Selling Solar: Lessons from More than a Decade of IFC's ... - Sunlabob

Selling SolarLESSONS FROM MORE THAN ADECADE OF IFC’S EXPERIENCE

© 2007 international finance corporationall rights reserved.International Finance Corporation2121 Pennsylvania Avenue, NWWashington, DC 20433 USATelephone: 202-473-3800Internet: www.ifc.org/EnviroPublicationsEmail: enviro@ifc.orgThe findings, interpretations, views, and conclusions expressedherein are those of the authors and do not necessarilyreflect the views of the Executive Directors of theInternational Finance Corporation or of the InternationalBank for Reconstruction and Development (theWorld Bank) or the governments they represent.rights and permissionsifc encourages use and distribution of its publications.Content from this document may be used freely andcopied accurately into other formats without prior permission,provided that clear attribution is given to theauthors and the original source and that content is notused for commercial purposes.printed in the united states of americaFirst printing, July 2007This publication is printed with soy-based inks onChorusArt, a process chlorine-free paper containing50% recycled and 15% post-consumer waste fiber.Printed by Balmar, Washington, DC (FSC certified)Designed by Marc Alain MeadowsMeadows Design O≈ce, Inc., Washington, DCwww.mdomedia.comacronyms and abbreviations used in this reportcepalco Cagayan Electric Power and Light Companyde distributed generationdgis Directorate General for International Cooperationebfp Environmental Business Finance Programee energy e≈ciencyeeaf Environmental Enterprises Assistance Fundeets Environ Energy-Tech Service, Ltd.eia Energy Information Administrationemt External Management Teamepa U.S. Environmental Protection Agencyfi financial intermediarygef Global Environment Facilitygtc GT Consulting, Inc.iea International Energy Agencyifc International Finance Corporationimf International Monetary Fundirc Investment Review Committeekrea Kenya Renewable Energy Associationkw kilowattkwh kilowatt/hourmdg Millennium Development Goalsmw megawattmwh megawatt/hourmbil Moser Baer India Ltd.ngo nongovernmental organizationoecd Organization for Economic Co-operation and Developmentone O≈ce National de l’Electricitepadgo Portfolio Approach to Distributed Generation Opportunitypv photovoltaicpvmti Photovoltaic Market Transformation Initiativere renewable energyreef Renewable Energy and Energy E≈ciency Fundsdc Solar Development Capitalsdf Solar Development Foundationsdg Solar Development Groupsef Sustainable Energy Facilityshs solar home systemsme small and medium enterprisespm Sunlight Power Maroc S.A.sri socially responsible investmentteri Tata Energy Research Institutetred Triodos Renewable Energy for Developmentseco Swiss State Secretariat for Economic Affairsvbard Vietnam Bank for Agriculture and Rural Developmentvwu Vietnam Women’s Unionwbg World Bank Groupwho World Health Organizationwp peak watt

contentsA Joint Message 4Executive Summary 6part 1: report overviewIntroduction: The Development Case for Rural Electrification 8Chapter 1: ifc’s Approach in the Early 1990s 10Chapter 2: ifc’s Experience in the Solar pv Market 12Chapter 3: ifc’s Lessons of Experience 16∫ Lessons about the solar pv market in general 16∫ Lessons about ifc’s financing programs 19∫ Lessons about what makes a successful solar pv company 22Chapter 4: ifc’s Approach Today 26part 2: selected case studies1. The ifc/gef Small and Medium Scale Enterprise Program (sme Program) 30∫ Grameen Shakti 32∫ Soluz Honduras S.A. de C.V. 35∫ Selco Vietnam, Ltd. 382. Photovoltaic Market Transformation Initiative (pvmti) 40∫ Muramati District Tea Growers Savings and Credit Cooperative, Ltd. 45∫ srei Infrastructure Finance, Ltd. 47∫ Sunlight Power Maroc S.A. 483. Solar Development Group 494. Cagayan Electric Power & Light Company 54References 58Acknowledgements 60

a joint messageThe Global Environment Facility (gef) is the largest funder of renewable energy in the developing world,supporting solar, wind, and other clean forms of energy. In 14 years, gef’s clean energy portfolio has grownto more than $1.9 billion in grants for projects in more than 60 developing countries, with a total value ofnearly $12 billion. These projects hold the promise of reduced greenhouse gas emissions, while also alleviatingpoverty through the provision of modern energy services using locally available resources.gef’s approach to renewable energy seeks to create conditions for growing commercially sustainablemarkets, catalyzing private investment and supporting government commitments to programs and policies,and widespread consumer acceptance. In promoting these aims, the private sector focus of the InternationalFinance Corporation (ifc) is particularly important and valuable. ifc brings an institutional commitmentand proven track record in promoting sustainable private sector investment in developingcountries. As gef seeks to define more e∑ective means of engaging the private sector as a partner in promotingsolutions to global environmental problems —a current priority on our agenda—ifc’s experienceand capacity will be increasingly important.As this study illustrates, despite the environmental appeal of solar pv and other renewable energy technologies,finding practical business solutions to their introduction has not been easy. gef has a responsibilitynot only to report these results, but also to avoid repeating mistakes and incorporate the lessons of experiencegoing forward. I am, therefore, very pleased by the effort ifc has made to review the outcome ofits projects and make them available for others working in this field of increasing worldwide interest. Thediscussion of how these lessons have influenced more recently established projects and mainstream ifc investmentsis an encouraging example of gef’s influence on the larger investment community. I congratulateifc for this study and trust it will be widely disseminated.Monique BarbutChief Executive Officer and ChairpersonGlobal Environment Facility4

Lack of access to electricity remains one of the key challenges in the developingworld. Renewable energy solutions including solar pv can make a significantcontribution to addressing the issue of rural electrification. However, finding away to employ these solutions on a commercial and sustainable basis in emergingmarkets has yet to be discovered.For almost 20 years, ifc has explored options for the commercialization ofsolar pv in the developing world. In the early to mid-1990s, ifc initiated severalsolar pv activities in partnership with other investors. Although these e∑ortscould not overcome all the complexities of the solar pv market, and have notalways lived up to their original expectations, they have provided valuablelessons of experience that are documented in this study.This report is in two parts. Part 1 describes the history of ifc’s approach tosolar pv and explains why ifc today has developed a di∑erent approach to addressingrural electrification. It will be of interest to those now in the field ascontext and background.Part 2 provides case studies on ifc’s solar pv financing initiatives, as well asexamples of some of the projects that these initiatives supported. These, wehope, will be useful to those in the business of exploiting the approaches to the nascent pv market inemerging markets.In light of these lessons of experience, ifc is moving toward a broad approach to market-based solutionsto rural electrification that supports a variety of technologies, including the commercialization of lowpowerlighting devices, and distributed power generation.ifc’s e∑orts in this field would not have been possible without donor support. The role of the GlobalEnvironment Facility (gef), the primary source of funding for renewable energy projects in developingcountries, was, and remains, invaluable in providing the resources required to cover the higher risks andnoncommercial costs of new business models and financing programs.Developing and sharing lessons of experience are a pivotal part of ifc’s strategy, and I hope this studywill be a useful tool to those who are currently operating in, or planning on entering, the solar pv marketin emerging markets. I also hope that this publication will contribute to the growing body of knowledgeregarding sustainable energy solutions for rural electrification.Rachel KyteDirectorEnvironment & Social Development Department5

executive summary1A set of eight internationaldevelopment goals for 2015,adopted by the international communityin the UN MillenniumDeclaration in September 2000,and endorsed by the InternationalMonetary Fund (imf), the WorldBank, and Organization forEconomic Co-operation andDevelopment (oecd).2This review does not discuss theexperience of the RenewableEnergy and Energy E≈ciencyFund (reef), due to an agreementamong participating investorsrestricting disclosure.Rural electrification has been a long-time focus for the World Bank Group (wbg) overall, and ifc, theWorld Bank’s private sector investment arm. The contribution of electrification to quality of life, throughimproved livelihoods and health, and increased education and productivity, is well documented. With approximatelyone-third of the world’s population living without access to electricity—mostly in rural areasof developing countries—rural electrification is a key means of achieving the Millennium DevelopmentGoals (mdgs) 1 and reducing poverty.Concern about the global environment in recent years has led to an increased focus on renewable energy(re) technologies. The Global Environment Facility (gef) was initiated as a pilot project in 1991 and thenformally established in its current form in 1994 with a mandate to address climate change and develop astrategy to support re investments. The World Bank Group became interested in exploring the solar photovoltaic(pv) market as a means to providing clean energy service in rural areas that had no access to thegrid. ifc recognized an opportunity to use gef funds to test various options for the commercialization ofsolar pv in emerging markets.This publication documents ifc’s solar pv experience. In total, ifc managed five gef-funded solar pv initiatives,of which four are discussed in this publication: the ifc/gef Small and Medium Scale EnterpriseProgram (sme Program), the Photovoltaic Market Transformation Initiative (pvmti), the Solar DevelopmentGroup (sdg), and the grid-tied solar power plant of the Cagayan Electric Power and Light Company(cepalco). 2 While ifc programs have been responsible for the installation of over 84,000 solar home systems(shs), these programs have been less successful from a financial standpoint, ifc having been unable tosignificantly transform markets and create sustainable businesses as originally anticipated.In some of the initiatives that were implemented, we have found the main challenge to lie, not in thetechnology of solar pv, but in accurately judging market reality and trends. ifc’s solar pv portfolio has beensignificantly impacted by well-documented market trends. These included a failed prediction that the priceof solar pv panels would come down, the decrease in supply of smaller modules, and a number of economicshocks. Hindsight shows that the initial beliefs of ifc and many market players about the solar pvmarket were overly optimistic. Through the implementation of its solar pv initiatives, ifc has learned agreat deal, not only about the solar pv market in general, but also about the type of financing required tosupport solar pv market growth and what it takes to develop a successful solar pv company. Perhaps one ofthe most important lessons that ifc has learned is that supporting the growth of the solar pv market is farmore complex than first envisioned, particularly due to the level of market segmentation that exists.ifc remains committed to addressing the issue of rural electrification in developing countries and is cautiouslyoptimistic that a self-sustaining solar pv market will develop there. However, in light of the lessons,we are moving away from focusing narrowly on solar pv as a way of addressing rural electrification. Instead,we are moving toward a broader approach, such as supporting a variety of technologies, the commercializationof low-power lighting devices, and distributed power generation.6 selling solar

key lessons from ifc’s experiences in solar pvı The issue of affordability cannot be addressed without segmenting the market.The rural unelectrified market in developing countries is large. To reach it, proper segmentation along incomelines, needs, and lifestyle are necessary. It was initially felt that if solar pv module prices could bebrought down to a certain level, or if business models could be structured to maintain low monthly payments,solar pv would become financially a∑ordable to the consumer and competitive with alternative energysources. Experience has shown that the definition of a∑ordability varies among market segments (relativeincome levels, market applications, etc.), and it remains a challenge for pv companies to identify theniche market segments where solar pv is the least-cost energy alternative for the consumer.ı While solar PV is a viable technical solution, it is not the only solution. Without some level ofsubsidization similar to that provided for grid-connected electricity, it often remains too expensive for theaverage rural consumer.Experience has also demonstrated that people are looking for a constant supply of electricity provided bygrid connection. It is important to note that, while solar pv is cheaper for governments than costly grid expansionin dispersed rural populations, grid connection has emerged as a key political tool in many developingcountries, and the grid has almost always been heavily subsidized. In addition, solar pv simply cannotprovide equivalent services to the grid, and it is also not the only technology available for addressingrural electrification demand. The high initial cost of acquiring a solar pv system makes solar pv considerablyless a∑ordable to the rural poor than alternatives, such as car batteries and kerosene.ı Private equity is not the most appropriate financial mechanism for financing solar PV activities indeveloping countries.An important lesson for ifc was that, while private equity and venture capital firms are heavily involved inthe manufacturing of solar pv for developed country markets, the risks and economics of solar pv in the developingworld mean that the returns that such o∑-grid investors typically look for are nonexistent.Profitable opportunities for solar pv utilization in the developing world lie further up the value chain, primarilyin the manufacture of solar pv modules for export to subsidized, developed world markets.ı Good government relations and support are a strong success factor.While there are examples of companies able to establish successful ventures without express governmentsupport, those companies fortunate enough to operate with a government concession for exclusive territorialrights to distribute solar pv systems (or with some form of subsidy or favorable pricing agreement)tended to be more successful than companies operating without explicit government support.executive summary 7

part 1Introduction: The DevelopmentCase for Rural Electrification3International Energy Agency(iea), World Energy Outlook 2004,oecd and iea, Paris.4Foley, 1995, p. 41.5Erickson, 1995, p. 1130.6Cabraal et al., 1996.7For more information on thewbg renewable energy strategy,visit www.worldbank.org/re.8ifc’s initiative with cepalco inthe Philippines was the only gridtiedinvestment made.More than 1.6 billion people, roughly one third ofthe world’s population, live without access to electricity.The vast majority of those without electricitylive in rural areas. The World Bank estimates that67 percent of the rural population in developingcountries is without electricity. As a result, low-incomehouseholds around the world spend billionsof dollars every year on expensive and environmentallydamaging energy sources such as charcoal, firewood,and disposable batteries, with an estimated$38 billion a year spent on kerosene alone. This isespecially true in Africa, where roughly two thirdsof households—more than 580 million people—dependon wood fuel for their daily cooking and heatingneeds. People in emerging markets who rely onthese fuels use much of their labor to gather woodfuel and are exposed daily to indoor air pollution.The World Health Organization (who) estimatesthat the pollution caused by using indoor biomasscook stoves is responsible for 1.6 million deaths peryear—mostly of young children and mothers.Access to modern, reliable energy is important forrural development and improved livelihoods. Energyis a major tool for poverty alleviation, incomegeneration, health, and other developmental agendas.The provision of clean electricity to low-incomehouseholds allows for increased opportunities forstudying in the evenings, as well as increased productivityof agricultural and micro-enterprise activities.Numerous studies have confirmed that thesocial benefits from electrification—the ability topower lights, radios, small appliances, and televisions—makea significant contribution to enhancingquality of life. This enhanced quality of life has,in turn, proved to be a key driver of increases in thedemand for energy.solar pv: an attractive technologyfor rural electrificationIne≈cient energy infrastructure, rapidly growing demand,and a general lack of generating capacitymean that many developing countries are unable tomeet their basic energy demands. The InternationalEnergy Agency (iea) estimates under its referencescenario that developing countries require a $300 billionannual investment for the electricity sectoralone. 3 In rural areas, where as many as four out offive people lack electricity, conventional gridconnectedelectricity schemes are often not feasible.Grid expansion can be extremely costly and has beendemonstrated numerous times to be far less coste∑ective than supplying shs. 4 Because of the highcost of extending electrical grid coverage in theseareas, non-grid-tied renewable energy technologies,such as solar pv, may be a least-cost solution.The availability of clean electricity not only helpshouseholds avoid the health risks associated withconventional forms of energy, such as kerosene,charcoal, and disposable batteries, but it also helpsthe global environment through the avoidance ofgreenhouse gas emissions and conventional air pollutionassociated with fossil-fuel-based forms of electricity.Renewable energy technologies also help governmentsgain energy independence and eliminatethe need for costly grid expansion to remote villages.Solar pv and shs are attractive renewable energytechnologies for many applications in o∑-grid areas.Most developing countries lie in areas with highsolar insolation levels—a “must” for solar pv—and,with the added benefit of shs being modular, shscan be installed to provide energy for one house,groups of houses, or an entire village. Other renew-8 selling solar: part 1

able energy resources (e.g., hydro or biomass) aretypically better suited for less dispersed populations,as they generally become economical only if they areable to provide energy to a more sizable population. 5Photovoltaic technologies already provide electricityin developing countries to an estimated500,000 to 1 million rural households lacking accessto electricity grids. shs is one of the most commonforms of solar pv application in rural areas. An shsusually provides electricity for two or three fluorescentlights; a radio, cassette player, or television; andother small appliances. Electricity is drawn fromrechargeable batteries charged through an electroniccharge controller by solar pv modules mounted on apole beside the house or on the rooftop. The totalcapacity of the unit is usually in the range of 30-100peak watts (wp), but can be less or greater. 6The direct economic benefits of shs include increasedconvenience and safety, improved indoor airquality, a higher quality of light than kerosene forreading, and the displacement of CO2 emissions.Improved lighting provides additional educationalbenefits, particularly for children, and can allow income-generatingactivities to occur beyond normalwork hours. Solar pv systems can power lights andvaccine refrigerators in medical clinics, run low-liftwater pumps, and allow for the operation of othervital systems.the world bank group’sinvolvement in solar pvWhen the Global Environment Facility (gef) was establishedin 1994, it made available a new source offunds to support projects that generated global environmentalbenefits. One of gef’s operational programssupports renewable energy activities that areunable to secure commercial financing elsewhere.The wbg was at the time particularly interested inutilizing gef funds to develop the renewable and energye≈ciency potential in emerging markets and togain experience in the solar pv market. 7ifc was particularly interested in exploring opportunitiesfor the commercialization of solar pv.With its mandate to further economic developmentthrough the private sector, ifc had been active in thesolar pv market since 1989, when it made a $3 millioninvestment (debt and equity) in Shenzhen ykSolar pv Energy Co., Ltd., a solar pv manufacturerin China. Although the investment, made using regularifc funds, did not meet its original expectations,it established an important precedent for investingin solar pv businesses in frontier markets.ifc has since used its skill and experience instructuring projects that target the private sector todevelop and implement a number of gef-fundedsolar pv projects, many of which operated acrosscountry lines. Today ifc, together with the WorldBank, is the largest financier of o∑-grid solar pv inthe developing world, having supported theinstallation of over 1.3 million solar pv systems. 8(See Table 1 below for details on the World BankGroup’s solar pvapplerelated projects.)TABLE 1: WBG SOLAR PV INITITIATIVESCOUNTRY NUMBER NUMBER OF SOLAR PV COST**OF PROJECTS SYSTEMS INSTALLED* CAPACITY (kWp) ($ MILLIONS)Argentina 1 30,000 2,843 36.0Bangladesh 1 198,000 9,900 91.4Bolivia 1 60,000 2,600 38.6Burkina Faso 1 8,000 300 3.0Cambodia 1 10,000 400 4.0Cape Verde 1 4,500 129 2.5China 1 400,000 10,000 144.9Ecuador 1 2,200 110 1.5Ethiopia 1 6,300 407 5.4India 1 45,000 2,500 24.0Indonesia 1 8,500 425 3.8Laos 1 4,000 160 1.3Madagascar 1 15,000 625 7.5Mali 1 10,000 420 5.0Mexico 2 37,000 704 12.9Mongolia 1 50,000 520 5.2Mozambique 2 9,800 1,096 13.5Nicaragua 1 6,000 215 3.0Pacific Islands 1 21,000 630 16.5Papua New Guinea 1 2,500 100 2.2Philippines 2 139,000 10,000 113.0Senegal 1 10,000 420 5.0Sri Lanka 2 104,400 4,176 36.1Swaziland 1 2,000 100 1.3Tanzania 1 40,000 2,500 30.0Uganda 1 90,000 6,300 67.7Multiple countries † 14 84,000 + 25.3Total 44 1,300,000 + ~58MW ~700.0Source: Anil Cabraal, 21st EU Solar PVSEC, 2006 (with update, January 2007).* Figures include both the number of systems installed and the target installation for projects currently underimplementation.** Cost includes only total investment of solar PV components/applications.†Includes projects of the SME Program in Bangladesh, Dominican Republic, Honduras, Tunisia, and Vietnam, and PVMTI inIndia, Kenya, and Morocco.introduction 9

chapter 1IFC’s Approach in theEarly 1990sthe emergence of solar pvIn the early 1990s, when ifc was in the process ofstructuring its first solar pv market initiative, therewas a great deal of anticipation about the future ofsolar pv in emerging markets. Overall industrygrowth had accelerated steadily since the early1980s. 9 In emerging markets, it was felt that therewas an opportunity for small-scale solar pv applications,such as shs, to replace diesel generation andto provide supplementary power to grid-connectedsystems. 10The first renewable energy initiatives that wereimplemented in emerging markets were largelydonor-led and focused on demonstrating the functionalityof the technology. Little to no attentionwas paid to commercializing the market to supportwider dissemination of the technology. Given thatthe majority of early initiatives were grant-basedprograms, many of them ended when funding wasexhausted, as the programs had not been designedwith financial sustainability or replication as keyconsiderations.By the mid-1990s, large capital flows were streaminginto solar pv businesses. Large players, such asBP, Shell, and Total Energie, were entering this market,manufacturing solar pv panels, and selling solarpv systems to rural electricity consumers in developingcountries. Anyone with the means and interestin electricity in emerging markets experimentedwith solar pv in what appeared to be a major emergingbusiness opportunity.sizable market and demandThe potential market was considered very attractive.An estimated 1.6 billion people were without electricity(roughly 400 million households), all ofwhich could be electrified with shs. The thoughtwas that if solar pv was made available to a community,demand would be similar to that experiencedwith the arrival of grid-connected electricity. Whilethe initial start-up cost of acquiring solar pv wasconsiderably more expensive than the alternatives,such as batteries or diesel generators, it was assumedthat a significant segment of the unelectrified populationwould opt for the superior quality of solar pv,could financially a∑ord it, and would purchase systemsif credit were available. The opportunity tobring about a substantial increase in the market sizewas present; in fact, some were predicting a similar10 selling solar: part 1

growth pattern to that experienced with personalcomputers and mobile phones.prices would fall and solar pvcould become more financiallyaffordableIt was widely perceived that the main barrier toscaling up the industry was rooted not in the technology,but in the financial a∑ordability of solar pv.Solar pv had proved to be una∑ordable for threekey reasons: (1) the overall price of solar pv moduleswas not considered competitive against alternativeelectricity sources, such as diesel generators;(2) there was no financing available to help solar pvconsumers with the large initial cost of acquiringand installing a solar pv system; and (3) subsidiesfor other forms of energy, such as grid-tied electricity,were distorting the market. The appropriatenessof solar pv technology itself was not called intoquestion, as it was largely believed to be the besttechnological solution to rural electrification, givenits scalability and fit for dispersed populations.At the time ifc was structuring the solar pv initiativesdiscussed in this report, it was widelythought that the prices of solar pv modules wouldcontinue to decline. The price of solar pv moduleshad decreased by a factor of over 50 since the early1970s, and it was expected that further price reductionswould continue to occur as a result of technicalprogress in materials, cell design, and manufacturingmethods, as well as economies of scale inmanufacturing. 11 It was expected that solar pv moduleprices would decline su≈ciently to allow solarpv to become a cost-e∑ective replacement for dieselfuel or kerosene. With cost expected to decrease, thelack of consumer access to financing was seen as themajor constraint. ifc considered that it could structuresolar pv initiatives to address this constraint.SIGNIFICANT SOLAR PV MARKET TRENDSDuring the period in which IFC’s solar PV projects were implemented, there were a number of well-documentedmarket trends that emerged in the global solar PV market that had a significant impact on solar PVmarkets in the developing world. These were the discontinuation of the expected downward trend in solarPV module prices, the increased demand for large solar PV systems in the industrialized world, and theglobal economic shocks that occurred in the late 1990s and early 2000s (the Asian and Russian financialcrises of 1997 and 1998, respectively, the Argentine economic crisis of 1999, and the 9/11 attacks). Whilethese market trends were not on their own responsible for the limited success of IFC’s portfolio, they didserve to further exacerbate existing obstacles.Prices did not decrease as expected and, in recent years, the exact opposite has occurred. According tothe United States Department of Energy’s Energy Information Administration (EIA), the average price forsilicon contracts increased by approximately 25 percent between 2004 and 2006. 12 As silicon is a key componentin the construction of solar PV panels, this has had a serious impact on the overall price of solar PVsystems. The main reasons for this increase were the continued tight supply of high-grade silicon, as wellas the increased demand for solar PV, fueled by subsidized programs in the industrialized world.It is currently estimated that as much as 50 percent of the cost of solar PV electricity is paid for throughtransitional subsidies. Most of this is for grid-connected systems, which currently represent well over threequarters 13 of the total solar PV market. In Germany, for example, the electric utilities are now paying customersa significant premium for any surplus solar PV power they sell back to the grid. This huge premiumhas resulted in a sizable increase in the global demand for solar PV systems.The increase in demand for solar PV in the industrialized world has affected solar PV markets in the developingworld, not only through increased prices, but also by shifting production away from the smallermodules. Load requirements in industrialized countries are significantly higher than those in developingcountries, and manufacturers have chosen to move away from the manufacture of smaller modules in favorof the increased profitability and steady cash flow associated with catering to the industrialized countrymarket. The lack of supply of smaller modules has led to increased working capital requirements forsmaller integrators, 14 as well as increased pressure on prices for smaller modules. In the period betweenmid-2005 and the end of 2006, the price of 40-watt panels has increased by 50 percent (36 percent for 20-watt panels). 159Jackson, 1999, p. 376.10Ahmed, 1994, p.7.11This review does not discuss theexperience of the RenewableEnergy and Energy E≈ciencyFund (reef), due to an agreementamong participating investorsrestricting disclosure.12http://www.solarbuzz.com.13Some estimates place grid-connectedsystems at over 90 percentof the total solar pv market.14Hande, 2006.15Hande, 2007.ifc’s approach in the early 1990s 11

chapter 2IFC’s Experience in theSolar PV Marketifc has learned many lessons from its experience inthe solar pv market. These lessons are summarizedin the following chapter. By far the most importantlesson that ifc has drawn from its experience is thatthere is not simply one target market for solar pv indeveloping countries, namely the entire unelectrifiedmarket, but many di∑erent target market segments.To be successful, solar pv ventures should be structuredwith a narrow, well-defined target market.The investment o∑ering can then adequately addressa relatively homogenous set of needs. This wasa relatively di≈cult lesson for ifc as, from the outset,the goal of many of our solar pv initiatives wassimply to provide services to the unelectrified. As aresult, the experiences outlined below led to varyingdegrees of success, both in their e∑ectiveness and intheir implementation.ifc approved its first gef-financed investment to aTABLE 2. IFC/GEF SOLAR PV INITIATIVESDATE TOTAL GEF SHAREOPERATIONS INVESTMENT OF TOTAL CURRENTPROGRAM BEGAN GOAL AMOUNT FUNDING STATUSSME Program 1995 Increase access to finance, $20 million* 100% The SME program wasbuild capacity, and increase ($2.7 million absorbed into thegrowth of markets for SMEs used for five Environmental Businessactive in the areas of climate solar PV- Finance program in 2004.change mitigation and related Some of the solar PVbiodiversity conservation. businesses) investments have beenclosed; others are ongoingand operating successfully.PVMTI 1998 Accelerate the sustainable $30 million 100% Ongoingcommercialization andfinancial viability of energyservices, based on solarPV electricity.SDG 2000 Deliver SHS to rural $41 million 25% Dissolved in 2004households in developing (SDF $12 million,countries.SDC $28.7 million)†CEPALCO 2002 Demonstrate solar PV $5.775 million 70% Operating successfullyeffectiveness in supplying(CEPALCOenergy during peak usageprovidedperiods, thus avoiding new$1.775 millionplant construction.in financing)* The SME Program received a total of $20 million in funding from the GEF in two stages; a portion of this funding was earmarked to finance solar PV-related projects.† For details on additional shareholders, see the SDG case study on page 49.12 selling solar: part 1

solar pv company in 1998 through the ifc/gef Smalland Medium Scale Enterprise (sme) Program. Thatsame year, ifc’s first solar pv-focused financing facility,the ifc/gef Photovoltaic Market TransformationInitiative (pvmti), also became operational. Over thenext five years, ifc supported three additional solarpv-related initiatives, all financed by gef, includingthe Solar Development Group (sdg), which includedthe Solar Development Foundation (sdf),Solar Development Capital (sdc), and the cepalcoSolar Photovoltaic Demonstration Project. 16 Table 2outlines ifc’s gef-funded solar pv initiatives.The programs and projects implemented by ifchave resulted in significant overall social and environmentalbenefits. Examined for nonfinancial returns,such as the number of households electrified,the displacement of indoor air pollution, genderempowerment, education, health, and increased income-generationopportunities for the end user, theifc solar pv portfolio has performed well, with over84,000 shs installed. However, from a financialstandpoint, performance has proved below expectations,as it has generally not met the initial projectionsof investee companies.the ifc/gef small and mediumscale enterprise programThe ifc/gef sme Program was established in 1995 asa $20 million initiative, financed entirely by gef. Itsought to increase access to finance, build capacity,and increase markets for smes active in the areas ofclimate change mitigation (energy e≈ciency and renewableenergy) and biodiversity conservationthrough the provision of concessional loan financing.The sme Program was the first gef-funded,nongrant, sme financing program targeting the privatesector. While the sme Program did notspecifically target solar pv companies, it became operationalat a time when there was considerable interestin the solar pv market. (See the ifc/gef smeProgram case studies, page 30.)Over its lifetime, the sme Program approved investmentsin five solar pv-related businesses (see Table 3for details). The sme Program’s experience in the solarpv sector was mixed, with one successful project(Grameen Shakti in Bangladesh) and other projectswith more limited success. In 2004, the operations ofthe sme Program were absorbed by the EnvironmentalBusiness Finance Program (ebfp), a $20 millionifc/gef initiative that was designed as a successor to,and based on, the experiences of the sme Program.TABLE 3. IFC/GEF SME PROGRAMS’S SOLAR PV PROJECTS PORTFOLIOCOMMITMENTCOMPANY COUNTRY (IN MILLIONS)Grameen Shakti Bangladesh $ 0.750Environmental Enterprise Assistance Fund (EEAF) Dominican Republic 0.075Soluz Dominicana*Soluz Honduras** Honduras 0.500E + Co Rex Investment † Tanzania 0.150Cogener Tunisia 0.500Selco Vietnam Vietnam 0.750* The SME Program lent to EEAF, which on-lent to Soluz Dominicana S.A., an SHS distributor.** Soluz Honduras S.A. de C.V. received a $400,000 loan, as well as a $100,000 equity investment.†The SME Program lent to E+Co fund, which on-lent to Rex Investment Ltd., a leading Tanzanian solar PV distributor.Perhaps one of the most important lessons ifclearned from the sme Program’s experience with thesolar pv market was that it was possible to o∑setsome of the risks associated with solar pv by investingin a number of di∑erent markets and sectors.Contrasting the experience of Grameen Shakti,which operated in densely populated Bangladesh,with those of Selco Vietnam and Soluz Hondurasshowed that economies of scale are harder to comeby in sparsely populated and remote rural areas.Economies of scale are vital to the success of solarpv companies, since they reduce the financial cost ofmonthly rental fee or payment collection, as well asan ongoing system service and maintenance. Withouta sizable service population, a private solar pvcompany simply cannot financially sustain the costof a service technician or collection agent, and ultimatelythis leads to collection issues and di≈cultiesin maintaining systems (as was the case with bothSelco Vietnam and Soluz Honduras).The sme Program experience highlights the importanceof local ownership and government support.Grameen Shakti’s ties to Grameen Bank and,through it, the local community proved to be invaluableand a major driver of Grameen Shakti’ssuccess. Soluz Honduras found that the lack of adefined exclusive government concession to defendgeographic service territories was problematic whenit was faced with unexpected grid expansions thateliminated large numbers of established customers.(See case studies on Grameen Shakti, Soluz Honduras,and Selco Vietnam in Part 2.)photovoltaic markettransformation initiative16This review does not discuss theexperience of the RenewableEnergy and Energy E≈ciencyFund (reef), due to an agreementamong participating investorsrestricting disclosure.The ifc/gef pvmti is a $30 million gef-funded initiativedesigned to accelerate the sustainable comifc’sexperience in the solar pv market 13

TABLE 4. PVMTI’S ACTIVE SOLAR PV PROJECT PORTFOLIOmercialization and financial viability of energy servicesbased on solar pv technology in India, Kenya,and Morocco. Launched in 1998, pvmti is expectedto continue to the end of its extended mandate inDecember 2009 and, by the beginning of 2007, hadcommitted over $18 million to 12 projects. (Seepvmti case study, page 40.)pvmti initially found it di≈cult to structure dealsas the extensive documentation required, small investmentsize, and long negotiation periods provedtoo burdensome for many small and thinly capitalizedsolar pv companies. It quickly became apparentthat companies would not be able to absorb and repaycommitted funds by the original end date ofDecember 2007. The extensive documentationprocess and small investment size also resulted inhigh operational costs for pvmti in relation to itsportfolio. In 2004, pvmti underwent a significantrestructuring that extended its operational mandateby two years and has resulted in an increase in theproportion of disbursements to commitments toroughly 80 percent.As pvmti is still an operational project, it is di≈cultto evaluate its overall performance (see Table 4for a summary of the pvmti portfolio). To date,pvmti financing has resulted in the installation ofover 60,000 shs units in previously unelectrifiedhouseholds. The Mid-Term Program Review, whichwas completed in July 2006, noted that pvmti willbe responsible for the displacement of an estimated109,466 tonnes of CO2 emissions over the lifetimeof the installed shss.The pvmti experience highlights the need forflexibility in program design. The initial $500,000minimum investment proved to be too large formost smes to absorb, and the extensive businessplans and other documentation proved too dauntingfor the small businesses active in the solar pvsector, particularly those in the lower density ruralCOMMITMENTCOMPANY COUNTRY (IN MILLIONS)Selco India India $ 1.10Eskom-Shell Solar Home Systems India 3.90Shri Shakti India 2.23SREI Infrastructure Finance, Ltd. India 3.50Barclays Bank, Kenya Kenya 2.00Equity Building Society (EBS) Kenya 2.10Muramati District Tea Growers SACCO Kenya 0.60Salafin S.A. Morocco 1.00Sunlight Power Maroc S.A. Morocco 1.075areas where solar pv was most needed. pvmti was,indeed, flexible, and ifc was able to restructure theprogram to better suit the needs of the market.pvmti’s experience in Kenya highlighted the needfor technical assistance funding. While the Kenyansolar pv market was well-established, with manyplayers and a true entrepreneurial culture, there wasno real structure to the market and no standards. Itquickly became apparent that there was a need forfunding to help strengthen the overall marketthrough the creation of performance standards andby securing government support. (See case studieson srei, Muramati Tea Growers sacco and SunlightPower Maroc S.A. in Part 2.)solar development groupThe Solar Development Group (sdg) was a $41 millioninitiative which became operational in 2000.The goal was to deliver shs to rural households indeveloping countries. sdg was comprised of two separateentities: (1) Solar Development Capital (sdc), a$28.7 million for-profit private equity fund that providedgrowth capital for private solar pv and solarpv-related businesses; and (2) Solar DevelopmentFoundation (sdf), a $12 million nonprofit entity thatprovided business development assistance and seedfinancing to support the establishment of new solarpv businesses. (See sdg case study, page 49.)sdc experienced problems very early in implementation.There simply were no viable opportunitiesin the solar pv market that would provide thereturns that private equity investors were seeking.Despite revisions of return expectations, sdc managedto approve only six investments, totaling $3.9million (of the approved projects, only $650,000was disbursed to three investments), before beingliquidated in 2004. In contrast, sdf, with its focuson the provision of early-stage working capitalloans, guarantees, and technical assistance grants,was largely able to meet its targets, making commitmentstotaling over $3.5 million to 54 companies in23 countries. With sdc’s liquidation, however, sdftransferred its operations to the Triodos RenewableEnergy for Development (tred) Fund and sdgceased to exist in April 2004.ifc learned a great deal from the sdg experiencerelating to the type of financing programs requiredfor solar pv. A key lesson was that a diverse shareholdergroup can be problematic. With over 15di∑erent investors from a wide range of institutions(nongovernmental, bilateral, and multilateral financ-14 selling solar: part 1

ing organizations, socially responsible investmentfunds, and private companies), as well as private individuals,it was very di≈cult to satisfy the shareholders’di∑erent objectives and expectations. Whenit became obvious that sdc required restructuring,reaching consensus on a new structure proved impossibleand sdc was eventually disbanded.The sdg experience also highlighted the necessityof focusing mainly on market development and capacitybuilding. Overall, sdg had only focused ondeveloping individual businesses, rather than on themarket as a whole.The sdg experience, more than any other ifc solarpv initiative, demonstrates ifc’s optimistic outlookon the market. During the initial planningstages for sdg, over 100 investment opportunitieswere identified for sdc. Ultimately, none of the opportunitiesidentified in the feasibility study receivedsdc support, as the market was not preparedfor equity-type investments, companies were not ina position to absorb so much capital, and the returnexpectations by most of the candidate businesseswere not met.cagayan electric power and lightcompanycepalco is a private electricity distribution companyon the island of Mindanao in the Philippines.In December 2002, cepalco received $4 million ingef funding (loan convertible to grant) from ifc tobuild a 1 mw distributed generation power plant,which was integrated into the 80 mw distributionnetwork of cepalco and operated in conjunctionwith an existing 7 mw run-of-the-river hydroelectricplant. The purpose of the project was to demonstratethe e∑ectiveness of solar pv in addressing distributionsystem capacity issues, thereby delayingthe need to construct a new hydroelectric plant.The solar pv plant operated through a conjunctiveuse application, whereby hydropower and solar pvresources were used jointly, for the first time, to increasethe capacity of the hydropower unit and convertthe solar pv plant’s power output to firm dispatchablepower, rather than an intermittentresource. 17 (See cepalco case study, page 54.)Fully operational since 2004, the cepalco planthas operated successfully and without incident sinceits inauguration, making a strong technical case forthe reliability of utility-scale solar pv power plantsand resulting in a significant reduction in greenhousegas emissions. It is important to note that thegef grant, which was provided through ifc, e∑ectivelysubsidized 70 percent of the construction andstart-up costs of the cepalco solar pv plant. Thus,the intended potential for replication is currentlysomewhat limited, since global solar pv prices remaintoo high. 18The cepalco experience highlights the importanceof a strong local presence and knowledge ofthe local market and its regulations. Although thecepalco plant was a small plant, the PhilippineGovernment did not make a distinction between itand the more conventional electrical plants.cepalco was still required to comply with the permitprocess required for much larger fossil-basedplants. As a result, over 50 permits and licenseswould have been required, many of which were inappropriatefor a small, clean RE plant. Without thesta∑ knowledgeable of local government processes,this would have been a very daunting process, potentiallycrippling the project’s implementation.The cepalco project also highlights the necessityof capital cost reductions for larger-scale grid applications.Without the gef grant, the cepalco plantwould not have been financially viable, as the priceof solar pv modules was simply too high to be competitivewith the capital cost of a 1 mw diesel generator.Perhaps most important, the cepalco projectdemonstrated the potential for conjunctive use applicationsof solar pv.17World Bank Group, RenewableEnergy for Development—The Roleof the World Bank Group, 2004.18Other large-scale solar pv powerplants are now being constructedin Europe and North America,due to generous subsidyprograms, financial incentives(e.g., tax credits), and the existenceof renewable energy portfoliostandards, some of whichrequire a certain percentage ofsolar capacity. Plans for a 40mwsolar pv plant were recentlyannounced in Europe (see boxrelating to Moser Baer, page 22).ifc’s experience in the solar pv market 15

chapter 3IFC’s Lessons of ExperienceThrough the implementation of the projects discussedin this report, ifc has learned a great dealabout the solar pv market in developing countries,the type of financing required to support solar pvmarket growth in those countries, and what it takesto develop a successful solar pv company. Perhapsone of the most significant lessons that ifc haslearned is that the solar pv market is far more complexthan first envisioned.This complexity is rooted in the fact that, despitethe apparent social and environmental benefits, solarpv remains una∑ordable to the majority of theunelectrified population in most developing countries.The mere fact that it is more economicallya∑ordable for a government to provide electricitythrough solar pv than through grid expansion doesnot, in itself, make solar pv financially a∑ordable tothe end user. It is important to explore the di∑erentsegments of a potential market, develop productsthat are suitable for di∑erent consumers, and respondto a range of needs and income levels, as wellas identify opportunities where solar pv is the leastcostalternative.lessons about the solar pv marketin generalWhile solar pv technology is a well-established technology,and technical advances have been significantduring the last several decades, the emergence of aconsumer market for solar pv is relatively recent inmost developing countries. When ifc first becameinvolved in the solar pv market, there was a greatdeal of excitement regarding the potential forgrowth in the market. As ifc quickly came to realize,however, these projections were overly optimisticand not supported by changes in market fundamentals.A∑ordability remains a key issue today;government support is still needed, but often lacking;and, overall, unsubsidized solar pv programs aredi≈cult to implement, particularly in light of thedegree of subsidization and political support for alternatives,such as grid extensions.Expectations Were Overly OptimisticIt has now become apparent that early solar pv programs,designed by ifc and others, su∑ered fromoverly optimistic outlooks on the solar pv market’sgrowth opportunities. With 400 millionunelectrified households, the potential marketseemed vast, but it rapidly became apparent that theactual potential demand did not equal the entireunelectrified population. While extensive marketstudies were carried out prior to the implementationof any of ifc’s solar pv programs, these studiesfocused more on identifying businesses than onevaluating end-user demand. Furthermore, uponimplementation, it became apparent that many ofthe opportunities that had been identified by themarket studies during the planning stage were notappropriate. In the case of sdc, for example, notone of the over 100 opportunities identified in thefeasibility study received sdc support. The initialmarket assessment had overestimated the maturityof the market; it simply was not yet ripe for the typeof equity investment sdc was seeking to provide.Solar PV Is Not the Only Answer to RuralElectrificationMost of ifc’s financing programs had a dedicatedfocus on solar pv, an optimistic and too restrictive“tunnel-vision” reflection of the general belief that16 selling solar: part 1

solar pv was the best solution in areas not connectedto the electrical grid. Small-scale solar pv systems donot o∑er the constant supply of electricity that mostpeople want, and without focused financing programs,it continues to be una∑ordable to the vastmajority of the unelectrified population. In the absenceof a grid connection market, consumers havecontinued to find other, more a∑ordable ways(diesel generators, disposable batteries, kerosenelamps) to meet their electricity needs.The Hard-to-Define and Very SegmentedSolar PV MarketThe market for solar pv in developing countries wasinitially defined to be the total number of unelectrifiedhouseholds, some 1.6 billion people. In reality,the market for solar pv is much smaller andsignificantly more complex to define; smaller largelydue to the issues of financial a∑ordability and perceivedvalue, and complex in the need to recognizethe di∑erent market segments, which are rooted inincome level, lifestyle, and numerous regional andgeographic di∑erences.Financial Affordability and Perceived ValueThe biggest barrier to widespread adoption of solarpv technologies is a∑ordability. At the time each ofthe ifc initiatives was structured, it was widely believedthat the price for solar pv modules wouldcontinue to decrease significantly as mass productionscaled up, thus making solar pv electricity morecompetitive. However, as indicated previously, theanticipated decrease in solar pv prices did not materialize;the price actually increased in 2004/2005.The issue of a∑ordability has been a key driver ofthe business models employed by solar pv companies.In order to keep the upfront cost of shs a∑ordable,companies have generally followed one of twoapproaches: they have provided shs on a fee-for-service,or rental model, charging a monthly fee for theservice provided, or they have arranged for financingto allow the consumer to pay the balance of theshs in monthly installments (lease/hire purchasemodel). Many businesses also engage in cash sales.Through Soluz Honduras, experience has shownthat the fee-for-service, or rental, model has ultimatelyproved unsustainable, since unexpected gridexpansion tends to jeopardize the customer base beforethe high upfront costs of acquiring and installingthe solar pv systems can be recovered by thesolar utility.The continued high price of solar pv modules isfurther o∑set by the issue of perceived value. Experienceby ifc and others has shown that a∑ordabilityis not exclusively linked to price or the availabilityof financing. A∑ordability is also linked to the perceivedvalue and opportunity cost of solar pv purchase.In many rural settings, the purchase of a solarpv system represents a very significant expense; inVietnam, for example, the systems provided byLESSON LEARNED: GOVERNMENT SUPPORT CAN MAKE A DIFFERENCESOLAR PV PLANT CONSTRUCTION IN THE CZECH REPUBLICThe IFC/GEF CEEF program* has issued a $1.08 million project guarantee for a 1.2MW solar power plant, thelargest solar PV power plant installed to date in Central and Eastern Europe. The solar PV power plantbecame operational on January 29, 2007. This installation, in Busanovice, southern Bohemia (Czech Republic),will decrease the country’s CO2 emissions by 743 tonnes annually, and will produce 620MWh/year.Electricity from solar PV, as well as other renewable sources, has the support of Czech national renewableenergy legislation, which guarantees feed-in electricity tariffs for 15 years from project commissioningand off-take obligation for the grid operator. For solar PV power, the feed-in tariff is 13.62 Czech crowns/kWh ($0.62/kWh).The plant has already been connected to the distribution network and started commercial operation onFebruary 1, 2007. The 2,660 silicon-based solar PV modules, which are spread across 6,170 m 2 , can ensureelectricity supply for 172 households. The maximum planned capacity of the power plant is 693 kW. Asolar PV panel with an area of 1 m 2 produces as much electricity in a year as 250 kg of coal and saves atotal of 750 tonnes annually of CO2 that would otherwise be discharged into the atmosphere by the operationof a coal-fired power station. By the end of 2007, the plant capacity will be ramped up to 1.2 MW.* IFC/GEF-CEEF, the Commercializing Energy Efficiency Finance Facility, operates in Eastern Europe with both GEF and IFC financing. The facilityprovides a partial guarantee for loans made by local financial intermediaries for EE/RE projects, as well as technical assistance.ifc’s lessons of experience 17

Selco Vietnam were sold for roughly a year’s income.As a result, the decision to buy even a smallsolar pv system often meant sacrificing a larger item.There is a trade-o∑, however: with their limitedelectricity supply capacity, solar pv systems often donot come out on top, when compared to possible alternativepurchases.Economies of ScaleSolar pv technology is well suited to rural, highlydispersed, sparsely populated areas, but it is preciselyin these areas that solar pv programs are mostdi≈cult to implement. A certain scale is required inorder to become profitable, but if the company seesa need to reach too far out of the central communityto increase the scale, it can become too expensiveto service the consumers, and the economies ofscale are then lost. Solar pv companies are morelikely to experience success when able to operate inmarkets with critical masses of potential consumersthat were geographically concentrated. It is onlywhen operating in a relatively dense market thatcompanies are able to take advantage of economiesof scale. Part of Grameen Shakti’s success, for example,is attributed to its servicing densely populatedareas; economies of scale were not an issue. Whilethere are examples of solar pv businesses that haveoperated successfully in more sparsely populated areas,it is certainly a more di≈cult venture.Government Support and Enabling EnvironmentA supportive legal environment is essential andshould include as many of the following elements aspossible: no import duties or tari∑s on shs components;incentives for solar pv energy or absence ofcompeting subsidized electricity; publicly disclosedlong-term government electrification plans; and alegal basis for enforcing loan collection. For example,the sme Program’s investment in Soluz Hondurashad significant issues surrounding unexpectedgrid expansion, forcing Soluz Honduras to removenewly installed systems at a considerable loss. Incontrast, pvmti’s investment with Sunlight PowerMaroc benefited from an agreement with the18 selling solar: part 1

national utility to provide shs under a subsidizedfee-for-service scheme within an exclusive geographicconcession.SubsidiesUnsubsidized solar pv programs have provedparticularly di≈cult to implement. Rural electrificationis heavily subsidized throughout the developingworld, as are solar pv and other RE technologies indeveloped countries, such as Germany, Japan, andthe United States. All programs examined in this reportreceived some form of subsidization, be it inthe form of financing, using gef funds providedthrough ifc at terms unavailable in the market, orthrough local government, as in the case of SunlightPower Maroc, which has benefited from an extensivegovernment program, subsidizing fee-for-service solarpv systems. The cepalco project received a grantequivalent to 70 percent of the overall constructioncost of the plant. While the plant makes a strongtechnological case for the reliability of utility-scalesolar pv power plants, from a financial standpointthe plant would not have been feasible withoutsome form of subsidy. It is widely acknowledgedthat given most current market conditions, someform of subsidy is necessary to maintain solar pvbusinesses in the developing world. 19 ifc has foundthat there is a particular need for continued technicalassistance funding as part of its financing programs.The question that remains to be answered,however, is whether technical assistance grants willprovide enough of a subsidy, or if more substantialsubsidies are needed and, if so, in what form?lessons about ifc’s financingprogramsWith gef funding, ifc has been able to implementa number of di∑erent solar pv financing programs.Despite the enthusiasm, with which the inceptionof the di∑erent financing programs was based, ifclearned very early in the implementation of its solarpv programs that the market reality was not whathad been envisioned. There was need for greaterflexibility and patience. Despite the risk-sharingtools o∑ered by ifc, in large part through gef support,financial institutions (fis) continued to perceivesolar pv to be risky, due to their inexperiencewith the technology, the nature of smes, and theeconomics of the solar pv market. In the end, it wasfound that many of ifc’s financing programs requiredtailoring to the specific needs of individualcountries and, therefore, the experience o∑eredlimited replicability at a general level. Mostprofitable opportunities in the solar pv market alsolay further upstream in the value chain.Need for Flexibility in Program DesignThe solar pv market is very much a developing marketand, as such, it requires substantial flexibility inproject design, as shown by the following examples.ifc’s experience in solar pv has demonstrated theneed to adapt initial project designs to a number ofareas: from the length of repayment schedules tointerest rates to the business plan and, particularly,with respect to the financial instrument used.restructuring. In the face of continued di≈cultiesin placing funds, particularly in Kenya, pvmti’sapproach to the market was restructured. The restructuring,allowing for more technical assistancefunding and longer repayment periods to be providedto Kenyan clients, had a significant e∑ect onpvmti’s ability to place funds. The di∑erent experiencesof sdf and sdc also highlight the importanceof being open to revising the original program design.sdf, with its more flexible funding options(providing working capital loans with minimal security,guarantees, and grants), was able to successfullymeet its financing targets, while sdc, with itsfocus on larger projects, greater return expectations,and exclusive focus on equity investments, had agreat deal of di≈culty placing its available funds.Had sdc been designed di∑erently from the outset,or if it had been able to reach consensus amongst itsshareholders concerning its attempted restructuring,it is possible that the overall experience of sdgwould have been much more positive.19Solar pv businesses are able tooperate sustainably without subsidieswhen they are the lowest costsource of power for a marketsegment.volatile market conditions. Lower incomerural populations are particularly susceptible to themacroeconomic situation of their country (economicshocks, currency changes, changes in theprice of crops), as well as natural disasters. They arealso subject to cash flow issues, as monthly incomesfluctuate according to seasonal harvests and sporadicincome-generation activity. These issues had astrong e∑ect on the ability to make payments bothat the consumer and project level, and ifc programshad to introduce some flexibility into the repaymentplans to accommodate these issues. Grameen Shaktisaw a full 90 percent of its operating area flooded in1998, when the worst flood in over a century hitBangladesh. People were focused on the essentials(food and shelter) and, as a result, sales were nonifc’slessons of experience 19

GRAMEEN BANKexistent and the default rate on collections was extremelyhigh. The sme Program loan to GrameenShakti had been structured with a two-year graceperiod, with payments to be made on an annual basisand, as a result, Grameen Shakti did not have issuesin servicing the sme Program loan, even thoughcollections from Grameen Shakti’s clients had to bedelayed as a result of the flood. Selco Honduras alsosu∑ered a major setback following the devastationbrought by Hurricane Mitch in 1998.simplified transaction process. In manyinstances, ifc’s lengthy and cumbersome deal approvalprocess proved too stringent for small solarpv companies. The sme Program, which had beenspecifically designed for small and medium companies,adopted a less burdensome deal approvalprocess that was more in line with the characteristicsof solar pv companies. Lacking an extensive focuson smes, both sdc and pvmti had cumbersomeGrameen Bank was established by Professor Muhammad Yunus in 1971 as a researchproject. By December 31, 1995, the bank had equity of $100 million. Notregulated by the Bangladesh Superintendent of Banking or any similar regulatorybody, Grameen Bank is notionally owned by its 2 million members, eachof whom owns one share. The majority of the bank’s clients are poor and landlessand live in rural areas, with 94 percent being women. The members elect9 of the 13 members of the Board of Directors.In 2006, Muhammad Yunus and Grameen Bank were awarded the NobelPeace Prize for their work on economic and social development among the poor.20This experience is somewhatunique to the developing world.Many private equity and venturecapital companies are involved insolar pv manufacturing projectslooking to sell solar pv in developedcountries, where biggersystems are being sold, prices arehigher, distribution problems lessdaunting, and needed economiesof scale are easier to come by.documentation processes, resulting in deals beingcanceled due to the fact that the time from approvalto disbursement was too long, as well as high operationalcosts in relation to portfolio size.Need for Flexibility in Investment OfferingsIn addition to flexibility in program design, flexibilityin investment o∑erings is necessary. Investmento∑ering needs vary significantly, based on marketsegment, specific country, or regional needs. Manyof ifc’s programs o∑ered one particular type offinancing—sdc provided private equity, and thesme Program provided concessional loans throughintermediaries—and, as a result, it was founddi≈cult to place their funds. sdc was eventually disbandeddue to its inability to disburse funds, as themarket was generally not ready for private equity investments.The sme Program, with its concessionalloan o∑erings, proved more flexible, and placed themajority of its solar pv-related funds directly, withoutthe help of intermediaries.Private Equity Is Not the Answerifc attempted to attract private equity to the solarpv market through sdc. It found, however, that privateequity funds, as a financial instrument, are notwell-adapted to the needs of the solar pv market inthe developing world 20 and the demanding requirementsof a private equity fund. The private equitymodel is premised on the concept of high risk/highreturn. The solar pv market in the developing worldhas certainly proved itself to be high risk; however,financial returns have generally been disappointing.In addition to poor returns, sdc found it di≈cult tomake equity investments, due to the small dealsizes, limited management skills, lack of financialaccounting standards, inadequate exit strategies, andthe time-consuming and costly administrative monitoringrequired for equity investments.Yet, solar pv companies are in need of capital.Launching a successful shs distribution companyrequires a significant upfront investment in order topurchase equipment, establish distribution channels,and raise consumer awareness. ifc experiencehas shown that it generally takes several years to recoupthis investment. Renewable energy and energye≈ciency projects often attract similar types of investors;however, it is important to keep in mindthat the payback periods for renewable energy projects,in particular solar pv projects, are much longerthan those for energy e≈ciency projects. To compare,energy e≈ciency projects often have paybackperiods of less than two years, while renewable energyprojects are almost always over three years. Inthe case of solar pv projects, payback periods can bein excess of 10 years. Patient capital and long-termloan commitments with modest financial return expectationsare what is needed. For small rural businesses,simple loan instruments with modest securityprovisions are most appropriate.Financial Institutions Still Find Solar PV to BeToo Riskyifc underestimated the conservativeness of localfinancial institutions as far as providing financing tosmaller solar pv companies. ifc believes that in orderto leverage local financial resources, it shouldnot only raise awareness and provide technical assistance,but also engage the local financial institutionsmore directly by devising risk-sharing products thatcan be deployed to finance renewable energy pro-20 selling solar: part 1

jects. Unfortunately, solar pv proved to be too riskyfor most fis. Solar pv is deemed risky for two reasons:(1) most solar pv companies are smes, and fishave generally been wary of financing smes, and (2)the economics of the solar pv market means thatthere are high risks and uncertain returns. GrameenBank’s support of Grameen Shakti (see GrameenShakti case study, page 32) was a notable exceptionto the reluctance of fis to support solar pv projectsand provides a successful example of how financialintermediaries can support the solar pv market.However, this is more representative of the uniquenessof Grameen Bank (see box, page 20) than it isindicative of the possible interest of most local fis toengage in solar pv.The sme Program, in particular, sought to workthrough financial intermediaries. However, while theprogram was able to work through intermediariesfor a number of the projects implemented throughit, when it came to solar pv investments, this provedto be a significant challenge. In Vietnam, for example,the sme Program had initially hoped to providefinancing to the Vietnam Women’s Union (vwu),which had taken on the sales and collections role forSelco Vietnam. However, the vwu viewed the risksof providing end-user financing as too high, and wasunwilling to take on the responsibility for thefinancing. In the end, the sme Program provided theloan directly to Selco Vietnam. In fact, E+Co andEnvironmental Enterprises Assistance Fund (eeaf),both nonprofit nongovernmental organizations(ngos) and financing institutions with environmentalmissions, were the only intermediaries supportedby the sme Program that agreed to commit concessionalloan financing to solar pv projects.Need for a Broader Technological FocusIn the ifc experience, investment facilities that wereexclusively focused on solar pv (sdc and pvmti) hadmore di≈culty placing funds than the programsthat had the flexibility to also invest in other sectorsand technologies (sme Program). Furthermore,those types of programs were able to o∑set some ofthe risks associated with solar pv.Need for a Broader Operational FocusIt is important to note that sdf, with its exclusivefocus on solar pv, was largely successful. However,this success can be linked to its pre-agreed scope ofoperation; unlike other projects, sdf was a nonprofitentity that provided a range of business developmentassistance and seed financing to establish newLESSON LEARNED: THE NEED FOR A MIX OF TECHNOLOGIESTHE PORTFOLIO APPROACH TO DISTRIBUTED GENERATION OPPORTUNITY (PADGO) PROJECTThe ultimate goal of the PADGO project is to reduce CO2 emissions by displacingcentral fossil-fuel-based generation in favor of a portfolio of renewable andclean fossil-based distributed energy (DE) generation technologies with wasteheat recovery (also known as combined heat and power). In order to achievethis goal, the project has been divided into two phases.Phase 1 of the project will focus on Sri Lanka, and will have three specificgoals. First, it will develop a performance framework that would enable risksharing between IFC and the local banks on their existing portfolio of mini-hydroinvestments. Second, PADGO will focus on introducing new technologiesto the DE mix by promoting complementary DE technologies (for example, reciprocatingengines, biomass, PV, wind). IFC will work with established privatesector players to do one or more pilot projects with a technology that has notbeen extensively implemented in Sri Lanka. Third, it will identify the key problemsthat the electricity grid may face with large-scale DE generation, and willdevelop key guidelines on how DE generation can be assimilated into an integratedresource planning effort at Ceylon Electricity Board.During Phase 2, IFC will incorporate into the risk-sharing framework thelessons learned from the introduction of the new technology-based pilot projectinitiated in Phase 1. The framework will thus be made more robust and applicableto a larger set of technologies. The financing process will move closerto an asset-backed securities approach, as larger volumes of transactions aretargeted. Significant progress is also expected during Phase 2 on the integrationof DE technologies into a mini-grid structure that allows for dispatch capabilityand value for capacity.solar pv ventures and support existing early-stagebusinesses. The type of flexible support sdf providedwas greatly needed and in high demand in thesolar pv market and, as a result, sdf was largely ableto meet its goals. By contrast, both sdc and pvmtiwere faced with having to undergo major restructuring,since their narrow initial investment terms wereout of touch with the market reality. sdc was eventuallydisbanded when management was unable toidentify a large enough number of investments toprovide the type of returns investors sought. pvmtiwas restructured to allow for longer repayment periodsand increased funding for technical assistance.pvmti also relied only on gef funding, while sdchad private capital, which was more demanding.Limited ReplicabilityThe experience of pvmti, in particular, demonstrateshow the same model can lead to di∑erent resultsin di∑erent countries because of specific countryconditions. pvmti operates in India, Kenya, andMorocco, three countries selected for their supportivepolicy environments and the presence of aifc’s lessons of experience 21

vibrant emerging solar pv market. pvmti experiencedconsiderably more success in India, where themarket for solar pv was widely established and enjoyedconsiderable government support, and therewere a large number of established solar pv companiesand relatively widespread knowledge about solarpv technology. In Kenya, pvmti found that therewas considerably more need for technical assistancefunding. A large number of solar pv systems hadbeen sold in Kenya on a pure cash basis by verysmall local companies, but there were no performancestandards, and the quality of many systemshad been called into question, undermining the initialprogress that had been made in the market. InMorocco, a stepped-up fee-for-service program,subsidized by the government, reduced the need forconsumer financing, while increased availability ofblack market solar pv modules put increased pressureon module prices.There Are Profitable Opportunities in theSolar PV Market, But They Lie Further Upstreamin the Value ChainAn important finding that emerged from ifc’s solarpv experience is that there are more viable opportunitiesfurther upstream (module manufacturing) inthe value chain than downstream (shs distributors).ifc’s investments lay primarily further downstreamin the value chain. One reason for this is that manufacturingcompanies are often able to obtain localcommercial financing more readily, because it isbacked with assets, unlike the cash-flow-backedfinancing provided to shs distributors, and ifc aimsto provide financing only when it is not availablethrough local sources. A second reason is that thevast majority of manufacturing activity in the solarpv industry is focused on export to subsidized westernmarkets (primarily Europe), and it was not appropriatefor ifc to use gef funds when it could notensure that they would be used to support renewableenergy in the developing country.Need for Technical Assistance FundsTechnical assistance grants are still needed in orderto help move the solar pv market forward. In orderfor commercial solar pv enterprises to be successful,technicians need to be trained, industry standardsneed to be developed, and local governments needto be lobbied for support. pvmti found that ta wasparticularly needed in Kenya. In the 2004 restructuring,the initiative approved additional ta grantfunding to support the training of solar pv technicians,create quality awareness in the market, supportthe Kenya Renewable Energy Association(krea), and establish a quality assurance programfor shs in the Kenyan market.Shareholder DiversityThe more diverse a shareholder group is, the moredi≈cult it is to manage expectations. The immaturityof the solar pv market means that financingprograms have to be flexible in order to respond tothe changing market realities. With a diverse shareholdergroup, this flexibility often does not exist.Both pvmti and sdc found very early on in the implementationstage that original targets should berestructured and reevaluated. pvmti, with its simpleshareholder base, was able to complete thesechanges; however, sdc, with its very diverse shareholderbase, found it impossible to reach consensuson changes, which forced its dissolution, ratherthan allowing for more flexible restructuring. Theshareholder diversity that had been much applaudedduring the initial structuring of sdc proved, in theend, to be one of sdc’s greatest constraints.LESSON LEARNED: MOVING UP THE VALUE CHAINMOSER BAER IN INDIAMoser Baer India Ltd. (MBIL) is the third-largest manufacturer of recordable opticalstorage media products (CDs and DVDs) in the world. MBIL is also an existingIFC client. Currently, MBIL is undertaking a two-year diversification programthat involves setting up an export-oriented solar PV cell and modulemanufacturing facility with an installed annual production capacity of 80MW inGreater Noida, India.IFC has recently approved a $22.5 million long-term loan to the company tosupport this $92 million project. This project, which has the potential to avoid80,000 tonnes of CO2 emissions annually, will also lead to the creation of about600 additional jobs.lessons about what makes asuccessful solar pv companyWhile there are some notable successes amongsome of the solar pv companies ifc providedfinancing for, the majority of them did not live upto their original expectations. Although ifc has notbeen able to identify a fully viable sustainable businessmodel for solar pv distribution companies indeveloping countries to replicate, its decade-longexperience working in the solar pv market has highlighteda number of key areas that companiesshould focus on and resources that need to be in22 selling solar: part 1

place in order to help ensure the successful operationof such businesses.Product Offering and Market SegmentationIn order to ensure that they are providing the rightproducts, solar pv companies should acknowledgethe di∑erent market segments that exist. Low-incomeconsumers are often looking for a solar pv systemthat will support a single light source, whilehigher income consumers might well be grid-connectedand be looking to purchase larger solar pvsystems that recharge back-up batteries to help ensurean uninterrupted power supply in the event ofpower outages. The needs of each of these di∑erentcustomers vary greatly, and solar pv companiesshould adjust both their product o∑ering and theirmarketing strategies in order to satisfy di∑erent consumermarket segments.In addition to income level, consumption prioritiesand lifestyle have proven to be key segmentationissues. These issues also vary from country tocountry. In Vietnam, for example, having a televisionwas seen as more important than having alight; therefore, people were more interested inlarger solar pv systems. Additionally, there was nodebt or consumer credit culture in Vietnam; i.e.,there was little market for smaller systems, and consumerstended to be from higher income groups,who could a∑ord to purchase systems outright (socalled“early adopters”).Provision of Consumer FinancingLESSON LEARNED: AFFORDABLE PRODUCT OFFERING TAILORED TO TARGET MARKET SEGMENTLIGHTING THE BOTTOM OF THE PYRAMIDThe Lighting the Bottom of the Pyramid program will leverage IFC’s global network,industry expertise, and regional experience, as well as donor funds to catalyzelocal and international lighting-related companies to offer the unelectrifiedpopulation in Kenya and Ghana greater access to modern and affordableoff-grid lighting products and displace fuel-based lighting products (kerosenelamps, candles).Current consumption of fuel-based lighting represents a large global market,mostly served by oil and gas companies, but largely untapped by lightingcompanies. Independent estimates indicate that worldwide spending on fuelbasedlighting in developing countries is $38 billion per year. In Kenya andGhana alone, IFC estimates the total spending on fuel-based, off-grid lightingto be $1.4 billion per year. Hence, IFC believes there is an opportunity to attractlighting companies to enter and compete in the fuel-based, off-grid lightingmarket with modern and affordable off-grid lighting products, harnessing theprivate sector profit-seeking motivation to increase access to modern lightingservices and reduce poverty.In order to seize this opportunity, IFC will facilitate the entry of local and foreignlighting companies to this market by helping firms (i) understand the market,including consumer behavior and preferences concerning lighting, acceptablepricing points, and distribution channels, and (ii) understand and mitigatethe perceived risks of entering into a new market in a region that, for most privatecompanies, is very challenging. The entry of several modern lighting companiesin this market and their competition for market share will bring unelectrifiedcitizens in Kenya and Ghana a variety of modern off-grid lightingproducts that will be better and more affordable than fuel-based lighting. Theirlower power requirements will also enable more cost-effective use of solar PVas a power source (e.g., solar PV lanterns).As of January 2007, 135 private companies and 63 stakeholders from 35countries had expressed interest in participating in this initiative. The projecttarget end-results for Kenya and Ghana are (1) to provide greater access to offgridlighting products that are more modern and affordable than fuel-basedlighting; (2) to reach 316,000 (low-end scenario) to 1,500,000 (high-end scenario)end users with modern, off-grid lighting products by 2015; and (3) to reduceCO2 emissions from fuel-based lighting from 782,000 tonnes (low-end scenario)to 3,909,000 tonnes (high-end scenario) by 2015.For more information, please visit www.ifc.org/led.tionship was a major contributing factor to the successof Grameen Shakti.Management and StaffingA consistent theme that emerged from ifc’s solar pvexperience is that the entrepreneur is absolutely criticalto the success of the project. Given that solar pvplayers tend to be small and operate in complexmarkets, it is documented that hands-on managerswho possess strong management skills are crucial.Consumer financing has often been seen as the keyelement to developing the solar pv market. Withoutfinancing, the large initial price tag of acquiring asolar pv system is simply una∑ordable to the vastmajority of the population without electricity. ifc’sexperience has shown that investors and banks donot like to finance solar pv purchases, as they perceivethem to be too risky. Also identified was thatsolar pv companies are generally more skilled at themanufacturing and commercial distribution of solarpv, and are not concerned with the risks associatedwith provision of consumer financing. Successfulsolar pv companies are those that have a way of engagingskilled providers of consumer financing, thusallowing them to remain focused on their core solarpv business. Grameen Shakti, an shs provider inBangladesh that received financing from the smeProgram, was well positioned to provide consumerfinancing because of its parent company, the wellestablishedand respected Grameen Bank. This relaifc’slessons of experience 23

Managers should be flexible in order to respondquickly to changing market realities. Establishing alocal presence through appropriately trained localsta∑ is mandatory for any company looking to be aplayer in the solar pv market. As the solar pv marketin each country di∑ers, the social and culturalknowledge of local sta∑ is essential to the developmentof a solid understanding of consumer needs.As mentioned above, Grameen Shakti benefitedsignificantly from its relationships with GrameenBank, as the bank’s presence in approximately36,000 villages provided significant local knowledgethat it was able to share with Grameen Shakti. Furthermore,Grameen Bank’s general manager dedicated20 percent of his time to the direction ofGrameen Shakti.The experience of cepalco in the Philippineshighlights the need for a strong and experiencedmanagement team and sta∑ who understand the localenvironment. Although it was only a small, 1 mwre project, local sta∑ had to obtain over 50 permitapprovals before construction on the cepalco solarpv plant could begin, a feat that would likely havebeen impossible without adequately knowledgeablelocal sta∑. Furthermore, the permitting process delayedconstruction and, as a result, strong managementwas required for the project to be completedon schedule.with local educational institutions to ensure that itremained on the cutting edge of technology. WhenSelco entered the Vietnamese market, it formed apartnership with the Vietnam Women’s Union(vwu) to support the financing of sales of shs inrural communities. The vwu had representatives inevery village in the country, and while the partnershipwas eventually dissolved, it was of great benefitto Selco Vietnam, as they began to gain a footholdin the country.Sales versus RentalThere are two primary models currently being employedin the solar pv market: sales (on a hire-purchasebasis) and fee-for-service rentals. Accordingto ifc’s experience, the sales model is much moresustainable, especially given current market conditions.Soluz Honduras began operations as a feefor-servicecompany, supplying shs to consumerson a rental basis. It became apparent after severalyears of implementation, however, that the rentalmodel was not financially sustainable. While arental model made it easier for the end user to ac-Local Partnerships and Government RelationsThe importance of strong government relations,particularly in terms of understanding planned gridexpansions, cannot be overstated. The Soluz Hondurasexperience showed the significance for stronggovernment relations to be backed by legally bindingconcession agreements. Soluz was operating inHonduras without any formal government concessionsand, as a result, found itself unprepared forcompetition from unexpected grid expansions.Grid expansions are a political tool in many developingcountries, and are often unpredictable interms of funding and timing. Operating withoutsome sort of binding legal agreement from the governmentputs companies at increased risk of customerloss.In addition to supportive government relationsand, in some instances, due to inadequate governmentrelations, the development of local partnershipsis pivotal to the success of a solar pv company.When Grameen Shakti first began operations inBangladesh, for example, local government supportwas lacking, and the company sought to partner24 selling solar: part 1

quire the systems, the equipment cost accumulatedwith the company and, as a result, it was not financiallysustainable for an sme start-up. The initiallarge outlays for capital equipment could not beo∑set by small monthly rental fees. Soluz Honduraseventually adopted a sales model.MarketingEven when a company has matched the right productsto the right market segment, a strong marketinge∑ort is crucial. There is still substantial misinformationand lack of understanding as to whatservices a solar pv system can reliably provide and atwhat cost. Furthermore, the presence of poor qualitysystems in some markets has resulted in solar pvbeing perceived as unreliable and even undesirable.Solid marketing strategies that include demonstrationprojects should be in place in order to educatethe general population. Grameen Shakti initiallyprovided shs free of charge to key people in a village,promoting a type of demand associated with“keeping up with the neighbors”. Selco Vietnam engagedthe vwu to help sell their systems, as the vwuhad presence and influence in just about every villagein the country. These unique marketing strategiescontributed significantly to overall sales.Entrepreneurial SpiritIn some countries, notably India and Bangladesh,there was a considerable entrepreneurial spirit tobe found among end users. This entrepreneurialspirit appeared to have some influence on the successof the solar pv company itself. In Bangladesh,one end user used solar pv panels to charge cellularphones, which he then rented out to people in hisvillage. His business was so successful that he wasable to purchase a larger solar pv system while alsoproviding a better education opportunity for hischildren. Where this type of entrepreneurial spiritexisted, there was greater demand for solar pvsystems, and consumers had a greater ability tomake their payments. It also demonstrated howthe provision of solar pv can lead to increasedincome-generating activities.ifc’s lessons of experience 25

chapter 4IFC’s Approach Today21This includes hydropowerprojects greater than 10mw perfacility.WORLD BANK GROUP COMMITMENT TO RENEWABLE ENERGYHaving extensively evaluated not just its own experience,but also the experience of several key players inthe solar pv business, ifc remains cautiously optimisticthat it is not a question of “if”, but of “when”the goal of a self-sustaining solar pv market in developingcountries will be met. Simultaneously, ifc recognizesthe current limitation of solar pv technologiesto address the issue of rural electrification. Tothat extent, it is currently exploring new ways to addressrural electrification through using a variety ofrenewable energy technologies as it moves awayfrom a specific solar pv focus to a more technologyneutralapproach.Although generally less heralded now comparedto the mid-1990s, solar pv as a technology continuesThe WBG has remained true to the commitment it made at the June 2004 InternationalConference for Renewable Energies in Bonn, Germany, to increaseits renewable energy and energy efficiency portfolio by 20 percent during afive-year period (2005–2009). During both 2005 and 2006, the WBG surpassedits Bonn target, having financed $668 million worth of EE/RE projects in 2006and $461 million worth in 2005. These commitments represent a 45 percent increasefor new RE and EE, more than double the Bonn 20 percent target. Overall,financial support to EE/RE 21 was $860 million in fiscal year 2006. Total WBGsustainable energy financing in 2006 supported 61 projects in 34 differentcountries.Among the various WBG institutions and units, IFC was the largest contributorto RE and EE, with $393 million in commitments, and contributions of $326million of its own funds for new RE and EE projects and $67 million for hydropowerprojects greater than 10MW. These increases suggest that the concertedefforts of the WBG to scale up support for new RE and EE are having apositive impact.to hold promise, having proved to be the most appropriateway of meeting the power needs (lighting,television, radio) of dispersed and remote ruralhouseholds. With the rising price of crude oil andnatural gas, the global commitment to the MillenniumDevelopment Goals and Kyoto Protocol, renewableenergy technologies, including solar pv,will become more economically viable. The privatesector can play an important role in making renewableenergy and solar pv services available, asdemonstrated by ifc through its recent investmentin Moser Baer in India (see box on page 22).The most important factor that will determinethe future role of solar pv in rural electrification initiativescontinues to lie in the ability of companiesto identify the niche market segments for which thistechnology is the least-cost alternative. The continueddecline in solar pv prices will help create moreof these opportunities. Increased manufacturing capacity,new materials that bypass the global bottleneckscaused by the limited supply of silicon, andnewer and higher e≈ciency solar pv materials andend-user devices (i.e., lighting via leds) are all contributingfactors that should reverse the recent upwardtrend in solar pv price.Continued Support to the MarketWhile solar pv is no longer a specific focus, ifc remainscommitted to it as a renewable energy technologyfor addressing the issue of ruralelectrification in developing countries. ifc hasmoved away from solar pv-focused initiatives, suchas sdg and pvmti, in favor of more technologicallyneutral programs. The corporation is increasinglyproviding its own funding on commercial terms,without reliance on donor subsidies, to support26 selling solar: part 1

larger utility-scale projects, through its InfrastructureDepartment, as well as solar pv module manufacturingcompanies, through its Global ManufacturingDepartment. (See Lesson Learned box onMoser Baer, page 22) Smaller solar pv initiatives,such as the shs distributors profiled in this study,may find ifc support through the following marketacceleration schemes:working through fis: the environmentalbusiness finance program (ebfp). The ebfp,a $20-million gef-funded facility, builds upon theexperiences of the sme Program, and is specificallyinterested in engaging financial intermediaries inthe financing of smes involved in activities thatbenefit the global environment. Designed to providefis with risk-sharing mechanisms that encourageintermediaries to provide financing to smes undertakingenvironmental projects, the ebfp alsoprovides technical assistance grants to develop andstrengthen an fi’s appraisal, risk management, andmonitoring and evaluation processes, as well as promotemarket development. Sustainable energy is akey focus of this program.direct investment to support a mix oftechnologies: the sustainable energyfacility (sef). The sef is a $14 million ifc/geffund that finances sustainable energy and energye≈ciency projects in Brazil, Central America,China, and Southeast Asia. Designed based on theexperience of previous ifc programs, including sdg,the sef structure has a more streamlined approvalprocess. A clear focus was placed on debt instrumentsover equity, with convertibility features totake advantage of any potential upside. Unlike someearlier initiatives, the sef has moved away from asingle focus on solar pv to a broader renewable energyfocus.focus on affordable off-grid lightingproducts: lighting the bottom of thepyramid. This initiative seeks to catalyze local andinternational lighting-related companies, o∑eringthe unelectrified population in Kenya and Ghanagreater access to modern and a∑ordable o∑-gridlighting products, and displacing fuel-based lightingproducts (such as kerosene lamps or candles). Theinitiative aims to facilitate the market entry of thelighting companies by helping firms, firstly, to understandthe market, including consumer behaviorand preferences concerning lighting, acceptablepricing points, and distribution channels and, secondly,to understand and mitigate the perceivedrisks of entering into a new market in a region that,for most private companies, has been very challenging.(See Lesson Learned box on the Lighting theBottom of the Pyramid initiative, page 23.)TABLE 5: WORLD BANK GROUP COMMITMENTS FOR RENEWABLE ENERGY AND ENERGYEFFICIENCY IN FISCAL YEAR 2006 (MILLIONS OF DOLLARS)SOURCE OF FUNDS NEW–RE HYDRO >10MW EE TOTALWorld Bank (IBRD/IDA) 135.7 118.6 115.3 369.5World Bank 54.7 6.0 1.2 62.0(GEF and Carbon Finance)IFC (own funds) 17.4 67.0 309.0 393.4IFC (GEF, Carbon Finance and 13.0 0.0 20.1 33.1other trust funds*)MIGA 0.0 0.0 1.8 1.8Total 220.8 191.6 447.4 859.8* The IFC’s “other trust funds” category includes the Environmental Opportunities Facility.focus on distributed generation projects:the portfolio approach to distributedgeneration opportunity (padgo). This projectaims to reduce CO2 emissions by displacingcentral fossil-fuel-based generation in favor of aportfolio of renewable and clean fossil-based distributedenergy generation technologies with waste heatrecovery (also known as combined heat and power).A key focus of the project is on developing a performanceframework to enable risk sharing betweenifc and local banks, and on piloting private companyprojects using new technologies. (See LessonLearned box on padgo, page 21.)Through the above market acceleration initiatives,the corporation aims to provide technical assistanceand financing to support renewable energytechnologies and practices that are commercially viablein certain applications, but whose market penetrationis hindered by the persistence of market barriers.These barriers may include high upfront costs;a lack of financing, successful business models, adequateproduct quality standards, and consumerawareness; limited managerial and technical skillsamong project developers, etc. By addressing thesebarriers through carefully designed market interventions,which may include transaction support, enterpriseand public education, and development ofquality standards, these initiatives seek to accelerateifc’s approach today 27

22While gef support is generallynot needed for small hydro initiatives,it is still an important componentof solar pv, geothermal,and wind initiatives.MAINSTREAMING SOLAR PV INTO IFC INVESTMENTSthe penetration of sustainable energy technologiesso that, at the end of project implementation, themarket is further ahead than it would have beenotherwise.Going forward, ifc expects to see an increasingnumber of opportunities for mainstream renewableenergy and energy e≈ciency investments, as renewableenergy technologies become more competitiveand regulatory frameworks are improved to encouragegreater utilization. Beyond these e∑orts tofinance mainstream sustainable energy projects, ifchas utilized and will continue to utilize limited—The projects described in this report were primarily financed with concessionalresources from GEF and other donor support mechanisms, because they didnot meet minimum eligibility requirements for IFC investments (although IFCalso did invest on its own account in the externally managed funds created underSDG). One measure of success in donor-supported programs is the processof “mainstreaming”; i.e., the ability to make similar investments on commercialterms without donor subsidies, an evolution that is occurring in the contextof IFC’s EE financing programs. As described in this report, the Corporation hasapproved an investment in a solar PV manufacturing facility in India (see LessonLearned box on Moser Baer, page 22) and has indirectly supported a bankloan for a grid-tied solar PV power plant in the Czech Republic through a cleanenergy finance program (see Lesson Learned box on solar plant constructionin the Czech Republic, page 17). For IFC to make additional fully commercial investmentsin solar PV production or enterprises, several conditions will have tobe met:and carefully targeted—concessional funding tosupport worthwhile projects that are likely to acceleratethe application of re and ee technologies indeveloping countries. Since gef was established in1991, a significant quantity of funds has been providedby gef under its operational programs to mitigatethe e∑ects of global climate change. ifc has receivedapproval for more than $200 million in geffunds to support climate change mitigation initiatives,including renewable energy. gef funding continuesto play a vital role in ifc’s continued supportof solar pv and other renewable technologies. 22 ifcalways attempts to minimize the use of concessionalfunding and targets the subsidies in a hierarchy ofpreferred market interventions that favor projectsthat are closest to commercial viability and preferablyinvolve a mainstream ifc investment.By building partnerships with key players in therenewable energy field and increasing confidence insolar pv technology among suppliers, governments,utilities, and end users in developing countries, ifc,as part of the World Bank Group, seeks to promotesolar pv and sustainable energy technologies. Ingeneral, it seeks to identify avenues to maximizefinancial leverage and experience and acts on particularopportunities to aggregate markets.∫ The investment should meet minimum size requirements to justify IFC’stransaction costs. While some latitude has been allowed for RE projects, dealsof less than $10 million are unlikely to be attractive.∫ The expected rate of return should be commensurate with the level of perceivedrisk, which may be an issue for the production and sale of solar PV cellsand modules currently being sold, primarily to satisfy short-term regulatorypolicies in a few industrialized countries, principally Germany, Spain, and theUnited States.∫ Other positive attributes that may increase IFC interest in a project includeopportunities to engage and influence government policy (e.g., through a solarPV purchase program), the leveraging of commercial finance from local FIs, localemployment and associated supply chain benefits of the investment, andexpected opportunities for further business growth.∫ The proposed financing should also meet standard IFC conditions, includingmaintenance of appropriate minimum debt service coverage ratios, projectedbusiness performance metrics, sponsor support, and security arrangements.Most importantly, the financing should be based upon a sound and financiallyviable business plan that addresses a quantifiable market opportunity and isguided by an experienced management team.28 selling solar: part 1

part 2Selected Case Studies

case studies 1The IFC/GEF Small and MediumScale Enterprise ProgramThe ifc/gef Small and Medium Scale EnterpriseProgram (sme Program) 23 was established in 1995.Financed with $20 million in gef funds and managedby ifc, the program’s goal was to improve accessto finance, capacity building, and markets forsmes active in the areas of climate change mitigation(energy e≈ciency and renewable energy) and biodiversityconservation. The program was the first geffunded,non-grant, sme financing program targetingthe private sector and the first gef program designedto receive capital reflows.backgroundThe objective of the sme Program was to encouragethe private sector to generate global environmentalbenefits. It provided loans of $500,000 to $1 millionto various intermediaries (financial, not-for-profit,ngo), and private companies for on-lending to smeswhose activities would conserve the global environment.The goal was to help these smes expand sothey would generate more environmental benefits.At the same time, the program sought to demonstratethat environmental benefits could be achievedthrough the private sector on a commercial basis,without the need for grants or subsidies.The intermediaries were selected by ifc on thebasis of their experience working with smes as wellas their financial viability, understanding of environmentalsectors, and technical capabilities in bothenvironmental and financial areas. The intermediariesidentified, assessed, financed, and monitoredenvironmental sme projects, assuming the risk inherentin these projects by providing loans to, ormaking equity investments with, the smes. Initially,the intermediaries typically received a long-term (upto 10 years), low-interest-rate loan (typically 2.5 percentper year) from the sme Program, combiningtheir own funds with other sources of funding tocomplement the financing requirements for the eligiblesme projects.Over its lifetime, the sme Program established asolid reputation and momentum, attracting the continuedinterest of intermediaries and other institutions.The initial $4.3 million pilot phase was replenishedwith $16.5 million in 1997 to expandoperations and reach additional smes. Over its lifetime,24 the sme Program approved $16.9 million to25 nontraditional financial intermediaries, ngos, orcompanies in 21 countries, which have providedfinancing to some 140 smes.While the sme Program was not designed tospecifically target the solar pv sector, it became operationalat a time when ifc had become interested inmaking solar pv-related investments. Ultimately, thesme Program financed six projects that involved solarpv businesses: Grameen Shakti, Soluz Honduras, andSelco Vietnam (all shs distributors); E&Co and eeaf(both nonprofit financing organizations with an environmentalmission) 25 ; and Cogener (a Swiss engineeringcompany that installed solar-powered advertisingpanels at a Tunisian airport). This study hasfocused on the three shs distributors:ı Grameen Shakti. A subsidiary of GrameenBank, Grameen Shakti works to develop and deliverrenewable energy systems to rural households andbusinesses in Bangladesh. The primary focus is onsolar pv shs.ı Soluz Honduras S.A. de C.V. A subsidiaryof Soluz, Inc., USA, Soluz Honduras sells and rentssmall solar pv systems to rural customers in Honduras.30 selling solar: part 2 / case studies

ı Selco Vietnam, Ltd. A subsidiary of U.S.-based Solar Electric Light Company, Selco Vietnamsells solar systems to unelectrified households inVietnam.Together, these three projects have installed over24,000 shs, for a total electrical capacity of over1.3mw at peak performance.what worked and what did notA “Hands-on” ApproachA key factor for the success of the sme Program wasthe small size of the program, which enabled themanagement to be very “hands on” and knowledgeableof the projects it financed. Unlike other ifcprojects with external management, the programwas managed by an internal sta∑ team. This “handson”management approach enabled the program torespond rapidly to restructuring.Diversity of Portfolio Offsets Risks of Solar PVIn contrast to other facilities with which ifc was involvedin the solar pv sector, the sme Program wasable to invest in smes working in sectors other thansolar pv. Given this flexibility, the program was ableto develop a diverse portfolio that was not dependenton one particular market, allowing it to o∑setthe risks of the solar pv market through smes workingin less risky markets. Many of the solar pv venturesoperational around the same period as thosefinanced by the sme Program had considerabledi≈culties. The solar pv market simply did not developas had been expected.The Need for Local Ownership and GovernmentSupport Proved VitalOne of the key lessons of the overall sme Programexperience was the importance of local country ownershipand government involvement. Evidence suggeststhat this was similarly the case among solar pvfocusedprojects, Grameen Shakti being the onlyproject that was locally owned and operated, whileboth Soluz Honduras and Selco Vietnam were subsidiariesof U.S.-based companies. Both Soluz andSelco were overly enthusiastic about the size of theirpotential markets, and both su∑ered from a lack ofgovernment support. While Grameen Shakti did nothave considerable support from the Bangladeshi government,it had the support of a widely recognized,respected, and reasonably well-capitalized organizationwith a similar client base that helped them todevelop networks to overcome that obstacle.Economies of Scale Are Hard to Come By inSparsely Populated AreasThe sme Program experience in solar pv highlightsthe importance of economies of scale to the solar pvmarket. Grameen Shakti, operating in densely populatedBangladesh, was a successful venture; Soluz,operating in Honduras where the rural populationwas more dispersed, was less so. While each entityoperated under a di∑erent business plan, both wereconfronted with having to reach a certain scale in orderto be profitable. However, that scale was far easierto attain in a densely populated area than withina dispersed population. While Grameen Shakti easilygained access to economies of scale serving manypeople in one community, Soluz Honduras, operatingin areas with more dispersed populations, founditself in a no-win situation. To increase its scale, itneeded to expand its area of operations, but when itdid extend, its service costs increased.conclusionAlthough the sme Program initially planned tofinance projects through fis, it ultimately financedmost of its solar pv-related sme projects directly. 26The program had found that fis had little interest infinancing solar pv projects (the two solar pv-relatedprojects that were financed involved fis with an environmentalmission). Many fis remained leery offinancing smes, and when smes were coupled with anascent technology like solar pv, fis (particularlycommercial banks) became even more reluctant.Grameen Shakti, the sme Program’s most successfulsolar pv project, far exceeded expectations interms of the number of solar pv systems installed.Grameen Shakti continues to perform well, with atotal of 77,000 shs installed, benefiting more than700,000 people in Bangladesh. 27 The performanceof other pv projects of the sme Program projects hasfailed to live up to original expectations.The sme Program proved overall to be quite successful,despite the mixed experience in the solar pvsector. Such experience gained has now been incorporatedinto the design of the ebfp. As previouslymentioned, the ebfp is an ifc/gef partnershipwhich targets smes working on projects that arebeneficial to the global environment. This programhas re, including solar, as one of its target technologiesand activities.23In the case of the sme Program,smes were defined as enterpriseswith assets valued at less than $5million.24In 2004 the sme Program wasabsorbed by the gef-funded EnvironmentalBusiness FinanceProgram.25Both E&Co and eeaf on-lentto solar pv-related projects.26The E&Co loan to Rex Investmentin Tanzania and the eeafloan to Soluz Dominicana representthe only solar pv projects toreceive sme Program fundsthrough fis.27http://www.grameen-info.org/grameen/gshakti/index.html,February 2, 2007.the ifc/gef small and medium scale enterprise program 31

grameen shaktiGrameen Shakti (meaning “village power” in Bengali),established in July 1996 by Grameen Bank,aims to support this bank’s poverty reduction missionby developing and delivering re systems torural households and businesses in Bangladesh.While the focus was on solar pv shs, GrameenShakti also supported wind and biomass projects,though on a much lesser scale.backgroundGrameen Shakti has a secondary mandate, that ofhelping connect rural areas to the world through(solar-powered) information technology. Althoughregistered as an ngo, it is run, for the most part, asa for-profit enterprise. The company’s solar pv programrepresents its largest business line, purchasingsolar pv panels and other systems components (i.e.,batteries) from a range of foreign and local suppliers;and assembling, selling, installing, and, wherenecessary, financing them.The Bangladeshi market for shs is considered tobe relatively large. Approximately 70 percent ofhouseholds do not have accessibility to electricity,and frequent floods and cyclones, low levels of urbanization,and a very slow political and economicreform process have made establishing a traditionalenergy network (or grid) very challenging. Furthermore,the density of the Bangladeshi populationmeans that, even in rural areas, there are significantconcentrations of potential consumers.In March 1998, Grameen Shakti was approved forfinancing from the sme Program. The $750,000 thatthe firm received permitted it to purchase shs inventory.The program loan called for the sale of 3,200systems within two years and provided a two-yeargrace period on repayment. This freed capital forGrameen Shakti to provide financing to customers,enabling them to overcome two considerable barriersto shs sales in Bangladesh, namely, high upfrontcosts and lack of consumer credit.An additional barrier to shs sales in Bangladeshwas lack of a strong field-based sales and servicestructure. The company, nevertheless, was fortunatein its ability to tap into the existing Grameen Bankbranch network. The latter, as indicated earlier, hasa presence in approximately 36,000 villages inBangladesh, and Grameen Shakti operates througho≈ces housed within the bank’s branch locations. 28Grameen Shakti o≈ces are established in locationswith high electricity demand and no access to thegrid. Each o≈ce is sta∑ed by two people, a managerand a technician, and is overseen by division managerswho report to the general manager ofGrameen Shakti.32 selling solar: part 2 / case studies

In addition to being able to tap into GrameenBank’s branch network to reach customers andquickly establish a local presence, Grameen Shaktialso benefits in other ways from the bank. Most notablyis the use of the Grameen name, a namewidely recognized and respected. Grameen Shaktialso makes use of the expertise of the general managerof Grameen Bank, who spends 20 percent ofhis time directing Shakti.obstaclesOverall, Shakti has experienced many issues in theBangladeshi market that are similar to those ofother shs companies in other world markets. Therehas been considerable skepticism concerning the viabilityand cost of shs systems, coupled with thelimited purchasing power of target end users, relativeto the large capital cost of imported shs equipment.Large volumes were needed to get unit costsdown to a financially sustainable level, but to e∑ectivelydo this, a large sales force was needed, and thecost of making individual sales with associated supportwas high.While Bangladesh is ideally suited to solar powerbecause of its higher than average solar radiation(ranging from 4.0 to 6.5kWh per square meter), solarenergy industry activity was minimal at the timeof Grameen Shakti’s founding. The local markethad not yet been established, and the BangladeshiGovernment o∑ered no support. Heavy importtaxes on internationally sourced solar panels and alack of local suppliers drove up prices. Furthermore,a general lack of awareness surrounding the technologywas a consequence for low demand.To overcome these obstacles, Grameen Shaktiplaced considerable focus on providing increasedvalue to its clients, while making a dedicated e∑ortto reduce costs and thus lower prices. An extensivewarranty package (which could be extended for asmall fee) included free maintenance for the firstthree years, training seminars for clients, routinesystem maintenance, and monthly inspections. Thiswarranty has enabled the establishment to manageits maintenance costs and has contributed to a highlevel of customer satisfaction. Additionally, thecompany o∑ers clients a 20-year money-back guaranteein the event that a client is, for any reason,unsatisfied with his/her system or the national grid,which is extended to service the client.Lacking government support, Grameen Shaktihad to rely on partnerships with other organizationsin order to stay ahead. Partnerships with educationalinstitutions and suppliers played an importantrole in pursuit of new technologies andidentification of trends. In partnering with educationalinstitutions, the company hoped to maximizeits resources and provide clients with the most upto-dateand e≈cient technology. The provision ofthe most market-applicable technologies allowedthe management to keep a step ahead of the competition.Partnerships with suppliers also proved usefulGRAMEEN SHAKTI AWARDSLike its parent, Grameen Bank, Grameen Shakti hasbeen widely recognized for its efforts in the internationalcommunity and has been awarded a numberof honors, including:∫ Energy Globe 2002—Best 50∫ European Solar Prize 2003—awarded by Eurosolarfor spreading RE through micro-credit∫ Best Organization Award 2005—awarded byInfrastructure Development Company, Ltd., ofBangladesh28Today there are over 160 o≈ces.grameen shakti 33

AN EXAMPLE OF AN INCOME-GENERATING ACTIVITYfrom both a cost and environmental managementperspective. Grameen Shakti, for instance, maintainsan agreement with one of its battery suppliersto take back, recycle, and adequately dispose of usedbatteries.Grameen Shakti has made strong e∑orts to raiseawareness about solar energy systems. A key marketingstrategy involved targeting the wealthier membersof a particular village; this approach was successfulin promoting a type of demand associatedwith “keeping up with the neighbors.” GrameenShakti also actively promoted the use of shs in income-generatingactivities.A mere 2.5 percent of sales were full cash remittances;most sales depended on a 36-month paymentplan. The seasonal cash flow of the economyproved to be a significant influence on customers’A shop owner in the Tangail district of Bangladesh purchased a 40Wp systemfrom Grameen Shakti. Demand for cell phone services had emerged in his villageand he saw an opportunity to provide solar-charged phone services.Business growth was significant: operations were extended by four hours aday, and the owner was able to add cell phone rentals to his product offerings.In only four months, income from the phone operations reached Tk2,000 ($30)per month—easily covering his payment installments of Tk470 ($6) a month toGrameen Shakti.ability to make payments, and the collection programwas adjusted accordingly. Most of the systemsprocured were for household application and qualityof life improvement (e.g., lighting and entertainment);however, some were used for income-generatingapplications (see box on this page for anexample).The worst flood in over a century hit Bangladeshin 1998, devastating two-thirds of the country. A full90 percent of the Grameen Shakti operating areawas flooded, and as a consequence, no sales wererecorded and defaults on collections soared, as people’sfocus shifted to the bare essentials of food,clothing, and shelter. The structure of the sme Programloan, with its two-year grace period and thearrangement of payments to be made on an annualbasis, enabled ifc’s client to remain current on itsloan, despite the delay in collections.The importance of Grameen Bank cannot beoverstated. The bank’s knowledge of the market,and its existing distribution infrastructure and clientbase, are key contributors to Grameen Shakti’s success.As was evidenced in the other solar pv investmentsunder ifc’s sme Program, a lack of localknowledge and presence is a major barrier to thesuccess of a solar pv enterprise. Furthermore,Grameen Bank was a valuable source of funding forthe firm during its early stages.moving forwardTABLE 6. GRAMEEN SHAKTI AT A GLANCE, 2007Number of villages covered 25,000Total beneficiariesMore than 700,000 peopleUnit office 227Total employees 1,135Total installation of SHS 77,000Installed power capacity 3.85 MWDaily power generation 16 MW-hrcapacityInstallation rateOver 2000 SHS/ monthInstallation of micro utility 1,000 systemsystemInstallation of biogas plant 500 (through October 2006)To date, Grameen Shakti has installed over 77,000solar pv systems, with a total installed capacity of3.85mw (a power generation capacity of 16mw perhour). This has considerably improved lives and hasprovided cleaner energy to 700,000 Bangladeshis.As rural communities have become electrified, thecompany has been able to work toward achieving itssecondary objective, that of connecting the rural areasof Bangladesh with the rest of the world throughthe service of information and communicationtechnology, as well as o∑ering computer educationand Internet access, provided by engineers at solarpoweredo≈ces. Computer education includes applications,such as Microsoft O≈ce and graphic design,as well as hardware installation and computerlanguage. This successful company has diversified itsoperations to include the construction of 500 biogasplants to provide improved energy solutions tocooking. (See Table 6 on this page for a summary ofGrameen Shakti’s achievements.)34 selling solar: part 2 / case studies

soluz honduras, s.a. de c.v.Headquartered in San Pedro Sula, Honduras, SoluzHonduras S.A. de C.V. is a subsidiary of Soluz, Inc.,USA. Soluz Honduras began operations in 1998,selling and renting small solar pv systems on a retailbasis to rural customers, who had no access to electricity.Soluz Honduras was one of the first solar pvcompanies to employ a rental model, which decreasedthe initial costs of acquiring a solar pv systemwithout having to depend on consumer subsidies,donor programs, or capital buy-downs.backgroundSoluz, Inc. was established in 1993 to further commercializesolar pv applications for rural areas,building upon the solar pv sales enterprise establishedby Richard Hansen in the Dominican Republicin 1986. Today Soluz, Inc. has two subsidiaries,Soluz Honduras and Soluz DominicanaS.A. in the Dominican Republic. 29 The solar pvrental o∑er was added to the existing cash and micro-creditsales o∑ers, first in the Dominican Republicin 1994, and later in Honduras, where SoluzHonduras had been selling solar pv on a wholesalebasis since 1994. Rolling out the solar pv rental o∑erwas a major focus aimed at penetrating rural markets.Customers were charged the equivalent of a$20 installation fee, as well as an average monthlyfee of $15 for rental and maintenance of the system(purchase of the battery was an additional cost incurredby the client). Along with rural households,Soluz Honduras sought to target small businesses,churches, schools, and health clinics, providing solarpv systems at monthly fees similar to the cost ofalternative energy sources (i.e., kerosene, dry cells,and automotive batteries). Despite an entirely ruralclient base and the devastating e∑ects of HurricaneMitch, 30 Soluz Honduras managed to attract 500solar pv rental clients in its first year of operations.By 2000, revenue totaled $144,556, with $100,499(70 percent) accounting for solar pv rental and$44,057 (30 percent) for solar pv sales.In late 2000, the sme Program approved a$400,000 loan and a $100,000 equity investmentin Soluz Honduras. The funds were to be used toexpand the Soluz customer base. With a further29Soluz Dominicana also receivedsupport from the sme Programthrough the eeaf.30Soluz Honduras was foundedjust prior to Hurricane Mitch,which devastated much ofHonduras and Nicaragua in 1998.The heavy rainfall associated withthe hurricane caused catastrophicflooding that was blamed for over11,000 deaths.31The Netherlands’ first greeninvestment fund screens its investmentsusing social and environmentalcriteria.$1 million from a co-investment transaction withthe sme Program and two other investors ($250,00each from Corporacion Financiera Ambiental,Costa Rica, and Triodos Bank’s Solar InvestmentFund 31 ), Soluz hoped to expand its solar pv rentalcustomer base to a break-even point of 2,500, a criticalstep on the way to a target of 5,000 customers.By July 2002, the number of solar pv rental customerswas stagnant at 1,500 and in danger of declining.Unexpected grid expansion and the inabilityto continue to pay for installed systems meantthat Soluz Honduras was forced to disconnect existingcustomers faster than it was adding new ones. Incontrast to the year 2000, revenue in 2002 totaled$406,772, with $227,762 (56 percent) accounting forsolar pv rental and $179,010 (44 percent) accountingfor solar pv sales. In an e∑ort to increase revenueand increase margins, Soluz increased its focus onsales (primarily by developing its dealer network).It became evident that the unsubsidized solar pvrental o∑er was not financially viable. While the upfrontcosts of installing a system decreased, themonthly charges remained too high for over 50 persoluzhonduras 35

cent of rural households. Furthermore, the solar pvrental o∑er did not succeed in creating an operatingmargin for Soluz Honduras, due to the high collectionand maintenance costs associated with such ahighly dispersed customer base. The local companythen adopted a more streamlined sales model, sellingwholesale through dealers and, thus, increasingsales and assuring their margins. The company beganto sell o∑ solar pv rental assets to meet lenderobligations and focus sta∑ time on increasing salesrevenue through an expanding dealer network. Italso continued to sell solar pv systems, on a cashand micro-credit basis, to rural customers and institutionsout of its three o≈ce locations.The Soluz business model (which included bothrental and sales o∑ers) resulted in more complicatedoperations than would typically be found in a solarpv business of equal size. The rental o∑er, in particular,required that Soluz Honduras devote considerableattention (at considerable cost) to collectionsand service in remote o∑-grid areas, where evenwell-established micro-finance institutions were unwillingto provide their financial products. Thecompany attempted to essentially build a microrentalfinancial product o∑ering in a very challengingenvironment, without the benefit of the developmentalsubsidies routinely provided to microfinanceinstitutions when establishing operations insimilar environments. All this led to high overheadand transaction costs that needed to be o∑set solelyby customer payments. An unsubsidized commercialapproach to establishing a pioneering solar pvrental operation was simply too challenging.In July 2003, Soluz found itself in severe financialdistress, with an $850,000 debt burden. When itwas clear that a change was needed for the companyto survive, investors o∑ered waivers of certain feesand payments, seeking only profitability. This suggesteda change in the business model of the company.Thus, an increased focus on sales, includingthe aggressive sale of used solar pv rental systems,was established to help meet lender payment obligationsand to build a viable operation. Managementproposed that, in the short term, Soluz Hondurascould reduce its solar pv rental fixed-asset inventorysignificantly, while at the same time increasing andstreamlining its sales e∑orts.obstaclesLike most other solar pv businesses around theglobe, the key problem faced by Soluz Honduraswas that of a∑ordability. The Soluz business modelwas designed specifically with this element in mind.In addition to cash and micro-credit sales, the companyplaced emphasis on fee-for-service or rental36 selling solar: part 2 / case studies

systems. By renting systems, customers avoided thehigh upfront costs of purchase, and the companymaintained solar pv system ownership to facilitaterepossession. The company, itself, however, had toraise significant capital to invest in solar pv unitsand to develop e≈cient micro-rental collection andservice operations in remote areas, a process that thecompany struggled with for several years. Whilemonthly fees collected from the limited number ofclients during startup simply did not o∑set thefirm’s operating costs, rollout peaked at 1,500 rentalunits, as “customer churn” reached high levels (3-5percent). The company’s financial structure was toohighly leveraged with debt for such a risky new venture.Total financing was $1.5 million, with $850,000in loans. Debt repayment obligations provided littlein the way of flexibility if the rollout of revenue andexpense did not stay to plan.Although Soluz Honduras succeeded in significantlylowering the upfront costs for customers,many found the monthly costs of renting a solar pvsystem from the company to be higher than thoseassociated with kerosene and batteries. Furthermore,when rural families lack the funds to purchasekerosene and batteries, it is not as crucial as beingobligated to pay for a rental contract for a solar pvsystem. Many households did rent a solar pv system,but later found that they were unable to continuepayments. Soluz Honduras attempted to pricerental fees to the equivalence of combined currentenergy costs (e.g., kerosene, dry cells, car batteries),but it estimated that, at this rate, it would require2,500 clients to break even. The firm instead disconnecteda significant number of customers who wereunable to make their payments, resulting not onlyin a loss of income, but also in costs related to thephysical act of disconnecting and remarketing thesolar pv system.The unpredictable timing of government ruralelectrification project timing, even when there werecommunication e∑orts with the local authorities,coupled with unexpected grid expansions, especiallydue to election promises, meant that Soluz oftenfound out about grid expansions just months beforethe grid reached a community. This required Soluzto remove hundreds of installed solar pv systems beforethe initial investment could be recovered. Fromthis, an interesting paradox emerged, further compoundingthe issue of unexpected grid expansion:residents wealthy enough to a∑ord solar pv systemsgenerally lived close to urban centers. Thus, theywere more likely to receive grid service in the nearfuture, while the available o∑-grid customer basecontinually became more dispersed and therebymore di≈cult to serve.Despite its di≈culties, Soluz Honduras was seenas a pioneer in the area of solar pv rental in the nation.The company was well-respected and had littlecompetition in the direct servicing of rural customers.When it came to larger sales, however, competitionwas strong, particularly in the areas of governmentbids and institutional sales.moving forwardOperations were further restructured in May 2005,in an e∑ort to overcome the large debt burden. Amemorandum of understanding was signed betweenSoluz Honduras and the sme Program, whereby asignificant portion of the debt would be forgiven,and a one-year repayment plan was designed for theremaining financing. Sta≈ng and all expenses werestreamlined. At this point, Soluz began to focus exclusivelyon sales, using the cash proceeds from thesale of solar pv rental assets to pay back its creditors(installed rental systems continued to be serviceduntil they were sold on short-term credit to existingor new customers).While the financial performance of Soluz Honduraswas disappointing as a result of this overlyambitious attempt to pioneer an unsubsidized solarpv rental o∑er, the environmental and social benefitsof the project should not be overlooked. To date,the company has installed well over 5,000 systems,and while the business model has proved di≈cult,the project did provide clean energy, leading to positivesocial and environmental impacts. The aggressivepenetration of a solar pv rental system createdwidespread awareness of solar pv in rural areas, thusincreasing demand. Employment opportunities andincome were provided for the micro-enterprises thatwere contracted to collect monthly fees.The move away from a rental o∑er to a cash andshort-term credit sales focus has resulted in increasedcash flows to Soluz Honduras, thus reducingits debt burden. With the reduced debt, the Honduranfirm expects to now have a viable businessmodel. Up to now, high debt on the books duringthe past three years of financial restructuring hasmade it di≈cult to purchase on credit. With a debtfreebalance sheet, however, a viable and sustainablefuture, in which the company will operate with lessrequired capital, will begin to emerge.soluz honduras 37

selco vietnam, ltd.AWARD FOR CORPORATE EXCELLENCESelco Vietnam, Ltd., based in Ho Chi Minh City, isa subsidiary of the U.S.-based Solar Electric LightCompany. Launched in 1998, Selco Vietnam soldsolar systems to unelectrified households in Vietnamand was the first 100-percent foreign-ownedcompany licensed to operate in the country.backgroundSolar Electric Light Company is a U.S.-based companywith o≈ces in India, Sri Lanka, and Vietnam.It grew out of the activities of the ngo Solar ElectricLight Fund, which was founded in 1990 to assistin the financing and installation of solar energysystem projects throughout a number of developingcountries. The for-profit Solar Electric Light Companywas launched in the late 1990s to scale up theprovision of solar electricity to households in developingcountries through a commercial marketbasedapproach. It operated through its subsidiarycompanies in India, Sri Lanka, and Vietnam (seebox below).Vietnam has a significant power shortage, withlittle capacity to meet urban demand and no infrastructureto distribute electricity beyond urban areas.The general lack of access to electricity, coupledwith the fact that even rural Vietnamese had higherdisposable incomes than those in neighboring countries,made Vietnam appear to be an attractive marketfor shs providers.Selco Vietnam’s primary focus in the country wason the sale of shs to households. However, it alsoprovided specialized applications, such as solarstreet lights, water pumps, and hot water heaters.An important part of Selco Vietnam’s businessmodel was its relationship with the vwu, throughwhich it had access to villages, and its partnershipwith the Vietnam Bank for Agriculture and RuralEstablished by the U.S. Secretary of State, the Award for Corporate Excellencerecognizes businesses that exemplify good corporate citizenship abroad.In 2001 the Solar Electric Light Company received the award for its work inreducing poverty and spurring economic development in rural Vietnam by supplyinghousehold electricity to families that lack access to the power grid.Development (vbard), which provides consumerfinancing.Selco Vietnam received a $750,000 loan fromifc’s sme Program in 1998. The loan was to enablethe company to provide collateral to third-partyfinancial intermediaries (specifically vbard) forworking capital financing up to $200,000, and tosecure loans for the company’s customers to financepurchases of shs. In addition, the loan agreementcalled for the sale of 12,000 shs in two years.obstaclesThe principal barrier faced by Selco Vietnam wasthat of a∑ordability. Vietnam had no debt cultureand, thus, no consumer financing availability, retailbanking, or home mortgage market. A general overalldistrust of the banking sector signified that mostwould rather do without than incur debt. In fact,until 2002, banks were controlled by local politicalinstitutions, and loans were approved not on creditworthinessbut, rather, on the recommendations ofthe local people’s committee, which was at oddswith the Selco Vietnam business model, since it reliedon the availability of consumer finance.Complicating the matter were the issues aroundthe mass publicity of grid expansion and the fluctuatingincomes common among the rural population,making consumers considerably price conscious.Unlike other countries, the Vietnamese didnot establish income-generating businesses based onsolar power (i.e., solar-charged cellular phonerental), a fact that also contributed to the price consciousnessof consumers.Vietnam presented a further complication: a televisionwas considered more important than lightingamong the targeted consumer group. As a result,larger solar pv systems, capable of powering a television,were in demand. In fact, most customerssought the largest and more expensive shs the companyhad on o∑er, which resulted in reducing thea∑ordability of shs even further.Most potential customers lived within 1 km of abattery-charging station. The average amount spenton batteries per month in Vietnam was the equivalentof between 66¢ and $2, considerably less than38 selling solar: part 2 / case studies

what was required for a solar pv system. The numberof those without access to electricity (60 percent)was initially taken as a measure of the prospectivemarket, but upon later review, it was determinedthat only about three percent of the unelectrifiedpopulation could actually a∑ord a solar pv system.There were also issues surrounding the policy environment.There was a heavy local politicalinfluence, yet it was not always supportive of Selco’sactivities. Seen as a U.S. company, Selco Vietnamdid not always get the same level of support that aVietnamese company might have received. Whileimport duties for solar modules and batteries werewaived for Selco Vietnam, the inconsistent subsidypolicy for electricity proved problematic.In an attempt to address the issue of consumercredit, ifc originally hoped to make the gef loan tothe women’s union, so that it could on-lend to consumers.However, the vwu was concerned with theliability issue, and ultimately the funds were loaneddirectly to Selco Vietnam, with the vwu agreeing toadminister them.Selco Vietnam’s expertise lies in the areas of solarsales and service, not in that of consumer finance.The initial intent had been to work through vbard(which would provide consumer financing withfunds guaranteed by the sme Program) and the vwu(which would handle the collections). Unfortunately,when vbard failed to make financing availableto potential Selco Vietnam customers, SelcoVietnam was forced to start providing consumerfinancing itself. Subsequently, when the vwu failedto provide proper collections service (its prioritieshaving shifted as a result of a pending election), thesolar pv company was forced to take on the addedresponsibility of a collections agent. But this provedto be a particularly challenging endeavor, as thecompany was unskilled in the areas of consumerfinance and collections, and was operating in a marketwith no debt history or consumer banking.moving forwardIn the end, Selco Vietnam was forced to accept thatdemand for lighting was considerably less than expectedin the country. During its first two years, theentity had planned on sales of 12,000 units, but itstruggled to reach 1,600. Sales targets were eventuallyrevised downward by 50 percent.In an e∑ort to lower the cost of the solar pvequipment, the company has now formed partnershipswith other Solar Electric Light Company subsidiariesaround the world in order to purchaselarger quantities from suppliers with volume discounts.Additionally, it has learned from the experiencesof other subsidiaries by sharing information.Selco Vietnam has struggled financially since itbegan operations. The management has undertakensome significant changes to improve financial performance,reducing operational and administrativeexpenses, and increasing its presence in the market.Since the end of 2006, the company has been operatingin survival mode, with only five employees focusedexclusively on cash sales. No credit is available,and the firm is not expected to make its smeProgram loan repayment deadline.selco vietnam 39

case studies 2Photovoltaic MarketTransformation InitiativeThe ifc/gef Photovoltaic Market TransformationInitiative (pvmti) is a $30-million initiative designedto accelerate the sustainable commercialization andfinancial viability of energy services, based on solarelectricity (solar pv) technology in India, Kenya,and Morocco. Funded by gef and managed by ifc,pvmti was based on the premise that private sectorproject design and financing on a commercial basiswould stimulate more sustainable ventures thangovernment or donor-financed solar pv procurementsalone. Launched in 1998, pvmti is still operationaltoday, and has committed over $18 million to12 projects.backgroundThe pvmti concept originated at a meeting held inPrinceton, New Jersey, in the early 1990s. The meetingwas attended by a large number of academics,ngos, the wbg, and others interested in the accelerationof the global market for solar pv and other retechnologies in developing countries. The originalproject concept, dubbed the "Green Carrot," wasbased on the same market transformation conceptsused by the United States Environmental ProtectionAgency’s (EPA) "Golden Carrot" program (see box,page 41). When o∑ering the “prize” was deemed unfeasible,the program was rebranded as pvmti, andmanagement of the program was passed from theWorld Bank to ifc.Although pvmti was restructured under ifc, thelessons learned from “market pull” initiatives (undertakenin North America and Europe, which employedfinancial incentives to engage the private sectorto encourage market adoption of new energyand re technologies such as the “Golden Carrot”program) continued to play a key role in the programdesign. While the initiative was originally conceivedto be a $60 million program that would involveglobal competitive procurement amongprivate sector companies, it was soon decided that itwould be more prudent to pilot the concept as asmaller $30-million initiative, targeting a smallnumber of countries.country selection processThe initial selection process began with theidentification of 35 potential countries. Of these 35countries, 30 were supplied with a summary of thepvmti program concept and a request for an expressionof interest. Of those that responded, Algeria,Argentina, Indonesia, Sri Lanka, Zambia, andZimbabwe were removed from consideration dueto planned World Bank and/or gef projects, or unfavorableeconomic conditions. The remainingcountries, Brazil, China, India, Kenya, Morocco,Pakistan, and Thailand, were visited for furtherconsultation during the first half of 1996.Three countries were ultimately selected for implementation:India, Kenya, and Morocco. Each ofthe three was considered to have an emerging solarpv market and a supportive policy environment inwhich the solar pv sector could grow. 32 In India,home to the largest solar pv market in the developingworld, pvmti was expected to stimulate investmentsin new commercial (not government-driven)sectors. In Kenya, which had a dynamic solar pvmarket with over 150,000 shs sold without any formalcredit facilities, pvmti was expected to provideworking capital and end-user financing in a marketdominated by small-scale enterprises. In Morocco,40 selling solar: part 2 / case studies

pvmti was expected to contribute to the demonstrationof the potential of private franchise models andguarantee facilities to finance alternatives to nonviablegrid extension, based on a commitment to solarpv by the national electric utility, O≈ce National del’Electricite (one).During the preliminary stage, some 25 potentialprojects totaling nearly $175 million in financingwere identified. Many of these projects were tosupport companies that sold and/or leased, distributed,installed, and serviced solar pv equipment.Other projects were focused on supporting the expansionof existing sales and distribution networksand entry into new markets, and a small numberof projects were identified working with fis to establishfinancing mechanisms to support end userpurchases.managementTHE GOLDEN CARROT PROGRAMThe Golden Carrot program is a federal program ofthe United States Environmental Protection Agencydesigned to support the commercialization of newenergy-efficient appliances in the residential sector.The “Golden Carrot” program offered a financialincentive to manufacturers to support advancesin energy efficiency. In the program, 24 utilitiespooled $30 million in the Super Efficient RefrigeratorProgram. That program then held a contest,and the manufacturer that built the most efficientCFC-free refrigerator at the lowest cost wasawarded guaranteed rebates from the pool to offsetthe incremental product costs. At the sametime, consumers got more affordable and environmentallyfriendly refrigerators.During the country assessment period, it becameapparent that a much higher level of engagementwould be needed over a longer period of time thanoriginally planned. Sector and country expertise,similar to that of an investment fund, was importantif pvmti were to succeed. Rather than establishin-house expertise, ifc management chose to selectan external manager. ifc decided to retain ImpaxCapital Corporation (now Impax Asset Management,Ltd.) 33 and IT Power, Ltd., 34 both firms experiencedin managing small, innovative renewableenergy portfolios, to serve as an external managementteam (emt). The emt also included local partnerorganizations in each of the countries—ITPower India (a subsidiary of IT Power UK), PipalLtd. (a Kenyan project management company inthe solar field), and resing (a Moroccan projectmanagement company)—which provided criticalsupport and local knowledge at all stages of theprocess.The role of the emt was to solicit, screen, andstructure solar pv business proposals and conductthe appropriate commercial, technical, and financialdue diligence. Once the emt felt it had a solid project,it would present the project to the ifc InvestmentReview Committee (irc) for final approval.After approval from the irc, the emt would be responsiblefor ensuring that all disbursement conditionswere met, as well as for reviewing progress ona regular basis subsequent to disbursement. Theemt is compensated in two forms: a set projectmanagement fee paid quarterly over the life of theprogram, and a series of performance bonuses, someof which are tied to performance and some of whichare tied to timing in the program. The set projectmanagement fees make up the majority of the emtcompensation.structurepvmti was financed with $30 million from gef,allocated as follows: $15 million for project financingin India, $5 million each for Kenya andMorocco, and the remaining $5 million to be usedfor technical assistance and project execution. Thetechnical assistance component of the programamounted to $3 million, or 10 percent of totalfunds. These funds were to be provided for noncommercialpurposes, such as technical assistance,training, the development of standards, and additionaluses as required by individual projects. Itwas expected that $13.5 million would be recoveredfrom investments and portfolio earnings, andwould be returned to gef at the end of pvmti’s operationalperiod.The financing terms o∑ered by pvmti’s were designedto be su≈ciently flexible to respond to theneeds of each project, and included debt, guarantees,and grant funding. Most investments were expectedto request debt at or below market terms.Additional financing tools, including partial guaranteesand equity, were also available, if the benefits ofsuch tools were deemed su≈cient to justify the increasedcomplexity.32Each country had, at the time,a large number of households ino∑-grid areas, an adequatefinancial services sector, and anexisting solar pv sales market(either subsidized or unsubsidized).33Impax Asset Management, Ltd.,is a specialist fund managementcompany focused on the environmentalsector. With £420m ($838million) in funds under managementin a combination of listedand private equity, Impax had thenecessary skills to assess potentialinvestments and to implement thesame.34IT Power, Ltd., is an energyconsulting firm with a specializationin re engineering and relatedeconomic, financial, commercial,and environmental considerations.IT Power has completedover 1,000 projects for both governmentand private sector clientsin over 100 countries.photovoltaic market transformation initiative 41

35It is important to note thatprojects approved through the ircare documented by the ifc legalteam. In some cases, projects wereapproved by the irc, but legaldocumentation was not finalizeddue to a variety of reasons.restructuringTABLE 7: PVMTI’S ACTIVE SOLAR PV PROJECT PORTFOLIOIn 2004, pvmti underwent a significant restructuring.The slow execution of deals in the early years ofthe pvmti program, due to the extensive documentation,minimum investment size, and long negotiationperiods, resulted in a lower-than-expected disbursementrate. With funds not being disbursed,the expected reflows (interest and principal payments)did not accumulate as expected, and pvmtiran into issues with insu≈cient cash flow to coverits ongoing administrative and operating costs. Therestructuring sought to extend the program implementationperiod by two years (from 10 years to 12),in an e∑ort to bring disbursements of committedfunds to approximately 70 percent, and to reclassify$1 million of pvmti’s investment funds to financethe cash shortfall on the administrative side of theproject.As part of the 2004 restructuring, pvmti alsoreceived approval for a grant of approximately$350,000 for a stand-alone technical assistancecapacity-building program in Kenya. This projectconsists of the development of training curriculumand the provision of training to solar pv technicians,creation of quality awareness in the market, establishmentof a quality assurance program for shs inthe Kenyan market, and provision of support tothe Kenyan Renewable Energy Association (krea).The total cost for these activities is estimated at$476,900, which includes about $115,000 of in-kindcontributions and co-financing for the project. Thegef funds used for this are all grant funds.COMMITMENTUNITS INSTALLEDPROJECT (IN MILLIONS) (APPROXIMATELY)Selco India 1.10 15,000Eskom-Shell Solar Home Systems 3.90 26,000Shri Shakti 2.23 2,000SREI Infrastructure Finance, Ltd. 3.50 15,000Total India 10.73 58,000Barclays Bank, Kenya 2.00 0Equity Building Society (EBS) 2.10 0Muramati Tea Growers SACCO 0.60 170Total Kenya 4.70 170Salafin S.A. 1.00 0Sunlight Power Maroc S.A. 1.075 1,700Total Morocco 2.075 1,700Total PVMTI 17.505 59,870In October 2006, pvmti sought, and was granted,approval from gef to increase the funds available fortechnical assistance, from a maximum of 10 percentof the overall program funds to a maximum of 20percent. This request was made in response to thefindings of the Mid-Term Review that suggested ifcshould explore the possibility of deploying uncommittedinvestment resources to grant-oriented technicalassistance activities, such as training, solar pvinformation dissemination, and capacity buildingfor the industry sector that will help advance theoverall objective of pvmti. This was suggested, inpart, because pvmti was one of the earliest markettransformation initiatives that ifc had managed, andlessons from subsequent projects seemed applicableto pvmti at the mid-term point. Since pvmti’s inception,ifc has managed a number of other programsthat have resulted in greater market impact thanpvmti is likely to have at the end of its life. This ismainly the result of a larger emphasis on technicalassistance and capacity building (as a portion ofoverall investments). The change in October 2006sought to correct this imbalance, while there wasstill ample time within the pvmti program to implementnew technical assistance and capacity building.performanceTo date, the irc has approved a total of 16 sub-projects(six in India, six in Kenya, and four in Morocco).35 The active pvmti portfolio is comprised ofnine projects with commitments of over $17.5 million(see Table 7 at left).Experience to date with the di∑erent projects hasbeen mixed, with some proving very successful andsome unable to make any progress due to a varietyof outside reasons. This report has focused on MuramatiDistrict Tea Growers Savings & Credit CooperativeLtd. (Muramati), srei Infrastructure Finance,Ltd. (srei), and Sunlight Power Maroc S.A. (spm).These projects were selected to be representative ofthe overall pvmti portfolio, as they represent projectsin each of the countries in which pvmti was active,projects involving both financial intermediariesand direct investments with pv companies, and projectsthat achieved varying degrees of success.42 selling solar: part 2 / case studies

Capacity Building and Technical Assistance May BeMore Important than Business FinanceAt the time pvmti was introduced, and for severalyears thereafter, the Kenyan market was not preparedfor the financial product and services thatpvmti o∑ered. The minimum deal sizes were toolarge for existing solar pv firms, and larger entities,such as fis, were not interested in pursing the ruralsolar pv market. Quality solar pv products and a reliablesolar pv service network were also lacking inthis market. In recognition of this, pvmti directedits e∑orts at providing technical assistance to raisepublic awareness of the merits of solar pv, upgradethe skills of local technicians, and foster an enablingenvironment for the establishment of high-qualitysolar pv products and service providers.Enabling Environment Is Criticalwhat worked and what did notThe pvmti experience was similar to that of othersolar pv projects undertaken by ifc and others, inthat it highlighted that solar pv projects are mostchallenging to implement, precisely in those marketswhere the demand for it, and the economicjustification for it, might be greatest. Often, rural,poor, and sparsely dispersed communities, who arefar from the grid and thus need solar pv, are unlikelyto generate the resources necessary for purchasingor maintaining these units without extensivesubsidies. The pvmti experience alsodemonstrates that there is an ongoing need for capacitybuilding and technical assistance, that investmentterms and management of solar pv focusedprojects should be tailored to the specific needs ofsolar pv, and that product quality is a serious issue.Success in the solar pv business, and the appropriatebusiness model to adopt, will depend to a large degreeon the enabling environment in which the firmoperates. India has the largest re financing e∑orto∑ered by any developing country. Governmentale∑orts to promote re, including solar pv, competewith pvmti, but also help open up the market andestablish solar pv as a viable technology. Additionally,the fact that the population of India is largeand densely populated means that service technicianscan economically serve a small geographicalarea (relatively inexpensive to reach potentialclients) with a critical mass of shs units. Furthermore,favorable tax, regulatory, and grid-extensionpolicies may help the development of the solar pvmarket in a given country.Product Quality Standards Are also CriticalMany of pvmti’s investments found the lack ofproduct quality standards to be detrimental to theiroperations. Muramati saw systems fail and installationsdelayed as a result of faulty batteries. spm sawincreased pressures on prices as a result of cheapercontraband product on the Moroccan market. Inhindsight, pvmti should have been more proactivein improving product quality and establishing quality-controlmechanisms. A portion of the grantcomponent would have been well spent investing inproduct innovation and quality control.IFC’s Typical Project Financing Requirements AreIll-Suited to Small Business Transactionsifc’s legal documentation and loan security docuphotovoltaicmarket transformation initiative 43

ments are suited to large project finance transactions.They can be extremely burdensome and timeconsuming for smes that are more accustomed tomuch simpler due diligence processes. As a consequence,investment transactions take months oryears to complete and, in some cases, market conditionswill change significantly between the investmentapproval and financial closure time frame. Afurther consequence is that the administrative costsare high in relation to investment size.Many solar pv businesses in the target countriesfound the $500,000 minimum investment to be toolarge. This was particularly true in Kenya, where investmentswere limited to fis and banks. Goingthrough the banks, however, proved to be cumbersomeand time consuming, since the banks did notsee financing shs as a main line of business, and itwas di≈cult to get many of them to move expeditiouslyon the projects.Furthermore, the small businesses and entrepreneurstargeted by pvmti found the extensive businessplans and other documentation required to be somewhatdaunting. While they had energy and ideas,many were not skilled in the writing of businessplans. This resulted in long negotiation periods forcustomized contracts. In some instances it took ayear from the date of review to the date of disbursal.Dedication to Solar PV and Provision of Value-Added Services Are Critical to SuccessAll of the firms that achieved modest success interms of utilizing pvmti resources and drawingdown their commitments were already in the solarpv business, or seeking to enter the business, whenthey received funding from ifc. pvmti found thatfirms that provided further value added, in particularservicing and maintenance, were more successful.Those who moved farther up the value chain,and were involved in the assembly of solar componentsand the installation of systems, seemed to dosignificantly better than firms that were merely engagedin consumer or producer financing.Firms that received pvmti financing that did nothave a particular focus on solar pv were significantlyless successful. Muramati, for example, wasdedicated to providing financing to people workingin the Kenyan tea sector, not to promoting solarpv. As a result, the financing of shs fell outside thecore business line, and proper resources were notdedicated.Decision Making Needs to Be Done by Those Closestto the ProjectThe pace of decision making was hindered by theadministrative structure adopted in this project. Alldecisions regarding investment commitment, loanclosure, disbursements, and acceptability of loancollateral were made by ifc sta∑ (Legal Departmentand Environment and Social Development Department)upon the recommendation of the emt. Thisstructure has resulted in significant delays in the administrationprocess, as those closest to the projects(the emt) were not making the decisions.conclusionpvmti has experienced considerably more success inIndia than in Kenya or Morocco. This success canbe attributed, in large part, to the high populationdensity in o∑-grid areas, the existence of establishedsolar pv companies, and the relatively widespreadknowledge about solar pv technology. In Kenyapvmti had initially set out to provide working capitaland end-user financing. However, the focus hassince shifted to providing more technical assistancefunding, particularly in the areas of training andquality assurance. In Morocco pvmti has continuedto rely on support from the national utility, one.Given that pvmti is still an operational project, itis di≈cult to come to any conclusions as to its overallperformance. While the program started slowly,disbursements have increased significantly since the2004 restructuring. With over $12 million in disbursements,pvmti is currently on track to meet therevised disbursement goals.pvmti was able to provide financing for a numberof businesses that otherwise would not have beenavailable. As a result of this financing, over 60,000previously unelectrified households now have electricity.The Mid-Term Review, which was completedin July 2006, noted that pvmti will be responsiblefor the displacement of an estimated109,466 tonnes of CO2 emissions over the lifetimeof the shs installed.44 selling solar: part 2 / case studies

muramati district tea growerssavings and credit cooperative, ltd.The Muramati District Tea Growers Savings andCredit Cooperative, Ltd. (Muramati), based inKenya, was approved for financing from pvmti inJune 2000. The funding received was to support theintroduction of a loan scheme to finance Muramatimembers in the purchase of shs.backgroundFounded in 1993, Muramati has grown to be one ofthe largest savings and credit cooperatives servingthe tea sector, with a current membership of over32,000. Muramati’s primary purpose is to providebasic savings and lending services to those involvedin the tea sector in Kenya.The concept for which the pvmti funds werededicated was relatively simple. Muramati wouldwork in partnership with a local shs supplier, whichwould supply and maintain the systems, while Muramatiwould market the systems and provide potentialcustomers with the financing to purchase thesystems. The customer would be required to pay adeposit on the system and maintain monthly payments.obstaclesProgress on the project was initially quite slow, asboth Muramati and the shs supplier seemed to bewaiting for the other party to drive progress. Underthe initial agreement, the shs supplier had agreed toestablish infrastructure in Muramati’s regions to servicethe shs installed through the project, however,given the low volume of shs orders, they were unwillingto make this investment and provided servicefrom their headquarters in Nairobi, four hoursaway. This resulted in delays in installing new shs,as the supplier would only install in batches, as wellas in delays in responding to maintenance calls. Afurther issue resulted when the supplier received afaulty batch of batteries, causing a number of systemfailures. Finally, there was a problem with thepricing systems. Muramati was assuming most ofthe financial risk in the arrangement, yet with teaprices having been stagnant over the past severalyears, tea growers were particularly aware of whatconstituted value for money, and the supplier’s systemswere seen as being particularly expensive.An additional reason for the slow initial progressof the project was rooted in that Muramati only hadto pay interest on drawn funds and, therefore, wasin no great rush to proceed with the disbursementsfrom ifc. Given this slow initial progress, pvmti hashad to restructure the terms of the Muramati loan,delaying the second and third disbursements, aswell as the overall repayment schedule.The credit terms initially o∑ered by Muramatiproved to be problematic. The short-term loan ofup to 18 months proved to be far too short to ensurea∑ordable repayment installments. Similarly, the 50percent down-payment requirement proved toomuch for many potential borrowers, and the interestrates set at 15 percent per annum were consideredtoo high.No market survey, unfortunately, had been undertakenduring the preliminary stages of the promuramatidistrict tea growers savings and credit cooperative 45

36ifc required that the disbursedfunds to the Muramati project beheld in a joint account (ifc andMuramati) at Barclays Bank inKenya. Muramati was to drawdown funds from this accountonly on approval from ifc. Thisprocess meant that Muramati hadtwo steps to receive funds: 1)request a disbursement from ifcto the ifc/Muramati account, and2) request approval of withdrawalsfrom the jointly held account.This process could take fourweeks or more.ject to determine the level of demand for shsamong Muramati members. Over the course of theprogram’s implementation, it became evident thatmany Muramati members were taking out loans topurchase shs on the open market, where they wereavailable for a competitive price. While there wasindeed a demand for shs, this did highlight the factthat the driving factor in purchasing decisions wasprice, not quality, thus leaving the systems o∑eredby Muramati at a distinct disadvantage.At the time of financial closure, there were concernsabout weak finances and internal controls. Asa result, a rather cumbersome system of accountswas established to manage the flow of funds. 36 Today,Muramati continues to expand its membershipand is in good financial health, despite the di≈cultconditions in the tea sector in recent years.moving forwardMuramati eventually did engage a new supplier toprovide systems at considerably lower prices thanthe original shs supplier for the project. The issueof maintenance was addressed through the use ofgrant funds from pvmti, used to train local freelancetechnicians to perform ongoing maintenance ofshs. Credit terms o∑ered by the cooperation werechanged; the maturity of loans was increased tothree years with a down payment no longer required;and interest rates were reduced to 12 percentper annum. These revised terms were, at the time,considered to be much more attractive to potentialborrowers.However, despite the above initiatives, the supplierpartnership has failed to properly materialize,and Muramati no longer engages in such partnerships.Until 2006, Muramati continued to providefinancing to its members to purchase shs, althoughmembers were free to select the system and installerof their choice. During the last four years, Muramatihas attempted to undertake marketing campaignsfor shs awareness to its entire membership.The results of this strategy have been limited. In late2006, Muramati asked to halt its pvmti program,and it has fully repaid its outstanding loan and unusedgrant funds. The company felt that the programwas too cumbersome to manage and that itwas ultimately outside its core business.46 selling solar: part 2 / case studies

srei infrastructurefinance, ltd.srei Infrastructure Finance, Ltd. (formerly srei InternationalFinance, Ltd.), is among the largest nonbankingfis in India. srei is engaged in the financingof construction and mining equipment, infrastructureprojects, and renewable energy systems. In February2001, srei received financing from pvmti toaddress two key issues facing the Indian solar pv industry,namely, the lack of after-sales services andmaintenance activities, and the lack of rural creditmechanisms.backgroundAlthough India has one of the world’s largest solarenergy programs, it still su∑ers from the same issuesthat face solar pv markets around the world: a∑ordabilityand reliability of technology. The srei projectsought to address these issues by developing afinancial model to provide unelectrified householdswith easy access to credit facilities in order to beable to access lighting options, and developing anetwork of solar service centers in the rural areas bybuilding on the existing infrastructure of systemsintegrators.The project involved a partnership between srei(providing management and a financing mechanismfor rural credit), Tata BP Solar India (India’s largestsolar pv cell and module manufacturer, which providedthe shs), The Ramakrishna Mission (an ngowith solar electrification experience and contactswith rural communities), and the Tata Energy ResearchInstitute (teri) (which brought project managementand quality assurance experience).obstaclesDuring the first few years of the project, implementationwas impeded by significant disputes betweenthe partners. The disputes centered aroundthe initial anticipation that srei would use loanfunds to establish The Ramakrishna Mission infrastructurein rural areas in order to install andmaintain shs. This endeavor never materialized,and eventually the partnership was dissolved.srei hoped to simplify operations and reducecosts by establishing a one-stop location where consumercredit, shs and spare parts sales, and aftersalesservice and maintenance were fully integrated.The concept, unfortunately, did not materialize,due to the reluctance of srei to establish rural infrastructure.Initial attempts to establish a rural credit mechanismwere unsuccessful. srei was reluctant to takeon rural consumer credit risk, despite its partialguarantee. During the early years of operation, saleswere almost 100 percent cash and carry, and the issuewas addressed by srei providing supplier credit,establishing partnerships with rural banks for creditand, eventually, by the provision of consumer loans.moving forwardToday, srei continues to source solar modules fromTata BP in India, and is working with a new ruralelectrification service provider, Environ Energy-TechService, Ltd. (eets). This new partnership has beenin place for over two years and is considered to beprogressing well.pvmti financing made it possible to provide eetswith working capital loans, enabling eets to introduceseveral small innovations to enhance customersatisfaction, such as theft insurance, damaged partsreplacement, regular visits by field technicians, andfive years of free service.The srei experience presents an interesting casestudy. srei focused on some of the more challengingareas of India, initially experiencing a great deal ofdi≈culty. Currently, however, srei performs well—the project has expanded beyond pvmti to otherprojects—and it has installed over 15,000 shs. Itsexperience highlights the need for patience, particularlyin a challenging solar pv market.srei infrastructure finance. ltd. 47

sunlight power maroc s.a.Sunlight Power Maroc S.A. (spm), in Morocco, receivedpvmti funding in December 2004. The fundingwas requested to finance infrastructure expansionand working capital requirements for afee-for-service project, as well as to create a newcredit business for solar pv sales.backgroundspm was originally founded in 1998 to provide installationand maintenance of shs in Morocco (primarilyin the northern regions of Taza, Swfrou,Taounate, and El Khemisset). The original businessmodel was based on a nonsubsidized fee-for-servicerental scheme for shs ranging from 20 to 80 wp.spm would maintain ownership of the systems, andthe households would pay a deposit followed bymonthly fees, depending on the size of the system.In early 2004, spm signed an agreement withO≈ce National de l’Electricité (one), to provideshs under a subsidized fee-for-service scheme. Theone scheme was established to provide solar pv electricityto the 15 percent of rural households (approximately300,000) that are not targeted for grid connection.The scheme, established in 2002, hasawarded six contracts for a total of 112,000 shs.Under the terms of the agreement, spm was givenexclusivity in the regions in which it was to installthe one systems, and it was provided with a sevenand-a-half-yeartime frame to complete the installation.one pays spm an upfront subsidy, and spm undertakesthe maintenance and repair of the systemover a 10-year period, in return for a monthly feepaid by the end user.When spm applied for financing from pvmti, thecompany had insu≈cient capital to meet leveragerequirements. To overcome this obstacle, ifc consideredhistorical equity contributions as cofinance tomeet minimum leverage requirements, and pvmtiwas able to recognize historical shareholder contributionsthat had funded the business prior to pvmtiinvolvement. As a result, no new cofinancing wasrequired for the project. As an exceptional case inthe pvmti portfolio, special approval had to begranted.obstaclesA number of market issues impacted negatively onspm’s initial success. Firstly, significant grid expansionhas taken place in recent years, shrinking thepotential market for recipients under the onescheme. Secondly, one has significantly increased itssubsidized fee-for-service scheme (which spmbenefits from), making credit schemes less desirablefor the public and impacting sales. Thirdly, there isincreased availability of cheap contraband solar pvmodules in Morocco, putting pressure on prices andmaking potential sales margins significantly narrower.spm had a similar experience to other pvmti projectsrelating to meeting the disbursement requirements.The cumbersome documentation process resultedin a two-and-a-half-year delay from ircapproval to financial closure.moving forwardDespite the above obstacles, spm has madesignificant progress under pvmti, and continues toincrease installations under the one scheme. spm hasattributed its ability to expand and establish threenew service centers to pvmti funding. It currentlyhas installed approximately 6,000 shs.spm has not yet managed to establish a creditbusiness, as it has focused e∑orts on establishing theone business. Furthermore, spm argues that consumerscannot yet a∑ord the monthly credit paymentsand, therefore, it has not put forth the necessarye∑ort to establish a credit business. Currently,no further e∑orts are being made to establish acredit scheme, as the company maintains its focuson installations under the one scheme.48 selling solar: part 2 / case studies

case studies 3Solar Development GroupThe Solar Development Group (sdg), a $41-millioninitiative, was created with the goal of increasing thedelivery of shs to rural households in developingcountries. Comprised of two separate entities, SolarDevelopment Capital (sdc) and the Solar DevelopmentFoundation (sdf), sdg provided financing toprivate sector companies involved in rural solar pvactivities in developing countries, as well as grantsfor business development services.backgroundMotivated by the enthusiasm about the potentialfor rural solar pv electrification exhibited by thoseinvolved in the solar pv industry during the mid-1990s, the charitable foundation community in theUnited States saw an opportunity to engage withthe wbg. In early 1996, a letter was sent from theRockefeller Foundation (on behalf of a number ofmembers of the charitable foundation community)to Jim Wolfensohn, then president of the wbg,proposing the creation of a solar energy investmentsubsidiary.The proposal called for the injection of “massive”amounts of money to launch the emerging marketsolar pv industry, through the development of aninvestment vehicle that would dramatically expandfinancing for commercial companies, so that theycould develop and provide rural energy services.The vision was grand, and at the earliest stage ofconcept development, investing up to $1 billion tocatalyze the solar pv market was discussed.By the end of 1996, a significantly scaled-downconcept paper for a $50-million “Solar DevelopmentCorporation” was being circulated. ifc was broughtin to work on the wbg-led initiative, given its experiencein private sector project finance. An externalconsultant, contracted in July of 1997 to develop afeasibility study and business plan, found that thesolar pv market showed tremendous potential, andidentified over 100 investment opportunities. InMarch 1999, Triodos Solar pv Partners 37 was appointedas advisor, and the fundraising and businessplanning process began.Originally envisioned as a single entity withboth financing and technical assistance components,the sdg, as the final initiative was named, consistedof two separate but closely related and supportiveentities, sdf and sdc. sdf was initiallyproposed as a $19.5 million ngo, o∑ering businessdevelopment and seed financing in the range of$10,000– $100,000 to assist solar pv companies inpreparing for private investment. sdc was envisionedas a $32 million for-profit private equityfund, providing growth capital in the range of$100,000–$2,000,000 for private solar pv and solarpv-related businesses in developing countries (seeTable 8 for an overview of sdg, page 50).sdg, through its two separate entities, in e∑ect respondedto two perceived problems in the market.sdc was designed in response to a view that therewas a cost-e∑ective business over the horizon, providedeconomies of scale could be achieved throughhigher volumes and greater commercial returnscould be realized through lower unit costs. In turn,sdf was designed to respond to the solar pv market(with high costs and underfunded entrepreneurs)requiring more of a nonprofit model, which becameknown as a patient capital approach. It soon becameevident that it was impossible to address both issuesunder one framework.The goal was to raise a total of $50 million from a37A nonprofit organization,Triodos Solar pv Partners wasformed by three organizations:Triodos International FundManagement (part of the TriodosBank Group), EnvironmentalEnterprises Assistance Fund(eeaf), and solar pv sector expertsGT Consulting, Inc. (gtc), ajoint venture of Soluz, Inc. andEnersol Associates, Inc.solar development group 49

TABLE 8: SOLAR DEVELOPMENT GROUP AT A GLANCESDFSDCObjective Help PV companies prepare for Provision of capital to PV-related SMEs,private investmentESCOs, banks, microfinance institutions andleasing companiesCompany Type Not-for-Profit For-profit private equity fund with venturecapital elementsNumber of Target: 75 enterprises in first five years Target: 27 investments in 15 companiesInvestments Made Achieved: Commitments totaling over totaling $18 million, including eight loans to$3.5 million in 63 projects (54 companies) FIs totaling $10 million$2.2 million disbursed by early 2004 Achieved: Six investments ($3.9 million)approved; three investments disbursed,totaling $650,000Total Capitalization Target: $19.5 million Target: $32 millionAchieved: $12 millionAchieved: $28.7 millionGeographical Scope Target: Global Target: GlobalAchieved: 23 countriesAchieved: Three countriesType of Investment Vehicles Loans (up to four years at 0 –10 percent Minority position capital investmentsinterest in local or U.S. currency)Provision of additional debt/quasi debtGuarantees (to facilitate local bank lending)Technical Assistance GrantsDate Began Operations March 2000 April 2001Date Closed and Managing March 2004 April 2004Agency Appointed Operations transferred to the Triodos Assets sold to TRED Fund, thereby liquidatingRenewable Energy for Development (TRED) the fundFund38Directorate General for InternationalCooperation (dgis) ofthe Netherlands Ministry ofForeign A∑airs, ifc, Swiss StateSecretariat for Economic A∑airs(seco), and wbg.39Cordaid, Joyce Mertz-GilmoreFoundation, Rockefeller Foundation,Rockefeller Brothers Fund,and Stichting Triodos-DoenFoundation.40gef, ifc, seco.41Cordaid and EnvironmentalEnterprise Assistance Fund(eeaf).42Calvert World Values InternationalEquity Fund, RabobankFoundation, and Triodos BankGroup.consortium of investors and donors for sdg. Ultimately,$41 million was raised from private sectorsolar pv businesses, individual private investors, ngos,multilateral organizations, bilateral organizations,and various socially responsible investment (sri)funds. sdf was supported, in large part, by multilateraland bilateral organizations, 38 as well as the charitableand ngo community. 39 (See Table 9 for detailon sdf shareholders, page 51.) sdc also received multilateraland bilateral support, 40 as well as supportfrom AstroPower (now GE Solar), a private individual,ngos, 41 and a number of sri Funds. 42 (See Table10 for detail on sdc shareholders, page 51.)Each entity (sdc and sdf) was established withits own mandate and separate board of directors.The separation of the two entities was necessary, aspreviously mentioned, not only to address the twokey perceived issues in the market, but also sincesdg had raised funds from foundations based in theUnited States. These foundations enjoy a tax-exemptstatus, provided they do not engage in profit-makingactivities, such as venture capital. As such, thecharitable foundations made up much of the shareholderbase for sdf, the ngo charged with the marketdevelopment activities and riskier seed capitalinitiatives, while the sri funds were focused on bysdc, the for-profit venture capital fund.implementationsdf began operations in early 2000 with $12 millionin commitments, and soon approved its first transactions.sdc concluded its fundraising phase inApril 2001 at $29 million, and approved its first investmentfive months later.It soon was clear that the investment opportunitiesidentified in the feasibility study had beengrossly overstated; in fact, not one of the over 100opportunities identified in the feasibility study ultimatelyreceived support from sdc. There was a ma-50 selling solar: part 2 / case studies

jor disconnect between where those involved in theinitial structuring of sdg felt the market was andwhere it actually was. The market was simply notripe for equity investments, the market assessmenthaving overestimated the maturity of the solar pvmarket and the number of business opportunities.Furthermore, market conditions were changing: theLatin American and East Asian financial crises andthe 9/11 attacks had an impact on emerging marketeconomies. Neither had the solar pv market movedas expected; in fact, the cost of solar pv had actuallyincreased rather than decreased, while the increaseddemand in the developed world was shifting the attentionof manufacturers away from the developingworld. In 2002, less than a year after sdc had begunoperations, Triodos International Fund Managementindicated that the existing investment guidelineswere unrealistic, given the nature of the market,and that new investment guidelines wereneeded. These issues, consequently, resulted inlower return expectations.Discussions on restructuring share ownershipwithin sdc began to take place. While restructuringwas necessary, the conflicting interests of the di∑erentshareholders were making it di≈cult to reachconsensus. Those shareholders, who looked at theirinvolvement more from a profitability standpoint,felt that the investment fund had no future andshould be closed. In contrast, those, such as ifc’sEnvironment and Social Development Department,which had a mandate to provide innovative projectfinancing, felt that sdc should be restructured.A revised implementation plan to expand thenumber of financing instruments o∑ered by sdcSUNLABOB RURAL ENERGY SYSTEM CO., LTD.SDG had an impact on the solar PV market beyondthe provision of financing. SDF developed a consumerfinancing handbook that provided modelingtools and guidelines for the assessment of creditrisk, which has proved effective.Sunlabob Rural Energy System Co., Ltd., usedthis handbook in the development of its businessplan for providing affordable and reliable solar energythrough rental services in Lao PDR.The World Bank awarded Sunlabob the DevelopmentMarketplace Award in 2005 for its work indeveloping a rental system that makes solar electricityaffordable for the majority of rural householdsor villages without the use of subsidies.TABLE 9. SDF SHAREHOLDERSSHAREHOLDERPERCENTWorld Bank 45.2International Finance Corporation* 4.8Global Environment Facility** 2.0Others 48.0*IFC–Environment and Social Development Department.**Represented by IFC’s Environment and Social Development Department.TABLE 10. SDC SHARREHOLDERSNUMBER OFNUMBER OFSHAREHOLDER A SHARES C SHARES PERCENTGlobal Environment Facility* — 10,000 34.8International Finance Corporation** 3,000 2,500 19.1Others 11,500 1,750 46.0*Represented by IFC’s Environment and Social Development Department.**IFC’s Infrastructure Department.and to lower the return requirements was presentedto and approved by a majority of the sdc board.The plan required larger shareholders buying outthe smaller shareholders. Despite the restructuringe∑ort, however, there were not enough viable investmentopportunities, and in June 2004, only threeyears after it began operations, sdc was disbandedas a legal entity. Assets were sold to the TriodosRenewable Energy for Development (tred) Fund.At the time of sale, 13 percent ($3.6 million) of the$29 million in funds committed to sdc had beencalled. Of this amount, $650,000 was disbursed toinvestments; the remainder went to operational anddeal-related expenses.Amid discussions relating to the restructuring ofsdc, the sdf Board of Directors solicited proposalsto manage the ngo in the event sdc were to fail. Inearly 2004, sdf transferred operations to the tredFund.performance and outcomesThe sdg initiative began in 1996 and ended in2004. During those eight years, $2.85 million wasdisbursed to solar pv projects in over 20 countries.Though not quite the billion-dollar order-ofmagnitudeinitiative that was initially called for, sdgwas certainly an experience that has provided manyvaluable lessons for the future.sdf, with its flexible, less risk-averse, and morea∑ordable funding, was able to meet most of its investmentobjectives. sdf financing was provided tosolar development group 51

43While perhaps not a directcause of sdg’s lack of profitabilityor overall success, the diversity ofsdg shareholders forced a managementstructure that was notonly cumbersome, but one inwhich the intended relationshipsdid not materialize.help create a more enabling environment, increasethe amount of consumer finance available, supportenterprise growth, and support innovation withinthe industry. Following the beliefs that the challengesfacing the market were more related to marketingand consumer financing than manufacturingand wholesale distribution, that retail operationsmore directly benefit the rural people, and thatstimulating demand at the retail level would benefitthe entire industry, retail distribution made up 80percent of the sdf portfolio.While sdf certainly had a positive impact on thesolar pv industry, the foundation was insu≈cientlysizable to produce the large-scale change that sdginvestors had envisioned and desired. Furthermore,sdf had been envisioned to be an entity to preparecompanies for sdc investment. This simply did notoccur.sdc was to have made 28 investments over a10-year period. Between 2001 and 2003, only six investmentswere approved by the sdc Board of Investors,totaling $3.9 million. Only $650,000 wasultimately disbursed to three countries: Kenya (investmentwent bankrupt in early 2005), Indonesia(investment terminated prematurely subsequent tothe cancellation of the World Bank solar pv subsidyprogram in the country), and Bolivia (investmenthas been moderately successful, although only halfof the approved funds have been disbursed). Whilethe sdc management took its fiduciary responsibilityvery seriously, the market was as yet unpreparedfor the equity investments sdc was looking to makeand, thus, it decided not to invest in unsatisfactorydeals. sdc ultimately was able to return money tosome of its investors upon project closure.Many of sdg’s shareholders had environmentalmandates that served as motivation for their participationin the initiative. Solar pv has been trumpetedfor many reasons, but one of the primary argumentshas consistently been the positive environmentalimpact achieved through the implementation of retechnologies and the resulting reduction in CO2emissions. Measuring the environmental impact hasproved di≈cult, however. In the case of sdg, fewshs systems were actually installed as a result of sdginvestment, and it is therefore safe to assume thatthe environmental impact of sdg was negligible.Furthermore, sdg never set out to determine thenumber of solar pv systems installed as a result ofsdg support; therefore, it would be impossible todetermine the actual amount of CO2 emissions thatwere displaced.what worked and what did notShareholder Diversity Proved ProblematicAlthough the diversity of the di∑erent shareholdersin sdg had been celebrated during the fundraisingstage, it rapidly became apparent during implementationthat it would be di≈cult to manage the variousinterests of such a diverse shareholder group.The separation of sdc and sdf had been designed toreflect the need to satisfy the di∑erent objectives ofthe for-profit investors, ngos, and foundations (15in total). With so diverse a composition (multilaterals,ngos, sris, private individuals), sdg found itnext to impossible to satisfy everyone’s needs.This is perhaps one of the greatest (non-marketrelated)reasons for the limited success of sdg, an issuefrom the very beginning. 43 The initial structureof sdg was designed so that sdf and sdc would providecomplementary services; sdf would “prepare”companies for sdc investment. Yet, with each entityhaving a separate board of directors, each wound upwith its own mandate, and not one single sdf investmentgraduated to sdc. In fact, sdf transcendedon the sdc project pipeline, evolving into more of asoft financing vehicle, providing working capitalrather than providing the seed capital and marketdevelopment assistance originally intended. When itbecame apparent that the structure and investmentguidelines were inappropriate, it proved impossibleto reach consensus among the shareholders on therestructuring, despite a year of attempts.Strategic Alliances Were Not DevelopedIt is interesting to note that while the diversity ofshareholders was di≈cult to manage, it also couldhave been used to advantage. Strategic alliancescould have been developed along a number of lines.In fact, most potential investees expressed that theywere equally interested in the contacts of sdg. Mostsolar pv companies are smes, often family owned,and have grown without the benefit of technicaltraining. The opportunity to consult with solar pvsector experts was very attractive to sdg investees;however, this type of relationship between sdgshareholders and sdg investees did not materialize,given that sdg shareholders did not stay invested inthe initiative long enough to cement relationships.Along these same lines, sdg omitted to takesignificant advantage of its relationship with theWorld Bank Group and other shareholders in thecreation of an enabling market environment (supportivepolicy and regulatory environment, avail-52 selling solar: part 2 / case studies

ability of end-user finance, knowledge and awarenessof solar pv). Furthermore, other groups (theUnited Nations, for example) were also involved insolar initiatives, although there was no coordinationbetween the di∑erent projects. An enabling environmentis necessary for sector development, but it iscostly and requires significant resources, as well asenthusiasm by local governments for solar pv. Theregulatory and policy environment will not alter untilthis occurs. While the World Bank is in a positionto impact and promote local government supportof solar pv initiatives, this relationship was nottaken advantage of by sdg.Need to Focus on Market Development andCapacity BuildingDespite the fact that there was, and is, a clear needfor market development and capacity building inthe solar pv sector, the reality was that sdg was focusedmore on individual businesses. Had more attentionbeen paid by sdf to develop the enablingenvironment, sdc may have found more investmentopportunities, and sdg would have had a greateroverall impact. Early recognition of market realitywould have led to a smaller loss of funds.Perhaps one of the greatest lessons that can bedrawn from sdg is one that resulted not from a failureto achieve investment goals, but rather from theresponse to this failure. sdc management was quickto recognize that the initial investment criteria weretoo stringent, and it worked with shareholders to revisethem. Rather than making bad deals, whichcould have jeopardized the reputations of those involved,it failed to make any deals, resulting in littleinvestment loss. Management was correct in makingthis decision, providing a lesson on the merit of restraint.When it became apparent that the marketcould not produce opportunities that met the revisedinvestment criteria, sdg was disbanded.the private equity fund, did not even come close.Solar pv markets simply were not mature enoughfor equity investments, and the family-owned natureof most solar pv smes further limited the possibilitiesfor equity investment. At this stage of marketdevelopment, di∑erent financing instrumentsand long-term patient capital was needed; sdc, withits ten-year fixed life and return expectations, couldnot provide this.Like many other projects implemented aroundthe same time, sdg grossly overestimated the marketand the number of business opportunities that existedwithin it. In hindsight, it is easy to state thatthe focus of the project was too narrow, that the projectshould have focused on other re technologies inaddition to solar pv, and that additional financinginstruments should have been provided. At the timethe project was implemented, however, those involvedin the industry truly did believe that the solarpv market was poised to take o∑. Had the solar pvsector actually performed as forecast, the sdg experiencewould surely have been quite di∑erent.conclusionsdg’s initial goal was to increase the delivery of shsto rural households in developing countries and tosupport the development of the solar pv market.While sdf is seen as having had a positive impact onthe solar pv industry, sdc failed to accomplish anyof its goals, and, overall, sdg came up short.The two entities that made up sdg had verydi∑erent experiences. sdf, the not-for-profit arm,which provided loans, guarantees, and grants, waslargely able to meet its investment goals, while sdc,solar development group 53

case studies 4Cagayan Electric Power &Light Company44After the solar pv plant becameoperational, ifc’s InfrastructureDepartment provided cepalcowith a local currency loan (thefirst provided in the Philippines)in an amount equivalent to $15million. The funds will allowcepalco to pursue its expansionprogram through a more stablefinancing structure.45Designed to ensure that thecompany remained competitive ina newly privatized environment,cepalco’s expansion plansincluded forming a partnershipwith a subsidiary of HawaiianElectric Industries, an internationalutility with private powerinvestment objectives in thePacific Rim, and planning aninvestment program gearedtoward e≈ciently expanding itssystem and attracting a largercustomer base.Cagayan Electric Power & Light Company(cepalco), a private electricity distribution companyon the island of Mindanao in the Philippines,received $4 million in funding from gef (throughifc) in 2002. The purpose of the project was todemonstrate solar pv’s e∑ectiveness (through a conjunctive-useapplication) in addressing distributionsystem capacity issues. The funds were used to builda 1 mw distributed generation solar pv power plant,which was integrated into the 80mw distributionnetwork of cepalco, and operated in conjunctionwith an existing 7 mw small, run-of-the-river hydroelectricplant.backgroundMindanao is the second largest island in the Philippines.The electrical grid on the island is well developed,but isolated, and it has a total generating capacityof 1,800 mw. Most of the generation capacityis from hydroelectric plants, with seasonal changesin generation capacity occurring. The transmissioninfrastructure on Mindanao is unable to transportlarge amounts of power over 200mw across largedistances without compromising system reliability.As a result, cepalco and other distribution companieson the island must obtain large portions oftheir power supply from local sources.In 1998, cepalco was in discussions with ifc forgeneral corporate financing 44 to assist with the company’sexpansion plans. 45 The environmental reviewundertaken identified the possibility for a solar pvbasedproject. A pre-appraisal mission took place inthe Philippines in December 1998 by the then ifc’sEnvironmental Finance Group, with the goal of investigatingsuitable applications for solar pv-basedgenerating equipment within the distribution networkof cepalco. Distributed solar pv technologies,which are by their very nature small and situatednear load centers, were believed to o∑er the potentialto address seasonal generation capacity issues.They were also considered compatible withcepalco’s needs and the geography of the transmissionsystem.Based on the findings of the December 1998 preappraisalmission, ifc proposed to gef a projectthat would use gef funds to partially finance theinstallation of a nominal 1 mw solar pv-based powerplant, which would be integrated with the cepalcopower distribution network. The stated objective ofthe project was to demonstrate the technical, operational,and economic feasibility of using solar pvelectricity supplies for supplementing and firmingup the productive capacity of an existing hydroproject.The solar pv plant was designed to operate inconjunction with the recently built 7 mwBubunawan Hydropower Plant Project, a smallrun-of-the-river hydroelectric power plant, whichwas already supplying power to cepalco’s system.The hydroelectric plant was to operate as a loadfollower, varying its output with the availability ofsolar pv power. The water saved, when solar pvpower was being produced, would be held in smallponds available at the Bubunawan plant, to be utilizedfor power generation during peak load periodswhen solar pv was not available. The addition ofthe solar pv plant would reduce the need forcepalco to purchase thermal energy during peakload periods, thus reducing CO2 emissions. In total,the 1 mw plant was estimated to lead to a reductionin CO2 emissions on the island of Mindanao54 selling solar: part 2 / case studies

of approximately 1,200 metric tonnes annually.The project was endorsed by the gef Council inDecember 2002. cepalco received a $4 millionconvertible loan from ifc, using gef funds, andfinanced the remaining $1.8 million in constructioncosts from cepalco’s own cash flow and interestsavings. The gef funds were initially provided as aloan, with the understanding that the loan wouldbe forgiven after five years of satisfactory projectoperation.In the period leading up to the final approval ofthe funding, cepalco issued an international requestfor proposals for the construction of aturnkey 1 mw grid-tied solar pv plant. The contractwas awarded to Sumitomo Corporation of Japan,which utilized solar pv modules supplied by SharpCorporation. Construction started in August 2003and was completed on schedule. President Arroyoof the Philippines inaugurated the plant in December2004. Today, the cepalco solar pv facilityrepresents the largest solar pv installation in thedeveloping world, and it has been operating withoutincident since its commission. It remains theonly solar pv power plant in the world that is operationalin a conjunctive-use application with a hydroplant.what worked and what did notThe ifc/gef project with cepalco has had a stronglocal presence, both through the in-country ifc representativeo≈ce in Manila and through the e∑ortsof the cepalco management and sta∑, which wascritical to the success of the project. The Philippinesta∑ at cepalco and the contractors from SumitomoCorporation did an impressive job, not only ofobtaining more than 50 permits and licenses inplace for the facility, but also of complying with ahost of project finance disbursement requirementsand, finally, learning to successfully operate the projectsince completion of construction.During project implementation, it quickly becameevident that both the local permit process andthe typical project finance structure were not gearedto a small 1mw-solar pv investment. On the permitside, the Philippine regulations did not make thedistinction between the solar pv plant and the muchlarger fossil-based plants, despite the obvious environmentalbenefits and the proven nature of solarpv technology. As a result, the cepalco solar pv projectmanagement team was required to performtasks and submit reports that were not relevant toan re project. The convertible loan structure (grantcagayan electric power and light company 55

convertible to loan) of the project called for a rigiddisbursement structure. It became apparent, however,that, in order to ensure timely completion ofconstruction, the prescribed disbursement structurewould have to serve as a guide only.It was widely recognized at the time the projectwas developed that there were more cost-e∑ectivetechnologies available to address peak load energysupply issues. In fact, the gef grant, providedthrough ifc, served to subsidize approximately 70percent of the construction and startup costs of thecepalco solar pv plant. The cepalco solar pv projectnever intended to compete head-on with otheralternative energy generation technologies on thebasis of cost e≈ciency, but it did intend to demonstratethat solar pv could be used as an e∑ective andtechnically reliable source of power that could becost e∑ective if solar pv prices declined su≈ciently.Furthermore, the cepalco project intended todemonstrate that there are technical advantages tothe operation of such a plant in conjunction withan existing small hydro plant with limited storagecapacity. The plant, which has operated without incidentsince its inauguration in 2004, appears todate to have been successful in proving solar pv tobe an e∑ective and technically reliable technology toaddress peak-load energy supply issues.conclusionThe cepalco experience is an interesting one. Unlikeother ifc/gef climate change mitigation programsand projects, many of which were designed asmarket transformation initiatives, cepalco was designedas a stand-alone experiment to demonstratethe appropriateness of the solar pv technology(through a conjunctive-use application) in addressingdistribution system capacity issues. To date, thecepalco solar pv plant has made a strong technicalcase for the reliability of utility-scale solar pv powerplants. Furthermore, by avoiding the need to purchasealternative thermal energy, the cepalco solarpv plant has resulted in a significant reduction ingreenhouse gas emissions. The plant is expected todisplace 24,000 tonnes of CO2 over its lifetime.When the financing was provided, it was expectedthat solar pv prices would decrease and thatsolar pv technology used on a utility scale wouldtherefore become more cost e∑ective. Had the priceof solar pv gone down as was expected (and it is stillpredicted to occur in the future), cepalco wouldhave been a project with a high potential for replication.However, with solar pv prices having increased,the potential for replication withoutsignificant subsidization is limited. Perhaps themost important demonstration value of the cepalcoproject is that solar pv works e∑ectively in a conjunctive-useapplication. It also illustrates the factthat the technical solution is not always the bestmarket solution.56 selling solar: part 2 / case studies

eferencesIntroductionBlanco, José Ma., 1995, “Photovoltaics, Wind and OtherDispersed Energy Sources,” in Energy as an Instrumentfor Socio-Economic Development, United NationsDevelopment Program, ed. by José Goldemberg andThomas B. Johansson: 49–59.Cabraal, Anil, M. Cosgrove Davies, and L. Schaeffer,1996, “Best Practices for Photovoltaic Household ElectrificationPrograms: Lessons Learned from Experiencesin Selected Countries,” World Bank TechnicalPaper, No. 324 (Washington, D.C.: World BankGroup).Erickson, Jon D., and Duane Chapman, 1995, “PhotovoltaicTechnology: Markets, Economics and RuralDevelopment,” World Development, 23 (7): 1129–1141.Flavin, Christopher, and Molly O’Meara, 1997, “ShiningExamples,” World Watch (May/June): 28–36.Foley, Gerald, 1995, “Photovoltaic Applications in RuralAreas of the Developing World,” World Bank TechnicalPaper, No. 304 (November), op. cit. note 1, EnergySeries (Washington, D.C.: World Bank Group).World Bank, Energy and Mining Sector Board, 2005,“Progress on Renewable Energy and Energy E≈ciency,1990–2004” (March) (Washington, D.C.: World BankGroup)._______, 2005, “Progress on Renewable Energy and EnergyE≈ciency, Fiscal Year 2005” (December) (Washington,D.C.: World Bank Group).Chapter 1Ahmed, Kulsum, 1994, “Renewable Energy Technologies:A Review of the Status and Costs of Selected Technologies,”World Bank Technical Paper, No. 240,Energy Series (January) (Washington, D.C.: WorldBank Group).Caldwell, James H., Jr., 1994, “Photovoltaic Technologyand Markets,” Contemporary Economic Policy, 12 (2):97–110.Chambouleyron, I., 1996, “Photovoltaics in the DevelopingWorld,” Energy, 21 (5): 385–394.Erickson, Jon D., and Duane Chapman, 1995, “ResidentialRural Solar PV Electricity in Developing Countries,”Contemporary Economic Policy, 13 (2): 98–109._______, 1995, “Photovoltaic Technology: Markets, Economicsand Rural Development,” World Development,23 (7): 1129–1141.Jackson, Tim, and Mark Oliver, 1999, “The Market forSolar PV Photovoltaics,” Energy Policy 27: 371–385.Kapadia, Kamal, 1999, “Off-grid in Asia: The Solar PVElectricity Business,” Renewable Energy World (November):23–33.Northrup, Michael F., Peter W. Riggs, and Frances A.Raymond, 1995, 1996, “Selling Solar PV: FinancingHousehold Solar PV Energy in the DevelopingWorld,” A Report Based on a Workshop at the PocanticoConference Center of the Rockefeller BrothersFund, Pocantico Paper, No. 2. (October 11–12) (NewYork: Rockefeller Brothers Fund).Wang, Xiaoping, 1998, “The Best Practices for O∑-gridSolar PV Energy—A Case Study on China” (School ofForestry and Environmental Studies, Yale University).Significant market trendsHande, Harish, 2006, “Does Solar PV Still Make Sensefor the Developing World?” RenewableEnergyAccess.com (February 16)._______, 2007, “Lighting the World,” Solar PV News(February): 14.Chapter 2International Energy Agency, 2002, “Financing Mechanismsfor Solar Home Systems in Developing Countries:The Role of Financing in the DisseminationProcess,” Photovoltaic Power Systems Program.Martinot, Eric, Anil Cabraal, and Subodh Mathur, 2000,“World Bank Solar Home Systems Projects: Experiencesand Lessons Learned 1993–2000” (Draft) (January).Martinot, Eric, Ramesh Ramankutty, and Frank Fittner,2000, “The GEF Solar PV Portfolio: Emerging Experienceand Lessons,” Monitoring and EvaluationWorking Paper, No. 2 (August) (Washington, D.C.:Global Environment Facility).Phasiwe, Khulu, 2004, “Lights Go Out for Off-Grid EnergyProjects,” Business Day. Available via the internet:http://allafrica.com/stories/200401130458.html.Rogers et al., 2006, “Innovation in Rural Energy Delivery:Accelerating Energy Access Through SMEs,”(Navigant Consulting/Soluz, Inc.).58 selling solar

World Bank, 2004, “Renewable Energy for Development—The Role of the World Bank Group” (Washington, D.C.:World Bank Group).Chapter 3Cabraal, Anil, Douglas Barnes, and Sachin Agarwal,2005, “Productive Uses of Energy for Rural Development,”AR Reviews in Advance (Washington, D.C.:World Bank Group).Carey, John, 2006, “What’s Raining on Solar PV’s Parade,”BusinessWeek Online, (February 6). Available via theinternet: http://www.businessweek.com.Mbogo, Stephen, 2001, “Solar PV Energy—The Potentialin East Africa,” Appropriate Technology, 28 (1):24–26.“Potential of Solar PV Energy in Rural Areas Untapped,”2001, Appropriate Technology, 28 (1): 21.“Power to the People,” 2006, The Economist (March 11):37–38.Saghir, Jamal, 2005, “Energy and Poverty: Myths, Links,and Policy Issues,” Energy Working Notes, No. 4 (May).“Sunrise for Renewable Energy?” 2005, The Economist(December 10): 18–20.Hansen, Richard D., 2005, “Strategic Review of the WorldBank-GEF Rural Renewable Energy Portfolio,” (Chelmsford,Massachusetts: Global Transition Consulting,Inc.).Jackson, Tim, and Mark Oliver, 2000, “The Viability ofSolar PV Photovoltaics,” Energy Policy, 28: 983–988.Mariyappan, Jason, and Dennis Anderson, 2002, “ThematicReview of GEF-Financed Solar PV ThermalProjects,” Monitoring and Evaluation Working Paper 7(Washington, D.C.: Global Environment Facility).Martinot, Eric, 2005, “Renewables 2005 Global StatusReport,” The Worldwatch Institute, Renewable EnergyPolicy Network (Washington, D.C.)Martinot, Eric, and Omar McDoom, 2000, “PromotingEnergy Efficiency and Renewable Energy: GEF ClimateChange Projects and Impact” (June) (Washington D.C.:Global Environment Facility).Pearce, Joshua M., 2002, “Photovoltaics—A Path toSustainable Futures,” Futures, No. 24 (September):663–674.Renewable Energy Access, subscriber-based dailyrenewable energy news. Available via the internet:http://www.renewableenergyaccess.com.“The Promise of the Solar PV Future,” 2005, Co-opAmerica Quarterly (Summer).SME Program Case StudiesEnerteam, 2001, “Di∑usion of Solar PV Home Systems inVietnam: Final Report” (Ho Chi Minh City, Vietnam).Wheeler et al, 2004, “Sustainable Business Approachesfor Solar PV Energy Enterprises in Developing Countries:Key Success Factors and Lessons Learned fromBangladesh, Honduras, and Vietnam,” Institute for Researchand Innovation in Sustainability (July) (Toronto:York University).Additional Bibliographic Resources ConsultedAllderdice, April, and John H. Rogers, 2000, “RenewableEnergy for Microenterprise,” (Colorado: National RenewableEnergy Laboratory).Alsema, Erik .A., and E. Nieuwlaar, 2000, “Energy Viabilityof Photovoltaic Systems,” Energy Policy, 28:999–1010.references 59

acknowledgmentsSelling Solar has been prepared by the SustainableBusiness Innovator Group of Environment and SocialDevelopment Department of ifc.This report would not be possible without the insightand guidance of Dana Younger.AuthorsNatalia Magradze, Alan Miller, Heather SimpsonReviewersAnil Cabraal, Bernard McNelis, Shilpa PatelEditorial TeamLouise Gardiner, Lucie Giraud, Margie Peters-Fawcett,Amy SweetingContributorsSachin Agarwal, Hany Assaad, Maurice Biron,Lisa Da Silva, Jon Forster, Devyani Hari, Terry Hart,Sandeep Kohli, Fernando Manibog, Shir Ashar Naveh,Fabio Nehme, Ashington Ngigi, Eluma Obibuaku,Pallavi Shah, Russell Sturm, Stacy Swann, JoannaWillottPlease address questions or comments to Alan Miller(amiller2@ifc.org) or Natalia Magradze(nmagradze@ifc.org).PhotographyAll images courtesy of IFC and World Bank Sta∑unless otherwise credited.Cover inset (woman pumping water): NationalRenewable Energy LaboratoryPage 32: Micro Energy International60