59000 The World Bank Asia Sustainable and Alternative Energy Program ASTAE Timor-Leste Key Issues in Rural Energy Policy December 2010 Copyright © 2010 The International Bank for Reconstruction and Development / The World Bank Group 1818 H Street, NW Washington, DC 20433, USA All rights reserved First printing: December 2010 Manufactured in the United States of America. Photo copyrights: ASTAE The findings, interpretations, and conclusions expressed in this report are entirely those of the authors and should not be attributed in any manner to the World Bank, or its affiliated organizations, or to members of its board of executive directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. Contents Preface and Acknowledgments ................................................................................................vii Abbreviations and Acronyms ....................................................................................................ix Executive Summary ...................................................................................................................xi Off-Grid Electrification ............................................................................................................................................. xi Household Energy Policies ..................................................................................................................................... xii Interfuel Substitution with Modern Fuels ........................................................................................................ xii Improved Stoves ............................................................................................................................................. xii Domestic Biogas Systems ............................................................................................................................. xiii Alternative Fuels: Biodiesel from Jatropha ............................................................................................................ xiv 1 Context and Background .......................................................................................................1 The Rural Energy Sector ...........................................................................................................................................1 Predominance of Biomass Fuels .......................................................................................................................1 Limited Access to Electricity Services ..............................................................................................................2 Institutional Responsibilities ..............................................................................................................................2 Key Issues in Rural Energy ................................................................................................................................2 Basis of Sound Rural Energy Policy ..........................................................................................................................3 Purpose and Organization of This Report .................................................................................................................3 2 Electricity Access Policies .....................................................................................................5 Off-Grid Electrification ..............................................................................................................................................5 Decentralized Mini-Grids ...................................................................................................................................6 Solar Home Systems.........................................................................................................................................7 Solar Battery Charging Stations ........................................................................................................................9 3 Household Energy Policies .................................................................................................11 Fuel-Wood: Key Issues ........................................................................................................................................... 11 Interfuel Substitution with Modern Fuels ...............................................................................................................13 Programs That Subsidize Fuel, or Equipment, or Both Without Income Tests .................................................14 Programs That Subsidize Fuel, or Equipment, or Both with Income Tests ......................................................14 Combination Programs of Direct and Voucher Subsidies ................................................................................14 Better Practice: Subsidize Equipment, Not Fuel ..............................................................................................15 Likely Continued Use of Fuel-Wood Stoves ....................................................................................................15 Improved Cooking Stoves as an Intermediate Approach ........................................................................................15 Initiating an Improved Stoves Program: Subsidy Issues .................................................................................16 Monitoring Indoor Air Pollution ........................................................................................................................17 Domestic Biogas Systems .....................................................................................................................................17 Costs and Benefits ..........................................................................................................................................17 Results of Analyses .........................................................................................................................................18 Sustainability Issues........................................................................................................................................18 Conclusions ............................................................................................................................................................19 iii iv Contents 4 Alternative Fuels: Biodiesel from Jatropha .......................................................................21 Land Use ................................................................................................................................................................22 Energy Balance.......................................................................................................................................................23 Cost of Supplying Jatropha Oil ...............................................................................................................................25 Subsidies for Jatropha Oil Growers .................................................................................................................26 Impacts on Foreign Exchange .........................................................................................................................27 Impacts on Farm Incomes and Employment ..................................................................................................28 Conclusions ............................................................................................................................................................28 Recommended Activities .......................................................................................................................................28 5 Conclusions and Recommendations .................................................................................31 Improving Access to Electricity ..............................................................................................................................31 Household Energy ..................................................................................................................................................31 Interfuel Substitution .......................................................................................................................................32 Improved Cooking Stoves ...............................................................................................................................32 Domestic Biogas .............................................................................................................................................32 Biodiesel from Jatropha .........................................................................................................................................33 Annex 1. Designing a Sustainable Solar Home Systems Program for Timor-Leste ............35 Program Phasing ....................................................................................................................................................35 Payment Scheme ...................................................................................................................................................35 Preparing Market Packages ....................................................................................................................................35 International Competitive Bidding ..........................................................................................................................36 Service Contract Terms ..........................................................................................................................................36 Options at End of Service Contract ........................................................................................................................36 Annex 2. Outline of a Program for Dissemination of Improved Fuel-Wood Cooking Stoves.................................................................................................39 Immediate Objectives ............................................................................................................................................39 Target Area and Market Size...................................................................................................................................39 Implementation Phasing ........................................................................................................................................40 Development of New Stove Models to Suit User Needs in Timor-Leste................................................................40 Possible Urban Stove..............................................................................................................................................40 Rural Stove .............................................................................................................................................................41 Production Strategy ................................................................................................................................................41 Urban Stove.....................................................................................................................................................41 Rural Stove ......................................................................................................................................................42 Organizational Arrangements .................................................................................................................................42 Marketing and Dissemination Strategy ..................................................................................................................43 Urban Stove.....................................................................................................................................................43 Rural Stove ......................................................................................................................................................43 Demonstration Improved Kitchens .........................................................................................................................43 Budget Requirements ............................................................................................................................................43 Annex 3. Economic Analysis of Domestic Biogas Systems ...................................................45 Case 1: Cooking Only, Three Households per Digester ..........................................................................................45 Case 2: Cooking and Lighting for One Household .................................................................................................46 Contents v Boxes 2.1 Experience with Solar Battery Charging Stations in Nicaragua .........................................................................8 3.1 Biomass Briquettes as a Substitute for Fuel-Wood .........................................................................................12 3.2 LPG Subsidies in India: Growth of Demand and Financial Liabilities ..............................................................14 3.3 Commercial-Scale Biogas Projects: Potential Contributions to Power Supply ................................................19 4.1 Jatropha-Based Fuels in the Context of Overall National Energy and Agricultural Policies..............................29 A3.1 The Opportunity Cost of Fuel-Wood in Rural Timor-Leste ...............................................................................47 Figures 3.1 Impacts and Efficacy of Programs to Promote Modern Cooking Fuels...........................................................13 4.1 Modeling Schematic for Commercial Jatropha Cultivation..............................................................................22 4.2 Net Energy Analysis of Jatropha Cultivation in Thailand ..................................................................................24 4.3 Jatropha Economics with Market-Determined Split for Growers and Processors ..........................................26 4.4 Government Subsidies for Growers Under Cooperative Ownership of Processing ........................................27 4.5 Government Subsidies for Jatropha Growers Under Typical Market Arrangements .......................................27 4.6 FX Impacts of Jatropha-Based Biodiesel .........................................................................................................27 A2.1 The Malawi Stove: A Portable Urban Stove.....................................................................................................40 A2.2 A Massive Rural Stove Built On-Site ...............................................................................................................41 Tables 1.1 Timor-Leste 2006: Useful Energy Share of Major Fuels for All Sectors ...........................................................1 1.2 Sources of Energy Used for Cooking ................................................................................................................2 2.1 Distribution of Potential PV Customers by District ............................................................................................6 3.1 Total Health Costs from Solid Fuels Use in Timor-Leste ..................................................................................12 3.2 Heating Replacement Value of Biogas ............................................................................................................18 4.1 Parameters for Economic Analysis of Jatropha Cultivation in Timor-Leste ......................................................25 A3.1 Case 1: Three Households per Digester, Cooking-Only Scenarios ..................................................................46 A3.2 Case 2: One Household per Digester, Cooking and Lighting Scenarios .........................................................46 Preface and Acknowledgments This report responds to a request made by the govern- (EASIN), who reviewed the draft and provided valuable ment of Timor-Leste for assistance in finalizing proposed . comments. The peer reviewers were Douglas F Barnes, policies for selected rural energy areas. The team that Rural Energy Specialist (Consultant); and Xiaoping Wang, prepared this report consisted of Frédéric Asseline, Senior Energy Specialist (LCSEG). Task Team Leader (EASIN); Ernesto Terrado, Renew- able Energy Specialist (Consultant); Donald Hertzmark, The team gratefully acknowledges the cooperation of Energy Economist (Consultant); and R. M. Amarasekara, and contributions made by Secretary of State for Energy Rural Energy Specialist (Consultant). The team received Policy Avelino Coelho da Silva and his staff. guidance from Charles Feinstein, Sustainable Develop- ment Leader for Timor-Leste, Papua New Guinea, and This study was made possible by the generous support the Pacific Islands (EASNS); and Dejan Ostojic, Energy of the Asia Sustainable and Alternative Energy program Sector Leader for the East Asia and Pacific Region (ASTAE). vii Abbreviations and Acronyms ADB Asian Development Bank Units of Measure CBO community-based organization /d per day EdTL Electricidade de Timor-Leste /y per year ERR economic rate of return bbl barrel FSD family-size digester GJ gigajoule FX foreign exchange ha hectare GDP gross domestic product kg kilogram GNI gross national income km kilometer IAP indoor air pollution kT kiloton ICS improved cooking stove kw kilowatt LPG liquefied petroleum gas kwh kilowatt hour MAFF Ministry of Agriculture, Forestry and Fisheries m3 cubic meter NER net energy ratio m3/d cubic meter per day O&M operations and maintenance MJ megajoule PV photovoltaic toe ton of oil equivalent REMP Rural Electrification Master Plan W watt SCBS solar battery charging station $ = US$ = United States dollar SHS solar home system ix x Executive Summary The primary goal of a sound rural energy policy is to pro- it is clear that under current prioritization criteria those mote measures that will enhance the quality of life of communities are unlikely to have access to electricity people in rural areas by improving their access to modern for at least the next decade. Government policy should energy services. The desired approach is one that is envi- be to provide at least basic electricity services to small, ronmentally benign and economic from both the coun- dispersed, and poor off-grid communities, through pro- try's point of view and for the welfare of individual rural grams implemented during the same timeframe as households. A key objective is to ensure that the imple- those directed at concentrated communities living near mentation of the government's rural energy programs the main electricity network. The implementation policy provides equitable distribution of benefits. should be based on principles of least-cost power plan- ning, and with the objectives of technical and financial In Timor-Leste the Secretary of State for Energy Policy sustainability. Where local energy resources are available is responsible for the design and implementation of the and potential users are sufficiently numerous and con- government's rural energy program. National energy poli- centrated, those resources could be exploited for decen- cies are approved by the Council of Ministers, and the tralized mini-grids. The government is already developing Secretary of State for Energy Policy takes responsibil- identified oil and gas seeps and hydro resources for this ity for developing legal and regulatory frameworks for purpose. all activities related to the use of energy resources. The Secretary of State for Energy Policy plays an essential The government's off-grid strategy is presently less role in ensuring coordination with other ministries and defined for providing individual households with stand- operators involved in the energy sector in Timor-Leste. alone solar photovoltaic (PV) systems. From 2007 to 2010, more than 2,000 PV systems for households and The purpose of this report is to assist the government of community centers have been provided at no cost to Timor-Leste, in particular the office of the Secretary of recipients, financed with infrastructure funds provided State for Energy Policy, to develop policies in key areas annually to the sucos (villages) by the central government. that would guide planning of the subsequent phase of its It is estimated that about 60,000 households nationwide ongoing rural energy programs. The selected key areas in could be provided access to electricity service only with the report cover off-grid electrification, household energy, PV. Based on the cost of investments benefiting grid-con- and the development of biofuels from Jatropha crops. nected households (average $640 per connection plus subsidized tariff), a substantial upfront cost subsidy for a The report does not prioritize policies in the selected rural modest-sized solar home system (for example, 50 watt- energy areas, nor does it aim to propose a comprehensive peak) may be justified in Timor-Leste on equity grounds. policy framework for rural energy solutions. A national However, it is best in any program to require PV recipi- rural energy program is currently being developed by ents to contribute some part of the system acquisition the Secretary of State for Energy Policy; in this context cost in order to instill a sense of ownership. this study proposes practical recommendations derived from lessons learned from international experience in the Timor-Leste does not yet have an environment that would areas of off-grid electrification, household energy, and the support an implementation mechanism for a government development of biofuels from Jatropha crops. program in which private PV dealers compete to make sales. The micro-financing system is still undeveloped and the current PV market consists almost entirely of Off-Grid Electrification households belonging to the bottom part of the income pyramid. Some type of simplified fee-for-service mecha- In Timor-Leste, conventional rural electrification through nism executed by the private sector but with a strong grid extension is being implemented based on a national government role would appear to be suitable for Timor- rural electrification master plan (REMP). While the REMP Leste, at least for the next 5 to 10 years. Using such a recognizes the special needs of off-grid communities, mechanism, competitive bidding for large numbers of xi xii Executive Summary installations grouped into several "market packages" Interfuel Substitution with Modern Fuels could result in unit acquisition costs much lower than Compared with fuel-wood, modern fuels, including LPG they are today in piecemeal, suco-by-suco transactions. and kerosene, have lower emissions and are more con- A suggested procedure is outlined in the report. venient and cleaner to handle. It is almost inevitable that demand for modern fuels, particularly LPG, will increase However, as with any model that serves very-low-income dramatically in Timor-Leste in the coming years, as eco- households and requires periodic payments to maintain nomic conditions improve and a middle class emerges operation, long-term sustainability is more likely only and expands. Most developing countries favor this mod- when combined with other programs that raise the ernization outcome and adopt policies to accelerate the incomes of the user community. transition from fuel-wood to modern fuels. It is not recommended that Timor-Leste directly subsidize Household Energy Policies the price of LPG or kerosene fuel for several reasons: (a) possible diversion of the fuel for noncooking uses, As found in a recent World Bank study, there is still abun- such as transport, can occur (this is likely to be more so dant biomass supply in most parts of the country, except with kerosene than LPG); (b) possible smuggling and in the Dili district.1 Because of this abundance, fuel-wood sale of the fuel outside of the country; (c) difficulties is, and for some time to come will continue to be, the in limiting benefits to low-income households or crafts cheapest cooking fuel in Timor-Leste compared with and professions; and, most important (d) the subsidy bur- liquefied petroleum gas (LPG), kerosene, and electricity, den to the government is likely to become unmanage- even after accounting for different cooking equipment able in the future. As experience worldwide has shown, efficiencies. it is extremely difficult to withdraw or even reduce sub- sidies once they become critical to consumer choices for There is no doubt, however, that the responsible policy for fuel use. the government to pursue must result in a reduction or eventual curtailment of household fuel-wood use to the It is often more effective, with fewer undesirable side extent possible. The rationale for this policy is based on effects, to subsidize access, not consumption. Subsidiz- (a) the certainty that the current positive supply-demand ing access calls for programs that subsidize the equip- balance will change in the future, with population growth ment needed to make the fuel switch (for example, and increased pressure on wood supplies as forest lands new stove purchases, deposits for LPG cylinders) and are converted to agriculture; and (b) the adverse impacts facilitating fuel distribution but keeping fuel prices at to public health from indoor air pollution (IAP) associated market levels. The financial exposure of the government with use of fuel-wood in traditional cook-stoves. A World for an equipment subsidy program can be determined Bank Country Environmental Analysis for Timor-Leste annually in advance, unlike fuel subsidy programs that conducted in 2008 estimated that the mean annual are entirely subject to the vagaries of fuel market fluc- morbidity and mortality cost of health effects from IAP tuations. The government may terminate the equipment associated with the use of solid fuels is between $5 mil- subsidy program at almost any time with minimal public lion and $20 million. The mean estimate is equivalent to inconvenience. about 1.4 percent of Timor-Leste's gross national income, or 3.5 percent of GDP2 in 2006. Improved Stoves The strategy, therefore, is to pursue programs that Promoting higher efficiency or "improved" cooking combine both demand-side management or interfuel stoves (ICS) should not be viewed as encouraging con- substitution (or both) with programs aimed at improved tinued use of fuel-wood but as a necessary intermediate management of fuel-wood supply. The latter programs approach--the only practical way for the government to are not only good for the country's forests in general, but reduce fuel-wood consumption on any significant scale they also buy time until modern fuels can make signifi- while the economy transitions to a future in which mod- cant inroads into the demand picture. ern fuels are more widely available and affordable for most households. The benefits associated with modern 1. S. Shum, E. Terrado, K. Openshaw, and V. Tuntivate, "Timor-Leste: , fuels, such as lower levels of harmful IAP higher energy Issues and Options in the Household Energy Sector--A Scoping efficiency, and cleaner kitchen environments, are also " Study, World Bank, 2007 . 2. Timor-Leste's GDP is $356 million (current 2006 dollars; World Bank 2007). Executive Summary] xiii obtainable to a certain degree with improved stoves. By economic analyses counted the benefits of producing their design, improved stoves burn wood more efficiently nitrogen fertilizer from the digester sludge. Using as a than traditional ones and thus have lower emissions. base case the mid-range values for capital costs and gas quality, the results clearly point to the low economic via- Promoting improved stoves in Timor-Leste is severely hin- bility of a program designed mainly to replace fuel-wood. dered by the continuing relative abundance and low cost The main reasons are the relatively high investment costs of fuel-wood supplies in the country, making it hard to per household and the low opportunity cost of the fuel- motivate households to economize. In addition, Timorese wood replaced. The results were somewhat stronger for households do not yet have a culture of purchasing the single household case mainly because of the higher stoves. The only way for improved fuel-wood stoves to replacement value of imported kerosene (for lighting). be disseminated widely in Timor-Leste is to substantially However, this case assumes that the single household subsidize their cost to users. Fortunately, compared with owns four to six cattle to supply the waste requirements other programs for reducing fuel-wood consumption, the of the FSD, thus limiting the coverage of the biogas pro- required subsidies per household are small and could gram to a relative handful of households. be terminated at any time by the government without immediate drastic welfare impacts on users (they simply Apart from the economics, it is also important for policy revert to traditional stoves). The potential benefits, even making to consider beneficiary issues related to the fact if just reducing IAP impacts, far outweigh the cost. An that current recipients of biogas systems are families IAP monitoring program should be carried out as part of that are undoubtedly better off than others in the com- any ICS program. munity. In the pilot or demonstration phase, there is little choice but to select such families for biogas installations. The report outlines procedures for a conceptual five-year However, in subsequent expansion phases of the biogas ICS program targeting 15,000 households in Dili and program, explicit consideration of the distribution of ben- Baucau. This amounts to about 25 percent of households efits should be factored into whatever financial or subsidy in the two districts and 8 percent of total households in mechanisms are employed. The Ministry of Agriculture, Timor-Leste. The ICS model would be patterned after the Forestry and Fisheries (MAFF) has an ongoing livestock higher efficiency, low-cost, portable "Malawi stove" or breeding and distribution program for cattle, buffalo, pigs, similar stoves already tested elsewhere. The cost of the goats, sheep, and poultry, and aims to establish large live- program over five years is estimated to be $300,000 to stock farms. When the livestock program has progressed $350,000, excluding the cost of international consultants significantly and more rural families are raising their own needed at the design and preparation stages. animals, it may become possible for the post-pilot phase of the biogas program to have a wider and more varied set of beneficiaries. Domestic Biogas Systems In Timor-Leste, a program has been proposed to provide In contrast with household-scale biogas digesters, the biogas systems to some 4,600 rural households, which is positive economics of feedlot-scale units has been the estimated number of households owning at least one established for a number of years. There are also ancil- to two head of cattle. The main purpose is to displace the lary benefits including reducing water pollution and run- fuel-wood presently used in these homes with a clean off and production of byproducts such as fertilizer and and renewable fuel substitute. In the initial configuration, fish farming on commercial scales. Consequently, as one small family-size digester (FSD) capable of producing larger livestock farms are established, government policy about two cubic meters per day (m3/d) of biogas would should make feedlot-scale biogas digesters a priority for be shared by three households, using waste from a total support. of four to six head of cattle owned by the families. The investment cost is estimated to range from $500 to Meanwhile, the current program to install FSDs has a $1,000 per FSD, depending on locality-specific prices for high capacity-building value and should be continued for construction materials. that purpose, limited for now to a pilot scale. A key objec- tive should be to monitor actual costs and other operat- Preliminary economic analyses were made of the above ing data from the pilot installations to guide decisions on configuration and of a case in which only one household the subsequent phase. uses the FSD and splits the output equally between cooking and lighting. In addition to the fuel benefits, the xiv Executive Summary Alternative Fuels: · Based on the overall costs of growing, transport- Biodiesel from Jatropha ing, and processing Jatropha, a crude oil price at or above $90 per barrel (bbl) is necessary to cover the The government has initiated a program of cultivation of costs of the various parties, and requires as well that Jatropha curcas, a drought-tolerant vegetable oil crop, the price component attributable to processors be to produce biodiesel and replace imported diesel oil for capped. In a more typical market setting, still higher electricity generation or transport. Currently, a few hect- crude prices, greater than $120/bbl, will be needed ares have been planted for a little more than one year by to provide adequate profitability for both the farmers small farmers in various places using marginal lands. The and the processors. For example, if Jatropha farmers overall concept is for farmers to harvest the Jatropha oil are guaranteed a $250/ha profit above estimated pro- seeds, bring them to the local cooperative for process- duction costs of $335/ha, the government will need ing, and be paid a fixed price for each kilogram of Jatro- to provide subsidies in excess of $2 million/year for pha seed. The seeds will be brought to a central place a B5 program at 2010 average oil prices and still more for pressing into oil. The oil will then undergo a transes- than $1 million/year at $110/bbl. Alternatively, if farm- terification reaction with methanol, using a catalyst. It is ers are paid on the net after costs for cultivation and envisioned that as many as 200,000 hectares that will be processing, then a "reasonable" level of profit will dedicated to Jatropha cultivation in the country. require oil prices consistently above $100/bbl, even in favorable production cases. With yields typical of Reliable and country-specific data are not available on marginal land and corresponding production costs, costs, yields, cultivation practices, and other factors even oil prices at $150/bbl will not produce accept- needed to assess and evaluate the costs and benefits able incomes, net energy ratios, or foreign exchange of the program. The study team used data from similar (FX) savings. programs in Thailand and India, reviewed the experience · With FX savings as an objective, the analysis must reported in the literature, and made a preliminary analy- consider the continued need for imported and oil- sis of the viability of the program under different diesel- based inputs, such as fertilizers and pesticides, blending scenarios and crude oil prices. The conclusions and diesel, methanol, and catalysts for processing reached are as follows: the Jatropha oil. Based on typical yields, estimates indicate that at current oil prices ($75 ± $20/bbl), FX · The land area requirements of the Jatropha program savings are negative (that is, Jatropha is a net FX are a significant consideration for a small country user), especially in light of the full costs of cultiva- like Timor-Leste. At an assumed yield of four tons tion, transport, processing, and blending. The poten- seed per ha, a B5 program (blend of 5 percent biod- tial FX savings become significant only at crude oil iesel and regular diesel oil) would require about 3.5 prices above $100/bbl, and the program represents percent of the country's cultivated land; a B10 pro- an effective use of FX only if Timor-Leste lacks the gram, 7 percent; and a B20 program, 14 percent. The ability to supply itself with domestically produced 200,000 hectares of Jatropha plantations that have hydrocarbons. Even synthetic crude from natural been proposed would be roughly equal to the size of gas is likely to be less expensive than the estimated total cultivated land in Timor-Leste today. costs of Jatropha oil processed for biodiesel blend- · The current Timorese program for Jatropha is predi- ing. The FX calculation includes neither forgone cated on the use of marginal lands. Studies else- export revenues from displaced crops nor additional where have consistently shown that good yields, low food or feed imports resulting from land diverted into unit production costs, and a positive net energy ratio use for energy crops. Thus, these FX impact calcula- (NER) for the final product are achieved only when tions are merely a first approximation and the costs Jatropha cultivation is approached in a commercial are probably substantially understated. manner using reasonably fertile land and appropriate · While Jatropha cultivation undoubtedly increases inputs, such as fertilizer, irrigation water, and pesti- rural employment relative to no programs to cides. To quote the findings of a recent international increase or improve output of agro-forestry crops, it conference on Jatropha, "Marginal lands will pro- is not clear that the cost is advantageous compared duce marginal yields. " with other means of generating rural employment opportunities in the context of Timor-Leste. Jatropha Executive Summary] xv cultivation needs to be compared with alternative that time, biodiesel substitution would make even programs to promote community-based forest man- less economic sense than it does under current agement and protection, and with agro-forestry ini- prices for imported fuels. The alternative of export- tiatives, for example. ing biodiesel is possible only if Timor-Leste enjoys a · Finally, the eventual local availability of petroleum clear comparative advantage over producers in other products from offshore oil and gas production needs countries, such as significantly lower production and to be considered. This could occur in a time frame processing costs and advantageous logistics with of fewer than 10 years, when significant biodiesel respect to potential customers. production facilities would just be gearing up. At 1 Context and Background Timor-Leste, a new country of just 1 million people, is The Rural Energy Sector a picture of contrasts. On the one hand, it is generating and accumulating substantial revenues from offshore oil Predominance of Biomass Fuels and gas resources developed in partnership with Austra- lia. On the other hand, the country is one of the least Table 1.1 provides an approximate comparison of current developed in the world with much of the population living consumption of all types of fuels used in Timor-Leste as in poverty. A decade after the period of violence and civil of 2006. The table shows that the consumption of bio- unrest in the late 1990s, slow but significant strides have mass energy (fuel-wood) exceeds that from all other been made in education, health, agriculture, infrastruc- fuels combined, including hydrocarbons, reflecting the ture development, and other sectors. Still, the majority still highly undeveloped and rural nature of the whole of the population lives in rural areas and most do not yet economy. have access to basic services. Table 1.1 Timor-lesTe 2006: UsefUl energy share of major fUels for all secTors consumption heating value Utilization efficiencya fuel type (tons/year) (mj/kg) (percent) Useful energyb (Ktoe) Percentage share Fuel-woodc 600,000 15.5 15 33.3 51.9 Dieseld 50,000 46.0 40 22.0 34.2 Gasolinee 22,000 47.0 30 7.4 11.5 Kerosenef 3,000 46.0 35 1.2 1.8 LPGg 500 45.0 60 0.3 0.5 Total 675,500 64.13 100.0 Source: Authors' compilation. Note: a. Fuel-wood, kerosene, and LPG efficiencies refer to cooking stoves; diesel to power generation; and gasoline to automobile engines. b. One ton of oil equivalent (toe) = 41.9 gigajoules (GJ). c. From Shum and others, op. cit. d, e, g. Ministry of Natural Resources and Environment (2006), Dili, Timor-Leste. f. From ADB Power Planning Study (2003) that estimated 3­4 million liters kerosene used per year. Asian Development Bank and Government " of East Timor, "Power Sector Development Plan for East Timor, September 2003, Dili. 1 2 Timor-Leste: Key Issues in Rural Energy Policy Table 1.2 soUrces of energy Used for cooKing consist mainly of rehabilitation of existing networks and old diesel plants, grid extensions, and establishment of new decentralized diesel generation. Implementation energy source Urban (%) rural (%) all areas (%) is under way, with the priority being the connection of district towns to the grid. For communities located in Fuel-wood 98.3 98.9 98.7 areas considered uneconomic for grid extension, the Agricultural residues 0.5 0.4 0.4 REMP recognizes the need to exploit local resources Charcoal 0.5 0.7 0.7 where available (mainly oil and gas seeps, and micro- hydro power) for providing electricity through isolated Kerosene 9.2 1.6 3.0 grids. The World Bank has supported the development Electricity 1.2 0.7 0.8 of a small pilot project to demonstrate the feasibility Source: Authors. of local power generation using a gas seep located at Note: The totals exceed 100 percent because some households use Aliambata on the southern coast. Norway has financed more than one type of fuel. a pilot micro-hydro power plant. For smaller, dispersed communities, the REMP has estimated the total number The largest consumer of energy is the household sector of households in each district whose only possibility of and the principal fuel used by almost all households is obtaining electricity service is through individual photo- fuel-wood. About 98.7 percent all households use fuel- voltaic (PV) systems. A consultant company from Por- wood as their main cooking fuel--with agricultural resi- tugal was commissioned by the government in 2008 to dues, charcoal, kerosene, and electricity constituting the carry out data collection and an in-depth assessment of negligible balance (table 1.2).3 About 80 percent of rural the potential for renewable energy applications, includ- households obtain their fuel-wood by collection while ing large-scale grid-connected electricity generation. The about 60 percent of urban households purchase their final report was delivered in May 2010, and it estimated fuel-wood. LPG is used by some higher income house- the nationwide hydro-electric generation potential at 252 holds in urban Dili, but most usage of LPG is by the com- MW, rising to 352 MW if pumped storage is applied. mercial sector (hotels, restaurants, and the like). National wind energy generation capacity was estimated at 72 MW, bringing the total potential for installed renew- Despite the overwhelming predominance of biomass able energy capacity in Timor-Leste to 451 MW. fuels, the 2007 World Bank scoping study found that, except for the Dili district, a positive fuel-wood sup- ply-demand balance is still indicated for most of the Institutional Responsibilities country. The planning and implementation of pilot projects for off- grid electrification with renewable energy, such as wind, Limited Access to Electricity Services micro-hydro, and PV, is the responsibility of the office of the Secretary of State for Energy Policy. The Secretariat Much of the electricity infrastructure was severely dam- of State for Energy Policy is also responsible for policies aged during the period of civil unrest in 1999. Today, the and activities related to the replacement of fuel-wood national electrification rate is about 20 percent but is esti- by alternative fuels (for household energy). Although the mated to be less than 10 percent for rural areas. The total intended use of biodiesel is not for rural energy supply load demand in the country is less than 50 megawatts but mainly to replace diesel fuel used in power plants, (MW). All generation uses diesel systems, including the the Secretariat of State for Energy Policy is carrying out 19 megawatt Comoro power plant serving the Dili dis- the ongoing pilot program to cultivate Jatropha crops. trict. In rural areas, 58 isolated grids provide about six hours of service at night. A systematic and comprehen- sive countrywide rural electrification master plan (REMP) Key Issues in Rural Energy was drawn up by Norplan (initial study in 2004, with an For the rural energy sector the key challenges to the gov- update in 2007). It is the blueprint being followed by the ernment lie in addressing the potentially serious health government for rural electrification and is being imple- and environmental problems associated with contin- mented with assistance from donors. The REMP tasks ued heavy dependence on biomass fuels, and providing improved access to modern fuels, particularly electric- 3. Shum and others, op. cit. A total of 865 households in urban Dili, ity, to all segments of the rural population. The govern- Baucau, and three selected rural areas in Liquica, Aileu, and Ainaro ment has initiated several pilot activities in rural areas districts were surveyed. Context and Background 3 to address these issues. As it proceeds to an expanded Purpose and Organization of phase of these activities, it wants to ensure that the This Report efforts are rooted in sound policies that are consistent with the country's overall goals of poverty alleviation and The government is already implementing several rural economic development. energy development programs in pursuit of the above- mentioned goals. The objective of this report is to help the government, particularly the Secretary of State for Basis of Sound Rural Energy Policy Energy Policy, to develop clear and coherent policies in key areas that would guide planning of the subsequent The primary goal of a sound rural energy policy is to pro- phase of its ongoing programs, the initiation of new pro- mote measures that will enhance the quality of life of grams, and the prioritization of projects competing for people in rural areas by improving their access to mod- a limited total budget. This requires analyzing the main ern energy services. In the context of Timor-Leste, part technical, economic, and social issues related to cur- of the policy is promoting the use of renewable energy rent programs and plans and identifying various policy resources that are indigenous to rural locations and are and strategy options. If data and needed information are environmentally benign. Another key part is promoting unavailable, the analyses use, as appropriate, informa- programs that replace fuel-wood with modern liquid tion from similar situations in other countries, drawing on fuels that are cleaner to handle and produce fewer harm- extensive experience in World Bank projects worldwide. ful emissions. In all cases, the desired approach is one that is environmentally benign and economic both from The present report focuses on three specific areas that are the country's point of view and for the welfare of indi- highly relevant to ongoing and planned programs of the vidual rural households. A crucial objective in the imple- Secretary of State for Energy Policy: (a) improving electric- mentation process is to ensure that the government's ity access for remote, off-grid populations; (b) addressing rural energy programs provide equitable distribution of household energy issues, particularly those arising from benefits, so that populations in the poorest and remotest the need to curtail fuel-wood consumption; and (c) devel- areas are accorded the same attention and opportunities oping biomass-based fuels, particularly Jatropha, cultivated as those living close to the electricity grid. in rural marginal lands, as a substitute for hydrocarbons. 2 Electricity Access Policies The official goal of the government is to achieve 80 per- Off-Grid Electrification cent electrification by 2025. Considering the current low electrification rate of about 20 percent and the damaged The REMP recognized that many communities and or substandard condition of the existing infrastructure, households located in areas far from the electricity net- the goal is ambitious. Nevertheless, a phased national work or across difficult terrain are not expected to be electrification plan--the Rural Electrification Master Plan reached by grid-extension programs for at least the next (REMP)4--was prepared in 2003 and later updated and 15 years. These communities are often small, dispersed, . initiated in 2007 The REMP was based on calculations and poor. Government policy should be to provide at of present and projected demand, inventory of available least basic electricity services to such off-grid communi- generation capacity, and rough estimates of potential ties, through programs implemented in the same time generation from local resources. The REMP systemati- frame as those directed at concentrated communities liv- cally prioritized districts, subdistricts, and sucos, based ing near the network. The implementation policy should on the principle that electrification should follow geo- be based on principles of least-cost power planning, with graphically the areas with the highest economic activity the objectives of technical and financial sustainability. and potential growth. The long-term program would add about 100,000 potential new rural customers at a total For sucos or parts of sucos considered in the off-grid cost of about $87 million. The first phase, estimated to category, access to electricity could be provided through cost about $40 million, would add about 50,000 poten- decentralized mini-grids powered by isolated diesel tial new customers. The time frame for the first phase gensets, local energy resources (if available), or by pho- would be five years or more, depending on availability of tovoltaic (PV) systems. In general, international experi- financing and development of local construction capacity. ence with diesel-based systems in remote areas has Electricidade de Timor-Leste (EdTL) has been implement- not been favorable. In Timor-Leste, as well, the diesel , ing the tasks defined in the REMP starting with district- systems managed by EdTL serving district capitals have level refurbishments and connections. Issues related to been found to be expensive and difficult to operate and financing, staffing, and other factors have severely lim- maintain, particularly with the low tariffs charged. The ited the capability of EdTL to meet targets. These issues problems experienced with large systems are likely to and others related to the conventional, grid-extension be exacerbated with small, suco-scale diesel systems part of the government's rural electrification program are scattered in relatively remote areas. For new decentral- being addressed by other assistance activities and are ized power systems, it is good policy to put priority on not discussed further in this report. harnessing locally available energy resources, especially environmentally benign renewables, wherever they are 4. S. Grongstad and B. Stenseth, "Update Rural Electrification Mas- available. " ter Plan Report, Norplan, 2007. 5 6 Timor-Leste: Key Issues in Rural Energy Policy Decentralized Mini-Grids Even when local energy resources are available within 5 kilometers or so of load centers, the least-cost option for The most likely local energy resources to be exploited in the government may be solar PV and not the develop- Timor-Leste are already-identified oil and gas seeps and ment of the local resource. This is often the case when (a) hydro resources.5 Initial estimates by Norplan indicate the potential users are highly dispersed and thus expen- that, out of a total 442 unelectrified sucos, about 157 sive to connect together, and (b) the most significant use have oil or gas seeps or hydro resources located within of electricity is only for lighting and small domestic pur- 5 kilometers, of which about 125 are hydro and 32 are poses (radio, small TV, and the like). Although the oper- oil and gas seeps. If screening criteria similar to those ating costs are low compared with diesel, developing a for grid-extension projects are applied, about 60 of these local energy resource, whether a gas seep or hydro, then sucos could be considered priorities for electrification, of building a power plant and constructing a mini- or micro- which 34 would be hydro and 20 oil and gas seeps. These grid network, normally results in a very high investment numbers are imprecise estimates but provide an idea of cost per user. It is easier to economically justify the proj- the size of the decentralized power needs of the rural ect when the total load demand is high, such as when a electrification program compared with the grid-connected large productive load, for example, a grain milling machine part. The individual scale of power generation from these or cold storage for fish and foodstuff, is envisioned to be decentralized resources and the number of connections built. However, care must be taken when assessing the possible depend on highly site-specific factors. Detailed true likelihood of an economic activity occurring in the localized investigations are obviously needed before any community when electricity becomes available. Many decision is made to finance individual investments. past consultant studies, particularly for micro-hydro, have proved overoptimistic in this regard, resulting in projects 5. In principle, biomass resources, such as animal wastes digested with significant unutilized power output, high generation to yield biogas or agricultural residues directly combusted or gasified costs, and high investment costs per connection. Estab- for power, are also possible but are unlikely to be significant for most lishing and sustaining a microenterprise requires some parts of Timor-Leste. Small wind power systems could also be used basic entrepreneurial and business skills on the part of but average wind speeds in Timor-Leste are generally not high. A period of systematic wind measurements at potential project sites the local residents who will manage the enterprise, even will need to be carried out first. when the establishment costs are fully borne by the Table 2.1 disTribUTion of PoTenTial PV cUsTomers by disTricT number of grid cost per grid number of potential district customers customer ($) PV customers PV installed in 2008 PV installed in 2009 Aileu 2,160 784 3,240 Ainaro 4,040 719 3,557 Lautem 5,574 847 5,427 58 Manufahi 3,490 603 5,111 Manatuto 4,670 628 2,893 94 Viqueque 4,775 263 5,990 240 772 Covalima 6,620 640 4,905 40 10 Baucau 7,320 415 8,649 6 Liquica 1,690 1,116 1,739 Ermera 5,825 645 7,735 96 585 Bobonaro 8,125 772 7,627 97 145 Oecussi 8,365 692 2,189 Total 62,654 59,062 473 1,670 Source: REMP (Norplan 2007). Electricity Access Policies 7 project. In most cases, these skills do not exist in remote from about $100 for the lantern system to about $1,500 communities of Timor-Leste. for the 100-watt SHS. For Timor-Leste, an SHS of about 50 watts, capable of powering two to three lamps for three to four hours a night, would be sufficient for the Solar Home Systems needs of most remote households and could be made Norplan estimates that about 60,000 households nation- the basis for program planning. This size has gradually wide cannot be electrified by grid extension or by local emerged in past Bank projects (for example, in Sri Lanka mini-grids in the foreseeable future and considers such and China) as sufficient for the majority of households households to be appropriate for stand-alone solar PV and most popular in price-performance ratio. The aver- systems. The breakdown of potential PV customers age international retail price, including installation, of a ranges from about 1,740 in Liquica to about 7,600 in Bob- 50-watt SHS is about $700­$800 today. This price could onaro. In 2008 and 2009, more than 2,000 PV systems be lowered significantly if the units were purchased in for households and community centers were provided large volumes and if installation of many systems in a by the government of Timor-Leste to sucos in Viqueque, remote area could be made in one trip by the provider. Ermera, Bobonaro, and a few other districts (table 2.1). subsidy issues To date, the PV installations have been fully subsidized The high upfront cost of SHS precludes its acquisition by by infrastructure funds provided annually to the sucos almost all households if not subsidized. In most past and by the central government. Suco leaders submit applica- ongoing PV projects financed by the World Bank, user tions for PV installations in their communities. The Secre- subsidies average about 50 percent of the upfront cost, tariat of State for Energy Policy reviews the applications with the actual level depending on the size of the sys- and recommends approval. When approved, the suco tem, the business model employed by the project, and leaders negotiate directly with private PV companies to the general policy of the government with regard to elec- install the systems. Recipient households do not contrib- trification of off-grid areas. In countries with high elec- ute to the purchase cost but are expected to operate and trification rates, very small unserved populations, and maintain their own systems. The Secretariat of State for robust sources of subsidy funds (for example, Mexico Energy Policy considers the ongoing effort to be a pilot and Argentina) a much higher level of subsidy has been and hopes to learn from the experience before embark- applied and sustained. Aside from the objective of finan- ing on a full-scale program for electrification with PV. cial sustainability of the program, however, a key purpose of requiring PV recipients to contribute part of the sys- costs tem acquisition cost is to instill a sense of ownership. A Most households in remote sucos require electricity wealth of evidence from past projects has shown that mainly for lighting, and for small domestic power appli- PV dissemination mechanisms in which systems are just cations such as radios, small TVs, and the like. These given away or fully subsidized fail within a short period. applications could be adequately met by individual PV systems. Moving from kerosene lamps to PV-powered In Timor-Leste, a policy for providing a relatively high level electric bulbs results, for the household, in a marked of subsidy for PV users may be justified on social equity improvement in the quality of lighting and a reduction in grounds (households in remote areas deserve attention indoor air pollution; for the country, it results in significant similar to that received by households living near the reductions in greenhouse gas emissions. However, even grid) and the small total number of households that need considering current household expenditures for kerosene to be provided individual systems (manageable total cost and candles for lighting and dry cells for small appliances, to the government). In Phase 1 of the REMP involving up the cost of shifting to PV can exceed the ability and will- to 63,000 potential customers, the average investment ingness to pay. cost per grid connection is estimated to be about $650, of which about $10 is paid by a majority of the customers PV systems for households range in capacity from about as a connection fee. In addition to the investment cost 10 watts for a portable solar lantern to 100 watts or more subsidy, households are charged a tariff of only $0.12 per for a large solar home system (SHS).6 The price can range kilowatt hour or $2.40 per month for a household con- suming about 20 kilowatt-hours. The tariff is well below 's EdTL cost of generation and distribution--about $0.27­ 6. An SHS comprises a solar PV panel, a battery, regulator, and lamps. The battery could be an ordinary car battery, which lasts about two $0.35 per kilowatt-hours. years or less, or a deep-cycle battery that could last for four to seven years. The lamps are typically 10-watt compact fluorescent lamps The above suggests that providing an upfront grant for PV but, recently, lower wattage light-emitting-diode lamps have been of about $640 per household while charging a monthly increasingly used. 8 Timor-Leste: Key Issues in Rural Energy Policy maintenance fee of about $3 (for battery replacement and Timor-Leste does not yet have an environment that would similar needs) would be roughly equivalent to the situa- support implementation of a dealer model: there are few tion of grid-connected rural households. The government or no companies that could be provided incentives to may decide under some circumstances to pay the full compete in selling PV; the micro-financing system is still upfront cost of a PV installation but, to ensure sustain- undeveloped; and most important, the current PV market ability, leave the full cost of operation and maintenance consists mainly of households belonging to the bottom (replacement of battery, lamps, and so on) as the user's part of the income pyramid. Some type of fee-for-service responsibility, to the extent possible. The appropriate mechanism executed by the private sector but with a level of capital cost subsidy for Timor-Leste will depend strong role for the government would appear to be more on the actual PV system costs at the start of implementa- suitable, at least for the next 5 to 10 years. A suggested tion of a full-scale program based on the mechanism dis- mechanism and implementation procedure for an SHS cussed below, the estimated total budget requirements, program are outlined in annex 1. and the source of funding decided upon by the govern- ment. Although other factors obviously need to be con- The approach would group several communities into a sidered, these policy considerations should be the basis package for competitive bidding, improving on the cur- for designing an implementation mechanism or model rent practice of letting the sucos negotiate prices of for the next phase of the government's PV program. installations individually with vendors, thus potentially reducing the total cost to the government. However, implementation options given the low implementation capacity of government Although several variations and combinations exist, the staff, technical assistance should be provided at the implementation mechanisms or business models for PV start to provide sufficient training to staff on the optimal dissemination that have been used in most countries gen- grouping of prospective communities, preparation of bid- erally fall into two basic categories: (a) dealer model (also ding documents, and conduct of the bidding process. called direct sales or open market), and (b) fee-for-service The mechanism outlined in annex 1 follows the typical model (energy service company or concession system). concession model but should be implemented in a highly In the dealer model, the consumer purchases the sys- simplified version, given constraints on local capacity. It tem either with cash or financing, and owns it. Beyond can be thought of as a simple extension of the winning warranty service, the consumer assumes responsibility supplier's responsibility to provide and install the equip- for all operational and replacement costs. In the fee-for- ment, then later to operate and maintain the equipment service model, the consumer is simply provided electric- in groups at the sites for an extended period. ity service. The company, which retains ownership of the equipment, is responsible for maintenance and providing The suggested mechanism results in a minimal monthly replacement parts over the life of the service contract. payment by each household to the service provider that box 2.1 exPerience wiTh solar baTTery charging sTaTions in nicaragUa In indigenous communities of Nicaragua's remote Atlantic Zone, seven SBCSs, each with a 2-kilowatt capacity, were installed in 2006. Each SBCS served some 50 households, and each family was provided a battery and lighting kit. The Nicaraguan government bore the capital cost of the stations and initial battery expenses. Beneficiary communi- ties were trained to operate, financially manage, and maintain the stations. Each family paid a monthly fee of $5 to cover weekly battery charging and contribute to a fund for buying replacement batteries. The original concept was to allow families to charge their batteries only when they had available cash (much like the retail buying of cooking oil or fuel-wood), but the concept proved unworkable in practice. To sustain the station busi- ness, however, each user family had to commit to regular monthly payments, which became a major stumbling block for this off-grid approach. Although community organizations managed SBCS operations well, the users--mainly poor subsistence farmers--eventually could not afford the monthly fees. The Off-Grid Rural Electrification Project (PERZA) is addressing this problem by working to raise farmers' incomes. For example, the project has developed a customized micro­business services program that assists in the bulk transport and marketing of crops and livestock and advises on agricultural matters. It has also arranged for noncash payment for battery charging. Electricity Access Policies 9 should be roughly equivalent to the amount the house- In countries where PV programs are based on the hold normally pays for traditional systems, for example, demand-driven dealer model for SHS, the market served kerosene lamps, dry cells, candles, and so on. Even with consists only of those in the middle and top portions of this small monthly responsibility, experience worldwide the income pyramid. In such countries, SBCS is often has shown that at some point many families become used as a way to address the needs of the poorest off- unable to pay. This is to be expected from any socially grid households, with the government paying for the full oriented energy provision program that serves the capital cost of the stations, initial batteries, and lighting lowest income communities. The only way to increase kits. A 2-kilowatt SBCS would cost about $20,000. the chances of sustainability of such energy provision pro- grams is to combine them with other agricultural or rural In principle, SBCS is also a fee-for-service PV business development projects that actively try to raise incomes in model. But compared with the SHS model previously the target communities. discussed, no solar panel is located at the user's house, and the user must manually carry the battery to the sta- tion weekly for charging. One advantage of SBCS is a Solar Battery Charging Stations small economy of scale in capital costs achieved when A solar battery charging station (SBCS) is an option that compared with acquiring several SHSs of the same total could be considered for electrification of remote house- capacity. A second advantage is that the mechanism does holds through PV. A typical SBCS would consist of an not rely on an external service provider. However, the array of PV panels totaling about 2 kilowatts of capacity resident local operator is often not capable of addressing and a special controller that enables the charging of up to any significant technical problems and still needs some eight batteries simultaneously.7 The station could serve external support. A requirement or disadvantage is that about 50 families, providing each with service equiva- client families must be situated relatively close to the lent to that of an individual 40-watt-peak SHS. The family station; otherwise, carrying the heavy battery back and would bring the battery for charging to the central station forth to the station becomes impractical. Where off-grid once a week. Normally, each family pays a fixed monthly situations can be found in Timor-Leste with sufficiently fee to cover the salary of the operator (a trained local concentrated but small numbers of poor households, the resident) and to amortize the cost of the next replace- SBCS option may be worth considering. Again, as with ment battery. any model that serves very-low-income households and requires periodic payments to maintain operation, long- term sustainability is more likely when combined with other programs that raise the income of the user com- . 7 Project Appraisal Document, Nicaragua Off-Grid Rural Electrifica- munity. (See box 2.1.) tion (PERZA) Project, World Bank, 2003. 3 Household Energy Policies This chapter addresses the use and replacement of such a policy. First, the heavy use of fuel-wood by house- fuel-wood for cooking through (a) substitution of mod- holds in very inefficient traditional stoves wastes tre- ern fuels, (b) an intermediate step of improved cooking mendous amounts of the country's biomass resources. stoves that still use fuel-wood, and (c) domestic biogas Second, population growth is certain to erode the current systems. overall positive supply-demand balance for fuel-wood, as a consequence of increased total wood consumption coupled with decreased area of forest lands resulting Fuel-Wood: Key Issues from conversion to agriculture. Development of a sound policy for fuel-wood use and Another important reason for the government to curtail its potential replacement by modern fuels must take into the use of fuel-wood relates to the public health impacts account a number of factors. Almost 99 percent of all of its use. Strong correlations have been found in many households in urban and rural areas in Timor-Leste use low-income countries between exposure to indoor air only fuel-wood for cooking, making this domestic use of pollution (IAP) from the use of solid fuels (smoke from biomass energy responsible for about half of all delivered traditional cook-stoves burning biomass) and acute respi- energy from all fuel types (including petroleum fuels) ratory infections, such as tuberculosis and pneumonia. used in all sectors.8 This reflects the still-abundant bio- For Timor-Leste, a World Bank Country Environmental mass supply in most parts of the country, except in the Analysis conducted in 200810 estimated that the annual Dili district.9 As a result of this abundance, fuel-wood is, morbidity and mortality cost of health effects from IAP and will be for some time to come, the cheapest cooking associated with the use of solid fuels is between $5 mil- fuel in Timor-Leste compared with liquefied petroleum lion and $20 million (see table 3.1). The mean estimate is gas (LPG), kerosene, and electricity, even after account- equivalent to about 1.4 percent of Timor-Leste's GNI, or ing for different cooking equipment efficiencies. . 3.5 percent of its 2006 GDP.11 There is no doubt, however, that the responsible policy The rationale for a policy curtailing fuel-wood use is clear for the government to pursue must be the curtailment but the issue is what realistically can be achieved in of fuel-wood use to the greatest extent possible. From Timor-Leste and at what cost. It may not be possible to the supply viewpoint, two main factors drive the need for . 10. J. Bojo and F Ruiz Nunez, "Timor-Leste Country Environmental 8. The use of biomass fuels in rural industries is insignificant in Timor- " Analysis, World Bank, 2009. Leste. 11. Timor-Leste's GDP is $356 million (current 2006 dollars; World 9. Shum and others, op. cit. Bank, 2007). 11 12 Timor-Leste: Key Issues in Rural Energy Policy Table 3.1 ToTal healTh cosTs from solid fUels Use in Timor-lesTe ($ Per year) lower bound central Upper bound ARI, Children < 5 Morbidity 277,283 1,651,832 3,495,129 Mortality 3,900,059 4,953,133 6,006,206 Total ARI 4,177,342 6,604,965 9,501,336 COPD, Adults 30 Morbidity 270,782 379,379 489,066 Mortality 533,901 5,457,434 10,380,966 Total COPD 804,684 5,836,813 10,870,031 TOTAL 4,982,025 12,441,777 20,371,367 % of GNI 0.58 1.44 2.35 . " Source: J. Bojo and F Ruiz Nunez, "Timor-Leste County Environmental Analysis, World Bank, 2009. Note: ARI = Acute Respiratory Infection; COPD = Chronic Obstructive Pulmonary Disease. significantly change the pattern of fuel-wood consump- The latter does not necessarily entail launching separate tion using substitution or demand-side management pro- efforts for managing supply of fuel-wood species but grams for at least the next decade. However, the surplus could be part of Ministry of Agriculture, Forestry and supply situation also suggests that it may be possible to Fisheries (MAFF) programs for management of forest buy time for the transition to other fuels with sustainable resources, in general. forest management programs that could enable contin- ued use of fuel-wood for some time, until modern fuels This report discusses issues and options for three dif- could make significant inroads into the demand picture. ferent approaches: (a) substitution of LPG and kerosene for fuel-wood, (b) promotion of higher efficiency cook- The strategy, therefore, is to pursue programs that com- ing stoves, and (c) promotion of domestic-scale biogas bine both demand-side management or interfuel substitu- systems. tion programs with programs for improved management of fuel-wood supply. box 3.1 biomass briqUeTTes as a sUbsTiTUTe for fUel-wood Biomass briquettes could be made from a variety of agricultural residues available in Timor-Leste, such as rice husks and coffee shells. The process is relatively simple, requiring only a mechanical or hydraulic press to compress the residues into a dense solid fuel that could be used in stoves. Programs for densification of biomass residues have been successful in regions that have severe fuel-wood shortages, such as parts of northern Vietnam and China. In such areas, many people have even resorted to using agricultural residues, such as rice straw, directly in stoves in their unprocessed or loose form. The problem with promoting biomass briquettes in Timor-Leste is not with the technology or availability of raw mate- rials for densification but simply the fact that the fuel of choice, wood, is still widely available for collection in almost all parts of the country. Even if biomass briquettes are sold in markets at a price much less than fuel-wood, it is highly unlikely to result in any significant shift by households to this alternative. It is appropriate for technical agencies and nongovernmental organizations (NGOs) in Timor-Leste to familiarize themselves with densification technologies and to conduct pilot activities, as is being done today. Such efforts, however, are more in the nature of preparing for a future of reduced fuel-wood supplies rather than expecting any possibility of having briquettes compete with fuel- wood now. Household Energy Policies 13 Interfuel Substitution with fuel-wood versus $48/GJ for electricity and $77/GJ and Modern Fuels $81/GJ for kerosene and LPG, respectively. In addition to the direct cost of fuel, which is greater for refined hydro- Compared with fuel-wood, LPG and kerosene12 are more carbon fuels, there is also the question of the cost of the convenient and cleaner to handle. LPG is the clean- equipment to use that fuel. Kerosene and LPG both intro- est, with the lowest amounts of pollutants emitted per duce significant costs of shifting, including the purchase kilogram burned, thus best for reducing indoor air pol- of new stoves or burners and the purchase of or cash lution.13 With both kerosene and LPG now sold at market deposits for storage cylinders for LPG. prices,14 however, those fuels are more costly per unit of delivered energy for households, even after account- There are four main options for promoting the increased ing for the significant efficiency differentials between adoption of modern fuels: wood stoves and kerosene or LPG stoves. According to the Bank's 2007 scoping study, in which consumers pur- 1. Subsidize the price of fuel (LPG and kerosene); chased wood, primarily in urban areas,15 the average cost 2. Subsidize the new equipment (stoves, cylinders); of delivered energy was about $43 per gigajoule (GJ) for 3. Provide vouchers to low-income customers only for fuel purchases; or 4. Provide vouchers to low-income customers only for 12. Modern cooking fuels usually include LPG, kerosene, and elec- equipment purchases. tricity. In Timor-Leste, even in more-affluent homes, electricity is used only in rice cookers. In most Asian countries, the growth of the middle class has resulted in rapid growth in the use of LPG. Either of the two fuel subsidization options can be com- 13. Kerosene actually has about the same level of pollution as fuel- bined with either of the two equipment subsidization wood, but because of its higher energy density and the higher effi- options, leading to a significant menu of policy choices. ciency of stoves that use kerosene, the amount of pollutants emitted for cooking the same meal is much lower. These choices result in significant differentials (see fig- 14. Before independence Timor-Leste was part of a national pricing ure 3.1) in the effects on beneficiaries, the cost to indi- scheme in Indonesia that provided significant subsidies to kerosene viduals and the government, and the likely efficacy of the users. subvention. 15. In rural areas 80 percent of wood supplies are collected by the household. figUre 3.1 imPacTs and efficacy of Programs To PromoTe modern cooKing fUels Favors poor consumers FS + VES Distribution of benefits VES only ES + VFS Slower Speed of implementation Faster VFS only ES only FS only FS + ES Favors more-affluent consumers Source: Authors. Note: FS = fuel subsidy; ES = equipment subsidy; VFS = voucher system for subsidized fuel; and VES = voucher system for subsidized equipment. 14 Timor-Leste: Key Issues in Rural Energy Policy Programs That Subsidize Fuel, or Programs That Subsidize Fuel, or Equipment, or Both Without Income Tests Equipment, or Both with Income Tests Programs that subsidize both fuel and equipment with no Programs that use income testing are identified as 3 and questions asked (identified as 1 and 2 in the list above), 4 in the list above. Voucher programs take a long time would obviously be the most effective and the fastest to to establish. The government will need to have accurate implement. Of the two options, subsidizing the fuel offers census and income tax information to ensure that only continuing value to consumers and has been shown in the targeted income groups receive the vouchers. Even other countries to result in significant and rapid switching if implemented properly, such programs will generally be from fuel-wood within a short time frame. It tends to be less effective than a no-questions-asked subsidy program. more effective because the process is convenient and The voucher options show up on the left side of figure the costs are hidden from consumers. Such programs 3.1. Voucher programs tend to be less successful in pro- establish expectations on the parts of consumers of a moting quick adoption of modern fuels because (a) they low--and stable--price for fuel. carry high administrative costs; (b) unlike a direct price subsidy, the financial costs of the program are apparent The main disadvantage is that many of the beneficiaries through the printing of the vouchers, thereby reducing of such unrestricted subsidies are often people who do the appetite of the government to expand it; and (c) the not need them. For example, in some countries (Indone- system is conducive to corruption--the emergence of a sia, for instance) subsidized kerosene has been diverted secondary market in vouchers is almost inevitable. for use as engine fuel, thereby benefiting car drivers, owners of small electricity generators, and industry.16 Subsidies without restrictions could create incentives for Combination Programs of smuggling the fuel or equipment out of the country to be Direct and Voucher Subsidies sold (or re-imported) at international prices. To better target the subsidies, some countries have attempted various combinations of approaches. Combin- 16. This is less of a problem with LPG because LPG requires a ing voucher programs with direct subsidies for equipment more complex engine conversion process to enable its use in lieu and fuel will tend to increase the administrative complex- of gasoline. ity and costs, possibly slowing the pace of adoption of box 3.2 lPg sUbsidies in india: growTh of demand and financial liabiliTies LPG was introduced in India as a domestic fuel in the 1960s. The demand for LPG grew from less than 200,000 tons in 1970­71 to about 5 million tons (MT) in 2000. The average growth rate in demand has been about 12 percent annu- .5 ally. In the 1990s, LPG demand was estimated to grow to 7 to 9.0 MT by 2001­02 and 10.0 to 12.0 MT by 2006­07 . The projected supply was only about 4 to 5 MT in 2001­02, leaving a gap of about 3 MT. The expected increase in demand is partly due to the expansion of the Indian middle class, which sees subsidized LPG as improving their lives. The introduction of differential subsidies and taxes on various products led to a misallocation of both private and pub- lic expenditures for selected petroleum products, leading to a burgeoning demand for the subsidized ones, supplied increasingly through imports. The large subsidy on LPG combined with that on kerosene and the historical subsidy burden contributed by diesel, together with infrequent adjustments to pooled prices and a mismanagement of the pool account, built up a deficit of $4.42 billion (184.4 billion rupees) by 1997­98, $6.5 billion in 1998­99, and almost $14 billion in 2000. In April 2002, the government announced that subsidies for all petroleum-based products would be phased out except for LPG and kerosene, which, the government pledged, would see their subsidies phased out within a three- to five-year period. LPG and kerosene are used as domestic cooking fuels by a large portion of the population. The issue, however, was not resolved and left to a later government. " Source: Center for Energy Economics, "LPG Subsidies in India, The University of Texas at Austin, Texas, 2005. Household Energy Policies 15 LPG or kerosene to less than what either voucher pro- verting, say, 15,000 households annually would amount gram alone might accomplish. to $1.5 million per year, about the same cost as would be incurred by the government to support Jatropha growers for a small scale (B5) biomass liquid fuel program. Better Practice: Subsidize Equipment, Not Fuel There would be no other government costs, and if the In rural electrification programs, the guiding principle for program were terminated after five years, overall costs sustainability (although not always observed) has been to .5 would be capped at $7 million, and about 75,000 house- subsidize only the construction of the system but not the holds, representing a significant proportion of the popula- purchase of electricity by consumers. The basic policy is tion, would have access to modern fuels. This represents to subsidize access, not consumption. In the household a modest public cost for potentially effecting a significant energy sector, the analogous policy--an alternative that change in the fuel mix. has worked well in some countries--is to subsidize the equipment needed to make the switch to modern fuels Likely Continued Use of Fuel-Wood Stoves but to keep fuel prices at market levels. This enables poor households to surmount the barrier of high first Even with full-scale switching to modern fuels, including costs but avoids recurring and increasing costs to the electricity, for cooking, experience in many countries indi- government. cates that fuel-wood stoves would continue to be used in a significant proportion of households. Fuel-wood stoves Although, as discussed above, the pace of switching by are often needed to supplement the LPG or kerosene households will be significantly slower than in a program supply when these fuels run out or when cooking costs that subsidizes the fuel price, it prevents the problem need to be trimmed. In other cases, it is simply a cul- of the government committing to a subsidy scheme in tural preference and households desire the flexibility of which costs are certain to grow very rapidly in the coming using multiple types of fuel. Nevertheless, it is expected years, as economic conditions improve and a middle class that both the volume of demand for fuel-wood overall, emerges and expands in Timor-Leste. The fiscal impact of especially in urban areas, and the rate of growth of wood the fuel subsidy requirement may become unmanage- demand, would tend to moderate as a modern-fuels able. At that point, as experienced in many countries, it program is fully implemented. becomes extremely difficult if not impossible for the gov- ernment to significantly reduce or remove the fuel subsi- dies because users would suddenly be faced with greatly Improved Cooking Stoves as an increased costs of a convenient fuel to which they have Intermediate Approach become accustomed (see box 3.2). Promoting higher efficiency or "improved" cooking In contrast, the financial exposure of the government for stoves (ICS) that still burn fuel-wood obviously does not an equipment subsidy program can be determined annu- have the same impact on fuel-wood use or IAP reduction ally, in advance. The government may stop the equip- as promoting full interfuel substitution with modern fuels. ment subsidy program at almost any time with minimal It should not be viewed as encouraging continued use of public inconvenience. The only negative effect would be fuel-wood but as a necessary intermediate approach-- a reduction in the number of new households switching the only practical way for the government to reduce fuel- to modern fuels. Those who are already using LPG or wood consumption on any significant scale while the kerosene would be unaffected. economy transitions to a future in which modern fuels are widely available and most households can afford It is not possible to quantify the costs and benefits of them. The benefits associated with modern fuels, such these different options at present. However, the finan- , as lower levels of harmful IAP higher energy efficiency, cial parameters of an equipment subsidy program can be established. Suppose, for example, that the costs of supplying a household with an LPG stove and an LPG from a food flavor standpoint. This has certainly been the experience elsewhere when the two fuels have been sold without subsidies. cylinder deposit is $100.17 In that case the cost of con- However, if lighting is a consideration, then support for kerosene equipment would be relevant. Also, if there is domestic produc- tion of natural gas, a small separation plant could be built for the . 17 It is assumed for the sake of this discussion that consumers who local market, something that is virtually unthinkable for crude oil or are receiving the equipment for free will prefer LPG to kerosene kerosene. 16 Timor-Leste: Key Issues in Rural Energy Policy and cleaner kitchen environments, are also obtainable to 25 percent of households in the Dili and Baucau districts, a certain degree with improved stoves. By their design, could be carried out at a cost of less than $350,000.21 improved stoves burn wood more efficiently than tradi- This amount is insignificant compared with even a frac- tional ones and thus have lower emissions.18 tion of the annual cost of health effects from IAP associ- ated with the use of solid fuels, as shown in table 3.1.22 A 100 percent subsidy to users in such a program could, in Initiating an Improved Stoves Program: theory, be more than justified, but is not recommended. Subsidy Issues The improved stoves should be marketed at a price that Ideally, improved biomass stoves programs should be is low enough to be affordable and attractive to buyers, based on sustainable commercial practices, that is, users but high enough that the stoves are not perceived as a buy the stoves at full cost, producers and marketers giveaway of no value. For the conceptual program out- make a profit and are able to produce and market more lined in annex 2, a retail price of about $2 for a higher effi- stoves, and so on. Subsidies provided to the stoves pro- ciency portable clay stove that costs about $7 to produce gram would be solely in the form of technical assistance, and market could be considered. training, and promotion. Stoves programs in some coun- tries have been implemented under this ideal principle; Given various implementation constraints, launching most other programs, however, could only be carried out a much larger ICS program is not possible. The initial by providing subsidies that reduced the final cost to the five-year, 15,000 "urban" stoves program outlined in users, at least to the amount that they normally would annex 2 would first focus on the Dili and Baucau urban have paid for their traditional, inefficient stoves. Govern- and peri-urban markets, where most families buy rather ments justify the subsidies on environmental (curbing than collect their fuel-wood and hence may be more deforestation) or public health (reducing IAP) grounds. easily motivated to switch to ICS.23 Depending on the In all cases, however, these successful large-scale pro- results, the program could be expanded to an additional grams were implemented in environments and cultures 20,000 units in the next 5 to 10 years, with expanded in which users were accustomed to purchasing their bio- geographic coverage. Initiating a program of even mod- mass stoves, and replacing them when they break by est scale as early as possible is good policy because it purchasing another one in the market. enables the necessary long-term market structure to be put in place slowly, and potential consumers can begin to In Timor-Leste, however, virtually no households pur- absorb information about the various benefits provided chase their fuel-wood stoves and, except for brief peri- by improved stoves. ods when some specially funded programs require it, no one produces stoves to be marketed.19 In addition, intro- After the initial five-year program, when the public has ducing improved stoves commercially is severely hin- become more familiar with improved stoves and their dered by the continued relative abundance and low cost benefits, the retail prices of the stoves will be raised of fuel-wood in Timor-Leste, making it hard to motivate by $1 per year so that at the 10th year all subsidies are households to economize. The only way for improved removed. Demand is expected to drop after prices are fuel-wood stoves to be disseminated widely in Timor- raised but is expected to stabilize at a certain level even Leste is to subsidize their costs to users substantially. after full removal of retail subsidies. Unlike subsidizing The questions for policy makers are whether the benefits LPG or kerosene fuel, which is not recommended in this outweigh the costs of subsidies, whether the size of the report, the total subsidy requirement for the proposed subsidies required is manageable, and whether there are improved stoves program over 10 years is a manageable less costly alternatives.20 amount and could be stopped at any time by the govern- ment without drastic welfare impacts on users.24 Fortunately, the cost to implement an ICS program is small compared with the cost of other types of interven- 21. Based on stoves cost only. Associated program costs should be tions such as subsidizing LPG and kerosene for domestic added. cooking. For example, a five-year improved fuel-wood 22. Obviously, the full health impact costs of IAP cannot be avoided stoves program targeting 15,000 households, or about by an improved fuel-wood stoves program because it only reduces, not replaces, wood use. 23. The initial five-year program will also include some 200 demon- 18. World Bank, "India: Household Energy, Indoor Air Pollution, and stration "rural" stoves, which are massive and must be constructed " , Health, ESMAP World Bank, Washington, DC, 2002. in place because of rural households' different cooking patterns. 19. Shum and others, op. cit. However, it is expected that demand for the portable urban stove 20. In the immediate term, because of the current overall fuel-wood exists in rural areas as well, so they will be actively promoted later in surplus situation, benefits related to curbing deforestation impacts these areas. See fuller discussion in annex 2. would be small and localized. 24. Unlike the case of withdrawing LPG and kerosene subsidies Household Energy Policies 17 Monitoring Indoor Air Pollution Under the proposed program, one small family-size digester (FSD) capable of producing about 2 cubic meters As part of the ICS program, a study should be carried per day (m3/d) of biogas would be shared by three house- out to evaluate the influence of stove type on kitchen holds, using wastes from a total of four to six head of air pollution levels and women's exposure to particulate cattle owned by the families. The FSD would be of Chi- matter from wood smoke. IAP levels and exposure levels nese fixed dome design with a volume of about 10 m3. would be measured in selected homes before and after Based on international experience, approximately 1 m3 of introduction of the improved stoves. The design of the biogas is required for the daily cooking needs of a family measurement regime, its execution, and interpretation of four. The average size of a Timorese family is almost of the data require specialized skills and equipment. It is twice this number and, in principle, a single family could recommended that the study be carried out by interna- use all of the output of the 2 m3/d digester. However, tional experts in this field. The results could indicate to the capital cost of this small digester is significant com- what extent the improved stoves are actually reducing pared with rural incomes and very few families in Timor- IAP impacts. Leste own enough cattle to supply the amount of waste required daily. Domestic Biogas Systems Costs and Benefits The production of biogas by anaerobic digestion of ani- To guide policy making in this area, an idea of the costs mal wastes has been practiced in various parts of the and benefits of different options for using family-sized world for many years, notably in China, India, and other biogas systems needs to be obtained. As yet, however, South Asian countries. Biogas can be used as fuel for no operational and cost data from the biogas installations cooking and lighting, and fertilizer can be produced as in Timor-Leste are available, so it is difficult to make a a byproduct from the sludge. In Timor-Leste, a program detailed quantitative assessment. Nevertheless, some has been proposed to provide biogas systems to some estimates can be made using information collected 4,600 rural households, which is the estimated number from the team's field mission and from international of households that own at least one to two head of cat- benchmarks. tle. The main purpose of the program is to replace the fuel-wood presently used in these homes with a clean The cases analyzed in this report consider the use of the and renewable fuel substitute.25 FSD in two ways: The technical feasibility of a biogas program is not in · The FSD replaces only fuel-wood for three house- question. The key issues are economic and distributional. holds, and Given the limited financial resources of the government, · The FSD replaces fuel-wood and kerosene (for light- and mindful of other programs to provide modern energy ing) in one household. to rural areas, can such a program, with an initial cost of at least $3 million, contribute more to the country's energy In addition to fuel benefits, the economic analysis counts supplies than some other use for that same money? the benefits of producing nitrogen fertilizer from the And will these funds go to those who can best use the digester sludge. The FSD will theoretically produce the resources for household betterment and improved stan- equivalent of 30­45 kilograms/year of urea fertilizer. dards of living? However, much of this fertilizer benefit is already gained by farmers who spread manure on their fields. Hence, after years of providing them and after the market has grown enor- the incremental benefit from this byproduct is relatively mously, withdrawing subsidies after several years from an improved small. The initial cost to construct the FSD is assumed to stoves program would have much lower negative impacts on house- range from $500 to $1,000, depending on locality-spe- hold welfare. When the subsidies are withdrawn, households are likely to simply revert to using traditional or self-made stoves. They cific prices for construction materials.26 may notice a cooking cost increase from increased fuel-wood use but the amount would not be significant in absolute terms because The methane fraction in biogas varies from 30 percent of generally low wood prices. This may even prod them to continue to 70 percent and the lower heating value of the 2 m3 buying ICSs at unsubsidized prices. 25. In addition to cooking, it is possible to use biogas for lighting, gas output will range from 28 to 51 megajoules per day. via a mantle lamp, though this use will obviously reduce the gas The amount of fuel-wood replaced depends on the qual- available for cooking. About 1 cubic meter of biogas could provide ity of the biogas and the heating value of the fuel-wood, light from a mantle lamp equivalent to a 60 watt bulb for about seven hours. 26. Information obtained from the local UNDP office. 18 Timor-Leste: Key Issues in Rural Energy Policy Table 3.2 heaTing rePlacemenT ValUe of biogas fuel-wood (15% conversion) biogas (60% conversion) mj consumed mj usefully consumed mj supplied mj usefully consumed Daily 348 52 50 30 Annually 126,870 19,030 18,320 10,990 " Source: David Ludington, "Calculating the Heating Value of Biogas, Ithaca, NY, not dated; and mission estimates. Household consumption figures from Shum and others, op. cit. Note: The World Bank and most other experts generally use 15 percent as the conversion efficiency of a simple wood fire. assumed here as about 15.5 MJ/kg. If about 50 mega- In Case 2, where only one household uses the FSD and joules of biogas per day is generated by the FSD, then splits the output equally between cooking and lighting, the project is capable of replacing about 58 percent of the base case yields a positive ERR of about 8 percent but the net wood energy used by the three households (table still a negative net present value of about $123 per house- 3.2).27 With the FSD shared by three households, the hold. Almost 67 percent of wood energy consumption is families will still need to supplement their fuel needs for replaced. The somewhat stronger economic results for cooking with fuel-wood or other alternatives. Case 2 are mainly due to the higher replacement value of imported kerosene. However, this case assumes that the The economic cost of fuel-wood replaced by biogas is single household owns four to six head of cattle to sup- assumed to be the opportunity cost of fuel-wood collec- ply the waste requirements of the FSD and thus further tion by families in rural areas, where about four to six limits the coverage of the biogas program. hours per week are spent gathering an average of about 55 kilograms of wood. Based on a GDP per capita of Moreover, the use of biogas for lighting with mantle about $440, the opportunity cost of fuel-wood for a rural lamps is considered by experts to be an inefficient use household would be about $12­$19 per ton. for the fuel and should be resorted to only if electricity is unavailable. A mantle lamp equivalent to a 60-watt elec- tric light bulb requires 0.11­0.15 m3 of biogas per hour. Results of Analyses The details of the economic analyses of the two cases Sustainability Issues are shown in annex 3. Using as a base case the mid-range values of capital cost and gas quality, the economic rate of The preliminary analysis has shown that investments in return (ERR) of Case 1 is nearly zero and the net benefits FSD systems generally produce poor or negative eco- per household are equal to negative $468.28 Less than 50 nomic returns. Benefits to household health, especially percent of wood energy consumption would be replaced. among women and children, resulting from IAP reduction A best case scenario, with an ERR of 32 percent, is pro- were not included in the analysis. To realize the health duced by the most positive assumption for each of the key benefits, the reduction in wood use through partial or full parameters: gas quality, initial cost, and the opportunity substitution with biogas in cooking must be sustained cost of the wood that is replaced. In this highly unlikely over a reasonably long period. In this regard, FSD sys- case the highest possible gas quality and lowest capital tems have generally fared poorly. Individual FSD systems cost was assumed for the FSD, and the opportunity cost have a poor track record for long-term performance, many of wood was assumed at twice the current opportunity being abandoned because of technical problems after cost of wood, that is, about $25­$40/ton. only a brief time of operation.29 This is especially true for systems for which users provided no significant contribu- tion to the investment costs. Although typical technical 27 Based on 7 kilograms of fuel-wood consumed per household . .5 problems (clogging, air in gas pipe, insufficient gas pres- per day, at 15.5 MJ/kg. sure, too much carbon dioxide in gas, and the like) are 28. The base case for biogas also uses a highly optimistic assump- tion of a 15­20 year economic life for the FSD. In practice, such lifetimes have been achieved in very few countries and five years or 29. The track record is better in China, India, and other South Asian less is a more normal FSD lifetime. Needless to say, the returns in countries where very-large-scale programs have been implemented even the most optimistic cases will be negative with such a short for decades, and a culture of livestock raising and waste recycling lifetime. exists. Household Energy Policies 19 often manageable, the tendency is simply to abandon Conclusions the biogas system and shift to alternative cooking fuels, such as fuel-wood, if these alternatives are still easily or The specific figures cited are less important than the cheaply obtainable. The need for continuous operating overall results that clearly point to the low economic and maintenance (O&M) assistance from the govern- viability of a biogas program designed mainly to replace ment to individual recipients of biogas systems and the fuel-wood, today. The current overall fuel-wood situa- cost of such assistance cannot be underestimated. tion in Timor-Leste does not yet present a significant supply problem to households in rural areas and results Worldwide, biogas systems that have operated sustain- in a low opportunity cost of fuel-wood. This, of course, ably over the long term are generally larger-scale, mul- does not suggest that the supply situation will continue tipurpose installations in livestock farms. They are built unchanged; efforts to prepare for a future deficit situation not simply to replace other fuels for cooking but also and higher fuel-wood costs are justified. to replace kerosene for lighting, produce higher quality fertilizers, generate electricity for decentralized use or The current program to install household biogas digest- export to the grid, provide process heat, nourish a fish ers has a high capacity-building value and should be con- farm, and so on. The improved overall economic return is tinued for that purpose but on a limited pilot scale. The due to economies of scale, higher value of the outputs main objectives of the pilot program would be to train of the varied applications, and the likelihood of operating installers and technicians, to demonstrate the technology continuously for 15­20 years as a result of better O&M to the public, and most important, to systematically col- procedures. lect cost and operational data that could be used to make realistic policy decisions for the next steps. Meanwhile, Conversely, experience with communal kitchens, whether as indicated by the analysis, the only way to improve the fueled by biogas or other energy forms, has generally not economics is to find digester designs and configurations been positive for a variety of social reasons. Although that reduce capital costs significantly, require less opera- economies of scale may also be gained by establishing tional intervention, and generate benefits other than just larger-scale biogas systems for communal cooking, the fuel-wood replacement. communal systems are not likely to be sustainable over the long term, thus failing to deliver potentially large IAP Apart from the economics, it is also important for policy reduction benefits. making to consider beneficiary issues. Presently, only a box 3.3 commercial-scale biogas ProjecTs: PoTenTial conTribUTions To Power sUPPly The Ermera district project integrates a 98-head (of cattle) feedlot, fish farming, fertilizer production, and a 15-kilowatt generator. Using the standard energy values for the expected gas output from this project, the village of 150 can be supplied with 15 kilowatts for approximately seven hours per day (29 percent load factor and 30 percent conversion efficiency). With initial total investment costs of about $250,000, and allocating 30 percent of the total costs to the electricity generation component, a the project earns an ERR of about 10 percent at an assumed value for electricity of 35˘/ kilowatt hour. Thus, at a cost that is competitive with small PV or diesel gensets, this integrated system can supply reliable electricity for both daytime and evening hours. Such a project, while not capable of contributing to the grid, can nevertheless provide a decentralized electricity supply for villages near large livestock farms. Larger biogas systems, based on commercial scale feedlots of 500­1,500 head of cattle, can support power plants capable of distribution grid integration (70 percent load factor), at sizes ranging from 36 kilowatts for the 500-head feedlot to 110 kilowatts for a 1,500-head feedlot. a. A 15-kilowatt generator of the type used at Ermera costs approximately $25,000­$30,000 for machinery, installation, controls, and distribution wiring. 20 Timor-Leste: Key Issues in Rural Energy Policy handful of rural families own sufficient livestock to sup- biogas program to have a wider and more varied set of ply the required amount of wastes for a biogas digester; beneficiaries. Meanwhile, poor rural families without live- these families are most likely to be better off than those stock could still be provided some basic energy assis- who do not own any livestock. Subsidizing the construc- tance through other government programs that are less tion of biogas digesters for only these few families may costly to implement as measured by investment cost per raise criticism that the program benefits the relatively household, for example, improved cooking stoves, solar better-off (less-poor) segment of the rural population. lanterns, and others. In some countries, this social issue has proved to be a major barrier to finding continued political support for As larger livestock farms are established, government household-scale biogas programs. In the pilot or dem- policy should make development of domestic biogas onstration phase, there is no choice but to use relatively systems a priority for support because the economics more well-off rural families because they are the only are very likely to be more attractive at large digester ones who have enough livestock. The beneficiary issue sizes. Unlike a cooking-only application, large-scale sys- becomes relevant in the subsequent or expansion phase tems could be used to provide other important services, of the biogas program, at which time it should be fac- such as power generation, process heating (for example, tored into the subsidy mechanism finally adopted. meat processing), and fish farming, that benefit the over- all economy. The Ministry of Agriculture, Forestry and Fisheries (MAFF) has an ongoing livestock breeding and distribu- The Ermera project (see box 3.3) could provide valuable tion program for cattle, buffalo, pigs, goats, sheep, and experience in this area. The government should take poultry, and aims to establish large livestock farms. advantage of the Ermera project to systematically eval- When the livestock program has progressed significantly uate actual costs and benefits, so that the economics and more rural families are raising their own animals, of similar subsequent integrated installations could be it may become possible for the post-pilot phase of the improved. 4 Alternative Fuels: Biodiesel from Jatropha The government of Timor-Leste has initiated a program of petroleum middle distillates? Is a positive net energy cultivation of Jatropha curcas (Jatropha), a drought-toler- ratio (NER) sufficient cause to continue the program? ant vegetable oil crop, to produce biodiesel and replace Conversely, is a negative NER sufficient reason to imported diesel for electricity generation. Currently, a few terminate the program? hectares (ha) have been planted for a little more than one · Cost. Can biodiesel be supplied at or below the cost year by small farmers in various places using marginal of imported diesel? If not, is the program worth con- lands. The overall concept is for farmers to harvest the tinuing based on potential benefits gained through Jatropha oil seeds, bring them to the local cooperative local employment opportunities? What types of sub- for processing, and be paid a fixed price for each kilogram sidies might be used to supply Jatropha at a cost of Jatropha seed.30 The seeds will be brought to a central that is competitive with that of petroleum fuels? Are location for pressing into oil. The oil will then undergo a such subsidies justified by employment creation or transesterification reaction with methanol, using a cata- other benefits? lyst. The product--biodiesel--has very similar properties · Need for a pilot phase. What critical data and infor- to petroleum-based diesel. Ultimately, it is envisioned mation should be collected during the pilot phase that as many as 200,000 ha will be dedicated to Jatropha to be able to fully support a decision to scale-up or cultivation in the country. terminate? · Foreign exchange saving and fuel security. Given the To address the issues raised in both the current propos- substantial oil revenues (about $60 million to $75 als and in examination of Jatropha cultivation in India and million per month) the country is earning, as well as elsewhere, the following issues and concerns must be current and potential sources of diesel fuel supply, examined: to what extent does the strategy of locally produc- ing a diesel substitute actually benefit the country in · Land use. Can Jatropha be cultivated on marginal foreign exchange (FX) savings and fuel security? lands? If so, is the question of the opportunity cost of land resolved or mitigated? At the present time Jatropha is not cultivated commer- · Energy balance. What is the energy balance for cially in Timor-Leste. However, Norges Vel has proposed the biodiesel fuel cycle, including blending into a pilot program that could scale to full commercial size for the cultivation, processing, and end use of Jatropha- 30. A price of $0.25/kg has been suggested as sufficient to provide based biodiesel. The question of Jatropha cultivation has farmers with "adequate" incentives to undertake cultivation of Jat- also been considered by the local office of the Food and ropha. However, most bioenergy schemes involve a percentage split Agriculture Organization. of the final price between growers and processors. That issue is addressed further on in this section. 21 22 Timor-Leste: Key Issues in Rural Energy Policy figUre 4.1 modeling schemaTic for commercial jaTroPha cUlTiVaTion Polyethylene bag Tillage and Irrigation by production ploughing hectare Jatropha seedling for planting Jatropha plantation Urea production, Jatropha seeds Potassium nitrate transport, and harvested from production, transport, application plantation and application Single super phosphate Jatropha seeds Irrigation water production, transport, to oil-- and application extraction facility Module key Ecoinvent 2.0 module Custom module Descriptive module for clarity " Source: Michael Whitaker and Garvin Heath, "Life Cycle Assessment of the Use of Jatropha Biodiesel in Indian Locomotives, NREL/TP-6A2-44428, March 2009, p. 24. Estimates of the costs of growing Jatropha, expected the use of fertilizer and irrigation. Because the crop has yields under different cultivation regimes, conversion to been cultivated commercially for only a few years, the biodiesel, and land use needs for various program alter- type of hybridization and selective breeding that produces natives in this report are derived primarily from studies of a uniform crop and seedpod has not yet occurred. Nor is Jatropha cultivation in India and elsewhere.31 The analyti- there a cultivar that has entirely predictable responses to cal schema for modeling Jatropha cultivation in the India fertilizer and irrigation use.33 study is shown in figure 4.1. The literature on Jatropha indicates the following: Land Use · At the present time no studies exist that use stan- dardized density, soil chemistry, fertilizer, and irriga- As a general rule, land with higher quality soil, more rain- tion plots with consequent yield figures such as do fall, and greater use of necessary pesticides and fertilizers exist for many other agricultural crops.34 Much of the will yield more crops of any kind than the same natural evidence on growing Jatropha remains anecdotal. conditions with fewer human and chemical inputs. The · Though Jatropha is drought tolerant (that is, it can same general rule applies to Jatropha. A number of crop stay alive for two to three extremely dry years), it studies consulted for this report32 indicate that Jatropha will not produce seedpods during such years, and responds similarly to many other crops with respect to subsequent production is reduced as well. · Jatropha responds to fertilizer and should be weeded regularly.35 31. Michael Whitaker and Garvin Heath, "Life Cycle Assessment of " the Use of Jatropha Biodiesel in Indian Locomotives, NREL/TP-6A2- 44428, 2009. 32. Pere Ariza-Montobbio, "Impact of Jatropha curcas Plantations for " Biodiesel on Livelihood and Food Sovereignty in South India, Insti- tute of Environmental Science and Technology, University Autono- 33. Ariza-Montobbio, op. cit., p. 6. mous of Barcelona (UAB), 2008. Roy Beckford, "Fundamentals of 34. Whitaker and Heath, op. cit., p. 4. Producing Jatropha curcas, University of Florida, November 2008. " 35. Whitaker and Heath, op. cit., pp. 4, 16, 20, 25. Alternative Fuels: Biodiesel from Jatropha 23 · Jatropha requires natural annual rainfall in excess of land (arable plus permanent crops) at an assumed yield 900 millimeters and total annual rainfall plus irriga- of four tons of seed per ha, a "normal" figure for a field tion in excess of 2,300 millimeters to produce seed receiving fertilizer and other cultivation in India. Expan- and oil continuously throughout a 15­20 year life. sion of the program to B10 or B20 will require 13,333 ha · Jatropha will not tolerate more than two days of and 26,667 ha, or 7 percent and 13.9 percent, respec- flood conditions,36 a significant issue as it relates to tively, of cultivated land. where to locate the Jatropha crops. · The plant is subject to a number of well-known plant Experience elsewhere indicates that use of otherwise pests, including aphids, weevils, and various boring unproductive land will not produce any salable amount insects. of seed for oil extraction in the absence of other inputs, · There are not yet any longitudinal studies indicat- including fertilizer, pesticides, cultivation, and irrigation. ing the longer-term impacts of growing Jatropha on Under such conditions Jatropha will compete with other either prime or marginal agricultural lands. permanent crops, including nuts, fruits, and fuel-wood, for marginal lands that surround farm fields. In other words, Jatropha has been promoted on the grounds that it can wherever Jatropha can be grown to yield a commercial be a "no-cost" supplement to farm incomes, with an volume of seeds it is likely that there are other tree and added FX benefit for the country with regard to fuel sup- shrub crops competing for the same land. Low- or no- plies. In particular, it is thought by some that Jatropha can input plantings of Jatropha on truly marginal lands are not be grown in areas adjacent to existing croplands, inter- likely to produce commercially salable seed quantities. If cropped with other crops, or grown on soils or slopes production of seeds is low in density, amount, and qual- that are not suitable for conventional cropping. ity, then the transport costs for moving small amounts to a distant processing center will be high, reducing the net Timor-Leste has about 1,231 square kilometers of arable income to the farmer for the sale of the Jatropha seeds. land, 8.2 percent of a 15,007 square-kilometer total land area. Cultivated land in the country, including permanent crops, totals 1,917 square kilometers, or 12.77 percent Energy Balance of the country's total land area. The area for permanent crops,37 4.5 percent of the country's land area, is low Two recent studies addressed the energy balance of relative to other countries in the region--more than 15 Jatropha cultivation, one for India and another study that percent of the total land area of both Malaysia and the used conditions in Thailand.39 Energy balance was gener- Philippines is devoted to permanent crops. The figure for ally found to be positive under conditions of high plant Indonesia is 7 percent.38 density, adequate rainfall and irrigation, good seed yields, and little efficiency decrement for the biodiesel mixtures In looking at the availability of "unused" or unproductive in the power plant.40 agricultural land for planting Jatropha, it is useful to con- sider the land-use requirements of a program that might The Thailand study found that the net energy ratio (NER) make a meaningful contribution to the country's energy for biodiesel production was generally greater than 1 supplies. Current consumption of middle distillate liq- (that is, energy value of the oil in a form that can be read- uid fuels in Timor-Leste is 2,300 barrels per day (bbl/d), ily mixed into diesel as a proportion of the total nonsolar equivalent to 133 million liters per year (123 kT/y). Stud- energy used to cultivate, transport, and process the Jat- ies of Jatropha in other countries used standard biodiesel ropha seeds). NER > 1 indicates a net energy contribu- blends of 5 percent, 10 percent, and 20 percent (B5, B10, tion from the proposed cultivation; NER < 1 indicates that and B20, respectively) for their analyses. Jatropha cultivation, transport, and processing consume more energy than is contained in the oil. Under "normal" A B5 program in Timor-Leste would use about 6,667 ha conditions, that is, assuming reasonable soil quality, rain- of land, roughly 3.5 percent of the country's cultivated fall and irrigation, fertilizer and pesticide inputs, and time spent cultivating the crop, the NER could be as high as 36. Beckford, op. cit., found that root rot was significant following 2.7 and averaged 1.42. This means that the energy value two days sitting in flooded conditions. of the output oil was on average 40 percent higher than . 37 Permanent crops are considered multiyear agricultural, viticul- tural, or horticultural activities and include fruit trees, nuts, vineyards, and perennial flowers. This figure excludes trees used for timber or 39. Whitaker and Heath, op. cit.; Kritana Prueksakorn and Shabbir H. paper products. Gheewala, "Full Chain Energy Analysis of Biodiesel from Jatropha 38. Arable land is considerably larger as a percentage of total land " curcas L. in Thailand, Environmental Science and Technology, 2008, area in both Indonesia and the Philippines (CIA, The World Factbook, 42 (9): 3388­93. June 2010). 40. Whitaker and Heath, op. cit., p. 47. 24 Timor-Leste: Key Issues in Rural Energy Policy the energy value of the inputs and as much as 170 per- For the conditions that were examined in Thailand and cent greater in some cases (the system configuration India, Jatropha can be a net energy contributor as long as used for this analysis is shown in figure 4.2). However, it is cultivated in a serious manner. This means, among under conditions of low yields, low shrub density, and other things, good soil, sufficient water, continual weed- poor conversion to a diesel-compatible oil, the NER could ing, and use of fertilizer and pesticides at appropriate drop as low as 0.53, meaning that the energy value of times. As a purely marginal activity on low-quality, dry the output oil was only about half as great as the energy lands Jatropha is not likely to show a feasible NER.41 value of the nonsolar inputs. Growing conditions sufficient to produce Jatropha oil on It may be reasonable to conclude from these studies that a feasible NER basis (NER > 1) should be examined from the NER can be positive as long as there is some effort to the standpoints of economics, rural incomes, and energy grow the crop on agricultural land and put some money supply considerations. Conditions under which Jatropha and time into the cultivation. Grown as a true marginal shows a net energy deficit (NER < 1) are unlikely to pro- crop on marginal lands, the NER is likely to be negative vide much in the way of income or energy supply ben- and strongly so. efits either. A source of renewable energy should be, on balance, a net energy contributor unless there are secondary con- siderations, including soil conditioning, wind breaks, and the like that provide a compelling economic case for the 41. The notation specifying the positive net energy condition can investment. A project with NER < 1 is unlikely to be eli- be found as Equation 15 in the UNFCC document at this location: gible for Clean Development Mechanism or other carbon http://cdm.unfccc.int/UserManagement/FileStorage/IXDFJ1ZG offset payments.41 IPXAWCCC2FKG605E4CFW0C. figUre 4.2 neT energy analysis of jaTroPha cUlTiVaTion in Thailand System boundary Jatropha T Wood Fuel stock farming Fruit T Energy/raw materials/auxiliary materials Emissions to air, water, and soil Oil extraction Peel and T Fuel stock, and refining seedcake fertilizer Oil Biodiesel T Use in diesel production Biodiesel engine T Glycerin Fuel stock T = Transportation " Source: Kritana Prueksakorn and Shabbir H. Gheewala, "Full Chain Energy Analysis of Biodiesel from Jatropha curcas L. in Thailand, Environmental Science and Technology, 2008, 42 (9): 3388­93. Alternative Fuels: Biodiesel from Jatropha 25 Cost of Supplying Jatropha Oil inputs and on inappropriate land the cultivation of Jatro- pha is not likely to be a net energy producer and will be The only information available on Jatropha cultivation that inferior to other uses of marginal lands, possibly includ- is specific to conditions in Timor-Leste is the Norges Vel ing fuel-wood. conceptual paper. However, the paper contains little that can be used to assess the economic feasibility of Jatro- For example, at a (farmer's break-even) seed price of pha cultivation and processing, much less its opportunity 8.3˘/kilogram, the raw material cost of 1 liter of oil will cost with respect to other agricultural activities. To make be 33.2˘. With minimal processing cost of 26˘/liter, the a preliminary economic analysis, this report uses yield minimal break-even price for refined diesel-compatible oil and cost data from other studies, primarily the studies is 57˘/liter, equivalent to diesel oil at $90/bbl. This result cited earlier.42 Using a series of prices paid to farmers is shown graphically in figure 4.3. If farmers are guaran- for Jatropha seed relative to different prices for imported teed 15˘/kilogram for seed, the minimal break-even price fuels,43 a pattern emerges with regard to economic cost rises to about 65˘/kilogram, consistent with diesel oil at and feasibility of Jatropha cultivation at different crude $110 to $120/bbl on a consistent basis. oil prices. Under conditions of active cultivation (use of fertilizer, The economic analysis used the following parameters: pesticide, weeding, irrigation, and so forth) profitable returns to the farmer will require crude oil prices in excess · yield of Jatropha seeds per hectare (tons), of $90/bbl. Even with an increasing proportion of the oil · yield of oil per kilogram of seed, price returned to growers as crude prices rise, farm profits · cost per hectare of Jatropha cultivation, from Jatropha planting are very modest until crude prices · crude oil price, reach $120/bbl.44 At no time does the price (actually the · processing cost, and netback value) for Jatropha approach the 25˘/kilogram · split between growers and processors of oil. level discussed by the Food and Agriculture Organiza- tion representative in Dili. Such a price, were it realized, Variations of these parameters will provide a range of would generate grower profits in excess of $500/ha, results. In the current circumstances, and in the absence but would also require a fuel price equivalent to almost of country-specific information, the values in table 4.1 $200/bbl as a break-even price for the processor. were assumed for a base case. 44. Grower-processor splits of the type displayed in figure 4.3 are The economic analysis assumes that purposeful cultiva- typical of a fully market-oriented grower-processor relationship, one tion activity of Jatropha takes place. As the net energy in which the processor would be responsible for blending with die- analysis and land use discussions have indicated, without sel or selling to a refined products jobber. Note also that even with 15˘/kilogram for the crop, net farm income at $100/bbl for oil is just $250/ha, hardly a dispositive case for Jatropha cultivation on normal 42. Whitaker and Heath, op. cit.; Prueksakorn and Gheewala, op. cit. croplands when other food or feed alternatives are likely to provide 43. The price split with processors was discussed previously. higher net income. Table 4.1 ParameTers for economic analysis of jaTroPha cUlTiVaTion in Timor-lesTe Yield of Jatropha seeds per hectare (tons) 4 Yield of oil (liters) per kilogram of seed 0.25 Cost per hectare of Jatropha cultivation $333 Crude oil price market Split between growers and processors of oil 50:50, rising to 65:35 at higher crude prices Processing cost (per liter raw oil, diesel-powered expeller including $0.17 equipment, profit, fuel, labor) Additional processing (transesterification per liter) $0.09 . Sources: Whitaker and Heath, op. cit., pages 26­27 Authors' assumption for split between growers and processors. Processing cost data from " Reinhard Henning, "Jatropha as a Tool to Combat Climate Change and for Poverty Reduction, UNCTAD (unofficial), 2006, p. 13. Cost of expelling process uses oil price of $75/bbl. Note: Processing includes transport to mill and subsequent transport of transesterified oil to blending plant. 26 Timor-Leste: Key Issues in Rural Energy Policy figUre 4.3 jaTroPha economics wiTh marKeT-deTermined sPliT for growers and Processors 0.18 400 Price for seed per kg Net income per ha 0.16 300 0.14 200 Net income per hectare 0.12 US cents per kg seed 0.10 100 0.08 0 0.06 ­100 0.04 ­200 0.02 0 ­300 40 50 60 70 80 90 100 110 120 150 US$/barrel crude oil Source: Authors. Subsidies for Jatropha Oil Growers If farmers are to cultivate Jatropha profitably and if pro- cessors are to breakeven on their costs, a minimum price With grower profits unlikely at current and expected of $110 to $120/bbl will be needed to support Jatropha future crude prices, one way to induce growers to switch cultivation and use in Timor-Leste. At least $90/bbl is to Jatropha is to guarantee a seed price sufficient to gen- required just for break even with no profits for proces- erate a reasonable profit, say, $250/ha. The cost of pro- sors and net farm income of less than $200/ha. Figure viding such a price floor at varying levels of output (B5, 4.4 shows what the government might be expected to B10, B20) and with the same variation in crude prices used spend subsidizing Jatropha cultivation and processing in in figure 4.3, was examined. a cooperative ownership situation for the oil processing. The figure indicates that guaranteeing a price to farmers The big change in assumptions in this case vis-ŕ-vis the of at least 15˘/kilogram will require significant subsidies case that resulted in figure 4.3 is that rather than share until the price of oil is a bit above $90/barrel. the profits from Jatropha sales through a market-deter- mined grower-processor split, it is assumed that farmers To have a meaningful impact on the energy sector, the will form a cooperative to own the processing facilities, government of Timor-Leste will need to look at Jatropha thereby effectively capping the costs associated with programs that are capable of providing a uniform fuel expelling oil and refining it up to blendable biodiesel quali- type throughout the country, thus one of the programs-- ty.45 Even under such conditions, the costs of processing B5 through B20--for biodiesel. Figure 4.4 shows the costs will vary directly with the price of oil. Both diesel (about 1 per year for providing growers with a profit of $250/ha by liter required for each 15 liters of oil expelled) and metha- guaranteeing a purchase price for the seed. A B5 program nol (about 0.4 liter required per liter produced) have vari- can be instituted at an annual subsidy of about $2 million able prices. for the current range (US$40­US$100/bbl) of oil product prices. The B10 and B20 programs are more expensive and scale up arithmetically. As figure 4.4 indicates the subsi- 45. Investment costs for such a plant are likely to be well beyond dies are very small at crude prices of about $90/bbl and what small farmers in the country can afford, with initial investment needs likely in excess of $20,000 for a small mill capable of servicing disappear entirely before crude prices reach $100/bbl, as the output of 50 hectares. long as the cost of processing is capped. Alternative Fuels: Biodiesel from Jatropha 27 Figure 4.4 government SubSidieS For growerS Figure 4.5 government SubSidieS For under Cooperative ownerShip oF Jatropha growerS under proCeSSing (that iS, CoSt Cap on typiCal market arrangementS proCeSSing) 14 B5 35 B5 B10 B10 12 B20 Annual subsidy (US$/millions) 30 B20 10 Annual subsidy (US$/millions) 25 8 20 6 15 4 10 2 5 0 0 40 50 60 70 80 90 100 110 120 130 40 50 60 70 80 90 100 Crude oil price ($/barrel) Crude oil price ($/barrel) Source: Authors. Source: Authors. Note: In this case growers and processors split 50:50 (grower: processor) for oil prices below $90/bbl and 65:35 for higher prices. Under market arrangements more typical of bioenergy The FX savings consist largely of the middle distillate fuel crops, the targeted break-even price for growers would that is not purchased as a result of the Jatropha program likely not be reached until crude prices hit $125/bbl or because there are FX costs for diesel used to run the more (see figure 4.5). Such a break-even price for growers expeller, methanol for transesterification, and machinery is based on price splits suggested for Jatropha cultivation for processing. Costs of transport of the fuel and seeds and processing by parties in Timor-Leste and in India. are not included in this calculation. Figure 4.6 shows the No present value analysis has been performed for this economic assessment because the data are insuffi- Figure 4.6 FX impaCtS oF Jatropha-baSed biodieSel cient to determine initial investment needs, land costs, investments in trucks and other equipment, and infra- 20 B5 structure improvements that must be made for the proj- B10 ect. Such information could be produced in the course 15 Dollars (million) B20 of a pilot-scale project as suggested at the end of this subsection. 10 5 Impacts on Foreign Exchange 0 Most of the inputs for growing Jatropha, including fertil- 40 50 60 70 80 90 100 110 120 130 izer, pesticides, and tools, will be imported, as will be the Crude oil price ($/barrel) equipment and chemicals for processing. The FX costs of each of the biodiesel programs will be dependent on Source: Authors. the amount of fuel produced and the cost of imported Note: These calculations do not include the FX that is forgone when inputs. These input costs are expected to vary in a gener- export crops are displaced by Jatropha cultivation or when food crops that currently compete with imported foods are displaced by Jatro- ally proportionate manner with crude oil prices because pha. Such impacts, requiring more extensive local analysis and data, most of the inputs are derived either directly or indirectly are likely to further disadvantage the arguments favoring Jatropha. from crude oil. 28 Timor-Leste: Key Issues in Rural Energy Policy net FX costs or benefits from the B5­B20 biodiesel pro- · The putative benefits of Jatropha cultivation on FX, grams as a function of oil prices. net energy, and farm incomes are highly dependent on crude oil prices. At the present time a major At low oil prices, the FX costs of importing inputs for investment in Jatropha cultivation and process- cultivation and processing are low relative to purchasing ing does not appear to be prudent unless oil prices diesel abroad, at about $1 million to $1.6 million annually achieve a plateau at or above $100­$120/bbl. for a B5 program when oil prices are in the $50­$60/bbl · At oil prices near current levels ($75/bbl ± $20/ range. However, at crude prices above $80/bbl, there are bbl) there could be measurable FX savings on noticeable FX savings, ranging from about $2 million per energy products only provided the cultivation year to just under $3 million per year when crude prices of Jatropha is approached seriously and not range from $80­$110/bbl for a B5 program. Substantial treated merely as a sideline crop for farmers. FX savings accrue once the oil price rises above $120/bbl · Farm incomes will require subsidies exceeding and for the larger-scale programs. $2 million/year for the B5 program at current oil prices and above $1 million/year at $80/bbl for The costs of imported inputs will remain a significant crude. Larger-scale programs would be propor- drain on larger programs, for example, inputs would cost tionately more costly. roughly $14 million annually for a B20 program at current · Agronomic studies have consistently indicated that (2010) crude oil prices. This expenditure of FX to produce cultivation of Jatropha on marginal lands without 27 million liters per year of refined Jatropha oil suitable much attention or input from the grower can result in for blending would save the importation of $12.5 million a commercially viable crop, or in one that is capable in equivalent diesel fuel, for a net FX cost to the govern- of supplying net energy, rural income enhancement, ment of $1.5 million annually. and FX savings to the country. To achieve the maxi- mum benefits of Jatropha cultivation, it will be nec- The results in this section do not change in an eco- essary to devote reasonably good land, cultivation nomic sense if the diesel is domestic or foreign in origin. time, fertilizers, pesticides, and irrigation water. However, from the standpoint of financial analysis a well- managed Jatropha cultivation program has the potential It is not possible at this time to calculate a rate of return to save FX and provide measurable net benefits to farm- for the local Jatropha projects. Fundamental data regard- ers only if the price of oil consistently exceeds $100 to ing investment and cultivation costs do not yet exist for $120/bbl. Timor-Leste. However, it is clear that achieving a positive rate of return would require opportunity costs for diesel that are well above current or projected crude oil prices. Impacts on Farm Incomes and Employment Indeed, at prices sufficient to make Jatropha a positive At this time it is impossible to gauge the impact of a contributor to the country's economy, the production of Jatropha program on farm incomes in Timor-Leste using synthetic crude oil from the country's offshore gas would domestically generated data. As has been demonstrated probably be a better investment (box 4.1). above, it is possible for Jatropha cultivation to be either a net energy drain or a contributor. However, using the NER > 1 standard for all feasible programs, a positive Recommended Activities increment to farm incomes will occur if oil prices rise well above current levels (and remain there) or if the gov- Certain basic information is needed for an intelligent ernment guarantees payments to farmers and growers assessment of the potential for Jatropha cultivation in that will compare in cost with importation of the same Timor-Leste to be made. It is suggested that ongoing volume of middle distillate fuel. pilot projects on Jatropha cultivation and processing be closely monitored to obtain actual field information that includes the following: Conclusions · yields of seed per hectare in relation to tree density, Available country-specific data are insufficient to make soil type, rainfall and irrigation, and fertilizer; firm conclusions about the economic viability of a future · cultivation labor requirements (weeding, pruning, investment program for Jatropha in Timor-Leste. How- bed maintenance, and so forth) in relation to yields; ever, based on agronomic work and other evaluations · susceptibility of Jatropha to pests common to of Jatropha projects in India, Florida, and Thailand, some Timor-Leste and appropriate and effective control tentative findings can be advanced for Timor-Leste: methods; Alternative Fuels: Biodiesel from Jatropha 29 box 4.1 jaTroPha-based fUels in The conTexT of oVerall naTional energy and agricUlTUral Policies As a significant producer of hydrocarbon fuels, and with new hydrocarbon development in the near term highly prospective, the devotion of substantial national resources to hydrocarbon substitutes needs careful examination. Subsidies to farmers to raise energy crops such as Jatropha compare in scale to the cost of simply purchasing middle distillate fuels. Ultimately, the arrival in the country of natural gas or gas condensates from offshore production is likely to make that source the fuel of choice for electricity generation. Such supplies could start within five to eight years, before Jatropha plantations pay back these investments just from sales to domestic energy markets. Once Timor-Leste begins receiving gas or condensate from offshore production, continued profitable operation of Jatropha plantations and processing plants would require export of the seeds or processed oil to some other country. Then the commercial viability of Jatropha vis-ŕ-vis other cropping alternatives would become a simple question of dollars and cents. Will Jatropha cultivation and export prove more profitable for farmers than use of the same land for other export-oriented crops? Such a question cannot be answered with much precision today, but it is crucial to land use and energy policies. The key to successful continuation of Jatropha cultivation would be commercial-scale planting with all output exported, possibly supported by carbon credits or similar mechanisms. Studies of Jatropha do not support the idea that the crop can contribute much, if at all, to farm incomes or energy supplies without serious effort at the farm or plantation level on productive, not marginal, land. Once domestically supplied hydrocarbons become widely available, it would not be prudent policy to continue a domestic biodiesel program that carried cash and other opportunity costs for the country's economy. · yields of oil per kilogram of seed in relation to cultiva- · comparisons of Jatropha impacts on rural incomes tion issues (first three bullets); with the impacts of other possible export crops · costs of establishment, cultivation, and stand main- using similar agricultural resources and inputs--land, tenance of Jatropha shrubs; water, fertilizer, pesticides, and labor. · costs of establishment and operation of processing plants, and investigation of scale economies, if any, It should take about three to five years for the current at each stage of operations; pilot programs to generate data needed to make deci- · optimal blending methods for B5­B20 programs, sions on the next phase. It is recommended that no fur- assessment of power plant output decrements, if ther expansion of the current programs be made until any, from biodiesel use; and such data are available. 5 Conclusions and Recommendations The primary goal of a sound rural energy policy is to pro- part of the system acquisition cost is to instill a mote measures that will enhance the quality of life of sense of ownership. A wealth of evidence from past people in rural areas by improving their access to modern projects has shown that PV dissemination mecha- energy services. The strategies to pursue are those that nisms in which systems are just given away or fully are sustainable, environmentally benign, and economic subsidized fail within a short period. from both the country's point of view and the welfare of · The government may decide under some circum- individual rural households. A key objective in implemen- stances to pay for the full upfront cost of a PV tation is to ensure that the government's rural energy installation but, to ensure sustainability, leave respon- programs provide equitable distribution of benefits, to sibility for the full cost of operation and maintenance the extent possible. (replacement of battery, lamps, and so forth) to the user, to the extent possible. · In the context of Timor-Leste, a fee-for-service Improving Access to Electricity mechanism executed by the private sector but with a strong role for the government would appear to be · Government policy should be to provide, at a mini- more suitable for a PV dissemination program for off- mum, basic electricity services to off-grid commu- grid areas. However, as with any model that serves nities, through programs implemented in the same very-low-income households and requires periodic time frame as those directed at concentrated com- payments by users to maintain operation, long-term munities living near the network. The implementa- sustainability is more likely when combined with tion policy should be based on principles of least-cost other agricultural or rural development programs power planning, and with the objectives of technical that raise the income of the target communities. and financial sustainability. · For new decentralized power systems, it is good policy to put priority on harnessing locally available Household Energy energy resources, especially environmentally benign renewables, wherever they are available. · Despite the still abundant biomass supply in most · Even when local energy resources are available parts of the country, the responsible policy for the within 5 kilometers or so of load centers, however, government to pursue must be the maximum cur- the least-cost option may be solar photovoltaic (PV) tailment of fuel-wood use by households. and not the development of the resource. This is · The rationale for this policy includes (a) the certainty often the case when the potential users are highly that the current positive supply-demand balance will dispersed and thus expensive to connect together, change in the future, with population growth and and the most significant use of electricity is only for increased pressure on forest land conversion to agri- lighting and small domestic purposes. culture; and (b) the adverse impacts to public health · Aside from the objective of financial sustainability, a from indoor air pollution (IAP) associated with use of key purpose of requiring PV recipients to contribute fuel-wood in traditional cook-stoves. 31 32 Timor-Leste: Key Issues in Rural Energy Policy · The strategy is to pursue programs that combine making it hard to motivate households to econo- demand-side management or interfuel substitution mize. In addition, Timorese households do not yet programs (or both) with programs for improved man- have a culture of purchasing stoves. agement of fuel-wood supply. · The only way for improved fuel-wood stoves to be disseminated widely in Timor-Leste is to substan- tially subsidize their cost to users. Fortunately, com- Interfuel Substitution pared with other programs for fuel-wood reduction, · It is not a recommended policy to directly subsidize the required subsidies per household are small and the price of liquified petroleum gas (LPG) or kerosene could be terminated at any time by the government fuel for three reasons: (a) diversion of the fuel for without drastic welfare impacts on users. other uses, such as transport, can occur (more likely · The potential benefits, even for just reducing IAP for kerosene than LPG); (b) the immediate benefi- impacts, far outweigh the costs. As part of an ICS ciaries of the subsidy would be households already program, an IAP monitoring program should be car- using modern fuels or having higher incomes; and ried out. (c) most important, the future subsidy burden to the government is likely to become unmanageable. Domestic Biogas · Inevitably demand for modern fuels, particularly LPG, will increase dramatically in the coming years, · Economic analysis of domestic biogas systems cur- as economic conditions improve and a middle class rently being installed clearly point to the presently emerges and expands in Timor-Leste. The burden low economic viability of a program designed mainly to the government of a fuel subsidy could become to replace fuel-wood resulting from the relatively huge and unmanageable. As experience worldwide high investment costs per household and the low has shown, it is extremely difficult to withdraw or opportunity cost of fuel-wood replaced. even reduce the subsidies at that point. · The current program to install household biogas · It is often more effective to subsidize access, not digesters has a high capacity-building value and consumption. This means subsidizing equipment should be continued for that purpose but on a lim- needed to make the fuel switch (for example, new ited pilot scale. stoves purchase, LPG cylinder deposits) and facilitat- · Apart from the economics, it is also important for ing fuel distribution but keeping fuel prices at market policy making to consider beneficiary issues related levels. The financial exposure of the government for to the fact that current recipients of biogas systems an equipment subsidy program can be determined are families that are undoubtedly better off than oth- annually in advance. The government may terminate ers in the community. It is clear that there is little the equipment subsidy program at almost any time alternative in the pilot or demonstration phase, but with minimal public inconvenience. in the subsequent or expansion phase of the biogas · The simplest way to provide subsidies must be program beneficiary considerations should be fac- adopted. Experience elsewhere has shown that a tored into the subsidy mechanism. voucher or means-testing system for applying subsi- · Compared with alternative rural energy programs, dies often carries high administrative costs and tends the coverage of the biogas program and its total to be more complex to implement than a direct price potential fuel-wood replacement benefits are subsidy mechanism. severely limited by the small number of rural families that own enough livestock to provide the daily waste requirements of a digester. Improved Cooking Stoves · When the Ministry of Agriculture, Farming and Fish- · The promotion of improved cooking stoves (ICS) eries' livestock breeding and distribution program should not be viewed as encouraging continued has progressed significantly and more rural families use of fuel-wood but as a necessary intermediate are raising their own animals, it may become pos- approach--the only practical way for the govern- sible for the biogas program to have a wider and ment to reduce fuel-wood consumption significantly economically more varied set of beneficiaries. while the economy transitions to a future in which · As larger livestock farms are established, govern- cleaner modern fuels are widely available and most ment policy should make the biogas program a pri- households can afford them. ority for support given that the economics are very · The promotion of improved stoves in Timor-Leste is likely to be more attractive at large digester sizes severely hindered by the continued relative abun- and benefits other than fuel-wood replacement can dance and low cost of fuel-wood in the country, be achieved. Conclusions and Recommendations 33 Biodiesel from Jatropha production costs of $335/ha, the government will need to provide subsidies in excess of $2 million · The program being implemented in Timor-Leste to annually for a B5 program at current oil prices and produce a biomass-based substitute for diesel oil above $1 million/year at $110 per barrel. has as its objectives fuel security, increased reliance · FX savings as an objective must take into account on domestic resources, and foreign exchange (FX) the continued need for imported and oil-based savings. These objectives make sense only if biofu- inputs, such as fertilizers and pesticides and metha- els can be produced locally with few imported inputs nol for the Jatropha oil processing. Based on typical and at reasonable costs in comparison with the price yields, estimates indicate that at current oil prices of imported diesel fuel. (about $75 ± $20/bbl), FX savings are negative (that · The land-area requirements of the Jatropha program is, Jatropha is a net FX user), especially in light of are a significant consideration for a small country the full costs of cultivation, transport, processing, like Timor-Leste. At an assumed yield of 4 tons seed and blending. The potential foreign currency savings per hectare, a B5 program (blend of 5 percent biod- become significant only at crude oil prices above iesel and regular diesel oil) would require about 3.5 $100/bbl, and represent an effective use of FX only percent of the country's cultivated land; a B10 pro- if Timor-Leste lacks the ability to supply itself with gram, 7 percent; and a B20 program, 14 percent. Up domestically produced hydrocarbons. Even synthetic to 200,000 hectares of Jatropha plantations have crude from natural gas is likely to be less expensive been proposed. This is roughly equal to the amount than the estimated costs of Jatropha oil processed of total cultivated land in Timor-Leste today. for biodiesel blending. · The current Timor-Leste program for Jatropha is · While Jatropha cultivation undoubtedly increases predicated on the use of marginal lands. Studies rural employment, it is not clear that the cost is elsewhere have consistently shown that good yields, advantageous compared with other means of gener- low unit production costs, and a positive net energy ating rural employment opportunities in Timor-Leste, ratio (NER) for the product are achieved only when such as programs to promote community-based for- Jatropha cultivation is approached in a commercial est management and protection and agro-forestry manner and with appropriate inputs, such as fertil- initiatives. izer, irrigation, and pesticides. To quote the findings · Finally, the eventual local availability of petroleum of a recent international conference on Jatropha, products from offshore oil and gas production needs "Marginal lands will produce marginal yields. " to be considered, which could occur in fewer than · Based on the overall costs of growing, transport- 10 years, just when significant biodiesel production ing, and processing Jatropha, a crude oil price at or facilities would be gearing up. At that time, biodiesel above $90/bbl is necessary to cover the costs of the substitution would make even less economic sense various parties, and requires as well that the price than it does under current prices for imported fuels. component attributable to processors be capped. The alternative of exporting biodiesel is possible In a more typical market setting, still higher crude only if Timor-Leste enjoys a clear comparative advan- prices, above $120/bbl, would be needed to pro- tage over producers in other countries, such as sig- vide adequate profitability for both the farmers and nificantly lower production and processing costs the processors. For example, if Jatropha farmers and advantageous logistics with respect to potential are guaranteed a $250/ha profit above estimated customers. ANNEx 1 Designing a Sustainable Solar Home Systems Program for Timor-Leste Program Phasing Payment Scheme The task of providing photovoltaic (PV) power to identi- As a matter of policy, beneficiaries should be required to fied communities and households in a more systematic contribute to the capital cost of their specific PV installa- and sustainable manner should be approached in phases, tions. The purpose is not merely to save the government with the first phase possibly covering about 15,000 homes some money but, as earlier stressed, to instill a sense of (about a fourth of total potential PV customers) in five ownership. For communities receiving PV equipment for years. The cost of this phase would be about $1.5 million public facilities, the contribution may be in kind, such as to $2.0 million. The program would include, as is being organizing a local means for routine maintenance of the done now, installation of PV systems for public facilities equipment or to maximize the community's use of the (schools, clinics, community centers, and the like). The equipment. Individual households that receive PV must idea is to select qualified private sector service providers be required to pay a portion of the upfront cost and agree through competitive bidding who would install the sys- to pay a monthly maintenance fee to the service provider. tems in the designated communities and maintain them The capital cost contribution could be in the range of for a contract period of five years. Although it is clear that 5­10 percent of actual costs. The monthly fee should be at this time providing PV systems to households must be sufficient to cover the O&M costs (replacements and the heavily subsidized, competitive bidding for large numbers like) for each user in the market package for the duration of installations grouped into several "market packages" of the service contract, as estimated by the company in would result in unit acquisition costs much lower than its bid.47 Based on international experience, this is gener- they are today in piecemeal, suco-by-suco transactions ally in the vicinity of $5­$7 per month. with providers.46 Based on the experience acquired in the first phase, the business model, subsidy levels, and other implementation details would be modified as needed Preparing Market Packages in the second phase, which could aim for an additional 20,000 off-grid households in years 6­10. The government's first step is to group planned installa- tions for public facilities and individual homes in unserved communities into several market packages of sufficient size to be commercially attractive to the private sector. Public facilities usually require much larger PV systems 46. The idea is to implement a highly simplified version of the "sus- and their inclusion increases the financial attractiveness tainable solar market packages" approach practiced in the Philip- pines and elsewhere. The lots in Timor-Leste will be small but, with the right publicity, could probably attract companies from neighbor- . 47 One criterion for a winning bid will be the monthly O&M fee the ing Indonesia and Australia. provider expects to charge customers. 35 36 Timor-Leste: Key Issues in Rural Energy Policy of the package. The grouping criteria relate mainly to the designated households in the service area. Before the physical location of the potential users: the areas must be installation phase, the company will submit equipment reasonably close enough to each other so that as many prototypes for inspection and acceptance by the govern- systems as possible can be installed and later serviced ment, train local technicians, and conduct site visits to periodically by the provider with a minimum number of meet with local officials and residents to promote the technicians and trips to the site. It may be necessary to project and provide basic information about the systems group together homes that belong to several communi- to be installed. Specific maintenance obligations will be ties or portions of other communities into one package defined in the contract, for example, responding to a cus- to meet the scale criteria, and to try and include in each tomer's call about malfunctioning equipment and fixing package some baseload public facilities. it within 15 days, and the like. However, the company will have flexibility in deciding how to organize its main- tenance response system, for example, one or two local International Competitive Bidding persons may be trained and paid to respond to relatively minor maintenance problems and the company might To maximize private sector interest, the packages should send a technician from headquarters for major cases. In be put up for competitive bidding simultaneously in a any case, the contract will require the provider to send multilot single tender, with the bidder having the option a qualified technician to inspect the installations at least to bid on one, some, or all of the packages. The bid docu- once every year. Finally, the provider will be required to ments will specify the experience required, the minimum organize a battery recycling system in its service area to technical requirements for equipment, and the desired ensure proper collection and eventual disposal of all dead outputs for each package. Assuming all conditions and batteries. eligibility requirements are met, the companies that quote the lowest cost will win the package or set of pack- Payments by the government will be tied to agreed mile- ages they bid on. stones, for example, 15 percent of total contract amount against pre-installation requirements, such as acceptance The bidding would be open to all interested parties, of prototypes; 70 percent against installations (paid on a domestic or international, and the invitation to bid would quarterly basis); and 10 percent against required annual be publicized widely. Because of the relatively small size maintenance visits. The government may also withhold 5 of the business and the geographical location of the coun- percent of the total amount until the end of the contract, try, international interest may be limited to neighbors in to ensure all obligations have been met. These are sug- the region, specifically Australia and Indonesia. Neverthe- gested figures and milestones only and should be modi- less, many qualified PV companies exist in these coun- fied for specific contracts, as needed. tries and it may be useful for the government to conduct a roadshow in those countries to maximize participa- tion. The success of the program will hinge on whether Options at End of Service Contract a sufficient number of qualified service providers could be attracted to participate in the bidding process. Donor At the end of the contract period, the government has agencies currently active in Timor-Leste may be asked to three possible options: (a) leave users on their own, (b) help publicize the project and the bid opportunity to pri- get a new service contractor through bidding, or (c) per- vate companies in their respective countries. suade the existing provider to continue for another term. The first option is feasible because it is expected that during the implementation of the first phase, the local Service Contract Terms PV industry would have gradually developed. At the end of five years, it is likely that a few more local companies The winning company will execute a contract with the would be selling PV equipment and accessories in the government to provide PV service to the subject area over country, and more technicians would be available. Fur- a period of five years.48 The company will install PV sys- thermore, solar home system users in the service areas tems for public facilities, if part of the package, and solar would already be familiar with their systems and would home systems of no less than 50-watt-peak capacity to only require sources of replacement parts, such as bat- teries and lamps. The only government action needed 48. At least one replacement of deep-cycle batteries is expected for would be to organize and train community leaders to most of the users. Designing a Sustainable Solar Home Systems Program for Timor-Leste 37 handle the new situation, and connect them with private In the third option, the current service provider would be technicians and sources of replacement parts. The sec- persuaded to remain for another five-year contract by ond option is only possible if the government provides the same upfront grant described in the second option. an upfront financial incentive to persuade prospective Keeping the same contractor would be an easier, less new service providers, because it is unlikely that users' costly option to the government than conducting a new monthly payments alone would be sufficient enticement. bidding process. ANNEx 2 Outline of a Program for Dissemination of Improved Fuel-Wood Cooking Stoves The purpose of the program described below is to · To construct on-site 200 "heavy mass" improved stimulate actions by relevant units of the government fuel-wood cooking stoves for use in selected rural to design a program for dissemination of improved fuel- homes. wood stoves.49 The proposed program is based on obser- · To develop an institutional mechanism for dis- vations made by the author during a visit to Timor-Leste semination. in July 2009, discussions with local authorities and non- · To strengthen the production capability of local pot- governmental organizations (NGOs) at that time, and on ters and artisans to enable realization of the above the findings of the 2007 World Bank report on household outputs. energy issues and options.50 The suggested implementa- · To measure and compare indoor air pollution levels tion procedures draw from the author's experience with in 25 homes before and after adoption of improved improved stoves programs in Sri Lanka and other coun- stoves. tries. The details provided must be viewed as guidelines · To construct 10 demonstration improved kitchens in and ideas rather than fixed prescriptions. A high degree urban and rural areas. of flexibility, based on their own knowledge of local con- ditions, must be exercised by government designers when formulating details of the final program. Target Area and Market Size The urban and peri-urban areas of Dili and Baucau are Immediate Objectives recommended as target areas for the pilot project. More than 60 percent of households in these areas purchase The concrete, quantifiable objectives might be as follows: their fuel-wood, compared with only about 20 percent in rural areas, suggesting that the urban households may · To produce and disseminate 15,000 portable be more receptive to ideas for economizing on fuel-wood improved fuel-wood cooking stoves in Dili and Bau- consumption. Within the two districts, the specific dis- cau within a period of five years. This amounts to tribution sites should be selected on the basis of local about 25 percent of households in the two districts fuel-wood deficiency, market price of fuel-wood, and and 8 percent of total households in Timor-Leste. household density. Another reason for choosing the two districts is that they are close to an existing pottery cen- ter in Manatuto district. 49. This annex was written by R. M. Amarasekara, President ARECOP . 50. Shum and others, op. cit. 39 40 Timor-Leste: Key Issues in Rural Energy Policy Implementation Phasing A large number of efficient stoves have been devised in many developing countries. The models used in those The project will consist of a preparation phase and a dis- countries probably cannot be directly introduced into semination phase. The preparation phase, with a duration Timor-Leste without some modification. In the context of eight months, will require initial assistance from inter- of socioeconomic standards in Timor-Leste, the following national experts, particularly in developing stove designs, considerations, in addition to user preferences, have to improving production infrastructure, and capacity-building be taken into account in developing suitable models: of project staff. The international assistance may consist of an energy and stove expert and a technician experi- · Considering the general poverty level, the stove has enced in stove construction, clay technology, indoor air to be cheap. pollution monitoring, and kitchen improvement activities. · Considering the lack of artisan skills and the primitive Their services will be provided intermittently within the technology currently used, the stove design has to be first year after initiating the project. The outputs of this simple so that its construction is not complicated. phase will be two proven stove models for the urban · The stove should use locally available raw materials and rural areas (discussed below), and the infrastructure and skills. needed to produce about 200 units a month, with capa- bility to be increased gradually according to demand. The second phase would be devoted to marketing and dis- Possible Urban Stove semination activities. For an urban stove suitable for commercial marketing, using or adapting a configuration similar to the portable Development of New Stove Models to clay stove developed for Africa by ITDG and being propa- Suit User Needs in Timor-Leste gated in Malawi by the GTZ-funded Integrated Food Secu- rity Programme is suggested (figure A2.1). The stove is Developing a stove design suitable for local conditions is reported to be capable of 40 percent fuel-wood savings. complicated and needs considerable amounts of patience The steps in constructing the stove are described on the and time. The stove not only has to be energy efficient Household Energy Network (HEDON) Web site (http:// but has to meet other consumer aspirations and socio- www.hedon.info/PortableClayStoveConstruction). economic requirements to achieve social acceptance and widespread dissemination. The design, if it is to be Prototypes of the stove should be constructed and sub- produced locally, has to take into account the production jected in the lab to basic tests for thermal efficiency, skills and infrastructure available to undertake production including water boiling and controlled cooking. To estab- of stoves. It is recommended that the assistance of inter- lish social acceptance, the stoves should be further tested national stove experts be obtained at this stage. in 15 to 20 households before large-scale dissemination. Two stove models are proposed to be developed for Timor- The Stove Project Implementation Organization (SPIO) Leste: (a) a single-pot portable clay stove to be produced in Bili Bala, Manatuto, for the urban market; and (b) a two- pot heavy mass stove, made of brick, cement, or mud, for rural kitchens to be made on-site by trained artisans. figUre a2.1 The malawi sToVe: a PorTable Urban sToVe The purpose of the small rural stoves component is to test the demand for this type of stove in rural areas, where cooking patterns may be different. It is highly possible that the portable "urban" stove will be preferred by more rural families than the more expensive, larger stove. The second five-year phase of the program will thus include dissemination of portable stoves in areas beyond Dili and Baucau. Source: http://www.hedon.info/PortableClayStove Construction. Outline of a Program for Dissemination of Improved Fuel-Wood Cooking Stoves 41 project staff have to be trained on how to conduct these lab tests. Sophisticated tools and equipment are not figUre a2.2 a massiVe rUral sToVe bUilT on-siTe required. The tests can be carried out in a room with good ventilation. The basic equipment needed includes · scale of at least 6 kilograms capacity and 1 gram accuracy, · heat-resistant pad to protect the scale, · heat-resistant gloves, · digital thermometer with 0.1 degree accuracy, · timer, · aluminum pots of 3 and 5 liter capacity, · small spatula, tongs to handle charcoal, · metal trays, · digital moisture-measuring meter, and Source: http://www.hedon.info. · digital thermocouple to measure up to 1,300 degrees C with probes may be required to monitor kiln performance. which are large, requiring greater cooking loads. A grate may also be added to improve combustion, if necessary, Rural Stove depending on user responses in the field-testing pro- cess. The pot seats may also require changes depending For the rural stove built on-site, the program may start on the size and shape of the pots used.51 with the two-pot heavy mass stove presently used in Sri Lanka and shown in figure A2.2. The stove has a thermal Under the proposed program, about 200 rural stoves efficiency of 22 percent and recorded fuel-wood savings would be built on-site. Including a heavy mass stove of 30­40 percent. The heavy mass stove must be built component in the program will enable local artisans to on-site by an artisan and obviously cannot be marketed be exposed to and trained in this particular improved like the Malawi portable stove. stove technology, assist them with identifying initial cli- ents, and potentially start them on the way to a long-term The totally different design is needed because urban business activity in their localities. After the project, the cooking and rural cooking have different characteristics. trained artisans are expected to find their own markets Urban cooking is mostly short-time cooking with a small and possibly develop other heavy mass designs using cooking load. In contrast, rural stoves are kept burning various materials to suit local user needs. for longer periods and the cooking load is larger because agricultural and livestock activities require that cattle food and the like be prepared. Accordingly, the rural stove Production Strategy needs to be heavy mass while the urban stove is light and uninsulated. The proposed rural stove is a two-pot Urban Stove stove that has to be insulated using bricks and mud or cement. The cost will vary according to the material used. A total of 15,000 urban stoves is proposed to be pro- If mud is used, the stove will be cheaper. In any case, it duced over a period of five years, with the local pro- is clear that the massive rural stove will cost more to duction capacity reaching about 500 stoves a month in build than the portable urban stove. Users selected for the fifth year of the program. To achieve this production the initial stoves would probably be households with rate the existing pottery production center in Bili Bala, many members, households with livestock, and small Manatuto, could be used. Bili Bala's major output now is rural food vendors. handicraft items, but it has some experience in producing clay stoves for burning coffee hull briquettes (for which A short period of field testing of prototypes should be there seems to be no consistent market). The current conducted before the regular installation program. One capacity of the center is severely limited by the small factor to determine is whether the volume of the combus- tion chamber is sufficient for Timor-Leste's rural families, 51. The UNDP has reportedly installed about 40 heavy mass rural stoves under a recent project. However, at the time of writing, the report for the UNDP project was still unavailable. 42 Timor-Leste: Key Issues in Rural Energy Policy number of skilled workers and the small kiln capacity. Rural Stove The ex-factory prices of stoves and handicraft items are The rural stove would be made of brick, cement, or mud very expensive compared with prices of similar items in depending on the availability and cost of raw materials. A neighboring countries. The introduction of improved pro- three-day training course will be conducted for selected duction techniques could potentially reduce costs con- artisans, covering both technical and social factors siderably. If used as a stove production center for the related to stove installation. Training will also be provided project, the factory will need technical and financial assis- to stove users on how to cook with and maintain the new tance. The available infrastructure is sufficient to meet stove. Selection of the users and artisan trainees will be the production targets and does not require any major done by the respective community-based organizations improvements other than moulds, drying boards, and (CBOs) after completion of awareness programs. Users the like. However, a new kiln design must be introduced would be expected to provide the materials for the stove to replace the current method of firing using a primitive and unskilled labor to assist the artisan. The construction bonfire. New drying techniques will also need to be intro- of the stove is estimated to take one full day and the duced. Clay supply appears to be of no concern because stove will take at least one week to dry before any cook- plenty of clay fields are available in close proximity to ing can be done on it. The artisan is expected to moni- the center. Staff training will be necessary to improve tor the stove's performance in the initial days of cooking. productivity. Two potters should be trained for about two With proper maintenance by the user, the lifetime of the months. The training will include clay preparation, throw- heavy mass stove is about four years. ing, drying, firing, and use of moulds. The total cost of constructing the stove is estimated to be An advantage to using the Bili Bala factory is the exis- about $25, of which about $6 will be paid to the installer tence of a traditional pottery village about 2 kilometers for services rendered. The balance will be mainly the cost away. These potters could be tapped for production of of materials, which the user is expected to provide. improved stoves. They presently produce pottery items such as pots and handicraft items using very primitive methods. With support to upgrade their skills and infra- structure, such as training in improvement of clay mix- Organizational Arrangements ing and throwing, and introducing kilns to fire stoves to replace the primitive bonfire methods, productivity In many countries where stove programs have been suc- could be significantly increased. At the fifth year of the cessful, NGOs and similar organizations have taken the proposed program, production capacity for urban stoves leading role. It is recommended that a suitable nongov- from all sources should be about 500 units per month. ernmental entity be selected for this purpose in Timor- About 300 units could be produced at the Bili Bala fac- Leste. First, it is proposed to form a national steering tory while 200 units could be produced by the potters committee consisting of a few representatives of rel- from the village. Stove production will begin at the Bili evant government agencies (for example, MAFF Minis- , Bala center because certain skills and infrastructure are try of Health, Secretariat of State for Energy Policy, and already available, then will be expanded to the village others), NGOs, and women's groups. The steering com- potters. mittee would then select the organization to implement the improved stoves program, the SPIO. The selected It is estimated that the materials for the Malawi urban organization does not need to be an existing NGO with stove would cost about $1 per unit. The stoves would experience in improved stoves work. It could be any NGO be sold ex-factory for $3 per unit, ensuring a reasonable or any organization already carrying out activities with profit margin for the producer. At a volume of 300 stoves strong community participation, in fields related to pov- per month, the total sales will be $900 per month and the erty alleviation, health, gender empowerment, or energy. profit $600 per month before labor costs. Producers will The SPIO staff will initially be trained by the international be provided with a 50 percent advance to help cover the experts, and will be responsible for the management, raw materials, transport, and labor costs for a month of financial administration, coordination, progress reporting, production. The estimated amount, $150, can be used as and overall execution of the project. a rolling fund. At the initial stages of production, produc- ers will be guaranteed that all stoves will be purchased The steering committee would monitor implementation by the SPIO. progress, help promote public awareness at the broader level, and help with the flow of funding. While direct Outline of a Program for Dissemination of Improved Fuel-Wood Cooking Stoves 43 involvement of the government in implementation would The portable urban stoves would be bought ex-factory for be limited, its commitment and interest are essential to $3 per unit by the SPIO. The SPIO would distribute the the success of the program. stoves to the retailers--the shops and the CBOs--at a price of $1 each. The retailers will be required to price the The SPIO will select or organize six or more CBOs to stoves at $2, thus making a profit of $1 per stove.52 promote, sell, or install the improved stoves at the suco or village level in the selected districts. To create demand for stoves, the CBOs will be trained and provided funds Rural Stove and promotional materials by the SPIO to conduct local The purpose of the rural stoves component at this stage awareness programs. Each CBO would aim to sell in its would be simply to demonstrate the technology and assigned area about 500 urban stoves and install about train a core of installers. The effectiveness of the dem- 40 rural stoves a year beginning with the second year of onstration stoves and user acceptance will need to be the program. A small profit would be made by the CBOs assessed before designing a financing mechanism for for every stove sold or installed. The SPIO will supply wider dissemination. stoves monthly to the CBOs. The SPIO will also select retail stores or shops where stoves could be sold. Demonstration Improved Kitchens Marketing and Dissemination Strategy In any effort to reduce indoor air pollution and related health impacts, provision of improved stoves is only one component of the solution. Improved ventilation and air Urban Stove circulation, orderly storage and processing of food, and As discussed in the main report, the urban improved other simple but ergonomic improvements should also stoves would be disseminated with a significantly sub- be promoted. The program should finance the construc- sidized retail price. tion of 5­10 demonstration kitchens incorporating these features to serve as models. The model kitchens could The stoves will be marketed through two different be used as sites for cooking demonstrations and other routes: promotional activities. Elsewhere, the demonstration of improved kitchen environments effectively promoted not · selling stoves through CBOs and women's groups only improved stoves but also improved kitchen behavior and in women. · selling stoves in public markets and hardware shops. Budget Requirements Involvement of CBOs is necessary to provide improved stoves to users who are relatively more isolated or far It is estimated that the proposed five-year program would from commercial markets. Initially, because of the lack cost between $300,000 and $350,000 excluding the cost of a well-developed market economy and the absence of international consultants needed at the design and of marketing channels, the SPIO has to play the role preparation stage. of distributing agent. Later, as the market develops, this role can be gradually transferred to private sector distributors. 52. Considering prevailing costs in neighboring countries, it is esti- mated that in an unsubsidized program, the selling price of the stove would need to be about $7 or a margin of about $4 from factory to retail store to cover all distribution costs and profit margins by the middleman (no longer a funded SPIO) and the retailer. ANNEx 3 Economic Analysis of Domestic Biogas Systems In Timor-Leste, a program has been proposed to provide wood = $52 (for 3.65 tons). Digester output of average- biogas systems to some 4,600 rural households, which is quality gas would be sufficient to supply each household the estimated number of households owning at least one with about 45 percent of its cooking needs. Full results to two head of cattle. The main purpose is to displace the for several ranges of the key parameters--gas quality, fuel-wood presently used in these homes with a clean digester cost, fuel-wood price--are shown in table A3.1. and renewable fuel substitute. In the initial configuration, one small family-size digester (FSD) capable of producing For a cooking-only program to be attractive from a country about 2 cubic meters per day (m3/d) of biogas would be economic point of view, three conditions are necessary: shared by three households, using waste from a total of four to six head of cattle owned by the families. · costs for the digester must be at the bottom end of the published range, or about $500; In this annex preliminary economic analyses were made · gas quality must be high so as to replace as much of the above configuration and of a case in which only fuel-wood as possible; and one household uses the FSD and splits the output · most important, fuel-wood prices need to reflect a equally between cooking and lighting. In addition to the growing scarcity resulting from population growth, fuel benefits, the economic analyses counted the ben- , increased GDP and consequent deforestation. efits of producing nitrogen fertilizer from the digester sludge. Using as a base case the mid-range values for Under such conditions, biogas can replace almost 60 capital costs and gas quality, the results clearly point to percent of fuel-wood consumption for a typical rural the low economic viability of a program designed mainly household, generate an ERR of 33 percent, and generate to replace fuel-wood. roughly $323 in net annual benefits for each household that participates. It may turn out that economic benefits based entirely on an optimistic set of assumptions would Case 1: Cooking Only, Three induce individuals to make such investments. Households per Digester However, under the base case set of assumptions (a) If three households use one biogas digester for cooking less than 50 percent of cooking needs are supplied by only, the economic rate of return (ERR) is negative, at biogas; (b) the ERR is negative at -0.23 percent; and (c) -0.23 percent; that is, if the initial cost = $750, annual negative net benefits in excess of $150 per participating operations and maintenance (O&M) = $40,53 annual (net) household are generated. Such a set of assumptions and fertilizer benefit = $10, and annual value of replaced fuel- outcomes would not prove sufficient to induce invest- ment from private individuals and would call into ques- tion a publicly financed program. 53. O&M expenditures are calculated at 2 percent of investment costs plus $25/year for labor, based on UN Biogas Manual, page 96. (Food and Agriculture Organization of the United Nations, "Biogas Technol- " ogy: A Training Manual for Extension, Nepal, September 1996.) 45 46 Timor-Leste: Key Issues in Rural Energy Policy Table a3.1 case 1: Three hoUseholds Per digesTer, cooKing-only scenarios (2 m3/d oUTPUT) case base (with outcome under given parameters base best worst low-cost digester) Costs (present value $) 1,141 848 1,316 848 Benefits (present value $) 673 1,816 375 673 Net benefits (present value $) (468) 968 (941) (175) ERR (%) (0.23) 32.31 (23.38) 4.97 Key parameter values Digester cost ($) 750 500 1,000 500 Gas quality (MJ/m ) 3 19.36 25.10 14.34 19.36 Wood price ($/ton) 18.90 43.00 12.61 18.90 Biogas as % of household cooking energy 44.57 57.77 33.01 44.57 Source: Mission estimates and UN data. Note: Includes fertilizer benefit; gas stove efficiency = 60 percent; fuel-wood efficiency = 15 percent. Case 2: Cooking and Lighting the results of four economic cases of single-household for One Household cooking and lighting end uses for biogas. Replacing hydrocarbon fuels used for lighting, and focus- The base case uses current fuel-wood and oil prices to ing on households that can support the proposed digester estimate benefits of substitution. With average gas qual- size by themselves for both cooking and lighting (house- ity and a digester cost of $750 per unit, the net benefits holds with four to six head of cattle), the economic cal- are negative at a 10 percent discount rate. However, if culations are decidedly more positive. Table A3.2 shows the digester can be installed for $500, net benefits are Table a3.2 case 2: one hoUsehold Per digesTer, cooKing and lighTing scenarios (2 m3/d oUTPUT) case base (with outcome under given parameters base best worst low-cost digester) Costs (present value $) 1,312 955 1,526 955 Benefits (present value $) 1,189 2,814 624 1,189 Net benefits (present value $) (123) 1,859 (902) 233 ERR (%) 8.02 43.59 (7.56) 14.74 Key parameter values Digester cost ($) 750 500 1,000 500 Gas quality (MJ/m ) 3 19.36 25.10 14.34 19.36 Wood price ($/ton) 18.92 43.00 12.61 18.92 Biogas as % of household cooking energy 66.89 86.71 49.55 66.89 Kerosene price ($/ton) 600 800 450 600 Source: Mission estimates and UN data. Note: Includes fertilizer benefit; gas stove efficiency = 60 percent; fuel-wood efficiency = 15 percent. Implied crude oil prices are $600/t <=> $72/bbl, $450/t <=> $55/bbl, and $850/t <=> $97/bbl. Economic Analysis of Domestic Biogas Systems 47 box a3.1 The oPPorTUniTy cosT of fUel-wood in rUral Timor-lesTe The price of fuel-wood in urban areas of Timor-Leste has been estimated at 10˘/kg ($100/ton) based on the 2007 World Bank Scoping Study (Shum and others 2007). This price comprises (a) the costs of gathering wood for sale to urban markets, (b) the cost of transporting the wood to those markets, and (c) a retail markup for the urban wood vendor. In rural areas, where biogas plants would be built, only the gathering cost is relevant, and it is lower than the gath- ering costs for urban markets. The team estimated the cost of supplying wood in rural areas using an hourly wage based on the country's current GDP/capita of about $440. The Bank Scoping Study estimated that rural households spend about six hours per week gathering an average of 55.37 kg of wood. Based on these figures, the opportunity cost of fuel-wood for a rural household is about 1.892˘/kg or $18.92/ton. A best-case estimate of the opportunity cost of wood, four hours/week of labor, results in an economic cost of $12.61/ton. In a few years, with continued economic growth (GDP rising to $750/capita) and increased deforestation (eight hours/week for gathering) the economic cost of fuel-wood in the countryside could rise to as much as $43/ton, just under half the current urban market price. positive and the ERR is 14.7 percent. Such a result might to the base case for the cooking-only scenario--a nega- be possible if there were a mass-production capability for .6 tive rate of return (-7 percent) and a negative net pres- digesters, thereby bringing the unit cost down. ent worth of more than $900. Unlike the cooking-only alternative, a program to provide both energy and light Improved gas quality, higher benefit estimates resulting could be done for a tolerable loss under a plausible set from rising oil prices and deforestation, and low instal- of circumstances. Inducing households to invest their lation cost will generate strongly positive returns. With own funds might require simply a subsidy of $250­$300 fuel-wood at $43/ton, oil at $100/bbl, and good gas qual- per digester unit, resulting in a rate of return close to 15 ity, the household can earn a rate of return of 44 percent, percent for the household (a financial rate of return for with a net present value of almost twice the initial cost the household, not the economic rate of return for Timor- (excluding livestock). The worst-case results are similar Leste as a whole). The World Bank The World Bank Group Asia Sustainable and Alternative Energy Program 1818 H Street, NW Washington, DC 20433 USA www.worldbank.org/astae