LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY ANIL CABRAAL WILLIAM A. WARD V. SUSAN BOGACH AMIT JAIN © 2021 International Bank for Reconstruction and Development / The World Bank The World Bank World Bank Office Dhaka 1818 H Street, NW Washington, DC 20433 Plot- E-32, Agargaon, Sher-e-Bangla Nagar, Dhaka-1207 USA Bangladesh Tel: 1-202-473-1000 Tel: 880-2-5566-7777 Fax: 1-207-477-6391 Fax: 880-2-5566-7778 www.worldbank.org www.worldbank.org/bangladesh This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, 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, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. 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 construed or 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 The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Attribution Please cite the work as follows: Cabraal, Anil, William A. Ward, V. Susan Bogach and Amit Jain. 2021. Living in the Light: The Bangladesh Solar Home Systems Story. A World Bank Study. Washington, DC: World Bank. 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. Views and opinions expressed in the adaptation are the sole responsibility of 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 reuse a component of this work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but not are limited to tables, figures, or images. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 1-202-522-2625; e-mail: pubrights@worldbank.org. Photo credit: IDCOL (Front cover photo, and page no. vii, xi, xvii, 26, 27, 78, 79, 105) David Waldorf/The World Bank (page no. 3, 7, 12. 30, 37, 44) Arne Hoel/The World Bank (page no. 65, 71) Graphic design and Illustration by M. Inamul Shahriar A World Bank Study LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY Anil Cabraal William A. Ward V. Susan Bogach Amit Jain TABLE OF CONTENT FOREWORD – GOVERNMENT OF BANGLADESH viii FOREWORD – THE WORLD BANK ix FOREWORD – INTERNATIONAL SOLAR ALLIANCE x ACKNOWLEDGMENTS xii AUTHORS xiv ACRONYMS AND ABBREVIATIONS xv ECONOMIC INDEXES xvi EXECUTIVE SUMMARY xviii 1 INTRODUCTION 1 2 THE SOLAR HOME SYSTEMS PROGRAM 9 3 ADAPTING TO REALITY 29 4 IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN 39 5 ECONOMIC AND FINANCIAL ANALYSIS 51 6 MAIN CONCLUSIONS AND LESSONS LEARNED 67 REFERENCES 73 APPENDIX 81 LIST OF FIGURES Figure 1. Organizational Structure of the SHS Program and Functional Roles (IDCOL) xix Figure 2. SHS Installation Progress xx Figure 3. Schematic of an SHS 2 Figure 4. Trends in Rural Electricity Access 1995–2018 4 Figure 5. SHS Program Functional Relationships and Roles 13 Figure 6. SHS Program Organization at IDCOL 14 Figure 7. Development Partner Financing for the SHS Program (US$, millions) 19 Figure 8. Annual and Cumulative SHS Installations 22 Figure 9. Number of SHS Installed from 2003 to 2018 by Size 22 Figure 10. Households with SHS as Percentage of Rural and Total Households 22 Figure 11. SHS Sales (by Color) and Market Penetration (% of Total Households) at the District Level 23 Figure 12. Timeline of Promotion and Outreach Activities 31 Figure 13. Trend in Weighted Average Size of SHS 33 Figure 14. Trend in Market Share of SHS by Capacity 33 Figure 15. Reduction of SHS Unit Cost with Cumulative SHS Installations 34 Figure 16. Trend in Unelectrified Rural Households and SHS Installations 40 Figure 17. BREB Grid Connection Rate Increased Rapidly after 2014 42 Figure 18. Sales of SHS and TR/KABITA Larger Public Service PV Systems 43 Figure 19. SHS Sales per PO under the SHS Program 44 Figure 20. Global Solar PV Module Cost Trend 45 Figure 21. Cumulative Stakeholder Benefits from the SHS Program, 2003–2042 62 Figure 22. ODA Loan Withdrawals and Repayments by GOB 2003–2054 62 Figure 23. Net GOB Loan Receipts by IDCOL Minus IDCOL Repayments 2003-2042 63 Figure 24. GOB Net on ODA Pass-Through to IDCOL 2003-2054 63 Figure 25. GOB Treasury Net Flows from SHS Program 2003 to 2054 64 LIST OF TABLES Table 1. Sources and Amounts of Financing to December 2017 for the SHS Program 18 Table 2. Lending Terms to POs 19 Table 3. Sources and Amounts of Grants for SHS per Agreement 20 Table 4. Actual Amount of Grants Provided for SHS 21 Table 5. Market Penetration of SHS in Bangladesh 2003–2018 24 Table 6. Total SHS Sales as % of Divisional Households (2003–2018) 25 Table 7. Trends in Unit Cost of SHS 2003–2017 34 Table 8. Debt Collection Efficiency of POs from SHS Customers, 2009 to 2019 42 Table 9. Example of Comparative SHS Costs in 2013 and 2016 45 Table 10. Strengthening Security Requirements for Loans to POs 46 Table 11. PO Loan Status 2009–2018 47 Table 12. Economic Analysis of the SHS Program 2003–2029 (Constant 2018 US$, millions) 54 Table 13. Financial Analysis of the SHS Program from the Household Stakeholder Group Perspective (Constant 2018 US$, millions) 57 Table 14. Stakeholder Indicative Net Financial Benefits from the SHS Program 2003–2042 (Constant 2018 US$, millions) 59 Table 15. Summation of Distribution of Estimated Net Financial Benefits to Indicated Stakeholders 2003–2042 61 Table D.1. Expected Useful Life of SHS Component Parts 94 Table D.2. Kerosene Cost Estimation: Landed Cost at Chittagong 2003–2018 (constant 2018 US$/liter) 96 Table D.3. Kerosene Cost Estimation: From Landed Cost at Chittagong to Official versus Actual Retail Prices per Liter, 2003–2018 (constant 2018 US$/liter) 96 Table D.4. Summary Survey Data (Grameen SHS CDM Project 2012) 97 Table D.5. Grameen Shakti Kerosene Avoided in CDM Application 97 Table D.6. Kerosene Offset (liters per year), Based on Grameen Shakti Survey Data 98 Table D.7. Calculation of Total Liters of Kerosene Use Avoided during 2003–2022 by SHS Program Installations 2003–2018 99 Table D.8. Kerosene Saved by the Bangladesh SHS Program, 2003 to 2022: Value at the Household Level, Aggregated SHS Households 99 Table D.9. Estimation of Kerosene Subsidy Saved by the GOB as a Result of SHS Program Electricity Output (constant 2018 US$) 100 Table D.10. Kerosene Distributor Profit Losses from SHS-Related Reductions in Kerosene Use 100 Table D.11. Financial and Economic Value of Grid Electricity Offset by SHS from 2022 to 2029 101 Table D.12. SHS Installation Data 101 Table D.13. SHS Program Costs by Year 102 Table D.14. Tax and Duty Information 102 Table D.15. Imputed SHS Approximate Conversion Factors by Year or Year Groups 103 Table D.16. Calculation of Tax Component of GOB Stakeholder Impact (constant 2018 US$) 103 Table D.17. Estimated Partner Organization Profits on SHS Sales and Installations 104 Table E.1. IDCOL Net Financial Benefit from Management of SHS Program, 2003–2042 106 Table F.1. ODA Lending Terms 108 Table F.2. ODA Loan Withdrawals and Repayments by Bangladesh Government 2003–2055 109 Table F.3. IDCOL ODA Loan Withdrawals and Repayments, 2003–2042 110 Table F.4. Comparison of Treasury versus IDCOL Stakeholder Impacts of Financial Structuring of ODA Pass-Through Funding of the SHS Program, 2003–2054 111 Table F.5. Net Impact on GOB of SHS Program in Present Value Terms 2003–2054 113 LIST OF BOXES Box 1. Solar Home System 2 Box 2. Bangladesh Solar Home Systems Program - World Bank Support 11 Box 3. Microfinance in Bangladesh and Its Role in SHS 15 Box 4. Case Studies - Improving Quality of Life 26 Box 5. Findings and Recommendations for Debt Problem Resolution 48 Box 6. Methodology Used to Compute Kerosene Cost 95 vi The solar home system has enabled us to break out of darkness and live in light, isn’t it good for us? Muktilia Bhrumo, SHS User in Thakurgaon District, conversation with Noara Razzak and others from BRAC University in 2012 vii FOREWORD Bangladesh accords a high priority to providing electricity access to our people. Our pledge to bring electricity services to all its people is enshrined in our 1972 Constitution as a fundamental principle of State policy. Access to electricity nation-wide reached 97 percent in 2020. By 2021, Bangladesh expects to achieve universal electricity access for its rural people, only forty percent of whom had electricity a decade ago. This contributes to realizing the social transformation of rural areas envisaged in the Constitution. Solar home systems technology has been an important instrument in achieving our universal access goal. Bangladesh completed the installation of 5.6 million solar home systems providing electricity to about 22 million rural people during my tenure. The Solar Home Systems Program led by the Infrastructure Development Company Ltd (IDCOL) was responsible for over 4 million of these installations from 2003 to 2018. This achievement required the mobilization of enormous financial and human resources, as well as the marshalling of the capabilities of Bangladesh civil society and the private sector, with the support of the Government. We acknowledge the technical and financial support of our international partners, especially the World Bank, which has been our steadfast partner for over 15 years and provided over US$400 million in financing towards SHS electrification. Building on the success of the SHS Program, the Government of Bangladesh has launched several other large-scale solar programs such as Solar Irrigation, Solar Mini-grids, Solar Parks, Roof-top Solar, Floating Solar, and Solar Boats. We expect that about 17 percent of Bangladesh’s electricity will come from renewable energy sources by 2041. Thanks in part to the enhanced prospects of renewable energy, Bangladesh is considering scrapping 90 percent of its coal power pipeline. We believe that coal power generation can be limited to 5GW, or about 12 percent of total generation capacity. This book tells a compelling story. It documents our experiences in deploying solar home system to bring electricity to our people. The book’s analysis of the SHS Program’s organizational effectiveness, how partners were mobilized, how risks were mitigated, and how financial resources were raised and deployed provide invaluable insights as we scale up our renewable energy use. Through this book, we share our experiences and lessons of solar off-grid electrification to help other countries around the world as they too mobilize to achieve universal access to electricity. NASRUL HAMID Honourable State Minister Ministry of Power, Energy and Mineral Resources Government of Bangladesh viii FOREWORD We are proud to partner with the Government of Bangladesh to increase access to clean electricity through solar power. Today, thanks to our joint MERCY MIYANG TEMBON efforts, along with other development partners, civil society and the private Country Director for sector, Bangladesh has one of the world’s largest domestic off-grid solar Bangladesh and Bhutan power programs bringing modern electricity services to its rural population. The World Bank The Bangladesh Solar Home Systems (SHS) Program supported by the Government of Bangladesh and the World Bank began in 2003 at a time, when only 27 percent of rural Bangladesh households had access to grid electricity. The 50,000 household pilot was implemented under the capable leadership of the Infrastructure Development Company Ltd (IDCOL). Thinking outside the box, IDCOL combined its expertise in infrastructure financing with Bangladesh’s pioneering work in micro-finance and early attempts at solar electrification by companies to build an off-grid electrification business model that ultimately brought electricity services to about 14 percent of Bangladesh population. Over the course of 15 years US$1 billion in financing was mobilized from international and domestic sources for SHS electrification. The SHS Program has demonstrated that hundreds of millions of dollars mobilized at the international level can flow efficiently down to the remotest corners of Bangladesh to offer loans in amounts of one hundred dollars or less that permitted a rural household to purchase a solar home system. The Bangladesh SHS electrification experience, as told in this book, convincingly demonstrates that off-grid electrification can be a mainstream GUANGZHE CHEN provider of electricity to a large segment of the population. Isolated Regional Director for Infrastructure communities no longer need to wait years or decades until the grid reaches South Asia Region their communities to reap the benefits of electricity. The World Bank Building on its success in using solar energy to provide electricity in rural areas, the World Bank is now extending considerable financial and technical support to Bangladesh to scale up other clean renewable energy options including solar irrigation, solar mini-grids, roof-top solar, and solar farms. The book offers practical lessons to other countries that are embarking on accelerated off-grid electrification programs. It is a steppingstone to achieving a clean energy future to benefit all humankind. The World Bank will use the insights in this book to help countries implement economically smart and tailored approaches that best suit their needs, and will continue to support technological, financial and policy innovations that can help accelerate the expansion of reliable and affordable electricity services and end energy poverty. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | ix FOREWORD The International Solar Alliance (ISA) has been conceived as a coalition of solar-resource- rich countries to address their special energy needs. The ISA is an action-oriented, member-driven, collaborative platform for increased solar energy technologies to enhance energy security and sustainable development and improve access to energy in developing member countries. The ISA has 72 countries have signed and ratified the ISA Framework Agreement, and another 89 countries have signed the ISA Framework Agreement. I was deeply honoured to take over as Director-General of the International Solar Alliance at inception to lead the global effort in bringing together countries with the collective objective of achieving their energy needs through solar alternatives. With energy independence becoming a part of the new global narrative, new partnerships, and scale involvement of the private sector will be paramount for ISA’s work. The lessons from the Bangladesh Solar Home System program reflect globally replicable results. Access to modern, clean energy is essential to give every person the opportunity for a prosperous, dignified, and healthy life. Today, as the world is battling the COVID-19 pandemic, access to modern energy is even more crucial and urgent to provide better health care access and improve peoples’ lives. With more than 800 million people still lacking electricity access, finding and deploying innovative ways to bring electricity services quickly, affordably, and sustainably to unelectrified communities is imperative. The Bangladesh Solar Home Systems Program represents one such way for the rapid and flexible provision of electricity access to individual households. Over fifteen years, in a successful partnership between the government, the private sector, civil society and the international community, it brought modern electricity services to remote and dispersed communities in Bangladesh, far faster and at a lower cost than would have been possible by extending the national electricity grid. The SHS Program contributed significantly to Bangladesh’s likely achievement of universal access to electricity by 2021. It did so with renewable energy resources and contributed to reducing greenhouse gas emissions. This book is a comprehensive description of the program. It documents the approach Bangladesh took, the partnerships it enabled, the challenges it faced, and the outcomes achieved. It is a useful reference to many other countries that wish to deploy off-grid renewable energy technologies to bring modern electricity services to their people. This book offers insights and lessons to other countries that strive to achieve some of the key objectives for ISA member countries and instil the confidence that they can succeed. This book is invaluable for us at ISA, to let leaders know what is working, where, when, and why. This includes shaping messages that reflect the promise and challenges faced in achieving our objectives by making stories, approaches, and data access, allowing diverse decision-makers to act and spur results. UPENDRA TRIPATHY Director-General International Solar Alliance x LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xi ACKNOWLEDGMENTS Born in 2002 when the early steps in using solar photovoltaics for off-grid electrification were being taken, the Bangladesh Solar Home Systems Program owes its success to the early visionaries and actors. They saw the promise of the technology and imagined what it would take to propagate it. Foremost among them were Dr. Masihur Rahman, Secretary, Economic Relations Division and Chairman, Infrastructure Development Company Ltd (IDCOL); Subramaniam Vijay Iyer, the World Bank Task Team Leader; and Fouzul Kabir Khan, the Chief Executive Officer (CEO) of IDCOL. Together, they conceived, launched, and guided the pilot program and began its scale-up. They were followed by the subsequent CEOs of IDCOL, up to Mahmood Malik the present CEO of IDCOL, and the World Bank Task Team Leaders who followed Vijay, including Raihan Elahi, Zubair Sadeque, and Amit Jain, and their teams who led its expansion. The success of the SHS Program is due to their drive for success and their pragmatic approach to problem solving. Crucial to the success was the building and nurturing of a public-private partnership by IDCOL. IDCOL staff played a vital role in the program implementation. Without them there would not be a successful SHS Program. Special mention is due to S. M. Monirul Islam, Deputy CEO, and to Md. Enamul Karim Pavel, Head of Renewable Energy of IDCOL, who directed the SHS Program on a day- to-day basis. Formanul Islam played an important role in the evolution of the SHS Program while he was Deputy CEO from 2012 to 2015 and previously as Head of IDCOL’s Legal Department. We are grateful to Rashed Rahman Khan, Manager, SHS Program, and Serajul Hossain, Vice President and Unit Head, SHS Program, at IDCOL without whose assistance in gathering data and providing insights into the SHS Program this report would not be possible. We deeply appreciate the willingness of IDCOL to openly share data and offer insights not only into what they did right but also into what could have been done better. Lessons are most valuable when they are based on open and honest discussion. Unless otherwise specified, all charts, tables, graphs, and figures are based on data provided by IDCOL. Over the course of many years, our visits to villages to meet solar home systems users provided firsthand testimony of their experiences and aspirations. Together with IDCOL, we learnt what would work, what would not, and how electricity positively impacted their lives. These visits, along with discussions with staff from partner organizations (POs), were invaluable and informative. Professor Rizwan Khan, Chairman of the independent Technical Standards Committee (TSC) and Vice Chancellor of United International University, played an important role from the onset to the present time in setting and enforcing quality standards for SHS. The SHS Program success owes much to the confidence consumers gained that the SHS would provide the promised services reliably over the long term. Contributions of the POs were crucial to the success of the SHS Program. Among their leaders were Hasna Khan and Asma Haque of Prokousholi Shangsad during the project launch stage. During xii implementation, we acknowledge the contributions of Dipal Barua, previously Managing Director of Grameen Shakti and then Chairman of Bright Green Foundation; Ruhul Kuddus, previously Deputy General Manager, Grameen Shakti, and then Executive Director of the Rural Services Foundation; Abdul Muyeed Chy, Executive Director, Bangladesh Rural Advancement Committee (BRAC); Mohammad Ibrahim, Executive Director, Centre for Mass Education in Science (CMES); and Hosne Ara, Chairman, Thengamara Mohila Sabuj Sangha (TMSS), to name but a few of many. World Bank Country Directors supported the SHS Program and committed the required financing and staff support. We acknowledge the valuable guidance from Demetrios Papathanasiou, Practice Manager, South Asia Energy Unit, the World Bank. We thank the other financiers who recognized the value of building on the SHS Program modalities and co-financed the SHS Program. Their contributions permitted the SHS Program to benefit tens of millions of people and build a renewable energy industry. They include the Asian Development Bank (ADB), Global Environment Facility (GEF), Global Partnership on Output-Based Aid (GPOBA), German Agency for International Cooperation (Deutsche Gesellschaft für Internationale Zusammenarbeit, GIZ), Islamic Development Bank (IsDB), Japan International Cooperation Agency (JICA), Kreditanstalt für Wiederaufbau (KfW), UK Department for International Development (DFID), and United States Agency for International Development (USAID). The feedback from the peer reviewers Md. Enamul Karim Pavel of IDCOL and Raihan Elahi of the World Bank, were valuable in improving the report and making it more understandable to a wider audience. Enamul Karim has had a long and deep engagement in the SHS Program and as the Head of Renewable Energy of IDCOL he was instrumental in guiding it to success. Raihan Elahi as a former Task Team Leader of the RERED Project, knew the project intimately and his suggestions are grounded in that deep knowledge. The authors acknowledge the data analysis support from Majid Hashemi at Clemson University. We acknowledge with thanks the multifaceted support from Sreyamsa Bairiganjan; editorial support from Selvaraj Ranganathan; Inamul Shahriar for artwork, typesetting and book design; and Mehrin A. Mahbub, Communications Officer at the World Bank Dhaka Office for guidance in producing this report. The findings, interpretations, and conclusions expressed in this work are solely those of the authors and do not necessarily reflect the views of the World Bank, its Board of Executive Directors, or the governments they represent; the Government of Bangladesh (GOB); or IDCOL. Any errors or omissions in this work are entirely the responsibility of the authors. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xiii AUTHORS ANIL CABRAAL, PhD He was with the World Bank from 1995 to 2010 where he developed and supported Formerly Lead Energy Specialist, off-grid solar electrification projects in Indonesia, Sri Lanka, China, Myanmar, India, Energy Unit, Energy,Transport and Water Bangladesh, Tanzania, and Liberia, among others. Between 2007 and 2019, as a Department, Sustainable Development World Bank team member supervising the SHS Program, he had close association Vice Presidency, The World Bank, with the Bangladesh SHS Program and its principal participants, including Washington, DC, USA. beneficiaries. He received the Professor Robert Hill Award for contributions to Photovoltaics for Development at the European Photovoltaics and Solar Energy Conference and Exhibition in 2005. Presently, he is a consultant working on World Bank renewable energy projects in Bangladesh, Myanmar, and elsewhere. WILLIAM A. WARD, PhD At various times during 1974–1980, he was with the World Bank and served as Professor Emeritus, John E. Walker Senior Economist, South Asia Agricultural Projects; Division Chief/Senior Lecturer, Department of Economics, Economic Development Institute; and Lecturer, Economic Development Institute. Clemson University, He continues to serve as a Project Economics Expert on the Panel of Experts of Clemson South Carolina, USA. the World Bank Inspection Panel. He is a Senior Project Economics Consultant for the Millennium Challenge Corporation. He taught project economics and microeconomic policy analysis to staff at the World Bank, Asian Development Bank (ADB), International Finance Corporation, Food and Agriculture Organization of the United Nations, and United Nations Development Programme (UNDP) and other aid officials. He is the coauthor of many peer-reviewed books and articles on economic analysis, including, among others, ‘The Economics of Project Analysis: A Practitioners’ Guide’ and ‘Cost-Benefit Analysis Theory versus Practice at the World Bank 1960–2015’. V. SUSAN BOGACH During her time at the World Bank from 1995 to 2012, she was the task team Formerly Senior Energy Economist, leader of energy projects and studies in Asia and Latin America, with a focus on The World Bank, rural electrification and renewable energy. Before her time at the World Bank, her Washington, DC, USA assignments included work with UNDP, the Canadian Development Agency, and the International Development Research Center in Canada. She is currently working as a consultant on World Bank energy projects, in Africa and elsewhere. AMIT JAIN, PhD He has a Ph.D. in solar, Fulbright Scholar at NREL, USA and Chevening Fellow at Senior Energy Specialist, King’s College, London. He works with the World Bank and is leading a $2 billion The World Bank renewable portfolio in India, Bangladesh and Maldives. It includes REWA and Charanka solar park, integrated by PM Shri Narendra Modi. The project has received the World Bank Group President Award for innovation and excellence and included in the Prime Minister’s book ‘A Book of Innovation: New Beginnings’. His previous assignments include the International Renewable Energy Agency (IRENA), Asian Development Bank (ADB) and the Clinton Foundation. Amit has authored two books on climate change policy and waste to energy and has several international peer-reviewed journal publications. His hobbies include scuba diving, badminton, and outdoor sports. xiv ACRONYMS AND ABBREVIATIONS AC Alternating Current HS Harmonized System ADB Asian Development Bank ICT Information and Communication Technology AIT Advanced Income Tax IDCOL Infrastructure Development Company Ltd AVT Advanced Trade VAT IEA International Energy Agency BIDS Bangladesh Institute of Development Studies IEC International Electrotechnical Commission BPC Bangladesh Petroleum Corporation IFC International Finance Corporation BRAC Bangladesh Rural Advancement Committee IRENA International Renewable Energy Agency BREB Bangladesh Rural Electrification Board IRR Internal Rate of Return BSTI Bangladesh Standards and Testing Institution IsDB Islamic Development Bank BUET Bangladesh University of Engineering and ISO International Standards Organization Technology JICA Japan International Cooperation Agency CBA Cost-Benefit Analysis JNNSM Jawaharlal Nehru National Solar Mission CD Customs Duty KfW Kreditanstalt für Wiederaufbau CDM Clean Development Mechanism LED Light-Emitting Diode CEA Cost-Effectiveness Analysis LGED Local Government Engineering Department CEIP Collection Efficiency Improvement Program MFI Microfinance Institution CEO Chief Executive Officer MIS Management Information System CF Conversion Factor MOPEMR Ministry of Power, Energy and Mineral Resources CFL Compact Fluorescent Light NGO Nongovernmental Organization CIB Credit Information Bureau NPV Net Present Value CIF Cost, Insurance, and Freight OC Operations Committee CMES Centre for Mass Education in Science ODA Official Development Assistance CUA Cost-Utility Analysis OGS Off-Grid Solar DC Direct Current OHSAS Occupational Health & Safety Management Systems DCF Discounted Cash Flow PA Participation Agreement DFID UK Department for International Development PAR Portfolio at Risk DSRA Debt Service Reserve Accounts PAYG Pay-as-You-Go EHS Environmental Health and Safety PE Private Entity EIRR Economic Internal Rate of Return PKSF Palli Karma-Sahayak Foundation ERD Economic Relations Division PMU Program Management Unit ERP Enterprise Resource Planning PO Partner Organization FIRR Financial Internal Rate of Return PV Photovoltaic FAO Food and Agriculture Organization of the United RD Regulatory Duty Nations RERED Rural Electrification and Renewable Energy FOB Free on Board Development (Project) FTL Fluorescent Tube Light RSF Rural Services Foundation GDP Gross Domestic Product SD Supplementary Duty GEF Global Environment Facility SEforALL Sustainable Energy for All GHG Greenhouse Gas SHS Solar Home System(s) GIZ German Agency for International Cooperation SMA Special Mention Account (Deutsche Gesellschaft für Internationale SREDA Sustainable and Renewable Energy Authority Zusammenarbeit) TMSS Thengamara Mohila Sabuj Sangha GOB Government of Bangladesh TR/KABITA National Social Safety Net Program GPOBA Global Partnership on Output-Based Aid TSC Technical Standards Committee GTZ German Agency for Technical Cooperation UNDP United Nations Development Programme (Deutsche Gesellschaft für Technische USAID United States Agency for International Zusammenarbeit) (replaced by GIZ) Development HHI Herfindahl–Hirschmann Index VAT Value Added Tax HIES Household Income and Expenditure Survey WTP Willingness to Pay LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xv ECONOMIC INDEXES 60.04 Bangladesh GDP Deflator (Index) 2003 85.69 US GDP Deflator (Index) 58.15 Exchange Rate (BDT/US$) 64.60 2004 88.04 59.51 69.15 2005 90.88 64.33 73.83 2006 93.67 68.93 80.56 2007 96.16 68.87 87.73 2008 98.05 68.60 92.48 2009 98.79 69.04 100.00 2010 100.00 69.65 111.40 2011 102.09 74.15 118.32 2012 104.05 81.86 127.23 2013 105.87 78.10 136.13 2014 107.88 77.64 144.56 2015 109.03 77.95 152.53 2016 110.22 78.47 161.23 2017 112.32 80.44 177.26 Source: Government of Bangladesh and IDCOL 2018 114.85 Fiscal Year is from July 1 to June 30 up to 2015 83.90 and thereafter, January 1 to December 31 xvi LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xvii EXECUTIVE SUMMARY The Bangladesh Solar Home Systems (SHS) Program is the largest national program in the world for off-grid electrification. Begun in 2003, SHS installations under the Program ended in 2018. It is the longest, continuously operating off-grid electrification program in the world. The SHS Program was led and implemented by the Infrastructure Development Company Ltd (IDCOL). Over a 15-year period beginning in 2003, over 4.1 million SHS were sold and supported using a competitive business model that offered consumers a choice of quality SHS, made affordable with financing. About 14 percent of the Bangladesh population (2011 Census), about 20 million people, obtained electricity services through the SHS Program. The SHS Program enabled one-quarter of the unelectrified rural population in 2003 to obtain electricity services far sooner than would have been possible with grid electricity. SHS were mainly used in rural homes for lighting, mobile phone charging, and powering TVs and radios. They were also used in about 200,000 rural businesses and religious facilities. The program led to SHS becoming a credible electricity source in Bangladesh and, more broadly, to the acceptance of solar photovoltaics (PV) as an electricity generation technology. Building on the credibility gained, SHS distribution to the poorest households under other government programs and commercial SHS sales picked up in later years along with IDCOL-financed sales. While the Bangladesh SHS Program will continue to 2021, this report covers the program from 2003 to 2018, describes its benefits and costs, and discusses how the program adapted to inevitable changes and risks over the 15-year period. It draws lessons that can help guide the development and implementation of other sustainable off-grid electrification programs. IMPLEMENTATION MODEL IDCOL mobilized partner organizations (POs) that were mainly nongovernmental organizations (NGOs) and microfinance institutions (MFIs) with rural networks. The POs competitively marketed, sold, financed, installed, and serviced quality-certified SHS to rural customers. Beginning with 5 POs in 2003, their number grew to 57 by 2015. The customers were mainly rural households, businesses, and religious institutions. The government obtained SHS Program funding from development partners, beginning with IDA funds from the World Bank. IDCOL obtained these funds from the government in local currency to refinance a portion of the loans given by POs to SHS customers. The POs sold SHS to customers on credit with payments spread out over a period of up to three years at interest rates of 12 to 16 percent. Small grants, declining over time from 19 percent of the cost in 2003 to 5 percent by 2017, were given to increase the affordability of the SHS and to help the POs strengthen their institutional capability. The customers repaid the loans to the POs which repaid their loans to IDCOL. IDCOL then repaid its loans from the xviii Provides foreign currency grants and loans on soft terms Government of Development Partners Bangladesh Repays loans Provides foreign Repays currency grants and loans loans on soft terms Independent PO Independent Selection Committee IDCOL Technical ap App Standards po lie Committee Ap in s fo pr tm r ov en es t Provides PO Repays s grants and Seeks Provides loans loans approval approval Seeks operational Supplies interventions equipment IDCOL Operations Partner Committee Organizations Suppliers Provides Pay for oversight, equipment guidance & solutions Sells & Pays for SHS service SHS Customers (Housholds) Figure 1: Organizational Structure of the SHS Program and Functional Roles (IDCOL) government which repaid the development partners. This satisfied. An independent Technology Standards Committee business model permitted loans of hundreds of millions of (TSC) was established by IDCOL to set and enforce quality dollars from international sources to flow through to give standards. The POs sourced SHS and components from microloans to millions of rural customers living in distant domestic and international suppliers that met quality and areas. The roles and responsibilities and the relationships performance standards. IDCOL established an independent between the principal entities are depicted in Figure 1. 1 PO Selection Committee to screen and qualify POs. Crucial to the successful program management was an Operations IDCOL led, managed, and supervised the overall program. Committee (OC) that met with the POs monthly to monitor IDCOL’s strong and committed management was crucial to progress, resolve problems, and share experiences and the program’s success, to ensure that all parties met their lessons. Technical quality audits, fields surveys, and financial and service obligations and that customers were consumer satisfaction surveys were conducted regularly. 1 Unless otherwise specified, all charts, graphs, figures, and tables are based on data provided by IDCOL. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xix EXECUTIVE SUMMARY ACHIEVEMENTS Customers could select from several SHS sizes and functional capabilities that ranged from 10 Wp for basic Beginning in 2003, SHS sales grew rapidly and peaked in lighting and mobile phone charging to as large as 300 Wp 2013 with over 861,000 SHS installed that year (see Figure 2). that could power TV, fans, and so on. At the beginning, Then the pace of installations began dropping. The decline SHS size averaged about 50 Wp. As more efficient and was due mainly to rapid grid network expansion beginning durable light-emitting diode (LED) lamps began replacing in 2015 and the National Social Safety Net Program (TR/ fluorescent tube and compact fluorescent lights (CFLs), KABITA) that, beginning in 2014–2015, gave away SHS to the average size of SHS decreased to about 30 Wp in 2013, poor households, PV systems for public services, and solar increasing affordability. Later, while SHS prices further streetlights. Also, owing to the credibility of SHS created declined due to a drop in PV module costs and reduction by the IDCOL SHS Program, commercial retail sales of in the minimum battery size requirement from three to two SHS began expanding about the same time. By 2018, days of autonomy, the average SHS size increased to 40 cumulatively, over 4.1 million SHS were sold under the SHS Wp as rural households’ income increased and they sought Program. The total solar PV capacity installed was 163 MW. additional services from the SHS. The average grant amount Over their useful lifetime (conservatively assumed to be 12 per SHS dropped from 19 percent of the retail price in 2003 years), SHS would supply about 2 GWh of electricity. to under 5 percent in 2017. In the latter stage, the grant was The share of rural households gaining access to electricity to buy down the SHS cost. services through the SHS Program grew steadily. It peaked Throughout this competitive, market-oriented SHS Program, at 16.2 percent of rural households by 2016, or 10.5 percent unit costs of SHS were comparatively low compared to of total households in Bangladesh. In comparison, the total other countries. In 2003, the installed unit cost of an SHS electricity coverage of the rural population in 2016 was 66 averaged US$12 per Wp. It dropped to about US$10 per Wp percent. The concentration of SHS in rural areas varied— in 2010 and to under US$5 per Wp by about 2017 (constant from a high of 39 percent of households in Barisal Division 2018 US$). What is particularly noteworthy was that the to 6 percent in Rajshahi Division. At a district level, SHS use price of an SHS included free maintenance services for three was as high with every two out of three households using years and a five-year warranty for batteries. Surveys done by SHS in one district, though, not unexpectedly, in more IDCOL found that 90 percent of these tubular plate deep- urbanized areas such as Dhaka District, it was a low 0.2 cycle batteries were operating within specifications even percent. after five years. Figure 2: SHS Installation Progress xx The drop in SHS Program sales after 2014 was caused watching TV, socializing, and visiting friends and mainly due to the start of rapid expansion of the grid. The neighbors after the adoption of SHS. Bangladesh Rural Electrification Board (BREB) increased TV, radio, and mobile phones enabled rural people to grid connections by 280 percent between 2015 and 2019, connect to the rest of the world and brought a great from 9.4 million at the beginning of 2015 to 26.5 million understanding of their rights. by the end of 2019, suddenly shrinking the number of unelectrified homes. The TR/KABITA Program began to Enterprise and social services: The SHS benefited nearly provide SHS to the poorest and supply PV systems for public 200,000 enterprise and social service customers with facilities and streetlights; initially it competed with the SHS better quality light, extended hours of operation, and power Program. By 2016–2017, IDCOL convinced the government for small appliances. These included offices (about 2,300), to let it manage TR/KABITA. IDCOL then used the SHS educational institutions (3,700), restaurants (270), retail Program infrastructure to run the TR/KABITA Program, shops (10,600), mosques (177,300), and other enterprises enforcing the same quality and service standards. TR/ (4,600). These beneficiaries accounted for about 5 percent of KABITA provided business to the POs to help overcome the the total SHS sold. drop in sales under the SHS Program. From mid-2015 to March 2019, the TR/KABITA Program supplied nearly 900,000 Kerosene savings: The SHS Program would have saved systems, 83 percent of which were SHS. about 4 billion liters of kerosene from 2003 to 2021. The value of kerosene saved by households between 2003 and 2018 at the retail price is estimated at US$908 million (in BENEFITS OF SHS PROGRAM constant 2018 US$ discounted at 10 percent). Additional Among the main benefits of the SHS Program in Bangladesh savings will continue to be obtained from SHS that are used were the following: beyond 2021—though these savings are smaller and accrue to avoiding mainly grid electricity use rather than kerosene Faster access to electricity: SHS consumers gained avoidance. access to electricity services far sooner than if they had to wait for a grid electricity connection. Before 2013, due to Technology improvements: The competitive business inadequate generation capacity expansion and the slow model permitted SHS consumers to benefit from pace of obtaining an electricity connection, it had been technology improvements, especially transitioning to more estimated that achieving universal access to electricity efficient LED lighting and direct current (DC) appliances. would have taken 30 years. Consumers benefited quickly from cost reductions due to increased appliance efficiency, price declines of solar Social impacts in rural families: Surveys conducted by modules, and the economies of scale of the program. BIDS found positive, statistically significant impacts on rural families using SHS: Solar enterprise development and employment: The program contributed to the development of the solar PV Brighter lighting allowed children to study longer hours. industry, including SHS retailers, service providers, Boys and girls with solar lights studied 10–12 minutes financiers, and manufacturers. The program led to backward per day longer on average than those without. These integration of the industry with Bangladesh extending few minutes per day sum to an additional 50–60 hours manufacturing from deep-cycle batteries and other per year of potential study time (based on 340 days of components, including in later years, to solar PV module SHS electricity availability per year). Importantly, surveys manufacture. At its peak in 2015, the POs had about 29,000 found that, especially in women-headed households, a staff in their SHS operations. There was, in addition, indirect strong motivation for acquiring an SHS was to improve employment created in the SHS supply sector and those children’s education. using the electricity available from SHS. SHS households enjoyed greater safety, comfort, and Environmental management: convenience compared to non-SHS households. Bright electric lighting afforded a greater sense of security. The SHS Program mandated that all battery SHS households had easier and lower cost access to TV, manufacturers adopt international standards for radio, fan, and mobile phone charging. battery manufacture and those facilities were regularly inspected by IDCOL. Four battery recycling centers were Though it cannot be directly attributed to SHS, supported, and all participating battery suppliers had to SHS households suffered less from several types of send their spent batteries for recycling. preventable illnesses such as general ailments, respiratory diseases, and gastrointestinal illness and reduced risk of The global environment was improved by the reduced fire. Households with SHS had lower fertility. kerosene combustion due to the reduction in CO2 and black carbon emissions. The CO2 emissions avoided SHS had a positive influence on women’s mobility, between 2003 and 2021 by kerosene offset by the SHS general and economic decision-making, and sense of are estimated at 9.6 million tCO2. security. Women spent more time tutoring children, LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xxi EXECUTIVE SUMMARY ECONOMIC AND FINANCIAL ANALYSIS The financial internal rate of return (FIRR) of the aggregate participating households when only the kerosene/electricity Cost-benefit analysis was conducted for the SHS Program savings benefits are accounted for is estimated at about 17.2 on an economic and financial basis over the period 2003- percent taking into account the grants, an average interest 2054.2 The economic analysis assessed the project from rate of 14 percent for the SHS loans, and the repayment a societal perspective, looking at its net contribution to defaults late at the end of the program. If there had been the country’s economy and considering the global impact no defaults, the FIRR would have been 13 percent since the of CO2 emission reductions. The financial analysis, on the households would have repaid more. If there had been no other hand, assessed the net financial benefits from the loans or grants, the households’ FIRR would have been 14.7 perspective of project participants (SHS users, POs, IDCOL, percent—though it is likely that far fewer households could kerosene dealers and the Government). have afforded an SHS. Before presenting the results, some simplifications and An estimate was made of the net financial benefits gained limitations of the analysis need to be acknowledged. First, from the activities of the SHS Program by the main household benefits are based on a simple measure in both stakeholders—households, POs, IDCOL, the government, the economic and financial analyses—the avoided cost of and kerosene dealers. The estimate showed that the kerosene and in later stages grid electricity for lighting. This undiscounted net benefits of all stakeholders were simple measure of avoided kerosene costs for lighting significant at US$1,702 million in constant 2018 US$. All greatly underestimates the benefits to households. As the main stakeholders benefited substantially, except highlighted in Section 2.9, use of an SHS has many other for kerosene dealers who lost profits because of reduced benefits including: (a) improved quality of life (for example, kerosene sales. On an undiscounted basis, the net financial more hours of study, household work or leisure, increased gain of rural households using SHS was US$1,348 million. safety, and more access to information through radio or Kerosene distributors lost US$47 million. POs gained TV); (b) other immediate financial benefits (for example, US$103 million and IDCOL gained US$54 million.3 The reducing cellphone charging costs or permitting extra hours government had cumulative net benefits of US$200 million of productive activity); and (c) valuable health and education from SHS taxes.4 All previous figures are on an undiscounted benefits in the longer term. These other benefits are excluded basis, in constant 2018 US$. On a cumulative present value from the analyses because they are difficult to estimate and basis discounted at 10 percent to 2018, the total net benefits the avoided costs for lighting alone justify the program in are estimated at US$1,852 million, of which SHS households economic and financial terms. Second, the analyses rest on gained US$745 million. Kerosene distributors lost US$56 several estimates and assumptions, for example, the amount million in profits. POs gained US$310 million while IDCOL of kerosene saved per household, the profits of participating gained US$379 million and the government net benefit was POs, and the losses of kerosene dealers. US$474 million (US$384 million in SHS taxes and US$90 Given these limitations, the results of the economic and million in kerosene subsidy savings). All discounted figures financial analysis must be used with care; they must not be are in constant 2018 US$. confused with the overall electrification and development The societal discount rate of 10 percent in constant terms impact of the SHS Program. The development impact can be is likely excessive from the perspective of individual seen by the fact that 4.1 million households purchased the stakeholders such as IDCOL and POs, as it is equivalent to SHS and were able to receive all the benefits described above 16 percent in current terms with the inflation rate of about 6 through the SHS Program, in advance of the arrival of the grid. percent over 2013–2018. As noted above, the total of IDCOL’s The economic internal rate of return (EIRR) of the SHS net benefit stream in undiscounted constant 2018 US$ Program is estimated at 20 percent in the base case, using is estimated at US$54 million. Since IDCOL’s opportunity the avoided cost of kerosene/grid electricity for lighting cost of capital is estimated at about 2.5 percent in to estimate benefits. When the additional benefit to the constant terms, the net present value (NPV) of the financial global community due to carbon emissions reduction is benefits from its perspective would more appropriately be added to the base case, the EIRR increases from 20 to 25 estimated at about US$139 million constant 2018 US$ when percent. Using an alternative approach that estimated the discounted at 2.5 percent. Similarly, the NPV of the POs net willingness to pay (WTP) of US$2.23 per kWh in 2018 US$ for gains would be estimated at US$262 million in constant the benefit calculation results in an EIRR of 51 percent. 2018 US$ discounted at 2.5 percent. 2 The economic analysis and the financial analysis of aggregate households extend from 2003 to 2029 when the last SHS installed in 2018 are assumed to stop operating. The financial analysis of stakeholder net benefits extends to 2042 when IDCOL makes the final repayment of loans to the government. The analysis of the impact of ODA financing on the government extends to 2054 when the Government repays the final concessional loan for the program. While current dollar figures are of limited value in assessing such a long term program, it is noted that IDCOL’s total net benefit in undiscounted current dollar terms was slightly negative at US$13 million (see Table E1 in Appendix E). The government was expected to benefit also from reduced kerosene subsidies due to reduced kerosene use for lighting. However, the kerosene subsidy effect of reduced kerosene use was actually to reduce government revenues slightly in undiscounted constant 2018 US$ because the official “subsidized” price of kerosene was lower than the kerosene supply cost over several years (see Section 5.3 and Section D.8 in Appendix D). xxii In addition to the net benefits gained by the program from SHS sales (Figure 2). When the SHS Program was launched SHS taxes and kerosene subsidy impacts, the Government in 2003, there were about 15 million unelectrified rural benefits from on-lending concessional funds to IDCOL households and rural electrification rate was under 27 on less favorable terms. From 2003 through 2054, the percent. The number of unelectrified rural households Treasury’s forecast net gain on IDCOL payments minus declined slowly to about 13 million by 2013. Then, the ODA repayments is US$1 million in constant 2018 US$ on pace of grid electrification accelerated and by 2018 over 80 an undiscounted basis and US$180 million in constant 2018 US$ when discounted to 2018 at 10 percent. On a cumulative present value basis discounted at 10 percent to 2018, the Treasury’s total net gain from the SHS Program 2012–2015 Cost of credit to POs increased by 1 percent was US$655 million, made up of US$384 million from and loan tenor dropped by 1 year, reducing taxes on SHS, US$90 million from savings due to avoided affordability of SHS to customers and/or POs kerosene subsidy, and US$180 million due to impact of ODA profit margins. pass-through. All are in constant 2018 US. 2013, 2015 The trend toward smaller systems sold to customers in more distant areas increased INVESTMENTS IN AND FINANCING OF SHS the cost of doing business and reduced POs’ PROGRAM margins. To reduce losses, operating costs, Total investment in the SHS Program during 2003–2018 is and overhead, POs lowered loan tenor which estimated at US$1,095 million (in current US$), to provide made the SHS less affordable. electricity services to about 20 million people, or about US$266 per household. Credit support came from four 2014 Declining SHS prices led to a situation where development partners among which the World Bank (IDA) customers could get a new SHS that cost less provided US$416 million in IDA credits or 69 percent of than the balance due on their old SHS loan. the total international credit support of US$602 million. Some customers abandoned paying for the Other credit financiers were the Asian Development Bank old SHS and got a new lower cost one. (ADB), Japan International Cooperation Agency (JICA), and Islamic Development Bank (IsDB), which provided US$185.6 2015 Political unrest shut down rail, road, and river million. Grant funds amounting to US$80.9 million were transport; reduced rural incomes; and led to received from the Global Environment Facility (GEF), Global a drop in demand for SHS. Partnership on Output-Based Aid (GPOBA), United States Agency for International Development (USAID), Kreditanstalt 2015 BREB began accelerating its pace and für Wiederaufbau (KfW), German Agency for International began connecting about 200,000–300,000 Cooperation (Deutsche Gesellschaft für Internationale customers monthly. Zusammenarbeit, GIZ), and UK Department for International Development (DFID). In total, development partners 2015–2016 The expectations of getting a free SHS provided US$683 million in loan and grant funds. The funds through TR/KABITA dampened demand for from the World Bank and other development partners SHS under the SHS Program. IDCOL took over leveraged private funding from users, POs, manufacturers, the TR/KABITA Program and integrated it into and distributors. Down payments by SHS customers, the SHS Program PO network, which helped equity investments by POs, and upstream investments by give alternative business to POs. manufacturers and distributors are estimated at US$412.5 million, of which user contributions for down payments for Private SHS sales increased, building on the 2015–2016 SHS are estimated at US$160 million up to December 2017. good reputation created through the SHS World Bank financing leveraged 161 percent more financing Program. They could sell at lower prices by from other sources. Moreover, the Rural Electrification and limiting after-sales service and warranties Renewable Energy Development (RERED) Project leveraged and selling through retail outlets selling the capabilities of the microfinance sector that Bangladesh many other products. pioneered and the nongovernmental and private sector capabilities to manufacture, distribute, finance, and service 2015–2018 As sales declined, POs began shutting solar and other clean energy products directly to the rural down sales and service centers (those not communities. engaged in TR/KABITA). This led to a decline in customer service and hampered debt collection. RESPONDING TO A DECLINING SHS MARKET After 11 years of sales growth, SHS sales began to decrease 2017 Devastating floods affected 32 districts of in 2014 as the market became saturated as a result of the the country, hurting SHS sales and further increasingly rapid pace of grid electrification coupled with hampering debt collection. competition from TR/KABITA Program and commercial LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xxiii EXECUTIVE SUMMARY percent of rural households had access to electricity. Among the quality of the POs’ loan portfolio with IDCOL, with the remaining unelectrified households, the expectation of below-standard debt reduced from BDT 11.9 billion in 2018 getting grid electricity rose and increased their reluctance (US$143 million) to BDT 2.4 billion by 2019 (US$28.6 million to invest in SHS. A series of factors compounded the market in 2018 US$). This is exceedingly small compared to the NPV challenges. of benefits accrued to the main stakeholders. The below- standard debt is now only about US$7 per SHS installed Among these events, the acceleration of grid expansion under the program. affected the SHS Program most significantly, while the TR/KABITA Program had a lesser impact. The impact of the TR/KABITA Program was mitigated by IDCOL taking CONCLUSIONS over the management of that program and using the PO This review of the Bangladesh SHS Program over 2003–2018 infrastructure to supply and service the systems supplied leads to several main conclusions about carrying out large- under the TR/KABITA Program. This retained the field service scale off-grid electrification programs in the long term: infrastructure and increased income for the POs which could help repay outstanding debts to IDCOL. Households value SHS highly and are willing to pay for its services; the sale of 4.1 million systems in a target market In 2011, IDCOL had estimated that the market for SHS was of 15 million rural households without electricity at the about 6 million households or about 50 percent of the start of the program indicates both the acceptance of the unelectrified rural households. At that time, the pace of grid SHS and the high value that households placed on the electrification was slow. Consequently, the government services obtained. sought additional financing for SHS, and the development partners responded with US$377 million in credits and The SHS Program was economically justifiable from the grants between 2012 and 2014, enough to finance an national and global perspectives, with an EIRR of 20 additional 2.7 to 3 million SHS. However, in 2015, BREB percent without considering global emission reduction began rapidly accelerating its grid electrification efforts benefits, and 25 percent with them, based only on with financial support from the government and many of benefits from savings in kerosene/grid electricity costs for the same development partners. The prospect of getting a lighting. grid connection soon dampened demand for SHS and led Households benefited substantially from the program on to some SHS customers defaulting on their loan payments. a financial basis, with an FIRR of 17.2 percent considering Consumers preferred grid electricity supply with the promise loan defaults by households to POs and 13 percent of unlimited access to electricity at subsidized low prices. if there had been no defaults based only on savings Better coordination between grid and off-grid electrification in kerosene/grid electricity use; the best evidence of planning could have lessened the problem to IDCOL and the benefits is the marketplace. POs caused by a sudden disappearance of their market. This The Government of Bangladesh (GOB) as the financier, coordination was necessary not only among government IDCOL, and the POs also reaped significant net financial agencies but also with development partners who were benefits from the program despite late-stage problems. simultaneously increasing financing to both grid and off-grid The SHS Program succeeded from 2003 to 2014 based on electrification. an implementation model including strong leadership Mainly because of the rapidly expanding grid, the SHS from IDCOL, POs with strong on-the-ground presence, market disappeared, the POs business profitability declined, a flexible and collaborative approach using the OC and debt collection suffered, leading to financial difficulties. an effective framework for controlling quality, and The POs’ inability to service their debt to IDCOL in turn enforcement of financial discipline of all parties. affected IDCOL’s financial position. The government, From 2015 onward, the program was hit by a perfect recognizing the positive development impacts of the SHS storm caused mainly by sudden and rapid grid expansion Program as well as its fiscal and other benefits, agreed to a that increased household connections by 280 percent proposal put forward by IDCOL to overcome the financial in five years; unintended consequences were the rapid difficulties by: (a) reducing the interest charges on IDCOL’s shrinkage in markets for SHS and defaults by some SHS loan with the government from 3 to 0 percent, effective from households on debt repayments. July 1, 2018, with IDCOL, in turn, waiving interest on SHS The damage to SHS Program sales from increased loans to POs, and (b) allowing a 10-year time for IDCOL to availability of the grid was compounded by the expansion build up a provisional amount for repayment from its future of the TR/KABITA off-grid program that provided SHS to revenue earnings. IDCOL will seek an exemption from the households at no cost and the expectations created of Bangladesh Bank from mandatory provisioning requirement getting a free SHS. for SHS loans. The sudden drop in SHS sales and reduction in collection IDCOL has also worked with the POs to restructure their rates of POs after 2015 created financial and operating debt and help them recover outstanding loan arrears from difficulties for IDCOL and the POs; the shrinking sales and customers. IDCOL extended to debt repayment from 2023 drop in collection rates meant that some of the POs were to 2026 in addition to waiving interest on SHS loans to POs. unable to fully repay their loans to IDCOL. IDCOL’s proactive efforts have succeeded in improving xxiv Recognizing the contribution made by the SHS Program Developing sustainable institutions to the GOB’s rural electrification goals as well as the Have a lead agency such as IDCOL to provide close and financial benefits reaped from the program, the GOB timely supervision as well as financial discipline. restructured its loans to IDCOL and supported IDCOL in restructuring its loans to POs in mid-2018. The GOB may Build on the strengths of existing organizations and need to further assist IDCOL and the POs as required to enterprises rather than creating new ones, where bring the program to an orderly end and ensure the long- possible. term sustainability of these organizations as well as the Provide responsive management taking advantage of SHS installed under the program. technological and business innovation. Better planning and coordination of electrification Ensure that participating businesses generate revenues to could have avoided the late-stage difficulties in the SHS cover costs and provide adequate returns. Program. The GOB was accelerating three major parallel efforts without such coordination: expanding the grid, Providing quality products and services promoting SHS under the SHS Program, and providing Ensure well-designed products, quality components and systems at no cost to the poorest households and public installations, and support services, for sustainability. institutions under the TR/KABITA Program. Adopt new technologies that offer better quality and In conclusion, the SHS Program made a significant more reliable services, for example, LEDs and flat-screen contribution to the government’s principle in the TVs and improve business practices such as pay-as-you- Constitution to transform rural areas by providing, among go (PAYG) technology, mobile pay, and computerized other elements, rural electrification. It provided electricity management information systems (MISs). in advance of the availability of the grid to around 20 Provide users with solid, practical information and million people through the provision of 4.1 million SHS. training in simple maintenance and safe operating It provided electricity service that was adopted by rural procedures. households cost-effectively and with net benefits to all participants except kerosene dealers while also reducing Overcoming the first cost barrier kerosene consumption by 4.4 billion liters and reducing greenhouse gas (GHG) emissions by 9.6 million tons. Do not try to compete with the grid’s promise of unlimited power at low tariffs. Provide access to finance on affordable payment terms— LESSONS LEARNED this is essential—and offer credit terms that approximate The experience with the SHS Program in Bangladesh, one household expenditure patterns. of the largest and most successful in the world despite Use grants and subsidies smartly, to build market challenges, offers lessons that may help other countries infrastructure or reduce capital costs of SHS to users. implement off-grid programs to complement grid electrification. Key lessons are summarized below. Rationalize duty and tax structures to level the playing field for SHS and alternatives. Planning the SHS Program Essential government and development partner support Design the program in relation to a clear goal such as deepening access or reaching universal access. The government needs to ensure that grid and off-grid electrification are promoted and coordinated to optimize Recognize that rural families value highly the access. electrification benefits of SHS. See the government and the private sector as Ensure coordinated planning of on-grid and off-grid complementary not as alternatives or competition. electrification, at the highest levels. Ensure that the government and development partners Evaluate the impacts on key stakeholders as well as the integrate new resources using coordinated approaches— overall economic returns when planning and justifying rather than running parallel initiatives with same/similar the program. objectives. Build an off-grid service and spare part supply Seek development partners support for technology and infrastructure that continues beyond the program. knowledge transfer and not just as source of money. Be flexible in implementation modalities while adhering Use development partner financing to leverage domestic to sound economic, technical, and business principles. financing to maximize the funds available. Have an exit strategy from the start, together with market monitoring, to adapt responsively and to share and As 2021 dawns and the GOB achieves its goal of bringing manage risks as the program winds down. universal access to electricity, IDCOL and its partners can be justly proud of their contribution toward achieving this vision. It is hoped that these successes and experiences from Bangladesh will spur and guide other countries to achieve the same goal. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | xxv INTRODUCTION 1.1 Overview 2 1.2 Evolution of Access to Electricity in Rural Bangladesh 3 1.3 View of SHS as a Rural Electrification Option 4 1.4 Benefits from Electrification 5 1.5 Early Support for Solar Home Systems in Bangladesh 5 1.6 Purpose of Report 6 1.7 Structure of the Report 6 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 1 INTRODUCTION 1.1. OVERVIEW The Bangladesh Solar Home Systems (SHS) Program is one of the largest national off-grid electrification programs.5 The SHS Program is a public-private partnership, led and implemented by the Infrastructure Development Company Ltd (IDCOL), with partner organizations (POs) that market, sell, finance, and service the SHS to households throughout Bangladesh. Some POs also assemble SHS from components sourced domestically and internationally. Box 1 has a brief description of a typical SHS used in Bangladesh. BOX 1: Solar Home System A typical solar home system in Bangladesh comprises a 10–300 Wp photovoltaic (PV) module (or two or more modules referred to as an array) mounted on a roof or a pole, tilted toward the south and facing the sun; a 12 V rechargeable battery for energy storage (mainly tubular plate lead-acid battery); a charge controller; and several lights (initially fluorescent tube lights [FTLs], later compact fluorescent lights [CFLs] and light-emitting diode (LED] lights). Photovoltaic Control Module/Array box LED lights Connection wires and switches Radio Battery Television Figure 3: Schematic of an SHS The SHS may also have power outlets for a television, radio, fan, mobile phone charger, or other low power-consuming appliance; switches; interconnecting wires; and mounting hardware (see Figure 3. Some may include an inverter to convert 12 V direct current (DC) electricity to 220 V alternating current (AC) to operate AC appliances. Both the solar module capacity and the availability of sunlight determine the amount of electricity available for daily use. There is seasonal variation in electricity available due to variations in the amount of sunlight. In Bangladesh, on average, about 3.5 Wh of electricity would be available daily for use from 1 Wp of solar PV capacity. For example, a 50 Wp SHS would generate enough electricity to operate four 5 W LED lamps (each equal to a 40 W incandescent lamp) for 5 hours a day and provide enough electricity to operate other appliances such as a 15 W TV for 5 hours a day. 2 The SHS Program began in 2003 with funding support from 16. Rural development and the World Bank under the RERED Project and continued agricultural revolution with additional tranches of financing from the World Bank and other development partners, extending to the RERED II The State shall adopt effective measures Project. Sales under the SHS Program ended in 2018. to bring about a radical transformation Cumulatively, 14 percent of the Bangladesh population (as in the rural areas through the promotion of 2011 Census), about 20 million people, or more than 25 percent of 15 million households that were unelectrified in of an agricultural revolution, the 2003, obtained electricity services through the SHS Program, provision of rural electrification, the some as early as 2003. The Program helped a significant development of cottage and other share of the rural population to obtain electricity services far sooner than would have been possible with grid electricity. industries, and the improvement The program led to SHS becoming a credible electricity of education, communications and source for off-grid communities. It has contributed to the public health, in those areas, so as development of the industry from service providers and progressively to remove the disparity financiers to manufacturers. It has convinced the government to support SHS and solar PV systems for public services to in the standards of living between the benefit the poorest. From a social and economic perspective, urban and the rules areas.” significant benefits accrued to the people and the country. In the early years, Bangladesh faced a tremendous 1.2. EVOLUTION OF ACCESS TO ELECTRICITY IN electrification challenge. To address this challenge, the government’s policy toward power sector development RURAL BANGLADESH was articulated in the Vision and Policy Statement on Bangladesh accords a high priority to providing electricity Power Sector Reforms in 2002 (Power Cell, n.d.) and then access to the population. The provision of rural access reinforced in the Perspective Plan of Bangladesh 2010–2021: to electricity is enshrined in the 1972 Constitution as a Making Vision 2021 a Reality (General Economics Division fundamental principle of State policy (GOB 1972): of Planning Commission 2012). The government was 5 In aggregate, global commercial sales of off-grid solar (OGS) products are much larger today than sales in any single program. The World Bank Group’s Lighting Global Program reports that between 2010 and 2019, about 180 million OGS units were sold worldwide. They comprise 150 million pico (less than 10 Watt) products and 30 million larger products. Between 2016 and 2018, Lighting Global estimates that total OGS commercial sales were about 25 million units in India, 4.4 million in Kenya, 3.5 million in Ethiopia, and 3.5 million in Uganda. The Lighting Global Program has facilitated the sale of 42.1 million OGS products by its affiliated companies, benefiting 52.4 million people in 60 countries as of August 2020. Lighting Global works with manufacturers, distributors, governments, and other development partners to build and grow the modern OGS energy market (Lighting Global et al. 2020). LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 3 1. INTRODUCTION 90 rate had risen to 61 percent and rural electricity access to 41 percent. The pace of grid electrification began 80 accelerating by 2014–2015. In the past four years, the pace Percent of Rural Electricity Access 70 of connections was about 300,000 a month. Consequently, rural electricity access had reached 70 percent by 2016 and 60 was estimated to be about 80 percent by 2018. The trend in rural electrification rate is shown in Figure 4. 50 In the early years, an electricity connection, especially 40 in rural areas, did not always mean access to electricity. Supply shortages led to brownouts and blackouts with 30 street protests against BREB. Bangladesh has increased 20 its generation capacity, and outages are now significantly reduced. Power generation capacity was 23,548 MW by June 10 2020 with an additional 1,160 MW of imports (Haque, 2020). Power generation capacity had risen sharply from 16,000 0 MW in 2018 due to the efforts of policymakers, public sector 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 investments, private sector participation, and support from international development partners. Generation capacity Year is expected to double in the next five years. Electricity Source: World Bank. consumption per capita was 510 kWh in 2020 (Haque, 2020), up from 375 kWh/capita in 2014 (IEA 2014). Access Figure 4: Trends in Rural Electricity Access 1995–2018 to electricity was 97 percent country-wide by 2020 (Haque, 2020). committed to achieving universal access to electricity by 2021 and to making the power sector financially viable, 1.3. VIEW OF SHS AS A RURAL improving sector efficiency, enhancing power system reliability, and making electricity service affordable. ELECTRIFICATION OPTION This led to the government’s acceleration in the pace of The traditional practice of providing grid-supplied electricity electrification and increased power generation as evidenced to rural areas is through a public utility that has a natural by the tenfold increase in government budget for the monopoly for supplying electricity while paying for power sector from BDT 26.8 billion in 2009 to BDT 262.9 household electricity service using lifeline tariffs supported billion in 2018 (Economist Intelligence Unit 2018). The by government subsidies. SHS-supplied electricity is a policy commitment backed by significant budget increases relatively late arriver as a viable alternative for avoiding the accelerated the pace of electrification, backed by generation high marginal cost of extending the grid to remote and/or capacity increases, beginning in about 2014–2015 and sparsely populated areas. Thus, SHS electricity is viewed continuing today. largely as a private good.6 In Bangladesh, the SHS Program permitted the government to leverage the capacity of the The Bangladesh Rural Electrification Board (BREB) was private and NGO sectors to use SHS to accelerate electricity founded in 1977 to help implement this principle of State access and ease pressure on grid-based electrification by Policy. In its early years, rural electricity access grew slowly. BREB. By 2000, only 20 percent of the rural population had access to electricity. When the Rural Electrification and Renewable In spite of its long list of potential benefits—including Energy Development (RERED) Project was launched in 2003, rapidly improving prospect for managing the cost and the national electricity access rate was 37 percent and the accelerating the spread of electricity access at the margins rural electricity access rate was 27 percent, with over 15 of the electrification market—SHS electrification is not fully million rural households without access to electricity. The integrated into systemwide planning for electricity provision urban-rural disparity and the health and safety impact on in Bangladesh, though such integration is beginning to the rural population, especially women and children, were occur in some other countries. The list includes Rwanda, great. Until about 2010, BREB was connecting about 500,000 Kenya, Mozambique, Nigeria, South Africa, Tanzania, consumers annually; though this was impressive, the World Uganda, and Myanmar—where average nationwide Bank estimated that even at that pace it would take over population densities tend to be much lower than that of 30 years to achieve universal access. After 2011, the pace Bangladesh. of rural electrification accelerated coupled with increased investment in generation, with about 125,000 households Bangladesh is not alone in according grid electricity obtaining electricity connections monthly. When the RERED significant advantages not afforded to off-grid options—for II Project was approved in 2012, the national electrification example, capital and operating subsidies and concessional 6 Merit and demerit goods concepts were codified in Musgrave’s (1959) classic ‘Theory of Public Finance’. Within that theory, merit goods are good for individuals and society and will be undersupplied by private markets and, thus, should have their production and/or consumption subsidized (encouraged). Conversely, demerit goods are bad for individuals and/or society and should be taxed (discouraged). 4 finance. Off-grid programs do receive grant support and 1.5. EARLY SUPPORT FOR SOLAR HOME concessional financing but are expected to graduate to SYSTEMS IN BANGLADESH full commercial operations without subsidies within a few years—partly the case in Bangladesh. In such situations, In 1996, Grameen Bank founded Grameen Shakti, which SHS dissemination is market driven—much like the sale embarked upon an SHS project with the installation of 20 of batteries and small generators—mostly divorced demonstration units. They were planning a second phase from government-funded and government-directed to begin testing their market and sales procedures. Their electrification projects. The risks are borne by consumers target market was the 350,000 Grameen members who and/or technology and service providers. have obtained housing loans from Grameen Bank as well as upper-income households in the villages. Grameen Bank’s The net result is that many of the rural households that experience in rural enterprise development and banking are the principal beneficiaries of off-grid electricity must and their rural outreach created a good institutional pay substantially more on a levelized cost basis per kWh arrangement for implementing an SHS electrification than urban and peri-urban households that have access to program. Grameen members could obtain loans from grid electricity. Interregional equity—and socioeconomic Grameen Bank to purchase the systems. Grameen Shakti efficiency—issues abound. Nonetheless, grid electricity also expected that microenterprises would be set up with is clearly preferred in Bangladesh and elsewhere for financing from Grameen Bank to sell and service the SHS as the promise it holds of providing unlimited amounts of well as manufacture components. electricity at low prices without the consumer having to invest in and run the supply infrastructure. About the same time, BREB, with assistance from the French government launched a PV pilot project to provide electricity services to consumers on an island on the 1.4. BENEFITS FROM ELECTRIFICATION Meghna River in Narsingdi District (Eusuf, n.d.). The project provided 795 SHS ranging in size from 5 Wp to 92 Wp and A survey by the UK Institute of Development Studies outlines comprised stand-alone SHS and charging stations. This the following positive impacts of SHS (Quak 2018)—most project was designed as a fee-for-service model. The SHS of which simply mirror the private and social benefits from were owned by BREB and the users were expected to pay government-subsidized, grid-provided electricity and, thus, an initial deposit and monthly tariffs for its use. BREB was in a world that is both economically efficient and socially responsible for the installation, maintenance, repair, and equitable would be accorded comparable financial and replacement of the SHS components. By 1998, the grid other advantages: had extended nearby, and the community clamored to More productive and longer work hours at home obtain grid electricity (called big electricity) instead of solar electricity (termed small electricity). BREB decommissioned Improved/increased opportunities for women the PV systems with the intention of installing them in other Better quality reading/study light for longer periods remote locations (Islam 2002). into the evening In 1996, the World Bank explored opportunities Improved health and safety from better indoor air for supporting solar PV in Bangladesh through a quality and reduced dangers from poisonous kerosene reconnaissance mission (Cabraal 1996). It noted, among (often stored in soda bottles) and burns from lamps others, that 85 percent of rural consumers did not have Savings on fuel-based lighting expenditures that may access to electricity, and it was likely that many rural be spent on food for a better, more balanced diet and consumers would not receive electricity services for many nutrition intake years. PV electrification for selected rural households in lieu More time for family to invite friends, eat together, and of grid service could bring electricity earlier and help make share experiences rural electrification more financially sustainable by reducing the pressure on BREB to extend grid service to uneconomic Reduction in greenhouse gas (GHG) emissions7 domestic consumers. With good to excellent solar resource Additional positive information and communication available throughout the country and throughout the year, impacts for the solar systems that provide lighting, there was good potential for PV use in unelectrified rural mobile phone charging or powering radio and TV homes if affordable products meeting consumer needs Renewable and efficient energy creating many times could be supplied and supported. It was recommended more jobs than nonrenewable energy systems do, that support for PV electrification be considered as part of a particularly for non-oil-producing countries least-cost rural electrification plan. The combination of more and better light, access Bangladesh Rural Advancement Committee (BRAC) began to information and communication technology investing in solar in 1997. By the end of 2000, the program (ICT), and awareness of solar technology increasing had installed more than 500 solar PV systems to serve its opportunities of marketing new services and branch offices, a few microenterprises, and government technologies to off-grid populations. offices (Islam 2002). BREB launched the Diffusion of 7 About 110 million tCO2 per year can be avoided by replacing all kerosene lamps with solar (Energypedia 2019). LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 5 1. INTRODUCTION Renewable Energy Technologies Program in 2002 to disseminate 6,000 SHS in remote locations. However, only 605 SHS were installed by April 2006 when the program ended (Marro and Bertsch 2015). Between 1997 and 2002, less than 10,000 solar PV systems were installed in Bangladesh by various public, private, and nongovernment entities. Total installed capacity was under 500 kWp. The World Bank’s first investment in SHS in Bangladesh was approved in May 2002 under the RERED Project that included a component to promote the use of SHS in remote rural areas (World Bank 2002).8 To prepare the RERED Project, the Global Environment Facility (GEF) provided a US$250,000 project preparation grant which included funds for a pilot to install 50 SHS by five POs to test the proposed project model. The RERED Project provided US$16.4 million in IDA funds and US$8 million in GEF grants to co-finance 64,000 SHS and fund related technical assistance, including a 50,000 SHS pilot program to be implemented by IDCOL. This was the dawn of the fruitful, ongoing relationship between the GOB, the World Bank, IDCOL, NGOs, and the private sector to bring solar electricity services to dispersed rural communities. 1.6. PURPOSE OF REPORT The purpose of this report is to draw generalizable lessons from the Bangladesh SHS Program to guide the development and implementation of sustainable solar off- grid electrification programs. It is believed that the lessons derived from evaluating this program would be useful to others embarking on similar off-grid electrification programs where many communities remain unelectrified and where grid expansion is costly and time-consuming. The study has four objectives: (a) Describe the SHS Program, including organizational arrangements, business and financial model, market trends, benefits and costs, financing, technology, and risks and measures taken to mitigate risks. (b) Show how IDCOL and other partners adapted to the changing business and market environment, technological evolution, and other unanticipated events. (c) Assess the costs, benefits, and distributional impacts of the SHS Program. (d) Derive lessons for other SHS programs. 1.7. STRUCTURE OF THE REPORT Within each chapter, the report attempts to draw out insights on why certain decisions were made and lessons learned. Chapter 2 describes the SHS Program in considerable detail, including the key features of the program—program management, finance, technology, market, and regulatory and policy aspects. Chapter 3 discusses how the SHS Program adapted to the reality of the business. An associated Appendix C, using a risk matrix, details how the SHS Program, by retaining a fair degree of implementation flexibility, adapted to changes in technology, financial conditions, changes in the off-grid market, policy and regulatory environment, and force majeure events. In Chapter 4, the impact of declining SHS sales and mitigation actions is discussed. Chapter 5 conducts an economic and financial analysis of costs and benefits, with certain simplifications and limitations, with supporting data in Appendixes D to F. The economic analysis examines the SHS Program from the perspective of the nation and global society while the financial analysis estimates net benefits from the perspectives of the main participants: SHS households, Government Treasury, kerosene suppliers, POs, and IDCOL. In Chapter 6, the lessons learned from the SHS Program are summarized. Several appendixes provide supporting data and analyses. 8The World Bank had previously carried out several preparatory activities that informed the development of the SHS component of the RERED Project. 6 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 7 8 THE SOLAR HOME SYSTEMS PROGRAM 2.1 Design Concept 10 2.2 Implementation Model 12 2.3 Organizations Involved in the SHS Program 13 2.4 Financing the SHS Program 17 2.5 Financing Terms from IDCOL to POs and Consumer Payment Terms 19 2.6 Loan Securitization 20 2.7 Subsidy Trends 21 2.8 SHS Program Results 21 2.9 SHS Benefits 25 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 9 THE SOLAR HOME SYSTEMS PROGRAM 2.1 DESIGN CONCEPT Market creation in rural areas faces issues of remoteness, low customer density, and low beginning asset stocks—typical traits of the population that off-grid SHS would be trying to reach in Bangladesh. In the case of the SHS Program in Bangladesh, many challenges needed to be overcome, including the following: Catalyzing markets for new technology to relatively unsophisticated customers. building capacity and skills to deploy and service the technology, and promote competition in this new market. The need to establish an ‘enforceable’ regulatory framework for quality while promoting innovation. The need for nontraditional actors. Commercial white-goods firms with banking relationships were not interested in the SHS business. It is a high transaction cost business with the need to collect many small monthly payments. Customers are rural, often with seasonal income, and mostly in unelectrified areas. Lenders view products as consumer goods not directly contributing to income generation. SHS was considered to not supply ‘real’ electricity (unlike grid electricity). Challenging financial attributes of the SHS market: • Selling a capital-intensive product to a market that is highly price sensitive with expectations of short payback periods. • Individual loan amounts are small and transaction cost is high. • Customers, whose income is often seasonal and uncertain, needed to make a financial commitment of up to three years to purchase the SHS, compared to flexibility of a traditional alternative such as kerosene (“if I don’t have money, I do without light”). • Limited ability to use the SHS as collateral as repossessing it is difficult in case of default. • SHS cost is declining while performance is improving so there is a risk of obsolescence or asset stranding. • Need for significant and continued capital increase to support a rapidly growing market served by companies with limited collateral to back additional borrowings. • Competition from subsidized substitutes—lighting using subsidized kerosene fuel, expectation of access to subsidized grid electricity, and expectation of getting a free SHS. SHS programs to offer electricity services to unelectrified communities can range from pure laissez-faire to a public sector approach. Commercial sales of SHS components and systems with no government intervention mean that consumers decide what they want from a range of products offered by private companies. Whoever can afford to pay can buy an SHS that can meet their needs—the consumer decides on the quality, type, and level of service based on the available information. Under this model, there is no expectation as to how many households or what percentage of a community will use SHS to gain access to electricity. At the other extreme is a public sector model, like many grid-based electrification schemes including BREB, where SHS products and services are provided to unelectrified customers as an integral part of the country’s rural electrification program, implemented by a public sector agency. The government or public sector agency sets the procurement terms, determines the quality and level of service, selects the consumers to receive the systems, and the government co- finances it. 10 The Bangladesh National Social Safety Net Program (TR/KABITA9) is an example where products, service providers, and customers are selected by a public BOX 2: Bangladesh Solar Home Systems Program - sector agency, and the government bears the full World Bank Support cost of an SHS. There are a few examples of off-grid electrification programs that are for the most part financed and implemented by government agencies: The World Bank’s program to support SHS in Bangladesh in Myanmar, the off-grid electrification program is run began in 2003 under the RERED Project. It evolved by the Department of Rural Development which will from GEF-funded project preparation work and a small electrify about 400,000 households using SHS;10 under pilot project with five POs to supply 50 SHS to test the the First Phase of Jawaharlal Nehru National Solar implementation model. While the RERED Project primarily Mission (JNNSM) in India, 200 MW of off-grid solar PV supported grid extension in rural areas, it had two components to use SHS to provide electricity services to systems were supplied;11 the Provincial Electricity households that were unlikely to be grid connected soon. Authority of Thailand included SHS as a electrification The SHS components were to option in achieving 99.98 percent electricity access by 2006 (Vechasart and Suttisom 2014); and Peru • Support BREB to develop a fee-for-service SHS program to began with a World Bank-assisted activity where serve 14,000 off-grid households and distribution utilities used SHS to provide electricity • Provide IDCOL with project development support and services to nearly 12,000 households that could not financing to offer loans and grants to finance SHS to be economically served by grid extension.12 This was 50,000 households using a competitive, microfinance- followed by a government contract with a private based sales program. company, Ergon, to provide about 220,000 SHS in BREB installed SHS in 11,796 households on a fee-for- isolated areas, including installation, operation, and service basis. It discontinued the program when it realized maintenance of the systems for 15 years. By mid-2019, that procurement of SHS took time and found it difficult to Ergon had installed 134,000 systems. The Government provide maintenance services to these dispersed units cost- of Peru aims to provide 500,000 SHS to households, effectively. It also found that many of the installed units schools, and clinics by the close of the program. were falling into disuse due to lack of interest on the part of users who had no ownership of the asset. The Bangladesh SHS Program was a hybrid version that combined elements of public sector and laissez- The market-based program led by IDCOL succeeded faire approaches to address then-known specific dramatically. The 50,000 SHS were sold within three years requisites and challenges of each. It was conceived with SHS primarily supplied, financed, installed, and jointly by the GOB and the World Bank, initially to supported by NGOs and microfinance institutions (MFIs) which were registered as IDCOL POs. Competition led to pilot test different implementation models for off-grid falling prices, and by utilizing unused loans, the number electrification. Based on the success of the pilot, of SHS installed rose to 236,000 by 2009. Two rounds of the program was scaled up with the World Bank additional finance for the RERED Project and support from continuing to support the program throughout its life, other development partners helped increase the project joined by other development partners (see Box 2). target to 994,000 SHS, and this was exceeded at project completion using further cost savings. By December 2012, 1.88 million SHS were installed, bringing 9 The Bangladesh Ministry of Disaster Management and Relief electricity to 6 percent of the nation’s population. Building launched the TR/KABITA Program in 2014–2015, to bring solar electricity to the poorest communities and households. The on the success of the RERED SHS Program and to maintain program paid 100 percent of the cost of SHS and solar PV systems momentum and continuity, the World Bank approved, for public facilities such as streetlights, schools, and clinics. In its at the government’s request, the RERED II Project in 2012 first year, it was run by local government institutions and supplied and additional financing in 2014 with a goal of reaching 328,000 SHS and other systems. This led to problems as the TR/ KABITA Program did not adhere to quality or service standards, 4 million SHS by 2021. Additional financing from other and beneficiaries had little recourse if the systems failed. The development partners was mobilized to complement government requested IDCOL to take over the management of the financing provided by the World Bank and domestic this program in 2016 and IDCOL utilized the infrastructure built for the SHS Program. The local government officers, and not IDCOL, sources. The goal was exceeded by June 2016, far ahead of selected the beneficiaries and decided what systems they would the target date, with over 4.1 million SHS installed by 2018. get. 10 World Bank, 2015, Myanmar Electrification Project. Project The RERED II Project was expected to end in 2021. It Appraisal Document. is being extended till 2023 to permit completing the 11 Energypedia, 2015. implementation of Improved Cookstoves Program. 12 The World Bank, Peru Second Electrification Project. Project However, loan repayment from the partner organizations Appraisal Document (2011), and the Implementation Completion which were supposed to end in 2023 have been extended Report (2018). As in the TR/KABITA Program, the PV subcomponent till 2026 because of rescheduling of some their loans was compromised by a massive parallel government household solar PV program that threatened to crowd-out the PV subcomponent. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 11 2. THE SOLAR HOME SYSTEMS PROGRAM 2.2 IMPLEMENTATION MODEL institutional capability. The customers repaid the loans to the POs, which in turn repaid their loans to IDCOL. In principle, the IDCOL-led business model is straightforward. But in practice, to ensure success, Working through POs that knew their customers was an considerable attention to detail and good implementation important feature of the SHS Program. The SHS customers’ oversight are necessary. ability to pay was evaluated based on their individual creditworthiness, unlike in a traditional MFI lending model IDCOL mobilized POs that are mainly NGOs and MFIs with where lending to one customer is guaranteed by a group. rural networks and with the experience required to market, The group lending model would not be suitable to the sell, finance, install, and service SHS to unelectrified rural objective of maximizing the number of SHS installed within consumers. In Bangladesh, the NGOs and MFIs were more a community, which in turn imposed the requirement of effective than traditional retail businesses in marketing and lowering the cost of doing business and effectively providing selling to these remote rural customers. spare parts and repair services. Financing was crucial to overcome the relatively high first Quality and reliability of technology, balanced by cost of SHS and to make the SHS affordable to lower- affordability consideration, are crucial. The POs source SHS income households in rural areas. The POs accessed and components from domestic and international suppliers financing from IDCOL to offer loans to their customers to that meet quality and performance standards established spread out the payments over a period of up to three years. by the SHS Program. Affordability was addressed by offering IDCOL sourced the requisite funds through the government, SHS of various capacities with different levels of service, to which sourced them from development partners. These give their customers choices that meet their ability to pay funds are then leveraged by POs’ equity and consumer and paying for the SHS over time. The SHS were backed by copayments. Mobilizing increasing amounts of financing performance warranties to increase the confidence of the was necessary to support steadily rising sales. Scaling up customers in these products. by using traditional financial sector instruments had not proven feasible for reasons discussed in greater detail in IDCOL must repay the loans they obtained from the Section 3.5. For example, IDCOL loans to POs, following government. IDCOL worked with the government the typical practice in Bangladesh of lending to MFIs, were to complement the government’s grid-based rural weakly collateralized, meaning that they cannot be readily electrification efforts led by BREB. IDCOL oversight and securitized and sold back into the financial system (capital close supervision of the overall program implementation markets) that, in the first place, were not as well-developed were crucial to ensure that all parties met their financial and as those in more advanced countries. Therefore, this technical obligations and that customers were satisfied. The demanded a rigorous and regular monitoring of the POs’ government eventually must repay the loans and credits financial performance and debt repayment.13 obtained from development partners. Small grants, declining over time, were given to increase the The details of the SHS Program, its implementation, finance affordability of the SHS and to help the POs strengthen their modalities, technology, and outcomes are described in the following sections. 13 The model IDCOL followed is similar to Palli Karma-Sahayak Foundation (PKSF), an apex development organization set up in 1990 for sustainable rural poverty reduction. PKSF loan recoveries have exceeded 98 percent. PKSF mainly works with poor and ultra-poor and women community to reduce poverty and other vulnerability including climate change risks. It works through partner implementation organizations to reach their beneficiary groups. 12 2.3 ORGANIZATIONS INVOLVED IN THE SHS PO debt collection and repayment to the government. PROGRAM Maintenance of records of SHS sales and issuance of regular reports on performance. The organizations involved in the SHS Program and their roles are depicted graphically in Figure 5. Ensuring of compliance with product quality, meeting service standards, physical verification, and inspections. 2.3.1 IDCOL Monitoring and verification of PO technical, environmental, social, and financial performance. IDCOL is the implementing agency for the SHS Program Procurement audits to verify POs are doing business with on behalf of the GOB. IDCOL was established in 1997 by responsible, reliable, and legitimate vendors and ensuring the GOB. It was licensed by the Bangladesh Bank as a cost competitiveness and proper inventory management. nonbanking financial institution in 1998 initially to finance large private sector infrastructure projects such as power Monitoring of consumer satisfaction and responding to plants telecommunication and ports. Later IDCOL began issues. financing small-scale SHS and other renewable energy and Support for setting up of testing and quality certification energy efficiency projects. Under the SHS Program, IDCOL facilities and product testing. provided grant and loan facility to its POs and manages Support for battery recycling including inspection of the overall program. IDCOL has a broad and crucial set of recycling centers of manufacturers. responsibilities, including Training of trainers, PO staff and technicians, and customer outreach. Program oversight and reporting to the government and the development partners. Promotion and awareness building for all stakeholders. Financial management including taking commercial Conducting of market assessment and other studies. risk for borrowing from government for on-lending to Hosting of the Technical Standards, PO Selection, and PO customers through POs. Operations Committees. PO loan appraisal, award, and supervision. Research and development to introduce new renewable energy technologies. Provides foreign currency grants and loans on soft terms Government of Development Partners Bangladesh Repays loans Provides foreign currency grants Repays and loans on loans soft terms Independent PO Independent Selection Committee IDCOL Technical ap App Standards Ap po lie Committee pr in s fo ov tm r es en PO t Provides Repays s grants and Seeks Provides loans loans approval approval Seeks operational Supplies interventions equipment IDCOL Operations Partner Committee Organizations Suppliers Provides Pay for oversight, equipment guidance & solutions Sells & Pays for SHS service SHS Customers (Housholds) Source: Based on Monirul (2019). Figure 5: SHS Program Functional Relationships and Roles LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 13 2. THE SOLAR HOME SYSTEMS PROGRAM SHS Program Unit Head Relationship Monitoring Unit Management Unit Regional Training and Collection Call Centre efficiency Officers Monitoring Team Inspection team Zonal Senior Manager Auditor Divisional Field Auditor Manager Accounts & Regional Adminstration Manager Deputy Regional Senior Technical Technical Manager Inspectors Inspectors Figure 6: SHS Program Organization at IDCOL The organizational requirements and costs were relatively Historically, NGOs and MFIs have played an important role modest for the SHS Program size.14 IDCOL established an in Bangladesh rural communities. Their local presence in SHS Division to manage the SHS Program, headed by a vice the communities and their ability to provide small loans to president who reports to the Head of Renewable Energy. At consumers, mobilize the communities, and manage local IDCOL headquarters, the division is staffed by seven officers labor forces made them the logical partner for IDCOL to comprising relationship and monitoring managers. During implement the SHS Program. Among them was Grameen 2003–2009, IDCOL conducted inspections of SHS from its Shakti, formed in 1996 to support clean energy and an SHS head office. With the increased number of SHS installations, pioneer in Bangladesh (Box 3). 2 regional offices were set up in Sylhet and Khulna in May 2010. Later, regional monitoring offices increased to 17. POs were selected by the independent PO Selection Each regional office is headed by a regional manager. A Committee on behalf of IDCOL using specific selection monitoring team of 172 diploma engineers and 46 regional criteria (for details see Appendix B): managers/divisional manager/zonal managers was based Legal registration in the regional offices (for all renewable energy programs). In total, the field force numbered 218 employees. The SHS Acceptable business plan Program’s management organization structure is shown in Satisfactory prior operational and financial performance, Figure 6. including in solar business if relevant Transparent and sound accounting, management 2.3.2 Partner Organizations information system (MIS), and internal audit system Currently operational with credit from selected domestic and Central to the implementation are the POs. POs are mainly international sources, a minimum number of beneficiaries NGOs, including MFIs (see Appendix A). and equity, and acceptable financial performance. 14 Total direct SHS Program management costs to IDCOL in 2007–2018 was US$6.6 million for 4.115 million SHS, less than 1 percent of total investments, or about US$1.60 per SHS. 14 BOX 3: Microfinance in Bangladesh and Its Role in SHS The microfinance concept was born in Bangladesh soon after the country gained independence and was designed to support health, education, agriculture development, and food security. In 1972, BRAC (formally known as Bangladesh Rehabilitation Assistance Committee) was founded by Sir Fazlé Hasan Abed. Initially, it focused on village development programs and vocational training for women. BRAC’s microfinance program began in 1974. In the mid-1970s, Prof. Mohammad Yunus and his team at Chittagong University began their ‘Jobra’ experiment to provide loans to poor households. Both used a solidarity-group-based finance delivery model where the group vouched for each other to guarantee repayment. Grameen Bank was formed in 1983. Originally, the loans were given to women- owned small businesses. The business model of group-based lending for small loans with exceedingly small weekly installment payments had wide appeal. Thus, even households that did not own businesses could use such loans for other purposes such as marriage, housing, and so on, and these became popular. Grameen Bank founded Grameen Shakti in 1996 as an independent enterprise to sell SHS, using the microfinance model. Recognizing the importance of consumer confidence, Grameen Shakti set up service centers and trained technicians, both men and women, to staff these centers. Grameen Shakti was initially financed by the International Finance Corporation (IFC) (US$100,000 concession loan), United States Agency for International Development (USAID) (US$1.5 million grant), Kreditanstalt für Wiederaufbau (KfW), and German Agency for International Cooperation (Deutsche Gesellschaft für Internationale Zusammenarbeit, GIZ) (previously GTZ). Applying this concept of microloans for financing SHS was a logical extension. The SHS was capital intensive but with loans, it could be affordable. The repayment could be made in small monthly amounts; the savings of households in the cost of buying kerosene and recharging batteries could go toward loan repayment. SHS financing did not use the traditional group lending model. In IDCOL’s view, the credibility of MFIs in Bangladesh, a stringent qualification process for selecting POs, and strong customer demand and willingness to pay (WTP) would offset the security that the group lending model would offer. In 2003, Grameen Shakti and BRAC became two of the five founding POs in the SHS Program. By 2005, Grameen Shakti accounted for 66 percent of SHS installations and BRAC 23 percent (in total 53,000 SHS). By 2010, Grameen Shakti continued to hold market share (63 percent), but BRAC share declined to 8 percent of over 705,000 SHS. BRAC ended its participation in 2013. By 2018, Grameen Shakti market share was 39 percent of 4.1 million SHS. Source: Wikipedia. 2012. BRAC. https://en.wikipedia.org/wiki/BRAC_(organization); Grameen Shakti 2012. Entry into the program was relatively easy with low entry instalments from the households, troubleshooting, and barriers. Starting with 5 POs, the SHS Program had 30 POs by training the households about the proper usage were also 2010, 46 by 2013, and 57 by 2015. Despite the large number the POs’ responsibilities. The POs were also responsible of POs, the SHS market was moderately concentrated with for providing after-sales services. The POs carried the the Herfindahl–Hirschman Index (HHI) at 1,845.15 commercial risk for loans they obtained from IDCOL by submitting disbursement applications to claim grants The POs obtained credit refinancing and grants from IDCOL, and refinancing from IDCOL. The POs were responsible for procured SHS from suppliers, participated in the monthly regular payment of interest and repayment of the loans. OC meetings and, as the principal customer-facing entity in the SHS Program, sold, financed, and serviced the SHS to households and other customers. 2.3.3 Technical Standards Committee POs signed Participation Agreements (PAs) with IDCOL The independent TSC has the following responsibilities: that laid out the roles and responsibilities of both parties. Set technical standards for solar system components (TSC The POs identified and qualified potential customers 2017). The standards cover individual components and the in their service area and informed them about SHS and system, certification requirements, installation practices, guided them in selecting the SHS model that matches documentation, packaging and delivery, and warranties. their requirements and affordability. The POs supplied SHS or components approved by the independent Technical Review and update the standards from time to time to Standards Committee (TSC), installed the systems, and ensure quality and consistency, introduce new technology, prepared Loan Agreements with the households. Collecting and support continuous improvement. 15 The HHI is a measure of market concentration. A score of 1,501–2,500 implies a moderately concentrated sector. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 15 2. THE SOLAR HOME SYSTEMS PROGRAM Approve products from manufacturers/suppliers based While all households that purchased an SHS did not have on these specifications. Suppliers, both domestic and access to electricity, there were significant differences international, that wish to have their products approved for among the characteristics of such adopter households use in the SHS Program must have a type-test certificate from according to a survey completed 10 years into the SHS an accredited testing and certification organization. For local Program (Asaduzzaman et al. 2013). Among the SHS users, products, a certification from a TSC-authorized institution nonagricultural occupations of household heads appear was acceptable. The approved products are posted on the to be much more prevalent than either the self-farming or IDCOL website and regularly updated (IDCOL 2020). worker category. SHS users had significantly higher financial Periodically monitor quality. and nonfinancial asset ownership and, in general, had better financial status and improved food consumption The TSC comprises members from engineering universities than non-adopters. These characteristics may be due to and representatives from the Power Cell, BREB, and Local the customer selection process where customers with Government Engineering Division. higher and more stable and predictable incomes may more easily obtain SHS loans from POs than nonagricultural households, and may have a greater appreciation of the 2.3.4 PO Selection Committee SHS benefits. An independent PO Selection Committee had the responsibility for selecting the POs. The committee A higher proportion of adopter households were female consists of representatives from the Ministry of Finance’s headed and often had more educated women, implying Economic Relations Division (ERD), Bangladesh Institute a greater role of women and especially educated women of Development Studies (BIDS), NGO Affairs Bureau, and in decision-making. More than 40 percent of households PKSF. PKSF is an apex NGO established to provide financial had a secondary or higher level of education compared assistance and institutional development support to to non-adopters with only half that. Around 70 percent create productive employment opportunities for the of households had at least one woman with primary moderately and ultra-poor, small, and marginal farmers and education compared to 60 percent among non-adopters, microentrepreneurs and to provide associated services. and about 20 percent of adopter households had at least one woman with secondary education compared to 10–12 percent among non-adopters. Adopter households on 2.3.5 Operations Committee average spent almost 50–80 percent more for children’s education than the non-adopter households, implying a An organizational element of singular importance to the greater appreciation of the role improved lighting (and success of the SHS Program was the Operations Committee. possibly, access to better communication) can play in their It permitted IDCOL to obtain timely information from the children’s education. field, get suggestions from the POs, convey consistent messages and directives to the POs, and efficiently manage Women played an important role in the decision to acquire the Program. It permitted the POs to learn from each other’s an SHS. The role of women in deciding to acquire an SHS experiences. The OC is chaired by the Chief Executive and the importance of SHS for women are illustrated by the Officer (CEO) of IDCOL and consists of representatives following quotes (Razzak, Mamun-ur-Rashid, and Biswas from all POs and IDCOL. The OC met regularly to look after 2012): the operational aspects of the SHS Program. It reviewed progress in SHS installations, implementation status of Halima Aktar, an assistant teacher in Gazipur, observed the decisions taken, collection efficiency and ‘portfolio the greater interest among women in obtaining an SHS: at risk’ (PAR) reports submitted by the POs and IDCOL inspectors, technical reports submitted by POs and IDCOL In this area, women are more technical inspectors, periodic submission of financial and other reports by the POs, and any other issues related to upfront about installing solar home the implementation of the program. In addition to the system, biogas plants or improved requirements in the PAs, decisions made in the OC meetings were binding on the POs. cooking system. They keep pressuring their husbands about availing these 2.3.6 The SHS Customers - Rural Households technologies. Moreover, in most of the Households were the principal customers for SHS. The households, the men live abroad, and households made decisions on whether to purchase an SHS, therefore, the women have to take what system to purchase, what payment terms to accept; learned about how best to take care of and use the SHS; decisions regarding these technologies, learned its benefits and limitations; paid for the product; to make their lives a bit easier. and contacted the POs or IDCOL in case of problems. Women are the beneficiaries of these Householders signed a Sales/Lease Agreement with a PO and paid a down payment. Then representatives from the PO technologies and these serve their installed an SHS and trained the customer in its use. needs.” 16 The appreciation of the benefits of SHS by women is further 2.3.7 Suppliers reinforced by observations they made during focus group Suppliers provide the POs, on a competitive basis, with discussions. products and components that are quality verified and Muktilia Bhrumo, a female adopter, observed, approved. The POs sign contracts with equipment suppliers that specify the obligations of the equipment suppliers. The suppliers provide warranties for the equipment. In the past, my children would burn These warranties are passed through by the POs to their their hands on the kerosene lamps customers. Battery suppliers are required to recycle expired batteries collected by the POs.16 while I was cooking. Sometimes, the lamps would run out of kerosene in 2.3.8 Development Partners the middle of a meal. At other times, Development partners provided the capital required to offer the children will run around and spill consumer financing and grants. Such financing is coursed kerosene oil on top of the food. The through the government to IDCOL and from IDCOL to the food would be spoiled for the day. Now, POs and then to consumers. World Bank IDA and GEF funds were the first to be provided for the pilot in 2003 and then I have no such problems. Everything is for scale-up. Building on success and experience, other so clear under the lights (powered by development partners gradually contributed both grants and loans. The development partners with IDCOL were GEF, solar energy). I no longer stay huddled Global Partnership on Output-Based Aid (GPOBA) (another with my children in fear of the dark. World Bank-managed trust fund), Asian Development Bank I can cook whenever I want to”. She (ADB), Japan International Cooperation Agency (JICA), USAID, KfW, GIZ, Islamic Development Bank (IsDB), and the continued, “The Solar Home System has UK Department for International Development (DFID). enabled us to break out of darkness and An important aspect of development partner support was live in light, isn’t it good for us? Now, my that they all adopted the implementation arrangements elder daughter can study well at school. and business model established by IDCOL without attempting to launch parallel efforts with different In the past, I had to work as a domestic requirements. In contrast, TR/KABITA, initially led by local help in different houses so that I could administration, imposed different requirements and was considerably disruptive to the SHS business guided by contribute into the household income. IDCOL.9 Smoother implementation resulted once IDCOL Even though, at times I was not feeling took over the administration of the TR/KABITA Program and well, or had a fever or a cold, I still had used the same POs, technical standards and warranties, and other features of the SHS Program. to go to work. My husband is a day laborer, and doesn’t earn enough for a 2.4 FINANCING THE SHS PROGRAM family of five. Now, I can make mats or Total investment in the SHS Program is estimated at umbrellas at home during nighttime. I US$1,094.93 million to provide electricity services to about can sell my products in the market and 20 million people or about US$266 per household. The I don’t have to work as a domestic help sources and amounts of financing are shown in Table 1 (Keystone Business Support Company Limited 2018). any longer.” The original RERED Project and two rounds of additional Nonetheless, the dominant role of men in deciding to get financing as well as the follow-on RERED II Project and one an SHS was acknowledged by a focus group discussion more round of additional financing raised US$416.3 million participant, “We all know that males are key persons, after for SHS from World Bank IDA resources. Other development all, there is no possibility of expansion of SHS to a single partners, building on the positive experiences and results of household without the consent and involvement of the the SHS Program, offered additional financing for grants and male members of the family.” Accordingly, with this implicit loans. Their funds were seamlessly integrated into the SHS bias, it was not surprising that marketing and advertisement Program using the same implementation modalities and drives were targeted predominantly toward men. POs. Down payments by users and equity investments by 16 POs are not permitted to sell a new battery to an SHS customer without collecting the expired ones. POs pay customers the salvage value of the battery. The battery manufacturers reimburse the salvage value to the POs. Subject to availability of funds, IDCOL pays US$5 equivalent as collection cost to POs. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 17 2. THE SOLAR HOME SYSTEMS PROGRAM POs further leveraged World Bank and other development the GEF, GPOBA, USAID, KfW, GIZ, and DFID, amounting partner financing. Responding to the demand for SHS to US$80.9 million. In total, international development systems and components, manufacturers and suppliers partners provided US$682.8 million in grants and loans. The invested in domestic manufacturing capacities, including funds provided by the World Bank and other development solar module manufacturing plants. partners leveraged considerable private funding from users, POs, manufacturers, and distributors. Down payments by Credit support came from four development partners SHS customers, equity investments by POs, and upstream among which the World Bank (IDA) provided 69 percent investments by manufacturers and distributors are of the credit support of US$601.9 million. Other credit estimated at US$412.15 million, of which user contributions financiers were the ADB, JICA, and IsDB, which provided are US$160.3 million up to December 2017. US$185.6 million in loans. Grant funds were received from Table 1: Sources and Amounts of Financing to December 2017 for the SHS Program Financial Contribution (US$, millions) Project Title Project ID Statusa Approval Date Total for Funds for SHS Component Project Credit Grant Equity Total RERED P071794 190.98 Closed June 25, 2002 RERED Additional P112963 130.00 Closed August 4, 2009 Financing (AF) 416.33 — — 416.33 RERED AF 2 P126724 172.00 Closed October 4, 2011 RERED II P131263 155.00 Active September 20, 2012 RERED II AF P150001 78.40 Active June 19, 2014 Associated or Related Project GPOBA: P119549/ 7.20 — 7.20 — 7.20 Closed March 11, 2010 Bangladesh SHS TF096551 GPOBA: TF098472 6.75 — 6.75 — 6.75 Closed November 6, 2011 Bangladesh SHS Other Development Partners ADB 2453-BAN (SF) 80.00 78.00 2.00 — 80.00 Closed November 16, 2011 ADB 3046-BAN(SF) 10.00 10.00 — — 10.00 Closed April 6, 2014 JICA BD P-75 89.38 81.06 — — 81.06 Closed May 9, 2013 USAID TF-15034 3.56 — 3.07 — 3.07 Closed July 10, 2013 KfW 2002 66 809 22.11 — 19.56 — 19.56 Closed December 22, 2005 GIZ 81169085 16.77 — 16.77 — 16.77 Closed December 1, 2013 IsDB BD-151 16.49 16.49 — — 16.49 Closed June 3, 2009 GEF P074040 8.20 — 7.00 — 7.00 Closed July 16, 2002 DFID 202976-107 28.35 — 18.55 — 18.55 Closed October 10, 2013 Sub-total 1015.19 601.88 80.9 682.78 Private Sector — 173.64 — — 160.29 160.29 — — POs — 227.07 — — 219.72 219.72 — — Manufacturers and Suppliers — 63.07 — — 32.14 32.14 — — (estimate) Sub-total 463.78 — — 412.15 412.15 Total — 1,472.22 601.88 80.9 412.15 1,094.93 — — Note: a. Active status as of July 2020. Sources: IDCOL and Keystone Business Support Company Limited 2018. 18 The US$416 million in World Bank financing leveraged concessionary terms and in foreign currency. The World 163 percent more financing from other sources. Moreover, Bank IDA interest rate was 0.75 percent and loan tenor was RERED leveraged the capabilities of the microfinance 40 years, including a 10-year grace period. The interest rate sector that Bangladesh pioneered and nongovernmental of JICA credits was 0.01 percent and the loan tenor and and private sector capabilities to manufacture, distribute, grace period were the same as for World Bank IDA credits. finance, and service solar and other clean energy products ADB funds were on-lent at an interest rate of 1–1.5 percent, directly to the rural communities. repayable in 25–32 years, including 5–8 years’ grace. IsDB loans carried a 0.75 percent service charge for a 25-year The development partners’ contributions toward the SHS tenor including a 6-year grace period. The government on- Program in grants and loans are shown in Figure 7. lent these funds to IDCOL in Bangladesh taka at an interest rate of 3 percent, while taking the foreign currency risk, 63%, 416.33, The World Bank repayable in 20 years with a 5-year grace period. 2%, 13.95, GPOBA IDCOL in turn refinanced loans that the POs made to SHS customers. The on-lending terms varied depending on the 14%, 90, ADB type and maturity and their level of lending of the POs as well as size, experience, and capability of the POs. IDCOL 12%, 81.06, JICA did not refinance the whole amount of the loans given by 0%, 3.07, USAID the POs to the customers. Increasingly, the loan terms were tightened to make them closer to commercial terms (see 3%, 19.56, KfW Table 2). 3%, 16.77, GIZ IDCOL reduced its interest rates to POs by 1–2 percent from 2%, 16.49, IsDB July 1, 2016, and then to 4 percent for outstanding balance from January 1, 2018. IDCOL reduced the interest rate from 1%, 7, GEF 4 percent per year to 0 percent on the SHS loan outstanding of POs with IDCOL with effect from July 1, 2018, concurrently Figure 7: Development Partner Financing for the SHS with the government eliminating interest payments by Program (US$, millions) IDCOL. More explanation on interest rate reductions are given in Chapter 4 and its financial implications are analyzed in Chapter 5. 2.5 FINANCING TERMS FROM IDCOL TO POS POs in turn financed SHS sales at a service charge of 12–16 AND CONSUMER PAYMENT TERMS percent (flat rate rather than on a declining balance basis) The investment financing provided to IDCOL by the repayable over 1–3 years. The down payment required from development partners through the government was on customers was typically 15 percent. Table 2: Lending Terms to POs Cumulative Refinance Loan Interest Rate Loan Tenor including Grace Period Years Amount (BDT, millions) Refinance (Percent per Year) Grace (Years) (Years) 2003–2008 — 6 10 2 Up to 500 6 8 2 2009–2011 500–1,000 7 7 1 Above 1,000 8 6 1 Up to 250 6 7 1 250–500 7 6 1 2012–2015 70–80% 500–1,000 8 6 1 Above 1,000 9 5 0.5 Up to 250 6 7 1 250–500 7 6 1 2016–2017 500–1,000 7 6 1 Above 1,000 7 5 0.5 Up to June 30, 2018 — n.a. 4 Interest rate on outstanding balance July 1, 2018 — n.a. 0 Retroactively renegotiated LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 19 2. THE SOLAR HOME SYSTEMS PROGRAM 2.6 LOAN SECURITIZATION shareholders, corporate guarantee, mortgage of land or bank guarantee to secure 20 percent of the outstanding Microfinance organizations were used as POs for financing loan, increasing DSRA balance to be equivalent to four SHS based on their success in microcredit activities and quarterly installment payments, and so on. widespread networks at the village level. Following practices by PKSF or similar funding sources, the POs were not However, none of the POs provided a legal mortgage of required to provide any security under the PAs executed land or bank guarantee. Also, they could not maintain the between IDCOL and POs except for maintaining a balance required DSRA balance but rather maintained a balance in the Debt Service Reserve Account (DSRA). This was equivalent to a maximum of two quarterly installments. equivalent to one semiannual installment payment that In addition, some POs registered as an NGO/society/ could secure less than 20 percent of the loan. Therefore, foundation expressed their inability to provide personal loans extended to the POs were almost collateral free. guarantees by the members of their executive committee as they did not own the organizations. In 2009, IDCOL executed an Amendment and Restatement Agreement to the PA with the POs that incorporated the Despite partially completed security documentation, IDCOL security package for the first time. This was the time when continued disbursement of loans to the POs to ensure IDCOL started to take SHS loans on its balance sheet. smooth operation of the SHS Program. Otherwise, POs Therefore, IDCOL decided that some additional security would not be able to continue installation of SHS and make should be in place to address the credit risk to IDCOL. The debt service payments to IDCOL due to liquidity problems. security package included a first charge hypothecation on all floating assets of POs, a lien on all project accounts, In March 2016, IDCOL made further changes to the security a demand promissory note, and a letter of continuity. package approved in 2011 considering their applicability However, this was not enough to significantly address the and status of the program. Mortgages of land or bank credit risk of IDCOL. guarantees to secure 20 percent of the outstanding loan were waived. POs were now required to maintain a DSRA In December 2011, the IDCOL Board included some balance equal to two quarterly installment payments additional securities considering the increased credit instead of four. For an NGO/MFI/society/foundation, a exposure as well as to achieve commercialization. These personal guarantee was required from one member of the included first charge hypothecation on all fixed and floating executive committee instead of all members. Most of the assets of POs, personal guarantee from the directors/ POs complied with these lesser security requirements. Table 3: Sources and Amounts of Grants for SHS per Agreement Amount of Grant Available per SHS No. of SHS Number of SHS Source Institutional Receiving Grant Financed Total Buy-down Grant Development Grant First 20,000 GEF 20,000 US$90 US$70 US$20 Next 20,000 GEF 20,000 US$70 US$55 US$15 Next 35,000 GEF 35,000 US$50 US$40 US$10 KfW 30,000 Next 88,160 EUR 38 EUR 30 EUR 8 GIZ 58,160 Next 35,000 KfW 35,000 EUR 36 EUR 30 EUR 6 KfW 135,000 Next 238,659 EUR 34 EUR 30 EUR 4 GIZ 103,659 KfW 103,000 EUR 28 EUR 25 EUR 3 Next 161,543 GPOBA 58,543 US$36 US$30 US$6 KfW 99,018 EUR 22 EUR 20 EUR 2 GIZ 24,359 Next 443,520 GPOBA 178,103 US$28 US$25 US$3 GPOBA 142,040 GPOBA 70,960 Nil Next 510,960 ADB 80,000 US$25 US$25 (US$3 will be paid to new POs only if funds remain) IDA 360,000 Total 1,552,842 20 2.7 SUBSIDY TRENDS declined from US$1.72 per Wp in 2003 to US$0.24 per Wp in 2018. As a percentage of SHS cost the grant declined sharply In addition to the loan funds, the SHS Program provided from 18 percent in 2003 to 4 to 8 percent from 2006 onward. grant funds. Grant funds came from several development From January 2012 onward, the grant was US$20 per SHS partners. The amounts of grant allocated per SHS declined for 30 Wp or smaller systems. There was no grant support for over time as the SHS installations grew. Two types of larger SHS. grants were offered: a capital buy-down grant to increase affordability and a small institutional development grant to help the POs establish the retail service infrastructure. The 2.8 SHS PROGRAM RESULTS sources and amounts of grant funds are shown in Table 3. Since the amount of grant is the same for all sizes of SHS, it is a progressive grant where grant support is greater for the 2.8.1 SHS Installations Under the Program smaller SHS. Since smaller SHS are demanded mainly by Beginning in 2003, POs began to market, sell, install, and poorer households, the grant benefits are skewed toward service SHS under the SHS Program. SHS sales began to the poorer SHS customers. grow—slowly at first, then accelerating, and levelling off over time as the market matured. The SHS market expansion The grant was to end when cumulative SHS sales appears to follow the classic model of market diffusion, reached 1,552,842. However, IDCOL had agreed with the and in this case, the market size decreased with time, as development partners that, if grant funds remained after grid connection expanded at a faster rate than household the SHS installations cap was reached, the funds would formation. The SHS market now exhibits the characteristics be disbursed for smaller SHS (under 30 Wp) and for of a saturated market. After slow growth in the early years, institutional development. The trend in the actual amount the pace of installation accelerated, peaking with 861,000 of grant provided to SHS is shown in Table 4. The grant Table 4: Actual Amount of Grants Provided for SHS Total SHS Total Wp Avg. SHS Cost Total Grant a Grant/SHS Grant/Wp Grant Share of Year Current Current Current Current SHS Cost (%) No. MWp US$/Wp US$, millions US$/SHS US$/Wp 2003 9,075 0.45 8.95 0.78 85.93 1.72 19.2 2004 18,499 0.94 8.49 1.45 78.52 1.55 18.3 2005 26,196 1.35 8.23 1.43 54.60 1.06 12.8 2006 35,731 1.98 8.59 1.43 40.05 0.72 8.4 2007 62,574 3.49 9.22 2.09 33.46 0.60 6.5 2008 100,640 5.58 9.98 3.85 38.21 0.69 6.9 2009 156,827 7.73 9.89 5.91 37.69 0.76 7.7 2010 295,597 14.70 8.39 9.33 31.56 0.63 7.6 2011 425,788 19.82 8.27 10.87 25.53 0.55 6.6 2012 612,373 25.63 8.00 13.89 22.68 0.54 6.8 2013 861,172 30.51 7.74 9.52 11.05 0.31 4.0 2014 726,512 23.54 5.38 8.32 11.45 0.35 6.6 2015 575,580 19.29 5.44 6.85 11.90 0.36 6.5 2016 175,990 6.31 4.29 2.23 12.68 0.35 8.2 2017 29,475 1.19 4.69 0.31 10.49 0.26 5.6 2018 3,455 0.13 3.25 0.03 9.25 0.24 7.3 Total 4,115,484 163 78 Note: a. Total grant of US$78 million is less than the grant provided by development partners, which was US$80.9 million, due to variations in the exchange rate. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 21 2. THE SOLAR HOME SYSTEMS PROGRAM installed MWp capacity perspective, they accounted for 43 percent of capacity. About 36 percent of installed capacity was for SHS that were 50 Wp to less than 75 Wp, which constituted 26 percent of total sales. The balance 21 percent of capacity was for systems in the range of 60 Wp to 300 Wp, which constituted 10 percent of the total number of systems installed. Total installed capacity was 163 MWp, with the average size of SHS being 40 Wp. Figure 8: Annual and Cumulative SHS Installations SHS installed in 2013. The pace of installation began dropping at an increasingly faster pace from 726,000 in 2014, 576,000 in 2015, 176,000 in 2016, 29,000 in 2017, and nearly 3,500 in 2018. About 4.115 million SHS were financed through the SHS Program (see Figure 8). 2.8.2 Sizes of SHS Demanded Figure 10: Households with SHS as Percentage of Rural Overall, 10 to 45 Wp SHS accounted for 64 percent of the and Total Households total number of SHS installed (Figure 9), though from an 2.8.3. Market Penetration and Regional Distribution of SHS The market penetration of SHS, as measured by the total number of SHS in use as a percentage of households each year, peaked in 2016. It was 16.2 percent of rural households (or 10.5 percent of total households), assuming that the SHS useful life was 12 years. In comparison, total electricity access of the rural population in 2016 was 66 percent. By 2014, one-fifth of all rural households that had electricity access were obtaining electricity services from SHS. By 2018, total rural electricity access reached about 80 percent with 13 percent of these households obtaining electricity from SHS. Increasing or decreasing SHS life within the range of 10–15 years did not make a significant difference in SHS market penetration (see Figure 10 and Table 5). The SHS Program installations are spread throughout the country’s off-grid areas as there were no restrictions as to where SHS Program sales could take place (see the map in Figure 11). The concentration of SHS installations is variable with Southern and Northeastern Divisions having the most SHS. Figure 9: Number of SHS Installed from 2003 to 2018 by Size 22 Figure 11: SHS Sales (by Color) and Market Penetration (% of Total Households) at the District Level LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 23 2. THE SOLAR HOME SYSTEMS PROGRAM Table 5: Market Penetration of SHS in Bangladesh 2003–2018 Rural HHs (Millions) Percentage of Rural Total HHs (Millions) Percentage of Total Persons per Rural Assuming 12-Year Percentage Rural Rural Population Persons per HH SHS Operating HHs with SHS HHs with SHS SHS Installed Population (Millions) (Millions) Year Life HH 2003 9,075 9,075 135.0 74.6 100.7 4.98 5.00 27.0 20.1 0.0 0.0 2004 18,499 27,574 137.0 73.9 101.2 4.91 4.95 27.7 20.4 0.1 0.1 2005 26,196 53,770 139.0 73.2 101.7 4.85 4.87 28.6 20.9 0.2 0.3 2006 35,731 89,501 141.0 72.5 102.2 4.77 4.80 29.4 21.3 0.3 0.4 2007 62,574 152,075 143.0 71.8 102.6 4.70 4.73 30.2 21.7 0.5 0.7 2008 100,640 252,715 144.0 71.0 102.3 4.63 4.66 30.9 21.9 0.8 1.2 2009 156,827 409,542 146.0 70.3 102.6 4.56 4.60 31.8 22.3 1.3 1.8 2010 295,597 705,139 148.0 69.5 102.9 4.50 4.53 32.7 22.7 2.2 3.1 2011 425,788 1,130,927 149.0 68.8 102.5 4.42 4.46 33.4 23.0 3.4 4.9 2012 612,373 1,743,300 151.0 68.0 102.7 4.35 4.39 34.4 23.4 5.1 7.5 2013 861,172 2,604,472 153.0 67.2 102.9 4.28 4.32 35.4 23.8 7.4 10.9 2014 726,512 3,330,984 155.0 66.5 103.0 4.21 4.26 36.4 24.2 9.1 13.8 2015 575,580 3,897,489 156.0 65.7 102.5 4.14 4.19 37.3 24.5 10.5 15.9 2016 175,990 4,054,980 158.0 64.9 102.6 4.06 4.11 38.4 25.0 10.5 16.2 2017 29,475 4,058,259 160.0 64.1 102.6 4.00 4.05 39.5 25.3 10.3 16.0 2018 3,455 4,025,983 161.0 63.4 102.0 3.93 3.98 40.4 25.6 10.0 15.7 Source: Population and rural population and electricity access data from World Bank Data Bank. Household size from CEIC (2020) estimated from Household Income and Expenditure Surveys (HIESs). Note: HH = Household. Barisal, Sylhet, Chittagong, and Mymensingh Divisions had more readily. Southeast is the Chittagong Hill Tracts. This the highest market saturation. In some districts more than area had been affected by conflict and insurgencies for a half to two-thirds of households use SHS, comparable to long time. Population density here is exceptionally low. grid electrification coverage during 2010–2015 (based on Grid electrification is challenging; hence solar would be an numbers of households per district from 2011 Census). Sales attractive power source. In the South and Southeast are the in the North and Northwestern parts of the country were low delta and the Sunderbans—the largest mangrove forest of (Table 6). the world. The whole Southern part of Bangladesh consists of thousands of small islands as the large rivers break into A total of 13 districts (20 percent) had over 1.5 million SHS hundreds of small tributaries as they fall into the Bay of installed, with SHS penetration of 30 percent or more of Bengal. Here too SHS would have been an attractive option. households in those districts. Not surprisingly, districts encompassing major urban areas that were substantially The maximum number of the unelectrified households electrified had low SHS market saturation (Figure 11 shows that were the target market for the SHS was about 15 the concentration of SHS sales at the district level). The million in 2003 when the RERED pilots started. By 2018, 4.1 number of households within a district is based on the 2011 million SHS were sold into this potential market. Assuming Population Census. Market penetration was highest in the that households with multiple SHS were rare, this would Northeast and Southern regions. The Northeast has small mean that about 25 percent of the maximum number of hills, tea gardens, and seasonal large water bodies (called unelectrified households bought and operated an SHS Haor or Bill), making it more difficult for grid electrification.17 between 2003 and 2018. This is a significant share of the People in these regions would have lower expectations of target market, especially of the better-off households that getting a grid connection and would have opted for SHS were the main customers. 17 Communication with Raihan Elahi, Lead Energy Specialist at the World Bank and former Task Team Leader of the RERED Project (August 11, 2020). 24 Table 6: Total SHS Sales as % of Divisional Households (2003–2018) Market Saturation SHS Sales as Share of District-wise Range Total Households (2011 Division 2011 Total Households Maximum SHS Minimum SHS Census) (%) Penetration Penetration Barisal 1,849,355 39 65.7 18.6 Sylhet 1,762,757 30 52.8 12.8 Chittagong 5,552,270 17 37.0 4.6 Mymensingh 2,528,321 14 25.0 11.2 Khulna 3,707,046 10 23.1 0.8 Dhaka 8,050,230 10 62.0 0.2 Rangpur 3,794,608 7 15.2 2.6 Rajshahi 4,461,096 6 10.4 1.4 Total 31,705,683 14 65.7 0.2 Note: IDCOL SHS sales database. Population data from Bangladesh Bureau of Statistics (2015). The reason that less well-off households did not participate Kerosene saving: The program is estimated to offset as strongly in the SHS Program appears to be mostly related about 4 billion liters of kerosene from its inception to to affordability issues (the need for a down payment and 2021. regular re-payments of loans for several years). This is Social impact: BIDS conducted impact assessment of supported by surveys completed in 2013 which found that IDCOL’s SHS Program, which estimated its effects on rural only 10 percent of households with under 2.5 acres of land families and communities as follows (Asaduzzaman et al. purchased an SHS whereas 25 percent of householders 2013) (see Box 4 for a few illustrative examples): with 5 acres or more land purchased an SHS. The SHS owner also earned an average of 80 percent more income • Study hours and schooling: Brighter solar lights allow than a non-adopter. Other factors influencing the buying children to study longer hours. Both the boys and girls decision were education levels of the user, quality of on average study 10–12 minutes per day longer with the house, and hygienic practices (which are related to solar lights than those without it. The year of schooling income). The presence of a strong microcredit institution completed was higher for children with SHS than those and geographically remote location also contributed to the without it and the differences are significant for both propensity to buy (Khander et al. 2014). boys and girls. • Safety and amenities: SHS households enjoy higher safety, comfort, and convenience compared to non-SHS 2.9 SHS BENEFITS households. For instance, SHS household members have a greater sense of security at night by replacing Benefits from the SHS Program have accrued to rural kerosene lamps with SHS light. In addition, the SHS households and to the country as well as the global households had easier and lower cost access to TV, community. Households have had access to the better radio, fan, and mobile phone charging services. quality and more extensive services that electricity can offer—far earlier than if they had to wait to obtain a grid • Impact on health: SHS household members suffered electricity connection. This section provides a broad, mainly less from several types of preventable illness such as qualitative overview of the benefits as follows: general ailments, respiratory diseases, gastrointestinal illness, as well as having reduced risk of fire. Households Coverage: The program has ensured supply of solar with SHS had lower fertility.19 Whether this is directly electricity to about 20 million rural people who previously attributable as an outcome of SHS use is debatable. consumed kerosene for lighting, which is equivalent to 14 percent of country’s total population in 2011.18 18 Based on estimates using survey data compiled by Grameen Shakti (2765 CPA CER Sheet Grameen Shakti) on kerosene fuel offset by SHS of various sizes and types and the number of kerosene lanterns replaced and daily hours of lighting. 19 The health impact appears to be a result of the process of information dissemination through the electronic media, TV, and radio. All family members, men and women and boys and girls, have experienced lower incidence of disease in SHS households compared to non-adopters. However, SHS adopters were economically and socially better-off than non-adopters. Hence, reduced disease prevalence may be due to a better economic situation and higher education. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 25 2. THE SOLAR HOME SYSTEMS PROGRAM • Benefits for women: SHS had a positive influence on women’s mobility and sense of security. Women spent more time tutoring BOX 4: Case Studies - Improving Quality of Life children, watching TV, socializing, visiting friends, and neighbors after the adoption of SHS. • Empowerment: TV, radio, and mobile phones Improving Quality of enabled rural people, especially women, to connect to the rest of the world and have Life and Income brought them ideas on various rights. Access to TV also enabled them to observe the customs and rights that women in other society practice and helped them reshape their rights and customs. Participation of women in different types of decision-making within the family had improved in terms of women’s freedom of mobility, participation in household, and economic decision-making. In the case of women’s freedom of mobility expressed as participation of decision-making, in visiting parental home, going shopping, visiting friends and relatives, and going outside the village, the women from SHS user households had shown greater involvement in decisions. The right to decide by herself was higher in SHS households than those for nonuser households. The tendency to make the decisions jointly with the father/husband is higher than in nonuser households. • Sense of security at night: Most of the SHS households confirmed that SHS connection enhances nighttime security. Replacing kerosene lamps by SHS lights provides better and, in most cases, cost-effective ways to provide lighting for longer durations at night. • Employment generation: IDCOL and each of the POs have created employment for rural communities through the establishment of the program. As of November 2018, IDCOL Mrs. Jorina Begum and her husband Nurul Islam collectively created 29,000 direct jobs through live together with one of their three sons in a the program. These are in addition to the medium-size village beside a river in Nalchity, in the employment created and income generated by Southern part of Bangladesh. Nurul Islam used to using the electricity available from SHS. drive an auto-rickshaw with which he could earn to serve his family’s needs. After he was injured in an The global environment is improved by the accident in his auto-rickshaw, the family income reduction in kerosene combustion due to the shrank. Nurul Islam could only function as a shop reduction in CO2 and black carbon (Bond et al. assistant but could not perform physical activities. 2011)20 emissions. The amount of CO2 emissions Both appreciated the benefits of using the SHS. They avoided between 2003 and 2021 by the 4 billion have purchased a 40 Wp SHS that operates three liters of kerosene offset by the SHS is estimated at lights and a small color TV and recharges their cell 9.6 million tCO2. phone. Mrs. Jorina uses the SHS to perform income- generating activities at night such as weaving handicraft and sewing clothes. Before the SHS, they used kerosene lamps, but the cost was higher, and illumination was poorer. The SHS has made a huge impact for them as they can keep one light on the entire night and lead a more comfortable life. 20 During its short atmospheric lifetime, 1 kg of black carbon produces as much positive forcing as 700 kg of CO2 does for 100 years. 26 Enhancing Safety and Improving Business Security Mr. Abdul Kader is a farmer in his 30s living on a small island in Godagari in Northern Bangladesh. He owns some agricultural land. He lives with his wife and five children, two of whom are married. He has a medium-size 50 Wp SHS to operate four lights and a small color TV and charge his cell phone. “It is very Mr. Abdul Halim is a traditional shop owner selling important to have the solar home system at night,” evening snack items at his small shop located in he says. “I felt I needed it for my family safety, and Kaunia, in the Northern part of Bangladesh. He now my family can socialize at night, they can go previously used to get electricity from his relative’s to the washroom without any fear and we are free house connection. He paid BDT 500 per month but from darkness.” Before he bought the SHS, he used a had no control over how much power he would kerosene lamp and sometimes a small battery to run receive and for how long. The SHS installed 15 a light bulb. The SHS has made a huge difference for months ago has changed his business quality and him, as he can leave two lights on all night. His family sustainabilty by a big margin. He can keep his shop is very happy using the SHS and they are willing to open for extended hours till 10 p.m., giving him upgrade to a large SHS with more options. additional income. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 27 28 ADAPTING TO REALITY 3.1 Continuous Adaptation of the Program to Changing Circumstances 30 3.2 Obtaining Timely Information/Data 30 3.3 Responding to SHS Technology Change and Consumer Choice 32 3.4 Responding to Environmental Management Needs 34 3.5 Attempts to Access Commercial Financing 35 3.6 Assuring Long-term Sustainability of SHS 35 3.7 Solar Industry Development in Bangladesh 36 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 29 ADAPTING TO REALITY 3.1 CONTINUOUS ADAPTATION OF THE PROGRAM TO CHANGING CIRCUMSTANCES A key feature of the SHS Program was retaining flexibility to adjust to changing technology, changes in consumer demands and affordability, and market conditions. To successfully adapt, IDCOL required timely information, quick decisions, and rapid implementation of changes. It also needed feedback to ensure that the changes were effective and, if not, to modify. In the SHS Program, a process of continuous adaptation naturally evolved during implementation, owing to the openness of management. This adaptation process as implemented by IDCOL was innovative and successful to a significant extent, but there is always room for improvement, as discussed in the following sections. The continuous adaptation of the SHS Program is illustrated using a detailed risk matrix to describe how IDCOL mitigated various risks over time (see Appendix C). The matrix demonstrates how the implementation features were modified based on feedback from the POs and suppliers, consumer surveys, inspections and monitoring, changing technology, changing costs and market conditions, and force majeure events. 3.2 OBTAINING TIMELY INFORMATION/DATA A crucially important aspect of the continuous adaptation process of the program was communication—between IDCOL and POs and suppliers, POs and their customers, and IDCOL and SHS users. Communication played several essential roles—ensuring POs complied with project requirements, carrying feedback from POs to help improve project implementation and address quality concerns as well as market development issues, providing consumer feedback, and enabling IDCOL and the POs to respond to their concerns. The OC payed an exceedingly important organizational role in such communication. 30 2003–2005 Awareness campaign in selected community Promotion of innovative design and quality 2006–2008 Mass promotion campaign to the rural community Promotion on product quality and standards 2009–2013 Promotion on special/incentive price offers Focus on door to door sales Limited promotion for non users 2013–2015 Promotion on SHS maintenance Promotion on quality components of SHS Promotion on collection efficiency improvement 2015 onwards Promotion on larger systems sales Figure 12: Timeline of Promotion and Outreach Activities 3.2.1 Consumer Outreach • In 2004, POs reported that some POs were making exaggerated claims about SHS performance, selling SHS From the onset, communication with customers, directly without taking down payments, erroneously claiming by IDCOL or through the POs, was important for market that their SHS was cheaper than that of other POs, development. It was used to inform consumers of how best and encouraging customers to return their system and to use the SHS (dos and don’ts), build confidence in the replacing them with system from another PO. The POs as products and in the POs, and obtain consumer feedback. a group agreed to stop such practices. Every few years, consumer surveys were conducted to obtain feedback on how the SHS were used, the benefits as • The OC approved policy guidelines regarding disposal perceived by the users, learn how technology performed, of warranty-expired batteries and POs agreed take back and PO services were viewed. old batteries for recycling when supplying replacement batteries. In the early years, IDCOL supported several mass-marketing • IDCOL and POs agreed to share the expenses for TV initiatives. These included billboards targeted to the commercial on an 80–20 percent basis. selected rural communities, drama and TV commercials • With the objective of lowering price and ensuring an broadcast during village fairs, customer orientation uninterrupted supply of solar panels to POs, IDCOL programs, and local cable TV. The major communications sought proposals for setting up a domestic solar panel efforts over the years are shown in Figure 12. assembling plant. • POs were to take immediate measures to disseminate 3.2.2 Operations Committee Meetings the stickers containing the call center numbers among the households. If any household was found without the Communications between IDCOL and the POs through the sticker after August 2009, the installation of SHS in that monthly OC meetings were crucial in identifying problems, household would be treated as noncompliant; grants and agreeing on solutions, and sharing lessons among POs and refinancing claims against that SHS would be temporarily with IDCOL. Issues discussed were many and varied as the withheld and disbursement would be made after following examples illustrate: compliance of the requirement. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 31 3. ADAPTING TO REALITY • To respond to increasing numbers of complaints, POs 3.2.5. Technical Audits agreed to train all customers using IDCOL’s user training IDCOL conducted third-party technical audits to ensure guidelines. accountability of suppliers and POs in terms of quality and • POs were advised on the Collection Efficiency after-sales services. The technical audits were conducted Improvement Program (CEIP) and proposed new models randomly to verify quality of the main SHS components, to improve the collection performance. that is, solar module, battery, charge controller, and CFL/ LED bulb, to ensure that suppliers were providing quality 3.2.3 Call Center products. In 2007, IDCOL set up a call center, open every day except The audit also verified the POs’ performance in terms of Fridays and government holidays, for customers to report installation, warranty, and after-sales support. Based on technical problems and seek after-sales service. IDCOL has the findings of the technical audit, necessary corrective since been receiving calls directly from customers regarding measures were taken including penalizing suppliers/ the problems not addressed by the POs for rectification. manufacturers for poor quality products. The TSC also made changes to the technical standards based on the feedback The complaints received by the call center were addressed of the technical audit. in two ways. The customer was given the contact details of the PO’s regional office to connect the two parties directly. Independent technical audits showed that most of the The concerned PO was also informed by the call center equipment supplied by manufacturers under the SHS about the complaint and was advised to take necessary Program was performing well. The audit identified problems actions. The record of the complaint was kept in a database. in some equipment in which case equipment approval A follow-up call was made by the call center to the customer was suspended or manufacturers were penalized for the to ensure the problem was resolved by the concerned PO equipment not meeting technical standards of the TSC. within a reasonable time. 3.2.6. Random Sample Testing 3.2.4. Verification of Operational Performance As a part of regular monitoring for quality assurance, the IDCOL teams visited and inspected a sample of SHS SHS Program randomly collected samples from suppliers’ installations. Release of grant funds and loan refinancing warehouses and installation sites and tested them at the was contingent on the installations passing the inspections. Bangladesh University of Engineering and Technology The quality inspectors visited customer households to (BUET), at a testing facility that had been set up with ensure the following: the support of the SHS Program. The test reports were submitted to the TSC for review. If there were deviations • SHS are installed within off-grid areas. from the technical standards, the TSC recommended • The systems meet technical and financial requirements as temporary suspension of the specific model. If retests were set by IDCOL. unsatisfactory, the TSC could delist the product. • Approved SHS components (solar panel, battery, controller, and so on) are used. 3.3 RESPONDING TO SHS TECHNOLOGY • After-sales maintenance and warranty support are CHANGE AND CONSUMER CHOICE provided to customers. Consumer choice was an important feature of the SHS • Customers are satisfied with the service. Program. Recognizing that consumers had differing abilities • Technical problems with SHS are identified for to pay and different priorities for electricity use, the SHS rectification by the POs. Program sold SHS of various capacities (10–300 Wp), though all had to meet quality standards. Though this increased the IDCOL prepared technical reports for each PO and provided cost of business, it was crucial to SHS acceptance. them to the PO unit offices for immediate action. IDCOL also followed up on the status of the problems with POs Initially, SHS of 30 Wp to 100 Wp solar module capacities and customers on a regular basis. Before 2013, IDCOL did were sold during the early years when the grant available not penalize POs for technical discrepancy of SHS as the for SHS was higher than in later years. Later, as more percentage of noncompliance was low. However, when efficient CFLs and then even more efficient LED lamps sales were at their peak, IDCOL observed that the share of became available, smaller SHS, including pico-SHS of 10 Wp noncompliant SHS installations began increasing, which capacity, were offered. As rural consumers became wealthier implied that POs were focusing more on installing SHS and their electricity needs increased, larger systems were rapidly than doing so properly. To reverse this, IDCOL started demanded. Most importantly, beginning in 2008, the lighting to deduct the noncompliant SHS from the monthly claim value of SHS electricity increased significantly with the request of POs. Such corrective measures made POs more introduction of far more efficient LED lamps. As the amount attentive in resolving problems, and gradual improvements of electricity needed for lighting declined, consumers were observed in PO performance. preferring lower cost of service purchased smaller-size SHS, and the average size of SHS sold declined. 32 2008‒LED lighting 60 introduced 2014‒Battery autonomy 55 56 55 reduced from 3 to 2 days, 51 52 reducing costs 50 49 50 50 47 42 2013‒Grants Cut Sharply 40 39 40 36 36 Size of SHS (Wp) 34 32 30 20 10 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Figure 13: Trend in Weighted Average Size of SHS In 2013, cost buy-down grants available for SHS declined The decline in SHS costs was driven in part by scale sharply, and the resulting price increase contributed to a economies and rapid reduction in the global cost of solar shift in demand to smaller SHS. In 2014, battery autonomy21 PV modules and intense competition (Table 7). The intense was reduced from three to two days, thus reducing the competition was mainly from the SHS market outside cost of SHS, and the average size of SHS demanded began the IDCOL SHS Program after 2015–2016. Though POs increasing again as rural consumers valued the additional competed with each other, they acted jointly to overcome applications that the SHS could power. The trend in average the competition from the unregulated market that was more size of SHS installed under the program is given in Figure 13. nimble and had no obligation to meet quality, warranty, or service standards. Changing trends in market share of SHS by capacity are also revealing. With the introduction of LED lighting (and grant Around 2007–2008, SHS costs started rising due to the reducing and then ending in 2013), there is a shift to smaller introduction of more efficient as well as more expensive capacity SHS (Figure 14). From 2013 onward, the market LED lights. In 2004, SHS cost averaged about US$8.50 per share of 45 Wp and smaller SHS exceeded 70 percent. Wp installed without subsidy. This included the supply and Figure 14: Trend in Market Share of SHS by Capacity 21 Battery autonomy measures the number of days that the SHS can supply the required amount of electricity if the battery was fully charged and there was no recharging of the battery. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 33 3. ADAPTING TO REALITY Table 7: Trends in Unit Cost of SHS 2003–2017 Total Wp SHS Cost without Subsidy a (Current US$/Wp) SHS Cost without Subsidy a (Constant 2018 US$/Wp) MWp 20 Wp 30 Wp 40 Wp 50 Wp 75 Wp 120 Wp 20 Wp 30 Wp 40 Wp 50 Wp 75 Wp 120 Wp 2004 — — 8.49 8.66 8.50 7.56 — — 11.07 11.30 11.09 9.86 2005 — — 8.23 8.66 8.10 6.95 — — 10.39 10.95 10.23 8.78 2006 11.73 — 8.59 8.60 7.69 6.95 14.39 — 10.53 10.55 9.43 8.52 2007 13.06 — 9.22 9.11 8.11 8.02 15.59 — 11.02 10.88 9.68 9.58 2008 12.12 — 9.98 9.91 8.55 8.81 14.19 — 11.68 11.60 10.02 10.32 2009 11.41 — 9.89 9.83 8.50 8.77 13.27 — 11.50 11.43 9.88 10.19 2010 9.99 — 8.39 8.41 7.54 7.82 11.47 — 9.64 9.66 8.66 8.98 2011 9.81 — 8.27 8.30 7.48 7.82 11.04 — 9.30 9.34 8.42 8.79 2012 9.81 — 8.00 8.04 7.14 6.79 10.83 — 8.83 8.87 7.88 7.50 2013 9.80 7.59 7.74 7.77 6.80 5.77 10.63 8.23 8.40 8.43 7.38 6.26 2014 7.66 6.04 5.38 5.14 4.74 3.83 8.16 6.43 5.73 5.47 5.05 4.08 2015 6.88 5.33 5.44 4.99 4.44 3.89 7.25 5.61 5.73 5.26 4.68 4.10 2016 6.14 5.23 4.29 4.27 3.50 3.18 6.40 5.45 4.47 4.45 3.65 3.31 2017 5.93 4.60 4.69 4.31 3.90 3.35 6.06 4.70 4.79 4.40 3.99 3.43 Note: a. Includes 3 years free maintenance plus 5-year battery warranty and 20-year module warranty. installation of the SHS and the warranties (20 years for solar module, 5 years for batteries, 1 year for controller and lights, and 3 years of free maintenance). In 2014, there was a sharp reduction in unit costs as shown in Table 7, due to reducing the size of the battery from three to two days of autonomy. The cost reductions appear to have experienced curve effects like that observed in ‘Swanson’s Law’22 for PV cost reduction, though applied in this case to SHS system-wide cost (see Figure 15). The average cost of an SHS (in constant 2018 dollars) dropped by about 21 percent for every doubling of cumulative number of SHS sold or for every doubling of cumulative MW of sales. 3.4 RESPONDING TO ENVIRONMENTAL MANAGEMENT NEEDS IDCOL realized as early as 2007 that the accelerating Figure 15: Reduction of SHS Unit Cost with Cumulative SHS demand for SHS meant that there would be many used Installations batteries that would need replacement every five years. While the number of such batteries would always be far less than batteries used in the transport sector, it would be important to ensure batteries are properly recycled. 22 Swanson predicted that cost of solar modules would drop 20 percent for every doubling of cumulative solar PV module shipment (The Economist 2012). 34 As agreed by covenant with the RERED Project, IDCOL made would be around 16–19 percent. In case of a consumer it mandatory for all battery manufacturers to adopt ISO23 loan, recourse is to the acquired asset and income of the 14001-2004 (Environmental Management Standard) and borrower. In this scenario, financing SHS would not be OHSAS24 18001:1999 by June 2012. After 2012, IDCOL did viable at all. not accept any battery manufacturer in the SHS Program • Investment in SHS is essentially rural financing. without having these certifications. Commercial banks and financial institutions do not Uniquely among SHS programs, IDCOL and RERED have much understanding of the related socioeconomic supported and required that POs collect used batteries aspects of these customers. and deliver them to approved battery recycling centers. By The report concluded that the best choices for handling 2018, there were 16 battery suppliers of which 15 were local and reaching out to the target population in the rural and the remaining one, Japan Solar Tech, sold imported areas are the NGOs/MFIs. Significant development and batteries. Four battery manufacturers (Rahimafrooz, Panna improvement of rural livelihood improvement have been Battery, RIMSO, and HAMKO) set up independent recycling achieved through these institutions and the SHS Program is facilities which were ISO 14001:2004/2005 and OHSAS no exception. 18001:2007 compliant. They have agreements with other battery suppliers to accept their old batteries for recycling. IDCOL inspectors regularly visit these recyclers, every three 3.6 ASSURING LONG-TERM SUSTAINABILITY months. IDCOL inspectors also visit the manufacturing OF SHS plants to ensure that the manufacturing process follows With module warranties extending 20 years and batteries environmental and safety standards and to verify that air 5 years, SHS users expect that their SHS will provide useful and effluent treatment comply with standards. service for a long time. However, as with any electrical equipment, failures can occur and customers will need access to spare parts and repair services, even after the 3.5 ATTEMPTS TO ACCESS COMMERCIAL warranty period ends. FINANCING Ensuring SHS customers had convenient access to repair The intention of the SHS Program was to eventually make services and spare parts is crucial. Such services were SHS financing fully commercial with the POs borrowing provided through the unit offices that the POs were required funds at market terms from commercial sources by the end to establish to provide spares and services. In 2013, during of the implementation. However, a commercialization study the peak time of SHS installations, POs had about 5,700 commissioned by IDCOL found that commercial banks unit offices throughout the country with over 29,000 staff were not interested in lending to this sector, either directly employed in 2015 (staff declined to 6,000 by 2018). By 2014, to consumers or to refinance POs (Alam 2013). The main a commercial SHS market had also evolved. Then, rapid factors hindering commercial lending in SHS financing cited grid expansion and commercial market development led to in the draft report were as follows: declining SHS sales under the SHS Program. The POs were • Although the investment improves living conditions and forced to close or consolidate some of their unit offices due generates indirect savings by avoiding alternate energy to lack of sales business. expenditures, lenders do not recognize that it yields direct IDCOL taking over the management of the TR/KABITA income and hence it fails the basic criteria of commercial Program and using the same POs to supply and service SHS financing. and public PV systems took up some of the slack caused • The target market is well outside the typical network of by the decline in SHS Program sales. This permitted unit a commercial bank and makes direct administration offices to remain functional and support both SHS Program impossible. and TR/KABITA SHS. As there is at least one PO nominated • Opposing business/financial dynamics: Typically, from in each upazila (subunit of a district) under TR/KABITA, the risk management and evaluation perspectives, these customers who purchased SHS under the SHS Program kinds of small loans are considered high risk as there can get their spare parts and repair services from these unit is no recourse to any asset of compensating economic offices even if the warranty period expires. This arrangement value such as a property or business. By the same logic, was effective. IDCOL call center records show that 89 percent the consumer or retail banking costs are higher. For of the 10,338 complaints received between November 2018 example, while a good standing corporate lending rate and April 2019 were successfully resolved within one month can be around 12–14 percent, consumer lending rate of receiving the complaint. 23 ISO = International Organization for Standardization. 24 OHSAS = Occupational Health and Safety Management Systems. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 35 3. ADAPTING TO REALITY As BREB grid electrification expands and universal grid electricity access is achieved, the TR/KABITA Program too will end, though warranty obligations will extend for another three years after installation. Once a household gets a grid connection, the SHS will be used mainly as backup (at least until the battery needs replacement, according to some customers) and for supplementary lighting. The few SHS that continue to be used will not justify operating dedicated SHS service centers. During the RERED Program, IDCOL has trained over 35,000 technicians to install, service, and repair SHS. In addition, nearly 500 trainers have been trained. IDCOL expects that some of the trained technicians who were previously employed by the POs will continue to provide fee-based repair services to SHS that continue to operate after the SHS Program and TR/KABITA Program end. Shops that sell SHS on a commercial basis are also available to provide spares and service. 3.7 SOLAR INDUSTRY DEVELOPMENT IN BANGLADESH The SHS industry component and systems manufacturing industry that was catalyzed by the SHS Program includes solar PV module manufacture, batteries, controllers, lights, and other appliances manufacture and pay-as-you-go (PAYG) technology. The SHS Program catalyzed the development of local expertise in these industries within the country. Components such as controllers, lights, and battery have been manufactured locally, supplemented with imported components. PV module manufacturing was also catalyzed by the RERED Project. Modules manufactured by six Bangladesh companies, Australia Bangladesh Solar Power, Electro Solar Power, Greenfinity Energy, Rahimafrooz Renewable Energy, Radiant Alliance, and Shouro Bangla, were approved for use in the RERED Project. These companies are now selling to the solar pumping and solar mini-grid markets and extending to the evolving grid- connected and roof-top solar market. Bangladesh tubular plate deep-cycle batteries continued to dominate the market as they have done from the SHS Program’s inception. Some companies that produced batteries solely for solar systems have evolved their product line by expanding into selling car batteries, UPS batteries, and other industrial batteries. In this way, the related industry component and systems manufacturing industry that was catalyzed by SHS program has evolved and maintained its sustainability. PAYG technology was introduced in the SHS Program as a pilot in 2016. Indigenous development took place when a suitable foreign supplier could not be found. Products of two companies, SDRS and SolShare, were approved. However, the commercial introduction of the technology at the late stage of the SHS Program when sales were declining proved to be difficult, especially since SHS users objected to retrofitting the units on their SHS. 36 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 37 38 IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN 4.1 Anatomy of a Market Decline 41 4.2 Impact of Competition on SHS Sales 42 4.3 Impact of Declining Markets on PO Operations 45 4.4 Strengthening Security Requirements for Loans to POs 46 4.5 Efforts to Overcome the Loan Default Challenges 46 4.6 IDCOL’s PO Loan Portfolio Quality and Measures to Improve It 47 4.7 Plan for IDCOL Soft Exit from the Program 48 4.8 Key Takeaways on SHS Market Decline 49 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 39 IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN SHS markets in most countries will eventually reach saturation. The rate at which market saturation happens and the SHS business declines to the point of non-viability will vary; it is often a function of factors beyond a project’s control. In Bangladesh, it was driven not by technological obsolescence but by the decline in the number of target market customers mainly due to grid electricity reaching them far sooner than anticipated. This is evident in Figure 16. When the SHS Program was launched in 2003, there were about 15 million unelectrified rural households. This declined slowly to about 13 million by 2013, and then the pace of the decline suddenly accelerated. By 2018, there were fewer unelectrified rural households than the number of SHS installations under the SHS Program and TR/KABITA, and the number was decreasing rapidly as BREB was connecting over 300,000 new consumers monthly. The market for SHS was disappearing rapidly and universal access to electricity was expected by 2021. Unelectrified Rural Households SHS Installations 18,000,000 16,000,000 14,000,000 Number of Rural Households 12,000,000 10,000,000 8,000,000 6,000,000 4,000,000 2,000,000 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Figure 16: Trend in Unelectrified Rural Households and SHS Installations 40 4.1 ANATOMY OF A MARKET DECLINE Beginning in 2014, about two years after the RERED II Project was approved with a significant injection of new funds, and one year after SHS sales peaked at over 861,000 systems in 2013, SHS sales began declining rapidly—dropping to under 3,500 in just four years. The market collapse was predominantly due to the rapid expansion of grid electricity access as described in the previous chapter. There were also other factors that contributed to a perfect storm of declining sales leading to reduced PO revenues, increasing defaults on loan repayments by consumers to POs and POs to IDCOL, and losses to POs as outlined as follows: 2012–2015 Cost of credit to POs increased by 1 percent and loan tenor dropped by 1 year, reducing affordability of SHS to customers or reducing POs’ profit margins. 2014–2015 Sales per PO declined from about 22,000 in 2013 to 8,000 in 2015. Demand trended toward smaller systems to customers in more distant areas. This increased cost of collection and reduced POs’ margins. To reduce their operating costs and overhead, POs reduced loan tenor, which made the SHS less affordable. 2014 With SHS prices declining sharply due to the reduction in minimum battery capacity from three to two days of autonomy per SHS, customers were offered new SHS that were less costly than the balance due on their loans for their old SHS. Some customers abandoned paying for the old SHS to get a new lower cost one. 2015 Political unrest shut down rail, road, and river transport; reduced rural incomes; and led to drop in demand for SHS. 25 2015 and After years of slowly connecting households to the grid, BREB accelerated its pace and began continuing connecting 200–300,000 customers monthly. This raised the expectations among unelectrified rural households that they would likely get an electricity connection soon. 2015–2016 The government’s TR/KABITA Program began giving away SHS to poor households as well as PV systems for public uses (for example, streetlights, schools, clinics). Implemented by local administrations, it supplied 328,000 SHS during 2015–2016. The quality of such systems was uncertain, no arrangements were made for service or warranties, and decisions on who obtained an SHS were politically driven. The expectation of getting a free SHS dampened demand for SHS under the SHS Program. IDCOL took over the TR/KABITA Program management and integrated it into the SHS Program PO network in 2016–2017, helping to give alternative business to POs and retaining the infrastructure to provide spares and service. 2015–2016 Private SHS sales picked up, building on the good reputation of SHS due to the SHS Program and creating more competition for the SHS Program. SHS of uncertain quality were sold through private shops; those without warranties were sold for 20 percent less than SHS sold under SHS Program. The POs could not compete on price as private sellers offered only on-demand repair services, with no requirement for quality assurance, shorter guarantees, and lower selling costs, as the SHS were sold through retail outlets offering many other products. 2015–2018 As sales declined, POs began shutting down sales and service centers not engaged in TR/KABITA. PO staff decreased from 29,000 in 2015 to 6,000 in 2018, placing enormous pressure on remaining staff. Some POs were losing 10 percent of staff monthly and some had increasing fraud at the field level (estimated by some POs at 5 percent of revenues) (GVEP International 2016). Shutting down sales and service centers led to a decline in customer service that further exacerbated debt collection. Collection efficiencies declined from 94 percent or more to the mid-high 80s up to 2015 and dropped thereafter to only 9 percent in 2018 (see Table 8). 2017 Further compounding the challenge were devastating floods that affected 32 districts of the country, hurting SHS sales and hampering debt collection greatly. 25 Wikipedia. 2015. Bangladeshi Political Crisis. https://en.wikipedia.org/wiki/2015_Bangladeshi_political_crisis. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 41 4. IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN Figure 17: BREB Grid Connection Rate Increased Rapidly after 2014 The perfect storm of rapid electrification, competition from 4.2.1 Impact of Grid Expansion on SHS Sales and TR/KABITA and private sales, and natural disasters not only Mitigation Actions reduced new sales of SHS but also dramatically lowered collection rates of POs as shown in Table 8. In 2011, the pace of grid electrification expansion was stagnating (Figure 17). Despite the government’s commitment to universal access to electricity by 2021, there appeared to Table 8: TDebt Collection Efficiency of POs from SHS be little prospect of it accelerating. The slow pace of new grid Customers, 2009 to 2019 electricity consumer connections by BREB from 2012 to 2014 bears out this expectation. Overall Collection Efficiency Year in the Year (%) Under these circumstances, IDCOL estimated in 2011 that the market potential for SHS was about 6 million 2003–2009 94 or better households, five times more than the 1.1 million installed 2010 92 up to the end of 2011. In 2012, there were about 13 million unelectrified rural households; therefore, the estimated 2011 88 market for SHS was about 50 percent of the unelectrified 2012 84 rural households. With the grid electrification pace stagnating, SHS appeared to be a logical alternative. When 2013 88 the planning and appraisal were being undertaken to 2014 87 commit additional financing in 2011–2013 for SHS, the SHS market expansion was robust (Figure 18). Between 2012 and 2015 88 2013, the number of SHS sold increased 41 percent. 2016 52 The government sought additional financing from the 2017 38 development partners to meet this anticipated demand for SHS. The development partners responded positively: 2018 9 • 2012: The World Bank approved the RERED II Project in August 2012, with a US$99.5 million credit. • 2013: Other development partners provided additional 4.2 IMPACT OF COMPETITION ON SHS SALES financing This section focusses on the impact of rapid grid expansion o ADB approved US$80 million (US$78 million credit and and competing SHS government giveaway programs and US$2 million grant). commercial SHS sales on the SHS Program. It discusses o JICA approved US$81.1 million credit. actions that were taken or, in hindsight, could have been o GIZ approved US$16.88 million grant. taken to mitigate such impacts. o DFID approved US$18.6 million grant. o USAID approved US$3.1 million grant. 42 Figure 18: Sales of SHS and TR/KABITA Larger Public Service PV Systems • 2014: The World Bank approved additional financing of It meant that IDCOL was attempting to expand SHS sales US$78.4 million credit for the RERED II Project. while the SHS market was shrinking, while access to information on the accelerating pace of grid expansion Total new financing from the development partners during was lagging. Coordinated planning of on-grid and off- 2012–2014 was US$377.6 million, which could have financed grid electrification efforts was needed at the ministerial about 2.7 to 3 million additional SHS or about half the level, with development partners and between BREB and potential market. IDCOL, but it was lacking. While the development partners Therefore, the decisions taken by IDCOL, the government, committed US$377.6 million in new financing for SHS, they and the development partners to commit additional were also committing even greater resources to BREB for financing for SHS appeared sound. IDCOL added 17 new expanding grid-based rural electrification.27 Electrification POs in 2013 to further support the accelerating SHS sales— efforts were accelerating rapidly on all fronts, with again a justifiable decision. expansion of grid extension, the SHS Program, and the TR/ KABITA Program, without high-level planning coordination. What was not foreseen was the acceleration in the expansion of grid electrification by BREB beginning in IDCOL’s independent Board of Directors presumably could 2014–2015 as shown in Figure 17. BREB accelerated its grid have assisted in this coordination but could not lead it. connections rapidly beginning in 2015, which is continuing The Sustainable and Renewable Energy Development to date. In the 30 years between 1978 and 2014, BREB Authority (SREDA), established in 2012, could have played connected 9.4 million consumers (270,000 connections the coordinating and policy-making role. But in 2013–2014 per year), but in the next 5 years, it connected 17.1 million when the coordination would have been most essential, consumers, with 3.6 million connected in FY2016/17 and SREDA was in its infancy. Similarly, better communication 4 million in FY2017/18. At the same time, reliability and and coordination among the development partners availability of grid power increased dramatically with the and their staff responsible for committing funds for increase in power generation capacity. Therefore, customer electrification investments was needed but missing. expectations of obtaining reliable and better quality grid electricity at low (subsidized) prices increased and the 4.2.2 Impact of Competing TR/KABITA Program and motivation to buy an SHS declined. As shown in Figure 18, Commercial Sales and Mitigation Actions SHS sales under the SHS Program declined steadily and rapidly from 861,000 in 2013, to 727,000 in 2014, and 576,000 The rapid decline in SHS Program sales was not due solely in 2015 while finally falling under 3,500 in 2018. to rapid grid expansion, though it was the dominant cause. Competition for SHS came from two other directions: A sudden acceleration in grid expansion combined with the government SHS giveaway program, TR/KABITA, was inadequate communication and coordination of planning creating the perception that households could get SHS for between grid and off-grid electrification is not unusual. free, and commercial sales were picking up. IDCOL took Similar situations have been observed in other countries.26 several initiatives to prop up SHS sales. Some succeeded. 26 This experience is not surprising as a similar phenomenon was observed in the Sri Lanka SHS Program (ended in 2012), where a greater-than-anticipated increase in grid expansion reduced the market for SHS to the point where it became saturated (IEG 2014). 27 By 2016, BREB completed 57 rural electrification projects costing US$ 1,735 million in financing from development partners (including nearly US$900 million from the World Bank) and government and own financing. Another 12 projects costing US$2,590 million were ongoing in 2016, increasingly with domestic financing, GOB and BREB financing US$1,885 million and US$46 million, respectively, and with US$659 million from development partners. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 43 4. IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN IDCOL takeover of TR/KABITA Program management. As SHS sales slumped, business volume and revenues of POs declined. SHS sales per PO declined from 20,000 in 2012 to 10,000 in 2015 to less than 100 per PO in 2018 (see Figure 19). IDCOL was able to successfully mitigate some of impacts on the POs of declining SHS business by convincing the government to permit IDCOL to take over managing the TR/ KABITA Program and use the POs as implementors. Since the end of 2016, almost all POs (except newly recruited 11) have been engaged in the TR/KABITA Program. from mid- 2016 and March 2019, the POs installed 883,346 SHS and larger PV systems financed by the TR/KABITA Program. IDCOL taking over the TR/KABITA Program management has other benefits to consumers. The SHS and PV systems had to comply with the SHS Program technical standards and warranty and service requirements. The existing infrastructure built for the SHS Program could be deployed to serve the TR/KABITA Program. It permitted the POs to employ their existing trained staff and field offices. The TR/KABITA Program did not have any marketing/ Figure 19: SHS Sales per PO under the SHS Program promotional expenses for the POs and there was no debt collection risk. In addition, about 18 percent of the systems were streetlights and larger, 300–1,500 Wp public service Competition from commercial SHS providers was increasing. systems with higher profit margins. The POs focused more Building on the reputation of SHS created through the SHS on the TR/KABITA sales and less on installations under Program, private SHS sales began accelerating in 2015.28 SHS Program. While installations under the SHS Program Low-priced products were sold component by component for declined, due to TR/KABITA business, the POs’ installations cash with short or no warranties or with variable commitment and profitability did not decline significantly. IDCOL and to provide after-sales services. Other than word-of-mouth, the government deserve considerable credit for taking the such systems and components had no credible quality initiative to transfer the TR/KABITA Program management to certifications, uncertain warranties, no service agreements, IDCOL, which had the added benefit of providing consumers and no means to receive/resolve customer complaints, unlike with better quality products and service. SHS sold under the SHS Program. They competed directly with PO SHS sales. IDCOL estimates these sales were about 30,000 to 40,000 a year—a tenth of the number of installations under TR/KABITA. Effectively regulating this commercial market is extremely difficult since it is impossible to police the retail market, and as many components are assembled locally, enforcement at ports of entry is ineffective. Nevertheless, IDCOL sought SREDA’s help to stop sales of substandard SHS by establishing national SHS standards. The TSC worked with SREDA to develop national standards for SHS. SREDA developed a policy to safeguard quality standards and issued consumer protection guidelines in 2016. Bangladesh Standards and Testing Institution (BSTI) issued national standards in 2017. The RERED Project also supported upgrading of testing facilities to include PV module, battery, and LED lamp testing in 2014–2016. However, establishing standards alone is ineffective without effective enforcement mechanisms. A publicity campaign would be effective in informing consumers of the advantages of purchasing good quality SHS. But this requires a quality label or mark that consumers can recognize. 28 A consequence of the success of the IDCOL brand image was that SHS commercial retailers began selling their products using the ‘IDCOL approved’ or ‘IDCOL Standard’ sticker. In early years, the TR/KABITA Program (before IDCOL was administering it) used these labels though its SHS products had not passed the SHS Program quality certification. 44 Table 9: Example of Comparative SHS Costs in 2013 and 2016 20 Wp Cost 50 Wp Price Cost Category 2013 2016 2013 2016 Change from Change from (Costs in Current BDT) 20 Wp Cost 20 Wp Cost 50 Wp Cost 50 Wp Cost 2013 to 2016 (%) 2013 to 2016 (%) Solar Module 1,710 1,083 −37 4,275 2,518 −41 Battery 3,506 2,280 −35 7,600 5,035 −34 Other Hardware 2,805 2,192 −22 4,294 3,556 −17 Transport 380 380 0 380 380 0 Overhead and Promotion 324 390 20 477 715 50 Taxes 458 328 −28 894 640 −28 Gross Profit and After-Sales 3,017 2,846 −6 11,980 7,147 −40 Service Total Cost 12,200 9,500 −22 29,900 19,990 −33 4.3. IMPACT OF DECLINING MARKETS ON PO OPERATIONS As sales declined and POs struggled to attract customers, there was also a decline in sales margin of POs, as evident from the SHS cost breakdown comparison in 2013 and 2016 shown in Table 9. Key observations from this table are as follows: • Solar module cost declined from US$1.09 per Wp in 2013 to about US$0.67 per Wp in current US$—nearly a 40 percent cost decline. During this period, international PV module costs hardly changed (Fu et al 2018; Regan 2018;). See the example in Figure 20. • Battery cost decline can be attributed to reduction in size of battery from three to two days of autonomy. • Even though other hardware costs (module support, Source: National Renewable Energy Laboratory 2018 wiring, controller, lamps, switches, and so on) are unlikely Figure 20: Global Solar PV Module Cost Trend to decline in price, their costs dropped 17–22 percent, indicating aggressive price reduction, especially during a loan defaults. Higher interest rates would have reduced period when general prices inflated 20 percent between demand, leaving POs and IDCOL with few or no good options. 2013 and 2016. IDCOL investigated the use of PAYG technology to reduce • Transport cost remained unchanged between 2013 and cost of debt collection and improve collections. The GVEP 2016 even though there were fewer and more dispersed International (2016) investigation calculated that the customers in 2016 compared to 2013 and more small manual debt collection cost was BDT 1,047 a month (47 capacity SHS were being sold. percent of the SHS price of a 30 Wp SHS). In contrast, they • Gross profit and after-sales service margins declined 6 estimated that collection cost for a PAYG system for the percent for the 20 Wp SHS and 40 percent for the 50 Wp. same 30 Wp SHS could have been 12 percent of the SHS price. Unfortunately, it was too late to introduce the PAYG The only cost component to increase was Overhead and technology to the SHS Program beyond a pilot scale. Promotion. However, the absolute amounts were small compared to other cost components. Overall, the SHS costs The manual accounting and financial control systems declined 22 percent for the 20 Wp and 33 percent for the 50 of most POs were unable to detect these losses in time. Wp SHS in current BDT terms in the three years, largely due IDCOL did require POs to use a mandatory enterprise to a decline in the margins of the POs. resource planning (ERP) system to improve transparency in accounting and financial control, but it was too late to The market collapse, the decrease in debt collection rates, reverse course. and increasing field-level fraud led to collection and service costs rising from 15 percent to 50 percent of the SHS sales As the profit margins of POs declined and their businesses price, resulting in losses for the POs and eating into their shrank, they found it increasingly difficult to repay their financial reserves (GVEP International 2016). Interest rates to loans to IDCOL. Collection efficiencies continued to drop consumers remained unchanged despite higher cost of debt (Table 8), the POs were increasingly unable to meet their collection due to a larger proportion of smaller loans to a repayment obligations on the original terms, and the quality smaller number of more dispersed customers and increased of IDCOL’s loans to the POs deteriorated. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 45 4. IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN Table 10: Strengthening Security Requirements for Loans to POs 2003–2008 • Maintaining a balance in the DSRA equivalent to one semiannual installment payment • First charge hypothecation on all floating assets of POs 2009 • Lien on all project accounts (Proceeds Account and DSRA) • Demand promissory note and letter of continuity • First charge hypothecation on all fixed and floating assets of POs • Personal guarantee from the directors/shareholders • Corporate guarantee from the concerned third parties • Letter of comfort from the governing board of the POs 2012 • Lien on all project accounts (that is, Proceeds Account and DSRA) • Maintenance of minimum required balance in DSRA equivalent to 4 quarterly installment payments • Mortgage of land or bank guarantee to secure 20% of the outstanding refinance amount • Demand promissory note and letter of continuity • First charge hypothecation over all fixed and floating assets of POs • Personal guarantee of all directors, if it is a limited company, or of the executive director/managing director/CEO/ chairman/key person of the PO to the satisfaction of IDCOL, if it is an NGO/MFI/society/foundation • Letter of comfort from the executive committee/governing board of the PO if the PO is NGO/MFI/society/foundation • Lien on all project accounts (that is, Proceeds Account and DSRA) 2016 • Maintenance of minimum balance in DSRA equivalent to 2 quarterly installment payments • Bank guarantee to secure 20% of the outstanding refinance amount (from the new POs enlisted in 2015) • Charge documents (demand promissory note and letter of continuity) • Undated cheques (from the new POs enlisted in 2015) • CIB undertaking (from the new POs enlisted in 2015) • Obtain corporate guarantee from the sister concern or concerned third-party of the PO, as applicable 4.4 STRENGTHENING SECURITY 4.5 EFFORTS TO OVERCOME THE LOAN DEFAULT REQUIREMENTS FOR LOANS TO POS CHALLENGES Unlike in the MFI lending sector where the loans were IDCOL was in a tough situation in the later stages; it had to often backed by solidarity group security, the loans continue disbursement to the POs to ensure operation of to the POs were unsecured, other than minimal DSRA their SHS Program. Otherwise, POs would not be able to requirement. Beginning in 2009, IDCOL strengthened the continue installation of SHS, further exacerbating the problem security requirements and then further strengthened of making debt service payments to IDCOL. Instead, IDCOL them in 2012 and again in 2016 (see Table 10). attempted other means to collect outstanding debt and ease the pain of making such payments. Efforts made by IDCOL Unfortunately, these measures were largely ineffective included the following: as PO collection efficiencies continued to drop. IDCOL had limited recourse to compel the POs to meet their • 2015: Implementing Collection Efficiency Improvement debt obligations. IDCOL loans to POs were only partially Program (CEIP). This was a joint effort of IDCOL and POs to securitized, and PO loans to consumers were securitized improve collection performance of POs. Its objectives were to only with the SHS asset. Most of the POs, 77 percent, increase collection from overdue customers, reduce employee were NGOs (foundations, societies, and MFIs). As their dropouts for POs, ensure regular customer visit by POs to executive committees are salaried personnel, their ensure after-sales service, strengthen relationships with local personal guarantees could not be legally enforced. administrations, and improve coordination between IDCOL Moreover, creation of hypothecation charge with the inspection teams and POs’ field forces. The CEIP had some Registrar of Joint Stock Companies was not possible. It impact on improving PO collection performance. But the was also not practical for IDCOL to take control over the principal problem remained—the high cost of collection and underlying assets such as SHS. The POs did not provide reduction in new sales had compelled POs to shrink their a legal mortgage of land or bank guarantee. None could operations in many areas. maintain required DSRA balance, as noted earlier. When • 2015: Seeking BREB support. At the request of IDCOL, BREB IDCOL strengthened the security requirements in 2016, advised its field officials to collect clearance certificate from many POs refused to comply. the respective PO before giving new electricity connection to Moreover, these securities were not enough to cover the an SHS customer. This proved ineffective as this requirement full exposure of the loan as SHS value had depreciated, was in direct conflict with the BREB objective of maximizing SHS costs fell sharply over time, and the repossession of electrification connections. SHS from homes was both impractical and costly. IDCOL • 2015–2016: Taking over administration of TR/KABITA could claim the debt from the POs, but obtaining the Program and enlisting the POs to supply and install SHS and amounts due was challenging. other public systems. This added business and revenues to PO operations as discussed previously. 46 Table 11: PO Loan Status 2009–2018 Standard SMA Substandard Doubtful Bad/Loss Year Amount Amount Amount Amount Amount No. (BDT, No. (BDT, No. (BDT, No. (BDT, No. (BDT, millions) millions) millions) millions) millions) 2009 10 1,140 — — — — — — — — 2010 13 5,727 — — — — — — — — 2011 22 10,743 — — — — — — — — 2012 29 17,156 — — — — — — — — 2013 37 22,073 — — 1 280 — — — — 2014 40 20,791 4 1,495 1 289 1 287 — — 2015 40 21,346 5 766 3 2,128 — — — — 2016 44 19,639 2 1,250 3 613 1 107 1 1,054 2017 39 14,563 9 3,607 1 990 — — 2 1,084 2018 38 5,725 8 10,456 1 170 1 232 2 1,084 (%) (76) (32) (16) (59) (2) (1) (2) (1) (4) (6) 2019 34 12,525 5 345 4 599 3 1,448 — — (%) (74) (84) (11) (2) (9) (4) (6) (10) • 2015–2017: Retrofitting PAYG meters to SHS. PAYG meters balance of BDT 2,950 million (US$36 million) which was the were introduced for retrofitting to existing SHS. It was only collateral under the program. Therefore, BDT 14,717 not until 2017 that their use was mandated. The delay million (US$179 million) of IDCOL loan was unsecured, was due to the PAYG technology having to be indigenized which was 33 percent of the total loan extended to the POs. when the foreign supplier withdrew as they felt their This was about 2.5 times the paid-up capital of IDCOL. This market in Bangladesh was too small. The product cost was equivalent to about BDT 3,600 (US$43) per SHS. BDT 2,000 (US$24) and was too costly for the small 20 Wp SHS (US$120) which dominated sales. SHS users objected; The number and amount of loans classified as Special they refused to pay a portion of this cost. Users were also Mention Account (SMA) or worse began growing from 2014 suspicious that this unit somehow used up part of their as shown in Table 11. The total amount of loans classified as electricity. The PAYG effort failed. SMA or worse in 2018 was BDT 11.9 billion (US$145 million). IDCOL is required to make provisions on loans to its POs in compliance with the requirements of Bangladesh Bank. This 4.6 IDCOL’S PO LOAN PORTFOLIO QUALITY includes Standard (1 percent), SMA (5 percent), Substandard AND MEASURES TO IMPROVE IT (20 percent), Doubtful (50 percent), and Bad/Loss (100 percent) which bears an adverse impact on the capital Under the SHS Program, as of December 2018, IDCOL adequacy ratio, in case of changing the status of loans.29 disbursed BDT 45.45 million (US$596 million) loans to the POs; a major portion was unsecured. As of December 2018, Since then, due to proactive efforts by IDCOL to improve POs repaid BDT 27,590 million (US$361 million) loans to the loan portfolio quality, the share of loans classified as IDCOL, which was 61 percent of the total loans disbursed. Standard (not at risk) increased to 84 percent of the total In addition, IDCOL received BDT 11,860 million (US$155 in 2019 from only 32 percent in 2018. IDCOL considered million) from the POs as interest. the situation and rescheduled 15 loan accounts as per Bangladesh Bank guidelines, including reducing interest The decline in sales, lack of SHS market, and the withdrawal rate on loans to 0 percent as per agreement with ERD. of POs from the SHS business and the consequent loan Repayment schedules of these POs were revised in line delinquency had a negative effect on IDCOL’s financial with their revenue stream from both the SHS Program and soundness. When collection efficiency of POs from TR/KABITA Program. Repayment duration of 11 PO loan customers dropped, POs’ loan repayments to IDCOL also accounts was increased by three years (extension from dropped. 2023 to 2026), resulting in reduced installment amount per quarter. Repayment duration and quarterly payments As of December 2018, the total amount of IDCOL loan of another account remained unchanged while the outstanding to the POs was BDT 17,667 million (US$215 durations of 3 other accounts were reduced while quarterly million). DSRAs maintained by the POs with IDCOL had a 29 Bangladesh Bank’s loan status classification: SMA - remain overdue for two to three months; Substandard - past due/overdue for three months or beyond but less than six months; Doubtful - past due/overdue for six months or beyond but less than nine months; and Bad/Loss - past due/overdue for nine months or beyond. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 47 4. IMPACT OF DECLINING SHS SALES AND MITIGATION ACTIONS TAKEN installments remained the same. As a result of these plan an orderly exit from the SHS Program (Box 5). actions, the value of the loan portfolio classified as Standard increased BDT 12.5 billion in 2019, compared to only BDT 5.7 The government and IDCOL acted on key billion in 2018. recommendations of the study by mid-2018. Recognizing the fiscal and other benefits that the IDCOL SHS Program Owing to IDCOL’s proactivity, as of 2019, the amount of has provided, the government reduced the interest rate outstanding loans that were classified as substandard on IDCOL’s loan from 3 percent per year to 0 percent per dropped to BDT 2,392 million (US$28.5 million in 2018 US$) year with effect from July 1, 2018. IDCOL in turn reduced or less than 5 percent of IDCOL’s total PO cumulative loan the interest rate from 4 percent per year to 0 percent per portfolio under the program of US$596 million. To put this year on outstanding SHS loans with the POs with the in another perspective, it amounts to about US$7 per SHS. same effectiveness date, as well as revising the repayment This potential loss is minimal in comparison to the benefits schedule as noted previously. IDCOL approached accrued to major stakeholder groups as detailed in Chapter 5. Bangladesh Bank to relax mandatory provisioning requirement for SHS loans so that IDCOL can build up adequate provision amount from the future revenue 4.7 PLAN FOR IDCOL SOFT EXIT FROM THE earnings against the classified loans. Bangladesh Bank PROGRAM advised IDCOL to request for specific loan accounts when those would become classified. IDCOL has restructured In 2017, IDCOL commissioned a study by the Bangladesh its SHS loans to match the cash flows from the TR/KABITA Institute of Development Studies (BIDS) to better Program and collection of installments. These actions have understand the underlying issues and determine other increased the probability of full collection of outstanding means to resolve the debt problem and enable IDCOL to loans that are not rated substandard or doubtful. BOX 5: Study to assess the SHS market situation and recommend an action plan for IDCOL’s soft exit from the program Findings: • There were about 1.2 million defaulters. Most owned connections by BREB and free distribution of SHS the small 20 Wp SHS, costing about US$120. The under the TR/KABITA Program ended up creating average default amount was US$110; defaulters have market distortions and harming the SHS Program. yet to pay 36.5 percent. • IDCOL’s financing under flexible conditions had • SHS loan repayment defaulters are slightly more created a moral hazard among the POs and wealthy than non-defaulters. About 65 percent of therefore POs pursued an aggressive and risky defaulters are willing to pay the due installments. marketing strategy for SHS, without filtering out bad customers. • Defaulters’ view: The main reason for default was financial constraints, followed by POs’ poor after- • IDCOL’s financial involvement with the POs was not sales service, higher price compared to open market fully securitized from the beginning. Refinancing SHS, and natural disasters. to the POs was made with rather loose terms and conditions. IDCOL may have the legal right to claim • POs’ view: Defaulters feel that the price of SHS in the from the POs, but practically obtaining the dues open market was less than what they have already can be extremely time-consuming and expensive. paid. They pressured POs to reduce prices, which reduced the profit margin from 12 to 2 percent the Options recommended: previous year. • Seek Bangladesh Bank approval for giving IDCOL a • There was a conflict between welfare and commercial longer time for provisioning for the default. objectives of the SHS Program. There should have • Seek repayment of IDCOL SHS loans to the been a body in place to coordinate policy among the government at a lower interest rate. stakeholders and provide guidelines for IDCOL and similar organizations. Though SREDA is that kind of a • Seek the government’s ‘no-objection’ to permit policy body, it has not been entrusted to do this. IDCOL’s largest debtor PO to participate in the TR/ KABITA Program. • Lack of policy coordination among the stakeholders such as BREB, IDCOL, the Ministry of Disaster • Agree on a time-bound rescheduled repayment Management and Relief, and so on meant that plan with the POs. Take legal action if an agreement positive interventions such as rapid expansion of grid cannot be reached or the agreement is violated. Sources: BIDS 2018 and IDCOL 2018. 48 4.8 KEY TAKEAWAYS ON SHS MARKET DECLINE KABITA Program, earlier and better coordination to ensure the use of common standards and prevent overlap could The SHS market could not sustain the multipronged have resulted in better outcomes and more efficient use competitive pressure, especially unexpectedly rapid of resources. grid expansion from 2015. The PO sales and collection performance significantly worsened. Installment collection • It is important to have a clear goal for any SHS program alone was not enough to sustain branch offices which had and to foresee its eventual end, in relation to this goal. to close, further affecting ability of POs to market and sell With the Bangladesh SHS Program, the goal was to SHS. Moreover, due to debt collection difficulties, IDCOL provide electricity to households in advance of the stopped providing new credit support in 2017. While the coming of the grid through a program that sold SHS program continues to operate, it is in its final stages with the on affordable but near commercial terms with credit. focus being on loan repayments of customers to POs and Coordinated planning by a rural electrification authority from POs to IDCOL. It is expected to close in 2021. After 2021, could have recognized that the market would be operation and maintenance support of SHS installed under saturated and the SHS Program would need to be brought the program will be supplied by any POs that still offer such to an orderly close while ensuring that the SHS installed services outside of the program or by commercial system under the program would continue to receive after-sales suppliers. service. This transition, while now under way, could have been foreseen with less pain to IDCOL and the POs in the Key takeaways from the period of market decline post 2013 process. are as follows: • If the intent of an SHS program is to deepen access to • A high-level oversight body within the government is electricity within communities, then a purely market- needed to take responsibility for planning and policy based program alone is not suitable as sales will be development of parallel on-grid and off-grid electrification to those with ability to pay. This was evident from programs to ensure that they complement each other Figure 11 which showed that SHS market penetration to achieve overall access goals, rather than competing. in some districts was above 50 percent, but in some From 2013 to 2018, the GOB was rapidly expanding predominantly unelectrified districts, it was barely 10 three parallel electrification programs without such percent. Even if smaller SHS with more limited service coordination: (a) BREB was accelerating grid extension, were offered, poorer households may have other priorities (b) the TR/KABITA Program was increasing provision of (for example, food), may not be willing to risk making a PV systems at no cost for public use and SHS for poor long-term financial commitment to buy on credit, or may households; and (c) the SHS Program was expanding not qualify for a loan. Incentives such as those offered sales of SHS to customers on near commercial terms for grid electrification may be needed to make SHS through the POs. All three programs were successful, but affordable to poorer consumers—similar in principle to the first two programs resulted in the disappearance of lifeline tariffs offered for grid electricity. The TR/KABITA the market for the SHS Program after 2015. While this Program is clearly one option to make SHS affordable to can be seen clearly in hindsight, the absence of high- poorer households, but it is subject to ‘leakage’ and SHS level planning and coordination meant that it was not may not reach the intended poorer consumers because of seen in 2013–2014 when major new resources were the inadequate selection process of recipients. being committed to the SHS Program. Two types of • POs diversifying to related business areas helped coordination are needed: them remain in business and meet their obligations to o Close coordination with and careful monitoring of grid customers and IDCOL. IDCOL achieved this by convincing expansion progress is necessary to adjust SHS sales the government to permit POs to also undertake TR/ expectations and plans to remaining market potential. KABITA installations with implementation supervision by In the extreme situation in Bangladesh, the market IDCOL. disappearance was so sudden that little could be done • Broader consumer awareness is needed even in a mature by the time it happened. Had grid expansion been more market to convince customers. A key success of the IDCOL gradual, POs could have been incentivized to market in SHS Program was establishing its brand image in rural districts where grid expansion would be delayed or where communities. Off-grid consumers were aware of the existing POs are not operating. quality of products approved by IDCOL. In other countries, o Close coordination and joint planning of competing in the absence of any form of quality recognition (such as off-grid electrification programs are also necessary. The Lighting Global),30 such broad information dissemination various development partners agreeing to adopt the as undertaken by IDCOL is necessary to sensitize same modalities of the SHS Program was of great benefit. customers. On the other hand, the TR/KABITA Program was initially run independently of the SHS Program. While the SHS Program eventually took over management of the TR/ 30 See Lighting Global (2020. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 49 50 ECONOMIC AND FINANCIAL ANALYSIS 5.1 EIRR: SHS Program Rate of Return to Society as a Whole 53 5.2 Financial Analysis of the Net Benefits of the SHS Program 56 5.3 Summary of Distribution of Net Financial Benefits Among Stakeholders 61 5.4 Analysis of ODA Flows for the SHS Program on the GOB 62 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 51 ECONOMIC AND FINANCIAL ANALYSIS This chapter uses financial and economic analysis of the SHS Program to estimate the indicative gains and losses to various stakeholders—society, consumers, service and technology providers, IDCOL, the government, and global society—over the half century from the start of the program in 2003 until the final concessional loan repayment is made by the Government in 2054.31 The chapter addresses three related questions: (a) Are the total economic benefits of the SHS Program enough to pay its total economic costs at the societal level? In other words, is the SHS Program desirable overall? (b) On a financial basis, did the households benefit sufficiently from the program to justify their costs? (c) What were the net financial benefits or costs to other key program participants and those affected by it—the government, IDCOL, the POs, and kerosene distributors? The first question is readily answered through a traditional World Bank economic internal rate of return (EIRR) calculation for the SHS Program. The second question is answered by financial analysis at the aggregate household level. The third question is answered by analyzing the program’s net financial impacts on the other program participants, some gaining and some losing. Methodological issues and assumptions and data are provided in Appendixes D through F. It is important to acknowledge from the outset that the economic and financial analyses of the SHS Program presented here are highly simplified and have important limitations. First, household benefits are based on a simple measure in both the economic and financial analyses—the avoided cost of kerosene and in later stages grid electricity for lighting. This simple measure of avoided kerosene/ grid electricity costs for lighting greatly underestimates the benefits to households. As highlighted in Section 2.9, use of an SHS has many other benefits including, among others, (a) improving quality of life (for example, more hours of study, household work or leisure, increased safety, and more access to information through radio or TV); (b) providing other immediate financial benefits (for example, reducing cellphone charging costs or permitting extra hours of productive activity); and (c) providing valuable health and education benefits in the longer term. These other benefits are excluded from the analysis in this chapter because they are difficult to estimate and the avoided kerosene costs alone justify the program in economic and financial terms. A second limitation is that the analyses rest on several estimates and assumptions, for example, the amount of kerosene saved per household, the profits of participating POs, and the losses of kerosene dealers. Given the underestimation of benefits to households, the results of the economic and financial analysis must be used with care. The estimates of net economic and financial benefits must not be confused with the overall development impact of the SHS Program. The other benefits identified above are significantly more important to the households and the government than the kerosene savings obtained from the program. Chapters 1 and 2 show that the SHS Program fully met the government’s development objective of bringing many of the benefits of electrification to a significant share of rural households in advance of the availability of the grid. 31 Together, the grace periods and repayment periods for official development assistance (ODA) loans and credits stretch out the overall net cash flow for the nation to 2054. 52 Section 5.1 presents the overall economic analysis of the Two variations of benefit estimates are made from the SHS Program. Section 5.2 contains financial analysis of base case. The first variation adds to the avoided cost of the impact of the SHS Program, first presenting a detailed kerosene/grid electricity, the global benefits of the GHGs financial cost-benefit analysis for households (Section avoided by not burning kerosene for lighting until 2022 5.2.1) and then estimating the net financial benefits for all and from avoided grid electricity generation after 2022. The stakeholders (Section 5.2.2). Section 5,3 analyzes the impact second variation estimates benefits based on an estimate of of ODA flows for the SHS Program on the government. the WTP of households for lighting that includes the avoided cost of kerosene and an estimate of the consumer surplus that is gained by the user through the improved quantity of 5.1 EIRR: SHS PROGRAM RATE OF RETURN TO light with the SHS (IEG 2008). Since this considers that the SOCIETY AS A WHOLE SHS provides more light, the benefit estimate based on WTP is higher than in the base case. The benefit to society as a whole is estimated using an economic cost-benefit analysis similar to that used by the The base case economic analysis in table 12, with benefits World Bank for conventional rural electrification projects. based only on the kerosene/grid electricity saved for The economic benefits and costs of the Bangladesh SHS lighting, shows that the SHS Program benefits Bangladesh Program are estimated for 2003–2029 for all participating society substantially with an EIRR of 20 percent. When households. The economic analysis is done from society’s global benefits due to GHG mitigation are added to the point of view; it excludes transfer payments such as grants, value of kerosene/grid electricity savings, the EIRR increases taxes, and subsidies and uses international or border prices to 25 percent. When the alternative benefit estimate uses for traded goods and ‘shadow prices’ for non-traded goods the WTP for lighting of US$2.23 in 2018 US$ per kWh in (see Appendix D for details). Bangladesh rural areas (see table 12 note), the EIRR is higher at 51 percent, recognizing the improved quantity and The economic analysis considers as costs the stream of quality of light.32 costs of the initial SHS cost to the households as well as the replacement costs of components over the life of the The EIRR of the SHS Program is robust. Switching value system. It assumes a 12-year life for the solar modules and analysis for the most conservative case where benefits are replacement of shorter-lived assets at regular intervals (see measured as savings in kerosene/grid electricity shows that Appendix D for details on component lives). Thus, the 2018 kerosene offset can be reduced by 25 percent before the tranche of SHS units can remain in service until 2029. EIRR drops to 10 percent (that is, average kerosene offset during 2003–2018 reduced to 0.23 liters per day per SHS The benefits in the base case are conservatively estimated from 0.31 liters per day per SHS). until 2022 as the avoided cost of kerosene for lighting to the households. That is, the benefits are estimated as the value The EIRR analysis clearly indicates that the total benefits of the kerosene saved for lighting. From 2022 onward, the easily pay for the costs of the SHS Program, even when the grid is assumed to be universally available and the avoided benefits are underestimated by valuing only benefits from cost is based on the electricity from the grid for lighting avoided kerosene/grid electricity costs of SHS households that is saved by using the SHS. As the avoided cost of grid for lighting and excluding the other important benefits electricity for lighting is much lower than kerosene, benefits identified in Section 2.9. drop from 2022 to 2029. It should be noted that the savings from the use of the SHS are estimated only for lighting. 32 This compares to estimates of WTP for lighting in other countries ranging from US$0.47 to US$3.37 (IEG 2008, 41). LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 53 54 Table 12: Economic Analysis of the SHS Program 2003–2029 (Constant 2018 US$, millions) SHS Costs Value of Kerosene and Base Case Variation 1: Variation 2: Cost Net GHG WTP to (Including Taxes on Kerosene and Kerosene Electricity Net Economic Net Economic Net Economic Year of Grants Mitigation Switch to Replacements) SHS Electricity Subsidies Economic Cost Benefits: Kerosene Benefits: Base Case + Benefits: WTP and Taxes Value SHS without Grants Saved Saving Benefits Savings only GHG Mitigation Columns 1 2 3=1-2 4 5 6=4+5 7=6-3 8 9=7+8 10 11=10-3 2003 7.88 0.84 7.03 0.46 0.03 0.49 (6.55) 0.10 (6.45) 1.41 (5.62) 2004 15.69 1.68 14.01 1.59 0.37 1.96 (12.05) 0.31 (11.75) 4.74 (9.27) 5. ECONOMIC AND FINANCIAL ANALYSIS 2005 21.68 2.32 19.36 4.11 0.34 4.45 (14.91) 0.62 (14.29) 11.65 (7.71) 2006 29.23 3.13 26.10 8.06 1.74 9.80 (16.30) 1.09 (15.20) 20.34 (5.75) 2007 52.41 5.62 46.80 14.79 1.83 16.62 (30.18) 1.95 (28.23) 34.47 (12.32) 2008 87.52 9.38 78.14 31.67 2.44 34.11 (44.04) 3.35 (40.68) 66.79 (11.35) 2009 114.83 12.30 102.53 35.85 (0.40) 35.45 (67.08) 5.48 (61.60) 86.56 (15.97) 2010 203.46 21.80 181.66 73.17 3.29 76.46 (105.20) 9.54 (95.66) 158.99 (22.67) 2011 250.56 26.85 223.72 146.98 35.81 182.79 (40.93) 15.46 (25.47) 302.53 78.82 2012 319.41 34.22 285.19 221.05 87.69 308.73 23.55 23.76 47.30 435.65 150.46 2013 379.59 40.67 338.92 307.15 44.55 351.70 12.78 34.95 47.73 573.23 234.31 2014 217.54 23.31 194.23 343.52 52.45 395.96 201.73 44.80 246.54 620.67 426.44 2015 149.38 16.00 133.37 265.76 (96.75) 169.01 35.63 53.20 88.83 478.96 345.59 2016 40.12 4.30 35.82 245.93 (130.37) 115.56 79.75 56.70 136.45 436.41 400.59 2017 7.36 0.79 6.57 283.95 (74.32) 209.63 203.06 58.41 261.47 498.56 491.99 2018 0.84 0.09 0.75 335.67 17.59 353.26 352.50 59.38 411.88 587.77 587.01 2019 — — — 329.26 17.25 346.51 346.51 59.26 405.77 547.14 547.14 2020 — — — 319.03 16.72 335.75 335.75 58.89 394.64 527.78 527.78 2021 — — — 304.23 15.94 320.17 320.17 57.56 377.73 494.68 494.68 2022 — — — 16.85 — 16.85 16.85 — 16.85 — — 2023 — — — 14.21 — 14.21 14.21 — 14.21 — — 2024 — — — 10.79 — 10.79 10.79 — 10.79 — — 2025 — — — 6.73 — 6.73 6.73 — 6.73 — — 2026 — — — 3.59 — 3.59 3.59 — 3.59 — — SHS Costs Value of Kerosene and Base Case Variation 1: Variation 2: Cost Net GHG WTP to (Including Taxes on Kerosene and Kerosene Electricity Net Economic Net Economic Net Economic Year of Grants Mitigation Switch to Replacements) SHS Electricity Subsidies Economic Cost Benefits: Kerosene Benefits: Base Case + Benefits: WTP and Taxes Value SHS without Grants Saved Saving Benefits Savings only GHG Mitigation 2027 — — — 1.02 — 1.02 1.02 — 1.02 — — 2028 — — — 0.18 — 0.18 0.18 — 0.18 — — 2029 — — — 0.02 — 0.02 0.02 — 0.02 — — NPV (10%) 779.72 83.54 696.18 916.58 19.66 936.23 240.06 143.13 383.19 1,685.47 989.29 EIRR 20.1% 25.2% 51.3% WTP per kWh of Lighting from SHS Discount rate (%) 10 5 NPV kWh used for lighting only (kWh) 754,644,574 1,430,174,623 NPV WTP for lighting (Constant 2018 US$) 1,685,466,283 3,058,926,698 WTP/kWh (2018 US$/kWh) 2.23 2.14 Based on methodology in Meier (2003). Lumen output per kupi lamp is 0.786 kilo-lumen/liter of kerosene and consumes, 0.0102 liters/hour; and from a hurricane lamp it is 1.513 kl/liter and consumes 0.0298 liters per hour of fuel (computed from data in Mills, 2011). Demand is assumed to be inelastic with an elasticity of 0.25. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 55 5. ECONOMIC AND FINANCIAL ANALYSIS Sensitivity to Kerosene Use Avoided Assumptions Average Kerosene Avoided/ EIRR SHS Size and Data Source SHS (Liters/Day) Kerosene Saved Only Kerosene + GHG Avoided WTP 0.10 BIDS -8.5% -7.7% 32.7% 0.19 To achieve EIRR of: 5% Sensitivity Analysis 0.23 To achieve EIRR of: 10% 0.28 GS CDM (40–74 Wp) 15.9% 20.3% 48.6% 0.31 This study (10–300 Wp) 20.1% 25.2% 51.3% 0.42 GS CDM (75–119 Wp) 34.8% 42.8% 60.0% Note: GS = Grameen Shakti; CDM = Clean Development Mechanism. Assumptions: BIDS/World Bank (2012) survey reported that average kerosene consumed per household was 2.91 liters per month before obtaining SHS, without discrimination by SHS size. UNFCC (2013) Grameen Shakti CDM application credited 40–74 Wp SHS would displace 2 kerosene lanterns that operate 4 hours per day for 340 days and use 47 liters per year, and 75–119 Wp SHS would displace 3 kerosene lanterns that use 143 liters per year. This investigation used Grameen Shakti Survey results: Weighted average kerosene offset - a 20 Wp SHS replaces 1 hurricane lamp and 1 kupi (bottle, open flame) lamp used 4 hours per day on average; a 40 Wp SHS replaces 2 hurricane lamps and 1.5 kupi lamps operating 3.8 and 1.9 hours per day, respectively, on average. A 50 Wp and larger SHS replaces 2.5 hurricane lamps and 1.8 kupi lamps, operating 4.5 and 2.3 hours per day, respectively, on average. SHS service is available 340 days per year. 5.2 FINANCIAL ANALYSIS OF THE NET BENEFITS 16 percent. The financial benefits to the households are the OF THE SHS PROGRAM avoided cost for kerosene for lighting until 2022 when the grid is considered to be universally available and then the This section addresses the net financial gains and losses avoided cost for grid electricity for lighting after 2022, both among the different stakeholders of the Bangladesh SHS based on the actual prices that would have been paid by the Program. The financial analysis is based on the real-world households including subsidies. transactions of the main stakeholders undertaken because of the SHS Program. The government made loans to IDCOL Table 13 summarizes the household-level financial benefits for the program, its kerosene subsidies were affected by and costs of the SHS Program during 2003–2029. For details the reduction in kerosene used by SHS households, and on costs and benefits, see Appendix D. The household-level it benefited from taxes on SHS and components. IDCOL analysis shows that avoided spending on kerosene and grid managed the program and made loans to POs, the POs sold electricity for lighting over the life of all the SHS units more SHS and made loans to households, households purchased than paid for the purchase of the SHS units and resulted in and operated the systems and repaid the SHS loans, and they a financial internal rate of return (FIRR) to the households purchased less kerosene as a result, which in turn reduced the of 17.2 percent. This evaluation accounts for households profits of kerosene dealers. Section 5.2.1 contains a financial defaulting on a portion of their loans, especially in the later analysis of the impact of SHS on the aggregate participating years as shown in Table 8. If there had been no defaults, households. Section 5.2.2 broadens the analysis to consider the FIRR would have dropped to 13 percent since the the net financial impact of the activities of the program on households would have made higher loan repayments. all stakeholders including households, POs, IDCOL, kerosene dealers and the Government. Two theoretical cases were also analyzed in Table 13. If there had been loans but no grants, the FIRR would have been 16.4 percent. Although the grants increased the FIRR by only 5.2.1 Aggregate Household-Level Financial Analysis 1 percent, it is likely that they nevertheless helped achieve early spread of SHS installations. Grants reduced the risks to This section first analyzes the financial costs and benefits the early adopters at a time when the SHS technology was to the households that purchased SHS. In the financial unfamiliar to them. If there had been no loans and no grants, analysis, all costs are based on the actual costs to the the FIRR would have been even lower at 14.7 percent (though stakeholders, including subsidies and taxes. The costs far fewer households would have been able to afford an SHS). are based on the actual costs to the households for the purchase of SHS and replacement parts, considering Households that purchased SHS from 2010 onward lost the any grant available on the initial cost. The systems were full benefit from avoiding kerosene use as the SHS replace purchased by the households with financing from the POs cheap electricity from the grid rather than kerosene after the including a 15 percent deposit and the remainder of the cost grid arrives in 2022. But, of course, the households switching financed at an average of 14 percent flat-rate interest for a to grid electricity got the benefit from potentially obtaining three-year period. Actual flat-rate interest varied from 12 to unlimited quantity of electricity at low prices. 56 Table 13: Financial Analysis of the SHS Program from the Household Stakeholder Group Perspective (Constant 2018 US$, millions) SHS Costs Household Equity HH Loan Household Benefits: Avoided SHS Theoretical Net Theoretical Net Loans from Year Including and Purchase of Grant Repayments at Cost with Grant Kerosene Purchases and Program Net Benefit with Grant Benefit with No POs to HHs Replacements Replacements 14% Interest and Loan Grid Electricity Saving HH Benefit but No Loan Grant or Loan (1) (2) (3) (4 = 1 − 2 − 3) (5) (6 = 2 + 5) (7) (7 − 6) (7 − (1 − 3)) (7 − 1) 2003 7.88 3.10 1.05 3.73 0.83 3.93 0.46 (3.48) (6.37) (7.42) 2004 15.69 6.21 1.89 7.59 3.35 9.56 1.59 (7.97) (12.22) (14.11) 2005 21.68 8.70 1.81 11.17 7.52 16.22 4.11 (12.12) (15.77) (17.57) 2006 29.23 11.83 1.75 15.64 12.66 24.49 8.06 (16.43) (19.41) (21.17) 2007 52.41 21.31 2.50 28.60 19.98 41.30 14.79 (26.51) (35.13) (37.62) 2008 87.52 35.54 4.51 47.47 32.73 68.27 31.67 (36.61) (51.35) (55.86) 2009 114.83 44.08 6.87 63.88 51.54 95.62 35.85 (59.77) (72.11) (78.98) 2010 203.46 78.32 10.72 114.42 79.63 157.95 73.17 (84.79) (119.58) (130.29) 2011 250.56 96.60 12.23 141.73 113.67 210.27 146.98 (63.30) (91.36) (103.59) 2012 319.41 128.57 15.33 175.51 149.44 278.01 221.05 (56.96) (83.03) (98.37) 2013 379.59 153.98 10.33 215.29 200.81 354.79 307.15 (47.64) (62.12) (72.44) 2014 217.54 87.80 8.86 120.88 215.00 302.80 343.52 40.71 134.84 125.98 2015 149.38 50.63 7.22 91.54 195.64 246.27 265.76 19.49 123.59 116.38 2016 40.12 13.54 2.32 24.25 81.76 95.30 245.93 150.63 208.14 205.82 2017 7.36 2.50 0.32 4.54 32.11 34.61 283.95 249.35 276.91 276.60 2018 0.84 0.29 0.03 0.53 3.19 3.48 335.67 332.19 334.85 334.82 2019 — — — — 0.73 0.73 329.26 328.53 329.26 329.26 2020 — — — — 0.12 0.12 319.03 318.91 319.03 319.03 2021 — — — — 0.01 0.01 304.23 304.22 304.23 304.23 2022 — — — — — — 6.28 6.28 6.28 6.28 2023 — — — — — — 5.29 5.29 5.29 5.29 2024 — — — — — — 4.02 4.02 4.02 4.02 2025 — — — — — — 2.51 2.51 2.51 2.51 2026 — — — — — — 1.34 1.34 1.34 1.34 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY 2027 — — — — — — 0.38 0.38 0.38 0.38 | 57 58 SHS Costs Household Equity HH Loan Household Benefits: Avoided SHS Theoretical Net Theoretical Net Loans from Year Including and Purchase of Grant Repayments at Cost with Grant Kerosene Purchases and Program Net Benefit with Grant Benefit with No POs to HHs Replacements Replacements 14% Interest and Loan Grid Electricity Saving HH Benefit but No Loan Grant or Loan 2028 — — — — — — 0.07 0.07 0.07 0.07 2029 — — — — — — 0.01 0.01 0.01 0.01 Total 1,897.50 743.00 87.72 1,066.77 1,200.72 1,943.72 3,292.07 1,348.35 1,482.30 1,394.57 FIRR with loan (14% flat rate interest and 3-year tenor (repaid 50% in Year 1, 100% in Years 2–3, 50% in Year 4) 17.2% 16.4% 14.7% 5. ECONOMIC AND FINANCIAL ANALYSIS Note: HH = Household. 5.2.2 Net Financial Impact of SHS Program on All Stakeholders universal grid electricity availability. With grid arrival, the kerosene-saving value of the SHS units drops, as SHS become grid-electricity-saving SHS units rather than kerosene- Starting from the net financial benefits to households detailed in Table 13, it is possible to saving SHS units. The transition in substitution values lowers the value of the 12-year life estimate the net financial impacts of the activities of the SHS Program on all stakeholders. of the SHS units installed from 2014 through 2018. This has been done in the following steps: The kerosene distributors and the POs net financial stakeholder accounts are the most (a) Allocate to the GOB the taxes paid by households on the SHS equipment (Appendix F). straightforward, although with considerable uncertainty in estimates. Not unexpectedly, (b) Estimate the kerosene subsidy impact on the government because of the reduced use the local kerosene distributors are net losers; their losses during 2003–2021 sum to of kerosene caused by the SHS Program (Appendix D). negative US$47.2 million (undiscounted) in constant 2018 US$. The POs, on the other (c) Estimate the net benefits to IDCOL from managing the program (Appendix E). hand, are imputed profits on sales/installations during 2003–2013, summing to (undiscounted) US$147.3 million in constant 2018 US$ (including profits from sale of (d) Allocate to the POs the profits made on the sale of the SHS to the households replacement parts); from 2013 onward, they are expected to break even. (Appendix D). (e) Allocate to the kerosene distributors the loss of the profit component of the kerosene The GOB and the IDCOL series are more complex. The GOB earns US$203.5 million price markup for the sales that were avoided by use of SHS (Table D.10). in taxes on SHS and loses US$4 million in kerosene subsidies, for a total net gain of US$199.5 million. The GOB SHS tax benefits were received up to 2018, with the result The net financial impact of the SHS Program activities on these stakeholders covers the that correcting for inflation to constant 2018 US$ and discounting as of 2018 increase 40-year period from 2003 to 2042, beginning with the receipt of the first funds for the SHS their value substantially (see Appendix D). Three negative and large yearly net benefit Program by IDCOL in 2003 and ending with the final repayment of concessional funds by flows—2015 through 2017—dominate the 2003–2029 series in the GOB kerosene subsidy IDCOL to the government in 2042 (see Table 14). The indicative net financial stakeholder account. These large deficit accounts occur because in some years the border price of impacts sum up to an overall positive total of US$1,702 million in constant 2018 US$ on kerosene was below the official price for kerosene, resulting in a loss in revenue to the an undiscounted basis. The SHS-using households have the largest net benefits, with government (Bangladesh Petroleum Corporation) when the SHS reduced kerosene use. US$1,348 million in constant 2018 US$ on an undiscounted basis from 2003 to 2029. The sudden drop in the household net benefit series in 2022 results from the expected Table 14: Stakeholder Indicative Net Financial Benefits from the SHS Program 2003–2042 (Constant 2018 US$, millions) Household Net GOB Net Benefits IDCOL Net PO Profits Kerosene Total Net Stakeholder Year Benefits with Taxes Collected Savings on Kerosene Total GOB Real Transaction Financial on SHS Distributor Benefits Grants and Loans on SHS Subsidies Account Benefits a Installations Foregone Profits (1) (2) (3) (4 = 2 + 3) (5) (6) (7) (8 = 1 + 4 + 5 + 6 + 7) 2003 (3.48) 0.84 0.03 0.88 (0.03) 0.95 (0.01) (1.69) 2004 (7.97) 1.68 0.37 2.06 (0.12) 1.88 (0.02) (4.17) 2005 (12.12) 2.32 0.34 2.66 0.57 2.60 (0.04) (6.33) 2006 (16.43) 3.13 1.74 4.87 1.72 3.51 (0.07) (6.41) 2007 (26.51) 5.62 1.83 7.45 4.62 6.29 (0.14) (8.29) 2008 (36.61) 9.38 2.44 11.82 9.28 10.50 (0.24) (5.25) 2009 (59.77) 12.30 (0.40) 11.90 15.16 13.78 (0.40) (19.32) 2010 (84.79) 21.80 3.29 25.09 23.58 24.42 (0.72) (12.42) 2011 (63.30) 26.85 35.81 62.65 27.30 30.07 (1.23) 55.49 2012 (56.96) 34.22 87.69 121.91 30.09 38.33 (1.91) 131.45 2013 (47.64) 40.67 44.55 85.22 34.35 7.59 (2.88) 76.64 2014 40.71 23.31 52.45 75.75 37.24 4.35 (3.78) 154.28 2015 19.49 16.00 (96.75) (80.74) 34.17 2.99 (4.58) (28.67) 2016 150.63 4.30 (130.37) (126.07) 35.50 — (4.97) 55.09 2017 249.35 0.79 (74.32) (73.54) 24.30 — (5.15) 194.96 2018 332.19 0.09 17.59 17.68 22.95 — (5.48) 367.34 2019 328.53 — 17.25 17.25 (2.72) — (5.38) 337.68 2020 318.91 — 16.72 16.72 (4.30) — (5.21) 326.12 2021 304.22 — 15.94 15.94 (4.30) — (4.97) 310.89 2022 6.28 — — — (3.24) — — 3.04 2023 5.29 — — — (3.80) — — 1.49 2024 4.02 — — — (29.76) — — (25.74) 2025 2.51 — — — (29.25) — — (26.74) 2026 1.34 — — — (28.79) — — (27.45) LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY 2027 0.38 — — — (28.37) — — (27.99) | 59 60 Household Net GOB Net Benefits IDCOL Net PO Profits Kerosene Total Net Stakeholder Year Benefits with Taxes Collected Savings on Kerosene Total GOB Real Transaction Financial on SHS Distributor Benefits Grants and Loans on SHS Subsidies Account Benefits a Installations Foregone Profits 2028 0.07 — — — (28.01) — — (27.94) 2029 0.01 — — — (20.51) — — (20.51) 2030 — — — — (16.59) — — (16.59) 2031 — — — — (16.06) — — (16.06) 5. ECONOMIC AND FINANCIAL ANALYSIS 2032 — — — — (10.86) — — (10.86) 2033 — — — — (5.61) — — (5.61) 2034 — — — — (2.95) — — (2.95) 2035 — — — — (1.80) — — (1.80) 2036 — — — — (1.77) — — (1.77) 2037 — — — — (1.54) — — (1.54) 2038 — — — — (1.51) — — (1.51) 2039 — — — — (1.49) — — (1.49) 2040 — — — — (1.47) — — (1.47) 2041 — — — — (1.45) — — (1.45) 2042 — — — — (0.71) — — (0.71) Undisc. Total 1,348.3 203.3 (3.8) 199.5 53.8 147.3 (47.2) 1,701.8 NPV at 10% 744.6 383.9 90.3 474.2 379.2 310.2 (56.1) 1,852.0 NPV at 5% 1,087.8 279.3 33.6 312.9 222.7 214.3 (51.0) 1,786.9 Note: a. Duration of net benefits varies with only IDCOL net benefits extending to 2042 due to debt service payments to the GOB. IDCOL’s net benefit stream (see Appendix E for details) shows significant benefits early in the program since its loan terms from the GOB are more favorable than the terms it provides to POs and then losses in later years when it needs to repay concessional loans to the GOB for an extended period beyond the repayments it receives from the POs. While the total of its undiscounted net financial benefit stream is US$54 million in constant 2018 US$, the NPV of its net benefits becomes more positive as the discount rate increases. It should also be noted that IDCOL’s opportunity cost of capital is around 2.5 percent in constant terms meaning that the appropriate NPV from IDCOL’s standpoint would be about US$139 million in constant 2018 US$, midway between the undiscounted value and the NPV at 5 percent. 5.3 SUMMARY OF DISTRIBUTION OF NET The kerosene subsidy savings impact is relatively small at FINANCIAL BENEFITS AMONG STAKEHOLDERS negative US$4 million undiscounted and US$90 million discounted at 10 percent to 2018. Taxes collected on SHS Table 15 summarizes estimated net financial benefits on an are significant at US$203 million on an undiscounted basis undiscounted constant 2018 US$ basis and presents NPVs at and US$384 million when discounted at 10 percent to 10 and 5 percent discount rates. A 10 percent economic return 2018. These taxes added about 12 percent to the cost of is the hurdle rate for projects used by both the GOB and the SHS to households; lowering them would have improved World Bank in Bangladesh, representing the opportunity affordability and increased demand for SHS and increased cost of capital to society (World Bank 2016). Since this is benefits to households. an analysis of financial rather than economic benefits, it is appropriate to look at the benefits from the viewpoint of the All three of the net financial benefit flows presented above individual stakeholders. The NPV at a 5 percent discount rate show that all stakeholders benefited from the SHS Program, in constant terms is useful for looking at the net benefits from except for kerosene dealers. Households consistently the perspective of stakeholders with lower opportunity cost of benefited the most, even though their benefits are capital, especially IDCOL and the POs. significantly underestimated by including only savings on kerosene/grid electricity for lighting. The NPV for household On an undiscounted basis, an estimated net benefit of benefits is zero at a discount rate of 14.7 percent. The US$1,702 million in constant 2018 US$ is ‘available’ to households took loans to purchase SHS with interest rates national stakeholders, while the NPV of this amount from 12 to 16 percent on a flat rate basis, sometimes at rates discounted at 10 percent would be US$1,852 million in higher than the 14.7 percent net financial return, indicating constant 2018 US$. While the overall net benefit looks the high value they placed on the SHS for reasons other than similar at different discount rates, its distribution among savings on lighting. stakeholders varies considerably (see also Figure 21). Households are the largest beneficiaries at any discount With respect to IDCOL, since its opportunity cost of capital rate (79 percent of undiscounted benefits and 40 percent of is around 2.5 percent in constant terms,33 the appropriate benefits discounted at 10 percent). The GOB is the second NPV from IDCOL’s standpoint could be about US$138 million largest beneficiary (12 percent share of undiscounted benefits in constant 2018 US$, midway between the undiscounted and 26 percent of benefits discounted at 10 percent), followed value and the NPV at 5 percent (see Appendix E). The POs also by IDCOL (3 percent undiscounted and 21 percent discounted likely have low opportunity costs of capital so that their NPV at 10 percent) and the POs (9 percent undiscounted and 17 would more appropriately be between US$147 and US$214 percent discounted at 10 percent), while the kerosene dealers million or 180 million in constant 2018 US$, remembering the are net losers with a steady loss of about 3 percent of net uncertainty in these estimates. benefits at any discount rate. Table 15: Summation of Distribution of Estimated Net Financial Benefits to Indicated Stakeholders 2003–2042 Net Financial Benefits 2003–2042a Constant 2018 US$, millions S. No. Stakeholders Net Present Value in 2018 Undiscounted 10% Discount 5% Discount Households’ net benefits from kerosene and electricity 1 1,348 745 1,088 savings, with grants and loan 3 GOB 200 474 313 4 Taxes collected on SHS sales 203 384 279 5 Net savings on kerosene subsidies (4) 90 34 6 IDCOL from on-lending to POs 54 379 223 7 POs’ profits on SHS sales 147 310 214 8 Kerosene distributors’ foregone profits (47) (56) (51) 9 Total (1 + 3 + 6 + 7+ 8) 1,702 1,852 1,787 Note: a. Duration of net benefits varies with only IDCOL net benefits extending to 2042 due to debt service payments to the GOB. 33 Return on equity of 8.5 percent in 2016–2018 according to IDCOL Annual Reports in current terms, reduced by 6 percent average inflation rate in the same period. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 61 5. ECONOMIC AND FINANCIAL ANALYSIS Figure 21: Cumulative Stakeholder Benefits from the SHS Program, 2003–2042 5.4 ANALYSIS OF ODA FLOWS FOR THE SHS development that are passed to and through IDCOL via the PROGRAM ON THE GOB GOB Treasury. In the SHS Program, IDCOL received loans from the GOB and passed them on to the POs. This section analyzes the structuring of the ODA funds pass- through to IDCOL for the SHS Program and the implications This section reflects the revision of the original financing for the GOB Treasury. The government received funds from terms between the government and IDCOL that took effect on international development partners for carrying out the SHS July 1, 2018 (see Section 4.7). Recognizing the fiscal and other Program on concessional terms. Such concessional loans benefits that the SHS Program provided, the government are designed to be beneficial to the receiving government. reduced the interest rate on IDCOL loans for the program from As with any concessional ODA loan that is passed on to a 3 percent per year to 0 percent per year with effect from July national recipient on less favorable terms, the concessional 1, 2018. IDCOL also concurrently reduced the interest rate loans for the SHS Program generated net financial benefits to from 4 percent per year to 0 percent per year on outstanding the government. SHS loans to POs with the same effectiveness date. Within the Bangladesh national economic development The ODA-GOB flows were positive from the GOB perspective system, IDCOL functions somewhat like a pre-1990s national- from 2003 until 2016 and then turn negative from 2017 level industrial development bank. Among IDCOL’s duties is through 2054 as the country repays the first of the to administer ODA loans and grants for targeted economic withdrawals following the grace periods provided by each Figure 22: ODA Loan Withdrawals and Repayments by GOB 2003–2054 62 Figure 23: Net GOB Loan Receipts by IDCOL Minus IDCOL Repayments 2003-2042 of the ODA organizations (see Figure 22, and Table F.2 in the GOB Treasury peaked in the 2017 repayment tranche of Appendix F). As the withdrawals build up and the grace period US$34.7 million and remain at the level of US$30–33 million expirations accumulate, the repayments will reach their peak from 2018 to 2028 (all in millions of current US$). annual amounts during 2033–2040 (annually exceeding US$24 million in current dollar values). After 2040, the annual Figure 24 and Table F.4 show the net impact on the GOB payments will recede until the final payment of US$1.37 Treasury of the flow of ODA funds by year, in constant 2018 million on the ODA loans in 2054.34 US$ values. During 2003–2030, the IDCOL repayments to the GOB Treasury exceed GOB Treasury repayments to ODA Figure 23 and Table F.3 show the flows of annual loan organizations in all years except for a few years (2003–2010) withdrawals of IDCOL (positive, IDCOL inflows) and annual when they cancel each other out for the most part. In repayments to the GOB Treasury (negative, IDCOL outflows) other words, during 2003–2030, the GOB Treasury suffers a during 2003–2042.35 The net flows of ODA financing to IDCOL negative net impact only for 5 years out of 28 on the ODA- (in constant terms) are positive during 2003–2016 and GOB-IDCOL pass-through of loans and credits to support negative during 2017–2042. IDCOL’s repayment obligations to the SHS Program. During 2031–2054, GOB repayments Figure 24: GOB Net on ODA Pass-Through to IDCOL 2003-2054 34Tables in Chapter 5 and in Appendix F are based on the information in Table 1 and the information in Appendix F. 35Repayments are calculated on a declining balance basis of combined principal and interest (at 3 percent per year in nominal terms). Calculations are made in current or nominal values in columns 3 through 4 of Table F.3 before conversion of the net flows to constant 2018 US$ in column 5. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 63 5. ECONOMIC AND FINANCIAL ANALYSIS to development partners exceed inflows from IDCOL. IDCOL net repayment liabilities are much reduced after 2033 and are scheduled to end in 2042, while the repayment liabilities of the GOB Treasury continue through 2054. Between 2010 and 2029, the GOB Treasury enjoys a steady margin on the difference between IDCOL repayments to the Treasury versus the Treasury payments to the ODA organizations—with that margin increasing from about US$1.6 million in 2010 to a high of US$24.2 million in 2018. The Treasury’s margin on repayment pass-throughs IDCOL-GOB- ODA range from US$10 million to US$21 million annually in constant 2018 values during 2014–2028. The GOB Treasury is also a leading gainer in the distribution of net financial benefits from the SHS Program activities (see Section 5.2), based on SHS taxes and savings in kerosene subsidies. The net financial impacts from taxes and subsidies and from ODA financing flows are combined in Figure 25 and detailed in Table F.5. The most negative years of the SHS program from the GOB Treasury’s standpoint were 2015 through 2017. During those years, the kerosene net subsidy turned against the GOB Treasury. During this period, the government’s fixed price was higher than the price at the border, meaning that a reduction in kerosene use reduced potential government revenues as it would have retained the difference as revenues. From 2003 through 2054, the Government Treasury’s forecast net gain on IDCOL payments minus ODA repayments is positive and totals US$1 million in constant 2018 US$ on an undiscounted basis and US$180 million in constant 2018 US$ when discounted to 2018 at 10 percent. On a cumulative present value basis discounted at 10 percent to 2018, the GOB Treasury’s total net gain from the SHS Program was US$655 million. This comprises US$384 million from taxes on SHS, US$90 million from savings due to avoided kerosene subsidy, and US$180 million due to impact of ODA pass- through. All are in constant 2018 US$. (see Table F.5 for details). Figure 25: GOB Treasury Net Flows from SHS Program 2003 to 2054 64 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 65 66 MAIN CONCLUSIONS AND LESSONS LEARNED 6.1 Conclusions 68 6.2 Lessons Learned 69 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 67 MAIN CONCLUSIONS AND LESSONS LEARNED The extensive analysis of the experience with the implementation of the Bangladesh SHS has led to several conclusions and lessons learned as described in this section. 6.1 CONCLUSIONS This review of the Bangladesh SHS Program over 2003–2018 leads to several main conclusions about carrying out large-scale off-grid electrification programs in the long term: • Households value SHS highly and are willing to pay for its services. The fact that 4.1 million systems were bought among a maximum of 15 million rural households without electricity at the start of the program indicates a market penetration rate well above 25 percent. This indicates both the acceptance of the SHS and the high value that households placed on the services obtained. • The SHS Program was economically justifiable from the national and global perspectives. The EIRRs demonstrate the net overall value of the program to the nation (20 percent without GHG emission reduction benefits) as well as to global society (25 percent with GHG emission reduction benefits). These benefits are underestimated since they are based only on the avoided cost of kerosene/grid electricity and do not include important other benefits such as improved quantity and quality of light; immediate financial savings from lower costs, for example, fewer batteries and free cellphone charging; and social and lifestyle benefits, for example, access to radio and TV and improved security. • Households benefited on a financial basis as well, with an FIRR of 17.2 percent considering loan defaults and a lower FIRR of 13 percent if there had been no such defaults. The most powerful evidence of household benefits is shown by the reality in the marketplace; the purchase of over 4 million SHS by rural households implies a strong willingness of households to pay for SHS when their ability to pay was enhanced by the availability of loans. • The GOB as the financier, IDCOL, and the POs also reaped financial benefits from the program despite late-stage problems. The government received substantial revenue from taxes on SHS sales and reduced subsidies for kerosene. Similarly, IDCOL benefited from the difference between its loans from GOB and the loans that it made to POs as well as from administration fees from ODA partners. These covered costs and permitted IDCOL to earn a return similar to that it makes on its other lending programs. The POs were able to earn profits in early years although these were much reduced in later years. • The SHS Program succeeded from 2003 to 2014 with the majority of SHS installed during this period (3.3 million of 4.1 million SHS). IDCOL’s strong leadership played an essential role in developing the implementation model with the following elements: implementation by POs with strong on-the-ground presence, a flexible, collaborative approach through the OC, an effective framework for controlling quality, and enforcement of financial discipline. • The SHS Program was hit by rapid grid expansion from 2015 onward that reduced suddenly potential markets. The financial viability of SHS at the household level increased with time as SHS technological progress reduced costs and improved performance. However, after 2015, the pace of grid expansion into areas served by SHS increased sharply, reducing the potential market of unelectrified households. An unintended consequence was that households could switch to grid electricity renege on SHS loan repayments with impunity. 68 • The impact on SHS sales decline under the program 6.2 LESSONS LEARNED with grid arrival was increased by the expansion of the TR/KABITA off-grid program that provided SHS How the SHS Program in Bangladesh was designed and to households at no cost. The arrival of the TR/KABITA implemented, how it adapted to changing conditions, Program could have been better planned and coordinated what were the benefits and costs of SHS electrification, its with IDCOL. IDCOL took over management of the TR/ successes and challenges, and how these challenges are KABITA Program and integrated it into its PO network, being overcome offer lessons to help other countries with providing additional business for the POs but this did not off-grid electrification to complement grid electrification fully ameliorate the damage to the SHS Program. efforts. Some of the main lessons learned that could be applied in other programs are summarized below. • The sudden drop in SHS sales and reduction in collection rates after 2015 created financial and Planning the program operating difficulties for IDCOL and the POs. The • Have a clear program goal. The SHS Program in shrinking sales and drop in collection rates caused the Bangladesh was a largely market-based program that POs to reduce operations, which in turn made it more aimed to provide benefits of electricity to rural households difficult to collect payments due on systems installed. in advance of the grid. If the intent of an SHS program Some of the POs were unable to fully repay their loans to is to deepen access to electricity within communities, IDCOL. then a market-based program alone may not be suitable • Coming to the financial aid of IDCOL and the POs to as sales will be to those with greatest ability to pay. ensure their sustainability is the proper follow-on role of Poorer households may not be served adequately unless the government, given late-stage problems. Recognizing additional incentives are offered to them. If an SHS the contribution made by the SHS Program to the GOB’s program is used to achieve a policy goal of universal rural electrification goals as well as the financial benefits access, then a more direct public sector intervention may reaped from the program, the GOB restructured its loans also be needed. to IDCOL and supported IDCOL in restructuring its loans to • Recognize that SHS users value the wider benefits POs in mid-2018. IDCOL has succeeded in renegotiating the from SHS. Customers have a strong willingness to buy outstanding PO debt to bring the substandard debt and bad SHS for the improvements observed in quality of life, debt down to US$28.6 million. It would also be appropriate access to information, safety, or other non-quantified for the GOB to further assist IDCOL and the POs, as required, benefits as described above, more than for savings in to bring the program to an orderly end and ensure the long- kerosene. Therefore, do not underestimate such attributes term sustainability of these organizations as well as the SHS in determining WTP. Nevertheless, ability to pay must be installed under the program. enhanced by making credit available so that payments for • Better planning and coordination of on-grid and SHS are affordable. off-grid electrification could have avoided the • Ensure high-level integrated planning of grid and off- late-stage difficulties in the SHS Program. The GOB grid electrification. Program planning must integrate was accelerating three major electrification efforts planning of grid and off-grid electrification at the highest simultaneously without foreseeing that the impact would level, based on economic principles. Be open with be to squeeze out the SHS Program. It was expanding information on electrification plans and consult and the grid, promoting SHS under the SHS Program, and cooperate with authorities responsible for grid extension providing systems at no cost to the poorest households and the SHS industry. The role of the government is and public institutions under the TR/KABITA Program. essential in ensuring that such high-level integrated SREDA, established partway into the SHS Program, is now planning takes place. an effective government authority to play the coordinating and policy-making role in future. • Consider distributional impacts in planning. There is a potential risk of focusing solely on the EIRR in project • In summary, the SHS Program made a significant appraisal as a justification, without closely examining contribution to the government’s efforts to meet the the distribution impacts to key stakeholders. The result stated Constitutional policy principle to transform of ignoring the distributional impacts is that program rural areas by providing, among other facilities, rural outcomes may not be realized if some key stakeholders electrification. It provided electricity in advance of the are disadvantaged. availability of the grid to 20 million people through 4.1 million SHS that that were purchased by rural households • Build long-term service infrastructure. SHS can have cost-effectively and with net benefits to all participants a long service life. However, inevitably, failures do occur except kerosene dealers, at an average cost of US$264 in and some components such as lamps, controllers, and constant 2018 US$ per SHS while also reducing kerosene batteries must be replaced at regular intervals. SHS consumption by over 4 million liters and reducing GHG program lifetimes are usually short (the 15+ years of the emissions by an estimated 9.6 million tCO2. IDCOL SHS Program is a rare exception). Therefore, it is LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 69 6. MAIN CONCLUSIONS AND LESSONS LEARNED essential to support building of a service and spare part Provide quality products and services supply infrastructure that continues after the program ends. • Ensure technical quality of SHS. Long-term sustainability demands well-designed products and • Retain flexibility in the program. Retain flexibility in the quality components and installations. Lower-capacity program implementation modalities, while adhering to but high-quality products should be offered to those sound economic, technical, and business principles. This customers with limited ability to pay. Costs should never is needed to adapt to changes in policy and economic be reduced by compromising quality or by decreasing conditions, technology evolutions, technology costs, support services. Where low-cost systems are offered, overall business environment, and even force majeure customers need to be fully aware of limitations. conditions. This requires good communications among the key stakeholders and ability of the main implementing • Adopt technology innovations. Adopt new technologies agency to effectively direct actions. that offer better quality and more reliable services. Examples include LED lighting, DC appliances, flat screen • Plan an exit strategy and monitor the market. A DC TVs, lithium batteries, and integrated SHS ‘plug-and- planned and orderly exit to a government-supported SHS play’ kits. Adopt technologies such as PAYG that can program needs to be foreseen from the beginning. The reduce cost of doing business and reduce financial risks. aim may be to continue the program until the SHS market Prevent barriers such as a singular focus on indigenizing is saturated or to transition to a commercial approach or high import duties from introducing imported new after the market is established. An exit strategy is needed technology. together with market monitoring so that risks related to all participants in the program are properly shared • Create consumer awareness. Do not oversell SHS and managed as the program ends. These risks include capabilities. User education is essential for an SHS business collapse, financial losses, and customers with no program’s success. Providing information and training recourse to repair/replacement services. on simple maintenance and safe operating procedures system is essential. Responsive and sustainable institutional infrastructure Overcome the cost barrier • Have a strong lead agency for program management. • Recognize that SHS cannot compete with grid A strong lead agency such as IDCOL is needed to provide electricity. It is nearly impossible for SHS sold on close and timely management and supervision; an a commercial basis to compete with grid electricity uncompromising attitude toward financial discipline is promising unlimited electricity at low tariffs. Even if essential. A successful SHS program needs well-qualified electricity tariffs are not subsidized, the cost of SHS sold managers and trained technicians. Adequate salaries commercially at the marginal cost cannot compete with and benefits are required. Technicians must be trained tariffs that are based on average costs to a large customer and with access to spares and tools to ensure responsive base rather than the marginal cost of new rural customers. repair and maintenance services. They need to be • Offer term credit and affordable payment schemes. adequately compensated or they will take their newly Due to the high first costs of SHS, offer multi-month- or acquired skills and move. multiyear-term credit that better approximates the • Build on existing organizations. Build on the strengths household expenditure patterns. Otherwise, only the of existing organizations rather than creating new ones, better-off households will be able to participate. where possible. But make the policy changes, identifiable • Take care with grants and subsidies. To ensure through the stakeholder analysis, needed for these sustainable programs, such assistance should be used to organizations to effectively transition to the changed build market infrastructure or limited equity to reduce the economy the program is designed to induce. capital costs. Operating costs should not be subsidized. • Provide responsive management. Take timely The analysis of stakeholder impacts will help identify advantage of technologies and business practices that groups whose transition to the SHS will be appropriately can reduce cost and improve financial and business assisted by grants and subsidies. management such as PAYG technology, mobile pay, and • Remove discriminatory taxes and duties. Level the computerized MISs for inventory, finance, and business playing field—governments should rationalize duties and management. taxes if these discriminate among electrification options. • Ensure financial sustainability. The business should While recognizing political issues in changing tax and generate revenues to cover costs and provide adequate duty structures, the analysis makes clear the drag that returns. Avoid destructive competition where profit government’s failure to make appropriate and timely margins are pared to the bone to gain market share. changes placed on the market creation objectives of the Ensure full cost recovery, select customers with ability to SHS Program. As noted previously, taxes on SHS systems pay or support rational incentives to enhance ability to added 12 percent to the price while kerosene subsidies pay, establish effective fee collection mechanisms, and are a disincentive to the poorer segment of the population simplify administration. to adopt SHS that offer far superior lighting services. 70 Government and donor support the government either did not act at all or acted inappropriately. This points to the need to coordinate off- • Ensure continuity and coordination. An important grid programs and grid expansion to avoid the late-stage factor in the success of the SHS Program was the problems in the SHS Program. continuity provided by the government, IDCOL, and the development partners agreeing to seamlessly integrate • Ensure development partner coordination and the additional resources into the SHS Program using technology transfer. Development partners should the same implementation modalities over the long coordinate their support with the government and other term. stakeholders and should deliver the messages outlined above in the process of discussing, designing, and • See government and the private sector as implementing such assistance. Development partners can complements, not alternatives. Market creation is help in technology transfer—not only physical technology not simply an alternative to government provision or but also organizational and institutional technology— government ‘interference’ in markets. Development beyond their role in financing investments in PV systems projects and programs introduce changes—whether as part of rural electrification and rural development major or minor—in the way the economy works. projects. The government holds a kingpin position in making those alterations. A key to the success of the SHS • Leverage scarce development partner funding. Ideally, Program was the government appointing IDCOL, an development partner financing should leverage domestic organization that straddles the public-private nexus, financing to maximize the funds available for the program. as the implementing agency, and letting it function IDCOL did ‘leverage’ financing from other development independently. partners and PO co-financing. Extending participation to the commercial market players may have leveraged more • The government needs to play strong coordination private funding, but it requires rethinking the business role. Electricity supply is often overseen by model and adapting to the changing business and the government. When markets fail because financial environment. of ‘coordination failure’, it usually is because LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 71 72 REFERENCES LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 73 REFERENCES Alam, Mahbub. 2013. Bangladesh Solar Home Systems: Transition to Commercial Financing. Draft report. Dhaka, Bangladesh: IDCOL. Asaduzzaman, M., M. Yunus, AKE Haque, AKM Abdul Malek Azad, S. Neelormi, and Md. A. Hossain. 2013. Power from the Sun: An Evaluation of Institutional Effectiveness and Impact of Solar Home Systems in Bangladesh. Dhaka, Bangladesh: Bangladesh Institute of Development Studies Report to the World Bank. https://www.sun-connect-news.org/fileadmin/DATEIEN/Dateien/ New/Bangladesh_Idcol_Assessment.pdf. Bangladesh Bureau of Statistics. 2015. Population and Housing Census 2011. Bangladesh Ministry of Planning, Dhaka: Bangladesh. http://bbs.dhaka.gov.bd/sites/default/files/files/bbs.dhaka. gov.bd/law_policy/6ed6b42c_2015_11e7_8f57_286ed488c766/Socio-Economic%20and%20 demographic%20Report%202012.pdf BIDS (Bangladesh Institute of Development Studies). 2012. Household Survey Data on Impact Evaluation of Solar Home Systems in Bangladesh. Report for World Bank. Washington, DC. Bangladesh Institute of Development Studies and International Institute for Sustainable Development. 2012. A Citizens’ Guide to Energy Subsidies in Bangladesh. Winnipeg, Manitoba, Canada: IISD Global Subsidies Initiative. https://www.iisd.org/publications/citizens-guide-energy- subsidies-bangladesh BIDS Team. 2018. “Assessing the Current Situation of Solar Home System (SHS) Program of IDCOL and Recommending Action Plan”. Principal Authors: Monzur Hossain, M. Asaduzzaman, Mohammad Yunus, Jamaluddin Ahmed, Kazi Ershadul Alam, Zabid Iqbal and Paritosh Kumar Roy. Report to Infrastructure Development Company Ltd, Dhaka, Bangladesh. Beaubien, J. 2019. “They Pump $15 Billion A Year into Bangladesh’s Economy — But at What Cost?” NPR. https://www.npr.org/sections/goatsandsoda/2019/06/03/722085193/they-pump-15-billion-a- year-into-bangladeshs-economy-but-at-what-cost. June 3, 2019. Bond T.C., C. Zarzycki, M. G. Flanner, D. M. Kock. 2011. “Quantifying Immediate Radiative Forcing by Black Carbon and Organic Matter with the Specific Forcing Pulse.” Atmospheric Chemistry and Physics 11:1505–1525. https://acp.copernicus.org/articles/11/1505/2011/ Cabraal, A. 1996. Bangladesh: Application of Renewable Energy. Back-to-Office Report to Per Ljung, Chief SA1EF. Washington DC, United States: World Bank. CEIC. 2020. Bangladesh HIES: Average Household Size. New York, NY, United States. https://www. ceicdata.com/en/bangladesh/household-income-and-expenditure-survey-number-of- household-by-size/hies-average-household-size. Dhaka Tribune. 2019. “Government to import 1.42m tons of Petroleum by June.” https://www. dhakatribune.com/business/2019/01/27/government-to-import-1-42m-tons-of-petroleum- by-june. January 27, 2019. 74 Economist Intelligence Unit. 2018. Can Bangladesh Overcome Its Energy Deficit? https://country.eiu.com/article. aspx?articleid=607103844&Country=Bangladesh&topic=Economy. Energia. 2019. Gender and Fossil Fuel Subsidy Reform: Findings from and Recommendations for Bangladesh, India, and Nigeria. https://www.energia.org/cm2/wp-content/uploads/2019/05/RA4_Gender-and-fossil-fuel-subsidy-reform_ without-Annex-2.pdf. Energypedia. 2019. The Reduction of Kerosene Lamp Emissions through Solar Lighting. 22 January 2019. https://energypedia. info/wiki/The_Reduction_of_Kerosene_Lamp_Emissions_through_Solar_Lighting. Eusuf, M. n.d. Dissemination of Solar PV Systems in Bangladesh: A Case Study: Narsingdi Solar PV Systems. UNDP Sustainable Development Networking Programme. http://www.sdnbd.org/pv_in_narsingdi.htm. General Economics Division of Planning Commission. 2012. Perspective Plan of Bangladesh 2010–2021: Making Vision 2021 a Reality. Dhaka, Bangladesh: GOB. https://policy.asiapacificenergy.org/sites/default/files/Perspective%20Plan_Final. pdf. GOB (Government of Bangladesh). 1972. Constitution of the Peoples’ Republic of Bangladesh: Part II Fundamental Principles of State Policy. Dhaka, Bangladesh: Government of Bangladesh. http://www.commonlii.org/bd/legis/const/2004/ part2.html GOB. 2012. Sustainable and Renewable Energy Development Authority Act No. 48 of 2012. https://www.dpp.gov.bd/upload_ file/gazettes/10720_39500.pdf Grameen Shakti. 2012. Renewable Energy Solutions in Bangladesh. https://static1.squarespace.com/ static/51bef39fe4b010d205f84a92/t/576936cfff7c5085d0a35c14/1466513104098/4_Grameen+Shakti.pdf. GVEP International and Insensus. 2016. PAYG Technology Introduction for IDCOL Market in Bangladesh. Report to IDCOL. London, United Kingdom: Global Village Energy Partnership (now Energy for Impact). Haque, M. A. 2020. Bangladesh Power Sector: An appraisal from a multi-dimensional perspective (part- 1). Dhaka, Bangladesh: EBL Securities. https://www.arx.cfa/-/media/regional/arx/post- pdf/2020/09/03/bangladesh-power-sector-an-appraisal-from-a-multi-dimensional-perspective. ashx?la=en&hash=0FA01602944BF9832225FFD52BAEE5B241C47861 IDCOL. 2006-2018. “Annual Reports”. Reports issued to shareholders, Dhaka, Bangladesh. http://www.idcol.org/home/an_ report. IDCOL. 2018. IDCOL Solar Home Systems Program: Present Status and Way Forward. Report to IDCOL Board, Dhaka, Bangladesh. IDCOL. October 2020. “SHS Program Updated Suppliers List”. Version as of October 1, 2020. https://idcol.org/home/ downloads/solar. IEG. 2014. ICR Review: Sri Lanka Renewable Energy for Rural Economic Development. Washington, DC: World Bank. http:// documents.worldbank.org/curated/en/830571474939221498/pdf/000180307-20141120032100.pdf. IEG (Independent Evaluation Group). 2008. The Welfare Impact of Rural Electrification: A Reassessment of the Costs and Benefits. Washington, DC: World Bank. https://openknowledge.worldbank.org/bitstream/ handle/10986/6519/454000PUB0978011PUBLIC10Mar06102008.pdf?sequence=1&isAllowed=y. Energypedia. 2015. Jawaharlal Nehru National Solar Mission. July 10, 2015. https://energypedia.info/wiki/Jawaharlal_Nehru_ National_Solar_Mission. IEA (International Energy Agency). 2014. IEA Statistics. https://www.iea.org/countries/Bangladesh/. Islam, M. 2002. “Utilization of Renewable Energies in Bangladesh”. In Shakti: Energy Website of Bangladesh. https://citeseerx. ist.psu.edu/viewdoc/download?doi=10.1.1.197.5569&rep=rep1&type=pdf. Islam, S. M. Monirul. 2019. “IDCOL Solar Home System Program.” Presentation at Pattaya, Thailand, revised by IDCOL and authors. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 75 REFERENCES Kaldor, Nicholas, “Capital Accumulation and Economic Growth,” in F.A. Lutz and D.C. Hague, eds., The Theory of Capital, St. Martin’s Press, 1961, pp. 177–222. Keystone Business Support Company Limited. 2018. Bangladesh Rural Electrification and Renewable Energy Development- MFD Analysis Report and Annex. Report to the World Bank. Dhaka, Bangladesh. Khander, S. R., H.A. Samad, Z.K.M. Sadeque, M. Asaduzzaman, M. Yunus and A. K. E. Haque. 2014. “Surge in Solar-Powered Homes: Experience in Off-grid Bangladesh,” In Directions in Development, Energy and Mining, Chapter 4. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/871301468201262369/pdf/Surge-in- solar-powered-homes-experience-in-off-grid-rural-Bangladesh.pdf. Lighting Global, GOGLA, and ESMAP, in collaboration with Vivid Economics and Open Capital Advisors. 2020. Off-grid Solar Market Trends Report 2020. Produced by Lighting Global, GOGLA, and ESMAP in collaboration with Vivid Economics and Open Capital Advisors. Washington, DC. https://www.lightingglobal.org/wp-content/uploads/2020/03/ VIVID%20OCA_2020_Off_Grid_Solar_Market_Trends_Report_Full_High.pdf. Marro, Peter, and Natalie Bertsch. 2015. “Making Renewable Energy a Success in Bangladesh: Getting the Business Model Right.” ADB South Asia Working Paper Series No. 41, Manila, The Philippines. https://sun-connect-news.org/ fileadmin/DATEIEN/Dateien/New/ADB-making-renewable-energy-success.pdf. Meier, Peter. 2003. Economic Analysis of Solar Home Systems: A Case Study for the Philippines. In Chapter 3 of Report to World Bank Asia Alternative Energy Unit, Washington DC. http://documents1.worldbank.org/curated/ en/820381468234277260/pdf/345410PH0Economic1Solar1Home.pdf Mills, Evan. 2003. Technical and Economic Performance Analysis of Kerosene Lamps and Alternative Approaches to Illumination in Developing Countries. Berkeley, CA, United States: Lawrence Berkeley National Laboratory, University of California. http://large.stanford.edu/courses/2011/ph240/machala1/docs/offgrid-lighting.pdf. Moazzem, K. G. and Mohammad Ali. 2019. “The Power and Energy Sector of Bangladesh: Challenges of Moving beyond the Transition Stage.” Paper presented at “Power and Energy Sector: Immediate Issues and Challenges,” Dhaka, Bangladesh: Centre for Policy Dialogue (CPD). https://cpd.org.bd/wp-content/uploads/2019/03/The-Power-and- Energy-Sector-of-Bangladesh.pdf. Musgrave, Richard A. 1959. The Theory of Public Finance: A Study in Public Economy. New York, New York: McGraw-Hill Book Company. National Renewable Energy Laboratory. 2018. U.S. Solar Photovoltaic System Cost: Benchmark: Q1 2018. Technical Report NREL/TP-6A20-72399, Golden, CO, United States. https://www.nrel.gov/docs/fy19osti/72399.pdf. Power Cell. n.d. Vision Statement. Ministry of Power, Energy and Mineral Resources, Government of Bangladesh, Dhaka, Bangladesh. https://powerdivision.portal.gov.bd/sites/default/files/files/powerdivision.portal.gov.bd/page/ f6d0e100_e2d8_47e7_b7cd_e292ea6395d3/4.%20VSPSPSectorReform.pdf. Quak, E. 2018. The Costs and Benefits of Lighting and Electricity Services for Off-grid Populations in Sub-Sahara Africa. K4D Helpdesk Report. Brighton, United Kingdom: Institute of Development Studies. https://assets.publishing.service. gov.uk/media/5af96657ed915d0df4e8cdea/Costs_Benefits_Off-Grid_Electricity_Lighting_Systems.pdf. Razzak, N., Md. Mamun-ur-Rashid, and Niloy R Biswas. 2012. Gender Responsive Social Assessment for Bangladesh Rural Electrification and Renewable Energy Development II Project. Dhaka, Bangladesh: BRAC University. Renewable Energy World. 2017. “As Pay-as-you-go Solar Proliferates, Here Are 7 Lessons for Business Success.” December 12, 2017. https://www.renewableenergyworld.com/ugc/articles/2017/12/01/bolstering-payg-agent-networks-7- lessons-learned-from-other-industries.html. Risad, M. 2017. Power Tariff to Hike Again from December 1st. Dhaka, Bangladesh: Energy Bangla. https://energybangla.com/ power-prices-hiked-again/. Regan, Ryan. 2018. “Innovations Spur Era of Rapidly Declining Solar Costs.” Solar Tribune, April 11, 2018. https://solartribune. com/residential-solar-cost-trends/. 76 Samad, Hussain A., Shahidur R. Khandker, M. Asaduzzaman, and Mohammad Yunus. 2013. “The Benefits of Solar Home Systems: An Analysis from Bangladesh.” Policy Research Working Paper 6724, World Bank, Washington, DC. http:// documents.worldbank.org/curated/en/656991467998808205/pdf/WPS6724.pdf. Susrut. 2019. “What Is the Cost of Shipping Oil by Tanker.” Cost Finder. January 26, 2019. http://cost-finder.com/what-is-the- cost-of-shipping-oil-by-tanker/. TSC (Technical Standards Committee). 2017. Technical Specifications for Solar Home System. Dhaka, Bangladesh: IDCOL. http://idcol.org/download/08da7f64cdb13269790a514f5a0cdd34.pdf. The Economist. 2012. “The Rise of Solar Energy.” December 20, 2012. https://www.economist.com/graphic-detail/2012/12/28/ pricing-sunshine. U.S. Environmental Protection Agency. 2014. “Emission Factors for Greenhouse Gas Inventories.” April 4, 2014. Washington, DC: EPA. https://www.epa.gov/sites/production/files/2015-07/documents/emission-factors_2014.pdf. UNFCC. 2013. “Grameen Shakti SHS Project CDM application.” 2765 CPA CER Sheet. https://cdm.unfccc.int/UserManagement/ FileStorage/MG04Y2DKQ7CXSZ9RTW6LENOJFAV3H8. Vechasart, R. and S. Suttisom. 2014. Rural Electrification in Thailand: Policy and Implementation. https://www.iitk.ac.in/ime/ anoops/for14/PPTs/Day%20-%205%20Bangkok/RE%20in%20Thailand%20-Policy%20and%20Implementation%20 -%20Mr.%20Reungvith%20Vechasart.pdf. Ward, William A. 2019. “Cost-Benefit Analysis Theory versus Practice at the World Bank, 1960–2015.” Journal of Benefit Cost Analysis 10 (1): 124–144. https://www.cambridge.org/core/journals/journal-of-benefit- cost-analysis/article/costbenefit-analysis-theory-versus-practice-at-the-world-bank-1960-to-2015/ C0F1838DEFCD0D44C6FCFDBB1BFD3072. Ward, William A., and Mehdi Al Bazzaz. 1986 Rev. Ed. “Containers Corporation of the Caribbean: Project Economic and Policy Analysis Case Study.” IE-5234P and IE-5234-S, EDI Case Studies and Exercise Series. Washington, DC: Economic Development Institute, World Bank. Wilson, Margaret, John Besant Jones, and Pierre Audinet. 2010. A New Slant on Slopes Measuring the Benefits of Increased Electricity Access in Developing Countries. ESMAP Report No. 53963-GLB. Washington, DC: World Bank. http:// documents.worldbank.org/curated/en/603961468152701050/pdf/539630ESW0Gray10Disclosed0414120111.pdf. World Bank. 2002. Bangladesh Rural Electrification and Renewable Energy Development Project. Project Appraisal Document. Washington DC: World Bank. http://documents.worldbank.org/curated/en/226581468741678811/pdf/multi0page. pdf. World Bank. 2011. Peru Second Electrification Project. Project Appraisal Document. Washington, DC: World Bank. http:// documents.worldbank.org/curated/en/550011468325218624/pdf/601540PAD0P1171OFFICIAL0USE0ONLY191.pdf. World Bank. 2015. Myanmar Electrification Project. Project Appraisal Document. Washington, DC: World Bank. http:// documents.worldbank.org/curated/en/149061468191334165/pdf/PAD1410-CORRIGENDUM-IDA-R2015-0237-2- Box393200B-OUO-9.pdf. World Bank. 2016. Technical Note on Discounting Costs and Benefits in Economic Analysis of World Bank Projects. OPSPQ. Washington, DC: World Bank. World Bank. 2017. Guidance Note on the Shadow Price of Carbon. Washington, DC: World Bank. http://documents.worldbank. org/curated/en/621721519940107694/pdf/2017-Shadow-Price-of-Carbon-Guidance-Note.pdf. World Bank. 2018. Peru Second Electrification Project Implementation Completion Report. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/427001520273463728/pdf/ICR00004397-03012018.pdf. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 77 78 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 79 80 APPENDIX APPENDIX A: Partner Organizations and Year of Appointment 80 APPENDIX B: PO Selection Criteria 81 APPENDIX C: Risks and Risk Mitigation in the SHS Program 82 APPENDIX D: Methodological Issues and Assumptions in Economic and Financial Analysis 94 APPENDIX E: Estimate of Indicative Net Financial Benefits to IDCOL from SHS Program Management 105 APPENDIX F: Impact on GOB Treasury from SHS Program Loan Transactions 108 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 81 Appendix A: PARTNER ORGANIZATIONS AND YEAR OF APPOINTMENT Partner Organizations and Year of Appointment 2002 (5 POs) 1. Thengamara Mohila Sabuj Sangha 4. Grameen Shakti 2. SRIZONY Bangladesh 5 COAST Trust 3 BRAC Foundation 2003 (4 POs) 6. Center for Mass Education in Science 8. Shubashati 7. Integrated Development Foundation 9. Upakulio Biddutayan O Mohila Unnayan Samity 2005 (5 POs) 10. BRIDGE 13. Padakhep Manabik Unnayan Kendra (PMUK) 11. Rural Services Foundation 14. Hilful Fuzul Samaj Kallan Sangstha (HFSKS) 12. Palli Daridro Bimochon Foundation 2009 (7 POs) 15. NUSRA 19. Rural Energy and Development Initiative 16. AVA Development Society (AVA) 20. RIMSO Foundation 17. Al-Falah Aam Unnayan Sangstha 21. InGen 18. Desha 2010 (9 POs) 22. SolarEn Foundation 27. Patakuri Society 23. ADAMS 28. Bright Green Energy Foundation (BGEF) 24. Green Housing & Energy Limited 29. Bengal Renewable Energy Limited 25. Resource Development Foundation 30. Risda Bangladesh 26. Shakti Foundation 2013 (17 POs) 31. Jagaroni Chakra Foundation (JCF) 40. Saif Powertec 32. Rural Development Sangstha (RDS) 41. Samaj Unnayon Palli Sangstha 33. Page Development Center 42. Sun Home Energy Limited 34. Sancred Welfare Foundation 43. Hamko 35. United Development Initiatives for Program Actions 44. MAKS Renewable Energy Company Limited 36. Pally Bikash Kendra (PBK) 45. Panna Rural Development Foundation (PRDF) 37. Friends in Village Development Bangladesh 46. Atmabiswas Souro Shakti Limited (ASSL) 38. SunRim Energy Ltd. 47. Polli Shakti Foundation Ltd. (PSFL) 39. Clean Energy Foundation (CEF) 2015 (11 POs) 48. Center for Development Innovation and Practices 54. National Development Programme (NDP) 49. Wave Foundation 55. Friendship 50. Uttara Development Program Society 56. Venus International Co. Ltd. 51. Gram Unnayan Karma (GUK) 57. Southern Renewable Energy Ltd. 52. ECO Social Development Organization 58. Voluntary Organization for Social Development 53. SKS Foundation 82 Appendix B: PO SELECTION CRITERIA Selection Criteria for POs (iii) Minimum total cash collection ratio of principal and interest on the current loan portfolio calculated on (a) The PO should have a satisfactory business plan a rolling 12-month basis of 95 percent approved by its Board of Directors. (iv) In case of an existing SHS loan portfolio, minimum (b) Operational and financial results should be available for total cash collection ratio of principal and interest at least the previous two years based on an acceptable calculated on a rolling 12-month basis of 95 percent audited report. The PO’s operations should be profitable (v) Minimum after-tax profit equivalent to 4 percent per for at least the past two years. However, in reaching year on revolving loan fund an assessment about potential profitability, IDCOL will also consider (i) forward-looking business prospects (vi) Where prospective business profitability is positive, and potential for profitable operations and (ii) if the PO the PO should be at least breaking even after is operating a solar program, the performance of their meeting operational expenses and debt service. solar business. However, in such cases, continued eligibility will be conditional on being able to meet the 4 percent per (c) The PO should furnish proof that its financial year after-tax profit criterion the following year performance is in conformity with the applicable financial criteria. (vii) Minimum debt service cover ratio of 1.25. (d) The PO must continue to meet the eligibility criteria, or Selection Criteria for Other Private Entities (PEs) its participation can be suspended or ended. (e) The PO will establish and maintain sound and (a) A lawful PE organized under the laws of Bangladesh, transparent accounting, MIS, and internal audit system. complying with pertinent laws and regulations regarding capital adequacy, classification of assets, (f) Accounts are audited by a reputable external auditor on nonaccrual of interest and provisioning, exposure limits, an annual basis. and so on In addition, participating MFIs must (b) A verification that PE meets satisfactory financial criteria, ratio requirements, and exposure limits (a) Must be registered with the appropriate registration (c) Capable of managing rural renewable energy program, authority to conduct microfinance services. as evidenced by the satisfactory business plan and (b) Currently be conducting microfinance services with soft operating results. loan funds from PKSF as a PO, Bank of Small Industries and Commerce Limited, or any other similar national or Criteria for Conversion of a Supplier PO/PE into a international funding source. Supplier and Lender PO (c) Have microfinance operations in project areas identified (a) A supplier PO may be converted into a supplier and in the priority list for the SHS Program. lender PO if these criteria are met: (d) Have at least 10,000 beneficiaries. (i) The supplier PO shall have signed a PA with IDCOL. (e) Be capable of managing rural renewable energy (ii) The supplier PO has installed minimum 1,000 SHS program. under IDCOL’s SHS Program. (f) Meet specific financial criteria. (iii) The loan recovery rate for those SHS shall not be (i) Minimum BDT 10,000,000 of equity less than 95 percent. (ii) Debt-to-equity ratio of the MFI less than 3.0 (iv) Satisfactory report from the auditors engaged by IDCOL. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 83 84 Appendix C: RISKS AND RISK MITIGATION IN THE SHS PROGRAM The matrix shows the risks classified by • Technical, • Financial, • Market, • Management, • Regulatory and policy, • Political and other, and • Environmental and social. For each identified risk, the matrix describes the mitigation measures adopted, the timeline for adoption, how effective they were, and options to be considered for the future. The future options are particularly useful for consideration during the design and implementation of future projects, wherever they may be. Risks Mitigation Timeline Effectiveness Future Options Technical risk: IDCOL has an independent TSC that prepares technical specifica¬tions First meeting • Surveys show a high level of consumer • Quality labeling (for example, Ensuring quality for systems and components and end-use equipment and approves held on satisfaction and effectiveness of warranty Lighting Global accreditation) products of manufacturers based on test reports from TSC-accepted test January 7, enforcement. • Adopting IEC standard at BDS centers. 2003 • Quality verification (initially no adequate standard (2017) Initially TSC comprises a professor, United International University testing facility centers; only documentation • Consumer awareness (former professor of BUET); chairperson and members from BREB; checking was done). • Use of barcode, QR code for professor, Electrical and Electronics Department, BUET; and professor, • Gradual development of the quality verification keeping track of inventory, Institute of Energy, University of Dhaka. system (PV testing lab at BUET in 2014) product tracking • Accreditation of testing Quality standards based on internationally accepted (International centers to ISO 17025 Electrotechnical Commission [IEC], ISO, OSHAS, and so on) standards cover battery, PV modules, fluorescent lights, charge controllers, and later LED lamps, DC fans and TV, and solar lanterns. Two more members added to the TSC: a representative from both September the Power Cell (technical arm of the Power Division) and the Local 16, 2013 Government Engineering Department (LGED) as per IDCOL Board decision to strengthen capacity of the committee. Risks Mitigation Timeline Effectiveness Future Options TSC’s terms of reference 2003 • Sets technical specification of solar equipment • Reviews the standards from time to time to ensure quality, consistency, and continuous improvement of the program • Enlists products to be used based on set specifications • Monitors quality control mechanism from time to time IDCOL issues approved products list. POs can use only the equipment 2004, updated approved by the TSC. monthly TSC banned the use of DC fan in the SHS Program since DC fan affects 2006 the longevity of battery. Six testing facilities at local universities approved by the TSC and 2007 accepted by IDCOL. SHS user manual and technical manual 2007 Standards updated to include CFLs and improved wiring; LED in SHS (<30 Wp) 2008–2009 Standards updated to include LED lights, lithium ion batteries, charge 2011–2012 controller set point adjusted Standards updated to include LED TV and DC fans 2013 Cable standardization for SHS and requirement of IEC certification of panel 2014 Technical specification of Li-ion battery 2015 Initiation of PAYG technology 2016 Technical specifications for DC and AC streetlights with solar PV 2016 SREDA developed regulatory policy to safeguard quality standards and 2016 consumer protection BSTI adopts IEC standards 2017 Testing facilities were upgraded to include module testing, LED light testing. 2017 Technical risk: Establishing warranty at component level 2003 Warranties enforced and honored by POs and Requiring warranty suppliers. Confirmed by audits and inspections. as a measure of Small SHS (10 wp to <30 wp) introduced, and warranty period for the 2007 quality and long- components of small SHS adopted: PV module - 12 years, battery - 3 term sustainability years, charge controller - 3 years, LED - 5 years LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 85 86 Risks Mitigation Timeline Effectiveness Future Options Technical risk: IDCOL web-based database is accessible online and allows IDCOL 2010 The database has given much more flexibility to Bar codes on SHS and use Inspection and regional offices and POs to access the database remotely. After IDCOL in managing the program and allows for of mobile app can improve verification to installation of the systems, the POs input data of the systems. The timely corrective actions to be taken. accuracy and ease data entry. ensure quality of software conducts initial screening of the POs’ entry such as duplicate installation systems, unapproved components, and so on. Regional monitoring offices: 17 regional monitoring offices with 142 Offices setup If problems found, POs are informed. PO solves At the beginning, monitoring technical inspectors. Staffing: Regional manager with 8–9 technical 2010 - 2 problem updates database for follow-up. Typical of SHS was conducted from inspectors to physically verify the systems throughout the year to ensure 2011 - 6 issues uncovered and then resolved by POs: long the head office. Later regional proper installation and after-sales services. 2014 - 12 cables, module shading, battery cell damaged, offices were opened at different • Inspectors check if systems were professionally installed and TSC- 2018 - 17 battery leakage, faulty charge controller, times based on concentration approved equipment are used, identify technical problems, and verify controller bypass, or no controller. of SHS to maximize inspection after-sales service. numbers as well as to reduce • Reinspection of the systems: random reinspection conducted to monitoring cost. substantiate findings of initial inspection. Process helped monitor quality and thoroughness of initial inspections. • Established for consumer feedback and complaints, the Call center 2007 The complaints received by the call center are • Dedicated short code has operates from 9 a.m. to 9 p.m. except Fridays and other government addressed in two ways. The customer is given been taken recently for the holidays to ensure customer service regarding any technical problem the contact details of the PO’s regional office to call center (16653). and after-sales service. connect the two parties directly. The concerned • TR/KABITA customers receive • Results discussed at OC meetings. PO is also informed by the call center of the a voice call or a text message complaint and is advised to take necessary every alternate month. measure. The record of the complaint is kept in • An integrated software is a database. A follow-up call is done by the call being developed through center to the customer to ensure the problem is which number of complaints resolved by the concerned PO. The results have resolved and complaint been found to be effective to some extent. resolution lead time can be tracked online. • Independent technical audit by third-party firm to ensure Audits Based on the findings of the technical audits, accountability of suppliers and POs for quality and after-sales services conducted in subsequent actions were taken. There are • Verifies quality of main system components and verifies POs 2003, 2009, instances of strict measures taken by IDCOL installation, warranty, and after-sales maintenance support 2012, 2014, which include penalization of the suppliers and • Corrective measures taken included penalizing suppliers/ and 2015 suspension of the approved models in cases of manufacturers and revisions to technical standards the unsatisfactory reports of the technical audits regarding the battery, charge controller, solar panels, and LED lamps. It see Same approach adopted for quality verification of TR/KABITA 2019 installations. BUET contracted to evaluate the quality, effectiveness, and service of the installed systems under TR/KABITA. Risks Mitigation Timeline Effectiveness Future Options Random sample test for quality assurance. Randomly collects samples Since 2009 12 models of solar battery have been delisted IDCOL has a testing facility from the suppliers’ warehouses and installation sites and test them and repeated due to unsatisfactory testing results. It was in Dhaka to conduct periodic from the testing facilities of BUET every year found effective. testing of equipment under TSC reviews test reports and, in case of any deviations from technical the program such as battery, standards, recommends temporary suspension of the specific model. charge controller, and so on. If the subsequent test result is found unsatisfactory, TSC recommends It also has testing equipment permanent suspension of the specific model of equipment. to test the performance of battery, charge controller, and so on at the field level. IDCOL financed to establish a PV testing lab at BUET for test and 2014 Took longer than expected to set up facility and Establish test centers earlier certification of modules, controllers, batteries, lights, and so on. train staff and/or enlist overseas test centers. Consider strengthening test capacity at institutions with core testing and certification responsibilities. Technical risk: • Ensuring warranty maintenance services 2003 onward Highly effective in early years. As SHS market Support trained staff of POs to Ensuring timely • Service centers in the locality declined and POs closed non-viable service set up private service and spare provisioning of • Institutional development grant (periodical change) centers, access to service and spares declined parts supply shops. service and spares and customer dissatisfaction increased. Technical System size (30 Wp to 130 Wp) with fluorescent tube lights 2003 Average size demanded in early years was about Because of rapid development risk: Offering 50 Wp which then declined to 30–40 Wp due to of rural economy, availability consumers’ choice Small SHS (10 wp to <30 Wp) with fluorescent tube lights 2007 introduction of efficient LED lamps and grant of efficient DC appliances such of usage and reduction. By about 2015, due to reduction as TV and fans, greater demand Small SHS ranging 10–15 Wp would be completely LED lamp based and 2008 service levels in minimum required battery size (reduced for higher capacity systems systems ranging from 16 Wp to 21 Wp would be CFL based where single autonomy from 3 to 2 days), cost declined and LED lantern can be used as options (warranty for LED being 5 years and demand for larger size increased. for CFL 1 year). Introduction of pico-PV system (below 10 Wp) with LED lamps 2015 As the cost of SHS dropped, increasingly larger systems were offered 2015 with additional connections for appliances (TV, fan). Technical risk: Consumer survey and gender response surveys conducted to 2009, 2012 Provided useful information to assess program Smart phone or SMS-based Consumer determine consumer satisfaction, appliance usage, kerosene displaced, effectiveness, justify additional finance short, more frequent surveys feedback improvements in health, education, security, and other outcomes mobilization, and market feedback to POs LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 87 88 Risks Mitigation Timeline Effectiveness Future Options Financial risk to At the beginning of the program, the PA did not require any security 2003 Till about 2014. This agreement, along with IDCOL has and is implementing IDCOL: Securitizing from the POs except for maintaining a balance in the DSRA equivalent regular supervision and OC meetings, was able several debt recoveries that are IDCOL financial to 1 semiannual installment payment. It adopted typical securitization to achieve collection efficiencies of 96% or ongoing and described later in exposure practice for MFIs in Bangladesh. better. But it started dropping to below 90% by this matrix. February 2015 and below 52% by 2016. After 6 years of operation, IDCOL executed Amendment and Restatement 2009 IDCOL is seeking government Agreement to the PA with the POs which incorporated the security assistance in overcoming the package for the first time. The security package was as follows: first risk. charge hypothecation on all floating assets of POs; lien on all project accounts (Proceeds Account and DSRA); and demand promissory note and letter of continuity. Revision of the terms and conditions for refinancing facility under the December 6, These measures were taken to secure the loan SHS Program aiming to secure its exposure of loan as well as to achieve 2011 exposure by taking additional securities from the commercialization POs. After that IDCOL was able to take additional securities from some POs. Amendment Agreement with the POs for additional securities, as July 2012 Mainly ineffective as collection efficiencies applicable: continue to drop to the 84%–88% range and then • First charge hypothecation on all fixed and floating assets of POs 52% in 2015 and down to 9% in 2018 and along • Personal guarantee from the directors/shareholders with it, the quality of the POs loans with IDCOL. • Corporate guarantee from the concerned third parties • Letter of comfort from the governing board of the POs No PO provided legal mortgage of land or bank • Lien on all project accounts (that is, Proceeds Account and DSRA) guarantee. • Maintenance of minimum required balance in DSRA equivalent to 4 (four) quarterly installment payments None could maintain required DSRA balance. • Mortgage of land or bank guarantee to secure 20% of the outstanding Only maintained DSRA balance equal to 2 refinance amount quarter installments • Demand promissory note and letter of continuity Some POs registered as NGO/society/foundation Despite partially complete security documentation status, IDCOL had unable to provide personal guarantees by the to continue disbursement to the POs to ensure smooth operation of members of their executive committee as they their SHS programs. Otherwise, POs would not be able to continue did not own the organization. installation of SHS and make debt service payments to IDCOL due to liquidity problem. Risks Mitigation Timeline Effectiveness Future Options Due to above deficiencies, the security package was again revised: 2016 However, a considerable time of the program In view of this, IDCOL • First charge hypothecation over all fixed and floating assets of PO period had elapsed without any or few management might consider • Personal guarantee of all directors, if it is a limited company, or of the securities. Later, IDCOL tried to introduce more change or waiver of some of executive director/managing director/CEO/chairman/key person of the security requirements which many POs did not the security requirements on a PO to the satisfaction of IDCOL, if it is an NGO/MFI/society/foundation agree. case-by-case basis. • Letter of comfort from the executive committee/governing board of the PO if the PO is NGO/MFI/society/foundation • Lien on all project accounts (that is, Proceeds Account and DSRA) • Maintenance of minimum balance in DSRA equivalent to 2 quarterly installment payments • Bank guarantee to secure 20% of the outstanding refinance amount (from the new POs enlisted in 2015) • Charge documents (demand promissory note and letter of continuity) • Undated cheques (from the new POs enlisted in 2015) • CIB undertaking (from the new POs enlisted in 2015) • Obtain corporate guarantee from sister concern or concerned third party of the PO, as applicable Personal guarantee, corporate guarantee, letter of comfort as well as other securities already obtained from the POs as per Amendment Agreement executed with the POs in July 2012 in addition to the proposed security package would be retained. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 89 90 Risks Mitigation Timeline Effectiveness Future Options Financial risk to Followed Bank of Bangladesh rules regarding loan status classification 2003 • Collection efficiency remained at over 90% • PAYG technology could have IDCOL: Monitoring and regular reporting to Board and taking timely corrective actions through 2010 and then dropped to mid-high made debt collection less and mitigating loan 80% till 2015 and then to 52 percent in 2016 risky, of lower cost, and more defaults Developed and implemented a CEIP to recover bad debts: February 2015 and finally down to 9% in 2018. effective but was introduced • To increase collection from overdue customers • Despite security package required from all too late in the program. • To reduce employee dropouts of POs POs, collection efficiency dropped. Could not • Could have adopted PKSF • To ensure regular customer visit by POs and thereby ensuring after- reverse declining quality of portfolio. principles and strategy, but sales service • Collection efficiencies decline coincided SHS scale of lending would have • To strengthen relationship with local administration market reduction with the acceleration of BREB made transactions costs • To improve coordination between IDCOL inspection team and POs’ grid expansion and connections beginning in high and risks greater single field forces. in 2015. POs had to close non-viable service product (SHS) than for PKSF’s centers and making debt collection more diversified portfolio. Restructured loans of some POs and aligned their debt service 2016–2020 expensive and less effective. obligations with the cash collection process under TR/KABITA and SHS • PAYG introduced first in 2015, to be retrofitted Program to existing SHS. Only required in 2017 as German supplier withdrew from market as Reduced interest rates charged on the POs from 6%–9% per year to July 2016 market was too small, and local alternative 6%–7% per year to support the POs had to be development. Local product cost of US$24 was too expensive for the small 20 Wp Introduce PAYG technology 2017 (US$120) SHS which dominated sales. SHS users objected—why pay a portion of cost, Further reduced interest rate to 4% per year for all POs January 2018 which would cut off supply if they did not pay? Users concerned that this unit used up part of their electricity. The PAYG effort failed. • In 2019–2020, IDCOL renegotiated outstanding debt payment schedule. of PO to reduce substandard debt to only US$28.6 million or only US$7 per SHS. Seeking government debt restructuring relief. IDCOL Board approved October 2018 Agreement reached with the GOB that effective the following proposals for considerations by the government: July 1, 2018, IDCOL would not have to pay • Waiver of interest rate from 3% to 0%, with effect from July 1, 2018, on interest on its SHS borrowings from the GOB and the loan outstanding of IDCOL with the government IDCOL, in turn, will that IDCOL will not charge interest from POs on waive interest on SHS loans to POs to 0%. their repayment of SHS loans also effective July • Allow 10-year time to IDCOL to build up provision amount from its 1, 2018. future revenue earnings. IDCOL will seek exemption from Bangladesh Bank from mandatory provisioning requirement for SHS loans during this period. Subject to approval of the above noted proposals by government authorities, IDCOL will prepare revised repayment schedules after adjustment of DSRA balance for the loans outstanding and allow IDCOL to continue working under TR/KABITA Program through its POs. Risks Mitigation Timeline Effectiveness Future Options Financial risk to POs used to give loans to households for up to 3 years before 2015 when 2003–2015 Debt repayment collections began dropping by Reduce loan tenor POs: Enhancing debt collections started to drop. about 2015. affordability while ensuring financial Only one-year loans offered to customers 2015 Higher monthly payments made loans less recoveries affordable to many of customers • Adoption of PAYG technology 2016 Currently virtually all sales are cash sales. Earlier adoption of PAYG • Cash sales only PAYG introduced too late in the program to be technology. But make sure SHS effective. service levels match customer requirements. Seek BREB support to require SHS customers who wish to obtain a grid 2017 Ineffective as it made the BREB objective of connection to certify they are current on SHS loan repayment. increasing connections more difficult Sustainable market IDCOL approved 17 new POs in 2013 (nearly 60% increase) and then 2013 and 2015 BREB accelerated its grid connections rapidly • Target new entrants to risk: Sustaining 11 more in 2015 (another 23% increase) to serve divisions and districts beginning in 2014 and continuing to date. This areas where markets are sales and service where SHS sales were extremely low. An IDCOL/World Bank study in raised expectations of getting a grid connection not saturated. Too much in a market 2014 estimated the market would be 6 million SHS compared to under 2 soon. It convinced potential customers to stop competition is as bad as too that is capacity million installed at that time. The optimism was fueled by the following: buying SHS and existing customers to return SHS little. constrained • The pace of SHS installations rising to over 860,000 in 2013 or default on loan payments • Closer planning • Good collection efficiencies and good customer loan recoveries Declining demand for SHS led to coordination with grid-based • Rapid drop in SHS prices • Increased competition and price cutting. electrification. • Government’s parallel TR/KABITA giving away large numbers of SHS • Reduced gross margins and revenues to POs. • SREDA to assist in clearer without coordination with IDCOL until 2015–2016 • Closing service centers. policy coordination. • BREB inability to expand grid connections fast and inadequate • Poaching each other’s staff. • Recognize that in densely generation capacity • With SHS price deflationary pressures, some populated countries like • Universal electricity access estimated to take 30 years. POs offering to sell new SHS to customers at Bangladesh, SHS is likely Cumulative SHS installations are now about 5 million and market shows lower cost than repaying old loans and added to be considered a pre- classic signs of saturation. to defaults. electrification option. Seek SREDA support to stop sales of substandard SHS by establishing 2017 Standards alone are ineffective without Consider adopting quality national SHS standards. enforcement or major publicity campaign. labeling such as Lighting Global. To ensure continued service and provision of spares to customers, 2019 In 2015, POs had over 29,000 employees which IDCOL could assess the interest access to such services and spares is needed in customers localities. dropped to 16,000 in 2017 and to 6,000 by 2018. of such employees or POs and Therefore, there are many experienced and identify their requirements to trained technicians, some of whom may wish to set up such enterprises. Either establish their own business that includes sales through IDCOL or through MFIs and service of SHS. Some existing POs may wish or banks, IDCOL could facilitate to undertake the same, especially if they are their entrepreneurs to establish also involved in related business such as solar such businesses. irrigation or solar mini-grids. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 91 92 Risks Mitigation Timeline Effectiveness Future Options Program • Engaged consultants for supporting POs’ collection process 2015-2017 These measures were found to be somewhat Closely monitor early warning management risk: • Appointed program managers in some POs to support their day-to-day effective in maintaining operational and signs and be more proactive in SHS installation operation under IDCOL’s payroll financial performance of the POs. corrective actions. Indicators: and collection • Engaged SHS POs in other solar-based programs of IDCOL • Collection efficiency begins dropped rapidly • Promoted solar-powered energy-efficient appliances with support dropping from CLASP (a US-based not-for-profit organization) • TR/KABITA SHS giveaway • Increased awareness of the community on the benefit of purchasing competition SHS from IDCOL’s POs. • BREB grid expansion accelerating. Avoid sharp increase in competition by mobilizing additional POs. Earlier adoption of PAYG. Incentivize markets in areas with continued low electrification rate and low. Regulatory/policy SREDA was established in 2012 as a statutory body. SREDA sets policy 2012 Earlier establishment of SREDA would have Such policy-making bodies are risk and coordinates renewable energy and energy efficiency issues of the been more effective. By the time SREDA was useful if properly staffed and government. operational, it had little opportunity for empowered coordinating with BREB, improving collection efficiency, and helping IDCOL in its dialogue with the government for debt forgiveness. Waiving tax on renewable energy and CE -certified products (inverter, DC 2017 New sales of SHS had already been dropped Extremely helpful for capital fan, TV, fridge) significantly. intensive products such as SHS Political and other Decline in sale of SHS under the program started at the end of 2013 2013 Movement of PO staff and SHS goods was Little could be done under risks when there was prolonged political turmoil for a couple of months that restricted then, which resulted in decline in these force majeure conditions. lasted up to January 2014. SHS sales to 52,000 per month. The number of sales gradually increased after that and average installation in 2014 was more than 60,000 SHS per month on average. This supports the considerable impact of political problem on the program. Political problem again started in January 2015 and lasted up to March 2015 This was another blow to the program because 2015. of which the number again came down to 46,000 per month. Installation rate continued to decline after that. Risks Mitigation Timeline Effectiveness Future Options Distribution of free SHS under the TR/KABITA Program was another 2015 The distribution of free SHS under KABITA/TR Important to coordinate serious setback for the IDCOL program as installations and quality were fund at the end of 2015 declined SHS sales to with not only grid expansion not verified and no arrangements for service and provision of spares had 20,000 per month. programs but also competing been made. SHS programs. IDCOL was able to convincingly argue to bring the TR/KABITA Program SREDA as a policy-making body under its management and use the infrastructure set up for SHS can play an important role in Program to implement TR/KABITA. It improved product quality and such coordination provided additional business to the POs which used portion of the revenues to repay debt to IDCOL. Environmental and EHS compliance of all IDCOL-enlisted battery manufacturing and 2011 All battery manufacturing and recycling Good model to follow social risks: Battery recycling plants. IDCOL conducts regular inspections monthly (monthly plants follow the compliance of Department manufacture, 2 manufacturing including recycling plants). of Environment, ISO 14001:2004 and OHSAS modules 18001:200, and submit EHS assessment report. manufacture and supply, and battery EHS compliance of all IDCOL-enlisted PV panel suppliers. IDCOL 2014 All local and Imported PV panel suppliers follow recycling conducts regular (Imported & local PV panel suppliers provide all the compliance of Department of Environment, compliance report on a yearly basis. Also, on a random basis, local ISO 14001:2004 and OHSAS 18001:200, and suppliers are monitored by IDCOL. submit EHS assessment report. Expired battery collection and distribution of new battery (collect 2014 IDCOL verifies collections before payments. monthly report from POs and battery recyclers and compile). IDCOL provides a payment of US$10 (US$5 for POs which collected the expired batteries and US$5 for recyclers) per battery that is recycled. Environmental Inspection and verification consultants confirm that outreach and Surveys Survey results indicate positive views of SHS and Survey design and instrument and social risks: training to users, including women, are provided by POs. Call center is conducted benefits to women and children documented can serve as a good model for Monitoring social accessible to all SHS users. in 2009, 2012 and quantified. others. impacts and gender and 2017 responsiveness Consumer surveys measure benefits to women and children as well as consumer satisfaction and technology service and performance. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 93 Appendix D: METHODOLOGICAL ISSUES AND ASSUMPTIONS IN ECONOMIC AND FINANCIAL ANALYSIS D.1 Methodological Issues the most important asset (the solar modules) and replaces shorter-lived assets at intervals outlined in Table D.1. Among the various revealed methods available for Thus, the 2018 tranche of SHS units is physically capable estimating benefit values, only the most conservative of remaining in service until 2029—though the kerosene approach to estimate the avoided cost of kerosene/grid savings benefits do not apply during the latter years of that electricity for lighting is used in the base case. As a result, period, thus forcing downward the expected financial and the immediate-term, household-level benefits presented economic returns. make up only a portion of the total benefits the household members receive over time from the SHS equipment. The impact of foregone global environment damages Nevertheless, these limited avoided cost benefits are shown related to reduced kerosene usage can be added to to pay the households’ costs for the SHS equipment— the Bangladesh society benefits to derive the global so long as the loan terms from the POs for buying the society benefits.39 Consumer surplus benefits accruing to equipment are kept reasonably attractive in real terms households is added to kerosene cost savings to assess the and/or grants are used to help overcome the effect of full value to households of switching to SHS. In this analysis, the government’s SHS tax policies and kerosene subsidy only the consumer surplus due to improved quality and policies. quantity of lighting obtained from switching to electric lighting from kerosene was considered, though the benefits The aggregated household-level analysis calculates the surveys noted previously elicited many other (less tangible) base case benefits and costs of the Bangladesh SHS benefits from the switch to SHS. Program as they accrue to participating households over 2003–2029.38 In analyzing the benefits and costs over the 2003–2029 lifetime of the SHS program assets, the following The base case for the analyses taking the Bangladesh and counterfactuals apply (that is, the without-program global perspectives uses 12 years as the economic life of alternative situation): Table D.1: Expected Useful Life of SHS Component Parts Component Period Duration Warranty Solar module (suggest 12 years as it is more likely other events such as grid 80% initial power output Years 12 arrival, some other physical damage, and so on happen before 20 years) available after 20 years Battery (Survey shows 90% of batteries had useful capacity exceeding 80% of 5-year warranty to minimum Years 5 initial capacity after 5 years) 80% of initial capacity Controller Years 3 3-year warranty Lights (assuming 4 hours/day of use) Fluorescent tube lights (used 2003–2008) Hours 1,500 1-year warranty CFLs (used 2005–2018) Hours 2,000 1-year warranty LED (CFLs began to be replaced with LED beginning about 2008 and were Hours 5,000 3-year warranty nearly wholly replaced by 2014) Balance of system Years 15 38 Assuming a 12-year life for the SHS equipment, the 2018 tranche of equipment will remain in service until 2029. 39 The reduced GHG emissions are valued as recommended in the World Bank Guidance Note on the Shadow Price of Carbon (November 12, 2017). 94 1) During 2003–2013, the counterfactual lighting source 3) During 2022–2029, the counterfactual to lighting by for the SHS households would be kerosene lamps and electricity produced by the SHS units would be lighting lanterns, and the source (at the margin) for the kerosene by electricity from the grid which by 2022 is expected would be incremental imports of already-refined to reach practically all SHS households. During this kerosene (see Box 6). third period, the kWh output of the SHS units is treated as a substitute for grid electricity, as informal surveys 2) During 2013–2021, the counterfactual lighting source of households have indicated they continue to use for the SHS households would be kerosene lamps and the SHS to save on purchased electricity, at least until lanterns, and the source (at the margin) for the kerosene the battery needs replacing. Whether the kWh savings would be Bangladesh-refined kerosene from imported during 2022–2029 are valued at lifeline tariff value in the crude oil (see Box 6). household stakeholder analysis and/or at the long-run marginal cost in the economic analyses, the impact on the FIRR and the EIRR is essentially the same (both are very small numbers in comparison with the 2021 value based on kerosene savings). BOX 6. Methodology Used to Compute Kerosene Cost The sales-weighted average size (in Wp) of SHS units is calculated by year, 2003–2018, from IDCOL data on SHS units sold and installed. This gives a single number for The society-level economic analyses of the SHS each year for (weighted average size) SHS units installed. Program (and the related stakeholder distribution To estimate the total liters of kerosene saved, the above analyses) add a kerosene costing model that is number of SHS units is integrated into the (author- developed as follows: corrected) method for estimating kerosene lighting replaced by SHS. Various studies reviewed and cited in the main text • The kerosene costing model starts with the of this report reveal that SHS downtime for bad weather cost per barrel of crude oil ‘free on board’ (FOB) and normal maintenance yields effective days of SHS40 Dubai, converted to US$ at constant 2018 functioning of about 340 days per year. Thus, all kerosene purchasing power. savings calculations are based on 340-day years. • Regression analysis over the period (2003–2019) provides the implicit kerosene crack spread D.2 Commonality Between Household-Level and between crude oil and kerosene values Society-Level Analyses (estimated as 1.22 to 1.0). • The model uses (a) the current standard The base case cost-benefit analysis takes the nautical shipping factor of US$1.00 per barrel kerosene cost savings and, later, the grid cost per 1,000 nautical miles for shipping petroleum products via large carriers and (b) the 3,155 savings as the benefits for the SHS Program—a nautical mile shipping route Dubai-Chittagong standard cost-benefit analysis method. The to calculate insurance and freight charges SHS provides electricity for more than offsetting to convert Dubai FOB to Chittagong cost, kerosene use for lighting; it offers the potential, insurance, and freight (CIF) values. depending on the capacity of the SHS, to recharge • To get a landed cost at Chittagong, port charges mobile phones conveniently, watch TV, listen to are added at a rate of 5 percent of the estimated insurance and freight, also known as insurance radio, and operate a fan, among others. However, and freight charges (Table D.2). in this simplified, more conservative analysis, the • Marketing and distribution costs of BDT 14.59 economic benefits from direct cost savings due per liter in 2012 values are converted to US$ at to switching from kerosene to electricity lighting constant 2018 value and added to the landed and the impact of reducing GHG emissions are cost to get the local market cost for kerosene considered. The implication of this assumption (Table D.3). is that the economic and financial rates of return The model uses the above-derived values computed are conservative estimates. The in computing the economic value of liters of simplifying assumption is used to avoid a debate kerosene saved by the SHS Program during 2003– 2022 and to estimate the stakeholder distribution on how to value intangibles related to a more of program impacts. comfortable living environment, watching TV, enhanced quality of home life, greater security, and improved communication. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 95 Central to the affordability of the SHS units is their primary status as consumer goods in the immediate term that are not expected to generate significant immediate additional flows of cash income for the households purchasing them. Thus, the SHS units must largely be purchased from existing assets and cash flows prevailing in the households’ countervailing situation—including remittances from family members working abroad. D.3 Kerosene Economic Cost Estimation Table D.2: Kerosene Cost Estimation: Landed Cost at Chittagong 2003–2018 (constant 2018 US$/liter) Freight & Insurance and Kerosene Landed Cost at Year Crude Oil FOB Cost (Dubai) Kerosene FOB Costa Port Charges b Chittagong 2003 0.22 0.27 0.02 0.29 2004 0.27 0.33 0.02 0.35 2005 0.38 0.47 0.02 0.49 2006 0.46 0.57 0.02 0.59 2007 0.50 0.61 0.02 0.63 2008 0.67 0.82 0.02 0.84 2009 0.44 0.53 0.02 0.55 2010 0.55 0.67 0.02 0.69 2011 0.73 0.89 0.02 0.91 2012 0.74 0.90 0.02 0.92 2013 0.70 0.86 0.02 0.88 2014 0.60 0.73 0.02 0.75 2015 0.33 0.40 0.02 0.42 2016 0.26 0.32 0.02 0.34 2017 0.33 0.41 0.02 0.43 2018 0.43 0.52 0.02 0.54 Note: a. 1.22 times crude oil cost estimated by regressing kerosene FOB cost versus Dubai crude oil FOB cost from 2003 to 2019. b. Calculated as 3,155 nautical miles from Dubai to Chittagong at a cost of US$1 per barrel per 1,000 nautical miles and 5 percent surcharge for port handling (http://cost-finder.com/what-is-the-cost-of-shipping-oil-by-tanker/) Table D.3: Kerosene Cost Estimation: From Landed Cost at Chittagong to Official versus Actual Retail Prices per Liter, 2003–2018 (constant 2018 US$/liter) Kerosene Landed Domestic Cost of Kerosene at Official Price of Year Subsidy b Cost at Chittagong Distribution Costa Retail Kerosene 2003 0.29 0.12 0.41 0.38 0.03 2004 0.35 0.12 0.47 0.36 0.11 2005 0.49 0.13 0.62 0.57 0.05 2006 0.59 0.13 0.72 0.56 0.16 2007 0.63 0.14 0.77 0.68 0.10 2008 0.84 0.15 0.99 0.92 0.08 2009 0.55 0.16 0.71 0.72 (0.01) 2010 0.69 0.17 0.86 0.82 0.04 2011 0.91 0.18 1.10 0.83 0.27 2012 0.92 0.19 1.11 0.67 0.44 2013 0.88 0.20 1.08 0.92 0.16 2014 0.75 0.21 0.97 0.82 0.15 2015 0.42 0.22 0.65 0.88 (0.24) 2016 0.34 0.23 0.58 0.88 (0.31) 2017 0.43 0.24 0.67 0.85 (0.18) 2018 0.54 0.26 0.80 0.76 0.04 Note: a. Domestic distribution cost of BDT 14.59 per liter in 2012 adjusted for inflation (Energia 2019). b. Excludes taxes and duties charged on kerosene which was in the 32–34 percent range and 18–20 percent for crude oil imports on CIF value, as per Bangladesh customs schedules. A ‘negative’ subsidy same as a tax. 96 D.4 Estimating Kerosene Fuel Displacement by SHS Grameen Shakti conducted a survey in 2012 of 441 SHS users with systems ranging in size from 20 Wp to 85 Wp to assess the number and types of kerosene lamps and hours of their use that were displaced by the SHS (UNFCCC 2013). Based on this survey data, and kerosene consumption per lamp type from Mills (2003), Table D.4 was compiled. Table D.4: Summary Survey Data (Grameen SHS CDM Project 2012) System (Wp) of Hurricane Lamps Used Lamps Used Sample Size Used (liters) Average No. Daily Hours Daily Hours of Usage of before SHS before SHS No. of Kupi Hurricanes Use (liters/ Use (liters/ Use (liters) Hurricane day) Kupi Kerosene Kerosene Kerosene Kerosene Average of Kupi Annual Usage Daily day) 20 1.0 2.0 4.0 4.0 1 0.12 0.12 0.24 81.60 40 2.0 1.5 3.8 1.9 67 0.23 0.04 0.27 92.81 50 2.5 1.8 4.5 2.3 165 0.34 0.06 0.40 135.96 65 3.0 2.2 5.1 2.8 122 0.46 0.09 0.55 187.46 85 3.6 2.5 5.4 3.2 86 0.58 0.12 0.70 237.41 Note: Calculated from survey data in 2765 CPA CER Sheet Grameen Shakti 28 Jun 13.xlsx. In their CDM application, Grameen Shakti also committed to offsetting a certain amount of kerosene (and therefore CO2 emissions). Their assumptions are presented in Table D.5. Table D.5: Grameen Shakti Kerosene Avoided in CDM Application SHS Range, Wp System (Wp) Sample Size 120 and 20 to 39 40 to 74 75 to 119 Above No. of kerosene lamps that would have Conservative assumption based 1 2 3 4 been used on Grameen Shakti Survey Default value as per AMS-I.A Average usage hours per day 3.5 3.5 3.5 3.5 methodology, version 14, EB 54 Specific fuel consumption (liter/hour) 0.04 0.04 0.04 0.04 Based on Grameen Shakti Survey Annual maintenance days 25 25 25 25 Assumption Annual days considered for savings 340 340 340 340 Calculation Annual saving of kerosene per lamp 47.6 47.6 47.6 47.6 Calculation (liter/year) Total annual saving of kerosene per SHS 47.6 95.2 142.8 190.4 Calculation (liter/year) Source: UNFCCC 2013. Most notably, the CDM credit is based on the following: (a) Number of lamps offset per SHS size in the CDM application is less than from their survey (for example 1 in application versus 3 in the survey for 20 Wp, 2 in application versus 3.5 in survey for 40 Wp) (b) Number of hours operating per day is less than from their survey (for example, 3.5 per day in CDM application versus 4 hours per lamp from the survey for 20 Wp, 5.7 hours for 40 Wp, and 6.8 hours for 50 Wp) (c) Kerosene offset per hour is more than from their survey (0.04 liters per hour versus 0.015 to 0.03 liters per hour in the survey (note also that there was an error by Grameen Shakti in computing kerosene used in the CDM application as the kerosene consumption per hour between hurricane lanterns and kupi lamps had been reversed). LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 97 D.5 Kerosene Offset Estimates from 2003–2021 Moreover, the authors assumed that the kerosene avoided for lighting in a household with a 45 Wp and larger SHS For the economic and financial analysis, the authors would not be more than for a user who purchased a 50 Wp decided to use the survey data in Table D.4 as it represents SHS as the extra electricity will be used for other purposes. actual consumer usage rather than a commitment that the These other purposes include mobile phone charging, CDM applicant must meet to get CDM credits. The latter TV viewing, and so on, which would not have consumed commitment needs to be conservative, as not meeting the kerosene. commitment means foregoing the CDM payments. Table D.6: Kerosene Offset (liters per year), Based on Grameen Shakti Survey Data Kerosene Offset (liters/ 82 93 136 136 136 year/SHS) Average Liters/ SHS Range (Wp) 10–21 25–40 45–65 70–90 100–300 SHS/Year a Representative SHS (Wp) 20 40 50 85 100 2003 — 261,900 720,718 123,858 5,574 123 2004 — 572,802 1,384,468 284,018 7,478 122 2005 — 681,386 2,107,633 453,966 1,767 124 2006 — 640,830 3,194,624 723,709 816 128 2007 — 1,044,724 5,783,415 1,193,446 136 128 2008 74,827 1,922,025 8,866,416 1,857,877 18,218 127 2009 1,989,979 3,141,689 10,814,434 2,581,314 8,158 118 2010 4,841,573 5,030,581 12,295,569 12,431,528 25,424 117 2011 9,356,256 7,146,943 23,460,641 8,325,436 44,459 114 2012 19,141,891 10,610,207 26,090,899 9,582,647 146,835 107 2013 37,152,480 15,170,166 23,903,050 8,343,791 711,200 99 2014 33,821,323 14,309,106 15,628,871 5,433,185 399,447 96 2015 23,820,427 14,040,524 12,953,066 4,489,767 554,712 97 2016 5,967,734 5,368,397 4,139,945 1,452,992 526,704 99 2017 791,602 967,787 783,802 304,140 182,729 103 2018 92,453 131,971 72,738 25,696 23,929 100 Weighted Average 105 Note: a. Analysis conservatively assumes that SHS larger than 50 Wp are purchased to provide services such as TV viewing and not more lighting than a 50 Wp SHS would provide. 98 Table D.7: Calculation of Total Liters of Kerosene Use Avoided during 2003–2022 by SHS Program Installations 2003–2018 Total Liters of Kerosene Use Saved 2003 to 2021 by Year Average Liters/SHS/Year No. of SHS Installed by Year SHS Installed in that Year 2003 123 9,075 13,394,700 2004 122 18,499 27,082,536 2005 124 26,196 38,979,648 2006 128 35,731 54,882,816 2007 128 62,574 96,113,664 2008 127 100,640 153,375,360 2009 118 156,827 222,067,032 2010 117 295,597 415,018,188 2011 114 425,788 533,938,152 2012 107 612,373 655,239,110 2013 99 861,172 767,304,252 2014 96 726,512 557,961,216 2015 97 575,580 390,818,820 2016 99 175,990 104,538,060 2017 103 29,475 15,179,625 2018 100 3,455 1,382,000 Total 4,115,484 4,047,275,179 Table D.8: Kerosene Saved by the Bangladesh SHS Program, 2003 to 2022: Value at the Household Level, Aggregated SHS Households Total Liters Saved by SHS Total Value of Kerosene Saved by Kerosene Price at Retail Year Households SHS Households Constant 2018 US$ Liters of Kerosene Constant 2018 US$ 2003 0.41 1,116,225 452,708 2004 0.47 3,373,103 1,591,890 2005 0.62 6,621,407 4,088,288 2006 0.72 11,194,975 8,054,024 2007 0.77 19,204,447 14,853,480 2008 0.99 31,985,727 31,808,530 2009 0.71 50,491,313 35,873,533 2010 0.86 85,076,162 73,131,236 2011 1.10 133,615,994 146,853,601 2012 1.11 199,139,905 221,417,864 2013 1.08 284,395,933 306,650,972 2014 0.97 354,141,085 341,821,510 2015 0.65 408,856,120 264,707,393 2016 0.58 424,022,252 244,498,673 2017 0.67 423,809,873 283,955,293 2018 0.80 419,581,805 335,825,674 2019 0.80 411,572,333 329,415,037 2020 0.80 398,791,053 319,185,133 2021 0.80 380,285,467 304,373,597 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 99 Table D.9: Estimation of Kerosene Subsidy Saved by the GOB as a Result of SHS Program Electricity Output (constant 2018 US$) Subsidy per Liter of Kerosene Total GOB Subsidy Saved Year Constant 2018 US$ Constant 2018 US$ 2003 0.03 1,051 2004 0.11 373,530 2005 0.05 341,938 2006 0.16 1,737,990 2007 0.10 1,833,423 2008 0.08 2,441,209 2009 (0.01) (399,150) 2010 0.04 3,294,865 2011 0.27 35,807,470 2012 0.44 87,689,034 2013 0.16 44,552,951 2014 0.15 52,446,940 2015 (0.24) (96,749,515) 2016 (0.31) (130,369,351) 2017 (0.18) (74,323,574) 2018 0.04 17,590,248 2019 0.04 17,254,464 2020 0.04 16,718,631 2021 0.04 5,942,816 Table D.10: Kerosene Saved by the Bangladesh SHS Program, 2003 to 2022: Value at the Household Level, Aggregated SHS Households Kerosene Retail Domestic Distribution Costa Kerosene Saved Profit Losses by Distributors Year Constant 2018 US$/Liter Constant 2018 US$/Liter Liters Constant 2018 US$ 2003 0.41 0.12 1,116,225 6,588 2004 0.47 0.12 3,373,103 20,961 2005 0.62 0.13 6,621,407 42,736 2006 0.72 0.13 11,194,975 74,791 2007 0.77 0.14 19,204,447 136,014 2008 0.99 0.15 31,985,727 242,004 2009 0.71 0.16 50,491,313 396,430 2010 0.86 0.17 85,076,162 718,299 2011 1.10 0.18 133,615,994 1,233,627 2012 1.11 0.19 199,139,905 1,912,847 2013 1.08 0.20 284,395,933 2,884,660 2014 0.97 0.21 354,141,085 3,776,109 2015 0.65 0.22 408,856,120 4,578,280 2016 0.58 0.23 424,022,252 4,969,268 2017 0.67 0.24 423,809,873 5,146,173 2018 0.80 0.26 419,581,805 5,483,953 2019 0.80 0.26 411,572,333 5,379,269 2020 0.80 0.26 398,791,053 5,212,217 2021 0.80 0.26 380,285,467 4,970,348 Note: a. Domestic distribution cost of BDT 14.59 per liter in 2012 adjusted for inflation (Energia 2019). 100 Table D.11: Financial and Economic Value of Grid Electricity Offset by SHS from 2022 to 2029 Total 2022 2023 2024 2025 2026 2027 2028 2029 (2018 US$) Quantity MWh/year 150,450 126,860 96,359 60,049 32,036 9,085 1,572 161 476,572 Financial value of Millions of 6.28 5.29 4.02 2.51 1.34 0.38 0.07 0.01 19.88 electricity offset US$/year Economic value of Millions of 16.85 14.21 10.79 6.73 3.59 1.02 0.18 0.02 53.38 electricity offset US$/year Electricity tariff 3.5 BDT/ kWh 41.72 2018 US$/MWh Risad 2017. (lowest block) Electricity cost from highest cost 8.4 BDT/kWh at HV from rental diesel plants See Table 2 and Figure 6 in Moazzem and Ali (2019). generator System losses 11.87% LRMC at LV 9.40 BDT/ kWh 112.00 2018 US$/MWh Note: Assuming 3.5 kWh/kWp of SHS output and 340 days per year availability, with SHS life of 12 years. LRMC at LV means Long Run Marginal Cost at Low Voltage. D.6 SHS Costs Table D.12: SHS Installation Data MWp Average Cost Total Initial MWp Installed MWh Supplied Average Size Year Number of SHS Operating in (constant 2018 Cost (constant in Year per Year Wp Year US$/Wp) 2018 US$) 2003 9,075 0.45 0.45 539 49.89 12.00 5,433,015 2004 18,499 0.94 1.39 1,654 50.64 11.55 10,823,556 2005 26,196 1.35 2.74 3,266 51.71 11.04 14,951,360 2006 35,731 1.98 4.72 5,622 55.42 10.18 20,156,264 2007 62,574 3.49 8.21 9,771 55.72 10.37 36,145,797 2008 100,640 5.58 13.79 16,416 55.48 10.81 60,359,933 2009 156,827 7.73 21.52 25,613 49.28 10.61 82,020,719 2010 295,597 14.70 36.22 43,104 49.72 9.89 145,327,667 2011 425,788 19.82 56.05 66,694 46.56 9.03 178,974,476 2012 612,373 25.63 81.68 97,195 41.86 8.65 221,812,910 2013 861,172 30.51 112.19 133,505 35.43 8.64 263,604,768 2014 726,512 23.54 135.73 161,518 32.40 6.42 151,068,465 2015 575,580 19.29 154.56 183,930 33.51 5.96 114,906,534 2016 175,990 6.31 159.94 190,329 35.88 4.89 30,857,755 2017 29,475 1.19 159.77 190,131 40.31 4.76 5,658,327 2018 3,455 0.13 157.93 187,935 39.05 4.81 649,130 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 101 Table D.13: SHS Program Costs by Year Total Cost of SHS SHS Cost Total Cost to Replacement Cost Year Installed without including Stock of Grant to Households Households of SHS Multiplier a Grant Replacement Partsa Installed with Grant PV in Constant 2018 Constant 2018 US$ Constant 2018 US$ US$ at 10% Discount 2003 5,433,015 1.45 7,877,872 1,049,138 6,828,734 2004 10,823,556 1.45 15,694,156 1,924,913 13,769,243 2005 14,951,360 1.45 21,679,472 1,935,058 19,744,414 2006 20,156,264 1.45 29,226,583 1,864,741 27,361,842 2007 36,145,797 1.45 52,411,406 2,491,036 49,920,370 2008 60,359,933 1.45 87,521,903 5,072,273 82,449,630 2009 82,020,719 1.40 114,829,007 6,792,034 108,036,973 2010 145,327,667 1.40 203,458,734 10,801,213 192,657,521 2011 178,974,476 1.40 250,564,266 12,832,039 237,732,227 2012 221,812,910 1.44 319,410,590 15,308,856 304,101,734 2013 263,604,768 1.44 379,590,866 11,380,928 368,209,938 2014 151,068,465 1.44 217,538,590 8,850,822 208,687,768 2015 114,906,534 1.30 149,378,494 7,205,686 142,172,808 2016 30,857,755 1.30 40,115,082 2,336,844 37,778,238 2017 5,658,327 1.30 7,355,825 321,844 7,033,981 2018 649,130 1.30 843,869 33,988 809,881 Sum 1,342,750,676 1,897,496,714 90,201,413 1,807,295,301 Note: a. PV at 10% discount rate for a stream of replacements for batteries (5 years), controllers (3 years), and lamps (2–3 years) that do not last as long as the solar modules and are replaced periodically over 12 years. D.7 Conversion Factors to Economic Costs of SHS previously. This was mainly due to differing ways POs had compiled SHS cost data. The data assembled by the authors Analysis of SHS component-level cost data indicated a are shown below. Given the inconsistence and not to wide range of tax and duty rates, often inconsistent and introduce spurious precision, a standard conversion factor inconsistent with Harmonized System (HS) codes shown (CF) of 0.89 was used in the economic analysis. Table D.14: Tax and Duty Information Bangladesh Tariffs by HS Codes (%) a HS Code Component/System 2011–2012 2015–2016b 2017–2018 2019–2020c 85013110 SHS (complete) 11.11 10.05 11.12 12.40 85414010 Solar panels 5.00 5.00 5.00 11.33 85072010 Batteries 37.23 60.02 58.69 60.31 85395000 LED lamps n.a. 60.02 43.08 44.53 85399021 CFL components 8.00 24.00 37.07 38.48 85393120 Fluorescent tube lamps 37.23 60.02 58.69 60.31 85363010 Other electronics 36.23 36.01 37.07 38.47 83119000 Wires, rods, and so on 37.23 36.01 37.07 38.47 85444900 Electrical wires 58.58 60.02 58.69 60.31 Note: a. Includes customs duty (CD), supplementary duty (SD), value added tax (VAT), advanced income tax (AIT), regulatory duty (RD), and advanced trade VAT (ATV). See https://www.scribd.com/doc/128218736/Calculation-of-Total-Tax-Incidence. b. For 2015–2016, LED lamp not specified. Instead used HS 8539110 (Energy saving lamp with 3 times efficiency of incandescent lamps. c. For latest tariffs, see http://www.bangladeshcustoms.gov.bd/trade_info/duty_calculator. 102 Table D.15: Imputed SHS Approximate Conversion Factors by Year or Year Groups Year Apparent Overall Average Tax Rate Applied for SHS Units (%) CF (Economic Value/Financial Value) 2003–2008 16.30 0.86 2009 20.00 0.83 2010 21.00 0.83 2011 10.0 0.91 2012 5.26 0.95 2013 5.26 0.95 2014 16.50 0.86 2015 2016–2018 5.26 0.95 Assumed value for all years (share of final cost) 0.89 Table D.16: Calculation of Tax Component of GOB Stakeholder Impact (constant 2018 US$) SHS Financial Costs (Including SHS Cost CFs Economic Costs Taxes on SHS Years Grants Replacements) without Grant (Economic/ Financial) Excluding Taxes Paid to GOB Constant 2018 US$ 2003 7,877,872 1,049,138 0.89 7,033,814 844,058 2004 15,694,156 1,924,913 0.89 14,012,639 1,681,517 2005 21,679,472 1,935,058 0.89 19,356,671 2,322,801 2006 29,226,583 1,864,741 0.89 26,095,163 3,131,420 2007 52,411,406 2,491,036 0.89 46,795,898 5,615,508 2008 87,521,903 5,072,273 0.89 78,144,556 9,377,347 2009 114,829,007 6,792,034 0.89 102,525,899 12,303,108 2010 203,458,734 10,801,213 0.89 181,659,584 21,799,150 2011 250,564,266 12,832,039 0.89 223,718,095 26,846,171 2012 319,410,590 15,308,856 0.89 285,188,027 34,222,563 2013 379,590,866 11,380,928 0.89 338,920,416 40,670,450 2014 217,538,590 8,850,822 0.89 194,230,884 23,307,706 2015 149,378,494 7,205,686 0.89 133,373,656 16,004,839 2016 40,115,082 2,336,844 0.89 35,817,037 4,298,044 2017 7,355,825 321,844 0.89 6,567,701 788,124 2018 843,869 33,988 0.89 753,454 90,415 Sum 1,342,750,676 1,897,496,714 90,201,413 1,807,295,301 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 103 D.8 Estimate of PO Profits for Stakeholder Two data sources are used in attempting to model PO Analysis profits: (a) a scattered sample of PO-stated markups as a percentage of SHS unit selling prices in the SHS cost There is considerable uncertainty in estimates of PO profits. breakdowns made available to the authors and (b) a The estimates are based on financial statements from SHS small number of PO financial statements. The most and cost breakdown data and two of the largest POs. The PO best data tended to be available from Grameen Shakti, profit on SHS sales is estimated at 12 percent of sales from the largest and earliest PO in the SHS program. But even 2003 to 2012, 2 percent from 2013 to 2015, and 0 percent those data were incomplete—and sometimes seemingly from 2016 to 2018. However, with over 30 POs, the variation inconsistent—within and between years. Company financial in net profits would be considerable. statements for the two largest POs were available for four of the later program years—2013/14 through 2016/17. Both the In the SHS Program, one would expect the early existence Grameen Shakti and the RSF financial statement summaries (2003–2008) of higher profits to attract POs into the SHS show overall profits for 2013/14 and losses for the other program, normal profits in the middle years (2008–2013), three operating years (see BIDS 2018). and perhaps negative profits overall in the later years as market saturation occurred (2014–2019)—and in the de Data issues such as this force an indicative rather than facto case, as some households fail to repay SHS purchase definitive analysis that employs stylized facts (Kaldor 1961) loans. rather than complete, audited discrete values (see Table D.17). Table D.17: Estimated Partner Organization Profits on SHS Sales and Installations SHS Costs without Grant PO Imputed Profit Rate PO Profit on SHS Sales Years (constant 2018 US$) (as % of SHS cost) (constant 2018 US$) 2003 5,433,015 12 651,962 2004 10,823,556 12 1,298,827 2005 14,951,360 12 1,794,163 2006 20,156,264 12 2,418,752 2007 36,145,797 12 4,337,496 2008 60,359,933 12 7,243,192 2009 82,020,719 12 9,842,486 2010 145,327,667 12 17,439,320 2011 178,974,476 12 21,476,937 2012 221,812,910 12 26,617,549 2013 263,604,768 2 5,272,095 2014 151,068,465 2 3,021,369 2015 114,906,534 2 2,298,131 2016 30,857,755 0 — 2017 5,658,327 0 — 2018 649,130 0 — Total 1,342,750,676 7.72 103,712,279 104 Appendix E: ESTIMATE OF INDICATIVE NET FINANCIAL BENEFITS TO IDCOL FROM SHS PROGRAM MANAGEMENT This appendix gives an indicative estimate of IDCOL’s IDCOL was able to restructure repayment terms of the net financial benefits from the management of the SHS outstanding debt such that only US$28.6 million was Program. It was expected that the SHS Program would considered substandard (risky). permit IDCOL to earn income to cover the cost of managing the program and earn a profit. IDCOL made loans to the Table E.1 shows indicative estimates of the net financial POs and earned revenue from repayment of principal benefits to IDCOL of managing the SHS Program. Data up to and interest by the POs and from administration fees42 2018 are from IDCOL, subject to assumptions as noted below. from development partners for managing the funds they From 2019 to 2021 when the SHS Program under the RERED provided for the SHS Program. IDCOL also earned additional Project closes, costs are assumed to continue at the same income from reinvesting reflows that arose from repayments level as in 2018. The table assumes POs’ outstanding debt of of SHS loans by POs, before IDCOL had to repay the loans to BDT 12,525 million (classified as Standard in 2019 in Table 11 the government.43 This was possible because originally the on loan status) will be repaid to IDCOL over five years at zero GOB required IDCOL to repay the loans over 20 years with a interest as agreed with the GOB. It also assumes repayment 5-year grace period, while loans to POs were repaid in 5–10 only of principal by IDCOL to the GOB from 2019 onward as years with 0.5 to 2 years of grace. There was also an interest per the same agreement with the GOB. rate differential—the GOB lent at 3 percent interest to IDCOL, The undiscounted IDCOL net financial benefit was estimated while IDCOL lent at 4–7 percent interest to the POs. The at US$54 million in constant 2018 US$ (note that in current lending terms to the POs varied over time (see Table E.1). US$ terms, there was a small loss). Discounted at the Costs to IDCOL included repaying the loan principal and the societal rate of 10 percent, the NPV in 2018 was US$379 interest for SHS loans refinanced by the government, direct million in constant 2018 US$. IDCOL’s NPV is sensitive to the SHS Program Management Unit (PMU) costs, and other discount rate; it has the unusual characteristic of increasing overhead and general and administrative costs. as the discount rate increases because the net benefits are Due to the difficulties explained in Chapter 4, not all POs larger in initial years than later years. Given that IDCOL’s were able to meet their full financial repayment obligations. weighted average cost of capital is about 3 percent and its The repayment obligations of POs for about US$143 million average return on equity was 2.5 per cent in constant terms were rated as questionable as of 2018 (see Table 11). (about 8.5 in current terms over 2016–201844 reduced by Consequently, the government agreed, retroactively from average inflation of 6 percent over 2016–2018), the midpoint July 2018 onward, to forgive IDCOL interest payments on of about US$138 million in constant US$ of 2018 between SHS loans that were refinanced and concurrently IDCOL the undiscounted net benefit and the net benefit discounted agreed to forgive interest payments by POs for outstanding by 5 percent is an indicative estimate of the NPV of the net debt from that point forward. Furthermore, in 2019–2020, financial benefits to IDCOL over the life of the SHS Program. 42 IDCOL received fees for administrating the SHS Program from KfW, GIZ, and ADB. Between 2007 and 2018, these amounted to BDT 590 million according to the financial statements in IDCOL Annual Reports. In addition, IDCOL was permitted by the GOB to retain 3 percent of the interest earned from loans to POs for administering the RERED Project for an additional BDT 339 million. For details, see IDCOL Annual Reports 2006–2008 to 2018. 43 No estimate was made of this income as it was not possible to segregate reflows from loans given to POs for financing SHS from reflows from all RERED loans. 44 IDCOL’s return on equity was taken from IDCOL Annual Reports 2016–2018. Gross domestic product (GDP) deflator from World Bank data. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 105 106 Table E.1: IDCOL Net Financial Benefit from Management of SHS Program, 2003–2042 Current US$, millions Constant 2018 US$, millions IDCOL Estimated Net Present Value to 2018 POs Principal Years Principal IDCOL Income SHS Program Other and Interest Admin. Fees IDCOL Net IDCOL Net and Interest from SHS Direct PMU Expenses for Repayments to Received Benefits Benefits Repayments to Program Cost Managing SHS Discounted 5% Discounted 10% IDCOL GOB Program 1 2 3 4=2+3–1 5 6 7=4−5−6 8 9 10 2003 0.01 n.a. — (0.01) — 0.02 (0.02) (0.03) (0.06) (0.12) 2004 0.07 n.a. — (0.07) — 0.02 (0.09) (0.12) (0.23) (0.44) 2005 0.10 0.56 — 0.46 — 0.02 0.45 0.57 1.07 1.96 2006 0.14 1.58 — 1.44 — 0.04 1.40 1.72 3.08 5.39 2007 0.20 3.58 0.62 4.00 0.06 0.07 3.87 4.62 7.90 13.18 2008 0.31 7.17 1.28 8.14 0.08 0.14 7.92 9.28 15.12 24.07 2009 0.55 13.01 0.95 13.41 0.13 0.24 13.04 15.16 23.52 35.76 2010 2.00 20.85 2.25 21.09 0.23 0.33 20.53 23.58 34.84 50.55 2011 4.25 28.26 1.06 25.06 0.39 0.41 24.27 27.30 38.41 53.20 2012 7.07 34.99 0.36 28.28 0.53 0.49 27.26 30.09 40.32 53.30 2013 11.02 43.81 0.21 33.00 0.80 0.54 31.66 34.35 43.84 55.32 2014 16.30 52.14 0.63 36.47 0.91 0.58 34.98 37.24 45.26 54.52 2015 22.79 56.12 0.31 33.64 0.60 0.60 32.44 34.17 39.56 45.49 2016 24.50 59.59 0.33 35.41 0.73 0.61 34.07 35.50 39.14 42.95 2017 34.72 59.93 0.14 25.35 0.97 0.62 23.76 24.30 25.51 26.73 2018 33.25 57.86 0.03 24.64 1.07 0.62 22.95 22.95 22.95 22.95 2019 30.72 29.64 n.a. (1.08) 1.07 0.62 (2.76) (2.72) (2.59) (2.47) 2020 32.40 29.64 n.a. (2.76) 1.07 0.62 (4.45) (4.30) (3.90) (3.55) 2021 32.24 29.40 n.a. (2.84) 1.07 0.62 (4.53) (4.30) (3.71) (3.23) 2022 32.63 29.16 n.a. (3.47) n.a. n.a. (3.47) (3.24) (2.67) (2.21) 2023 33.06 28.92 n.a. (4.14) n.a. n.a. (4.14) (3.80) (2.98) (2.36) 2024 33.02 n.a. n.a. (33.02) n.a. n.a. (33.02) (29.76) (22.21) (16.80) Current US$, millions Constant 2018 US$, millions IDCOL Estimated Net Present Value to 2018 POs Principal Years Principal IDCOL Income SHS Program Other and Interest Admin. Fees IDCOL Net IDCOL Net and Interest from SHS Direct PMU Expenses for Repayments to Received Benefits Benefits Repayments to Program Cost Managing SHS Discounted 5% Discounted 10% IDCOL GOB Program 2025 33.02 n.a. n.a. (33.02) n.a. n.a. (33.02) (29.25) (20.79) (15.01) 2026 33.06 n.a. n.a. (33.06) n.a. n.a. (33.06) (28.79) (19.48) (13.43) 2027 33.16 n.a. n.a. (33.16) n.a. n.a. (33.16) (28.37) (18.29) (12.03) 2028 33.30 n.a. n.a. (33.30) n.a. n.a. (33.30) (28.01) (17.19) (10.80) 2029 24.81 n.a. n.a. (24.81) n.a. n.a. (24.81) (20.51) (11.99) (7.19) 2030 20.42 n.a. n.a. (20.42) n.a. n.a. (20.42) (16.59) (9.24) (5.29) 2031 20.12 n.a. n.a. (20.12) n.a. n.a. (20.12) (16.06) (8.52) (4.65) 2032 13.84 n.a. n.a. (13.84) n.a. n.a. (13.84) (10.86) (5.49) (2.86) 2033 7.28 n.a. n.a. (7.28) n.a. n.a. (7.28) (5.61) (2.70) (1.34) 2034 3.89 n.a. n.a. (3.89) n.a. n.a. (3.89) (2.95) (1.35) (0.64) 2035 2.42 n.a. n.a. (2.42) n.a. n.a. (2.42) (1.80) (0.79) (0.36) 2036 2.42 n.a. n.a. (2.42) n.a. n.a. (2.42) (1.77) (0.74) (0.32) 2037 2.14 n.a. n.a. (2.14) n.a. n.a. (2.14) (1.54) (0.61) (0.25) 2038 2.14 n.a. n.a. (2.14) n.a. n.a. (2.14) (1.51) (0.57) (0.23) 2039 2.14 n.a. n.a. (2.14) n.a. n.a. (2.14) (1.49) (0.54) (0.20) 2040 2.15 n.a. n.a. (2.15) n.a. n.a. (2.15) (1.47) (0.50) (0.18) 2041 2.15 n.a. n.a. (2.15) n.a. n.a. (2.15) (1.45) (0.47) (0.16) 2042 1.08 n.a. n.a. (1.08) n.a. n.a. (1.08) (0.71) (0.22) (0.07) Total 590.9 586.2 8.2 3.5 9.7 7.2 (13.4) 53.8 222.7 379.2 Note: Administrative fees from the RERED Project are 3 percent of interest earned from loans to POs which are included in revenues from POs. Other expenses in 2003–2004 are assumed to have been the same as in 2005. 2018 Other Expenses are assumed to remain constant from 2019–2021 until the SHS Program ends in 2021. Similarly, PMU cost in 2018 is assumed to remain the same from 2019–2021. The table excludes income from investing in reflows of funds back from POs. The analysis assumes POs’ outstanding debt of BDT 12,525 million, classified as Standard in 2019, will be repaid to IDCOL over five years at zero interest as per agreement IDCOL made with the GOB on forgiving interest. It also assumes repayment only of principal to the GOB from 2019 onward. IDCOL agreed to reschedule the PO debt repayment to extend it from 2023 to 2026. This rescheduled repayment is not reflected in this table. LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 107 Appendix F: IMPACT ON GOB TREASURY FROM SHS PROGRAM LOAN TRANSACTIONS Table F.2 is compiled by IDCOL based on ODA lending terms drawing rights (SDR) (Yen in case of the JICA loan) and shown in Table F.1. then converted into US dollars at different repayment dates considering actual/projected exchange rates of The withdrawal and ODA loan repayment schedules for SDR/US$ on those repayment dates. each loan/credit shown above were compiled by IDCOL finance and accounting staff. They are based on actual (b) Repayments from IDCOL to the GOB: Repayment disbursements under the program as recorded by IDCOL schedules have been prepared in BDT and then at different times under various development partner converted into US dollars at different repayment dates accounts. Historical exchange rates have been used considering actual/projected exchange rates of BDT/US$ to calculate projected exchange rates among different on those repayment dates. currencies. The IDCOL to GOB repayment schedule reflects the (a) Repayments from the GOB to development partners: agreement reached that no interest will be charged on Repayment schedules have been prepared in special IDCOL repayments to the GOB from July 1, 2018, onward. Repayments are typically twice a year. Table F.1: ODA Lending Terms Service Commit- Principal Principal Charge on ment Fee Grace Repayment 1 Repayment 2 Loans/ Project Approved Closing Disbursed on Undis- Period Credits ID Date Date Repay Repay Balance bursed (Years) Period Period per Year per Year (%) Balance (%) (Years) (Years) (%) (%) World Bank June 30, 2008 June 25, RERED P071794 (revised to 0.75 0.50 10 10 2.000 20 4 2002 December 31, 2012) August 4, December RERED AF P112963 0.75 0.50 10 10 2.000 20 4 2009 31, 2012 RERED October 4, December P126724 0.75 0.50 10 10 2.000 20 4 AF 2 2011 31, 2012 December 31, 2018 September RERED II P131263 (revised to 0.75 0.50 10 10 2.000 20 4 20, 2012 December 31, 2021) December 31, 2018 RERED June 19, P150001 (revised to 0.75 0.50 10 10 2.000 20 4 II AF 2014 December 31, 2021) Other Development Partners ADB 2009 Closed 1.24% n.a. 7 21 4.762 N/A 0 JICA 2013 Closed 0.01% n.a. 10 30 3.333 N/A 0 IsDB 2011 Closed 0.75% n.a. 6 19 5.263 N/A 108 Table F.2: ODA Loan Withdrawals and Repayments by Bangladesh Government 2003–2055 Constant 2018 Current US$, millions US$, millions Years Loan Loan Repayments Net Loan Minus Net Loan Minus Withdrawals Commitment and Interest Principal Total Payment Payment Payment 2003 1.63 — — — 1.63 2.19 2004 4.44 0.59 — 0.59 3.85 5.03 2005 5.25 0.31 — 0.31 4.94 6.25 2006 8.10 0.33 — 0.33 7.77 9.53 2007 12.43 0.34 — 0.34 12.09 14.44 2008 21.98 0.36 — 0.36 21.62 25.32 2009 14.26 0.49 — 0.49 13.77 16.01 2010 78.97 1.27 — 1.27 77.70 89.24 2011 41.12 1.51 — 1.51 39.61 44.56 2012 112.50 2.49 0.55 3.05 109.46 120.82 2013 93.57 3.00 1.09 4.10 89.47 97.06 2014 104.78 3.43 1.08 4.51 100.26 106.74 2015 75.23 3.62 1.00 4.62 70.62 74.39 2016 26.58 3.75 1.91 5.67 20.91 21.79 2017 (2.82) 4.16 4.83 8.99 (11.81) (12.08) 2018 (0.46) 4.16 4.89 9.05 (9.51) (9.51) 2019 — 4.06 6.08 10.14 (10.14) (9.97) 2020 — 4.01 6.84 10.85 (10.85) (10.48) 2021 — 3.95 7.94 11.89 (11.89) (11.28) 2022 — 3.88 9.55 13.43 (13.43) (12.53) 2023 — 3.79 14.95 18.74 (18.74) (17.19) 2024 — 3.69 15.63 19.32 (19.32) (17.42) 2025 — 3.58 16.32 19.90 (19.90) (17.63) 2026 — 3.47 16.37 19.84 (19.84) (17.27) 2027 — 3.35 16.42 19.77 (19.77) (16.92) 2028 — 3.24 16.47 19.71 (19.71) (16.57) 2029 — 3.13 17.28 20.41 (20.41) (16.87) 2030 — 3.00 18.10 21.10 (21.10) (17.14) 2031 — 2.88 19.25 22.13 (22.13) (17.67) 2032 — 2.74 20.41 23.15 (23.15) (18.17) 2033 — 2.59 22.49 25.08 (25.08) (19.34) 2034 — 2.43 22.22 24.65 (24.65) (18.69) 2035 — 2.26 22.94 25.20 (25.20) (18.78) 2036 — 2.09 23.00 25.09 (25.09) (18.37) 2037 — 1.91 23.07 24.98 (24.98) (17.98) 2038 — 1.74 23.13 24.87 (24.87) (17.59) 2039 — 1.57 22.74 24.31 (24.31) (16.90) 2040 — 1.40 22.81 24.20 (24.20) (16.54) LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 109 Constant 2018 Current US$, millions US$, millions Years Loan Loan Repayments Net Loan Minus Net Loan Minus Withdrawals Commitment and Interest Principal Total Payment Payment Payment 2041 — 1.24 19.80 21.04 (21.04) (14.13) 2042 — 1.11 18.78 19.89 (19.89) (13.13) 2043 — 1.00 17.75 18.75 (18.75) (12.17) 2044 — 0.89 17.81 18.69 (18.69) (11.92) 2045 — 0.78 17.86 18.64 (18.64) (11.68) 2046 — 0.67 17.91 18.58 (18.58) (11.44) 2047 — 0.55 17.97 18.52 (18.52) (11.21) 2048 — 0.44 18.02 18.46 (18.46) (10.98) 2049 — 0.33 16.47 16.80 (16.80) (9.83) 2050 — 0.23 14.92 15.15 (15.15) (8.71) 2051 — 0.14 12.66 12.81 (12.81) (7.23) 2052 — 0.08 10.40 10.47 (10.47) (5.82) 2053 — 0.03 4.46 4.49 (4.49) (2.45) 2054 — 0.01 1.36 1.37 (1.37) (0.73) Table F.3: IDCOL ODA Loan Withdrawals and Repayments, 2003–2042 IDCOL Loan Repayments to Government (No Interest Payments after July 2018) Current US$, millions Constant 2018 US$, millions Years IDCOL Loan IDCOL Loan Net Loan Receipts Net Loan Receipts NPV in 2018 NPV in 2018 Withdrawals from GOB Repayments to GOB Minus Repayments Minus Repayments at 10% at 5% 2003 1.67 (0.01) 1.67 2.24 9.34 4.65 2004 4.53 (0.07) 4.45 5.81 22.06 11.50 2005 4.94 (0.10) 4.85 6.12 21.14 11.55 2006 7.97 (0.14) 7.83 9.59 30.11 17.23 2007 12.56 (0.20) 12.36 14.76 42.10 25.24 2008 21.12 (0.31) 20.81 24.38 63.23 39.71 2009 14.34 (0.55) 13.79 16.03 37.80 24.87 2010 78.97 (2.00) 76.96 88.39 189.48 130.59 2011 41.12 (4.25) 36.86 41.47 80.81 58.35 2012 112.50 (7.07) 105.44 116.38 206.18 155.96 2013 93.57 (11.02) 82.55 89.55 144.22 114.29 2014 104.78 (16.30) 88.47 94.19 137.90 114.49 2015 75.23 (22.79) 52.44 55.24 73.52 63.94 2016 26.58 (24.50) 2.08 2.16 2.62 2.38 2017 1.83 (34.72) (32.89) (33.63) (36.99) (35.31) 2018 (0.46) (33.25) (33.71) (33.71) (33.71) (33.71) 2019 — (30.72) (30.72) (30.20) (27.45) (28.76) 2020 — (32.40) (32.40) (31.30) (25.87) (28.39) 110 IDCOL Loan Repayments to Government (No Interest Payments after July 2018) Current US$, millions Constant 2018 US$, millions Years IDCOL Loan IDCOL Loan Net Loan Receipts Net Loan Receipts NPV in 2018 NPV in 2018 Withdrawals from GOB Repayments to GOB Minus Repayments Minus Repayments at 10% at 5% 2021 — (32.24) (32.24) (30.61) (23.00) (26.44) 2022 — (32.63) (32.63) (30.45) (20.80) (25.05) 2023 — (33.06) (33.06) (30.32) (18.83) (23.76) 2024 — (33.02) (33.02) (29.76) (16.80) (22.21) 2025 — (33.02) (33.02) (29.25) (15.01) (20.79) 2026 — (33.06) (33.06) (28.79) (13.43) (19.48) 2027 — (33.16) (33.16) (28.37) (12.03) (18.29) 2028 — (33.30) (33.30) (28.01) (10.80) (17.19) 2029 — (24.81) (24.81) (20.51) (7.19) (11.99) 2030 — (20.42) (20.42) (16.59) (5.29) (9.24) 2031 — (20.12) (20.12) (16.06) (4.65) (8.52) 2032 — (13.84) (13.84) (10.86) (2.86) (5.49) 2033 — (7.28) (7.28) (5.61) (1.34) (2.70) 2034 — (3.89) (3.89) (2.95) (0.64) (1.35) 2035 — (2.42) (2.42) (1.80) (0.36) (0.79) 2036 — (2.42) (2.42) (1.77) (0.32) (0.74) 2037 — (2.14) (2.14) (1.54) (0.25) (0.61) 2038 — (2.14) (2.14) (1.51) (0.23) (0.57) 2039 — (2.14) (2.14) (1.49) (0.20) (0.54) 2040 — (2.15) (2.15) (1.47) (0.18) (0.50) 2041 — (2.15) (2.15) (1.45) (0.16) (0.47) 2042 — (1.08) (1.08) (0.71) (0.07) (0.22) Total 601.24 (581.23) 20.01 124.22 782.90 433.96 Table F.4: Comparison of Treasury versus IDCOL Stakeholder Impacts of Financial Structuring of ODA Pass-Through Funding of the SHS Program, 2003–2054 Net Present Net Present ODA→GOB Loan GOB→IDCOL Loan GOB Net on ODA GOB Net on ODA Years Value in 2018 at Value in 2018 at Minus PMT Minus PMT Pass-Through Pass-Through 10% 5% Current 2018 US$, millions Constant 2018 US$ 2003 1.63 1.67 (0.03) (0.04) (0.17) (0.08) 2004 3.85 4.45 (0.60) (0.68) (2.59) (1.35) 2005 4.94 4.85 0.10 0.11 0.36 0.20 2006 7.77 7.83 (0.05) (0.05) (0.17) (0.10) 2007 12.09 12.36 (0.26) (0.28) (0.78) (0.47) 2008 21.62 20.81 0.81 0.82 2.14 1.34 2009 13.77 13.79 (0.02) (0.02) (0.05) (0.03) 2010 77.70 76.96 0.74 0.74 1.58 1.09 2011 39.61 36.86 2.74 2.69 5.24 3.78 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY | 111 Net Present Net Present ODA→GOB Loan GOB→IDCOL Loan GOB Net on ODA GOB Net on ODA Years Value in 2018 at Value in 2018 at Minus PMT Minus PMT Pass-Through Pass-Through 10% 5% Current 2018 US$, millions Constant 2018 US$ 2012 109.46 105.44 4.02 3.86 6.84 5.18 2013 89.47 82.55 6.92 6.54 10.53 8.35 2014 100.26 88.47 11.79 10.93 16.00 13.28 2015 70.62 52.44 18.18 16.67 22.19 19.30 2016 20.91 2.08 18.84 17.09 20.68 18.84 2017 (11.81) (32.89) 21.08 18.77 20.65 19.71 2018 (9.51) (33.71) 24.20 21.07 21.07 21.07 2019 (10.14) (30.72) 20.58 17.61 16.01 16.77 2020 (10.85) (32.40) 21.55 18.13 14.98 16.44 2021 (11.89) (32.24) 20.36 16.83 12.64 14.54 2022 (13.43) (32.63) 19.20 15.60 10.66 12.84 2023 (18.74) (33.06) 14.32 11.44 7.10 8.96 2024 (19.32) (33.02) 13.69 10.75 6.07 8.02 2025 (19.90) (33.02) 13.12 10.12 5.19 7.19 2026 (19.84) (33.06) 13.23 10.03 4.68 6.79 2027 (19.77) (33.16) 13.38 9.97 4.23 6.43 2028 (19.71) (33.30) 13.59 9.95 3.84 6.11 2029 (20.41) (24.81) 4.41 3.17 1.11 1.85 2030 (21.10) (20.42) (0.68) (0.48) (0.15) (0.27) 2031 (22.13) (20.12) (2.01) (1.40) (0.41) (0.74) 2032 (23.15) (13.84) (9.31) (6.36) (1.67) (3.21) 2033 (25.08) (7.28) (17.80) (11.96) (2.86) (5.75) 2034 (24.65) (3.89) (20.77) (13.71) (2.98) (6.28) 2035 (25.20) (2.42) (22.78) (14.78) (2.92) (6.45) 2036 (25.09) (2.42) (22.67) (14.46) (2.60) (6.01) 2037 (24.98) (2.14) (22.84) (14.32) (2.34) (5.67) 2038 (24.87) (2.14) (22.73) (14.00) (2.08) (5.28) 2039 (24.31) (2.14) (22.17) (13.42) (1.81) (4.82) 2040 (24.20) (2.15) (22.05) (13.12) (1.61) (4.49) 2041 (21.04) (2.15) (18.88) (11.04) (1.23) (3.60) 2042 (19.89) (1.08) (18.81) (10.81) (1.10) (3.35) 2043 (18.75) — (18.75) (10.59) (0.98) (3.13) 2044 (18.69) — (18.69) (10.38) (0.87) (2.92) 2045 (18.64) — (18.64) (10.17) (0.78) (2.72) 2046 (18.58) — (18.58) (9.96) (0.69) (2.54) 2047 (18.52) — (18.52) (9.76) (0.62) (2.37) 2048 (18.46) — (18.46) (9.56) (0.55) (2.21) 2049 (16.80) — (16.80) (8.56) (0.45) (1.89) 2050 (15.15) — (15.15) (7.58) (0.36) (1.59) 2051 (12.81) — (12.81) (6.30) (0.27) (1.26) 2052 (10.47) — (10.47) (5.06) (0.20) (0.96) 2053 (4.49) — (4.49) (2.13) (0.08) (0.39) 2054 (1.37) — (1.37) (0.64) (0.02) (0.11) Total (110.04) 10.34 (120.38) 1.25 180.39 138.05 112 Table F.5: Net Impact on GOB of SHS Program in Present Value Terms 2003–2054 Net Present Value in 2018 (10% Discount) Cumulative Net Present Value in 2018 Total GOB Kerosene Taxes on ODA Pass Treasury Net Total GOB Total GOB Year Subsidy Kerosene Total GOB Impact of SHS Through from SHS GOB Taxes ODA Pass Treasury Net GOB Taxes Treasury Net Saved Subsidy Subsidy ODA Pass Program on SHS Through from SHS on SHS from SHS Saved Saved Through Program Program Constant 2018 US$, millions Col. 1 2 3 4−1+2=3 5 6 7 8=5+6+7 9 10 11 12 = 9 + 10 + 11 2003 0.84 0.03 (0.04) 0.83 3.53 0.13 (0.17) 3.49 3.53 0.13 (0.17) 3.49 2004 1.68 0.37 (0.68) 1.37 6.39 1.42 (2.59) 5.22 9.91 1.55 (2.76) 8.70 2005 2.32 0.34 0.11 2.77 8.02 1.18 0.36 9.56 17.93 2.73 (2.39) 18.27 2006 3.13 1.74 (0.05) 4.82 9.83 5.45 (0.17) 15.11 27.76 8.18 (2.56) 33.38 2007 5.62 1.83 (0.28) 7.17 16.02 5.23 (0.78) 20.47 43.78 13.41 (3.35) 53.85 2008 9.38 2.44 0.82 12.64 24.32 6.33 2.14 32.79 68.10 19.75 (1.21) 86.64 2009 12.30 (0.40) (0.02) 11.88 29.01 (0.94) (0.05) 28.02 97.11 18.80 (1.26) 114.66 2010 21.80 3.29 0.74 25.83 46.73 7.06 1.58 55.37 143.84 25.87 0.32 170.03 2011 26.85 35.81 2.69 65.34 52.32 69.78 5.24 127.33 196.16 95.65 5.55 297.36 2012 34.22 87.69 3.86 125.77 60.63 155.35 6.84 222.82 256.78 250.99 12.40 520.17 2013 40.67 44.55 6.54 91.76 65.50 71.75 10.53 147.78 322.28 322.75 22.93 667.96 2014 23.31 52.45 10.93 86.68 34.12 76.79 16.00 126.91 356.41 399.53 38.93 794.87 2015 16.00 (96.75) 16.67 (64.07) 21.30 (128.77) 22.19 (85.28) 377.71 270.76 61.12 709.59 2016 4.30 (130.37) 17.09 (108.98) 5.20 (157.75) 20.68 (131.87) 382.91 113.01 81.80 577.72 2017 0.79 (74.32) 18.77 (54.77) 0.87 (81.76) 20.65 (60.24) 383.78 31.26 102.45 517.48 2018 0.09 17.59 21.07 38.75 0.09 17.59 21.07 38.75 383.87 48.85 123.52 556.24 2019 — 17.25 17.61 34.87 — 15.69 16.01 31.70 383.87 64.53 139.53 587.93 2020 — 16.72 18.13 34.85 — 13.82 14.98 28.80 383.87 78.35 154.51 616.73 2021 — 15.94 16.83 32.77 — 11.98 12.64 24.62 383.87 90.33 167.15 641.35 2022 — — 15.60 15.60 — — 10.66 10.66 383.87 90.33 177.81 652.01 2023 — — 11.44 11.44 — — 7.10 7.10 383.87 90.33 184.91 659.11 2024 — — 10.75 10.75 — — 6.07 6.07 383.87 90.33 190.98 665.17 2025 — — 10.12 10.12 — — 5.19 5.19 383.87 90.33 196.17 670.37 2026 — — 10.03 10.03 — — 4.68 4.68 383.87 90.33 200.85 675.04 2027 — — 9.97 9.97 — — 4.23 4.23 383.87 90.33 205.08 679.27 LIVING IN THE LIGHT: THE BANGLADESH SOLAR HOME SYSTEMS STORY 2028 — — 9.95 9.95 — — 3.84 3.84 383.87 90.33 208.91 683.11 | 2029 — — 3.17 3.17 — — 1.11 1.11 383.87 90.33 210.03 684.22 113 114 Net Present Value in 2018 (10% Discount) Cumulative Net Present Value in 2018 Total GOB Kerosene Taxes on ODA Pass Treasury Net Total GOB Total GOB Year Subsidy Kerosene Total GOB Impact of SHS Through from SHS GOB Taxes ODA Pass Treasury Net GOB Taxes Treasury Net Saved Subsidy Subsidy ODA Pass Program on SHS Through from SHS on SHS from SHS Saved Saved Through Program Program Constant 2018 US$, millions 2030 — — (0.48) (0.48) — — (0.15) (0.15) 383.87 90.33 209.87 684.07 2031 — — (1.40) (1.40) — — (0.41) (0.41) 383.87 90.33 209.47 683.66 2032 — — (6.36) (6.36) — — (1.67) (1.67) 383.87 90.33 207.79 681.99 2033 — — (11.96) (11.96) — — (2.86) (2.86) 383.87 90.33 204.93 679.13 2034 — — (13.71) (13.71) — — (2.98) (2.98) 383.87 90.33 201.95 676.14 2035 — — (14.78) (14.78) — — (2.92) (2.92) 383.87 90.33 199.02 673.22 2036 — — (14.46) (14.46) — — (2.60) (2.60) 383.87 90.33 196.42 670.62 2037 — — (14.32) (14.32) — — (2.34) (2.34) 383.87 90.33 194.08 668.28 2038 — — (14.00) (14.00) — — (2.08) (2.08) 383.87 90.33 192.00 666.20 2039 — — (13.42) (13.42) — — (1.81) (1.81) 383.87 90.33 190.19 664.38 2040 — — (13.12) (13.12) — — (1.61) (1.61) 383.87 90.33 188.57 662.77 2041 — — (11.04) (11.04) — — (1.23) (1.23) 383.87 90.33 187.34 661.54 2042 — — (10.81) (10.81) — — (1.10) (1.10) 383.87 90.33 186.24 660.44 2043 — — (10.59) (10.59) — — (0.98) (0.98) 383.87 90.33 185.27 659.46 2044 — — (10.38) (10.38) — — (0.87) (0.87) 383.87 90.33 184.39 658.59 2045 — — (10.17) (10.17) — — (0.78) (0.78) 383.87 90.33 183.62 657.82 2046 — — (9.96) (9.96) — — (0.69) (0.69) 383.87 90.33 182.93 657.13 2047 — — (9.76) (9.76) — — (0.62) (0.62) 383.87 90.33 182.31 656.51 2048 — — (9.56) (9.56) — — (0.55) (0.55) 383.87 90.33 181.76 655.96 2049 — — (8.56) (8.56) — — (0.45) (0.45) 383.87 90.33 181.32 655.52 2050 — — (7.58) (7.58) — — (0.36) (0.36) 383.87 90.33 180.96 655.16 2051 — — (6.30) (6.30) — — (0.27) (0.27) 383.87 90.33 180.69 654.89 2052 — — (5.06) (5.06) — — (0.20) (0.20) 383.87 90.33 180.49 654.69 2053 — — (2.13) (2.13) — — (0.08) (0.08) 383.87 90.33 180.41 654.61 2054 — — (0.64) (0.64) (0.02) (0.02) 383.87 90.33 180.39 654.59 Totals 203.30 (3.79) 1.25 200.77 383.87 90.33 180.39 654.59 5% Discount 279.32 33.63 138.05 450.99 10% Discount 383.87 90.33 180.39 654.59 The authors tell a remarkable story: • A remarkable story where over 15 years, 20 million This book is a must read for any rural people in Bangladesh obtained access to government, business or NGO that modern electric lighting and basic electricity wants to develop a substantial solar services using solar home systems, far sooner than they could have, had they waited for the electric energy market in rural areas. It brilliantly grid to arrive. These families, with 10 million describes the business model and roles children, enjoyed far better quality of lighting, a and responsibilities of the different cleaner and safer home environment, and access to the wider world through communications players, it provides a superb analysis technology. Eventually, their children will gain of the financing model and the social, upward mobility through improved education and economic, and environmental benefits, health due to electricity services from SHS. and describes the implications of rapid • A remarkable story of how a government, partnering with the Infrastructure Development grid expansion on the project. Company Ltd, delivered over US$600 million of Prof. Sir Robert Watson CMG FRS financing to rural families in small increments Former Chairman, Intergovernmental Panel on Climate of about US$100 per transaction. The story Change, United Nations Framework Convention on continues— these experiences are leading IDCOL Climate Change to finance investments in larger-scale roof-top and ground-mounted solar projects and solar irrigation pumping. • A remarkable story of a partnership with Bangladesh microfinance institutions, nongovernment organizations, and private Since 1954, the International companies to deliver solar home systems and services and provide access to finance to make SHS Solar Energy Society and its members affordable to the rural people. At its peak, more have undertaken technical research, than 29,000 people were employed in this new product development and advocacy for industry. the growth of solar and renewable energy • A remarkable story where the World Bank with other development partners, local investors, technologies. With the boom in the grid and households invested US$1,095 million, of connect market, solar home systems for which US$81 million was grants, and brought the unelectrified is an application which international best practice to this new industry. often gets overlooked. This book not • A remarkable story where Bangladesh avoided burning 4 billion liters of kerosene that was only shows how Bangladesh successfully previously used for home lighting and avoided over implemented a program providing solar 9 million tCO2 emissions. Rural people also avoided home systems to millions of people via the risk of home fires as well as respiratory illnesses from breathing kerosene smoke. a micro-credit facility the end user could • A remarkable story where costs of solar home afford, but is also an inspiration for others systems dramatically dropped, and quality of to learn from as different countries and products increased enormously over these 15 years organisations work towards meeting and these benefits were transferred to the rural people. SDG7 to provide renewable electricity This book documents the achievements, the to communities and accelerate the approaches, the successes, the challenges, and the transformation with solar energy for lessons. The off-grid solar technology and business everyone used wisely and efficient. have advanced greatly in these past 15 years. But the principles, lessons, and insights gained from what Prof. Dr. Klaus Vajen, President was then the world’s largest off-grid electrification International Solar Energy Society program will endure. In this decade of Sustainable Development for All, as the global community accelerates its efforts to achieve universal access to electricity by 2030 while reducing carbon emissions, the findings of this book will be invaluable. The insights and lessons learned in this fact-filled and deeply analytical book will be a useful reference for other countries and organizations that intend to embark on a similar journey. The main beneficiaries will be the 800 million people who are yet to experience the benefits of electricity.