Privatesector P U B L I C P O L I C Y F O R T H E The World Bank FPD Note No. 6 April 1994 Solar energy answer to rural power in Africa Robert van der Plas In Kenya, more rural households get demand and enough money to afford their electricity from solar energy than it. More households could afford it if from the official rural electrification the Government removed “obstacles” program (REP). The spread of the photo- like high duties which make the solar voltaic systems has been recent, rapid systems more expensive than they and market driven. The service is as need to be. Their removal would be good as the grid for the low power loads good rural pre-electrification policy. that prevail in many rural areas, and With easier access to solar power, the the price is competitive with other op- Government could achieve higher liv- tions for low loads. Customers can buy ing standards more quickly for more it “off the shelf”. They don’t have to Kenyans than it can through the exist- wait the years it may take for the REP ing rural electification program. to come to their own area. REP expan- sion is slow because funding is tight. Thriving market The program consists of extending the Today more than 1 MW of photovol- already existing national electricity grid, taic power has been installed in Kenya. establishing decentralized generation Around 20,000 households have pur- capacity in combination with a local chased solar energy for their homes grid, or helping auto producers of pow- compared to the 17,000 connected to er to start serving surrounding house- the official rural electrification pro- holds and small commercial enterpris- gram. Demand for photovoltaic sys- es. New connections already require tems (PV) has grown exponentially substantial government and donor since the mid 1980s when Kenyan outlay. The rate at which the program entrepreneurs realized that photovol- reaches households would have to in- taics could meet rural demand to op- crease ten-fold for it just to compensate erate electric lights, radios, televisions, for current rural population growth. At or stereos often at a lesser cost than present only 0.5% of rural households grid connections, systems driven by have access to the grid. The chances of generators, or by using kerosene and getting a grid connection through the (drycell) batteries. The market grew as REP look remote. Therefore, solar power mainly rural-based electricians linked is a serious and fast option for those up with urban businessmen, and rural households with limited power formed business agreements with solar Industry and Energy Department Vice Presidency for Finance and Private Sector Development electric suppliers primarily in Nairobi. There are Rural grid electricity supply is not without its own presently about eight companies based in Nairobi set of problems either, and power outages or brown- who supply this market. Each company has scores outs occur rather frequently. However, the differ- of agents based in rural areas who market, install ence is that the latter happen at irregular times while and maintain systems. Solar electric systems are suc- the photovoltaic electricity supply users know in cessful despite taxes and import duties, which com- advance when their supply is running out and that bined amount to more than 30% of the final price they should use electricity more sparingly. users pay, and despite the absence of credit facili- ties. At present, the household systems range from From an efficiency point of view, PV systems score $150 for a 10 W one-light system to $1500 for a 100 well because they don’t depend on an energy source W five or six-lamp system with radio and television other than the sun, and the lamps they use are in- connections. Consumers can buy their systems (one variably the latest technology (tubular or compact or more photovoltaic panels, battery, charge control- fluorescents) available. These lamps are some five ler, and lamps, wires & switches) straight off the times more efficient than the incandescent lamps shelf. The systems are flexible in design and can be normally used in rural areas on the grid. Although configured to the user’s wishes. When the electrical fluorescent lights can be used with the grid also, this output is not sufficient more solar panels can be pur- does not happen in practice because of the high chased at any time. cost and the problems of voltage drop associated with the grid connection. Solar power makes sense Solar power can match the grid service partly be- Solar systems are price competitive with the rural cause rural households do not consume much elec- grid and the genset (a small kerosene/petrol gen- tricity—at least for the first few years. Household erator set) assuming a low level of electricity con- electricity usage for those just connected to the grid sumption and an “equal” level of service. Like the is often limited to a few lamps (3–6), a radio, and/or grid, the genset option is capable of satisfying a television, or 30–60 kWh per month. These services higher demand for power and energy than the can also easily be provided by solar systems. House- rural user normally makes. The capital cost is sim- hold demand remains low for at least a couple of ilar to a 50 W solar electric system, but gensets years, after which households will start to add high- need imported fuel to operate and their service er power consuming appliances, such as rice cook- life is much shorter. Annualized costs of a 750 W ers, tea kettles, or refrigerators if the grid capacity genset (amortized over 5 years) plus fuel and main- allows. Though solar electric systems are modular tenance (to deliver 40 kWh/month, at the pump and more photovoltaic modules can be added, ther- price) amount to $195. The annualized cost to the mal applications (such as cooking and heating) are consumer of a grid connection with a higher watt- not feasible as this would be too expensive. However, age capacity (amortized over 30 years) and elec- thermal applications are often not possible either tricity consumption of 40 kWh/month at the pre- with grid based rural electricity systems as their vailing tariff amounts to approximately $51. By generation capacity is normally limited. comparison, the cost is about $61 for the 50 W solar electric system, assuming a service life of 20 Only in periods of extended rains or cloudy condi- years and the purchase of a new battery every tions that last entire days will the solar supply falter. other year. Rural households would easily spend Although under these conditions photovoltaic sys- this amount on drycell batteries and kerosene for tems do generate electricity, it is considerably less lighting—it represents less than 5 liters of kero- than on sunny days and users must be aware that sene and 10 drycell batteries per month—but get they should not draw down too much electricity so a lesser service for it. they avoid permanent damage to their batteries. Therefore, in the case of a light load, a grid connec- 2 tion and a solar electric system are comparable cent. People who use candles or kerosene therefore from a users’ point of view. For lighting only, solar typically limit themselves to using only enough light power is the more attractive option. The cost of points to enable walking around the home; such lighting (the average incremental cost per Klmh, practices usually do not provide enough illumination or kilo.lumen.hour) is some 10% lower for the solar to read comfortably. Hence, although people in ru- electric system than for the grid: 2.0 vs 2.2 KSh/ ral areas do have access to modern forms of energy, klmh. The cost of providing equivalent lighting from their use of these conveniences is limited because kerosene wick lamps is 10 times higher than either of this expense. The levels of services many rural the grid or the solar systems. households “enjoy” now is only barely distinguish- able from that of medieval Europe. Rural households Since most rural households will not have their homes will never obtain a level of service even comparable connected to the electricity grid in the forseeable to the urban population unless they get access to future, photovoltaics will often be the next best al- electricity in a different form than drycell batteries. ternative. This remains true only so long as house- holder appetite for electricity remains limited to a Obviously, households with a solar electric system few light points plus a radio and/or TV connection. at the moment are not among Kenya’s poorest. They The following table illustrates the circumstances in are part of the middle income class. But they are in which solar power is a viable option. exactly the same target group that receive large sub- sidies under the REP. The REP, run by Kenya Power and Lighting Company (KPLC), involves a subsidy for each new rural connection. These subsidies are Electricity Options largely funded by the Kenyan Government and aid load/demand donors. provided by low medium high The cost of bringing power to the vicinity of new rural customers under the REP is on average PV yes maybe no KSh 55,000 (US$ 720). New rural customers only pay KSh 2,500 ($38) each as a connection fee, though genset no yes yes new customers have to for pay the step-down trans- former (possibly shared with neighbors) and the cost of the distribution wires to their homes as well grid no yes yes as in-house wiring. This policy of reduced connec- tion fees would be justifiable if there were a provi- sion in the electricity tariff to clawback the subsidy. However, that is not the case. Rural consumers have Better policy a preferential electricity tariff, cross-subsidized by Better access to electricity would make a big differ- urban consumers. This public policy preference ence to the quality of rural life. Many rural people position—subsidies for the REP and taxes for PV do consume conventional modern energy, but they equipment—does not make sense. do so in small amounts. They also pay for it dearly: drycell batteries provide electricity for about $3 to The REP is a slow-track method of electrification. $10 per kWh. A candle or a kerosene wick lamp While the KPLC has succeeded in electrifying all does give a high quality light, but households need, 42 district headquarters, little rural electrification respectively, about 60 or 20 of them to obtain the has occurred outside of these towns. In fact, less same amount of light emitted by a single 60 W in- than 30% of urban and 0.5% of rural Kenyan candescent lamp or a single 12W compact fluores- households have access to electricity. In other 3 individual households. It would make sense to do Kenya energy consumption (1989)— so, particularly given the low population density in low share for electricity combination with the small demand for power and energy in rural Africa. The solar power option is an woodfuels 71% effective first step. And it can be market driven rath- er than depending on cash-strapped governments or aid donors. Governments should be encouraging this option rather than penalizing it by taxing im- ported components. Future Action petroleum fuels 22% Rural householders would benefit if the Kenyan coal 1% renewables 2% electricity 4% Government removed at least four significant im- pediments to the solar power market. First, a more Source: Ministry of Energy, Kenya rational (comparable to rural electrification program equipment) import duty and tax regime should be applied. Second, competition should be further en- couraged by opening the market to all potential words, the outlook for rural connections under ex- investors. Investors are presently hampered by re- isting policy is fairly bleak since even urban electrifi- strictions on access to foreign exchange. Third, cation has not progressed very far. financing mechanisms should be put in place to make solar electric equipment more accessible to a Although the actual numbers and details differ, simi- larger share of the population. Fourth, technical lar circumstances are found in many other African standards for solar electric systems should be es- countries. Despite substantial amounts of money tablished and applied. Solar electric companies that have been invested in the African power sector, have been compromising technical standards to off- only a minority of African households are able to set the increase in price after the recent devaluations enjoy modern services provided by electricity. The of the Kenyan Shilling. More frequent equipment majority of rural households have no grid-based failures may have a long term negative impact on electricity, and will not receive it in their lifetime the development of the market. either. Government least-cost electricity extension programs usually exclude looking seriously at alter- native approaches, such as (low load) electricity Robert van der Plas, Energy Planner, Finance and Private Sector Devel- generation through photovoltaic panels placed in opment This series is published to share ideas and invite discussion. It covers financial and private sector development as well as industry and energy. The views expressed are those of the author(s) and are not intended to represent an official statement of Bank policy or strategy. Comments are welcome. Please call FPD Note line to leave a message: 202-458-1111. Suzanne Smith is the editor and production manager. Her address is Room G8105, The World Bank, 1818 H Street, NW, Washington, DC 20433. 4 5