 0.84 n.a. n.a. n.a. n.a. Source: South Africa’s Department of Energy. Note: Prices assume full inflation indexing over a 20-year contract. CSP = concentrated solar power; kWh = kilowatt-hour; PV = photovoltaic; REBID = renewable energy bid; REFIT = renewable energy feed-in tariff; n.a. = not applicable. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 165 The bidders who preferred the second window also offered superior local content terms, with average local content for solar rising from 28.5 percent to 47.5 percent, wind rising from 21.7 percent to 36.7 percent, and CSP from 21 percent to 36.5 percent. The deadline for financial closure for round two was extended from the end of 2012 to March 2013. The remaining 1,167 MW was made available in the third bid round in May 2013. While prices have fallen in South Africa, they are not necessarily as attractive as those achieved in other countries. For example, Maurer and Barroso (2011) reports that in Brazil average auction prices for wind power fell from 9.8 cents/kWh in 2009 to 8.5 cents/kWh in 2010 to 6 cents/kWh in 2011. The same source quotes 6.9 cents/kWh for wind and 12 cents/kWh for PV in Peru. South African prices might be higher because of local content and economic development cri- teria. Interviews also suggest that the initial bidding round involved high transac- tion costs in terms of advisors and financing. These costs fell in round two (along with equipment prices) and are likely to fall further in subsequent rounds. Impact of Renewable Energy Tariffs on the Consumer Renewable power has been contracted at prices higher than the average Eskom generation cost and higher than the marginal cost of new coal-fired power. Eskom has signed 20-year PPAs with RE IPPs. The costs of these contracts are blended in with the costs from its other power stations. In its most recent application to the NERSA, Eskom estimated that power purchase costs from IPPs (mainly renew- able IPPs) would add 3 percent to the tariff, on average, over the next five years. There are no direct fiscal subsidies for grid-connected renewable power. Customers are paying the additional costs of renewable power. At present these additional costs are relatively modest and there has not been much public oppo- sition. But as the proportion of RE increases, and as consumers continue to face above-inflation tariff increases, this could become a more sensitive political issue. Conclusions The South African REIPPP program is not only the largest RE program in Africa, it is also the largest IPP program of any African country and probably the most complex public-private procurement ever run on the continent. According to Bloomberg New Energy Finance, South Africa ranked in the top 10 counties investing in clean energy in 2012, ahead of Canada, Brazil, Spain, and France. This is all the more remarkable, given South Africa’s previously dismal record in IPPs and the dominance of its national utility. Eskom, on the government’s instructions, had attempted to run a number of IPP procurements before, all of which failed. Ultimately, the Department of Energy and National Treasury had to wrest control of the REIPPP from Eskom. Although projects still have to achieve commercial operation, the South African REIPPP program can be considered a success in terms of attracting a multitude of private project developers and investors. In its second round, The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 166 Case Study: South Africa the REIPPP has also fostered competition with consequent, and impressive, falls in prices, which would in all likelihood not have happened in a REFIT program. And it has achieved this in record time: bids closed three months after the issuing of the RfP, preferred bidders were announced within a month, and contract sign- ing and financial closure were achieved 10 months later—even though as many as 28 projects, employing different technologies of different sizes at different sites, had to be processed in parallel. The following elements have contributed to this success: • The procurement process was well designed. Recognizing that there was little institutional capacity to run a sophisticated, multiproject, and multibillion dollar international competitive bid process for RE, South Africa’s Department ­ of Energy sought the assistance of the Public Private Participation Unit in the National Treasury who, in turn, relied extensively on local and international transaction advisors. • High standards were set and, apart from necessary clarifications, the government stuck to the announced schedule and core bid requirements (although the deadline for financial closure slipped a few months as the government finalized financial security arrangements). Despite a tight time schedule and tough qual- ification criteria, the REIPPP program attracted 58 bids in round one and 79 in round two. A significant number of these met the minimum qualification thresholds: namely, 28 in round one and 51 in round two. But it should be noted that the announcement of the REFIT two years before the launch of the REIPP contributed to early market interest, and a number of bidders had already identified sites and begun resource measurements. Prior to the issuing of the RfP, the DOE had also issued an earlier Request for Information from prospective project developers, which confirmed significant market readiness. • The design of subsequent bid rounds was flexible, allowing lessons to be incorpo- rated and thus improving the competitiveness of bids and prices. For example, it became apparent that the capacity made available in round one exceeded the capacity of the market to deliver, and tendered capacity was subsequently reduced in round two to induce more competition. • The RE sector is potentially highly competitive, given the diversity of energy sources, the modular nature of most of the technologies, and the number of project developers. When South Africa ran its first competitive tender for IPPs—two large gas turbine peaking plants—it received only two bids, one of which subsequently withdrew. It is, perhaps, no accident that the first success- ful international competitive tender for power in South Africa has been in RE. • Subsequent bid rounds have also incorporated more stringent thresholds, as well as target criteria for local content objectives. • Initial investment was significant. The total investment in the REIPPP’s 3,725 MW of RE will approach $15 billion. The local capital market has The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 167 responded positively to this opportunity. Commercial banks have been willing to finance construction and some are on-selling debt to insurance. But local banks will be stretched to further fund the REIPPP’s programs (given compet- ing demands in other infrastructure sectors in South Africa); other sources of funding (such as pension funds) will need to be mobilized. • Equity returns in round one were close to 17 percent, in real terms (in local ­currency), that was envisaged in determining the original REFIT tariffs. Equity returns dipped slightly in round two for wind, and probably more substantially for PV. Dollar returns in the range of 12−13 percent have been reported. • Project bidders are required to incorporate a tax of 1 percent of project revenues that will go into a government RE Fund to fund subsequent procurement programs. General Lessons But in hindsight, some areas could have been better designed and managed: • The size and readiness of the local RE market was initially overestimated, resulting in less capacity being bid than was made available. There was thus limited competition in round one, and bid prices were close to the price cap. The single price offer (rather than a dynamic reverse auction—as employed, for example, in Brazil) also restricted competition. • The size and complexity of the REIPPP meant that available legal and finan- cial advisory services were stretched to the limit. Some firms were permit- ted to offer advisory services to both government and private ­ bidders and funders, provided they created adequate “Chinese walls” within their firms. Some bidders complained that legal and finance firms were offering a “one size fits all” service that was not always appropriate for specific projects. • The above two points suggest that it may have been more prudent to start smaller, and then gradually ramp up the program, with larger blocks of capac- ity being offered in subsequent rounds. • All of the successful bidders in round one have reached financial closure and have commenced construction. It remains to be seen what proportion of preferred bidders in round two will achieve financial closure. The aim of the ­ REIPPP is lower prices, but projects must still be bankable. A successful bid- ding process should have a low attrition rate of preferred bidders. Bid prices need to be realistic. • Specifications on what constitutes local content could be improved, including more focus on those parts of the value chain that maximize local employment. • A balance needs to be struck between the promotion of economic develop- ment and prices. Already the economic development thresholds and target The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 168 Case Study: South Africa criteria are more stringent than in any other domain (and, indeed, more stringent than previous PPPs in South Africa). The South African RE market is ­ small by international standards, and investment in local manufacturing capa- bility is not necessarily competitive. International benchmarks indicate that South African RE prices are high. • In some areas, there is inadequate transmission grid capacity and otherwise viable and attractive projects have to compete for access. There have also been complaints about the lack of responsiveness of Eskom transmission planners. Integration of planning, procurement, and contracting functions in an inde- pendent transmission, system, and market operator will make it easier to resolve these constraints. • The transaction costs for the REIPPP were high for both the government and bidders (certainly higher than a REFIT program). The government has had to rely on external transaction advisors. But there is the potential to transfer these skills and experience in future procurement rounds and to ­ arket build capacity in the proposed independent transmission, system, and m operator. • The levelized energy costs that were calculated for the initial REFIT tariffs served as the departure for the REIPPP program. It should be noted that some other countries such as Tanzania have used avoided costs as their starting point. • In October 2012 the minister of energy announced that an additional 3,200 MW of renewable power projects would be bid out, with a target of COD between 2017 and 2020. South Africa’s power market continues to be shaped by centrally managed power-planning and procurement processes. But there are growing political and stakeholder concerns around rising electricity prices. Demand growth is also lower than predicted. The sustainability of the REIPPP program is dependent on volumes and predictable procurement processes. But its sustainability will depend also on the rate at which RE prices fall and ­ compete with alternatives. Note 1. An exchange rate of R8 per $1 has been used throughout this chapter. Bibliography Integrated Resource Plan (IRP). 2010. “Integrated Resource Plan for Electricity 2010– 2030.” http://www.energy.gov.za/IRP/irp%20files/IRP2010_2030_Final_Report​ _20110325.pdf. Maurer, Luiz, and Luiz Barroso. 2011. Electricity Auctions: An Overview of Efficient Practices. World Bank Study, Washington, DC. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: South Africa 169 NERSA (National Energy Regulator of South Africa). 2009. Renewable Energy Feed-in Tariff. Phase 2, NERSA Consultation Paper. http://www.nersa.org.za/Admin/Document​ /Editor/file/Electricity/REFIT%20Phase%20II%20150709.pdf. ———. 2011. Review of Renewable Energy Feed-In Tariffs. NERSA Consultation Paper. http://www.nersa.org.za/Admin/Document/Editor/file/Electricity/Consultation​ /­Documents/Review%20of%20Renewable%20Energy%20Feed-In%20Tariffs%20 Consultation%20Paper.pdf. South Africa Department of Energy. 2002. White Paper on Renewable Energy. http://www​ .energy.gov.za/files/policies/whitepaper_renewables_2003.pdf. ———. 2011. “South African Perspective on New and Renewable Energy.” Presented by Ms. Nelisiwe Magubane, Director General. http://www.energy.gov.za/files/media​ /­presentations/2011/20111206_DOE_MagubaneNewAndRenewableEnergy.pdf. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 7 Case Study: Tanzania Sector Background Tanzania’s energy supply is dominated by biomass-based fuels, providing an ­ estimated 90 percent of the national primary energy requirements. These fuels are mainly used in the informal sector, for household and small industrial use, drawing on the nation’s 35.5 million hectares (ha) of forests and home gardens to provide a cheap and accessible source of energy for the population of 45 ­million people. The main commercial forms of energy used are petroleum, gas, and electricity. Electricity is mainly produced from hydropower plants, but in recent years, sub- stantial thermal power generation, too, has been required to meet the growing demand and to improve supply security. Petroleum and hydropower account for about 8 percent and 1 percent of the primary energy supply, respectively. Another 1 percent of the primary energy requirement is met with coal, solar, and wind power. The estimated primary energy consumption was 22 million tons of oil equivalent (TOE) in 2003, amounting to a per capita consumption level of about 0.7 TOE. The main indigenous sources of energy are (a) biomass and agricultural waste, (b) hydropower, (c) natural gas, (d) coal, and (e) other forms of renewable energy (RE) such as wind and solar power. These sources are available in abundance, but so far, large-scale developments have been only in the hydropower and natural gas subsectors. While a large hydroelectric potential remains to be developed, exploration for natural gas and petroleum is ongoing. Coal use for electricity generation is limited to a small power plant, while there is no significant use of other renewables for electricity generation in a commercial scale. The national energy policy (Ministry of Energy and Minerals 2003) identifies the following challenges: (a) increased demand for electricity supply and distribu- tion (with demand tripling over a period of 20 years) will accelerate the need for investment in all elements of the electricity industry including private sector participation; (b) development of the petroleum sector to sustain gas production and increase gas and oil exploration will save foreign currency spent on the import of petroleum products; (c) improved regional and international interconnection will The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   171   172 Case Study: Tanzania support growth and improve reliability of electricity supply for mutual benefit; (d) greater rural electrification will make electricity available for economic activi- ties in rural areas, townships, and commercial centers and balance socioeconomic growth; and (e) the reaching of more rural households, providing energy supply to replace kerosene used for lighting and improve the efficiency of wood-fuel use will better the household environment and reverse deforestation. With regard to the supply of electricity, the relevant policy statements declare that (a) competition, as a principle to attain efficiency, will be applied to the electricity market, (b) generation of electric power will be fully open to both private and public investors, (c) there will be open access to the grid, (d) regional cooperation and integration will be given priority to ensure reliability and to exploit low-cost power, (e) priority shall be given to developing domestic power generation capacity, (f) strategic partnerships with technically suitable and finan- cially strong partners will help develop a competitive market in generation and distribution, (g) Tanzania will conduct research and participate in international research on commercially viable large-scale technologies for renewable sources of electricity generation, (h) support will be given to ownership contracts to ensure competition and a high level of investment, and (i) a new governance system shall be established, differentiating the roles of policy making, regulatory functions, and operational functions. Tanzania is continuing with the exploration of liquid petroleum but has not had any positive results so far. Therefore, all liquid petroleum products are imported. The total annual demand for petroleum products exceeded 1.5 million tons/year, and cost over $300 million in 2005. Petroleum products are used in transport (45 percent), manufacturing (25 percent), agriculture (10 percent), households (10 percent), and commerce (5 percent). Petroleum fuels are used for power generation as well. Since 2003 a 100 megawatt (MW) fuel-oil-burning power plant has been in operation, and was used heavily to address the power crisis of 2006. A few small diesel-burning power plants are used in the main grid as well as the mini-grids. The petroleum supply industry is fully liberalized and several players are in the market. The proven natural gas reserves are located offshore near the Songo Songo island in the Indian Ocean. The important gas discoveries have been in Songo Songo (30 billion cubic meters, bcm) and Mnazi Bay (15 bcm). Discovered reserves are limited and used for electricity generation, industrial applications, and petrochemical industries. A gas-fired power plant (Songas) has been in operation for several years, and presently has a total capacity of 190 MW. To address the ongoing generating system crises that started in 2006, gas-fired gen- eration was increased in the system on a short-term basis. More gas-fired generat- ing plants are also under construction. As of 2013 gas-fired power plants generating a total of 244 MW are in operation. A pipeline has already been built to deliver gas to Dar es Salaam for use in power generation. Coal reserves are found in Mchuchuma, in southwestern Tanzania near the northern end of Lake Nyasa. Some studies indicate that the Mchuchuma coal deposits can provide fuel for 400 MW generation capacity for up to 35−40 years. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 173 A small coal mine at Songwe-Kiwira started production in 1988. A small coal- fired power plant with an effective capacity of 1.5 MW is in operation at Kiwira and sells electricity to the Tanzania Electric Supply Company (TANESCO). Coal reserves in Tanzania are being used for industrial applications, but this major resource is yet to be exploited at its full potential. A small amount of electricity is being produced but there are plans to build larger power plants to use coal. Hydropower is the main form of RE used in Tanzania for the supply of com- mercial energy, that is, electricity. The country presently has an installed capacity of 561 MW of hydropower across six power plants. A few off-grid small hydro- electric power plants are in operation. Tanzania’s total technical hydroelectric energy potential is reported to be in excess of 4,700 MW of installed capacity or about 3,200 MW of firm capacity. Of this potential installed capacity, only about 12 percent has actually been developed. The economic potential—when the costs of developing hydroelectric capacity are compared with gas and coal-fired thermal power generation—is yet to be established (table 7.1). Research and measurements are being conducted on wind energy potential in various parts of the country. Tanzania has large reserves of indigenous energy resources, including enough natural gas, coal, and hydroelectric potential to meet the demand of the power sector for many years. There is also an undetermined potential of geothermal energy. Tanzania is likely to heavily use natural gas for electricity production in the foreseeable future, owing to the limited access to other sources of energy for electricity generation. Other uses of natural gas (such as for petrochemical indus- tries) would also emerge, establishing competing uses of gas. Liquid petroleum products might need to be imported in the foreseeable future, or until the ongo- ing exploration yields any positive results. The prospects of using coal are promis- ing, in the face of increasing prices of petroleum products and competing uses for natural gas, particularly in a scenario where Tanzanian coal is most suitable for use within Tanzania owing to the relatively lower heat content when com- pared with internationally traded coal. The potential for further developments in Table 7.1 Hydro Candidates Average energy, Firm energy, Plant/site Installation, MW GWh GWh River Kakono 53 404 335 Kagera Upper Kihansi with addition at Lower Kihansi 120 69 99 Rufiji Mpanga 144 955 646 Rufiji Masigira 118 664 492 Ruhuhu Ruhudji 358 1,928 1,333 Ruhudji Rumakali 222 1,475 908 Rumakali Rusumo 62 (21 Tanzania) 463 425 Kagera Songwe 340 1,669 1,045 Songwe Steiglers Gorge to Phase 3 1,200 5,259 3,227 Rufiji Source: Based on Power System Masterplan Study, TANESCO, 2008. Note: GWh = gigawatt-hour; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 174 Case Study: Tanzania hydropower, in both large and small projects, is high. Other renewable sources for commercial energy supply (particularly wind energy and biomass) also remain high, and more research and feasibility studies are required to establish their potential and economic justification. Sector Institutions The electricity supply industry in Tanzania is structured as follows: (a) TANESCO—a vertically integrated utility owned by the government—is in the business of generation, transmission, distribution, and supply of electricity through the main national grid, operating at 220 kilovolt (kV), 132 kV, 66 kV, 33 kV, 11kV, and 400 V; (b) there are 12 mini-grids, owned and operated by TANESCO, served with diesel power plants; and (c) there are distribution companies in Zanzibar, Resolute (mining area), and Kahama (mining area), which purchase from TANESCO and distribute in their own areas. TANESCO owns and operates the national grid. Generation into the grid is predominantly hydroelectric, but recurrent droughts in the past eight years have caused the thermal generation share to be significant. Demand Forecasts The forecast growth in demand for electricity is significant. Table 7.2 provides the forecast used in the most recent master plan study. The total national installed capacity on the grid is 1,438 MW (January 2013). Efforts are being made to increase power generation from local resources (namely natural gas, coal) and RE sources (namely geothermal, solar, wind, and biomass). With about 900,000 TANESCO customers, electricity is available to an estimated 18 percent of the population. A target of reaching an access level of 30 percent by year 2015 has been stated by TANESCO. Renewable Energy Development Prior to 2008 there are no reported incentives for generation of electricity using RE. It has been reported that some prospective investors in RE-based power plants have been negotiating with TANESCO for several years, on tariffs and power purchase agreements (PPAs). Table 7.2 National Demand Forecast 2006 2006 Increase Growth Growth Actual Unconstrained (%) 2016 (%) 2031 (%) National sales (GWh) 2,784 3,400 22 8,600 9.7 23,100 6.8 National losses (GWh) 806 1,100 36 2,100 6.7 4,000 4.4 National generation (GWh) 3,590 4,600 28 10,700 9.0 27,100 6.4 National sum of peak demands (MW) 800 1,700 7.8 4,800 7.2 Source: Power System Masterplan Study, TANESCO 2008. Note: GWh = gigawatt-hour; MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 175 In 2008 Tanzania formalized the incentives for RE-based electricity genera- tion by announcing (a) a procedure for the development of RE for electricity generation, (b) a standardized power purchase agreement (SPPA) and a stan- dardized tariff (a feed-in tariff, FIT), and (c) initiating a series of incentives for development of off- and on-grid RE power plants, and associated mini-grids. Of these, establishment of (a) and (b) was supported by the World Bank through technical support, with the expectation that the small power producer (SPP) program would operate independently, with the FIT being the key instrument. Incentives for project development, such as matching grants for feasibility studies and matching grants for project implementation, continue to be supported by the World Bank’s project and are financed by a number of agencies. The Small Power Producer Program SPPs eligible to sign a standardized power purchase agreement (PPA) are defined as those whose: (a) primary source of energy is either an RE source or waste heat, (b) net export is less than or equal to 10 MW, and (c) agreements and FITs are standardized and nonnegotiable. SPPs are accepted both for main grids and mini- grids (either existing or new). But the FIT is only applicable to TANESCO-owned mini-grids. SPPs on the main grid are nondispatchable, must run for the SPP, and are a must-take for TANESCO. This means that for the main grid, TANESCO cannot refuse to p ­ urchase power at any time (irrespective of the generation eco- nomics at this time), except in a case where TANESCO is constrained from pur- chasing power (such as when the transmission line to the SPP is interrupted). Three SPPs were the first additions under the SPPA introduced in 2008. Three power plants that existed at the time of the introduction of the SPP program have since been grid connected under the SPPA: TPC (9 MW), operating on bagasse (waste sugarcane); TANWAT (1.5 MW), operating on wood waste from the leather tanning industry; and the Mwenga (4 MW) hydropower plant. Many new, small power plants are being developed by the private sector. Renewable Energy Targets Tanzania has not published a quantity target for RE-based electricity, as have most of the other case study countries. As such, Tanzania is free to allow the most economically optimal (mix of) sources of RE to be developed for electricity production. Design of Incentive Schemes FITs for SPPs serving the main grid and mini-grids are based on two formulae that use the calculated avoided cost of the respective grids. The SPP tariffs for the main grid were first calculated for 2007 and 2008. From 2009 onwards, ­ tariffs were calculated for both the main grid and existing isolated mini-​ grids. Tariffs are revised every year, based on an Energy and Water Utilities Regulatory Authority (EWURA)–approved methodology, which considers a The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 176 Case Study: Tanzania number of parameters including the projected long-run marginal costs of the TANESCO grid and forecast costs of thermal power generation (both from TANESCO and from non-TANESCO sources in the subsequent year for main-grid-connected SPPs, and the cost of a generic diesel-fired off-grid gen- erating facility for SPPs connected to an isolated mini-grid). Grid-Connected Tariffs The system for setting SPP FITs followed by EWURA is based on avoided costs to TANESCO, and the tariffs are technology neutral. This means that (a) TANESCO pays SPPs a price that reflects what it costs to produce or pro- cure electricity from other sources, and (b) there is no special preference or price incentive given to any specific technology. If the resource is renewable or from cogeneration and the SPP intends to export the electricity produced from 10 MW or less of export capacity, then electricity generated from that power plant qualifies to pay the FIT announced for each year. With varying conditions of hydropower available in TANESCO’s main power plants, oil and gas prices forecast the hydro-thermal mix in the ensuing year. The SPP FITs announced by EWURA for the main grid, based on estimates of TANESCO’s long- and short-run marginal costs in Tanzania, are listed in table 7.3. Mini-Grid Tariffs The system for setting up SPP FITs for supply to mini-grids is similar: (c) the average of the avoided costs to the mini-grid is calculated by the average cost of generation from a diesel power plant and the avoided cost of the main grid, and (b) the tariffs are technology neutral. This means that (a) TANESCO pays SPPs a price that reflects what it costs to produce or procure electricity from typical diesel power plants serving the mini-grids, with the objective that it will someday be connected to the main grid;1 and (b) there is no special preference or price incentive given to any specific technology. If the resource is renewable or if it is Table 7.3 SPP Tariffs for the Main Grid, 2008–12 Year Season Price offered 2008 2009 2011 2012 Dry season T Sh/kWh 120.5 115.33 145.36 183.05 In equivalent cents/kWha 9.36 9.66 10.74 12.06 Wet season T Sh/kWh 90.4 86.5 109.02 137.29 In equivalent cents/kWha 7.02 7.25 8.06 9.05 Weighted average T Sh/kWh 100.43 96.11 121.13 152.54 In equivalent cents/kWha 7.80 8.05 8.95 10.05 1287.50 1193.55 1352.92 1,517.65 Source: For 2008, Ministry of Energy and Minerals (MEM) reports; from 2009 onwards, EWURA (2010, 2011a, 2011b, 2012a, 2012b). In year 2011 no FIT was announced, and the 2009 FIT is presumed to be operational in 2010 as well. Note: The dry season is August to November, the wet season is from January to July, and December. kWh = kilowatt-hour; SPP = small power producer. a. Exchange rate used in the tariff calculations of each respective year. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 177 Table 7.4 Mini-Grid Feed-In Tariffs, 2009–12 2009 2011 2012 Min-grid FIT 334.8 380.22 480.50 T Sh/$ 1,193.55 1,352.92 1,517.65 Cents/kWh 28.05 28.10 31.66 Source: EWURA (2010, 2011a, 2011b, 2012a, 2012b). Note: No FIT was announced in 2010, and the 2009 FIT is presumed to be operational in 2010 as well. FIT = feed-in tariff; kWh = kilowatt-hour. from cogeneration and the SPP intends to export the electricity produced from 10 MW or less of export capacity, then electricity generated from that power plant qualifies to pay the FIT announced for each year for mini-grids. Unlike for the main-grid FIT, the tariff is not seasonal. Mini-grid FITs announced in the recent past are summarized in table 7.4. The rationale for offering a high FIT for TANESCO mini-grids is that they are served by diesel generators that cost around 40 cents/kilowatt-hour (kWh) to produce, and that any SPP producing below this production cost would off- set or completely eliminate expensive diesel generation. Mini-grids are gradu- ally absorbed into the national grid, and when that happens, the mini-grid SPP automatically converts into a main grid FIT. There would be a significant reduc- tion in the FIT but possible improved dispatch of the SPPs’ output, which now would not be limited by the customer load profile in the mini-grid. In an SPP operating in a mini-grid, TANESCO can purchase only what it can dispatch to customers. Transparency The SPP process in Tanzania was established in year 2007 through a series of stakeholder consultations, involving policy makers (at the ministry level), the utility (TANESCO), prospective investors, prospective lenders, the regulatory authority (EWURA), and academics. The guidelines for project development, optional methods to calculate FITs, and the conditions of the SPPA were widely discussed (at not less than five workshops held over 2007−08) before EWURA first announced a public consultation on the first FIT proposed for the year 2008. The publication of EWURA consisted of (and continues to include in subsequent revisions) (a) the standardized tariff methodology document including data sources and (b) detailed tariff calculations including actual data used from various sources. Separate publications address grid-connected and mini-grid ­ SPPs. After a comments period of three weeks, the EWURA board makes a determination and issues the tariff order. This procedure for review and opening for public comments and subsequent decisions was followed in the years 2008, 2009, 2011, and 2012. But there was no tariff announced for 2010. Furthermore, at the time of writing (May 2013), the FIT for 2013 had not been announced by EWURA. Therefore, while the incentive scheme is transparent (by way of a standardized procedure and tariffs), there are delays in announcing the FIT each year. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 178 Case Study: Tanzania Other documents associated with the SPP process and useful for developers have been published by EWURA, too, such as (a) guidelines for developers, (b) guidelines for grid interconnection (in three parts), and (c) the SPPA. These are available on the EWURA Web site as approved/recommended documents. Accordingly, information about the incentives provided through the SPP proce- dure is easily available to any developer or stakeholder. The information appears to be reasonably up to date. But there are limitations to the information available, especially regarding the actual procedure to be followed in project development. A procedural document has been developed, but is not available in the public domain as yet. As such, a developer does not have easy access to project development guidelines. Additionally, the procedure to secure a letter of intent from TANESCO is some- what long, although documented. Implementation Issues Tanzania’s government decided, at an early stage of policy formulation, that the SPP program would establish a tariff on the basis of avoided costs to the grid and to mini-grids. Therefore, in principle, there is no additional financial burden to the offtaker (TANESCO) owing to purchases from SPPs. But several constraints were observed that may have caused apparent losses to TANESCO and prospec- tive investors. The two key issues raised are as follows: • Of the first three power plants to be grid connected, two had been in negotia- tions with TANESCO for several years to sell their surplus to the grid at prices in the range of 5 cents/kWh, whereas the SPP process and the FIT com- menced at 7.8 cents/kWh in 2008 and is currently at 10.05 cents/kWh. Thus, TANESCO views the first three SPPs as reaping undue profits, as electricity would otherwise have been purchased at lower prices and benefits passed on to customers (or used to cushion the losses of TANESCO). • Developers of greenfield (that is, new) hydroelectric power plants see the FIT as inadequate to meet their cash-flow requirements and achieve a reasonable return on equity (ROE), while prospective developers of power plants who use other RE sources (such as wind and biomass) see no prospects at all of developing power plants at the FIT offered. The apparent additional purchase price (the FIT, which in turn is higher than a possible negotiated price) has since been effectively passed on to customers or added to TANESCO’s losses.2 But TANESCO has had to sign up for emergency power plants operating on diesel to boost the supply to the grid, which was suf- fering from lower hydropower inputs from TANESCO’s own power plants as a result of recurring droughts in the period 2008–12. Thus, one may argue that the inputs from the three initial SPPs would have otherwise been produced using diesel at costs much higher than the FIT, although the SPPs are viewed to be reaping windfall profits (see figure 7.1). But the numbers of such preexisting The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 179 Figure 7.1 Announced FIT and Surplus Power from Existing Power Plants, 2007–12 12 10 Tariff to grid (equivalent USc/kWh) 8 6 Perceived negotiated 4 price range for surplus power from existing power plants 2 0 2006 2007 2008 2009 2010 2011 2012 2013 Year Main grid FIT based on avoided costs Source: EWURA 2010, 2011a, 2011b, 2012a, 2012b. Note: FIT = feed-in tariff; kWh = kilowatt-hour; USc = U.S. cents. power plants using RE, and qualifying for the SPP program, are not many. Thus, this issue will be resolved, accepted a necessary concession and the unavoidable reality of standardization. On the positive side, these three projects—all of which involved lengthy negotiations—were quickly grid connected and now deliver power. Meanwhile, TANESCO has delayed payment to these SPPs, an issue that will be dealt with later. Developer Cash Flows The issue of lower returns on equity and negative cash flows in the initial year (previously highlighted) remains a barrier to the development of mini-hydro SPPs, the only type of SPP that is possibly viable at the range of the avoided-cost-based FIT (7.8−10.1 cents/kWh) announced in recent years. Figure 7.2 illustrates the typical situation of such an SPP hydropower developer. Given that the lending rates in Tanzania are about 16 percent (16.3 percent calculated in this example, on the basis of Bank of Tanzania assessments), an internal rate of return (IRR) of 7 percent pretax is not adequate if a mini-hydro SPP costs $2,000 per kilowatt (kW) to build, in spite of the fact that this example assumes a good site with a capacity factor of 50 percent. In the initial years, the cash flow is negative for the example considered, an inherent problem with any SPP program incentives based on avoided costs (see figure 7.3). It is assumed that the guaranteed price for all future purchases will The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 180 Case Study: Tanzania Figure 7.2 A Hypothetical Case, Illustrating Negative Returns in the Initial Years Investment = 2000 US$/kW, Debt:equity = 60:40, 300 Interest rate = 16.3%, Capacity factor = 50% Cashflow requirement (TZS/kWh) 250 200 Feed-in tariff 2012 150 Result: Pre-tax IRR to equity = 9% 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Year of operation Interest Loan principle Other expenses O&M Note: This figure does not represent any particular small power producer, but illustrates a typical situation. IRR = internal rate of return; kW = kilowatt; O&M = operation and maintenance. Figure 7.3 Cash Flow Profiles for a Mini-Hydro SPP 800 Minihydro investment = 2000 US$/kW, Capacity factor = 50% FIT of 2012, Pe-tax IRR = 9% 600 Gross return to equity (TZS million) 400 FIT of 2011, Pe-tax IRR = 3% 200 0 –200 –400 0 2 4 6 8 10 12 14 16 Year of operation Note: SPP = small power producer; FIT = feed-in tariff; IRR = internal rate of return; kW = kilowatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 181 Figure 7.4 Cash Flow Profiles for a Mini-Hydro SPP with a 5 Percent Growth in FIT 1,600 Minihydro investment = 2000 US$/kW, Capacity factor = 50% 1,400 Gross return to equity (TZS million) 1,200 1,000 FIT of 2012, 800 Escalates at 5%, Pe-tax IRR = 16% 600 400 200 0 –200 0 2 4 6 8 10 12 14 16 Year of operation Note: SPP = small power producer; FIT = feed-in tariff; IRR = internal rate of return; kW = kilowatt. be the floor price, which in the Tanzania SPP program is the FIT of the first year. But as seen in the five-year period since 2008, the FIT has increased at a com- pound average growth rate of 10 percent per year. Consideration of a moderate 5 percent increase in dollar terms yields a significantly improved ROE of percent, which is not adequate (as seen in figure 7.4). Therefore, mini-hydro 14 ­ SPPs that cost significantly lower than $2,000/kW (say $1,500/kW) and with a capacity factor of 50 percent, are more likely to be viable under the FIT regime presently operational in Tanzania. Additionally, the perception of SPP developers that the FIT is low (7.8−10.1 cents/kWh) compared with the costs of incremental generation from short-term emergency power contracts (about 30 cents/kWh) is also a negative factor. While it is granted that SPP developers under the 15-year SPPA enjoy a purchase guarantee on a nondispatchable basis over a relatively long period, and that short-term emergency generation has to be operational on demand, this has not been considered an adequate explanation for the large gap between the FIT and the price of emergency purchases. A possible means of addressing the lower or negative cash flow to equity in the initial years is to offer a tiered tariff, by which the first six to eight years of operation are awarded a higher FIT (possibly a certain percentage higher than the calculated announced avoided costs), in return for a FIT that is lower than the avoided costs in the second half of the SPPA. In Tanzania such a policy and The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 182 Case Study: Tanzania Figure 7.5 Optional Approaches to Improve the Cash Flow of Mini-Hydro SPPs 300 Cashflow requirement (TZS/kWh) Approach 1 250 Incentive 200 Approach 2 Incentive Avoided cost-based feed-in tariff 150 100 Recovery 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Year of operation Interest Loan principle Other expenses O&M Note: SPP = small power producer; FIT = feed-in tariff; kWh = kilowatt-hour; O&M = operation and maintenance. options have been discussed but not yet implemented. Figure 7.5 shows two possible approaches by which over the period of perceived debt repayment by the SPP (a) an incentive is paid that is recovered in the later years or (b) a non- recoverable incentive is paid. A common misconception is that a site has only one capacity (MW) rating, one capacity factor, and nothing else. This position is incorrect. If a site is found to be financially unviable at the FIT offered, a cautious developer with compe- tent professional advice would examine whether it can develop the same site at a lower installed capacity (hence lower investment), which will yield a higher capacity factor (hence an improved utilization of the asset) using lower-cost equipment (hence lower investment). Developers and other analysts often speak of a site as an x MW site. Once optimized to the offered tariff, the site may be rated lower or higher than x. After all the options to optimize the site have been investigated and found to be uneconomical, then it should not be developed. There will be other sites proposed by other developers that can meet their own profitability criteria, and such sites should be developed first. Sites that can- not be made viable at the offered FIT, in principle, should wait until avoided costs (and hence the FIT) increase further (for example, if fossil-fuel prices increase in the future). Until then, developers have to wait. This is a common situation in many countries with an SPP or similar program. In the same manner, if the FIT is inadequate to ensure the commercial viability of other technologies, such as wind and solar, these, too, should be postponed until avoided costs (and hence the FIT) increase to higher levels. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 183 Tariff Requirement for Project Viability A more reasonable assessment would be to examine the required FIT to make a generic SPP project viable so it achieves an equity IRR of (say) 22 percent—the minimum IRR expected by a Tanzanian investor. When the increase in dollar terms is assumed to remain at 5 percent per year, table 7.5 shows the break-even points at which the project becomes viable to the investor, when the year 1 FIT is changed. Results indicate that small hydro SPPs would be viable at the 2012 FIT, provided they (a) can be built at a cost not exceeding $1,600/kW of capac- ity, (b) the site has a good flow uniformity so that a capacity factor of 50 percent is optimal, and (c) the FIT is increasing by 5 percent per year in real terms. But in Tanzania, developers claim that owing to long transport distances for project equipment and construction material, and difficult access to sites from main roads, the project capital costs are not in the range of $1,500/kW, but higher. The FIT offered is adequate for the project to be profitable if developers respond to the incentive by initially developing the potentially better sites (those with a lower specific investment and a higher capacity factor) or sites for which the parameters are such that an optimized design to achieve an equity IRR of 22 percent would yield a viable project. Conclusions Since its first introduction in 2008, three existing power plants quickly went through the SPP process and connected as main-grid SPPs: TPC (an existing thermal 9 MW power plant, which previously served only the sugar company’s mini-grid), TANWAT (which previously served the tanning company’s mini-grid and was closed down for about one year because the main grid had reached the factory, but then was restarted to feed 1.5 MW to the grid), and Mwenga (an existing small hydropower plant that is grid connected and serving 4 MW to the grid). Actual energy delivered to the grid has not been published as yet; any sig- nificant inputs are likely to be from year 2012 onwards. The payment record of TANESCO and its adherence to contractual condi- tions have not been encouraging. There are reports that payments to the three operational SPPs do not come on time, or even several months later. A dispute Table 7.5 Required Additional Incentives to FIT for Project Viability Investment ($ per kW) 1,500 1,600 1,700 1,800 1,900 2,000 Capacity factor (%) 45 45 45 45 45 45 Base year FIT (T Sh/kWh) 153 153 153 153 153 153 Escalation of FIT in real terms (%) 5 5 5 5 5 5 Incentive above FIT (%) 0 0 0 0 0 0 Equity IRR (%) 24 22 20 19 17 16 Incentive above FIT (%) 0 0 7 12 19 24 Equity IRR (%) 24 22 22 22 22 22 Note: FIT = feed-in tariff; IRR = internal rate of return; kW = kilowatt; kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 184 Case Study: Tanzania with a thermal independent power producer (IPP), ongoing for several years, has not improved investor confidence in the SPP process or the SPPA. While there are reasons for TANESCO to be short of cash, a constrained cash flow has caused a loss of confidence among prospective SPP developers, in addition to the anxiety caused by lower FITs. A significant development was the biomass power plant on Mafia island (off- shore), where a TANESCO mini-grid is operational. The power plant project was developed and made operational. Being on a mini-grid displacing diesel, the project is offered the higher mini-grid tariff, with no risk to the investor of the mini-grid being absorbed into the main grid, which may cause tariffs to drop. As there is no centralized clearinghouse for applications for SPPs, it is difficult to know the number of projects being developed by investors, and their viability (or its lack) under the FIT. Indicative estimates can be obtained on the basis of the letter of intent issued by TANESCO and the environmental licensing authority. By providing the concession of offering an SPPA and a FIT, Tanzania is fulfill- ing the country’s policy objective of encouraging the private sector to invest and operate power plants to contribute to grids and mini-grids. By limiting such concessions to SPPs (RE), the desire to develop RE is also being fulfilled. The SPP program had been running for five years (2008–12) at the time of this writing. Initial delays in making the SPP process and the SPPA acceptable to the govern- ment, TANESCO, prospective lenders, and the developer community probably constrained the number of power plants built and made operational in five years. Notes 1. TANESCO plans to gradually connect its mini-grids on the mainland of Tanzania to the main grid by 2017. 2. TANESCO has been reporting losses for several years, owing to an increase in produc- tion costs and an inability to raise tariffs to reflect such costs. Bibliography EWURA (Energy and Water Utilities Regulatory Authority). 2010. The Electricity Act (CAP 131): The Electricity (Development of Small Power Projects) Rules. Dar es Salaam, Tanzania. http://www.ewura.com/pdf/SPPT/PROPOSED%20RULES/The%20 Electricity%20(Development%20of%20Small%20Power%20Project)%20Rules​ -2010.pdf. ———. 2011a. Detailed Tariff Calculations for Year 2011 for the Sale of Electricity to the Main Grid in Tanzania under Standardized Small Power Purchase Agreements in Tanzania. Dar es Salaam, Tanzania. http://www.ewura.go.tz/pdf/SPPT​ /PROPOSED​ %20GUIDELINES/PROCESS%20GUIDELINES/2011%20SPPT%20Calculation​ %20for%20Main%20Grid.pdf. ———. 2011b. Detailed Tariff Calculations for Year 2011 for the Sale of Electricity to the Mini-Grids in Tanzania under Standardized Small Power Purchase Agreements in The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Tanzania 185 Tanzania. Dar es Salaam, Tanzania. http://www.ewura.go.tz/pdf/SPPT/2011%20 SPPT%20Calculation%20for%20Mini-Grid.pdf. ———. 2012a. Detailed Tariff Calculations for Year 2012 for the Sale of Electricity to the Main Grid in Tanzania under Standardized Small Power Purchase Agreements in Tanzania. Dar es Salaam, Tanzania. http://www.ewura.go.tz/pdf/SPPT/2012/2012​ %20SPPT%20Calculation%20for%20Main%20Grid.pdf. ———. 2012b. Detailed Tariff Calculations for Year 2012 for the Sale of Electricity to the Mini-Grids in Tanzania under Standardized Small Power Purchase Agreements in Tanzania. Dar es Salaam, Tanzania. http://www.ewura.go.tz/pdf/SPPT/2012/2012%20 SPPT%20Calculation%20for%20Mini-Grid.pdf. Ministry of Energy and Minerals. 2003. National Energy Policy. http://www.mem.go.tz​ /wp-content/uploads/2014/02/0001_17022013_National_Energy_Policy_2003.pdf. TANESCO. 2008. Power System Masterplan Study. Dar es Salaam, Tanzania. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 8 Case Study: The Arab Republic of Egypt Sector Background Egypt’s power sector reforms have involved both the unbundling and rebun- dling of its state-owned power entities, combined with the shuffling of respon- sibilities for policy and regulatory oversight and corporate governance. A major institutional reform was undertaken in 2000 through Law 164, which formed the Egyptian joint stock (holding) company under the name Egyptian Electricity Holding Company (EEHC). In July 2001 more restructuring took place through the unbundling of generation, transmission, and distribution activities into 13 companies (5 generation, 1 transmission, and 7 distribution). An internal wholesale power pool was also created in 2002 under Law 164 to replace the previous dispatch processes. Under the pool provisions, the generators provide bids for dispatch, and their generating units are scheduled for dispatch on the basis of these bids. The bids are based on costs, however, and so the pool has never operated as a genuine market-clearing exchange. Moreover, the EEHC has retroactively adjusted the cost-based prices in the pool to maintain substantial cross-subsidies among the pool members, which has further blunted any com- petitive pressure to improve efficiency. Further unbundling took place in 2002 with the division of one of the distri- bution companies into two companies, and again in 2004 with the division of another of the distribution companies into two companies. Currently, as illus- trated in figure 8.1, the Egyptian electricity market is composed of government- owned utilities (6 generation, 1 transmission, and 9 distribution) under the direct management of the EEHC; three independent power producer (IPP)–owned projects; one wind-generating company, New and Renewable Energy Authority (NREA) within the Ministry of Electricity and Energy (MOEE); and about 12 small isolated and/or semiconnected independent service providers in either generation or distribution. All of these reforms took place without an independent regulator. The regula- tory agency was established in May 2001 and started operations in early 2002. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   187   188 Case Study: The Arab Republic of Egypt Figure 8.1 Structure of the Egyptian Power Sector Ministry of Electricity and Energy EEHC Six generation companies NREA EETC: The transmission company and the single Three IPPs buyer entity New IPPs Nine distribution companies Consumers Source: Vagliasindi and Besant-Jones 2013. Note: EEHC = Egyptian Electricity Holding Company; EETC = Egyptian Electric Transmission Company; IPPs = independent power producers; NREA = New and Renewable Energy Authority. In addition to ensuring an adequate supply of electricity to meet demand at equitable prices, the NERA’s mandate covers for competition in the power mar- ket. The NERA’s powers fall short of a truly independent regulatory agency, however, because it does not have tariff-setting power. Moreover, its rulings are under government influence: its board is chaired by the minister of electricity and energy. Between 1992 and 2004 there were no changes to the tariffs in nomi­ nal terms, even though a substantial decline in real terms resulted. Then in October 2004 the cabinet of ministers approved nominal tariff increases of approximately 5 percent per year for the next five years, with the aim of covering costs by 2009. In August 2007 the government announced a three-year plan to remove subsidies from natural gas and electricity tariffs for energy-intensive industries. In June 2008 the tariff increases under this plan were accelerated and implemented immediately. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 189 Egypt has embarked on a program of targeted energy pricing and fuel subsidy reforms to achieve cost-recovery and to replace untargeted subsidies with tar- geted social safety net programs, which were off limits in the past. This is a sig- nificant first step toward achieving fiscal sustainability in the sector. It should be noted that this program was launched during the postrevolutionary period, which was marked by greater political and economic instability and when reforms were typically difficult to implement. An overall 15 percent electricity tariff increase for households and commercial consumers was implemented in two steps in November 2012 and January 2013, together with a tariff increase of up to 50 percent for energy-intensive users implemented in January and July 2012. These reforms of tariffs (mainly those for commercial and energy-intensive users) are expected to increase economic efficiency for the targeted sectors, improve the financial sustainability of the EEHC, and send a price signal to save energy. The EEHC organized the first tender for private power generation in 1996 and awarded a contract under a power purchase agreement (PPA) in 1998. The PPA provided for power to be supplied from a gas-fired steam generator of 682.5 ­ megawatt (MW) capacity for a period of 20 years under build, own, oper- ate, transfer (BOOT) arrangements with project financing. Subsequently, the EEHC quickly concluded two more BOOT projects for generating plants under similar contract terms and the same set of conditions extended by the govern- ment. Between 1996 and 2003 the private sector added 2 gigawatts (GW) in new power capacity in the form of three gas-fired, steam-generating plants of equal rated capacity of 682.5 MW, accounting at that time for about 10 percent of the country’s installed capacity. Debt financing was provided by local and foreign banks as well as by institutional investors and multilateral agencies. The EEHC has continued to work on five-year development plans—­ particularly for generating capacity. The EEHC concluded its first fast-track power generation program for adding 4,500 MW of gas-fired, combined-cycle generating capacity during its fifth five-year plan for 2002–07. The EEHC then implemented a second fast-track power generation program during its sixth five-year plan for 2007–12, which consisted of 7,240 MW of new generating ­ capacity (6,500 MW gas-fired plant and 600 MW of wind-power capacity). The EEHC is planning to add about 15,000 MW of new capacity during its seventh five-year plan for 2012–17. Nevertheless, the EEHC’s generation reserve margin is expected to remain tight for some time because of the expected growth in power demand. In conclusion, the government’s strategy for meeting Egypt’s demand for electricity has managed to expand the power supply impressively, helped by ­ the discovery of large reserves of natural gas that provide low-cost power ­ generation while expanding access to electricity to nearly the whole population. The government has kept electricity prices down to help low-income house- holds afford it, and energy-intensive industries remain competitive. These achievements have, however, come at a cost. The subsidies imposed substantial burdens on the government’s fiscal resources and weakened the financial The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 190 Case Study: The Arab Republic of Egypt structure of the state-owned enterprises involved in supplying electricity and energy and in financing the energy sector. This outcome has impeded the govern- ment’s intentions to privatize power supply entities and attract private invest- ment to the sector. Egypt faces challenges in terms of security of supply. As figure 8.2 illustrates, since 2009 Egypt has become a net oil importer, after reaching a peak in oil pro- duction in 1993. In terms of oil reserves, figure 8.2 shows a stable reserve of around 4 billion barrels for the past three decades, indicating that no more large oilfields are expected to be discovered in the future. Alongside a growing popula- tion, oil consumption is expected to increase, and with decreasing oil production and flat oil reserves, Egypt will become more dependent on oil imports and more vulnerable to fluctuations in international oil prices. This in turn means more burdens on the already cash-strapped national budget. The situation is less dramatic in the case of natural gas, where production is still greater than consumption, as shown in figure 8.3. Whereas consumption is Figure 8.2 Evolution of Oil Production and Consumption and Reserves a. Evolution of oil reserves 5.0 4.5 4.0 3.5 Billion barrels 3.0 2.5 2.0 1.5 1.0 0.5 0 1980 1985 1990 1995 2000 2005 2010 2015 b. Evolution of oil production and consumption 400 350 300 Million barrels 250 200 150 100 50 0 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Oil consumption Oil production Source: NREA. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 191 Figure 8.3 Evolution of Natural Gas Production and Consumption and Reserves a. Evolution of natural gas production and consumption 80 60 Billion m3 40 20 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 b. Evolution of natural gas reserve 2.5 2.0 1.5 Trillion m3 1.0 0.5 0 1980 1985 1990 1995 2000 2005 2010 2015 Natural gas consumption Natural gas production Source: NREA. Note: m3 = cubic meters. monotonically increasing, production came to a halt in 2009. If both production and consumption keep up their 2012 rate of growth, consumption will surpass production in just four years. Natural gas reserves’ rate of increase has been decreasing since the beginning of the millennium, and the reserve value almost peaked in 2010 (as shown in figure 8.3). The Egyptian government has opened areas for international exploring companies in the deep water of the Mediterranean Sea, where there are great hopes of finding large natural gas reserves, as has been done for neighboring countries. It is worth mentioning (regardless of the fact that production is greater than consumption) that about 85 percent of fossil fuel supplied to power plants is from natural gas, and that since 2010 power plants have experienced production The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 192 Case Study: The Arab Republic of Egypt disruption, especially during summer months, because of frequent fluctuation in the pressure of natural gas pipes. The frequent power outages of the past three years are a new experience for Egyptians, and the Ministry of Energy (MoE) has been mitigating the effect by rotating blackouts. The installed capacity of thermal power plants—which rely primarily on natu- ral gas, not heavy oil—constitutes the major part of the total installed capacity in Egypt, followed by hydropower and a very minor contribution from wind power and one solar thermal power plant. In terms of percentages, figure 8.4 shows how this installed capacity is distributed among the different primary energy resources. This figure also illustrates that dependence on thermal power plants has grown over the past decade. Similarly, as shown in figure 8.4, the total electrical energy production has increased during the past decade by 76 percent—from around 89,000 gigawatt- hours (GWh) to 157,000 GWh, most of it obviously coming from the thermal power plant. The consumption evolution of different primary energy types used for electricity production (in percentages) involves an increasing share of thermal power at the expense of a decreasing share of hydropower (figure 8.4). Taking two snapshots at 2002−03 and 2011−12 shows the percentage of electricity production by different types of primary energy. Regardless of the increasing wind and solar projects developed in the past decade, Egypt is becoming more dependent on fossil fuel now than before. Rapidly growing demand is a key feature of the power sector in Egypt. This demand is driven by population growth, development of energy-intensive indus- tries, and increasing use of electrical appliances, especially air-conditioners in residential sectors. The residential and industrial sectors are by far the largest ­ consumers, and together account for 70−75 percent of total electricity consump- tion. Peak electricity demand increased from 15,678 MW in 2005 to 21,330 MW in 2009, and to 24,400 MW in 2011—a 14 percent increase in just two years. The growth in demand has outstripped growth in the supply capacity, leading to some disconnections during the peak summer seasons in recent years and raised public concerns about energy security. Although the annual demand growth slowed to approximately 5 percent during the political crisis, the EEHC forecasts demand growth to rebound to previous levels (6.4 percent) in the foreseeable future. Renewable Energy Development Among the six known renewable energy (RE) resources, Egypt enjoys hydro- power through the Nile River, wind energy in some specific locations where it is economically feasible, solar energy almost all over the country, and a very minor amount of geothermal energy in the Sinai Peninsula. Although currently there is a national program to exploit biomass resources in the Ministry of Environment, it is not targeting electricity generation but rather biogas and natural fertilizer production for rural development. In this section, we will explore the efforts made in the hydro, wind, and solar energy sectors, as they are the major contribu- tors to the national RE targets and plans. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 193 Figure 8.4 Evolution of Fuel Mix Production and Installed Capacity, FY2002/03 to FY2011/12 a. Production 160,000 140,000 120,000 100,000 GWh 80,000 60,000 40,000 20,000 0 3 2 4 5 6 7 8 9 0 1 /0 /0 /0 /0 /0 /0 /1 /1 /1 /0 11 03 04 05 06 07 08 09 10 02 20 20 20 20 20 20 20 20 20 20 b. Installed capacity 30,000 25,000 20,000 MW 15,000 10,000 5,000 0 03 04 05 06 07 08 09 10 11 12 02/ 03/ 04/ 0 5/ 06/ 07/ 0 8/ 0 9/ 10/ 11/ 20 20 20 20 20 20 20 20 20 20 Thermal Hydro Wind Solar thermal Source: NREA. Note: GWh = gigawatt-hour; MW = megawatt. Hydro Hydropower was, historically, the first RE resource to be exploited in Egypt. Its generation started with the building of the Aswan Dam 1 (322 MW) in 1960, followed by the High Dam (2,100 MW) in 1967. Table 8.1 and figure 8.5 illus- trate the historical evolution of the installed capacity of hydropower plants The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 194 Case Study: The Arab Republic of Egypt Table 8.1 Major Dams on the Nile River Name Installed capacity (MW) Year Aswan 1 Dam 322 1960 High Dam 2,100 1967 Aswan 2 Dam 270 1985 Esna Dam 87 1993 Naga Hamady Dam 64 2008 New Assuit barrage 32 2017 Source: NREA. Note: MW = megawatt. Figure 8.5 Evolution of Hydropower in the Arab Republic of Egypt, 1955–2025 Naga 3,000 Hamady Aswan2 2,500 New Esna Assuit 2,000 MW 1,500 High Dam 1,000 500 Aswan 0 70 20 60 80 90 00 10 55 65 75 85 95 05 15 25 19 20 19 19 19 20 20 19 19 19 19 20 20 20 19 Source: NREA. Note: MW = megawatt. in Egypt, while figure 8.6 shows the amount of generated electricity from each dam in the past five years. Egypt’s hydro resources are almost entirely developed, and not available for further development. Finally, in the hydro sector there are two micro dams (0.68 MW and a 0.8 MW), located in the Fayoum governorate, that were commissioned in 1991 and 2003, respectively. The latest reports of the Ministry of Electricity do not show further plans for new micro dams in Egypt. Wind In 2005, and in a joint venture between the Danish RISO laboratories and the Egyptian Meteorological Authority, a national wind atlas was issued showing that Egypt enjoys some excellent wind regimes in both onshore and offshore regions. Onshore, the wind speed in the Suez Gulf region reaches 10.5 meters per second (m/s), making it one of the best places in Egypt, followed by large regions to the east and west of the Nile River, where the speed ranges from 7 m/s to 8 m/s. The first wind farm (5 MW) erected in Egypt was in Hurghada in 1993. This was followed by a series of wind farm projects in Zafarana (northern of Hurghada) that extended from 2002 till 2010. Table 8.2 shows the eight wind The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 195 Figure 8.6 Evolution of Hydropower in the Arab Republic of Egypt, FY2007/08 to FY2010/11 16,000 11,000 GWh 6,000 1,000 –4,000 2007/08 2008/09 2009/10 2010/11 Naga Hamady Esna Aswan Dam2 Aswan Dam1 High Dam Source: NREA. Note: GWh = gigawatt-hour. Table 8.2  Wind Projects Electric capacity Number of Turbine power Provider company Project name (MW) turbines (MW) (type) Financing country Zafarana-1 30 50 0.6 Nordex (N43) Netherlands Zafarana-2 33 55 0.6 Nordex (N43) Germany Zafarana-3 30 46 0.66 Vestas (V47) Netherlands Zafarana-4 47 71 0.66 Vestas (V47) Germany Zafarana-5 85 100 0.85 Gamesa (G52) Spain Zafarana-6 80 94 0.85 Gamesa (G52) Germany Zafarana-7 120 142 0.85 Gamesa (G52) Japan Zafarana-8 120 142 0.85 Gamesa (G52) Netherlands Total 545 700 Source: NREA. Note: MW = megawatt. farm projects that were implemented in Zafarana along with the installed capac- ity of each project, the number of turbines, and the power of each turbine. Collectively, the total installed capacity in Zafarana is 545 MW with a total of 700 turbines installed, making it the largest wind farm in the Middle East and North Africa (MENA) region and on the African continent. Figures 8.7 and 8.8 illustrate the historical development from 2002 till 2012 of the wind-installed capacity, electricity generated, fuel saving, and emissions reduction during the construction of the different projects at Zafarana. Concentrated Solar Power Egypt is one of the sunniest countries in the world, with a large potential of solar energy. Egypt issued its solar atlas in 1991 indicating that the average direct nor- mal solar radiation ranges between 2,000 (north) and 3,200 (south) k ­ ilowatt-hour 2 per square meters per year (kWh/m /yr), with very few cloudy days and an aver- age sunshine duration of between 9 (winter) and 11 (summer) hours/day. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 196 Case Study: The Arab Republic of Egypt Figure 8.7  Wind-Installed Capacity and Production, FY2002/03 to FY2011/12 a. Wind-installed capacity b. Electricity generated from wind energy 600 1,600 500 1,400 1,200 400 1,000 GWh MW 300 800 200 600 400 100 200 0 0 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 2 20 /03 20 /04 20 /05 20 /06 20 /07 20 /08 20 /09 20 /10 20 /11 2 /1 /1 / / / / / / / / / 02 03 04 05 06 07 08 09 10 11 02 03 04 05 06 07 08 09 10 11 20 20 Source: NREA. Note: GWh = gigawatt-hour; MW = megawatt. Figure 8.8  Fuel Savings Due to the Implementation of Wind Energy Projects, FY2002/03 to FY2011/12 350 300 250 Thousand TOE 200 150 100 50 0 3 9 0 1 2 4 5 6 7 8 /0 /0 /1 /1 /1 /0 /0 /0 /0 /0 02 08 09 10 11 03 04 05 06 07 20 20 20 20 20 20 20 20 20 20 Source: NREA. Note: TOE = tons of oil equivalent. Egypt started to exploit its vast solar energy resource with the Integrated Solar Combined Cycle (ISCC) power plant in Kuraymat, which is one of three similar projects in the world (the other two are in Morocco and Algeria). The plant started its operation in July 2011 and its total installed capacity is 140 MW—­ 20 MW is from the concentrated solar power (CSP) solar field and the remaining 120 MW from a combined cycle gas and steam turbine. Table 8.3 lists technical information about the project, while figure 8.9 shows a schematic diagram of the ISCC power plant. Renewable Energy Targets As the cost of electricity production from wind energy is the closest to conven- tional resources among the different RE technologies, wind energy has the highest priority right now in Egypt’s national plans and targets (figure 8.10). ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 197 Table 8.3 Technical Specifications of the Kuraymat Solar Field Site Longitude 31.25° E Latitude 29.27° N Altitude 60 m Annual direct normal solar radiation 2,441 kWh/m2 Field dimensions 678 m (N–S) 921 m (E–W) Performance Annual insolation on solar field 317,975 MWh Optical efficiency 79.74% Thermal efficiency 65.6% (peak) 64.4% (annual) Source: NREA. Note: E–W = east to west; kWh/m2 = kilowatt-hours per square meter; m = meter; MWh = megawatt-hour; N–S = north to south. Figure 8.9 Integrated Gas-Steam-Solar Combined Cycle Stack exhaust 100˚C 393˚C Steam 540˚C. 100 bar HRSG Air and Steam turbine vapor Solar HX No Sun: 36 MW G~ Cooling 50 MJ/s Sun: 59 MW tower Air Air 293˚C Condenser Parabolic Exhaust trough field 600˚C G~ Electricity to the grid Gas turbine 74 MW No Sun: 104 MW Sun: 126 MW Solar island Combined cycle island Source: NREA. Note: HRSG = heat recovery steam generator; HX = heat exchanger; MJ = megajoules; MW = megawatt. A national strategy for renewable energy that was approved in February 2008 (and that aims to cover 20 percent of the electricity generated by renewable energy in the year 2020) sets a specific target of 12 percent of the electricity generated to come from wind energy in the target year. This target is translated to a total installed capacity of 7,200 MW of grid-connected wind energy in 2020. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 198 Case Study: The Arab Republic of Egypt Figure 8.10 CO2 Emissions Reduction Due to the Implementation of Wind Energy Projects, FY2002/03 to FY2011/12 1,000 Thousand ton CO2 800 600 400 200 0 2 3 8 9 0 1 5 6 4 7 /1 /0 /0 /0 /1 /1 /0 /0 /0 /0 11 02 07 08 09 10 04 05 03 06 20 20 20 20 20 20 20 20 20 Source: NREA. 20 Note: CO2 = carbon dioxide. Figure 8.11 The Egyptian Solar Plan Approved by the Cabinet in July 2012 3,500 3,000 2,500 2,000 MW 1,500 1,000 500 0 6 7 8 9 0 1 2 3 4 5 6 7 /1 /1 /1 /1 /2 /2 /2 /2 /2 /2 /2 /2 15 16 17 18 19 20 21 22 23 24 25 26 20 20 20 20 20 20 20 20 20 20 20 20 Source: NREA. Note: MW = megawatt. In a move to put solar energy on the national plan of the energy mix, the min- isterial cabinet approved the Egyptian Solar Plan in July 2012. The plan, which starts in 2015, targets installing 3,500 MW of solar energy by 2027 (­figure 8.11). The target amount is divided into 2,800 MW of CSP and 700 MW of photovol- taic (PV). The plan also addresses the enhancement of the relevant local industries that can feed into the targeted technologies. It is worth mentioning that the plan relies on a 67 percent share of private investment to implement the required solar projects, revealing a large opportunity for national and international investors to play an essential role in the future of Egypt’s solar projects. One of the first projects under the Solar Plan is a 100 MW CSP power plant, with four hours’ storage, proposed for Kom Ombo in Upper Egypt. The pro- posed financing arrangements for this project, and the extent to which the high The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 199 Table 8.4  Future PV Projects Installed capacity Expected year Location MW Funding agencies of operation Status Hurghada 20 Japan International 2016 A feasibility study was completed by the end of 2012. Cooperation Agency Kom Ombo 20 French 2017 A grant was signed with the French Development Development Agency in May 2012; €800,000 will fund a feasibility Agency study of the project. A land plot of 15 square kilometers (km2) was designated for the project in November 2012. A consultant contract was expected to be signed in 2014. Source: NREA. Note: MW = megawatt; PV = photovoltaic. incremental costs can be bought down by concessional financing, is discussed below in some detail. Finally, the NREA is planning the first two large PV projects in Hurghada and Kom Ombo (of 20 MW each), which are expected to start operations in 2016 and 2017, respectively (table 8.4 provides more detail on each project). Production Costs Table 8.5 shows the levelized economic production costs of the generation alter- natives in Egypt (as considered in the EEHC master plan). The costs for CSP are taken from the recent feasibility study (FS) for the proposed Kom Ombo CSP project (Fraunhofer and Lahmeyer International 2012). The resulting levelized costs of around 22 cents/kWh imply an incremen- tal cost of 18−19 cents/kWh, when measured against the alternative of natural gas at $3 per million British thermal units (mmBTU).1 The incremental cost of wind power is much lower, at around 5.6 cents/kWh. The economic value of gas is certainly quite low, but was derived in a detailed 2007 study of Egypt’s gas resources, and is based on the long-run marginal costs of production, plus a depletion premium (Economic Consulting Associates 2007), which is consistent with the valuation of gas in two recent gas projects financed by the World Bank (Helwan and Giza North). But more recently Egypt has been able to export natural gas to Jordan at a price of $6/mmBTU. At this higher price of gas, the levelized economic production cost increases from 3.8 cents/kWh to 5.7 cents/kWh, with a corresponding decrease in the incre- mental costs of renewable energy. Defining a plausible thermal generation counterfactual is not straightforward. Although the indicated thermal alternative is a gas combined-cycle gas turbine (CCGT), Egypt suffers from a periodic natural gas shortage associated with sup- ply infrastructure and transportation bottlenecks. CCGTs are therefore designed to run with diesel oil as a supplementary fuel, while steam cycle gas projects The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 200 Case Study: The Arab Republic of Egypt Table 8.5 Production Costs: Generation Alternatives Steam Steam CCGT OCCT plant plant Wind CSP CSP Fuel Natural gas Natural gas Natural gas HFO Wet Dry Installed capacity MW 750 250 650 650 80 100 100 Operating life Years 25 20 30 30 25 25 25 Overnight construction cost $/kW 800 500 1,076 1,076 2,000 7,233 7,355 Construction period Years 3 2 4 4 2 2 2 Construction period adjustment factor $/kW 1.202 1.144 1.308 1.308 1.144 1.144 1.144 SCF ratio 0.946 0.946 Local portion percentage 19.2 19.1 $/kW 1,588.8 1,607.0 SCF adjustment $/kW 85.8 86.8 Capacity credit ratio 0.6 Capacity cost $/kW −244.8 Economic cost $/kW 961.6 572 1,407 1,407 2,533 8,189 8,327 Capital recovery factor ratio 0.110 0.117 0.106 0.106 0.110 0.110 0.110 Annualized capital cost $/kW/yr 105.9 67.2 149.3 149.3 279.0 902.2 917.4 Fixed O&M $/kW/yr 16.0 9.0 3.0 4.0 76.0 30.5 29.2 Total fixed cost $/kW/yr 121.9 76.2 152.3 153.3 355.0 932.7 946.6 Variable cost Efficiency percentage 54.0 34.0 40.3 39.8 Heat rate BTU/kWh 6,319 10,035 8,475 8,575 Fuel cost $/mmBTU 3 3 3 14.5 $/kWh 0.019 0.030 0.025 0.124 Nonfuel variable O&M $/kWh 0.0002 0.003 0.0004 0.0004 0.007 0.007 Total variable cost $/kWh 0.019 0.033 0.026 0.125 0.000 0.007 0.007 Total cost Capacity factor ratio 0.75 0.20 0.85 0.85 0.43 0.51 0.50 Annual generation kWh 6,570 1,752 7,446 7,446 3,758 4,504 4,369 Total cost/kWh $/kWh 0.038 0.077 0.046 0.145 0.094 0.214 0.223 Incremental cost over CCGT $/kWh 0.04 0.01 0.11 0.06 0.18 0.19 Source: World Bank 2013. Note: CCGT = combined-cycle gas turbine; CSP = concentrated solar power; kW = kilowatt; kWh = kilowatt-hour; mmBTU = million British thermal units; O&M = operation and maintenance; OCCT = open-cycle combustion turbine; SCF = statement of cash flow; HFO = heavy fuel oil. (like Helwan) are designed to use heavy fuel oil (HFO) as the supplementary fuel (Mazout). But if, in fact, gas shortages were to occur, then the gas that is available would be used at the most efficient projects (at CCGTs) and curtailed at gas-steam plants. Therefore the counterfactual is a combination of 20 percent HFO (at steam cycle projects) plus 80 percent natural gas (at CCGTs). There are great hopes that the high capital costs of CSP can be significantly reduced over present levels. But as shown in table 8.6, even if costs were just half of what they are today, and taking into account the avoided local environmental health damages of gas (and HFO) generation, there remains an economic incre- mental cost of $127 million. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 201 Table 8.6 Impact of Capital Cost Reductions Capital cost CSP capital ERR + local Incremental cost ERR + local + GHG reduction cost ERR damage cost (economic) (1) damage cost ($30/ton) [%] $/kW [%] [%] $ million [%] 0 7,233 −3.3 −1.4 −412 0.1 10 6,510 −2.6 −0.7 −355 0.9 20 5,787 −1.8 0.2 −298 1.9 30 5,063 −0.8 1.2 −241 3.0 40 4,340 0.3 2.4 −184 4.3 50 3,617 1.8 3.9 −127 6.1 Source: World Bank 2013. Note: (1) Incremental cost to the Arab Republic of Egypt includes avoided local environmental damages. CSP = concentrated solar power; ERR = economic rate of return; GHG = greenhouse gas; kW = kilowatt. Table 8.7 Incentive Mechanisms Item NREA Competitive bidding Feed-in tariff Program size 2,200 MW 2,500 MW 2,500 MW Single wind farm size Large (100–400 MW) Large ten modules (each of Medium and Small below 250 MW) 50 MW Developer NREA Private (most probably Private (focus on local) international) Finances Governmental and soft Commercial finance Commercial finance financing from international development agencies Tariff setting Proposed by EgyptERA and According to the bid Proposed by EgyptERA and approved by the Cabinet of outcome approved by the Cabinet Ministers of Ministers Contracting period 20 years Long-term PPA mostly for 20/15 (under study) years 20 years Off taker Grid Grid Grid or distribution system O/M NREA Developer Developer Construction responsibility NREA through EPC Developer Developer Source: NREA. Note: EPC = engineering, procurement, and construction; EgyptERA = Egyptian Electric Utility and Consumer Protection Regulatory Agency; IPPs = independent power producers; NREA = New and Renewable Energy Authority; MW = megawatt; PPA = power purchase agreement. Design of Incentive Schemes The governmental wind projects are developed, owned, and operated by the NREA. These projects are financed by multilateral and bilateral financing agencies as well as national government concessional financing and grants, and are open to public bidding. The commercial wind program consists of two ­ components: a competitively bid large-scale IPP commercial wind program and a commercial wind program for small-scale IPPs benefiting from a feed-in tariff (FIT). The key difference between the three schemes is described in table 8.7. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 202 Case Study: The Arab Republic of Egypt The competitively bid commercial wind program for large-scale IPPs (which is currently approved and in the planning phase) plans to select experienced IPPs through competitive bidding to build, own, and operate (BOO) wind power plants for a term of 20−25 years on predetermined sites (on the shores of the Gulf of Suez and the east and west of the Nile River). The Egyptian Electricity Transmission Company (EETC) will purchase the energy generated from the wind power plant throughout the duration of the agreement, according to the terms and conditions of the PPA. These particular IPP projects benefit from newly approved government incentives. The commercial wind program for small-scale IPPs (benefiting from a FIT) is currently planned but not yet in effect, pending the passing of legislation. It will be applied to wind farms of up to 50 MW to be executed either on predeter- mined sites allocated by the Egyptian government or on private sites owned by the developers. The EETC/distribution companies are obliged to purchase all the generated energy from the RE power plant through a declared tariff, which allows the investor to achieve a predefined return on equity. This tariff is divided into blocks: the first is constant for all projects under the FIT, and the second depends on the sites’ capacity factors to achieve the predefined return on equity. A third-party scheme is also included. It is similar to the self-supply approach that served as a catalyst for wind financing and uptake in Mexico. The scheme includes a bilateral agreement between the IPP wind power proj- ect and its direct customers, while the EETC provides third-party access to transfer power from the power plant to its customers. Additionally, the EETC will purchase any excess wind power and provide supplemental energy to cus- tomers during low wind production time (NREA 2010). The first of these projects is to be undertaken by Italgen, the energy generation arm of Italian cement giant Italcementi. Italgen plans to invest €140 million for a 120 MW facility to be constructed along the shores of the Red Sea in the Gulf El Zeit area and supply energy to the group’s Suez Cement Plant. The success of future self-supply in Egypt will depend upon pending legislation, as well as the ability to access government-controlled land where high wind speeds make wind power development feasible. Wind capacity installed to date has been provided by NREA-led government projects. With the first government phase of wind development under way, Egypt is now focusing on its first phase of commercial IPP business models as it continues to build RE capacity (tables 8.8–8.10). These power projects benefit from the following government incentives approved by the Supreme Council of Energy: • All permits for land allocation already obtained by the NREA. • Land-use agreements signed with the investor against payment equivalent to 2 percent of the annual energy generated from the project. • Environmental impact assessments (including bird migration studies) pre- pared by the NREA in cooperation with international consultants and financed by the German Development Bank (KfW). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Table 8.8  Wind Projects Currently under Development by the Egyptian Government Installed Expected year Location capacity (MW) Funding agencies of operation Status Gabal El Zayt 200 KfW, EIB, European Commission April 2014 • All construction contracts were signed and the project is under construction. Gabal El Zayt 220 Japan n.a. • A soft loan agreement was signed with Japan in March 2010. • An Environmental Impact Assessment study has been finalized. • Bidders submitted their proposals on January 28, 2013. 120 Spain End of 2016 • A €120 million loan was signed in February 2008 with the Spanish government. • The project is exclusive to the Spanish market. • The feasibility and environmental impact assessment study has been finalized. • In November 2012 the Spanish government appointed a consultant to assist in preparing tender documents. • The tender will be issued in the first half of 2013. Gabal El Zayt 200 KfW, EIB, French Development End 2015 • A grant of €10 million from the Neighborhood Investment Facility is secured. Agency, European Union • The German government has agreed to provide €140 million. • The feasibility study is expected to be finalized by September 2013. Gabal El Zayt 200 Masder • The project will be financed by Masder and the NREA, $220 million each. • A fund of $1 million from the CTF was signed in February 2012 to finance a feasibility study. • The feasibility study is expected to be finalized by September 2015. Gabal El Zayt 200 French Development Agency, KfW 2016 • n.a. West of Nile 200 Japan 2017 • A 4,242 m2 lot west of the Nile River is designated for the project. • In August 2010 a Japanese consultant was appointed to carry out a feasibility and environmental impact assessment. • Ten measuring stations at 80 meters height were installed in the area for wind speed measurement and the study will be finalized in July 2013. Source: NREA. Note: CTF = Clean Technology Fund; EIB = European Investment Bank; KfW = German Development Bank; m2 = square meters; MW = megawatt; NREA = New and Renewable Energy Authority. n.a. = not applicable. 203 204 Case Study: The Arab Republic of Egypt Table 8.9  Wind Projects That Will Be Built and Operated by the Private Sector on a BOO Basis to Supply the National Electricity Network Installed capacity Expected year of Location (MW) operation Status Gulf of Suez 250 Mid-2015 • Ten developers were shortlisted in December 2009. • Measurement studies will be completed by mid-2013. Gulf of Suez 500 n.a. • Prequalification document for the second competitive bidding were to be announced in the second half of 2013. Source: NREA. Note: BOO = build, own, operate; MW = megawatt. n.a. = not applicable. Table 8.10  Wind Projects to Be Built and Operated by the Private Sector for Self-Consumption or to Directly Sell to Consumers Installed Expected year Location capacity (MW) Owner of operation Status Gulf of Suez 120 Italgen 2014 • Agreement is signed to build a wind farm to feed the Suez Cement Company. • Environmental study was finalized in April 2010. • In June 2012 the land usufruct agreement was signed. Gulf of Suez 600 Not determined • The NREA announced the availability of 6 × 15 km2 yet pieces of land to establish 100 MW wind farm projects in each on an auction basis. Source: NREA. Note: km2 = square kilometers; MW = megawatt; NREA = New and Renewable Energy Authority. • All RE equipment and spare parts exempted from customs duties and sales taxes. • Long-term PPAs of 20−25 years, signed. • The Central Bank of Egypt to guarantee all financial obligations of the EETC under the PPA. • The project to benefit from carbon credits. • The project company to receive licenses for power generation from the Egyptian Electric Utility and Consumer Protection Regulatory Agency (EgyptERA). Despite the social and political revolution of early 2011 and a lack of final legislation, Egypt moved forward in launching its first 250 MW BOO IPP project and part of the first tranche of a 2,500 MW procurement competitive bidding scheme. This is the first private sector power producer venture in renewable energy in Egypt, and the first in which project developers benefit from ministry- approved government incentives. A unit to be established within the EETC will be responsible for the sale of the certified emission reduction (CER) credits of the IPP projects. Given that the environmental attributes of the IPP projects remain the property of the Government of Egypt, the proceeds of the CER credits sale remain within the government treasury and do not contribute to the overall IPP financing package. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 205 As part of the RE strategy legislation, and to encourage investors to establish RE power plants, the fund might cover: • Full or partial deficit between the RE cost and market prices. • Exchange rate risk, in case the cost is transferred—whether fully or partially— to consumers. • Guarantee of the transmission company payments. • Financial support to pilot projects. • Research and development for renewable energy. The main sources of the fund include: • Subsidies currently given to fossil fuels used in power generation. • The state budget. • Donations. • Investment of the fund money. The current selling price of the electricity produced from wind energy is 17.6 piaster/kWh (including 2 piaster/kWh from fuel saving), while the average production cost is 38 piaster/kWh—hence, the need to fund the substantial incremental cost. Carbon Accounting As noted above, the avoided costs of carbon are critically dependent on the ther- mal alternative under consideration. Table 8.11 shows the avoided cost of carbon for the Kom Ombo CSP, assessed against a thermal alternative consisting of a mix of gas and HFO at varying gas prices. At $6/mmBTU (that is, the current export price of Egyptian gas to Jordan) at a 20 percent HFO share, the avoided cost of carbon is $267/ton. This may be contrasted to the avoided cost of wind in Egypt, shown in table 8.12. For the same thermal alternative, wind is a win-win! There is a net economic benefit to wind against a set of thermal generation assumptions (indi- cated by negative [shaded] values in the table). Table 8.11 Avoided Cost of Carbon: Concentrated Solar Power Gas price, $/mmBTU HFO share (%) 3 4 5 6 7 8 9 10 0 448 429 411 393 374 356 338 319 10 363 349 335 321 307 293 279 264 20 300 289 278 267 256 245 235 224 30 251 243 234 226 218 209 201 192 40 213 206 200 193 187 180 174 168 50 181 177 172 167 162 157 152 147 Source: World Bank 2013. Note: HFO = heavy fuel oil; mmBTU = million British thermal units. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 206 Case Study: The Arab Republic of Egypt Table 8.12 Avoided Cost of Carbon: Wind Gas Price, $/mmBTU HFO share (%) 3 4 5 6 7 8 9 10 0 76 57 39 21 3 −16 −34 −52 10 46 32 18 4 −10 −24 −38 −52 20 24 14 3 −8 −19 −30 −41 −52 30 8 −1 −9 −18 −26 −34 −43 −51 40 6 −12 −19 −25 −32 −38 −45 −51 50 −17 −22 −26 −31 −36 −41 −46 −51 Source: World Bank 2013. Note: HFO = heavy fuel oil; mmBTU = million British thermal units. Figure 8.12 Impact of CDM on Incremental Costs of Wind and CSP 14 12 First-year tariff, USc/kWh (%) 10 8 6 gas:HFO:5.5USc/kWh 4 2 0 0 10 20 30 40 50 $/ton CO2 CSP Wind gas:HFO Source: World Bank 2013. Note: CDM = clean development mechanism; CO2 = carbon dioxide; CSP = concentrated solar power; HFO = heavy fuel oil; kWh = kilowatt-hour; USc = U.S. cents. Figure 8.12 shows the impact of potential clean development mechanism (CDM) revenues on the financial internal rate of return (FIRR) and incremental costs. In the case of CSP, under an optimistic case of an Emissions Reduction Purchase Agreement (ERPA) covering 70 percent of carbon dioxide (CO2) reductions for 14 years, and CER revenue at $20/ton CO2, the tariff decreases from 12.7 cents/kWh to 12.1 cents/kWh; but in the case of wind under the same assumptions, the tariff falls below that of the gas or HFO alternative. Without carbon revenue the wind tariff is slightly above that of the thermal alternative, so CDM additionality could likely be demonstrated. But for the more expensive CSP, CDM revenues (if they were available) would do little to buy down the incremental costs. While Egypt derives a financial The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 207 Table 8.13 Incremental Financial Cost to the Arab Republic of Egypt, $ million (as NPV) CER price, $/ton CO2 CSP, $ million Wind,$ million 0 −194 5.8 10 −185 14.8 20 −176 23.8 30 −167 32.8 40 −158 41.8 50 −149 50.8 Source: World Bank 2013. Note: CER = certified emission reduction; CO2 = carbon dioxide; CSP = concentrated solar power; NPV = net present value. surplus of $5.8 million for wind even without the CER revenues (under the same concessionary financing package as provided to the CSP), for CSP the financial balance is significantly negative. At $30/ton CER revenue, Egypt would still incur $167 million in incremental costs—to be carried either by the government or the consumers. The ability to buy down the incremental costs by concessionary financing is discussed in more detail below (table 8.13). Incremental Costs and Their Recovery Problems in Traditional Financial Analysis The conventional financial analysis encountered in project appraisals of RE proj- ects is generally unsatisfactory. Most often one finds a calculation of the project financial return more or less following the format of the economic analysis—­ adding back in taxes and duties, and then comparing this to some calculation of the weighted average cost of capital (WACC). If the resulting FIRR exceeds the WACC, the project is declared financially feasible. Only rarely does one find an assessment of the annual incremental financial cash flows that must be covered either by tariff increases or by governments through the state budget. The problem with a WACC calculation is that it looks only at interest rates and not the tenor of loans. But the tenor of loans is rarely coincident with the presumed financial life—and in the case of project financing the WACC tells us nothing about the impact of short loan tenors in typical commercial financings. Worse, the tenors of concessionary loans (and notably carbon finance) may be as much as 40 years, which significantly affects the actual cost of capital. A WACC calculation for a large utility (like Perusahaan Listrik Negara [Indonesian State Electric Utility Company] [PLN] in Indonesia), with a complex mix of debt financing and bond issuances, may be meaningful, but for a project-financed IPP (or subsidiary project company that is supposed to run along commercial lines) one needs to look at the actual proposed financial structure to make informed judgments about financial costs. Comparisons of average levelized costs can be similarly deceptive since they are very sensitive to the discount rate used. Thus to make a realistic assessment of the financial implications of an RE project, the best approach is to compare the actual financial flows of the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 208 Case Study: The Arab Republic of Egypt proposed RE project against the actual financial flows of the thermal alternative.2 So in the case of the Kom Ombo project, we compare the cash flows of the CSP under the proposed financing scheme with the cash flows of the thermal genera- tion alternative (in this case a gas combined-cycle combustion turbine [CCCT]) under a plausible financing scheme for that alternative (in this case financed by the International Bank for Reconstruction and Development [IBRD]—a plausi- ble counterfactual since the World Bank has indeed recently financed a gas CCGT in Egypt). Such a financial model calculates the tariff necessary to achieve a given equity return, and reveals a first-year tariff requirement of 4.12 cents/ kWh. The year-by-year revenue requirements for the fossil-fuel project then provide the yardstick against which the year-by-year tariff requirements for the RE project can be measured. Who Pays? Who pays for the incremental cost of renewable energy is conveniently displayed in table 8.14, in which the financial costs and benefits are reconciled among the stakeholders. The columns represent the various stakeholders and the rows, indi- vidual transactions—here under the assumption of domestic financing of the Kom Ombo CSP. The net impact on the stakeholders is listed in the bottom row of the table. In the case of the CSP company (whoever that may be), it is assumed that the financial surplus (return) is passed back to the government as dividends, so the net impact on the CSP is always shown as zero. All entries are in million dollars, expressed as lifetime present values at a 10 percent discount rate. In column [12] we also show the environmental benefits, namely the sum of the local avoided environmental damage costs and the avoided GHG emis- sions (valued at $30/ton CO2). The cost to the consumer (at the current assumed retail tariff of 3.5 cents/ kWh) is shown in row [2]—$95 million. The consumer, however, derives a ben- efit of $170 million (which is what he or she would pay for the equivalent amount of electricity in the absence of the subsidy): the difference of $75 million is the consumer surplus.3 This underscores the impossibility of Egypt investing in a CSP (or any other expensive RE project) without the assistance of the international community. There is a small gain to the domestic banks (because the assumed interest rate of 12 percent exceeds the discount rate). The government gains from income taxes; but even if there were an income tax exemption, there is no change in the net result for Egypt: the total net financial impact on Egypt is $464 million (figure 8.13). Of course, when the environmental benefits are added back in, the ­ net result is less negative (minus $373 million), but the inclusion of global social environmental costs is unlikely to impress ministries of finance. CDM revenue would be real cash, and is relevant to actual financial flows (but as we have seen in section “Carbon Accounting,” the prospects of significant CDM revenue in the next few years are poor and dwindling, and therefore not considered here). The required CSP tariff in the absence of international financial institution (IFI) and concessionary finance is 25.1 cent/kWh. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Table 8.14 Reconciliation of Economic and Financial Flows, $ Million, as NPV at 10 Percent Discount Rate (Domestic Finance Only) Consumers CSP Government Domestic banks EU CTF IBRD KfW EIB AfDB Total finance Env Adjusted total [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] 1. Benefits (cost of CCGT) 170 0 0 0 0 0 0 0 0 0 170 0 170 2. Consumer cost of −95 0 95 0 0 0 0 0 0 0 0 0 0 electricity 3. Tariff revenue, CSP 0 781 −781 0 0 0 0 0 0 0 0 0 0 4. Grants 0 0 0 0 0 0 0 0 0 0 0 0 0 5. Loan disbursements 0 456 0 −456 0 0 0 0 0 0 0 0 0 6. Principal repayments 0 −373 0 373 0 0 0 0 0 0 0 0 0 7. Interest repayments 0 −120 0 120 0 0 0 0 0 0 0 0 0 8. OPEX 0 −63 0 0 0 0 0 0 0 0 −63 0 −63 9. Equity 0 114 −114 0 0 0 0 0 0 0 0 0 0 10. Construction costs 0 −571 0 0 0 0 0 0 0 0 −571 0 −571 11. Income tax 0 −104 104 0 0 0 0 0 0 0 0 0 0 12. Dividends 0 −121 121 0 0 0 0 0 0 0 0 0 0 13. Local environmental benefits 0 0 0 0 0 0 0 0 0 0 0 40 40 14. GHG benefits 0 0 0 0 0 0 0 0 0 0 0 51 51 15. Total 75 0 −576 37 0 0 0 0 0 0 −464 91 −373 Source: World Bank 2013. Note: AfDB = African Development Bank; CCGT = combined-cycle gas turbine; CSP = concentrated solar power; CTF = Clean Technology Fund; EIB = European Investment Bank; EU = European Union; Env = environmental benefits (GHG+local); IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NPV = net present value; OPEX = operating expenses. 209 210 Case Study: The Arab Republic of Egypt Figure 8.13 Stakeholder Impacts, No Foreign Assistance 200 100 0 –100 US$ million –200 –300 –400 –500 –600 –700 s s t RD D v P B EU l F l W en er nk cia ta En CS EI CT Af Kf um IB to m ba an rn ed ns in tic ve st lf Co es Go ju ta m Ad To Do Source: World Bank 2013. Note: AfD = French Development Assistance; CCGT = combined-cycle gas turbine; CSP = concentrated solar power; CTF = Clean Technology Fund; EIB = European Investment Bank; EU = European Union; Env = environmental benefits (GHG + local); IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NPV = net present value; USc = U.S. cents. Salient indicators Indicator Value Tariff, USc/kWh 25.1 Actual avoided cost of carbon, $/ton 267 Financial package Private finance only Share of incremental costs Egypt, Arab Rep. (Govt + consumers + domestic banks) 100% ($464 million) International community 0% Buying Down the Incremental Costs with Concessionary Finance From this starting point we can now assess the degree to which the incremental costs can be bought down by the international community. Table 8.15, taken from the CSP feasibility study, shows the likely sources of IFI assistance for Kom Ombo. For example, if the entire debt ($579 million) were taken up by the IBRD, the CSP tariff (to achieve 5 percent FIRR) falls dramatically to 12.2 cents/kWh, and the impact on Egypt falls to $183 million (figure 8.14).4 Note that in this sce- nario, the CCGT is also assumed to be financed by the IBRD. Of course this leaves open the question of why the IBRD would make this magnitude of resources available for a GHG benefit worth $51 million (at $30/ton CO2). The IBRD can thus be said to potentially buy down the incremental financial cost by half! The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 211 Table 8.15  Financing Options Interest rate, % Grace, years Repayment period, years Currency Availability CTF 0.25 10 30 $ 100 IBRD 2.75 6 21.5 $ 170 AfDB 2.75 6 20 $ 170 KfW 3.00 4 15 Euro 174 EIB 3.15 3 20 Euro 100 AfD 3.70 4 20 Euro 50 NIF(1) grant Euro 25 Source: CSP-FS, tables 8.2 and 8.3. IBRD and CTF terms as in the World Bank Egypt Wind Development Project. Note: AfDB = African Development Bank; AfD = French Development Assistance; CTF = Clean Technology Fund; EIB = European Investment Bank; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NIF = EU Neighbourhood Investment Facility. Figure 8.14 Impact of IBRD Financing 200 100 0 US$ million –100 –200 –300 –400 –500 s s t RD D v ve SP B EU F l l W en er nk cia ta En EI CT Af Kf C m IB to m ba an u rn d ns in tic te lf Co es s Go ju ta m Ad To Do Source: World Bank 2013. Note: AfD = French Development Assistance; CCGT = combined-cycle gas turbine; CSP = concentrated solar power; CTF = Clean Technology Fund; EIB = European Investment Bank; EU = European Union; Env = environmental benefits (GHG + local); GHG = greenhouse gas; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NPV = net present value; USc = U.S. cents. Salient indicators Indicator Value Tariff, USc/kWh 12.2 Financial package Private finance only Share of incremental costs Egypt, Arab Rep (Government + consumers + domestic banks) 39% ($183 million) International community 61% ($281 million) The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 212 Case Study: The Arab Republic of Egypt Table 8.16 shows the application of funds proposed in the CSP-FS—which includes a €25 million grant from the European Union (EU) Neighbourhood Investment Facility (NIF). The table also shows the more recent proposal of the World Bank, which increases the carbon finance funding (Clean Technology Fund, CTF) by $43 million, to $123 million (with a corresponding decrease in the IBRD financing). Table 8.17 shows the impact of all the various financing options, assuming (for sake of comparison) that the entire debt (80 percent of the total investment cost) is assumed by each IFI. With financing by domestic banks, Egypt bears 100 ­percent of the incremental cost ($464 million, as net present value [NPV]). Carbon finance is the most effective in buying down the cost: if 100 percent of the debt were of the CTF, the CSP tariff would fall to 7.4 cents/kWh, and the Table 8.16 Proposed Application of Funds As per FS Revised CTF % $ million % $ million Equity 19.8 143.2 19.8 143.2 Domestic debt 0.0 0.0 0.0 0.0 CTF 10.6 76.7 17.0 123.0 IBRD/AfDB 28.2 204.0 21.8 157.7 KfW 20.2 146.1 20.2 146.1 EIB 10.6 76.7 10.6 76.7 AfD 6.2 44.8 6.2 44.8 Grants 4.4 31.8 4.4 31.8 Total 723 723.0 Source: CSP-FS (Fraunhofer and Lahmeyer International 2012), table 8.3. Note: AfDB = African Development Bank; AfD = French Development Assistance; CTF = Clean Technology Fund; EIB = European Investment Bank; FS = feasibility study; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank. Table 8.17 Comparison of Effectiveness in Buying Down the Incremental Costs CSP tariff Egypt, Arab Rep. Others Total Cents/kWh $ million % $ million $ million Domestic debt 25.1 464 100 0 464 CTF 7.4 60 13 404 464 IBRD/AfDB 12.2 183 39 281 464 KfW 17.1 291 63 172 464 EIB 16.5 277 60 186 464 AfD 16.8 286 62 178 464 Proposed 13.2 204 44 260 464 Revised 12.7 194 42 270 464 Source: World Bank 2013. Note: AfDB = African Development Bank; AfD = French Development Assistance; CTF = Clean Technology Fund; EIB = European Investment Bank; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 213 share of the incremental cost carried by Egypt (by some combination of con- sumers and ­ government) is just 13 percent, or $60 million. As noted, if the IBRD accounted for all the debt, it would be the second-most effective in bring- ing down the CSP tariff to 12.2 cents/kWh (as noted above). Impact of the Proposed Financing CSP Packages The financing package proposed in the FS brings is little different to 100 percent IBRD finance—the low cost of CTF is offset by higher finance costs from KfW, EIB and AfD. The required first-year CSP tariff is slightly higher at 12.7 cents/kWh. Table 8.18 and figure 8.15 show the resulting distribution of costs and benefits. While the above reconciliation of financial flows shows the impact as lifetime NPVs (and therefore subject to the problems of choosing discount rates, as dis- cussed in chapter 2), from the Government of Egypt’s perspective what matters are the actual incremental financial flows required each year to cover the differ- ence between purchasing CSP power and purchasing gas power. These are sum- marized in table 8.19: the 10-year cost is $348 million, starting in 2017 with an additional subsidy requirement of $32.5 million. It is clear that subsidies of this magnitude are unlikely to be acceptable to the government, and that a much higher proportion of grant is required. But even with a grant of 50 percent of the total (higher than the presently proposed $31.8 million–$362 million, with the balance financed just by the IBRD and CTF), Egypt’s incremental cost is still $46 million in NPV terms, with a ten-year (undiscounted total of $263 million). The annual subsidy requirement is ­ $6.9 million in 2017, increasing to $7.9 million by 2026 (table 8.20). At this level of concessionary aid and grants, the burden of incremental costs to Egypt falls to a more reasonable level—but the likelihood of grants and CTF of this magnitude are close to zero. Conclusions Avoided Cost of Carbon We have already noted in chapter 2 (table 2.8) that the carbon valuation for CSP is so much higher in Egypt ($267/ton), than in South Africa ($115−$155/ton CO2, depending on the technology configuration and storage provided): it is a simple consequence of what fossil fuel is displaced: in Egypt natural gas, in South Africa coal. The GHG emissions factor for coal is three times higher per kilowatt-hour than for a CCGT. In short, whether CSP has a reasonable (and ­ affordable) avoided cost of carbon depends on the technology against which it competes, and on the economic cost of fossil fuel. Concessionary Finance Without concessionary finance, expensive RE technologies impose significant incremental costs on the host country—and it is very hard to argue that given Egypt’s current political and economic situation, it should bear a significant share of the incremental costs of CSP. It is even more difficult to explain why Egypt The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 214 Table 8.18 Reconciliation of Economic and Financial Flows ($ Million, as NPV at 10 Percent Discount Rate): Revised Finance Package Consumers CSP Government Domestic banks EU CTF IBRD KfW EIB AfD Total finance Env Adjusted total 1. Benefits (cost of CCGT) 170 0 0 0 0 0 0 0 0 0 170 0 170 2. Consumer cost of electricity −95 0 95 0 0 0 0 0 0 0 0 0 0 3. Tariff revenue, SP 0 395 −395 0 0 0 0 0 0 0 0 0 0 4. Grants 0 25 0 0 −25 0 0 0 0 0 0 0 0 5. Loan disbursements 0 433 0 0 0 −97 −124 −115 −61 −35 0 0 0 6. Principal repayments 0 −128 0 0 0 9 23 55 26 14 0 0 0 7. Interest repayments 0 −61 0 0 0 2 25 17 10 7 0 0 0 8. OPEX 0 −63 0 0 0 0 0 0 0 0 −63 0 −63 9. Equity 0 113 −113 0 0 0 0 0 0 0 0 0 0 10. Construction costs 0 −571 0 0 0 0 0 0 0 0 −571 0 −571 11. Income tax 0 −29 29 0 0 0 0 0 0 0 0 0 0 12. Dividends 0 −115 115 0 0 0 0 0 0 0 0 0 0 13. Local environmental benefit 0 0 0 0 0 0 0 0 0 0 0 40 40 14. GHG benefit 0 0 0 0 0 0 0 0 0 0 0 51 51 15. Total 75 0 −269 0 −25 −86 −76 −44 −25 −14 −464 91 −373 Source: World Bank 2013. Note: AfDB = African Development Bank; CCGT = combined-cycle gas turbine; CSP = concentrated solar power; CTF = Clean Technology Fund; EIB = European Investment Bank; EU = European Union; Env = environmental benefits (GHG+local); GHG = greenhouse gas; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NPV = net present value; OPEX = operating expenses. Case Study: The Arab Republic of Egypt 215 Figure 8.15 Revised Finance Package 200 100 0 US$ million –100 –200 –300 –400 –500 s s t RD D v P B EU F l l W en er nk cia ta En CS EI CT Af Kf um IB to m ba an rn ed ns in tic ve st lf Co es Go ju ta m Ad To Do Source: World Bank 2013. Note: AfD = French Development Assistance; CCGT = combined-cycle gas turbine; CSP = concentrated solar power; CTF = Clean Technology Fund; EIB = European Investment Bank; EU = European Union; GHG = greenhouse gas; IBRD = International Bank for Reconstruction and Development; KfW = German Development Bank; NPV = net present value; USc = U.S. cents. Salient indicators Indicator Value Tariff, USc/kWh 12.7 Financial package Proposed package Share of incremental costs Egypt, Arab Rep. (Government + consumers + domestic banks) 42% ($194 million) International community 58% ($270 million) Table 8.19 Incremental Financial Flows for Tariff Support, Revised Financial Package Total 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Energy sold GWh 449 447 445 443 440 438 436 434 432 430 Cents/kWh 12.7 13.0 13.2 13.5 13.7 14.0 14.3 14.6 14.9 15.2 $ million 610 57 58 59 60 61 61 62 63 64 65 Gas: HFO tariff Cents/kWh 5.5 5.6 5.7 5.8 5.9 6.0 6.2 6.3 6.4 6.5 $ million 263 24.5 24.9 25.3 25.7 26.0 26.4 26.8 27.2 27.6 28.0 Net cost to government $ million 348 32.5 33.0 33.5 34.0 34.5 35.0 35.5 36.1 36.6 37.1 Source: World Bank 2013. Note: GWh = gigawatt-hour; HFO = heavy fuel oil; kWh = kilowatt-hour. The total is not discounted. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 216 Case Study: The Arab Republic of Egypt Table 8.20 Tariff Support, 50 Percent Grant Total 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Energy sold GWh 449 447 445 443 440 438 436 434 432 430 CSP tariff Cents/kWh 7.0 7.1 7.3 7.4 7.6 7.7 7.9 8.0 8.2 8.4 $ million 337 31 32 32 33 33 34 34 35 35 36 Gas: HFO tariff Cents/kWh 5.5 5.6 5.7 5.8 5.9 6.0 6.2 6.3 6.4 6.5 $ million 263 24.5 24.9 25.3 25.7 26.0 26.4 26.8 27.2 27.6 28.0 Net cost to government $ million 74 6.9 7.0 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 Source: World Bank 2013. Note: GWh = gigawatt-hour; CSP = concentrated solar power; HFO = heavy fuel oil. The total is not discounted; kWh = kilowatt-hour. should pay for CSP when the same quantity of GHG emission reductions can be achieved at a third of the incremental cost by wind—and indeed Egypt has some of the best wind resources anywhere in the world, with annual plant factors in excess of 40 percent. Of course, it is true that the presently high cost of CSP can only be brought down by a global commitment to the technology, but it is hardly an argument that Egypt’s (poor) consumers should carry the costs of this technology development. Notes 1. Million British thermal units. 2. One often hears the argument that a presentation of the project financial return is better because it is independent of the financial package that may be developed for the project, the precise details have yet to be negotiated, or (where the project is proposed by a state-owned utility) the returns on equity have no meaning. These are all feeble rejoinders, and particularly so in the case of an RE project with high incre- mental costs: the specifics of the financing package are central to project feasibility. 3. Of course, if the subsidy on electricity were eliminated, the consumer surplus would decrease. But this is more than offset by a reduction in production costs, the difference being the deadweight loss (see box 5.3 for explanation). But subsidies on fossil fuels should be reduced whether or not a CSP is implemented. 4. Needless to say, higher FIRR have significant impacts on the tariff. For the same sce- nario with 10 percent equity return, the required IPP tariff increases from 12.2 Cents/ kWh to 13.8 cents/kWh. Bibliography Economic Consulting Associates. 2007. Egypt: Economic Costs of Natural Gas. Final report, Economic Consulting Associates, Washington, DC, February. Fraunhofer and Lahmeyer International. 2012. Draft Feasibility Study for the Kom Ombo Concentrated Solar Power Project. Washington, DC: Fraunhofer and Lahmeyer International. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: The Arab Republic of Egypt 217 NREA (New and Renewable Energy Authority). 2010. Annual Report, Cairo, Egypt. Vagliasindi, Maria, and John Besant-Jones. 2013. Revisiting Standard Policy Recommendations on Market Structure in the Power Sector. Washington, DC: World Bank, Directions in Development. World Bank. 2013. Kom Ombo Project Appraisal Document. World Bank, Washington, DC. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 9 Case Study: Brazil Sector Background Brazil offers a special case for renewable energy (RE) because of its innovations to make its renewable market competitive based on the predominance of hydro- electricity. Brazil relies on hydroelectricity for most of its power supply, but the proportion of total supply from hydropower has declined steadily from over 90 percent in 1998 to 80 percent in 2012. Before 1995 the power sector was predominantly government controlled with vertically integrated companies. The federal company Eletrobrás and sev- eral state companies owned and operated most of the generation, transmission, and distribution in the country. The reforms of the electric power sector in Brazil were set in motion by the ratification of the Electricity Concession Law No. 9074 early in 1995. Eletrobrás retained the ownership of the transmission grid, the Brazilian part of the binational Itaipu Dam and hydroelectric power station, the nuclear power plants, and the Centro de Pesquisas de Energia Elétrica’s (CEPEL’s) research and development activities. The law provided for the unbundling of the sector—principally the functions of the dominant power generator and transmitter, Eletrobrás. Between 1995 and 1998 key institutions were created, including the Brazilian Electricity Regulatory Agency (ANEEL) as an independent regulatory entity, the National Energy Policy Council (CNPE) to propose national energy policies, and the system operator (ONS) to control power generation and transmission activities in the interconnected power system through a tight pool dispatch system, the Wholesale Electric Energy Market (MAE), which was created to promote the accounting of agents’ transactions in the multilateral short-term market under market rules. During 2001–02 Brazil suffered one of its worst droughts, which forced the government to implement a strict rationing program for nine months to reduce the load in 80 percent of the country by 20 percent. Special authority was given to an emergency committee in charge of the program. The country went from a situation of power supply scarcity to one of surplus, helped by some emergency generation capacity installed during the drought. In 2004 the government The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   219   220 Case Study: Brazil implemented a second wave of power market reforms, known as the new model, to address some of the problems associated with incentives for installing new generation capacity, improving competitive conditions, and strengthening the institutional framework. The main characteristics of this new model included an emphasis on the forward contract market to induce additions of new generation capacity, the strengthening of the regulatory agencies, and the requirement of mandatory energy auctions for distribution companies to cover 100 percent of all loads. The last requirement meant that distributors could acquire energy only through auctions for contracts of three to five years, which reduced risks for generation investors and promoted competition. Benchmark prices were used for passing through wholesale power costs to consumers procured under the new energy auctions, as the supply costs reflected the average price of all contracts. The two models differed in significant ways. The original reform model (implemented in 1995) was characterized by the opening up of the power mar- ket (with emphasis on the privatization of all the companies) and system expan- sion (to be achieved through short-term price signals and contracting obligations). In the new model (implemented in 2004) the emphasis was on coexistence between state-controlled and private companies, with the subsidiaries under Eletrobrás holding 69 percent of total transmission lines and about 68 percent of Brazilian distribution assets controlled by private sector companies (Vagliasindi and Besant-Jones 2013). Renewable Energy Development Brazil has the second-largest proven oil reserves in South America (12.9 trillion cubic feet of proven natural reserves), but remains a net energy importer. Brazil’s RE power capacity, including large-scale hydropower, is the fourth largest in the world. Its biomass power capacity is the second largest. The 4.8 gigawatts (GW) of biomass cogeneration plants at sugar mills generated more than 14 terawatt- hours (TWh) of electricity in 2009, nearly 6 TWh of which was excess fed into the grid. Also, 606 megawatts (MW) of wind farm capacity was installed, with another 450 MW under construction. Other than for small hydro (for which a detailed master plan is available [PECC1 2001]) and agricultural waste (biomass, which can be reliably inferred from official data on agricultural production), other RE resources suitable for grid-connected projects are either largely unknown (if not quite speculative, as in the case of geothermal), too small to make any signifi- cant contribution (such as landfill gas), or vastly overestimated in light of existing evidence (as in the case of wind, where estimates of “physical potential” have little practical meaning). The hydropower sector is highly developed in Brazil. This (including small- scale hydro) is the RE sector that requires the least amount of financing. The 10-year Energy Research Corporation (EPE) plan predicts that installed capacity from hydroelectric plants will rise from just less than 85 GW at present to more than 115 GW. The principal contributor to the increase in hydropower will come from the extra capacity generated by the proposed Belo Monte dam, to be built The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 221 on the River Xingu through a public-private partnership (PPP), which will com- mence power generation in January 2015. Belo Monte will be the world’s third- biggest hydropower plant. Brazil has an estimated 140 GW of total hydropower potential, with an estimated 40 percent remaining untapped, making it a valu- able resource for future electricity generation. Recent measurements carried out in 2008 and 2009 from the Brazilian wind atlas indicate that the real potential for wind power in Brazil is 350 GW. This is more than double the initial predictions from 2001 of 143 GW, positioning Brazil as one of the future global wind energy leaders. The Brazilian wind market has expanded tremendously since its commencement and now boasts several key market players. Latin America, led by Brazil, is expected to develop 46 GW of total installed wind capacity by 2025; the Brazilian market is expected to repre- sent 69 percent of the total installed capacity in Latin America by then. Brazil is the third-largest producer of biomass electricity behind the United States and Germany, thanks to large amounts of sugarcane waste that cover most of its needs for solid biomass electricity production. Because of the country’s location, levels of solar radiation (particularly in northern Brazil) are among the highest recorded in the world. The Amazon is the sunniest region in Brazil, with an average radiation level of 6,000 kilowatt-hours per square meter (kWh/m2). Solar energy potential is estimated at 114 GW. According to the Global Energy Network Institute, total installed capacity of solar photovoltaic (PV) energy is estimated at 12−15 MW and is primarily used to supply telecommunications and rural installations. In 2009 Brazil had approximately 5 million m2 of solar panels installed; the government plans to ­ triple the area by 2015. Solar hot water technologies are becoming widespread and contribute significantly to hot water production. Brazil led the market for newly installed c ­apacity worldwide during 2009, when its capacity increased 14 ­ percent, bringing total existing capacity to nearly 3.7 GW thermal (5.2 million m2). Geothermal remains the least-tapped energy sector in Brazil, with only 1.84 GWh produced in 2005. Despite there being a potential for exploiting geothermal energy, particularly in southern Brazil, investment is currently not being pursued. Renewable Energy Targets The Government of Brazil established formal targets for RE in the Program for the Promotion of Renewable Energy (PROINFA), introduced in 2002. The tar- gets, which were to be reached by 2006, are given in table 9.1. Although no new formal or national targets for RE have been established since the PROINFA, the government produces a 10-year indicative generation expan- sion plan periodically that provides key guidelines for system expansion without imposing a commitment on developing projects or technologies. The 10-year expansion plan released in 2010—covering the period 2010–19—prioritizes the development of RE with the special objective of complementing the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 222 Case Study: Brazil Table 9.1 Targets under the PROINFA Technology Target to 2006 (MW) Wind 1,100 Small hydro 1,100 Bioelectricity 1,100 Source: ANEEL. Note: MW = megawatt; PROINFA = Program for the Promotion of Renewable Energy. development of large hydroelectric capacity. But the expansion plan provides only a reference and does not establish any official targets for RE penetration in the country. The reference milestones are summarized below: • Wind energy: Brazil hit the 1 GW milestone in May 2011, but plans to have to 11.5 GW by 2020. • Small hydro: An increase from 3.8 GW in 2010 to 6.4 GW in 2020. • Biomass: An increase from 4.5 GW in 2010 to 9.2 GW in 2020. In total, wind, small hydro, and biomass are expected to reach 27 GW by 2020, compared to 9 GW in 2010. Investment plans to reach such reference points are as follows: • R$70 billion ($44.5 billion) for RE sources excluding large hydro. • R$96 billion ($60.7 billion) for large hydro plants. • R$25 billion ($15.8 billion) for fossil-fuel projects. In late 2010 Brazil enacted a decree targeting its carbon dioxide (CO2) emis- sions. The decree requires a 1.3 billion ton reduction in emissions by 2020 (UNEP, BNEF, and FS 2012). Brazil aims to maintain or increase the existing share of RE in total energy (44 percent in 2010) and in electricity generation (85 percent in 2010) through 2030, and this policy goal is broken down into a number of technology-specific goals. For wind the government has set a goal of achieving 11.5 GW of production capacity by 2020. Design of Incentive Schemes The Experience with Feed-In Tariffs The Brazilian government uses several tools to promote RE. In 2002 the govern- ment launched the PROINFA to encourage the use of RE sources such as wind power, biomass, and small hydropower. The program was intended to be imple- mented in two stages. By 2008 the PROINFA 1 was to add 3,300 MW of elec- tricity capacity stemming from RE sources (divided equally among wind, biomass, and small hydropower) to the interconnected system and establish a minimum national business participation rate of 60 percent (that is, equip- ment and services of national origin). In the second phase—which was never The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 223 implemented—the program called for a 90 percent national business participa- tion rate and established a target for RE supply at 15 percent of total annual electricity consumption. The chosen subsidy instruments were technology-­ specific feed-in tariffs (FITs) with a cap on the number of supported MW. The program was operated by Electrobrás, which bought energy at preset preferential prices (different for each of the three sources) and marketed the electricity. The cost of subsidies and incentives was covered by the Energy Development Account, funded by end-use consumers through an increase in energy bills. Low- income sectors were exempt from this increase. The PROINFA was expected to generate 150,000 jobs and leverage private investments of around $2.6 billion. The PROINFA 1 was completed in 2008 with 3.3 GW installed. Wind farm capacity increased from 22 MW in 2003 to 606 MW in 2009, as part of 36 private projects; another 10 projects with a capacity of 256 MW were under ­ construction, while 45 additional projects with a capacity of 2,140 MW were approved by ANEEL. The capacities (MW) of the supported biomass projects were far below the original target: the FIT for biomass projects was too low, mak- ing it more favorable for new biomass plants to sell directly to the wholesale market. The incentives in the PROINFA included a technology-specific FIT and a purchase obligation on final consumers. The FIT level was established by the Ministry of Mines and Energy (MME) and designed as an adjustable 20-year tariff, indexed to inflation. The FIT was designed as a function of the plant’s capacity factor (CF) with the aim of: (a) promoting the development of wind- based generation in different geographic locations and avoiding possible transmission capacity bottlenecks and (b) avoiding an overcompensation of elec- ­ tricity generation at good wind locations (that is, minimizing the producer sur- plus for plants with high CFs). The incentive was limited to 220 MW per state, also with the intention of avoiding geographic concentration and bottlenecks in the electrical grid (for example, the best wind conditions can be found in the northeast of Brazil, where the transmission network is less developed). Wind farms with lower CFs received a higher compensation per energy unit than wind farms with higher CFs. The design therefore included a FIT that increased lin- early as the CF l­ owered (that is, from R$180.18/MWh for a CF of 41.93 percent to R$204.35/MWh for a CF of 32.40 percent). The minimum and maximum CFs were fixed by the MME. The CFs of dif- ferent plants were verified periodically to adjust the compensation level in the 20-year power purchase agreement (PPA). The average price paid to each tech- nology (2010 prices) is provided in table 9.2. The PROINFA was indeed the first step toward scaling up RE in Brazil, but it has been criticized for the lack of economic rationale behind project allocation procedures and for the imposition of rules that have created various bottlenecks to RE development. The first criticism is that the allocation of the targeted amount of 3,300 MW in equal shares of 1,100 MW to each source did not pro- mote the least-cost expansion of RE capacity in the system. Project selection— within the technology-specific quotas—was also not based on a least-cost approach. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 224 Case Study: Brazil Table 9.2 Average FIT Levels under the PROINFA Technology PROINFA FIT ($/MWh) Wind 154 Small hydro 96 Bioelectricitya 77 Source: ANEEL (2010 values). Note: Exchange rate: $1 = R$1.85. FIT = feed-in tariff; MWh = megawatt-hour; PROINFA = Program for the Promotion of Renewable Energy. a. Price includes taxes. Projects were selected based on the dates relevant to environmental permits being issued. The older the permit, the closer the project was in the merit order for contracting. This ended up creating a “black market” for environmental licenses. In fact, the issue of permitting and licensing became a bottleneck to the introduction of new capacity in general, creating serious economic distortions and high transaction costs, leading to lengthy court cases. In addition, the mini- mum national business participation rate of 60 percent required by the PROINFA became a bottleneck to wind generation development, given that Brazil had just one local wind manufacturer at the time. As a result, not all tech- nologies could reach their quotas, and some volumes of capacity were transferred from one technology to the other to achieve the total target of 3,300 MW. The PROINFA was also very much criticized for its management of the clean devel- opment mechanism (CDM) revenues. Under the program, Eletrobrás was responsible for managing the CDM revenues to reduce program costs, which were supposed to be passed on to consumers. But Eletrobrás was unable to prepare and submit the CDM projects as required by formal international ­ ­ procedures, and therefore could not collect the corresponding carbon revenues. The PROINFA target was reached four years later, in 2010, and with the introduction of the auction-based approach, the program was closed without ­ entering into a second phase. From Feed-In Tariffs to Auctions Under the regulatory structure introduced in Brazil in 2004, most new power projects participated in auctions for long-term PPAs with energy distributors who were required to enter into long-term contracts for all of their electricity demand via a reverse auction system. The energy auctions were carried out by ANEEL through a delegation from the MME. There were specific auctions for both existing energy sources and for new energy sources. Existing plants were offered short- to medium-term contracts (from a few months to eight years), while new energy initiatives were offered long-term contracts (15−30 years). The clearing price of existing plants was lower than the clearing price of new energy. Auctions for RE plants targeted specific energy sources and large hydropower- project-specific sites. The tenders fixed maximum price caps and had penalties built in for developers who signed contracts they could not uphold.1 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 225 The procurement of new generation projects is carried out regularly through two public auctions every year (see table 9.3): one for electricity delivery three years ahead and another one for electricity delivery five years ahead (usually referred to as A-3 and A-5 auctions). Each auction offers long-term energy con- tracts (15-year-duration contracts for thermal plants and 30-year-duration con- tracts for hydro plants). The auction contract can be of two types: (a) standard financial forward contracts, where generators bid an energy price and (b) energy call options, where generators bid an option premium ($/MW) and an energy strike price ($/MWh). In the call option proposal, the consumer “leases” the plant from the investor, paying a monthly fixed amount (to allow recovery of investment and fixed costs) for its availability and reimbursing the plant’s owner on its declared variable operating costs whenever the plant runs. In this case, the consumer is responsible for the cost of trading on a spot basis. Since spot prices tend to be low most of time, the option contract is very attractive. The contract auctions are organized by the government as a centralized process, carried out jointly to meet the total load increase. The objective of the joint auction is to allow smaller distributing companies to benefit from economies of scale in the new energy contracting. But the government does not interfere on the demand Table 9.3 Renewable Auctions Date Name Technology 18/06/2007 1º Leilão de Energia de Fontes Alternativas Biomass, wind 26/07/2007 4º Leilão de Energia Nova Hydro 16/10/2007 5º Leilão de Energia Nova Hydro 10/12/2007 Leilão da Usina de Santo Antônio Hydro 19/05/2008 Leilão da Usina de Jirau Hydro 14/08/2008 1º Leilão de Reserva Biomass 17/09/2008 6º Leilão de Energia Nova Hydro, natural gas 30/09/2008 7º Leilão de Energia Nova Hydro, biomass 27/08/2009 8º Leilão de Energia Nova Biomass 14/12/2009 2º Leilão de Energia de Reserva (eolic) Wind 20/04/2010 Leilão da Usina de Belo Monte Hydro 30/07/2010 10º Leilão de Energia Nova A-5 Hydro 25/08/2010 3º Leilão de Energia de Reserva (Fase 1) Biomass 25/08/2010 3º Leilão de Energia de Reserva (Fase 2) Biomass 25/08/2010 3º Leilão de Energia de Reserva (Fase 3) PCH, biomass, wind 26/08/2010 2º Leilão de Fontes Alternativas Eolic, biomass 10/12/2010 9º Leilão de Energia Existente (A-1) Hydro, biomass 17/12/2010 11º Leilão de Energia Nova (Hídrica A-5) Hydro 17/08/2011 12º Leilão de Energia Nova Hydro, wind, biomass 18/08/2011 4º Leilão de Energia de Reserva Biomass, wind 30/11/2011 10º Leilão de Energia Existente (A-1) Hydro 20/12/2011 13º Leilão de Energia Nova (A-5) Wind, biomass, hydro 14/12/2012 14º Leilão de Energia Nova (A-5) Wind, hydro Source: ANEEL. Note: PCH = pequeñas centrales hidroeléctricas (small hydro). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 226 Case Study: Brazil forecast (which is directly declared by the distribution companies) or the energy contracts (each winning generating company signs separate—private—bilateral ­ contracts with each of the distribution companies, in proportion to their fore- casted loads). The auction mechanism follows a hybrid design, combining an iterative descending clock auction with a final pay-as-bid round. Finally, in the regular new energy auctions, all technologies compete jointly. Candidate genera- tors require either a concession (in the case of medium and large hydropower facilities) or an authorization (for all other plants). Authorizations and conces- sions are granted by the MME. Concessions are also granted through auctions, after the EPE studies a relevant site and the MME approves the project. ANEEL held the first biomass-only reverse energy auction in 2008, contracting 2,379 MW produced by 31 thermoelectric plants (using sugarcane and napier grass) with supply beginning in 2009 and 2010, and contracts extending for a 15-year period. The final average price was $32/MWh. With a baseline of 554 MW, ANEEL contracted an additional 191 MW in 2010 and 60 MW in 2011. The first wind energy auction was carried out in December 2009, resulting in 1.8 GW being contracted from 71 wind power plants scheduled to start opera- tions by July 2012. In August 2010, 89 projects—representing 2.9 GW of installed capacity and involving R$26.9 billion ($15.2 billion) in investments— were contracted from biomass and wind farm developers. Biomass projects with a capacity of 713 MW were contracted at an average price of R$144/MWh, or $83.50/MWh, while the 2.1 GW generated from wind power were contracted at an average price of $74.4/MWh. In December 2012, 12 wind and hydro proj- ects of 574 MW were contracted for an average price of R$91.25. As shown in figures 9.1−9.3 the use of auctions resulted in significant savings over time due to a sharp decline in prices, particularly in the case of wind. Discounts on Transmission and Distribution Tariffs Law 9427/96 sets specific incentives for the sale of RE through contracts in the free market. These incentives take the form of discounts on transmission and distribution (T&D) tariffs for consumers who purchase energy through contracts signed with nonconventional RE developments of up to 30 MW. Although it was introduced in 1996, the incentive was confirmed in ANEEL’s Resolution No. 247 of 2006, which established the regulations on the commercialization of RE-based generation (small hydropower, wind, biomass, and solar initiatives with capacities below 30 MW). This resolution also extends the incentive to regulated ­ consumers with loads greater than 500 kilowatts (kW) for whom the wire tariffs are high: as the price reflects distribution costs at lower voltages, an RE contract produces substantial savings. Also, for some types of RE-based generation, this option is economically more attractive than are energy auctions. At present, Brazil has about 50 trading companies, and the number is rapidly growing. Although it is difficult to estimate the amount of RE-based capacity that has been attracted by the incentive, given the small scale and distribution of initiatives, it is estimated that more than 500 MW of both small hydro projects (SHPs) and bioelectricity have been installed under the scheme. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Fo 200 R$ per MWh R$ per MWh nt 7 En es 1 er 0 20 40 60 80 100 120 140 160 Al Le gi 2 t 0 20 40 60 80 100 120 140 160 i a d 00 20 ern lão e 92 Source: ANEEL. Source: ANEEL. 07 ati de Re 5 v se Le Us En L as rv ilã in er ei g a( o Case Study: Brazil ad ia lão En eo de e 200 No de er lic Sa 7 va nt Le gi ) ad 2 Note: MWh = megawatt-hour. Note: MWh = megawatt-hour. o ilã A o e 010 20 ntô da Re 3 Us 08 nio se L in Le rv ei a ( lã 20 a d ilão Fa o d 08 e J da ir se e En 7 L au 3) er ei Us g lã Fo 201 in 2 ia N o d nt 0 En a 0 de 10 ova e es 2 er Al Le gi 2 Bel Leil te ilã a N 01 o M ão rn o ov 0 1 on da at de En a( 0 t iv er H Le e 20 as gi 11 a d 2 ídric ilão e 010 a A de Re 3 - 1 se L 5) En 2 L rv ei er ei 2 a l gi lão Fo 01 (Fa ão aN d nt 0 se de Figure 9.1 Price Evolution through Wind Auctions, 2009–12 En es 2 3 ov e a er A L ) gi 2 lter eilã 0 aN 1 n o En 201 ov 0 1 ativ de er 1 En a( 1 a gi 4 L s Figure 9.2 Price Evolution through Hydroelectric Auctions, 2007–11 er ad L gi 2 Híd eilã e eilã a N 01 ric o Re o ov 0 1 a A de se de a ( 1 -5) rv En Hí Le 20 a er 20 dric ilão gi 1 a d En 12 a E 0 9 A- e xi 5 er 14 L ) gi 20 sten eilã a N Le 11 te o ov ilão 1 (A de a( d En 2 -1 A- e En 2 erg Leil ) 5) er 01 ia ão gi 1 a E 10 Nov de xis L a 2 ten eilã En 011 te o d er 13 (A- e The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 gi a N Le 1) ov ilão a( d A- e 5) 227 228 Case Study: Brazil Figure 9.3 Price Evolution through Biomass Auctions, 2007–11 160 140 120 100 R$ per MWh 80 60 40 20 0 08 iva e se de No de se e se e se e A- e ov e se de A- e at d Fa d a d 2 (Fa ã d Fa d e( d aN d a( d va a s a 1) 2) 3) 1) a 5) Re o ov ilão rn ã Re lã rg ilã a ( ilã a ( ilã nt lã gi lã rv rv e eilã te il a il i te ei er ei Al Le Le e rv Le rv Le rv Le a N Le L L L ad L ia es 1 1 En 7 3 3 se 3 xis 9 En 2 gi 4 er 13 1 nt 07 08 Re 0 Re 0 Re 10 aE 0 er 1 e e 01 e 01 gi 01 11 En 201 En 11 se se Fo 20 20 20 0 gi ad 2 ad 2 er 2 20 20 e En gi gi gi er er er En En En Source: ANEEL. Note: MWh = megawatt-hour. Another strength of Brazil’s RE development strategy is that it emphasizes the employment and regional development potential of the RE sector. The Brazilian Development Bank (BNDES) plays a central role in RE finance countrywide. Its funds are often passed to regional banks, which help build the capacity of the more local financing institutions. BNDES is the favored channel for funding for international donors and finance partners, such as the German Development Bank (KfW), which provides a credit line to BNDES for SHPs, supports pilot projects in biogas, and works on grid-connected PV pilot projects. BNDES’s overall RE lending amounted to $6.4 billion in 2009. Moreover, the government uses a number of instruments to ensure that RE investments support the creation and growth of national businesses. To benefit from subsidies and from BNDES financing, projects must fulfill national content requirements. Law 10762 mandates a minimum nationalization of 60 percent in total construction costs, as well as regionalization criteria, where each state has maximum limits of 20 percent of total capacity for wind and biomass and 15 ­percent for small hydro. Foreign manufacturers of RE and energy-efficiency technology, moreover, face a 14 percent tax surcharge on imports. The 60 percent national content requirement has led to significant installed production in Brazil. Major industry companies (such as Siemens, GE, Vestas, Suzlon, and Führlander) have now started production in Brazil or are actively seeking local presence there. Regional banks, such as Banco de Nord Este, are also active in RE finance, but generally work with BNDES’s funds that are passed on to the regional level. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 229 The conditions offered by BNDES to potential projects under the PROINFA were adjusted in 2005 (table 9.4) to support the actual deployment of RE-based capacity, as the combination of incentives (FIT offered by the PROINFA and soft loans offered by BNDES) was not sufficient to trigger investments in RE. BNDES was therefore the main investment development bank for renewables in Brazil under the PROINFA. In general, the support of BNDES is aligned with the federal government’s programs and still plays an important role in the financ- ing of RE capacity. For instance, in 2008 BNDES approved loans for 11 biomass cogeneration projects, 2 wind-based power plants, and a landfill gas initiative with a combined capacity of 532 MW (BNDES 2013). Today, special financing conditions are given by BNDES to different types of generation capacity. These are released directly by the bank through its own programs and regulations. Table 9.5 illustrates the financing conditions applied today, considering the risk spread is upper bound. The rural electrification program “Light for All” has a strong RE component. It assumes that (a) the use of approximately 130,000 PV systems is the most economically efficient electrification option for about 17,500 localities with small populations in the Amazon territory; and (b) a further 2,300 villages with about 110,000 buildings could be equipped with a mini-grid based on PV or biomass sources, 680 additional medium-sized communities could be supplied Table 9.4  Financing Conditions Offered by BNDES under the PROINFA Financial conditions 2002 Adjustment in 2005 Debt share (depending on Up to 70 percent Up to 80 percent nationalization quota) Amortization period 10 years 12 years Interest rates TJLP plus BNDES charges About 13.25% (TJLP plus BNDES charges) Source: ANEEL. Note: BNDES = Brazilian Development Bank; PROINFA = Program for the Promotion of Renewable Energy; TJLP = long-term interest rate. Table 9.5  BNDES Financing Conditions Offered to Generation Projects Equity (%) Amortization Interest rates Small hydro 30 14 years 100% TJLP* + 2.8% spread Biomassa 20 14 years 100% TJLP + 2.8% spread Wind 30 14 years 100% TJLP + 2.8% spread Thermalb 30 14 years 50% TJLP + 50% currency basket + 2.8% spread Source: BNDES 2013. Note: BNDES = Brazilian Development Bank. a. It is considered that biomass can finance the whole project, resulting in 20% of equity. b. For liquefied natural gas (LNG) power plants. *TJLP: Brazilian long-term interest rate; as of June 2010, its value was 6%. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 230 Case Study: Brazil on the basis of hybrid systems, and 10 larger communities could be provided with power based on conventional diesel generators or hybrid systems. Brazil is a successful promoter of CDM projects, accounting for 40 percent of all CDM projects in South America and for 44 percent of contracted certified emission reduction (CER) credits up to 2012. Brazil’s National Fund on Climate Change is an example of a holistic fund concept with a strong RE component. It aims to mitigate the environmental impact of oil production by allocating a por- tion of the state’s revenue from oil to support projects, studies, and enterprises relating to climate change mitigation and adaptation. The law establishing the fund was adopted in December 2009. At that time, the government pledged $113 million, part of which would come from oil industry revenues. The fund has already started supporting mitigation and adaptation programs and projects involving a wide range of activities. These activities include capacity building, climate science, adaptation and mitigation projects, projects aimed at reducing carbon emissions from deforestation and forest degradation (particularly in vul- nerable areas), development and dissemination of technologies, research and development (R&D), development of products and services that contribute to mitigation and adaptation, payment for environmental services, establishment of agro-forestry systems that contribute to reducing deforestation and carbon sinks, and the rehabilitation of degraded areas. In August 2012 ANEEL announced two new pieces of regulation to support the solar industry: (a) a net metering for micro generation up to 1 MW, and (b) a tax break of 80 percent for installations up to 30 MW. ANEEL also announced that it would launch an auction for solar projects between 1 MW and 3 MW, but no details are available yet. Financing of Incremental Costs Eletrobrás was in charge of administering the PROINFA and of transferring the expenditures to consumers in proportion to their consumption (with the excep- tion of the residential low-income subclass, or those with consumption levels below 80 kWh/month). Thus a specific “levy” was applied to recover the incre- mental costs associated with RE. For new projects, the PROINFA system has been replaced by ANEEL’s energy auctions, which also changed the way the incremental cost of RE is financed. Acquired power is fed into the power pool at the contracted price, raising the averaging pool price. The increase is subject to a politically fixed maximum: the average price of energy for end consumers can increase up to a cap of 0.5 percent annually and 5 percent over 20 years. Wind became one of the cheapest sources of power in Brazil as a strong cur- rency and slowing global demand for turbines drove down costs. Developers agreed to deliver electricity generated by new wind farms at an average price of R$99.54/MWh ($55.99/MWh) in a government-organized auction in August 2011. This was cheaper than two natural-gas thermal-electric plants and a hydro- electric plant expansion that participated in an energy auction a day earlier, The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 231 and 33 percent cheaper than contracts awarded in the country’s first auction for wind power in December 2009. The average price in the A-5 2012 auction on December 14 was R$87.94/MWh ($42.16/MWh). This was 9 percent below the lowest price contracted in the 2011 auctions, 12 percent below the average prices in 2011, and 21.5 percent below the R$112/MWh “reference price” set by the Brazilian Government Energy Agency, EPE, which manages Brazil’s energy auctions. Taking into account the fall in the value of the Brazilian real against other currencies since the 2011 auctions, the low prices are even more concern- ing. If inflation and exchange rates are taken into account, the prices should in fact have been around R$122/MWh. Just 281.9 MW of wind energy was ­ contracted in 10 projects ­scheduled for completion by 2017. This stands in sharp contrast to the August 2011 auctions, which saw 1.9 GW of wind power contracted for completion by 2016. ­ Prices for wind energy in Brazil, currently the lowest in the world, may rise at least 15 percent due to government policies designed to make the nation’s power grid more reliable. Developers must install as much as 15 percent more generat- ing capacity at new wind farms to compensate for the variable output from turbines. They also face new restrictions on where they can build. The second policy requires developers to build their own power lines to connect wind farms to the grid or install turbines near existing cables, sites that may not have the most wind. Under previous policies, the government auctioned the right to build power cables that linked wind farms to the grid. The two measures were sched- uled to apply to wind projects that participated in an August 2013 government- organized auction to sell power (and perhaps in another auction that same year). The rules were expected drive up the cost of power from wind farms that require more than 150 km of power lines. There are about 600 MW of wind turbines installed in Brazil’s northeast (where the breezes are the best) that aren’t con- nected to the grid because the power distributor responsible for building the transmission lines is behind schedule. Conclusions The Brazilian experience with FITs and in recent energy auctions is revealing. Auctions have proved to be an interesting way to support the implementation of RE at a minimum cost, for a given portfolio of technologies and renewable quo- tas defined as part of the energy policy agenda. An auction is perhaps an indirect way to achieve a FIT price discovery. As with FITs, long-term contracting reduces risk aversion and facilitates project financing. In principle, auctions maintain the advantages of FITs (income certainty) while also minimizing costs to consumers, thanks to the exercise of a competitive process. The technology-targeted energy auctions have catalyzed the RE market and provided: • A reliable policy framework for investors. • Involvement from public and private investors. • Development of a local RE industry. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 232 Case Study: Brazil Brazil provides an excellent example of the implementation of creative policy measures, which, in combination with financial and risk mitigation support, have been able to increase the national RE capacity. Table 9.6 shows a comparison of the PROINFA with RE-specific auctions in terms of their respective resulting prices, volumes, and costs. The main observa- tion is that although the annual costs of both mechanisms are practically the same (around $1 billion), the energy auction scheme is expected to deliver 20 percent more total capacity, with an average energy cost, and an expected tariff 60 percent lower in the case of wind. In the case of bioelectricity, plants acquired through the auction scheme exhibited higher efficiencies. The RE-specific auctions being implemented in Brazil can facilitate the intro- duction of specific projects while avoiding speculative behavior in auction participation. But the low prices achieved in the wind auctions have raised the fear that projects will not be implemented due to foreseen financial insolvency. On the other hand, if all projects were implemented, the low prices obtained in the wind auctions might have paved the way to a direct competition between wind and other sources. This could avoid the organization of specific auctions for this tech- nology, and wind power could start competing in the regular contract auctions organized by the distribution companies, in which all technologies participate on a level playing field without discrimination. Indeed, policies seeking to promote the introduction of RE in an economically efficient way must take into consideration the costs of RE generation (in relation to the avoided social cost of generation), resource availability in relation to sea- sonality, as well as the technical conditions of the system (for example, capacity of T&D lines to absorb specific volumes of RE). An assessment of policy effi- ciency in this context requires complex modeling coupled with the use of other Table 9.6 A Comparison of the PROINFA and RE-Specific Auctions Technology-specific auction (“reserve PROINFA energy” auction) MW GWh/year $/MWh MW GWh/year $/MWh Wind 1,423 3,740 154 1,800 6,596 80 Small hydro 1,191 6,260 96 — — Bioelectricity 779 2,661 77 2,379 4,800 84 Small hydro Impact on costs Total capacity (MW)a 3,393 4,179 Total energy (GWh/year) 12,661 11,397 Average cost ($/MWh) 109 80 Sources: Eletrobrás, EPE, Aneel, ONS, and PSR. Note: Exchange rate: $1 = R$.1.85. Values as of April 2010; prices with taxes. Gross cost is total (fixed) cost paid by the consumers. For the auction case, the net cost includes estimates of yearly spot revenues collected by consumers. GWh = gigawatt-hour; MW = megawatt; MWh = megawatt-hour. PROINFA = Program for the Promotion of Renewable Energy; RE = renewable energy. — = not available. a. Installed capacity includes self-consumption. In the auction case, energy values correspond to the excess energy sold to the grid at the auction. More excess energy from the new plants is available to be sold to the free market or at future auctions. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Brazil 233 tools to help analyze the adequacy of the institutional structure in place and governance issues. Nevertheless, and despite the stepped-up tariff, the design of the PROINFA did not explicitly promote the least-cost introduction of RE, since it established equal targets for different types of technologies (a limit of 220 MW of RE per state) and introduced restrictions in the form of a “minimum national parti­ cipation rate,” which became a bottleneck to the development of wind-based capacity. The PROINFA also did not provide any signal for technology improvement; for instance, the extra energy or surplus produced due to technology upgrades or efficiency improvements was not considered under the program (for example, degression factors in FIT design based on empirically derived progress ratios). Signals for economic efficiency—in terms of least-cost expansion of RE—were poor due to the administrative setting of different prices for different technolo- gies. But beyond the design features of the PROINFA, the management of the program also hindered the introduction of best-performing sites, as projects were selected based on the dates environmental permits were issued. The PROINFA also centralized the management of CDM revenues under very inefficient oversight. On the other hand, the auctions—as competitive mechanisms—seek to stimu- late the introduction of least-cost generation. But the first “reserve energy ­ auction” for wind, carried out in 2009, concluded in very low—perhaps artificial—prices and raised concerns about the risk of delays in construction or ­ of no wind plants being constructed at all. It is also interesting to note that the 2009 auction did not result in a clear correlation between CFs and prices. It is of course too early to assess the merits of the auction scheme in deploying RE with economic efficiency. In terms of economic efficiency, the reserve energy auctions meant to speed the introduction of RE may have other caveats: (a) they are not technology neu- tral, and (b) the type of technology and contract volume is discretionary (that is, the government has the prerogative to call an auction to contract a given volume of energy even if it is not contemplated in the demand forecasts prepared by the distribution companies). Note 1. For more details of the auction system, see Maurer, Barroso, and Chang (2011). Bibliography BNDES (Brazilian Development Bank). 2013. “Corporate Presentation.” http://www​ .bndes.gov.br/SiteBNDES/export/sites/default/bndes_en/Galerias/Download/AF​ _­DEPCO_english.pdf. Maurer, L., L. Barroso, and J. M. Chang. 2011. An Overview of Efficient Practices. Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 234 Case Study: Brazil PECC1 (Power Engineering Consulting Joint Stock Company 1). 2001. “Small Hydro Masterplan.” PEEC1, Hanoi. UNEP, BNEF, and FS (United Nations Environment Programme, Bloomberg New Energy Finance, and the Frankfurt School). 2012. Global Trends in Renewable Energy Investment. http://fs-unep-centre.org/sites/default/files/publications​ /­globaltrendsreport2012.pdf. Vagliasindi, M., and J. Besant-Jones. 2013. Revisiting Standard Policy Recommendations on Market Structure in the Power Sector. Washington, DC: World Bank, Directions in Development. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 10 Case Study: Turkey Sector Background Turkey is moderately endowed with primary energy resources, mainly ­ hydropower and lignite with some natural gas, and therefore relies on imports for about 70 percent of its energy needs. But it lies at major international cross- roads of energy trade between the gas- and oil-rich regions of the Middle East and Central Asia and the major European demand centers, which enables Turkey to diversify its sources of imported energy and to profit from extensive transit trade in energy products. The country has abundant wind and solar resources that promise to decrease its dependency on imported fossil fuels while reducing greenhouse gas (GHG) emissions. In 1993 the Turkish Electricity Authority (TEK) was split into two separate public utilities, which were corporatized: (a) the Turkish Electricity Generating and Transmission Corporation (TEAS), responsible for both generation and transmission activities, and (b) the Turkish Electricity Distribution Company (TEDAS), responsible for distribution and retail sale activities (Vagliasindi and Besant-Jones, 2013). In 2001 the Electricity Market Law (EML) No. 4628 was passed; its aim, among others, was to ease the burden of the power sector on the public budget. The provisions of the EML were designed to be in line with the European Union’s (EU’s) Energy Acquis, as part of Turkey’s ambition to join the European Union (EU). The law overhauled electricity legislation and set the foundation for a radi- ­ ramework in both the design and regulation of the Turkish electric- cally different f ity market. The law provided for the unbundling of state-owned electricity assets, opened the market above a certain level of electricity consumption, and allowed third-party access to the grid. The EML required the creation of a bilateral con- market, complemented by a residual balancing mechanism. All generation tracting ­ capacity was to be sold to wholesalers, retailers, and consumers either directly or ­ via a spot market. In response, the TEAS was unbundled into three separate ­state-owned entities: • The Electricity Generation Company of Turkey (EUAS) for generation. The EUAS directly owns most hydropower units and acts as the holding company for six The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   235   236 Case Study: Turkey portfolio generation companies with thermal power units and some hydropower units. In addition, several private sector generating units estab- ­ lished under build, operate, transfer (BOT), build, operate (BO), and transfer of operating rights (TOOR) contracts supply power to the grid on the basis of power purchase agreements (PPAs) guaranteed by the government.1 There are also a few privately owned independent power producers (IPPs). Industries with captive generating units (autoproducers) and privately owned renewable energy (RE) units also supply to the grid. • The Turkish Electricity Transmission Company (TEIAS) for transmission and ­dispatch. The TEIAS also operates the balancing market, which complements the bilateral free market, and acts as a settlement agency. • The Turkish Electricity Trading and Contracting Company (TETAS) acts as the single buyer of electricity sold under the PPAs for BOT, BO, and TOOR units, and on-sells this electricity to the distribution companies. Although corporatized with separate accounts, these entities remain subject to government decision making and have little managerial autonomy. Distribution is handled by 21 regional distribution companies, 20 of which are the holders of operating rights for their franchise areas from the TEDAS. The remaining one (Kayseri) is a privately owned distribution company. The EML also established the Energy Market Regulatory Authority (EMRA) as an independent and finan- cially autonomous regulator of power, gas, petroleum, and liquefied petroleum gas, to be supervised by the Energy Market Regulatory Board. In 2004 the strategy paper “Road Map of the Market Reform and Transition” was approved by the Higher Planning Council. It outlined the steps for further liberalization of the electricity sector. It covered procedures for privatizing distri- bution and generation assets with the introduction of transitory vesting contracts through which generation—either from existing contracts (via the TETAS) or from public companies—would be allocated to distribution companies based on their weighted share in total demand (to compensate for the demand of captive consumers). The strategy paper also provided the basis for determining the ­ revenue requirements of the regional distribution companies ex ante. Any pos- ences between the ex ante revenue requirements of the distribution sible differ­ companies and the real incomes collected via the tariff in force were expected to be reimbursed by means of a price equalization mechanism. The paper also envisaged the implementation of a national tariff. As set out in the strategy paper, the TEDAS, with its 20 regional distribu­ tion companies, was transferred to the Privatization Administration (PA) on April 1, 2005. A competitive wholesale electricity market went into operation in 2006. A balancing and settlement system was developed and started operating as a branch of the TEIAS. By 2010 approximately 400 private companies—­ dispatching about 30 percent of total electricity supply, on average—were trading power in this mar- ket. The EMRA issued new balancing and settlement regulations to improve the functioning of the wholesale electricity mar­ ket in April 2009. In December 2009 the market moved from monthly settlement to hourly s ­ ettlement (figure 10.1). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 237 Figure 10.1 Evolution of the Power Market in Turkey: Key Phases Final phase First phase Day ahead market Second phase Balancing power market Balancing Day ahead planning mechanism Hourly settlement Balancing power Demand side participation Monthly 3 market periods Enables market splitting Hourly settlement settlement Base-load futures offered Intra-day market underway Aug. 2006–Nov. 2009 Dec. 2009–Nov. 2011 Dec. 2011–Nov. 2013 Source: Vagliasindi and Besant-Jones 2013. Renewable Energy Resource Endowment According to the Wind Energy Potential Map of Turkey (REPA), the high-­ efficiency wind energy potential in Turkey is nearly 19,000 megawatts (MW); high-potential fields are located in Aegean, Marmara, and the coastal part of the Eastern Mediterranean regions. On the other hand, the REPA study showed that the technically feasible installed capacity potential in regions having a wind speed between 7.5 to 8 meters per second (m/sec) is 29,259 MW, while the potential in more than 9 m/sec wind-speed regions is only 196 MW. That is, Turkey has a 48,000 MW mid-high-efficiency wind energy generation potential and an annual average wind speed of 7.5 m/sec and higher. Renewable Energy Development Turkey’s installed power generation capacity in 2012 consisted of 10,100 MW of lignite- and coal-fired plants, 17,600 MW of gas- and oil-fired plants, and 13,900 MW of hydroelectric plants, with 600 MW of geothermal and other types of capacity (see figure 10.2). Annual generation of electricity was 198.6 terawatt-hours (TWh) in 2008, of which about 66 percent was from thermal power generation and 33 percent from hydroelectric gen­ eration (the remaining 1 percent was from geothermal and wind power). This amount of power was supplied to 29.52 million consumers. The country is nearly entirely electrified, mostly from these power networks. Renewable Energy Targets The Electricity Market and Supply Security Strategy Paper, issued by the High Planning Council in 2009, set the following targets: (a) wind electricity genera- tion capacity to be increased to 20,000 MW by 2023, (b) the known geother- mal capacity of 600 MW suitable for electricity generation and all technically possible hydroelectric capacity to be fully utilized by 2023, and (c) the share electricity generated using renewable sources to be increased to at least of ­ The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 238 Case Study: Turkey Figure 10.2 Evolution of the Power Market in Turkey: Generation and Installed Capacity Percent a. 2012 electricity generation (TWh) b. 2012 installed capacity (MW) Hydro, 57.84; Geothermal, 24 162.2; 0 Other renewable, Wind, 2,260.5; 158.5; 0 4 Wind, 5.85; 3 Geothermal; 0.85, 0 Hydro, 19,619.7; Thermal, 174.54; 35 Thermal, 34,870.6; 73 61 Source: TEIAS. Note: MW = megawatt; TWh = terawatt-hour. 30 percent of the total electricity generation. Turkey’s first wind power plant was set up in 1998 at Cesme-Alacati with an installed capacity of 1.5 MW, according to the BOT model. Design of Incentive Schemes Following the enactment of the EML in March 2001, the process for the installation of RE plants was tailored according to the law, and the process gained pace by the enactment of the Renewable Energy Law (REL). The REL (No. 5346) enacted on May 18, 2005, introduced certain advantages with respect to floor price and priority dispatch. The law included wind, solar, geo- thermal, biomass, biogas, wave, stream, and tidal energy resources; canal and river-type ­ hydroelectric-generating facilities; and hydroelectric generation facilities with a reservoir area of fewer than 15 square kilometers (km2). A Renewable Energy Resource Certificate (RER certificate) was introduced so investors could benefit from these advantages. But this law did not get the desired results, as the declared floor price was found very low by the investors and/or the lenders. Therefore, initially the Turkish Average Wholesale Electricity Price was used to promote all types of RE, and then a floor price of 5e cents/ kilowatt-hour (kWh) and a cap price of 5.5e cents/kWh were also applied. The REL has been amended at various times; the most recent comprehensive amendment became effective on January 8, 2011. According to the REL and related regulations, a “renewable pool” was introduced. Renewable generation facilities are supported by distributing the total cost of the electricity supplied to the pool among all the suppliers selling energy to final consumers (rather than only to the direct purchaser of the energy generated by each facility). The tariff is applied for a period of 10 years from the first operation date (if the commissioning date is between May 18, 2005, and December 31, 2015). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 239 Other critical provisions of legislation are summarized as follows: • The support scheme is valid for the facilities commissioned until the end of 2015. • For the later period, beyond the commissioning dates, the feed-in tariffs (FITs) will be determined by a Council of Ministers decree, which in any case will not be higher than the FIT for the first period. • The facilities that prefer using the FIT cannot sell energy to the market for the current year. • The total support amount is distributed among the suppliers who sell energy to consumers directly. • For solar and wind license applications, site measurements are required. • Solar and wind license applications can be submitted only on the dates determined by the EMRA Board. ­ The promotion of RE sources in the electricity market was assigned to the EMRA by the EML. Specifically, the Electricity Market Licensing Regulation (LR) assigned the EMRA with the responsibility to encourage the utilization of renewable and domestic energy resources, and to initiate actions with relevant agencies for the provision and implementation of incentives in this field. In the LR, generation facilities based on RE resources are defined as those power plants that utilize wind, solar, geothermal resources, waves, tidal movements, biomass, biogas, and hydrogen; river or canal-type hydroelectric generation facilities; and hydroelectric generation facilities with a reservoir area smaller than 15 km2 or with pumped-storage hydropower plants. The EML and LR specified that in case of more than one application for the same region and/or the same transmission substation (in case substation connection capacity is limited), the licensed entity shall be qualified through an ­ auctioning process executed by the TEIAS with respect to the maximum contri- bution fee per kilowatt-hour. To promote generation from renewable sources, electricity generation from power plants less than 500 kilowatts (kW) based on renewable sources is exempted from license obligation and, unlike other market activities, the owners of such plants do not have to establish a company. The WPP projects licensed by the EMRA between September 3, 2002 (when the market was opened), and June 4, 2004 (when the license applications for wind power plants were sus- pended), were mainly the old BOT projects that had been developed earlier. Some of those projects’ owners resigned their existing contracts and became license holders. The connection capacity of these projects had already been allo- cated by the TEIAS; therefore, there was no problem getting connections. In addition to these old BOT projects, there were several license applications submitted to the EMRA for new wind projects. There were no ­ predetermined wind project sites marked by the public authorities and no published informa- tion about the regional or substation-based transmission system connection capacities. Therefore, companies were determining project sites according to The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 240 Case Study: Turkey their own evaluations and were proposing connection points. But those applications could not be concluded by TEIAS with respect to connection and ­ use of the system. Also, there were criticisms regarding possible problems due to the intermittent behavior of wind energy and its possible effects on system operation, and the limited connection capacity. Furthermore, the regulations did not make provisions for how to select one among several applicants for the same project site. Accordingly, on June 4, 2004, the EMRA board decided to suspend all license applications for WPPs and to stop review, evaluation, and granting processes for six months until the TEIAS issued the maximum annual WPP capacity to be connected to the grid. But the TEIAS was unable to issue these projections, and the EMRA upheld the suspension for more than three years. Due to public pressure, however, the EMRA decided to reopen the applications on November 1, 2007; 751 appli- cations were received by the EMRA for a corresponding 78,000 MW capac- ity. Most of the applications were made for the same ­ project sites. But as expected the EMRA was not able to take a decision on the applications without inputs from the TEIAS. This lack of necessary information on the evaluation and selection of the applications was followed by another pro- longed period of inactivity. To address this problem the EML was amended on July 9, 2008, to introduce an auctioning process for cases in which more than party applies for the same plant site, or total requested capacity exceeds substation capacity. Meanwhile, new regulations were issued about a pre-elimination of the projects by the Electrical Power Resources Survey and Development Administration (EIE) and the auctioning process by the TEIAS. The TEIAS also decided, in February 2010, that the total capacity to be connected to the grid would be 8,450 MW compris- ing a total of 142 substations. Accordingly, the EMRA informed the license applicants about the TEIAS’s decision and asked for the installed capacities to be revised downwards. The applicants who did not reduce their original installed capacity figures within 10 days’ time were disqualified by the EMRA without further notice. The remaining applications were reviewed and evaluated in technical terms by the EIE, and the applied capacity of nearly 78,000 MW was finally reduced to 31,268 MW. Of this capacity, 1,378 MW were single applications, and the owners were granted licenses, while the remaining (having more than one appli- cant) were subjected to the auctioning process by the TEIAS. The TEIAS auctioning process (according to the maximum contribution fee) was started in 2010 for 13 different groups of applicants and concluded in July 2011. A total of 149 projects were qualified through those auctions with a total installed capacity of about 5,500 MW. The weighted average of the contribution fees per kilowatt-hour was realized as TL 1.91, and the highest fees of TL 6.52, TL 5.60, and TL 5.25 were offered to the Antakya, Can-Canakkale (Dardanel), and Izmir substations, respectively. The results of the auctions according to the classification of contribution fees are given in figure 10.3. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 241 Figure 10.3 The First TEIAS Wind Capacity Auction: Capacity Allocations Percent 5≥ krs/kWh; 481 MW; 9 4–5 krs/kWh; 728 MW; 13 0–0.5 krs/kWh; 2,215.7 MW; 40 3–4 krs/kWh; 816.6 MW; 15 2–3 krs/kWh; 225.5 MW; 4 0.5–1 krs/kWh; 427.8 MW; 8 1–2 krs/kWh; 605.5 MW; 11 Source: TEIAS. Note: krs= kurus (1/100 of Turkish Lira); kWh = kilowatt-hour; MW = megawatt. This period lasted more than three years, and the installation of wind power plants also took a considerable amount of time. The allocated capacity and ­ general status of wind projects as of end March 2013 are given in table 10.1. At the moment, no new license application is being accepted by the EMRA. The EMRA will be issuing a new date for applications, and these will be pro- cessed according to the procedures and rules discussed in subsequent sections. The development of installed wind capacity is given in figure 10.4, which shows that the development of wind energy gained pace after the electricity reform and the REL was enacted in 2005. It is interesting to note that, although the auctioning process was concluded in 2011, nearly 50 percent of the eligible projects have either not been licensed yet or, even if licensed, the project companies have not signed connection agree- ments with the TEIAS. The main reason for this slow realization appears to be the high and unrealistic bid prices during the auctioning. Considering the FIT level or the market prices, it is very difficult to find financing for projects whose auction prices are as high as 3−5 cents/kWh. One can expect that high bids for the contribution fee indicates the operator’s or project’s efficiency—that is, bid prices would normally be based on the ­ bidder’s feasibility studies. More efficient projects can achieve greater revenues, and hence these will be bid for at higher prices. In the previous tenders, though the bid prices for some projects were as high as 4−5 cents/kWh, neither the ­ support price of 7.3 cents or the market price of 9−10 cents would suffice to make these projects The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 242 Case Study: Turkey Table 10.1 The Results of the First TEIAS Wind Capacity Auction By type MW Before November 2007 3,761 For single applications 2,027 For coinciding applications 5,371 For R&D 6 Capacity increases in 2013 752 Total allocated capacity 11,917 In operation 2,337 Positive connection opiniona 2,371 Without connection agreementb 4,737 With connection agreement 2,472 Total allocated capacity 11,917 Source: Dilli 2013. Note: MW = megawatt; R&D = research and development. a. To be licensed. b. Connection agreement with the TEIAS is pending. Figure 10.4  Development of Installed Wind Capacity, 2000–12 2,500 2,000 1,500 MW 1,000 500 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Source: TEIAS. Note: MW = megawatt. feasible. If bidders are careful, then the bidding price will indicate the merits of the project. But unfortunately, past experience shows that it is not always so. Delays in project realization are mainly due to project trading in the Turkish mar- ket, and the creation of a secondhand market where projects are bought and sold. There were about 37,000 MW project capacities in the Ministry of Energy and Natural Resource (MENR) at different stages of project development when the new regime was started in 2001 under the EML. It soon became that, under current economic conditions, even 10 percent of this capacity could not be The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 243 ­ealized with take-or-pay provisions and treasury guarantees. As such, many r artificial project capacities were disqualified in time. The main path followed by ­ affected project developers was to apply to the MENR for the same project when a new project application was announced, or to develop a project similar to that planned by related public institutions. The new EML introduces a new challenge for those project developers who have a provisional license. In the provisional license period, license holders are not allowed to sell the companies. It was thought that this would be useful for stopping the project trade, or that at least only those projects that fulfilled their provisional license obligations could be transferred to other parties. RE generators have two distinct ways to sell their electricity2: • Through bilateral contracts or in the day-ahead market. In this case the compa- nies can benefit from incentives except the FIT, and the price will depend on the wholesale market price or bilateral contract price. They are treated as any other generator and would bear the risk of generation imbalances, which might be quite high for wind generation. Since the wholesale market price level is higher than the FIT, most of the companies prefer this and try to hedge their risks by establishing generation portfolios with some thermal and hydro generation. With the introduction of an intraday market (currently in a trial ­ phase), the imbalance risk could be better managed. • In the “renewable pool.” Previously, retail sellers were required to purchase an amount of energy equal to a certain percentage of the electricity that they had sold in the previous year from entities holding an RER certificate. Accordingly, they were required to sign bilateral agreements with the RER certificate holders. But the REL set forth a new method for the performance of suppliers, as opposed to that of retail sellers only. According to the REL, instead of executing separate bilateral agreements for each sale transaction between a supplier and an RER certificate holder, the said obligation was to be performed through a program in which all suppliers were obliged to share the cost of energy generated by all willing RER certificate holders in the renewable pool. The regulation for RE resources was issued through a provi- sion of the REL—the RER Support Mechanism (YEKDEM). According to the provision, RER certificate holders are entitled to participate in the YEKDEM on a yearly basis for the first 10 years of operation, provided their power plants are commissioned on or before December 31, 2015. Once they participate, the RER certificate holders cannot terminate their participation during that year, and the option to participate is available only at the beginning of each calendar year. Those who do not wish to participate in the YEKDEM may sell electricity in the free market and may sign bilateral ­ agreements. In such cases, however, they would not be entitled to benefit from the purchase and price guarantee incentives of the REL. The wind capacities of producers that participated in the YEKDEM in 2011, 2012, and 2013 were 469 MW, 688 MW, and 76 MW, respectively. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 244 Case Study: Turkey The licensing process is represented in figure 10.5. To receive an opinion on the connections, the EMRA sends the applications to the TEIAS. If there is no capacity constraint and the connection proposal of the company is deemed appropriate by the TEIAS, the TEIAS shall approve the project. In case the capacity of the application exceeds the substation capacity, the TEIAS notifies the EMRA, and the EMRA sends a notification to the applicant to decrease a portion of the installed capacity. If there is more than one applicant for the same connection capacity or for the same connection region, the TEIAS will organize an auction to determine the qualified applicant(s) to be connected. In the auctioning process, the bidders who offer the highest price per megawatt (that is, ­ contribution fee) will become eligible to connect to the grid until the available capacity is reached. The offered amount will be paid in the first three years of operation. Figure 10.5 Licensing Process • Required documents are published on EMRA’s webpage • All applications must include an initial collateral of “10.000 TRY x Application capacity MW ” • Project assessment and auction (based on fee/kWh in case of multiple applications) for hydro by DSi before license application When the application phase is completed, the projects is assessed for approval: • Conditions for grid connection are determined by DSO and/or TSO and approved by EMRA Project • Technical assessment of the project is done by YEGM for wind assessment and solar • If there are multiple applications for the same site and/or connection point (exceeding total available capacity) then TSO makes auctions (based on fee/kWh) and winners are licensed If the assessment phase is finalized positively, then the project is approved and some obligations are determined for licensing: • Environmental Impact Assessment (EIA) • Contract with TSO or DSi if a tender was made Project • Increasing the limit of the initial collateral (details on EMRA approval website) • Required amendments in the main status of the company • Capital increase for the company (to 20% of the total investment for all projects that have been approved) (Obligations must be fulfilled within 90 days. If a comprehensive EIA is required then 300 days apply for EIA only) Licensing If the obligations are fulfilled then the project is licensed Source: EMRA. Note: DSI = Directorate of State Hydraulic Works; EIA = Environmental Impact Assessment; EMRA = Energy Market Regulatory Authority; TSO = transmission system operator; TRY = Turkish Lira; YEGM = Yenilenebilir Enerji Genel Müdürlüğü (General Directorate of Renewable Energy). The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 245 An Environmental Impact Assessment (EIA) is required before the granting of the license. In this respect, projects with installed capacities of more than or equal to 75 MW, are obliged to obtain a positive EIA. Projects having an installed capacity of more than or equal to 10 MW are to be checked to decide whether an EIA study is required for them or not. Before the granting of the license the applicants should submit the positive EIA or the decision stating that an EIA is not necessary to the EMRA (although there is no obligation for a positive EIA decision for projects less than 75 MW, investors generally prefer to have it due to lenders’ requirements). Unlike for hydro projects, wind power projects tend to receive public approval. Although some projects have experienced minor problems (particularly for transmission line construction), no major problems have been reported so far. ­ The new EML limits the provisional license period to 24 months. The legal entity fulfilling the requirements indicated in the LR will be granted a provisional license by an EMRA Board decision. After the granting of the license, the ­ following obligations should be met by the licensee during the provisional license period: establishing of land-usage rights; signing of the Connection and Use of System Agreement; and acquiring of public works permits (from local authori- ties), road permits, and technical interaction permits (from the Turkish General Staff and National Intelligence Agency). The critical provision in the new EML is that share transfers during the provisional license period are prohibited. After the above-listed procedures and obligations are completed, the provi- sional license is granted to the owner. The project company then has to submit project design documents to the MENR. After approval of these technical docu- ments, construction work can start. On completion of construction, the project company applies to the MENR for a Project Compliance Approval. A team formed by the MENR will evaluate the plant site to see if all necessary controls are in place, check whether all permits are obtained, and whether the project is in line with the project technical documentation. After commissioning tests by the team, the plant is officially commissioned and registered. The EMRA issued a Communiqué on Wind and Solar Measurements on October 11, 2002, to open licensing applications for wind projects. Although the requirements are not detailed, for the licensing application to be accepted by the EMRA, measurement results for one year should accompany the application. The communiqué alone could not increase the number of wind projects’ licensing applications without a long-term support scheme for these projects. That support scheme was introduced with the enactment of the REL. Meanwhile, the EIE issued the REPA showing the country’s wind potential at different heights, and investors are able to purchase detailed data of specific locations. Following these developments, and also due to pressure from investors, the EMRA removed the measurement requirement by cancelling the communiqué on January 19, 2006. But after a chaotic experience with wind license applications in November 2007, the EMRA had to issue a Communiqué on Measurement Standards for Wind and Solar Energy License Applications on February 2012. Following this, the Directorate of Turkish State Meteorological Service (SMS) also issued the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 246 Case Study: Turkey Communiqué on Implementation of Wind and Solar Measurements to be performed for Wind and Solar Energy License Applications on July 10, 2012, to ­ regulate principles and procedures of measurement and evaluation of data. According to the new EML for unlicensed wind projects of up to 1 MW capacity, measurement is not required. The first version of the REL provided the same FIT for every source of renew- able electricity generation. After the amendments, the FITs were differentiated according to sources (wind, hydro, solar, and so on), as is reported in table 10.2. In addition to the basic FIT, a bonus is provided in case the electromechanical equipment is manufactured in Turkey (a local content premium), as reported in table 10.3. Other incentives include: • Exemption from the compulsory 1 percent turnover payment for operating businesses on immovable assets of the treasury. • A 99 percent exemption from the licensing fee and annual license fees for the first eight years of operation. • Priority in system connection. • Value added tax (VAT) exemption for domestic equipment for Investment Support Certificate holders. • VAT, customs tax, and Resource Support Utilization Fund payment exemp- tions on imports for Investment Support Certificate holders. • Research and development (R&D) deduction (that is, deduction of R&D expenditures from the corporate tax base at a rate of 100 percent). • Income tax exemption (of 80 percent of salary income for eligible R&D and support personnel). • Social security premium support for five years. • Stamp tax exemption. All RE generators can benefit from these incentives whether they are ­ participating in the YEKDEM or not. Table 10.2  Feed-In Tariff Plant typea Schedule Ib Schedule II c 1.Hydro 7.3 2.3 2. Wind 7.3 3.7 3. Geothermal 10.5 5.8 4. Biomass 13.3 2.7 5. Solar (PV) 13.3 6.7 6. Solar (CSP) 13.3 9.2 Source: EMRA. Note: CSP = concentrated solar power; PV = photovoltaic. a. Before Law No: 6094, feed-in tariff was € cents 5−5.5 per kilowatt-hour for all of the renewables (Law No: 5346). b. 10 years for plants to be commissioned until December 31, 2015. c. Incentive for local content—5 years for plants to be commissioned until December 31, 2015. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 247 Table 10.3 Proposed Premium for Use of Equipment Manufactured Locally € cents/kWh Hydropower 1. Turbine 1.0 2. Generator and power electronics 0.8 Generation facility based on wind power 1. Blade 0.6 2. Generator and power electronics 0.8 3. Turbine tower 0.5 4. All of the mechanical equipment in the rotor and nacelle groups 1.0 (excluding payments made for the blade group and generator and power electronics) Photovoltaic (PV) solar power 1. PV panel integration and solar structural mechanics manufacturing 0.6 2. PV modules 1.0 3. Cells forming the PV module 3.0 4. Inverter 0.5 5. Material focusing solar rays on the PV module 0.4 Concentrated solar power (CSP) 1. Radiation collection tube 2.0 2. Reflective surface plate 0.5 3. Solar tracking system 0.5 4. Mechanical equipment of heat energy collection system 1.0 5. Mechanical equipment of the system collecting solar rays on the 2.0 tower and producing steam 6. Stirling engine 1.0 7. Panel integration and solar structural mechanics manufacturing 0.5 Biomass 1. Fluidized bed steam boiler 0.6 2. Fluid or gas-fired steam boiler 0.3 3. Gasification and gas cleaning group 0.5 4. Steam or gas turbine 1.5 5. Internal combustion engine or Stirling engine 0.7 6. Generator and power electronics 0.4 7. Cogeneration system 0.3 Geothermal 1. Steam or gas turbine 1.0 2. Generator and power electronics 0.5 3. Steam injector or vacuum compressor 0.5 Source: EMRA. Note: kWh = kilowatt-hour. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 248 Case Study: Turkey Green or Concessional Funds In 2004 the government disbursed a $202.03 million loan from the International Bank for Reconstruction and Development (IBRD) and the World Bank, and opened a special purpose debt facility (SPDF) to finance privately owned RE generation facilities. The SPDF—a term-lending facility—was operated by two financial intermediaries: the Turkish Industrial Development Bank (TSKB, privately owned) and the Turkish Development Bank (TKB, government owned). ­ The SPDF was designed to leverage equity investment from local private RE developers, export credit financing, or other means for the construction and operation of qualified RE projects. The SPDF operated in the period 2004–08. Under the special loan structure, the financial intermediaries required a mini- mum equity of 25 percent, and were able to offer maximum maturities of 12 years including a 4-year grace period. In 2009 the government disbursed a new $500 million IBRD loan, complemented by a $100 Million Clean Technology Fund (CTF) concessional loan to replenish the facility. The design of the SPDF required modifications during the implementation period, including the following (as described in the World Bank’s Implementation Completion Report of March 2010): (a) the capacity limit on hydro power plants was increased from 50 MW to 100 MW, (b) the maximum loan size allowed for each subproject was increased from $20 million to $40 million, and (c) the international competitive bidding (ICB) threshold for civil works was raised from $8 million to $15 million, and a maximum of $15 million was allowed to finance civil works carried out by a sponsor-related construction firm (that is, in the original procurement guidelines, the financing of construction by a firm or subsidiary affiliated with the RE project’s sponsor was not allowed). With the ratification of the Kyoto Protocol, Turkish power plant operators will now have the right to engage in the trade of various emissions-related financial products after 2012. Before this, Turkish activities were limited to the voluntary carbon markets. Conclusions The experience of Turkey offers remarkable lessons. The first notable observation is that around only 5 percent of the total technically feasible potential has been utilized in the past few decades. To reach the 20,000 MW target in 2023, roughly 1,700 MW should be commissioned each year till 2023 (excluding the period 2005–08, the existing added capacity was roughly 500 MW/year). As an approxi- mation, excluding the required transmission investments, a total of $28 billion is needed in the upcoming 10 years for the realization of the wind-based capacity target. These disappointing results can be mostly attributed to the lack of consen- sus between different stakeholders. The Turkish government renewed its com- mitment to RE in 2002, but the selected support scheme was only enacted in 2005 and did not meet expectations. The REL and other support mecha- nisms, the electricity market reform, the new trading mechanisms (which were The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Case Study: Turkey 249 implemented in this reform process), and the high electricity prices in the market increase the attractiveness of RE generation. But market prices alone were not sufficient for RE investment decisions and financing. A support mechanism is required to provide long-term certainty and decrease the risk of investment by providing a certain revenue stream for the project. Provided that relevant risks (including the imbalance risks) can be hedged, market reform and competition can also play a role in wind development (new trading mecha- nisms such as the day-ahead and balancing market provide risk-hedging pos- sibilities, and higher market prices for electricity increase its attractiveness). Implementation of the intraday market (which is currently in the trial phase) also might decrease the imbalance risk of wind power plants. Although the FIT level set by the last amendment of the REL was not found attractive, it is unlikely that this level will be increased. As mentioned before, projects with a high contribution fee will probably be cancelled and those capaci- ties will be reissued to the market. Projects are mostly being financed by Export Credit Agency (ECA) credits, international financial institution (IFI) lending (such as from the World Bank and the European Bank for Reconstruction and Development [EBRD] through local banks), and some contribution via voluntary carbon-trading mechanisms. Still, the most important bottleneck is financing. The TEIAS is working to strengthen regional connection capacities for new projects. The intention is to determine new capacity to be utilized each year, starting from 2014. The unused capacities allocated by previous auctions are being determined and included in a new capacity list. The chaotic experiences of the past have provided valuable lessons for the administration as well as investors. Companies are now much more careful when selecting projects. Under current regulations progress will be slower, but will encompass the realization of feasible projects by more sophisticated investors. Significant problems reported by investors relate to the power of incentive mechanisms to cover risks (such as the currency risk, the uncertainties of local equipment support, the lack of purchasing guarantees, and the imbalance risk in the free market), legal uncertainties (related to the reactive power control obliga- tions and contribution to grid investments), inexperienced investors (improper wind measurements and evaluation, improper project planning, lack of turn-key contracts, incorrect turbine selection, misleading financial analyses), deficiencies in infrastructure (power limitations, financial and administrative weaknesses of the transmission company, low reimbursement value for grid investments ­ realized by licensees, lack of long-term grid investment planning), licensing issues (high contribution fees, lack of an annual license application program, weak coordina- tion among public institutions and lengthy permit procedures, priority of mining fields in licensed sites), and financial issues (lack of an insurance mechanism for local equipment, small number of local banks that can extend IFI loans). More specific lessons are reported below: • The private sector and the government authorities should be well informed about the challenges of wind power projects. A sustainable and an applicable support scheme The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 250 Case Study: Turkey should be defined at the outset of the process. It will be useful if the related public organization performs measurements and provides measurement results to the market for constructing wind power plants in suitable sites. The grid company should make necessary studies for calculation of required grid capaci- ties for connections. The specifications of these sites, together with the available connection capacities, should be announced. • In the case of more than one application, an auctioning process should be performed to select the successful applicant; but tendering should be done among equals. The limited connection capacity should not be allocated to an inefficient or unfeasible project owner just because it proposes the highest contribution fee. Implementation of a preselection method, which is based on technical merits and financial capability, would be useful. • To integrate wind power to the system without causing system reliability problems, the system operator should be equipped with wind-forecasting tools and control mechanisms. Projects should be developed according to international technical and financial requirements. The capacity of the transmission system operator, the TEIAS, to integrate increasing volumes of wind and other intermittent renewable sources more effectively into the Turkish power system needs to be built up. More transmission investments and control/dispatch tools (supervi- sory control and data acquisition, or SCADA) are required for reliable system operation. Notes 1. BOT, BO, and TOOR refer, respectively, to build, operate, and transfer; build and oper- ate; and transfer of operating rights. 2. For details see Dilli (2013), “Wind Based Energy Development in Turkey,” from which this chapter draws. Bibliography Dilli, B. 2013. “Wind Based Energy Development in Turkey.” Mimeo, World Bank, Washington, DC. EMRA (Energy Market Regulatory Authority). Annual Report. Several years, Istanbul, Turkey. TEIAS (Turkish Electricity Transmission Company). Annual Report. Several years, Istanbul, Turkey. Vagliasindi, M., and J. Besant-Jones. 2013. Revisiting Standard Policy Recommendations on Market Structure in the Power Sector. Washington, DC: World Bank, Directions in Development. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Chapter 11 Summary and Conclusions For policy makers seeking to design renewable energy (RE) support mechanisms suited to the needs of developing countries, the main lessons are clear and inescapable. Successful RE policies: ­ • Will only be effective once the state-owned utilities that are the buyers of grid- connected RE are themselves in good financial health (in all of the case study countries, the power utilities were under financial duress). • Need to be grounded in economic analysis and the application of market prin- ciples to ensure economic efficiency. • Require a sustainable, equitable, and transparent recovery of incremental costs. The Financial Health of Power Utilities The first and arguably most important lesson relevant to sustainable incentives for RE is that the power utilities involved need to be in good financial health. In most of the countries represented in this study, the utilities are in poor financial health, resulting in cash-flow problems. Such problems affect the rural distribu- tion companies that are the typical buyers of RE. This is the case in Indonesia, where the consumer tariff is just 50 percent of PT Perusahaan Listrik Negara’s (PLN’s) cost, in Sri Lanka because of the unfortunate historical dependence on oil for power generation, and in Vietnam due to the reluctance of the govern- ment to raise consumer tariffs. The effects of the power utilities’ financial status are most clearly illustrated in the case of Sri Lanka: with coal displacing oil, the Ceylon Electricity Board’s (CEB’s) revenue requirement per kilowatt-hour has already begun to decline, offering the opportunity for consumer tariff reductions. But the CEB is still not in good financial health, and consequently still opposes having to absorb the incremental costs of RE. Until such time as a nation’s utilities are in good financial health, and operate under a transparent regulatory system that sets electricity tariffs on a sustainable The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   251   252 Summary and Conclusions basis—and that allows for the incremental costs of RE to be passed to the consumer—utilities will continue to oppose what they see as unnecessary costs ­ that worsen their already poor financial situation. The idea that the incremental costs for RE can be recovered on a sustainable basis when utilities are in financial distress is unrealistic. Power purchase agreements (PPAs) with RE producers require payment in cash within 30 days, which means that cash-flow management is the first priority of utilities that may have a significant number of RE generators. Such utilities are often concentrated in relatively small geographic areas—so it is the rural distribu- tion companies, not those in large urban areas, that feel the most pain. This is well illustrated in the case of Vietnam, where the additional costs of building up the 115 kilovolt (kV) transmission network for power evacuation to the main 220/500kV grid fell on the distribution companies, and where assurances that the tariff methodology would eventually reimburse them for these incremental costs were greeted with great skepticism by entities faced with short-term ­cash-flow problems. Widespread Consumer and Political Support In countries in the Organisation for Economic Co-operation and Development (OECD), governments have been able to impose RE policies not only because of widespread consumer and political support (as in Germany), but because power companies are mainly in private hands and in good financial health (indeed, critics claim they are in too good a financial condition, at the expense of consumers), and tariff regulation is most often decided by independent regulators, reasonably free of direct government interference. If it in fact so ­ chooses, an OECD government may make the case that a greater share of RE citizens’ long-term interests, and then adapt to the challenges it is bound is in ­ to face (see box 11.1). Box 11.1 Lessons from Germany: Coping with Higher Shares of Renewables An increasing share of RE generation does not react to market price signals. How then can we ensure efficiency of the market for plant dispatch? The benefits of a feed-in premium (see figure B11.1.1 for the case of Germany) in addition to feed-in tariffs (FITs) are that: (a) price signals reach RE generators, so they have incentives to adjust to market prices; (b) it helps in efficient market integration; (c) incentives improve diagnosis and balancing; (d) it makes available more players for developing innovative solutions for pooling or demand-side management; and (e) it opens new markets for RE (balancing). Possible drawbacks of a feed-in premium vis-à-vis a FIT are that: (a) wind and photovoltaic (PV) have limited abilities to react to market signals and (b) higher risks also imply higher costs. box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 253 Box 11.1  Lessons from Germany: Coping with Higher Shares of Renewables (continued) Figure B11.1.1 The German Feed-In Premium Income = Market revenues + support premium Support premium Technology (calculated weighted Feed-in Technology specific monthly at the benchmark for tariff management premium end of the market month) revenues Individual tariff of each Monthly average market Covers the cost of plant value for the specific balancing the renewable technology portfolio The costs of a FIT can be lowered by: • Introducing an annual degression right from the beginning. • Installing a cap when growth becomes too fast too soon. • Keeping tariff adjustment away from long parliamentary processes. • Enhancing flexibility of the system and market integration of RE, by enhancing grid expansion, demand-side management, and storage (see figure B11.1.2). Figure B11.1.2 Enhancing System Flexibility in Germany 20% EE 35% EE 50% EE 65% EE 80% EE Grid Grids Netzausbau fũr großrãumigen Stromaustausch (Baustein 1) expansion Flexible thermal Demand side Generation Flexible thermische Kraftwerke (2) power plants; Reduktion “must-run”(4) feed-in Einspeisemanagement Wind & PV (6) management Flexible Nachfrage durch Lastmanagement (3) Demand side/ Load management Power-to-heat statt Einspeisemanagement (reduziert “must-run” ) (5) Power to heat Storage Pump storage Pumpspeicher D/Alpen/Norwegen (5) Power to gas Power-to-gas (5) box continues next page The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 254 Summary and Conclusions Box 11.1  Lessons from Germany: Coping with Higher Shares of Renewables (continued) In the case of Germany an automatic degression was linked to the newly installed capacity of PV without a cost-assessment study and without undergoing political pressure. The key elements of the system are illustrated in figure B11.1.3. • The basic annual degression: 11.4 percent until 3.5 gigawatts (GW) was newly installed + 4 percent automatic degression for each gigawatt installed on top of the 3.5 GW. • Degressions come into effect monthly; to avoid seasonal sales they are based on growth in the last 12 months. • There is an overall cap of 52 GW solar PV. • Expiration of the Renewable Energy Sources Act (Erneuerbare-Energien-Gesetz, EEG) PV support, but continuation of priority feed-in. Figure B11.1.3 Another View of System Flexibility in Germany Per year: Per month: 29% above 7.500 MW 2.8% 26% above 6.500 MW 2.5% 23% above 5.500 MW 2.2% 19% 1.8% above 4.500 MW +1.000 MW 15% 1.4% above 3.500 MW 2.500–3.500 MW Target corridor 1% 11.4% –500 MW 9% up to 2.500 MW 1.75% 6% up to 2.000 MW 1.5% 0% up to 1.500 MW 0% –6% up to 1.000 MW –0.5% Seite 30 Source: Lauber 2013. Note: MW = megawatt. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 255 By contrast, developing-country governments have a fundamental conflict of interest: they not only set RE policy but they themselves own their (often technically insolvent) power companies—which require extensive government subsidies. In the battle between advocates of RE (typically in ministries of environment, or of energy) on the one hand, and the Ministry of Finance (MoF)—which foots the bill for the electricity subsidy—on the other, it is generally the MoF that wins, on grounds that subsidies are already unsustain- ­ able. According to this view, such subsidies should not be made worse by any additions attributable to RE—this lacks a clear economic rationale. In Vietnam the prime minister’s office did not support the concept of a renewable fund dependent on even a low consumer levy. In Sri Lanka the National Energy Policy expressly states that the achievement of RE targets shall have no impact on consumer tariffs. And in Indonesia the MoF points out that the new geo- thermal feed-in tariff (FIT) violates legislation that geothermal prices must be subject to competition, and that the massive subsidies presently paid to ­ support low consumer tariffs are already unsustainable and cannot carry yet further increases. Setting Renewable Energy Targets To set an economically rational target for a given year, decision makers need to estimate (a) the RE supply curve for that year, and (b) the expected avoided social cost of thermal generation. Where these two variables intersect defines the target. Such an exercise is subject to several uncertainties: the cost of RE technologies may change, the world price of fossil energy may change, and new estimates of environmental damage costs may become available. But as shown in the case of the Croatia RE study (Appendix A), a simple analytical framework can deal with these uncertainties (by comparing the losses of setting targets too low against the losses of setting them too high). Although these economic principles for setting RE targets have been advo- cated by numerous World Bank studies,1 few countries have in fact set targets on this basis. Among our case study countries, most RE targets are simply political statements: Sri Lanka’s target of 10 percent of total energy by 2015 was unsup- ported by any economic analysis (table 11.1). In Vietnam the targets for RE electricity generation in power development plans have no analytical basis: the plans for additional RE installed capacity are, quite simply, arbitrary, and the incremental costs were not even costed. The only exception is Brazil (where there are no official targets but indicative benchmarks derived by a 10-year expansion plan). The lack of intellectual rigor in setting RE targets lies at the heart of the slow uptake of RE generation in most of the case study countries. Targets that bear no relationship to the economic realities of the incremental costs of RE are rarely achieved; even worse are those targets (and associated support tariffs) issued in the complete absence of knowledge about the magnitude of incremental costs The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 256 Summary and Conclusions Table 11.1 Renewable Energy Targets in the Case Study Countries Country Target Rationale Vietnam 1,600 MW by 2020, set in the 7th Power Development Unsupported by any detailed economic Plan. analysis. Indonesia 9,500 MW geothermal by 2025; 3,967 MW by 2014. Unsupported by economic analysis, or understanding of the incremental costs. Mainly a reflection of wishing thinking. Sri Lanka 10% renewable energy by 2015. Political statement, unsupported by economic analysis or incremental costs. South Africa 10,000 GWh. Tanzania n.a. Egypt, Arab Rep. 20% of the electricity generated by renewable energy by 2020, with a specific target of 12% coming from wind. This target translates to a total installed capacity of 7,200 MW of grid-connected wind energy in 2020. Brazil Wind, small hydro, and biomass are expected to reach Supported by a generation expansion 27 GW by 2020. plan. Turkey 20,000 MW in 2023 (of which 600 MW geothermal) with 30% share of electricity from renewable energy. Note: GWh = gigawatt-hour; MW = megawatt. n.a. = not applicable. implied (the most notable recent example of which is the Indonesian geothermal tariff). The rationale for targets for biomass-based, grid-connected power generation needs to be substantiated. From the perspective of reducing greenhouse gas (GHG) emissions, it does not matter whether rice husk is burnt for grid-­ connected generation, or whether it provides process heat at rice mills (as is the case in much of Vietnam): it is only important that it is not burnt or left to rot in fields (again, as in Vietnam) or discarded into waterways. The Difficulty of Predicting Unexpected Consequences The inability to anticipate unexpected consequences is a major problem. Consider the example of biomass in Vietnam: A major technical assistance program sup- ported by a bilateral donor seeks to promote biomass use (and rice husk in particular) for power generation—for which no precise rationale has ever been ­ provided except the presumably self-evident presumption that biomass power generation must surely be desirable. Presently rice husk is used as a fuel in rice mills, in ceramics kilns, and even brick making—where it replaces oil. Rice husk used for power generation will displace the most expensive fossil fuel, which is combined-cycle gas turbine (CCGT) (at projects where gas prices are linked to international fuel oil prices). So rice husk used for power generation will not reduce GHG emissions, it will increase GHG emissions. In any event, the marketplace and technology innovations are overtaking the attempts of governments to intervene. Developments in pelletizing technology The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 257 have now led to an emerging export trade in rice husk pellets to Japan and the Republic of Korea—with the result that rice husk prices have already increased to as much as $30/ton, making rice husk power generation in Vietnam even more expensive. When this was pointed out at a recent consultation workshop, the reply was that it was a good thing for gas to be conserved, since it is in “short supply.” It is true that much domestic gas is sold to Electricity of Vietnam (EVN) at not much more than its production cost (around $3.5 per million British thermal units [mmBTU]), and that there is concern about future gas shortages if no new fields are found soon. But the way to fix that problem is not to provide yet another subsidy to biomass electricity producers, but to reform gas pricing (in this case by adjusting the gas price for an appropriate depletion premium). To correct one subsidy by advocating another is poor economic policy. Risk and Reward The fundamental problem with administered pricing for RE—and production- cost-based FITs in particular—is the lack of recognition given to the relationship between risk and reward. All projects of a given RE technology are assumed to be of equal risk, and subject to a single estimate of what is a “fair” or “reasonable” return on equity. But as seen most clearly in the case of the Indonesian geother- mal tariff, the risk at the tendering stage is very unevenly distributed across ­ various areas according to the amount of information about the field being bid— with the result that the final electricity price may have to be renegotiated once the resource has been more precisely identified. And herein lies the benefit of an avoided cost tariff (ACT): the marketplace will determine which projects are economic and which are not at the issued tariff, without governments having to guess what rate of return is necessary; the same is true of auctions (whether by price, or by amount of subsidy required for access to a fixed price). Institutional Barriers While policies to provide stable price support are obviously necessary to induce private investment in renewables, other institutional and regulatory barriers may be just as critical. The importance of standardized power purchase agreements (SPPAs) is no better illustrated than in the contrast between Vietnam and the Lao People’s Democratic Republic. In Vietnam, largely as a result of a published tariff and a standardized power purchase agreement (SPPA), there is 750 MW of small hydro in place or under construction. In Lao PDR, where proposals to introduce a published tariff and an SPPA have been rejected by vested interests (there is a lively trade in memorandums of understanding, MOUs), the existing small hydro capacity is not much more than 30 MW. The introduction of the Indonesian geothermal FIT is another example of a false premise, and lack of attention to institutional barriers. It was assumed by the The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 258 Summary and Conclusions government that an inadequate price was the main barrier to achieving ­ geothermal targets. But in reality, the main barriers to bringing geothermal projects to financial closure are interminable delays in environmental permitting, the lack of an SPPA and contractual documents (though there is some hope that this is now resolved), and uncertainty about guarantees. Moreover, since the announced FIT was not clearly interpreted its introduction made the problem worse, not better. The Avoided Cost of Carbon The avoided cost of carbon is a useful indicator for the development of low- carbon emission options. Table 11.2 shows a comparison of such values, taken from the case studies in this report and other recent World Bank reports. There are several reasons for the variation in estimates: • As noted in chapter 2, such calculations depend on the dominant fuel in the least-cost case. For example, where concentrated solar power (CSP) competes against CCGTs, as in the Arab Republic of Egypt (which has low emissions per net kilowatt-hour), the avoided cost is much higher than where it competes against coal (whose emissions per net kilowatt-hour are three times that of gas), as in South Africa. • The quality of the RE resource matters greatly. Wind in Egypt has annual plant capacity factors in excess of 40 percent, compared to Vietnam, where the planning assumption is 27 percent—and so the avoided cost in Egypt ($24 ton) is much lower than in Vietnam ($124/ton). Table 11.2 Avoided Cost of Carbon Sri Lanka Vietnam Indonesia Egypt, Arab Rep.c South Africad Alternative CEB Least-Cost Plan — — Gas CCGT Medupi coal Hydroa 37 — — — 7e Supercritical coal 7 — — — — LNG 86 — — — 105 Wind — — — 24 124 NCREb 87 — — — — CSP — — — 300 115−55f Nuclear — — — — 67 Underground coal gasification — — — — 223 CCGT (gas oil) — — — — 275 Geothermal — — 20−30 — — Note: CCGT = combined-cycle gas turbine; CEB = Ceylon Electricity Board; CSP = concentrated solar power; LNG = liquefied natural gas; NCRE = nonconventional and renewable energy; — = not available. a. Hydro candidates are not in the least-cost plan but have been proposed in the past, and may be economic at given carbon prices. b. Sri Lanka’s composite renewables scenario to meet the 10 percent target. c. World Bank 2013. This assumes CSP replaces 80 percent gas and 20 percent heavy fuel oil. d. World Bank 2010. e. South Africa’s share (2,350 MW) of the 4,360 M Inge-III project in the Democratic Republic of Congo, including the cost of the 3,000-kilometer transmission line to South Africa. f. Depending on the amount of storage provided. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 259 There is widespread confusion about what value to use to reflect global envi- ronmental externalities. This is partly because of the uncertainty about future damage costs (and the discount rate to be used in the analysis of the global social avoided cost of carbon). The result is that every analyst uses whatever value seems appropriate. Definitions of Renewable Energy The question of what constitutes an RE technology is uncontroversial except for hydro, for which many countries establish size thresholds when stating RE tar- gets. The rationale for setting this threshold, and the value of the threshold, show large variations (table 11.3). Arbitrary thresholds between “good” renewables (small hydro) and “not good” (or even “bad”) renewables (large hydro) are not rational. There are many exam- ples of poorly executed small hydro projects with significant environmental Table 11.3 Size Thresholds for “Small Hydro” Projects MW threshold Comments Brazil (PROINFA) 1−30 Utah (United States) 1 Nepal 5 Sri Lanka 10 Set at the maximum level the CEB was willing to permit private developers, with the aim to break their earlier monopoly on power generation. Thailand (VSPP) 10 The new VSPP is based on net metering. Oregon (United States) 10 UNFCCC 15 Simplified CDM rules apply below this threshold. India 15 India defines projects up to 100 kW as “micro,” 101 kW−2 MW as “mini,” and 2−15 MW as “small.” European Union 20 Threshold for imports of project-based credits into the ETS.a Vietnam 30 Set at the level for mandatory participation in the new competitive generation market (projects less than 30 MW being exempt). Indonesia PSKSK, 1995 30 (Java Bali) 15 (Other) China 50 Thailand (SPP) 90 The threshold was set at a high level so as to include many large gas-fired (and even coal-fired) cogeneration plants. Note: Bold: Case study countries. CEB = Ceylon Electricity Board; CDM = clean development mechanism; kW = kilowatt; MW = megawatt; PROINFA = the Brazilian Program for the Promotion of Renewable Energy; PSKSK = the Indonesian avoided cost tariff; SPP = small power producer; UNFCCC = United Nations Framework Convention on Climate Change; VSPP = very small power producer. a. The European “Linking Directive” that allows for the import of project-based credits into the European Union Emission Trading Scheme (EU-ETS) establishes special conditions for hydropower projects above 20 MW that involve the construction of a dam and reservoir: “In the case of hydroelectric power production project activities with a generating capacity exceeding 20 MW, Member States shall, when approving such project activities, ensure that relevant international criteria and guidelines, including those contained in the Report of the World Commission on Dams will be respected during the development of such project activities.” The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 260 Summary and Conclusions problems attributable to road construction and poor construction practices, just as there are good large hydro projects that meet all safeguard policies and have excellent environmental and social management plans. Indeed, even the EU includes large hydro in its RE targets (for electricity generation), so there is no rational reason for exclusion of large hydro in the RE targets of developing countries. Indeed, it is quite rare for such thresholds to be set on the basis of specific environmental reasons. The 10 MW threshold in Sri Lanka was set not with any specific environmental concern, but due to the CEB’s monopoly on hydro gen- eration: 10 MW was the maximum it would allow any private sector developer entering the market. Indeed, a better definition might be power density (watts/ square meters [m2] of reservoir area), which although also controversial (as are all thresholds) at least has some explicit link to efficiency. It is by no means clear how 30 small hydro projects of 10 MW each have environmental impacts that are smaller than a single 300 MW project, especially if the latter falls under internationally recognized safeguards procedures (as in the case of the World Bank Safeguards Policies, or the “Equator” principles). The Transparency of Tariffs There are wide differences in practice, in part dictated by very different legal traditions among countries and in part by the presence or absence of an indepen- dent regulator. The range of practices can be summarized as follows, in order of increasing transparency: • Publish nothing except the tariff itself (Vietnam wind FIT, Indonesian geother- mal tariff). • Publish the methodology (Vietnam and Sri Lanka ACTs). • Publish the data assumptions (the Philippines) (table 11.4). • Publish the spreadsheet used for calculations (Sri Lankan 2009 FITs). The Philippines provides a particularly interesting example insofar as it ­ ighlights the impact of alternative assumptions, and illustrates the concep- h tual problems of government administrators having to make judgments in the face of wide information asymmetries. Of course it is true that developers will always complain that administered tariffs are too low, and governments will always worry about “windfall profits” if the tariffs are too high: which is why the best way to set RE tariffs uses the social avoided cost of thermal energy. Transparency is important; private developers and their lenders need to fore- see the evolution of the tariff in the future, and need to understand the method- ology of its derivation so that they can themselves make an assessment of future cash flows. For example, in the case of Vietnam, the government issued a regula- tion that described in detail the rationale and methodology, and how the tariff would be calculated each year. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 261 Table 11.4 Tariff Assumptions in the Philippines (for Biomass) Proposed by developers DoE estimate NREB estimate Representative size MW 8.3 8.3 8.3 Project cost $/kW 3,191 2,600 3,076 EPC cost $/kW 1,982 2,324 2,366 Net capacity factors % 72 72 72 O&M cost $1,000/unit/year 1,645 987 987 Fuel cost PhP/ton 1,297 1,464 1,297 $/ton 31.6 35.7 31.6 Fee rate kWh/ton 576 800 730 Equity IRR % 22 16 16 After-tax WACC % 12 10.2 10.2 Tariff PhP/ton 8.22 6.09 6.55 Cents/kWh 20.0 14.9 16.0 Source: Philippines NREB. Note: Exchange rate $1 = PhP 41. DoE = Department of Energy; EPC = engineering, procurement, and construction; IRR = internal rate of return; kW = kilowatt; kWh = kilowatt-hour; MW = megawatt; NREB = National Renewable Energy Board; O&M = operation and maintenance; WACC = weighted average cost of capital. We cannot conclude that transparency in setting and adjusting a support tariff will necessarily support its acceptance. But we can say that the issuance of an opaque tariff makes it very unlikely that it will be successful. The recent wind FIT in Vietnam and geothermal tariff in Indonesia were both issued without any indication of how the tariff was derived, and in the case of Indonesia, without any commentary about how (or even if) it might be adjusted in the future. Neither has been successful. In short, transparency is a necessary—though not a sufficient—precondition for a successful RE support tariff. Auctions The examples of Brazil, Turkey, and South Africa provide several important les- sons for the design of incentive systems: • Competitive bids are a viable alternative to FIT programs for RE, and poten- tially offer better price outcomes with fewer risks of excessive rents being appropriated by RE suppliers. • The core rationale for introducing FITs in developed industrialized countries was to create market certainty and simplify and lower transaction costs to stimulate production and innovation in climate-change-mitigating RE tech- nologies and markets, thus bringing down prices over time. But this rationale does not apply in many developing countries, especially in Africa, where the market for RE technologies is much smaller. Indeed, for small developing countries with low carbon footprints, the argument for greater use of more expensive RE technologies needs to be balanced against other development priorities. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 262 Summary and Conclusions • While FITs are potentially an attractive alternative to competitive biddings, transaction costs are high and many small developing countries may not have the resources or capacity to run such complex and expensive procure- ment processes. Competitive bid programs are generally simpler, although the requirements for good design and evaluation should not be underesti- mated. Development assistance programs, including those from develop- ment finance institutions, should consider carefully the costs and benefits of competitive bids versus FIT regimes. Ultimately, it will be more cost- effective to fund the higher initial transaction costs if lower power prices are likely. The above lessons apply, in the main, to auctions for RE power. Competitive bids generally incorporate a weighting of price and nonprice factors while auctions are awarded solely on the basis of lowest price (sometimes after a ­ number of rounds) among qualified bidders. Running effective auctions might ­ require even more time, expenditure, transaction costs, expertise, and capabilities than tenders. Auctions may also encourage underbidding, with the risk of subse- quent contract failures. Meanwhile, the experience of dynamic reverse auctions—such as for wind energy in Brazil—has been positive: competition has driven prices down dra- matically. But the low prices achieved in the wind auction have raised the fear that projects will not be implemented due to foreseen financial insolvency. On the other hand, if all projects were implemented, the low prices obtained in the wind auction might have paved the way to a direct competition between wind and other sources. This would make specific auctions for this technology unnecessary, and wind power could start competing in the regular contract auctions organized by the distribution companies, where all tech- nologies participate on a level ­playing field without discrimination. Indeed, policies seeking to promote the ­ introduction of RE economically must take into consideration the costs of RE generation (in relation to the avoided social cost of generation), resource availability relative to seasonality, and the tech- nical conditions of the system (for example, the capacity of transmission and distribution lines to absorb specific volumes of RE). An assessment of policy complex modeling coupled with the use of efficiency in this context requires ­ other tools that can help analyze the adequacy of the institutional structure in place as well as governance issues. The poor experience with wind conces- sions in China (unrealistic prices, projects not delivered) prompted a switch to FITs. Note 1. As, for example, in the RE studies for Croatia, Serbia, Vietnam, South Africa, and China. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Summary and Conclusions 263 Bibliography Lauber, Volkmar. 2013. “Can Germany’s Energiewende Still Be Slowed Down?” Presentation made at 8th Internationale Energiewirtschaftstagung (IEWT), Technical University of Vienna, Austria, February 2013. Philippines NREB (National Renewable Energy Board). “Feed-In-Tariff (FIT) Disbursement and Collection Guidelines.” Manila, Philippines, http://www.erc.gov.ph/Notice​ /NoticeDownload/464. World Bank. 2010. Economic Analysis of the Medupi Coal Project. Washington, DC: World Bank. ———. 2013. Economic and Financial Analysis of the Kom Ombo CSP Project. Washington, DC: World Bank. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Appendix A Dealing with Uncertainty in Setting Renewable Energy Targets: Croatia This report argues for a supply curve analysis as a rational basis for setting renew- able energy (RE) targets: where the supply curve intersects with the avoided social cost of thermal generation is the optimal quantity of RE. But neither this RE supply curve, nor the avoided social cost of thermal energy, is known with any certainty; both are subject to a range of assumptions, many of which are entirely beyond the control of national decision makers. This appendix provides a practical example of how to deal with such uncer- tainties, based on a World Bank study of RE options for Croatia (Frontier Economics 2003). Table A.1 shows the results of such an analysis of RE targets under three sets of input assumptions: • Unfavorable assumptions (for renewables). Higher than expected capital costs for wind turbines, lower valuations of local damage costs, and avoided social costs based on combined-cycle gas turbines (CCGTs) (that have the lowest emissions of local air pollutants per kilowatt-hour generated). • Expected assumptions. Those that are seen by the government as the most likely. • Favorable assumptions (for renewables). Low capital costs for wind turbines, high valuations of local environmental damage costs, and avoided social costs based on coal. The result is a wide range of potential targets, ranging from 37 megawatts (MW) to 1,337 MW! The range is so large because of the high uncertainty in many of the input assumptions, such as the damage cost of thermal generation (which varies by a factor of 4.5). Given such wide ranges in the value of the target, how should one proceed? Such ranges in uncertainty exist in many planning problems, and one approach to making a decision is to ask about the robustness of the decision. Suppose we choose the 317 MW target based on an assessment of what is most likely and make investments to reach that target: renewables replace a The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   265   266 Dealing with Uncertainty in Setting Renewable Energy Targets: Croatia Table A.1 Economically Optimal Quantity of Renewables Unfavorable assumptions Expected (most likely Favorable assumptions (for renewables) assumptions) (for renewables) Local externality value € cents/kWh 0.35 1 1.6 Wind turbine capital costs €/kW 675 600 525 Technology replaced Gas CCCT Gas CCCT + coal Coal Net benefits, 2010 € million 4.5 13.5 63 2010 target GWh 175 1,070 3,340 2010 target MW 37 317 1,335 Source: Frontier Economics 2003. Note: CCCT = combined-cycle combustion turbine; GWh = gigawatt-hour; kW = kilowatt; kWh = kilowatt-hour; MW = megawatt. mix of gas combined-cycle combustion turbine (CCCT) and coal, and wind turbine capital costs fall to €600/kW, which brings about estimated annual economic benefits of €13.5 million. But suppose that having settled on and built the 317 MW target, the future brings unfavorable conditions—wind replaces only gas CCCT, and capital costs fall to only €675/kW. What then are the net benefits? And what are the net benefits of the more favorable assumptions? Indeed, for the three scenarios por- trayed above, there are nine combinations of assumptions and futures. The various outcomes of this analysis, with three choices and three actual outcomes, can be displayed in a 3 x 3 matrix, as shown in table A.2. The entries in columns 1, 2, and 3 represent the net benefits that correspond to each choice (represented by the rows). For example, if we choose the 317 MW target, but the actual outcome is unfavorable, then there is a net loss of €4.1 million, or if we choose the 1,334 MW target, and the actual outcome is favorable, there is a net benefit of €63.1 million, and so on. How one makes a decision on the basis of these estimates of costs and benefits then depends upon the: • Judgments about the probability of different outcomes. • The decision maker’s risk aversion. Suppose all three outcomes were thought to be equally likely (that is, with a probability of 33.3 percent, as in table A.2). Then we may compute the expected value of the three alternative decisions, as shown in column 4. For example, the expected value, E, for the 317 MW target is: E{expected assumptions} = −4.1 × 0.333 + 13.5 × 0.333 + 29.9 − 0.333 = €13.1 million. Similar calculations are shown for the unfavorable and favorable assumptions. These expected values—shown in Column [4], would be the basis for making a choice for a risk-neutral decision maker: in which case the target selected should be 317 MW, because it has the highest expected value (€22.9 million). On the other hand, if the government is risk averse, then an alternative crite- rion is the Mini-Max decision rule, which calls for choosing the option that has The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Dealing with Uncertainty in Setting Renewable Energy Targets: Croatia 267 Table A.2 Payoff Matrix (Net Benefits in 2010, in € Million) Actual outcome Decision criterion Risk neutral Risk averse Unfavorable Expected Favorable [expected value] [mini-max] [1] [2] [3] [4] [5] Probability of outcome 33.3% 33.3% 33.3% Target (MW) Assumption 37 Unfavorable 4.5 6.7 10 7 4.5 317 Expected −4.1 13.5 29.9 13.1 −4 1,334 Favorable −7.4 13.1 63.1 22.9 −7.4 Source: Frontier Economics 2003. Note: MW = megawatt. Table A.3 Revised Payoff Matrix: Future Coal Plant Unlikely € million Actual outcome Decision criterion Risk neutral Risk averse Unfavorable Expected Favorable [expected value] [mini-max] Probability of outcome 30% 65% 5% Target (MW) Assumption 37 Unfavorable 4.5 6.7 10 6.2 4.5 317 Expected −4.1 13.5 29.9 9.0 −4.1 1,334 Favorable −7.4 13.1 63.1 9.5 −7.4 Source: Frontier Economics 2003. Note: MW = megawatt. the best worst outcome. Column 5 of table A.2 shows the worst outcome for each target; based on this criterion the 37 MW target is optimal, since it has the best worst outcome of €4.5 million. The assumptions favorable to RE are based on coal being the fossil fuel being displaced, but given the government’s policy not to build a new coal plant, a lower probability may be assigned to this scenario. For example, if the favorable scenario (with coal as the avoided cost) is given only a 5 percent chance of occur- ring, then the payoff matrix will appear as shown in table A.3. Now the gain in expected value by choosing the optimistic scenario over the mid-level scenario (from €9.0 to €9.5 million) is quite small, particularly when faced with a possible €7.4 million loss if the unfavorable future occurs. Such analysis may well require many additional assumptions, but it has the advantage that it forces decision makers to be explicit about their risk prefer- ences, and makes the connection between assumptions and the robustness of decisions more transparent. Bibliography Frontier Economics. 2003. Cost-Benefit Analysis for Renewable Energy in Croatia. Report to World Bank, Washington, DC. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Appendix B Multi-Attribute Decision Analysis and Trade-Off Plots Multi-attribute decision analysis (MADA) is a set of techniques designed to go beyond the single objective cost-benefit analysis as a basis for making decisions.1 These offer some practical help in coming to better decisions in the face of mul- tiple objectives where not all variables of interest can be monetized—by provid- ing better insights into the problems, by forcing clarity about goals and risks, by facilitating understanding (if not agreement) among diverse stakeholders, and by assisting decision makers in making trade-offs (Hobbs and Meier 2000). In the assessment of renewable energy (RE) technologies, greenhouse gas (GHG) emis- sions are often treated as a separate attribute precisely because they are so diffi- cult to value, and invite a distracting debate about discount rates. Undiscounted lifetime GHG emissions are thus an oft-encountered attribute. The World Bank Study of Sri Lanka (Economic Consulting Associates 2010)​ —presented in chapter 4—used the following non-monetized attributes to complement the usual economic efficiency variable of total system cost (as generated by the Wien Automatic System Planning [WASP] model): • Local air pollution impacts. Population and stack-height-weighted sulphur dioxide (SO2) emissions. • Energy security (diversity). The Herfindahl Index of generation mix (an index used in economics to measure the concentration of firms in an industry)2: H= ∑s n 2 i where si is the share of generation from the i-th supply source (the lower the value of H, the greater is the diversity of supply). • Consumer impact. Levelized average consumer tariff, Rs/kilowatt-hour (kWh). • Undiscounted lifetime GHG emissions. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   269   270 Multi-Attribute Decision Analysis and Trade-Off Plots When framing such attributes, the first priority is to make sure that the attribute is a meaningful indicator of the underlying goal. For example, the simplest proxy for local air emissions is tons emitted per year—now a routine output of most power systems planning models that supposedly provide information about envi- ronmental impacts. But as noted, in fact tons of emissions say very little about actual impacts on human health, or about the costs—fiscal, social, and other—of health care. In the case of GHG emissions, it matters not where in the world the emission takes place, but in the case of local pollutants such as particulate matter, where and at what height the emission takes place is of crucial importance. One kg of PM103 emitted at ground level by a diesel bus in the center of Colombo has an impact on human health several orders of magnitude greater than a kg of PM10 emitted from a tall utility stack in a remote and sparsely populated area4 (and where most emissions are in any event blown out to sea). The difference between gross emissions, and population-weighted SO2 emissions as a more meaningful proxy for actual damage costs, is illustrated in figure B.1. When loca- tion is taken into account, even though gross emissions increase (with the addi- tion of many new coal projects), damage costs may decrease as the location shifts to less densely populated areas. Trade-off curves are simply XY plots of attributes, two at a time. Typically one shows quadrants relative to the baseline, into which fall the options that may be defined as perturbations of that baseline. Figure B.2 shows such an (illustrative) plot. Each quadrant contains different types of projects: • Quadrant I contains solutions best described as “lose-lose”—options that have higher emissions and higher costs. Typical options in this quadrant would be those involving fossil-fuel price subsidies (assuming the baseline is at ­economic prices), or not subcritical coal units (if the baseline includes ­ supercritical units). • Quadrant II contains solutions involving trade-offs—costs decrease, but emis- sions increase. No flue gas desulphurization (FGD) or pumped storage (PS) are two options that typically occupy this quadrant. • Quadrant III contains solutions that are “win-win,” of which demand-side management (DSM) and reduction in transmission and distribution (T&D) losses are typical examples. Here both attributes improve—that is, there are lower emissions and lower economic costs. • Quadrant IV again contains options that require a trade-off—emissions decrease but only at an increased cost. RE options and the substitution of coal by liquefied natural gas (LNG) are typical options to be found here. Figure B.2 also shows the “trade-off curve.” This is defined as the set of non- dominated options. Option B is said to be dominated by option A, if option A is better than B in both attributes. Thus, in figure B.2, DSM dominates the baseline—and because it is better in both attributes, a rational decision maker ­ would never prefer the baseline over DSM. Intuitively, one may say that options The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Multi-Attribute Decision Analysis and Trade-Off Plots 271 Figure B.1 Emissions vs. Stack Height and Population Weighted Index, Sri Lanka a. Total SO2 emissions 80 SO2 emissions, 1,000 tons 60 40 20 0 20 1 20 2 20 1 22 09 10 13 14 20 9 20 23 24 20 5 16 20 5 26 17 20 8 27 28 1 1 2 1 1 2 1 20 20 20 20 20 20 20 20 20 20 20 20 20 b. Stack height and population-weighted 250 200 Pop-weighted SO2 emissions 150 100 50 0 20 1 12 20 1 22 09 10 20 3 14 20 9 20 23 24 20 5 16 20 5 26 17 20 8 27 28 1 2 1 1 1 2 1 20 20 20 20 20 20 20 20 20 20 20 20 20 Resid Furnace oil Auto diesel CoalTrinco Coal LNG Source: World Bank 2010. Note: resid = residual oil-fired projects (with no sulfur controls, typically burning high sulfur oil); coalTrinco = coal-fired projects with flue gas desulphurization (FGD) on Trincomalee Bay on the eastern coast, sparely populated; coal = coal projects with FGD sited north of Colombo on the west coast; LNG = liquefied natural gas; SO2 = sulphur dioxide. that lie on this trade-off curve are “closest” to the origin, but they all require trade-offs. If, as in this illustrative example, there is a sharp corner in the trade-off curve (the so-called “knee set”), the option that occupies that corner (or one that may be close to it) would receive special attention. In this example, “no pollution controls” has greater emissions than DSM, but only a very small cost The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 272 Multi-Attribute Decision Analysis and Trade-Off Plots Figure B.2 Illustrative Trade-Off Plot 9.0 Trade-o Lose-lose 8.5 Renewable energy Economic cost, USc/kWh 8.0 Fossil fuel price subsidies 7.5 Baseline DSM 7.0 No pollution controls Win-win Trade-o 6.5 200 250 300 350 400 GHG emissions, million tons Trade-o curve Note: GHG = greenhouse gas; kWh = kilowatt-hour; USc = U.S. cents. advantage—so a decision maker would have to give enormous weight to cost and almost no weight at all to emissions to choose this option. Similarly, “RE” (as drawn here) has only slightly lower emissions, but a much higher cost than DSM—so again, to prefer RE over DSM would require that huge weight be given to emissions, and not much to cost. Not all trade-off plots have such knee sets, or even any win-win options, in which case decisions are more difficult to make. Figure B.3 shows a trade-off plot for Vietnam. Trung Son is a World Bank– financed 260 megawatt (MW) hydro project (World Bank 2011). The baseline in this case, which defines the quadrants, is the least-cost capacity expansion plan without Trung Son. The system cost and GHG emissions are plotted relative to the baseline: negative amounts indicate improvements to the objectives (cost reductions, GHG emission reductions). In the lose-lose quadrant are scenarios in which the assumed availability of domestic gas in the baseline is not fulfilled, and must therefore be replaced either by imported LNG or coal plus PS to meet the intermediate and peaking demand of the system. In the trade-off quadrant IV is wind—which in Vietnam is very The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Multi-Attribute Decision Analysis and Trade-Off Plots 273 Figure B.3 Power Sector Options in Vietnam 2 Trade-o Lose-lose Wind No gas [LNG] NPV (system costs), US$ billion 1 No gas [coal+PS] Baseline 0 Renewables to pecon Trung son DSM Win-win Trade-o –1 –40 –20 0 20 40 NPV (GHG emissions), million tons Source: World Bank 2011. Note: DSM = demand-side management; GHG = greenhouse gas; LNG = liquefied natural gas; NPV = net present value; PS = pumped storage. expensive (because the wind regime is at best modest), though it does of course reduce GHG emissions. Trung Son is in the win-win quadrant by virtue of lower lifetime power pro- duction costs, and lower GHG emissions since it displaces gas-fired combined- cycle plants. Also in the win-win quadrant is non-wind renewables. “Renewables to Pecon” refers to the point at which the avoided social cost of thermal generation intersects the RE supply curve, which defines the optimal level of RE. DSM (demand-side management and efficiency improvement) is also in this quadrant. Both DSM and renewables (mainly small hydro) are also being financed by the World Bank.5 Notes 1. At the World Bank, the use of MADA was first elaborated in an Environment Department Research study of Sri Lanka (Meier and Munasinghe 1995) and subse- quently adopted in 1998 for a major World Bank study on environmental issues in the Indian power sector (World Bank 1999), which included detailed state-level The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 274 Multi-Attribute Decision Analysis and Trade-Off Plots assessments for the states of Rajasthan and Karnataka (World Bank 2004a, 2004b). This followed the introduction of such techniques in the 1990s into the Integrated Resource Plan procedures adopted by many utility regulatory commissions in North America, and the pioneering work of Ralph Keeney and Howard Raiffa (1993). The most recent applications in the Bank include a 2009 study of alternatives to coal- based power generation in Sri Lanka (whose results are described in chapter 4), and an economic analysis of the controversial Medupi coal-fired project in South Africa. The academic literature on MADA applications has grown rapidly since 2000: Wallenius and others (2008) found 267 MADA studies in the energy and water resources literature. 2. The quantification of energy security is one of the more difficult issues. In the case of the United States, increasing energy security is arguably a matter of reducing imports. On the other hand, for Nepal, which is dependent entirely on hydro resources, increasing security (and reducing exposure to hydrology risk) is a matter of increasing imports of electricity and fossil fuels for power generation. In the case of Sri Lanka, where for the past 20 years major additions to power generation have only been based on imported auto-diesel, importing coal diversifies supply sources and also improves energy security. 3. Particulate matter (no greater than 10 microns in diameter). 4. In a study of damage costs in six large cities in the developing world, the average dam- age cost in $/ton per 1,000,000 population per $1,000 of per capita income was estimated at $42/ton for particulate matter (no greater than 10 microns in diameter) emitted from high-stack power plants, compared to $3,114/ton from low-level stacks (standby diesel units, diesel buses). See Lvovsky and others (2000). 5. In other words, DSM is not, strictly speaking, a mutually exclusive option (in the sense of the old OP10.04 guidelines for economic analysis): rather, it is a complement to supply-side options, and is part of any portfolio of win-win options. Bibliography Economic Consulting Associates. 2010. Sri Lanka: Environmental Issues in the Power Sector. Report to the World Bank, Washington, DC, May. Hobbs, B., and P. Meier. 2000. Energy Decisions and the Environment: A Guide to the Use of Multi-Criteria Methods. Boston, MA: Kluwer Academic Publishers. Keeney, R., and H. Raiffa. 1993. Decisions with Multiple Objectives. New York: Cambridge University Press. Original edition published by Wiley, New York, 1976. Lvovsky, K., G. Hughes, D. Maddison, B. Ostrop, and D. Pearce. 2000. “Environmental Costs of Fossil Fuels: A Rapid Assessment Method with Application to Six Cities.” Environment Department Paper 78, World Bank, Washington, DC. Meier, P., and M. Munasinghe. 1995. “Incorporating Environmental Concerns into Power- Sector Decision-Making: Case Study of Sri Lanka.” Environment Department Paper 6, World Bank, Washington, DC. Wallenius, J., J. S. Dyer, P. C. Fishburn, R. E. Steuer, S. Zionts, and K. Deb. 2008. “Multiple Criteria Decision-Making, Multi-Attribute Utility Theory: Recent Accomplishments and What Lies Ahead.” Management Science 54: 1360–49. World Bank. 1999. “India: Environmental Issues in the Power Sector Manual for Environmental Decision Making.” Energy Sector Management Assistance Program (ESMAP) Paper 213. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Multi-Attribute Decision Analysis and Trade-Off Plots 275 ———. 2004a. “Environmental Issues in the Power Sector: Long-Term Impacts and Policy Options for Karnataka.” ESMAP Paper 293. ———. 2004b. “Environmental Issues in the Power Sector: Long-Term Impacts and Policy Options for Rajasthan.” ESMAP Paper 292. ———. 2010. Sri Lanka: Environmental Issues in the Power Sector. World Bank, Washington, DC. ———. 2011. Trung Son Hydropower Project. Project Appraisal Document, Report 57910- VN, Washington, DC. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Appendix C Estimating Incremental Costs from Renewable Energy Supply Curves As discussed in chapter 5, increasing the old geothermal tariff ceiling of 9.7 cents/ kilowatt-hour (kWh) to 12.5 cents/kWh raised the question of the potential impact of the new ceiling on an additional subsidy payable by the Ministry of Finance (MoF). The incremental costs can be visualized by looking at the inter- section of the supply curve with the old and new ceiling prices, and calculating the areas in the relevant segments of the curve. This visualization methodology was first used in the China Renewable Energy Scale-up Program (CRESP) in 2003 (Spencer, Meier, and Berrah 2007).1 The supply curve for geothermal projects in Java and Sumatra is shown in figure C.1, which enumerates a total capacity 2,432 megawatts (MW) of ­ geothermal projects. Also shown in this figure are the estimated Perusahaan Listrik Negara (Indonesian State Electric Utility Company, PLN) avoided costs (6.7 cents/kWh),2 and the former 9.7 cents/kWh ceiling price. If only the projects whose costs are below the 9.7 cents/kWh ceiling were built, then 1,949 MW would be built. The other 483 MW of geothermal ­ projects in the supply curve exceed the ceiling and would not be built. The incremental costs associated with this level of geothermal development are ­ represented by the (roughly triangular-shaped) area A under the curve. This area represents the incremental costs, i.e., the subsidy that must be paid to PLN by the MoF. For the costs as shown here, this comes to $120 million per year once all 1,949 MW that have ceiling prices below 9.7 cents/kWh have been built—assuming the bid tender prices (or negotiated prices for the old, legacy Wilayah Kerja Pertambangan Panas Bumi [geothermal work areas as known in Bahasa, Indonesia] [WKP])3 were at the levelized cost of energy as reflected in the supply curve. Figure C.2 shows the potential impact of raising the ceiling to 12.5 cents/kWh— which intersects the supply curve at 2,362 MW. Now the incremental costs increase by the additional amount represented by the areas C+B ($104 m ­ illion), for a total subsidy of $214 million per year once all 2,362 MW have been built. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7   277   278 Estimating Incremental Costs from Renewable Energy Supply Curves Figure C.1 Castlerock Supply Curve, Java and Sumatra (with the Old Ceiling Price) 16 Production cost, USc/kWh 14 12 10 9.7 USc/kWh ceiling 8 A 6 PLN avoided cost (6.7 USc/kWh) 4 0 500 1,000 1,500 2,000 2,500 Installed capacity, MW Source: Meier, Lawless, and Randle 2014. Note: kWh = kilowatt-hour; MW = megawatt; PLN = Perusahaan Listrik Negara (Indonesian State Electric Utility Company); USc = U.S. cents. Figure C.2 Impact of a 12.5 Cents/kWh Ceiling Price 16 14 Production cost, USc/kWh 12.5 USc/kWh, revised ceiling 12 9.7 USc/kWh ceiling B 10 8 C A 6 PLN avoided cost (6.7 USc/kWh) 4 0 500 1,000 1,500 2,000 2,500 Installed capacity, MW Source: Meier, Lawless, and Randle 2014. Note: kWh = kilowatt-hour; MW = megawatt; PLN = Perusahaan Listrik Negara (Indonesian State Electric Utility Company); USc = U.S. cents. Such visualizations have proven useful in communicating the concept of supply curves and incremental costs of renewable energy to stakeholder con- sultation groups. They are easily calculated in simple spreadsheets, and easily presented for different implementation scenarios of bidding behavior (see table 5.8).4 The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Estimating Incremental Costs from Renewable Energy Supply Curves 279 Notes 1. Recall figure 2.4 (in chapter 2). 2. PLN now pays the international price for coal. Wilayah Kerja Pertambangan Panas Bumi (geothermal work areas in Bahasa, 3. Indonesia). 4. The amounts calculated here correspond to column [4] of table 5.8 (i.e., tender bids at the levelized cost of energy [LCOE]). Bibliography Meier, P., J. Lawless, and J. Randle. 2014. Indonesia Geothermal Tariff Reform: Tariff Methodology Report. World Bank and Asian Development Bank, Jakarta, March. Spencer, R., P. Meier, and N. Berrah. 2007. Scaling Up Renewable Energy in China: Economic Modelling Method and Application. ESMAP Knowledge Exchange Series #11, Washington, DC, June. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, the Publishing and Knowledge Division lever- ages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. ­ The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. Whenever possible, books are printed on 50 percent to 100 percent postconsumer recycled paper, and at least 50 percent of the fiber in our book paper is either unbleached or bleached using Totally Chlorine Free (TCF), Processed Chlorine Free (PCF), or Enhanced Elemental Chlorine Free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4. The Design and Sustainability of Renewable Energy Incentives  •  http://dx.doi.org/10.1596/978-1-4648-0314-7 Joining a debate often dominated by widespread misconceptions, this book introduces a rigorous and objective economic perspective on current renewable energy support mechanisms and an empirical analysis of their strengths and weaknesses. It complements the analysis with operational advice on how the regulatory design may need to be modified to minimize the impact on the budget and be affordable to the poor, as well as how to identify and fill the financing gap. The proposed analytical framework illustrates tradeoffs between thermal electricity generation and renewable energy supply with local, regional, and national impacts in the short and in the long run; studies distributional impacts; captures externalities; and compares alternative projects based on equivalent output and cost. Unsurprisingly, the book stresses the need to get the economic, financial, and institutional basics right for the deployment of renewable energy. The study also integrates renewable energy subsidies with fossil subsidies, bringing important questions to the fore, such as the following: to reduce carbon intensity in developing- country economies, is it more efficient to deploy renewable energy or implement alternative options, such as eliminating subsidies on fossil fuels? A representative sample of countries based on energy endowments (coal, natural gas, and hydro-based systems) and policy incentives (from feed-in tariffs to auctions) are examined in the book: Brazil, the Arab Republic of Egypt, Indonesia, South Africa, Sri Lanka, Tanzania, Turkey, and Vietnam. These case studies compare the incremental cost of renewable energy with the average cost of generation and determine the impact that alternative support has on the government budget and residential consumers. The main lessons emerging from The Design and Sustainability of Renewable Energy Incentives are that, to be successful, such incentives • will be effective only once the state-owned utilities who are the buyers of grid-connected renewable energy are themselves in good financial health, • need to be grounded in economic analysis and accompanied by the application of market principles to ensure economic efficiency, and • require a sustainable, equitable, and transparent recovery of incremental costs. ISBN 978-1-4648-0314-7 SKU 210314