57700 Energy Sector Management Assistance Program ENERGY AND POVERTY Special Report 007/ l 0 August 2010 Peru: National Survey of Rural Household Energy Use Peter Meier Voravate Tuntivate Douglas F. Barnes Susan V. Bogach Daniel Farchy Contents Exchange Rate ix Abbreviations and Acronyms xi Foreword xiii Acknowledgments xv Executive Summary xvii Introduction Geographical and Socioeconomic Diversity in Peru 2 Electricity Sector Structure 4 Rural Electrification to Date 6 Key Rural Electrification Issues 7 2 _Household Energy Use and Expenditure 9 Household Energy Use · 9 Energy Expenditure 13 Comparison of Households with and without Access to Grid Electricity 14 Socioeconomic Characteristics 14 Energy Use 16 Energy Expenditures 17 Conclusions 18 3 Electricity from the Grid 21 Access to Grid Electricity 21 Service Reliability 22 Overall Electricity Use and Expenditure 23 Electricity Tariff by Usage Level 23 Electricity Usage for Lighting 24 Household Appliances 26 Conclusions 28 4 Off-Grid Electricity 31 Car Batteries 31 Dry Cell Batteries 33 Small Generators 35 Solar Home Systems 36 Conclusions 37 iii Use 5 Benefits of Rural Electrification 39 Background on Rural Electrification Benefit Estimation 39 Measuring the Benefits of Better Lighting 41 Benefits of Communications 49 Television 49 Radio 50 Refrigeration 51 Benefits of Education and Health 53 Education 53 Health and Environmental Benefits 54 Benefits to Home Business 54 Willingness to Pay for Electricity in Enterprises 55 Conclusions 58 6 Policy Implications of Survey Results 59 Connection Rates in Electrified Villages 59 Distribution of Electricity Consumption by Village 60 Growth of Electricity Consumption 61 Pricing Policy 62 POSE 63 Tariff Structure 64 Targeting Performance of FOSE 68 Efficient Lighting 70 Issues for Further Research 72 Conclusions 72 Annexes 1 Survey Design and Methodology 75 2 Survey Results . 83 3 Estimating the Benefits of Rural Electrification 115 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas 119 100. Characteristics of House & Household 120 200. Characteristics of Household Members 122 300. Sources of Energy (Only for the Head of Home or the Spouse) 126 400. Productive Equipment 148 500. Time Use 150 600. Household Income 151 700. Attitude· 155 800. Business Module 156 900. Opinion and Attitude on Energy and Business 159 Bibliography 161 Boxes 5.1 Comparison of Survey Results with an NRECA Study 52 6.1 Rate of Return for Replacement of Incandescent Lighting with CFLs 71 Figures 2.1 Application of Kerosene Users (Users Only) 11 2.2 Expenditure Differences Across Regions: Fraction of Households in Each Expenditure Quintile by Region 11 2.3 Households Using Modem Energy by Expenditure Quintile 12 2.4 Households Using Traditional Fuel by Expenditure Quintile 13 2.5 Household Monthly Spending on Energy by Region and 'I)rpe (Soles/Month) 14 iv Contents 2.6 Percentage of Households Maintaining Kerosene. ?fld .Candles to Supplement Electric Lighting 17 2.7 Household Energy Expenditure as a Percentage of Total Expenditure by Region 18 2.8 Household Expenditures on Electricity and Other Lighting Fuels/Energy by Expenditure Quintiles (Soles per Month) 19 3.1 Percentage of Households with Access to Grid Electricity by Region 22 3.2 Percentage of Households with Access to Electricity by Expenditure Quintile 22 3.3 Months of Service per Year 23 3.4 Hours of Service per Day 23 3.5 Lighting versus Total kWh by Quintile 25 3.6 Appliance Use in Electrified Homes 26 3.7 Appliance Ownership in Electrified Homes by Expenditure Quintile 27 3.8 Television Ownership in Electrified Households by Region 27 3.9 Fan Ownership by Region 28 4.1 Percentage of Households Reporting Use of Dry Cells 34 4.2 Dry Cell Watt-Hour Consumption by Expenditure Quintile 35 5.1 Demand Curve for Lighting (Theoretical) 41 5.2 Demand Curve for Lighting (Actual) 44 5.3 The Impact of Assumptions 45 5.4 Demand Curve Based on Price and Quantity of Energy Use by Income Class 47 5.5 Energy Source Differences Between Electrified versus Unelectrified Enterprises 56 5.6 Change in Monthly Expenditure with Electrification 57 6.1 Breakdown of Villages with Average Consumption above and below 22 kWh/Month by Region 61 6.2 Distribution of Years of Electricity Service 62 6.3 Monthly kWh of Consumption versus Age of Connection 63 6.4 Components of the Tariff 64 6.5 Typical Tariff Curves 65 6.6 Average Cost/kWh, Electrosur 68 6.7 Proportion of Households Reporting Given Levels of Consumption 69 6.8 Effective Rate of FOSE Discount 69 A.1.1 Monthly Household Electricity Expenditure by Region (Users Only) · 79 A.1.2 Monthly Kerosene Expenditure by Region (Users Only) 80 A.1.3 Kernel Density Estimation of Total Monthly Household Cash Expenditure without Energy Expenditures (Logarithm) 82 A.3.1 Demand for Lighting 115 A.3.2 Comparison of Estimates Used for Kerosene Lighting 116 Tables 1.1 Population by Region and Area 2 1.2 Poverty Incidence in Rural Areas (% of Households) 4 1.3 Residential Subsidized Tariffs (Soles/kWh) 5 1.4 Latin American and Caribbean Region Electricity Coverage, by Percentage of Coverage 6 2.1 Percentage of Households that Use Each Type of Energy by Region 10 2.2 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles (Users Only) 15 2.3 Total Household Expenditure and Education by Electrification Status 15 2.4 Percentage of Households that Use Each fype of Energy by Electrification Status 16 2.5 Household Monthly Cash Expenditure on Electricity and Lighting Fuels/Energy by Electrification Status (Users Only) 17 3.1 Household Electricity Consumption, Expenditure, and Average Effective Price per KWh by Region 24 3.2 Household Electricity Consumption, Expenditure, and Average Effective Price per kWh by Expenditure Quintiles 24 v Special Report Peru: Nalional Survey of Rural Household Energy Use 3.3 Number and Type of Electric Lights Owned by Level of Usage 25 4.1 Use of Car Batteries (% of Households) 32 4.2 Car-Battery Statistics by Expenditure Quintile 33 4.3 Uses of Dry Cell Batteries (% of Users Only) 33 4.4 Dry Cell Battery Costs 34 4.5 Small Generator Users, Cost Data 36 4.6 Percentage of Households that Use Solar PV Systems by Electrification Status and Expen(!.iture Quintile 36 5.1 Lumen Output for Lighting Devices 42 5.2 Lighting Ownership and Hours of Utilization by Households (Unweighted) 43 5.3 Percentage of Households Reporting Use of Lamps 44 5.4 WTP Estimates 46 5.5 Number of Households Using Lighting Energy by Income Class, 2005 47 5.6 Estimates of Lighting Service and Price 48 5.7 Increases of Consumer Surplus by In:come Class (Soles/month) 48 5.8 WTP per kWh (Alternate Method) 49 5.9 Cost and Viewing Hours for Television 50 .5.10 Cost per Radio Listening Hour Based on Energy Source 50 5.11 Price and Quantity of Radio Listening 51 5.12 Use of Refrigerators in Unelectrified Households 51 5.13 Average Number of Hours per Night Household Members Read/Study (Weighted) 53 5.14 Percentage of Children in the Household Attending School (Weighted) 53 5.15 Home Business Incidence by Major Lighting Type 54 5.16 Distribution of Households with Home Business by Major Lighting Type 54 5.17 Energy Sources in Rural Enterprises 55 5.18 Average Energy Expenditures in Rural Enterprises 56 5.19 Enterprise Willingness to Pay for Electricity 57 5.20 Net Benefits of Grid Electrification (per HH/month) 58 6.1 Connection Rates by Region 60 6.2 Major Reasons Cited for Households Lack of Grid Access: Electrified Villages versus Unelectrified Villages(% of Respondents) 60 6.3 Average Consumption in Electrified Villages 61 6.4 FOSE Subsidy Rates 63 6.5 Number of Connections Benefiting from FOSE, 2004 64 6.6 Fixed Charges 65 6.7 Sample Electricity Bills from Electrosur, April 2005 67 6.8 Targeting Performance for POSE Transfer 70 6.9 Average Monthly Electricity Bill, Soles/Month 72 A.1.1 Distribution of the Sample Size 75 A.1.2 Weighted versus Unweighted Estimates of kWh/HH/Month 76 A.1.3 Comparison of Monthly Energy Expenditure and Total Cash Expenditure between the ENAHO and the Survey 78 A.1.4 Monthly Household Electricity Expenditure by Region (Users Only) 80 A.1.5 Monthly Household Kerosene Expenditure by Region (Users Only) 81 A.1.6 Monthly Household Kerosene Expenditure by Region 82 A.2.1 Percentage of Households that Use Each Type of Energy, by Region 83 A.2.2 Percentage of Households that Use Each Type of Energy, by Expenditure Quintiles 84 A.2.3 Total Household Monthly Cash Spending on Energy by Region, in Soles (Users Only) 85 A.2.4 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles, in Soles (Users Only) 86 A.2.5 Total Household Monthly Cash Spending on Energy by Region, in Soles (All Households) 87 A.2.6 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles, in Soles (All Households) 87 vi Confen ts A.2.7 Percentage of Households that Use Each Type of Energy by Electrification Status and Region ' · 88 A.2.8 Percentage of Households that Use Each Type of Energy by Electrification Status and Expenditure Quintile 89 A.2.9 Comparison of Total Household Monthly Cash Spending on Energy between Households with and without Access to Grid Electricity by Region, in Soles (Users Only) 90 A.2.10 Comparison of Total Household Monthly Cash Spending on Energy between Households with and without Access to Grid Electricity by Expenditure Quintiles (Users Only) 91 A.2.11 Number and Percentage of Households Using Kerosene for Lighting and Cooking by Electrification Status and by Region (All Households) 92 A.2.12 Number and Percentage of Households Using Kerosene for Lighting and Cooking by Region (All Households) 93 A.2.13 Number and Percentage of Households Using Kerosene for Lighting and Cooking by Electrification Status and Expenditure Quintiles (All Households) 94 A.2.14 Number and Percentage of Households Using Kerosene and Candles for Lighting by Electrification Status and Region (All Households) 95 A.2.15 Number and Percentage of Household Using Kerosene and Candles for Lighting by Electrification Status and Expenditure Quintiles (All Households) 96 A.2.16 Household Monthly Expenditure on Kerosene for Lighting and Cooking by Electrification Status and Region.(Users Only) 97 A.2.17 Household Monthly Expenditure on Kerosene for Lighting and Cooking by Region (Weighted-Users Only) 98 A.2.18 Household Monthly Expenditures on Candles for Lighting by Electrification Status and Region (Users Only) 98 A.2.19 Household Monthly Expenditures on Kerosene for Lighting and Cooking by Electr_ification Status and Expenditure Quintiles (Users Only) 99 A.2.20 Household Monthly Expenditures on Kerosene for Lighting and Cooking by Expenditure Quintiles, in Soles (Users Only) 100 A.2.21 Household Monthly Expenditures on Candles for Lighting by Electrification Status and Expenditure Quintiles, in Soles (Users Only) 100 A.2.22 Household Monthly Expenditures on Kerosene and Candles for Lighting by Electrification Status and Region, in Soles (Users Only) 101 A.2.23 Household Monthly Expenditures on Kerosene and Candles for Lighting by Region (Users Only) 101 A.2.24 Household Monthly Expenditures on Kerosene and Candles for Lighting by Electrification Status and Expenditure Quintiles (Users Only) 102 A.2.25 Household Monthly Expenditures on Kerosene and Candles for Lighting by Expenditure Quintiles 102 A.2.26 Household Monthly Expenditures on Lighting and Electricity by Electrification Status and Region (Weighted-Users Only) 103 A.2.27 Household Monthly Expenditures on Lighting and Electricity by Region, in Soles (Users Only) 104 A.2.28 Household Monthly Expenditures on Lighting and Electricity by Electrification Status and Expenditure Quintiles, in Soles .105 A.2.29 Household Monthly Expenditures on Lighting and Electricity by Expenditure Quintiles 106 A.2.30 Percentage of Households with and without Access to Grid Electricity by Region 106 A.2.31 Percentage of Households with and without Access to Grid Electricity by Expenditure Quintiles 107 A.2.32 Household Electricity Consumption, Expenditure in Soles, Effective Price per kWh, and Electricity Used for Lighting by Region 107 A.2.33 Household Electricity Consumption, Expenditure Effective Price per kWh and Electricity Used for Lighting by Expenditure Quintiles · 108 vii Special Report Peru: National Survey of Rural Household Energy Use A.2.34 Type and Number of Electric Lights Owned by Household by Expenditure Quintiles (All Households with Grid Connection) 108 A.2.35 Type and Number of Electric Lights Owned by Region (Users Only) 109 A.2.36 Type and Number of Electric Lights Owned by Expenditure Quintiles (Users Only) 109 A.2.37 Electricity Usage for Lighting by Lifeline Rate 109 A.2.38 Type and Number of Electric Lights Owned by Lifeline Rate 110 A.2.39 Television Ownership by Region 110 A.2.40 Television Ownership by Expenditure Quintiles 110 A.2.41 Plug-in Radio and Television Ownership by Region 111 A.2.42 Plug-in Radio and Television Ownership by Expenditure Quintiles 111 A.2.43 Electric Appliance Ownership by Region 112 A.2.44 Electric Appliance Ownership by Expenditure Quintiles 112 A.2.45 Households with Photovoltaic (PV) Systems 113 A.2.46 Small Generator Users, Cost Data 114 A.3.1 Statistical Comparison of User Estimates of Kerosene Consumption versus Average Lamp Consumption 117 A.3.2 Assumptions and Results, Willingness to Pay for Lighting per Month, by Quintile 117 A.3.3 Cross-Country Comparisons of WTP Calculations 118 viii Exchange Rate (Effective 2007) Currency unit Nuevos Soles (s/.) 1 U.$. dollar 3.23 Nuevos Soles ix Abbreviations and Acronyms AER Areas de Empadronamiento Rural ADINELSA Electric Infrastructure Administration Enterprise (Empresa de Administraci6n de lnfraestructura Electrica, S.A) B&W black and white (TV) CFL compact fluorescent lamp (also known as compact fluorescent light bulb) CIER 1 Comisi6n de Integraci6n Electrica Regional CV coefficient of variation DEP Executive Project Directorate ECLAC Economic Commission for Latin America and the Caribbean EDElNOR. Empresa de Distribuci6n Electrica de Lima S.A. ELP ELECTROPERU-;t'ublic Electric Enterprise of Peru ENAHO National Household Survey (Encuesta Nacional de Hogares) FONAFE National Financing Fund for State Enterprise Adivity (Fonda Nacional de Financiamiento de la Actividad Empresarial del Estado) FONCODES National Fund for Compensation and Social Development of Peru FOSE Social Tariff for Electricity Consumption (Fonda de Compensaci6n Social El€ctrica) GoP Government of Peru HH household ICT information and communication technologies ·IDA International Development Agency INEI National Institute of Statistics and Information Technology (Instituto Nacional de Estadistica e Informatica) kWh kilowatt-hour LCE Law of Electricity Concessions (Ley de Concessiones Electricas) LPG liquified petroleum gas MEM Ministry of Energy and Mines (Ministerio de Energia y Minas) MHS micro hydroelectric systems xi Special Report Peru: National Survey of Rural Household Energy Use NRECA National Rural Electric Cooperative Association OSINERG Supervisory Commission for Energy Investments (Organismo Supervisor de la Inversi6n en Energia) PPIAF Public Private Infrastructure Advisory Facility (World Bank) PV photovoltaic WTP willingness to pay All monetary amounts are US dollars unless otherwise indicated. xii Forevvord In order to gain a better understanding of the existing and potential users of electricity in rural areas of Peru, the National Survey of Rural Household Energy Use was carried out in seven regions of the country, the Coastal North, Central, and South regions, the Andean North, Central, and South regions, and the Amazon region. The Survey provided data on rural household energy use and expenditures, use by rural households of electricity from the grid, and use by rural households of off-grid electricity. The Survey also provided information for an analysis of the economic benefits from electricity use in rural areas in Peru. Finally, the data were analyzed to provide implications for further development of rural electrification policies in Peru. It is important to note that the report represents the situation with respect to rural electrification in Peru in 2005-2006. The Survey was initiated during the preparation of the World Bank-GEF-assisted Peru Rural Electrification Project. It provided socioeconomic and energy data to inform the design of the Project and also assist in improving policies for rural electrification in Peru. The preliminary data from the Survey were used to prepare the economic and financial analysis for the Peru Rural Electrification Project. The main conclusion of the survey is that rural households in Peru have a significant desire, willingness, and ability to pay for electricity. Households without electricity from the grid frequently pay more for energy of lesser quality from kerosene lamps or batteries than they would pay for electricity service. However, the need to pay the connection cost is a significant barrier, and 25 percent of households living in areas with electricity service are not connected. Use of car batteries by 18 percent of rural households without electricity is a strong indication of unsatisfied demand for electricity in areas near to the grid. The Survey report provides data for the planning of rural electrification in the context of Peru, including estimates of the benefits, which are particularly important for the economic analysis of Projects. However, we believe that the survey report will also be useful to other countries as an example of a comprehensive effort to collect and analyze original data on rural household energy use. xiii Ackno'4Vledg111ents The Survey was carried out with financing from the Ministry of Energy and Mines for the survey fieldwork and from the World Bank's Energy Sector Management Assistance Program (ESMAP) for the survey design and the preparation of the final report. The work was completed in 2007, it then went through a series of internal and external reviews. It is the intention of the authors that you find this material interesting and insightful. This report has also been published in Spanish by the originating unit within the World Bank. The report can be found as: Peru: Encuesta Nacional de Consumo de Energia a Hogares en el Ambito Rural, 54286-PE. The authors would like to thank the former and current authorities of the Ministry of Energy and Mines for their help and support, as well as former Vice Minister of Energy Mr. Juan Miguel Cayo Mata. The Technical Directorate of Demographics and Social Indicators of the National Institute of Statistics and Information Technology (INEI) deserves special mention for designing the sample and the survey, as well as for organizing and supervising the field survey work under difficult conditions, and entering and verifying the data. INEI's team was led by Arturo Arias, and the sample design was done by Juan Valverde Quesada. Patricia Ormeno supervised and coordinated the survey fieldwork on behalf of the Ministry of Energy and Mines. Laura Berman assisted in coordination among the Ministry of Energy and Mines, INEI and the World Bank, and drafting the Introduction to the Report and Annex 1. Voravate Tuntivate was responsible on the World Bank side for the sample design, development of the questionnaire, providing guidance and supervision of the initiation of the fieldwork, deaning and analyzing the data, preparing the tables in Annex 2, and drafting a preliminary report. Peter Meier prepared the complete draft of the first report, as well as the benefits estimates in Chapter 5 of the report, and assisted in review and finalizing the data. Daniel Farchy prepared the second draft of the report and the first draft of the executive summary. Paula Tarnapol Whitacre edited the final draft of the report. Thomas Haven assisted in reviewing and editing the final draft of the main report. The authors are grateful to the following reviewers for their valuable comments: Dana Rysankova, Kyran O'Sullivan, and Eduardo Zolezzi Douglas Barnes provided guidance to the task team throughout the preparation of the Survey and the report..The task team leader was Susan V. Bogach. Special thanks to Shepherd, Inc. for copyediting the final report and to Marjorie K Araya (ESMAP) for coordinating the publication process. The findings, interpretations and conclusions expressed in this report are entirely those of the authors as individuals. The report was completed under the guidance of the Energy Unit of the Sustainable Development Department of the Latin America and Caribbean Region of the World Bank. xv Executive Surn111ary Peru is a country of extreme diversity, both in its Program (ESMAP), to obtain information on the geography and the socioeconomic conditions of its demand and use of electricity in rural areas of Peru. citizens. This makes it a challenge for the Government The Survey covered 6,690 households with and of Peru (GoP) to extend access to basic infrastructure without electricity in rural areas of Peru. To represent services, including electricity, to the dispersed the target population for rural electrification, rural population living in rural areas. Plans and targets areas were defined as those populations living have been. in place for rural electrification since the in aggregations of 1,000 households or less. (This early 1970s, but by 2005, only 39 percent of rural definition is different from that used by the Institute households had electricity service. Peru has one of the of Statistics and Information Technology [Instituto lowest rural electrification rates in Latin America. An Nacional de Estad1stica e Informatica, INEIJ in the estimated 6 million people in the predominantly poor Census, which defines rural population centers rural areas of Peru did not have access to electricity as those with less than 100 dwellings grouped in2005. contiguously.) The sample was large enough to The MEM initiated a World Bank and GEF- provide reliable estimations about the survey assisted Rural Electrification Project in August 2006 to population at seven regional levels: Coastal North, assist local distribution companies in reaching rural Central, and South regions, the Andean North, populations (World Bank 2006). The project aims to Central, and South regions, and the Amazon region. 2 supply electricity services to about 160,000 currently The survey was conducted through the National unserved rural households, businesses, and public Institute of Statistics and Information Technology facilities, such as schools and health clinics (serving (Instituto Nacional de Estadistica e Informatica, about 800,000 people), using both conventional grid INEI), together with experts on household energy extension and renewable energy sources. surveys provided by the World Bank. The information Detailed data were required in order to prepare collected includes general socioeconomic information the design of the Peru Rural Electrification Project, ,on households, as well as detailed information on as well as to improve the rural electrification their current energy use, energy expenditures, and program and to analyze the economic and financial ability/willingness to pay for electricity services. Until aspects of rural electrification. Consequently, it was now such data have not been available. decided to implement the National Survey of Rural This report presents the main results of the Survey, Household Energy Use (referred to as the Survey in and shows how Survey information can contribute to this publication),1 with the assistance of the World the analysis of important policy issues in developing Bank's Energy Sector Management Assistance an improved rural electrification framework in Peru. 1 In Spanish, Encuesta de Consumo de Energ{a a Hogares en el Ambito Rural. 2 The expected standard deviation in each region ranged from 0.021 to 0.050 (see Annex 1). Special Report Peru: National Survey of Rural Household Energy Use Household Ene!l'gy !UJse a1nitdl IE.xpendnture financially better-off and poorer households. The Survey compares energy usage among Energy expenditure represents a heavier burden for households in different regions, expenditure quintile households in the three Andean regions than for classifications, and categories of households with and households in other regions of the country. without access to grid electricity. Although poor households spend less on energy than nonpoor households, their energy spending Vo11riations in Energy Use accounts for a larger portion of their income. Households in the lowest quintile spend about Rural households in Peru, like rural households elsewhere in the world, rely on various sources 17 percent of their total monthly expenditures on energy, while households in all other quintiles of energy for lighting, cooking, and appliances. spend less than 10 percent. Part of the reason for More than 84 percent of rural households rely on this discrepancy is that the poor often lack access to fuelwood for cooking, while 24 percent use dung and 11 percent use agriculture residue. Liquified relatively cheap grid electricity. petroleum gas (LPG) is used mainly for cooking Households with grid electricity are financially better off than households without access (average by 14 percent of all households. An estimated 74 percent of all households use dry cells for small 430 soles/month versus 317 soles/month). Yet, appliances such as radios and flashlights, and the Survey also found that households with electricity spend only marginally more on grid about 60 percent of all households use candles electricity and electricity substitutes (16.3 soles and kerosene for lighting. Electricity is used by 39 percent of all households. A surprisingly high per month) than households without electricity 11 percent of households use car batteries to run spend on substitutes alone (15.4 soles per month). electric appliances, indicating a high, unmet Households without electricity are paying demand for electricity services. A tiny fraction, comparable amounts for much-lower-quality 0.6 percent, have generators, and 0.5 percent have services. This indicates that they would be able solar home systems. to pay for electricity if it were available. There is a high degree of regional variation in these figures, particularly between the richer Coastal Eledrncuty from the Grid Regions and the Andean and Amazon Regions that As already noted, the Survey showed that only contain significant indigenous populations. The 39 percent of rural households currently had percentage of households in the Andean regions with access to grid electricity. In addition to regional access to grid electricity ranges from 22 percent in variations, access is strongly correlated with the north to 52 percent in the central region. In the exp~nditure quintile: 28 percent of households in Coastal regions, coverage of grid electricity ranges the poorest quintile have access, compared with from 35 percent in the north to about 71 percent in 49 percent in the top quintile. Electricity usage the south. Electricity access is lowest in the Amazon, among rural households in Peru is relatively low, at 18 percent. at an average of 27 kWh per month. This may be due to several factors, including a high tariff, Variations in Energy Expenditures unavailability of inexpensive appliances, and high The total monthly cash expenditure for all types prevalence of poverty.· of energy used in the household is estimated to be As a result of fixed charges, the average effective 25 soles per month, on average. This amounts to about rate for households that use small amounts of 9.7 percent of total household cash exp'enditures each electricity is relatively high. Currently, about 70 month. However, househol~ energy expenditure percent of households with a grid connection use varies significantly among regions and between less than 30 kWh per month. These households' xviii average effective electricity price is 0.76 soles per The use of dry cell batteries for specific uses kWh. However, the average effective price per kWh is very common among both grid and off-grid for households that use more than 30 kWh per month households in rural areas. Often, such batteries is only 0.46 soles per kWh. fulfill an energy niche that cannot be entirely met The proportion of total electricity used for though the use of grid electricity. However, it is also lighting is strongly dependent upon expenditure evident .that households with grid electricity are less quintile. The bottom quintile uses 39 percent of reliant on batteries for their electricity needs than total electricity consumption for lighting, while households without access to it. As a consequence, the top quintile uses only 21 percent. As income they save having to pay for what is a very expensive (expenditure) increases, the ability to purchase form of energy. expensive electric appliances increases, and thus a greater fraction of electricity is used for color TV, Small Generators and Solar Home Systems sound equipment, and refrigerators. Small generators and solar home systems in rural Radios are by far the most common type of Peru are uncommon. Overall, 0.6 percent of rural household electric appliance, with 66 percent of households, or an estimated 13,100 households, use electrified households owning one or more, followed small gasoline or diesel generators. The estimated by black-and-white televisions (37 percent of cost of using a generator is much lower than the households), color televisions (33 percent), and electric cost of using a car battery, and it would give far irons (25 percent). Appliance ownership variations by better service levels. It is likely that a significant region are in line with regional income disparities. barrier to the adoption of generators is their high upfront costs. Off-Grid Electricity Solar photovoltaic (PV) systems represent an People often assume that households without access option for providing electricity to households in to the electricity service from the grid do not use remote rural areas, where the costs of grid extension electricity. This is not the case. The electricity are particularly high. The use of solar systems may cost them more and they may use less of it, is quite rare in rural Peru because of a lack of but almost all households have some form of off- promotion of the use of such systems. Most of the grid electricity use. This is evidence of a pent-up households that would use a solar PV system now consumer demand for electricity and an indication use car batteries. Solar systems are estimated to be that people are willing to pay high prices for small present in 0.8 percent of all households, or about amounts of it. 16,700 rural households. Car and Dry tell Batteries Benefits of Rural Electrification Close to one-fifth of households in rural Peru without The benefits of electricity consumption can be electricity use car batteries for televisions and lights. broken into two categories: direct and indirect. This is an important indication of the very high value Direct benefits include improvements to lighting of electricity for people in rural areas, as the work and television viewing. Indirect benefits include and expense involved in charging car batteries is improved educational outcomes for children in homes not trivial. Battery costs vary across the expenditure with electricity and improved income-generation quintiles, with the poor paying higher prices per opportunities. Most of the quantitative work kilowatt-hour than the more wealthy households. described in the literature relates to estimating the The poorest quintile seems to purchase batteries of direct benefits. However, there is evidence that some significantly lower capacity, while paying similar direct benefits, such as improved lighting, give rise prices as the richer quintiles do for better batteries. to indirect benefits. xix Special Report Peru: National Survey of Rural Household Energy Use There are two principal methods for estimating Benefits of /Elee:trricity Use to Business direct benefits, or the willingness of consumers to About 13 percent of sampled houses reported a home pay for services: avoided cost and demand curve business, with a higher proportion in grid-electrified estimates. The former tends to underestimate value. households (18.3 percent) than households without This study uses demand curves to estimate the electricity (Z7 percent). Although the small number benefits of lighting, television viewing, income- of households using electricity from car batteries generation, and other services. have a similar proportion of home businesses as those connected to the grid (16.1 percent), it is clear Benefits of Electricity Use to 11.ighting that home businesses are concentrated in households Although various forms of energy are used by all income connected to the grid. groups, it is primarily the poor who depend on high-cost Willingness to pay (WTP) for electricity in non- and less-efficient alternatives to grid electricity, such as household applications may be estimated from the candles and kerosene, to provide lighting.. results of the business survey, which sampled 192 Using consumer surplus calculations, the report rural enterprises. Sixty-nine (69) percent had access shows the benefits in switching to different forms to the grid. Even a simple consideration of energy of lighting. Benefits from improved lighting range sources suggests that WTP for electricity for business from 17 to 90 soles/month/household, depending on is much higher than for domestic applications; 26 expenditure level and assumptions. The estimates have percent of unelectrified businesses use car batteries high variance, but even at the low end of the range, and 24 percent use small generators. the economic benefits are substantial. Not only do Total energy expenditures remain largely households with electricity enjoy much greater levels unchanged: 154 soles per month for electrified of service, but also they obtain a real income gain since enterprises versus 155 soles per month for unelectrified their total expenditure on lighting service decreases. enterprises. These energy expenditure data do not take into account the dramatic difference in enterprise Benefits of Electiricity Use to Communications incomes. The average monthly turnover (gross sales) in electrified enterprises is 3,520 soles/month, For radio, most basic calculations suggest that as opposed to 1,140 soles/month in unelectrified households without electricity would save 4.6 soles per enterprises. month if they were to use grid electricity. For television viewing, demand curve calculations find a total benefit Policy Implications or willingness to pay of 24.2 soles per month. Chapter 6 uses data from the Survey to consider policy issues relevant to the creation and sustainability of Benefits of Eledricity Use to Education and rural electrification programs: Health The Survey shows that children aged 6 to 18 in • Connection rates in electrified villages. Almost one- households with electricity read or study 65 minutes quarter of households without electricity are in per night, compared to 51 minutes for those without villages that are electrified. The most common electricity. Although school enrollments for children reason given for nonconnection in these villages aged 6 to 12 with and without grid access are is the upfront costs of connection, wiring, and comparable, school enrollment of children aged 13 to equipment. The financial sustainability of projects 18 in households with electricity is 82 percent, versus is strongly influenced by connecting as many 62 percent in households without access. households as possible, from which follows Although not quantified, the health benefits from that connection costs, perhaps including house reduced burns and respiratory effects from kerosene wiring, should be part of the overall cost eligible are major benefits of rural electrification. for subsidy. xx Execulive • Variations in electricity consumption levels. An . kWh, notwithstanding the POSE mechanism. indicative consumption threshold of 22 kWh/ Households in the lowest quintile capture only household/month is used in this report for 7.7 percent of the total POSE subsidy received by whether most rural electrification schemes all rural households, yet this quintile constitutes would be financially viable. Although the 20 percent of all households. The highest quintile average consumption in 374 electrified v~llages captures 32.6 percent of the benefit. In short, is 35 kWh/household/month, these averages the targeting performance of the POSE could show significant variation across regions. In be improved. Improvements in the targeting the Andean South region, the average is only performance could be achieved by further 15 kWh/household/month, and only 23 percent of lowering the POSE cap. If the 50 percent discount villages had consumption levels above 22 kWh/ were limited to 15 kWh/month and phased out household/month. This suggests that there is likely at 25 kWh/month, the share of benefits going to to be a significant problem with financial viability the lowest quintile would be 19 percent, while the of rural electrification in the Andean South. richest would receive less than 10 percent. • Growth of electricity consumption. One of the • Efficient lighting. The economic case for linking important assumptions in making financial future rural electrification projects with an projections of the viability of rural electrification efficient lighting program using compact projects is the rate of growth in consumption. At fluorescent lamps (CPLs) is compelling. Rural least based on the experience of those communities electrification costs per household are between prioritized by the current scheme (often the US$445 and $600, so an additional US$8-$9 poorest and most lacking in infrastructure access), for three CFLs per household would have little there is no evidence that annual consumption impact on rural electrification project budgets. growth per connected household would be much higher than the commonly assumed 0.5 to 1.0 Supplementary lnformc:ation percent per year. Therefore, the Survey results The Annexes provide additional information about suggest that these rates continue to be used in the Survey design and methodology (Annex 1), more projections. detailed findings (Annex 2), additional details about • Pricing policy. Those who consume small amounts estimating benefits (Annex 3), and the Survey itself of electricity pay relatively high prices per (Annex4). xxi 1 Introduction Peru is a country of extreme diversity, both in its both conventional grid extension and renewable geography and the socioeconomic conditions of its energy sources. citizens. This makes it a challenge for the government Detailed data were required in order to prepare the of Peru (GoP) to meet its targets to extend access to design of the Peru Rural Electrification Project. Data basic infrastructure services, including electricity, to were also needed to improve the rural electrification the dispersed population living in rural areas. Plans program and to analyze the economic and financial and targets have been in place for rural electrification aspects of rural electrification. The information since the early 1970s, but by 2005, only 39 percent of needed includes general socioeconomic information rural households had electricity service. Peru has on households, as well as detailed information on one of the lowest rural electrification rates in Latin their current energy use, energy expenditures, and America. ability/willingness to pay for electricity services. Until An estimated 6 million people in the now, such data have not been available. Consequently, predominantly poor rural areas of Peru do not it was decided to implement the National Survey have access to electricity. Together with the scarcity of Rural Household Energy Use (referred to as the of other infrastructure services, lack of electricity Survey in this publication),3 with the assistance of results in high costs for basic energy services, a lower the Energy Sector Management Assistance Program quality of life, poor medical care and education, and (ESMAP), to obtain information on the demand and limited opportunities for economic development. The use of electricity in rural areas of Peru. extremely high incidence of poverty in rural areas The Survey was conducted through the National of Peru highlights the importance of investing in Institute of Statistics and Information Technology provision of basic infrastructure, such as electricity, (Instituto Nacional de Estadfstica e Informatica, INEI) as part of the national rural development agenda. and experts on household energy surveys provided The MEM initiated a World Bank- and GEF- by the World Bank. INEI's Technical Department of assisted Rural Electrification Project in August 2006 Demographics and Social Indicators executed the to assist local distribution companies in reaching fieldwork and data processing from April through rural populations with well-targeted subsidies, July 2005 in the 24 departments (departamentos) of Peru. aiming at financing projects that would be financially It is essential to point out that the definition of sustainable after receiving a subsidy of a substantial rural population center that is used in the National part of the capital costs (World Bank 2006). The Survey of Rural Household Energy Use is different project aims to provide financing for investments from that used by INEI in the census. The definition in subprojects to supply electricity services to used by INEI for the purpose of the census is that about 160,000 currently unserved rural households, rural population centers are those with less than 100 businesses, and public facilities, such as schools and dwellings grouped contiguously. The definition used health clinics (serving about 800,000 people), using in the National Survey of Rural Household Energy Use 3 In Spanish, Encuesta de Consumo de Energia a Hogares en el Ambito Rural. Special Report Peru: Notional Survey of Rural Household Energy Use for rural population centers are those with less than Table 1.1 1000 dwellings grouped contiguously, a definition Population by Region and Area that better represents the target population for rural electrification programs. This difference in definition Population Percentage of rural population centers means that the data from Total. Living in Living in Region Population Rural Areas Rural Areas this survey cannot be directly compared with data Coastal North l 12 951, 147 24.3 from the census of other surveys conducted by INEI. Coastal The Survey covered 6,690 households with and Central 1,846,606 315,465 17.1 without electricity in rural areas of Peru. Rural Coastal South 713,042 173,413 24.3 areas were defined as those populations living in Andean North 2,270,580 2,057,476 90.6 aggregations of 1,000 households or less. The sample Andean was large enough to provide reliable estimations Central 4,096,006 2,445,860 59.7 about the survey population at seven regional levels: Andean South 3,632,728 1,885,401 51.9 Coastal North, Central, and South Regions, the Amazon 3,836,036 2,080,865 54.2 Andean North, Central, and South Regions, and the Lima Amazon Region. The expected standard deviation in Metropolitan 8,228,084 0 0 each region ranged from 0.021 to 0.050 (see Annex 1). Total 28,537,394 9,909,628 This report presents the main results of the Source: INEI, Enoho 2004-1, II, Ill ond IV rounds. Survey, and shows how Survey information can contribute to the analysis of important policy issues population, located mostly in urban areas. Only about in developing an improved rural electrification one-quarter of its population lives in rural areas, yet framework in Peru. 58 percent4 of those living in rural areas are poor. Income in the region is generated mainly through fishing, agriculture, and mining. Agricultural Geographical and Socioeconomic products include citrus fruit, corn, and potato. Diversity in Peru The Coastal Central region, including Lima, Each of Peru's seven geographical regions has unique contains both the largest percentage of the country's geography and distinct socioeconomic realities. Key to population and greatest share of its economy. Its understanding the slow pace of progress in bringing economy consists of mostly industrial production, as electricity access to rural households in Peru is an well as services, agriculture; fishing; livestock; lead, appreciation of the impact of this geographic and zinc, and silver mining; and tourism. The geography socioeconomic diversity. The country's geography is mostly flat with arid conditions, yet there are also ranges from the high-altitude Andean mountains, fertile valleys. The Andean chain borders to the east. through dense, lush Amazonian tropical rainforest, The Coastal Central region is also mostly urban, to .the dry, flat coastal desert plains. It is estimated with only 17 percent of its population living in rural that about 65 percent of Peru's rural population live areas. It has a rural poverty rate of 29 percent, which is in the Andean regions, while about 20 percent live low in comparison to the national rural po~erty average in the Amazon and 15 percent in the Coastal regions of 55 percent. :: Four percent live in extreme poverty, (Table 1.1). . much lower than the national average of 26 percent in The Coastal North is made up mostly of desert rural areas. Rural households in the Coastal Central and beaches, although there are also fertile valleys region spend around 744 soles per month, which is with citrus fruit cultivation. The Ecuadorian border much higher than the average expenditure of 482 soles lies to the north. This region has Peru's third-largest per month for rural households across all regions. 4Poverty figures for this report were calculated using the Encuesta Nacional de Hogares (ENAHO) 2004-I, II, III and N rounds, compiled by the Instituto Nacional de Estadfstica e Informatica (INEI). · 2 The Coastal South region is also flat and desert- where. m1;1ch of the population lives, making it difficult like, with some fertile irrigated valleys, the Andes to to raise any crops other than potatoes. Although the east, and the Chilean border to the south. It has there is a thriving tourism industry (mainly from the both the lowest population of Peru's seven regions and Cuzco-Machu Picchu area) and a large percentage the smallest number of people living in rural areas. of its income derives from natural gas production Approximately one-quarter of its population lives in (Camisea), the region also contains Peru's two poorest rural areas, and it has the lowest rural poverty index of departamentos, Huancavelica and Huanuco. The the regions, at 21 percent. The Coastal South also has majority of the rural population generates income the country's highest rural household expenditure, through agriculturalproduction, mainly potato, and about 775 soles per month. The economy depends subsistence farming. A little more than one-half of on fishing; copper mining; agriculture such as corn, the total population lives in rural areas. The Andean potato, and asparagus; production of wine and Pisco; South has the second highest poverty incidence in and production of poultry and other livestock. the country. Nearly 70 percent of rural households The landscape of the Andean North is a mixture are poor, and about 38 percent are extremely poor. of high peaks, plateaus, and deep gorges and valleys, The Amazon makes up 60 percent of Peru's total which makes the provision of basic infrastructure to land area. It is covered with thick tropical forests in the these areas very difficult and expensive. This region west and dense tropical vegetation in the center and also has one of the most expansive land areas of the east. As a result, the region remains largely unexplored regions. (The entire Andean region covers 30 percent and undeveloped. This makes the infrastructure, such of Peru's total land area.) The Andean North is the as grid-connected electricity, expensive. The Amazon world's sixth largest producer of gold, as well as a is one of the most populated regions, with 54 percent major producer of livestock, such as cattle and sheep, of its population in rural areas. Although not as poor and associated products, such as milk and cheese. It as the Andean regions, 58 percent of rural households also produces agricultural goods such as com, potato, in the Amazon are poor and 26 percent live in extreme and rice, and has some tourism. However, much of its poverty. The Amazon region mainly produces citrus rural population continues to depend on subsistence fruit and coffee, and also generates income through farming. It has the highest percent of population living tourism. It also produces rice and yucca, and there is in rural areas, at 90 percent, and the highest index of some petroleum mining. poverty in rural areas. Household income and expenditures are positively The Andean Central region has some of the correlated with urbanization and density of population. highest peaks in the world, particularly in the White The coastal regions are the most· commercialized, Andean chain, as well as valleys, gorges, and rivers. urban, and prosperous. Almost one-third of the As in the Andean North, this difficult geography country's population lives in Lima, but only 3 percent hinders the provision of public infrastructure such of its population lives in extreme poverty. Extreme as roads and electricity. Income in this region is poverty rates among urban populations range from generated from lead, zinc, and silver mining and 4 percent in the Central and South areas to 15 percent smelting. Potato and other root crops are other major in the Coastal North. This contrasts sharply with sources of income. Much of the rural population is conditions in the North and Central Andes, where . dedicated to subsistence farming. This Andean region a predominantly indigenous population engages in has the largest population of the country (excluding traditional lifestyles. In the Andean regions, between metropolitan Lima), with 60 percent of its inhabitants 38 and 47 percent of all households live in extreme living in rural areas. Sixty-eight percent of these rural poverty, and rural households on average have less households are poor, and 44 percent live in extreme than one-quarter of the average annual income per poverty. household in Lima (INEI 2005; World Bank 2005) (see The Andean South is characterized by high Table 1.2). According to data from the World Bank's altitudes, harsh winters, and strong winds in the areas Peru Poverty Assessment (2005), indigenous households 3 Special Report Peru: National Survey of Rural Household Energy Use Table 1.2 Electricity Sector Structure Poverty Incidence in Rural Areas (% of Households) Until the late 1970s, the GoP did not have systematic rural electrification policies or programs. Apart from Monthly Household a few pilot projects in rural communities during the Extreme Expenditure late 1960s, extension of electricity access was generally Region Poverty(%} Poverty(%} (Soles} 6 ad hoc and politically driven, aimed at gaining Coastal North 57.8 15.4 583 political support in rural areas. Rural electrification Coastal projects were neither clearly defined nor prioritized Central 29. l 4.2 744 according to potential rural electricity demand or Coastal South 20.9 4.1 755 financial viability. Andean North 77.8 47.2 271 Starting in the late 1970s, the government Andean introduced measures to try to increase the population's Central 68.5 44.1 343 access to electricity services. Early efforts focused on Andean South 69.3 37.6 292 urban and peri-urban areas, especially along the Amazon 58.3 26.4 471 more densely populated and prosperous coast, where Average 55.0 26.0 482 connection costs are lower and communities could be Sources: INEI, 2004, used for poverty figures, and 2005, used for easily connected to the national interconnected grid. household expenditure figures. From the early 1970s, the electricity sector in Peru was run by the public enterprise ELECTROPERU (ELP). Recognizing the enormity of the task to are, on average, poorer, less educated, and less healthy provide electricity access to rural areas, ELP created a than nonindigenous households.5 Directorate of Rural Electrification projects in 1976 to Access to infrastructure reflects income differences develop a national rural electrification plan, and the and geographical challenges. In 2003, 62 percent of Ministry of Energy and Mines (MEM) declared rural rural households had access to water and 49 percent electrification as a key goal (Carrasco 1989). to sanitation services. However, only 28 percent of In 1982, the General Electricity Law was passed. rural households had access to an unpaved road in One of its objectives was to expand the access to rural good condition (13 percent to a paved road), and areas at the least cost. The model adopted by ELP 9.3 percent of villages had a public phone. was to connect rural areas to the national network The association of poverty with geographical wherever possible through the construction of mini- dispersion is particularly pernicious. The poorest grid systems. Between 1980 arid 1986, ELP constructed region, the Andean North, has a population density approximately 42 mini-grid systems, mostly located of less than 0.2 inhabitants per square kilometer, in peri-urban areas (Carrasco 1989). To finance the compared to well over 4,000 in the richest part of the electrification projects, a tax was established on country around Lima (INEI 1993). This means, in effect, 25 percent of energy consumption above 160 kilowatt- that the places that are most expensive to reach for hours (kWh) per month, 50 percent of which was infrastructural services are also the least able to afford earmarked for rural electrification. Despite this effort, to pay for these services-with serious implications for bureaucratic complications and inadequate project the sustainability of electricity systems in these areas. information, combined with poor site selection and 5 In 2000, 70 percent of indigenous households lived in poverty versus 54 percent for the total population. The secondary school completion rate in 2003 was 27 percent for indigenous peoples and 48 percent for non- indigenous peoples. The under-1 mortality rate in 2000 was 54 per 1,000 live births for indigenous people, versus 34 for the total population. In addition, wasting and stunting levels in 2000 were roughly twice as high for indigenous households as for the total population (World Bank 2005). 6 Annual household expenditure: Conversion to dollars calculated using a rate of 3.23 soles/US$. 4 1 Introduction prioritization, made it difficult for ELP to reach its concession areas concentrated in small areas around project goals. A failure to properly train local staff urban centers and have an obligation to meet service also resulted in poor administration, operation, and requests only within 100 meters of the existing management of the mini-grid systems. network. There is thus no incentive for either public In 1992, a new legal and regulatory framework for or private companies to extend service to households the electricity sector was put in place through the Law outside these concession areas,. Areas with electric of Electric Concessions (Ley de Concessiones Electricas, concessions in Peru are dwarfed by areas with no LCE). In line with then-President Alberto Fujimori's service from a distribution company. Connected areas focus on economic reform, the LCE envisaged the are heavily concentrated in urban coastal areas, such private sector as the principal actor, with the public as Lima, while the majority of the rural population sector playing mainly a regulatory and supervisory remains unserved. role. As with many other Latin American countries In the 1992 restructuring of the sector, the during this period, the vertically integrated model was electricity tariff scheme was predicated on a full-cost replaced with a new structure in which generation, recovery. This situation prevailed until the middle transmission, and distribution were unbundled, with of 2001, with no explicit subsidies to electricity competitive markets operating in the generation and rates. In July 2001, the government announced commercialization markets, while transmission legislation establishing a "social tariff" for electricity and distribution was regulated, based on free-entry consumption (known as the Fonda de Compensaci6n and open access. A privatization program was Social Electrica, FOSE). Since July 2004, the level of established to break up the vertically integrated ELP ·subsidy has consisted of tariff reductions for monthly ,and transfer the assets into private hands. consumption up to 30 kWh, set at 25 percent for Prices for small retail users (known as regulated urban users supplied by the interconnected system users) were regulated, while a free market was and 62.5 percent for rural users supplied by isolated created for large industrial and commercial customers systems. For consumption between 31 and 100 with demand above 1,000 kW (free users). Price kWh, the reduction is gradual, from a maximum of setting was done on the principle of a reasonable 31.25 percent for rural users supplied by isolated return to compensate the costs of an "economically systems to a minimum of 7.5 percent for urban users efficient" service provider. The main regulatory supplied by the interconnected system. Consumers body, responsible for tariff setting, supervising, and who use more than 100 kWh per month pay a cross- monitoring the legal and technical regulations for the subsidy in proportion to their energy consumption electricity sector, was the Supervisory Commission above 100 kWh/month to finance the FOSE discount. for Energy Investments (Organismo Supervisor de la Statistics show that about 33 percent of all Inversion en Energia, OSINERG). residential users consume less than 30 kWh per month Despite attempts to extend privatization and another 35 percent have monthly consumption throughout Peru, factors such as high capital costs, low demand, and difficult geography have discouraged private investment outside Lima. There are two principal private distribution companies, EDELNOR Table 1.3 and Luz del Sur, created when ELP was privatized in 1994. They serve approximately half of the total electric Residential Subsidized Tariffs (Soles/kWh) market in Peru, primarily in the areas around Lima. Consumption Lima Rural Twelve other regional electric distribution kWh/month Consumer Consumer companies provide service in Peru, as well as the few Less or to 30 kWh 0.242 0.201 smaller-scale municipality electric companies-all From 31 to 100 kWh 0.322 0.402 of which are publicly owned. These companies hold Source: INEI, 2005. 5 Special Report Peru: National Survey of Rural Household Energy Use between 31 and 100 kWh. This means that 68 percent Table 1.4 of all residential consumers receive some electricity Latin American and Caribbean Region Electricity price subsidy. This cost of the subsidy represents Coverage, by Percentage of Coverage a surcharge of 3 percent cost of electricity to the users providing the subsidy (those with monthly Population Population, Electricity w/o consumption over 100 kWh.). Table 1.3 shows 2005 Coverage Electricity electricity tariffs, including the FOSE subsidy, for a (millions} (%) (millions) residential user with monthly consumption up to Nicaragua 5.5 54 2.5 lOOkWh. Bolivia 9.2 69 2.8 It should also be noted that rural tariffs vary by Honduras 7.2 69 2.2 location, based on the tariff calculated by OSINERG Peru 28 78 6.3 for the areas of each distribution company. The El Salvador 6.9 82 1.2 price paid per kWh from the Survey (including Guatemala 12.6 83 2.1 fixed and variable, as well as other charges such as Panama 3.2 86 0.5 lighting), varied from a low of 0.47 soles/kWh in the Paraguay 6.2 87 0.8 Coastal South Coast region to a high of 0.83 soles/ Ecuador 13.2 89 1.5 kWh in the Andean South region. The fixed charge Brazil 186.4 92 15.8 for connection, until recently paid by the customer, averaged about 320 soles per connection. Under Venezuela 26.6 92 2.2 the 2006 Rural Electrification law, the distribution Mexico 103.2 93 6.8 company will pay the connection charge. The Colombia 41.5 94 2.7 connection facilities (wire drop and meter) will be Argentina 38.7 95 2.1 owned by the distribution company and will be Uruguay 3.5 95 0.2 recovered through the distribution value-added Chile 16.3 98 0.3 charge as part of the tariff. Costa Rica 4.3 99 0.1 The National Financing Fund supervises state- Total 512.4 90 50. l owned distribution companies for State Enterprise Sources: CIER, ECLAC, Official statistics in the case of Colombia, Activity (Fondo Nacional de Financiamiento de la Actividad Mexico, and Chile. Empresarial del Estado, FONAFE). FONAFE is a state organization that holds assets, sets policies, and directs Directorate (DEP) within the Ministry of Energy and investments of regional distribution enterprises. Mines (MEM) as a project implementation branch whose principal objective is to extend electricity ·access, mainly in rural areas. The primary function Rural Electrification to Date of the DEP is to define and implement the rural Overall, electricity coverage rates are lower than in electrification plan, financing or cofinancing the most countries in Latin America, at 78 percent. In majority of these projects and directly implementing comparison, the coverage rate is 89 percent in Ecuador, them by contracting with construction firms. which has roughly the same per capita income (see The DEP prepares a national rural electrification Table 1.4). As noted earlier, an estimated 6 million plan that sets out a list of projects to be developed, people in the predominantly poofrural areas of Peru annual investment budgets, and sources of funding. do not have access to electricity. The plan has a 10-year horizon and is updated The low level of rural electrification in Peru reflects annually, reflecting program progress, new policies, the fact that the framework under the Electricity Law and prioritization and allocation of economic failed to address rural electrification. To fill this gap, resources. The Plan Nacional de Electrificaci6n Rural a 1993 Supreme Decree created the Executive Project 2006-2015 aims to increase the national coverage rate 6 1 Introduction from 78 percent in 2006 to 93 percent in 2015, at a total The activities of the electricity distribution cost of US$929 million (MEM 2007). companies within their concessions, and of the The DEP ·performs all of the administrative, DEP and social funds such as FONCODES in rural technical, and/or financial activities required areas, have increased national coverage levels from to develop projects (directly or through service 57 percent in 1993 to 78 percent in 2006. Although contracting), including prefeasibility and feasibility coverage is at approximately 94 percent in urban ·studies, procurement, contracting, execution- of the areas, it is still only about 39 percent in rural areas works, supervision, and inspection until the service (INEI, 2005). The total investment by the DEP to begins. Financing for these projects comes from the 2004 was just over US$600 million, with an annual central government's budget. The constructed systems average of about US$50 million. The DEP completed are later transferred to distribution companies or to 608 projects during this time period. About 4.8 the Electric Infrastructure Administration Enterprise million people benefited from these projects (ADINELSA), a government holding company, as (1 million households). The average amount of described later in this chapter. kilowatt-hours (kWh) consumed per month by each The first step in the DEP process is for the household that benefited from the DEP projects is community, local or regional government to submit around 20 kWh. a letter of request. to the DEP. The DEP evaluates the Once DEP or FONCODES projects are project based on technical criteria (actual project commissioned, ownership of the fixed assets is state, electric infrastructure, provincial electricity transf~rred to distribution companies. Where these coefficient), economic criteria (actual social net assets cannot be transferred to electricity distribution value, investment/capita), and socioeconomic criteria companies-usually in areas located outside the (poverty index, geographic location). An engineer geographical limits of the regional electricity then visits and evaluates the site and draws up the companies-they are transferred to the Electric technical plan for project implementation, followed Infrastructure Administration Enterprise (Empresa by the preparation of prefeasibility and feasibility de Administraci6n de Infraestructura Electrica S.A., studies. The DEP then, through a bidding process, ADINELSA), a state company formed to administer contracts the construction of selected projects. the fixed assets of the DEP program and supervise After the project is constructed, administration is the operation of the isolated rural electricity systems. transferred to the primary electricity distributor in ADINELSA is in charge of administering the the region or, if it is an isolated system, to ADINELSA. electricity installations and delegates the operation In addition to the DEP, the National Fund for and maintenance of the facilities to concessionary Compensation and Social Development of Peru enterprises or municipalities. (FONCODES) also played a part in rural access extension in the 1990s. Created in 1991 as a temporary autonomous, decentralized agency that reports Key Rural Electrification Issues directly to the executive branch of the GoP, it was The first and most important issue for the rural designed to improve the living conditions of the poor, electrification program is adequate financing. There create jobs, help to meet the basic needs of the poor, is a need for sustained and predictable financing and encourage the poor to take part in their own of the subsidies required. Funding for the rural development. Between 1991 and 1996, FONCODES. electrification projects constructed by the DEP or invested more than US$57 million in 1,733 energy FONCODES has come almost entirely from the infrastructure projects. FONCODES was originally Treasury, with some contributions from other state given funds to cofinance rural development projects, entities and regional and local governments. The but since decentralization, the funds are given directly levels of investments have dropped significantly from to regional governments. a peak of US$135 million in 1996 to about $40 million 7 Special Report Peru: National Survey of Rural Household Energy Use per year in 2004-2005. Although it may be unrealistic companies and to broaden the involvement of additional to expect to reach the GoP's target, which would actors in project development. Aside from a few require almost a doubling of current budget levels, exceptions, the municipalities have also not participated mobilizing cofinancing from distribution companies, in electric service provision. Instead, they have taken as well as from local and regional governments, on the role of lobbying on behalf of local demands could help. for obtaining electricity service and contributing to A second key concern for rural electrification financing the electricity projects (Aragon 2004). projects has been financial and technical sustainability In the Rural Electrification Plan of 2007, the GoP during the operation of the projects. Projects that are aims to increase national coverage from 78 percent transferred to distribution companies, and especially in 2006 to 88.5 percent in 2011 and 93 percent in 2015. ADINELSA, often have costs for operation and To meet this commitment, investments benefiting maintenance that are higher than the revenues from 4.8 million people and totaling US$929 million the tariff. ADINELSA, for example, must continuously between 2006 and 2015 are planned (MEM 2007). subsidize the operators of its projects, and is, as Most of these investments are planned in rural areas. a consequence, facing increasingly heavy losses, To accomplish this ambitious task, the GoP will with operating losses at US$2.8 million in 2003 and need to improve the rural electrification framework US$4.7 million in 2004. to increase economic efficiency and attract broader Part of the problem occurs because the weighting participation and financing from communities, factors for project selection have resulted in priority regional governments, and electricity service being given to areas with low provincial electricity providers. Congress passed the General Law of Rural coverage and a high incidence of poverty as opposed Electrification on July 1, 2006. The General Law creates to criteria such as economic efficiency, minimum a Rural Electrification Fund and provides a base from subsidy, or maximum economic. benefit. This which specific regulations can be developed for an undermines the long-run project sustainability and improved strategy. imposes an excessive burden on the distribution The MEM initiated a World Bank- and GEF- companies or ADINELSA, which must then subsidize assisted Rural Electrification Project in August 2006 projects whose operation and maintenance costs are to assist local distribution companies in reaching higher than tariff revenues. rural populations with well-targeted subsidies, A third key issue is that the DEP and FONCODES aiming at financing projects that would be financially have followed centralized processes with very limited sustainable after receiving a subsidy of a substantial participation of distribution companies in the process part of the capital costs (World Bank 2006). The of identification, selection, and development of project aims to provide financing for investments projects. The distribution companies, for their part, in subprojects to supply electricity services to have generally lost interest in participation in the about 160,000 currently unserved rural households, extension of rural electricity service, as there have businesses, and public facilities, such as schools and been no incentives available to them· to cover the health clinics (serving about 800,000 people), using capital costs for grid extension. both conventional grid extension and renewable Despite MEM's significant achievements in energy sources. The Project also includes a component improving electrification in Peru, limitations on fiscal aimed at increasing productive uses of electricity. It is budget allocations and problems with the existing hoped that lessons learned during the implementation approach suggest the need for an overhaul of the of this Project would assist the Ministry to develop a model. An improved strategy is required to promote more sustainable and cost-effective strategy for rural the involvement of public and private distribution electrification. 8 2 Household Energy. Use and Expenditure Knowledge about existing energy use and expenditure kerosene for lighting. Electricity is used by 39 percent patterns of rural households is essential for of all households. A surprisingly high 11 percent formulating energy policies and programs to enhance of all households use car batteries to run electric living standards and alleviate poverty in rural Peru. appliances, indicating a high, unmet demand and It enables energy planners to determine the potential willingness to pay for electricity services. LPG is willingness and ability of rural households to pay for used mainly for cooking by an estimated 14 percent modern energy, such as electricity, kerosene, liquefied of all households. A tiny fraction of households, petroleum gas (LPG), and off-grid electricity sources 0.6 percent, have their own generators; and 0.5 percent such as car batteries. It also facilitates assessment have solar home systems. of the potential demand for such modern and clean There is a high degree of regional variation in energy sources. these figures, particularly between the richer Coastal This chapter presents the information from the regions and the Andean and Amazon Regions Survey on current energy use and expenditures in that contain significant indigenous populations. rural households in Peru. It compares energy usage Electricity use is highest in the Coastal Central at among households in different regions, different 60 percent and Coastal South at 71 percent, and household expenditure quintile classifications, and lowest in the Amazon at 18 percent. Similarly, LPG different categories of households, with and without use is also highest in the Coastal Central and South access to grid electricity. It should be noted that the areas at 63 and 53 percent, and lowest in the Andean report uses total household monthly expenditure as North and Amazon Regions at 5 and 7 percent. Car a proxy for household monthly income. battery use is concentrated in the Coastal North and Central Regions. The use of dung is concentrated in the Andean South, and to a lesser extent in the Household Energy Use Andean Central region. The Survey shows that rural households in Peru, More households in the three Andean Regions use like rural households elsewhere in the world, rely candles and kerosene than households living in the on various sources of energy for lighting, cooking, three coastal regions. This is expected because fewer and appliances, including agriculture residue, households living in the mountains have access to grid fuelwood,. animal dung, candles, kerosene, electricity, electricity than households on the coast. The percentage liquid petroleum gas (LPG), dry cell batteries, car of households in the Andean Regions with access. to batteries, generators, and even solar home systems grid electricity ranges from 22 percent in the North to (Table 2.1). More than 84 percent of households 52 percent in the Central region. In the Coastal regions, rely on fuelwood for cooking, while 24 percent use coverage of grid electricity ranges from 35 percent in dung and 11 percent use agriculture residue. An the North to about 71 percent in the South. estimated 74 percent of all households use dry cells Kerosene is used by57 percent of households for for small appliances such as radios and flashlights. lighting and cooking, although the overwhelming About 60 percent of all households use candles and majority of kerosene consumers use it only or mainly 9 Peru: National of Rural Household Use Table .2.1 Percentage of Households that Use Each' Type of Energy by Region Coastal Regions Andean Regions All North Central South North Central South Amazon Regions Grid Electricity 35 60 71 22 52 44 18 39 Fuelwood 85 74 68 94 92 64 95 84 Dry cell battery 71 51 55 78 66 74 91 74 Kerosene 71 32 31 71 44 52 73 57 Candles 47 53 60 56 69 66 46 60 Car 31 23 13 9 8 7 15 11 LPG 28 63 53 5 17 10 7 14 8 7 5 5 18 13 3 11 0.4 0.5 15 3.6 26 65 0.1 25 Solar PV 0.3 0.1 0.1 0.4 0.9 0.5 Small generators 0.9 0.2 0.9 0.6 All househplds (OOOs) 156.4 75.3 27.8 362.0 634.2 565.0 383.4 2,204.2 Source: INEI, 2005. for lighting. Among these kerosene users, about for. dry cell batteries. This is due primarily to the 83 percent use it exclusively for lighting. Only unique, portable nature of the dry cell battery. 4 percent use it for cooking and 3 percent use it for Over 80 percent of households in rural Peru rely on both lighting and cooking. The remaining 10 percent fuelwood for cooking. Not surprisingly, use of fuelwood of kerosene users use it for other purposes including varies by region, reflecting availability differences. starting a fire, for home appliances, and home Almost 95 percent of households in the Amazon region business purposes (Figure 2.1). use fuelwood for cooking due to its abundance. In Significantly more households in the three contrast, fuelwood use is lowest in the Andean South Coastal regions use LPG than in other regions. and Coastal South regions, at 64 and 68 percent. These households use LPG almost exclusively At the bottom of the fuel ladder are agriculture as cooking fuel. Less than 1 percent of those. residue and animal dung, both of which are used by households that use LPG report using it for lighting. a significantly smaller proportion of households than This result is not surprising, given that LPG is fuelwood. However, 65 percent of households in the more expensive than other fuels. Furthermore, LPG Andean South and 26 percent in the Andean Central requires better roads to distribute it to end users. regions use animal dung as a cooking fuel (Table The Coastal regions have better road networks than 2.1). These two regions have a high share of poor the rest of the country. and indigenous households. Agriculture residue and Dry cell batteries are used extensively in rural animal dung are widely available and are typically the households, despite a very high cost per equivalent fuel of choice for the poor, since family members can kWh. The percentage of households using dry cell collect these fuels. Furthermore, due to their terrain batteries is highest in the Amazon at 91 percent, where and topography, fuelwood is less abundant in these grid electricity penetration is lowest, and lowest in regions than in other parts of the country. the Coastal Central and South regions, where grid Many of the differences across regions can be electricity penetration is highest. The availability of explained by differences in income. As shown in grid electricity lowers but does not eliminate demand Figure 2.2, the Coastal regions have a lower proportion 10 2 Household Energy Use and Expenditure Figure 2.1 Application of Kerosene Users (Users Only) (3.0%) Lighting & Cooking (83.0%) Lighting (4.0%) Cooking (10.0%) Other Purposes Source: INEI, 2005. Figure 2.2 Expenditure Differences Across Regions: Fraction of Households in Each Expenditure Quintile by Region c 0 ] Coastal Coastal Coastal Andean Andean Andean North Central South North Central South Amazon 5 (Richest) 38% 55% 54% 8% 18% 11% 32% 4 33% 31% 28% 16% 18% 16% 25% 3 19% 10% 13% 22% 18% 23% 21% 2 7% 2% 4% 30% 20% 27% 11% 1 (Poorest) 4% 2% 2% 25% 27% 42% 10% Source: INEI, 2005. of poor households, while the proportion of poor Andean regions are in the poorest quintile, compared households is much higher in the Andean Regions. with only 2 to 4 percent in the Coastal regions. For instance, 38 percent of households in the Coastal These patterns reflect the findings of the World North are in the top expenditure quintile, compared Bank's Poverty Assessment (World Bank 2005a), with only 8 percent in the Andean North. Similarly, which noted that poverty in rural Peru is higher . between 25 and 27 percent of households in the in the Andean and Amazon regions than in the 11 Peru: National of Rural Household Use Figure 2.3 Households Using Modern Energy by Expenditure Quintile 100 I (/) "O 80 ------ --------' ----------orv c61I sattefy I I 1 I 0 .r. Q) I I 1 I I ~ 60 -t--- l ---- - J_ - _,_ - -l- - - - - - - I I -1- 1 1 I ::c I 1 1 I 0 I Grid Electricit~ 1 & I _J... _ _ _ _ - I ~ 40 - -1- - - - - - - - - L Q) LPG I e if. I I 20 .,. - - t- ---- - _,,_~:-at.-----;,- cir "Battery - ~ I 1 Sdlar PV, Genset : 0 1 (Poorest) 2 3 4 5 (Richest) Expenditure Quintile Source: INEI, 2005. three Coastal. Regions. 7 The report also pointed including dung and agriculture residue (so-called out that most of the regiop.al variations in poverty inferior goods). rates can be attributed to variations in household 3. Traditional energy forms that show remarkably characteristics and in access to basic services and small variation across expenditure quintiles, such road infrastructure, rather than to geographical as candles, kerosene, and fuelwood. differences per se, such as altitude and temperature. In other words, observationally equivalent households For the modem energy forms whose use increases ·have similar probabilities of being poor irrespective with income, the results are consistent with worldwide of the geographic characteristics of their region of experience. For example, only 27 percent of households residence. in the lowest expenditure quintile use grid electricity, In relation to expenditures, energy used in rural compared with 38 percent and 50 percent in the middle areas can be classified in three main categories: and top quintiles, respectively. LPG and car batteries exhibit similar trends, but at much lower initial levels. 1. Modem energy forms (or energy such as grid LPG use jumps from 1 percent of households in the electricity that requires higher income to poorest quintile to 27 percent in the richest Car battery purchase the appliances needed to utilize it), such use goes from 4 percent in the poorest quintile to as LPG, car batteries, and electricity, whose use 19 percent in the richest. Use of solar photovoltaics increases with increasing expenditures/income (PV) and small generators are both extremely low at (so-called normal goods). all expenditure levels (Figure 2.3). 2. Traditional energy forms whose use falls The use of dung drops significantly from significantly with increasing expenditure levels, 31 percent of households in the poorest quintile to 7Comparing Table 1.2 and Figure 2.2, it can be seen that Quintiles 1 and 2 correspond to households living in extreme poverty in all · regions; Quintile 3 in all regions and Quintile 4 in Coastal North and Amazon regions correspond to households living in poverty; and the remainder correspond to households that are not living in poverty, i.e. Quintile 5 in all regions and Quintile 4 in all regions except Coastal North and Amazon. 12 2 Household Energy Use and Expenditure Figure 2.4 Households Using Traditional Fuel by Expenditure Quintile 100 I I I • I I I • • I I I I I I 80 -i- - - -1 I --------1---- ___ -~ _______ ,,. fir.e.wgQd___ ~ "' "C :g I I, I 1 I I I I Q) I I I f I "' 5 60 ~a~C'!:nd~e- -i- :r: 0 . Q) ' , :~Kerosene Cl I ~ 40 -~- - ----r---------r Q) I I !: ~ I .1 l I I 2.0 -1- - - - - - -1- - - - - - r- - t- ~ 1 : ~--oung 1 • 1 ----·it----....,·it-----4.ilAg. residue I 0 1 (Poorest) 2 3 4 5 (Richest) Expenditure Quintile Source: INEI, 2005. 15 percent of households in the richest (see Figure 2:4). following section provides a descriptive analysis of The use of candles increases slightly with income rural household energy expenditure in Peru, by region levels, while kerosene use declines modestly. Yet, even and then by expenditure quintile. in the top quintile, 51 percent of households report Household expenditures on energy are highest using kerosene. This reflects the substantial number of in the three Coastal regions and lowest in the unelectrified households in the top quintile, as well as Andean North (Figure 2.5). Energy expenditures for the fact that kerosene is still used for lighting in many households living in the Coastal Central and South electrified households (although as shown below, the regions are about 2.5 times higher than those for quantities of kerosene used in electrified households households living in the three Andean regions and and the corresponding expenditures are very small). the Amazon region. Regional disparities are partially explained by the fact that households from different regions rely on Energy Expenditure different types of fuel, which have different prices and Household surveys generally show that energy varied availability (Figure 2.5). The biggest differences expenditures by households, for lighting, cooking, come from spending on kerosene, fuelwood, LPG, and and appliance usage, account for 5 to 10 percent of electricity. Household spending on these four types all household expenditures. Based on the Survey, the of energy is much higher in the three Coastal regions. total monthly cash expenditure for all types of energy Energy expenditure represents a heavier burden used in the household is estimated to be 25 soles per for households in the three Andean regions than month, on average. This amounts to about 9.7 percent for households in all other regions of the country. of total household cash expenditures each month. Although monthly expenditure of households in However, household energy expenditure varies the Andean regions is significantly lower than that significantly among regions and between financially of households in the Coastal regions, their energy better-off households and poorer households. The expenditure accounts for 10 to 12 percent of total 13 Special Report Peru: Notional Survey of Rural Household Energy Use Figure 2.5 Household Monthly Spending on Energy by Region and Type (Soles/Month) I I I I I I I I I I I I I I I I I I I I I I I I 60 ---, I I I - I ---1- I - --,- I 1 ---1 I I I I I I I I I I I I I I I I I I I I I ·I I I I I I ____ 1____ -L-- 40 ---L-- . I 1 I --L--- 1 _J. __ I - I I I I I I I I I ·I I I I I I I 0 Region Coastal N. Coastal C. Coastal S. Andean N. AndeanC. Andean S. Amazon all Regions I!! Candle 1.29 2.81 2.66 1.65 2.34 2.3 i.7 2.05 Ill Kerosene 10.34 7.33 8.18 5.37 3.67 4.13 6.89 5.28 !!I Gen-Set 0.38 0.44 - - - 0.05 0.2!'.l 0.1 0 Dry Cell Batt 3.18 3.22 2.9 3.39 3.57 3.24 6.79 3.97 Ii! Car Battery 1.89 2.83 1.14 0.51 0.43 0.39 0.99 0.72 El Grid Electricity 6.5 15.71 16.63 2.34 6.91 3.84 2.66 5.03 0 LPG 8.55 23.19 19.34 1.68 5.63 2.73 2.31 4.64 0 Firewood 5.09 3.02 4.2 3.81 3.87 3.21 1.2 3.29 Source: INEI, 2005. household expenditure. Conversely, the energy (see Table 2.2). Although poor households spend less expenditure of households living in the Coastal on energy than nonpoor households, their energy regions accounts for only 8 to 10 percent of total spending accounts for a larger portion of their income. household expenditure. The World Bank report Households in the lowest quintile spend about Opportunity for All: Peru Poverty Assessment (World 17 percent of their total monthly expenditures on Bank 2005a) pointed out that poverty in rural Peru is energy, while households in all other quintiles higher in the Andean and Amazon region than in the spend less than 10 percent. Part of the reason for three Coastal Regions. Therefore, the financial burden this discrepancy is that the poor often lack access to of energy expenditure on households in the Andean relatively cheap grid electricity. regions further exacerbates poverty conditions. A comparison of household energy spending among households in different expenditure quintiles Comparison of Households with shows a positive relationship between household and without Access to Grid Electricity energy spending and household financial well being for all fuel types. Households in the lowest Socioeconomic Chall'Cllderistks expenditure quintile spend an average of 9 soles Households with access to grid electricity are per month on energy. Energy expenditures for the financially better off than households without access s~cond, third, fourth, and richest quintile average to grid electricity. As already mentioned, the average 15, 21, 31, and close to 49 soles per month, respectively monthly expenditure for grid-connected households 14 2 Household Table 2.2 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles (Users Only) 1 (Poorest) 2 3 4 5 (Richest) All < 113 Soles 113-201 Soles 201-321 Soles 321-533 Soles > 533 Soles Grid Electricity 7.4 8.5 10.4 14.2 22.5 13.6 Candle 2.7 23.0 3.6 3.6 4.1 3.4 Kerosene 4.9 6.6 8.8 11.7 16.2 9.3 13.0 29.0 38.1 33.2 3.4 4.5 5.3 6.0 7.3 5.4 5.2 5.1 5.8 6.7 7.4 6.5 LPG 20.4 20.6 26.2 30.1 37.1 32.6 Fuelwood 13.6 17.9 22.9 27.7 36.0 26.6 All energy 9.4 15.3 20.6 31.l 49.1 25.1 % of total 17.1% 9.9% 8.2% 7.4% 5.8% 9.7% Source: INEI, 2005. · Table 2.3 Total Househ()ld Expenditure and Education by Electrification Status Grid Eledricity With Access Without Access All Areas No schooling 10% 16% 14% Primary education 51% 63% 58% 30% 18% 23% education 9% 3% 5% Population 839,581 1,326,075 2, 165,656 Total Household Exp. (Soles/Month) 430 317 361 Population 851,510 1,352,705 2,204,215 Total Users (Electricity & Electricity Substitutes) 845,522 1,340,491 2, 186,013 Source: INEI, 2005. is 430 soles versus 317 soles for households without or ethnic minority between households with and a grid connection. Another disparity between without access to grid electricity. Almost all children households with and without access to grid electricity between ages 6 and 18 are attending school, regardless is the educational level of the head of household. of their household's electrification status. Howev:er, Thirty-nine percent of grid-connected households studies have shown that electricity enhances are _headed by someone with at least a secondary children's education. For example, electricity allows education, compared with only 21 percent of children to study and/or do homework at night, unelectrified households (Table 2.3). allows schools to use modern educational equipment, There are no differences in household size, and enables children to gain access to computers and number of children at home, education of children, the Internet. These benefits are further discussed in 15 Special Report Peru: National Survey of Rural Household Energy Use Table 2.4 Percentage of Households that Use Each Type of Energy by Electrification Status Electrification Status Electrified (%) Unelectrified (%) All Households (%) Candle 51 65 60 Kerosene 20 80 57 Small 0.0 0.6 Dry cell battery 55 86 74 Car battery 0.7 18 11 LPG 28 6 14 Fuelwood 81 86 84 Solar PV 0.0 0.8 0.5 12 10 11 26 24 25 851.5 1,352.7 Source: INEI, 2005. Chapter 5. Children living in households without the two groups. These differences reflect the fact that access to grid electricity would be at a disadvantage. electrified households are financially better off than unelectrified households. Therefore, a higher Energy Use percentage of electrified households use LPG As previously noted, only 39 percent of rural · and lower percentage use fuelwood relative to households have access to grid electricity. Rural unelectrified households. In rural Peru, LPG is used households without electricity rely on traditional fuels primarily as a cooking fuel, although a tiny fraction such as candles and kerosene for lighting. Among the of households use it for lighting. 8 Availability of 1.3 million rural households without electricity, the · LPG is still limited in many rural areas, because overwhelming majority-80 percent-use kerosene. it requires a good transportation network for Although kerosene can be used for both lighting and distribution and high upfront costs, including a cooking, households without electricity that consume deposit for the LPG cylinder. kerosene use it primarily for lighting. Similarly, about Over half of households with electricity continue to 65 percent of unelectrified households use candles for use kerosene, candles, or both to supplement electricity lighting (Table 2.4). lighting, with significant differences by region (see None of the rural households with access to Figure 2.6). This is likely a result of interruptions in electricity use it for cooking. This is similar to rural electricity service in some rural areas (see discussion households elsewhere, since the use of electricity of small generators). The percentage of households for cooking is still more expensive than traditional with access to grid electricity who use candles and or fossil fuels. Switching to electricity for cooking kerosene lamps ranges from 33 percent in the Coastal usually takes decades, and households in rural Peru Central region to around 60 percent in the Coastal have not yet made this transition. North, Andean Central, and Andean South regions. At A comparison of LPG usage between households 61 percent, the Amazon region has the highest with and without access to a grid electricity proportion of grid-connected candle and kerosene connection shows significant differences between lamp users. 6 While about 27 percent of household surveyed, or 6,000 households, reported using LPG, only 9 households in the survey reported using LPG for lighting. 16 _.,' 2 Household Energy Use and Expendilure Figure 2.6 Percentage of· Households Maintaining Kerosene and Candles to Supplement Electric Lighting 00,-~~~~~~~~~~~~~~~~~~~~~~~--'-.~ 60 60% 61o/~ (I) 58% 58% 56%,. "C 0 ..c:: (I) 50% .. (/) ::l 0 40 44% ::r: 0 33%' :,e 0 20 Oc.__-'--~~~-'-.L_L.-'.L..L....L......L..j_J__j_L.:_j__J_J__:_L..J~~~L_LJ_J Coastal North Coastal South Andean Central Amazon ·National Average Coastal Central Andean North Andean South Source: INEI, 2005. Table 2.5 Household Monthly Cash Expenditure on Electricity and Lighting fuels/Energy by Electrification Status (Users Only) · With Electricity Without Electricity All Households (Soles/month/HH) (Soles/ month/HH) (Soles/month/HH) .. Electricity (Grid) 13.63 13.63 Candles l.32 4.49 3.43 Kerosene (light only) 3.92 7.98 7.67 LPG 18.26 16.24 17.05 Small generators . 28.50 33.31 33.20 cell batteries 3.68 6.04 5.36 Car batteries 5.72 6.61 6.60 All Expenditures (Electricily and electricily substitutes) 16.26 15.44 15.76 Source: INEI, 2005. Energy Expenduturres (Table 2.5). In other words, households without Since electricity is not used for cooking in rural electricity are paying comparable amounts for much- households in Peru, this section focuses primarily lower-quality services. on household expenditures for noncooking energy There is much greater variation in energy use. The most important finding is that households expenditures as a fraction of total expenditures in with electricity spend only marginally more on grid households with electricity compared to households electricity and.electricity substitutes (16.3 soles per without electricity. In households with electricity, month) than households without electricity spend energy expenditures as a percentage of total on electricity substitutes alone (15.4 soles per month) expenditures range from a low of ZO percent in the 17 Peru: National of Rural Household Use Figure 2.7 Household Energy Expenditure as a Percentage of Total Expenditure by Region 12.9% 9.9% • 9.5% 8% 7.5% electrified 7.0% 6% '-'~~--'-~~---'-~~-'-~~--'--~~~~~~~~~~~'-' Coastal North Coastal South Andean Central Amazon National Average Coastal Central Andean North Andean South Source: INEI, 2005. Amazon region to a high of 12.9 percent in the Ande~ The average household with grid electricity Central region. In households without electricity, the · spends 84 percent of its noncooking energy budget range is between 7.8 percent in the Coastal North to on electricity, while the remainder is spent on 10.9 percent in the Andean Central region (Figure 2.7). supplemental lighting fuels like candles, kerosene, However, a comparison of spending by quintile and dry cell batteries. For households without grid shows that, with the exception of households in the electricity, the largest portion of noncooking energy lowest quintile, all households without access to grid spending is for candles and kerosene fuel for lamp electricity spend slightly more on lighting fuels/ lighting. The average monthly expenditure for candles energy and other sources of electricity (Figure 2.8). and kerosene lamp lighting among households This suggests that households that have no access to without access to grid electricity is close to 12 soles grid electricity have the ability to pay for monthly per month. The remaining 7 or 8 soles are spent on electricity services by reallocating their lighting fuels/ dry cell batteries and other sources. Households with energy monthly budget to electricity. no access to grid electricity that use car batteries for Households without access to grid electricity home electricity supply spend as much as 7 soles per spend about 15 soles per month for lighting fuels and month for car battery recharging fees alone. electric energy sources including candles, kerosene for lamp lighting, LPG for lighting, dry cell batteries, car battery recharging fees, and diesel or gasoline fuel Conclusions for generators for electricity supply (see Table 2.5). Rural households in Peru still have limited access Households with access to grid electricity spend about to modern fuels. The majority of lower-income 16 soles per month for electricity and supplemental households still rely on traditional fuels (kerosene, fuels for lighting such as candles, kerosene, and fuelwood, and agriculture residue) for lighting and · LPG, as well as supplemental sources of electricity cooking. Higher-income households rely more on including dry-cell batteries, car batteries, and diesel/ modem fuels such as grid electricity, car batteries, gasoline fuel for generators. and LPG. Since grid electricity is only available to 18 2 Household Energy Use and Expenditure Figure 2.8 Household Expenditures on Electricily and Other Lighting fuels/Energy by Expenditure Quintiles (Soles per Month) 30 I I 25 -r- ---;-- -i--- ---- - -t-- - I I I I I I l I 20 l------ I ----~------ -~--~- _,_ I :S c ' I 0 I ~ I I I ____ -~i~h9~t_G!i.<;i !-:::c:_e~s' Q) 0 _L __ 15 - en I I I I I 10 ----r- --------,-- ---- - -r I I . I I I I· I I 5 '----'~~~~~~'--~~~~~-'--~~~~~-'--~~~~~_i._I 1 (Poorest) 2 3 4 5 (Richest) Expenditure Quintile Source: INEI, 2005. less than half of rural households, the majority of the lowest to highest quintile, and similarly from rural households are still using kerosene lamps and 17 percent of all expenditures in the lowest quintile candles for lighting. to less than 6 percent in the highest quintile. Energy For cooking, fuelwood is the preferred fuel expenditures also vary by region. Households in the choice for almost all rural households. Households Coastal Central and South regions spend about twice in the Andean South region also use animal dung as much on energy as do households in the Andean widely as a cooking fuel. However, LPG is becoming and Amazon regions. However, the share of energy popular as a cooking fuel among higher-income expenditure to total household expenditure is slightly households. Kerosene is used primarily for lamp lower in the Coastal regions. The disparity in energy lighting, but some higher-income households use it expenditure and relative burden of energy costs are as their cooking fuel. a result of household fuel choices, availability of fuels About 39 percent of all rural households have and energy sources, prices of energy sources, and access to grid electricity, ranging from a high of income levels, all of which vary across regions. 71 percent in the Coastal South to a low of Households with electricity spend only marginally 18 percent in the Amazon. To substitute for the lack more on grid electricity and electricity substitutes of grid electricity, 11 percent of all households use car (16.3 soles per month) than households without batteries to supply electricity, especially in the three electricity spend on electricity substitutes alone Coastal and in the Amazon regions. Due to the cost (15.4 soles per month). In other words, households of batteries and the recharging fee, the majority of car without electricity are paying comparable amounts battery users tend to be financially better-off rural for much-lower-quality services. This implies that, households. These households represent a significant on average, households without electricity could unmet demand for grid electricity among households afford to pay for monthly electricity service if it were that can certainly afford to pay for the service. to become available. Households without electricity Rural household energy expenditure shows could reallocate their current expenditures on lighting significant variation, varying from 9 to 41 soles from fuels to an electric bill. 19 3 Electricity from the Grid As noted in the Introduction, access to electricity This chapter provides detailed characteristics services brings important benefits to rural areas. of electrified rural households. It also provides an Lighting with electricity improves the quality of life, assessment on how rural households utilize and extends the time available for productive or leisure benefit from electricity. activity, and increases the time available for study and learning by students (Barnes 2002). Eledricity . improves health in homes by reducing indoor pollution Access to Grid Electricity associated with lighting from kerosene and lowering . Electrification varies significantly across regions. As the number of bum injuries, especially among chilruen, shown in Figure 3.1, the Andean North and Amazon from fires caused by kerosene lamps and candles. It regions have the lowest rural electrification rates (22 and also improves health in rural communities through 18 percent, respectively). The next lowest rate is the the improved efficacy of refrigerated vaccines and the Coastal North at 35 percent In contrast, the more densely lighting of rural health clinics. Home and community populated and more easily accessible Coastal Central security is enhanced by illumination and the provision and South regions have achieved the highest rural of public lighting. Finally, electricity also empowers electrification rates, at 60 and 71 percent, respectively. the rural poor by increasing access to information There is a direct positive relationship between a and communication technologies (ICT). Electricity household's financial well being and access to grid infrastructure is vital for development, alleviation of electricity. The vast majority of the poor-measured poverty, and improvement in the living conditions of in terms of total household expenditure-do not rural populations. have access to grid electricity, while the vast majority Typically, electricity service from the grid is first of financially better-off households do have access. extended to financially better-off households living Access to electricity is strongly correlated with in more densely populated rural areas that can afford expenditure quintile: only 28 percent of households in to connect to the grid and pay for electricity. However, the poorest quintile have access to electricity, compared as rural electrification expansion progresses and grid with 49 percent in the top quintile (Figure 3.2). electricity is extended further and further, poorer Poverty-measured in terms of low expenditure, low households eventually gain access to grid electricity. income, and low access to basic services-is a way In Peru, grid electricity service has been provided to of life for the majority of rural Peruvians. The lack only 39 percent of rural households, partly because of access to basic infrastructure services, including of the difficult geography and topography of the electricity, not only exacerbates poverty conditions, ·country. but also hampers efforts to alleviate poverty. 21 Special Report Peru: National Survey of Rural Household Energy Use Figure 3.1 Percentage of Households with Access to Grid Electricity by Region 80% 71% 60% 60% UJ "'O 0 52% ..c: \],) UJ ::J 40% 44% - - 0 :r: 39% 0 35% ";/?. 0 20% 18% 0% Coastal North Coastal South Andean Central Amazon National Average Coastal Central Andean North Andean South Source: INEI, 2005. Figure 3.2 Percentage of Households with Access to Electricity by Expenditure Quintile 55% ' ' ' 50% -1- ----~----------r---------~ I I I I ~ I I 0 I I ..c: 45% -1- --------4- --------~- Ol ! I 3 0 I ' I I I :r: I 0 40% -r I I --------,--------- I r- --,---- I - -,- ' \],) 'I ' I f rf 35% ' ___ _ _L I I - J ___ _ I ' ' - -'- '' ' 30% ' I- - - - -~ ---- -- I -!------ ----+--- - - _,_' 1 ' ' 25%L_~~~~~~~~~~~~~~~~~_;___~~~~~--' 1 (Poorest) 2 3 4 5 (Richest) Expenditure Quintile Source: INEI, 2005. Service Reliability than 91.6 percent of households report year-round availability, and a further 3.7 percent report 11 months Two ways of looking at reliability of electricity service of service. Around 12,000 households (1.6 percent) are the number of months of service per year and the experience only one or two months of service per number of hours of service per day. The first relates year. Although the question of generation type was to seasonal availability. As shown in Figure 3.3, more not specifically asked in the Survey, it is reasonable to 22 -~ '. Figure 3.3 electricity usage. 9 Electricity usage among rural households in Peru is relatively low, at an average of Months of Service per Year 27 kWh permonth, compared to other rural households 1000 . . - - - - - - - - - - - - - - - - - - , in countries such as Thailand, the Philippines, and 0 0 "' 800 91.6% Lao PDR. This may be due to several factors, including 0 :::!. a high electricity tariff, unavailability of inexpensive .,, "' 0 ..: 600 electric appliances, and high prevalence of poverty :l\ :J 0 in rural areas. :r .... 0 400 Electricity usage among rural households ... OJ .!J E ioo varies significantly among regions. As expected, · ::l z rural households living in the Coastal Regions use 0.8% 0. % 0.7% o. % significantly more electricity than households living 0 1 2 3 4 5 6 7 8 9 10 11 12 Months of Service in the Andean and Amazon Regions. Within the Source: INEI, 2005. three major regions there are further disparities. For instance, electricity usage in the Coastal Central and South Regions is between 54 and 61 percent greater · Figure 3.4 than in the Coastal North region (Table 3.1). Hours of Service per Day Households with a grid electricity connection 1000 r------------------~ spend on average 14 soles per month on electricity. As with the amount of electricity usage, the amount of g "' 0 800 93.2~ . money spent on electricity varies significantly across .,, "' 0 600 regions and expenditure quintiles, reflecting different ""' :l\ ::I 0 usage levels as well as prices (Table 3.1 and Table 3.2). ;: 40:0 l- o ... ., -" E 200 1- ::1 z 1.0% Electricity Tariff by Usage Level a to n: 14 16 u 20 22 24 Hours of Service per Day Aside from income and regional disparities that have a Source: INEI, 2005. direct impact on variation of electricity consumption, assume that households with service just a few months electricity tariff structure, and ownership of electric per year are served by small hydro systems. In terms of appliances also play important roles in determining hourly service reliability, 93.3 of households reported the level of consumption. As expected, the average 24-hour service, and another 2.7 percent reported effective electricity price per kWh each household has 12-hour service (Figure 3.4). However, only 80 percent to pay depends on the level of usage. Larger electricity of households reported 24-hour service throughout the users that live in the Coastal Central and South year (i.e., 24-hour service in every month) regions pay an average of only 0.49 and 0.47 soles per kWh, respectively. By contrast, smaller users, who tertd to be poorer customers who live in the Andean Overall Electricity Use regions, pay about 0.60 to 0.80 soles per kWh. The and Expenditure variation of average effective electricity price per kWh In Peru, the interaction of regional factors, income, and is due directly to the tariff structure, which includes a price play an important role in determining household fixed charge, maintenance charge, and public lighting 9Of the 3,098 households sampled that reported access to grid electricity, only 977 reported their quantity of electricity used. About half of electrified households in the sample reported only average monthly expenditure, making it possible to calculate usage when average tariff data are known. For the remainder of electrified households in the sample (637 households), neither quantity nor expenditure data are known. Of those 637 households, 274 are served by municipal utilities not regulated by OSINERG, meaning that tariff schedule information is not available. 23 Special Report Peru: National Survey of Rural Household Energy Use Table 3.1 Household Electricity Consumption, Expenditure, and Average Effective Price per KWh by Region Coastal Regions Andean Regions North Central South North Central South Amazon All kWh used/ month 38.3 61.7 59. l 21.7 26.9 16.7 31.6 27.2 Soles/month on electricity 19.8 27.0 24.7 10.9 13.4 9.4 16.0 13.6 Avg. price/kWh (soles) 0.57 0.49 0.47 0.60 0.62 0.83 0.71 0.67 % kWh used for lighting 28.0 24.0 24.2 43.7 41. l 54.6 38.5 42.9 kWh for lighting month 7.0 10.4 9.3 6.4 7.7 5.8 6.9 7.1 Source: INEI, 2005. Table 3.2 Household Electricity Consumption, Expenditure, and Average Effective Price per kWh by Expenditure Quintiles 1 (Poorest) 2 3 4 5 (Richest) All KWh used per month 11 .7 14.64 19.96 28.66 48.51 27.19 7.36 8.54 10.38 14.2 22.52 ' 13.63 0.83 0.76 0.69 0.62 0.55 0.67 Source: INEI, 2005. fee that apply to all customers. Although a large is very large. However, the effect of fixed charges on number of distribution companies charge a public average effective electricity price becomes smaller and lighting fee for smaller users (those who consume smaller as monthly electricity usage becomes larger. less than 30 kWh per month), that is lower than for Given the level of monthly electricity usage among larger users, the ove:i:all fixed charges still play a rural households in Peru, the impact of fixed charges significant role in the retail price of electricity. As a on average effective electricity price is quite high. For result of fixed charges, the average effective electricity example, the average effective price for a household for households that use small amounts of electricity that uses less than 10 kWh per month is about one each month is relatively high, even though the overall sol (1.03 soles) per kWh. However, the impact of fixed fixed charges for consumers using less than 30 kWh charges on average effective price is minimized as per month is lower than those using more than 30 consumption reaches 50 kWh per month. kWh per month. Currently, about 70 percent of households with grid electricity connection use less than 30 kWh per Electricity Usage for Lighting month. These households' average effective electricity The proportion of total electricity used for lighting price is 0.76 soles per kWh. However, the· average is strongly dependent on expenditure quintile effective price per kWh for households that use more (Figure 3.5). The bottom qu,intile uses 39 percent of than 30 kWh per month is only 0.46 soles per kWh. total electricity consumption for lighting, while the The impact of fixed charges among households top quintile uses only 21 percent. The explanation is that consume small amounts of electricity per month simple: As income (expenditure) increases, the ability 24 3 Figure 3.5 Lighting versus Total kWh by Quintile 50...--.~~~~~~~-.-~~~-.-~~~-r--~~--,.---~~--..-, I _ JI __ I ____ l, ___ ..,. - _J __ 40 - ~- I I I I I l I I I ~ -~---- I I. I I • 30 --i----- -1- - - - - -1- ~ I I i I I I 20 -f- - I I I -1- - ---,- I - _,_ - I - • 10 I - -1- l 1 (Poorest) 2 3 4 5 (Richest) allHH e nghting 5.1 6.4 6.5 8 9.7 7.7 •total 13.1 14.8 22.2 28.5 46.5 27.1 Source: INEI, 2005. Table 3.3 Number and Type of Electric Lights Owned by Level of Usage Usage per Month Incandescent Fluorescent Compact Fluorescent All Electric Lamp Lighting ::; 30 kWh/month 2.6 1.9 2.0 3.1 > 30 kWh/month 2.9 2.6 2.6 4.6 All Levels of Usage 2.7 2.2 2.2 3.5 Source: INEI, 2005. to purchase expensive electric appliances increases, report no lamps at all, the individual distributions and thus a greater fraction of electricity is used for reveal that 46 percent of all households with color TY, sound equipment, and refrigerators. electricity have no fluorescent lamps, and therefore Income differences largely explain the regional have only incandescent lamps. However, 23 percent of variations as well. The Coastal South and Central households have no incandescent lamps, and therefore regions have the highest proportion of upper quintile have only the more efficient, but also more expensive, households. Therefore, the fraction of electricity used fluorescent lights. for lighting in those regions is lowest. For example, Households that have exclusively fluorescent households in the Coastal Central region use 16percent lights are disproportionately in the upper quintiles, of their total electricity usage on lighting. In contrast, and, not surprisingly, households with only inefficient households in the less prosperous Andean South incandescent lights are disproportionately in the region use 35 percent. bottom expenditure quintile. On average, households The median rural household in Peru has three that consume less than 30 kWh of electricity per lights. However, this aggregate distribution masks month have a greater share of incandescent bulbs significant differences by lamp types. Although there as a percentage of total lights than households that are a negligible 1,000 or so electrified households that consume more than 30 kWh (Table 3.3). 25 Special Report Peru: National Survey of Rural Household Energy Use Figure 3.6 Appliance Use in Electrified Homes Percentage of Households 0% 20% 40% 60% 80% Radio B&WTV Color TV Electric Iron Sound Equipment Refrigerator F:======: 11 % Video/DVD !:::::::---'-'----' 11 % Electric Fan Microwave Oven Electric Motor Electric Sewing Machine Washing Machine Electric Drill Domestic Water Pump 0% Electric Pump Irrigation 0% Electric Saw 0% Electric Stove 0% Source: INEI, 2005. Note: See Table A.2.43 in Annex 2 for data by region. Household Appliances quintiles. Black-and-white TVs are clearly replaced by color TVs. Appliance ownership variations by Typical electric appliances used in rural households region are in line with regional income disparities (see can be classified in three major categories: (1) radio, Table A.2.44 in Annex 2 for the complete data). television, and other entertainment appliances; Radio and television are two of the most (2) refrigerators, fans, and other appliances that can important home appliances for both urban and rural be used for cooking or domestic work; and (3) electric. households. For a large portion of rural households in appliances directly used for income-generating Peru and the rest of the world, radio and television are activities. The number and type of electric appliances the only means to gain access to news and information in grid-connected rural households provide a good beyond their community. Radio and television are also indication of living standards improvements made a key source of entertainment in rural communities. possible by electricity. The Survey reveals that about 15 percent of rural Figure 3.6 summarizes major appliance use households with electricity have neither plug-in radios in electrified homes. Radios are by far the most nor plug-in television sets at home. Although plug-in common type of appliance, with 66 percent of radio and television are iilexpensive to use, especially electrified households owning one or more. Radios are in comparison to radio and television powered by followed by black-and-white televisions (37 percent of dry cell or automobile batteries, these households households), colortelevisions (33 percent), and electric are unable to take full advantage of grid electricity. irons (25 percent). Of the households with neither plug-in radios nor Ownership of almost all types of electrical TV, more than 60 percent are in the bottom two appliances goes up as household income increases expenditure quintiles. The Survey also reveals that (Figure 3.7). The only exceptions are black and white the vast majority-80 percent-of households without televisions and radios to some degree, which show plug-in radio and/or television lives in the Andean drops in ownership between the fourth and fifth and Amazon regions. 26 3 Electricily from the Grid Figure 3.7 Appliance Ownership in Electrified Homes by Expenditure Quintile 80% I radio , I color TV ' J ____ ----'- 60% - - - - -'- - - - - - - L I - ' I I I I I I I 40% - /- - .... -----r- 1 I I 20% 0% 1 (Poorest) 2 3 4 5 (Richest) Expenditure Quintile Source; INEI, 2005. Figure 3.8 Television Ownership in Electrified Households by Region 93% 92% 0.8 87% ti) -c 0.6 64% 65% 15 60% 62% .c: 58% 533.22 S/month S/month S/month S/month S/month All With access to 0 0 0 0 0.1% 0.0% Without access to grid electricity 0.3% 0.1% 0.8% 0.4% 2.9%. 0.8% Source: INEI, 2005. consumes about 1.2 liters/kWh, so 19 liters generates Solar Home Systems 15.8kWh/month, or 190kWh/year. Assuming a 10-year Solar photovoltaic (PV) systems represent an option life and no maintenance costs, the capital cost (at a for providing electricity to households in remote 12 percent discount rate) is 301 soies/year, or Sl.58/ rural areas, where the costs of grid extension are kWh. The cost of fuel is 3.06 soles/liter, or 3.68 soles/ particularly high. For whatever reason, the use of kWh, for a total of 5.26 soles/kWh. This is 10 times solar systems is quite rare in rural Peru. Most of the the typical cost of grid electricity. households that would use a solar PV system now The estimated cost of using a ge~erator is much use car batteries. lower than the cost of using a car battery, and it Solar systems are estimated to be present in would give far better service levels. It is likely that 0.8 percent of all households, or about 16,700 rural a significant barrier to the adoption of generators households. Of this total,· 13,345 are in households is their high upfront costs, which are 10 to 15 times without electricity service from the grid, while 3,373, higher than purchasing a car battery. Also, in other or 20.2 percent of the total, are in households with countries, generators are used mainly by families that electricity service from the grid. As shown in Table either have a business or can provide electricity to a 4.6, almost all solar systems are in households in the small shop to help pay for the operation of the system. top expenditure quintile. Solar home systems are 36 4 Off-Grid Electricity concentrated in the Andean Central, Andean South, high in the rural areas surrounding Lima, and car and Amazon regions. battery recharging would be relatively easy due to Of the 42 households sampled reporting PV the presence of good roads. The two other Coastal systems, only 22 households reported use in the regions also have very high levels of car battery use previous month. Of the 20 systems not used, 14 of the in households without electricity-47 percent in the households have grid access. Since only four of the North and 37 percent in the Sou~h. In the Amazon, systems were reported to be installed before the year about 18 percent of households use car batteries, 2000, this indicates that the programs installing the while in the Andean regions, usage levels among systems did not target well to ensure that the solar households without electricity range from 11 to home systems were destined for areas that would 16 percent. Costs are high, estimated at 5 to 13 soles not be connected to the grid. Five systems appear to per kWh for operation, plus 10 soles per kWh for be out of service since the households have no grid amortization of the battery. access and report no. use in the previous month. Both households with and without electricity Three of the systems in use were more than 12 years use dry cell batteries, but their use is more prevalent old, and the oldest was installed in 1982. The four in households without grid electricity. Less than systems reported to be installed before 2000 were still 1 percent of all households have either a generator or a in use, including the one installed in 1982. In terms solar home system, and these are mainly concentrated of appliance use, most of the systems are used for in households with higher incomes. lighting and communications, especially radio and · The off-grid use of electricity in rural Peru is both black and white TV, while no uses for color TVs or pervasive and expensive. A significant proportion VCRs were reported. Only one of the systems operated of households that are not receiving electricity from · during the last month also had grid access. the grid consume electricity that is available from other energy sources, such as car batteries, small generators, solar home systems, and dry cell batteries. Conclusions Such electricity is generally of lesser quantity and The main types of off-grid electricity used in rural poorer quality than that available from the grid, and Peru are car and dry cell batteries. In addition, a small has much higher cost per energy unit. The common number of people use generators and solar home use of this rather expensive electricity by households systems. An estimated 18 percent of all households with no connection to the grid is an indication that that do not have grid electricity are using batteries as they place a high value on the services provided by their main source of electricity, and this amounts to electricity. This is a testament to the strong desire for approximately 240,000 households. More than half of electricity in rural Peru due to the benefits it can bring households without electricity in the Coastal Central to rural households. In the next section, the benefits region use car batteries. Incomes are comparatively of grid electricity are quantified. 37 5 Benefits of Rural Electrification The benefits of rural electrification are well recognized. and how much they ar~ paying for these services, so However, there are few empirical studies that provide comparisons become possible. a firm economic quantification of these benefits, particularly in rural areas. In part, this is because of Background on Rural Electrification . the difficulties of quantifying benefits that may take decades to be realized-as in the case of improved Benefit Estimation educational outcomes from better study habits or The goal of determining the benefits of projects is improved income generation opportunities. Long-term hampered by the fact that "benefit" has no natural outcomes are further blurred by migration from rural measure. Psychologists, sociologi.Sts, and economists to urban areas. Difficulties in quantifying benefits are may imagine a measure, such as the "value" or also due to the intrinsic difficulty of quantifying health "utility." Yet, no physical meter or device can measure and safety benefits, such as measuring the benefits of the increased value or utility enjoyed by individuals avoid,ed burn injuries to children. Gathering reliable and households that results from the significant information in remote areas through household change in lifestyle that occurs once grid electricity is surveys also presents challenges. If establishing delivered into a home. reliable estimates of household and energy expenditures The benefits of electricity consumption can is difficult where there is at least the common basis be broken into two categories: direct and indirect. of money, establishing quantitative estimates of the Direct benefits include improvements to lighting diversity of services provided by electricity, such as and television viewing. Indirect benefits include lighting, television viewing, and refrigeration, is more improved educational outcomes for children in difficult still homes with electricity and improved income- In this chapter, we examine the benefits of generation opportunities. Most of the quantitative providing electricity to people in rural areas. This is work described in the literature relates to estimating not a trivial task, and it involves both understanding the direct benefits, and most of this section is devoted demand behavior and making some assumptions to assessing these direct benefits using techniques abouthow households without grid electricity will that have been increasingly used in similar studies in change their behavior once they have access to it. other countries. However, there is evidence that some Implicit in this work is that electricity is valued not of the direct benefits, such as improved lighting give in and of itself, but rather, for the services that are rise to indirect benefits, such as improved education provided. In some cases, these services already are or school attendance. Thus, even though measuring provided though the use of other fuels. Candles and and quantifying indirect benefits may be problematic, kerosene are used for lighting before households some of them may be embedded in some of the direct gain access to electricity. In other cases, there are benefits measured. As an example, lighting may allow new uses that are just not possible without electricity. children to read in the evening and parents may then Fortunately, we have evidence from households that ·see a long-term benefit of sending their children to already have electricity concerning how they use it school because they will perform better. 39 Special Report Peru: National Survey of Rural f:iousehold Energy Use There are two basic approaches for estimating To circumvent the underestimation problem of the direct economic benefits of rural electrification. the "avoided cost" method, another direct benefit The first-which is well-established in the applied calculation method can be used that involves estimating economics literature-is to set benefits equal to a demand curve. However, this is generally difficult the avoided costs of the various devices that are because few actual data points may be available replaced by electrification, including kerosene, to accurately determine the shape of the curve. In diesel generation for auto battery charging, candles, many cases, there are only two points: one of which and dry cells batteries. This avoided cost method is corresponds to the quantity consumed and price easily applied because it needs only expenditure paid by households with electricity, and the other information. It also has the advantage that the for the quantity/price combination of households estimates of benefits are empirical and demonstrable. without electricity. Despite the additional uncertainty, For example, if electricity displaces a certain amount this method of estimating the demand curve and of kerosene and candles, then it is reasonable to the formalization of willingness to pay is generally assume that the monetary benefits of electrification accepted as a more realistic measure of the benefits of will be at least equal to those avoided costs. electrification than the avoided costs method and has The avoided cost method generally underestimates been widely adopted. the actual benefits, for two reasons. The first is that the Economic theory holds that the total benefit of quality of service from electrification is far superior to consuming a given quantity of a good at a given price that from most alternative devices: The illumination is equal to the area under the demand curve. Such a derived from a compact fluorescent lamp is far demand curve is illustrated in Figure 5.1 for the case of superior to that provided by candles or kerosene lamps. lighting: For the purpose of this illustrative example, Moreover, electric lighting eliminates many harmful it is assumed that lighting in households without side effects, such as smoke, odor, and the risk of fire and electricity is provided by kerosene lamps only, and is injury. Individuals are prepared to pay more for high- represented by the point x on the demand curve. The quality service, which is to say that they value a given quantity of service consumed is therefore QKERO' at number of lumens from an electric bulb much more the price P KERO' Thus, the total household expenditure than the equivalent number of lumens from candles on lighting is Q KERO X P KERO' equal to the area B + D. and kerosene. For this reason, the benefits are greater The total willingness to pay (WTP) for the service than those that may be inferred from replacement at level QKERo is the total area under the demand curve costs alone. to that level of consumption (i.e., areas A + B + D). The second reason why benefits can be This is the total benefit to the consumer. However, underestimated is that it is well established that the cost is area B + D, and therefore the net benefit of individuals are prepared to pay very high prices for the consuming QKERO' also called the consumer surplus, is first few kWh of electricity (or lumens). The evidence the difference between the two, namely, the area A. is that people commonly use kerosene and dry cell After electrification, the level of service (in the batteries with a very high cost per kilolumen hour. case of lighting, the number of lumen-hours) typically They are also prepared to pay high prices for enough increases substantially and is represented by the point electricity to power a small television. But the amount y. Consumption therefore increases from QKERo to QE, they are prepared to pay, for example, for the tenth and but the price paid for the electricity service also falls eleventh compact fluorescent lamp (CFL) will be much (typically) from P KERO to Pr Now the household's less than that which they are willing to pay for the first expenditure for electricity is PE X QE, equal to the and second CFL. This demand curve-the representation area D + E. At this level of consumption, the total area of quantity demanded as a function of price-is therefore under the demand curve to QE (i.e., the total benefit), downward sloping, and the total benefits from some level is now the area A + B + C + D + E. Therefore, the of consumption is given by the area under the demand net benefit, or consumer surplus, after subtracting curve (to that level of consumption). the cost D + E, is A + B + C. Thus, it follows that the 40 5 Benefits of Rural Electrification F!gure 5.1 Demand Curve for Lighting (Theoretical) price D E Q, seivice level Source: INEI, 2005. net economic benefit of electrification is the increase in associated with new prices. One must recognize that consumer surplus, which is the area B + C. the demand curve shifts outward with increases in Areas B, D, and E are readily calculated income. For most so-called normal goods at a given from knowledge of consumption before and after price, higher income would mean an increase in electrification, from the household budget for demand. However, in the case of an inferior good, kerosene (and battery charging), and from electricity consumption decreases with increase of income.12 It tariffs. In other words, given knowledge of the two will be seen later in this chapter that radio listening points on the demand curve, x and·y, the areas B, D, decreases when households with electricity start and E are immediately calculable. Area C is more using television. But for lighting, it is possible to difficult to estimate, since it requires knowledge of the estimate the benefits by cakulating the increase in shape of the demand curve between points x and y. · consumer surplus from the relevant demand curve. The most convenient assumption is that the demand Thus, the methodology requires estimates of curve is linear. However, as shown next in the case of the cost and quantity of a service before and after the demand for lighting, for which several points on electrification, described in the following sections the curve are available (each representing different for a number of the important services provided by steps in the lighting ladder), the shape is concave. electricity-lighting, television-viewing, radio, and This approach of estimating changes in welfare refrigeration. by consumer surplus has a number of issues and limitations that are rarely acknowledged, and needs Measuring the Benefits to be applied with some caution. Consumer surplus is an approximation of real benefit increase that lies of Better Lighting somewhere between the area under the demand Idefllly, to measure the benefits of electrification, as measured by old prices and benefit increases one would collect data on lighting utilization 12 These limitations are discussed further in Annex 3. 41 Special Report Peru: National Survey of Rural Household Energy Use (and corresponding expenditure data) for a set of Table 5.1 households prior to grid connection and compare this with electric lighting utilization (and expenditure Lumen Output for Lighting Devices data) of the same set of households after the households Type of Lighting Lumen-Flux (Im) have been connected to the grid. Unfortunately, Incandescent collecting such time-series data is time-consuming 10 watts 50 and expensive, and is simply not practicable for 15 watts 100 large numbers of households. Instead, the standard 25 watts 230 approach is to simplify data collection through a 50 watts 580 cross-sectional survey of a randomly selected sef of 75 watts 1,080 households with and without electricity. 100 watts The underlying behavioral assumption is that, Fluorescent once connected to the grid, the households currently 10 watts 600 without electricity will utilize it in a manner 20 watts analogous to those households currently with 40 watts {straight) 1,613 electricity-all other things equal (such as income). The Peru Survey was designed to measure the 22 watts {circular) 1,480 amount of light purchased through various sources 32 watts {circular) 1,506 of energy. As indicated, households in rural Peru light Compact fluorescent lamp their homes with candles, kerosene, batteries, and · 10 watts 600 electricity. Households use various types of lamps to 12 watts change the energy in fuel to light. Lamps have varying 18 watts efficiencies in converting energy into light, which is 20 watts 1,480 measured in lumens. For instance, kerosene lamps are 25 watts very inefficient in producing light. Electric lamps, by Kerosene contrast, provide as much as 100 times more lighting 11.4 than a kerosene lamp. l hurricane lantern 32.4 By asking households how many hours they used l pressurized kerosene lamp these various types of electric and nonelectric lamps {Petromax) 2,040 per day, we are able to estimate the total lumen-hours Candle of lighting that a households uses from the different 1 candle weight 30-50 gram sources (candles, wick lamps, etc.). In this way it is Candle use 0.5 kg. 1 kilolumen hour possible to compare the level and cost of lighting Source: The Netherlands Energy Research Foundation (ECN), in households that rely on various types of energy Rural Lighting Services: A Comparison of Lamps for Domestic Lighting in Developing Countries, ECN-CX-98-032, July 1998. sources. The vast majority of rural households without access to grid electricity rely on candles and kerosene for lighting. Small numbers of households also use hours of light. In contrast, a 10-watt compact car batteries, LPG, solar PV home systems, and small fluorescent lamp provides 600 lumens, and a simple generators. kerosene wick lamp provides about 1011.:!mens of light. Understanding the assumptions of the lighting Survey questions establish the inventory of estimates employed in this analysis is important. As lights present in a home (number and wattage), and indicated, the quantity of lighting is most commonly typically ask how much each device was used over measured by the lumen-flux, the measure of light the last 24 hours. Analogous calculations are made intensity. As shown in Table 5.1 a 10-watt incandescent for homes without electricity. For example, in the bulb provides 50 lumens, so 1 hour of use requires case of candles, one establishes how many candles 10 watt-hours of electricity and provides 50 lumen- were bought (used) over the last month. Using the 42 5 Benefits of Rural Electrification Table 5.2 Lighting Ownership and Hours of Utilization by Households (Unweighted) Number of Electric Average Number Number of Survey Lamps for Lighting of Hours per Day Used Households Using Lamp HH without Grid Access Candle NA 2.05 195 Kerosene wick kerosene 1.7 3.36 Hurricane wick kerosene 1.5 4.27 Pressurized kerosene 1.17 2.04 109 1.92 4.44 2,519 Car Incandescent 1.4 2.3 156 Fluorescent 1.2 2.2 61 1.6 2.4 25 of 1.4 2.4 227 Grid-electrified HH Incandescent 2.7 4.3 Fluorescent 2.4 5.2 2.5 4.7 1 148 3.8 6.5 Source: INEI, 2005. conversion factors in Figure 5.1, each 40-gram candle in rural Peru. It is necessary to take several things may be said to provide 0.08 klmh (kilolumen hours) into consideration when constructing this demand of light, from which follows the total klmh provided curve. The first is that there is more variation each month from candles. between households with different sources of energy The distribution of lighting appliances and their than within them. For instance, the kerosene price estimated average usage in households with electricity differences for households at different income levels in rural Peru is fairly similar to other countries, with are largely explained by the relationship between the the exception of greater use of more efficient electric size of purchase and cost. The average price per liter of lamps (Table 5.2). There is significant overlap of the kerosene bought in small bottles (between one quarter figures in the table: for instance, many households and one third of a liter) is 3.25 soles per liter. However, with electricity still use kerosene fo; candles as a kerosene bought in 1-liter bottles costs 2.84 soles per backup source oflighting. However, the findings show liter, and is only 2.60 soles per liter when purchased that a significant number of households use candles, in larger gallon containers. kerosene, and grid electricity. A significant finding The procedure for estimating the lighting costs is is that a relatively high percentage of households similar for all classes of lamps. The survey provides without electricity use car batteries for both lighting information on both the use of the lamps and the and television. The results confirm a high demand for total energy use for all households. But households lighting services in all expenditure quintiles. use energy for multiple types qf activities within To construct a demand curve for lighting, we the household. For instance, kerosene is used for examine all of the lighting sources for households both lighting and cooking and electricity is used 43 Special Report Peru: National Survey of Rural Household Energy Use for lighting and to power many other household These are shown in Figure 5.2: the five points shown appliances. As a consequence, the lighting cost is first in the figure for each device represent the points for. estimated based on the hours of use of various types . the five expenditure quintiles. of lamps for the households, and then cross-checked The lighting demand curve based on the individual against the total quantity of energy used by the steps of the lighting ladder match the theoretical households. The prices used for lighting are derived shape extremely well, as shown in Figure 5.2: from the actual prices that households paid for each the overall shape is clearly concave, not linear. type of energy. Nevertheless, the estimates of WTP depend The process begins with users of simple wick- critically on assumptions about the shape of the lamps, the most common form of kerosene-based lighting device (Table 5.3), for which lumen-hours Table 5.3 and price per lumen-hour are estimated.13 Percentage of Households Reporting Use of Lamps The next step on the lighting ladder is hurricane lamps. Hurricane lamps provide more lumen-hours Wick Hurricane Petromax 49.1% 30.2% 1.0% in all quintiles and have a cost advantage that derives from a somewhat higher efficiency. Similar 2 55.4% 30.6% 0.8% calculations can be made for lighting from car 3 59.3% 26.7% 2.2% batteries and petromax lamps, though compared to 4 56.2% 31.3% 3.0% wick and hurricane lamps they are used by only a 57.6% 39.7% 5.3% small number of households. Finally, there are the Lu mens 11.4 32 points that represent grid-connected households. Source: INEI, 2005. Figure 5.2 Demand Curve for Lighting (Actual) 4~~~~~~~~~~~~~~~~~~~~~~~~~ 3 DEMAND CURVE TOP QUINTILE GRID 100 200 300 400 lighting, kLmh/month DEMAND CURVE BOTTOM QUINTILE Source: INEI, 2005. 13 The quantity of lumen-hours provided by candles is even smaller than that provided by wick lamps, and the cost per lumen hours is more than four times than that of kerosene (see Figure 5.6). 44 S Benefits of Rural Eleclrification demand curve between individual points. Small The nature of the problem of the shape of the changes in the degree to which the actual curve is demand curve is illustrated in Figure 5.3. If one concave will result in large changes in the resulting includes the car battery point, curve A might apply. consumer surplus estimates. There are many issues of But if one excludes the car battery point, then curve B data reliability, since the calculation of service demands (which corresponds to a much higher constant price (such as lumen-hours) depend on numerous behavioral elasticity of around -1.1) results in much lower and technical assumptions in addition to the usual estimates of consumer surplus (with the area under problems of survey variance. The estimates for the costs the curve representing less tlian 10 percent of the area of battery charging are particularly uncertain because of the linear triangle, rather than around 33 percent it is difficult to establish the costs associated with in the case of curve A). transportation (to the charging station), and the small If one takes the conservative stand and excludes number of users, particularly in the low expenditure the car battery point, and assumes constant price quintiles, raises the question of whether the car battery elasticity between the points between kerosene users point can validly be included in the curve. and grid-electrified households, then the estimates of Only 1 percent of unelectrified households in consumer surplus range from US$1.54 (for the lowest the lowest expenditure group use car batteries (with quintile, with a price elasticity of-1.3), to $1.23/kWh 15 observations), as compared to 79 percent (and (for the highest quintile, with a price elasticity of-1.1) 1,094 observations) for kerosene. However, in the (Table 5.4). Further details of these calculations are Coastal North regions, between 37 and 56 percent provided in Annex 3. of all households without electricity service use car This methodology is consistent with that used batteries, indicating a high degree of acceptance. Car in similar studies in other countries. However, battery use is greatest in these regions where access recognizing the uncertainties of the shape of the to grid electricity is relatively closer than in the demand curve, an alternative approach was also mountain and Amazon areas and incomes are higher. applied to estimate the benefits associated with Figure 5.3 The Impact of Assumptions 3 ~~~~~~~~~~~~~~~~~~~~~~~~~~ ... ... ... ... ... ... ... ... 2 ... ... ... .c ... ... E ... ... -l ... ci5 ... ... ... ... ... ... A ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... Petromax °' ...... ,grid 50 100 150 kLmh/month Source: INEI, 2005. 45 11ITobfie5.4 l WTP Estimates Unit 1 (Poorest) 2 3 4 5 (Richest) QKERo [wick-lamp} klmh 0.8 1.1 1.1 1.2 l.7 klmh 11 l.9 129.5 141.9 205.6 323.5 Pmo [wick-lamp} S/klmh 3.0 2.9 2.8 2.8 2.7 S/klmh 0.061 0.053 0.048 0.034 0.026 Results -1.3 -1.2 -1.2 -1.2 -1.1 23.9 26.2 26.4 29.0 38.0 Net Benefit s 17.1 19.3 19.6 21.9 29.7 kWh 4.8 5.6 6.5 7.4 9.6 WTP/kWh S/kWh 5.0 4.7 4.1 3.9 4.0 US$/kWh 1.54 1.46 1.26 1.21 l.23 Source: INEI, 2005. the curve that includes the car battery point. The 79 percent of off-grid HH in the poorest category differences in approach are as follows: use kerosene as their principal source of lighting, and 1.1 percent use car batteries as their principal • In place of the conventional expenditure quintile source. disaggregation of households (as used elsewhere in this report), a three-way categorization of HH The demand curves for the three income groups is used (increasing the number of HH in each are depicted in Figure 5.4: as expected, they shift category). outward with increasing expenditure (income): for • The assumption of constant elasticity between example, in the case of grid-~onnected households, kerosene lighting and grid electricity lighting is the highest expenditure group consumes three times relaxed, and each segment of the demand curve, the quantity of lumen-hours of the lowest group). In including that for car batteries, is separately each linear segment, the area under the corresponding calculated. curve is estimated not by the corresponding (linear) • Households relying mainly on candles alone triangle, but by the lesser area corresponding to a was added as a category (and petromax lamp concave curve of demand elasticity of around -0.65.14 user deleted, in view of the small numbers of Table 5.6 shows the corresponding values for price households that use petromax lamps). and quantities. • In place of estimating prices and quantities for Table 5.7 shows the results of the consumer each observation (so a household using more surplus calculat~ons. For each income group the than one lighting method is represented in each increase in consumer surplus is shown for each step lighting category), households were first classified in the lighting ladder. For instance, a household with a according to their principal lighting method. This car battery would only have an increase in consumer categorization is shown in Table 5.5: for example, surplus for the demand segment from the car battery 14 This value is suggested by a statistical analysis for the approximately 900 households with electricity bills: the relationship is: ln (kilolumen-hour consumption per month) = .66 + 0.05*region + 0.32*ln (expenditures per month) -0.63 ln (price per kilolumen-hour) + error. The regional variable is not significant, and the overall R2 is 0.45.. 5 Benelils of Rural Eleclrilication Table 5.5 Number of Households Using Lighting Energy by Income Class, 2005 Low Medium High Total Expenditure per Month Candle 128.69 360.37 899.34 401.22 Number of households 274 245 170 689 and candle 128.03 353.26 826.3 355.27 Number of households 1,094 .762 531 2,387 (% of off-grid HH) 79% 70% 65% Car 154.95 381.25 1 199.97 823.86 Number of households 15 81 122 218 (% of off-grid HH) 1.1% 7.4% 14.8% All Off.grid Households 128.46 356.94 896.78 395.89 Number of households 823 Grid Electric 140.14 372.93 973.81 572.26 Number of households 739 l 1 14 Total Households 131 .75 364.71 946.7 481.09 Number of Households 2, 159 2, 158 2, 159 6,476 Source: INEI, 2005. Figure 5.4 Demand Curve Based on Price and Quantity of Energy Us~ by Income Class 15.--~~~~~-..,---~~~~~-.,.~~~~~~-.,..,~~~~~~-. 11 Candles 11 I ... :I 0 10 -~ . . I ........- .................--........................................................................................................................................................................................ ! i ! .c ; I ; E ~ E i! i i ~ i ! j i ; i Iii (.I I I I --~- -------+- -------+------- - - ·;:: D.. 5 . - - - - --- \ ~ Kerosene I I l l\.. Car batteryI I I 0 t.:::~----=::::::::==~~~·-~::;;:;;~o:~i;:::;;::;;:;;;;;;;;;;;;;;0iiiiiji~G~-nd;_________J 0 100 200 300 400 Lighting service, kLmHour/month Source: INEI, 2005. to grid electricity. Only a household switching from on lighting decreases (e.g., by four soles/month for a candle to grid electricity would gain the consumer a low income household moving from kerosene to surplus of all segments. Households also experience grid electricity), even though the quantity of service a real income gain because the total expenditure (Kilolumen-hours) increases. 47 Special Report Peru: National Survey of Rural Household Energy Use Table 5.6 Estimates of Lighting Service and Price Low Expenditure Medium Expenditure High Expenditure Q klmhrs Price Q Klmhrs Price S/klmh QKLmh Price S/klmh Candles 0.48 12.28 0.68 13.0l 0.83 13.45 Kerosene 3.16 3.69 5.76 3.41 10.28 3.03 11.17 l.35 14.37 1.2 22.03 1.23 107.66 0.048 183.64 0.031 313.48 0.021 Source: INEI, 2005. Table 5.7 Increases of Consumer Surplus by Income Class (Soles/month) Increase in Ending Expenditure Difference in Expenditure Class Consumer Surplus on Lighting Expenditures Low Expenditure (<236.115/.) Candle to kerosene and candle 8 7 2 Kerosen and candle to car 10 13 6 Car battery to electric grid electricity 34 3 -10 Total: Candle to Grid cs 52 3 -4 Medium Expenditure (236.11-511.33 S/ .) Candle to kerosene and candle 14 12 4 Kerosene and candle to car 15 17 5 · Car battery to electric grid electricity 51 4 -13 Total: Candle to Grid cs 79 4 -4 Candle to kerosene and candle 23 18 8 Kerosene and candle to car battery 20 22 4 Car 88 5 -17 Total: Candle to Grid 132 5 -5 Source: INEI, 2005. When the results are weighted by the proportion of Despite the uncertainties and the substantial households in each starting category, the calculations range of the estimates of benefit, the conclusion is for average WTP per kWh and total monthly benefits that households switching from candles, kerosene, or are as shown in Table 5.8. These estimates of benefit car batteries to grid electricity for lighting enjoy high are significantly higher than those derived in Table 5.4 economic benefits. The estimates have high variance, (e.g., US$2.25/kWh for the lowest expenditure group, but even at the low end of the range, the economic as opposed to US$1.54).15 · benefits are substantial. Not only do electrified 15One of the issues in these calculations is that the expenditure, defined by Average kLmhrs consumed x Average price/kLmhr, does not equal the estimate of expenditure derived from the expenditure data. When the kLmhr and price/kLmhr estimates are scaled to expenditure estimates, the resulting values of consumer surplus would be about 20 percent lower. 48 S Benefits of Rural Electrification Table 5.8 WTP per kWh (Alternate Method) Low Medium Net benefit Soles/month 46.5 67.5 110.2 74.7 Soles/month 3.0 4.0 5.0 Soles/month 49.5 71.5 115.2 kWh/month 6.82 10.00 10.42 S/kWh 0.44 0.4 0.48 S/kWh 7.26 7.15 11.05 2.25 2.21 3.42 Annual Net Benefit S/year 558 810 1,322 897 $/year 173 251 409 278 Source: INEI, 2005. households enjoy much greater levels of lighting television vi~wing in a similar manner to household service, they also obtain a real income gain since lighting. The survey finds that about 158,000 rural their total expenditure on lighting service decreases. households without electricity in Peru-or about 12 In the interest of using conservative assumptions, percent of all rural households without electricity-are the economic analysis of the new rural electrification still using black-and-white (B&W) television. Table 5.9 approach has used the lower values of WTP, but the shows the viewing hours and costs of the three main results suggest that actual returns may be greater. television types: B&W powered by car batteries, plug- in B&W, and plug-in color. With these input assumptions, the total monthly Benefits of Communications benefit of television viewing is estimated at 24.2 soles/ Radio and television are among the most important month. Alternatively, the benefit associated with sources of communication and entertainment for 212 viewing-hours per month of plug-in color television averages to 0.11 soles/viewing-hour. The rural households. Typically, after electric lighting, plug-in radio and television are the most common net benefit, namely, the increase in consumer surplus, appliances in households with electricity. Without is 14.7 soles/month. electricity, the cost to operate radio and/ or television These estimates of net benefit may be overstated, is extremely high, and the total hours listening to radio because they include only the lower cost of electricity and viewing television tend to be relatively low or itself. The cost of acquiring a color television is limited. significantly greater than that of the black-and-white television that it replaces. Assuming a price difference of US$100 (cost of a new color television less the cost of Television selling the existing B&W television on the secondhand Television viewing is one of the most desired aspects market), and assuming a five-year life, the transition of electrification. Almost 20 percent of households to color television translates into an additional cost of without electricity in rural Peru have a television set 0.0226 soles/viewing hour,16 or 5.17 soles per month. operated by car batteries, which requires significant However, it should be kept in mind that this is the money and time (because battery recharging is often value of moving from using car batteries for watching at considerable distance to the home). Television television to using grid electricity. The benefits of viewing (and particularly color TV) is a normal good, television viewing for households without car batteries so it is possible to estimate the consumer surplus of (that do not have television) would be higher. 49 Special Report Peru: Nalional Survey of Rural Household Energy Use Table S.9 Cost and Viewing Hours for Television Unit Car Battery, B&W TV Grid, Plug-In B&W Grid, Plug-In Color TV Viewing hours Hours/Day 2.81 2.59 6.83 Hours/Month 87 80 212 Power rating of TV . Watts 24 48 75 KWh kWh/Month 2.09 3.85 15.88 Cost 13.58 2.312 9.528 Cost hour 0.16 0.0288 0.0450 Source: INEI, 2005. Table 5.10 Cost per Radio Listening Hour Based on Energy Source Plug-in, Grid Unit Cell Car Connected hours 4.64 3.68 2.87 141 112 87 Power of device Watts 3 9 18 kWh 0.42 1.01 1.57 164 6.5 0.6 Cost per month 69.4 6.6 0.9 Cost per listening hour 0.49 0.06 0.01 Source: INEI, 2005. Note: Price per kWh for car battery and dry cell use is derived in Chapter 3. Radie cell battery radios are used even after electrification. Radios are the most widely used non-lighting However, it is reasonable to assume that almost appliance among rural households. Table 5.10 and all of the households using car battery-powered Table 5.11 show the time spent listening to the radio, radios-126,200 households, or 9 percent of the total as well as costs. Even though dry cell batteries cost rural households without electricity, according to the more than 25 times more than car batteries and Survey-would switch to grid plug-in models once 270 times more than grid electricity per kilowatt-hour, grid electricity became available. they are still the energy source of choice for radio The· analysis used for lighting that estim.ated listening in rural households. changes in consumer surplus from downward- The apparent anomaly of the most expensive sloping demand curves cannot be used given the shift form of radio listening being used the most is simply from radio to television viewing, since this would a reflection of the mobility of dry cell-powered require a multivariate function to properly model radios (Table 5.11). Indeed, as noted previously, dry both goods. Radio listening is arguably an inferior 16 310 soles divided by (229 viewing hours per month X 12 months/year x 5 years 13,740) = 0.0226. 50 5 Benefits of Rurol Electrification Table 5.11 Price and Quantity of Radio listening Price and Quantity Value Unit Radio by Sources of Electricity prrtc 0.49 Cost per listening hour (soles) Radio using dry cell batteries Q,dc 4.64 listening hours per day pcrh 0.06 Cost per listening hour (soles) Radio using car battery Q,_b 3.68 Listening hours per day r;ne 0.01 Cost per listening hour (soles) Radio using grid electricity Qrne 2.87 listening hours per day {for plug-in radio) Source: INEI, 2005. Note: Assumption of wattage for radio is as follows: dry cell batteries is 3 watts; car battery is 9 watts, and plug-in radio is 18 watts. Assumption of dry cell battery capacity: size C and D size supplies 6 Wh of electricity; AA and AM supplies 4 Wh. good, with evidence that consumption decreases with Table 5.12 increases in income, as households generally prefer Use of Refrigerators in Unelectrified Households to increase television viewing. Hence, the benefits of the transition from car battery to grid radios are best LPG Kerosene estimated as the financial savings per listening hour Number of actually 8 11 of switching to grid electricity. For the 2.87 listening households hours/day for grid-connected households, the benefit Number of households 1,422. 1, 100 per listening hour is simply the difference in cost {weighted) between a car battery (0.059 soles/listening hour) As % of unelectrified 0.1% 0.07% and grid radio (0.011 soles/li.stening hour), namely households 0.048 soles/listening hour. Source: INEI, 2005. Refrigeration expenditure) is very small. Therefore, expenditures The Survey reports a very low proportion of of unelectrified households on refrigerators may be unelectrified households using refrigerators ignored in the overall benefit estimation of potential (Table 5.12). Households using small generators or car rural electrification projects. Lack of refrigeration in batteries were not asked whether they were used to unelectrified households also makes it impossible to power refrigerators, and it is unlikely that they would estimate a demand curve and undertake consumer do so. With fewer than 0.1 percent of unelectrified surplus/willingness-to-pay calculations. households using refrigerators, even if there NRECA reports an estimate of WTP for were a sufficient number of households for which refrigeration of US$0.86/kWh (3 soles/kWh) in the expenditure data could be used with confidence, Amazon and Coastal regions (and zero in the Andean the impact on the overall average willingness- regions). However, neither the data nor the details to-pay calculation (i.e., once the expenditure is of the calculations that underlie these estimates are weighted by the fraction of households incurring this available (Box 5.1). 51 Special Repart Peru: National Survey of Rural Household Energy Use Box 5.1 Comparison of Survey Results with an NRECA Study In 1999, the Notional Rural Electric Cooperative Association (NRECA) estimated willingness-to-pay (WTP) values based on a limited survey. It used a methodology for estimating demond curves similar to that used iii this chopter. The NRECA estimates of willingness-to-pay are significanty higher than those suggested by this Survey. The reasons appear to be in the sample of households surveyed: NRECA surveyed areas closer to better-off urban areas, resulting in higher consumption estimates. The average monthly consumption of households surveyed by NR.ECA is 58.SkWh/ household/month, over twice the average consumption in the rural areas recorded by this Survey {TClble A). Table A Average Monthly Consumption, kWh/HQuseholds/Month NRECA Current Lighting Radio& TV Refrigeration Other Total Survey Andean 7.3 5.4 23.4 36.1 15-26 Amazon 9.2 5.4 13.5 20.2 48.3 31 Coastal 8.8 5.4 22.5 58.5 95.2 39-59 All 8.4 5.4 10.7 34.3 58.8 26 NRECA reported significant consumption for refrigeration, whereas in this Survey, the proportion of households reporting refrigeration is very small, making demand curve estimates impossible. The NRECA estimates for WTP in soles/kWh are shown in Table B. However, there is little supporting evidence for these values in the NRECA report. The NRECA estimate of willingness-to-pay for radio i;md television is 3.5 soles/ · kWh and for lighting is 4.6 soles/kWh. The comparison of total monthly WTP against the results ofth.e Survey show the NRECA results to be comparable for TV, and within the overall range for lighting. Table B NRECA Aven;ige WTP Estimates, Soles/kWh Lighting Radio& TV Refrigeration Other Andean 4.8 3.3 Q.O 0.5 Amazon 4.9 3.1 3.0 0.5 Coastal 3.2 4.8 3.0 0.5 All 4.6 3.5 3.0 0.5 Total monthlyWTP, soles/month NRECA 38.6 18.9 Survey, low estimate 24-38 19.0 (color TV)* estimate 40-90 *Ad justed for cos.I of color N. Source: NRECA, EstralegJa Integral de Electrificaci6n Rural. Limo. September 1999. 52 5 Benefits of Rural Electrification Benefits of Education and Health Table 5.13 This section describes the indirect benefits of Average Number of Hours per Night Household electrification, including education, health, and Members Read/Study (Weighted) environmental benefits. Given the difficulties noted No Grid Grid previously, a formal quantification of these benefits Access Access is not attempted. Indeed one of the major problems Children Aged 6 to 18 0.86 1.09 in such a quantification is double-counting: for Attending School Read/Study (Hours/Night) example, it is likely that households internalize the of households 750,283 496, 154 benefit associated with the reduction in kerosene lamp-related burn injuries to children in their WTP All Household Members 0 ..33 0.47 Read (Hours/Night) for electric lighting. Number of households 1,339,829 845,03 Source: INEI, 2005. Education Empirical data on the indirect economic benefits Table 5.14 of electrification for education are not as well documented as the direct economic benefits of Percentage of Children in the Household Attending School (Weighted) ·· education. It is clear that electricity extends evening lighting hours, making it easier for children to study, Without Grid do homework, and read. The survey found that Access Grid Access children aged 6 to 18 in households with electricity Children 6 to 93.6% 92% 12 Years Old who are currently attending school spend an average of 65 minutes per night reading and/or studying, Population 415, 112 695,010 whereas in households without electricity, the Children 13 to 62% 82% 18 Years Old figure is only 51 minutes (Table 5.13). The increase Population 517,277 326,547 of 27 percent in reading/study time is statistically significant. However, caution should be used when Source: INEI, 2005. interpreting this result since the correlation could be due to a third variable, such as household income, with modern teaching equipment and information and not necessarily demonstrate a causal relationship. and communication technologies, especially access The survey shows that there is no difference to the Internet. in levels of school enrollment of children aged 6 Unlike at the primary school level, school to 12: almost all children aged 6 to 12 are reported enrollment at the secondary level is significantly higher to be attending school regardless of their home for households with a grid electricity connection. The electrification status. This appears to confirm that survey reveals that school enrollment of children the educational campaign in Peru during the past aged 13 to 18 from households with a grid connection decade is working. Undoubtedly, electrification will is about 82 percent, which is 20 percent higher than reinforce this success and make it sustainable, since in households without electricity (Table 5.14). Thus, it is ·reasonable to assume that electrification gives there is strong evidence that having electricity in rural children more flexibility in choosing when to do households involves educational benefits. There is a schoolwork. Empirical evidence elsewhere has also strong likelihood that these educational benefits are shown that children who are doing well at school or already quantified as part of the consumer surplus can keep up with their peers are more likely to stay in for household lighting. As a consequence, we do not school longer than those who do not do well at school. make an attempt to quantify those benefits for this Similarly, electricity enables schools to be equipped chapter. 53 Peru: National of Rural Household Use Health airnd Environmenfh::d Benefits 17 However, itis virtually impossible to separate home Seventy-three percent of rural households (98~000 and business use of lighting in homes that do report a households) use kerosene for lamp lighting. Based home business. Moreover, as shown in Table 5.16, among on the Survey data, it is estimated that about 3 liters home businesses, only 4.2 percent are in households of kerosene per household per month are used predominantly lit by car batteries, as opposed to specifically for lamp lighting, or 2.96 million liters of 28.4 percent without electricity. Whatever are the kerosene per month for the entire rural population. causalities, it is certainly clear that home businesses The negative health effects associated with this level are concentrated in households connected to the grid. of kerosene use are significantly greater than the health effects associated with grid-generation. Even with increases in the share of gas-fired generation Table 5.15 from the large electricity generation companies or Home Business Incidence by Major Lighting Type small diesel sets for isolated systems, the health damages caused by the emissions from such facilities Major Lighting Total Sampled are two orders of magnitude smaller than those from Home Business HH burning kerosene in wick lamps inside the home. Number Percent Number Grid electrification will directly contribute Candle 59 8.6% 689 to a reduction of respiratory illness among the Kerosene and candle 179 7.5% 2387 rural population, reducing both public and private Total unelectrified 238 7.7% 3076 healthcare costs. Although there is no specific documentation for Peru, in other countries, the Car 35 16.1% 218 avoidance of children's burn injuries from kerosene Grid 566 18.3% 3098 lamps-particularly from generally unsafe simple Total 839 13.1% 6392 wick lamps-is a major benefit of rural electrification.18 Source: INEI, 2005. Benefits to Home Business Slightly more than 13 percent of all sampled households Table 5.16 reported a home business (Table 5.15). However, the proportion is much greater in grid-electrified Distribution of Households with Home Business by Major Lighting Type households (18.3 percent) than in unelectrified households (7.7 percent) .. For car-battery electrified Number Percent households, the proportion (16.1 percent) is close to Candle 59 7.0% that of grid-electrified households-which suggests Kerosene and that what is important is electrification, rather than candle 179 21.3% whether electricity is provided by the grid or by car Total unelectrified 238 28.4% batteries. This is surprising, given the difference in cost: Car 35 4.2% as noted in the previous section, electrification by car Grid 566 67.5% battery costs an average 24 soles/kWh, as against an Total 839 100.0% average of 0.6 soles/kWh for grid electricity. Source: INEI, 2005. 17The analysis of kerosene fuel savings is based on weighted survey results. · For example, in Sri Lanka, in D~cember 2000 a new bums unit was opened at the Lady Ridgeway Children's Hospital north of 18 C~lombo. By 1?-~gust 200.1, 176 children had been tr.e~ted fo~ burn wounds. The majority of the victims were from the rural villages without electricity of Ch1law, Puttalam, and Karap1hya, which are dependent on kerosene for lighting (Sri Lanka Sunday Observer, September 23, 2001). 54 5 Benefits of Rural Electrification Willingness to Pay for Electricity the corresponding average expenditures for rural enterprises in electrified and unelectrified in Enterprises households. Total energy expenditures remain Willingness-to-pay (WTP) for electricity in largely unchanged: 154 soles per month for nondomestic applications may be estimated from electrified enterprises versus 155 soles per month the results of the business survey, which sampled for unelectrified enterpris~s. 192 rural enterprises. Of the 135 rural businesses However, the overall pattern of expenditure sampled 93 had access to grid electricity (69 percent), does change. In an electrified enterprise, the average considerably greater than the 40 percent of households additional expenditure for electricity (50 soles per sampled. Even a simple consideration of the incidence month) is offset by sharp decreases in expenditures · of energy sources (Table 5.17) suggests that WTP for wood (Figure 5.6). In short, LPG and electricity for electricity is much greater than in domestic replace wood and kerosene. Electrification also brings households. Specifically, 26 percent of unelectrified increased expenditures for self-generation. Businesses businesses use car batteries (versus 11 percent of used to the availability of electricity are willing to households), and 24 percent use small generators pay the high costs of small generators to secure their (compared to just 2 percent of households). businesses against a lack of electricity service. The differences in fuel sources betwe.en These energy expenditure data do not take electrified and unelectrified enterprises are similar into account the dramatic difference in enterprise to that encountered in households: significant incomes. The average monthly turnover (gross sales) decreases in electricity substitutes (e.g., 45 percent in electrified enterprises is 3,520 soles/month, to 10 percent for kerosene use), and significant as opposed to 1,140 soles/month in unelectrified increases in LPG use (from 2 percent to 20 percent), enterprises. Therefore, even though average energy as shown in Figure 5.5. expenditures are roughly the same (155 soles/month), The average electrified enterprise consumes the energy expenditure per unit of sales is much lower 94 kWh per month of electricity. Table 5.18 shows in electrified enterprises. Table 5.17 Energy Sources in Rural Enterprises Unelectrified Enterprises Electrified Enterprises Number of Enterprises Percentage of Total Number of Enterprises Percentage of Total Kerosene 19 45.2 9 9.7 Candles 24 57.1 36 38.7 21 50.0 26 28.0 11 26.2 3 3.2 LPG 1 2.4 19 20.4 Solar 2 4.8 2 2.2 Electric generator 10 23.8 16 17.2 Fuelwood 9 21.4 13 14.0 2 4.8 2 2.2 residuals 1 2.4 1 1.1 Charcoal 0 0.0 0 0.0 Coal 0 0.0 1.1 Total 42 93 Source: INEI, 2005. 55 Special Report Peru: National Survey of Rural Household Energy Use Figure 5.5 Energy Source Differences Between Electrified versus Unelectrified Enterprises UJ (]) UJ ·;:: e- ~ 1J "' "" ·c 0 .S! "' =22 kWh/Month Coastal North 36 31 70 Coastal Central 54 70 85 Coastal South 55 51 93 Andean North 23 16 44 Andean Central 25 26 41 Andean South 15 14 23 Amazon 24 15 47 All 35 223 60 Source: INEI, 2005. Figure 6.1 Breakdown of Villages with Average Consumption above and below 22 kWh/Month by Region 1.2 ~------------------------~ 0.8 70% 0.6 0.4 0.2 0 Coastal North Coastal South Andean Central Amazon National average Coastal Central Andean North Andean South Source: INEI, 2005. the breakdown of villages with consumption above . household electricity consumption. It is generally and below 22 kWh per month by region. supposed that consumption increases with household income (or with household expenditure, the proxy used in this study), and, therefore, all other things Growth of Electricity Consumption equal, consumption per connection should increase One of the important assumptions in making over time. financial projections of the viability of rural However, it is widely reported that growth rates electrification projects is the rate of growth in in consumption per connection in rural areas of 61 Special Report Peru: National Survey of Rural Household Energy Use Figure 6.2 Distribution of Years of Electricity Service 16% Years of Electricity Service Source: INEI, 2005. Peru are very small-often less than 1 percent per schemes require careful scrutiny. At least based on connection per year. This observation is consistent the experience of those communities prioritized with much slower economic growth in rural areas by the current scheme (often the poorest, and most than in urban areas, where electricity consumption lacking in infrastructure access), there ii;; no evidence growth per domestic connection grows much faster. from the Survey that annual consumption growth Inferring growth rates from a cross-sectional per connected household would be much higher than survey requires caution. Nevertheless, one could the commonly assumed 0.5 to 1.0 percent per year. reasonably hypothesize that all other things equal, monthly consumption would be higher the greater the age of the electrical connection. Figure 6.2 shows the Pricing Policy age distribution of electrical connections: 67 percent of The electricity tariff faced by low-income rural rural connections are less than 10 years old, testimony households is a complex nonlinear function of to the rural electrification efforts over the past decade. monthly consumption, a consequence of the approach The peak rate of rural electrification appears to have to rate-making adopted by the regulator, Organismo been achieved four years ago. Supervisor de la Inversion en Energia (OSINERG), and There is little correlation between the age of the cross-subsidy system adopted to finance the a connection and the kWh/household consumed lifeline tariff rate, the Fonda de Compensacion Social (Figure 6.3); the trend line shown is not statistically Electrica (POSE). significant.21 Nevertheless, as noted in Chapter 3, the broad Projections of consumption growth in rural areas pattern is clear: Those who consume small amounts that are presented in proposed rural electrification of electricity pay relatively high prices per kWh, The ordinary least squares relationship is: [kWh/household/month]= 28 + 1.105 [age of connection]; R' 0.04 which suggests every year 21 of connection increases monthly consumption by 1.1 kWh. However, the relationship is not statistically significant. 62 6 Policy Implications of Survey Results Figure 6.3 Monthly kWh of Consumption versus Age of Connection Age of Connection, Years Source: INEI, 2005. Table 6.4 FOSE Subsidy Rates Discount for Households Discount for Households Consumption of 30 kWh/ Consuming Between 30 and Sector Month or Less 100 kWh/ Month Interconnected System Urban 25% of energy charge 7.5 kWh/month Urban-rural & rural 50% of energy charge 15 kWh/month Isolated Systems Urban 50% of energy charge 15 kWh/month Urban-rural & rural 62.5% of energy charge 18.75 kWh/month Source: OSINERG Ayuda Memoria FOSE. notwithstanding the POSE mechanism. Households FOS~ in the lowest expenditure quintile use, on average, The rationale for the POSE is regional equalization of 12 kWh/month, at an average price of 0.83 soles/ · tariffs for those at the lower levels of consumption, kWh, while households in the top quintile use with the general objective of reducing the differential on average 49 kWh/month at an average price of between the high tariffs of the outlying provinces and 0.55 soles/kWh (Table 3.1). Without FOSE, the .· the lowest tariff in Lima. The FOSE subsidy rates are average price for the lowest quintile would increase shown in Table 6.4 most households sampled in the to 1.3 soles/kWh. 22 Survey lie in rural and urban-rural zones. 22 If one subtracts typical fix~d charges from the average bill (10.34 soles -1.88 Soles (basic fixed charge)-1.0.soles (public lighting)- 0.64 Soles (connection maintenance)= 6.8 soles/month for the energy charge; divided by 13.25 kWh= 0.514 soles/kWh. Since this reflects at least a 50 percent FOSE discount, without FOSE the monthly bill would be 6.8 soles higher; therefore without FOSE the average price per kWh (10.34 + 6.8)/13.25 = 1.3 soles/kWh. 63 Special Report Peru: National Survey of Rural Household Energy Use Table 6.5 Number of Connections Benefiting from FOSE, 2004 Monthly Interconnected Consumption FOSE Benefit System Isolated Systems Total Participation (%) 0-30 kWh Yes 1 08 1 11 60.1 30-100 kWh Yes 1,071 ,963 86,026 1, 157,989 > 100 kWh No 1,486,349 59,825 1,546,174 39.9 Total 3,584,420 295,227 3,879,647 100 Source: OSINERG Ayudc Memoric FOSE. Figure 6.4 Components of the Tariff 1 ·~~~~~~~~~~~~~~~~~--,-~~~~~~~--, 0.8 -- - --- -- -- - -,- I ,---------------- I ' I I I 0.6 --,------ I .r:. ~ 0.4 ENERGY CHARGES 0.2 FIXED CHARGES 0 0 50 100 150 Monthly Consumption, S/kWh/month Source: INEI, 2005. Sixty percent of all customers benefit from lighting (typically between 0.8 and 1 soles per month FOSE (Table 6.5). The total FOSE transfer in 2004 was for consumption below 30kWh/month, and 2.8 to US$18 million (compared to a total consumer bill of 3.0 soles per month for households with consumption US$600 million). The recovery mechanism increases greater than 30 kWh/month), and a connection bills to all consumers with consumpti9n greater than maintenance charge (typically between 0.58 and 100 kWh/month by 2.5 to 3 percent.23 0.64 soles per month). These tariff components are illustrated in Figure 6.4. Tariff Structure For a given nominal energy tariff, the resulting The tariff consists of an energy charge, as adjusted by tariff curves have the shape shown in Figure 6.5. POSE, and a fixed charge (cargo fijo), a charge for public However, in practice, urban and rural areas will in 23The FOSE accounts are rebalanced every quarter. The sum of the FOSE subsidy paid is recovered from consumers (in the regulated market) with consumption greater than lOOkWh/month as an ad valorem surcharge. At the time of the survey in June 2005, the "FOSE Factor de Recargo" was 1.026. 64 6 of Resulls Figure 6.5 Typical Tariff Curves 1.2 . - - - - - - - - - - - - - - - - - - - - - - - - - - - - , I I I ---~----- ------ ---~-- -- -- --- -- I 0.8 - - '1 - - - ~1- - - - I I I 0.6- I r 0.4 ---- ---,--I - - -r- - - I RURAL 0.2 ~-------~--------~----------' 0 50 100 150 Monthly Consumption, S/kWh/month Source: INEI, 2005. Table 6.6 Fixed Charges Public Lighting, .Public Lighting, Connection Fixed Charge <30kwh/Month >30 kWh/Month Charges Electrosur 1.87-2.00 0.79 2.37 0.64 ElectroCentro '1.87 1.01 3.04 0.64 ENOSA 1.87 1.05 3.15 0.64 Hidrondino 1.88 0.87 2.58 0.67 ElectroOriente 1.87 1.24 3.81 0.64 Source: INEI, 2005. general have different nominal energy tariffs, so the across the country. This is in part due to the complexity two curves do not in fact converge as suggested here. 24 of tariff setting, in which every major area has its own Table 6.6 shows the fixed charges for five tariff based on the assignment of sector tipicos for distribution companies for which a sample of actual distribution costs and for various power generation consumer bills have been obtained. costs. There are five sectors tipicos based on the degree Although there is little variation in fixed charges, of urbanization. For example, sector 1 applies to Lima the nominal energy charge shows large variation (high-density urban) while sector 5 is for rural areas. 245ee Figure 6.6 for Electrosur, for example. 65 Special Report Peru: National Survey of Rural Household Energy Use Thus, the OSINERG Annual Report lists 166 tariffs ¢ FOSE discount applicable to the first 30 kWh for the 20 regulated distribution companies, based on of monthly consumption 133 distinct tariff schedules. a Tariff variable charge, soles/kWh (before To cross-check the results from the Survey, a FOSE discount) sample of actual consumer bills were obtained from Q1 Monthly consumption if less than or equal to five distribution companies. Table 6.7 shows the data 30 kWh, zero if greater than 30 kWh from 12 such bills from Electrosur for April 2005 Q2 Monthly consumption if greater than (a few months before the Survey was conducted in 30 kWh/month, zero otherwise June 2005). from which follows the average cost per kWh, C, as Several points may be noted: C = T QI+ Q2 • The 12 bills in the sample fall into one of two The result for the two tariff regimes reflected in tariff categories: the first B households have the billing data of Table 6.7 is shown in Figure 6.6. ·a fixed charge (cargo fijo) of 1.87 soles/month The actual sample points fall almost exactly on the and a variable charge of 0.3504 soles/kWh, and predicted tariff curves. Such variation as remains is represent urban households; the second four have due to the small miscellaneous adjustments shown a fixed charge of 2.00 soles/month, and a variable in Table 6.7. charge of 0.449 soles/kWh, representing rural This example, and the way that the tariff is households. structured, raise several issues for pricing policy: • The FOSE discount in the first eight bills is a 25 percent discount on the first 30 kWh, and • The jump (in both rural and urban tariffs) at the 30 a 7.5 kWh discount for consumption in the kWh/households/month consumption threshold 31.100 kWh/month range. The discount in the is due not to FOSE, but to the way in which public rural bills (columns 8-12) is a 50 percent discount lighting is billed. Twenty-three percent of all on the first 30 kWh, and a 15 kWh discount on the households have monthly consumption between subsequent tranche from 31 to lOOkWh/month. 20 and 40 kWh. These households are potentially • Cost recovery for public lighting is a step affected by the tariff increment at 30 kWh/month. function: Bills with more than 30 kWh of monthly • In the range of 31 to 100 kWh/month, the consumption are charged2.37 soles/month; those incentives in ru:ral and urban areas are quite with less than 30 kWh/month are charged 0.79 different. In urban areas, the average costs per soles/month. kWh decline with increasing consumption, while in rural areas they increase. The corresponding tariff curves are readily • The size of the step increase at 100 kWh/ calculated. The monthly bill (including 19 percent month-due to the FOSE-is far greater in rural VAT) Tis areas than in urban areas. Less than 3 percent T = 1.19[F +A+ M + (l -100 is whether the tariff jump at 30 kWh/month-a 66 --- ---------- Table 6.7 Sample Electricity Bills from Electrosur, April 2005 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] consumption [kWh] 94 18 50 14 59 13 93 16 18 78 12 37 first 30 kWh. [S/kWh] 0.2628 0.2628 0.2628 0.2627 0.2249 0.2249 first 30 kWh [SJ 7.888 7.88 7.8793 7.8787 6.75 6.75 31-100 kWh [S/kWh] 0.3504 0.3504 0.3504 0.3503 0.4498 0.4497 FOSE discount [S] 2.63 1.58 2.63 1.21 2.63 1.14 2.63 1.4 4.05 6.74 2.7 6.74 cargo fijo [SJ 1.87 1.87 1.87 1.87 1.87 1.87 1.87 1.87 2 2 2 2 energy [SJ 30.31 4.73 14.89 3.68 18.04 3.42 29.95 4.2 4.05 28.34 2.7 9.9 public lighting [S] 2.37 0.79 2.37 0.79 2.37 0.79 2.37. 0.79 0.79 2.37 0.79 2.37 Mont. y Reposic [S] 0.64 0.58 0.58 0.64 0.58 0.64 0.64 0.58 0.64 0.64 0.64 0.64 de C interes [S] 0.03 0.05 0.06 0.04 0.05 0.12 0.11 compensatorio AjALPUB [S] 0.06 0.13 0.04 0.06 total [S] 35.25 7.97 19.87 7.07 22.86 6.78 34.83 7.48 7.53 33.53 6.24 14.91 VAT [%) 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% 19.0% [S] 6.70 1.51 3.78 1.34 4.34 1.29 6.62 1.42 1.43 6.37 1.19 2.83 AjALPUB [S] 0.02 0.01 comp.Int I/I [S] -0.33 lnt.Moratorio [S] 0.01 -0.20 total facturado [S] 41.95 9.48 23.65 8.41 27.20 8.07 41.12 8.90 8.99 39.90 7.24 17.74 average S/kWh [S/kWh] 0.45 0.53 0.47 0.60 0.46 0.62 0.44 0.56 0.50 0.51 0.60 0.48 °' 'if Source: INEI, 2005. .Q' Notes: Mantenimiento y Reposici6n Conexi6n (Connection Maintenance and Replacement) refers to the charge for maintenance and replacement {at the end of its service life) of the service drop and meter that belongs to the customer. lnteres Compensatorio (Compensation Interest) refers lo interest paid in case of delay in payment of the bill. Al.Al.PUB refers lo ... 3 an adjustment in the public lighting charge. t· "' a. en c ~ m i" Special Report Peru: National Survey of Rural Household Energy Use Figure 6.6 Average Cost/kWh, Electrosur 0.7 .s::. s: ~ en 0.6 a) D RURAL Ct ____ ... Ql O> f.t:! 0.5 ~ ID .... .._,/ ....., URBAN 0.4 0.3 ~--~--~--~---~--~--~--~ a 20 40 60 80 mo 120 140 kWh/month Source: 1NEI, 2005. Note: The square boxes represent the urban households in, the diamonds the rural households. consequence of the way that public lighting is billed- marginalized end-users who consume less than significantly discourages households from increasing 15 kWh/month (the .average monthly consumption their consumption from less than (or equal to) of the poorest quintile is 13.3 kWh) benefit much less 30 kWh/month to more than 30 kWh/month. in terms of the value of the subsidy in Soles than the Figure 6.7 suggests that, indeed, the consumption in the average consumer with 25 to 35 kWh/month, because 30 to 35 KWh/month interval is smaller than would be at low consumption levels it is the fixed charges that expected from a smooth regression of consumption. dominate the bill. The fixed charges are not subject However, when the deviations between 25 To toFOSE. 35 kWh/month are compared to other parts of the Indeed, recent research raises some questions frequency distribution curve, similar deviations from about the targeting performance of cross-subsidies a smooth consumption curve (for example between 40 such as FOSE to achieve lifeline rates (Komives et and 60 kWh/month) make it more difficult to argue al. 2005). The effectiveness of a subsidy to reduce that the first deviation is actually caused by the tariff income inequality can be gauged by the proportion jump at 30kWh/month. In short, there is no evidence of each sole of subsidy that reaches the poor. Since that the present billing practice limits consumption the ,Survey does not cover urban areas, the overall to below 30 kWh/month:. targeting performance of the FOSE cannot be assessed. However, one can assess the targeting Targeting Performance of FOSIE performance within rural households as a whole There is no question that FOSE significantly reduces by calculating how much of the total FOSE transfer the energy charge to rural consumers (by 50 percent reaches the poorer rural households. Using the tariff if supplied by the interconnected system, 62.5 percent curves discussed above, the FOSE benefit received by in the case of isolated systems) (Figure 6.8). However, households in each quintile can be assessed and the 68 6 Policy Implications of Survey Results Figure 6.7 Proportion of Households Reporting Given Levels of Consumption 14% I I l --r I I ACTUAL l 1-- I I r-- -, l I 12% l l I l ___ L___ L_ __ L --------L--------- I I I 1 1 HIGHERTHJ;1.N1 I l 10% I l : EXPECTED?. : l l _ _ _ !_ __ _ L-- -L 1 I l 8% '- - I - 1---- l -'- l ~--- 1 1 6% EXPECTED? : l ---f---- -A---- --- ~-- --t- --- --- 4% 2% --r LOWER THAN l - : EXPECTED? : -V-- I l I I r - r --r ;--- I I I 0% 10 20 30 40 50 60 70 Monthly Consumption, kWh/HH/month Source: INEI, 2005. Figure 6.8 Effective Rate of FOSE Discount 0.6 c: ::l 0.4 8 .!!l Cl w (/) 0 ~ u.. i 0.2 0 ·1 Q) «! ~ er: ~ ' ! +:> 0 0 ~ 0 10 20 30 50 70 00 90 100 110 120 Monthly Consumption, kWh Source: INEI, 2005. 69 Peru: National Ta'ble 6.8 : Targeting Performance for FOSE Transfer Average FOSE Expenditure . Average FOSE Benefit, Benefit, Soles/Month/ % of Total FOSE Quintile Average kWh/Month Soles/kWh Household Benefit 14.5 0.27 3.9 7.7 2 19.1 0.25 4.8 12.2 3 30. l 0.21 6.4 22.0 4 37.8 0.18 6.7 25.5 58.1 0.11 6.4 32.6 36.8 0.27 5.9 100 Source: INEI, 2005. total benefit that accrues to each expenditure quintile increase. The top quintile's share would decline from can be calculated.25 33 to 23 percent and the bottom quintile's share would The results of the targeting performance increase from 8 to 11 percent. calculations are shown in Table 6.8. Households in Additional improvements in the targeting the lowest quintile capture only 7.7 percent of the total performance could be achieved by further lowering POSE subsidy received by all rural households, yet the FOSE cap. If the 50 percent discount was limited this quintile constitute 20 percent of all households. to 15 kWh/month and phased out completely at The highest quintile captures 32.6 percent of the 25 kWh/month, the share of benefits going to the benefit. Indeed, households in the top quintile get an lowest quintile would be 19 percent, while the richest average benefit of 6.4 soles/month, versus 3.9 soles/ quintile would receive less than 10 percent. month in the bottom quintile. In short, the targeting performance of the FOSE is poor. These results are consistent with the findings Efficient Lighting of a recent OSINERG study (Gallardo and Bendezu As noted earlier, the poor use electricity very 2005). The study found that a significant proportion inefficiently for lighting. Because they cannot afford of the total subsidy reaches the nonpoor, though it the higher cost of fluorescent lights, most of their also found that errors of exclusion and inclusion26 are lighting is provided by incandescent bulbs that lower in rural than in urban areas. consume four times the kWh per lumen than do The impact of a revenue-neutral adjustment on fluorescent lamps. In fact, 46 percent of electrified rural targeting performance is readily simulated. If the total households in Peru report only incandescent lights. POSE subsidy to rural households is held constant, The economic case for linking future rural the FOSE discount on qualifying households could eledrification projects with an efficient lighting be raised. For example, if the POSE were phased out program is compelling. The rate of return on at 50 kWh/household/month rather than 100 kWh but replacing an incandescent bulb with a CPL, under the rate of the discount was held constant, the fraction conservative assumptions, is 100 percent (Box 6.1.) of the subsidy captured by the lower quintiles would This is significantly higher than the economic returns 25 This is done by: (1) subtracting from each monthly expenditure the fixed charges; (2) estimating the average energy charge; (3) applying the rate of FOSE discount to each household's monthly kWh consumption-which is either 50 percent or 62.5 percent of the average energy charge for rural customers, depending upon whether the customer is served by the interconnected system or an isolated system; (4) calculating the aggregate amount of the FOSE transfer received; and (5) sorting the FOSE amounts by expenditure quintile. 26 The error of exclusion is the fraction of the poor that do not benefit from a subsidy; the error of inclusion is the fraction of the nonpoor that do benefit from a subsidy. 70 ., ·~ ' 6 Policy Implications of Survey Results Box 6.1 Rate of Return for Replacement of Incandescent Lighting with CFLs on rural electrification per se (which is in the range of electricity bills than those with incandescent lights 26 to 59 percent for individual schemes, according to (i.e., the savings from efficient lighting would be spent the World Bank's Peru Rural Electrification Project on nonenergy items), or whether household energy (2005b) economic analysis). It necessarily follows expenditure would be roughly the same (i.e., savings that including an efficient lighting component would from efficient lighting would be spent on more TV, improve the aggregate economic returns. radio, or appliance use). The impact of an efficient lighting scheme on the Table 6.9 shows that households with only finances of distribution companies would also have fluorescent lamps spend more per month on their total to be analyzed. The key question would be whether electricity bill than households with incandescent households with fluorescent lights would have lower lamps only-a finding that is common to all quintiles. 71 Peru: National of Rural Household Use Table 6.9 Average Monthly Electricity Bill, Soles/Month Households with Both Households with Fluorescent Fluorescent and Incandescent Households with Incandescent Lamps Only Lamps Lamps Only 1 (Poorest) 9.0 8.1 7.6 2 9.7 10.3 7.4 3 13.2 12.7 10.2 4 16.4 15.5 11.3 5 25.7 23.4 14.1 Source: INEI, 2005. In other words, households that use the more developing policies of rural electrification. Examples efficient lighting may value electricity more (and of areas for further research could include the therefore spend more of their income on electricity) following: than households that use inefficient incandescent lighting. • Reasons for low electricity consumption. These could In urban areas, where consumption levels are include cultural preferences, lack of resources to much higher, worldwide experience with efficient buy appliances, lack of easily available appliances lighting programs shows a reduction in consumption in rural areas, or lack of promotion of electricity (and peak demands). Indeed, the energy conservation use in rural areas by the distribution companies impact is the principle rationale. But in poor rural and authorities. areas where household budget constraints limit • Reasons for low number ofsolar home systems in rural electricity use, the evidence of the Survey suggests areas. These could include a lack of promotion that this might not necessarily be true. by government authorities, donor agencies and But does this finding negate the argument of commercial solar photovoltaic companies. high economic returns, which are premised on energy • Ways to promote productive electricity use. Given savings? The answer is no: There is clearly an increase the low level of usage of income-generating in household welfare if for the same number of equipment reported in the Survey, it would be kilowatt-hours of electricity, higher levels of service useful to investigate how productive uses of (more lighting, more TV viewing, more radio, or other electricity could be best be promoted, including uses) are obtainable. experiences in other countries. Clearly, the incremental capital costs of providing CFLs to consumers as part of a rural electrification scheme are small. Rural electrification costs per Conclusions household are between US$445 and $600, so an This chapter uses data from the Survey to consider additional US$8 to $9 for three CFLs per household policy issues relevant to the creation and sustainability would have little impact on rural E;?lectrification project of rural electrification programs: budgets. • Connection rates in electrified villages. Almost one- quarter of households without electricity are in Issues for Further Research villages that are electrified. The most common The results of the Survey suggest a number of areas reason given for non-connection in these villages that would benefit from further research, to assist in is the upfront costs of connection, wiring, and 72 6 Policy Implications of Survey Results equipment. The financial sustainability of projects higher than the commonly assumed 0.5 to 1.0 is strongly influenced by connecting as many percent per year. households as possible, from which follows that • Pricing policy. Those who consume small amounts· connection costs, perhaps including house wiring, of electricity pay relatively high prices per should be part of the overall cost eligible for subsidy. kWh, notwithstanding the FOSE mechanism. • Distribution of electricity consumption by village. An Households in the lowest quintile capture only indicative consumption threshold of 22 kWh/ 7.7 percenf of the total FOSE subsidy received by household is used in this report for whether most all rural households, yet this quintile constitutes rural electrification schemes would be financially 20 percent of all households. The highest viable. Although the average consumption in 374 quintile captures 32.6 percent of the benefit. In electrified villages is 35 kWh/household/month, short, the targeting performance of the FOSE is these averages show significant variation across poor. Additional improvements in the targeting regions. In the Andean South region, the average performance could be achieved by further is only 15 kWh/household/month. As in the case lowering the FOSE cap. If the 50 percent discount of individual households, the regional differences were limited to 15 kWh/month and phased out in consumption can be explained by regional at 25 kWh/month, the share of benefits going to differences in income. the lowest quintile would be 19 percent, while the • Growth of electricity consumption. One of the richest would receive less than 10 percent. important assumptions in making financial • Efficient lighting. The economic case for linking projections of the viability of rural electrification future rural electrification projects with an · projects is the rate of growth in consumption. At efficient lighting program using compact least based on the experience of those communities fluorescent lamps (CFLs) is compelling. Rural prioritized by the current scheme (often the electrification costs per household are between poorest and most lacking in infrastructure access), US$445 and $600, so an additional $8 to $9 for there is no evidence that annual consumption three CFLs per household would have little growth per connected household would be much impact on rural electrification project budgets. 73 Annex 1 Survey Design and Methodology This annex has two main sections relevant to the design Table A.1.1 and methodology of the Peru National Survey of Rural Household Energy Use (henceforth known as the Survey). Distribution of the Sample Size The first section discusses the survey sample, weighting Expected and estimation procedures, questionnafre design, and Standard implementation of the Survey. Annex 4 contains the Sample Sample Deviation complete questionnaire. The second section compares the Region Conglomerates Houses (CV) Survey with the National Household Survey (Encuesta Coastal North 64 960 0.032 Nacional de Hogares, ENAHO). Coastal Central 64 960 0.050 Coastal South 48 720 0.038 Andean North 66 990 0.021 Survey Design Andean Central 68 1,020 0.029 It is essential to point out that the definition of rural Andean South 68 population center used in the National Survey of Rural Amazon 68 Household Energy Use is different from that used by INEI in the census. The definition used by INEI for the purpose Total 446 6,690 of the cemH1s is that rural population centers are those Source: INEI, 2005. with less than 100 dwellings grouped contiguously. The definition used in the National Survey of Rural Household 401-1000 houses, semi-rural aggregations of fewer than Energy Use for rural population centers are those with less 401 houses, and dispersed rural aggregations of population than 1000 dwellings grouped contiguously, a definition centers located in the interior of so-called Areas de that better represents the target population for rural Empadronamiento Rural (AER). 27 Once the stratifications electrification programs. This difference in definition of were made for each study region, sample conglomerates rural population centers means that the data from this were chosen. Each conglomerate is a geographic area survey cannot be directly compared with data from the with approximately 100 houses. The selection of the census qf other surveys conducted by INEI. conglomerate sample in each stratum was random, and The Survey covered 6,690 electrified and nonelectrified proportional to the number of houses in the stratum. households in rural areas of Peru. This sample is large Finally, in each conglomerate of the sample, 15 houses were enough to allow for reliable estimations about the survey randomly selected.28 Table A.1.1 depicts the distribution of population. The fieldwork was conducted in seven regions: the sample size by region.29 the Coastal North, Coastal Central, Coastal South, Andean Consequently, the sample type is probabilistic, North, Andean Central, Andean South, and Amazon stratified by areas, two-staged, and independent in each regions. In each of these study regions, a stratum was region of the study. In light of the fact that the number assigned that was proportional to the number of houses. of households selected in each study region is also There are three stratifications: peri-urban aggregations of proportional to the total population of that geographic This stratification is only utilized for sampling effects and not necessarily to obtain results for these levels. The AER is the geographic 'LI area conformed by a group of semidispersed houses that has, on average, 100 independent houses. These houses are grouped in one or more than one population center. 211 These conglomerates can be made up of one or several communities, depending on the size of the population. 29 The final database contains 6,476 records, 214 records having been found nonresponsive or otherwise unusable. 75 Peru: National of Rural Household Use area, the Survey data are representative at the regional The difference between weighted and unweighted and national level. data at the regional level is very small for the three Coastal The questionnaire and survey methodology were regions (2 percent or less), but much higher for the others designed to obtain information on the demand and use of (2-13 percent). The difference in the national average, electricity in rural areas of Peru and collect detailed data however, is 35 percent. The explanation for the large for an analysis of the economic and financial aspects of difference in the national average is simple: The Andean rural electrification in Peru (see Annex 4) Data Weighting regions (with low consumption) are underrepresented Procedures. in the sample, while the Coastal regions (with high The estimation methodology to process the Survey consumption) are overrepresented. Hence, the lower data involved the usage ofa weight or expansion factor that national average reflects the dominance of the Andean is multiplied by all the data of each register in the database. regions in the total population, giving them greater weight The final factor for each register has two components: the in the overall average. Basic Expansion Factor and the Adjusted Factors for no response. Estimation of Standard Errors The Basic Expansion Factor (Wi) for each sample home is determined by the sample design and equivalent to the and Confidence Interval inverse of the probability of final selection: For the current Survey, the sample errors of the estimations of the principal survey variables were calculated using Wi = 11/ the Variation Calculation System (CENVAR), which In order for the estimations derived from the Survey provided the estimators for sample variation for population to be representative of the population, it is necessary to parameters for the different regions of the estimation. The multiply data from each sample home contained in the precision of the estimation was measured through the database by the weight or expansion factor calculated sampling error, calculated statistically from the sample according to the sample design. Likewise, it is important data, and determined by the standard deviation: to adjust the expansion factors keeping in mind the T magnitude of the nonresponse. Given that the expansion c factors are calculated at the level of each conglomerate, it is advantageous to adjust the expansion factors to this level. A simple manner to interpret the sample error of an Table A.1.2 shows a comparison of weighted versus estimation performed from the Survey is presented in unweighted data for total kilowatt-hour (kWh) consumption. terms of the confidence interval. The confidence interval An obvious question is why (with one exception) the of the Survey is 95 percent and it was calculated in the weighted estimates are lower not just for the national following manner: average, but even within each region. p ± 1.96 * s The standard error was also used to obtain the variation Table A.1.2 coefficient (CV), also known as standard relative error. The CV allows the user to evaluate the precision of the estimator Weighted versus Unweighted Estimates of in relative terms (see Table A.1.1 for the estimated variance kWh/HH/Month of the sample by region). Unweighted Weighted Difference Confidence limits are calculated from the standard [1] [2] [3] form: Coastal North 38.4 38.3 -0% Coastal Central. 60.8 61.7 2% Coasta I South 59.6 59.1 -1% where n is the sample size, xis the estimated sample inean, Andean North 23.0 21.7 -6% and s is the sample standard deviation. Andean Central 27.4 26.9 -2% Andean South 18.9 16.7 -13% Questionnaire Design and Conduct Amazon 34.2 31.6 -8% of the Survey All Households 38.9 27.2 -35% The questionnaire was developed with input from the Source: INEI, 2005. National Institute of Statistics and Information Technology 76 Annex 1 Survey Design and Methodology (INEI). The Technical Directorate of Demographics and • Section 2: Use of Kerosene Sodal Indicators of INEI was responsible for executing • Section 3: Use of Candles the fieldwork from April thr~ugh July 2005 in the 24 • Section 4: Use of Dry Cell Batteries departamentos of Peru. The Technical Directorate was also • Section 5: Use of Car Batteries in charge of the data processing. • Section 5: Use of LPG The questionnaire was drafted and revised by both the • Section 7: Use of Solar PV Home System World Bank team and the MEM. Adjustments were made • Section 8: Electric Generator Set to the health, energy use, time use, and income sections • Section 9: Use of Firewood of the initial survey, which was then tested in the field. In • Section 10: Use of Agriculture Residue April 2005, the questionnaire was piloted in rural villages • Section 11: Animal Dung in three different geographic regions of Peru: lea (Coastal), • Section 12: Use of Cooking Stove and Cooking Huancayo (Andean), and Chanchamayo (Amazon). • 400. Productive Equipment Discussions were held with the surveyors during the pilot • Section 1: Electric Pumps survey implementation to identify problematic questions • Section 2: Diesel Pumps and difficulties that arose in the field and to eliminate any • 500. Time Use confusion. The questionnaire was then revised based on • 600. Household Income feedback from the surveyors and lessons learned in the • Section 1: Income from Work field. The final survey questionnaire was designed to collect • Section 2: Income from Agricultural Activities very detailed in formation on energy use and consumption. • Section 3: Income from Livestock Activities In May 2005, the World Bank held a training session • Section 4: Income from Fisheries with the survey supervision team at INEI, which was • Section 5: Other Income followed by a six-day workshop held in Lima by INEI for • Section 6: Household Expenditures the surveyors hired to execute the final Survey. A total of • 700. Attitude 75 surveyors and 24 supervisors (one for each departamento • 800. Business Module of Peru) were selected to implement the survey. Generally, • Section 1: Basic Characteristics of the Business teams included surveyors who were native to the area or or Establishment who had prior work experience in the area to which they • Section 2: Financing Sources for Business were assigned. Regional INEI offices and the regional • Section 3: Uses of Motors (Motive Power) in supervisors provided additional support to the survey Business teams. The central INEI office in Lima was responsible • Section 4: Income from Business for providing the statistic and cartographic information, • 900. Opinion and Attitude on Energy and Business providing technical and logistical support, and delivering the final survey results. Comparison between the National Completed survey forms were sent to the regional INEI office in Arequipa, where the data entry was performed. Survey of Rural Household Energy Once completed, the survey results were sent back to Lima, Use and the National Household where INEI took additional steps to check the accuracy of Survey data entry. The final data editing and preparation for data analysis was performed and completed in Washington, DC. This section compares the estimates of energy expenditures and total household expenditures obtained by the National Survey of Rural Household Energy Use (the Survey) Household Energy described in the main text of this report with those from Questionnaire Outline the ENAHO, a national household survey done by INEI every year to estimate socioeconomic characteristics and The following is an outline of the survey questionnaire. The poverty conditions in Peru. full survey questionnaire is included in Annex 3. This exercise is performed even though it is clear that the Survey and the ENAHO are not strictly comparable. • 100. Characteristics of the House and Household First, the ENAHO is carried out continuously throughout • 200. Characteristics of the Household Members the year, while the Survey was implemented during June • 300. Sources of Energy and July of 2005, which creates seasonality issues that are • Section 1: Use of Electricity from Interconnected Grid beyond the scope of this study. In addition, the ENAHO data and Isolated System used here are from 2004. Second, the sampling design and 77 Special Report Peru: National Survey of Rural Household Energy Use sample size are different. In particular, as noted under the cost of house rental for families who own a house. For Survey Design section, the definition of rural populations these reasons, it was decided that the income figures in is different. Although the ENAHO accumulates information the ENAHO and the Survey are not strictly comparable. from 8,240 households throughout the year, the Survey A more reliable estimate of household socioeconomic well interviewed 6,776 households. Third, and perhaps most being in both surveys is total household expenditure. importantly, the questionnaire designs differ because each The aggregate estimates for energy expenditure in the survey pursues a different objective. The Survey contains Survey-that is, considering all types of energy used by much more detail on energy use, expenditures, and demand. the household-results in 34.3 soles per month, while the As a result, the Survey estimates of energy consumption are ENAHO estimate is 14.6 soles per month. This difference of more accurate. Due to these differences, the two studies are approximately 20 soles between the two surveys is mainly compared to identify similar trends, rather than to prove or explained because the Survey captured almost 4 soles more disprove actual values. of electricity expenditure, 6 soles more of LPG expenditure Income information from the Survey was not included (three times than ENAHO estimates), and 2.2 soles more in the analysis because the Survey did not collect as of fuelwood expenditures. In addition, almost 5 soles of detailed information as the ENAHO, meaning that the difference comes from energy expenditures that only the Survey income figures are lower than the ENAHO ones. Survey collects-that is, dry cell batteries, car batteries, and The ENAHO collects both monetary and nonmonetary electric generators (small generators). income data and imputes many values that are not collected The details of the estimates are summarized in in the Survey. For example, a significant component of Table A.1.3 As seen in this table, when looking at each ENAHO's income estimation comes from the opportunity of the total energy expenditure components, there are Table A.1.3 Comparison of Monthly Energy Expenditure and Total Cash Expenditure between the ENAHO and the Survey ·ENAHO Survey Expenditure Expenditure Expenditure Expenditure Type of Energy % Users All % Users All Expenditure (EE)lll Users N Mean N Mean Users N Mean N Mean 31.6 13.8 4.4 45.6 2,954 18.27 8.33 Kerosene 52.6 4,338 7.7 8,240 4.1 50.5 3,271 11.9 6,476 6.00 LPG 10.5 862 28.0 8,240 2.9 26.9 1,745 34.4 6,476 9.26 Candle 23.3 4.9 1.1 56.6 3.6 2.02 Coal 0.7 61 20.3 0.2 Fuelwood 9.3 765 17.1 1.6 13.1 849 29.6 3.88 Diesel 2.5 205 9.0 0.2 Gasoline 0.2 15 41.5 8,240 0.1 68.4 4,432 5.2 6,476 3.58 13.8 894 7.9 1.1 Small generator 0.4 26 36.4 6,476 0.15 EE whole sample including zeros or 8,240 14.6 34.3 8,240 375.1 6,476 375.6 8,240 389.6 409.9 Source: ENAHO 2004, Four quarterly rounds, INEI, 2005. (1) Households with missing values in any of the variables of the Energy Expenditure variables were recoded with zero. (2) Includes cash expenditure from food, dress, clothing, investment in HH, furniture and equipment, house, conservation, health core and medical services, transportation and communication, leisure and entertainment, and others. Does not include selkonsumption, self-supply, expenditure in agriculture activities, and livestock. (3) Result of adding EE and Other Cash Expenditures. 78 Annex 1 Survey Design and Methodology similar tendencies in the average values and in the Energy Expenditures (Electricity, proportion of users from both surveys. Perhaps the most intriguing difference is the share of LPG and candle users; Kerosene) by Region In ENAHO, only 10.5 percent of the households mention Electricity and kerosene are two of the most commonly LPG expenditures, while in Survey, almost 30 percent of used energy sources for rural households, so it is important the households mention LPG expenditures. By contrast, for policy makers to analyze their consumption in detail. in ENAHO, 23 percent of households reported spending However, the question arises about whether similar trends money on ·candles versus 56.6 percent of households in electricity and kerosene expenditures as those observed in Survey. These facts contribute to the higher Survey at the national rural level between the two surveys maintain estimates of total energy expenditures. when desegregating the data at the regional level. This One way to verify that the energy expenditure section performs these comparisons. Results confirm estimates in the Survey are accurate is to discount similar regional behavior for electricity and kerosene the energy expenditure estimates from the total cash expenditures. In addition, the average difference in the expenditure in both surveys and compare the results. The absolute values of the variables is only 4.4 soles lower for resultant variable is called "Other Cash Expenditures." As electricity expenditure· and 4.1 soles lower for kerosene shown in Table A.1.3, the average estimate of this variable expenditure in ENAHO. This variation is expected since the in both surveys is almost the same: 375.1 soles in ENAHO Survey has more detailed questions to collect informatfon and 375.6 soles in the Survey. In short, it is possible to about energy expenditures. In addition, part of the confirm that the Survey has collected valuable and accurate difference could be explained by issues such as seasonality information concerning energy consumption and also and distinct sampling design objectives. r.eliable information about total household expenditures Electricity. Figure A.1.1 helps to visualize the· when excluding energy expenditures. At the same time, electricity expenditure trends between the two surveys. In these results provide support for the construction of the general, it is possible to confirm that estimates move in the indicator: share of energy expenditures to total household · same direction when data are desegregated by region. In expenditures, which is commonly used to perform country respect to the absolute values, two regions, Central Coastal comparisons. and Southern Coastal, have the smallest difference, 0.6 soles and 0.5 soles, respectively, whereas ENAHO has the lower Figure A. 1 . 1 Monthly H.ousehold Electricity Expenditure by Region (Users Only) Northem Central Southem Northern Central Central South Amazon Coastlll Coastal Coastal Mountain Mountain Region -.i.- E NAHO - NR ES Source: ENAHO 2004, four quarterly rounds, INEI, 2005. Note: The NRES refers to the Survey. 79 Peru: National of Rural Household Use monthly electricity expenditure figures. This difference difficulties involved for surveyors to reach the most isolated can be visualized in Figure A.1.1 as the gap between the households in the Amazon region. 30 two lines. Table A.1.4 shows detailed information for estimated The Amazon region has the highest disparities values and percentage differences from the two surveys. between the electricity estimates. The monthly expenditure It is important to note that in the ENAHO survey, there for the ENAHO survey is 2.8 soles lower than the Survey. is a higher frequency of households in regions with low One possible explanation is that the Survey may have levels of electricity expenditure, while the inverse is true surveyed more semirural households-those who use for the Survey. These different frequencies arise due to the electricity more intensively-due to the geographic distinct sampling design of each survey. Consequently, Table A.1.4 Monthly Household Electricity Expenditure by Region (Users Only) ENAHO Region Mean (soles) N Mean N Difference Coastal North 17.5 259 19.9 371 2.3 Coastal Central 26.5 192 27.1 565 0.6 Coastal South 24.6 117 25.1 445 0.5 Andean North 10.5 123 11.4 311 0.9 Andean Central 11.8 764 13.6 540 1.8 Andean South 9.5 669 10.3 453 0.8 Amazon 14.2 483 17.0 269 2.8 Total Rural Peru 13.8 18.3 2954 4.4 Source: ENAHO 2004, four quarterly rounds, INEI, 2005. Figure A.1.2 Monthly Kerosene Expenditure by Region (Users Only) Northern Central Southern Northern Central Central South Amazon Coastal Coastal Coastal Mountain Mountain Region -11r- ENAHO -11- NRES Source: ENAHO 2004, four quarterly rounds, INEI, 2005. Note: The NRES refers to the Survey. 30 Isolated Amazon areas are also very dangerous not only because of drug trafficking, but also because the communities themselves are reluctant to have contact with people outside their communities. A few months prior to the Survey implementation, four health professionals from the Ministry of Health were assassinated by native people from Tagkijap community (El Comercio, 05/21/2005). 80 Annex 1 Survey Design and Melhodology there is a difference in the total rural electricity expenditure Total Household Cash Expenditure estimates of 4.4 soles. Kerosene. Results of the kerosene expenditure At the beginning of this section, it was mentioned that comparison are very similar to electricity expenditures. The when excluding energy expenditures, the total household average variation in the absolute values is a lower kerosene expenditures are similar in both surveys at 375 soles per expenditure of 4.1 soles in the ENAHO. Differences at the month. However, the mean is only a summary measure of one regional level can be observ:ed in Figure A.1.2. The estimates variable, which could hide great differences, especially when for this variable are especially close in the Coastal North comparing two variables--"in this case, total expenditure in and South regions as well as in the Amazon and Andean ENAHO versus the Survey. It is important to look with more North regions. Disparities for these regions are less than detail and make a comparison throughout the distribution of 1.5 soles. In contrast, the Central Coastal and the Andean values from both variables. This section attempts to address South regions present the highest disparities. Despite these this issue. Results confirm that total cash expenditure excluding differences, it is clear that similar tendencies exist among energy expenditure in both surveys is very similar, not only as the regions for kerosene expenditure estimates. an average, but also throughout the entire distribution ofvalues. Table A.1.5 shows detailed information for ·estimated In order to do the comparison, a nonparametric values and percentage disparities from the two surveys. As estimation was performed to construct densities for each of in the case of electricity, the same patterns can be observed the two variables. Put simply, these densities, generally known here: the total average kerosene expenditure is 4.1 soles as kernel densities, show the concentration or the frequency lower in ENAHO mainly because the households from this that the given variable takes on a certain numerical value. 31 survey are concentrated in regions with lower expenditures Table A.1.6 summarizes the numerical values of the as a result of the survey's distinct sampling design. Indeed, variables analyzed in this report. Logarithm values were when looking at the regional level, only the Coastal Central used in the estimation in order to smooth the distribution region presents a difference that is dose to the total rural of the values and also avoid distortions that would originate expenditure (4.5 soles). from outliers. Table A.1.5 Monthly Household Kerosene Expenditure by Region {Users Only) ENAHO Region Mean (soles) N Mean (soles) N Difference (soles) Coastal North 12.7 448 14.3 642 l.5 Coastal Central 18.9 87 23.4 294 4.5 Coastal South 23.9 68 25.0 205 1.1 Andean North 6.0 664 7.5 611 1.5 Andean Central 6.1 902 8.6 428 2.5 Andean South 5.3 814 8.2 459 2.9 Amazon 8.0 9.0 632 1.0 Total Rural Peru 7.7 4,338 11.9 3~271 4.1 Source: ENAHO 2004, four quarterly rounds, INEI, 2005. 31Following Chapman and Hall (1995), the kernel density estimator for given a random sample X1,. •• ,Xn with a continuous univariate I Yf. den s't ' is: f(x,h)= 1 " nhLK (x T -X ) w1"th kerneI K and bandwid th h. In this case it was performed a Epanechnikov kernel function, 1-1 (1 2 )' which is K(x,p) when p = l, where K(x, p) = -x l{I x kl}, with B(a,b) f(a)r(b) 22 •* 1 B(p+l,p+l) r(a+b) 81 Special Report Peru: National Survey of Rural Household Energy Use Table A.1.6 Monthly Household Kerosene Expenditure by Region Variables N Mean Std. Dev Min Max Absolute values ENAHO 8,208 376.55 366.46 6,837.67 6,452 377.02 331.31 2 4,688.00 8208 5.50 1.04 0.00 8.83 6452 5.57 0.94 0.69 8.45 Source: ENAHO 2004, four quarterly rounds, INEI, 2005. Note: There are fewer observations because only positives values were kept due to the fact that logarithms exist only in this domain. Figure A. 1.3 Kernel Density Estimation of Total Monthly Household Cash Expenditure without Energy Expenditures {logarithm) "' 'I'. ;;;:;. -~(".) ID 0 Qi EN ID , :.:: '": a 0 2 4 6 8 10 Nuevo Soles Log Source: ENAHO 2004, four quarterly rounds, INE!, 2005. Note: NRES refers to the Survey. The resulting densities can be observed in information about energy expenditures. Other reasons Figure A.1.3. The main conclusion from this exercise is could be the seasonality involved in the different surveys' that both distributions are close: in other words, for each implementation and the different objectives of the sampling log value of total expenditure, both the ENAHO and the designs of each survey. Specifically, the Survey was Survey show values whose concentration levels are very implemented one year after the ENAHO survey and most similar. Perhaps the most important difference between importantly, this former Survey was implemented specifically the two surveys occurs to the right of the average, in which in rural areas with an inference level for each rural region. the Survey estimate is slightly higher than that of ENAH 0. · Second, when excluding energy expenditures from total household expenditures, the average expenditure is almost the same for both surveys. Moreover, the similarities Survey Comparison Conclusions maintain not only on average, but also throughout the entire The comparison of the National Rural Energy Survey with distribution of values. the ENAHO survey has generated three important findings. Third, when desegregating electdcity and kerosene First, energy expenditures are higher in the Survey. expenditures at the regional level, despite some variations, Reasons behind this difference are due mainly to the fact there is a clear similarity in the tendencies of the estimates that the Survey is more detailed for questions that collect from both surveys. 82 Annex2 Survey Results All the tables shown in Annex 2 summarize the results of the National Survey of Rural Household Energy Use (INEI, 2005). The results reported in this annex are after applying the weighting factors discussed in Annex 1, Section 4, unless otherwise specified. Table A.2.1 Percentage of Households that Use Each Type of Energy, by Region · Coastal Region Andean Region All North Central South North Central South Amazon Candle 47% . 53% 60% 56% 69% 66% 46% 60% Kerosene 71% 32% 31% 71% 44% 52% 73% 57% Small generator 0.9% l.3% 1.0% 0.2% 0.9% 0.6% Dry cell battery 71% 51% 55% 78% 66% 74% 91% 74% Car battery 31% 23% 13% 9% 8% 7% 15% 11% Grid 35% 60% 71% 22% 52% 44% 18% 39% LPG 28% 63% 53% 5% 17% 10% 7% 14% Fuelwood 85% 74% 68% 94% 92% 64% 95% 84% Solar PV 0.3% 0.1% 0.1% 0.4% 0.9% 1.1% 0.5% 8% 7% 5% 5% 18% 13% 3% 11% 0.4% 0.5% 15% 3.6% 26% 65% 0.1% 25% All Households 156,419 75,314 27,787 362,029 634,240 565,024 383,403 2,204,216 100% 100% 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 32 Source for Annex 2 is: Peru National Survey of Rural Household Energy Use, INEl, 2005. 83 Special Report Peru: Nalional Survey of Rural Household Energy Use Table A.2.2 Percentage of Households that Use. Each Type of Energy, by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All Candle 54% 60% 60% 62% 63% 60% Kerosene 61% 61% 57% 54% 51% 57% 1.1 % 0.1% 0.3% 0.5% 1.1% 0.6% 64% 80% 77% 74% 76% 74% 4% 7% 10% 16% 19% 11% 27°/o 34% 38% 45% 50% 39% 1% 5% 8% 21% 37% 14% 87% 86% 84% 83% 81% 84% Solar PY 0.3% 0.1% 0.5% 0.2% 1.5% 0.5% 16% 12% 10% 9% 8% 11% 31% 34% 26% 19% 15% 25% All Households 441,398 441,612 440, 132 440,247 440,827 2,204,216 ,100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 84 Annex 2 Results Table A.2.3 Total Household Monthly Cash Spending on Energy by Region, in Soles (Users Only) Coastal Region Andean Region All North Central South North Central South Amazon Candle 2.74 5.34 4.43 2.97 3.41 3.49 3.67 3.43 Number of households 73,657 39,680 16,723 201,768 434,878 372,737 177,456 1,316,898 Kerosene 14.57 22,80 26.06 7.51 8.39 8.02 9.39 9.30 Number of households 111,037 24,201 8,721 258,658 277,913 290,873 281, 155 1,252,557 Small 41.37 37.20 20.28 33.57 33.20 Number of households 1,431 887 1,278 3,353 6,949 Dry cell battery 4.48 6.26 5.29 4.34 5.42 4.40 7'.45 5.36 Number of households 110,934 38,683 15,200 282,821 417,450 417,082 349,836 1,632,008 Car 6.18 12.57 8.65 5.53 5.76 5.95 6.76 6.60 Number of households 47,704 16,956 3,719 33,871 50,252 36,970 56,522 245,994 Grid 19.82 27.04 24.66 10.87 13.36 9.44 16.03 13.63 Number of households 51,328 43,751 18,739 78,070 327,738. 229,695 63,624 812,945 LPG 30.90 36.87 36.48 32.70 32.70 28.40 33.68 32.60 Number of households 43,299 47,376 14,734 18,567 109,233 54,359 26,274 313,843 Fuelwood 35.60 38.71 48.48 27.13 21.62 29.06 30.18 26.58 Number of households . 22,355 5,884 2,407 50,898 113,628 62,483 15,286 272,941 All Energy 37.22 58.55 55.06 18.76 26.42 19.89 22.83 25.09 7.6% 9.5% 9.6% 9.9% 11.9% 9.3% 7.4% 9.7% Number of Households 156,419 75,315 27,787 362,029 634,240 565,023 383,403 2,204,215 Source: Authors' calculations, 2005. Note: Expenditure on car battery only includes recharging fee. 85 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.4 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles, in Soles (Users Only) 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All Candle 2.71 2:95 3.62 3.64 4.13 Valid N 235,931 265,881 265,618 270,536 278,933 1,316,898 Kerosene 4.89 6.61 8.80 11.36 16.19 9.30 Valid N 270,397 269,364 249,869 238,264 224,663 l,252,557 Small 13.00 28.97 38.14 33.20 ValidN 775 1 Dry cell battery 3.39 4.50 5.34 6.00 7.30 5.36 Valid N 280,933 352,093 339,003 324,723 335,256 1,632,008 Car 5.19 5.10 5.83 6.73 7.39 6.48 Valid N 18,351 31,570 42,297 70,281 83,495 245,994 Grid 7.36 8.54 10.38 4.20 22.52 13.63 ValidN 113,534 143,193 160, 146 187,719 208,353 812,945 LPG 20.42 20.64 26.20 30.08 37.07 32.60 Valid N 3,649 l 9 ,861 33,750 93,032 163,550 313,843 Fuelwood 13.63 17.87 22.94 27.73 35.98 26.58 ValidN 17, 184 46,243 56,207 74,155 79, 153 272,941 9.41 15.33 20.59 31.08 49.06 25.09 % of Total 17.1% 9.9% 8.2% 7.4% 5.8% 9.7% Valid N 441,398 441,612 440, 132 440,248 440,826 2,204,215 Source: Authors' calculations, 2005. 86 Annex 2 Results Table A.2.5 Total Household Monthly Cash Spending on Energy by Region, in Soles (All Households) Coastal Region Andean Region All North Central South North Central South Amazon Candle 1.29 2.81 2.66 1.65 2.34 2.30 1.70 2.05 Kerosene 10.34 7.33 8.18 5.37 3.67 4.13 6.89 5.28 0.38 0.44 0.05 0.29 . 0.10 3.18 3.22 2.90 3.39 3.57 3.24 o.79 3.97 1.89 2.83 1.14 0.51 0.43 0.39 0.99 0.72 6.50 15.71 16.63 2.34 6.91 3.84 2.66 5.03 LPG 8.55 23.19 19.34 1.68 5.63 2.73 2.31 4.64 Fuelwood 5.09 3.02 4.20 3.81 3.87 3.21 1.20 3.29 All Energy 37.22 58.55 55.06 18.76 26.42 19.89 22.83 25.09 As% of Total Expenditure 7.6% 9.5% 9.6% 9.9% 11.9% 9.3% 7.4% 9.7% Number of Households 156,419 75,315 27,787 362,029 634,240 565,023 383,403 2,204,215 Source: Authors' calculations, 2005. Table A.2.6 Total Household Monthly Cash Spending on Energy by Expenditure Quintiles, in Soles (All Households) 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month . S/month S/month All Candle 1.45 l.78 2.19 2.24 2.61 2.05 Kerosene 3.00 4.03 5.00 6.15 8.25 5.28 Small generator 0.02 0.11 0.39 0.10 Dry cell battery 2.16 3.59 4.11 4.43 5.55 3.97 Car battery 0.22 0.36 0.56 1.07 . l.40 0.72 Grid l.89 2.77 3.78 6.06 10.64 5.03 LPG 0.17 0.93 2.01 6.36 13.75 4.64 Fuelwood 0.53 1.87 2.93 4.67 6.46 3.29 All 9.41 15.33 20.59 31.08 49.06 25.09 % of Total Spending 17.1% 9.9% 8.2% 7.4 5.8 9.7 V~lid N 441,398 441 ,612 440, 132 440,248 440,826 2,204,215 Source: Authors' calculations, 2005. 87 Special Report Peru: Nalional Survey of Rural Household Energy Use Table A.2.7 Percentage of Households that Use Each Type of Energy by Electrification Status and Region Coastal Region Andean Region Amazon Ali North Central South North Central South Region Regions With Access to Grid Candle 50% 31% 49% 41% 57% 52% 48% 51% Kerosene 31% 15% 21% 12% 19% 20% 26% 20% 0.2% 0.0% 46% 34% 42% 50% 52% 61% 77% 55% Car battery 0.3% 0.2% 3.5% 1.0% 0.4% 0.8% 1.5% 0.7% Grid electricity 100% 100% 100% 100% 100% 100% 100% 100% LPG 58% 74% 65% 18% 27% 14% 29% 28% Fuelwood 70% 66% 60% 89% 89% 74% 87% 81% Solar PY 0.2% 0.0% 4% 4% 4% 1.6% 17% 14% 5% 12% 0.1% 0.4% 12% 1.7% 25% 54% 0.7% 26.2% All Households 54,585 45,378 19,651 80,260 332,084 250,561 68,990 851,509 100% 100% 100% 100% 100% 100% 100% 100% Without Access to Grid Candle 46% 85% 88% 60% 81% 77% 46% 65% Kerosene 93% 58% 57%. 89% 71% 76% 84% 80% Small l.4% 3.2% 2.0% 0.4% 1.0% 1.0% Dry cell battery 85% 77% 85% 86% 81% 84% 95% 86% Car battery 47% 56% 37% 12% 16% 11% 18% 18% Grid electricity LPG 11% 46% 2% 6% 6% 2% 6% Fuelwood 93% 85% 86% 95% 95% 57% 97% 86% Solar PY 0.4% 0.3% 0.5% 0.5% 1.6% 1.3% 0.8% residue 10% 12% 8% 6% 19% 13% 3% 10% Animal dung 0.5% 0.5% 21% 4% 28% 73% 24% All Households 101 136 281,769 Source: Authors' calculations, 2005. 88 Annex 2 Results Table A.2.8 Percentage of Households that Use Each Type of Energy by Electrification Status and Expenditure Quintile 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 All Expenditure S/month S/month S/month S/month S/month Quintiles With Access to Grid Candle 45% 49% 55% 52% 53% 51% Kerosene 18% 21% 18% 20% 21% 20% 48% 58% 57% 52% 58% 55% 0.6% 0.3% 0.3% 0.7°/o 1.3% 0.7% 100% 100% 100% 100% 100% 100% LPG 9% 16% 36% 58% 28% Fuel wood 85% 85% 84% 81% 75% 81% Solar PV 0.1% 0.1% Ag. residue 17% 18% 13% 10% 7% 12% Animal dung 38% 39% 31% 20% 13% 26% All Households 118,912 149.,335 168,910 195,931 218,422 851,510 100% 100% 100% 100% 100% 100% Without Access to Grid Candle 57% 66% 64% 69% 74% 65% -Kerosene 77% 81% 81% 82% 80% 80% Small 1.5% 0.2% 0.5% 0.8% 2.1% 1.0% Dry cell battery 70% 91% 90% 92% 94% 86% Car battery 6% 11% 15% 28% 36% 18% Grid electricity LPG 0.5% 2.4% 2.7% 9% 16% 6% Fuelwood 88% 86% 85% 84% 88% 86% Solar PV 0.3% 0.1% 0.8% 0.4% 2.9% 0.8% residue 15% 10% 7% 8% 8% 10% Animal dung 28% 32% 24% 19% 16% 24% All Households 322,486 292,276 271,222 244,317 222,404 1,352,705 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 89 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.9 Annex 2 Results Table A.2. 10 Comparison of Total Hot.!sehold Monthly Cash Spending on Energy between Households with and without Access to Grid Electricity by Expenditure Quintiles (Users Only) 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 All Expenditure S/month S/month S/month S/month ·s/month Quintiles With Access to Grid Electricity Candle l .18 1.10 1.31 1.28 1.56 1.32 Valid N 53,630 73,649 92,432 l 01,956 115, 184 436,850 Kerosene 3.94 4.57 7.07 11.94 16.51 9.88 ValidN 21,863 31,729 30,902 38,947 46, 105 169,545 Small 28.50 28.50 Valid N 158 158 Dry cell battery 2.66 3.12 3.66 3.85 4.39 3.68 Valid N 55,887 87,221 96, 118 100,864 126,562 466,652 Car 3.00 6.00 8.00 4.46 6.89 5.72 ValidN 672 516 504 1,329 1,904 4,925 Grid 7.36 8.54 10.38 14.20 22.52 13.63 Valid N 113,534 143,193 160, 146 187,719 208,353 812,945 LPG 25.71 19.40 24.62 30.43 36.96 32.63 Valid N 12,878 0 70, l 33 127,334 238,799 Fuelwood 14.50 19.32 22.23 27.37 34.17 27.00 ValidN 7,337 24,440 30,518 56,273 56,331 174,899 All 10.89 16.38 21.82 37.43 58.75 32.41 Valid N 118,912 149,336 168,909 195,931 218,422 851,510 Without Access to Grid Candle 3.15_ 3.66 4.86 5.08 5.93 4.49 Valid N 182,301 192,231 173,186 168,581 163,749 880,048 Kerosene 4.97 6.88 9.04 11.25 16.10 9.20 Valid N 248,535 237,635 199,317 178,558 ' 1,083,012 Small 29.02 38.14 33.31 Valid N 3.58 4.95 224,374 264,873 242,381 208,694 l, 164, 180 Car battery 5.28 5.20 6.13 6.83 7.49 6.61 Valid N 30,359 39,556 68,436 80,653 236,683 Grid ValidN LPG 13.69 22.93 31.88 29.02 37.48 32.49 Valid N 1,606 6,983 7,340 22,899 36,216 75,044 Fuelwood 12.99 16.25 23.79 28.85 40.46 25.83 ValidN 9,847 21,803 25,689 17,881 22,821 98,042 All 8.86 14.79 19.83 25.99 39.55 20.48 ValidN 322,486 292,276 271,223 244,316 222,405 1,352,705 Source: Authors' calculations, 2005. 91 Special Report Peru: National Survey of Rural Household Energy Use Table A.2. 11 Number and Percentage of Households Using Kerosene for Lighting and Cooking by El,ectrification Status and by Region (All Households) Coastal Region Andean Region Amazon All North Central South North Central South With Access to Grid Cooking l, 188 3,743 3,335 334 14,520 12,901 551 36,572 2.2% 8.2% 17.0% 0.4% 4.4% 5.1% 0.8% 4.3% 7, 125 972 233 4,070 24,447 27,915 12,714 77,476 2% 1% 5% 7% 11 % 18% 9% .0 263.0 139.0 181.0 672.0 2.7% l.3% 0.2% 0.7% 0.3% 0.0% 0.6% 1,077 235 4,815 9,315 2.4% l.2% 6.0% 6.9% 3.7% 5.9% Not used 38,383 70% 85% 79% 88% 81% 80% 74% 80% Total 54,583 45,378 19,652 80,261 332,085 250,561 68,990 851,510 100% 100% 100% 100% 100% 100% 100% 100% 278 11 115 0.3% 5.3% 17.9% 0.0%. 0.6% 1.9% 0.0% 0.8% Lighting 92,906 12,093 2,505 233,750 193,485 191,304 231,423 957,466 91% 40% 31% 83% 64% 61% 74% 71% 325.0 2,740.0 380.0. 1,454.0 8,850.0 11,749.0 6,993.0 32,491.0 0.3% 9.2% 4.7% 889 793 314 (%) 0.9% 2.6% 3.9% 5.0% 3.3% 9.8% 7.9% 6.1% Not used 7,437 12,731 3,480 32,468 88, 184 74,393 51,001 269,694 7% 43% 43% 12% 29% 24% 16% 20% Total l 01,835 29,937 8, 135 281,768 302, 155 314,463 314,412 1,352,705 100% 100% 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 92 Annex 2 Survey Results Table A.2. 12 Number and Percentage of Households Using Kerosene for Lighting and Cooking by Region (All Households) Coastal Andean Region Amazon All North Central South North Central South Region Regions All Areas ·(Electrified and Unelectrified) Cooking 1,467 5,323 4,790 334 16,235 18,986 551 4(,686 !%) 0.9% 7.1% 17.2% 0.1% 2.6% 3.4% 0.1% 2% Lighting 100,031 13,066 2,738 237,820 217,932 219,219 244, 137 1,034,943 (%) 64% 17% 10% 66% 34% 39% 64% 47% Lighting and cooking 905 3,943 643 1,593 11,031 12,420 6,993 37,528 0.6% 5.2% 2.3% 0.4% 1.7% 2.2% 1.8% 1.7% . Other purposes 8,634 l,869. 549 18,911 32,715 40,247 29,474 132,399 5.5%- 2.5% 2.0% 5.2% 5.2% 7.1% 7.7% 6.00% Not used 45,382 51,114 19,066 103,371 356,327 274, 151 102,248 951,659 (%) 29% 68% 69% 29% 56% 49% 27% 43% Total 156,419 75,315 27,786 362,029 634,240 565,023 383,403 2,204,215 100% 100% 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 93 Special Report Peru: National Survey of Rul"al Household Energy Use Table A.2.13 Number and Percentage of Households Using Kerosene for Lighting and Cooking by Electrification Status and Expenditure Quintiles (All Households} 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All With Access to Grid 6,336 11 131 1.1% 1.6% 3.8% 5.7% 7.0% 4.3% Lighting 15,430 20,478 13,440 13,480 14,648 77,476 (%) 13% 14% 8% 7% 7% 9% 47 2,523 0.0% 0.0% 0.8% 0.5% 1.2% 0.6% Other purposes 5,071 8,831 9,725 13,268 13,565 50,460 4% 6% 6% 7% 6% 6% Not used 17 681 (%) 82% 79% 82% 80% 79% 80% Total 11 0 100% 100% 100% 100% 100% 100% Without Access to Grid Cooking 252 813 4,341 2,737 2,972 11,115 0.1% 0.3% 1.6% 1.1% 1.3% 0.8% Lighting 232,222 211,492 187,956 170,400 155,397 957,467 {%) 72% 72% 69% 70% 70% 71% Lighting and cooking 2, 111 8, 144 4,635 8,291 9,310 32,491 0.7% 2.8% 1.7% 3.4% 4.2% 2.4% Other purposes 13,950 17, 185 22,035 17,889 l 0,879 81,938 (%) 4% 6% 8% 7% 5% 6% Not used 73,951 54,641 52,255 44,999 43,846 269,692 23% 19% 19% 18% 20% 20% Total 322,486 292,275 271,222 244,316 222,404 1,352,703 100% 100% 100% 100% 100% 100% All Households l 13 187 (%) 0.4% 0.7°/o 2.4% 3.2% 4.2% 2.2% Lighting 247,652 231 ;970 201 ,396 183,880 170,045 1,034,943 56% 53% 46% 42% 39% 47% Lighting and cooking 2, 111 8, 191 6,036 9,359 11,832 37,529 0.5% 1.9% 1.4% 2.1% 2.7% 1.7% Other purposes 19,021 26,016 31,760 31, 157 24,444 132,398 (%) 4% 6% 7% 7% 6% 6% Not used 171 ,000 172,248 190,263 201,984 216,163 951,658 39% 39% 43% 46% 49% 43% Total 441,397 441,612 440, l 32 440,248 440,826 2,204,215 100% 100% 100% 100% 100% 94 Annex 2 Survey Results Table A.2. 14 Number and Percentage of Households Using Kerosene and Candles for Lighting by Electrification Status and Region (All Households) Coastal Region Andean. Region Amazon All North Central South North Central South Without Access to Grid Do not use candles/kero 22,822 30,226 9,850 44,838 133,034 105,537 26,993 373,300 (%) 42% 67% 50% 56% 40% 42% 39% 44% Kerosene only 4,516 932 214 2,367 9,427 ' 15,055 8,849 41,360 (%} 8% 2% 1% 3% 3% 6% 13% 5% Candles only 24,056 12,976 9,306 31,214 172,423 116,437 29,284 395,696 (%) 44% 29% 47% 39% 52% 47% 42% 47% Candles and kerosene 3, 190 1,244 281 17,200 13,532 41 154 6% 3% 1% 2% 5% 5% 6% 5% Total 54,584 45,378 19,651 80,261 332,084 250,561 68,991 851,510 100% 100% 100% 100% 100% 100% 100% 100% Without Access to Grid Do not use candles/kero 3, 173 812 266 5, 101 2,576 10,263 36,576. 58,767 3% 3% 3% 2% 1% 3% 12% 4% Kerosene only 52,251 3,664 734 107,954 54,325 61,432 133,530 413,890 51% 12% 9% 38% 18% 20% 43% 31% Candles only 5,431 14,291 4,984 41,462 97,245 101,146 39,421 303,980 5% 48% 61% 15% 32% 32% 13% 23% Candles and kerosene 40,979 11, 169 2, 152 127,250 148,010 141,621 104,886 576,067 (%) 40% 37% 27% 45% 49% 45% 33% 43% Total 101,834 29,936 8, 136 281;767 302, 156 314,462 314,413 1,352,704 100% 100% 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 95 Special Report Peru: National Survey of Rural Household Energy Use table A.2.1 5 Number and Percentage of Household Using Kerosene and Candles for Lighting by Electrification Status and Expenditure Quintiles (All Households) 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All With Access to Grid [)o not use candles/kero 58,073 65,549 69,816 94,580 373,299 49% 44% 41% 44% 43% 44% Kerosene only 7,209 10, 137 6,661 8,695 8,658 41,360 (%) 6% 7% 4% 4% 4% 5% Candles only 45,408 63,262 84,252 96, 103 106,670 395,695 38% 42% 50% 49% 49% 47% Candles and kerosene 8,221 10,387 8, 180 5,853 8,513 41, 154 (%) 7% 7% 5% 3% 4% 5% Total 118,911 149,335 168,909 195,932 218,421 851,508 (%) 100% 100% 100% 100% 100% 100% Without Access to Grid Electricity Do not use candles/kero 15,071 12,417 12,885 10,775 7,620 58,768 (%) 5% 4% 5% 4% 3% 4% Kerosene only 125, 113 87,628 85,152 64,961 51,035 413,889 39% 30% 31% 27°/o 23% 31% Candles only 73,081 60,223 65,746 54,851 50,077 303,978 23% 21% 24% 23% 23% 23% Candles and kerosene 109,220 132,008 107,440 113,730 113,672 576,070 34% 45% 40% 47% 51% 43% Total 322,485 292,276 271,223 244,317 222,404 1,352,705 (%) 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. 96 ,'·· Annex 2 Results Table A.2. 16 Household Monthly Expenditure on Kerosene for Lighting and Cooking by Electrification Status and Region (Users Only) Coastal Region Andean Amazon All North Central South North Central South Region Regions With Access to Grid 27.19 33.69 32.70 25.20 23.54 19.21 3.45 .23.72 Valid N 1, 188 3,743 3,335 334 14,520 12,901 551 36,571 5.50 8.27 4.79 4.54 5.11 4.33 5.75 4.98 Valid N 7, 125 972 233 4,070 24,447 27,915 12,714 77,476 Lighting and . 8.26 30.00 33.50 30.00 14.83 12.00 18.71 Valid N 580 l 263 139 181 672 Other 5.67 11.86 6.31 4.85 8.35 2.94 6.46 6.50 Valid N 7,746 1,077 235 4,815 22,794 9,315 4,479 50,460 Total 7.22 26.16 29.63 5.81 10.78 7.96 5.86 9.88· Valid N 16,640 6,995 4,065 9,358 63,941 50,803 17,744 169,545 Without Access to Grid 16.00 30.96 33.94 13.09 21.70 Valid N 278 1 11 115 15.60 18.24 16.40 7.55 7.16 6.53 9.66 8.72 Valid N 92,906 12,093 2,505 233,750 193,485 191,304 231,423 957,467 Lighting and 90.00 32.88 25.07 3.30 12.77 27.07 13.43 20.27 Valid N 325 2,740 380 1,454 8,850 11,749 6,993 32,491 Other purposes 16.64 11.50 19.98 8.41 8.62 9.14 8.29 8.84 Valid N 889 793 314 14,096 9,921 30,931 24,995 81,938 Total 15.87 21.43 22.94 7.58 7.67 8.04 9.63 9.20 Valid N 94,397 17,206 4,655 249,300 213,971 240,070 263,411 1,083,012 Source: Authors' calculations, 2005. 97 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.17 Household Monthly Expenditure on Kerosene for Lighting and Cooking by Region (Weighted-Users Only) Coastal Region Andean Amazon All North Central South North Central South All Areas (Electrified and Unelectrified} 25.07 32.88 33.18 25.20 24.64 17.25 3.45 23.25 Valid N 1,467 5,323 4,790 334 16,235 18,986 551 47,686 Lighting 14.88 17.50 15.41 7.50 6.93 6.25 9.46 8.44 Valid N 100,031 13,066 2,738 237,820 217,932 219,219 244, 137 1,034,943 37.58 32.00 28.52 5.63 13.18 26.25 13.43 20.06 Valid N 905 3,943 643 1,593 11,031 12,420 6,993 37,529 14.13 . 7.51 8.43 7.71 8.01 7.95 549 18, 911 32,715 40,247 29,474 Total 14.57 22.80 26.06 7.51 8.39 8.02 9.39 9.30 Valid N 111,037 24,201 8,721.00 258,658 277,913 290,873 281, 155 1,252,557 Source: Authors' calculations, 2005. Table A.2.18 Household Monthly Expenditures on Candles for Lighting by Electrification Status and Region (Users Only) Coastal Region Andean Region Amazon All North Central South North Central South Region Regions With Access to Grid Electricity Candle 0.91 1.62 1.66 1.18 1.37 1.26 1.49 1.32 Valid N 27,247 14,220 9,587 33,055 189,623 129,969 33, 149 436,850 Without Access to Grid Candle 3.81 7.42 8.14 3.32 4.99 4.68 4.17 4.49 Valid N 46,411 25,460 . 7, 136 168,713 245,255 242,767 144,307 880,048 Source: Authors' calculations, 2005. 98 Annex 2 Results Table A.2. 19 Household Monthly Expenditures on Kerosene for Lighting and Cooking by Electrification Status and Expenditure Quintiles (Users Only) 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All With Access to Grid 10.37 16.19 11.83 23.19 31.34 23.72 1,361 2,374 6,336 11,131 15,370 36,571 3.66 3.55 6.08 5.77 6.64 4.98 15,430 20,478 13,440 13,480 14,648 77,476 7.50 16.35 14.54 22.00 18.71 47 1,401 5,038 3.08 3.80 3.99 8.55 9.35 6.50 Population 5,071 8,831 9,725 13,268 13,565 50,460 Total 3.94 4.57 7.07 .94 16.51 9.88 Population 21,863 31,729 30,902 38,947 46, 105 169,545 Without Access to Grid Cooking 12.00 12.75 15.96 18.47 36.33 21.70 252 813 4,341 2,737 2,972 11,115 4.93 6.67 8.68 1 l.17 14.51 8.72 232,222 211 ;492 187,956 170,400 155,397 957,467 12.21 8.82 18.31 14.28 38.43 20.27 Population 2, 111 8, 144 4,635 8,291 9,310 32,491 Other 4.50 8.21 8.86 9.54 14.23 8.84 Population 13,950 l 7, 185 22,035 17,889 l 0,879 81,938 Total 4.97 6.88 9.04 11.25 16.10 9.20 Population 248,535 237,635 218,967 199,317 178,558 1,083,012 Source: Authors' calculations, 2005. 99 Special Report Peru: Nalional Survey of Rural Household Energy Use Table A.2.20 Household Monthly Expenditures on Kerosene for Lighting and Cooking by Expenditure Quintiles, in Soles (Users Only) 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All ·All Areas (Electrified and Unelectrified) Cooking 10.62 15.31 13.51 22.26 32.15 23.25 1,613 3, 187 10,677 13,868 18,342 47,686 4.85 6.40 8.51 10.77 13.83 8.44 Population . 247,652 231,970 201,396 183,880 170,045 1,034,943 Lighting & Cooking 12.21 8.81 17.86 14.31 34.92 20.06 Population 2, 111 8, 191 6,036 9,359 11,832 37,529 Other Purposes 4.12 6.71 7.37 9.12 11.52 7.95 Population 19,021 26,016 31,760 31, l 57 24,444 132,398 Total 4.89 6:61 8.80 11.36 16.19 9.30 Population 270,397 269,364 249,869 238,264 224,663 1,252,557 Source: Authors' calculations, 2005. Table A.2.21 Household Monthly Expenditures on Candles for Lighting by Electrification Status and Expenditure Quintiles, in Soles (Users Only) 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All With Access to Grid Candle 1.18 1.10 1.31 1.28 l.56 1.32 ValidN 53,630 73,649 92,432 101,956 115, 184 436,850 Without Access to Grid 3.15 3.66 4.86 5.08 5.93 4.49 Valid N 182,301 192,231 173,186 168,581 163,749 880,048 Source: Authors' calculations, 2005. 100 Annex 2 Survey Resull5 Table A.2.22 Household Monthly Expenditures on Kerosene and Candles for lighting by Electrification Status and Region, in Soles (Users Only) Coastal Region Andean Region Amazon All North Central South North Central South Region Regions With Access to Grid Kerosene only 2.56 3.86 8.36 3.74 5.45 2.73 4.65 3.86 Valid N 4,516 932 214 2,367 9,427 15,055 8,849 41,360 Candles only 0.94 1.57 1.54 1.17 1.41 1.26 1.19 1.31 Valid N 24,056 12,976 9,306 31,214 172,423 116,437 29,284 395,696 Both candles and kerosene 5.87 8.19 14.44 6.55 4.79 4.64 8.10 5.38 Valid N 3, 190 1,244 281 1,842 . 17,200 13,532 3,865 41, l 54 Total l.66 2.25 2.06 l.62 1.89 1.73 2.56 l.88 Valid N 31,762 15, 152 9,801 35,423 199,050 145,024 41,998 478,210 Without Access to Grid Kerosene only 11.16 13.55 22.25 6.55 5.74 5.65 7.60• 7.32 Valid N 52,251 3,664 734 107,954 54,325 61,432 133,530 413,889 Candles only 6.26 8.21 9.13 4.68 6.74 6.42 8.21 6.64 Valid N 5,431 14,291 4,984 41,462 97,245 101, 146 39,421 303,980 Both candles and ke·rosene 21.75 24.39 17.78 10.49 10.82 9.94 11.93 11.80 Valid N 40,979 11, 169 2, 152 127,250 148,010 141,621 104,886 576,069 Total (All) 15.29 15.08 12.72 8.08 8.57 7.90 9.32 9.15 Valid N 98,662 29, 124 7,869 276,667 299,580 304, 199 277,836 1,293,938 Source: Authors' calculations, 2005. Table A.2.23 Household Monthly Expenditures on Kerosene and Candles for Lighting by Region (Users Only} Coastal Region Andean Region Amazon All North Central South North Central South All Areas (Electrified and Unelectrified) Kerosene 10.48 11.58 19 .11 6.49 5.70 5.07 7.41 7.00 Valid N 56,767 4,596 948 110,322 63,752 76,487 142,378 . 455,249 Candles 1.92 5.05 4.19 3.17 3.33 3.66 5.22 3.63 Valid N 29,488 27,267 14,289 72,676 269,668 217,583 68,704 699,676 Both candles and kerosene 20.60 22.77 17.40 10.43 10.19 9.48 11.79 11.37 Valid N 44, l 69 12,413 2,433 129,092 165,210 155,153 108,751 617,223 Total 11.97 10.69 6.81 7.35 5.91 5.91 8.43 7.19 ValidN 130,424 44,276 17,671 312,090 498,630 449,223 319,834 1,772, 148 Source: Authors' calculations, 2005. 101 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.24 Household Monthly Expenditures on Kerosene and Candles for Lighting by Electrification Status and Expenditure Quintiles {Users Only) 113.26- 201.01- 321.1.4- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All With Access to Grid Kerosene 4.55 6.10 8.04 10.20 11.33 7.32 Valid N 125,113 87,628 85, 152 64,961 51,035 413,889 Candles 3.95 5.81 6.85 8.07 9.74 6.64 ValidN 73,081 60,223 65,746 54,851 50,077 303,980 Both candles and kerosene 7.57 8.85 11.49 13.51 17.86 11.80 Valid N 109,220 132,008 107,440 113,730 113,672 576,069 Total 5.48 7.34 9.17 11 .31 14.42 9.15 Valid N 307,414 279,859 258,338 233,542 214,784 1,293,938 Without Access to Grid Kerosene 3.64 2.54 3.75 5.34 4.16 3.86 Valid N 7,209 10, l 37 6,661 8,695 8,658 41,360 Candles 1.24 1.12 1.20 1.26 1.58 1.31 Valid N 45,408 63,262 84,252 96, l 03 106,670 395,696 Both candles and kerosene 3.87 4.42 6.10 5.06 7.53 5.38 Valid N 8,221 10,387 8, 180 5,853 8,513 41, 154 Total (All) 1.88 1.70 1.78 1.78 2.17 1.88 Valid N 60,839 83,787 99,093 110,651 123,841 478,210 Source: Authors' calculations, 2005. Table A.2.25 Household Monthly Expenditures on Kerosene and Candles for lighting by Expenditure Quintiles 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month S/month S/month S/month S/month All All Areas (Electrified and Unelectrified) Kerosene 4.50 5.73 7.73 9.63 10.29 7.00 ValidN 132,323 97,765 91,813 73,656 59,693 455,249 Candles only 2.91 3.41 3.68 3.74 4.19 3.63 Valid N 118,490 123,486 149,998 150,954 156,748 699,676 Both candles and kerosene 7.31 8.53 11. 11 13.09 17.14 11.37 ValidN . 117,441 142,395 115,620 119,582 122, 185 617,223 Total 4.89 6.04 7.12 8.25 9.94 7.19 Valid N 368,253 363,646 357,431 344, 192 338,625 1,772, 148 Source: Authors' calculolions, 2005. 102 Annex 2 Results Table A.2.26 Household Monthly Expenditures on Lighting and Electricity by Electrification Status and Region (Weighted- Users Only) Coastal Andean Region Amazon All North Central South North Central South Region Regions With Access to Grid Candles 0.91 1.62 1.66 1.18 1.37 1.26 1.49 l.32 Valid N 27,247 14,220 9,587 33,055 189,623 129,969 33,149 436,850 Kerosene (Light only) 3.62 5.12 8.52 4.39 4.40 3.04 4.55 3.92 Valid N 7,706 2, 175 496 4,209 26,628 28,587 12,714 82,514 LPG l.05 2.00 20.67 18.26 Valid N 100 47 1 1 180 Small generator 28.50 28.50 Valid N 158 158 Dry cell battery 2.94 4.62 4.03 3.47 3.88 3.36 4.11 3.68 Valid N 24,898 15,530 8,272 39,849 172, 192 153, 117 52,795 466,652 Car battery 3.52 12.00 10.56 3.24 6.00 3.93 8.13 5.72 Valid N 149 81 651 807 516 . l 0 812 19.82 27.04 24.66 10.87 13.36 9.44 16.03 13.63 Valid N 51,328 43,751 18,739 78,070 327,738 229,695 63,624 812,945 All Expend (Lighting and electricity) 21.40. 28.59 26.95 13.25 16.46 11.83 19.70 16.26 Valid N 53,461 45, 1l8 19,390 79,041 331,055 248,678 68,781 845,522 Without Access to Grid Electricity Candles 3.81 7.42 8.14 3.32 4.99 4.68 4.17 4.49 Valid N 46,411 25,460 7, 136 168,713 245,255 242,767 144,307 880,048 Kerosene (Light only) 14.28 16.89 14.56 7.12 6.64 6.24 8.33 7.98 Valid N 93,230 14,833 2,886 235,205 202,335 203,053 238,416 989,958 LPG 12.67 13.07 18.00 16.24 ValidN 437 154 1t162 1,753 Small 41.37 37.20 20.28 33.82 33.31 Valid N 1,431 887 1,278 3, 195 6,791 Dry cell battery 4.93 7.37 6.80 4.48 6.50 5.01 8.04 6.04 Valid N 86,036 23, 153 6,929 242,973 245,259 262,790 297,041 l, 164, 180 Car 6.19 12.57 8.24 5.59 5.76 6.06 6.74 6.61 Valid N 47,556 3,027 32,730 46,295 35,060 55, 140 236,683 Valid N All (Lighting and 22.65 28.76 21.16 12.50 14.74 12.85 17.57 15.44 electricity) Valid N 101,149 29,804 8, 135 280,593 300,612 309,684 310,512 1,340,491 Source: Authors' calculations, 2005. 103 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.27 Household Monthly Expenditures on lighting and Electricity by Region, in Soles (Users Only) Coastal Region Andean Region Amazon All North Central South North Central South All Areas (Electrified and Unelectrified) Candles 2.74 5.34 4.43 2.97 3.41 3.49 3.67 3.43 Valid N 73,657 39,680 16,723 201,768 434,878 372,737 177,456 1,316,898 Kerosene (Light only) 13.47 15.38 13.68 7.08 6.38 5.85 8.14 7.67 ValidN 100,936 17,009 3,381 239,414 228,963 231,640 251, 130 1,072,472 LPG (Light only) 12.67 8.32 2.00 20.67 18.00 17.05 ValidN 437 254 47 1,032 1, 162 2,933 Small generator 41.37 37.20 20.28 33.57 33.20 Valid N 1,431 887 1,278 3,353 6,949 4.48 6.26 5.29 4.34 5.42 4.40 •7.45 5.36 110,934 38,683 15,200 282,821 417,450 415,907 349,836 1,630,832 Car 6.18 12.57 8.65 5.53 5.76 5.95 6.76 6.60 Valid N 47,704 16,956 3,678 33,537 46,811 36,970 55,952 241,608 19.82 27.04 24.66 10.87 13.36 9.44 16.03 13.63 Valid N 51 ,328 43,751 18,739 78,070 327,738 229,695 63,624 812,945 All (Lighting and electricity) 22.22 28.66 25.24 12.66 15.64 12.39 17.96 15.76 Valid N 154,610 74,922 27,525 359,633 631,667 558,362 379,293 2, 186,013 Source: Authors' calculations, 2005. 104 Annex 2 Results . Table A.2.28 Household Monthly Expenditures on Lighting and Electricity by Electrification Status and Expenditure Quintiles, in Soles 113.26- 201.01- 321.14- <113.25 201.00 321.13 533.22 >533.22 S/month $/month S/month S/month S/month All With Access to Grid Candles l.18 1.10 1.31 1.28 1.56 1.32 Valid N 53,630 73,649 92,432 101,956 115, 184 436,850 Kerosene 3.31 3.01 3.70 4.61 5.17 . 3.92 Valid N 15,430 20,525 14,841 14,547 l 7, 171 82,514 LPG 35.00 5.97 1.05 18.26 Valid N 516 563 100 1, 180 Small 28.50 28.50 Valid N 158 158 Dry cell battery 2.66 3.12 3.66 3.85 4.39 3.68 Valid N 55,887 87,221 96, 118 100,864 126,562 466,652 Car battery 3.00 6.00 8.00 4.46 6.89· 5.72 Valid N 672 516 504 1,329 1,904 4,925 Electricity (Grid) 7.36 8.54 10.38 14.20 22.52 13.63 Valid N 113,534 143,193 160, 146 187,719 208,353 812,945 All 9.38 11.03 13.11 . 16.82 25.55 16.26 Valid N 117,418 148,819 168,760 194, 164 216,360 845,522 Without Access to Grid Candles 3.15 3.66 4.86 5.08 5.93 4.49 Valid N 182,301 192,231 173, l 86 168,581 163,749 880,048 Kerosene 4.74 6.14 7.93 9.99 12.90 7.98 Valid N 234,333 219,636 192,592 178,691 164,707 989,958 LPG 12.75 16.95 16.24 Valid N 298 1 Small 13.00 29.02 38.14 33.31 Valid N 775 1,466 4,550 6,791 3.58 4.95 6.02 6.97 9.06 6.04 224,374 264,873 242,381 223,858 208,694 1, 164, 180 Car 5.28 5.20 6.13 6.83 7;49 6.61 Valid N 17,679 30,359 39,556 68,436 80,653 236,683 Valid N All (Lighting and 8.15 12.14 15.17 19.40 26.14 15.44 Valid N 316,738 290,086 221,481 Source: Authors' calculations, 2005. 105 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.29 Household Monthly Expenditures on Lighting and Electricity by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Quintile> S/month S/month S/month S/month S/month All All Areas (Electrified and Unelectrified) Candles 2.71 2.95 3.62 3.64 4.13 3.43 Valid N 235,931 265,881 265,618 270,536 278,933 1,316,898 Kerosene (Light only) 4.65 5.88 7.63 9.59 12.17 7.67 ValidN 249,763 240, 160 207,433 193,238 181,877 1,072,472 LPG 26.85 5.97 15.93 17.05 Valid N 814 563 1,556 2,933 Small 13.00 28.97 38.14 33.20 Valid N 775 l ,623 4,550 6,949 3.40 4.50 5.35 6.00 7.30 5.36 280,261 352,093 338,499 324,723 335,256 l,630,832 Car battery 5.19 5.22 6.16 6.78 7.48 6.60 Valid N 18,351 30,875 40,060 69,765 82,557 241,608 Electricity (Grid) 7.36 8.54 10.38 14.20 22.52 l 3.6:J Valid N 113.,534 143, 193 160,146 187,719 208,353 812,945 All (Lighting and 8.48 11.76 14.38 18.25 25.85 15.76 Valid N 434, 156 438,905 437,990 437, 120 437,841 2, 186,013 Source: Authors' calculations, 2005. Table A.2.30 Percentage of Households with and without Access to Grid Electricity by Region Coastal Region Andean Region Amazon All North Central South North Central South Region Regions With access to grid 35% 60% 71% 22% 52% 44% 18% 39% Without access to 65% 40% 29% 78% 48% 56% 82% 61% All Households 156,418 75,315 27,786 362,029 634,240 565,023 383,403 2,204,214 100% 100% 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005·. 106 Annex 2 Results Table A.2.31 Percentage of Households with and without Access to Grid Eleetricity by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All With access to grid 27% 34% 38% 45% 50% 39% Without access to 73% 66% 62% 56% 51% 61% All Households . 441,398 441,612 440, 132 440,247 440,827 2,204,216 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. Table A.2.32 . Household Electricity Consumption, Expenditure in Soles, Effective Price per kWh, and Electricity Used for Lighting by Region Coastal Region Andean Region Amazon All North Central South North Central South kWh 38.30 61.73 59.06 21.68 26.87 16.66 31.56 27.19 ValidN 51,328 43,751 18,699 78,070 327,738 229,695 63,624 812,904 month 19.82 27.04 24.66 10.87 13.36 9.44 16.03 13.63 Valid N 51 78,070 229,695 63,624 812;945 Effective price kWh 0.57 0.49. 0.47 0.60 0.62 0.83 0.71 0.67 ValidN 51,328 43,751 18,699 78,070 327,738 229,695 63,624 812,904 % of electricity used for 28.00 24.01 24.16 43.71 41.07 54.61 38.55 42.87 Valid N 50, 109 42,503 17,731 74,548 316, 196 224,429 57,883 783,398 kWh for lighting per month 6.99 10.38 9.32 6.44 7.70 5.82 6.92 7.10 51,328 43,751 18,498 76,739 319,949 229,695 61,330 812,904 Source: Authors' calculations, 2005. 107 Peru: National of Rural Household Use Table A.2.33 Household Electricity Consumption, Expenditure Effective Price per kWh and Electricity Used for lighting by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month All kWh 14.64 19.96 28.66 48.51 27.19 l 193 1 146 month 7.36 8.54 10.38 22.52 Valid N 11 193 l 146 Effective price kWh 0.83 0.76 0.69 0.62 0.55 0.67 Valid N 11 193 1 146 187,719 kWh for lighting month 4.37 5.28 6.46 10.06 7.10 Valid N 11 146 l 160,872 128 81 Source: Authors' calculations, 2005. Table A.2.34 Type and Number of Electric Lights Owned by Household by Expenditure Quintiles (All Households with Grid Connection) 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All Incandescent lamp 2.1 2.0 2.2 2.0 1.9 2.0 Valid N 118,912 149,336 168,909 195,931 218,422 851,510 Fluorescent tube 0.4 0.5 0.5 0.9 1.4 ·a.a Valid N 118,912 149,336 168,909 195,931 218,422 851,510 Compact . fluorescent lamp 0.3 0.4 0.5 0.6 1.1 0.6 ValidN 118,912 149,336 168,909 195,931 218,422 851,510 All electric 2.7 2.9 3.2 3.5 4.5 3.5 ValidN 118,912 149,336 168,909 195,931 218,422 851,510 Source: Authors' calculations, 2005. 108 Annex 2 Results Table A.2.35 Type and Number of Electric lights Owned by Region (Users Only) Coastal Region Andean Region Amazon All North Central South North Central South Incandescent lamp 2.64 2.61 2.90 2.50 2.74 2.60 2.25 2.64 Valid N 37,351 25,755 13,885 56,807 . 259, 160 227,446 36, l 80 656,584 Fluorescent tube 2.21 2.73 2.61 2.24 2.26 1.86 2.34 2.24 Valid N 29, l 00 25,266 6,838 31,099 135,123 50,698 30, 182 308,307 Compact fluorescent 2.31 2.74 2.88 2.38 2.05 . 2.03 2.21 2.22 Valid N 23,354 22,686 6,911 28,868 99,824 31,433 34,989 248,065 All electric 3.97 4.38 3.97 3.49 3.67 3.00 3.35 3.50 Valid N 54,584 45,293 19,651 80,261 332,084 250, 167 68,351 850,3?3 Source: Authors' calculations, 2005. Table A.2.36 Type and Number of Electric lights Owned by Expenditure Quintiles (Users Only) 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All Incandescent lamp 2.5 2.5 2.7 2.6 2.9 2.6 ValidN 100, 145 118,748 141,047 149,687 146,958 656,584 Fluorescent tube 1.8 1.8 1.9 2.2 2.7 2.2 ValidN 24,438 40,263 46,540 79,254 11 Compact fluorescent lamp 1.5 2.0 2.1 2.0 2.6 2.2 ValidN 21,762 29,711 39,314 61,583 95,695 248,065 All Electric 2.7 2.9 3.3 3.5 4.5 3.5 Valid N 118,912 149,336 167,877 195,931 218,337 850,393 Source: Authors' calculations, 2005. Table A.2.37 Electridty Usage for lighting by lifeline Rate kWh Used/mo for Lamp % of Electricity Used for Average Effective Price per Usage. per Month Lighting Lighting kWh (in Sf month) <=30 kWh/mo 5.9 53% 0.76 Valid N 555,784 555,784 578,065 10.3 19% 0.46 Valid N 3 3 234,840 All Levels of 7.2 43% 0.67 Valid N 783,398 783,398 812,904 Source: Authors' calculations, 2005. 109 Special Report Peru: National Survey of Rural. Household Energy Use Table A.2.38 Type and Number of Electric Lights Owned by Lifeline Rate All Electric Lamp Usage per Month Incandescent Fluorescent Compact Fluorescent Lighting <=30 kWh/mo 2.6 1.9 2.0 3.1 Valid N. 469, 108 163,054 145,713 577,587 > 30 kWh/mo 2.9 2.6 2.6 4.6 Valid N 137,738 8 All Levels of Usage 2.7 2.2 2.2 3.5 Valid N 629,352 300,792 238,331 812,427 Source: Authors' calculations, 2005. Table A.2.39 Television Ownership by Region Coastal Region Andean Region Amazon All North Central South North Central South Region Regions NolV . 13% 7"/o 8% 40% 38% 42% 36% 35% B&WlV 28% 25% 29% 34% 32% 38% 21% 32% Color lV 55% 63% 54% 22% 26% 16% 38% 28% Color and B&W lV 4% 5% 9% 4% 5% 5% 5% 5% All Households 54,585 45,377 19,652 80,261 332,084 250,561 68,990 851,510 Note: Television refers to plug-in television. Table A.2.40 Television Ownership by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 E~nditure Quintile> S/month S/month S/month S/month S/month All NolV 75% 50% 37% 22% 13% 35% B&WlV 17"/o 38% 42% 38% 23% 32% Color lV only 7% 9% 17% 33% 57"/o 28% Color.and B&W lV 1% 3% 4% 7% 8% 5% All 118,911 149,335 168,910 218,420 851,507 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. Note: Television refers to plug-in television. 110 Annex 2 Results Table A.2.41 Plug-in Radio and Television Ownership by Region Coastal Region Andean Region Amazon All North Central South North Central South No Radio/TV 8% 4% 2% 24% 11% 21% 21% 15% Radio only 5% 4% 6% 17% 27% 21% 15% 20% TV 46% 28% 19% 23% 14% 15% 27% 19% Radio and TV '42% 65% 73% 37% 49% 43% 37% 46% All Households 54,583 45,377 19,651 80,260 332,085 250,561 68,990 851,507 Source: Authors' calculations, 2005. Note: Television refers to plug-in television. Table A.2.42 Plug-in Radio and Television Ownership by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 S/month S/month S/month S/month S/month All No 39% 21% 13% 8% 5% 15% 36% 29% 25% 14% 8% 20% 10% 13% 16% 22% 28% 19% Radio and TV 15% 37% 47% 56% 59% 46% All Households 118,911 149,336 168,909 195,931 218,422 851,509 100% 100% 100% 100% 100% 100% Source: Authors' calculations, 2005. Note: Television refers to plug-in television. 111 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.43 Electric Appliance Ownership by Region Coastal Andean Region Amazon All North Central South North Central South Region Regions Iron 52% 60% 46% 21% 22% 14% 33% 25% Fan 12% 11% 5% 0.8% 1.2% 0.1% 8% 3% Refrigerator 29% 41% 33% 6% 7% 4% 19% 11% Video/DVD 22% 21% 23% 5% 9% 8% 15% 11% ·Microwave 2.4% 2.6% 1.9% 1.1% 0.9% 0.1% 0.6% 0.9% Stove 0.5% 1.0% 0.8% 0.2% 0.3% 0.1% l.9% l.8% 1.4% 0.2% 0.5% 1.3% 0.6% 1.9% 2.1% 1.7% - 0.3% 0.3% Electric motor 0.8% 1.0% 0.2% 1.7% 0.8% 0.4% 0.7% 0.8% machine 2.0% 1.0% 0.9% 0.7% 0.4% 0.7% 0.6% Electric drill 1.1% 0.2% 0.9% 0.4% 0.1% 0.4% 0.5% 0.4% 0.2% 0.3% 0.2% 0.2% 0.1% 0.4% 0.7% 0.3% 0.1% 0.3% 0.2% All Households 54,584 45,377 19,652 80;260 332,084 250,561 68,990 851,508 Source: Authors' calculations, 2005. Table A.2.44 Electric Appliance Ownership by Expenditure Quintiles 1. Poorest 2 3 4 5. Richest <113 113-201 201-321 321-533 >533 Expenditure Quintile> S/month S/month S/month S/month S/month All Iron 5% 12% 23% 39% 58% 25% Fan 0% 0% 1% 3% 10% 3% 0.9% 3% 5% 17% 37% 11% Video/DVD 3% 6% 8% 14%· 28% 11% Microwave 0.1% 0.4% 0.4% 4.3% 0.9% Stove 0.1% 0.3% 0.4% 0.1% Washing machine 0.0% 0.3% 0.0% 3.1% 0.6% . Dom water pump 0.1% 0.3% 1.5% 0.3% Electric motor 0.5% 0.6% 0.7% 0.5% 1.8% 0.8% machine 0.9% 0.0% 0.8% 1.7% 0.6% Electric drill 0.1% 0.1% 0.3% 1.7% 0.4% Electric saw 0.1% 0.0% 0.4% 0.1% Irrigation pump 0.1% 0.1% 0.0% 0.9% 0.2% 0.2% All Households 193, l 17 204,630 164,044 150,843 138,876 851,510 Source: Authors' calculations, 2005. 112 Annex2 Results Table A.2.45 Households with Photovoltaic (PY) Systems Panel Used Monthly Year Annual Departamenlo Grid Access LastMonlh Panel is:lll Rental System Cost Installed Maint. Cost. Size [Soles] [Soles] [Soles] !Watts] Yes Tumbes Yes Tumbes Yes Piura Yes 2500 2001 La Libertad la Yes owned 1400 2003 lea Yes Lima Yes 2004 Yes Yes Yes Yes owned 550 1986 34 Yes Cajamarca Yes 5000 2004 Yes owned 900 2004 85 Cajamarca Yes rented 18 2005 500 12 Yes Huanuco Pasco Yes Yes Yes Cusco Yes owned 2500 2002 Pu no Yes owned 650 2002 75 Pu no Yes 1200 2000 Pu no Yes owned 1500 2004 100 Pu no Yes owned 2800 1982 Pu no Yes owned 1800 1992 12 San Martin Yes Loreto Yes Yes owned 704 2004 77 Loreto Yes Cajamarca Amazon as Yes owned 1300 2000 Amazonas Yes owned 8000 1998 70 Amazonas Yes 500 2005 75 Amazonas Loreto Yes 120 1999 145 Loreto Yes owned 85 2001 135 .600121 Loreto Yes 2002 60 600121 Loreto Yes owned 230 2004 .10 600121 Yes 600 2005 75 Source: Authors' calculations, 2005. Notes: (1) No systems were reported as leased. Only two systems are reported as given. Whether the inference is that systems reported as owned were in fact bought {by the reporting householders) is not clear. (2) It seems likely that there is a data entry error for those households reporting 600 Watt systems, given cost estimates of 85 to 230 soles. The question was asked as "panel cost." Whether this was interpreted by the enumerators as total system cost is not clear. 113 Special Report Peru: National Survey of Rural Household Energy Use Table A.2.46 Small Generator Users, Cost Data Owned Small Maintenance and Generator Cost Diesel Fuel Cost Gasoline Fuel Cost Repair Cost Soles Soles/M~nth Soles/Month Soles/Month Coastal North 17 78 12 Coastal Central 1,716 44 59 Coastal South Andean North Andean Central Andean South 928 18 3 · Amazon 2, 140 25 60 All 29 53 7 Number of Households Actually Sampled Coastal North 7 2 5 4 Coastal Central 7 7 2 0 Coastal South 0 0 0 0 Andean North 0 0 0 0 Andean Central 0 0 0 0 Andean South 3 0 3 Amazon 6 3 3 0 All 23 12 13 5 Source: Authors' calculations, 2005. (1 l Average cost of gasoline generators = 1703 soles. 114 Annex3 Esti111ating the Benefits .of Rural Electrification Many issues arise in ·the estimation of demand curves, Figure A.3.1 and in the use of changes in consumer surplus to measure benefits. These are discussed in this Annex, and the Demand for Lighting methodology and results are compared with other studies. price The Demand Curve Assumptions about the shape of the demand curve are critical when only few points are available. Many of the early studies of the benefits of rural electrification recognized ·that demand curves were likely concave (as drawn in Figure A.3.1), but then used a linear curve anyWay.33 Unfortunately such an assumption will lead to an overestimation of the area C, and of the net benefits of electrification, because the empirical evidence is that the demand curve is much more likely to have a concave D E shape of the type shown in Figure A.3.1. Recognizing this problem, some studies (e.g., the World Bank's Solar Homes o.,,. project in Bolivia) take_ the area C as one third of the area service level determined by a linear demand curve. Source: INEI, 2005. Others have made the assumption of some particular functional form, which has the advantage of calculation This has the merit of simplicity, but some shortcomings of the area under the demand curve by definite integrals. remain. The implication of such a curve is that at zero One commonly used approach is to use a curve of constant price, the quantity consumed would be infinitely large. elasticity /J: Yet, in the face of very dramatic electricity price decreases (as are achieved for example by grid-connection vis-a-vis Equation A.3.1 car battery use), in the short run, consumption will be constrained by the stock of appliances required to actually For which the corresponding area C (i.e., between QKER0 use larger quantities of electricity.35 and QE) can be calculated from the corresponding definite This approach of estimating changes in welfare by integral. 34 consumer.surplus has a number of issues and limitations 33 For example, a linear demand curve was used in the Philippines study of rural electrification benefits (Barnes, D. F., A. Domdom, V. Peskin, and H. Peskin, Rural Electrification and Development in the Philippines: Measuring the Social and Economic Benefits. ESMAP formal Report 255/02. Washington, DC: World Bank, 2002). 34 For details, see P. Meier, Economic Analysis of Solar Home Systems: A Case Study for the Philippines, World Bank, 2003. 35 P. Choynowksi (Measuring Willingness to Pay for Electricity, Asian Development Bank, Economics Research Department Technical Note Series #3, July 2002) has therefore proposed an alternative functional form of the general specification lnq = a + Pp . This has the advantage that the upper bound of electricity demand (when the price is zero) is given by e•, which captures the fact that consumption is bounded by the stock of electrical appliances. The price elasticity is given by Pp, which varies with price: the greater the price, the greater the elasticity of demand. The implication is that at very high prices (typical of the equivalent price of kerosene), demand is more elastic than at low prices (typical of the price of grid-electricity). 115 Special Report Peru: National Survey of Rural Household Energy Use that are rarely acknowledged in many studies of rural for those who state 100 percent of household kerosene electrification benefits. Therefore, this approach needs consumption is for lighting), the calculations presented in to be applied with some caution. 36 One must recognize this box use the typical average consumption rates cited in that the demand curve shifts outward with increases Jones et al (2005) (0.01 liter per hour for a simple wick lamp, in income (for a so-called normal good for a given price, 0.03 liters per hour for a hurricane lamp, and 0.07 liters per higher income implies a greater demand). However in the hour for a petromax lamp). This approach also provides case of an inferior good-of which radio listening is a good an estimate of kerosene consumption that is intrinsically example-consumption decreases with increase of income. consistent with estimates of lumen-hours, which is vital for the willingness-to-pay calculations. Indeed, as shown in Table A.3.1, estimates of cost/kLmh based on average Data Issues consumption of lamp-types have much lower variance than based on user estimates. Estimating the points required to establish a demand curve for some service such as lighting is not straightforward. In the Peru survey, respondents were asked to estimate the proportion of kerosene used for the various uses (such as Estimates of Willingness· to Pay lighting, and cooking), which was cross-checked against Table A.3.2 shows the benefit calculations for each estimates derived from information collected about the expenditure quintile. As described in Section 5 of this hours of use and type of kerosene lamps. As shown in report, QKERO refers to quantity of kerosene consumed by Figure A.3.2, these estimates show very different results. unelectrified households, PKERO is the price of kerosene, Because it is likely that user estimates of the fraction QE is the quantity of electricity consumed (by electrified of kerosene devoted to lighting are unreliable (except households), and PE is the price of electricity. The table Figure A.3.2 Comparison of Estimates Used for Kerosene lighting .... :::J 40 0 ~ (/) 0 ~ 0 c: 0 20 0 a E :::J (/) 0 0 c: 0 0 (I) c: (I) 30 @o (/) E (I) 0 00 0 .>t!. (I) CJ) f.!! ~ 10 :E' 0 c: 0 0 ~ 0 ~ 10 30 40 litres/month, based on percentage use estimate Source: INEI, 2005. 36 The idea of measuring changes in consumer surplus by the area under the (uncompensated) demand curve is attributed to Marshall in 1890. Thus, changes in utility are measured by a monetary amount. But the impact of electrification is to dramatically decrease prices for lighting, 'IV viewing, and other services previously provided at high cost by electricity substitutes-in some cases, by an order of magnitude. Hence, as price falls, the consumer's real income (though not monetary income) rises. In other words, one should measure the area under a constant real income demand curve, not a constant monetary income demand curve. 116 ' ." ~: Annex 3 Estimating the Benefits of Rural Electrification displays the values (i.e., B, C, D, and E) that correspond to higher price for kerosene lighting, and therefore benefit the different areas under the lighting demand curve shown from electrification proportionately more than the upper in Figure A.3.1. The table then sums these values to obtain quintiles. However, the total value of benefits increases with the estimated willingness to pay for each quintile. increasing expenditure. The calculations show that the willingness of The willingness-to-pay results for Peru are consistent unelectrified households to pay for grid electricity with those obtained in other countries (Table A.3.3). Average ranges from 24 to 38 soles per month (B + _C + D + E), WTP in Peru is US$8 per household per.month, compared depending on the expenditure quintile. The net benefit with US$11 in Bolivia and US$12 in Laos, but significantly (after subtracting existing benefits from kerosene lamps) lower than the Philippines (US$38/household/month). This ranges from 17 to 30 soles per month. Average willingness is due to the use of a linear demand curve in the Philippines to pay per kilowatt-hour ranges from 3.9 to 5.0 soles/kWh calculations. When a constant elasticity demand curve is (or US$1.23 to 1.54/kWh). The WTP/kWh decreases with used, WTP in the Philippines falls to US$7/household/ increasing expenditure because the poor pay a much month. Table A.3.1 Statistical Comparison of User Estimates of Kerosene Consumption versus Average Lamp Consumption Based on User Estimates of Proportion of Based on Average Hourly Consumption Kerosene Used for Lighting, (Soles/klmh) of Lamp-Types (Soles/klmh) Mean 0.93 0.73 Standard deviation 4.39 1.72 Coefficient of variation 4.69 2.35 Source: INEI, 2005. Table A.3.2 Assumptions and Results, Willingness to Pay for Lighting per Month, by Quintile Unit 1 (Poorest) 2 3 4 5 (Richest) Assumptions QKERo [wick-lamp] klmh 0.8 1.1 1.1 1.2 1.7 QE klmh 111.9 129.5 141.9 205.6 323.5 PKERO [wick-lamp] S/klmh 3.0 2.9 2.8 2.8 2.7 PE S/klmh 0.061 0.053 0.048 0.034 0.026 Results Elasticity [] -1.3 -1.2 -1.2 -1.2 -1.1 Areas: B s 2.5 3.1 3.2 3.3 4.7 c s 14.5 16.2 16.4 18.6 25.0 D s 0.1 0.1 0.1 0.0 0.0 E s 6.8 6.9 6.8 7.1 8.3 Total WTP s 23.9 26.2 26.4 29.0 38.0 Net Benefit s 17.1 19.3 19.6 21.9 29.7 Average kWh kWh 4.8 5.6 6.5 7.4 9.6 Average WTP/kWh S/kWh 5.0 4.7 4.1 3.9 4.0 US$/kWh 1.54 1.46 1.26 1.21 1.23 Source: INEI, 2005. 117 Peru: National Rural Household Use Table A.3.3 Cross-Country Comparisons of WTP Calculations Unit Peru Bolivia Laos QKERO klmh/month l.13 7 4.1 20 QE klmh/month 142 90 204 435 PKERO $ per klmh 0.89 0.48 0.36 0.195 PE $ per klmh 0.015 0.04 0.0075 0.003 Results -1.l 8 -1.03 * -0.74 Total WTP (per $U.S. household/month) 8.17 12.24 38.18 11.20 Source: Peru results for QKERO are for wick-lamp klmh/only. Bolivia data from Annex 9, ERTIC Pro;ect PAD, 2003. Philippines data from ESMAP, Rural Electrification and Development in the Philippines: Measuring The Social and Economic Benefits, Formal Report 255/02, May 2002. Laos data from PAD, 2nd Southern Provinces Rural Electrification Proiect, 2004 kLmh = kiloLumen-hour. * Based on linear demand curve. 118 Annex4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Survey Consumption of Energy Households in Rural Areas ~ ~ Household Module INEl§£µ,j Confidential Questionnaire Amparado por el Decreto Legislativo N° 604 ECEHAR.01 Contains: Characteristics of the house, household and the household members. Household energy sources. Opinions about the use of electric energy use. Time used. Household income. 11. House address 119 Special Report Peru: Nalional Survey of Rural Household Energy Use I 15. Functior:iaries of the Interview , I Offic:;e Cod.; ·9. Names & last names .'. fInterviewer: !foccll supe~viso~:. [Nqtional super.visor:· i ! 16. Total # people ' . ObservatiOns. I [ r:egjstered in chapter 200 .· 100. Characteristics of House & Household House Data I l101: .· Tipo ~e yivienda: · f 104. The predomincJ.nt mate~ial in the roof is; 1 Independent house Concrete 2 Apartment in building 2 Wood 3 House in villa 3 Tiles 4 House in vicinity house (Alley, or yard) 4 Calamine/fiber of cement 5 Hut or cabin 5 Bamboo or rustic mat with mud 6 Improvised house 6 Rudimentary mats 7 Local not fit for human habitation 7 Palm leaf/thatched 8 Other_ _ _ _ _ _ _ _ _ (specify) 8 Other material _ _ _ _ _ _ (specify) ~ ~ j H)2. The predominant material iri .the outer walls i.~:· l 1 Brick or cement block Excluding kitchen, bathroom, garage and storage. 2 Stone or sillar with lime or cement 3 Adobe (sun-dried brick) 4 Quincha (cane with mud) N°o 5 Stone with mud !106. · Do·you use a spa~~ i~ the·~rf~rm: <:In 6 Wood ! activity that pro:vi~~s incc:iitj~ to !H~ ho.me: · .:..· · 7 Rustic mat Yes l 8 Other material_____ (specify) No 2 [1-0?'~-:Th";);;;i~m!~.ant materi~IIn. the flbor i~,-~ Interviewer: If code l (Yes) is circled in question l 06, fill out chapter l Parquet, polished wood 800 "Business". 2 Vinyl or asphalt strips 3 Ceramic tiles 4 Bare wood planks 5 Cement 6 Earth, sand 7 Other material (specify) 120 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas f;1o7~· :.·rijcii:.hci~~· th~fyi>Qt' b'oJs~.h~ta .~ctipiefi~: '. .. io~~ .. ·The.sanit~tic>11··~~$t~m..o'f thi~ fi9u~~ii61a1i~s\ig~·; ·:;:v~;j What is the .:·'·::·:-;;cqi)ne,e:tec:J!to::.:., <.:::_, .:.): ,'.··:.=:o.r:· . ;;·,,S/. "<' monthly l Public network within the house amount? 2 Public network outside the house but within Rented? }s1 . _ _ I_ _ _..,,.....! building Own, totally paid? 2 3 Pit toilet (treated) Own, by invasion? 3 4 Pit toilet/latrine (untreated) Own, buying it on credit? 4 5 River, stream or canal Yielded by the work center? 5 . 6 None Yielded by another home or institution? 6 Other way? 7 (Specify) j_-~ __ • -;.--,- __ ~-,"· ,. ".' ."·-;~,.-- ,-t;_ -,,, .. ·.·._.-•';~_-,: '< -. S-_::;f,-----,;:_-· .. -~,,.-:-·~- ·' ---,,.,,--_"· __ :·;.\·-, ,•,' ·0 t~fO,?• ·,Jtte W9tef,'.5,llRP~ tQ: q~i.~k~rad::io:priiiJPre f·· , ' Cpd~ .. •' .:Yes '..No .. .:Yes' :N •• .. •• 1~ ' Namec. :: ' 1 1 l 2 1 2 1 2 2 1 2 1 2 1 2 3 4 1 1 2 2 {f~ l l 2 2 5 1 2 l 2 1 2 6 1 2 1 2 1 2 7 1 2 1 2 1 2 8 1 2 l 2 1 2 9 1 2 1 2 1 2 10 1 2 1 2 1 2 11 1 2 1 2 1 2 12 1 2 1 I 2 1 2 122 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Do you attend some How many hours center or do you read or Which is the last year or program study during the How old (Name} degree of studies and of regular night at home? Sex are you? Considers themself: level that approved? education? Frequency: Native quechua? Without level 1 Daily 1 Native aymara? 2 Initial education 2 Every other day 2 Native amazonico? 3 Primary incomplete 3 Weekly 3 African-peruvian or _Primary complete . 4 Monthly 4 black? 4 Oriental or of asian Secondary incomplete 5 origin? 5 White or of european Secondary complete 6 origin (caucasian}? 6 Mestizo? 7 Superior. Nonuniversity. Incomplete. 7 Superior. Nonuniversity. Complete. 8 Superior. University. Incomplete. 9 Superior. University. Complete. 10 Postgrado 11 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 123 Peru: National Rural Household Use Interviewee N° 200. Characteristics of Household Members ·;1or:~ ·: '• , . ! )' , ·' .·•,, . .. :, , , . ,. .21,:ii. •., ·,, .. ; ,· «:.'. .. . .... ,·, .. ., ; ,, ; .. : .. .· .. 215. ' .. , .. ··'' :216:··.. ?,-. •v .. }·;·. ~w, 2·11::0 r)'.,' _ "j ' . w ".. ; . . •. . " ; . , .. r. \ •.. • '.. , , .. , , .··. <, " .,· ·,,. ·. '',, . ·.· .\;•; .,. . ·; .'.)·,· ( •. , , :·;; ·;; .· . ·P(lra.14 aiiosy-mas edqtk):\ : . : ·~ .. ' : '•f ·.· • i. ·.:. .,... What is your business, What is the organization principal or enterprise occupation dedicated to for You work in your N° of that you your principal principal occupation Ord. Last week, from ............. to .............. perform? occupation? or business as: Perform some type of work? 1 Employer or patron? 1 Perform some task for money? 2 Independent work~r? 2 Did not work but has job? 3 Employee? 3 Helping on the farm, store or family Blue collar worker? 4 business without being paid? 4 Was looking for work before? 5 Unpaid family worker? 5 Was looking for work, first time? 6. Household worker? 6 Was taking care of home, Other? without work? 7 go lo {Specify) 7 Was studying & without work? 8 Chopt. 300 , & apply Living off pension or retired & Section l {go to Chapt. 300 & without work? 9 Chapt. 600 apply Chapt. 600 Section 1) Living off rents and without • work? 10 Other? {Specify) 1l . -·-·- -.~:_<:r:-:::· It:{ .· ,\\i. '· .'' .;,,, .·" \:.:,;.· OCI~': . . ·. ,'. ,c.. ·; .; '· .·· .. • '····· .. 'sBQ<:!fy .... ..... '.:'. . --"t'_:;--,_ -L. .. < , :~ Sp~ifY·' &· . •. : ,. ·.':'./~oc:le, - ·-~:_· 'I~ ·F •, ,.,• ,,;, >:~ l 2 3 4 5 6 7 8 9 10 11 12 124 Annex4 Questionnaire: of Households in Rural Areas The type of pay or income your receive from your Agricultural principal occupation is: Monetary income income Livestock Fishing income (Choose an alternative) section section income section section Paycheck? ~~~ 2 Commision? 3 Pay for unit (piecework)? 4 Tip 5 Grant? 6 Professional Honoraria (with R.U.C)? 7 Income for business or service? 8 Income (earnings) for Agricultural Act? 9 Income (earnings) for Livestock Act? 10 Income (earnings) for Fishing Act? 11 Others? 12 l 2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 8 8 8 8 9 9 9 9 10 10 10 10 11 11 11 11 12 12 12 12 125 Peru: National of Rural Household Use 100. Characteristics of House & Household 1 06. Under this question, there should be an instruction for interviewer to make sure that he/she fill in small business questionnaire module as well as complete the rest of household questionnaire. 2000 Characteristics of Household Members Please delete questions 212, 214 and 215. Please also include codes for question 213 (principal occupation) that include codes for rural and urban 300. Sources of Energy (Only for the Head of Home or the Spouse) . .. . i 30.l" 'Are the. followir:i9: ene~gy souri:.es used in yo~r home? .. ..,. . ·Yes' .:No ..1 ' 1 Electricity from interconnected g or isolated system 1 2 2 Kerosene 1 2 ~batteries 1 2 1 2 5 Car batteries 1 2 6 LPG 1 2 7 Solar PV home system 1 2 8 Firewood 1 2 9 Animal dung 1 2 10 Crop residue 1 2 11 Electric generator set 1 2 12 Charcoal 1 2 13 Coal 1 2 14 Other, specify 1 2 SECTION 1: USE OF ELECTRICITY FROM INTERCONNECTED GRID AND ISOLATED SYSTEM .302.. hcive Yes Go to 304 No 2 · Go to 303A 303~·.'. .tfy:oui ~~me hci$ !10 ele~fri~.ity~ pl~~e. in(U~a.'tewhether _the ~qllowing statements are. mhajo' r,",m. inor 0.t not a r~sori l I I . . . . ·to expJam w~y.the ~ousehol~ 1s not CO!'!f1.&cted;to thegrrp? Code: Major Reason = 1 · . .. . , . · ··.. .· . ·. · . . · · .. ··, Minor Reason .. 2 Not a Reason = 3 Not Applicable = -7 No Reason Minor Reason Major Reason 2 3 2 3 3. Our household can't pay the cost of house wiring 2 3 2 3 5. Our household can't afford to buy electrical equipment 2 3 6. We are satisfied with source 2 3 7. We do not see any application of electricity 2 3 8. Other reason 2 3 '":. .• . • ',; .1 'h' ,.··:,· •. ; 21.1. 1. 1.1. l. 1.1 Yes 1 Go to 326 No 2 Go to 326 126 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Code: Write down name of the company 21.1.1.1.1. l. l.2 Yes 1 21. l. l. 1. 1.1.1.3 No 2 Go to 308 !;3oz..·:· :nyhou~~n~M; 9.r§ c6rfn~ct~:i~~tbe;s9fu~.~1~ itj~fer. i~~~oditf9.)'~~rs~:: Code: Number of homes or /1 l" if the responding household is the only home that connect to electric meter · ··:399.::•.l-!o~mciriydciy~ pl!r it{C,~t~CJ~~S: "3JO{ :o~ri~Sthe.la~:fa gionlhs;hoY/. · ... .) y:.·.yQur.hou~sehoJd !YPi!=a!ly hay~.··. .. - • titany; months.ha~.yol!r.home. . .e,l~tr:icit)''~ei:vi~e;i9'.your~olfle?, .ha_d el~ctrid.Y s'ervice?... , Code: Hours per day of service Code: Days per month of service Code: Months with service for the last 12 months . Don't know -8 Don't know -B Don't know -8 .. .~ '·:i·: , L't 3 Go to 315A 4 Do not 5 Go to 315A 127 Special Report Peru: National Survey of Rural Household Energy Use Per KWH used (amount of units consumed shown in the meter) How much does HH No. of days pay for each billing? per billing period By the number of bulbs, fluorescent tubes or electrical apparatuses 2 i Fixed charge or flat rate 3 Others----------------- 4 Go to 315A (Specify) r~.a. .·.J·l·;:·:".~~?;·~.~t.~~1im~'.f.~.~~~.·:~:.tu·b·!r.~r~~~;.·~Gy.:d.r.t.·.~:~i.f;u.'.m.r;.4~.·.~.~~· ~.~ .·~ ~.-~.:~;fb.}~.l.~.~:.•~ .·.f. ';.;.;· ·; a.·:~1.~: .~.nl.y·\~~.o.~..d.:~~H,..jj .• . ! . .:.: .u,slJge:and cost of; ekictr,11;i1f¥ excl.ud1pg·m~tallat1op; feezDo;notirclu~e iostollati9n. fee . ·may be.!n~lu9tiq· m . .. ·· .• 1 h'·'.; \;t. ·th~:~lll/_,;:.:'7 ~!" ::, :..·;;·':...:.'.•_e.,~ ; :,', . ~: ~:.::, /·· ::/> .: ..).r ·: . . ,. ~ . ,.::;"/·; ·~:. , ; '· :;:, ;;: '" :;. ,.. ·1 Date of the previous reading Dote of the last reading G. KWH H. Cost A Day B. Month C. Year D. Day E. Month F. Year Usage (S/.J Bill #1 Bill #2 Bill #3 r~t~~~f~'.ff.re~~~~-~~t'~~~ffti!;~11o~··e:~\'ip~;Ei1~ti~~(iY;·b1i1,_:~h~1·'is·t11~~e~a~e;~~~fit.~Q':~rie·~~h~li;ta~:ctti~)·pf ..:;\.;q [;{:-'.:!:~: >"~·-. --~~l!¢t.fip~·$e~~f!~-k~~:~:_r_):;--i~;-L ~~ --~~--~ ;:t;::~~---\: .:·:\~ ,:-:"4._\-~--~-'~~;:~~-£>'.f>-:y-:~_-~~:-~-·.> .~- ~-·-,. "~ :- '.; _ i~---·" ~<;-~~J:~~-,;~r- "t;:\'~1:/·;-/":> _:;C~-~:-_~_::k::'"-:,._.:~-~-'~·:~- :·:_;:i~-,~~~- - -~:~7:1~-t:-;>d Code: Enter month. Does not know -8 Code: Does not know -8 No 2 Does not know -8 Type and size N of light bulb Yes No No. of incandescent N6. of hours No. of minutes 1 25 Watts 2 3 50 Watts 2 4 75 Watts 2 5 100 Watts 2 128 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas ·· · ·.. ).l6~:· ::!:Wh!lt:is:JG~:s"'til~;al!. •· '· ·.·····: JO.r~all.•J>ulbs2usedd1f . . t: .• last:24'hour.f:>~nocl?. "'~; ~ ;'~ '': ,- -~.,_~}": ~:;; _,': _ / ,::_:_< /:--_~!- ,,- i;~--;_-~::;t~~-\~::__,_ Note~to. enumerafors:~k:;t \~spo~clenfd~ottitheus~;of:e· bulb' .in waJt .classes 6f !.:>u18.s:i householckand scl.rri'·th~:fotal: h th~ b.u~lbs·:~t~ ~.s~~hniih~.J&st; Type and size of N light fluorescent Yes No No. of fluorescent No. of hours No. of minutes 10 W (Straight) 2 2 20 W {Straight) 2 3 2 4 22 W (Circular 4 32 W (Circular) t~317A:-\~~. yo~.r:~~Ysehoi~:.~se 6~f·:" 13]:1B., .:·;H9~;.a1~r1y.,tubes·~. th~s. ci~;s r ~()f,t.~~.follow1~g.energy:si::i:ving · : :does~t~~ho~seholCI ~set· / ··,'·317e~· .•)iV~~t i~ th~ ·s~T)f~~ll·~~~r~~ :·· · . : f9r:all•9ulbs•y~.ea~99r:ing:the ~t \"::light,~til~s? •~ ~' .. · ·· ., · · · · · ... . . last pt hCl,u.r .~ri()~?. . . ~ . ~· ~: ...\r ·:::<·· ·:f. .' ·. . ·. . ...f~t~t$::~f\~~tH~. '~~'.·.·•·.···· . -.;.i ~teJo.·e.n~m··· e . J':'n:~sponde.nt apo0f . . . f:~. ' .ecti:1c1t~<•;.~ • . ·· · ""·' ; : . .,-. :. ,.;t_ :·.::~ .. :::··· '•.· , '1 Code: Reason Minor Reason No Reason 1. For entertainment 2 3 2. For information and/or the news 2 3 2 3 2 3 2 3 2 3 2 3 2 3 131 Special Report Peru: National Survey of Rural Household Energy Use SECTION 2: USE OF KEROSENE Yes 1 No 2 Go to 330 :a~78:;t''l;lo~+m~~y units:i>f .·• :• .~·/Y>. 5ene·Clo.you use .gu1rchci~~l(et()sl:lne1:~A'1~/ { 1" p ~l)qthi·f' . ';' , ~~;~, P~itJ~fers :.fa ~~e•ot·., .. . . . . answered in.A.: : .. ·.. ··., , ., J_:poi~f·. ~I~~~ . Jfi'a'fr'~,,.~· 9 6Jlbn ~r • ·· --"=-'--'4~"-'~-'--'-'-"-::..-.;.~~-'--~"'""--'-"-...-=""-"--'"'-.._;-'-"-'~'-'---"-' Code: Code: Enter number of units Code: Enter price in S/. per Code: Amount in S/. of 1 =Gallons of kerosene used in a unit answered in A. monthly spending. 2 = Liters month. 3 Other _ __ (Specify) Code Number Quantity S/.per unit S/.per month Code: "O" if none and percentage if applicable Does not know -8 Not applicable -7 Yes No Percent Does not know 1. To start firewood 2 -8 2. Lamp lighting 2 -8 3. Cooking 2 -8 4. Appliances 2 -8 5. Home Business 2 -8 6. Other (specify) 2 -8 Total 100% 132 Annex 4 Sunray Gueslicnnaire: Ccnsumplion of Energy Households in Rurul Areas -.whaH~ithe'sdrn of' 3290:; · -W~cit}i~1ihe.s4m of·. 1 :p·111~_ours'f~.u:.C;1IJ;,_)':, ' · · · qll :d.ay~:fcir-alL: . , · ··.11_!sep:du~ing·.iheilijst · :~seClr~urlng:the'lpst · 243~Qu,r;pe~i~~(:~; : •· .:·"ni~j1tht?. · ·.. ::~ · .·; ·NC?te:· Asktbeire~poDd~nt ::~~te: Ask'the··r~s~ondent ·about the use of each·>"'.. . oO'out'.tne'.us~ of'Eidch . .. . . : in .t.he .ho~sehol~ ~rid liu;n·: .' i;Hhe ~Ko~;-~hold and suai dHe fotal:.hc>u·rs,thet:tbe!.. ·th.a :t9td'I '8dys.that,ih~: .. , ·ar~ used in: the .last'2:4 < dJ.9' o~~·Tn,!h:e.Jlait3nontb .. · :;h~.~r,~:, ,/ . . ... . . ·. .. . ,. ~- ... Code: Enter number Yes No Quantity No. of hours of days 1 . Simple wick lamp 2 2 ... Hurricane lantern 2 3. Petromax lamp 2 4. Wick stove 2 5. Pressurized stove 2 6. Refrigerator 2 7. Freezer 2 8. Other, specify 2 t .,· .· . .... . ... • ~ '• .. . • •. • . . . .... . • ... • . . ·. °"' .... ! 330.. · .~l~::~he,J>a5t month,- d~~ :r:~i!f.'hc)u~~hQI~ '4!e ca,~dl.~s;:for.,ilJ~n:1irici_tiOn? Yes 1 No 2 Go to 333 j 33.t:J\::"J,J:iow;m~~Y~~aildli!.;' What is-,tl~~(pri,ee;9f ·. .·whcit:is .the c;ivertJg~: :33.1 p,:. wo· __.91·1·_·hj~otuhr'es·. fso"_·rm_:__a·~.1......·.·. lj f.ha I · ·.i~!d you'r 1lfo.~s~hold 'each"caiidle{. · · · . :,monthly E!xpenditµre .. ~ f·.:~ 1 '~-JU:se~~in_1th~'.1pCsf. \, ·._of,theih~use.~?ldon .·~a~dle.s·used du~mg:•l .1 • · • 1"1onth?: / . · c:andles .. • ·t~e !ast24.-hour. · ·--·j i'-, -'.~ peri9d~ > \ . · ··. . · j . . . ! . . Note: :Ask :tbe respondent -1 about the use 'of ea2h caridlel in.the :h~)us~h61d.~nd sum . ·tbe to~al hoiJ~s ;that }hf!,. · .. ci:irldles :are-used' in the. last 24 hours .... · . · . · · ;j Code: Enter number Code: Enter price in S/. Code: Enter monthly Code: Enter number of candles. expenditure in S/. of hours/min. Hours Minutes 133 Peru: of Rurol Household Use . 3.32A..~;;; youi;~h~us~hold, iJse ·~an~les~f9rJgi.1c?.,.,in9. ··· - -3~3-2B-. Ii} g~nerdl,'.What P,;fcentage~~f.can~les dQe~ · ·. .:-·~ l l . . Code: purposes? . . ,, / · .· "O" if none and percentage if applicable .: , the •nouseh.old,:1,1se ·each month· for tile f()!lowing purpos~s? \• ~ , ·., ~ ! ' l Does not know Not applicable -8 -7 ~:rj Yes No Percent Does not know l. Home use 2 -8 2. Family business use 2 -8 3. Other 2 -8 Total 100% SECTION 4: USE OF DRY CELL BATTERIES Yes 1 No Go to 336 r·~34A.. D~syour · . ·.•.. In a.typical.Jll1>nth~. ~34(:. Wh1;Jt was the pric~ . :3340~ .• li1 the 1a.st monthj ! . · household u5e how mqny 4ry : . of ecic;h.battery of . . hov( much· ~id •the I I batteries . . . of . the · · :bqtteries 6f.~ •. ..: did. 'site, •••? .h'ou~holp spend·on t j . .f()llowing sizes? · < your ho_usehol~ u~e ;batteries for ~ch' L .... .. .in the past month? .size.· .•?· < -, ' Code: Enter number of dry Code: Enter price in S/. of Code: Enter monthly Yes No cell batteries. battery. expenditure in S/. l. Large (Size D &CJ 2 2. Small . (size AA &AAA) 2 l 335A~ : O()es:you.i:h~!!Seh&ld usedry.c;el!batferies'f9rthe·· · 335it 1: ·< · ': following purpos~s? ·. · · . -_ . · How many'hours~'per;~6y'~~e~ your.~pusehol~:'.u~:1 the ..•?· .-/ ', ·. · · ,_. ·· ·-.i Code: Enter number of hours used per day; do not use any enter "O" Yes No Hours Minutes 1. Radio 2 2. Clock 2 xxxxxxxxxxxxxxxxxxxxxxxxx 3. Flashlight 2 xxxxxxxxxxxxxxxxxxxxxxxxx 4. Others? 2 Yes 1 No 2 Go to 344 --~, Code: Enter number of car batteries. I I 134 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Code: Enter C:ost in S/. of Code: Enter voltage of car battery, Code: Enter ampere of car car battery if no.battery enter -7 battery if do not know enter -8 Batt Do not Not No. 6V av 12 v 24 v Other know applicable 6 8 12 24 -8 -7 2 6 8 12 24 -8 -7 3 6 8 12 24 -8 -7 Code: Enter number of months previous battery lasted. Does not -7 '3J.aJ.: ..iH~~~o/ '3':181?.wh~ti~ ·/( · · ··'::. :rechqrg'es/ the co~tof 'fl:>i: all•:c;lr ' • .ro't1114 trip•:: · '.C.l>afterie~ · . tmjt'spoi't~tion . · •"-d~s.yc?ur~ ·per ~harge? ',ht~· to. ,enur\ierator:s: As~ the rtespqn~el').t about th.e •"'~e 'df,eac:h ~ 'l;iulb: in. watt dbss~s ol lbulbs in the! . househ()ldand sum :the,total hodr~:ihat tlie·bulb~ are us~<;! i~ t~.~ last 24 h~urs: Code: Enter the number, or "-7" for Code: Enter hours of use with Type and size do not use fraction., or "-7" for do not use of light bulb Yes No · Hours Minutes < 10 Watts 2 2 15 Watts 1. 2 3 25 Watts 2 135 Peru: National of Rural Household Use How mailyt11bes fhisciass .· l40C~ ·-What. is the sum of'all· hour~ ••. ;:\! does the ho~seh 0ld use? . . . • Jor :all bQlbs ~s~~dui'ing th~.. J . last 24 houl'·period? • · . :i Nqte !o'. E\num;rati;>rs: .. JAs~ _the.. •• · •j re~p€>nd,ent ·about the ·u~e qf eoeh ..· .• ' .:.1 golb.in·W(JHc!qs~es9f b:ulbs in the , j hol:lsehold and s.um t.he total hours that.· 1 tne buJbs.·are.«u.sed -lnJ6~ foit 2.(Lh'aur~; J Code: Enter the number, or -7" for 11 Code: Enter hours of use with · Type and size do not use fraction., or "-7" for do not use of fluorescent tube y N Hours Minutes 10 W {Straight) 1 2 2 20 W (Straight) 2 3 · 22 W (Circular) 2 ·.341A, .,D()Elsygud~ou~hold 1tse. · · , .341.B. 1-i<>W m?nyligllf bulbs of this .341C. 'Wnat i~ the su~ ofall_.hours. -! any qf the .following en~rgy: ' class does the householp use? ·fOr all .~ulbs used du ting t~e , I saying light ~~lbs, which are· ' 'la~:2~· hou(periOd? ~· . · l ·energized. by cijr batteries( · ' . . '.! • -- _y - - ~: :- ; " :;. · ·. Note.to enuirieraiats; Ask the: · ! , respondent a·gp~f'.fhe~u$e of each.· .' »I .bu lb')'n waif classes of bulbs in lhe. \I household and so,;,:the'total hour$ th6t 1 lhe .- -- bulbs -- ' - ar~-· used ·frl /)' . tfie'. lcist ii!: kours)l ,;; ', ', ,, - - ,,. -·, Code: Enter the number, or -7" for 11 Code: Enter hours of use with Type and size do not use fraction., or "-7" for do not use of light bulb y N Hours Minutes 7 W or less 2 2 9 Watts 2 3 12 Watts 2 4 18 Watts 2 5 20 Watts 2 342A. . Does ;youi->h~U!iehold ,!,15e.~.'-~ar ~itery for the,. 3428•. ·In gef!e.ralt Wfi.J~i~~~pur~~.s···~;-•?.-.. 346~. ' 'In gene~C~hat. pere~nklg;-of_1_PG do~s your-·.-.-~·-·1 1 r.··346;\••' -. i ,hciuseh'old use each. month for theJoll9wing ' . !" · · purpo~es? . · · I Code: "O" if none and percentage if applicable Does not know -8 Not applicable -7 Yes No Percent Does not know 1. lamp lighting 2 -8 2. Cooking 2 -8 3. Appliances 2 -8 4. Home business 2 -8 5. Other (Specify) 2 -8 Total 100% 1-13-47_/j._:_·_D_cie_s~t-he_h_o"-u-se_li_o_ld~~-~ 'a47B. H~w'.nia~y·of egch .34.7C. · Wh~t is·the sum 3470. What is the ~urh .. use the following gas · appliance does · of all hours for all of~li days for all I f , appliance? . your household have? · ·•.• us~ cl.uring · the last 24 hour . ·... used.during the last month? i . pf!riod?, · · · ·· ;' ',' ~oh~;.Ask the respo~de.~t N~t~: Ask the respondent . about the use·of each'. . about the use of each • , . ·in the hdusehold·and sum in the household cind· sl.mi' thefotal days th.anhe... the t~)tal hours'that the . are use.din t~e ldst. ;. : ·.'are used in th~ last 24 houFs. ;: . Code: Enter number of Code: Enter number of Code: Enter number of appliances hours, or if do not days, or if do use any enter "-7". not use any enter "-7". Type of Appliance Hours Minutes 1. Gas lamp 2. LPG stove 3. LPG stove & oven 4. Refrigerator 2 5. Freezer 2 Yes 1 ·No 2 Go to 360 Owned? leased? 2 Go to 351 Given to the hh? 4 Go to 352 Not -7 138 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas .·:· ,il~Y!necf [:~hqf'Qqs ~citc~~t.RqiCl in;;Ca~h'.;f~i ih~ sAl9rjpy ·fi~111~:~m·liHc;l6a~ cih:ihe;~~P.6!1~~f!lf :· ' :·~;;q Code: -7 Sf. Go to 352 /3§J~(flf leas~~,·~qw·rJi~blj is the·· · !\ · ~~tr.c:>nthly paym,entt t· · · Code: Ente~ the amount of monthly Code: Enter the number of initial Code: Enter the number of payments. payment. payments in Sf., or if initial Not applicable -7 Not applicable -7 payment is not required enter "O" Not applicable -7 Sf. Sf. Sf. Code: Enter year the household obtained it (i.e. 1990) Not applicable -7 Does not know -8 Code: Enter repair cost in Sf., or 110 11 for no spending on repair Code: Enter size of solar PV in Wp. Does not know -8 1:v••..,,..~•..?~0es·y~ti1-·11~~s~H~l(us~~~nf;.: ' 3S5B~ •. 1:19~ P\any·UghtbuJ~s·irltnis· .· •· :sssc,. :.whari& 'th~"suiri of all'.h~uii ~f tli~.foU~~ing il:!~af!de~enf :c;lass doesthehousehold use? '. 'for tilLbulbs. use~. during the · " {ight{),ulbs,·which:are·= . \· :: .· last 24 hour period?. · ·: :;~~ergi,z~d by solqr.P'f~s~tc:l!fi,? · · ·· ~· ·.· ·. ···. · · :::· ~ .. · · · · :Nofe:to~nu!Tlerator.s: Ask:the : ;..•·..·.~.·.·..·.:•~..·'·.·.• ,:.' .·.·.~·,'.... · ·,~f;~ponde~t 'Cibqutthe. u$e~·6f edch · bul&'in watt'.'closses of:bolos.ln 'tfie housel1okjand ~~m the)qtci houis . the b.ulbs·are used in. the.last 44· hours. Code: Enter the number, or -7 for 11 11 Code: Enter hours of use with Type and size do not use fraction., or 11-7" for do not use N of light bulb y N Hours Minutes 10 Watts or less 2 2 15 Watts 2 3 25 Watts 2 139 Peru: National of Rural Household Use Type and size Code: Enter the number, or "-7" for of fluorescent do not use tube y N Hours Minutes 10W (Straight) 2 20W 2 (Straight) .2 22W 3 (Circular) 2 11 Code: Enter the number, or -7" for Code: Enter hours of use with Type and size do not use fraction., or "-7" for do not use of light bulb y N Hours Minutes 7 Watts 1 or less 2 2 9 Watts 2 3 12 Watts 2 4 ] 8 Watts 2 5 20 Watts .1 2 Code: "O" if none and percentage if applicable Does not know -8 Not applicable -7 Yes No Percent Does not know 1. Lamp lighting 2 -8 2. Cooking 2 -8 3. Appliances 2 -8 4. Home business 2 -8 5. Other _ _ _ __ 2 -8 (Specify) Total 100% 140 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas !Jsld u~e !he . ·3598. How !llaiJ¥ 359G. . ~at i~~the' ·.~59f ..Wf1 ·' .· .the.~ll•fr!?(~ll~~~u~s ·.~. o·:.·11? .• .: ·:~wh1~h ..· ·., . . .:.w ... ··.·.f ri·.·g· · ;e.le . .•. i.are . i·c ctr··.·. .. ~i.•.a·.~·. ..:•.·P . ·.·.qe poweill··,_ -·": "- ;,'."'._ -~;:.:.··:" : :.>~'_,,_,;:(; ;;.;,::',0J\':'_·_,,--,__ 3~'5.B; :·Ho~}~ClnjuriitS .~:.;: . .~'·::' ';Xoff~~l·111e11tlcfoecf ;X· ;· : ;itir369kdi~-y~ur· ..:; ) hou~hold Jseffii'.:,:r · .9~,,:sliflost;f.n~ri,~?~- Yes No No. of units Type of unit S/. per unit S/. per month Diesel 2 S/. 2 Gasoline 2 Type of Unit Gallon ..... 1 Liter ........ 2 Code: Enter repair cost per month in S/., or "O" for no spending on repair. S/. Code: Enter the number, or "..:.7" for Code: Enter hours of use with Type of do not use fraction., or "-7" for do not use light bulb y N Hours Minutes 25 Watts 1 2 2 50 Watts 1· 2 3 75 Watts 2 4 100 Watts 2 142 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Code: Enter the number, or - -7 for 11 11 Code: Enter hours of use with Type and size of do not use fraction., or "-7" for do not use fluorescent y N Hours Minutes 10 W {Straight) 2 2 20 W (Straight) 2 3 40 W (Straight} 4 22 W (Circular) 2 5 32 W (Circular) ·H~~ri\;;ny'ligfif b~l$~··i•ft~1(.:- :~~?«£.'··wfj~(i~ ~tt ·s~m_of ·all..hi>ul"~ .. · iclds~does the1hou~~hold use?-'i· . · _·...•• ;· for~alf~.Ll,lbs: ~s~'Cluring fhe . '! ·· ·- '· .. : '-~')a~t ~4.h~~{perjodt· ._ _ . :•_!. '"'" ~ ,, ' (N ote·i~ eriurrier~forsf A~k· the •-• _,·1·· ,r~~pg~~-e~t abciVt.theuse·9f ~c'h"'· b,t:1lb irJ,,Yatt·clqssesofbul!:)sinth~ C . th h?lc!~a~d;sorn the)otal bours"thqt·,. Jbsdre·used in the,:lqst"24 Hoors. _ 1 11 Code: Enter the number, or '-7" for Code: Enter hours of use with Type and size of do not use fraction., or "-7" for do not use light bulb y N Hours Minutes 1 Less than 12 W 1 2 2 12 Watts 2 3 18 Watts 2 4 20 Watts 2 5 25 Watts 2 ·aios:: . ~~3ZOA··• Q~es.;yoor:~oy~e~qid ,us~kl~t~ic· gkil.~i:at~~'set-lQ~ :. ·. lii.~eiieraf:~al p~n;eiitage . of: tour it;,usehold - -· k:-- :thefpll~~mgpu.rp~~-es?;. , '<..,:; . ,. · ... ;,. . "1.ont~~Yi.~~ndin~~n:eJ~,~icg~nef(Jto~_set·isfof.the_l ti;' •.'.!£ ~;;~·::':,\~ '.::'~. · )9llP,WJOQ,;Pf,!!'f>O$eS., ·{ · · y .. j Code: 0 if none and percentage if applicable 11 11 Does not know -8 Not applicable -7 Yes No Percent Does not know 1. Lamp lighting 1 2 -8 2. Cooking 1 2 -8 3. Appliances 1 2 -8 4. Home business 1 2 -8 5. Other 1 2 -8 {Specify) Total 100% 143 Special Report Peru: National Survey of Rural Household Energy Use -"-. ---~--·---------~- 3i1A: D~~s'!l~ h.ouse~old. 'us.e. t~e . ~718. Ho.Wm .. any -37'.1 c. 'What~- 3i1o:--w~at ist~;~~; of alhhOu_rs ~1 · following el.ectric appliances, of each. . . average for all appliances used:during ·.! r j, • . : which (Ire power~ by . . . . appliance . . wattage . the last 24 ~our pe~iod?- t electridty fr?m generator ;.does .your . . rating. I . set?· · household · of the' . Note to enOmerator.s: If the household [ ·· have? appliance? ncis more thcin·on~ appliance of this_ . l .type', (JSk}he r~sporlcJenf about the use j !· ;Note: Estimate the of each appliance in the household , · i I average wa~age 'jf rriore than one . ,. and surn • the aP.P. .. . . total hours . . that the liances: are.used _in the last . j i ·i l I- a'ppllance in use; .2~ hours. 1 Code: Enter Code: Enter the Code: Enter the number of hours of number of average use with fraction or if do not appliances number of use enter "-7" or if do not watts of use enter appliances u_711 or if do not use eriter Yes No "-7" Hours Minutes l Radio 2 2 Sound equipment I 2 TV black and 3 white 2 4 TV color 2 5 Video recorder 2 6 DVD 2 7 Others 2 (Specify) Yes No 2 Go to 376 Go to 375A Purchase and collect 3 Other _ _ _ __ 4 (Specify) THE FOLLOWING ARE QUESTIONS FOR PURCHASED FIREWOOD ----- f 374A. HowrmJch-did yoq 3748; · · How ma11y totar. .374c; What was the. 3740. ·:How long did'i! 'take. ~. · " .spend pqi:ing the · day~ will'this .. one·waydistance- · · fo tiavel 01_1e"wurehase-last? traveled· (in meters) · · ;make thispuri:ha·s~. to make·this of firewood? pur'chase? Code: Enter amount of Code: Enter number of days Code: Enter distance in km Code: Enter time in hours money (in S/.) spent firewood lasted. traveled, use fraction and minutes. last time. for less than one km. *Don't include transportation Does not know -8 cost Total S/. Decimal Hours Minutes Adult Male Adult Female Child 144 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Code: Number of collection Code: Enter number of days firewood Code: Enter distance in meters lasted. traveled, use fraction for less than one meter Does not know . . . -8 '.:~l~ci).:ij~~~i·~:~6 :9~~t~~9o~:·er1t1ti!·:f~11~w1r19•~~ih~r~i.~.·~':1 Code: Enter number of hours or "O" for not spending Code: Enter hours of use with fraction., or "-7" any time for do not use Not applicable -7 Use Type Hours Minutes Adult Male Adult Female Children SECTION 10: USE OF AGRICULTURE RESIDUE Yes 1 No 2 Go to 378 Code: Number of collection Code: Enter number of days Code: Enter distance in meters agriculture residue lasts. traveled, use fraction for less than one meter Does not know -8 Code: Enter number of hours or "O" for not spending any time Use Type Hours Adult Male Adult Female Children SECTION 11: ANIMAL DUNG Yes No 2 Go to 380A 145 Special Report Peru: National Survey of Rural Household Energy Use Enter number of days dung Enter distance in meters lasted. traveled, use fraction for less than one meter. $?;~~~.:J~:we;l~siw~~t·li.Jw;~Gch. · ~Jr:i~.. ,~~k) -~~s, . J~ecj5ifi'~61l~ci1J19~Juri?J ~~.tfr~ f9H~fo9,,.. ~~~~-:-w1"" Code: Enter number of hours or "O" for not spending any time Not -7 146 ' SECTION 12: USE OF COOKING STOVE AND COOKING Code: Code: Code: Code: Code: l =Open fire l = Outdoors 0= None 0=. None 1 =Firewood e.g. three stones 2= Semi-enclosed l = One only 1= Firewood 2= Crop residue or wood 2 =Traditional 3= Separate 2= Two or more 2= Crop residue or wood chips stove no kitchen chips · 3= Dung cakes chimney 4= In living area 3= Dung 4= Charcoal 3 =Traditional 4= Charcoal 5= Coal stove with 5= Coal 6= Kerosene chimney 6= Kerosene 7= LPG 4 =Gas/kerosene 7 = LPG 8= Electricity stove 8= Electricity -7=Not applicable Code Number Code Number Code Number Code Number Code Number Code Number 1. 2. 3. Special Report Peru: National Survey of Rural Household Energy Use 400. Productive Equipment SECTION 1: ELECTRIC PUMPS Code: Number of pumps Enter 0 if none and go to Section 405 Yes 1 No 2 Yes No Agricultural activities 2 2 2 j403J\. ·.Wh~t is t~fn.uP1~er6f the<~l~tficit}' ·•. •4038. ' What is th~ l: ..• .·.: . . :·.;' . . ·\. : .· :: .. . . . IIf.. j ~l~tric'oiltis{riot.available,,focik fo)ithe mqter . · hwmbe~· qt tlie. ei~dt~icity.mef~r. / '·~ ~ 1 ·· L..:::;· • ,· ' • ·.L . . ~· ' •. : ' "·. Code: Enter number. Code: Enter size of pump (KW) Code: Enter electricity charges S/ per year 0 Do not use -7 Not applicable -8 No meter 21 . 1. 1. l . 1. 1. 1.4 21 . 1. 1. 1 . 1. 1. 1.5 Pump# 1 21 . 1. 1. 1. 1. 1. 1:6 21.1.1.1.1.1.1.7 21.1.1.1.1.1.1.9 Pump# 2 21.1.1.1 .1.1. l .8 21.1.1.1.1.1.1.10 21.1.1.1.1.1.1.12 Pump# 3 21.1.1.1.1.1.1.11 148 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas 21 . l .1 . l . l . 1. l . 16 Pump # 2 21. l. l. l. 1.1. l. 17 21.1.1.1.1. l.l .18 21. l. l. l. l. l. l.l 9 21. l. l. l. l. l.1.20 21.1.1.1.1.1.1.21 Pump# 3 2 l. l. l. l. l.1.1.22 21.1. l. l. l. l.1 .23 21. 1. l. l. 1. l. 1.24 21.1. l. l. l. l. l.25 SECTION 2: DIESEL PUMPS Code: Number of pumps Enter 0 if none and go to Chapter 500 Yes 1 No 2 Yes No Agricultural activities 2 Livestock (including poultry farm) 2 Other, specify 2 C"Wti~t,is~tli~:si~e'lh'horse p6~el- ' WhatJs the total yearly cost of ,' :'. jf1~t9f,each:pumr>.t~ 1 '' , fur ·Clieserfuel • ·.,' each pumR? Code: Code: Enter cost of diesel in 0 Do not use S/ per year -7 Not applicable -8 No meter 21.1. l 1.1.1.1.26 21 . 1. l . 1. 1. 1. l .27 Pump# 1 21 .1 . l . l . l . l .1 .28 21. 1. 1. 1.1. 1. 1.29 21 . 1. 1. 1. 1. 1. 1 .31 Pump# 2 21 :1 . 1. l . l. 1. l .30 21.1.1.1. l. l. l.32 21 . 1. 1. l . l . l . 1.34 Pump# 3 21. l .1.1 .1.1.1.33 149 Special Report Peru: Notional Survey of Rural Household Energy Use I. I·. Last y®r/h.ow many"hours per Clay and days pe{ month, pump ~t numb~r . .y• • Last year/ho; .many hours per day and· days permorith~ pump set nu111ber .. .: .. I ... ..was'operated duripg'the, 6 month 'l{(Js·opera~ during the 6 month · I Diese~ pump .set number.·· . . · . . peric;KI between · · · periOd b~tween . . · § months period ·· 6.months· period•. April - 5eptem.ber .. ·October..:. March · N~. (if' hours No:of days 21. l.1 . 1. 1. 1. 1.37 Bomba # 1 21 . 1 . 1 . 1 .1 .1 .1 .3 8 Bomba # 2 21 . 1 .1 .1 .1 .1 .1 .39 21 .1 .1.1.1 .1.1.40 !21 .1 .1 . 1.1 . 1. 1.41 I 21 .1 .1 .1 .1. 1.1 .42 21 . 1 . 1 . 1 . 1 . 1 . 1 .43 Bomba # 3 21 . 1 . 1 . 1 .1 . 1 .1 .44 21 . 1 . 1 . 1 . l . 1 . 1 .45 21 . 1 .1 . l . 1 . l . 1 .46 21 .1 .1 . l . 1 .1 . 1 .47 500. Time Use r~re;;se indioote the ~mbe~;-of h;,;spent o~ various activities by ~aie and female household member and children in th~l hou~~hold yesterd~¥" in Jto~rs ·and fra~tions. All an.sw~~s should: be for: a 2~·hour P;eriod. · ·. : · · . ·~ I , Note: The total number of hours for all acti:vities'must add up· t9 24 hours. · . · · ! Code: Enter number of hours, or fraction for less than one hour or "O" for do not spend time on that activity category. -7 for not applicable (i.e., no children in the family, or no spouse of head of the household. (Enter 1st (Enter 2nd Woman Man Person ID Person ID (Head or (Head or number, see number, see Spouse of Spouse of ID number in ID number in Head) Head) Section 200) Section 200) 21.1.1.1.1.2 Activities for last 24 hours Hrs Hrs 1 . Sleeping (night sleep) 2. Bathing and beautifying yourself 3. Preparing meal/cooking 4. Farming, gardening, animal husbandry, fishing 5. Income earning activities such as, doing handicraft, tending shop 6. Taking meals 7. Processing food and/or preparing cheese & butter. 8. Water fetching and collecting fuels 9. Other household chores such as, washing clothes & house cleaning 10. Repairing clothes, basket, machineries, equipment, tools, and etc. 11 . Religious practices such as, praying, reading bible, and etc. 12. Reading/studying 13. Watching TV/listening to radio/resting 14. Visiting neighbors/socializing/entertaining guests 15. Other leisure activities 16. Shopping 1 7. Other, specify Totol (24 hours per person) 150 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas 600. Household Income SECTION 1: INCOME FROM WORK Code: Enter income in S/. No income earned 0 Person Person Person Not applicable -7 Number Number Number l.f>(>,1;·, >Dqrijt9:n1$:.1~i :1 ~· ~~ntljs;""'~;i, 1$, )'ouf<~;;~s~hQl~. ·.· . ., ·. •· .... , ___ -\-,_, __ :: ";_,,,_/:_-,;_,,; -,~: !,; . ·· :. ·~rt ~qsh'.!hc()jn~ 'f".9~, !~~!!,s!:~saJ(J~iEJs1 ~~d.•: {; · ·/:·, .,l sf.!nco,me; 3rd• 11l~~rrie . 4th. Income· !• ;e>~~l'l1!!?~.~. ;." ' :;~"'.·; . . •.,:·; ,' . . ' '> :.r'. . ~qrner: \ ;;:~i~~r: .. :;.< E,qrn.~r : (include wages and salaries from government, private company and selling labor} Total :;t'.. ,';::"~>~/,.:~·~·;.: ~ ,; '::7:'::. ·\~.•... '.;. " ; ;., . .: 'Ith: '.· . ~u;f1ng th~ l~~t 11~ nionths1 what is the !ncorn~ youi. • Income T6tal l ho~~~~qlCI r~~iv~Jf:Qi11 t~~:fcillcwing~sQy~~~: .· · "•'· ,. ·:~1;1m~r·:.:~~ 1s:11.;i 1 . Christmas and independent day holiday bonus 2. Vocation bonus 3. Profit shoring 4. Compensation for service lime 5. Other bonus income (specify) 6. Other bonus income (specify) Total f6~3: ,ovf;iF~ ~'1;:1~,~' 1·2,~o"'"~~:d;~ r~u( ~~~;~~:.~;_ .~· j .. ~nalhc:C!~e;.' 3rd.l~c;ome . TotOI·.~ f.•:.. rec;eiye in<;OIJI~ from the,·f(!llown:ig sou.rc~s.: . .:.: .· : · .~al'l)~r. :· . ,.Ead1el'· (S;l}, I l . Income from divorce, separation and alimony 2. Pension from being widow or surviving family member 3. Retirement pension 4. Remittance Total 151 Special Report Peru: National Survey of Rural Household Energy Use 1'<>04:· pµrj~g:th~ia~.1.2.; mo~ths,1 plea~ i~dicate !~~ ~tatairio~nt of irico~~ &ur h9~~hold:r~c~Mid fr()rn. the fcillowii:ig ·1' r agricultural adiy1ties: ' Q .. . . .' ' . . . .' . . . . . ' . . . '· /, I·.Ehumerat~n I . . . ~ . . (l.)· Ask for the gtoSS·incorne ear~ed fn:>~ sal7s of' each cigricultural prodJJC!. {2) ·i;~ter."O'( for no·in~<:>1z1e.ear11ed;i.n thaLc<;itegor:y. .· ; . ·.· · .. .· . ·. . . .~ J 4 1 ; ·... (3) To.odd up ince>me 'from agriculture,. first ~urn up all of the gr?ss .income frc;im epch ag'rkult1.1ral activily.in the I· ·· last column and enter result in row a. . .· ·.· " :· · . . ·. .. . · . . ,. · ·· !', (4) Enter"total expenditur-e 'for. agricultural.atti~ities in' row 9 last cqlumh. Be sure _to include di) types ~f . I .. 1, exp.enditure.• · . . . · . . . ,' . . · • . . · .. (5)-:p~duct expenditure in row .,9 ,and eijter r1et income frpm c;igricultu~e in tow l Q; • • ,. •· . . · ,· •. Code: Enter income in S/. Enter land use for cultivation in Hectares, use fraction for less than one Ha No income earned 0 Not applicable -7 Income From Agricultural Activities Indicate name of crops that you grow during the past 12 months. For example, com, yucca, wheat, coffee, cotton, sugar cane, fruits such as, orange, lime, apple, melon, grape mango, and etc. I6().4A. Tyfpe.· 604.~· rtQI . ·. ,' · 604~.:. What il t~e 601fJ·;·. Aipou.~t ~o(d 6c)4~_}F!Ht~; :: ~~t 'fqlcJl ·,:J !'· prc#luCt· · · .production 'eq~iV~lence · . -~ ' · .. · . ,1.• . \ : ~i";u~.i{i> .~--~ ~}!n~,!P~! [ '·. · ,. · · .· ·.. ·. in; Kilos? \' ., . . ... • . Jk1lo) .. :·1 ''.> : ($.1 )c-..,· J Quantity Unit Quantity Unit Quantity Unit (total) measure (total) measure (total) measure l. 2. 3. 4. 5. Total 152 Annex 4 Survey Queslionnoire: Consumption of Energy Households in Rural Areas r6os: ··:·,~,ea~indic~te;.ihJ total.nf: L ·· ,:q~imal•IJnlfciwls.. L:,..·:. ~tljat,y~ni.rciis~- ....· k' ·. c::i~d·sold over-th~·· . I.:: . ·.fA15t1i morith~.}. i: r ·. . ~qmple•6f animal .. F.· . ottivesiock·afe .... ! ·llg!nai .Olpaca, goat, !: .~• .... ·~-~~e,9uJ~ea;pig,;, rqb~1t, C()W p19. 1 ~ •" ·· •agd~etc•..ExarJ,1>1~ . i', .. . oUowls cire hen,. I-<·· r~cister;.~uc:k; ,and·· C· :,tuf~~y,·~'. ···\.·. l. 2. 3. 4. 5. 6. 7. 8. 9. 10. Total 153 Special Report Peru: National Survey of Rural Household Energy Use r,.69s.. ln~ome &~· by·pr;iu~fofliy;s;~k l?;tMties tmd product from ani.;;;i~~-;'''" ' · /Indicate nc:l111e 9f liv&stock by· product. t~at were sold over the past l~ me>ilths. : . 0 ·:-:-:-~-.· -·1 '· ·Exa~pl~·q~ -~y~pro~~~ts a~e wool; m_~l~t:__ch~se, ~u~r, and·etc~. · ' $ l !'..605.F; Ty~~s ~f .pr9dudS, 605G. Totof #produced 605H; .6os1: . Total Revem1e ... : /from sale of eacn· . · I l· ' .. .. . procJ~cfS/ . l Total, q4dntity: . ,U,nit Measure . 1. 2. 3. 4. 5. Total .6Q.5J... TotaHnco~e. from li~~~f9ck activities. · Expenditure :6QSK... Plecise}n,dicfaie the. to~i .exper1diture for liye'st0ck activities inCludii'!g· la rid rental r .. ..fee, hired labor, fodgers or f~stock, yacciliati!:m, medicine~, V(ater, and etc;; . ,. Income from fisheries I {,()QB •.. Please: i~dieate th.e tQ~l ·;~hdit~r~ fpr fisfierje~ inclJlding bo~t r~pair arid ., mainfenance;.'fue,1.s,fishingnet,,ariQ relat~ equipment; · · . .: · · i667:: !. ' r?.uring the· past' 12· mc!;n,ths •·. . ·fc> ·~ • '. following~ou~~s~ ' ' aid your household receive incorn~ from the ' ,., . ' r ' ' .• ,, ' ' •, •,: '· ' l . House, apartment or room rental 2. Income from renting agricultural land, or animal 3. Income from renting machinery and vechicle 4. Income from dividend stock and bond 5. Interest from savings or lending 6. Reward and prizes 7. Other Total l . Food for household members 2. Household expenditure for water, telephone, and transportation 3. Home maintenance and repair 4. Household expenditure for personal hygine-soap, detergent, shampoo-and clothing 5. Recreation activities, entertainment, cultural services Total 154 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas ''tii'e•rOlal(fi6usehhld·s··~·;. ,.,, ,.. , .. , ...•., •.,,· ._,,, ''"{,.,P,,, .1 . Expenditure for health care including medicine doctor fee and hospitalization. 2. Education of household members 3. Transfer expenditures (pensions, remittances to other family members, etc.) 4. Clothing and shoes for household members Total f·~9~~r;:c.16P~ Effi9~ttii11J~~~.~v)~~y)~.~[qf·w~~'.th~1.!~lfh l . Furniture & cooking utensils (furniture repair, electro domestic appliances, etc.) 700. Attitude Strongly disagree l Disagree 2 Agree 3 Strongly agree 4 Does not know 5 Disagree Agree Does not know Electricity is very important for the children's education. 1 2 -8 2 With electrical light the children can study at night 1 2 -8 3 At the moment, it is easy to read at night in the home. 1 2 -8 4 Reading with electrical light is better than with the light of candles or lamp. 2 -8 5 Our household is happy with with the lighting system that we have in our home. 2 -8 6 To use kerosene or oil is harmful for the health. 2 -8 7 A car battery is a good source of electricity. 2 -8 8 A solar PV home system is a good source of electricity. 2 -8 9 Electricity helps with domestic tasks and care of the children. 2 -8 10 Today, the quality of life of my household is better than it was 1 0 years ago. 2 -8 11 The monthly electric bill is or would be a financial burden for my family. 2 -8 12 Monthly spending for non-electric energy sources is/was a financial burden for my family. 2 -8 13 I feel safe in my house in the evening. 2 -8 14 I feel safe outside my house in the evening. 2 -8 15 The electricity makes it easy to have information and the news. 2 -8 16 Watching TV provides my household with great entertainment. 2 -8 -- 17 News and information from radio and television provide good information relevant for conducting business. 2 -8 18 News and information From radio and television provide useful information about agricultural activities. 2 -8 19 News and information from radio and television provide good knowledge on family health issues. 2 -8 155 Special Report Peru: National Survey of Rural Household Energy Use 800. Business Module SECTION 1: BASIC CHARACTERISTICS OF THE BUSINESS OR ESTABLISHMENT tJ1~'r~ri~ci~cifllpijit,fti~7of)~i~J~IJsiries~7~jij~l1 ~rit~tPris;i.····~··· 1 ·· Owner 1 Relative 2 Employee ·3 -~ '->'\<\ -4;, ! ', ----,/- -,~-J No formal education 1 Initial education 2 3 4 5 6 7 8 Superior University Incomplete 9 Superior University Complete 10 11 Does not know -8 Male Female 2 .1894.;). · · Production/extraction (fishing, mining, etc.) of some possession Commerce and sale of merchandise? 2 Describe 156 Annex 4 Survey Questionnaire: Consumption of Energy Households in Rural Areas Code: Enter and month in numeric ·What was Financing Financing the total Source Source amount Yes No Yes No financed? 1. Own resources? 2 1 . Own resources? 2 2. Personal loans from 2 2. Personal loans from 2 relative/family/friend? relative/family/friend? 3. Partnership? 2 3. Partnership? 2 4. Commercial Banks? 2 4. Commercial Banks? 2 5. Nongovernmental 2 5. Nongovernmental 2 organization organization 6. Money lenders? 2 6. Money lenders? 2 7. Cajas Rurales? 2 7. Cajas Rurales? 2 (Type of rural (Type of rural agricultural bank) agricultural bank) 8. Others? 2 8. Others? 2 (Specify) (Specify) SECTION 3: USES OF MOTOR (MOTIVE POWER) IN BUSINESS jThe folloiting ques~orisrefer ttfinotor(s) uSecl'iri th~ business• · . . .· .· . · ·.. .· · ' . ·. ·. .· ... . .. . . · .rd:l i.:· ~.. ·r:~.·.~:·~.·. ;.~! £::·· 0 .~'. ·~t·m.~p.'~·/·. ;a:'·':~.·~ .·~.·1:e· ~::.·;.~:.f.· ·.·.··~.e."~fJ°~~ 9 . V~.ed···t~:=.'. °J~~. ;:·g·~·'a~."f.1~~. r.d. ;. m~,:~' 9 . !'o.· .. .•. . ·mo···' ks.o9: ..,11.Ye>ur:·pu~iness d9:fou us~-rn~t9r~::to' 9ri~~,machin~rt for.any of th~ ~c,11owins ·aJ>~ncations~ .. . . · . : _:.,·;· · • . Yes1 No 2 l. 2 2. Cutting/drilling 2 3. Fan (exclude fan that is typically used for cooling in 2 the household) 4. 2 5. 2 6. Other 2 7. Other (Specify) 2 Note: If there are no motors used for any purpose in the household, go to section 4. 157 Peru: Nalional of Rural Household Use t~.,~~~~,i~~'.~i>ll!i~ati911~.m~ntioii!?d:?~o~~i~~tti~·'119~&a.rf~~~·6i~a6*'·qr~f~ie~l·iji¥>~ 11 11 Code: Enter 0 1 if answer "No" for all of the questions above. Number of electric Number of diesel motors or gasoline motors 21. l. l. l. l.2 21. 1. l. 1. 1.2 21 .1 . 1. 1. 1. 2.1 .6 21 . l .1. 1. l .2. l .7 21.1.1. l. l.2 21. 1.1. 1. l .2 21.1. l. l. l.2. l.l 0 21.1.1.1.1.2.1.11 21. l. 1. l. 1.2 21.1.1.1.1.2 21. 1. l. l. 1.2. l. 13 21.1.1.1.1.2.1.14 SECTION 4: INCOME FROM BUSINESS . Amount in (S/.l ,...,.-___,,....---,.~-,,.,.~~~-,,,-,.,,.~~~,...-;--~~-,,.~~ !.'.~;jq~ ;; ;p~·rLIJs'th~Ia·sj~~~fri6~~~s:Wli~~i51h~;t~t41:9~~,~~xf> \fg~ yq°'f; ~Jsiit~ss'(;~~·;·,:,(~; ... : Amount in (S/.l 158 Annex4 Queslionnaire: of Households in Rural Areas 900. OPINION AND ATTITUDE ON ENERGY AND BUSINESS ·~~!6JisHcifssia~~rit!~~oht~r~i~g,,~1J;tlJy]vi~and~.6~~ri~~u~~:~.1:.~~ulclJik~ ·y9u't~ '~il:·111~,if.yqu ;~: '...:;! O . ·. .··· ··• . ·,·:· · . . ·. · · . · :t~~s~;stat~iJ!~nfs~n4·~c;;yt~st,l:ol)g•your-fe:ling~'cire.··.:; :· ....:" ·1, · f~u;iing s~a:~~i'e~·'.4 ~n~ b~ · ·~.?~~~:¥·r~sp~68e~t·W.~~t~er:h~(lh~:~~i~:,or.pis~w~e;a~d:. . ., .h1sYJiedeelmg•1s·., · •·. •,. ••' ;·;·::.:;;,:;:,;:,.:~.,,,.·. ,:., 'Z • •• •• :.,. ·., :» .... :•. ,.· · ··l Disagree 1 Agree 2 Does not know 3 Disagree Agree Does not know 1 News and information from radio provide good knowledge 1 2 -8 for conducting business activities. 2 The use of electricity has allowed or will allow me to keep my 1 2 -8 business open for longer hours. 3 The low quality of the electrical service can be harmful for 1 2 -8 my business. 4 The quality of the electrical service has gone down during 1 2 -8 the last 2 or 3 years. 5 The cost of electricity at the prevailing rate is quite 1 2 -8 reasonable for my business. 6 The purchase of diesel for my business is not a problem 1 2 -8 to me. 7 With greater availability of credit (loan), I would buy more 1 2 '-8 electric appliances, 8 With electricity I could/can make more money from my 1 2 -8 business. 9 Electricity would help me run my business efficiently. 1 2 -8 10 lighting with Solar PV Home System is the next best thing to 1 2 -8 electric lighting from the grid. 159 Bibliography Barnes, D. F., A. Domdom, V. Peskin, and H. Peskin. 2002. Rural Electrification and Development in the Philippines: Measuring the Social and Economic Benefits. ESMAP Report 255/02. Washington, D.C.: World Bank. Bogach, S.V., Papathanasiou, D., Zolezzi, E. H. 2007. "Electricity Sector" in An Opportunity for a Different Peru. Washington D.C.: World Bank. Carrasco, Alfonso V. 1989. La Electricidad en el Peril.: Politica estatal y electrificaci6n rural. Tecnologia Intermedia IIDG, Lima. Choynowksi , P. 2002. Measuring Willingness to Pay for Electricity, Asian Development Bank, Economics · Research Department Technical Note Series #3. ADB. CNE (Comisi6n Nacional de Energia). 2005. Programa de Electrificaci6n Rural. Santiago, Chile. Available at http://www.cne.cl ECN (Netherlands Energy Research Foundation). 1998. Rural Lighting Services: A Comparison of Lamps for Domestic Lighting in Developing Countries, ECN-CX-98-032, July. Gallardo, J. and Bendezu, L. 2005. Evaluaci6n del Fonda Social de Compensaci6n Electrica. OSIN"ERG, Oficina de Estudios Econ6micos, Documento de Trabajo #7, Lima. INEI (Instituto Nacional de Estadisticas e Informatica). 1993. Censo Nacional de Poblaci6n y Vivienda, Lima, Peru. - - . 2000. Factores que Determinan elingreso de los Hogares en el Peril.: Formulaci6n de un modelo Estadfstico. Lima. Available athttp://wwwl.inei.gob.pe/biblioineipub/bancopub/est/lib0385/indice.htm - - . 2004. Encuesta Nacional de Hogares: Condiciones de Vida y Pobreza. Rondas I, II, III y rv, Lima. - - . 2005. Encuesta de Consumo de Energia a Hogares en el Ambito Rural 2005. Base de Datos. Lima, Peru. Komives,K., Foster, V., Halpern, J., and Wodon, Q. 2005. Water, Electricity and the Poor: Who Benefits from Utility Subsidies.'' World Bank, Washington, DC. MEM (Ministerio de Energia y Minas). 2004. Plan Nacional de Electrificaci6n Rural. Lima. MEM. - - . 2005. Plan Nacional de Electrificaci6n Rural 2005-2014. Lima. -.- - . 2006. Plan Nacional de Electrificaci6n Rural 2006-2014. Lima. NRECA (National Rural Electricity Cooperative Association). 1999. Estrategia Integral de Electrificaci6n Rural, Lima. Meier, P. 2003. Economic Analysis of Solar Home Systems: A Case Study for the Philippines, World Bank. - - . 2007. Estimating Benefits of Rural Electrification from Household Energy Surveys: A Guidebook for Practitioners. World Bank. Waddle, Dan. 2005. Propuesta para un Nuevo Marco General para la Electrificaci6n Rural en Peru. NRECA. World Bank. 2004. Project Appraisal Document Second Southern Provinces Rural Electrification Project. Washington D.C. - - . 2005a. Opportunities for All: Peru Poverty Assessment. Report No. 29825-PE, Prepared by Poverty Reduction and Econom.ic Management Sector Unit, Latin America and the Caribbean Region. Washington D.C. 2005b. Peru Rural Electrification -Project: Economic and Financial Analysis. Washington D.C. (unpblished). - - . 2005c. Project Appraisal Document Peru Rural Electrification Project. Washington D.C. Available at www.worldbank.org. 161 Special Report Series RENEWABLE ENERGY THEMATIC AREA East Asia and Pacific Region (EAP) Sustainable and Efficient Energy Use to Alleviate Indoor Air Pollution in Poor Rural Areas in China (002/07) Latin America c:md the Caribbean Region (LCR) Nicaragua Policy & Strategy for the Promotion of Renewable Energy Resources (003/07) Global (GLIB) Considering Trade Policies for Liquid Biofuels (004/07) ENERGY POVERTY THEMATIC AREA Globai (GLB) Risk Assessment Methods for Power Utility Planning (001/07) South Asia Regeon (SAR} Restoring Balance: Bangladesh's Rural Energy Realities (006/09) Latin America and the Caribbean (B.CR) Peru: National Survey of Rural Household Energy Use (007/10) ENERGY SECURITY THEMATIC AREA G!obal (Gl!.B) Coping with High Oil Price Volatility (005/08) 163 Energy Sector Management Assistance Program (ESMAP) Purpose The Energy Sector Management Assistance Program is a global knowledge and technical assistance program administered by the World Bank and assists low-income, emerging and transition economies to acquire know-how and increase institutional capability to secure clean, reliable, and affordable energy services for sustainable economic development. ESMAP's work focuses on three global thematic energy challenges: • Energy Security • Poverty Reduction • Climate Change Governance and Operations ESMAP is governed and funded by a Consultative Group (CG) composed of representatives of Australia, Austria, Canada, Denmark, Finland, France, Germany, Iceland, Norway, Sweden, The Netherlands, United Kingdom, and The World Bank Group. The ESMAP CG is chaired by a World Bank Vice President and advised by a Technical Advisory Group of independent, international energy experts who provide informed opinions to the CG about the purpose, strategic direction, and priorities of ESMAP. The TAG also provides advice and suggestions to the CG on current and emerging global issues in the energy sector likely to impact ESMAP's client countries. ESMAP relies on a cadre of engineers, energy planners, and economists from the World Bank, and from the energy and development community at large, to conduct its activities. Further Information For further information or copies of project reports, please visit www.esmap.org. ESMAP can also be reached by email at esmap@worldbank.org or by mail at: ESMAP c/o Energy, Transport, and Water Department The World Bank Group 1818 H Street, NW Washington, DC 20433, USA Tel.: 202-473-4594; Fax: 202-522-3018 THE WORLD BANK Energy Sector Management Assistance Program 1818 H Street, NW Washington, DC 20433 USA Tel: 202-458-2321 Fox: 202-522-3018 Internet: www.esmop.org Email: esmop@worldbonk.org