Mozambique - Household electricity utilization study

10~ ~ ~ ~ ~ ~ ~~t/S -V3 lb Bo&sector a tp 0~~~~~Z- 11 -t0 C liej""" - ~Mozambique Household Electricity Utilization Study Report No. 113/90 -Il~~~~~~~~~~~~~~~~ , . t .~~~~~~~~~~~~C C'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  !- { ij 27 ' cfz i I e ; ' 1 - - . ' *~~~~~~~~~~ ~ ~~~~~~~~~~~~ K ,) t:[ X S | :'~~~~ '' J--~~ i U "'  ACRONYMS BADEA Arab Bank for Economic Developmet in Africa DANIDA Danish ILtei;atonal Developmen Authority DOE Depatment of Energy EDM Electricidade de Mozambique FSMAP Energy Sector Mnagement Assistce Program ODP Gross Domesdc Product SIDA Swedish International Development Authority IDA -Iational Devdopment Association UNDP United Nations Developmen Program ABBMEVATIONS CR one-hundredth of a Rand (i) g gram GWh Gigawatt-hour (million kWh) kCal kilocalorie km klHometer kV kilovolt .',ouand Ivolts) kVA kilovolt-ampere kW kilowatt kWh kilowat-our I litre LPG Liquid Petroleum Gas in5 cubic meter MCal Megacalorie (million calories) MT Meticais (Mozambican currey) MW Megat (thousad kW) MWh Megawatt-hour R Rand (South African currency) toe tons of oil (equivalent) v volt ,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  0 / o t)~~~~~~~~ , _, , - 2Us'. Ofica ra. a.o.Mrh199 '>~~~~~~~~~~~~~~~~~~~~~~~~~~~~- C 0 .,~~~~~~~~~~~ a~~ . . 'I~~~~~~~  MOZAMRIQUE: HOUSEHOLD ELECrRICIrY UTIIZATION STUDY TABLE OF CNF1 L SUAONDURY ........ .............................. 1 IL I 1RODUCIION ............................................. S CountrySituation ............................... S Enery Profile and Issues * * * * .......................... . 6 Household EnerVyPriodties andStrateV .............. ... 7 III OVERALL PROJECT OBJECTIVES AND IMPLEMENTATION ..... 8 IV. ElECTRICTIY SUBST1TUON FOR COOKING: THE DEMAND SIDE 9 Household Energ Profe,Study Methodology .............. 9 Mvfajor Fndinis ........ ...................... 11 Estimated Household Fuel Costs and Income Levels ......... . 12 Comparative COSt of Cokin Fuel Options ..... ........... 13 Co diton EistoEety SucltsSubtuiutin.................... 13 V. ELEC`TRCITY SJBSIITUTION FOR COOKING: THE SUPPLY SIDE 18 Study Approach .................................. 18 New NetworkDesign and Panning ..................... 19 Eiting Network Requirements ........................ 21 Estimated InvestmentCost .......Co................... 22 VL FOLLOW-UP ACTIONS ................................ 24 VIL CONCLUSIONS AND RECOMM]ENDATIONS ..... .............. 25 TABLES Tble 4.1 Comp"atiVe COst ofFuel forCooking .................... 14 Table 5.1 Estimated Ranges ot Houshold Conmtions and Elect Consumption ........................... 20  q- l~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ANNESES Annex 1: Household Ener Consumption Su ......................... 29 Annex 2: A. Energy Coefficients Used in This Work ................ 38 B: Energy Sources and Use . ......................... 39 C: Household Eergy Consumption in Mozambique .......... 40 D: Average nerg CoDsumption per Household Per Month ..... 41 Annex 3: Compartive Costs of Household Fuels: Basic Assumptions ....... 42 Annex 4: Price list of Equipment Required for Distribution System Renforcement and Expansion .................. 44 Annex 5: Estimated Ranges of Consumers, Electricity Consumpon and Related Distribution Systems Ivmetments by Area ....... 45 Annex 6: Estimated Electricity Consumption Levels and Investment Requirements for the Matola Network ................ . 46 A I~~~~~~~~~~~~~~~~~~~.  L SUMMARY 1.1 This report on the Household Electricity Study for Mozambique (bereafter known as *tie ESMAP Study") was prepared under the Joint UNDP/World Bank Energy Sector Management Assistance Program (ESMAP), and consolidates the flndings of (a) a survey on household energy use to ame the demand conditions for electricity substitution; and (b) a preliinary evaluation of requirements and costs for reinforcing and expanding the electricity distribution system to meet increaed electricity demand. The higher electricity demand is expected to result primarily from the expanded access of households to the electricity grid and the promotion of electricity use for cooking, to replace traditional fuels such as wood and charcoal. -1.2 This project emerged from an initial finding in the Mozambique Energy Assessment 1/ that electricity, and in certain cases coal, v"ere the least cost fuels for meeting household energy needs in Mozambique. Subsequently, the Government of Mozambique requested further assistance from ESMAP to evaluate, in more detail, the conditions for electicity substitution. The findings of 'this evaluation were to be incorporated in the Government's program for increasing the availability and reducing the cost of household fuels. The Govermnent of Sweden, tbtough the Swedish Inteaional Development Authority S[SIDA), agreed to provide the funds required to meet te direct cost of the study. 1.3 Some of the findings of the ESMAP Study lave already been incorporated in the on-going Inteational Development Agency's (IDA) Urban Household Energy Credit which supports a multi-fuel project to assist the Government in the implemenation of its strategies for meeting basic household fuel requirements. In addition, the International Development Association (IDA) and other cofinanciers have provided a credit for the implementation of a project to be known as the 'Urban Household Energy T*oject" during the implementation of which it will be possible to take action on more of the study's recommendations. 1.4 Chapter II gives a brief overview of the country situation, profiles the energy situation and describes the Government's household energy priorities and strategy. The key features of this section are: (a) The economy, strongly dependent on agriculture, has declined during the 1980s, constraining the availabilty of foreign exchiange for imported energy along with the spare parts and equipment needed to restore domestic ii "Mozambvque I*sues and Optionsim the Enerr Sectoe, Jwanuy 1987. World Bank Energy Dewnnt Report Na 6128-MOZ  -2- energy supply facilSties. These have been ravaged by internal armed conflict. (b) Households play a major role in the overall energy picture. They consume over 90% of the total energy used in Mozambique, mostly in the form of woodfuel (fuelwood and charcoal). This resource is badly degraded (particularly around urban areas) and cannot sustain present levels of urban consumption in the long term. (c) Altough 80% of the population lives in rural areas, accelerating rral-to- urban migration combined with deforestation, and disruption of energy supplies due to armed conflict, has led to severe and Increasing urban household energy shortages, estimated at 10-30% of consumption in Maputo, the capital. (d) The Government recognizes that an expanded woodfuel supply program alone will not solve the problem of energy shortages. Hence the Government has developed a multifaceted fuel strategy designed to improve access to all forms of energy in the short term, in order to ensure that there is sufficient supply to-meet basic household energy needs, while at the same time concentrating on the development of the most cost-effetive energy options-electricity and coal-for the long term. 1.5 Chapter m outlines the specific objectives of the ESMAP Study. Tfhese objectives, which concentrate on the feasibility of expanding the use of electricity for cooldng, were to (a) -conduct a survey of existing and potential domestic electricity consumers, gathering data on specific conditions which would have an impact on electricity use and (b) evaluate the condition of the distribudon system and estimate investment requirements needed for expansion to meet increased demand. 1.6 Chapter IV, based in part on the household energy survey, concenaes on the demand side of electricity substitution for cooking. It highlights (a) survey findings on household energy consumption pattern; (b) the comparative cost of fuels; and (c) the burden of energy coms in relation to income levels. The details of the study methodology are given in paras. 4.1 to 4.8. 1.7 Ihe findings of the household survey cover (a) fuel-specific consumption patterns cross-referenced to sample survey are by status of linkage to the grid; (b) cooking pracdces and tpes of equipment used; (c) specific conditions surrounding the potential subtitution of electricity for fuelwood and other fuels for cooking (i.e., ownerShip and use of electrical appliances in households already linked to the grid, familiarity of households not linked to the grid with electricity in the home); and (d) fuel prices. Information - war gathered by direct  -3- questioning of households in three general geographic areas, (a) Maputo's so-called "formd Cement City" (hereafter referred to as the Cement City), the older, well-etablished sections of Maputo, including the higher- and middle-income sections; (b) 'Informal Maputo (ereafter referred to as Informal Maputo), the newer, low-income housing areas of the city where standards of building constrution vary widely from concrete houses with permament roofs to dwellings with walls of cane and roofs of thatch; and (c) Quelimane. This city consists predominantly of dwellings similar to those in Informal Maputo. The income levels and standards of living of the two areas are similar. In the Cement City nearly all households are connected to the electricity grid but gas is the predominant fuel used for cooking. In Informal Maputo, woodfuel Is the main is the main cooking fuel but 85% of households use other fuels as well. In Quelimane, woodfuel is used almost exclusively. .8 The comparative economic cost of various household fuels indicates that electricity and coal are the least-CoSt fuel options. However, the immediate financial cost of electricity is high, requiring an initial outlay equivalent to about US$450 for electricity connection and stove purchase. In investigating means by which the access of households to electricity could be improved, IDA has estimated that a loan program could reduce the annual cost of cooking with electricity to the equivalent of US$94, including amortized cost of connection and fixed-rate financing for stove purchase. By comparison, the estimates of cooking with traditional fuels range. from US$96 to US$230. These latter costs (n local currency) are wholly subject to inflation, which in the Mozambican economy is currently running at 30%. While even the projected lower cost is still quite high given the country's official per capita income of US$170, a strong underground economy is known to exist which would presumably alleviate some of the cost burden There is no reliable data on income from the parallel sector. 1.9 Chapter V discusses the supply side of household eiectrification, i.e., the reinrcement and expansion of the electricity distribution system. A range of possible consumption figures has been determined and estimates of the upper and lower range of investments required have been made. The high forecast implies increases in current consumption (about 7,400 MWh per annum in the sections of Maputo surveyed) amounting to (a) 10% of the present level from existing consumers and (b) a near doubling of the present level from the connection of new consumers. According to the findings on load levels and investment requirements, it would be possible to increase the present energy consumption by 25% with an investment of US$50 per MWh. Furthermore, a load increase of 50% would require investments of US$75 per MWh while investments of US$90-100 per MWh would allow almost an unlimited increase over the present load. The electrification target zones are intended (at least initially) for the middle- to upper-middle income levels (skilled manual workers) of the present non-electrified households, with the Maputo subdivision of Matola as the first pilot project for the distribution otimization plan. 1.19 Chapter VI briefly outlines follGw-up actions in response to the study's findings, which are being carried out the Government of Mozambique under the Urban Household Energy  4- Project financed by a credit totalling U$50 million provided by IDA, the Nordic Development Fund, The Arab Bank for Economic Development in Africa (BADEA) and the Danish Inational Development Authority (DAN)IDA). (See para. 6.1 and 6.2). 1.11 Chapter VII presents the conclusions and recommendations of the ESMAP Project. The main conclusion is that expansion of access to electricity and promotion of its use as a substitute for woodfuel in cookling will have important benefits, the most significant of which is lowered energy costs, in addition to Increased fuel availability and reduced envitonmental degradation. Possible constraints to the successfil implementation of this program are: the initial cost of electricity connection and stoves; the present condition of the distribution system; the fuel preferences of electrified and non-electrified households; and the fact that a signiflcant share of non-electrified households are not familiar with electricity use in the home. The Government, under the Urban Household Energy Project, is taking steps to minimize some of these constraints by: (a) creating a loan program for financing the eosts of electricity connection and stoves (b) reinforcing the power distribution system; (c) reviewing and developing low-cost safety criteria to modify dwellings for electrification; (d) identifying initial priority areas based on the ability of target households to pay for electricity; and (e) the launching of an extensive Government program designed to educate the population to the advantages of cooking with electricity. 1.12 It would be worthwhile to do some "fine-uning" of the demand evaluations to take account of the above concerns. This is particularly important for households cooking mainly with fuelwood. Becase of the importance of cooking with electricity in maintaining planned load forecast leels, it would be usefl for the Government, perhaps under a cm t of the Urba Household Energy Project, to conduct additional survey work to take account of specific questions on electric stove use or non-use, cooking habits, utensils employed, and staples consumed in order to fine-tune its education campaign, which is crucial to the success of electricity subsitution for cooking.  1L INTRODUCTION 2.1 Mozambique 1s a low-income country with an annual per capita income of about US$170 and a population of some 14 million. Situated on Africa's largest- southern coastal plane, Mozambique occupies 799,380 square kIlometers Oarger than England, France and Portugal combined). It is bordered by Tanzania on the north, by Swaziland and the Natal Province of South Africa to the south; and on the west by Zimbabwe, Zambia and Malawi as well as the Transvaal province of South Africa. The eastern part of the country lies on the Indian Ocean (Mozambique Cbannel), wbicb separates the mainland from Madagascar. 2.2 The economy of Mozambique s based on agriculture whicb accounts for nearly half of the cointry's gross domestic product (GDP) and 70% of foreign exchange earnings. More than 80% of Mozambique's population lives in rural areas but urban growth bas accelerated in recent years. The population of Maputo, the capital, bas increased from 375,000 in 1970 to well over one million today. The rest of the country's cities are thinly populated; only two other cites, Beira and Nampula, have populations over 100,000. 2.3 The economy has experienced extreme disruption during the past decade - between 1980 and 1986 overall production fell by 30%, exot declined by abou 75%- ad imports decreased by about one-third. Former high Import levels, combined with collapses in commodity and service exports, led to large balance of payment deficits. In order to finance these deficits, Mozambique resorted to foreign borrowing and seriously eroded its foreign reserves. Several major factors bave inhibited the country's economic growth inctuding (a) a legacy of former colonization (high imports of consumer goods); (b) a severe 4op in traditional foreign exchange remttances from Soith-Africa; (c) continuing interna i wfare since independence in 1975; and (d) natal disastert affecting agricultural production. Inadequate development policies, instiutional weaknesses, and higb!y disorted prices frther contributed to the deterioration of the economy. 2.4 In order to improve the country's economic condition, the Government launched an economic recovery program in 1984. The principal focus of the program was on immediate, short-term measures to allocate scare foreign exchange to sectors of the economy which would produce a quick production response and foreign exchange earnings (agriculture, consumer goods, etc). At the same time, the program included preparatory measures for long-term sustained economic growth such as: (a) improvements in the institutional and economic policy environment, (especially for small-holder agriculture); (b) critical assessment of public enterprise operation; (c) a review of the country's manpower and training needs, and (d) development of a public  -6- expenditure program that would facilitate a more efficient allocation of resources and reduction of deficits. E=gy Proflle and Issues 2.5 The decline in economic actiAity during the early eighties led to stagnation in energy demand. In 1984, the latest year for which total energy consumption figures are available, total energy consumed was around 3 million tons of oil equivalent, about the same level as in 1981. Households are the largest consumers of energy in Mozambique, accounting for more than 90% of total energy consumption, most of which is in the form of woodfuels (fualwood and charcoal). Woodfels are in short supply in urban areas and command high prices due to extensive deforestation around the major cities as well as the fear of armed attack, which has hindered fuelwood gathering in forested areas. Furthermore, the country's economic situation has limited the availability of alternate fuels. In particular, shortages of foreign ex:hange have limited petroleum supplies as well as the spare parts and equipment needed to rehabilitate distribution systems for petroleum and electricity. Lack of maintenance and attacks on the local population and the electric power infrastructure have disrupted the country's two main facilities for domestic commercial energy production - the Cahora Bassa hydropower stadon and the coal mines at Moatize - which are virtually inoperative because of lack of resources to transport the energy produced. 2.6 The Mozambique Energy Assessment Report highlighted the 'ollowing issues in the energy sector, which the Government agrees needi to be addressed urgendy: (a) Disruption of energy supplies, costing the economy about US$25 million anually i foregone foreign exchange earnings from exported energy (hydropower and coal); (b) A backlog of maintance and rebabilitadon work on energy facilities due to the lack of sufficient foreign exchange for spare parts and equipment; (c) The diminished effectiveness of energy institutions due to skilled manpower shortages, deteriorating energy facilites, and the diversion of management expertise away from dealing with long-term management concerns to handling daily crisis situations; and (d) Forest degradation and the constraints to increasing production of domestic alternative fuels, such as electricity and, in certain areas, coal, leading to the need for costly imports of kerosene and LPG to meet basic household energy needs in the short term.  -7 2.7 The Government of Mozambique is addressing the task of improving the reliability and efficiency of the supply system for electricity and petroleum products. Its efforts are being assisted by a credit provided by IDA and known as the Energy Technical Assistance and Rehabilitation Project. This project also is providing assistance to improve fuelwood and charcoal production capacity and reduce the costs of transporting woodfuels. At the same time the Project is eamin various alternatives for fuelwood supply such as woodlots, small-holder planting, large plantations, and the improved management of natura woodlands. During 1989 the Project completed an aertdphotog,aphic survey of wood resources. The results of this survey and other related studies will be incorporated into the household energy strategy described in the following paragraphs. Household Anergy Priordties and Strategy 2.8 The disruption of energy supplies, combined with increasing deforestation, has led to severe and growing household energy shortages, estimated at 10-30% of energy needs in Maputo. A major priority for the household energy sector is to increase the availability of household energy to urban areas: particularly woodfuels, but also LPG, kerosene, and electricity. The Government is committed to this priority and is developing programs to increase fuel availability and reduce overall fuel costs for a large number of households. 2.9 The household energy supply strategy is multi-faceted. Recognizing that an enhanced woodfuel supply program alone will not solve the problem of household shortages in the near tem, the Govenment is planning a rapid increase in household electrification, the test- marketing of coal stovs, and improved efficiency in the use of traditional fuels such as fuelwood and charcoal. The objectives of the Household Electrification Project, are oudined in para 3.1.  8 - m. OVERALL STUDY OBJECTvES AND MLEMETATION 3.1 Given fuelwood shortages and the need to reduce Imported energy over tie long term, the Government's household energy strategy is concentrating on two domestic sources of supply which appear to be low-cost options: electicity and, in some areas, coal. As part of a stegy to investigate these options, the Government requested assistance fom the Joint UNDP/World Bank Energy Sector Management Assistance Program (ESMAP) to focus on the feasibility of the dectricity option, by preparing a Household Electricity Utlization Study which would (a) conduct a survey of potential domestic electricity consumers; and (b) evaluate the condition of the electric distribution system and estimate the investments required for expansion and reinforcement in support of anticipated increaes in household electricity demand. 3.2 The Government of Sweden, through SIDA, provided the funds required to meet the direct costs of the study. Accordingly ESMAP arranged for the execution to begin in June 1988. 3.3 Tle study was implemented in two phases. First, a survey was conducted of household energy consumption patterns, with particular attention to the characteristics of fuel use which would have an impact on potential electricity substitution for other fuels in cooking. In parallel, a report was prepared which estimatd investments required for the electricity ditribution system based on the energy demand patterns reported in the household energy surey. Tbe second phase consisted of evaluating the information, preparing a supplementary analysis within the Bank on comparative fuel costs, and consolidating the findings into a single completion report on the Project.  9- IV. ELECTRICIIY SUBSTITUTION FOR COOKING: TH DEMAND SIDE Household Enery Profile. Study Methodology 4.1 The Household Energy Survey ZI was carried out in July arid August of 1988 in the cities of Maputo and Queliman. Its main purpose was to investigate household energy consumption patterns with a view towards the potential substitution of woodfuel for electricity in cooking. The methodology consisted of a household survey and the preparation of an analytical report on the data obtained. The survey was composed of 48 questions designed to investigate energy types and sources, cooking equipment, pricing and energy use preferences of inhabitants in the survey. an English translation of the questionnaire for the households surveyed is given in Annex 1. 4.2 The study surveyed the consumption of firewood, charcoal, gas, kerosene, and electricity. Electricity use was estimated from the month's utility bill. To obtain data on other fuels, the households were asked how much of each were consumed, and these figures were converted to monffly values in terms of their calorific content. The energy content of firewood and charcoal consumed was determined by weighing typical bundles purchased and converting the weight in kilograms to calorific content at a rate of 4350 kcal/kg for wood and 6900 kcal/kg for charcoal. a/ Survey samples were allocated to each of the selected areas in direct proportion to their respective populations, which were estimated from 1970 census figures and adjusted for population growth, including recent migration patterns. 4.3 The demographics of the survey concentrated on two cities: Maputo and Quelimane. Total sample sizes for Maputo were 193 households in the Cement City and 461 households in Informal Maputo. Quelimane's sample was composed of 98 households, resuting in an aggregate total sample of 752 households surveyed. The average household size consisted of seven people. There was no questionnaire module estimating household income or expenditure, becase of the presence of a strong underground economy. Respondents were reluctant to provide information which would facilitate developing useful ata on the contribution of the underground economy to disposable income. Income levels were therefore implie4 by the type and physical ciistcs of the dwellings surveyed. Within the city of Maputo itself, two y/ -HoiehoWd Energy Cosmpto In Urban Mozambique' by Dr. Aldo Bens, ESM4P Consdiant November 198 i/ Taken fom "ENERGM. Baancos (1980'86) e Projecoes (198-890)', by Jose Lopes, Depaym of EneV, MJnfsty of Indsy and EnerV, Maputo, MOZambiqu4 Ju4' 1987. See also Anna 11 A.  - 10- distinct sections of the city were selected: the upper middle income, or Cement City; and six low-to-middle income "informal" neighborhoods. Maputo was selected because it is by far the most important city in Mozambique and is represemntative of the rapid rural-to-urban migration and related problems of fuel scarcity, high prices and the deforestation of the surrounding savannah. A further consideration in the selection was the capacity to implement alterate energy supply policies, particularly electricity substidtion. Quelimane, the other city in which surveys were undertaken, was chosen as being represntative of provincial cities in Mozambique and additionally for its ability to provide Electricidade de Mozambique (EDM) with insttutional support to conduct the survey. 4.4 Thn areas chosen for the study were those thought to be representative of locations with gxod potential for electrification as well as for improving distribution of commercial fuels. These areas would therefore not be representative of the overall position in these cities. 4.5 Due to time, personnel and data constraints, the survey did not rely on a pure probability sampling form. All available listings of shanty town dwellings were incomplete, many were overlapping, and none was accessible witbin a reasonable time frame. Aerial photographs of the city were more than a decade old and were used to assess the accuracy of maps used. Students would have been helpful as enumerators in the study but they were unavailable at the time. The electric utility (EDM) could only offer the services- of meter readers during the afternons of off-peak weeks and part of the weekend. Because the survey was carried out during the winter, when daylight hours are short, daytime surveying was restricted to about three hours per day. Since lighting conditions were poor in many areas and there was the danger of armed attack after dusk, the survey was not conducted in the evening. 4.6 Given the above restrictions and time constraints the survey used the combination of a cluster and a systematic probability sample. All sampled areas were selected by EDM and the Department of Energy (DOE) of the Ministry of Industry and Energy to reflect both long- term energy planning interests and medium-term electrification capacity. The number of neighborhoods selected by survey areas were as follows: four in the Cement City, six in Informal Maputo and five in Quelimane. 4.7 The total intended questionnaires were allocated to each of the selected neighborhoods in proportion to its estimated population. Homogeneous geographical clusters were randomly selected within each indicated neighborhood and interviews were apportioned among the clusters based on dwelling density, estimated by in situ inspection. Within each cluster, the dwellings themselves were selected at random, through a clustering point and a sampling interval designed to exhaust the sample.  i ~~~~~~- 11 - 4.8 The findings of the household energy study consit of data on (a) fuel-specific consumption patterns cross-referenced to geographical areas by status of Lnkage to the electricity grid; (b) cooking practices and types of equipment used; and (c) specific conditions relating to the potential substitution of electricity for other fuels in cooking, I.e., fuel demand patterns, use of electrical appliances in households already electrified, the degree to which households not connected to the grid were familiar with electricity, etc. 4.9 The detailed data in support of the major findings are attached, for reference, as Annex 2, and are arranged in the following categories: A. Energy Co-Efficients; B. Energy Sources and Use; C. Household Energy Consumption; D. Average Energy Consumption per Household per Month. This annex summarizes, from the household energy survey data, general trends relevant to further electrification and the salient features of household energy consumption patterns in the selected areas. 4.10 The average total monthly consumption of primary energy in the households surveyed in Maputo is about 1760 Mcal per month compared to 600 Mcal in Quelimane. The higher consumption in Maputo is believed to be due to higher incomes, larger family sizes and under-recording of fuelwood freshly collected in Quelimane. 4.11 Due to the unreliability of fuel supply, households in Maputo tend to use several types of stoves and a variety of fuels. The choice of fuel depends to a large extent on its availability. Although most households surveyed were able to specify their main cooking fuel, there is a strong possibility that the wide range of cooking alternatives in Maputo may have affected the accuracy of replies to questions on the specific amounts of fuels used. 4.12 The fuel consumption patterns by survey area may be summarized as fullows: (a) In the Cement Cit, although nearly all households are connected to the electricity grid, the majority of residents (55%) cook mainly with gas while about 36% use mostly electricity for cooking. Fuel substitution occurs mainly between gas and electricity, though about one third of the population resorts to the use of charcoal as a secondary cooking fuel. (b) In Informal Maputo, firewood is the main cooking fuel, used by about 60% of the population. At the same time, almost 85% of the population uses other fuels as well. About 48% of households are linked to the electricity grid, but only 15% of households cook mainly with electricity. Charcoal is the main substitute fuel for non-electrified households. Gas is the main cooking fuel for 14% of the population. Kerosene is rarely ised for cooking, and is the main cooking fuel for only 22% of the Informal Maputo population.  - 12- (c) In Ouelimane, only 13% of households are connected to the electricity grid and households cook with firewood almost exclusively (97%). Only one household reported cooking primarily with electricity, and one other family reported cooking mainly with kerosene. The main substitu fuel is charcoal, disclosing a heavy dependency overall on woodfuels for cooking. 4.13 There is an apparent anomaly which should be noted, in the comparison of firewood consumption for households in Informal Maputo and Quelimane. Average monthly firewood consumption in Informal Maputo, for a household not connected to the grid and cooking mainly with firewood (sometimes substituting other fuels), is 1348 Mcal. This figure 1i more than double the amount used by the average household In Quelimane cooking almost exclusively with firewood (624 Mcal). The household energy survey takes note of this anomaly and recommends further research to determine the reason behind the differeace. Possible explanations for the discrepancy include (a) the underreporting of f&elwood consumption in Quelimane by the underestimation of wood gathered freely' (i.e., not purchased in the marketplace); Lb) fuel price differentials in the two areas; (c) regional differences in the moisture content of the woods used in the two areas; (d) regional differences in dietary and cooking habits; and (e) larger income and family size in the city of Maputo, leading to higher levels of consumption. Estimated Household Fuel Costs an Income Levels 4.14 Available data indicate average monthly fuel costs in Informal Maputo to be 13,000 to 14,000 Meticais (MT). Data indicated monthly fuel costs for Quelimm'ne to be somewhat less - from 5,000 to 7,000 MT. The lower cost of fuel in Que1imane is believed to result from a number of factors including smaller family size, less use of total energy, lower use of charcoal, the highest-priced fuel and particularly the gathering of fuelwood at no financial cost. 4.15 'rhe household energy survey did not assign income levels to the suey respondents which would have given some indication of the consumers' ability to pay for various fuels. One reason why income levels were omitted is the existence of a strong underground economy for which there is no reliable data on eanings and initial investigations indicated that those surveyed would be reluctant to provide information related to the underground economy. In the country's per capita income and the salary levels given in the latest available economic report on Mozambique (1984), the average monthly fuel costs for Informal Maputo, given in para 4.14, would amount to about US$220 on an annual basis (in 1989 terms) which exceeds thfs current per capita income of US $170. In addition the 1984 economic report cites salaries ranging from 2,100 MT per month for the average worker to 10,000 MT per month for university graduates. Even if the formal meticais income of the average worker has increased tenfold since 1984, fuel costs seemingly would absorb most of it. Presumably income from the parallel sector assists in easing some of this burden.  - 13 - Comaratve Cost of Cknf Fuel Otons 4.16 A comparison of the economic cost of various fuels for cooking confirms the findigs of the Mozambique Energy Assessment that coal and electricity are the least-cost fuel alternatives for cooking to the economy (Table 4.1). Electricity from domestic hydropower actualy is the least-cost alternative over the long run. A comparison of financial costs shows the actual annual cost to the consumer of cooking with electricity to be substantially lower than cooking with the traditional woodfuels, which are currently widely used. It is also somewhat lower than cooking with ksene and LPG, which are often not available. The assumptions used in preparation of these cost comparisons are attached as Annex 3. 4.17 Calculation of the financial cost of electricity use to the consumer must also take into account the high capital cost of household connection, wiring and stove purchase, estimated to be in the order of US$450 equivalent, of which US$300 would be costs to the power company for system reinforcement and the balance of US$250 payable by the householder. This figure includes intal wiring, sockets, lamp bulbs, a stove, appropriate cooking vessels and about $100 of EDM's costs for the meter, meter box and service connection. In order to reduce the initial high cost, it is estimated that loan programs would bring the annual cost of electricity for cooking to US$94 (including the cost of energy as well as US$40 towards loan amortization) compared to US$96-230 per year required for woodfuels. In addition to the cost advantage, electricity has the further advantages of convenience, versatility in use Oighting, cooking, refrigeration) and lack of adverse impact on the enviroment. Coal is another viable fuel alternative and its financial cost to the consumer, as indicated in Annex 3, is actually the lowest among the various household fuels. Coal use currently is very limited but the testing of coal stove utilization has met with some success. However, a large-scale exp~nsion of coal use probably will be limited by environmental concerns and the low current rate of stove assembly. 4.18 The comparison of cooking fuel costs therefore indicates electricity to be a viable option for reducing household fuel costs, providing a clean, versatile source of enery and for utilizing Mozambiq4e's vast supply of low-cost domestic hydropower over the long term. The following section will highlight some of the main factors, aside from cost, which must be considered in a program for expanding electricity supply. Conditions for Electricit Substitution 4.19 Overview. The household energy survey revealed certain characteristics of household energy consumption and other conditions of electricity substitution for cooking which merit careful consideration in the preparation of programs for electricity expansion. These conditions relate to: (a) the ownership and use of stoves in electrified households; (b) increases in electricity consumption with stove use; (c) familiarity with electricity; (d) preferred uses of electricity once households are connected to the grid; (e) the opndition of dwellings relative to  -14- COHPSWARXVZ COSTS OF FUELS FOR OOW R Economi Cost ax Wuel Anna nLt cout * quivalent Appliace Effectiv cost per per t.o.e. UffLeieny cost per t.o.e eusehold Power - fre Cahore "sa 1. 3 cIIkWh 70 65 105 8 - from B8CCt 6.2 *RIkWh 323 65 498 40 (cost at Maputo city gate) Coal - from South Africa 53 Rlton 36 20 182 1S (cost at )oputo) - local coal 48 USD/ton l 76 20 380 S0 (cost at Beira) LPG - ex-distrtbutor $400/ten 376 45 836 67 terosene - edLtributor $23I/ton 233 25 932 75 Financial Cost of Fuels Annual aitCost BEqulvalent Applane Effeotive Cost per per t.o.e. Effioiency cost per t.o.e Household Uls$ Xu (US8}8S) Power 20.9 Nt/kWh 433 65 666 54 (price at the household) Coal (delivered to the bairro) 20 T/kg8 56 20 280 22 LPG - ex-distributor 210 NT/kg 334 45 742 59 Kerosens - ex-distributor 110 HS/liter 233 25 932 75 Wood - ex-distributor 22 NT/kg 97-12S 8 1210-1560 96-125 Charcoal - ex-distributor 107 MT/kg 270-345 12 2235-2885 180-230 Source: Staff Appiaisal Report: "osambique Eousehold Energy Project, Way 12, 1989. Report No. 7611-MOZ. _ Notes: Exchange rate used US$1 M NT 380. (a) All costo are based on October 1988 prices. (b) Annual iost is based on fuel use for cooking. - Not lncludlng the cost of power eonnemtionldistribution, lMhc. is BestLmated at about US$40 per annum extra. )~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  standards for dectricity connection; and (f) degree of occupant stability in dwellings, an important factor in the consideration of a credit program to cover the cost of investment in electricity connections. 4.20 EL_;tw Th ere is a fairly low level of familiarity with electricity in the non-dectrified households of normal 1aputo and Quelimane - only 44% of the population in Informal Maputo and 27% of the population in Quelimane claim to know a close relativ or fMrend who has electricity at home. However the household energy did not take into account the average worker's exposure to electricity at the workplace. An extensive education campaign by the Government was launched in May of 1989 to encourage electricity use for cooking, and this is expected to greatly enhance the exposure and familiarity of the average household to electricity use, particularly with regard to cooking. 4.21 Cokig. In the Cement City, which is almost totally electrified and where income levels are presumably higher than those in Informal Maputo, only 36% of households actually use electricity as their main cooking fuel although 70% own electric stoves. The indications are that contibutory factors to the relatively low usage of installed electric stoves include (a) unavailability of spares and (b) low system voltage during the hours in which most cooking is done. In Informa Maputo only about 15% of households linked to the grid currently cook mainly with electricity. This is roughly equal to the share of the population owning stoves. One apparent reson for the low level of electricity use for cooking (considering that about half of the households are connected to the grid) has been the extremely high cost of the stoves (US$120) which were frmerly purchased from Tanzania in accordance with a bilateral trade agreement. The equivalent stove retails in the United States for US$40. In addition, spare parts for the stoves were unavailable and the cooking utensils employed were unsuitable because they burned out. These and other considerations, such as the quality of the electricity supply, must be taken into account if fiuther survey work is carried out which will investigate the reasons for the low levels of electricity use for cooldng. Households connected to the grid in Informal Maputo cooking mainly with electricity use an average of 198 Mcal of electricity per month compared to 105 Mcal per month in households which primarily use other fuels for cooking. Therefore the switch to cooking with electricity may effectively double electricity usage in the hosehold sector. 4.22 Buiding Stanards. Depending on the sandards to be decided on by the Government to define the suitability of buildings for electricity connection, some dwellings in Informal Maputo and a large number of the dwellings in Quelimane may require modification to qualify for electricity connection. In frmal Maputo, 62% of houses not linked to the grid have wails made of clay bricks. Most other dwellings have walls made of cane. However, nearly al dwellings in nformal Maputo have permanent roofs made of zinc. in Quelimane, 90% of houses not linked to the grid have adobe walls and 93% have roofs made of palm leaves. Apparently well-preserved adobe walls do not prove a hazard to electricity connection since 54% of the electrified houses have adobe walls. By contrast, 38% of electrified households and 03% of non- electrified households -have palm-leaf roofs. Hence it may be advisable to review low cost  - 16 methods of modifying these dwellings for safe electricity service in order to maximize the extension of electrification in these areas. 4.23 End-Use Preference. For the households in Informal Maputo and Quelimane not connected to the grid, the first preference for electricity use would be for illumination (89%) and secondly for cooking (66%). However, as indicated in para. 4.26 and 4.27, more households with electricity own refrigerators than stoves. Interestingly enough, the munber of households responding to the question on cooking was only 64% of those responding to the question on illumination. And the number of households responding to the possibility of using electricity for refrigeration was less than half (47%) of those responding to the questions on ilumination. Even more iesting is the actual data on the ownership of electrical appliances, which reveals a larger share of households with refrigerators and electric irons than those with electric stoves (Annex 2, Sections C.5, C.6, and D.ll). This seems to indicate that refrigerators have priority in user preference over electric stoves, and candidacy to a loan package which included rfrigerators afwr stove purchase repayment has been completed could enhance the incentive to households for increased overall use of electricity. 4.24 Lighting. A survey of illumination in Maputo and Quelimane revealed that non- electrified households used kerosene and electrified households used electricity &3 their main source of lighting. Field experience indicates that wick lanterns are most frequently used for kerosene illumination and these are owned by 97% of households in Quelimane, 78% of those in Infmal Maputo, and 33% of those in the Cement City. The high rate of kerosene use, particularly since nearly all of the Cement City and half of Informal Maputo are electrified, was reportedly due to unavailability of replacement lamps and bulbs. 4.25 Most households use incandescent lamps in the range of 40 to 60 watts although more than half of the Cement City reported having at least one fluorescent lamp and a quarter of the households claimed to have more than one. Location has no apparent impact on the power of light bulbs used but there are differences in the total power installed for lighting, indicated by the number of electric light bulbs instlled per household. Over half the households in the Cement City have six lamps per house, more than twice the amount in Informal Maputo. Since the survey indicated that dwellings in the Cement City were not significandy more spacious than those in Informal Maputo, the former apparently desire and can afford higher illumination. 4.26 Refrieration and Air-Conditioning. The use of refrigerators and air conditioners i8 concentratd mainly in the Cement City - about 76% of dwellings have electric refrigerators and 11% have air conditioners. In comparison, 19% of dwellings in Informal Maputo and 10% of those in Quelimane own refrigerators while ownrship of air conditioners is negligible. 4.27; For the purpose of electrification, it is important to note that the rate of refrigeator ownership is significandy higher than stove ownership In all of Maputo and Quelimane. Thus it appears that use of electricity for refrigeration is preferreid to its use for  g~~~~~ -17- cooking. Furthermore, ther ae not many other alternative fuels for tefrigeration, except for a very sml mm*er \ of refigerators opated by krose ,which are used only by about 4% of dwellings In ioimal Mpo hus in pling for further electrification It should be noted that households which use electricity for cookg hisorically have first sed electricity for referaion; and that this preferece by the losal populace may merk some ation in providing ap inenive plan with appropriat credit faciities aer initial requirements of sove pu,rcb and use ase d. satfed Q3 *)~~~ "I\~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- 3- . KY~~~! * 7~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,  - 18 - V. ELEUFRICITY SUBSTITON FOR COOKINGs THE SUPPLY SIDE A=at9ch 5.1 The results obtained from the household energy survey indicated that the next steps in assessing the electricity option for cooking should be to: (a) evaluate the electrical power distribution rehabilitation and expansion requirements, along with related investment costs and (b) determine the cost of using electricity for cooking relative to the costs of other fuels. Accordingly additional investigations these areas were undertaken by a separate team of consultants during August and September, 1988. 4/ 5.2 Cost estimates were prepared for the extension and reinforcement of Mozambique's power distribution network and a new optimized network for sections of the Matola area which are not yet electrified. A sample load forecast was prepared for selected areas of Maputo and for Quelimane. The forecast assumed that the total number of households using electricity would consist of all households presently electrified plus all the remaining households now within reach of the distribution network but not yet connected to the grid. 5.3 The study selected a total of 7,012 households for Maputo, consisting of 3,824 existing household connections and 3,188 potential connections of which a little more than one -third would be in the new, optimized network proposed for the Matola area. The households included in the survey correspond to about 14% of all urban households. For Quelimane, the forecast covered 85 existing consumers and 621 potential consumers. Projected consumption was extrapolated from available data supplied by Electricidade de Mozambique (EDM). The data indicated an average annual consumption of 1,900 kWh per domestic consumer. Of the seven areas selected for Maputo, the Liberdade area in the Cement City was found to have the highest average consumption, about 2,300 kWh, where presumably there is some use of electric stoves and refrigerators E/. At the same time, the Lnhagoia area was taken as representative _of "other areas' which had neither stoves nor refrigerators, with an average annual consumption of 1,800 ji Report enti `Mocambique JlotseholdElectck Utiation Project by SwedPowerABA Stockhol* Swedenm March 1989. This level appears low for the use of both stoves and refrigerators by househol The low figure may be due to the fact that a number of households wth eleciic stoves use them not at all or only part of the tima The electyic consumpton forecast assumes a maimum annual electricity consunm per household of about 4,000 k#h based on elctriciy being the main fuel for cooking.  - 19 - kWh. This figure appears to be high for just lighting and must include the use of other electrical appliances. 5.4 The load forecast developed a range of consumptior. figures in order to estimate the upper and lower limits of investment required to reinforce and expand the network. The average annual consumption assumed is in a range of 1800 kWh on the low side to 4,000 kWh on the high side. The estimated number of consumers in each area and the range of their consumption are shown in Table 5.1. With the exception of the area selected in Matola, which is not electrified, the table indicates three cases of electricity consumption: (a) the present consumption level; (b) incteased household consumption by existing connections alone; and (c) increased consumption of existing and new consumers combined. However the demand projections and calculations of investment requirements are not based on the expected maximum household demand of 4,000 kWh for all possible new consumers as it is considered unlikely that such high average loads wIll occur in the near future. 5.5 The investment requirements for expanded electricity use were evaluated in each of the seleced areas separately. For areas which had not been electrified, such as Matola, newly optimized low-voltage networks were designed. According to the evaluation, all electrified areas would require reinforcement. The cost comparison of various design alternatives, as well as the final calculation of investment requirements, was based on estimated average prices for equipment prevailing in the electricity industry at the time the report was prepared in 1988 and is shown in Annex 4. A major component of the standard equipment package, the twisted overhead low- voltage cable, is commonly used in many countries with three-phase 400 V low-voltage systems. This type of cable presently is being used in the outskirts of Maputo and one area of Matola. It may be possible to manufacture the twisted overhead cable in the Beira cable factory but a reliable price for a locally manufactured cable could not be obtained for use in the investment calculations. It may also be possible to manufacture other components locally but given the uncertainties surrounding production capacity and prices for locally manufactured equipment, investment costs were based on international prices which include assembly, erecion, and other related costs. New Network Design and Planning 5.6 For new networks, an area of Matola (Unidade H) was selected as the model for the detailed design of a new, optimized distribution network. The optimization program takes the present values of the investment costs and the costs of losses for a 13-year planning period, representing about half the economic life of a typical distribution system. The loss evaluation is based on the assumption that the load will increase by 10% annually until the thirteenth year, when the load is expected to reach a maximum consumption level of about 3,000 kWh annually per household. Sensitivity of cost comparisons to variations in the average annual consumer  '20' - n _ ''' ~~~~~~~TABUS 5. ]'J ; ESTIM&TID RANGES OF HOUSEHOLD CONNECTIONS AND ELECTRICITY CONSUMPTION Number of Average Annual Total Area Notes Consumers Consumption Consumption ...kWh--- -.-NWh--- Chamanculo A/ 594 1,800 - 1,069 594 4,000 2,376 Q/ 779 4,000 , 3,116 Inhagoia / 817 1,800 1,471 , k/ 8117 4,000 3,268 1,343 -4,006 5,372 Liberdade A/ 782 2,300 1,799 - hI 782 4,000 3,128 816 4,000 3,264 Masxaquene J/ 478 1,800 860 47878 '4,000 1,912 .551 4,000 2,204 Polana A/ 595 1,800 . 1,071 *I ( j/595 4,000 2,380 1,780 -4,000 7,120 Xiamanine .. 558- 1,800 1,004 558 4,000 2,232 LI 635 4,000 2,540, Matola A/ 1,108 0 0 hI ~~0 0 -~---0 L W 2 A / ~1 108 4,000 4,432 OUELIMAE 8/ 85 - 1,800 153 J , / o 85 4,000 - 440