ID!E3 1E l 3 El O E 1 t!.1 C W El l a El- El El 0 E3 El El ES} m O U*2 U m 1E22 ElS m El El E} 2=3 [a1: 0 E _w_ C}EI - EINHE elij ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAM PURPOSE The World Bank/UNDP/Bilateral Aid Energy Sector Management Assistance Program (ESMAP) was launched in 1983 to complement the Energy Assessment Program which had been established three years earlier. The Assessment Program was designed to identify the most serious energy problems facing some 70 developing countries and to propose remedial action. ESMAP was conceived, in part, as a preinvestment facility to help implement recommendations made during the course of assessment. Today ESMAP is carrying out preinvestment and prefeasibility activities in about 60 countries and is providing a wide range of institutional and policy advice. The program plays a signiitcant role in the overall international effort to provide technical assistance to the e-nergy sector of developing countries. It attempts to strengthen the impact of bilateral and multilateral resources and private sector investment. The findings and recommendations emerging from ESMAP? country activities provide governments, donors, and potential investors with the information needed to identify economically and environmentally sound energy projects and to accelerate their preparation and implementation. ESMAP's policy and research work analyzing cross-country trends and issues in specific energy subsectors make an important contribution in highlighting critical problems and suggesting solutions. ESMAP's operational activities are managed by three units within the Energy Strategy Management and Assessment Division of the Industry and Energy Department at the World Bank. - The Energy Efficiency and Strategy Unit engages in energy assessments addressing institutional financial, and policy issues, design of sector strategies, the strengthening of energy sector enterprises and sector management, the defining of investment programs, efficiency improvements in energy supply, and energy use, training and research. - The Household and Renewable Energy Unit addresses tech-nical, economic, financial, institutional and policy issues in the areas of energy u-e by urban and rural households and small industries, and includes traditional and modenm fue; supplies, prefeasibility studies, pilot activities, technology assessments, seminars and workshops, and policy and research work. - The Natural Gas Development Unit addresses gas issues and promotes the development and use of natural gas in developing countries through preinvestment work, formulating natural gas development and related environmental strategies, and research. FUNDING The ESMAP Program is a major international effort supported by the World Bank, the United Nations Development Programme, and Bilateral Aid from a number of countries including Australia, Belgium, Canada, Denmark, Finland, France, Iceland, Ire!and, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal Sweden, Switzerland, the United Kingdom, and the United States. FURTHER INFORMATION For further information or copies of the completed ESMAP reports listed at the end of this document, contact: Energy Strategy Managen.ent OR Division for Global and Interregional Programs and Assessment Division United Nations Development Programme Industry and Energy Department One United Nations Plaza The World Bank New York, NY 10017 1818 H Street N.W. USA Washington. D.C. USA 20433 INDONESIA URBAN HOUSEHOLD ENERGY STRATEGY STUDY TECHNICAL APPENDICES FEBRUARY 1990 VOLUME II . 1 - APPENDIX I INDONESIA: URBAN HOUSEHOLD ENERGY SURVEY FEBRUARY 1990 -2- ABBREVIATIONS AND ACRONYMS BPS Biro Pusat Statistik BAPPENAS Badan Perencanaan Pembangunan Nasional (National Planning and Development Board) BAKOREN Badan Koordinasi Energi (Ministerial Energy Coordination Board) DJLEB Direktorat Jenderal Listrik dan Energi Baru LEMIGAS Lembaga Minyuak dan Gas (Oil and Gas Research Institute) LKE Liters of Kerosene Equivalent MIGAS Minyak dan Gas (office of Oil and Gas) MME Ministry of Mines and Energy NUDS National Urban Development Strategy PERTAMINA Perusahaan Tambang Minyak Negara (National Oil Company) PLN Perumahan Listrik Negara (National Electricity Company) FLN-LMK National Electricity Company Laboratory and Testing Facility PrE Persatuan Teknik Energi (Technical Committee on Energy) SUSENAS Social and Economic Survey UHESS Urban Household Energy Strategy Study YLKI Yayasan Lemabaga Konsumen Indonesia (Indonesian Consumers Union) GWh GigaWatt hour k thousands kWh KiloWatt hour M mlllion MJ MegaJoules TPA tons per annum EXCHANGE RATE 1700 Rupiah = US$1 CONVERSION FACTORS EMPLOYED Heating values: I liter kerosene = 35.2 Megajoules 1 kilogram LPG = 45.77Megajoules 1 kilogram wood = 14 Megajoules 1 kilogram char. = 25 Megajoules 1 kilowatt-hour = 3.6 Megajoules Substitute rations for cooking: 1 kilogram LPG = 1.69 liters of kerosene equivalent 1 kilogram wood = .235 liters of kerosene equivalent -4 - TABLE OF CONTENTS THE INDONESIA URBAN HOUSEHOLD ENERGY SURVEY ................... 6 Survey Objectives .............. .................................... 6 Summary of Past Urban Household Energy Surveys ......................... 6 Organization of the UHESS Sutvey .................................... 10 Definition of Urban Area and Estimation of Urban Area Size .... ...... 11 Sample Frame and Sampling Procedure ........................... 11 Sample Size ............. ................................... 12 Questionnaire .............................................. 12 Survey Execution: Pretest, Training and Field Work ................. 13 Data Validation ............................................. 13 Data Base Development ....................................... 14 Recommendations for Continuation of Survey Work ....................... 16 ANNEXES Annex 1: Proposed Household Energy Module for SUSENAS ..................... 18 Annex 2: 1988 Urban Household Energy Questionnaire .......................... 20 Annex 3: UHESS Summary Tables ......................................... 36 Annex 4: SUSENAS Tables ............................................... 54 LIST OF UHESS SUMMARY TABLES Table 1: Number of Respondents by City Size and Household Expenditure Level ...... 37 Table 2: Percentage Distribution of Sampled Households by Household Expenditure Level and Urban Area Size ............................................ 37 Table 3: Household Size and Average Expenditure by Expenditure Group .... ....... 37 Table 4: Percentage of Households by Fuel-Use Combination for Each Expenditure Group .............................................. 38 Table 5: Average Household Fuel Consumption by Fuel-Use Combination for Users in Expenditure Group .................................... 39 Table 6: Average Household Fuel Consumption by Fuel-Use Combination for All Households in Expenditure Group ............................. 40 Table 7: Percentage of Households in Expenditure/Income Category ..... .......... 41 Table 8: Percent of Households Using Specified Fuel Source by Expenditure Group .... 42 Summary Tables - Urban Area Size and Energy ........................ 43 Table 9: Percentage of Households in Urban Area Type ......................... 43 Table 10: Average Household Fuel Consumption by Fuel-Use Combination for Users in Urban Area Size Category .... .......................... 44 -5- Table 11: Average Household Fuel Consumption by Fuel-Use Combination for All Households in Urban Area Size Category ....................... 45 Table 12: Average Household Expenditures on Fuel-Use for All Households in Urban Area Size Category ....................... 46 Table 13: Percent of Households Using Specified Fuel Source by Urban Area Size ............. ................................ 47 Table 14: Household Member with Major Influence on Purchase and Use of Device Mentioned ............ ................................ 48 Table 15: Reasons Non-Electrified Households Do Not Use Electricity ..... ......... 48 Table 16: Percentages of Households Electrified by "Income" and Urban Area Size ..... 48 Table 17: Pattern of Ownership and Use of Selected Electric Appliances ..... ........ 49 Table 18: Ownership of Other Electrical Appliances ............................. 50 Table 19: Summary Electricity Data by Source of Electricity ....................... 50 Table 20: Bulbs per Thousand Households by Type and Use-Level .................. 51 Table 21: Distribution of Electricity Use (kWh) for Lighting by Bulb Type and Use-Level ............. ................................... 51 Table 22: Average Household Cooking Fuel Use and Mix by Urban Area Size and Expenditure Group ........................... 52 Table 23: Main Reasons for Use/Non-Use of Cooking Fuel ....................... 53 LIST OF SUSENAS SUMMARY TABLES Table 1: Urban Household Characteristics by Province and Year .................. 54 Table 2: Average Fuel Use of Using Urban Households ......................... 55 Table 3: Total Fuel Use of Urban Households by Province and Year ..... .......... 56 Table 4: Percentage of Urban Households Using Fuel by Province and Year .... ...... 57 Table 5: Avcrage Fuel Use of Urban Households by Province and Year .... ......... 58 Table 6: Urban Household Characteristics by Expenditure Group and Year .... ...... 59 Table 7: Number of Urban Households Using Fuel by Expenditure Group All Indonesia . .................................................. 60 Table 8: Average Fuel Use of Using Urban Household .......................... 61 Table 9: Total Fuel Use of Urban Households by Fuel and Expenditure .... ......... 62 Table 10: Percentage of Urban Households Using Fuel by Expenditure Group ........................................... 63 Table 11: Average Fuel Use of Urban Households by Fuel and Expenditure .................................... 64 Table 12: Average Fuel Expenses by Using Urban Households by Expenditure Group ... 65 Table 13: Average Fuel Expenditure by Urban Households by Expenditure Group ...... 66 Table 14: Percent of Total Expenditure on Fuel by Urban Household Expenditure Group .............. ................................ 67 -6- THE INDONESIA URBAN HOUSEHOLD ENERGY SURVEY Survey Objectives 1. The urban household energy survey (UHESS) was undertaken both to support the formulation of an urban household energy strategy, and to provide a continuing source of information for policy decisions. The funding and effort expended in past urban household energy surveys (see below) indicate the importance numerous organizations attach to understanding household energy use patterns. The UHESS survey attempts to move towards a consolidated survey instrument, capable of addressing the diverse concerns of the numerous organizations involved in household energy issues. Many of these concerns are discussed in the bodv of this report. 2. The strategy development, around which this report centers, relies heavily on the survey data. For example, the proposal for a kerosene stove program draws on survey data concerning cooking habits. The proposal for an electric appliance conservation program draws on appliance ownership and use data. The analysis of kerosene-LPG substitution draws on fuel choice and use data, as well as inter-fuel substitution ratios derived statistically from the survey data. All of the strategy components utilize energy demand projections based on relations between energy use patterns anl household characteristics identified in the survey. The summary tables distilled from the survey data also provide important contextual information for all of the strategy components. Indeed, during the course of the strategy analysis, nearly all of the survey data was used in one form or another. 3. In addition to this one-time analysis of the survey data, future policy decisions will require alternative treatments of the data. During the course of developing the information campaigns for the kerosene stove and electric appliance programs, for example, it will be necessary to re-access the data in order to analyze which of the mass media will best reach the target households (e.g. refrigerator owners, or household with factory stoves), and who in the household is likely to be deciding on the type of energy-using device to be purchased. In order to facilitate the continuing use of the data, a data base was developed. Summary of Past Urban Household Energy Surveys 4. During the first half of the 1980's, in the course of more than five independent survey efforts, energy-use surveys were administered to over 10,000 urban households in Indonesia. In 1982, the State Gas Company (PGN) enlisted the aid of the Central Bureau of Statistics (BPS) to administer a survey of almost 3,000 households in nine major cities in order to ascertain the potential demand for piped gas. At about the same time, the Directorate General for Oil and Gas (MIGAS) administered a survey of over 2,000 preponderantly urban households in Java, in the attempt to improve their understanding of household kerosene demands and requirements (MIGAS), 1983). In 1984, the Department of Public Works commissioned a survey of energy use - 7 - in over 2,000 urban households in Java, in order to help assess the potential market for charcoal imported from transmigration areas. In 1985, the State Electric Company (PLN) administered a survey of almost 2,0Z0 electrified households in 19 major cities in Indonesia, in order to assess their appliance ownership and electricity use patterns (PLN,1985). Also in 1985, the Directorate General for Electricity and New Energy (DJLEB), working with a U.S. consulting firm (E/DI), oversaw a detailed survey of over 1,600 urban households in nine major cities in Indonesia (DJLEB-E/DI, 1985). Although some of these surveys concentrated on a particular fuel or policy issue, the overlap was considerable. 5. Unfortunately, the lack of systematic sampling procedures or survey instruments makes comparisons very difficult. Only the results of tbe 1985 DJLEB-E/DI survey, whose objectives were somewhat similar to the UHESS survey, will be summarized here. The DJLEB-E/DI survey was envisaged as a "pilot effort", undertaken as part of a broader attempt to collect "baseline energy consumption data" in support of the energy planning activities of DJLEB. It was designed to collect information on: (a) household characteristics (e.g. size and income); (b) household fuel consumption by end-use (including transport); (c) fuel availability and price; (d) energy-using device ownership and usage patterns. 6. A summary of the results is provided in the Tables below. The households surveyed were from nine of Indonesia's largest cities, including 7 in Java: Jakarta, Bogor, Bandung, Semarang, Yogyakarta, Surabaya, Malang, Ujung Pandang, and Medan. The sample was selected from households previously surveyed for the BPS-administered SUSENAS survey. The fieldwork was undertaken by University students, and supervised by staff of the Institute of Technology at Bandung, with the assistance of foreign (U.S.) consultants. -8- Table A1.1: END-USE FUEL CONSUMPTION PER HOUSEHOLD (tOM PER YEAR) 1985 DJLEB-E/DI SURVEY Income Group Low Middle High Average Fuel/End Use Cooking Kerosene 2.9 3.4 3.1 3.1 LPG 0.0 0.1 0.7 0.2 Wood 0.3 0.1 0.1 0.2 Charcoal 0.2 0.1 0.1 0.1 Lighting Electricity 0.1 0.3 0.5 0.3 Kerosene 0.2 0.2 0.2 0.2 Appliances Etectricity 0.1 0.2 0.4 0.2 M/ Low Income: < 20,000 Rupiah per ann (for hcusehotd). W Middtle Income: > 20,000 Rupiah and < 40,000 per anrum. g/ High Income: > 40,000 Rupiah per annum. : The inmome figures are extremely low, suggesting either a biased sample or incorrect reporting. Table A1.2: PERCENTAGE OF HWSEHOLDS CONSUMING SPECIFIED FUEL 1985 DJLEB-E/DI --AVEY Income Grouw Low Middle High All Fuel/End Use Electricity 56 69 78 66 Kerosene 91 95 78 90 LPG 1 6 28 a Wood 8 1 2 4 Charcoal 29 16 9 20 - 9 - Table A1.3: ELECTRIC APPLIANCE OWNERSHIP 1985 DJLEB-E/DI SURVEY (% OF HOUSEHOLDS OWNING) Income Group Appliance Low Middle High Ali Incandescent Bulb 77 92 94 87 Fluorescent Bulb 67 80 89 77 TV 60 76 89 72 Iron 39 69 81 59 Radio Cassette 30 53 63 45 Refrigerator 6 29 59 26 Cassette 13 19 33 20 Water Pump 8 17 47 19 Radio 10 8 14 12 Video 1 7 29 9 AC 0 1 8 2 Note: The original report (DJLEB-E/DI. 1985, pages 15-16) claims that 66% of the sample was electrified, contradicting many of the figures above. It is suggested that this results from reporting and data entry errors (page 53). More likely, it is the result of househotds sharing etectric meters: the electrification rate is based on billing information. This would suggest that about 20% of the households surveyed did not have their own connection, but did use electricity. 7. In addition to more detailed tabular presentation of these results, an attempt was made to extrapolate the results to the whole urban population in Indonesia. Extrapolating the survey estimates gives an overall urban household energy consumption of 27 mboe, of which 20.7 mboe are kerosene, 3.1 mboe are electricity, 1.3 mboe are LPG and 2.0 mboe are biofuel (DJLE13-E/DI, 1985, page 14). Given the bias towards large cities, one would expect overestimates of electricity, LPG, and possibly kerosene. For LPG and electricity, this appears to be the case. PLN estimates urban household electric sales at roughly 2.4 mboe (PLN projection spreadsheet). Total LPG sales in 1985 were roughly 1.3 mboe, of which .8 mboe was sold to industrial users. Kerosene is more diffilcult to asscss. Total kerosene sales were about 43 mboe, but it is uncertain how much went to rural households and other users. The large-city bias can also be expected to lead to underestimates of biofuels, but there is no means of obtaining supply-side estimates. 10 - Table A1.4: REASONS GIVEN FOR NOT USING PARTICULAR FUELS (PERCENTAGE OF RESPONDING NON-USERS GIVING SPECIFIED REASON) 1985 DJLEB-E/DI SURVEY Reason for Not Using Number of Investment Lack of Fuel Responses Fuel Price Costs Supply Inconvenience Electricity 161 18% 70% 15% 4% Kerosene 131 21% 14% 6% 59% LPG 1462 28% 63% 5% 4% City Gas 1621 14% 40% 43% 3% Wood 1511 2% 2% 14% 82% Charcoal 1275 4% 1% 7% 88% 8. Inter-urban comparisons of energy use are also made. Broad similarities exist but, according to the report, "significant variations are also evident" (page 23). Unfortunately, the cities are not grouped into different types, and no statistical analysis of inter-city differences was performed. As a result, it is not possible to determine which of the variations are significant. 9. Urban-rural comparisons are made, drawing on past rural surveys. As would be expected, electricity and LPG are shown to be urban fuels, while wood is predominantly rural. Kerosene dominates urban cooking and rural lighting, and only accounts for a small share of rural cooking and urban lighting. Again, the large-city bias is likely to exaggerate these differences. 10. It is claimed that among all surveyed households, energy expenditures averaged approximately 17% of total annual expenditures. This figure is far in excess of the estimates provided by SUSENAS surveys, and is likely to include expenditure; under-estimation. 11. Viewed as a sources of information on urban household energy, the DJLEB-E/Dl survey is a useful addition to a meager data set. Viewed as a pilot survey, it reveals two significant dangers: (a) under-representation of small urban areas, and; (b) serious mis-estimation of income. Both of these problems reappear in all of the other urban household energy surveys noted above. The UHESS survey was purposely designed to avoid these pitfalls. Organization of the UHESS Survey 12. The survey was executed by The Central Bureau of Statistics (BPS), with assistance and supervision from ESMAP and DJLEB. An inter-ministerial committee was convened to guide the questionnaire development 1/ and a steering and technical committee of DJLEB and BPS staff was created to supervise the overall progress of the survey. DJLEB staff and the ESMAP Project Manager participated actively with BPS staff in all stages of the survey design and execution. The LI This committee was later supplanted by thle PTE sub-committee on household eneVy. 11 X survey represented the first cooperative effort involving DJLEB and BPS. The validated data was sent to Lemigas for further processing and analysis by the project team. Definition of Urban Area and Estimation of Urban Area Size 13. For the purposes of the UHESS, an urban household is one residing in what BPS identifies as an urban "desa" (village). For each desa, BPS assigns a "rank" score between one and ten for each of three sub-criteria: population density, percentage of agricultural households, and number of urban facilities. The ranking system is summarized in the Table below. A village with a score over 20 is urban, while marginal villages with scores of 19 and 20 are delineated according to further sub-criteria involving distance to closest town and growth prospects. Agricultural households are defined as those which make their living primarily from agriculture, though in practice the principal occupation of the household head is used as a short cut (Rietveld, 1988). Urban facilities include educational, medical, industrial, transport, electricity, and even some retail facilities. Tabte A1.5: RANKING SYSTEM FOR IDENTIFYING URBAN DESA Population X of Agricultural Number of Urban Rank Density Households Facilities Value (P/WA2) < 500 > 95 - 1 500- 999 91 - 95 0 2 1000 - 1499 86 - 90 1 3 1500 - 1999 76 - 85 2 4 2000 - 2499 66 - 75 3 5 2500 - 2999 56 - 65 4 6 3000 - 3499 46 - 59 5 7 3500 - 3999 36 - 45 6 8 4000 - 4999 26 - 35 7 9 5000 + < 26 8 10 14. To estimate the size of each. households' urban area, the functional definition of the National Urban Development Strategy (NUDS) was employed. NUDS created a list of over 800 functional urban areas in Indonesia, by combining contiguous urban desa (with some exceptions), regardless of intervening administrative boundaries. NUDS estimated the size of these urban areas based on the 1980 Census, and projected their population up through the turn of the century. This data was made available by The Directorate General of Human Settlements. Sample Frame and Sampling Procedure 15. The UHESS sample was sub-selected from the urban households surveyed for the 1987 BPS Social and Economic Survey (SUSENAS). Descriptions of the SUSENAS sampling procedure are available from BPS. Briefly, Census blocks are selected systematically from an ordered spiral of - 12 - blocks extending from the center to the outskirts of each urban area (at least for those urban areas of sufficient size to create a spiral). Households within selected blocks are pre-surveyed, ordered in terms of expected income class, and households are taken at fixed intervals from the resulting list. The number of households selected within a block is proportional to the size of the block, but generally lies between five and twenty. 16. For reasons of cost, the UHESS survey was administered only in Java. Given the choice between restricting the survey to large cities and restricting it to Java, the latter was considered preferable. The results confirm that variation among different sized urban areas is far more significant than inter-island differences, though in future surveys it would be preferable to include all the island groups. 17. To select UHESS households from the SUSENAS sample, the SUSENAS-Java sample was first stratified by 1980 urban area size, as defined by the National Urban Development Strategy Study (NUDS). The UHESS blocks were selected to ensure proportional representation of each urban area size category. Over sixty urban areas are represented in the UHESS sample. 18. In case of houses no longer in the Sample Frame location, it was agreed that the new residents (or, failing that, a neighbor) would be administered both the UHESS survey and the SUSENAS expenditure module. Sample Size 19. The sample included 2,702 households, spread out over about seventy urban areas and two hundred and twenty blocks (slightly less than half of the SUSENAS blocks). As is made evident in the data analysis, this number of households and blocks is sufficient for the purposes envisaged. Users should understand the limits to the data set, however. For example: (a) The sample size is not sufficient for city specific analysis of energy use in any but the largest cities; (b) The sample size is not sufficient for analyzing the characteristics of appliances owned by less than about 5% of the population, such as air conditioners, and other luxury white goods; (c) The number of blocks are insufficient for the relatively rare and regionally clustered energy technologies, such as city gas and high-power electric tariffs (i.e. R3 and R4). Questionnaire 20. A translation of the final household questionnaire is provided below. The questionnaire was developed in November and December of 1987, and pre-tested on thirty households in Bogor in mid-January of 1988. BPS, DJLEB, and ESMAP were all involved in the questionnaire design, and comments were received from a multi-Ministerial task force. Most of the questions are directed - 13 - at household members (the principal man and woman of the household). Surveyors were instructed to observe appliances, to corroborate household responses, and to engage in direct measurement in certain cases. Specifically, wood-fuel consumption and kerosene consumption for secondary purposes were physically measured over a two day period for all surveyed households. Disregarding the direct measurement (which required multiple visits), the questionnaire could be finished for almost all households in less than one hour, and often in as little as half an hour. 21. To complement the household data, a small data set was collected by a supervisor for each of the blocks surveyed. The questions concerned fuel supply and availability. Survey Execution: Pretest. Training and Field Work 22. Surveyor instructors and other BPS staff administered the pre-test surveys to thirty households in Bogor in mid-January 1988, with the help of DJLEB and ESMAP staff. As a result of the pre-test, several questions were dropped, and improvements were introduced. The most important finding of the pre-test was the importance of inter-household electricity sales, with the metered households selling to their neighbors. 23. From the questionnaire and the experience fo the pre-test, a manual for surveyors and supervisors was developed. Eight instructors were trained February 7-9, 1988. One-hundred-and- forty-one surveyors and sixty-nine supervisors were trained between the 11th and 20th of February, with separate sessions in each of the five provinces (Jakarta, West Java, Central Java, Yogyakarta, and East Java). The surveyors and supervisors were BPS professionals. 24. The field work was carried out in March, 1988. Since the surveyors live in the vicinity of the households to be surveyed, the precise dates were determined by the surveyors and their supervisors. Inspection teams, including DJLEB and/or ESMAP staff, were sent out from the BPS central office to each province. Data Validation 25. BPS was charged with the preliminary data processing. The questionnaires were reviewed by supervisors, according to procedures outlined in the supervisors' handbook. Surveyors were requested to revisit household when anomalies arose. Regional officers also reviewed the questionnaires. 26. Tne data was entered directly into Personal Computers. BPS developed a data validation program designed to detect both intermal inconsistencies and unreasonable values. The Questionnaires failing the validation tests were reviewed, and the flagged inconsistencies resolved. Three iterations were required before BPS considered the data to be suitable for analysis. The data were then transferred to Lemigas, where further data validation was undertaken using interactive software. The majority of the errors uncovered at this stage involved unrealistic technology characteristics, particularly with regard to electric appliances. - 14 - Data Base Development 27. The data base development was undertaken by Lemigas. Seven files were created, in both SPSS and @Base versions. Each fie has the same first household identifier variables, including summary fuel-use data, and household characteristics likely to be relevant to most aspects of energy use. The remaining variables depend on the file topics, which may be functional or fuel-specific: (a) File One - General; (b) File Two - Cooking, (c) File Three - Lighting; (d) File Four - Electricity; (e) File Five - Kerosene; (f) File Six - LPG; and (g) File Seven - Biofuel. 28. A list of the variables included in each file is provided in the UJHESS Data Base Development Report by Lemigas. There is considerable overlap between .he files. To the extent that the file organization is successful, there should be a single file providing all of the variables relevant to any given substantive issue. Different government agencies are likely to be interested in a different subset of files. 29. In many cases the variables were not directly from the validated survey data, but involved computations. All energy consumption figures are expressed on a daily basis in the data-base, while the questionnaire includes a variety of units and periods. Most conversions were straightforward, and involved simple linear transformations, but in some cases more elaborate estimation procedures were required. 30. The most complex data processing was for the estimation of electricity use by appliance. Three sources of information regarding appliance energy use were employed: (a) survey-based estimates of wattage and hours of use available for every household; (b) crude technical estimates of expected appliance use for "average" appliances sold in Indonesia; and (c) statistical analysis of the relation between appliance use and recorded electricity consumption. The first source alone was insufficient. To a rough approximation, hours of use times appliance wattage equals watt- hours of electricity consumption. However, there are appliances for which hours of "use" are not equivalent to 'on-time" (e.g. automatic water pumps and refrigerators), the wattage information coLlected is not appropriate (e.g. it may be maximum wattage) or systematic mis-estimation occurs (e.g. it is more likely for households to omit lamps that exist than to include ones that don't). For - 15 - these reasons, whenever the survey based watt-hour estimates were contradicted by statistical analysis, efforts were made to improve the original estimates on the basis of technical or statistical analysis. 31. To detect mniestimation statistically required several steps: (a) selecting heaseholds with metered electricity and no commercial uses of electricity; (b) regressing electricity consumption on the estimates of individual appliance electricity use; (c) identifying appliances whose regression coefficients were significantly different from one. For televisions, refrigerators 2/, and irons, the statistical analysis corroborated the original watt-hour estimates. For pumping, air conditioning and cooking, the statistical analysis indicated significant errors in the original estimates. For water pumping, technical estimates of expected appliance electricity consumption also indicated that the original estimates were in error. Furthermore, there was an obvious source of error in the original estimates: households are often not aware of how long their pumps operate. For air conditioners, a technically based estimate lay somewhere between the original estimate and the statistical estimate. Again, there was a possible source of error: confusion over the treatment of second and third air-conditioning units, with a resulting loss of information. In both cases, the household appliance estimates were multiplied by a constant designed to ensure that the average consumption per appliance corresponded to the technical estimate. The possible bias in the cooking estimates were ignored given the relative unimportance of elecfic cooking, and the likelihood of omitted variable bias in the OLS coefficient. ./ 32. For lighting, the statistical analysis suggested a 26% underestimate of watt-hours in the survey data. Furthermore, average household electricity use with the sum of ioentified appliance use indicated an overly large residual for "other electric appliances". In response, all household lighting estimates were increased by 1.36 to compensate for this underestimation. More recent analysis suggests that the lighting underestimation was actually limited to households with a large number of lamps. Further detailed analysis of lighting should extend this procedure to account for this observation. Z/ Refngerts we asnumed to operae 65% of the tine. / Households wih ie-coorkes are also more likely to own appliances omitted for the equation, thereby biasing the coeficiaent upws. The same bias may akso exi for aiconditioner. - 16 - 33. The equation below summarizes the OLS results following these adjustments: TOTELEC a 4.6 + 1.0*11 + 1.0*FL + 1.0*IR + 1.1*TV, 1.I*RE + 1.O*PU + 3.4*AC + 3.1MCO T =(1.6)(14.5) (8.9) (4.2) (6.3) (15.7) (8.9) (15.6) (7.8) R squared u .83 F 744 Where: TOIELEC: total (metered) electricity consumption IL: electricity used by incandescent lamp. FL: electricity used by fluorescent lamp. IR: electricity used by iron. TV: electricity used by television. RE: electricity used by refrigerator. PU: electricity used by pump. AC: electricity used by air conditioner. CO: electricity used for cooking. 34. As well as data processing, the creation of the data base involved adding information from other sources. The two principal additions were urban area size and household expenditure. Urban area size was transferred from a NUDS file, including population estimates for all cities up to the turn of the century (1988 estimates were interpollated). Expenditure was taken from the SUSENAS expenditure module. As most households in the sample were from the 1987 SUSENAS sample frame, their 1987 total expenditure was inflated by 10% to account for the increase in nominal personal consumption per capita in urban areas. For the 10% of households which had to be replaced, 1988 SUSENAS data was available. Neither 1988 nor 1987 SUSENAS expenditure data was available for a final 10% of households from the 1987 SUSENAS sample frame, as their SUSENAS questionnaires had not yet been processed. Recommendations for Continuation of Survey Work 35. By analyzing both the UHESS data for 1988 and SUSENAS data for 1981, 1984, and 1987, it was possible to gain a number of important insights into the urban household energy situation, and to develop a useful data base. Unfortunately, the energy data from SUSENAS only includes expenditures and quantities, and for electricity and fuelwood the quantity data is either low quality or not collected. Alternatively, the UHESS survey covers only urban households in Java, and does not provide a portrait of changes in fuel use patterns over time. By adding a small energy-module to the SUSENAS expenditure survey, a three-yearly data collection effort could become institutionalized. Given the experience of the UHESS survey it should be possible to develop a two page module, ensuring that the most critical information is collected. While some work would be required to ensure its applicability to the rural households, the resulting information would be invaluable for the formulation of energy policy. A draft 2-page household energy module for SUSENAS is included below. 36. By incorporating a revised and tested version of this energy module in the SUSENAS expenditure survey, the advantages of the UHESS and SUSENAS surveys could be combined to - 17 - provide a dynamic picture of the energy transition currently in process among urban households. The implementation of a coherent urban household energy strategy would be greatly facilitated. Perhaps more important, an important part of the information needed to develop a rural household energy strategy would become available. Finally, policies, such as kerosene pricing, which affects both urban and rural households, could fnally be based on a solid foundation of knowledge concerning both who bears the cost of kerosene price changes, and how they react. - 18 - Annex 1: Proposed Household Energy Module for SUSENAS Check if fuel is used for: Use Cookina Lishtins Water H. Space Heat Com*ercial Other Electricity l I I I l l l Kerosene j l l l l l LPG l l I l l l 1 Charcoal t l l l l l l Wood/Biofuel t l l l l l l City Gas j _ _ I | I Electricity If don't use, why? _ non-electrified area _ other. Connection/Source: _ PLN direct _ Other household's meter Battergy (volts __) _ Non-PLN Electricity Distributor If use for Commercial purposes, specify - (code for type) If PLN: Meter Number KVA: KWh/Mo (most recent period available): Electric Appliances use (Number): Fan Kettle Color TV Other Electric Oven Window Air Conditioner Pump Video Freezer Refrigerator Sound System Central Air Conditioner Radio 8&W TV Hair Dryer Blender/Mixer Electric Stove Washing Machine Iron Toaster Rice Cooker Electric Lighting: Incandescent: Fluorescent: Number of Bulbs: Number of Bulbs: Total Wattage: Total Wattage: Kerosene If do not use, is kerosene available? Source: _ Delivered _ "Warung" _ Dealer _ Other. Price: _ Rp/liter Use during past week: _ liter For cooking: _ liter For lighting: _ liter Other: _ Code If cook: Stove 1: wicks - days used in last week: Stove 2: wicks - days used in last week: Stove 3: wicks - days used in last week: if light: Lanp 1: code for type - days used in last week: Laup 2: Lamp 3: Lamp 4: Price of Kerosene: Source: _ Deliver _ Shop _ Dealer _ Other - 19 - Wood (Thfs sectfon should be both pretested and calibrated.) if use for commercl purpose, specify _ (code for type) Average daely use during pest week: _ kg Of which kg home cooking & water heating kg commercial use kg other Extra use during cold months _ kg/day for _ months (approximate) Sources of wood/biofuel used in pest week: I Principal Source/Type. 2 Used daily In large quantity. 3 Used at least one day in large quantity. 4 Used daily In small quantities. 5 Used at least one day in small quantities. Land Tvye Buy Own Neighbor Marginal Forest Fallen/Dead 8ranches Rice Nusks uPruwed" Branches ("pollarding') "Lops and Topse Leaves Coconut tree parts Tree fetled for fuel Cassava Stalks Other Agricultural LPG If do not use, Is LPO available? _ Source: _ Deliver _ Shop _ Dealer _ Other if use for commercial purposes specify _ (code for type) Number of 11 kg bottles ouned: _ Of which _ back-up; Uses include: code Consumption in Kg/wk: (WMRK SPACE PROVIDED FOR CACULATING FROM LENGTH OF TIME BETWEEN REPLACEMENTS) Price of contents (1ikg): - Rupiah Number of 45 kg bottles owsed: _ Of which _ back-up; Uses include: code Consumption in Kg/wk: (WORK SPACE PROVIDED FOR CALCULATING FROM LENGTH OF TIME BETWEEN REPLACEMENTS) Price of contents (45kg): _ Rupfah Stove: _ burners Ovnn: -_ Y/I Water Heater _ Y/N Other: _ Y/N City Gas Meter No: _ Monthly co_rnu tion durng last period metered: _ cubic meters. Stove: _ burners Om.n: _ Y/N Water Heater: _ Y/N Other: _ Y/N Charcoal Price: _ Rp/kg Use over past week: _ kg Of which - kg ironing - kg sate - kg other cooking _ kg commercial purpose *20 - Anr*4 : 1988 Urban Household Energy Questionnaire Republic of Indonesia Central Bureau of Statistics Confidential 1. IDENTIFICATION OF LOCATION 01 Province l li 02 Regency/Municipality Iii 03 Sub District |__ 04 Village ll_I 05 Region URBAN 06 City Classification 1980 LI 07 Enumeration District Number 08 Block Census Number l_|} 09 SUSENAS Satple Code Nurber L i It. EXPLANATION OF ENUMERATION 01 Number of Enumerator 03 Enumerator Signature 05 Date of Supervisor 02 Date of Enumeration 04 Name of Supervisor 06 Supervisor Signature 111. GENERAL INFORNATION OF THE HOUSEHOLD 01. Name of the head :.......................... 02. Sex of the head : Male -1 Female -2 03. Age of the head: . .......... years LJI 04. Number of family members: . .......... persons LJJ 05. The highest education graduated by the household head Never attended school -0 Not Graduated from Elementary School - 1 lemtentary School - 2 Junior High School (general) - 3 Junior High School (specialist) . 4 LI Senior High School (general) - 5 Senior High School (specialist) - 6 Diploma I/Diploma It - 7 Academy/Diploma III - 8 University/Diploma IV - 9 06. Level of household members in neis/magazine reading habits, radio listening habits, TV watching habits. Encircle one of the suitable codes in coluin below: Activity Head Husband/wife Another family mebers (1) (2) (3) (4) 1. News/Magazine reading habits 1 2 3 0 1 2 3 0 1 2 3 l_ Ii lI 2. Radio listening habits 1 2 3 01 2 3 01 2 3 3. TV watching habits 1 2 3 0 1 2 3 0 1 2 3 L Code: Don't have husband/wife/other family menbers - 0 Regularly . 1 Sometimes - 2 Never . 3 - 21 - IV. TYPE OF BUILDING 01. The living home status : Own - 1 Credit - 4 Contract - 2 Government - 5 Rent - 3 Other * 6 02. The building utilization : Only for living * (go to question 4) L For livfng and working - 2 03. If the answer to question 2 is for living and working, what kind of work: Shop '1 Barber/Hai" dresser * 2 Laundry - 3 |_ Tailor - 4 Household industry - 5 Battery charging - 6 Others - 7 04. How long has the family been living in this building : .. ... year I_I_ 05. How many rooms are there in the building ....... rooms Li 06. Physical building : Single building - 1 Single building with more than one floor - 2 Semi-detached building single story - 3 Semi-detached multistory - 4 Single story row building - 5 Appartment - 6 07. Room/place for cooking Multi-purpose cooking space - 1 Separate kitchen - 2 LI Open space - 3 * 22 - V. CHOICE OF FUEL 01. Which fuets were used in this household during the post month: a. Electricity : yes - 1 no -2 2 b. Piped gas : yes - 1 no -2 2 c. LPG : yes - 1 no -2 2 d. Kerosene : yes - 1 no -2 2 e. Fuelwood : yes - 1 no - 2 f. Charcoal : yes - 1 no -2 2 g. Others : yes -1 no -2 2 02. If the answer of question la is code 2 (do not use any electricity). What are the reasons and the principal reason: All Principal Reasons Rsmasn Electricity supply is not yet available 1 1 The installation procedure is toc difficult 2 2 |__ |_ The installation cost is too high 4 3 Electricity is too expensive 8 4 Prefer kerosene lightning 16 5 Others (. .... ) 32 6 03. If the answer of question lb is code 2 (do not use piped gas). What are the reasons and the principal reason: All Principal Reasons Reason Piped gas is not yet available 1 1 Objection to piped gas installation work in the house 2 2 Equipment cost is too expensive 4 3 Piped gas is too expensive 8 4 L1LU |_ Safety considerations (afraid it will explode/cause fires) 16 5 Do not prefer piped gas for cooking 32 6 Others ( .....) 64 7 04. If the answer of question 1c is code 2 (do not use LPG). What are the reasons and the principal reason: All Principal Reasons Reason LPG is not yet available 1 1 The road to the house is too narrow for the LPG distributor 2 2 LPG equipment cost is too expensive 4 3 L 'l| LPG is too expensive 8 4 Safety considerations (afraid, it will explode/cause fires) 16 5 Do not prefer LPG for cooking 32 6 Others ( .... ) 64 7 * 23- 05. If the answer of question Id is code 2 (do not use kerosene). What are the reasons and the principal reason: All Principal Reasons ~Reas Its too difficult to get I 1 Equipment cost is too expensive 2 2 Kerosene is too expensive 4 3 Kerosene is too dirty 8 4 I_I_|_' lI Using kerosene is ineonvenient 16 5 Safety consideration 32 6 Others t ........... ....) 64 7 06. If the answer of question 1e is code 2 (do not use fuelwood). What are the reasons and principal reason: All Principal 8aasmna Reason It's too difficult to get 1 1 Price of fuelwood is too expensive 2 2 fuetwood is too dirty 4 3 |_|_| | Do not have special kitchen 8 4 using fuelwood is inconvenient 16 5 Others t . ... ) 32 6 07. Can one get fuelwood from your yard or garden yes l1 no -2 2_ 08. If LPG was sold in 3 kg bottles at the sane price per kg as now in the 11 kg bottles, would you buy it: yes I do not understandV 5 no - 2 stfIlt doubtful - 3 -24 - VI. EXPLANATION OF ELECTRICITY USAGE AND EQUJIPMENT 01. Did this household use electricity last month: Yes - 1 No - 2 (Go to block VII) 02. If yes, 'rom: PLN - I Battery/accu - 8 Non-PLN - 2 Others I_t_ Own generator - 4 t .. ....) -16 03. When did this household start using electricity (year) ... L 04. Electric power installed ............. VA . 05. Electric voltage : ................... Volt 06. Average electricity usage per month last year :......... Kwh LI.i.J.i l 07. Average monthly cost of electricity last year Rp .L..... 08. Is your electricity used by your household only: Yes -1 no -2 L 09. Electricity is used in this household for: Cooking - 1 Leisure appliance - 4 others Lighting - 2 Economic Activity - 8 t .) - 16 LI 10. If electricity is used for economic activity, what kind of economic activity, a. Activity: Tailor/confection - 1 food industry - 8 barber/hair dresser - 2 Accu charger -16 LU Laundry - 4 Other (. ) -32 b. The electricity usage : .......... watt L 11. If the household uses battery/accu: a. battery/accu voltage :..Volt LU... tt b. Amp-hour rating :..AH.... I _|_ c. How many days the accu/battery must be electrified/charged : ..day 1... da d. Cost per charge :. . Rupiah LLLLI 12. If the household uses their own generator a. The generator power :. watt ... w1a b. How many days it was used during the last month :.. days ... day c. The average daily usage :. . hours l_} 13. The fuels and lubricant usage during the last month for the own generator: Kind of fuels Volume Value and lubricant (litre) (Rp) (1) (2) (3) a. Benzine ___ ______ b. ADO l____|| c. Kerosene r..f 1 d. Lubricant t IIj| i Total t 14. Are they any problems in your electric se;pty Yes - 1 No - 2 (go to q. 16) | - 25 - 15. What kind of pbIMtem : The electricity is often cut I The voltage is not stable - 2 The voltage is sub-normal - 4 Other ( ... ) - 8 16. Number of lamps used for lighting in this household by capacity and length of usage: Kind of lanp Length of use used (watt) 1 1-2,9 3-5,9 6-11,9 12-23,9 24 (1) (2) (3) (4) (5) t6) (7) (Incandescent) Bulbs: 5 1 1 _1 11 1. 1-1- 1_1. 1-1 10-25 U LI L._. LU U 40-60 LI H_} 1_ Ii I i_i 70 + _1 11 1_ 1| 1_ _ (Fluorescent) Tubes: 10-19 1 _t || _1' l 20-39 1 _§ l0_ _|l l_ l 40 + 1_ _l l|_ _|} l_ _ 17. The electric kitchen utensils: Kind of electric Power Capacity Length to Equipment utensils (watt) unit use per day age (hour) (year) (1) (2) (3) (4) (5) 1. Electric stove hit I | pit I_ 2. Rice cooker L LI Lt 3. Electric Oven 1 L Lt 1_1 1_ 4. Electric Iron LLLI x X LI LI U_I LI 5. Color TV III LLI inc 1L 6. B/W TV L_._. L.. inc 7. Refrigerator 111. _[ I t LL LLI LI 8. Air Condition JJJj, J 8TUWh III J H 9. Washing Machine I.. I Kg L.._. I.. , ._. 10. Water heater 11111 tIll Lt III III LI (Storage) 11. Water heater 111 1 Lt/mt _ II 12. Electric Kettle 1111 1 Lt I_._. III LI 13. Water pump | | | | Lt/mt IU l LI 14 . ............I|l||l}'} __ _ Is5. -............_, ll III 11111 16.. ............ LL.. I I 18. If the household has a refrigerator: a. Does it have freezer: Yes -1 No -2 b. If yes, does it have special door: Yes -1 No -2 c. What kind of defrost system: Auto-defrost -1 Button-defrost -2 II Hanual defrost - 26 . 19. Other electric appliances owned by the family: Kind of other Kind of other Electric appliances Number Electric appliances Number (1) (2) (3) (4) 1. Blender L a 8. Vacuun cleaner 2. Mixer [ 9. Sewing machine II 3. Toaster jj 10. Hair dryer 4. Radio/Tape recorder J 11. Shaver 1.1 5. Sound System LI 12. Freezer LI 6. Video H 13. Others 7. Fan II 20. Does the household have electric entertainment appliances Yes - 1 No - 2 (go to q. 22) 21. Influence of household members in time of buying, type of appliances, price and usage of the electric appliances for entertainment : Influence level of Household head HusbandVwife Other family members (1) (2) (3) (4) 1. Buying time 1 2 3 0 1 2 3 0 1 2 3 H L 2. Type of appliance 1 2 3 0 1 2 3 0 1 2 3 7 L I_ 3. Price 1 2 3 0 1 2 3 0 1 2 3 LI 4. Usage 1 2 3 0 1 2 3 0 1 2 3 L I Code : Do not have husband/wife other family menbers - 0 Have a large influence - 1 Have a little influence - 2 Do not have influence - 3 22. Does the household have kitchen electric appliances (see q. 19) Yes - 2 No - 2 (go to q. 24) 23. Influence of household members in time of buying/type, price and usage of the kitchen electric appliances: Influence level of Household head Husband/wife other family members (1) (2' (3) (4) 1. Buying time 1 2 3 0 1 2 3 0 1 2 3 U U LI 2. Type of appliance 1 2 3 0 1 2 3 0 1 2 3 LI L 3. Price 1 2 3 0 1 2 3 0 1 2 3 | 4. Usage 1 2 3 0 1 2 3 0 1 2 3 U LI Ul Code : Do not have husband/wife another family mebers - 0 Have a large influence - 1 Have a little influence - 2 Do not have influence - 3 24. What has the household done to use electricity more efficiently No action - 0 Cook more carefully - 1 Use the electric appliances more carefully - 2 Buy the more efficient electric appliances - 4 Use the lighting more carefully - 8 Others (. .... ) -16 - 27 - VII. EXPLANATION OF PIPED GAS USAGE 01. Did the household use piped gas last month : Yes - I No - 2 (Go to block VIII) LI 02. When did the household obtain its connection (year) IH 03. The piped gas is used for : Cooking - I Boiling water 2 Lighting - 4 I_I_ Working -8 Others ( . ....) -16 04. Why does the household use piped gas for cooking All Principal Reasons Reason The piped gas network was already connected when the household took possession of this house 1 1 It's easier to take piped gas 2 2 Piped gas is cheaper 4 3 Cooking with piped gas is faster 8 4 t Piped gas does is not dirty 16 5 Others ( .... ) 32 6 05. Piped gas using in this household is for Using During a month aao Durina a vear ago Code ) Volume (N3) Value (Rp) Voltume (l3) Value (Rp) (1) (2) (3) (4) (5) ') Column (1) Code : Cooking - 1 Working 8 Boiling water - 2 Other ....) -16 Lighting - 4 06. Kitchen appliances that use piped gas : Kind of appliances Unit Capacity Type (1) (2) (3) (4) 1. Gas stove 11_1. 2. Oven I_I_I *) Type code for 3. Water heater (Storage) gas stove 4. Water heater (direct) HII Open 1 5. Other ( ...... ) I close/cover - 2 07. Efforts to use piped gas more efficiently by this household No action -0 Cooking more carefully - 1 Use the gas stove more efficient - 2 Use the piped gas equipment more efficient * 4 Insulated the water heater - 8 Other ( ... ) -16 -28- VIII. EXPLANATION OF LPG USAGE 01. Did the household use LPG tast month : Yes - I No - 2 (Go to block IX) LI 02. When did the household begin using LPG (year) ..... 03. The LPG is used for : Cooking I Working - 8 Boiling water - 2 Others 16 Lighting - 4 L.LI 04. The household reasons in using LPG for cooking : All Principal Reasons Reason When the household took possession of the house, the gas stove was already installed 1 1 It's easy to get LPG 2 2 LPG is cheaper than other fuels 4 3 Can cook quickly with LPG 8 4 L.L LI LPG does not make utensils dirty 16 5 others (. .... ) 32 6 05. This household buys LPG from : LPG agent/dealer - I LPG delivery - 4 Shop seLling LPG - 2 Others ( . ) - 8 L.. . 06. Size of bottle, price household pays and LPG usage Bottle Size LPG price Used for* Length of usage** (net) (code) (day) (1) (2) (3) (4) 11 Kg t_|B l wat 2 Lnt o 11 kg ____ __ __ 11 kg ____ _ _ 45 kg _____ _ __ 45 kg _____ _ _ ................................... . ... ..................... ............................ ... *)cotumn (3) code : Cooking - I * Exotanatfon of cotumn (4): Boiling water - 2 Length of usage is lighting - 4 calculated from the time working - 8 the bottle is first used others - 16 until it is empty. 07. Other kitchen appliances that use LPG: Kind of atoliances Unit Capacitv Tvze (1) (2) (3) (4) 1. Gas stove pit I_i-i LI 2. Oven litre *) Type code for 3. Water heater (Storage) litre 1111 gas stove : 4. Water heater (direct) It/mt liii Open - 1 5. other ( .......) I close/cover - 2 08. Efforts of the household to use LPG efficiently: None - O Cooking more carefully - 1 Used the gas stove more efficiently - 2 Use other fuels for cooking . 4 Insulate the water heater 8 Others ( .... ) - 16 - 29 - IX. EXPLANATION OF KEROSENE USAGE 01. Did the household use kerosene last month : Yes - 1 No - 2 (so to block X) lI 02. When did the household start using kerosene : l_|_| 03. Kerosene is used for : Cooking 1 Working - 4 LL Lighting - 2 Others C...) - 8 04. Wh, does this household use kerosene for cooking : At. Principal Reasons Reason Kerosene is easy to get 1 1 It's cheaper 2 2 Cooking with kerosene is easy 4 3 jjj l_| Kerosene doesn't dirty utensils 8 4 Others ( ..... ) 16 5 05. Where does the household buy kerosene : From the kerosene agent/dealer - 1 From the shop/warung near the house - 2 l_ From the kerosene door-to-door salesman - 4 06. What quantity of kerosene does the household commonly buy each purchase: ....... itre 07. The price per litre is : .......... rupiahs ...I..J 08. Kerosene lamps and frequency of usage: Frequency of usage Type of lamp __ ______ Everyday Sometimes Never (1) (2) (3) (4) 1. Petromax lamp U I L| 2. Semprong lamp LI L| 3. Sentir/Teplok lamp lI I lI 4. Others ( ....... ) LI 09. If the household uses kerosene lamps, who influences when to purchase, type, price and usage: Influence in Head Husband/ Other wife household members (1) (2) (3) (4) 1. Buying time 1 2 3 0 1 23 0 1 2 3 L L L 2. Type of lamp 1 2 3 0 1 23 01 2 3 LIt l _ 3. Price 1 2 3 0 1 23 01 2 3 I L LI 4. Usace 1 2 3 0 1 23 01 2 3 LI | Code : Do not have husband/wife other household members - 0 Have a large influence - I Nave some influence - 2 Do not have influence - 3 - 30 - 10. The kerosene stove by type and frequency usage : Frequency of Use Kind of stove Everyday Sacetimes Never (1) (2) (3) (4) 1. Have 20 wicks and more |_ |_ |_| 2. 2 - (19) wicks |_ |_ L 3. One fuse (Asbestos) |_ 4. Pressure |_ |_| 5 1 S. Others ( ......-.)|_ |||_ 11. If the household uses a kerosene stove, who influences the time of purchase, type, price and usage: Influence in Head Husband/wife Other household members (1) (2) (3) (4) 1. Buying time 1 23 0 1 23 0 1 2 3 | | 2. Type of stove 1 23 0 1 23 0 1 2 3 |_| LI |_| 3. Price 1 23 0 1 23 0 1 2 3 LI L 4. Usage 1 23 0 1 23 0 1 2 3 |_ L Code : Do not have husband/wife/Other family mtmbers - 0 Have large influence - 1 Have some influence - 2 Do not have influence - 3 12. How much the kerosene was used last week : ......i..... ttre/week 13. How much the kerosene is used daily : ....... ....... litre/day 14. Measured kerosene consumption over two days: (fill in this question on the second day) End use Total (litre) The average per day (litre) (1) (2) (3) 1. Cooking|_|_|||| 2. Lighting LI |_ |_| |_| 3. Working |__ Li LI5 1 LI 4. Others ( .. ) LL. LI LI. Total LLL L L 1 1 Date of first visit : ........n ......... Date of second visit : .......on :......... 15. Efforts of the household to use kerosene more efficiently None - O Cooking with kerosene more carefully - I Use the stove more efficiently - 2 Use other fuels for cooking - 4 |_|_| Use kerosene for lighting more carefully - 8 Others ( ..... ) -16 - 31 - X. EXPLANATION OF FUELWOOD OR OTHER BIOFUEL USAGE 01. Did the household conesue fuel woods or other blofuel last month: Yes - 1 No - 2 (go to blockXl) LX 02. If yes, how do you get it: Purchased - 1 Collected - 2 Purchased and collected - 3 03. If purchased, how do you buy it: Delivered to the home 1 Purchased in the market - 2 Both purchased and delivered -3 L 04. When did the household start using fuelwood (year) L.Il 05. Why does the household use fuetwood for cooking : All Principal Reasons Reason It's easy to get the fuelwood 1 1 Wood fuels are cheap/free 2 2 LLI LI-8 The kerosene stove is expensive 4 3 Food cooked with wood fuels is tastier 8 4 Fueluood gives a hotter flame 16 5 Others ( ..... ) 32 6 06. fuetwood price in this area: a. In the ralny season: per bundle- .rupiahs (One bundle . Kg) b. In dry seasons: per bundle .rupiahs LLLLI (one bundle .Kg) 07. If the wood fuels are collected, where is it from: Own land - 1 Neighbor's land - 2 Forest - 4 I_I_I The "waste" from a building project - 8 Others (.....) - 16 08. Kind of fireplace that's used in this household: Open - 1 closed - 2 t_ 09. Now miuch fuelwood did the household use last week : kg t l 10. Average daily fuelwood consuption: kg 11. Measured fuelwood use for cooking over 2 days. (filled in on second visit). a. Total of fuelwood use: ...... kg l b. Time of first visit : .at c. Time of second visit .at . 12. Efforts of the household to conserve fuelwood: None - 0 Cooking with fuelwood more carefully - 1 Using fireplace more efficiently - 2 Use other fuels for cooking - 4 Use a lower quantity of fuel - 8 Others ( .... ) - 16 - 32 - XI. EXPLANATION OF CHARCOAL USAGE 01. Did the household consume charcoal Last month: Yes -1 No -2 (go to block XII) jj 02. If yes, how do you get it: Buy it at a shop nearby - 1 Buy it from the seller 2 2 Others (.. .... ) 4 03. The charcoal price per kg :........ Rupfiahs L IJ 04. The charcoal consuwption is for Ironing - 1 Grill Sate - 2 Cooking - 4 L1 Business - 8 Others (.......... ) -16 05. Total of charcoal consumption: a. During last eek : .............. kg ILl b. During last month: .............. kg J - 33 - XII. EXPLANATION OF COOKING HABITS 01. Who cooks in this household: Head - I Husband/wife 2 Other family memter - 3 servant - 4 02. If the cook is head/husband/wife/other family memter, does she/he also work (paid) outside the home Yes - 1 No - 2 03. How many times does the household cook per day: Once - I Twice -2 L2 Three Times and more - 3 04. How much water is boiled for drinking per day : ............. litre liLt 05. Model and diameter of the main water boiling appliance : a. Model of the appliance: Flat pan - 1 Electric kettle - 4 Rice steamer - 2 Others ( ......) - 5 Non electric kettle - 3 b. The diameter: ........ ....... cm LLI 06. How much rice is cooked per day: . litre L..i 07. Model and diameter of the main rice cooking appliance: a. Model of the appliance: Flat pan - I Other C....) - 4 Rice steamer - 2 Electric Rice cooker - 3 b. The diameter: .......... cm LL 08. Model and diameter of the main vegetable cooking appliance: a. Model of the appliance : Flat pan - 1 Wok -2 L2 Others t ..... ) - 3 b. The diameter: .......... cm 09. Diameter of the main frying appliance: ...... cm 10. How does the household boil rice: Only boiled - 1 Steamed -2 L2 Steamed while boiling drinking water - 3 11. If use a kerosene stove for cooking, the stove is, made by: Factory - I Home industry -2 2 Do not know - 3 12. Upon purchasing a new kerosene stove, what characteristic are inmortant: The price is affordable - 1 The fuel use is more economical 4 4 The brand/model is better - 2 Others ........) - 5 The flame is better - 3 LI 13. Has the household changed the main cooking fuel during the past five years: Yes - 1 No - 2 (Go to O. 15) Li 34 - 14. If yes, fron which fuel to which fuel: _ Kind of fuel From To (1) (2) (3) Fuetwood I 1 Kerosene 2 2 LPG 3 3 I l I Piped Gas 4 4 Electric 5 5 Others ( ...... 6 6 a. If they changed from fuetwood to kerosene, what are the reasons: Att Reasons Principal Reason Move to another house/place I I Became diff. to get fuelwood 2 2 Became wealthy enough to buy kerosene 4 3 LII Li Fuelwood became more expensive 8 4 Others t. .... ) 16 5 b. If they changed from kerosene to fuelwood, what are the reasons: All Reasons Princihal Reason Move to another house/place I I It's easier to get fuelwood 2 2 Kerosene is more expensive 4 3 Fuetwood became cheaper 8 4 Others (. .... ) 16 S c. If they changed from kerosene to LPG, what are the reasons: All Reasons Princioal Reason Move to other house I I It's easier to get LPG 2 2 Became wealthy enough to buy LPG 4 3 Kerosene became more expensive 8 4 Others (. .... ) 16 5 d. If they changed from others fuels, what are the reasons: All Reasons Princial Reason Moved to other house 1 1 Change of fuels stocks/avail. 2 2 Wealth changed 4 3 Price change 8 4 Others 16 5 15. without regard to price or availability, which fuel do you think is the best fuel for cooking: Fuelwood 1 Kerosene 2 LPG -3 LI Piped gas - 4 Other ( .... ) -5 16. What efforts has the household done to use energy more efficiently: None .0 Prepare food before turning on the stove - I Turn off the stove after the water has boiled - 2 I...LI Make the burner more efficient - 4 Use a multipurpose pan -8 Keep the pan and stove clean -16 Use a cover when cooking -32 Use an insulated box -64 - 35 - ATTACHMENT : A2 Ouestionnaire UHESS - CO for Block Information 1988 Urban Household Energy Survey CONFIDENTIAI 1_|_ 5jj 1. IDENTIFICATION OF LOCATION 01 Province 6_1_1 02 Regency/Nunicipelity 81.ji_ 03 Sub District 101t11 04 Village 12i_1__1 05 Region URBAN 15|_| 06 City Classification 1980 161_1 07 Enumeration District Number 08 Block Census Number 171 1 1 1 09 SUSENAS Sample Code Nuder 20_1_1_1 10 Series Nuwber of households sample 231_|_1 11 Series Numer of household selected It. EXPLANATION OF ENUMERATION 01 Name of Enmerator 02 Date of Enmueration 03 Enumerator Signature 04 Name of Supervisor 05 Date of Supervision 06 Supervisor signature 111. CENSUS BLOCK INFORNATION 01. Number of households of this block base on SSN 87 - S2 sample : ........... household l_| 02. Number of mtched household with SSN 87 - S2 v.......ol....hs.. ehol_d| 03. Number of unmatched household with SSM 87 - S2 : household hold 04. Percent of households with own electricity installation (customers) in this block : . ......... % L l 05. Average market price of a litre of kerosene in this block: Rp.. |___ 06. Is there an LPG agent in this block/easy to get LPG (by LPG delivery vehicle): Yes - 1 No - 2 Difficult -3 |_| If "yes", what is the average market price of LPG for the 11 kg size (charge only): Rp..| What is the price of the 11 kg bottle: Rp ........... 07. Is it easy enough to get firewood in this block: Yes - 1 No - 2 Difficult -3 lI If "yes", what i8 the average market prfce for a kg : Rp ..... - 36 - Annex 3: UHESS Summafy Tables Java Urban Households 1988 - 37 . SAMPLE CHARACTERISTICS Table 1: Nutber of Respondents by City Size and Household Expnditure Level Level of Income/Expenditure C000 Rp/month) city Size Missins 4 75 75-120 120-185 185-295 m 295 All Respondents 1. 1 " 118 142 221 328 237 150 1196 2. 500 T -I 66 106 86 73 50 30 411 3. 200 T 500 T 18 69 62 38 26 13 226 4. 100 T 200 T 17 65 37 24 19 9 171 5. 50 T 100 T 5 27 42 30 1S 13 128 6. 20 T 50 T 19 135 82 51 20 21 328 7. 10 T 20 T 16 110 51 33 19 13 242 All Respondents 259 654 581 577 382 249 2702 Table 2: Percentage Distribution of Sampled Households by Household Expenditure Level and Urban Area Size Level of Income/Expenditure ('000 Rp/month) City Size Missing < 75 75-120 120-185 185-295 > 295 All Respondents 1. >1M 4% 5% 8% 12% 9% 6% 44% 2. 500 T -I N 2% 4% 3% 3% 2% 1% 15% 3. 200 T -500 T 1% 3% 2% 1% 1% 0% 8% 4. 100 T -200 T 1% 2% 1% 1% 1% 0% 6% 5. 50 T -100 T 0% 1% 2% 1% OX 0% 5% 6. 20 T - 50 T 1% 5% 3% 2% 1% 1% 12% 7. 10 T - 20 T 1% 4% 2% 1% 1% 0% 9% All Respondents 10% 24% 22% 21% 14% 9% 100% Table 3: Household Size and Average Expenditure by Expenditure Group Expenditure per Household - '000 Rupiah per Month Nissing <75 75 - 120 120-185 185-295 >295 All Ave. Household Size 4.2 3.5 4.8 5.1 5.7 6.1 4.7 Ave. Exp./Hsehold NA 51 96 149 228 478 156 - 38 - Tabts 4: Percentage of Households by Fuel-Use Cobirnation for Each Expenditure Group Expenditure per Household - 000 Rupiah per Nonth Missing 475 75 - 120 120-185 185-29S 295 All Electricity 86X 65X 85X 9SX 9TX 100X 85X Cooking 1X 0X 1X 3X 3X 14X 2X Lighting 85X 63X 83X 94X 96X 98X 84U TV 40X 22X 47X 67X 73n 894 52X Ironing 37X 16X 34K 57X 68X 80K 44K Refrigerator 7X 1X 3X 9X 21K 46K 11X Air Cond. 0X 0X 0X 0X 1K 2X 0X Pumping 6a 1K 3K 7X 16X 35K 9X washin N. 0X 0X 1K OK 1K 11K 1X "Sell" 10X 10X 13X 17X 15X 6a 12X Kerosene 81K 87X 87X 92X 91K 78K 87X Cooking 73X 53K 78X 894 90K nx 75X Lighting 17X 39X 18X 7X 5 2X 17X Econ.Act 2X 3K 2X 2X 3K 2X 2X Other 0X 2X 1K 0X 0X 1S 1K L P G 2X 0X 2X 4X 8X 29X 5X Cooking 2X 0X 2X 4K 8X 294 5K W.Heater 0X 0X 0X 0X 0X 4K oX siofuel 14K 46X 26K 11X 8X 7X 22X Charcoat 17X 33X 39X 24K 16K 12X 27X Cooking 2X 4X 2X 2X 3K 3K 3K Ironing 15X 30X 36K 22X 13K 8X 24K Other 1K 2X 2X 1K 1K 2X 1K - 39 - Iable 5: Average Household Fuel Contuiptfon by Fuel-Use Combination for Users In Expenditure Group (original units) Expenditure Group ('000 Rupiah per Nonth per Household) Nissing <75 75 - 120 120-185 185-295 >295 All Electricity 47.4 25.7 38.7 58.4 88.5 172.6 64.2 (kWhlno) Cooking 8.8 4.5 7.1 6.8 12.5 20.3 14.7 Lighting 25.4 21.3 26.8 32.7 38.8 60.2 32.6 TV 8.4 6.1 7.3 9.0 10.3 12.7 9.2 Ironing 6.9 6.2 7.0 7.9 9.6 11.0 8.4 Refrigerator 54.4 36.9 47.3 47.6 44.5 52.5 48.9 Air Cond. 0.0 0.0 0.0 0.0 97.9 272.3 222.5 Pumping 16.7 17.4 23.2 26.8 30.0 42.2 32.3 Washing M. 0.0 0.0 8.0 4.5 7.7 23.8 20.0 "Sales" 22.1 17.6 22.3 28.0 34.1 38.5 26.1 Kerosene 1.1 0.9 1.1 1.3 1.4 1.6 1.2 (tt/day) Cooking 1.0 0.9 1.0 1.2 1.3 1.4 1.1 Lighting 0.5 0.4 0.6 0.5 0.6 0.6 0.5 Econ.Act 2.5 2.6 1.9 3.9 1.8 6.5 2.8 Other 0.3 0.2 0.2 0.1 0.1 0.8 0.2 L P G 0.7 0.6 0.7 0.7 0.8 0.8 0.8 (kg/day) Cooking 0.7 0.6 0.7 0.7 0.8 0.8 0.8 U.Heater 0.0 0.0 0.0 0.0 0.0 1.4 1.4 Biofuel 3.8 4.5 4.3 4.1 4.5 4.4 4.4 (kg/day) Charcoal 0.3 0.2 0.2 0.2 0.3 0.2 0.2 (kg/day) Cooking 0.7 0.6 0.6 0.8 0.5 0.6 0.6 Ironing 0.1 0.1 0.1 0.1 0.2 0.1 0.1 Other 1.2 0.7 1.2 0.6 0.5 0.9 0.8 -40 - abte6: Average Household Fuel Consumption by Fuel-Use Coabination for Att Households in Expenditure Group Expenditure per Month per Household ('000 Rp) Missing 75 75 - 120 120-185 185-295 >m95 All Electricity 40.6 16.6 32.9 55.6 85.7 171.9 54.8 (kWh/month) Cooking 0.1 0.0 0.0 0.2 0.3 2.8 0.4 Lighting 21.6 13.5 22.3 30.8 37.2 59.2 27.4 TV 3.3 1.3 3.4 6.0 7.5 11.3 4.8 Ironing 2.6 1.0 2.4 4.5 6.5 8.7 3.7 Refrigerator 4.0 0.3 1.5 4.1 9.4 24.0 5.2 Air Cond. 0.0 0.0 0.0 0.0 0.5 5.5 0.6 Puiping 1.0 0.2 0.6 1.9 4.7 14.9 2.7 Washing N. 0.0 0.0 0.0 0.0 0.1 2.6 0.3 "Salesn 2.1 1.7 2.9 4.7 5.0 2.5 3.2 Residual 5.8 -1.5 -0.4 3.3 14.5 40.4 6.5 Kerosene 0.8 1.1 1.0 1.2 1.3 1.2 1.0 (tit/day) Cooking 0.7 0.5 0.8 1.0 1.2 1.1 0.8 Lighting 0.1 0.2 0.1 0.0 0.0 0.0 0.1 Econ.Act 0.0 0.1 0.0 0.1 0.0 0.1 0.1 Other 0.0 0.0 0.0 0.0 0.0 0.0 0.0 L P 0 0.0 0.0 0.0 0.0 0.1 0.2 0.0 (kg/day) Cooking 0.0 0.0 0.0 0.0 0.1 0.2 0.0 W.Heater 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Biofuel 0.5 2.1 1.1 0.5 0.4 0.3 1.0 (kg/day) Charcoal 0.0 0.1 0.1 0.1 0.0 0.0 0.1 (kg/day) Cooking 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Ironing 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Other 0.0 0.0 0.0 0.0 0.0 0.0 0.0 - 41. Table 7: Percentage of Households in Expenditure/lncome Category Having Device Mentioned Level of Income/Expenditure ('000 Rp./mo.) Missing 75 75-120 120-185 185-295 > 295 Total Kerosene (Petromax) 8a 15X 8X 6X 4X 4X 8a (Wick) 19X 401 201 131 71 71 201 (Hurricane) 141 251 151 81 81 7% 14X - Stoves Single Wick 5S 1X 31 61 81 18x 6X 2-19 wicks 641 461 63X 78X 75X 58s 641 20 + wicks 81 11X 171 141 161 161 14% Pressure 0 01 1X 1X 1% 0 0 LPG -Cooking 2X 0 21 4X 8a 271 5 - W. Heating 0 01 0 0 01 21 0 - 42- Tsble 8: Percent of Households Using Specified Fuel Source by Expenditure Group LEVEL OF INCOME / EXPENDITURE ('000 Rp./mo) missing * 75 75-120 120-185 185-295 > 295 Total Electricity Percent Using - 86% 65% 85% 96% 97% 100% 85X Source Breakdown: PLN 55% 50% 66% 82% 90% 94% 73% PLU Indirect 43% 47% 32% 16% 9% 5% 26% Non PLN 0% 2% 0% 1% 1% 1% 1% Battery 1% 2% 2% 1% 1% 0% 1% 100% 100% 100% 100% 100% 100% 100% Kerosene Percent Using - 81% 85% 86% 91% 90% 78% 86% Source Breakdown: Delivery 22% 7% 10% 23% 29% 29% 16% Shop 71% 91% 85% 68% 65% 43% 65% Agent 6% 2% 5% 10X 6% 6% 5% 100% 100% 100% 100% 100% 100% 100% LPG Percent Using - 2% 1% 3% 5% 9% 28% 6% Source Breakdown: Delivery 50% 0% 28% 36% 53% 46% 43% Shop/Agent 25% 100% 67% 50% 33% 43% 46% Both 25% 0% 6% 14% 14% 10% 11% 100% 100% 100% 100% 100% 100% 100% Wood Percent using - 14% 47% 26% 11% 8% 7% 23% Source Breakdown: Purchase 31% 22% 43% 42% 19% 44% 31% Collect 44% 62% 39% 41% 68% 44% 35% Both 25% 15% 18% 17% 13% 11% 16% 100% 100% 100% 100% 100% 100% 100% - 43. Summary Tables - Urban Area Size and Energy Table 9: Percentage of Households fn Urban Area Type by Fuel-Use Combination Urban Area Size Small Medium Large V. Large All Electricity 67X 75X 89X 95f 85X Cooking 1X 2X 2X 4X 2X Lighting 63X 74X 88X 94K 84X TV 36X 42X 52X 62K 52X Ironing 25X 34X 43X 56X 44X Refrigerator 3K 5X 7K 18X 11K Air Condition OK OK OK 1X OK Pumping 3X 3X 5K 15K 9K Wash Machfne OX OX 1X 2X 1K Econ. Ac 2X 3X 6K 8X 6X uSell Elec. 5K 9X 9K 18X 12X Kerosene 82X 91K 79X 93X 87X Cooking 49K 76K 69K 91X 75K Lighting 40X 29X 14X 6K 17X other 1K 4X 1X OX 1X LPG 1X 3X 4X 8X 5X Cooking 1K 3K 4K 8X 5S Water Heater OK OX OX OX OX Biofuel 55 30K 25X 3X 22X Charcosl 33X 48X 32X 17X 27X Cooking 4K 2X 7X OX 3X Ironing 29X 36K 26X 17X 24K Other AC 1X 3X 2X 1K 1X -44 - Table lO: Average Household Fuel Consumption by Fuel-Use Coabination for Users in Urban Area Size Category Urban Area Size SntIl Medium Large V Large All Electricity-kWh/mo 35.2 45.2 46.0 86.8 64.2 Cooking 4.7 13.7 9.1 17.5 14.7 Lightino 29.6 32.8 30.7 34.5 32.6 TV 7.0 8.2 7.5 10.6 9.2 Ironing 6.8 8.5 6.4 9.6 8.4 Refrigerator 43.4 28.0 41.1 52.3 48.4 Air Conditioning 0.0 0.0 93.8 243.9 222.5 Pumping 23.8 28.5 25.0 34.4 32.3 Wash Machine 13.3 10.3 4.8 23.0 20.0 "Sold" 25.5 19.6 24.3 24.3 27.1 Kerosene-It/day 1.0 1.1 1.1 1.3 1.2 Cooking 1.1 1.1 1.0 1.2 1.1 Lighting 0.5 0.4 0.5 0.6 0.5 Business 0.1 0.2 0.2 0.9 0.2 LPG-kg/day 0.6 1.0 0.9 0.8 0.8 Cooking 0.6 0.9 0.8 0.7 0.8 Water Heater 0.0 0.5 4.1 1.4 1.4 Biofuet-kg/day 5.0 4.2 3.6 3.3 4.4 Charcoal kg/day 0.2 0.2 0.3 0.1 0.2 Cooking 0.6 0.6 0.6 0.4 0.6 Ironing 0.1 0.1 0.1 0.1 0.1 Other 1.2 0.8 0.8 0.4 0.8 . 45 - Tabte II: Average Household Fuel Consumption by Fuel-Use C.oubInation for Alt Households in Urban Area Size Category Urban Area Sfze Smtll edium Large Very Large Total Electricity-kWh/mo 23.5 34.0 40.8 82.6 54.9 Cooking 0.0 0.2 0.2 0.6 0.4 Lighting 18.7 24.2 27.1 32.6 27.4 TV 2.5 3.5 3.9 6.6 4.8 Ironing 1.7 2.9 2.7 5.3 3.7 Refrigerator 1.3 1.3 2.9 9.3 5.2 Air Conditioning 0.0 0.0 0.1 1.2 0.6 Pumping 0.6 1.0 1.1 5.1 2.7 Wash Machine 0.0 0.0 0.0 0.5 0.3 Sold 1.4 1.7 2.2 4.5 3.4 Residual -2.9 -0.8 0.5 17.3 6.7 Kerosene-It/day 0.8 1.0 0.9 1.2 1.0 Cooking 0.5 0.8 0.7 1.1 0.8 Lighting 0.2 0.1 0.1 0.0 0.1 eusiness 0.1 0.1 0.1 0.1 0.1 LPG-kg/day 0.0 0.0 0.0 0.1 0.0 Cooking 0.0 0.0 0.0 0.1 0.0 Water Heater 0.0 0.0 0.0 0.0 0.0 Biofuel-kg/day 2.7 1.3 0.9 0.1 1.0 Charcoal kg/day 0.1 0.1 0.1 0.0 0.1 Cooking 0.0 0.0 0.0 0.0 0.0 Ironing 0.0 0.0 0.0 0.0 0.0 Other 0.0 0.0 0.0 0.0 0.0 - 46 - tgble 12: Average Household Expenditures on Fuel-Use for All Households in Urban Area Size Category RVPIAH/HouseholtdMonth Urban Area Size Small Mediun Large V. Large All Electricity 2,471 3,260 3,976 8,296 5,509 Cooking 2 22 23 72 40 Lighting 2,029 2,393 2,864 3,298 2,828 TV 276 353 410 64S 480 Ironing 168 261 298 502 357 Refrigerator 126 125 283 953 529 Air Conditioning 0 0 26 77 0 Pumping 60 96 119 478 261 Wash Machine 5 3 3 58 28 Residual (194) 7 48 2,151 917 Kerosene 4,767 5,943 5,182 7,277 6,106 Cooking 3,240 4,929 4,269 6,625 5,168 Lighting 1,112 792 432 219 521 Business 415 223 480 433 417 LPG 161 593 648 1,084 732 Cooking 161 562 589 1,043 696 water Heater 0 0 59 41 36 Biofuel 3,165 1,630 1,405 349 1,325 Charcoal 458 486 489 189 349 Cooking 106 55 195 2 75 ironing 257 277 194 162 202 Other 87 147 89 14 62 All Energy 11,022 11,912 11,700 17,195 14,021 Total Expenditures 106,627 137,731 141,919 192,488 156,012 (energy/total) (10.3X) (8.61) (8.21) (8.9%) (9.01) .47 - tatle 13: Perecnt of Households Using Specified fuel Source by Urbcn Area Size Urban Area Size Sath Nediuw Large V.Large Att Electriecity Percent using: 67X 75S 89X ffX 85X MMrc Brefkdomin PLN 68K 68X 65K 79X 73X Non PLN 3X 0X oX 1X 1X Battery 4K 1X oX oX 1X PLN Indirect 25X 31K 34K 20K 25K Total 10OX 100 100K 100K 100 Kerosene Percent using 84K 88K 86% 87K 86X Source Breakdown: Agent 7X 5X 4K 7% 6a Shop 86S 89X 81% 59X 73% Delivery 5K 4X 5K 34X 18X coebinations 3K 2% 10K 0X 3X Total 100 100% 100K 100 100 LPG Percent using 2X 4K 5X 9X 6S Source Breakdown. Delivery 15X 8K 52X 48X 43X Shop 77K 42X 45K 43X 46K 3oth 8X 50X 3X 10X 11K Total 1OOX 100K 100X 100X 100K Wood Percent using 55K 30X 26X 3X 23K Source Breekdom: Purchase 28X UX 34K 7K 31X Collect 58X 36S 46K 80K 53K Purchase & Col 14K 20K 20X 13X 16X Total tOOK lOOK 100X ioOn 100X Breakdown of Collectiln Sournes: O,n Land 44 48X 50X 11K 43X Others Land 13X 10X 11 3X 11 Forest 11K 0X oX 0X 6X Const.Projec 1X 0X 4X 22X 4X Others 3X 16K 10K 11 7X Cobfinations 2 26K 25K 54K 29X Total 1008 100K IonK 100K 100l - 48 - Miscellaneous Tables table 14: Household Iubner with Major Inftuence on Purchase and Use of Device Nentioned Electric Leisure App. Kerosene Lamp Kerosene Stove Type Use Type Use Type Use Pr. Female 16% 28% 53% 54% 84% 86% Pr. Kale 68% 44% 44% 36% 12X 6% Other 16% 29X 4% 10% 5% 8% Total 10GX 100% 100% 100% 100% 100% Tbtle 15: Reasons Non-Electrified Households Do Not Use Electricity Expenditure per Household ('000 Rp) All City Size Missing 4 75 75-120 120-185 185-295 > 295 Respondents 1. No Elec. Grid 14% 38% 9% 8% 0% 50% 26% 2. Inst. Procedure 3% 6% 10% 4% 8% 0% 6% 3. Inst. Payment 43% 30% 43% 52% 33% 50% 36X 4. Expensive Bill 22% 6% 7% 12% 8% 0% 8% 5. Taste 3% 4% 1% 0% 0% OX 3% 6. Other 15% 16% 31% 24% 51% 0% 21% Total Non Users 100% 100% 100% 100% 100% 100% 100% (No of non-users) (32) (230) (88) (25) (12) (2) (394) Table 16: Percentages of Households Electrified by mInco e and Urban Area Size Level of Household Expenditure ('000 Rp/month) All Urban Area Size MissIng < 75 75-120 120-185 185-295 > 295 Exp. Levels 1. > IN 93% 88% 90% 97% 99X 100% 95% 2. 500 T - I M 91% 80% 93% 97% 98% 100% 91% 3. 200 T - 500 T 67% 77% 84% 95% 92% 100% 84% 4. 100 T - 200 T 71% 58% 76% 83% 79% 89% 71% 5. 50T - 100 T 80% 70% 71% 93% 91% 100% 81% 6. 20 T - 50 T 74% 37% 77% 88% 95% 100% 65% 7. < 20 T 63% 49% 78% 100% 95% 100% 69% All Urban Areas 86% 65% 85% 96% 97% 100% 85% TabLe 17: Pattern of Ownership and Use of Selected Electric AppLiances (Fraction Owning in Expenditure Group and Monthly Electric Consumption for Users) Household Expenditure Level ('000 Rp/Month) Alt Electric Missing < 75 75 - 120 120 - 185 185 - 295 > 295 Respondents Appliances % kWh/month X kWh/month % kWhtmonth % kWh/month % kWh/month % kWh/month % kWh/month Bulb 85% 25.36 63% 21.42 83% 26.77 94% 32.74 96% 38.82 98% 60.21 84% 32.30 Range 0% 0.00 0% 0.00 0% 3.00 0% 0.00 0% 0.00 0% 28.20 0% 24.00 Rice Cooker 1% 8.75 ox 4.50 0% 9.30 2% 6.79 3% 12.54 13% 13.80 2% 11.51 Iron 37% 6.95 16% 6.20 34% 7.04 57% 7.85 68% 9.57 80% 10.96 44% 8.39 Color TV 15% 11.66 3% 8.88 9% 12.18 21% 14.38 34% 14.45 66% 14.72 19% 13.87 B & W TV 25% 6.51 19% 5.62 38% 6.14 47% 6.54 39% 6.47 24% 6.71 33% 6.31 Refrigerator 7% 54.40 1% 36.92 3% 47.29 9% 47.58 21% 44.51 46% 52.48 11% 48.86 Air Conditioner 0% 0.00 0% 0.00 0% 0.00 0% 0.00 1% 97.92 2% 272.34 0% 222.51 Washing Machinm 0% 0.00 0% 0.00 1% 8.00 0% 5.00 1% 8.00 11% 23.85 0% 20.14 Water Heater 0% 0.00 0% 0.00 0% 0.00 0% 0.00 0% 0.00 1% 67.67 0% 67.67 Electric Kettle 0% 0.00 0% 0.00 0% 0.00 0% 0.00 0% 0.00 1% 43.00 0% 30.33 - 50 ^ Table 18: Ownership of Other Electrical Appliances (Percentage of Respondents in Expenditure Group Owning Specified Appliance) Household Expenditure ('000 Rp/Month) All Missing < 75 75-120 120-185 185-295 > 295 Respondents 1. Radio/Tape 39.4% 20.8% 42.0% 56.8% 61.8% 78.3% 45.9% 2. Fan 15.8% 2.3% 10.3% 22.2% 30.1% 45.8% 17.5% 3. Water Pump 5.8% 1.4% 2.6% 7.3% 15.4% 34.9% 8.4% 4. Mixer 5.0% 0.8% 3.3% 8.1% 14.1% 28.5% 7.7% 5. Video 1.2% 0.9% 1.9% 3.6% 10.5% 25.7% 5.4% 6. Sound System 3.1% 0.3% 1.7% 1.9% 5.0% 14.5% 3.2% 7. Sewing Machine 1.9% 1.1% 2.6% 3.6% 3.4% 10.4% 3.2% 8. Hair Dryer 1.2% 0.2% 1.0% 1.9% 6.3% 16.1% 3.1% 9. Blender 1.2% 0.3% 0.5% 2.8% 5.5% 14.9% 3.0% 10. Razor 0.8% 0.0% 0.9% 0.3% 0.5% 5.6% 0.9% 11. Toaster 0.0% 0.2% 0.7% 0.5% 1.0% 2.4% 0.7% 12. Vacuum Cleaner 0.4% 0.0% 0.2% 0.5% 0.3% 5.2% 0.7% Table 19: Sumary Electricity Data by Source of Electricity Non PLN PLN Total All Elec Indirect Direct Elec Households 1. Sanple Size 393 589 1676 2309 2702 2. Share of NH In Survey (%) 15% 22% 62% 86% 100% 3. Share of Connex in Survey(X) 100X 100% 100% 4. Family Size 4.1 3.8 5.2 4.7 4.6 5. Average Consump. (kWh/No) 0 15 84 64 55 6. Number "selling" electricity NA 335 335 335 7. Estimated sales/seller (kWh/Mo) NA 26 26 26 8. Average net Cons. (kWh/No) 0 15 79 60 52 9. Fam. Expenditure (Rp/Mo) 69,820 103,285 193,844 156,012 143,476 10. Rp/Mo for Electricity 0 2,360 8,098 6,447 5,509 S1. IIabtAe : Bulbs per Thousand Households by Type and Use-Lewl (Includes non-electrif led households) <1 HR 14HR 340W 1 10 24 20 9 0 64 TOTAL 40 232 1293 1751 1158 102 4575 Tabte 21: Distribution of Electricity Use (kWh) for Lighting by Sutb Type and Use-Level d1 HR 1HR<3 370W 0.00% 0.02% 0.31% 0.61% 0.52% 0.00% 1.46% Fluorescent 10-19W 0.00% 0.10% 2.07% 4.74% 3.98% 0.54% 11.44% 20-39W 0.00% 0.2,;% 4.12% 7.45% 4.73% 0.52% 17.11% >40W 0.00% 0.17% 0.93% 1.64% 1.11% 0.00% 3.85% Total 0.06X 1.39% 17.17% 41.67% 35.34% 4.36% 100.00% S52- Table 22t Averag Nouhbold Cooking Fuel Use and Nix by Urban Am Sie and penditure Group Urban Size Samll lNdiuw Lare V Large Alt (Liters of Kerosene Euivalent per Household per Day) Expenditure Very LoW LKE W 1.11 0.84 0.89 0.93 0.98 % Kerosene 21% 54% 37X 94% 49% % LPG OX OX OX 1% OX % Wood 78X 46X 43X a 51% Low IME 1.14 1.19 1.00 1.11 1.10 % Kerosene 54% 66X 69X 93% 74% % LPG 0% 1% 2X 3% 2% % Wood 46% 32% 30% 4% 24% Noderate LKE 1.32 1.30 1.09 1.17 1.20 % Kerosene 68% 85% 82% 94% 87% X LPG 1% 1% 5% 5% 4% % Wood 31% 14% 13X 1% 9% High LKE 1.36 1.50 1.21 1.42 1.38 % KIerosene 71% 65% 77X 92% 85% % LPG 3% 13% 18 6% 9% % Wood 25% 22% 5% 1% 6a very High LKE 1.62 1.92 1.31 1.52 1.53 % Kerosene 64% 77% 83% 70X 72X % LPG 8% 18X 15X 30% 23% % lood 28% 5% 2% 1% 5% All LKE 1.20 1.18 1.02 1.22 1.18 % Kerosene 4% 69% 71% 89% 73% % LPG 1% 5% a 9% 6% % Wood 54% 27% 23X 2% 21% A/ Units are liters of kerosene equivalent (LIE) per household per day calculated on the basis of the cooking fuel substitution ratios provided at the end of the Tables. A consuption level of one liter of kerosene equivalent met by, for example LPG, implies that were the household to use kerosne, one liter of kerosene would be used. The substitution ratio depends on both power and efficiency of use. The substitutions ratios were calculated statistically on the basis of the UHES survey results. The LPG-kerosene substitution ratio conform to expectations based on technical characteristics of the kerowene and LPG stoves tested in the laboratory for the UHESS: part of the efficiency advantage of LPG is lost because of the hfgher power of LPG stoves. Alternatively, the fuelwood substitution ratilo can be taken to sugest that wood is used more efficiently or at lower power in urban than In rural areas. jM: Households for .ihich expenditure data are missing are not lncluded. * 53 - Tkale 23: Main Reasons for Use/Non-Use of Cooking Fuel Main Reasons for KEROSENE NON USE Too difficutt to obtain 3 1X Equipment too expensive 44 12X Kerosene too expensive 75 21X Kerosene too dirty 51 14X Kerosene too time conwming 54 15X Safety consideration 36 10X Other 90 251 Totat Non-Users 353 1001 Main reasons for USING KEROSENE for cooking Easy to get kerosene 1004 48X Kerosene is cheap 272 13X Cooking with kerosene is easy 643 311 Kerosene Is clean 115 61 Other 55 31 Total Users 2089 1001 Nain reasons for FUELWOOD NON-USE Too difficult to get 407 191 wood too expensive 42 21 Wood is too dirty 582 281 Don't have proper kitchen 316 151 Wood is too time consuing 672 321 Other 79 41 Totat Non-Users 2098 1001 Main reasons for USING FUELWCOD for cooking Easy to get fuelwood 252 411 Cheap/free 223 371 Kerosene stove is expensive 17 31 Wood cooked fuel nicer 48 81 Wood gives hotter ftlame 46 8X Other 22 41 Totat Users 608 1001 Main Reasons for LPG NON-USE There is no LPG distributor 410 161 Road too narrow for dist. 55 2X LPG equipment too expensive 883 341 LPG too expensive 596 231 Safety considerations 277 11X Don't tike LPG for cooking 101 41 Other reasons 242 91 Total Non-Users 2564 1001 Main Reasons for LPG use for cooking Stove came with house 3 2X Easy to get LPG 18 131 LPG is chuap 20 141 LPG cooks quickly 71 51X LPG is clean 26 191 Total Users 138 1001 .54. Annmw: SUSENAS Tables Tabte 1: Urbeo Household Characteristics by Province and Year Number of Household Nousehold Nousehoids Size Expenditure (Rp/Io) 1987 Sumatra 1,314,938 5.4 167,700 Java 6,319,322 4.7 150,234 Musa Tanggara 264,972 4.8 129,761 Kalimantan 367,883 5.0 191,680 Sulawesi 442,818 5.0 133,889 Hat. & Irian 92,948 5.5 203,877 Indonesia 8,802,883 4.9 153,703 1984 Sumatra 1,196,064 5.4 130,863 Java 5,460,629 4.8 120,465 Musa Tanggare 224,387 4.9 102,098 Kalimantan 323,736 5.2 152,732 Sulawesi 395,265 5.2 120,765 Mal. & Irian 54,414 5.7 169,238 Indonesia 7,654,496 5.0 123,278 1981 Sumatra 980,475 5.7 94,911 Java 4,736,521 5.0 88,685 Musa Tanggara 182,641 5.3 76,479 Kalimantan 285,074 5.2 100.416 Sulawesi 313,863 5.4 89,097 Nal. & Irian 50,921 5.3 118,323 Indonesia 6,549,495 5.2 90,038 55 - Table 2: Averao Fuel Use of Using Urban Households by Province and Year Electricity LPG Kerosene Charcosl Wood kMh/mo kg/mo lt/mo kg/mo kg/mo 1987 Sumtra MA 18 25 4.1 140 Java MA 18 29 3.4 140 Nusa Tanggara NA 15 17 3.4 140 Kalimantan MA 17 22 7.6 140 Sulawesi MA 16 22 2.2 140 Mal. & Irian MA 0 26 2.1 140 Indonesia MA 18 28 3.6 140 1984 Sumatra MA 22 30 3.3 140 Java MA 19 35 2.9 140 Nusa Tanggara MA 0 19 2.7 140 Kalimantan MA 0 24 3.3 140 Sulaweui MA 20 30 5.0 140 Nal. & Irian MA 0 43 3.0 140 Indonesia MA 19 33 3.0 140 1981 Sumatra MA 17 34 3.0 140 Java MA 16 42 2.9 140 Nusa Tanggara MA 26 24 3.0 140 KaLimantan MA 18 28 3.2 140 Sulawesi MA 23 35 3.8 140 Hat. & Irian MA 0 48 3.1 140 Indonesia NA 16 40 2.9 140 -56 - tablet3: Total Fuel Us. of Urban Households by Province and Year Etectricity LPG Kerosene Charcoal Wood kWh/mo kg/mo lt/mo kg/mo kg/mo 1987 Sumatra NA 634 30,975 1,415 44,579 Java MA 5,934 172,510 3,452 197.741 Nusa Tangp ra NA 62 3,840 90 16,936 Kalimantan MA 53 7,960 126 17,071 Sulawesi NA 417 8,871 23 20,568 Mal. & Irian NA 0 2,338 16 4,623 Indonesia NA 7,099 226,494 5,123 301,518 1984 Sumatra NA 196 34,232 869 28,160 Java MA 1,867 181,619 2,774 151,924 Nusa Tan3gara NA 0 3,985 56 14,569 Kalimantan MA 0 7,569 19 11,485 Sutawesi MA 65 11,091 48 14,429 Nal. & Irian MA 0 2,285 23 1,949 Indonesia NA 2,128 240,781 3,788 222,516 1981 Sunatra NA 81 32,217 954 25,809 Java NA 955 197,763 4,196 121,759 Nusa Tanggara NA 9 4,316 62 11,585 Kalimantan NA r 7,805 41 12,529 Sulavesi NA 24 10,711 41 12,516 Mal. & Irian NA 0 2,409 20 1,244 Indomesia NA 1,075 255,221 5,314 185,442 - 57- Table 4: Percentage of Urban Households Using Fuel by Province and Year Electricity LPG Kerosene Charcoal Wood 1987 Sumatra 692 32 942 26X 242 Java 762 5 932 16X 22X Nusa Tanggara 64X 22 86X 102 462 Kalimantan 692 12 972 42 332 Sulawesi 752 62 902 22 332 Mal. & Irian 572 02 96X ax 362 Indonesia 742 42 932 162 24X 1984 Sumatra 562 12 97? 222 172 Java 622 2X 96% 182 202 Nusa Tanggara 472 02 952 9X 462 Kalimantan 682 02 992 22 252 Sulawesi 632 12 942 22 262 Mat. & Irian 54X 02 972 122 262 Indonesia 61X 12 962 162 212 1981 Sumatra 452 02 982 322 192 Java 472 1% 982 312 182 Nusa Tanggara 362 OX 972 112 452 Kalimantan 572 02 962 52 312 Sulawesi 462 02 97X 32 282 Cal. & Irian 522 OX 982 132 17? Indonesia 47X 12 98X 282 202 - 58 - Table 5: Average Fuel Use of Urban Households by Province end Year Electricity LPG Kerosene Charcoal Wood kWh/mo kg/mo lt/mo kg/mo kg/mo 1987 Sunatra NA 0.48 24 1.1 34 Java NA 0.94 27 0.5 31 Nusa Tanggara NA 0.23 14 0.3 64 Kalimantan NA 0.14 22 0.3 46 Sulatwesi MA 0.94 20 0.1 46 Mal. & Iruan MA 0.00 25 0.2 50 Indonesia NA 0.81 26 0.6 34 1984 Sumatra NA 0.16 29 0.7 24 Java NA 0.34 33 0.5 28 Nusa Tanggara NA 0.00 18 0.2 65 Kalimantan NA 0.00 23 0.1 35 Sulaiesi NA 0.16 28 0.1 37 Mal. & Irian NA 0.00 42 0.4 36 Indonesia NA 0.28 31 0.5 29 1981 Sunatra NA 0.08 33 1.0 26 Java NA 0.20 42 0.9 26 Nusa Tanggara NA 0.05 24 0.3 63 Katimantan NA 0.02 27 0.1 44 Sulawesi NA 0.08 34 0.1 40 Nat. & irNan NA 0.00 47 0.4 24 Indonesia NA 0.16 39 0.8 28 .59 - tabte 6: Urban Household Characteristics by Expenditure Group and Year Number of Housdold Household Households Size Expenditure Rp/No Expenditure 1987 Very Low 1,819,674 3.0 48,193 Low 2,275,458 4.4 88,257 Moderate 2,123,929 5.2 136,068 High 1,554,686 6.0 209,472 Very High 1,029,136 6.6 473,282 All 8,802,884 4.9 157,932 1984 Very Low 1,699,104 3.2 38,525 Low 1,938,981 4.4 72,389 Moderate 1,864,054 5.4 111,364 High 1,325,915 6.2 171,805 Very High 826,442 6.9 365,939 All 7,654,479 5.0 123,278 S _ 1981 Very Low 1,578,911 3.4 28,788 Low 1,634,232 4.7 52,885 Moderate 1,592,805 5.6 81,887 High 1,055,305 6.5 126,855 Very High 688,242 7.2 281,181 All 6,549,495 5.2 90,038 - 60 - Tabte 7: Nuiber of Urban Households Using Fuel by Expenditure Growp Atl Indonesia Electricity LPG Kerosene Charcoal Wood Expenditure 1987 Very Low 604,715 1,271 1,715,899 316,905 871,624 Low 1,549,371 3,628 2,150,671 528,759 667,923 Moderate 1,742,273 28,691 2,039,731 361,191 384,802 High 1,428,893 76,141 1,470,002 146,055 169,651 Very High 974,685 282,743 801,460 68,936 59,698 Alt 6,499,938 392,474 8,177,763 1,421,846 2,153,698 1984 Very Low 449,516 0 1,614,737 263,901 712,921 Low 989,736 846 1,871,587 496,409 451,853 Moderate 1,313,197 6,919 1,817,902 382,140 268,779 High 1,133,287 20,341 1,296,419 160,711 114,501 Very High 784,440 82,457 757,872 49,808 41,344 All 4,670,175 110,563 7,358,518 1,352,969 1,589,398 1981 Very Low 256,090 0 1,554,102 321,755 565,997 Low 534,876 167 1,613,176 604,093 349,793 Noderate 867,119 417 1,568,806 558,434 251,343 High 802,265 13,493 1,036,024 248,119 117,858 Very High 590,526 51,296 650,935 104,021 39,595 All 3,050,876 65,373 6,423,043 1,836,422 1,324,586 - 61 - Tabte 8: Average Fuel Use of Using Urban Households by Fuel and Expenditure Electricity LPG Kerosene Charcoal Wood kgftn kWh/mo kg/uo lt/Mo kg/mo Assumed Expenditure 1987 Very Low MA 11 16 2.7 140 Low MA 11 24 3.0 140 Moderate NA 14 30 3.8 140 High MA 17 36 5.1 140 Very High MA 19 42 8.1 140 All NA 18 28 3.6 140 1984 Very Low MA 0 19 2 140 Low NA 26 30 3 140 Moderate NA 13 35 3 140 High MA 14 41 4 140 Very High NA 22 46 6 140 All NA 20 33 3.1 140 1981 Very Low MA 0 28 2 140 Low MA 27 38 3 140 Noderate MA 14 43 3 140 High MA 13 46 4 140 Very High MA 17 54 4 140 All NA 16 40 2.9 140 .62- Table 9s Totat Fuel Use of Urban Households by Fuel and Expenditure Electricity LPG Kerosene Charcoal Wood kWh/mo kg/mo lt/mo kg/mo kg/mo Expenditure 1987 Very Low MA 14 27,552 as8 122,027 LoW MA 40 51,452 1,600 93,509 Moderate MA 391 60,954 1,357 53,872 High MA 1,274 53,129 742 23,751 Very High MA 5,359 33,590 559 8,358 All MA 7,077 226,676 5,116 301,518 1984 Very Low MA 0 30,808 638 99,809 Low MA 22 56,784 1,304 63,259 Moderate NA 92 64,484 1,220 37,629 High MA 286 53,458 681 16,030 Very High MA 1,791 35,241 312 5,788 Allt A 2,190 240,775 4,155 222,516 1981 Very Low NA 0 43,754 646 79,240 Low MA 5 61,008 1,572 48,971 Moderate MA 6 67,576 1,682 35,188 High MA 181 47,981 969 16,500 Very High MA 883 34,895 447 5,543 All MA 1,075 255,214 5,317 185,442 - 63 - Table lOs Percentage of Urban Households Using Fuel by Expenditure Group Electricity LPG Kerosene Charcoal Iood Expenditure 1987 Very Low 44X 0.0 94X 17X 48X Low 68X 0.2X f5f 231 29X Moderate 82X 1.42 96X 17X 18X High 921 4.9X 95f 9X 11X Very High 95f 27.5X 78X 71 6X Atl 74X 4.5X 93X 16X 242 1984 Very Low 26X 0.0 951 16X 42X Low 51X 0.01 971 262 23X Moderate 701 0.4X 961 21X 141 High 852 1.5X 961 121 9X Very High 952 10.01 921 6X SX ALL 61X 1.4X 961 181 211 1981 Very Low 162 0.01 961 202 36X Low 332 0.01 99X 371 212 Moderate 542 0.01 98X 35X 16X High 76X 1.31 961 242 11X Very High 862 7.52 951 15X 6X All 47X 1.0 986 281 20X -64- Tale 11: Averae Fuel Use of Urban Households by Fuel and Expenditure Electricfty LPG Kerosene Charcoal Wood kWAh/so kg/mo Lt/mo kg/m kg/mo Expenditure 1987 Very Low NA 0.0 1S 0.5 67 Low NA 0.0 23 0.7 41 Moderate NA 0.2 29 0.6 25 High NA 0.8 34 0.5 15 Very NHgh NA 5.2 33 0.5 8 All NA 0.8 26 0.6 34 1984 Very Low NA 0.0 18 0.4 59 Low NA 0.0 29 0.7 33 Moderate NA 0.0 35 0.7 20 High NA 0.2 40 0.5 12 Very High NA 2.2 43 0.4 7 All NA 0.3 31 0.5 29 1981 Very Low NA 0.0 28 0.4 50 Low NA 0.0 37 1.0 30 Moderate NA 0.0 37 1.1 22 High NA 0.2 45 0.9 16 Very High NA 1.3 51 0.6 8 All NA 0.2 39 0.8 28 * 65 - Table 12: Average Fuel Expenses by Using Urban Households by Expenditure Group Etectricity LPG Kerosene Charcoal Wood RpImo Rp/mo Rp/mo Rp/mo Rp/mo Expenditure 1987 Very Low 2,428 4,875 3,259 432 2,660 Low 3,411 4,608 4,883 542 3,254 Moderate 4,630 4,968 6,162 642 3,452 High 6,807 6,223 7,537 885 4,042 Very Nigh 13,940 7,230 8,766 1,094 5,604 All 5,942 6,837 5,719 605 3,176 1984 Very Low 1,949 0 3,474 323 1,989 Low 2,890 4,300 5,588 362 2,459 Moderate 3,884 4,936 6,546 462 2,923 High 5,397 4,993 '(4' 630 3,290 Very High 10,982 6,761 8,;41 1,041 3,394 All 5,047 6,302 6,060 439 2,411 1981 Very Lou 1,217 0 1,577 212 1,324 Low 1.866 1,400 2,094 279 1,77 Moderate 2,435 0 2,379 344 2,107 High 3,475 3,774 2,568 417 2,356 Very High 5,791 7,025 2,983 645 3,066 All 3,156 6,295 2,205 326 1,736 -66- Table 13: Average Fuel Expenditure by Urban Households by Expenditure Group Electricity LPG Kerosene Charcosl Wood Rp/mo Rp/mo Rp/mo Rp/mo Rp/mo Expenditure 1987 Very Low 1,074 3 3,073 75 1,274 Low 2,323 7 4,615 126 955 Moderate 3,798 67 5,918 109 625 High 6,256 305 7,126 83 "I Very High 13,203 1,986 6,827 73 325 Alt 4,387 305 5,313 98 777 1984 Very Low 516 0 3,302 50 834 LoW 1,475 2 5,394 93 573 Moderate 2,736 18 6,386 95 422 High 4,613 77 7,529 76 284 Very High 10,424 675 8,034 63 170 All 3,079 91 5,826 78 501 1981 Very Low 197 0 1,552 43 475 Low 611 0 2,067 103 380 Moderate 1,326 0 2,343 121 332 High 2,642 48 2,521 98 263 Very High 4,969 524 2,821 97 176 All 1,470 63 2,162 92 351 - 67. Tabie 14: Percent of Total Expenditure ia Fuel by Urban household Expenditure Group Electricity LPG Kerosene Charcoal iood All Expenditure 1987 Very Low 2.2% 0.0% 6.4% 0.2% 2.6% 11.4% Low 2.6X O.OX 5.2% 0.1% 1.1% 9.1% Noderate 2.8X 0.0% 4.3% 0.0% 0.5% 7.7% Nigh 3.0% 0.1% 3.4% 0.0% 0.2% 6.8% Very High 2.8% 0.4% 1.4% 0.0% 0.0% 4.7% Alt 2.8% 0.2X 3.4% 0.0% 0.5% 6.9X 1984 Very Low 1.3% 0.0% 8.6% 0.1% 2.2% 12.2% Low 2.0% 0.0% 7.5% 0.1% 0.8% 10.4% Moderate 2.5% 0.0% 5.7% 0.0% 0.4% 8.7% High 2.7% 0.0% 4.4% 0.0% 0.2% 7.3% Very High 2.8X 0.2% 2.2% 0.0% 0.0% 5.3% Att 2.5% 0.0% 4.7% 0.0% 0.4% 7.8X 1981 Very Low 0.7% 0.0% 5.4% 0.1% 1.6% 7.9% Low 1.2% 0.0% 3.9% 0.2% 0.7% 6.0% Moderate 1.6X 0.0% 2.9% 0.1% 0.4% 5.0% High 2.1% O.OX 2.0% 0.0% 0.2% 4.4% Very High 1.8% 0.2% 1.0% 0.0% 0.0% 3.1% Att 1.6X 0.0% 2.4% 0.1% 0.4% 4.6% .68 - APPENDIX I KEROSENE STOVE PERFORMANCE AND USE IN INDONESIA: PROPOSAL FOR A PILOT KEROSENE STOVE IPROVEMENT PROGRAM FEBRUARY 1990 * 69 . TABLE OF CONTENTS I. BACKGROUND AND INTRODUCTION ........ ......................... 71 II. STOVE TECHNOLOGY ........................................... . 73 Introduction ..................................................... 73 Stove Types ..................................................... 74 Power and Efficiency: Laboratory Stove Tests ....... ..................... 74 Safety ..... 77 Modifications ............................................. 79 HI. THE COOKING PROCESS .. ......................................... 81 Controled Cooking Tests ........................................... 81 Household Monitoring ..................... ........................ 83 IV. HOUSEHOLD STOVE CHOICE ...................................... 87 V. KEROSENE STOVE PRODUCION AND MARKETING ................ .... 89 Factory Production ................................................ 89 Design .................................................... 89 Production Technology ........................................ 90 Economics ................................................. 90 Adaptability . . . .............. 90 Artisanal Production ................... 91 Design.. ... ............... 91 Production Technology ........................................ 91 Economics . . . .............. 91 Adaptability . . . ..............92 Stove Market .................... 92 VI. PROPOSED KEROSENE STOVE PILOTPROGRAM ...................... 94 Standards, Endorsements and Factory Stove Improvement ................... 95 Two Region Program (First 18 months) ........................... 96 Expand to all Cities over 500,000 in Population (Second 18 Months) ..... 97 Artisanal Stove lnprovement ........... ............................. 97 Kerosene Stove Research ............ ............................... 98 Consumer Awareness/Involvement .................................... 99 VII. EVALUATION OF THE PROPOSED KEROSENE STOVE PILOT PROGRAM .. 100 Summary of Findings .............................................. 100 Kerosene Use Projection ............................................ 100 Program-specific Assumptions ........................................ 102 Program Budget ............................................. 104 Economic Evaluation ............................................... 105 - 70- Conversion Factors .......................................... 105 Costs .................................................. . 106 Beneefts ................................................... 106 Government Budget and Balance of Payments ...................... 106 Financial Analysis ................................................. 108 VIII. LPGAND BIOFUELSTOVE POLICY ................................ 109 ANNEXES Annex 1: Quality and Test Procedures for Kerosene Wick Stoves ........ ........... 111 Annex 2: The Kerosene Stove Market in Jakarta ............................... 114 TABLES Table 5.1: Estimated Share of the Different Stoves Sold on Jakarta Markets and the Number of Different Types in Each Group ..... ......... 93 Table 7.1: Urban Stove Market Projections ................................... 103 Table 7.2: Stove Improvement Program Budget (US$ millions) ..................... 104 Table 7.3: Conversion Factors . ............................................ 105 Table 7.4: Economic Evaluation of Kerosene Stove Pilot Program .................. 107 Table 7.5: Financial Evaluation of Kerosene Stove Pilot Program ................... 108 FIGURES Figure 2.1: Power and Efficiency of Indonesian Stoves ........................... 76 Figure 2.2: Efficiency and Price of Indonesian Stoves ............................ 77 Figure 2.3: Burner Configuration and Efficiency of Indonesian Stoves ..... .......... 78 Figure 2.4: Effects of Simple Improvements on Efficiency and Power ................ 80 Figure 3.1: Controlled Cooking Test Results: Specific Fuel Consumption ..... ........ 81 Figure 3.2: Stove Power and Cooking Time ................................... 83 Figure 3.3: Household Members and Rice Consumption ......................... 85 Figure 3A: Household Members and Kerosene Use ............................. 86 Figure 4.1: Present Value of Kerosene Conservation ............................ 87 - 71 - I. BACKGROUND AND INTRODUCrION 1.1 An effective program to upgrade existing kerosene stoves is economically desirable, and would have a significant impact on both household (particularly women's) welfare and national oil consumption. The stove analysis undertaken for this project has uncovered important technical possibilities for stove improvement. There are numerous non-technical obstacles to improving kerosene stoves, however. The challenge facing any kerosene stove program will be in translating the large technological and economic opportunities into actual improvements in the cooking stoves. 1.2 In 1986, roughly one seventh of Indonesia's consumption of refined oil products was consumed in kerosene stoves, most of which were used by urban households. At a price of US$18 a barrel, these 135 Petajoules of kerosene per annum would be worth over US$400,000,000. This is more, in both energy and value terms, than the total consumption of petroleum products for electricity generation in 1986 by the public electricity company (PLN). 1.3 Kerosene is used as a cooking fuel by about 75% of urban households in Indonesia. For most urban households, kerosene is a major expenditure item: many low income households probably spend more on kerosene than on any other commodity except rice. Between 1981 and 1984, in response to economic pressures, the Government of Indonesia (GOI) more than doubled the real price of kerosene, bringing it more in line with international prices. Following these price increases, national kerosene consumption declined 6% in 1984,3% in 1985 and remained constant through 1986. 1.4 There are four reasons why the fuel use of kerosene stoves can still be expected to be uneconomically high. First, consumers purchasing a stove cannot ascertain its fuel-economy, either through physical inspection or reputation. Second, stove producers are also ignorant of their stoves' fuel economy, and are generally too small scale to engage in research and development to improve fuel-economy. Third, the habits of both consumers and producers developed under low price conditions, and have not fully adjusted to existing prices. Fourth, kerosene remains subsidized. 1.5 Stoves are produced both in factories and by household industry, or artisans. Factories produce between 5,000 and 100,000 stoves per annum, and generally use presses to make many of the stove parts from new materials. Artisans can produce as little as three hundred stoves per annum, though artisanal complexes also exist, producing up to about 5 thousand. Artisans typically use old metal sheeting, and can be severely constrained by the availability of these materials. The artisans rarely use a press. 1.6 Test results indicate that the efficiency and quality of the kerosene stoves used in Indonesia varies considerably, although the majority are wick stoves of similar design. Many stoves are of low efficiency, and there is little relation between stove price or producer type and efficiency. - 72 - The technical changes required to save kerosene and/or increase the power of most existing kerosene stoves are simple 1/, and need not be expensive. 1.7 The economics of kerosene cooking suggest that a 5% change in a stove's kerosene consumption would be worth over Rp 4,000 to a typical consumer, assuming a discount rate of 10%. The additional production costs of achieving such a change through efficiency improvements would, for most stoves, be considerably less. For example, a 300 Rp (or less) air-flow controller omitted in most artisanal stoves could lead to efficiency improvements of up to 19%, leading to an average kerosene conservation of between 6 and 3% (and in the latter case faster cooking). 1.8 Simple behavioral changes can save considerable amounts of kerosene, but may also require some facilitating technical changes. For example, many households boil and steam rice at the same high power, although, for a wide power range, power has no perceptible effect on the speed of rice steaming or the quality of the cooked rice. Turning down the power during the rice steaming phase could save up to 12% of daily kerosene used in cooking. Assuming this turn down operation required a minute of the cooks time, the return would be 1000 Rupiah per hour: 5 times the average wage. On the other hand, if it required that the stove be monitored for the whole 15 minute steaming phase because of flame instability at low power, the return would only be 65 Rupiah per hour. In short, the ease with which low power cooking can be carried out is intertwined with cooking behavior and fuel economy. 1.9 The effects of technical and behavioral changes are also interdependent. For a typical meal, an increase in efficiency from 41% to 45% will save 3% of the kerosene if the stove continues to be used at the same power setting. If the power is reduced by 10%, so that the heat delivered to the pan remains as it was before the efficiency increase, this same efficiency increase will save 6% of the kerosene. Cooking tests performed in the laboratory indicate that the smaller savings are probably more realistic, although it is possible that long run adjustments would lead to larger savings. Except where otherwise mentioned, efficiency changes will be assumed to have the smaller effect. 1.10 Prior to this study, research on kerosene stoves already indicated the diverse performance of stoves on the Indonesian market. With this study, it has been possible to better identify some of the key determinants of stove performance. Furthermore, a better understanding of household needs and desires has been obtained. It is now appropriate to move beyond the laboratory and the survey, and to work with the producers and users of kerosene stoves. ./ An increase in the tchnolog efficiency of a stove means mom heat is delivered to the pan for a given amount of kerosene. If the powerof stove operation is not adjusted down commensurately, thefuel savings will not mefea thefult incrase in efficiency, especiallyforopetonssuch assteamning wherein higherpowergoes agely to produce moa waste heat. - 73 - II. STOVE TECHNOLOGY Introduction 2.1 This chapter reviews generic kerosene and LPG stove features, summarizes the results of the stove tests undertaken for the UHESS, and describes the stove modifications undertaken in an effort to improve kerosene stove performance. The kerosene stove analysis demonstrates the large potential for improving kerosene stoves with minor modifications. 2.2 The nature of kerosene stoves is such that both power and efficiency can vary significantly among stoves. LPG stoves, on the other hand, can be expected to vary considerably in power, but relatively little in efficiency. For this reason, and given the dominance of kerosene as the major urban cooking fuel, the bulk of the work was on kerosene stoves. 2.3 Past work on testing kerosene stoves in Indonesia already supplied information about the characteristics of the types available in the market. Kerosene cookers, used in about 80% of the urban Indonesian kitchens, all have a comparable basic design. Differences are not essential and relate to cooking speed and type of material used. But these apparently minor differences cause the performance to vary widely. Fires due to exploding stoves are common and can be prevented by better designs. Past tests also reveal the existence of a wide range in power and efficiency. The wide variation in their fuel use implied by these results suggests a large scope for improvements. 2.4 However, prior to the UHESS, these test results were never thoroughly analyzed with respect to how performance is influenced by the stove design. In the work described in a recent World Bank report v/ a first attempt was made to link design features to performance for 20 stoves from all over the world. For the UHESS, the stove work was aimed at extending this insight into the determinants of the performance of Indonesian stoves. 2.5 Lemigas conducted extensive testing on the performance of 18 kerosene and 2 LPG stoves. The work consisted of a detailed description of the stove followed by a series of performance tests. First the power of the stove was determined by weighing the fuel use in half an hour. This was followed by an efficiency test, where the percentage heat which is utilized is obtained. Each efficiency test lasted one hour and was repeated three times to decrease errors. Finally, safety was evaluated by a five hour long test during which the temperatures of different stove parts, including the kerosene, were monitored. This resulted in an overall picture of the quality of each stove, concerning both fuel use and safety aspects. The wide variation in performance was again confirmed. 2.6 An analysis of the stove test results indicates that, controlling for pan size, more compact burners are more efficient. Alternatively, pan supports with rings touching the pan appear to result in lower efficiency. On the other hand, no strong relationship appears to exist between: i/ Sowre: Test Results on Kerosene and OrherStovesforDeveloping Countnes: Energ Depatment PaperNo.27, August 1985, The Wotfd Bank - 74. 1) efficiency and power; 2) price and efficiency; 3) producer type and efficiency, or; 4) age and efficiency. 2.7 In collaboration with a visiting stove expert, several attempts were made to improve the kerosene stoves tested. Two modifications were found to improve the efficiency of a number of stoves. First, restricting the amount of air entering through the inner flameholder was found to increase the efficiency of those stoves which did not already have a built-in flow-restrictor. Second, inverting the pan support was found to increase the efficiency of stoves whose pan supports had rings touching the pans. The average percentage change in efficiency resulting from these modifications were 8% and 13% respectively 3/. 2.8 The LPG stove analysis was limited to two stoves. One was deemed typical of current LPG stoves used, while the other was chosen because of its low cost and expected low power. A wide power variation was indeed found, with the high power LPG stove being more than twice as powerful, at maximum power, than the average for kerosene stoves. The low power gas stove, on the other hand, had a power comparable to most common kerosene stoves for reasons discussed below. Stove Types 2.9 Cotton wicks transport the kerosene from the tank to the flameholder. The kerosene evaporates and burns in between the two cylinders of the flameholder. These cylinders are perforated to let air enter, causing the kerosene vapor to burn with an almost invisible blue flame. A wick adjustment lever allows the user to change the height of the wicks, thereby regulating the power. 4/ A shield surrounds the flameholder, both to prevent accidental touching and also to decrease radiation losses. The pan is placed on top of the frame, often on a separate pan support. 2.10 Most gas stoves have two identical burners, one of which is shown in this figure. The gas flow can be changed with a regulator before it enters the venturi where it is mixed with air. With this particular type, the burner had two rings of holes. The gas supply to the outer rings automatically ceased at low power. Power and Efficiency: Laboratory Stove Tests 2.11 The amount of fuel used over a certain period of time is called the stove power and is related to cooking time required. The higher the power the shorter the time to fulfil certain cooking tasks such as boiling water. For other dishes the cooking time is more or less fixed and does not depend on the power. Each stove has a range over which the power can be adjusted by regulating the fuel supply. When too much fuel is supplied, red flames appear, resulting in black pans. The highest position giving blue flames is called the (blue flame) maximum power. As the t/ All changes in effidency in this to ar relauiw changes. E.8, a 10% effiiency inemase means changingftrwm 40)% to 44%. 1/ Actuao power adjustments dwing cooking ar somev*at difficdt. Not only ls it Sfficult to obsenve theflane, but at low power the buwing process oten becomes igdrar and instable. - 75 . (blue flame) maximum power setting changes when a pan is placed on the stove, it is somewhat difficult to reproduce. A minimum power setting is even more difficult to reproduce. (This accounts for the vatying stove power data, depending on the source.) 2.12 Most households use their kerosene stoves at approximately maximum power. When the food reaches boiling point, the power is normally not lowered. Therefore, the efficiency measurements are done with the maximum power setting. Users are reluctant to decrease the power because the flames are difficult to see. Also the stability of the flames over time is less at low power. There is a common, but generally misguided, belief that high power leads to shorter cooking times, even if the food is boiling already. For the dishes requiring a more or less fixed cooking time, fuel consumption becomes directly related to stove power if it is not turned down. 2.13 The percentage of the heat from the fuel actually utilized to heat the pan contents is given by the efficiency. In the water-boiling efficiency tests conducted at Lemigas, a pan with cold water is brought to boil and is kept at that temperature for the rest of the hour. The heat absorbed by the pan can be calculated from the temperature rise and the amount of evaporated water. The ratio of this absorbed heat and the total energy content of the fuel used during this test is defined as the stove efficiency J/. 2.14 After knowing a stove's power, a suitable pan size can be chosen. A procedure was adopted whereby the volume of the pan is linearly related to power. This results in roughly the same time to bring the water in the pan to boiL If all the stoves were tested with the same pan size, an artificial disadvantage would be given to the high powered ones. High power in combination with a small pan will lead to relatively large losses and thereby to a low efficiency. By using different pan sizes efficiencies of stoves in all power categories remain comparable. 2.15 Each efficiency measurement has its inherent uncertainty caused by measurement errors and variable testing circumstances. It can be calculated that the measurement errors are about 2-3% of the efficiency. As an example: a measurement error of 2.5% when the measured efficiency is 40% results in an expected efficiency between 39 and 41%. Circumstances are never perfectly reproducible. Differences in room temperature, draft, and setting of the wicks lead to varying efficiency results. The standard error in different measurements of one stove is about 5%. To account for both effects, efficiency was measured three times for each stove. Most of the efficiency values given in this report are averages based on three measurements. 2.16 Results of the efficiency and power tests are given in figure 2.1. The power ranges between 0.8 and 3 kW, and efficiencies between 29 and 64%. LPG stoves rank highest with an average efficiency of 60%. Artisanal kerosene stoves have an average efficiency of 40% and factory types 43% V/. Most artisanal stoves have a power less than about 1.5 kW. 2.17 As can be seen in figure 2.1, the range of efficiencies of kerosene stoves is large (29- 5 1%). This indicates considerable potential for decreasing fuel use. It should be clear however that g For more details on efciny calcutons see 'ioesene and LPG Stows in Indonesia Perfonnance, Safet Aspects, and Impromme, m by E. Sangen, 1988 or Tests and Evaluation of Kensene and LPG Stoves, by LAmigas, 1988 This diffeance is not staaly signipcant (Probabily = 0.22). -76- because of limitations in the ability to turn down the power, one can not expect that an increase in efficiency wil lead to an equal decrease in fuel use. 2.18 Cheap artisanal stoves are made from scrap materials and their dimensions are less standardized. When efficiency is plotted against stove price (see Figure 2.2), it can be seen that high efficiency stoves can be cheap as well as expensive. The variation in quality for the cheaper stoves is much higher than for the more expensive ones. It would appear that the higher the pr- ice, the smaller the chance that the efficiency wil be very low. Figure 2.1: Power and Efficiency of Indonesian Stoves Lemigas kerosene stove tests Power and officiency 0.6- 0.5 +0+ + a ~~+ >1 0.4 i t + + i- O0.3 _ 0.2 0.1 0 - , . , , I 0 1 2 3 4 Power In kW o Artisanal + Factory o Gas 2.19 Data available on the 20 tested stoves made it possible to gain some insight into the determinants of stove efficiency. It was already known from other stove tests in and outside Indonesia that the configuration of pan and stove plays a crucial role. That is the reason why the size of the pan has been chosen depending on the power of the stove. Factors, which were found to influence stove efficiency include the distance of the flame holder to the pan and the diameter of the flame holder compared to the diameter of the pan. The results show that efficiencies . 77 - improve somewhat if the distance of the pan to the flame decreases. In general, stoves with a large flame holder diameter have a lower efficiency. In figure 2.3, efficiency is related to the ratio of pan and flame holder diameter. For low ratios (relatively large flame holders) efficiencies are lower, on average. In this figure it is also possible to see the effect of the pan support ring. Stoves having such a ring which touches the pan have a lower efficiency, on average. Figure 2.2: Efficiency and Price of Indonesian Stoves Efficiency - price relation _.7 Lemigas stove tests x 0.5 0.5 o 0 >' 0.4 - 0 03 Y 0 v 4- 0.3 - w 0.2 - 0.1 0 20 40 60 (Thousands) Price in Rupiah 0 Factory o Artisanal x Gas stoves Saety 2.20 In Jakarta, fires caused by kerosene stoves rank highest in number after those caused by electricity. The existing national standards for kerosene stoves prescribe characteristics which should ensure safe use (see Annex I). No flame leakage to parts other than the flame holder is allowed. The temperatures of the different parts should be below certain limits. The temperature of the kerosene in the tank is especially important. If its temperature is higher than about 40 degrees celsius (flash point) a spark can ignite the kerosene vapor. Temperature of the kerosene in the tank is therefore considered the main indicator of safety. Temperatures of the other parts are less important and also more difficult to measure. * 78- eigure 2.3: Burner Configuration and Efficiency of Indonesian Stoves Effect of compactness of burner on the stove efficiency 0.68 0.5 _ 0 Do Y~~~~~~~~~ XO 0.4 0 X 0 ~~~~~~~~~~x 0.3 - '- 0.2 - 0.1O- 0 - I p III 0 1 2 3 4 Ratio pan/burner oiameter o no ring x ring touches pan 2.21 Each kerosene stove was tested for a five hour period, during which it burned at maximum power. Temperatures of the different parts were measured each hour. Flame stability and leakages were noted. Tests started with a half full tank, so that at the end of the five hour period about 10% of the original fuel volume was left over. The results of the temperature measurements are shown in Annex m. (On the horizontal axis the percentage is given of the fuel which is higher than the 50°C limit allowed according to the safety standards at the time of the measurement.) For a few stoves the safety tests had to be stopped because temperatures became so high that the kerosene began to evaporate and started burning from the air vents. 2.22 The large variation in the temperature curves suggests that the current method of burning until 10% of the fuel is left over is not optimaL Approximately three to four hours after starting, most curves are more or less horizontaL To improve reproducibility it might be better to use a three hour period in combination with a lower temperature, e.g. 45 degrees celsius. 2.23 Some of the other requirements for the standards were fulfilled by all the stoves. For all tested stoves, the thickness of the tank material is higher than the prescribed minimum of - 79 - 0.27 mm. Also the stability of the construction was adequate for ali the stoves. Only one stove (a factory made type) had flames which leaked to the ventilation holes. The majority had a stable flame. In total, eleven out of the eighteen stoves did not comply to all the requirements in the SII standards. Two of these eleven currently bear the SII mark. Modifications 2.24 The above described test results suggested certain improvements which might lead to higher efficiencies. Three sorts of modifications were explored. It was already known from other tests that decreasing the distance between pan and burner can increase the performance. One stove was modified by lowering this distance by 1 and 2 cm. Another modification which was explored was limiting the amount of air entering the flameholder. A flow restrictor is used in some but not all of the currently sold kerosene stoves. A third improvement was suggested by the effects of the ring of the pan support; if it touches the pan the efficiency is much lower. 2.25 Decreasing the distance between pan and flame holder allows the flames to touch the pan better. If there is some draft in the kitcnen its negative effect on the transfer of heat towards the pan is reduced. Less heat is spoiled because of hot gasses being carried away by the draft. To measure this effect the pan support of a large artisanal stove was lowered from 57 to 47 and then 37 mm. Contrary to what was expected, the efficiency decreased in both cases by about 1%, which is within the measurement error. It remained unclear why decreasing this distance did not have the expected outcome, but further work on this modification was stopped. 2.26 When too much excess air is supplied during the burning process, the hot flue gasses become unnecessarily diluted. The resulting lower temperature will have a negative impact on the heat transfer to the pan. Some of the factory made stoves sold in Indonesia already have a thin sheet of metal with holes in the inner flame holder. This flow restrictor limits the amount of incoming air. As an important side effect, it also reduces the radiative heat transfer from the flame holder to the tank, leadi'g to lower kerosene temperatures. 2.27 Four stoves were modified by adding flow restrictors. In a trial and error process, the appropriate sizes of the holes were found. All the modified versions showed an increase in efficiency, varying from a 1% to 19% and, on average, 8%. Adding this device to a stove is estimated to cost at most a few hundred Rupiah, and is strongly recoi.1mended both for efficiency as well as safety improvements. 2.28 Some of the factory made stoves have a removable pan support made of thick steel wire. More or less by accident, it was found that if the pan support has a ring to connect the four legs the position of this ring affects the stove efficiency. There are two possibilities: the ring touches the pan or the ring is a few mm below the pan. In the latter case the flow of hot gasses along the pan is less disturbed, resulting in a higher efficiency. For six stoves which have a removable pan support, the efficiency increased between 5 and 25% when the pan support was inverted. The average increase was 13%. 2.29 The results of the three different modifications are shown together in figure 2.4. Each dot represents one efficiency measurement. On the horizontal axis the change in power is given as a ratio of modified to unmodified. For example, 1.1 means that the modification results in a 10% increase in power. From this figure it can be seen that increases of power are about as * 80. common as decreases. In one case, which is not shown in this picture, both the pan support was changed and a flow restrictor added. The combined effect resulted in an efficiency increase of 28%. Figure 2.4: Effects of Simple Improvements on Efficiency and Power Lemigas kerosene stove tests Results of modifications 1.3 e~~~~~~~~~~~~~~~~~? VV 0.9_ r X l . 8 a 0.7- 0.7 0.9 1.1 1.3 Power change a pen list o pan sup. v flow res Figure 2A. Results for three types of modification: changing the distance between pan and burner, turning the pan support, and adding a flow restrictor. Horizontal: the relative change in power. Vertical: relative change in efficiency. - 81 - IIL THE COOKING PROCESS Controlled Cooking Tests 3.1 Laboratory boiling water tests, result in technical parameters such as the efficiency and power of the stove. These results do not show how much fuel is actually used for a certain cooking task. To obtain an insight into how efficiency and power are related to fuel use, controlled cooking tests were conducted during which the same standard meal was prepared on different stoves. A typical meal was chosen, suitable for a five person household. It consisted of steamed rice, cooked vegetables (Cap Cay), and chicken. Figure 3.1: Controlled Cooking Test Results: Specific Fuel Consumption Fuel consumption per kg of food in controlled cooking tests 10 9 - 8 7 13 e9 6- i 5 5 004. en 4 o ° 3 O + A 2 1 0 1 2 3 4 Power CkW) o ker/unc + gas/unc o ker/coo a gas/coo Figure 3.1. Energy consumption per kg of food given as a ratio of cooked and uncooked fcod. - 82 - 3.2 Rice was prepared in two stages. First it was brought to boil with a sufficient amount of water. It was allowed to boil for about five minutes until all the water was absorbed. A steamer takes the place of the pan. After the water started boiling, the rice was put in the upper compartment of the steamer, where it remained for about 20 minutes 2/. A wok was used to prepare both vegetables and chicken in a process which is a combination of frying and boiling. 3.3 The energy needed for cooking the meal can be presented as a ratio of the total energy divided by the weight of the food, to correct for small differences among the ingredients. Two ratios can be calculated, one based on the amount of dry food and the other on the amount of cooked food. Both are shown in figure 3.1 for the results of eight tests. From this graph it can be seen that the higher the power the higher the fuel consumption. Though the gas stoves have a somewhat lower energy consumption, they also show this relation of higher consumption with higher power. As expected, the data using the dry food ratio are above those for cooked food. 3.4 Contrary to original expectations, It was found that fuel use did not depend on the efficiency. A graph of the data showed no trends, and this was confirmed by statistical analysis V/. The absence of a dependence on the efficiency is caused by the fact that this dish, like most dishes, require a more or less fixed cooking time. From the in-depth household interviews, it was found that it is not common to turn down the power of a stove when the contents boil. If used at a const- ant power, the fuel use of a stove becomes directly related to the power. 3.5 Figure 3.2 shows that the cooking time depends only slightly on the power. For the kerosene stoves, with power ratings varying by a factor of 2, the cooking time varies only about 20%. Both gas stoves allowed for fast cooking. Though this could be inherent to gas stoves, it might also be caused by the expectation of the cook to be able to cook faster. 3.6 Besides the meals also water is heated on the stove, not only for drinking but some- times also for bathing. Because heating water is essentially what happens during the water boiling efficiency tests these data could be used to calculate an average daily energy consumption in case kerosene is used for cooking. It was assumed that daily amount of food cooked is twice the amount used in the controlled cooking tests. This meal is prepared once, after which it is reheated. Calculating the daily energy needs results in: Etot = 6.6 + 3.8 * Power + 3.2 / eta [MJ/day] Eta is here the efficiency. For typical values of efficiency and power, about 1/3 of the energy consumption is constant, 1/3 depends on the power and another 1/3 on the efficiency. For example an efficiency increase of 30% can be expected to result in a decrease of 10% of the daily cooking energy consumption. When this efficiency improvement is accompanied by a 30% decrease in power, a daily fuel consumption decrease of 20% can be expected. The cooking time will remain Z/ The household survey indicates that this is the most common way to prepac nice in wban areas. About 70% of households use a steaner, while 24% use only a pan. I/ Multiple regression on the data of the sikerosenestoves resudted in an insignificant value (t = -066 probabi'ity = 0.56) of the pammeter for the efficiency. *83 - the same in this example because the amount of energy which enters the pan does not change when efficiency and power change proportionally. Figure 3.2: Stove Power and Cooking Time Cook i ng t i me mea I vs. power controlled cooking tests 0 0 0 4 ~~~~~~~~+ + 6T 'D~~~~~~~~~~ 1 2 -+ 2- I - 0- l I I I I I 0 1 2 3 4 Power CkW) o kerosene stoves + gas stoves Figure 3.2. Cooking time for the preparation of a complete meal as a function of the power of the stove. Household Monitoring 3.7 To complete the picture of Indonesian cooking practices, a number of households wel c closely monitored over a period of two weeks. Nine kerosene using households of roughly middle income were selected. When they were first visited, questions were asked about their current stoves. Over one week, the cooks were asked to fill in a form, describing all the meals, the number of persons who ate those meals, and the time required for cooking. For measuring the kerosene consumption they were given a measured fuel supply. Four of the households got new -84- stoves, the remainder used their own. After one week, the stoves were collected and exchanged for other types. The monitoring ended with a interview, asking the users about their experiences and preferences. 3.8 The majority of the stoves in use by the chosen households were artisanal types. Only three out of the original group of twelve households which were visited used factory made types. Two third of the artisanal stoves were bought from travelling salesmen for prices between Rp. 3,500 and Rp. 6000. Some people had an agreement with this salesman that the stove could be exchanged if it is not good enough e.g. not giving a nice blue flame. Stoves in these homes averaged three and five years old for artisanal and factory types. 3.9 In laboratory tests, it was discovered that the position of the pan support affects fuel efficiency considerably. One household was asked to use the same stove for two weeks, with the only difference that in the second week the position of the pan support was reversed. Kerosene use in the second week was 15% lower (when corrected for the number of times household members were present during both weeks the consumption is 9% lower). This corroborates the laboratory finding of a 16% higher efficiency. However, the pan support position giving the lowest efficiency was preferred because the traditional dandang was not stable in the other position. A slightly inclined pan support (in the center lower than in the perifery) can overcome this problem. This illustrates the need for trying out modifica.ions first in households before promoting them on a large scale. 3.10 Safety is an important aspect for both consumers and the authorities. Accidents with kerosene stoves is one of the main causes of fires in households. Some of the participants had actual experiences. One forgot to close the fuel regulator in her fixed wick stove, causing large flames. She never used it again. In another household the small stove which was given to them exploded, luckily causing no fire. But, according to the cook, this stove was safer than her own, because when that one exploded it caused only a small fire. It is clear that safety considerations should remain of primary importance in standards. Support for modifications increasing fuel efficiency have to be accompanied by changes leading to a higher safety. 3.11 The preferences for certain stove characteristics was one of the main aims of questions which were asked to the cooks. By far, the most important characteristic is the possibility to cook fast, being mentioned by all the households. This is an interesting finding in relation to the generally low power of the stoves in use and the easy availability of stoves with a higher power. However, most respondents made clear that they understood that the higher the stove power the higher their fuel consumption will be. 3.12 Other preferences were mentioned by at most two respondents. A turning knob was not valued high over a lever mechanism. Surprisingly, only one respondent answered that she wanted a nice blue flame. Maybe this is thought to be so obvious for a good stove that it is not worth mentioning. One person said that she would like to have a stove which saves fuel and another wanted one which is easy to light. The person using the stove with the 'auto-lighting' device (no need for matches) was not really enthusiastic about this feature. 3.13 To make energy consumption in different households comparable to each other one has to look for a measure of the amount of food cooked during the week. Just the amount of * 85 - household members is not sufficient because it is common for people to eat outside the home. Two parameters were measured, the amount of rice cooked and the number of people present during each of the meals. When the average number of persons multiplied by the number of meals is plotted against the average daily rice consumption, a linear relation can be found as shown in figure 3.3. Because rice is eaten in almost each meal the data approximate a straight line. The number of meals per day is assumed to ve a good measure of the amount of food consumed during the week. Figuare 33: Household Members and Rice Consumption Household members and rice consumption 26 dur ing household monitoring 24- 22-or 20- 16- 14-C 0 0 'fl 12 X 10 0 0 8~~~~~~~~ 4 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 rice consumption Ckg/d) 3.14 One expects the kerosene consumption to increase when the amount of food cooked each day increases, although not in the same proportion because of economies of scale. When the amount of meals is plotted against the average daily kerosene cunsumption for each household, figure 3.4 results. No relation can be found, which suggests that the energy consumption does not depend on the size of the household. The expected errors in combination with the small size of the sample does not allow to draw hard conclusions from this data. ,86- Eigure 3.4: Household Members and Kerosene Use Household members and kerosene use 2. during household monitoring 1.9 1.8 1.7 1.6 1.3 1.4 1 0 2- l 1. 0.1 4 00 0~ 0 0970 0.4 0.3 0.2 0.1 0 48 112 1'6 2024 meal-persons per day 3.15 Further statistical analysis proved to be fruitless. Energy consumption is not related in a simple manner to stove power and efficiency. A high (negative) correlation between household size and stove power makes it difficult to draw conclusions. The simple model to calculate household kerosene consumption for cooking described above could not be validated. The average daily consumption of 0.85 liters of kerosene per day (for an average household size of 5.3) is somewhat higher than the simple model predicts (0.71 l/d). This can be explained by the fact the some of the respondents used their stoves also to heat bathing water, which is not included in the modeL 3.16 A gas stove was given to one of the households who used a kerosene stove in the first week. The average daily energy consumption went down from 39.6 MJ/d to 21.6 MJ/d. This is caused by both the much higher efficiency (64% compared to 37%) as well as the possibility to turn down the gas stove. Her own fixed wick kerosene stove could not be regulated. - 87 - IV. HOUSEHOLD STOVE CHOICE 4.1 Women are not only the major stove users, but also exert the most influence on when, what type, and what price of stove is purchased. In fully 85% of households, the principal woman determines stove purchases. 42 As noted above, stove-efficiency is not perceived as an important criterion in purchasing a stove. Only 12% of the households surveyed in Java selected fuel-economy as a factor influencing their purchase of a new stove. Even fewer (10%) claim it is the major factor. In-depth interviews revealed that when fuel consumption is a consideration, it is primarily through its relation to power. High-powered stoves are recognized as fuel-guzzlers, and for at least some Figure 4.1: Present Value of Kerosene Conservation PRESENT VALUE TO A HOUSEHOLD OF 5% DECLINE IN A STOVE'S KEROSENE USE '12 5 YR LIFE 10 - 3 YR LIFE '-'I-' +J0 4)~~~~ 2 0 I ' i' I~ l ,1| SX 25X 45X 65X 85X 15% 35% 55% 75% 95% D I SCOUNT RATE Figure 4.1. Financial savings achieved by a 5% decline in kerosene use. - 88 - households the choice of stove is guided by a trade-off between the faster cooking of a high- powered stove and the lower kerosene use of a low-powered stove. 2/ 4.3 Superficially, purchasers' neglect of efficiency (and more generally, fuel economy) appears economically irrational. At the extremely high discount rate of 100%, 10Q/ the present value of a 5% decline in the kerosene consumption is over 2000 Rupiah, even if the stove only lasts two years. This is roughly oiie third the cost of a moderately priced stove. Given the wide v3riation in efficiency of stoves on the market, this would suggest that efficiency should be a very important parameter in choosing a stove. Even under pessimistic assumptions, a 15% change in stove efficiency should lead to a more than 5% change in kerosene consumption, and such differences are commonplace. 4.4 Rather than irrationality, however, the most obvious explanation is a lack of information. It is not possible to ascertain a stoves efficiency simply by examining its features 11/. Once in use, the large variation in the amount of cooking done easily hides a 5% change in average kerosene consumption. Even when the Project provided measuring devices to monitor households' kerosene consumption, differences between stoves were swamped by other factors, along with remaining measurement problems. Briefly, under existing circumstances it would not be rational for a purchaser to use efficiency as a criterion for selecting stoves since s/he cannot know the efficiency. 4.5 Three important policy-relevant observations can be made from this analysis of household stove choice. First, attempts to promote more efficient stoves will not be supported by consumers unless something is done to help consumers ascertain which stoves are efficient. Second, consumer preferences vary considerably, and no single stove design can satisfy all users. By implication, a program to improve existing stoves is likely to be more successful than one aimed at propagating particular models. Third, consumers are likely to be attracted to efficiency improvements in part because they allow faster cooking, although faster cooking dissipates some of the potential fuel economy of a more efficient stove. This has important implications for the promotion of efficient stoves. 2/ This trade-off only etis because power adjustmew is not easy for most kerosene stoves. j/ A 100% discount rate approimates tight cash flow constraints on many households a. exhibited by comnnon stove purchasing pattens. L/ All other things being equai a bue flame is more Mefflcient"than a yellow flame, and blue flpames aoprefemnd. Flame color is not a good indicator of stove efficiency, however. - 89 - V. KEROSENE STOVE PRODUCTION AND MARKETING 5.1 For a stove program it is indispensable to know something about how stoves are produced, on what scale, and in what quantities. Furthermore, it is important to understand the stove marketing system. In this chapter, the differences in technology used by f-.ctories and artisans are described, and related to the opportunities for improving stove quality. This is followed by a discussion of the stove market. The discussion is limited to locally produced stoves, imports being insignificant. 5.2 Both factory and artisanal stove production are amenable to the modifications discussed in Chapter II. Factories, being both larger scale and having more ready access to capital, are a more obvious target. Even the smallest artisanal producer should, however, be able to add a flow-restrictor and/or produce efficient pan supports. Furthermore, artisanal producers are often clustered in one area, so;netimes all making the same model of stove. 5.3 Materials are the major cost for both factories and artisans. For artisans, reliability of material supply is also crucial. This could provide an opportunity for government assistance, linked to a stove improvement program. (Factory production, being more standardized, could be affected through official stove standards.) 5.4 A small survey of the Jakarta market indicates that artisanal stoves account for about 80% of sales, or, given shorter lifetimes, about 70% of the stoves used. The UHES survey shows that 50-60% of the stoves used are artisanal types. More than half of the artisanal stoves are sold door-to-door, making Standard Product Information sheets inappropriate. The payment schemes for the door-to-door sales tend to involve very high implicit interest rates, suggesting that costly improvements are unlikely to be popular among artisanal stove buyers. Factory Production Design 5.5 An overwhelming variety of stove designs can be found in Indonesia. Round or square frames, using sheet metal or aluminum, and luring the potential buyer with a finish of paint or enamel. Though largely irrelevant for the performance, these features affect consumers' decisions on which type to buy. Factory made stoves have a standardized design, made possible by the use of machinery such as presses. They can also be distiniguished from artisanal types by the use of new materials, since most artisanal stoves are fabricated from scrap material. 5.6 l he design characteristics which are most relevant to the performance are dimensions of the flameholder and the form of the pan support. Both are important for the fuel efficiency. Existence of a flow restrictor influences both efficiency as well as safety via lower temperatures of kerosene in the tank. Some of the currently available stoves already have high efficiencies, making them difficult to improve on. Others only need small modifications in the design such as changing the pan support. The rest might need more fundamental changes. Results from work on modifications in this project suggest that most kerosene stoves can be improved by small design modifications. * 90- Production Technology 5.7 Hand tools can be sufficient to make a good quality stove. However, to produce homogeneous series with the same dimensions and appearances, machines are indispensable. Certain stove parts, especially the flame holders, have critical dimensions and need to fit well to ensure a steady blue flame. Consumers are supposed to prefer stoves with a nice appearance, giving the impression of a higher quality than stoves made by hand tools. 5.8 Machines are mainly used for tasks which can not be done by hand, such as pressing the frame parts. The use of dies in these presses guarantee that the sizes of the different parts are exactly the same for each stove. Some factories have a small workshop, allowing for the production of dies. A typical workshop is equipped with a lathe, drilling and welding machines and hand tools. A spot welding machine is sometimes used to make the flameholder, and to fix the plate which supports the wicks and the wick setting mechanism. Finishing is done by paint spray installations. 5.9 A considerable part of the work in a stove factory, such as the assembling and seaming of the wick pipes, is still being done by hand, not much diff-zresn from artisanal production. An increase of the labor productivity is probably not the most important stimulus for mechanization. In the three visited factories the share of the labor cost was only 10-15% of the production costs. The daily output is between 3 and 4 stoves per employee, which is only about double the daily amount made by an artisanal producer. Economics 5.10 From two of the three factories visited, data were obtained to estimate production costs. In the case of a 16 wick stove made of thin sheet metaL with a retail price of Rp. 6,000, the share of the material cost was about 50% and the share of labor 15%. The costs for distribution, overhead, profits and retail margin is 35% of the total. For an aluminum stove with a retail price of Rp. 9,500 the share of the material cost is about 63% and of labor 10%. Prices of aluminum rose about 50% in 1988, making it difficult to produce an aluminum stove at a competitive price. Adaptability 5.11 Modifications of stoves to improve safety and efficiency can be grouped into three classes. First, the modifications which only require a minor change in the production process but involves no extra material. An example of this is making a pan support for which the ring does not touch the pan. This can be accomplished with the tools which are already in u,'. The second group are the modifications requiring only low cost adaptations. An example would be inserting a flow restrictor, which will not change the total production costs very much. It will add only a small amount to the material costs, and requires no different technology, because it is essentially comparable to the flameholder. The third group consists of more fundamental changes in the design, leading to extra material cost or increasee use of machines. This will be the case when dimensions of critical parts need to be changed. - 91 - 5.12 Risk aversion and the wish to minimize costs favor the design modifications which only require low extra costs. During this project, two low cost modifications were identified, giving a considerable improvement in efficiency and safety. These should have the highest priority in implementation, because of their large chance of being successful. In the longer term, these should be complemented with modifications requiring more fundamental changes. The latter should be accompanied by research and support from outside. Producers will probably only be interested if there is a clear, effective demand from the public for these modified stoves. An extra incentive could be some financial support for extra investments. Artisanal Production Desien 5.13 Almost all of the stoves produced by artisans have a cylindrical shape and all have ten or more wicks. Designs are simple, omitting less crucial parts to keep the costs as low as possible. Artisans are often concentrated in certain areas where they all produce more or less the same type. The variety in shapes is much less than the variety available in factory made stoves. Because they often bear no brand name, and are painted either blue or green, artisanal stove types are difficult to distinguish between each other. Production Technolo 5.14 Extremely simple tools can produce reasonably good looking stoves. A hammer and chisel, a dye for the wick holder and some soldering equipment are the most important tools. Most of the stove parts can be made from oil drums and old cans. The use of these scrap materials and the absence of machine tools cause the dimensions of certain parts of the stove to vary considerably within a series of one stove maker. It is common knowledge among users that the quality, especially of the flame, can differ widely for the same stove type. This is corroborated by efficiency tests of four stoves of the same type, showing that the highest efficiency is 1.46 times higher than the lowest 1/. Economics 5.15 In the artisanal complex Penggilingan, the 16-wick type is sold to distributors for about Rp. 3,500 a piece. The largest share of these costs are the material cost (36%) and the labor cost (34%). Overhead takes 16% and profits 14% of the total. The retail price of this type is about Rp. 5,000. These cost shares are assumed to be typical for artisanal production, though on the average the overhead is probably somewhat lower. 5.16 As could be expected, the share of the material cost is smaller compared to the factories, but it is still the main cost. Together with the new materials, the price of the scrap materials continue to rise, causing problems for the small producers. A regular, guaranteed supply for reasonable prices would benefit their business. 21/ See: Zests and Evaluation of Kerosene and LPG StovesT b Lemigas. .92. Adaptability 5.17 Performance improvements can be expected from design modifications comparable to the ones mentioned for the factory stoves. Adding a flow restrictor, which is generally absent at the moment, is the main change which should be focused on. Large differences in fuel efficiencies of stoves of one stove maker suggest that a good design is necessary but not a sufficient condition for a satisfactory performance. A well-designed stove can have low quality flames because critical parts such as the flameholder cylinders are not well fitted. Therefore it is preferable to lay the major emphasis on production quality improvements with some additional minor design changes dealing with safety. 5.18 One of the steps in quality improvement is the use of suitable materials. Artisans are interested in a guaranteed supply of basic materials of which the quality is not too variable. Currently most of the handicraft stoves are made from scrap. The flame holder, which is the most critical part when performance is concerned, is often made from whatever is available. There exists a national program in which large industries support small ones (the 'anak angkat' program). Within this program the delivery of constant supply of basic materials could be more or less guaranteed. This would be most easily executable if the artisans worked together in cooperatives. 5.19 At the moment, most artisans use a minimal amount of tools. A simple set costing about one million Rupiah can increase appearance of the stove, its quality and also the productivity. The high cost and added risks have to be overcome by capital support, preferably via the existing credit programs for the support of small industries (KIK/KMKP). A research and testing institution can play the role of intermediary between credit suppliers and the artisans. Stove Market 5.20 Almost all stoves used in Indonesia are produced locally. Imported kerosene stoves are seldom seen, although Japanese names and labels are often used to promote sales, especially for stoves from Surabaya. 5.21 In Jakarta, most people purchase their stoves from a market or from a travelling salesman. Buying from a salesman has three advantages: a) you can exchange the stove in case of unsatisfactory performance; b) it relieves women from carrying the sometimes heavy stoves; c) salesmen are more inclined to offer credit. Some people agree on two installments, others are so short of cash that they pay in up to 30 installments. The price, up to 50% higher, is partly a compensation for the need of daily collection of these small amounts of typically only Rp. 200. 5.22 The stove market is very much divided in regions, each area having its own factories and groups of artisanal producers. Many factory stoves are also available in Jakarta. Stocks of stoves available in six market places in Jakarta were counted to get a first impression of market shares of the different types. Under the assumption that the rate of sales is directly related to the amounts in stock, the shares are given in table 5.1. In total 923 stoves were counted and 60 different types were identified. For details on assumptions, the calculation procedure, and a list of stoves, see the Annex II. The crudeness of the method cause the results in Table 5.1 to be only rough estimates. v 94 - VI. PROPOSED, KEROSENE STOVE PILT PROGRAM 6.1 This chapter describes a kerosene stove program intended to improve the kerosene- economy of urban cooking stoves. Rather than attempting to identify and promote the "best" stove, the objective of the proposed program is to raise the quality of existing stoves. This strategy was selected because: 1) there are numerous obstacles to marketing a new stove; 2) tastes in stoves vary so that no single stove can satisfy the diverse requirements; and 3) simple modifications should be able to raise the efficiency of most existing stoves. 6.2 The kerosene stove program described here-in moves beyond kerosene stove analysis, towards direct involvement with kerosene stove producers and users. The preceding chapters demonstrate the large technical potential for low cost improvements in kerosene stoves. Furthermore, they indicate a need for government to bridge the 'information gap', and enable kerosene producers and users to make informed, rational choices affecting kerosene use. An effective program must provide both technical support for improving kerosene stoves, and the information to enable users to demand better stoves. 6.3 The proposed program involves four components: 1) Factory stove improvement; 2) Artisanal stove improvement; 3) Kerosene stove research; and 4) Consumer awareness/advocacy, promoting high performance kerosene stoves and kerosene-saving cooking techniques. No radical stove design changes or cooking changes are envisaged. Rather, the program emphasizes minor changes which would almost certainly be implemented given better informaticn. As described below, the program focuses on kerosene economy. Safety aspects should be added to the program, but their evaluation is beyond the scope of this study. 6.4 Guidance and monitoring for the energy aspects of the program would be provided by the Directorate General for Electricity and New Energy (DJLEB), based on its prior experience with energy conservation. The Ministry of Industry would be responsible for the technical side of the program. The regional offices of the Ministry of Industry (Kanwil) would foster contacts with producers. The Ministry's laboratories (Balai Besar Penelitian Bahan dan Barang Teknik), which have considerable experience testing kerosene stoves for safety, would provide the facilities. The Indonesian Consumer Organization (YLKI) and local womens' Education for Family Welfare groups (Pendidikan Kesejahteran Keluarga) would assist in disseminating information, and provide feedback. A research group could be located at one of the technical universities. 6.5 Work with factories would emphasize standards and endorsements, whereas work with artisanal producers would emphasize direct support in improving the production process. There are several reasons for distinguishing between artisanal and factory stoves. Artisans produce the majority of kerosene stoves, but are not readily amenable to standards, and are of too small scale to receive individual attention. Work with artisans would, at least initially, be limited to the large complexes of artisans, which may already be involved in government supported activities. Factories are more amenable to standards and endorsements, and produce on a larger scale. That some factories are already willing to go to some effort to obtain the little publicized "SII" suggests that they will be very interested in endorsements promoted through a publicity campaign. - 93 . Table 5.1: Estimated Share of the Different Stoves Sold on Jakarta Markets and the Number of Different TIpes in Each Group Group Market Share U types Surabya made (Butterfly and Toyo series) 7X 16 Twc targe Jakarta producers 21X 5 Rest of the factory iWde stoves 8X 26 Artisanat production 64X 13+ Totats: 100X 60+ 5.23 From the responses of 20 households, lifetimes of stoves were estimated to be three years for the artisanal and five years for the factory types. These are rough estimates and are not true for all stoves; sturdy artisanal types can last for 10 years. Using these lifetimes and the assumption that on average a household owns one and a half stove, the yearly number of stoves which are bought in Jakarta can be calculated to be about 473,000 (see Annex II). Based on sales data of the two large Jakarta producers, the total sales of ali factory types was estimated to be 93,000 per year. This results in the following percentages: 20% of the stoves sold in Jakarta are factory made, 30% are handicraft types bought on markets and the remaining 50% are handicraft stoves obtained from travelling salesmen. The finding that most artisanal stoves are sold in door- to-door sales is corroborated by the in-depth household survey result that only 4 out of the 11 visited households bought their artisanal stove on the market. 5.24 From the table in A.nnex II it can be seen that about 30% of the artisanal stoves sold in markets bear brand names. When it is assumed that this is also the case for the door-to-door sales, then about 44% of the total amount of kerosene stoves sold :n Jakarta have a brand name. After taking into account the estimated differences in lifetime it. .an be calculated that 51% of the Jakarta population use factory or artisanal stoves with brand names. This can be compared with the survey result of 40 - 50%. * 95 - 6.6 A sub-program to improve consumer awareness, and solicit consumer participation in the program, will also be necessary. Improved consumer awareness will be very important. Without pressure from consumers it will be far more difficult to convince producers to improve stove quality. Standards and endorsements must be understood by the consumers as vell as producers, so that consumers can assess the costs of purchasing a sub-standard stove. Simultaneously, the knowledge and preferences of consumers (or consumer representatives) must be tapped to insure that the standard/endorsement program is appropriate to their needs. 6.7 Applied research based on "trial and error" would be an integral part of the effort to advise on kerosene stove improvements. In addition, it would be worthwhile to support a small scale research program, complementary to the overall stove program. This research would draw on the stove test results, but its objective would be to better understand the general determinants of stove performance, rather than possible improvements in particular models. Though useful, this research would not be strictly necessary to the stove-improvement program. Standards. Endorsements and Factory Stove Improvement 6.8 Efficiency tests should eventually be incorporated into existing kerosene stove standards so as to ensure minimum efficiency. In addition, high efficiency stoves should receive special endorsement. The Ministry of Industry organizes the testing of kerosene stoves from factories wishing production licenses. Research organizations (the "Balai Besar") regularly test stoves according to the "Standard Industry Indonesiae (SII). 13 Currently the Standard is concerned only with safety. Within the next four years, efficiency should be added to this standard. Also, the existing standard should be reviewed and simplified so as to ensure that the tests employed are reproducible. Prior to the inclusion of efficiency in the "SI", regional programs should be implemented to improve kerosene stoves in these regions. In the course of these regional programs, the necessary experience would be gathered to facilitate the adoption of efficiency standards and endorsements nationally. 6.9 Currently, few stoves meet the 'SIP, and even fewer are entitled to bear the "SIP" 1aLal, indicating they comply. Partly as a result, the standards have not been strictly enforced. Tests undertaken at Lemigas 14/ illustrate two additional problems inherent in stove standards. First, the performance of different stoves from a single producer can vary significantly. This has been clearly demonstrated for two artisanal stove producers, and some variation can also be expected among factory stoves from a single producer. Second, measurement difficulties, especially for such parameters as the temperature of different stove parts, make the "SII" tests difficult to reproduce. As an example of these difficulties, neither of the two stoves tested at Lemigas and bearing the "SIP" label passed the Lemigas "SIl" tests. Similar difficulties apply to efficiency measurements, as mentioned above. Under such circumstances, it will be necessary to improve the reproducibility of the Standards and/or to include second round stove tests for stoves within a / Details of this Standarepni *ed in Annexi, and wem used to guide the Lemigas Stove Tesdng undeflaken for the UHESS. 1/ Test and Evaation of Kvsene and LPG stoves, Lemigas, Jakata, September 1988. - 96 - specified range of the allowable performance. Fvrthermore, some criteria should be developed to identify the worst offenders. 6.10 The producer side of the program would thus consist of two phases. During the first phase, the program would be limited to two regions, and would include a re-working of the "SIIT", as well as stove testing and extension work to improve the kerosene stoves produced in these regions. During the second phase, the efficiency standards and/or endorsements would be included in the revised "SIP to be applied nation-wide. Two Region Program (First 18 months) 6.11 The two region program would be analogous to a pilot project: by the end of two years a systematic and reproducible procedure for kerosene stove testing and improvement should have been developed. The two region program would include, for example, Jakarta and Surabaya. Many of the stoves tested for the UHESS were made in Jakarta, and the modifications uncovered are definitely appropriate, making Jakarta an obvious program area. Surabaya is the source of many of the better known factory stoves. The timing could be simultaneous or sequential. The six activities envisaged during this first phase are described below. Revision of Standards 6.12 First, a new "SIP" standard would be developed. This new Standard would: 1) incorporate efficiency (via water boiling tests) and/or efficiency endorsements; 2) simplify the safety tests and ensure repor&ucibility; and 3) allow stoves to be ranked with respect to safetv. Creation of Standard Product Information Data Sheet 6.13 The format of a Data Sheet, eventually to be provided to consumers with each factory stove, would be developed. The data would include information on maximum power, efficiency and safety. The information would, however, be in a form accessible to users. Thus, for example, efficiency could be summarized in terms of the kerosene required to bring one liter of water to boil, or combined with power estimates to calculate an "average" daily kerosene requirement. Purchase of Stoves 6.14 Sample stoves would be purchased directly from all factories and artisanal complexes in each region, informing the producers of the purposes of the program. Simultaneously, summary data would be collected on production levels by stove type and technology employed. Stove Testing 6.15 Up to 40 stoves would be tested in each region, according to the new standard, informing producers of the results and providing them with a sample Standard Product Information data sheet for their stove. - 97 - Modification of Selected Stoves 6.16 Starting with the more popular low quality stoves, the tested stoves would be examined for possible improvements (especially air flow restrictor and pan holder, at least in initial stages). The suggested improvements would be discussed with the producers. Producers failing to meet the standards would be notified that standards will be enforced in the near future. Tests of any modifications made by a producer would be made free of charge. For stoves for which significant improvements seem possible, and the producer is agreeable, household tests would also be undertaken free of charge if deemed necessary. During this activity, applied research would also be undertaken to extend the list of possible stove-improvements. Also, the adoption of proposed modifications would be monitored, and barriers to adoption would be reassessed. Revise Proposal 'for SII and Second Phase of Program 6.17 Based on the work of Phase one, the proposed Standard for kerosene stoves would be formalized. Also, the details of the second phase of the program would be proposed, systematizing the procedures employed. Phase II Exand to al Cities over 500.000 in Population (Second 18 Months) 6.18 By 1990 there will be about 14 cities over 500,000 in population, accounting for almost half of the urban population of Indonesia, and well over half of the large scale kerosene stove producers. During Phase 2, the program would be cxpanded to include these cities. Adopt New "SII' on National Scale 6.19 By the end of the second phase of the program it should be possible to adopt the revised stove standards and endorsements on a national scale. Artisanal Stove Improvement 6.20 While standards and endorsements are not appropriate to artisanal stoves, artisans should receive support in improving their stoves. Not only do artisans produce the majority of stoves, but their production is more labor intensive, thus providing more employment per stove. Also, any attempt to coerce artisans into abiding by stove standards will risk simply breaking-up existing artisanal stove complexes, and further reducing their opportunities for improvement. 6.21 The standard-endorsement program should therefore include extension work with artisanal producers. This would involve testing and advising on the more popular, low quality, artisanal stoves. It is expected that introducing flow-restrictors would have a large impact on the kerosene-economy of many artisanal stoves. -98 - 6.22 Government programs already involving kerosene stove producers should also be expanded to include efficiency and power considerations. The 0Anak Angkat" program could be a useful vehicle: a state company whose waste includes used sheet metal (e.g. used oil barrels) could "adopt" a complex of stove producers, ensure a continuous supply of raw material, and solicit advice on stove improvements from the group working on factory stove tests. (Alternatively, experience in modifying artisanal stoves could be obtained by working with the prisoners who are already producing kerosene stoves in significant quantities in, for example, West Java's Cirebon regency county prison.) More generally, a training program could be developed following the first phase of factory stove improvement. 6.23 The improvement of artisanal stoves should include modifications of the production process itself. As noted above, one of the disadvantages with artisanal stoves is the large variation in the performance of stoves, even from the same producer. Especially with safety, the range of quality is very important: it is not enough that the "typical" stove from a given producer is safe. Technical suppoi t aimed at improving the production process is expected to have a positive effect on quality. Kerosene Stove Research 6.24 The work with producers above inherently involves applied research. Developing of improved stoves can draw on and extend the "trial and error" research described briefly above, and in more detail in a working paper. .L/ However, important aspects of the burning process are still insufficiently understood, providing a barrier to stove improvement. As such, there is a need for a small research project as an integral part of the stove program. The aim of this additional research will be to identify and quantify the determinants of kerosene stove efficiency, power and safety. This will enable stove improvements to be identified more rapidly and systematically. 6.25 Close cooperation will be necessary between the stove-testing group and the research group. Ideally, the research group would work directly with stove testers, so as to ensure the transfer of practical knowledge on available stoves to the research group, in return for the stove testers gaining insight into the fundamentals of the burning process. The research would be small scale (no more than one senior expert and one or two assistants). It would be directed towards applications, though not to particular stoves. Testers would have to obtain regular reports on results, and provide input into the research program. 6.26 A possible base for this proposed research group is a technical university. Low cost and simple construction make kerosene stoves a good subject of research for students. The laboratory requirements include balances, temperature sensors, a micro-computer for data logging and analysis, and a location with adequate ventilation but minimal drafts. W 'Kimsene and LPG stows hu lIonesia' Einew SanFgu, Ausw 19088 . 99 . Consumer Awareness/Involvement 6.27 For the success of the kerosene stove improvement program, it is important both for consumers to be made aware of how to obtain a more efficient (and safer) stove, and for the program to be made aware of consumer preferences. If consumers "demand" improved stoves, producers will be far more inclined to adopt modifications. Furthermore, only consumers themselves can ever be fully aware of the advantages and disadvantages of the modifications. 6.28 Consumer awareness should emphasize knowledge currently not available to kerosene stove users. Ideally, thiis knowledge would be disseminated via local television: 65% of the persons with a major influence on purchasing kerosene stoves watch television regularly. It should be oriented to women, who account for 85% of these influential persons. The awareness campaign should be linked to the program of standards and endorsements, initially in the two regions selected and later expanding to the rest of Indonesia. Information would cover both how to interpret the endorsements, Standard Product Information data sheets, as well as details on the new kerosene stove Standards for factory stoves. Also minimal information would be included, in a form which most users can understand, to demonstrate that, for example: (a) Flow Restrictors can improve stove performance, leading to lower kerosene use and faster cooking. (b) Pan Supports with rings touching the pan lead to more kerosene consumption and slower cooking. (c) Steaming rice at high power wastes kerosene & water. 6.29 Consumer and Women's organizations (Yayasan Lembaga Konsumen Indonesia, YLKI) and Pendidikan Kesejahteran Keluarga (PKK)) should both spread this information, obtain feedback, and act as consumer advocates. While household tests can be undertaken within the stove modification program, alone this would be insufficient. The Indonesian Consumer Organization already has some experience with kerosene stoves, and should eventually play a major role in providing critical information on kerosene stoves. The program for family welfare (PKK) has neighborhood groups, and could provide a much-needed direct link with kerosene stove users. - 100 - VII. EVALUATION OF THE PROPOSED KEROSENE STOVE PILOT PROGRAM 7.1 The first section of this chapter summarizes the findings of the evaluation. This is followed by a discussion of the projected urban kerosene use and the presentation of a tentative total kerosene use estimate. The third section contains a presentation of the various other parameters used, some of which were estimated by the project. The fourth and last section gives a more detailed overview of the evaluation of the stove programme. This assessment covers the potential value of the program to (a) the national economy and to (b) the government in terms of its budget and the balance of payments as well as (c) to the consumer. Summary of Findings 7.2 The program is a very attractive proposition from an economic point of view. Discounted at the opportunity cost of capital (10%) net returns over the program period of ten years would be about six times the cost. In the base case, the net present values (NPVs) of the flows of gross benefits, costs, and ne-t benefits are calculated to be US$16.6, 2.2 and 14.4 million respe^tively. In terms of the government budget, the program is expected to be more than budget- neutral by showing a net benefit NPV-discounted at a 'soft' loan rate of 10% of US$2.7 million. The balance of payments is expected to benefit quite significantly, over the program period. The program is estimated to have an import oantent of about 40%. This cost is dwarfed by the inflow of funds generated by the sale of saved kerosene in the international market. Over the program, net benefits are accumulated with an NPV of about US$12.3 million (at 10%). 7.3 The financial analysis, which looks at the program from the consumers' point of view, has a lower flow of benefits because the border value exceeds the financial price of kerosene by about 30%. However, it has also significantly lower costs because the program costs are not included as the consumer is not expected to pay for them. As a result, the returns are similar to those of the economic analysis. Kerosene Use Projection 7.4 Since the stove program is aimed at the urban population, the relevant projection concerns the urban use of kerosene for cooking purposes. The UHESS household survey has found that in 1988 kerosene was the main, or only, fuel used for cooking by about 70% of urban households. However, this share is subject to change as kerosene has several substitutes including LPG, fuelwood, town gas and electricity. A relative change in consumer access to any of these fuels will change their position in the urban market. 7.5 Access may change as a result of a change in both the (physical) availability of particular fuels and purchasing power because of adjustments in price and/or income levels. Town gas is at present a very minor source of cooking fuel largely as a result of its very limited availability. Electricity, on the other hand, is widely available in urban areas. In Java, some 85% of the urban households presently use electricity for lighting and virtually all urban households are expected to have electricity by the turn of the century. However, the use of electricity for cooking requires both a considerable investment in cooking equipment and a fairly drastic change in cooking - 101- practices. For these reasons neither town gPs nor electricity is expected to change its current marginal position, at least within the time horizon of the project. 7.6 The use of fuehwood is expected to actually decline as cities continue to grow In size and income increases. The household survey has found an inverse correlation between city size and the importance of fuelwood as a fuel source for cooking. This is something one would expect to see. Widening city limits encroach on agricultural land, making it scarcer and as a result fuelwood production declines. As cities grow in size, fuelwood becomes a more cumbersome, less practical fuel. Distances increase and fewer dwellings accommodate the use of fuelwood. Also, as incomes rise more people adopt lifestyles in which reducing the time spent on cooking chores becomes more important. 7.7 Kerosene and, to a much lesser extent, LPG will be the two main cooking fuels. Urban kerosene use for lighting is expected to fall as electrification progresses. Its use for purposes other than lighting and cooking, is minor and likely to fall. As income levels rise, more households will switch to LPG; providing access of consumers to LPG also increased. 16/ At lower income levels, however, kerosene will replace fuelwood. As more people are expected to switch at the upper end of the income scale than at the lower end, a decline in the use of kerosene for cooking .ould be expected. However, as incomes rise, the average household use of kerosene also rises. Overall, in fact, the use of kerosene is expected to increase in urban areas. 7.8 Fuel use projections are illustrated in Chapter HI of the main report. Kerosene use is projected to grow at 3.4% per year over the next 12 years. Fuelwood use is expected to grow marginally. LPG is expected to double its share of urban cooking fuel demand and triple in absolute terms over the period. Population Growth Assumption 7.9 The urban population of 45.3 million is expected to increase to 82.3 million by the year 2000; implying a growth rate averaging 4.1% over the period concerned. At present the urban household consists on average of 4.85 persons and, for simplicity, is assumed to stay constant. Share of Househu.ds Cooking With Kerosene 7.10 At present almost 70% of the urban households use kerosene as their main fuel for cooking. As mentioned earlier, growing cities and rising incomes are expected to bring about a substitution of kerosene for fuelwood as weli as an increase in the kerosene use per average urban household. The combined effect is estimated to increase the share from 70% to slightly more than 77% of all urban households by the turn of the century. Many urban households possess more than one stove. The average for urban kerosene using households is 1.6 stoves. This average would rise, but only very slightly, with the rise in income. The present average use of 250 It of kerosene per stove per year is not expected to change. 1W Not only in tenns of phjkal availability but also in tenns of the initial investment required -i.e. in bottles and cooking equipment. It woudd seem that what keeps potentid consumer awayfmm LPG is nor so much the pnce of LPG itself but nwtherthe prce of the inital (set oj) LPG bottles, which ponedly go forpsices considerably higher than the official pices, and the ptice of the LPG stow. -102 - Kerosene Price and Cost of Supply 7.11 Kerosene prices are set by government. It is subject to a uniform pricing system at the (Pertamina) central depot level. From the depot to the end-user, a system of fixed regional transport and uniform trade margins regulates price. The government subsidy on the average re- tail price and the border price of kerosene were calculated based on the third quarter 1989 Singapore posted price as shown in Chapter of II of the main report. The average retail price of Rp 203 per liter comes from the household survey. At this price, kerosene includes an average subsidy of Rp 92 per liter. Prouram-specific Assumptions 7.12 Of the kerosene stoves used by urban households most have been made by artisans. Only some 30% are estimated to be factory-made. This has implications for the average lifetime of the stoves as the factory-made stoves are thought to be more durable. A factory stove will need to be replaced on average every five years. The lifetime of artisanal stoves is estimated to be not more than three years. This reduces the average replacement period of urban stoves to 3.9 years. 7.13 The program design includes a start-up period of 18 months with two units; one each in two large urban areas in Java. A unit would comprise two local experts and two local tech- nicians. After this first phase the number of units in these two areas is planned to double and four additional units would be established to cover four urban conglomerates in the outer islands. Du- ring both phases of the program an expatriate expert would direct and co-ordinate the various units. 7.14 Each unit is expected to test on average 55 stoves per year: i.e. 20 factory-made stoves and 35 models produced by (large complexes of) artisans. Half of factory stoves and artisanal stoves are expected to offer useful opportunities for effective low cost design improve- ments. It is anticipated that the response of factory owners will be such that half of the improve- ments wii be adopted. With regard to artisanal stoves the expected "success rate" is about two- fifths. More specifically, each unit is expected to bring on average 5 factory stove models and 7 artisanal stoves up to "improved" standards every year. The effect of the modifications made should result in an average efficiency improvement of 15% for each improved stove. A 15% improvement in efficiency is expected to lead to a savings in kerosene use of 5%. 17/ 7.15 The average production run of factory stoves is estimated to be 5,000 units per year. Artisan complexes are thought to average 2,000 units. The design improvements emanating from one stove improvement unit could therefore lead to 25 thousand factory and 14 thousand artisanal improved stoves being put on the market every year that the program operates. With two teams in the first one-and-a-half years of the program, and 8 units in the following years, the numbers involved would be some 78 thousand improved stoves in year two, an additional 200 thousand or so units in year three and another 300 thousand stoves by the end of year four. By the end of the fourth year there could be a stock of improved stoves in the urban households of close to 600 thousand units. Jj/ A gmaterstwve efiicincymay leadhouseholds to cookat lowerpowersenings and increasefuelsavings. However, this reladonship is farfrom clear, and such savings have remfom been ignored in the evauation of the progiwn. - 103- 7.16 However, work will be needed to maintain the level reached after the first few years. Sustainable upper limits of the shares in the two segments of the annual market for stoves that could be obtained by factory stoves and artisanal stoves were considered to be 25% and 15% respectively. 7.17 As Table 7.1 shows, improved factory and artisanal stoves, together, would manage to capture 18 percent of the total urban household stock of stoves by the tenth and final year of the program. In terms of annual sales, factory stoves are expected to obtain, and maintain, a share of 25% of their market segment by the fourth year of the program. The share of artisanal stoves is expected to reach 15% in year 8. Table 7.1: Urban Stove Market Projections 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total stock 11.7 12.3 12.9 13.5 14.1 14.7 15.3 15.9 16.5 17.2 17.8 18.5 19.3 of which improved: 0.62 2.52 5.32 9X 12X 15X 16X 172 172 172 Factory 0.05 0.23 0.45 0.69 0.94 1.16 1.26 1.32 1.38 1.45 Artisinal 0.03 0.13 0.34 0.63 0.97 1.30 1.60 1.81 1.94 2.03 Annual Purchases New entrants 0.53 0.55 0.61 0.63 0.65 0.68 0.71 0.73 0.76 0.79 0.82 0.85 0.89 Replacement 2.80 2.94 3.10 3.26 3.44 3.62 3.81 4.00 4.20 4.41 4.63 4.86 5.10 Total 3.33 3.49 3.70 3.89 4.09 4.30 4.51 4.73 4.97 5.21 5.45 5.71 5.98 of which: Factory 0.71 0.74 0.79 0.83 0.87 0.91 0.96 1.01 1.05 1.10 1.16 1.21 1.27 Artisinal 2.62 2.75 2.91 3.06 3.22 3.38 3.55 3.73 3.91 4.10 4.30 4.50 4.72 7.18 There are an estimated 150 or so factory stove models at present, produced in some 30 to 40 stove factories. At the end of phase IH of the program, another 30 stove models and/or another half dozen stove factories will be in production. By then, the program will have tested more than 200 factory stove models. However, the total number of original models looked at is expected to be considerably smaller. Many models will need to be worked on by a unit for more than one year. At first success will come relatively easy but, as time goes by, returns are expected to fall fairly rapidly and more effort will be required. 7.19 This is in fact the reason for putting an upper bound of 25% on the sustainable level of market share. Following the first two relatively easy phases, most of the effort will go into sustaining past achievements. It seems inevitable that some producers as well as consumers will drop out. Making up for these numbers will take up much of the time of the stove units. They are not, therefore, expected to increase the number of improved stove models -that are put into production- by more than a few each year during the last half of the programme. . 104- 7.20 Ir. the artisanal sector of the stove market there are many more models than in the factory sector. There will be no dearth of artisanal stove models for the program to test. Because of the much more individual approach to stove-making, progress is expected to be significantly slower than in the factory segment. Capturing 15% of this sector after about 7 years, however, should he possible. Program Budget 7.21 Total base cost of the first two phases of the program is estimated to be US$1.3 million in constant 1988 US dollars. Some 60% would cover the cost of project personnel. About 12% would be spent on transportation and 14% would go to renting space. A publicity campaign (5%), support of fundamental research activities (3%), sundry expenses (3%) would account for the rest of the budget. Table 7.2: Stove Improvement Program Budget (US$ millions) Item Unit Unit 1990 1991 1992 Total Buildings Rent 1.0 month 24.0 60.0 96.0 180.0 Equipment 11.0 33.0 4.0 Scale 2.0 4.0 12.0 16.0 Computers 2.5 5.0 15.0 20.0 Other equipment 1.0 2.0 6.0 8.0 Transport 25.0 75.0 100.0 Van 10.0 20.0 60.0 80.0 Motorbike 2.5 5.0 15.0 20.0 Personnel 174.0 255.0 336.0 765.0 Local experts 1.5 mm 36.0 90.0 144.0 270.0 Foreign experts 10.0 mm 120.0 120.0 120.0 360.0 Technicians 0.75 mm 18.0 45.0 72.0 135.0 Publicity capaign 10.0 20.0 30.0 60.0 Sundry 15.0 15.0 15.0 45.0 Household testing Artisan training Stove and fuel purchases Research 15.0 15.0 15.0 45.0 Institutional support 4.0 4.0 4.0 12.0 Transport Om 7.0 17.5 28.0 52.5 Van 6.0 15.0 24.0 45.0 Motorbike 1.0 2.5 4.0 7.5 Base cost 285.0 494.5 524.0 1303.5 Contingencies 10X 28.5 49.5 52.4 130.4 Total cost 313.5 544.0 576.4 1433.9 105 - 7.22 Phase I and II would include 720 man-months of local experts/technicians and 36 man-months of expatriate expertise. Transport would include 8 light vans and 8 motorbikes. Annual operating and maintenance costs were estimated to be 30% and 20% of the initial invest- ment cost of vans and motorbikes, respectively. Contingencies were included at a rate of 10% to cover possible cost increases of, particularly, she publicity campaign, institutional support and su- ndry expenses. 7.23 The program design covers a period of ten years in all. The expatriate coordinator is not expected to need more than three years to help set up -'Ie program and transfer his exper- tise to one or more of his local experts. As mentioned before, much of the program's efforts during the following seven years will need be concentrated on program "maintenance", ie, keeping up market share. The cost of the "Phase IIIP is estimated to be US$300 thousand per year (in current 1988 US$). After ten years this need is not expected to have been fully met, but the period in- volved is long enough for it to have beer. considerably reduced. Some testing, however, will still be required for a long time afterwards. 7.24 The ten year time horizon for the program was chosen because after a full decade the program ought to be completely institutionalized. Moreover, because costs and benefits are discounted, extending this period further would affect the cost benefit analysis only slightly. Economic Evaluation Conversion Factors 7.25 Conversion factors (CFs) are used to convert financial values into economic values. Such factors adjust financial prices by correcting for prevailing market distortions and excluding transfer payments such as taxes. For instance, the border price calculation implies a CF for kerosene of about about 1.43, as the economic cost (the border price) of kerosene is 1.43 times its average financial price. Except for household fuels such as kerosene, the estimation of CFs would clearly be beyond the scope of the present project. Therefore, use has been made of available estimates presented in Table 7.3. The official exchange rate (OER) used in the analysis is US$1.00 = Rp 1,700. For the sake of convenience, economic values have been denominated in US dollars rather than Rupiah. Table 7.3: Conversion Factors Standard conversion factor 0.9 Rural unskilled labour: Java 0.7 Elsewhere 0.8 Industrial plant and machinery 0.9 Other traded manufactures 0.8 Construction 0.8 Non-traded services 0.8 Transport 1.0 General project costs 0.8 Opportunity cost of capital, .a. 10% - 106- Costs 7.26 There are two cost sources. In addition to the cost of operating the program, there is a cost incurred by implementing the recommended design improvements. The latter is expected to be only a sma'l cost on average because (a) the program is expected to focus specifically on low cost improvements, and (b) in many cases design improvements should be input-neutral, i.e. just using the same amount of inputs in a different way, even when (b) more inputs are required, the cost increase involved should be quite minor as both material and labour costs are rather low. The average cost of bringing a stove model up to the improved standard was estimated to amount to Rp 250 per stove produced. The CFs applied to program costs and additional stove production costs, respectively averaged out at 0.80 each. Benefits 7.27 The benefits generated by the program are the savings in kerosene used valued at the border price of this fuel. Savings of kerosene are expected to reach a level of some 40 million litres by the tenth year of the program, or an economic value of about LUS$2.7 million per year. The value of the flow of net benefits, after accounting for costs, over the ten year period of the program discounted at the opportunity cost of capital (10%) amounts to LJS$ 16.6 million. This is about six times cost (of US$ 2 million, discounted). In fact, with an outcome of this magnitude the result of the internal rate of return (IRR) calculation is not meaningful. This is a very attractive program as it would still show a reasonable return even if only half of the expected benefits were obtained (a savings of only 2.5% instead of 5%) and the costs would double. Both possibilities are quite remote. If anything, costs would be expected to be somewhat lower than budgeted, particularly after year 3. Government Budget and Balance of Payments 7.28 The government budget is affected in two ways. It has been assumed that the GOI will be able to obtain a soft loan to cover program expenditures (their financial value). Therefore, a discount rate of 10% has been used. Secondly, by not selling the kerosene (domestically) the GOI need not pay out the subsidy on kerosene (of Rp 92 per liter). In the base case, the GOI budget would show a net benefit of US$ 2.75 million. 7.29 The balance of payments effect consists of the import requirement emanating from the program and the revenues from the international sales of kerosene saved. The import element of the various budget items was estimated to average about 40%. In the base case, the balance of payments would improve by almost US$ 12 million over the period concerned. - 107 - Table 7.4: Economic Evaluation of Kerosene Stove Pilot Program 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 HH stock of stoves (Milliois) Total 12.9 13.5 14.1 14.7 15.3 15.9 16.5 17.2 17.8 18.5 improved 0.0% 0.6% 2.5X 5.3% 8.6% 11.9% 14.72 16.22 17.0% 17.2% Sales of stoves (Millions per wanum) Total 3.8 4.0 4.2 4.3 4.5 4.7 4.9 5.1 5.3 5.5 Improved 0.0% 2.0% 6.6% 10.0% 12.32 14.4X 16.42 17.12 17.1% 17.12 Costs USS (Mitlions) Program cost 0.25 0.44 0.46 0.30 0.30 0.30 0.30 0.30 0.30 0.30 Increased stove cost 0.00 0.01 0.03 0.05 0.06 0.07 0.06 0.04 0.03 0.02 Total cost 0.25 0.44 0.49 0.35 0.36 0.37 0.36 0.34 0.33 0.32 Kerosene saved Liters (Millions) 0 1.0 4.4 9.8 16.4 23.6 30.3 34.9 37.9 39.9 Foregone Fuel Supply Costs USS (Millions) 0.00 0.17 0.75 1.67 2.80 4.03 5.17 5.96 6.47 6.80 Net Economic Benefits -0.25 -0.28 0.26 1.32 2.43 3.66 4.81 5.61 6.14 6.48 Economic Internal Rate of Return 128.3% Government Budget Expenditures 0.31 0.54 0.58 0.38 0.38 0.38 0.38 0.38 0.38 0.38 Income 0.00 0.05 0.24 0.53 0.89 1.28 1.64 1.89 2.05 2.16 Net -0.31 -0.49 -0.34 0.16 0.51 0.90 1.27 1.51 1.68 1.78 Balance of Payments imports 0.13 0.23 0.25 0.16 0.16 0.16 0.16 0.16 0.16 0.16 Exports 0.00 0.13 0.60 1.35 2.25 3.24 4.16 4.79 5.20 5.46 Met -0.13 -0.10 0.36 1.19 2.09 3.08 4.00 4.63 5.04 5.30 Summary Evaluation of Kerosene Stove Pilot Program Net Present Value O 10% Annualized NPV (USS Millions) (USS Millions) Kerosene Saved 16.65 2.71 Program cost 2.03 0.33 Increased stove cost 0.21 0.03 Totat Costs 2.25 0.37 Net Economic Benefits 14.41 2.34 Gowvenment Bisdt 2.74 0.45 Balance of Pa)nts 12.30 2.00 1700 Rupiah = I US Dollar. Economic Costs of Kerosene Supply (Rupiah/liter): 290 Government Kerosene Subsidy (Rupiah/liter): 92 Export Value (Rupiah/liter): 233 - 108- Table 7.: Financial Evaluation of Kerosene Stove Pilot Program 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 (US$ mitlions) Increased stove costs 0.011 0.040 0.064 0.078 0.086 0.081 0.058 0.040 0.028 Kerosene savings 0.131 0.591 1.329 2.228 3.217 4.142 4.805 5.262 5.586 5.850 Net Financial Benefits 0.120 0.551 1.265 2.149 3.130 4.061 4.747 5.222 5.558 5.850 Met Present Value at 10X (US$ millions) Increased Stove Costs 0.3 Kerosene Savings 13.4 Net Benefits 13.1 Financial Analysis 7.30 The financial analsis concerns the people who actually save kerosene, i.e. the con- sumers. They have to pay a little more on average to obtain an improved stove but once they have one, their expenditure on kerosene is expected to fail by 5%. Table 7.5 illustrates that this is an excellent proposition as over the decade involved, net returns, on a discounted basis, are well over thirty times costs. The financial benefits are a little lower than the economic benefits, largely be- cause of the fuel subsidy, but the difference in cost almost makes up for this. The cost is consider- ably lower because the consumers are not charged the cost of the program. To the consumer, the program cost is a GOI subsidy. In short, the consumer would capture most of the benefits of the program, but only pay a small share of the costs. - 109 - VIII. LPG AND BIOFUEL STOVE POLICY 8.1 Kerosene stoves deserve and receive the focus of this report. However, both LPG and biofuel are significant household fuels. LPG will grow in importance, and is already used by over 10% of the households in Jakarta. Woody biofuel is used by about 22% of the households in urban Indonesia, and is especially important among low income households residing in the smaller urban areas. For both these fuels, the quality of the stoves will be important to future developments, and there are possibilities for small scale programs. In this chapter, the two fuels are reviewed in turn, and the most important policy issues identified. Ltiquified Petroleum Gas 8.2 LPG is now used by roughly 5% of the urban households in Indonesia, with a total consumption of over 150 metric tons per year. The market has been growing rapidly, despite a shortage of bottles and attempts to constrain demand by raising fuel and bottle prices. One of the strategy options examined in the UHESS is the promotion of LPG as a household fuel. The attractiveness of this option depends upon the amount of kerosene displaced by LPG. This depends, in turn, on the quality of the kerosene and LPG stoves. 8.3 As discussed in Chapter II, the current LPG stoves in use in Indonesia are high powered when compared to typical kerosene stoves. This dissipates a large part of the savings resulting from the higher efficiency of LPG, if the higher power is used during such operations as steaming rice. An econometric analysis of households using LPG and kerosene for cooking among upper income households (see main report) suggests that one kilogram of LPG substitutes for 1.7 liters of kerosene. This implies that LPG cooking requires about one quarter less energy. By way of contrast, looking only to efficiency (which averages about 50% higher for LPG), one might expect LPG cooking to require one third less energy. 8.4 Especially if LPG is to be promoted, this observation has important policy implications. First, consumers should be made aware of the costs incurred by using high power LPG stoves. Ihis information could be disseminated via Standard Product Information sheets and a small publicity campaign. Second, any obstacles to local production of low power stoves should be removed X8/. In particular, low power stoves should be demonstrated to local LPG stove producers who currently model their stoves on high powered stoves suitable only for very high income consumers. At least one producer willing to market low-power stoves vigorously should be identified, and, if necessary, assisted. a/ Tlis power issue is especialy imponan for one and tKw burmer LPG stoves which do not often have the low power buner supplied with the four bwner models. . 110 - Biofuel 8.5 As incomes rise, biofuel may eventually cease to be a sinificant urban cooking fuel. However, the results of the UHESS survey clearly indicate that biofuel is not an exclusively rural cooking fuel Furthermore, biofuel being more scarce in urban areas, urban users can be expected to be partulary attracted to improved stoves. While there is no justification for an urban wood- stove program, there is a clear opportunity for extending a rural stove program to urban areas. - 111 - Anna 1-: Qualy nd Test Prcedures for Kerosene Wlck Stoves Indonesian Industrial Standard SII0135-76 1. Introduction: These standards consist of regulations about the quality and testing procedure of kerosene wick stoves. 2. Definition: A kerosene wick stove is a stove having one or more wicks, and using kerosene as fuel in a tank which is not pressurized. 3. Quality regulations 3.1 Temperature 3.1.1. Fuel: The maximum temperature of the kerosene in the tank is 50 °C 3.1.2. Stove surface temperatures: * Maximum surface temperature of frequently touched parts is 80 ec. * Maximum surface temperatures for other parts, except the fuel tank is 94 oc. 3.2 Construction 3.2.1. Fuel tank and fuel transport system: * All connections below the fuel level have to fit neatly to prevent leakages. * If the tank is made from steel plate with an anti-corrosion layer, the minimum thickness is 0.27 mm. * If the tank does not have an anti-corrosion layer, the minimum thickness of the steel plate is 0.35 mm. * If the flame holder is not directly connected with the tank, the fuel transport system should be firm and not leaking * The fuel tank should have a ventilation hole which is situated far from the flame holder. 3.2.2. Flame regulator: The flame regulator should be made as follows: * Not leaking. * It should be possible to adjust the height of the wicks fast and smoothly. * The wick should be sufficiently tight in the wick tubes to prevent falling. * The construction should be firm. * It should be easy to light. 3.2.3. Stability: If placed on a plane with an inclination of 15 degrees the stove should remain stable in all directions, without spilling kerosene. 3.3. Flames: During the time the stove is used it should have stable flames, which do not leak to other parts and which does not give black smoke on the position of the wick regulator. - 112 - 4. Test procedure 4.1. The kerosene used for the test should satisfy the following conditions: * Distillation 200 ec: 40-60% based on volume. * Higher boiling point: 240-260 ° C. * Distillation residue: 1.5% based on volime. * Sulphur content: 0.1-0.15% based on weight. * Flash point (Abel): higher than 100 ° Fahrenheit. Caloric Value: 18750 BTU/lb. * Color: no color. 4.2 Testing circumstances: * Room temperature during tests: (26 + /-2) °C. * Starting temperature of the kerosene: (26 + /-2) "C. * The room should be as free as possible from drafts, which could disturb the tests. 4.3 Temperature 4.3.1 Measuring the kerosene temperature: Tank is filled. Next the flames are regulated as high as possible but still burning well. The temperature is measured each hour until 90% of the kerosene is used. For measuring the temperatures a quick-silver thermometer should be used with a scale of 0-100 ° C. The quick-silver reservoir should be below the kerosene surface (if possible, use a cork). The temperature when 90% of the fuel is used is not allowed to be higher than 50 °C. 4.3.2 Measuring stove surface temperatures: Measuring the temperatures of the frequently touched parts and the other parts is done according to 4.3.1, using an instrument for measuring surface temperatures. Measurements are conducted on aforementioned parts which are expected to have high temperatures. The highest allowed temperature of frequently touched parts is 80 " C, and for the other parts except for the tank 94 ° C. 4.4 Construction 4.4.1 Testing the construction of the kerosene tank and transport system such as mentioned before in the safety regulations can be done based on visual inspection of the testers. The thickness measurements should be done with a screw-micrometer. 4.4.2 Testing the wicks except for the wick holders as mentioned in the regulations can be done based on visual inspection during the whole test. Test of the wick holder: During the kerosene temperature test the wicks are suddenly lowered to the lowest position giving a good flame. The wicks are not allowed to slide downwards or fall in the kerosene, which can cause fires. 4.5 Flames: This test consists of observing the flames when the stove is lighted. The test starts with a full tank and the flames are regulated to give the highest power as advised by the producer. It is observed if the flame is stable and does not leak to other parts - 113 - until 90% of the fuel is used. This test is repeated with flames dimmed to half power. - 114- Annex 2: Tbe Kerosene Stove Market In Jakarta A large range of different stove types are available in Indonesia. To get an impression of the variety and to make an estimation which part of them are handicraft, six large Jakarta markets were visited. Besides the travelling salesmen, these are the popular places to obtain stoves from. Shops inside the main market building and the surrounding streets mostly sell the factory stoves, while the stalls sell the artisanal types. In addition to the markets, the two largest stove factories in Jakarta were visited. From the Sri Jaya factory about 95% of the output consists of the 16 wick type Kujang. Total production was estimated to be 100,000 stoves per year, of which 40% are sold in Jakarta. In the Iriga Gelang' factory, total yearly sales were reported at 20,000 of both types of stoves produced. To calculate the total sales of factory made kerosene stoves in Jakarta from market shares and known production rates P of the Kujang-16 and Tiga Gelang the following equation was used to obtain an estimation: Stock of all factory stoves PF = E: p *P W Stock of type i With PF the yearly sales of factory stoves, P, the yearly sales for type L and W1 a weight according to relative market share (Kujang W, = 0.6; Tiga Gelang W2 0.4). Using the stock data results in an estimated yearly sales of 130,000 factory stoves. The yearly sales of artisanal stoves can be calculated from an estimated total minus the share of the factory stoves. It was assumed that roughly 80% of the 1.6 million households in Jakarta have on average 1.5 kerosene stoves. These are assumed to have lifetimes of three years for the artisanal and five years for the factory types. This leads to yearly sales of 370,000 artisanal stoves. Results are summarized in the Table below. Estimated Yearly Sales of Kerosene Stoves in Jakarta Type of stove Yearly sales Market Share Factory made 130,000 26X Artisanal sold via market 190,000 38X Artisanal sold house-to-house 180,000 36X Total 500,000 100X - 115. APPENDIX Im A POUCY AND PROGRAM STRATEGY FOR MORE EFFICIENT HOUSEHOLD ELECICITY USE IN INDONESIA 1/ FEBRUARY 1990 1/ Prepared by Lee Sd*p, L.www Betey LAbcwtoy. - 116 - TABLE OF CONTENTS I. THE PATTERN OF HOUSEHOLD ELECTRICITY CONSUMPTION IN I)ONESA ..................................................... 118 Introduction: The Household Energy Survey ............................ 118 Main Findings: Household Electricity Use in Urban Java ... ............. ... 120 International Comparisons ............... ...................... 129 EL PRESENT SITUATION IN INDONESI: EXISTING AND NEW APPLIANCES ..... 132 Survey: Existing Appliances ......................................... 132 Lighting ................................................... 133 Refrigerators and Freezers ..................................... 133 Air Conditioners ............................................ 136 Washing Machines ........................................... 136 Television ................................................. 136 WaterPumps ............................................... 137 Irons . ................................................... 137 Fans . ..................................................... 137 Rice Cookers ............................................... 137 Other Appliances ............................................ 138 Issues in the Appliance Market ....................................... 138 Appliance Manufacturers in Indonesia .................................. 139 The Role of Foreign Producers of Electric Appliances Sold in Indonesia ... 140 AIL POTENTIAL FOR IMPROVING THE EMCIENCY OF ELECTRICITY USE ...... 142 Technical Conservation Potential ................................ 143 Behavioral/Management Strategies ............................... 149 Barriers to More Efficient Electricity use in Indonesia ...................... 152 Information ................................................ 152 Buying Habits and Environment ................................. 152 First Cost and Quality ........................................ 153 Electricity Prices ............................................ 154 Other Problems ............................................. 154 IV. REGULATIONS: TESTING, STANDARDS, AND NORMS FOR ELECTRIC APPLIANCES .................................................... 155 Regulation and Standards ...................................... 155 Regulations in Indonesia that Apply to Household Appliances .... ...... 156 Establishing Testing and Regulatory Standards for Electric Appliances in Indonesia ................................................ 157 V. AN ELECTRICITY EFFICIENCY PROGRAM FOR INDONESIA ..... ........... 162 Introduction . ..................................................... 162 Proposed Electricity Conservation Program for Indonesia .................... 169 - 117- Information Interventions ....................................... 170 Available Information on Appliances .......... .. ................. 170 Testing and Labelling of New Appliances ......... .. ............... 170 Metering of Actual Appliance Energy Use .......... ............... 173 Labelling and Information Requirements ......... ................. 173 Consumer Outreach: Increasing Consumer Appliance Literacy .......... 174 Demonstrations and Government Procurement ......... ............. 175 Information Program: Summary and Time Frame ........ ............ 176 Technical Interventions: Producing and Selling More Efficient Appliances ........ 177 Voluntary Agreements: Targets for Efficiency .......... ............. 177 Minimum Efficiency Standards ............... ................... 178 Research in Indonesia .................. ...................... 179 Technical Interventions: Summary and Time Frame ....... ........... 180 Incentive Programs ................................................ 180 Indicative Benefits and Costs of the Efficiency Program ......... ............ 181 TABLES Table 1.1: Urban Residential Electricity Consumption by Tariff Category .... ......... 119 Table 1.2: Electricity Consumption by Expenditure Group ........................ 121 Table 1.3: Electricity Consumption per Appliance by Expenditure Group .... ......... 123 Table 2.1: Percentage of Households in Tariff-source Category Using Each Device ...... 132 Table 5.1: Components of Future Electricity Use and Potential Savings ..... ......... 163 Table 5.2: Urban Electricity Conservation Program Costs and Benefits ..... ......... 183 FIGURES Figure 1.1: Appliance Ownership On Java, UHESS Survey ........................ 122 Figure 1.2: Monthly Urban Residential Electricity Use and Expenditure, UHESS Survey . 124 Figure 1.3: Urban Residential Electricity Use By Major End Use ................... 126 Figure 1.4: Household Electricity Use For Common Appliances, UHESS Survey .... ... 127 Figure 1.5: Household Electricity Use For Uncommon Appliances, UHESS Survey ...... 128 Figure 1.6: Urban Residential Electricity Use in Indonesia and Major OECD Countries . . 130 Figure 2.1: Appliances Ownership by Expenditure Group, UHESS Survey .... ........ 135 Figure 3.1: Wattage vs. Size of 2 Door + Refrigerators .......................... 146 Figure 3.2: Energy Efficiency Ratio vs. Capacity of Room Air Conditioners .... ....... 147 Figure 3.3: Wattage vs. Size of Color Television in Indonesia ..................... 148 Figure 3.4: Electricity Use of Common Appliances in Indonesia: Present Stock and Advanced Models ........................................... 150 Figure 3.5: Electricity Use of Uncommon Appliances in Indonesia: Present Stock and Advanced Models ........................................... 151 Figure 5.1: Impact of More Efficient Appliances on Future Electricity Use .... ........ 166 Figure 5.2: Household Electricity use on Java: Potential Impact of Greater Efficiency in 2008 ............................................ 167 REFERENCES * 118 - I. THE PATrERN OF HOUSEHOLD ELECTICITY CONSUMPTION IN INDONESIA Introduction: The Household Energy Survey 1.1 In order to quantify the potential for more efficient household energy use in Indonesia, it was necessary to carry out a detailed survey of energy and e.lectricity use, under the Urban Household Energy Strategy Study (UHESS). While this survey is described in detail in Appendix 1, we review some of its important features as they pertain to electricity use, efficiency, and conservation potential. 1.2 The UHIESS survey was based on a sample of 2700 households chosen at random from the existing SUSENAS sample frame. The sample was stratified by urban area size so as to be representative of all of Urban Java. (Previous urban surveys reached only larger cities.) The main questions covered household size, total household expenditures, information about cooking and other energy uses for fossil fuels or biomass, and detailed questions about electricity using equipment and consumption patterns. These questions covered electricity tariff type, appliance ownership and characteristics (size, age, features like no-frost), average wattage and/or hours of use, why appliance bought, and who decided about buying the appliance. The survey also recorded total electricity consumption (as well as consumption of other forms of energy) over the previous year. The answers to all of these questions varied significantly among sample households. 1.3 Other aspects of consumption that were investigated include hours of use of appliances, wattage of appliances, age of appliance, and whether nor not the interviewed household engaged in sharing of electricity (as a buyer or seller) with neighbors. Estimates of usage and age were fraught with uncertainty, but these were used as clues to the average electricity use for each appliance. 1.4 Table 1.1 shows important characteristics of the different consuming households on Java. According to the UHESS, approximately 15% of surveyed households had no source of electricity, while 22% obtained electricity from neighbors. 5% were connected to PLN through the "social" tariff, while the majority of the remaining homes were connected to PLN through one of the tariffs R1-R4. (Tariff types are explained below). About 1% of homes surveyed either bought electricity from non-PLN generation sources, or used batteries. Thus the present survey goes beyond previous efforts through coverage of a significant number of homes that are officially "unelectrified" yet obtain electricity indirectly from the grid. 1.5 Weighting the sample to the entire population proved difficult. The correct weightings of the numbers of consumers in each tariff group that yield total urban electricity use were elusive, since previous surveys, as well as PLNs own definitions of rural and urban used for their own billing, used different definitions of urban and different choices of cities. Since consumption varies greatly between different tariff groups, total electricity consumption for the entire population, which depends on the weightings of the different tariff groups, is also uncertain. As a result, the extrapolation of the results from this survey to obtain total consumption on Java may be uncertain, giving a total electricity use that is 10% higher or lower than the total sales as recorded by PLN. In fact, differences in the relative numbers of customers by tariff group probably Table 1.1: Urban Residential Eleccricity Consumption by Tariff Category PL PLN Totat All Non-Etec Indirect Si RI R2 R3 p R4 Direc Elec Households 1. Sauple Size 393 589 146 1,279 242 8 1 1,676 2,309 2,702 2. Share of HH in Survey lx) 14.5X 21.8% 5.4% 47.3% 9.0% 0.3% 0.0% 62.0% 85.5% 100.0X 3. Share of Conmex in Survey (X) - - 8.7X 76.3% 14.4% 0.5% 0.1% 100.0% 1 100.0% 4. FamiLy Size 4.1 3.8 4.6 5.1 5.9 6.4 6.0 5.2 4.7 4.6 S. Averase Consumption (kWh/no) 0 15 35 60 183 1,190 2,527 84 64 55 6. Number TSelling8 Electricity - 0 37 274 24 0 0 335 335 335 7. Estimated Sales/Seller (kWh/Mo) - 0 17 24 66 0 0 26 26 26 8. Average Net Cons. (kUh/io) 0 15 31 55 177 1,190 2527 79 60 52 9. Family Expenditure (Rp/No) 69,820 103,285 156,964 171,248 316,336 599,571 976,061 193,844 156,012 143,476 v 10. Rp/No for Electricity 0 2,360 2,824 5,238 18,146 172,032 382,739 8,098 6,447 5,509 11. Rp with Fixed Charge - - 4,924 7,338 20,246 175,112 385,419 - - - 12. Share of Exp. for Electricity - 2.3% 1.8% 3.1% 5.7% 28.7% 39.2% 4.2% 4.1% 3.8% 13. Share with Fixed Cost - - 3.1% 4.3% 6.4% 29.2% 39.5% - - - 14. Maxmumn Load, kVA - - <0.20 0.25-0.5 0.5-2.2 2.2-6 >6 - - - 15. Fixed monthly Charge - - 2,100 2,100 2,100 3,080 2,680 16. Rps/kWH, 1988 Tariff - - 0 70.5 84.5 126.5 158 p At least 1 R3 Household and the R4 Household, were engaged in business activities. - 120 - accounts for a great deal of the difference between the present results and those in an earlier PLN survey, which found significantly higher electricity use per customer. But the present survey still describes in detail the consumption patterns of different households which taken together, represent the situation in 1988 on Java. 1.6 The number of homes using PLN electricity as supplied by neighbors is very large, nearly 22% of all households surveyed. Table 1.1 shows which households "sold" electricity, about 20% of Si and Rl households, and 10% of R2 households (It is likely that R3 and R4 households are found only in neighborhoods of virtual saturation of electricity connections, hence there a, e no "customers" for these households. Sales by SI and Ri are limited by the low value of connected load that these customers are permitted. The consumption of the buyers/users was estimated, not billed, but was thought to be very low, around 15 kWh/mo. This electricity, which was billed to the Si, RI, and R2 customers, is reallocated to the 25% of households with electricity who buy from PLN customers in our own analysis. This means that electricity use per home with electricity is approximately 25% less than electricity consumption per electrified PLN customer, another factor accounting for differences between the present survey and previous surveys. Indeed, when this factor is applied, the consumption per official connection, which first appeared high by international standards, falls close to levels experienced in other urban LDC areas. It appears. therefore, that selling of electricity between households has an important impact on the interpretation of the survey results. Main Findings: Household Electricity Use in Urban Java 1.7 We examined first the consumption in households as a function of the tariff they use. Consumption among PLN customers, as shown in Table 1.1, varies greatly between tariff groups, although the majority of users consume less than 75 kWh/mo. The number of high (Tariff groups R3 and R4) consumers is very small, but their consumption is very high, which raises the average per electrified household to 85 kWh/mo among PLN customers. This value is high by international standards for urban LDC consumers. The high standard deviation is not unusual and is intrinsic to the way residential electricity use differs even among similar households But the small number of R3 and R4 households and the large scatter among R2 consumers limits our ability to probe patterns further using tariff category as an important strata. 1.8 The ownership of appliances varies significan.ly among these tariff groups. This is not surprising, since average household expenditures, which indicate a household's ability to afford appliances, vary significantly from one tariff group to the next. Among households obtaining electricity from neighbors ("non-PLN" in Fig. 1.1), appliance holdings are very low. Ever. among Si (social) customers, only TV was found in more than 50% of homes (mostly BW), while irons and fans were found in less than 30% of homes, and heavy appliances (water pump, refrigerators, A/C, washers) in less than 2% of homes. In RI-type households, small appliances are more common, but large appliance holdings are still limited (refrigerators, 10%; pump, 9%). Recall, too (Table 1.1) that the connected load for SI and RI customers is very limited; these homes cannot have many large appliances running at any one time. Ownership of appliances in the smaller R2 group is significantly higher than in RI. But appliance ownership among households surveyed that used the - 121 - R3 tariff (or among those from an earlier PLN survey that sampled households on the R4 tariff) resemble patterns in S. Europe in the 1960s and 1970s. respectively. TabJl 1.2: Electricity Consumption by Expenditure Group Average Monthty Expenditures Avg/Tot Unknown <75 75-120 120-185 185-295 2495 Rupiah (t000) /IN/No Sample Size 2702 259 654 581 577 383 248 Share of NH in Survey t00.1X 9.62 24.22 21.5X 21.4X 14.2X 9.2X Share of Connections 85.7 8$6.0X 65.02 85.02 96.02 97.02 100.02 (Share in PLN Survey) - 55.02 50.02 66.02 82.02 90.02 94.02 kWh/Year 59.9 47.4 25.7 38.7 58.4 88.5 172.6 Family Size 4.7 4.2 3.5 4.8 5.1 5.7 6.1 Setters 12.52 10. 10.02 13.02 17.02 15.02 6.02 Sates, kWh/mo a/ 3.2 22 18 22 28 34 39 Consumed, kWh/mo 56.7 45.2 23.9 35.8 53.6 83.4 170.3 Average sales is averaged over alt households, while expenditure categories are average sates per selling household. 1.9 The households in the survey can be grouped by the household's monthly expenditures for all items. This averaging allows study of the variation of consumption with expenditures (closely correlated with income), and allows us to compare electricity use across incomes, which in turn is useful for forecasting. At the same time the grouping allows us to look at consumption as a function of a continuous variable, income, rather than as a function of the tariff class. As Table 1.1 shows, the sample contains very few R3 subscribers, and only one from R4. The results of this expenditure stratification are shown in Table 1.2, which also shows the structure of selling and buying of electricity. Grouping by expenditures has the effect of separating the high users from others in the R2 category, because these consumers are in general those with the highest levels of expenditures. 1.10 Not surprisingly, appliance ownership and electricity consumption are both strongly dependent upon income, as represented by the level of household expenditures. Figure 1.2 suggests that the relationship between expenditures and electricity consumption is almost linear. The ownership of appliances among the two top expenditure groups resembles that for Italy in the 1960s, although the level of expenditures, approimately $350/mo, is lower than in Italy in the 1960s. Interestingly enough, selling of electricity occurs most frequently among the middle-expenditure groups, but is not absent from the highest group. Even after subtracting selling, however, consumption per household appears high by international standards, given the levels of expenditures and the stocks of appliances. - 123 - 1.11 To better understand how billed electricity use per household is distributed over the appliances held, the survey estimated monthly consumption per appliance from the size or capacity of each appliance and from its estimated daily usage. Regression analysis was then applied to the entire survey. This procedure, descrbed in Appendix 1, yielded estimates that are shown in Table 1.3 for customers within each income class. For some appliances (water heaters, air conditioners, pumping) monthly electricity use was judged on the basis of international experience. 1.12 When these estimates are compared with figures from other countries: (a) For their size and features, refrigerators consume an usually high amount of electricity. The high levels imply that refrigerator motors run most of the time, and that the refrigerators are inefficient; (b) Electricity use for ironing, TV, and lighting is high, which may represent both high levels of activity (i.e., hours of ironing or TV) as well as inefficiency of the devices; Table 1.3: Electricity Consumption per Appliance by Expenditure Group (kWh/mo) Unknowin <75 P/ 75-120 120-185 185-295 >295 Average Lights 305 256 322 392 466 722 391 Color TV 140 107 146 173 173 177 167 Black & White TV 78 67 74 78 78 81 76 Video 100 100 125 150 150 150 145 Iron 83 74 84 95 115 132 101 Refrigerator 653 443 567 571 534 630 586 ,ump 200 209 278 322 360 506 387 Fan 24 36 48 48 72 72 48 Washer 0 0 96 60 96 286 242 Freezer 0 0 0 0 0 8400 700 Rice Cooker 105 54 112 81 150 166 139 Hot Plate 100 100 100 100 100 100 100 Oven 106 54 85 82 150 244 172 Water Heater 0 0 0 0 0 812 812 Air Conditioner 2667 0 0 0 0 1175 2667 A/ Expenditure categories are average monthly expenditures in thousands of Rupiah/Household/Month. Figure 1.1: Appliance Ownership on lava, UHESS Survey SHARE OF HOMES OWNING 70% 60%- 60%- 40%- 30%- 20%- 10%- 0%- A Cir TV SW TV Video Iron Fridge WaterPump Fan APPLIANCE TYPE NON F PLN ZALL JAVA Figure 1.2: Monthly Urban Residential Electricity Use and Expenditure, UHESS Survey MONTYHLY ELECTRICITY USE (EXCL SALES) 200 160 100- 60- 0 _ 0 100 200 300 400 600 MONTHLY EXPENDITURES, 100ORp - 125- (c) Unit consumption for most purposes is likely to be higher in higher income (upper tariff) households, because these house-holds own larger appliances. 1.13 It is always possible that these unit consumption levels are overestimated. If this were true, however, the "miscellaneous" or unaccounted-for consumption would have been much higher than is shown in the summary tables (Appendix 1). But this would imply that either the survey omitted important end-uses, which is not the case, or that the few miscellaneouses uses were grossly underestimated. Both of these possibilities are ruled out by experience in other countries (Schipper et al. 1987). We believe therefore that these unit consumptions estimates are reasonable. The picture presented here then gives an accurate portrayal of household electricity use in urban Java and allows us to conclude that electricity use in homes in Indonesia is unusually high. 1.14 From the survey and the estimates of unit consumption we can derive the shares of total electricity use devoted to the major end-uses. These are illustrated in Figure 1.3. It can be seen that for all households in the survey, lighting is the most important use, followed b", ironing, then TV, then refrigeration. Use for water heating and cooking is small because few households have major appliances that provide these services; use for washing is also minor today because the few washing machines utilize ambient water with no heating. But this aggregate picture is misleading. If we estimate the shares of electricity used for each purpose for each expenditure group, the shares of use for each purpose vary considerably (Fig. 1A). For low income homes, lighting and TV still dominate, as expected. For the highest incomes, refrigeration, pumping, and "other", which includes cooking and water heating and cooling (shown in Fig. 1.5) increase in importance. (Water heating appears to be a minor end use, but we expect it to increase as it has in other warm developing countries, such as Brasil). A conservation program must be aimed both at uses that predominate today, as well as uses that will be important in the future. 1.15 Using average electricity prices in each tariff group, we calculated the ratio of expenditures on electricity to total expenditures (Table 1.1). This ratio rises rapidly from the level of non-PLN households (2% of expenditures) to R2 households (6%). One reason is that the ownership of appliances rises rapidly as households approach the level of expenditures of R2 households. Electricity consumption patterns among R2 consumers represent those of homes that have had connections long enough to accumulate appliances. The patterns of the R3 group are heavily influenced by the use of air conditioning (even more so for the R4 customers surveyed previously by PLN; the one R4 in the present survey is an anomaly, and may have been running a business). High consumption in R3 is suggestive of elevated levels of use for lighting (2400 kWh/year, or twice the U.S. level) and refrigeration (1000 kWh/yr per refrigerator, close to the U. S. level). Such high levels imply considerable room for more efficient use of electricity. 1.16 The comparisons across tariff group are reflected by comparisons across expenditure groups. The R2 group has an average expenditure level midway between those of the two highest expenditure groups (as tabulated by the present survey). Appliance ownership levels in these homes are comparable to those in many other urban areas of developing countries. This comparison suggests that the R2 group, or alternatively, the groups with the highest expenditures, represent models for consumption patterns we can expect to arise in Indonesia as incomes grow. That is, household electricity use will grow rapidly during a certain period in which a home is rapidly acquiring minor and then major appliances. A doubling of household income in Indonesia (14 years growth at 5%/annum), if applied to all levels of incomes/expenditures, would allow the "moderate" Figure 1.3: Urban Residential Electricity Use by Major End Use, UHESS Survey LIG(I II-S 49'.1 1 V/E3VII; I C) tgOTT /ER N ~~10.6 A/C, FANS IRONING \-- WAl1ER P'MP 13.9 FREFRIG. 6.1 8.7 Cooking, Water Heating, Miscellaneous Figure 1.4: Household Electricity Use for Common Appliances, UHESS Survey ELECTRICITY USE/MO 150- 100 50 156 Unknown 61 96 149 228 478 MONTHLY EXPENDITURES, 100ORp - LIGHTS I T TV L T IRONING M REFRIG I --- PUMP LI ALL OTHER Figure 1.5: Household Electricity Use for Uncommon Appliances, UHESS Survey kWh/mo 20- 15 ........ 10 - I i. ixx.:: W Fi 166 Unknown 51 96 149 228 478 MONTHLY EXPENDITURES, 100ORp MI VPHHER IZCOOKING EVQER HEAT A/C 10U1FAN 8a1es of Electricity Excluded - 1Z9 - income households in the present survey (as shown in Fig. 1.4 with mean expenditures of 149,000 Rp/HH/mo) to reach a level of expenditures midway between the two highest expenditure groups. But the electricity use of 'high and very high" income households is more than twice that of "moderate" income households. In other words, use appears to grow more rapidly than expenditures (or income). Consumption patterns of the two highest expenditure groups should therefore be included as "targets" for any program that aimed to increase electricity-use efficiency. International Compaisons 1.17 An international comparison of these usage figures puts the Indonesian patterns in perspective. Fig. 1.6 shows estimates of average annual electricity use by purpose for urban Java, the United States, Japan, W. Germany, and Italy (from Schipper et aL 1987 with space heating excluded). By the European standards, electricity consumption per household in Indonesia is high, particularly according to Geller, (private communication) new 2001 refrigerator/freezers in Brasil were tested to consume nearly 1200 kWh/year, almost 3 times what we would expect from experience in Japan or Europe. This finding makes the high electricity consumption recorded in the survey households plausible. It appears that Indonesian appliances may be less efficient than those available in Japan or European countries. 1.18 The hours of usage of irons, lights, and TV may be unusually high compared to european households. Much of this appears to occur at night. The load profile for all of Java shows a sharp increase from 2000 to 2500 MW near dusk, falling off rapidly towards midnight (as the TV stations shut down). The shape is strongly suggestive of a large residential night-time load. It appears, however, that every family keeps at least one el"tric light burning through the night. A single 40W bulb left on for 12 hrs per day will consume 175 kWh/yr. On the RI tariff, this consumption costs 12,270 RP/year. not a trivial amount. These high usage factors may represent poor management or careless behavior. 1.19 Other factors tend to increase household electricity use in low income countries. Household size (4.7 members, vs. approx 2.5 in Europe) and presence of servants accounts for some of the high consumption in Indonesia. Among the households in the UHESS, family size increases with expenditures, from 3.5 people/hh at 51,000 Rp/month to 6.1 people/hh at 478,00ORp month. (Household expenditures tend to increase as the members age and advance within their jobs; the number of children increases over time as well, which accounts for some of the relation between household size and expenditures.) Family routines and traditions, such as sharing electricity among neighbors via connections or by allowing neighbors or inviting other family to use appliances in Figure 1.6: Urban Residential Electricity Use in ndonesia and Major OECD Countries ELECTRICITY USE PER HOUSEHOLD, MWH /YEaR 10 ...... 4 2 _ _ JANA MIDDULJA HIGHESTJAVA ALW GERMANY SOUS 86 JAPAN 86 ITALY 86 VER HEAT - COOKING ] LIeHTS REFRIGERATION lL A/C OTHER APPLIANCES - 131 - households that are electrified also increases household use. It is also likely that some "homes" may be businesses taking advantage of lower rates as residential customers. But it is diffi'cult not to conclude that on average, residential electricity use in Indonesia is high compared with the number and kinds of appliances present in homes. This means that use is inefficient. 1.20 An efficiency program that encouraged more efficient appliances (and more careful use) could affect electricity use by changing the first two of these three components of high electricity use, technology and behavior/management. Gradual evolution of lifestyles in Indonesia, as families become smaller and more consumers have their own electricity and appliances, will likely reduce high use caused by the third factor. In the following chapters, we investigate in detail how such a program might be structured. 1.21 The present survey addresses electricity use in urban Java alone. Since Java represents the greatest concentration of wealth (and appliance availability) in Indonesia, the survey results do not reflect consumption patterns in other parts of Indonesia, or in rural areas. Indeed, the PLN urban survey confirms that homes in other parts of Indonesia tend to use less electricity than homes in Java, even controlling for Tariff class. But the appliances are basically similar in all of Indonesia, although size and features may vary around the country. The basic uniformity of the appliance stock, however, implies that the same inefficient use that apparently characterizes urban Java may apply to other regions. Similarly, efforts to improve the stock of appliances in Java would doubtless have an impact on other islands. Therefore, the thrust of our discussion, while based primarily on survey results and other information from Java, will have general validity in all of urban Indonesia. * 132- H. PRESENT SrrUATION IN INDONESIA: EXISTING AND NEW APPLIANCES SuRvey: Existing Appliances 2.1 Large household appliances are still uncommon in urban Java. According to the survey (Table 2.1), refrigerators are found in 12.5% of all urban households in Java with electricity, Table 2,1: Percentage of Households in Tariff-source Category Using Each Device Non- PLN Total Total Etectrified Indirect S1 RI R2 R3 R4 Elec. Sampte Radio/Tape Recorder NA 32.32 47.92 60.0X 77.32 100.02 100.02 53.7K 45.92 Iron NA 13.22 23.32 62.52 89.3X 87.5X 100.0X 49.32 42.22 Black & White TV NA 22.22 43.22 48.32 21.52 12.5X 0.0 38.52 32.92 Color TV NA 2.52 11.62 24.02 73.12 87.52 100.02 22.82 19.52 Fan NA 5.12 11.02 22.42 53.3X 87.52 100.02 20.5X 17.52 Refrigerator NA 1.42 2.1X 10.22 57.02 87.52 100.02 12.52 10.72 Water Pump NA 0.32 1.42 8.72 45.02 50.02 100.02 10.02 8.52 Mixer NA 1.42 9.62 8.22 31.82 37.52 100.02 9.12 7.72 Video NA 0.2X 0.72 4.42 33.12 62.52 100.02 6.32 5.42 Sewing Machine NA 0.72 0.02 4.02 12.02 25.02 100.02 3.82 3.2X Sound System NA 0.22 1.42 4.02 12.02 25.0X 100.0X 3.72 3.2X Hair Dryer NA 0.2X 0.7X 2.42 19.82 50.02 0.02 3.7X 3.12 Blender NA 0.32 0.02 2.72 16.92 37.52 100.02 3.62 3.02 Rice Cooker NA 0.02 0.72 1.62 16.52 37.52 100.02 2.92 2.52 Washing Machine NA 0.02 0.02 0.22 9.9X 12.52 100.02 1.32 1.12 Freezer NA 0.2X 0.02 0.92 7.02 12.52 0.02 1.32 1.12 Toaster NA 0.02 0.02 0.52 3.7X 25.02 0.02 0.82 0.7X Air Conditioner NA 0.0 0.02 0.02 2.1X 25.02 0.0 0.32 0.32 Water Heater NA 0.02 0.02 0.02 1.22 12.52 0.02 0.22 0.22 Electric Oven NA 0.02 0.02 0.02 0.8K 12.52 0.02 0.12 0.12 * 13i - washers are found in 2.5% of all homes with electricity, A/C in 0.3% of homes, TV in 58%, and water pumps in 10% of all homes with electricity. Fans, irons, and small kitchen appliances make up the other group of appliances found in 20% or more homes. Present residential electricity use in Java is dominated by lighting, TV, and irons. 2.2 "his pattern is changing rapidly. Sales of major appliances are up. In the long run, the ownership e.f the heavy appliances should continue to grow, and the electricity consumption of these devices wzill come to dominate total household electricity use. It is important, then, to consider appliances that are becoming popular, not simply those already in many or most homes. Establishing the pattern of present use, as well as the sales volume and characteristics of new appliances, is a necessary step in finding the most important targets for improving efficiency of electricity use. 2.3 To evaluate the present and future appliance market, five major Japanese/Indonesian appliance companies were interviewed, as was Philips. Official sales data were obtained (estimated from imports and production in a variety of Indonesian trade and production statistics), complemented by data provided by NationaL by Sharp (covering ali Japanese companies in Indonesia), and from a survey commissioned b: a non-Indonesian company contemplating entering the market there. These sources showed little agreement on the numbers of appliances manufactured or sold in Indonesia, nor did they indicate the share of purchases by households (as opposed to small stores, restaurants, or other establishments). Thus the figures reported here are approximate. One important trend revealed by the non-Indonesian source is that the size of the market shrank between 1982 and 1985, presumably in response to the lower oil earnings in Indonesia. 2.4 Table 2.1 and Figure 2.1 review ownership patterns of major appliances in urban households. In comparison with the figures for total stock, the sales of refrigerators, freezers, washers, pumps, and A/C are significant: ownership is growing rapidly. Improving the efficiency of most existing appliances is difficult, although proper maintenance and careful use will reduce electricity needs. Improving the efficiency of new appliances relative to those already in place will have a more significant effect on future electricity consumption, because (i) the new appliance stock is growing rapidly, and (ii) new appliances tend to be considerably more efficient than older ones already in place. Lighting 2.5 Lighting is found in virtually every electrified home. It appears that most homes have at least one fluorescent light, as well as 4-5 incandescent lights ranging from 25W to 40W each. According to Philips (with 50% of the market), fluorescent lighting is more common in rural areas where grid connection wattage is low, and even PL and SL type lamps are found. The survey indicates that at least one light is kept on for 12 hours by each household. Refrigerators and Freezers 2.6 Refrigeration equipment is found in about 12% of urban Java homes (about 700,000 homes), according to the UHESS Study (but nearly 39% of homes with PLN electricity in all urban Indonesia, according to the earlier PLN survey.). In the R2 tariff class, saturation of refrigerators * 134- jumps to 53%, and is near 90% in the R3 class, but for most homes saturation is 10% or less. However, the most popular models today are small (< 150 liters) and selling to households in the lower tariff classes. Sales of these one-door models approached 70,000 in 1987. Thus refrigerator ownership is growing rapidly. Even among 2-door models ownership is growing rapidly. Freezers are only found in 1% of Urban homes in Java, but with sales of almost 10,000 units/year (some to commercial establishments), their importance is also growing. The total market appears to be around 140,000 units in 1987, down somewhat from 163,000 in 1985. 2.7 Although their share in total electricity use is still small, refrigerators are the most important electricity consumer among maor appliances in Indonesia. In ger.eral, one door refrigerators (many with small freezer compartments), one-door freezers under 175 liters, and two- door refrigerator-freezers under 200 liters are assembled in Indonesia. One-door models range in size from 50 liters to almost 200 liters, and range in price from 450,000 Rp (the best selling 135 liter Amaryllis of National) to over 4 million Rp for the best and latest 3- and 4-door 400 liter types imported directly from Japan. The one-door models often have small freezer compartments, UHESS which keep food for a few days, while the larger locally produced two-door types include full freezers, some of which are rated at three stars, an indication that the freezer can freeze fresh food and store it safely for many months. Japanese-made models that have up to 5 doors are also available; each door covers a compartment at a certain temperature and humidity. The market is dominated by the one-door types, however. The typical new refrigerator sold in Indonesia in 1988 holds roughly 150 liters of food, with one-third of all models having two doors and full freezers. By comparison, average new models sold in Europe today lie between 300 and 400 liters, and nearly 2/3 are the two-door type with full (three star) freezers. Refrigerators up to 200 liters in size are assembled in Indonesia. 2.8 The average lifetime of a refrigerator is less than 10 years, according to National. The first owner keeps it for an average of 5 years. It is then sold back to a service center, or traded for new one, and then resold until spare parts are no longer available, i.e., < 10 years. This pattern applies to other major appliances. 2.9 The one-door machines have small freezer compartments, but these cannot freeze fresh food, and can only keep already frozen food for a limited time. The two-door machines, by contrast, have true freezers, particularly the better models. Two, three, and four door models are imported from Japan, but in 1985 these appeared to represent only 10% of the total market, a share confirmed by figures supplied by National for 1987. In the UHESS, respondents indicated if they had a freezer either as part of their refrigerator or as a separate device. But they did not say whether the freezer within the refrigerator was contained in a separate compartment with its own door or not. 2.10 Manufacturers have introduced 'electricity saving" features on their better two-door models. These features include rotary compressors (now with 3 year guarantees and increasingly with circuitry that protects compressors from changes in line voltage), capacitors to limit the initial spike of current when the machine turns on, and a fan system to circulate air better (the so-called jet system of National). Figure 2.1: Appliances Ownership by Expenditure Group, UHESS Survey SHARE OF HOMES OWNING 100% 80% - 60%- 40%- 0% LEILLF IL Cir TV BW TV Video Iron Fridge Pump Fan APPLIANCE TYPE, HOUSEHOLD EXPENDITURES I!iJ AVERAGE HES (156Rp) M LOWEST (475Rp) EDI MIDDLE (120-185Rp) HlGHEST ,295Rp - 136- Air Conditioners 2.11 Air conditioners are found in only 0.3% of UHESS survey homes (3% of Urban Indonesian homes, according to PLN). Like refrigerators and freezers, however, smaller A/C are selling rapidly (with a significant but unknown number sold to smaller stores, offices, and other non- residential buildings with localized air conditioning needs). Recently, National introduced a small model drawing only 300 W which, in spite of its low cooling power (3000 BTU/HR, the smallest on the market), is selling well to households with limited wattage. Given the warm climate in Indonesia, we expect small and medium sized A/C units to grow rapidly in popularity. 2.12 Two kinds of A/C are sold and used in Indonesia: window type systems that cool one room only, and split systems that may cool more than one room. In a split system, the fan and controls are mounted in the wall, or on a console on the floor adjacent to the wall. The compressor sits outside the house, connected to the fan by a hose. The window-type systems are locally made, of considerably older technological vintage, relatively low-cost, and have energy efficiency ratios (EERs) between 7 and 11. The split systems, most of which are imported from Japan, are of very recent vintage, more expensive (but featuring many sophisticated controls including timers), but have EERs that range from 10 to nearly 15. Most important, the split systems are far quieter (indoors) than the window units. Typically a unit runs all day and has a rating of 750 W. 2.13 The quality of A/C is improving. Window units dominate the market, but the split types are growing in popularity. The lifetime is 4-5 years. Most window-type A/C are assembled in Indonesia, while the more advanced (and generally efficient split-type) are imported. The compressor and electrical components are imported. Washing Machines 2.14 Most washing machines sold and used in Indonesia are of the Japanese type, top loading and connected to ambient temperature water supply. No energy is used to heat water. Electricity powers the agitator/centrifuge and pumps. The centrifuges spin at up to 1200 RPM in the best models, considered optimal by manufacturers since this speed removes about 50% of all moisture in clothes. However, about 13,000 washing machines were imported from Italy in 1985. These presumably have coils to heat incoming water, the practice in Italy. They may become more popular in the future, which could more than quadruple electricity use for a single wash. 2.15 Washing machines were found in only 1.5% of the surveyed households, almost exclusively in those with R3 connections. Since approximately 45,000 washers were imported or assembled in 1985, however, their numbers are growing significantly. Most experts referred to the low cost of household servants as the main reason why washing machines were so uncommon. Energy consumption was estimated at over 100 kwh/yr by PLN; by contrast, similar machines are thought to consume far less electricity in Japan. However, much of the difference between these figures could be caused by differences in the number of washes per day. Television 2.16 In Western countries, television is not a significant component of household electricity use. In Indonesia, however, television is responsible for 7% of consumption, and video - 137 - for another 1% (Figures from PLN are even higher). TVs are found in 37% of homes (B/W) and 23% of homes (color) respectively, video in 6% of homes. National and other companies explained that the wattage of TV has decreased by nearly 50% during the past 10 years. Offsetting this trend has been a gradual switch from B/W to color. In 1985, net purchases of TV lay at approximately 500,000 for BW and 230,000 for color, but according to National, the BW market has shrunk considerably while the color market has held steady. Color TVs consume at least 2-3 times more electricity than B/W. But B/W TV still commands an important share of the market for rural consumers (low wattage 14' TV can run from batteries or even solar photovoltaic systems) and consumers wishing to limit their connected loads for other reasons. 2.17 Most television sets are assembled in Indonesia. The average size is 20'; for B/W 14" is the most common size, as it can be run from a 12-volt car battery in homes without electricity. In 1988, 14" color units make up 50% of total sales. Following the world-wide trend, TV electricity consumption has fallen. A 14' model needed 80 Watts 5 years ago; now 57 W; a 21 inch model required 130 W 5 years ago; now 79 W. The big market for small, efficient TV is the group of households with the <450 VA connections. Water Pumps 2.18 Water pumps represent an important use of electricity, accounting for nearly 6% of kWh in the UHESS sample. Unfortunately, no figures were available on total sales. Irons 2.19 Irons represent the most widespread of all significant small appliances, being found in over 50% of surveyed households and nearly 2/3 of those on the RI tariff. According to the UHESS survey, power consumption ranges from 200-750W. If ironing requires 200kWh/yr, then ironing claims a 13% share of household electricity use. Reasons for popularity and high use center around the importance of cotton and silk fabric, which require ironing. Official sources give the 1985 production/imports as near 200,000, but National set the market at nearly 700,000. Fans 2.20 Because of their low cost, room fans are far more popular than A/C in all classes of households. Connected wattages are usually well below 100 W. Our best estimate of average annual kWh consumption is around 50 kWh. According to official data, almost 1 million were produced in Indonesia in 1985 (780,000 according to National). Rice Cookers 2.21 Although only present in 3% of homes, rice cookers are also gaining in popularity. We estimate rice cookers to use less than 100 kWh/yr, giving them a small share (<0.5%) of total kWh sales. But in most other Asian countries these have become very popular among urban families, so their presence in homes in Indonesia can be expected to grow. Sharp imports its models from Thailand; other companies are importing models from Japan and Korea. - 138 - Other Appliances 2.22 The market for other appliances is small. A few thousand water heaters are sold every year (from Australia, Italy, Japan, and the U.S.), all as imports. Most run on gas, but electric storage heaters were found in stores. A few hundred dishwashers and dryers are also imported. Microwave ovens are now available and will probably become popular among upper-income households, in part because they use less energy than other cooking devices, and thereby heat up kitchens less. Sales of video recorders and audio tape decks are also growing, some of the latter assembled in Indonesia. While water heaters, dryers, dishwashers, and cookers represent important energy uses in the distant future should they become popular and affordable, video, audio, and other small appliances are not expected to ever be significant users of electricity. Issues in the Appliance Market 2.23 This review of the appliance market has raised several important issues that must be confronted if the electricity efficiency of appliances sold and used in Indonesia is to be increased. The most important issue, the role of the manufacturers, is treated later in this chapter, but several other issues remain. 2.24 The Two Market System. There are two appliance markets in Indonesia. Expatriates and well-to-do Indonesians buy high quality imported appliances, many of which are among the most energy-efficient in the world. Because of their high ownership of large appliances, these families have the highest consumption of electricity. By contrast, middle-class Indonesians face a simpler market dominated by locally made appliances of clearly lower quality and lower prices than is the case for imports. In the National Gobel Catalogue, for example the popular Amaryllis refrigerator (1 door, 135 1) sells for less than 500,000 Rp. National's imports, while holding less than three times the volume, cost roughly 5 times as much or more. The Amaryllis resembles a Japanese refrigerator of 1970s vintage, but the two-door models imported from Japan by most companies resemble more recent models sold there. Even the two-door locally made models look old, a fact confirmed by discussions with company representatives. 2.25 Quality and Durabflity. It was clear from direct observation that the imported models are not only bigger, they are better made, and probably use less electricity per unit of volume than locally made models. (Our only available yardstick, connected wattage, confirmed this suspicion, but wattage itself is a poor measure of average electricity consumption.) Companies report that locally made refrigerators (and other appliances) "last" about 5 years in first owners' homes. Thereafter, these are either traded back to dealers for new models (and resold), or sold directly as second-hand models. The used models then are resold for up to five more years, beyond which spare parts become difficult to obtain. The short residence time with first owners suggests that quality is low and deteriorates rapidly, although more models are offered with guarantees. 2.26 Energy Efficiency. As noted elsewhere, measures of energy efficiency of TV and A/C were obtainable from catalogues. For refrigerators, freezers, pumps, and washing machines, no measures were available. Our overall impression, however, formed by inspection of appliances in stores as well as catalogues, discussions with local representatives, and discussions with Japanese manufacturers, suggests that the efficiency of appliances assembled in Indonesia is lower than those imported from Japan and, most important, lower than necessary. - 139- 2.27 Discussions with the local manufacturers did suggest that appliances made today in Indonesia are more efficient than those made five or 10 years ago. If, as it appears, most appliances made in Indonesia follow designs developed in Japan 10 years ago, then this trend towards greater efficiency should continue, albeit at a slow pace. At the same time, new appliances tend to be larger and offer more features (color TV vs Black and White; larger freezing compartments, through the door ice making). Thus some or all of the impact of electricity-saving improvements of new appliances may be swallowed by other changes in appliances. But these changes merely make improvements in electricity use efficiency more attractive to consumers as the intrinsic service delivered by each appliance grows. 2.28 The efficiency of electric appliances made in Indonesia is improving slowly, but is still low by international standards. Increases in size and features of most appliances tend to raise appliance electricity use, but also increase the economic attractiveness of improvements that increase efficiency. Improvements that could accelerate the slow process of efficiency improvements in new appliances in Indonesia are discussed in the next chapter. First, however, the structure of the appliance manufacturing industry in Indonesia is briefly reviewed. Anpliance Manufacturers in Indonesia 2.29 There are two markets: local manufacture from outsourced components or knock down, and imports, mostly of upper range of market. Main local manufacturers of electric appliances are licensed by Japanese companies. In order of market importance, these companies are: GobeL NationatlNatsushita Rf C. P. W, TV, EL Yosanta Sharp Rf C, W, TV, EL Sanyo Saryo R, C, P, W, TV, EL Lippo Nelco Mitsubushi Rf C, TV, EL WitiSaptma U. Toshiba Rf C, TV, EL Argaska Kesuma Hitachi Rf C, TV Dfaikin Dalkin C w Polytron Polytron/Phitips TV, EL some Rf U Note: R, refrigerator; C, A/C; P, pump; WU ctotheswasher; EL, electronics. 2.30 The first seven companies are Japanese, while Polytron is a collaboration between Philips and a local company. Ariston and Modena (Italy) gas and electric cookers are sold here. Rheem (Australia) electric and gas water heaters are also available. Recently, Gold Star (Korea) and Taiwanese companies have entered the market. A few large American products (Philco, General Electric) are available; we found an attractive display of large American appliances in a well-cooled marketplace in the center of Jakarta, for example. According to information from Gabungan Elektronika, however, the Japanese companies dominate the market. 2.31 The two markets referred to above are important for understanding the evolution of appliances in Indonesia. The three largest companies produce most of their products in Indonesia, particularly one- and two-door refrigerators, freezers, washing machines, window-type - 140- A/C, and TV. Components come from all the Asian countries, not just Japan. The designs are Japar.ese, often of relatively older vintage. (For example, small one-door refrigerators are very uncommon in Japan today.) 2.32 The market shares of each company vary by appliance type. For example, National sold nearly 50% of all one-door refrigerators in 1987, but only 10% of all large (> 200 1) two- and three-door refrigerators, a market that imports, and Mitsubushi dominate. However, about 110,000 1 door refrigerators (and two-door types smaller than 200 1) were sold in 1987, vs. 11,000 large two- door types. Thus National dominates the overall refrigerator market, with Sanyo running second, Mitsubishi third, and Sharp fourth. In the A/C market, National had 45% of the market in 1985, with Sanyo, Daikin, and Mitsubushi around 10-15% each, and Sharp at 8%. 2.33 Local assemblers increasingly are choosing a variety of countries for components. Rotary compressors (from Japan, Taiwan, Singapore) have been introduced in the better refrigerators and A/C. Circuits for TV come from Taiwan, Korea, Singapore, and Thailand. These countries are also providing an increasing share of "Japanese" imported products. The Role of Foreign Producers of Electric Appliances Sold in Indonesia 2.34 The most important characteristic about the electric appliances sold in 1988 in Indonesia is that they are imported or produced (ie., assembled) from imported parts. Almost all of the brands imported are controlled by Japanese companies, although products from Taiwan, Korea, Singapore, and Thailand are gaining in importance. Some devices are imported from Italy (cookers and washing machines), Sweden (small hotel refrigerators), and other European countries, and even from the United States. We discussed the Indonesian market with three major companies, one in Europe, and two in Japan. 2.35 From discussions with these manufacturers, it is evident that progress in improving efficiency of appliances in Indonesia does depend on the market approach, planning, and technical approach of the overseas companies that control the appliance markets in Indonesia. This control was described in somewhat negative terms by a senior appliance expert in Indonesia, who maintained that it was very difficult to "Indonesianize" Japanese-controlled appliances. The reason for the difficulty was lack of interest on the part of the technical divisions of the Japanese companies. 2.36 Another important conclusion arises from contrasting the European company (Electrolux) with the Japanese companies. Electrolux enters a market when it can sell European levels of quality in appliances, while the Japanese companies appear to search their older catalogues for models (and costs) appropriate to current consumption and style in Indonesia. Philips, a competitor to Electrolux in Europe, had planned to introduce a refrigerator assembly line in Indonesia, but backed out at the last minute and moved the apparatus to Algeria. Probably, the reason was that Philips decided it would be difficult to manufacture and sell European-standard equipment in Indonesia or other neighboring countries.) The comparison is not meant to fault the Japanese companies, only to highlight the importance of understanding company strategies as these directly or indirectly affect electricity efficiency of products made or sold in Indonesia. * 141 - 2.37 The role of foreign companies is both a barrier and a potential booster to electricity efficiency of appliances in Indonesia. Until recently, the appliance companies did not appear to put much emphasis on electricity use in Indonesian appliances, simply passing on efficiency improvements as they arrived from Japan. More recently, "hemat listrik" ("save electricity") has begun to appear in brochures and on stickers on models seen in stores, but little or no checking of the effectiveness of the technologies offered (or indeed the meaning of the label) has been done. The slow approach to electricity saving appears conscious on the part of manufacturers. 2.38 At the same time, a stepped up campaign to foster a climate where increased efficiency is rewarded (testin& labellin& etc.) or even required might act as a meaningful stimulus to accelerate savings in Indonesia. This is because the most important suppliers have already learned how to make efficient appliances in other countries (ie., Japan). If the rewards for accelerating these improvements in Indonesia were clear--information, voluntary standards, etc.-- the foreign companies could draw immediately upon their experience with products that save energy. Purely local firms would have a more difficult time learning new technologies. (The author found that the same was true with energy saving in industry and commerce. Multi-national firms in Kenya and Thailand are well supplied with energy saving ideas and technologies, while local firms in these places have considerably less access to knowledge, hardware, and the appreciation of expertise.) The importance of foreign firms is thus both a hindrance to greater electricity-use efficiency, but potentially a blessing in the long run. - 142 - m. POTENTIAL FOR IMPROVING THE EFFICIENCY OF ELECTRICITY USE 3.1 There are three components of the electricity requirements for each end-use. One is the technical design efficiency of a device, such as the ratio of electricity required to heat removed by a compressor in a refrigerator. The second is the actual operating efficiency of each appliance, which may be influenced by maintenance or by ambient conditions. The third is the utilization factor, i.e., the number of hours per year the device is used. A fourth factor, family routine, is discussed elsewhere but not considered here as an "efficiency' factor. This chapter focuses on technical potential, but potential savings from other changes are also discussed. After reviewing these potentials, the main barriers to improved efficiency are discussed. 3.2 Electricity prices do play an important role in our considerations. Prices affect the economic attractiveness of a particular electricity saving option. They affect manufacturers' perceptions of this attractiveness, their willingness to invest in improvements and to market them. Electricity prices affect consumers attention with regard to choosing more efficient appliances, and affect consumers' behavior with regard to use and management of electricity use. For example, electricity use per household for lighting in Norway is at least 2 times higher than in Denmark; electricity prices are at least 2 times higher in Denmark. The low electricity price in Norway blunts consumer interest in advanced lighting or even in fluorescent bulbs. In the following discussion, we will assume that average prices will rise at least to those that the ADB finds reflective of present- day marginal costs, about 50% higher than current average residential prices. 3.3 Experience in Brazil supports this reasoning. Geller compared refrigerators made in Brazil with those produced in Europe and Japan and estimated that better American models consumed less electricity than smaller but typical Brazilian models. Geller et al. (1988) noted that the most efficient models cost 10-15% more than average models, but estimated that for Brazil, this extra cost paid back in less than three years. He found this strategy far less costly than supplying the equivalent amount of electricity from new powerplants. Since we have essentially made the same kinds of comparisons, we believe that our estimates of conservation potential and cost- effectiveness will prove to be reasonably accurate. Significant to the economic comparison, Brazilian residential electricity prices in December 1986 (at then exchange rates) varied from $0.048/kWh to $0.053/kWh for monthly consumption of 30-100 and 100-200 kWh/month, respectively. These prices were lower than those in effect in Indonesia at that time. Since wage rates in Indonesia are lower than those in Brazil, it is likely that the extra costs of greater efficiency in Indonesia will also pay back against Indonesian residential electricity prices. 3.4 There are many studies that document past improvements and future potential improvements in electricity-use efficiency. Schipper et al. (1987) found that changes in household electricity use in OECD countries between 1972 and 1983 (currently being updated to 1986), pointed to improved efficiency as a major factor in reducing growth in demand. Thorough study of the United States (Turiel, 1986; Geller, 1988), experimental appliances built by Noergaard and colleagues in Denmark (Noergaard 1987), and experience in Brazil (Cepel, 1985; Geller 1986; Gelier, Goldemberg et al. 1988) quantified both the electricity savings that could arise if specified improvements are made in the appliances that characterize the U.S., Europe, and Brazil, respectively. For Japan, the improvements since the early 1970s in refrigerators, air conditioners, and TV are well documented (Energy Conservation Center, 1988). Since the UHESS indicates that - 143 - appliance unit consumption (electricity/device/year) in Indonesia is high by international standards, we believe the studies of other countries can be applied to Indonesia. 3.5 Although lack of measured consumption data from Indonesian appliances makes precise quaitification of the cost of improving them and the savings that would result, our inspection of available specifications, and comparison with appliances sold elsewhere convince us that large savings could be designed into Indonesian appliances at low cost. Geller (1988) and Turiel (1986) addressed the U.S. situation in generic ways that apply to appliances made or sold anywhere. In general, the kinds of improvements that we recommend are features that both of these experts find cost effective and technically practical today in the United States. Many of these features have been introduced in Japan or Europe without meeting consumer resistance. That is, purchase of the better Japanese appliances today in Indonesia would lead to electricity saving, compared to locally made appliances with similar size and features. Presently, however, the imports tend to be larger or more powerful. But the features that make the imports more efficient could be applied to simpler domestic models. 3.6 Some of the potential for improved electricity use in Indonesia is immediate. The best selling air conditioners in Indonesia, for example, are considerably less efficient than the most efficient ones available, but few consumers know how to buy efficiency. For refrigerators, confidential information made available by the Japanese parent of one of the major producers of appliances in Indonesia documents the decline in electricity use in that company's refrigerator- freezers since the early 1970s. Since the models they manufacture and sell in Indonesia resemble Japanese models from the mid-1970s, it is likely that their Indonesian-made models could also be improved rapidly. Therefore, we believe that the potential described here for increasing the efficiency of appliances sold in Indonesia is realistic, even in the near term. 3.7 Based on the above considerations, we present below our estimates for the technical potentials for increased electricity-use efficiency in the major appliances likely to be used in Indonesia in the next twenty years. These estimates were made by comparing unit consumptions (electricity/household or per device) of existing appliance in Indonesia with those in other countries, notably Japan, W.Germany, and the United States, and careful consideration of technical and economic studies of conservation potentials in these three countries. The estimates are based on material by the Lawrence Berkeley Laboratory for the U.S. appliance efficiency program and surveys of appliance use in major OECD countries, from material Geller has collected for the U.S., and Europe, and from Geller's own estimates of conservation potential in appliances in Third- World countries. (See Schipper et al. 1987). Technical Conservation Potential 3.8 Consumers can reduce electricity costs today by choosing more efficient appliances from among those being offered, and manufacturers can improve the efficiencv of all appliances in the next few years. In this section we review briefly the technical options. 3.9 The refrigerators and freezers imported from Japan appear to be as efficient as those sold in Japan. Those assembled in Indonesia appear to be considerably less efficient. Unfortunately, available information (wattage) does not give a reliable estimate of total energy use. -144 - (Fig. 3.1 shows wattage vs. size for 2+ door refrigerators.) However, a consumer would get a better, more efficient machine if he/she: (a) bought one with a rotary compressor; (b) bought one with at least 2" of insulation in the door and sides; (c) bought one with a low wattage skin heater. 3.10 These features tend to be found in imported appliances or in the best locally made appliances. If all locally made refrigerators had rotary compressors and improved insulation, for example, electricity use in locally made refrigerators would be reduced by at least 20%. 3.11 To improve efficiency of all refrigeration equipment, all boxes should be made with rotary compressors (now being added on better models). The efficiency of motors and the coefficient of performance of the compressor itself should be improved. Insulation levels should be increased to 3", the door seals should be improved, and temperature controls improved as well. Manufacturers should consider putting the compressor/motor above or at least towards back of the device and increasing access of air to these elements (for better cooling). Claims that "flat back", or covered evaporator coils, save energy, may have some merit, since the benefit of using the heat from these coils to drive away moisture (illustrated in a Sharp catalogue) is increased by setting these in thermal contact with entire back of the refrigerator. (On the other hand, fresh air cannot easily blow across the evaporator coils.) Other more advanced improvements are listed in the LBL analysis for the U.S. Department of Energy.) In summary, we believe t}at electricity use in refrigerators could be reduced by 40% using the best designs available today, and 60% by the year 2000, using advanced technology. For freezers we put these figures at 50% and 65%, respectively. 3.12 Careful consideration of characteristics of other appliances offers important opportunities for saving electricity: 3.13 Air Conditioners: Currently the efficiency of A/C, as expressed by the EER, can be determined from catalogues. Buyers can switch to the highest EER in each size category and save 10-50% of their electricity needs for a given amount of cooling. (Fig. 3.2 shows the EER vs. capacity for most models sold in Indonesia. Note that EER appears to fall with increasing model capacity, after peaking among the 7000-9000 BTU/hr models.) Even among locally made window A/C, the buyer can save electricity by buying the best EER available, either by choosing an import or by choosing only from locally made products. Electricity-saving features il clude timers and thermostats, rotary compressors, and the ability to operate the fan without using the compressor. Split-type systems cost more than window-type, but tend to use significantly less electricity. 3.14 To improve the efficiency of A/C manufactured in Indonesia, all models should have rotary compressors, thermostats and timers, and, by the mid-1990s, variable speed motors. The heat exchangers could also be improved. It appears that the average EER today lies around 9.5 as measured in Japan. This value should be raised to at least 13-14 by making the improvements recommended here. The fact that EER of larger models is lower than that of smaller ones suggests a significant potential for improved efficiency in these models, too. In the U.S., EER tends to increase with larger capacity. Thus much can be done to improve air conditioners sold in Indonesia. - 145 - In summary, installation of the best akr conditioners avaibble In Indonesia would reduce unit consumption by 40% compared to today's estimated values for households using air conditioning; installation of advanced technologies after the year 2000 would permit a 60% reduction. 3.15 Washers: Improved motor efficiency and continuous water level adjustments (for part loads) should appear on all models. The water use should be reduced, particularly in models that use heated water. But reducing water use in non-heated models saves electricity by reducing the load that the agitator has to move. To avoid need for drying, the spin cycle in washers should reach 1200 RPM. (Manufacturers of European washing machines reduced electricity use for the motorsbra by about 33% between 1973 and 1985, and reduced electricity (or gas) requirements for heating water by more than 50% in the same time frame. Similar improvements were made in dishwashers.) In summar, the best models today would permit a 20% reduction in electricity use; advanced models ready after 2000 would use 35% less electricity than those in use today. 3.16 Television: Currently buyers can choose a 14" TV using as little as 42W (rather than 70W); similar choices are available in all size categories. (Fig. 3.3 shows the variation in wattage among popular models, as a function of screen size.) Instant on features should be avoided, remote-control on/off chosen instead. In summary, the most efficient color TV would use 25% electricity than the average sets used today; advanced designs will reduce electricity use by 50%. 3.17 Water Heaters: As the popularity of storage water heaters grows, these should all have heavy insulation, at least a level of R-9 (in U.S. units), which corresponds to about 10cm of mineral wool. A side benefit of insulation is that the tank heats the house less. In summary, the best models for water heaters offer a 20% savings compared to typical models. Advanced models, which combine solar heat with electricity, wil offer a 70% reduction over present day demand. 3.18 Electric Ovens: Convection-type ovens use less electricity than conventional ovens; all ovens should be insulated to the standards of European or N. American self-cleaning ovens. In summary, the best models will allow a 25% reduction in electricity use over typical models. Advanced technology, combining microwave ovens and conventional ovens, will allow 50% reduction in electricity use for ovens. 3.19 Irons: Our inspection of irons suggests that the predominant kinds do not use water, and transfer heat poorly to clothing. This impedes the ironing and slows the process considerbly. Higher quality irons have water fill and good thermostats, which reduces both electricity consumptior and the time needed to carry out the ironing. Improved irons will require 15% less electricity than today's; By the end of the century, we believe these savings will Increase to 25% 3.29 Lighting: The UHESS found that nearly 35% of all bulbs in place were fluorescent types. These bulbs are underrepresented among those burning for more than 6hrs/day; thus even in existing homes, there appears a potential for saving by moving bulbs around where this is practical. However, the wattage of fluorescents bulbs, according to the survey, is high compared with the wattages given for incandescent bulbs, given the greater output and efficacy of fluorescents vis-a-vis incandescents. Figure 3.2: Energy Efficiency Ratio vs. Capacity of Room Air Conditioners in Indonesia Energy Efficiency Ratio, BTU/WTT-HR 14~~~~~~4 12 - +-4 10 -_ . 2~~~~~~~ + + + +~~~~~~~ + + 8 i + 6 0 6 10 1i 20 25 Capacity, '1000 BTU/hr Figlre 3.1: Wattage vs. size of 2 Door + Refrigerators in Indonesia RATED WTTAME 260 200- 44 160 _ 1 100 4-41-4-+ t++ + loo ~ ~ - ++ ++ + 60 -~~~~~4 60 0 100 200 300 400 500 SIZE, Litre Figure 3.3: Wattage vs. Size of Color Television in Indonesia RATED VOTTAGE 140 120 - 100 * 80 * 60~~ ~ ~~~~~~~~~~~~ ;0 40_ 20_ 0 5 10 is 20 26 30 SIZE, Inches Diagonal - 149 - 3.21 The economics of switching from incandescents to fluorescents are very attractive, as are the incentives to switch all the way to compact fluorescents (as noted in a previous World Bank report), but the higher first cost of fluorescents, particularly the compact types, deter consumers. Compact fluorescents allow as much as 70-80% reduction in electricity use relative to incandescents, depending on the consumption of the ballast. The compact fluorescents have moved very slowly into the residential markets of Europe and the U.S., principally because of their high cost, but their long lifetime (5000 hrs vs 1000 hrs for incandescents) provides as great an incentive to using them as does their electricity saving. Geller reports improvements that may both lower costs and improve performance. In summary, inreased use of conventional flourescent lamps, and use of one compact flourescent In the most Important lamp in each home could reduce electricity use for lighting by 25%; by the end of the century, compact flourescents could account for more than half of the bulbs and 80% of the light In Indonesian homes, resulting In a 50%o reduction in electricity use for lighting. 3.22 Available data make a precise estimate of the savings from greater efficiency in new electric appliances impossible. Nevertheless, we summarize in Figures 3.4 and 3.5 our best estimates of present day use for key appliances, along with the two levels of use that would obtain from higher efficiency. Again, the first level is based on estimating what the best appliances currently available in Indonesia (or in countries exporting to Indonesia) would consume relative to the actual mix of appliances in use in Indonesia, while the second estimate incorporates our judgement as to improvements in appliances that could occur in the next twenty years. We have also made estimates of potential savings for a few other appliances (video, fans, rice cookers, pumps), mainly through use of better motors and electronic controls. There seems little doubt that an enormous potential for electricity conservation exists in Indonesia. 3.23 The estimates we have made are based on assuming that the appliances consumers use, and the services they receive, wou'd remain constant over time; only the number of households and the penetration of these appliances would increase. We have not accounted for the likely increases in the levels of services demanded by future Indonesian consumers, i.e., larger refrigerators and water heaters, hot water in washing machines, wide-screen, high resolution TV, increased water pumping. Taking these changes into account would raise the base-line of electricity use, and raise the benefits of greater efficiency as well. These estimates yield considerable savings for most appliances, even before we consider future technologies. If future ownership levels are considered, savings are even larger (see Chapter V). 3.24 The testing and labeling program that we recommend later will allow analysts to more carefully determine consumption levels in present-day new models. By comparing the actual components of these models with those in more advanced models in other countries, analysts can then make more precise comparisons of incremental costs and savings as well. Behavioral/Management Strategies 3.25 Purchasing an efficient appliance is the most important way to evoid high electricity costs. But there are important management strategies that also reduce operation costs in existing (and even new, efficient) appliances by 10-20%. These include: cleaning and maintenance of refrigeration and cooling equipment, particularly removing dust from evaporator coils; turning ' ff lights, A/C, fans, TV, etc. when not in use (and using timers); using irons, washers, and other Figure 3.4: Electricity Use of Common Appliances in Indonesia: Present Stock and Advanced Models CONSUMPTION PER APPLIANCE OR HOME,KWH/Y 700 00 '88 Unit Consumption B Beat '88 Appllanoe. 600 ] Advanced Appliances. 600 400- 300- 200 100 Lights Clr TV SW TV Video Iron Fridge Pump APPLIANCE TYPE Figure 3.5: Electricity Use of Uncommon Appliances in Indonesia: Present Stock and Advanced Models kWh/device 3600 3000 - 2008, U8S BEST. 2600E 2008, ADiRNCED TECH 2000. 1600 1000 600 0 Washer Fresr Rlce Cooker A/C Oven Wbter Heat APPLIANCE TYPE * 152- equipment to fullest capacity (turning on an iron to iron one piece of clothing may double the electricity required to iron; running a clotheswasher with a full tub of water and only one or two pieces of clothing uses almost as much electricity as running the washer full). 3.26 For air conditioning, terhnology and behavior combine in ways that can yield greater savings. For example, use of a timer and thermostat allows the consumer to turn back the A/C when asleep or when gone for a short time, as well as to turn the A/C off when gone for a long time, knowing that the A/C will come back at the desired time. Equipping A/C with a variable speed motor allows extra capacity for rapid cooling, allowing consumers to turn A/C off even for shorter periods of absence. 3.27 An additional management effort can improve the efficiency if all equipment using motors and compressors, namely better control of line voltage. If PLN improves its voltage regulation, or homes are equipped with voltage regulators, then motors and compressors will be strained less when voltage drops. (According to National, this voltage drop causes serious problems in Indonesia.) Barriers to More Efficient Electricity use in Indonesia 3.28 From observations of the way in which electricity is used in Indonesia, and the way in which appliances and other electricity-using technologies are marketed, we offer the following summary comments on barriers to more efficient electricity use. Information 3.29 lIpical consumers have little information on how much electricity appliances use, particularly how much one model will use compared with another. Stickers on appliances, or slogans in catalogues and advertisements make claims about electricity savings which, whether true or false, cannot be substantiated or evaluated as to their cost worthiness by buyers. Well educated (and presumably well-off) appliance buyers and users can usually figure out which devices or use patterns save electricity, but the majority of users (particularly future users) probably lack the sophistication to deal with the meager information available in Indonesia today. The labelling .nd testing programs we recommend should help this situation greatly, as would the "appliance literacy" campaign we suggest. Buying Habits and Environment 3.30 Studies at the Lawrence Berkeley Laboratory showed that in the United States, appliance buyers tend to discount future savings of electricity by very high rates. The higher the discount rate, the shorter the time horizon. In other words, their time horizons are short, far shorter than wolnld be justified by simply applying store or credit-card rates of interest (and certainly shorter than justified by the savings rates offered by banks). Part of this situation is caused by the nature of the buying environment. Salespersons are uninformed about energy, buyers are in a rush, stores are noisy and crowded, men buy appliances but women use them, etc. Consideration of these problems weighed heavily in the U.S. decision to place minimum efficiency - 153 - standards on products such as refrigerators, because it was felt that pushing manufacturers was easier than pulling consumers. Nevertheless, labels on appliances, and information in stores and advertisements, would improve the buying environment greatly. (One informal survey carried out by LBL compared sales of two similar refrigerators by the same company. One refrigerator used 33% less electricity than the second, but cost approximately $50 more. In a state where labelling of appliances for their electricity consumption was required, the more efficient model was chosen by 50% of buyers; elsewhere, only 33% opted to invest in the more efficient model The payback time was less than 3 years.) First Cost and Quality 3.31 In most analyses, more efficient appliances have a slightly higher first cost. This is both because some improvements do cost more (larger heat exchanger surfaces, more copper instead of other materials, etc.) and because more efficient appliances tend to be of a higher quality. (Indeed, we observed that there were really two appliance markets in Indonesia, one for the well-to-do, one for the majority of consumers.) But some manufacturers turn the marketing of quality to their advantage by 'gold plating". This means that the higher efficiency features are only available on top of the line models, even though the costs for the efficiency improvements in particular may be minor. Goldplating is an understandable marketing strategy, but goldplating causes serious problems for electricity efficiency when consumers'incomes for appliance purchasing are iimited. 3.32 Whether efficiency improvements are built into all models or not, greater efficiency is tied to better quality in the long run. As consumers' incomes rise, they will demand bet, -r (and more efficient) appliances. To accelerate these improvements, one must face slight increases in first cost of appliances to cover the costs of energy-related improvements (including the industry's costs for learning and retooling). These costs will probably meet with some consumer resistance, and certain resistance from marketing managers. Yet the kinds of improvements we discuss herein (or the kinds consumers can enjoy from simple changing their choices of what to buy) pay back in a few years and "cost" less than expansions of the electricity-supply system. It is therefore very important to move both consumers and manufacturers of appliances beyond the first-cost barrier that has so often hindered application of improved technology to electric appliances. 3.33 The acceleration of progress towards greater efficiency through programs, whether voluntary or mandatory (as explained below) does entail the risk that manufacturers and suppliers will raise prices. Manufacturers may justify such price increases by using costs of greater efficiency as an excuse (i.e., beyond what is actually justified by the changes in manufacturing costs and equipment features). Or real cost increases may be larger than necessary because of the cost of retooling manufacturing equipment before its useful life has expired. Both these cost problems can be minimized by publicity that allows consumers to compare costs (and energy use) of similar models that are more and less efficient, by encouraging competition in the marketplace, and by allowing producers sufficient time to retool and reposition themselves in the marketplace. The government could offset some of these real and illusory cost increases by permiting more favorable tax treatment on production and sales of the most efficient models. -154 - Electricity Prices 3.34 According to recent World Bank analysis, most private consumers pay lower prices than would be justified by costs of maintaining or expanding the system. If this is true, then a powerful incentive to more efficient choice and use of appliances is weakened. Proper pricing of electricity is an important element in any attempt to increase the efficiency of electricity use. Other Problems 3.35 One potential problem we have discussed is the fact that most decisions about the technologies used in household appliances are made by foreign companies, often by experts in Japan or other countries where parent companies reside. These companies may resist efforts on the part of GOI to accelerate efficiency improvements. However, if a program is presented to companies who are given at least a year or two of warning about testing, efficiency goals, incentives for efficiency, demonstration programs, or even standards, the manufacturers will have time to adjust their production and marketing strategies to the new goals of electricity-use efficiency. 3.36 Such a presentation implies that good lines of communication are open among the actors who play a role in determining how appliances are made and used. While we found no open antagonism among the actors, we found that appliance companies, PLN/LMK, DJLEB (or other government groups), YLKI, and appliances dealers had little interaction among each other. An electricity-use efficiency program must foster greater communication, or else an individual group can either ignore the problem ("someone else takes care of that"), or 'solve" the problem on its own without fitting itself into the fabric that holds all of these groups together. (The importance of weaving energy efficiency into this fabric, while strengthening the fabric itself, was underscored in a study of how residential energy use became so efficient in Sweden [Schipper et al., 1985].) - 155 - IV. REGULATIONS: TESTING, STANDARDS, AND NORMS FOR ELECTRIC APPLIANCES Regulation and Standards 4.1 In many countries, regulation of electric appliances through "standardsc has been extended to electricity-using properties such as efficiency. This option may be important for Indonesia. It is therefore worth considering what "standards" mean for electric appliances and related goods. 4.2 Standards are laws that affect properties of equipment which are deemed important to safety, durability, noise or other undesirable properties of equipment. Standards can also place requirements on overall performance. A given piece of wire should be able to carry a certain current before melting, or withstand a given strain, for example. Standards also specify tolerance limits relating allowable typical performance of products (or their components) to promised or design performance. For example, a standard could require that a given piece of wire withstand a certain temperature beyond its design tolerance, or be able to suspend a certain amount of weight. In other words, a standard should define the margin by which randomly chosen samples of any product could miss the stated performance criterion. 4.3 Standards can used to specify that certain properties of materials or certain characteristics of products be present in equipment. For example, standards can require certain amounts of thermal insulation be present in walls and doors of refrigerators. Such a standard could either specify thermal resistance values, or require that a certain thickness of commonly used material be present. Standard-setting through specification of detailed properties of a component or system is called "prescriptive", while the specification of overall properties of systems is usually referred to as "performance" standards. 4.4 Standards can also be employed to define how products are labelled and classified. In the United States, for example, there are different grades of beef, different kinds of drinks made with oranges, and, as in many countries, different sizes of TV. "Standardsc specify how the beef is to be graded, how the orange drinks are to be labelled by orange content, and how the TV screen is to be measured. Within limits, however, these standards do nothing more to the product (although other regulations may apply to health and safety aspects, for example). Thus orange juice makers can choose from products with little orange content to those made only from oranges. Only the latter may be labelled "orange juice". Currently there is an analogous, widely recognized labelling procedure for refrigerator/freezers that classifies the freezing ability of the freezer compartment according to temperatures attainable. This procedure leaves the manufacturer free to choose temperatures, but forces him to reveal his choice in a way that makes comparison among products easy. 4.5 It is crucial that the standard setting body specify what properties of materials or equipment are to be measured, how the measurements are to be carried out, and how the results of measurements are to be applied. For example, standard-setting bodies could require that manufacturers of washing machines test the capacity of each machine (in kg of dry clothes), the amount of water and electricity consumed for a given wash, and the amount of water left over in - 156- the clothes when the wash cycle is complete. (In the examples given above, the procedure for testing content of orange juice or the indoor temperature of the refrigerator compartments must be specified.) These measurement procedures are also commonly referred to as standards, but they really delineate methods and reporting procedures. No regulatory process of standards can succeed without such methods and procedures. 4.6 Standards that specify minimum performance parameters of household goods are not uncommon. For example, a performance standard could require that a bed be able to withstand a certain temperature before catching fire, without specifying what materials the bed were made from. Examples of performance standards that could be applied to electric appliances are requirements on size, volume, or capacity. For example, the government could specify temperature ranges that refrigerator and freezer compartments of refrigeration equipment had to attain. Or the government could state minimum electricity performance standards for appliances, which might specify the maximum energy use allowed for an appliance of a given size and performance. Clearly it is easier to specify prescriptive standards for simple wiring than for complex devices such as electric appliances, but in most instances it is easier to specify performance standards for complex devices than to try to prescribe the properties of each important component, particularly when energy or electricity is concerned. Regulations in Indonesia that Apply to Household Appliances 4.7 There are few regulations affecting properties or performance of electric appliances in Indonesia. Testing and standards do affect house wiring and light bulbs, however, and the government plans tests for ballasts for fluorescent lights. Electronic ballasts developed in Europe and the United States will be an important element in reducing electricity use for fluorescent lamps. For appliances, however, the government at present can only rely on test results from producers/importers, which may arise from local testing or from tests carried out in the country of origin. 4.8 One standard, SLI 005, could regulate properties of electric appliances, but nothing is yet applied to appliances. Some appliances are already tested for safety by LMK or by Lembaga Elektronik National LPMB (Ministry of Trade) tests quality of materials, while LIPI/LMK tests lamps, following the IEC convention. According to the Ministry of Trade, there is no specific regulation, quality control, or standardization of electric appliances for import or export, only for wires and light bulbs. The general view is that manufacturers are making their own tests for quality control. Since foreign firms dominate the appliance market, they are not easily "controllable" from Indonesia. 4.9 Thus there is little regulation of electric appliances at present. However, there is an arrangement for testing equipment that could provide the framework for either information/labelling or the background for efficiency standards. The laboratories that are used to test wire, the manufacturers themselves, or LMK could be asked to perform certain tests on appliances that would cover safety, performance and energy use. (We return to this proposal in the final section.) The history of standards for ordinary wire are worth reviewing. - 157 - 4.10 In the late 1970s, after a series of house fires, concern that poor quality wiring had contributed to a series of house fires led to establishment of performnce standards for wires. These were promulgated in 1980, with the manufacturers given approximately 3 years to comply. 4.11 Enforcement is carried out by random testing of wire. The government buys samples at ordinary outlets. They remove any means of identifying the manufacturer, then pass three samples to factories for testing. If the tests come back below a certain standard, then further tests are ordered at a national laboratory. Manufacturers themselves have to pay the costs of these tests. 4.12 Suppose a particular brand of wire is found not to comply with the standards. First, the company is issued an oral warning. If their standards do not improve, they are issued a written warning. If there is still no improvement, the written notification is circulated among the association of cable manufacturers. This third step is usually effective. 4.13 How could this procedure be extended to cover household electric appliances? For appliances, it is impossible to "de-identify" the manufacturer. Nevertheless, it is possible to have manufacturers test each other's products (which is often done in other countries out of competitive pressure) as a short run measure, while introducing a new institution to test appliances in the long run. In the final chapter we discuss this new institution more fully. Establishing Testing and Regulatory Standards for Electric Appliances in Indonesia 4.14 Minimum efficiency standards can be an effective tool for reducing electricity demands in new electric appliances, as the experience of individual states in the United States has shown. If the GOI were to pursue regulatory options that involve setting appliance efficiency standards or goals, GOI would have to establish quantitative standards covering important properties of electric appliances, establish or empower a recognized testing authority to carry out testing (or to pass on information from recognized test procedures carried out elsewhere), and ensure that a reporting system were established to both warn the government of products that do not meet standards as well as show consumers the amount by which acceptable products exceed standards. The government would then prohibit products that did not meet standards from reaching the market. 4.15 The key ingredient in any country's program for effilciency standards is the organization of authority for information and testing. In most countries, responsibilty for the elements of a testing/labelling/standards program are dispersed among different authorities. Each country has its own electrical equipment safety lab, without whose approval it is almost impossible to sell anything that connects to the mains. For electricity consumption, however, the "authority" lies vested (or simply has sprung up) in different kinds of groups. In Sweden, for example, Konsumentverket (KoV), an official body supported mainly by government (but which accepts a fee from manufacturers in order to carry out unbiased tests), tests some appliances (washing machines) for many consumer features including energy consumption. For refrigerators, KoV specifies tests to be carried out by manufacturers and reports these results. In Germany, Stiftung Warentest tests all devices for most properties of interest to consumers, while in the United States, Consumer's Union accepts manufacturer's energy-related tests, makes its own evaluation of other properties, and then publishes all results. The energy-related tests were initially established by the U.S. - 158 - Department of Commerce, National Bureau of Standards, but these have been superseded by those promulgated by the U.S. Departmert of Energy. 4.16 The U.S. example illustrates an important step to be considered by GOT. Indvidual companies using procedures established by authorities (i.e., the test procedures established by the U.S. Department of Energy, or by the Swedish Konsumentverket (for refrigerators) can carry out tests. The government or consumer authority must police these tests from time to time. The advantage of this approach is that appliances do not need to be bought and then resold. Since most manufacturers test appliances through quality control for some properties in Indonesia, extending such quality control to energy or electricity consumption should present no problems. The advantage of using an existing consumer/technical authority, on the other hand, is that such a group tends to be both well established technically, well respected by both producers and consumers and, most important, relatively independent of influence by these groups (or by the government itself). In short, an independent test authority and/or set of procedures are necessary for credible implementation of energy and electricity efficiency standards. But this step costs more than manufacturer testing. In the conclusion, we will recommend that the manufacturers initiate testing, to be taken over by a consumer or electricity authority once procedures are well established, understood, and proven to be practicable. 4.17 Although there are no appliance testing programs in Indonesia, the government, through DJLEB, is proposing certain safety standards for refrigerators. This proposal could be extended to cover electricity consumption properties. DJLEB would provide a list of components that must be specified by the manufacturer or seller/important of major appliances. These components include insulation thickness, wattage of the compressor, the fan(s), the skin heater, etc. of refrigerators or freezers, the efficiency of motors (refrigerators, freezers, washers), etc. These properties do not reveal directly the exected or actual electricity use, but they are useful to many observers who wish to compare devices. On the other hand, no performance testing is necessary; only a check to see if a motor of a particular rated efficiency, or insulation of a given thickness, is actually present in each device. 4.18 Actual electricity consumption, however, is the most important information for the consumer. In order to make performance comparisons meaningful, an authority must specify test conditions under which washers, refrigeration equipment, and other appliances would be tested in the factory. Such procedures are well established in virtually every country, although the conditions vary from country to country. These tests only approximate actual conditions under which new appliances will perform in homes. However, the tests (like the testing of automobiles for fuel economy) are usually reliable enough to show the relative performance of different machines. Information from these tests can be used to both follow progress within the industry, as well as to provide consumers with reliable indicators of the expected relative performance of different devices on the market. At present, the only device for which an indicator of energy performance is provided is A/C. 4.19 Two properties of electricity consumption are important to consumers. The first is the maximum power consumed by the device (in watts), and, where important, the average power as well. Maximum power is uf great concern in most households in Indonesia, because most homes pay for tariff that provides a relatively low total wattage to the consumer. (Overloading of household circuits is common.) Related to maximum power is the output under such conditions: - 159 - cooling capacity, washing load, TV picture and sound output, etc. The second figure of interest to consumers is the total electricity consumption under either typical or standardized operating conditions. This is because wattage is not always directly related to electricity consumption. 4.20 The most important measurements by appliance are summarized below. In addition, other properties that should be specified by manufacturers and tested occasionally are noted: (a) For television, maximum wattage is important to the consumer on a limited wattage connection to PLN, but average wattage is a better measure of energy use: average wattage times viewing time gives total use. Fortunately, average wattage tends to be proportional to maximum wattage, so a ranking of TV by size and maximum wattage is meaningful. (b) For refrigeration equipment, maximum wattage is not a reliable indicator of energy use, because different components (compressor, fans, light, skin heater, emullion heater [between the refrigerator compartment and the freezer compartment]) operate for different periods of time during the day. This means that each refrigerator must be tested to see how much is consumed. The energy consumption tests must be developed carefully. For refrigerators, for example, authorities specify the design temperatures for the refrigerator compartment, the freezer compartment (combined or separate), and the temperature of the room. A certain number of door openings per hour can also be specified. The machine is then run for a period of up to several days to measure its electricity consumption. (At present, there are no data on electricity consumption of refrigerators available in Indonesia, although careful comparison of imported models with Japanese catalogues allows comparison of these models tested under Japanese conditions.) The test allows the manufacturer to specify kWh/month of expected electricity use. (c) Manufacturers of refrigerators should also specify motor efficiency, skin-heater wattage, defrost cycle electricity consumption, compressor coefficienct of performance (efficiency), insulation thickness and values for walls and doors. (d) For washers, a test specifies the quantity of water and weight of clothes that are to be run through certain cycles. The manufacturer then specifies kWh/cycle, and estimates a standard number of cycles per month in order to arrive at monthly electricity consumption. If the washing machine includes a water-heating element, electricity consumption for heating water through each cycle should also be measured. Manufactuters should also specify motor efficiency, and water consumption for each cycle. (e) For air-conditioners, the test measures the amount of heat a given device removes from a given room under specified outdoor and indoor temperature and humidity conditions. Further tests show consumers how quickly a room of a given size can be cooled down from a given initial temperature under a given outdoor temperature. Finally, the noise level under different performance conditions can be measured. - 160- Manufacturers should also specify compressor coefficient of performance, motor efficiencies, and heat-exchanger efficiency under specified temperature and humidity conditions. (f) For water pumps, the test should measure electricity and power requirements to pump water from a given depth to the level of the pump, and then from below the pump up to a given height (the maximum depth and height are given in pump catalogues). The test specifies the diameter of the water pipe. Manufacturers should also specify the efficiency of the pump motor. (g) For light bulbs, the most important information concerns the efficacy (output in lumens, divided by power in watts), and the lifetime. These should be tested and specified together, since they tend to be inversely related. Because the line voltage varies significantly in Indonesia, bulbs should also be tested for lifestyle and efficacy under considerably higher and lower than normal line voltage. (h) Several tests can be applied to cooking appliances. For electric ovens, tests should measure the amount of electricity required to heat the oven to a specified temperature, and the time required. Then, the amount of electricity required to keep the oven at the specified temperature should be measured. A similar test should be carried out on rice cookers, measuring the consumption and time required to bring water to a boil and then to sustain the cooking rice for a given time. Manufacturers should specify the insulation values of the walls and door of the oven. (i) For water heaters, tests must measure the amount of electricity consumed to keep the water at a constant temperature. Then, after a given amount of water is withdrawn, the test measures the electricity consumed to heat up incoming water to the desired temperature. Manufacturers should specify these two quantities, the maximum wattage of the unit, and the level of insulation in the water heater. 4.21 At present, there is enough information available for consumers to choose wisely among A/C and TV, but no reliable information on refrigeration equipment, washing machines, or pumps. As noted above, the wattage of TV sets is published in catalogues. For A/C, manufacturers already specify the energy-efficiency ratio (EER) of virtually every model sold in Indonesia. This ratio is defined as cooling power (in BTU/hr) over wattage. Values for imported models are taken from manufacturers' specifications--one company even notes the EER with an outdoor temperature of 32°C and with one of 35°C. Since we found no evidence that local models are tested for cooling power, we assume that EER figures for these model--mostly window types--are either guessed from calculations or derived from specifications for the Japanese models on which the local models are produced. Information on efficacy and lifetime of lightbulbs is not easily obtainable. In no cases were any of these data supported by tests carried out in Indonesia. 4.22 Other appliances can be compared from existing product information and tested further. For irons, the wattage depends both on the model and the setting, so a test to determine energy use is not really meaningful. However, tests can compare how much of the electricity consumed is converted to heat on the bottom of the iron. For light bulbs, locally made incandescent bulbs have a poor reputation for lifetime and reliability. A test of output per unit of - 161- power (usually expressed in lumens/watt) would be useful, as this measure, called efficacy, can vary by as much as 20% among bulbs of a similar type. For water storage water heaters, manufacturers in the U.S. express the efficiency as the amount of energy lost in keeping the water in the tank at a certain temperature. The same figure of merit is used for ovens (keeping the oven at a certain temperature). In both cases, the level of insulation in the device determines the amount of energy required. 4.23 There are certain special conditions that characterize Indonesia that should ultimately be reflected in the tests carried out on appliances. First, ambient indoor conditions (temperature, humidity, etc., which affect energy use, product lifetime, rusting, etc.) are different from those in Japan or many other countries where products are made. These conditions require that some products be tested in Indonesia. Second, habits of Indonesian consumers (refrigerator door opening, washing machine use, particularly choice of detergents) should be reflected in energy- related tests. Finally, the relative instability of electric power voltage (i.e., both fluctuations and brown-outs) has important effects on appliance components, light bulbs, and, most important, energy consumption and lifetimes of most equipment. Tests should be established that "torture equipment to see how serious the effects of power instability really are. 4.24 As noted elsewhere, little of this testing is carried out in Indonesia, either for electricity use or for other consumer features. However, private companies do carry out their own testing of some features in order to maintain quality control and most of their products are tested and rated by their mother company in Japan. In requiring testing, one issue that arises is the treatment of smaller companies, particularly those that import, but do not manufacture or assemble products Indonesia. These companies would have to call on the services of an independent laboratory to have their products tested. An advantage of adhering to certain well-established testing norms, such as JIS (Japan International Standard) is that imported products from a variety of countries would likely be tested in country of origin (usually Japan, but possibly other Asian countries as well). If this were the case, companies importing products that are tested and labelled would only need to give the information available from the country of origin. - 162 - V. AN ELECTRICITY EFFICIENCY PROGRAM FOR INDONESIA Introduction 5.1 Data from ,.., present household energy survey, and examination of the appliances currently on the market in Indonesia, suggest that electricity is used inefficiently in Indonesian homes. This inefficiency has two components, 'uninformed behavior" and inefficient technology. This inefficiency costs Indonesia millions of dollars, much of which could be saved through a carefully designed set of policies and programs that aim to increase electricity use efficiency. In this chapter we propose such a program, review the benefits, and sketch out actual program elements. Similar programs have been implemented in Brazil (Procel, 1985. "Energy Conservation in Home Appliances". Informe No. 20. Rio de Janeiro: CEPEL), and proposed for Jamaica (World Bank 1988). This chapter discusses how elements of these and other programs could be applied to Indonesia. 5.2 Why is there concern about electric power in energy-rich Indonesia? PLN's capacity expansion plan calls for 12% growth in installed capacity through the year 2000 and will require roughly US$ 135 billion (constant 1988 dollars) in capital expenditures each year for investments in generation, alone. PLN foresees a nearly four-fold increase in the number of residential customers on Java by the year 2000 (to 8 million), and a similar increase out of Java to nearly 4 million. They also foresee a very slow increase in use/household, because the new customers will be small users. But appliance prices are decreasing in most parts of the world; it is possible that PLN has underestimated the impact of increases in appliance holdings in newly electrified homes, and therefore future household electricity demand. Meeting this demand will be expensive. Actions that reduce these outlays and save consumers money as well could therefore be beneficial to Indonesia's economy, as Geller argues for electricity-rich Brazil (Geller 1986). 5.3 The residential sector is a particularly important sector on which to focus efficiency programs. For one thing, residential consumers accounted for about 1/3 of PLN's sales in 1987. Yet most homes did not have refrigerators, washers, waterpumps, or A/C. But appliance ownership is growing rapidly. Acquisition of these appliances will increase consumption per customer. Moreover, appliances are getting larger and more homes are getting their first appliances, which will also boost consumption. Air conditioning, which is just beginning to grow, will add to a daytime peak load. Finally, more marginal consumers (i.e., small load) being added to the network, will eventually increase their own consumption significantly. For example, consumers in two of the highest expenditure categories use about two times as much electricity/hh as the average of all consumers, and three to four times more than those in the two lowest categories. The appliance holdings of the highest expenditure households in Java resemtble those in urban areas in Brazil, Thailand, Singapore, and even SenegaL Since most households in Java will reach levels of income similar to those in these countries by the next century, we expect that the two highest expenditure groups serve as a model for the typical household in urban Java in 20 years. As real appliance prices fall, there is even more reason to expect that appliance ownership will increase rapidly in Java. Setting in motion a program that increases the efficiency of new appliances will assure that this growing group of consumers acquires efficient devices in the future. Not doing so could be very costly as residential electricity use grows. - 163 - 5.4 Based on the savings figures estimated in Chapter m, a 15% reduction in the demand of a typical customer today; or a savings of 100 kWh/yr could be obtained through appliance efficiency improvements. When multiplied by 4.6 million households on Java with electricity, this calculation yields 0.46 TWH/yr. The percentage saving in a home in the two highest expenditure categories is considerably higher, of course, because such a home is likely to have several of the major appliances. Since appliance ownerhsip is expected to grow rapidly, it is therefore more instructive to examine the impact of more efficient electricity use on a likely mix of appliaoces in the future. Table 5.1: Components of Future Electricity Use and Potential Savings A: Saturation; 3: Savings, Unit Consumption, and Use/Houshold Saturation UEC Use/HH Savings UEC Use/NH Save UEC Use/NH 1988 2008 1988 2008 Best'88 2008 2008 Adv Adv 2008 Lights 83.9% 100% 391 391 75% 293 293 50% 196 196 Color TV 19.5% 80% 167 133 75% 125 100 50% 83 67 Black & White TV 32.9% 10% 76 8 80% 61 6 60% 45 5 Video 5.4% 20% 145 29 80% 116 623 65% 94 19 Iron 43.9% 90% 101 91 85% 76 68 75% 66 59 Refrigerator 10.7% 35% 586 205 60% 352 123 40% 234 82 Pump 8.4% 25% 387 97 80% 310 77 60% 232 58 Fan 21.0% 90% 48 43 90% 43 38 80% 38 34 Washer 1.3% 10% 242 24 80% 193 19 65% 157 16 Freezer 0.1% 10% 700 70 50% 350 35 35% 245 25 Rice Cooker 2.3% 25% 139 35 80% 112 28 65% 91 23 Hot Plate 2.0% 10% 100 0 0 0 - 0 0 Over 2.7% 10% 172 17 75% 129 13 50% 86 9 Water Heater 0.1% 15% 812 122 75% 609 91 30% 244 37 Air Conditioner 0.3% 20% 2667 533 60% 1600 320 40% 1067 213 Other - 9.5% - - - - - - Total, kWh/yr - - 681 1798 - - 1236 - - 840 Total, kWh/mo - - 57 150 I 103 - - 70 Notes: Saturat ons: Shows ownership levels for 1988 (UHESS) and estimated 2008. UEC: Shows the Unit Consumptions, or kWh/year/device. Savings: Reductions in Unit Consumptions under "best", "Aovanced" cases. - 164 - 5.5 To do this, appliance ownership levels are projected to those that would be obtained if appliance ownership levels reached the average of those for the two highest consumption expenditure categories (see Table 5.1). (Exceptions: Lighting reaches 100% of connections, color TV, 90%, freezers and ovens, 10%, water heaters, 15%, air conditioning, 20%). Using these future ownership levels, consumption for each major appliance and for all uses are calculated under the assumption that: (i) consumption/appliance remains constant; (ii) consumption/appliance falls with the introduction of appliances that today are among the most electricity-efficient on the market; and (iii) consumption/appliance falls to levels deemed achievable with technologies that will be common after the year 2000. Figure 5.1 illustrates electricity consumption per household when each of the two sets of conservation "potentials" discussed earlier are achieved, and compares these results to present-day consumption and to the level that would be obtained if efficiencies were frozen. (Note the important growth in refrigeration and "other".) The figure shows that more efficient electricity use could reduce household electricity costs significantly. 5.6 To estimate the overall impact of greater efficiency on electricity sales in 2008, Figure 5.2 presents present consumption (based on 4.6 million households using electricity on Java in 1988), and future consumption based on an estimate of 12 million electricity-using households in 2008 (PLN estimates 8.1 million subscribers in 2000; our extrapolation assumes continued growth and includes non-subscribers using elecricity as well). The calculation, as shown in the figure, shows that increases In electricIty-use effciency permit a savings of 7.4 TWH If today's best models are chosen, and 11.5 TWH Uf advanced models are chosen. (About half of these savings will be realized by 2000.) Put another way, sales to the residential sector grow by 10%/year if efficiency remains constant, but by only 8% and 6%/year respectively if the two efficiency levels are reached. The difference between the base case and either of the two efficiency scenarios for 2008 shows the important role greater efficiency plays in reducing future growth. As noted previously, increases in appliance size and features could not only drive the baseline consumption up even more than illustrated here, but also increase the gap provided by savings. While the calculations are only illustrative, they show that the overall impact of greater efficiency can be large: PLN expects to sell around 18.5 TWH by the year 2000. More efficient household electricity use could reduce these sales by a considerable amount. As an important additional benefit, these savings will likely be replicated in the commercial building sector, which buys a significant number of refrigerators, room A/C, televisions, water heaters, and other appliances. 5.7 These savings affect both connected load (in watts) as well as total use. Room A/C load could fall to less than 500 w instead of 1000 w, yet still cool one or more rooms. Color TV could require under 50 W instead of closer to 100 w, while the connected load of lights could be reduced by about 100 w per household. The 500 MW night time peak of Java appears to be composed of 60% residential and 40% non-residential; it is conceivable that the residential peak load could be reduced by 30-40% (cooling, lights, TV). 5.8 The economic impact of more efficient electricity use is important to Indonesia. The marginal cost of electricity production is about 150 Rp per kWh, compared with average revenues from residential customers of less than 100 Rp. It has been estimated that the cost of saving about 3040% of the electricity used in new appliances would be considerably less than 100 Rp/kWh. 2/ The savings from greater efficiency could reduce much of the growth in residential kWh. Thus an intervention that promoted more efficient household electricity use could allow a given investment in PLN + efficiency to stretch further, i.e., electrifying more customers for a given - 165 - capacity. Unfortunately, however, there is much evidence that residential customers miss many or most opportunities for saving electricity (Ruderman et al 1987; Geller 1986; Schipper 1976), and that manufacturers of equipment, in turn, will not innovate further than consumers' buying choices dictate (Schipper et al. 1987). Thus, a program that accelerates efficiency improvements in electricity use may be called for. Such programs have stimulated efficiency improvements in the United States and other countries. 5.9 Higher electricity prices appear to be an unavoidable element in the future development of the electric power system in Indonesia, given the disparity between prices and costs today. Moreover, higher prices are an indispensable element to a successful strategy to cut waste in electric power usage. But higher electricity prices alone cannot "fix" a situation in which consLmers appear to be underinvesting in conservation and efficiency. This is because consumers tend to have very poor information about which appliances will cost more or less to use, particularly in Indonesia. Since consumers must also choose from other options not related to energy, such as color, features, etc., they may not be able to select the models they want and also select energy efficiency. Programs that improve marketplace information tend to give energy cost considerations more weight in the buying decision process. 5.10 Even if information were perfect, a more important problem tends to determine the outcome of market choices. More efficient appliances often cost slightly more than less efficient ones, but pay back the incremental costs in a few years. Consumers must therefore weigh higher first cost against lower running costs. But consumers have notably short time horizons. American consumers, for example, tend to select more efficient appliances with a 2-3 year payback or less in mind (Ruderman et al. 1987). This selection reveals an implied discount or interest rate of 40% real or more, far greater than the cost of consumer loans, interest rates offered in banks, or the rates of return required by investments in utility capacity. This revealed capital scarcity might be overlooked, but the capital to provide electricity capacity to cover the extra consumption required because consumers forego efficiency investments is a problem in Indonesia (as well as in the United States). Thus a package of programs that lengthen consumers' time horizons vis-a-vis the purchase of efficiency characteristics in new appliances w benefit any electric uti'ity short of capital. Geller shows that for Brazil, these potential capital savings are worth billions of US Dollars over the next decade or so. Similar benefits are possible for Indonesia. 5.11 There are other market imperfections that constrain consumers from buying efficient appliances and discourage manufacturers from producing more efficient devices. Appliance users are not always purchasers, particularly where well-to-do consumers live in furnished housing or employees use company flats. Consumers who already own appliances that break down usually rush to replace these with little time for considering alternative purchases. The nature of the market in Indonesia--many smaller stores, each with limited selections and often located in areas difficult to reach conveniently--makes it unlikely that consumers can shop around vis-a-vis energy, since they can bargain for a device in any store until the price is mutually satisfactory. 5.12 Manufacturers are hindered from improving their products by many factors, too. For one thing, the majority of assemblers/ manufacturers are saddled with old designs dictated by foreign-dominated companies. At the same time, there are almost no local innovators who could enter the market with applianccs that were truly adapted to the Indonesian climate, market, FiurS.w1: Impact of More Efficient Apphances on Future Electicity Use kWh/Month/use 160 140- 120 100 800toL 60 -~~~~~~~~~~~~~~~ 40 - 20 0 1988 2008 2008 Conservatlon 2008 dwanoed EFFICIENCY SCENARIO LieHT S TV/VIDEO IRON FRIDGE [11 PUMP El OTHER Flgures reflect ownership Ionia Figure 5.2: Household Electricity use on Java: Potential Impact of Greater Efficiency in 2008 Total Sales, TWH (Thousands) 25 20 0 ALL USES 20 16 10 6 0 1988 2008 2008 Best from 1988 2008 Advanced CONSUMPTION SCENARIO - 168 - electricity grid, and, most important, utilization patterns, all of which have an important impact on electricity needs. As long as the information about actual electricity use in any device is missing or incomplete, manufacturers have no need to improve their products, since there is no measuring stick against which to compare improvements. (Some improvements that allegedly save electricity are noted in some catalogues found in Indonesia, but the buyer cannot ascertain whether these features really save electricity.) If manufacturers see no real return from improving efficiency significantly, they have no rationale for doing so. 5.13 Hence, the marketplace for appliances in Indonesia is so imperfect that neither consumers nor manufacturers can be expected to take up many or most of the kinds of efficiency improvements that would be justified by the current (or future) price of electricity. A carefully designed electricity conservation/efficiency program could remedy many of these failures. 5.14 A reasonable policy goal is to intervene in such a way as to cause changes in electricity-use efficiency that are worth more (to customers, to PLN, to Indonesia) than the total cost of intervention. These interventions must be aimed both at consumer and at those who produce electricity-using equipment. The cost of intervention is both the cost of actual investments or behavior changes, the direct cost to the government and other institutions of the program, and the indirect political costs. However, many of the investments in greater efficiency would be undertaken over 30 years anyway, as products and consumers become more sophisticated. For example, appliance testing and labelling takes place in most developed countries, and is being implemented or proposed for developing countries as well (Brazil, Philippines). Therefore, the overall impact of a conservation program would be to accelerate the maturation processes by which manufacturers and consumers move towards more efficient appliances. 5.15 The choice of policy elements for Indonesia depends on the time frame. In the very near term, higher electricity prices would encourage consumers to use their equipment more effectively, perhaps reconsider how many lights should be left burning at night, etc. Although lower income consumers might feel the burden of higher prices more immediately, upper income consumers, whose electricity use patterns suggest wide margin for immediate savings, would doubtless respond as well to higher prices. Authoritative institutions--the Government, PLN, YLKI, could certainly advise consumers on ways to cut back on consumption by better use of equipment. 5.16 In the medium and long term, however, there is actually scope for greater savings. For example, information programs (as discussed below) could encourage consumers to purchase more efficient lamps, appliances, and other equipment. Such changes could take place within one or two years in Indonesia. If consumers responded, manufacturers would see a larger market for more efficient devices. And in the longer term (3-5 years and beyond), this response could be further reinforced by both new products developed by manufacturers in response to competition, as well as by publicized efficiency targets or even standards, which could be related to improvements realized in appliances sold each year. If such improvements were deemed to fall short of expectations (as defined in part by the costs of electricity and the investment requirements of PLN), the improvements could be accelerated by setting of minimum efficiency standards. Finally, consumers could be encouraged to choose more efficient appliances if the government itself, and other visible institutions (hotels, restaurants, etc.) also invested in more efficient lights, appliances, and other electricity-using devices. * 169 - Proposed Electricity Conservation Program for Indonesia 5.17 Implementing a conservation program requires important considerations of phases, timing, and actors. In this section, we suggest three separate but related kinds of interventions, break these into phases, discuss timing in broad terms and suggest some-but not all--of the actors who ought to be involved. 5.18 Following Wilson et al. (1988), we classify the proposed interventions as follows: (a) Information Intervention: Inform consumers about product choices and use. Test the consumption of new appliances; measure actual consumption in consumers' homes, to relate tested consumption to actual use patterns; display the tested consumption of new appliances prominently in catalogues, advertising, and labels that are actually attached to new appliances and the cartons in which they are packed and sold. Prepare a wide publicity campaign to inform consumers, officials, and manufacturers about the elements of the efficiency program, as well as educating consumers, sellers, etc., about efficiency. Information programs can be implemented immediately in theory; in practice in Indonesia several years of negotiation and testing are required before substantially complete information about appliances can be provided to consumers. (b) Technical Interventions: Change the kinds of products on the market. Essentially this approach represents a calculated investment strategy, by which authorities encourage or require that new appliances be more efficient, which usually raises their prices slightly. Start with voluntary targets for efficiency improvements in new appliances. Announce mandatory efficiency standards to take effect after several years of the voluntary program has been in effect. For Indonesia, such interventions must await the testing and labelling described above, since, as we describe below, establishing targets or standards (c) Incentives: Change the conditions of the market. Adjust prices to reflect the cost of providing electricity. Design subsidy programs to modify market conditions for electric appliances. These programs give rebates (either through utilities or through appliance dealer associations) to customers for buying the most efficient models. 5.19 These strategies interact: higher prices justify interest and investment in efficiency. If electricity prices rise, consumers gain greater benefits from investing in more efficient appliances, and consumers also see greater returns from modifying how they use appliances (or lights). At the same time the producers, importers, or sellers of appliances see a larger and safer market for improved, efficient appliances. If the market for efficiency increases, manufacturers tend to be willing to invest more of their own resources in improving the efficiency of products and of marketing these more efficient products. Information advises consumers how to react to higher prices; conversely, higher prices make consumers more interested in finding out about ways to save electricity. In the following pages we review elements of each of these phases of intervention. - 170 - Information Interventions 5.20 Little can be done to improve the efficiency of electricity use without information about current patterns of use and the characteristics of existing and new end-use devices, ie, appliances, water heaters, lights, cookers, etc. In every country where a program has been launched to spur increased efficiency of electric appliances and other electricity using-devices, an effort has been made to fill this information gap by testing new appliances for energy consumption (and other properties as well), and publishing the results as widely as possible. A submetering experiment helps authorities pin down end-uses where consumption appears excessive because of careless use or where improvements in efficiency are justified by high levels of activity (such as more extensive use of lighting, ironing, water boiling, etc.). This information is used by consumers to compare products; by government and utilities to evaluate overall improvements in actual efficiency of devices on the market, and by manufacturers as goals against which to compete and to measure progress. Good information also makes it more difficult for manufacturers to make vague claims about electricity saving features that may be questionable, rewarding those manufacturers who can demonstrate that their designs do save electricity relative to those of the competition. Available Information on Appliances 5.21 Currently there are no tests of electricity consumption in new household appliances carried out in Indonesia. The only major appliances for which important and useful data related to electricity consumption are available are air conditioners (A/C) and televisions (TV). For these devices, cooling power and size, respectively, can be compared with wattage to provide a useful measure of energy consumption per unit of useful output. 5.22 However, there is no actual checking of this information. For refrigerators, freezers, washers, ovens, water heaters, pumps, and microwave ovens, each appliance must be run over a period of time under well-defined conditions to establish energy consumption. Some additional information about appliance electricity use may be readily available in Indonesia from Japanese producers of imported appliances. We found in the catalogue from one Japanese manufacturer a clear indicator of electricity use for refrigerators and freezers, expressed in kWh/month. We also believe that information on kWh/kg of clothes for washers is available from Japanese importers. We believe that for most appliances imported into Indonesia, Japanese test results not already stated could be found. Testing and Labelling of New Appliances 5.23 The first step towards helping consumers obtain information about energy consumption would be a government requirement that all major appliances be labelled as to their tested energy consumption. Initially, the government could accept testing according to JIS, Japan International Standard, which many countries in the Far East use. Where locally assembled products that are identical with products in other countries are available in Indonesia, consumption data from these other countries must also be provided. 5.24 This first step is only an interim one. Different countries test appliances under different conditions. It is important, therefore, that the GOI, through a ministry or PLN and in consultation with appliance manufacturers and importers, establish: (a) procedures for adjusting - 171 - results of other countries tests to Indonesian conditions; and (b) tests to be carried out in Indonesia for all products sold in Indonesia. By following these steps, imported products could be compared fairly with those locally made, using information that reflects Indonesian conditions. We found, for example, that one Japanese manufacturer of imported A/C stated energy efficiency ratios with outdoor temperatures of 32°C (appropriate for much of Japan) and 35°C (more appropriate for Indonesia). At this higher temperature, the EER is slightly lower than at the lower temperature. Similar adjustments must be made to refrigerator and freezer tests to account for high indoor temperatures. 5.25 The next step is to promulgate a list of appliances that must be tested in Indonesia. At present, the most important electricity-using devices are light bulbs, TV, irons, and refrigerators. However, air-conditioners, freezers, and water pumps are growing rapidly in importance, and washing machines could become important as well. In terms of use per appliance (or application), however, ironing is considerably less important than other appliances, while testing of TV would add little information to that already available. Furthermore, the slight differences in energy consumption/output between lightbulbs of a given type are far less important than the differences among the major types. Thus the testing of appliances should focus initially on A/C, refrigerators, freezers, and pumps. Water heaters and clothes washers could be considered for testing at a later date. But TV and iron wattage, and water heater insulation levels should be included in labelling to be attached to each device. 5.26 The choice of test conditions and procedures that reflect local conditions and a consensus of how appliances are actually used is important. Such tests have now been developed in Brazil, and have been proposed for Jamaica, as part of a World Bank Program. For refrigerators, for example, door openings, ambient temperature and humidity, variations in local voltage, etc., should be considered in the tests. Initially, each seller/manufacturer should be required to test at random a certain number of each model for sale of an appliance on the test list. Any new model introduced must also be tested. After a given period, no appliances of the type on the test list will be permitted to be sold (or imported and sold) unless they have been tested in Indonesia. 5.27 Test results appear in catalogues of each appliance that list, among other things, tested energy use. In some cases, these results are given in manufacturers' catalogues with technical information or even on the same page where each appliance is presented. In the United States, a label estimating energy use of each appliance, with comparisons to other models, must appear on every new appliance sold. (We found some of these tags still displayed on new American appliances sold in Jakarta.) In many European countries the results are included as part of overall evaluations that note other important qualities of interest to consumers. 5.28 In Japan, the U.S., and most European countries, tests are carried out by either manufacturers or by a combination of government or quasi-government labs, private consumer organizations, or other recognized, independent groups. Manufacturer testing has proven successful in some countries, such as the U.K, the U.S., or (for some appliances) Sweden. Additionally, however, the U.S. Consumers' Union makes its own evaluation after purchasing appliances from its local outlets. By purchasing the appliances, they avoid the risk that units supplied by manufactures might not be representative of the average unit. In Germany, Stiftung Warentest, an independent group with some government support, carries out this function, using independent - 172 - laboratories for the actual tests. In Jamaica it is proposed that the manufacturers test, but the Jamaican Bureau of Standards oversees the function. In all countries, test conditions are worked out between government, manufacturers' representatives, consumer groups, and the actual testing lab. Both approaches--manufacturer testing or independent testing--have advantages and disadvantages. 5.29 Consider first requinrng manufacturers in Indonesia to test appliances. This is the least costly approach, since devices unpacked and tested can still be used as display models in stores or repackaged. Indeed, manufacturers already carry out spot checks on compressor power consumption of refrigerators and other properties of appliances. Requiring energy consumption testing means that the appliances actually checked will have to remain under test for several days, so manufacturers will have to dedicate space and personnel to this task. This procedure must be policed by an appropriate government body, who must make spot checks to insure that results are obtained correctly and recorded faithfully. (Curtently there is no widely recognized central testing facility or Bureau of Standards in Indonesia.) And the government will want to purchase a limited number of devices to perform their own checking, a function that could be carried out by LMK. But manufacturer testing may enccunter resistance from consumer groups, who may have reason to doubt the objectivity of such tests. 5.30 The alternative path is for a recognized "neutral" body in Indonesia, or a newly established testing institution, to acquire a sufficient number of each kind of appliance so as to be able to carry out statistically credible tests themselves. These appliances could later be sold as "almost" new, or used in government housing projects. Alternatively, the government could require that the manufacturers/sellers "lend" the appointed testing group the required appliances, which would be returned to the manufacturers to be used as demonstration devices or sold at a fair discount over "new/in box" models (although again the manufacturer might preselect the best model available). If this path were chosen, the most likely group for testing electricity consumption would be LMK. However, it might be desirable to examine other properties of appliances of interest to consumers, such as noise, vibration, and other characteristics related to quality. For these purposes, it might be more desirable to work with other testing labs and YLKI to develop more thorough procedures for testing all major consumer appliances. Such testing is already carried out in most industrialized countries, so the effect of choosing this path would be to start such activities at an early stage in Indonesia's development. 5.31 Comparing the two choices, it is clear that the requirement for manufacturers testing Is the simplest and least costly, but the requirement for establishment of an independent test leads to other consumer benefits In the long run, benefits now available from testing of products in most industrialized countries. It Is sensible to start with manufacturer testing to gain experience at low cost, and then proceed after a few years to independent testing, i.e., to develop a laboratory to take over this work after a number of years. Such a laboratory could be designed to test appliances for a number of countries in the region. 5.32 As a result of these tests, buyers of household appliances will have reliable information on the relative energy-related performance of each product. Additionally, manufacturers will have a framework of information against which to compete on "energy-saving" properties. These approximate figures will reflect the relative energy use of different appliances under actual conditions. - 173- 5.33 The process of working out test conditions (ie, reference temperatures, wash cycles, etc.) may take many months; in the EEC, disagreements over such details have all but sabotaged any EEC-wide labelling. We expect that publication of draft standard test conditions will lead to some difficulties that will lead to revisions, although Indonesia will be able to draw on the experience of many other countries. The sub-metering experiment discussed below is aimed in part at quantifying actuaL rather than theoretical or laboratory patterns of use, in order to clarify test conditions. Some of the information that would be required, such as frequency of use, can be obtained today from the UHESS survey. Additionally, however, the proper label the format of information in catalogues, etc., must be determined. 5.34 Costs for these steps thus far are difficult to determine; for Jamaica, implementation of testing and labelling was estimated to cost close to $US 1 million. Indirect costs include dedication of time of %government officials and engineers, and the time spent organizing both the necessary institutional structure as well as the research into use patterns (and other countries' experiences). Two to three person-years for the first two years, less thereafter, appear to be sufficient. Metering of Actual Appliance Energy Use 5.35 The next vital step in the information program is to perform a submetering experiment. In several hundred homes, LMK should put meters on the major appliances of concern to test consumption over a period of not less than three months. Using small electronic meters, consumption could be recorded without bothering occupants or influencing their consumption patterns. Alternatively, electro-mechanical meters could be used, but these would have to be read regularly, a disturbance that might influence consumer usage. 5.36 This step gives the appliance manufacturers, PLN, and consumers themselves valuable information on how appliances are actually used, both as a function of the tested appliance efficiency, and as a function of family size, routines and habits, and special conditions. PLN wili improve its own knowledge of customer profiles and end-use structure. And manufacturers will gain better insights on how well their appliances actually work. Labeling and Information Requirements 5.37 The final step in the appliance labelling and testing program is to design labels and characterize the information that will be required in catalogues, advertising, and on the cartons in which appliances are packaged. Because of its broad appeat we favor the U.S. labeL This label states the averaged tested consumption of each appliance under the assumed (and stated) conditions, and shows clearly the range of consumption (from highest to lowest, shown on a linear scale, i.e., a straight line). The label also states the average yearly cost of using the appliance for a range of electricity prices. Thus the label "rates' the appliance against all others available. A statement of the consumption alone might be misleading to the individual consumer, all that can be guaranteed is the relative energy and monetary cost of using different machines. The label should of course be designed to reach contemporary Indonesian families, perhaps borrowing (with authorization) the face of some familiar cartoon character or other widely recognized symboL In most countries, for example, the name of the conservation program should be prominently displayed. Consumers then come to recognize the label and its purpose immediately. - 174 - 5.38 Establishment of testing and labelling procedures has an important indirect benefit. The results of the program allow authorities to monitor the changes in efficiency of new appliances actually sold in Indonesia. This is done by weighing the efficiency of each size or class of appliance (size of refrigerator, capacity of clothes washer, etc.) by the number actually sold, to give the sales weighted efficiency factor, or SWEF. The SWEF will show how the overall stock of appliances is improvinS even if the size or output rating of a particular type is increasing. SWEFs provide a useful way of establishing goals for improvements in efficiency, much as the CAFE (for corporate average efficiency) standards are used to set goals for the improvements in automobiles in the U.S. SWEFS were used in the U.S. and the Federal Republic of Germany as the basis for voluntary agreements. These agreements, in the case of Germany, were extremely effective in provoking improvements in all products produced in that countly. The appliance association, ZVEI agreed that its members would improve the average efficiencies of each appliance sold. The targets for 1985, defined as percentage improvements in the SWEFs of key appliances, were agreed to in the late 1970s for 1985. These targets were fulfilled well ahead of time. Consumer Outreach: Increasing Consumer Appliance Literacy 539 Another useful part of an information program is to reach out to consumers and others in the chain between manufacturer and user. Energy consumption is certainly not the only criterion for choosing an electric appliance. And the provision of information on energy consumption, while potentially simple, may not be sufficiently undecstood to be effective. What is required is a broader program to increase consumer literacy over how appliances work, how they are to be maintained, and how quality varies among appliances. The Swedish Board of Consumer Affairs attempts to judge as many as 10 features of washing machines, including energy use, before giving an overall rating. To inform consumers about appliances, particularly the first-time buyers or users who will make the majority of purchases over the next twenty years, will take time and patience. 5AO One way of increasing "appliance literacy' is for either consumer groups (such as YLI) or PLN to promote appliance clinics, at which proper use of appliances is demonstrated. Some utilities in the US. and Europe have done this through their home economists, others through their own sales outlets for appliances. Whfle there may be a potential conflict of interest here-the utility is selling both electricity and know-how on eectricity savings--the utility has the ubiquitous presence required to reach a maximum number of households. YLKI could also hold such clinics, although they are wary of being seen promoting a particular product. The Trade associations themselves could organize such clics, which would also tend to boost sales. Some might see a selleres clinic as simply a promotional stunt, however. 5A1 The second element in improving appliance literacy requires careful production of attractive written material covering features of appliances including energy consumption. This means that results of energy consumption tests must be displayed prominently in catalogues and price lists The best display features consumption test results along with other material in each catalogue. Inclusion of energy consumption data with a detailed list of other technical specifications is important, too, but the energy figures should also be seen in the same view as the picture of the model itself. Finally, the government could consider requiring that consumption test results be displayed prommnently in printed and media advertising - 175 - 5.42 Written material on appliances will include information on energy use and other aspects of quality. For example, the Swedish Board of Consumer Affairs produces a brochure that examines all popular washing machines for energy and other important features. They test for noise, moisture content after washing and centrifuging, and other properties of interest to consumers. Most prospective buyers read the booklet before buying. Alternatively, corresponding groups in the U.S. and Germany describe test results in their monthly publications whenever new tests are performed. The lesson here is again that energy consumption information alone is seldom interesting to consumers, but should be given a prominent place alongside other information important for good consumer choice. 5.43 The sellers of appliances must also be targets for information. They must be taught what the various figures in catalogues and labels mean to consumers; how to work energy efficiency into sales arguments; where to turn for more information, particularly when buyers return with operational questions. Seminars for these vitai actors, as are proposed for Jamaica, should be organized. 5.44 Some governments have tried to "educate" consumers, beyond information on how to buy and use efficient appliances. This "education" urges not simply care in using electricity, but rather self-denial. For example, the Japanese Energy Conservation Center once distributed hand- held fans that were inscribed with messages saying "Do not turn on the Room A/C until the temperature reaches 26°C." Few Japanese colleagues paid heed to this message by itself except in times of clear danger to energy supply, such as in the summer of 1979. Similarly, U.S. utilities do report some success in getting customers to use less A/C or other major appliances during periods of capacity shortage or extreme heat. Efforts in France and other countries to "suggest" less comfortable indoor temperatures also failed, and similar efforts in the U.S. to urge consumers to sacrifice for saving energy produced few tangible results, other than resentment for the words "energy conservation." Similar efforts in Asian developing countries to discourage A/C, lighting, and even elevator use had little impact except during widely-perceived crises. Since the clear and present danger is not a part of the Indonesian environment, it is not likely that the behavioral stick can have much impact in Indonesia. 5.45 Nevertheless, if information on how to achieve electricity savings is provided well in advance of announced increases in electricity prices, consumers will probably heed some of the advice in order to cut costs. Indeed, when Taiwan raised the price of residential electricity in 1980/81 after many years of constant or falling prices, sales growth, which had run at nearly 10% a year, was slashed. 5.46 Carrying the whole message to sellers and consumers will require substantial start- up publicity. Manufacturers could pay for labels (as in Jamaica); the government, through PLN, could support activities of LMK, PLN, and YLKI. Funding must come from GOI. The Jamaican proposal suggests that these publicity costs are small compared to the total volume of appliance sales, as well as compared to the value of the energy saved under the most pessimistic assumptions. Demonstrations and Government Procurement 5.47 One parallel step the government can initiate immediately is to revise its own procurement procedures so as to favor purchase and installation of energy efficient equipment. In - 176 - the years following the first oil price shock, many oil-importing countries realized that they could control as much as 15% of household electricity use because they controlled the housing, either because the housing was built by public authorities, or because governments support families who occupy the homes. Additionally, governments purchase hundreds (or thousands) of home appliances that are used in offices, in military establishments, or to equip government-financed projects like hotels, schools, or hospitals. The government can demand that all purchases of appliances be from a select list of the most efficient models available, and that employees or other users be instructed in proper use and maintenance. We noticed several government offices where a room air conditioner was running while windows were open enough to almost neutralize the effect of the A/C. 5.48 In the near term, this procurement program will have to be limited to A/C and lighting, where the choices between more and less efficient are relatively straightforward. The government could establish as a goal the purchase of A/C with an EER of 12.5 or greater (14.5 is the highest available in Indonesia, according to Japanese tests). Over time, this goal could be raised. As information became available from tests, the government could make similar requirements for TV, water heating, cooking, washing, and refrigeration equipment, all of which will be needed for schools, hospitals, barracks, and, where joint ventures are planned, for hotels and restaurants as well. Additionally, the government could assure that most, if not all, new lamps and fixtures it bought would be fluorescent or compact fluorescent. The latter would be appropriate for existing fixtures that hold incandescent lamps. In the longer term the Government could move to purchase only compact fluorescent bulbs. Information Program: Summary and Time Frame (a) Available information on A/C and TV could be collected rapidly by PLN or YLKI and reproduced in a brochure to be made available to customers in appliances stores, as well as throug.i contacts both PLN and YLKI have with consumers. This could be implemented by 1990. (b) Information on careful use of existing major appliances can be made available within one year. To this end, a committee composed of representatives of manufacturers, PLN/LMK, and YLKI should develop a list of simple strategies for major appliances and lighting. Information from the HES will be helpful. The submetering experiment should also be implemented. These steps can be implemented in 1990. (c) Information on tested appliance electricity consumption in producer's home countries, where such information has been produced, could be required of all product importers, again by 1990. This information would be used in determining efficiency goals (see below), as well as in comparing foreign and proposed Indonesian test results. PLN could collect such information through the Appliance Association. (d) Test procedures (and the organizations nominated for testing) will require some development, estimated at as much as two years. This particular phase will require at least two or three independent engineers who will represent the utility and consumers in negotiations with manufacturers, judge manufacture capability for testing, study foreign experience with testing, etc. A reasonable target date for - 177- required testing, if a program began in 1990, would be preliminary tests 1991, results made public by late 1991, and requirements for testing published in 1992. By the end of 1992, it would be illegal to sell an appliance in Indonesia without revealing tested energy use. (e) Even as manufacture testing is beginning, the search for an appropriate institution for independent testing should begin. We estimate that within a year of the start oi manufacturing testing, the appropriate independent institution could take over the testing function. A board composed of representative parties would oversee the functions of the independent institution. (f) With required testing of new appliances, information could be available to consumers in catalogues and brochures within a few months of the testing. This could take place in 1992 if preliminary testing efforts (by manufacturers) began in late 1990 or early 1991. PLN and YLKI should also prepare a brochure for each type of major appliance, listing models avaiable, tested energy use, and other properties deemed important to consumers. (g) PLNlConsumer Awareness outreach can be implemented by 1991. This program will initially use leaflets to inform consumers and buyers of the cost of using major appliances. After better information from testing is available, clinics that show consumers and consumer groups how to choose and use appliances effetively can be launched. (h) Government procurement can be implemented immediately for key appliances like A/C. By the time testing and labelling is completed, the government will be able to choose the most efficient appliances for almost all major types. Technical Interventions: ProJucing and Selling More Efficient Appliances 5.49 Technical interventions alter the make-up of products in the marketplace. Such interventions are aimed at the manufacturers and importers of electricity using products. These interventions complement the information and price signals we discussed above. Voluntary Agreements: Targets for Efficien9t 5.50 The first step in changing the mix of products available on the market in Indonesia is to approach the manufacturers and sellers of appliances with a program of targets for energy efficiency. Initially the government, through PLN/LMvfK. together with representatives of consumers, manufacturers and sellers, should establish "voluntary' efficiency targets for 1992 as "Gentlemen's Agreements." Such "gentlemen's agreements! in Japan and W. Germany (Wilson et aL) stimulated manufacturers to stop selling their least efficient models, and at the same time to increase the efficiency of all models. The agreements also send a strong signal to appliance buyers/users that efficiency had become important. It took the German government less than one year (1977) to negotiate such agreements, which effectively covered targets for the years 1978 - 1980 - 178 - - 1985. By all estimates, these targets were reached. Manufacturers in other countries improved their products to sell in Germany, which caused tLe German program to lead to improved efficiency in most countries in Europe. 5.51 How are 'Gentlemen's Agreements" implemented? Efficiency targets can be designed progressively: a certain percent improvement within 2 years, a greater improvement widtin 5 years. The SWEFs (see above) for Indonesia, or at least the energy consumption of individual appliances that sell particularly well, could be compared with appliances sold in japan, Europe, and North America to establish goals for improvement in the near term (2 years) and longer term (4- 5 years). Testing of energy use in appliances currently manufactured or imported would allows determination of SWEFS for major appliances sold in the starting year. Yearly progress towards meeting these targets is measured using the SWEFs. The manufacturer's association tabulates the data by weighting the tested consumption of each product by the sales of that product by each company. (A careful system would have to be established to insure confidentiality.) By offering manufacturers a goal and offering the public a measuring standard, the market response to improve efficiency and to buy efficiency would be enhanced. Minimum Efficiency Standards 5.52 If the information from testing and labelling shows that progress is not being made towards goals, then the government can promulgate minimum efficiency standards. This approach was taken by many states in the United States, most notably California, and finally adopted by the United States as a whole, to take effect in 1990-1992 (depending on the appliance). Put simply, the National Appliance Efficiency and Conservation Act makes it illegal to sell major appliances (both gas and electric) after 1992 that do not meet certain minimum performance efficiency standards. (The kinds of criteria were discussed in the Chapter on Regulation and Standards.) The energy performance standards are based on well established tests of consumption developed for labelling programs. These standards will be progressively tightened, giving manufacturers near term targets that are easier to achieve, but long term targets that are more ambitious. These standards considered analyses of the payback to consumers from buying more efficient appliances than otherwise. I/ 5.53 Which approach is better for Indonesia? The answer is both. If testing and labelling begins now, a voluntary program along the lines of the Gentlemen's agreements, can begin in a few years. If this program succeeds in improving the efficiency of appliances sold in Indonesia, strict energy norms can be put off. If the voluntary program does not succeed, then the strict program can be put into place. The size of the Indonesian market is potentially so large that producers and the countries that represent them would probably find it worthwhile both adjusting the mix of products they market in Indonesia to reflect concerns over efficiency and to develop efficient LI 7These standards were passed by the U.S. Congress after neatly a decade of debate, dunng which many large states, mcludingCaifornia, implementedeffcieystandardson cerai applances. Thesestatestandanfswerestmngenough to fomee as much as one-third of availabl produas out of lumaive markets ake California For a time, national cataoues that featued appliances sinp( labeled somne 'not available in California.T The economic importance of that state (12% of the U.S. GNP) and the gwing list of others with similarstanda-dsfowed the man ufactumes to tum to the government to ask for a latively strong national standard, rmther than produce separate products for different states. - 179 - models that are truly built for Indonesian climate and conditions. Alternatively, the government could implement the voluntary program for a period of five years and within the first two years announce standards that would be mandatory by the end of the fifth year of the program. 5.54 In implementing standards, the authorities must recognize that while relatively mild requirements that simply prohibit the sale of the most inefficient appliances could be implemented rapidly (i.e., with one or two year's notice), more ambitious targets must be timed to allow manufacturers to adjust their production and marketing accordingly. Indeed, it is advantageous for the authorities to announce both near and longer term standards, since some manufacturers may find it advantageous to plan for the more ambitious targets sooner than required if they are otherwise planning a new product line or retooling. These concerns argue for a set of voluntary targets today that conclude with mandatory standards at the end of the voluntary period, because the manufacturers have the most time to meet the standards, as well as goals for improving efficiency voluntarily up to the time that the standards take effect. Research in Indonesia 5.55 Developing efficient products in Indonesia is not solely the responsibility of foreign companies. It would be possible for the Government, acting through BPPT or PLN, to offer research funds for development of efficient, locally produced appliances, possibly through a series of competitions. Such an approach was used in Sweden to encourage housing and equipment companies to develop a new generation of technologies for reducing heating in that country (Schipper et al., 1985). And such research at the Technical University of Denmark (Noergard and Pedersen, 1987) has yielded a series of energy efficient appliances, one of which (the "Low Energy Refrigeratort" of Broederne Gram A/S, Denmark) is now under production for the Danish Market. Judging from the experience in Denmark, five to seven years will pass between the first experiments on any given appliance and its entrance into the marketplace. 5.56 One L -:icularly interesting candidate for development in Indonesia would be a combined solar/ele.ric water heater. Currently, water heaters are uncommon except in high- inmcme households; most Indonesians bathe in ambient-temperature water. Gas-fired water heaters have ventilation and safety requirements that make them expensive, and LPG, the most logical fuet is expensive. Instant, electric point of use water heaters are cheaper, but these are difficult to use in Indonesia because of the wattage requirements (as much as 3000 w). Storage water heaters have the disadvantage that the water they store, as well as their heating elements when active, heat the bathroom and effectively the rest of the house. 5.57 Ir, anticipating that Indonesians will turn to heated water (as have Brazilians living in similar climates), therefore, the government could sponsor development of low-cost systems that rely on solar energy for most heat, use outdoor storage, but provide electric boosts when conditions do not permit full use of the sun. Since water heaters are much simpler than most other appliances, their development and testing in houses could take place rapidly, within two or three years. Being relatively simple to make, these water heaters could be produced by local companies. 5.58 The considerations above reflect knowledge that most appliances manufactured in Indonesia are controlled by powerful and technologicatly advanced companies, some of which recognize the need for greater Indonesian participation in design as well as assembly. A research - 180 - program would give these companies incentives to utilize their Indonesian partners' capabilities more fully, developing greater technical know-how that any country the size of Indonesia would need anyway. Technical Interventions: Summary and Time Frame (a) Efficiency Targets: Gentlemen's Agreements could be developed by late 1990, or roughly one year after testin& labeling, and information programs have been launched. The targets should be specified to take effect one or two years after announcement, and be tightened four or five years after announcement. (b) Standards: Minimum efficiency standards could be announced by 1991 to take effect in 1994 or 1995, ie., three or more years after the Gentlemen's Agreements were in place. Two years' minimum notice to manufacturers would be required before standards could become binding; at the same time, a second phase should also be announced that were stiffer than those in the first phase. (c) Alternatively, standards for 1994 or 1995 could be announced in 1992, ie., after enough testing has been undertaken to determine the actual and optimal features of products on the market in Indonesia. The Gentlemen's Agreements serve as a bridge from the present production and market mix of appliances to those determined by the first set of standards. If the standards are exceeded by the actual product mix sold in 1993 or 1994, manufacturers will have a head start towards the second round of standards, which should follow the first round by 2 to 3 years. (d) Development of New Indonesian Appliances. A concerted effort along these lines in Indonesia, using eNisting local appliance companies, could probably develop marketable indigenous appliances by 1995 if research began in 1989. Incentive Programs 5.59 in the United States, many electric utilities are giving consumers Ix,nuses or rebates if they purchase appliances from the most efficient available. This incentive focuses attention on the importance of buying a better appliance, often at a higher initial cost, in order to pay less to use that appliance. Even with relatively high electricity prices (higher than those in Indonesia), American consumers do not always feel the incentive of financial savings from purchase of more efficient appliances. The utility carrot offers them additional savings. These subsidies are still popular in the U.S. 5.60 The subsidies reviewed by Wilson et aL seem to have functioned more to heighten consumer interest than to significantly lower investment costs of energy saving strategies that pay anyway. And the subsidies do attract free riders. But the subsidies solidify the market for proven technologies. Subsidies can be implemented rapidly, provided that the necessary market research has shown that: (a) the products being subsidized save energy and money; (b) these products aren't being bought; (c) the widest conceivable acceptance of the program would not have undue - 181 - consequences for revenue and (d) the overall distribution effects of the subsidy are considered acceptable. 5.61 We believe, however, that such an approach would not work in Indonesia. First, the market for new appliances is laced with bargaining and other informal elements that make it difficult to couple a formal rebate to the buying process. Second, most American consumers have grown up with the major appliances affected by programs, and they have bought these many times over. The program only exists to tilt the next purchase in the direction of greater efficiency. But in Indonesia, such a rebate could seriously alter the size (and most likely the electricity consumption properties) of major appliances. This is because the size and features of appliances on the market are seriously limited by what most consumers could afford. A subsidy for efficiency could easily be converted into a subsidy for larger appliances, without having the effect of encouraging manufacture and purchase of more efficient appliances. Finally, the cost of improving present-day appliances is probably less than the size of the subsidy that would be required to encourage sufficient consumers to choose more efficient models than otherwise. 5.62 Nevertheless, subsidies to lower the price of the most efficient light bulbs may be appropriate for Indonesia. Efficient lighting offers significant electricity savings (van der Plas and de Graff (1988: World Bank; see also Geller, 1986). Unfortunately, the most efficient lamps are simply too expensive in initial cost for most consumers to buy outright, although they pay back, relative to ordinary incandescent lamps, over their lifetime. Subsidies that both reduce the initial price of these lamps and increase the demand for them (thereby lowering their unit costs) are probably cost-effective compared with expansion of the utility system. We judge that for the most efficient compact fluorescent lamps, a subsidy of 25% of the purchase price of the lamp could be tested in selected districts. The Stockholm municipal utility has started such a program. PLN could initiate such a program in a variety of localities. Indicative Benefits and Costs of the Efficiency Program 5.63 There are many important benefits of increased electricity-use efficiency. Consumers save money: The present stock of new appliances appears so inefficient that a considerable (25- 35%) savings of electricity will pay back any initial extra costs consumers have to pay in less than 3-5 years. This benefit will be realized whent ver a more efficient appliance replaces a less-efficient one, or when a consumer is influenced to choose greater efficiency in the first purchase. 5.64 Additional benefits accrue indirectly to consumers through reductions in costs to the utility. PLN saves money by reducing future borrowing requirements for capacity. This applies both to base-load and peaking units. Presently, &3 -ity L affected significantly by lighting and TV, which peak at night. Refrigerators and water heaters, whose importance is growing, operate around the clock .nd thus boost base load capacity needs. As air conditioning grows in importance in Indonesia, its contribution to a daytime peak will increase, along with its potential to disrupt electricity services. A conservation strategy would thus reduce the size and disruptive influence of the A/C peak and reduce base load growth as well. A conservation program also ameliorates the impact of the inevitable increase in the price of electricity that will pay for this capacity - 182- expansion. If capacity grows more slowly, prices rise less rapidly. If the utility is seen as an active participant in such a program, both the credibility and image of the utility can be increased. 5.65 There are several kinds of costs of greater efficiency that have to be considered, but these will be minor compared with the benefits. Actual investments in greater efficiency might increase the cost of major appliances (refrigerators, freezers, air conditioners, pumps) by as much as 5%. The trade off between incandescent and compact fluorescent lamps is an exception, but the energy use is reduced by 80% and lifetime extended 3-5 fold. The savings for a home with the four major appliances, TV, and lights might cost US$100 (or Rp 170,000) in incremental first cost. The benefits (savings of 700 kWh @ 85 Rp/kWh - 60,000 Rp/yr), give a simple rate of return of nearly 33%. 5.66 There are also public administrative costs. Ener$ave for Jamaica estimates the cost of the testing and labelling program at less than US$ 2.00 per appliance sold in the first year of that program (based on manufacturer testing). Similar costs could be expected for Indonesia. Other costs that arise, including publicity costs, would have benefits beyond energy saving, since these costs would promote better consumer awareness of all aspects of owning and using appliances. That is, many of the activities recommended herein will become desirable in Indonesia for a variety of purposes. Clearly, a more detailed examination of costs of the programs suggested here is desirable, but perusal of programs in other countries suggests that the administration and start-up costs are small compared to the total benerits to both consumers, PLN, and GOI. 5.67 In the final analysis, the conservation strategies recommended amount to an investment strategy for Indonesia that will make more electricity available than similar investments in electricity supplies. This conservation investment program must be developed in parallel with efforts to expand electricity supplies. Benefits to Indonesia accrue both directly, in the form of lower appliance operating costs, and indirecty, in the form of avoided costs for expanding the supply system more than necessary. 5.68 An indicative economic evaluation of the costs and benefits Indonesia could expect from an electricity-use efficiency program as described above is shown in Table 5.2. Using conservative estimates of electricity savings and market penetration, this evaluation shows that robust net benefits could be expected to accrue to the national economy from such a program. - 183- Table 5.2: Urban Electricity Conservation Program Costs and Benefits 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Costs (Uss millions) -Test facilities 0.3 1.5 1.5 -Training, Personnel, and Administration 0.3 0.5 0.5 0.5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 -Info campairn 0.3 0.5 0.5 0.5 0.4 0.3 0.3 0.3 0.3 0.3 0.3 -Testing + labelling Refrigerators a uss 2 / unit 0.45 0.50 0.55 0.61 0.67 0.74 0.82 0.90 0.99 1.10 1.21 Air-conditioners a uss 5 , unit 0.04 0.05 0.05 0.06 0.07 0.08 0.09 0.11 0.12 0.14 0.16 Television sets a USS 2 / 10 units 0.22 0.24 0.26 0.28 0.31 0.34 0.37 0.40 0.44 0.48 0.52 Total Program Cost 1.61 3.28 3.36 1.95 1.75 1.76 1.88 2.01 2.16 2.32 2.50 Market Potential Share g/ Savings h/ Benefits (Gwh Saved) 0 9.6 26.0 55.9 90.2 129.1 166.1 207.5 253.8 305.3 362.5 Lighting 101 301 0 5.4 14.5 30.9 49.4 69.8 88.8 109.5 132.3 157.1 184.0 Refrigeration 351 25X 0 3.7 10.0 21.6 35.2 51.0 66.3 83.9 103.8 126.5 152.1 Air-conditioning 351 251 0 0.2 0.7 1.6 2.9 4.4 6.1 8.1 10.4 13.1 16.1 Television 201 5X 0 0.3 0.8 1.7 2.7 3.9 4.9 6.0 7.3 8.7 10.2 Program Effectiveness g/ 0% 5X 131 25X 38X 501 602 70M 801 901 100X Total Value of Savings (US mdtllions a 99/kWh) 0 0.85 2.29 4.93 7.96 11.39 14.65 18.31 22.40 26.94 31.98 Lighting 0 0.48 1.28 2.73 4.35 6.16 7.83 9.67 11.67 13.86 16.24 Refrigeration 0 0.32 0.88 1.91 3.11 4.50 5.85 7.40 9.16 11.16 13.42 Air-conditioning 0 0.02 0.06 0.14 0.25 0.39 0.54 0.71 0.92 1.15 1.42 Television 0 0.03 0.07 0.15 0.24 0.34 0.43 0.53 0.64 0.77 0.90 Met Program Benefits (USS millions) -1.61 -2.43 -1.07 2.99 6.21 9.63 12.78 16.30 20.24 24.62 29.49 Economic Internal Rate of Return: 72.01 aJ Targetted market share of efficient appliances by the year 2000. b/ Kwh savings of improved appliances expressed as percent of kWh used in standard models. S/ Percent of year 2000 improved appliance market penetration target. Suary Evaltation of Urban Electricity Conservation Program Met Present Value a 101 Anr alized NPV (US$ Millions) (US$ Millionis) Total Progra Cost (USS mIllions) 14.7 2.3 Electricity Conservation (G*h) 729 112 Lighting 384 59 Refrigeration 295 45 Air-conditioning 28 4 Television 21 3 Economic Value of Savings (USS millions) 64.3 9.9 Lighting 33.9 5.2 Refrigeration 26.1 4.0 Air-conditioning 2.5 0.4 television 1.9 0.3 et Economic Benefits 49.6 7.6 . 184- Table 5.2: URBAN ELECTRICITY CONSERVATION PROGRAM ECONOMIC COSTS AND BENEFITS (con't) UHESS Urban Applianee Sales Projections Annual sales Annual 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Growth Light bulbs (millions) 5.9X 39.1 41.3 43.8 46.3 49.0 51.9 54.9 58.1 61.6 65.2 69.0 73.0 Refrigerators (thousands) 10.4X 204.0 225.2 248.6 274.5 303.0 334.6 369.4 407.8 450.2 497.0 548.7 605.7 Air-conditioners (thousands) 15.0X 7.0 8.1 9.3 10.6 12.2 14.1 16.2 18.6 21.4 24.6 28.3 32.6 Television sets (millions) 9.3X 1.0 1.1 1.2 1.3 1.4 1.5 1.7 1.8 2.0 2.2 2.4 2.6 UNESS Urban Electricity Demand Projections for Urban Indonesia (GIAIyr) 1988 19R9 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Incandescent lighting 1,966 2,122 2,280 2,440 2,603 2,768 2,937 3,110 3,288 3,471 3,659 3,854 4,056 Fluorescent lighting 932 1,013 1,096 1,181 1,268 1,357 1,449 1,545 1,643 1,746 1,852 1,963 2,079 Refrigeration 653 710 772 839 912 990 1,074 1,165 1,263 1,369 1,484 1,606 1,739 Ironing 400 425 451 479 509 540 573 608 645 685 726 770 817 Water pumping 305 328 352 378 406 436 469 503 540 580 622 668 716 Color television 292 312 333 355 379 404 431 460 490 522 557 593 632 SW television 225 236 247 259 271 284 297 311 325 340 356 372 389 Sales 368 384 397 406 411 413 412 403 377 350 323 295 268 Others 426 599 792 1,003 1,232 1,479 1,744 2,026 2,325 2,640 2,972 3,319 3,683 Total 2,669 2,994 3,344 3,719 4,119 4,546 5,000 5,477 5,966 6,487 7,039 7,624 8,243 - 185 - REFERENCES CEPEL, 1985. Energy Conservation in Home Appliances. qInforme", No. 20 (April). Available from Electrobras, Rio De Janerio, Brasil Ener$ave Jamaica, 1989. Program Plan. Final Report of Phase One Program Development, Jamaica Energy Efficiency Standards and Labelling Project. March 17. Geller, H., 1986. End-Use Electricity Conservation: Options for Developing Countries. Energy Department Paper No. 32. Washington, D.C.: World Bank Energy Dept. Geller, 1988. Efficiency of Residential Equipment. Prepared for the Office of Technology Assessment, U.S. Congress. Washington, D.C.: Am. Council for an Energy Efficiency Economy. Geller, H., Goldemberg, J., Moriera, J. R., Hukai, R., Scarpinella, C., and Ysohizawa, M., 1988. Electricity Conservation in Brazil: Potential and Progress. Energy, Vol. 13, No. 6, pp 469-483. Noergaard, J. and Pedersen, P. B., 1987. Low Electricity Household of the Future. Lyngby, Denmark: Energy Group, Physics Laboratory HI. See also Noergaard, J., and Gulbrandsen, T., 1987. Potentials for Electricity Savings Using Known Technology. Lyngby: Energy Group. Rudermann, H., MacMahon, J., 1987. Behavior of the Markeplact for Residential Appliances. Energy Economics. Schipper, L., 1976. Towards More Efficient Energy Utilization. Annual Review of Energy, VoL 1. Palo Alto, Ca: Annual Reviews Incorporated. Schipper, L., Ketoff, A., Meyers, S., and Hawk, D., 1987. Residential Electricity Consumption in Industrialized Countries: Changes since 1973. Energy, the International Journal. Schipper, L., and Meyers, S., 1983. Energy Us.. md Conservation in Kenya. Energy Policy (Sept.). Schipper, L., Meyers, S., and Kelly, H., 1985. Coming in from the Cold: Energy Wise Housing from Sweden. Washington, D.C.: Seven Locks Press. Sudja, N., Widodo, S., and Ibrahim, S., 1988. Indonesia's Load Profile toward the year 2000: An Approach of Load Forecasting Methodology in a Developing Country. Preprint, available from PLN, Jakarta. Turiel, I., 1986. Design Options for Energy Efficiency- Improvement of Residential Appliances. LBL 22372. Berkeley: Lawrence Berkeley Laboratory. (Updated in 1989). van der Plas, R., and de Graaff, A.B., 1988. A Comparison of Lamps for Domestic Lighting in Developing Countrics#.. Industry and Energy Department Working Paper, Energy Series No. 6. Washington, D.C.: World Bank. - 186. Wilson, D., 1989. Report from Interationa Energy Studies, Lawrence Berkeley Laboratory (in preparation) for the U.S. Dept. of Energy. This report describes efficiency activities in Sweden, Japan, France, and the Federal Republic of Germany.