70236 KENYA Rural Electrification Access Expansion Study June 2006 Christophe de Gouvello , AFTEG (TTL) With contributions from : J. Arungu Olende (consultant), Joel Maweni (LCSQE), Josphat Sasia (AFTTR) Document of the World Bank 1/81 ABBREVIATIONS AND ACRONYMS ASAL Arid and Semi-Arid Lands BOT Build Operate Transfer CBS Central Bureau of Statistics CDF Constituency Development Fund CFL Compact Fluorescent Lamp COFREP Coffee Farmers Electrification Project DDC District Development Committees ERB Energy Regulatory Board ERSWEC Economic Recovery Strategy of Wealth and Employment Creation ESCO Energy Service Company GEF Global Environment Facility GOK Government of Kenya IFAD International Fund for Agricultural Development IFC International Finance Corporation IPPA Interim Power Purchase Agreement km kilometer KPLC Kenya Power and Lighting Company KSh Kenya Shilling kWp Kilowatt peak LDC Least Developed Countries MoE Ministry of Energy MSEP Multi-Sector Energy Program MWh Megawatt hour NSSF National Social Security Fund O&M Operations and Maintenance PPA Power Purchase Agreement PVMTI Photovoltaic Market Transformation Initiative SACOS Savings Cooperatives SWARE Single Wire Return by Earth Wp Watt peak 2/81 SUMMARY Executive Summary ___________________________________________________________ 5 I. Introduction: Rural Electrification and the ERSWEC _________________________ 11 II. Demand Analysis _______________________________________________________ 13 1. Current level of access to electricity in Kenya ______________________________ 13 1.1. Electrification rate and Geographical distribution of connected customers ____________ 14 1.2. Geographical distribution of non connected households __________________________ 19 2. Energy and capacity requirement to achieve access goal _____________________ 22 2.1. Estimate of the number of additional rural households to be served until 2010 _________ 22 2.2. Estimate of the domestic needs for electricity services of rural households and additional productive and social uses that would be triggered by the corresponding extension of access to electricity. ____________________________________________________________ 22 2.3. Calculation of corresponding energy to be served and additional capacity required in rural areas to reach the goal of 20% of access in 2010. _______________________________ 23 3. Electricity needs in productive and social sectors in rural areas _______________ 24 3.1. Methodology____________________________________________________________ 24 3.2. Sector programmes _______________________________________________________ 24 3.2.a. Agriculture _____________________________________________________________ 25 Fish industry ____________________________________________________________ 25 Livestock 26 3.2.b. Education ______________________________________________________________ 26 3.2.c. Health _________________________________________________________________ 30 3.2.d. Information and Communications ___________________________________________ 31 3.2.e. District development plans _________________________________________________ 34 3.2.f. Water and Irrigation ______________________________________________________ 34 3.2.g. Cooperatives ____________________________________________________________ 37 3.2.h. Micro and small scale enterprises ____________________________________________ 37 Posho mills _____________________________________________________________ 38 Bakeries 38 Hotels and restaurants ____________________________________________________ 38 3.2.i. Jua Kali sub-sector _______________________________________________________ 39 III. Structure, Organization and Performance of the Rural Electrification Program ___ 39 1. Current structure of the Power Sector ____________________________________ 39 2. Institutional Setting for Rural electrification _______________________________ 40 3. Expenditures and Sources of Financing ___________________________________ 41 4. Outcomes of the RE Program ___________________________________________ 44 IV. Solar Photovoltaic Electrification __________________________________________ 45 1. The market is still driven by low cost small and low quality products ___________ 46 2. Fee-for-service arrangements as a way to make quality and size affordable _____ 48 3. Main lessons of the Kenya solar experience: _______________________________ 49 3/81 4. Integrating photovoltaic systems and grid extension in large rural electrification projects ______________________________________________________________ 49 V. Analysis of Options for Faster Access Expansions ____________________________ 50 1. Global Institutional Framework under the National Energy Policy ____________ 50 2. Options for Detailed Design of the Institutional and Financial framework_______ 51 2.1. Revised modalities for customer financial participation : _________________________ 51 2.1.a. A new tariff structure that takes into account operational costs and new customers willingness to pay ________________________________________________________ 52 2.1.b. A new connection policy that maximizes both the number of beneficiaries and revenues _ 52 2.2. A new Financing Mechanism: ______________________________________________ 54 2.3. Private Public Partnerships to leverage additional implementation capacity ___________ 55 2.4. Ensure that the regulatory framework provides for integration of new technical solutions 57 2.5. Set up a strong institutional champion for expanding access _______________________ 58 2.6. Multi-Sectoral Partnerships for maximizing ratchet effect on rural development _______ 59 2.7. Options for Public Private Partnership to scale up access to electricity _______________ 59 2.7.a. Options to increase access rate inside “KPLC domain� as defined by the French consultant proposal: _______________________________________________________________ 60 Option a: Inject more financing into current KPLC electrification procedure: _________ 60 Option b: Improve the efficiency and the pace by preparing larger volumes of projects awarded to same contractor in the same area: _________________________ 60 Option c: Delegate to larger private sector entities the densification of whole compact areas __________________________________________________________ 60 2.7.b. Options to increase access rate outside “KPLC domain� as defined by the French consultant proposal: _______________________________________________________________ 61 Option a: Turn key contracts for clusters of RE schemes: _________________________ 61 Option b: Built Operate Transfer (BOT) contracts for electrifying large compact areas: _ 63 Option c: Distribution licenses for large rural areas _____________________________ 61 VI. Conclusions and Recommendations ________________________________________ 64 Bibliography and Data Sources_________________________________________________ 68 Annex 1 – Projection of Population per District 2000-2006 __________________________ 70 Annex 2 - Distribution of households by main type of lighting _______________________ 76 Annex 3 – Field Visits : Main Findings __________________________________________ 77 Annex 4 – Map of the Kenya Electricity Grid _____________________________________ 81 4/81 Executive Summary Rural Access and Economic Recovery Strategy for Wealth and Employment Creation The Government of Kenya adopted in 2004 an Economic Recovery Strategy for Wealth and Employment Creation (ERSWEC), which recognizes three main pillars for economy recovery namely: (i) strengthening economic growth; (ii) enhancing equity and reducing poverty; and (iii) improving governance. The ERSWEC reiterates that the achievement of the three pillars is dependent on adequate and reliable access to least-cost energy. Since agriculture continues to be the mainstay of Kenya’s economy, ensuring adequate access to electricity in rural areas is an important component to achieving the objectives of the ERSWEC. This is confirmed by investigations made by this study regarding specific energy needs for the different sectors of productive and social activities in the rural areas, for agriculture, livestock, fishery, tea and coffee cooperatives, telecommunications, water pumping and health and education services. The Government of Kenya has adopted a National Energy formulated in the Sessional Paper No 4 of 2004 consistent with the ERSWEC, which set double target of a 20% access rate to electricity in rural areas and 40% in 2020. Breadth of the challenge to attain the target of 20 % of rural access by 2010 On the basis of existing data issued by the Central Bureau of Statistics (CBS) on demography and households amenities surveys and most recent KPLC statistics on domestic customers, it has been estimated that the current rate of direct access to any kind of electricity services would be around 14.5 %, of which about 8.6% is provided by the KPLC grid. Since the second number relies on population projections that may suffer from some bias, it should be used with caution. It is therefore recommended that the monitoring and evaluation plan of future access development projects should integrate mini-census to improve the quality of this estimate. However, even taking into account such intrinsic uncertainties, a large gap remains between these two figures, due to: (i) illegal and/or secondary connections, (ii) connection to private mini-grids, (iii) individual gensets and (iv) use of individual solar PV systems. Although some households manage to get some electricity services, this also means that this access is not regulated, that is, they do not benefit from any regulatory protection regarding safety standards, quality and continuity of the supply, or price charged to them for these services. Field observations and interviews of users of such alternative has led to the conclusion that these non regulated options are generally proportionally very expensive, work only a limited number of hours per day and are not reliable. Individual homes or shops served by local private mini-grid are charged around 500 Ksh per month for one single 40W bulb used only six hours per day, when the corresponding bill would be less than 90 Ksh per month (A0 category). Individual 5/81 gensets are even more expensive. PV solar systems sold on the market are of poor quality and fail quickly. The number of additional customers that should be served to reach the 20 % target in rural areas by 2010 has been estimated on the basis of Population Projections for 2010 by CBS. If all households are to be connected as new legal customers, about one million new rural connections should be done by this date. While these figures seem impressive, the characteristics of the Kenya population settlement may help a lot to achieve large volumes of connections. Kenyan population is concentrated in less than one third of the national territory, as a result of high density of population in the western part of the country. This density is mainly due to the historically high concentration of population in the rural areas of the highlands facilitated by the exceptional climatic conditions and quality of soils; in certain districts, the rural density is more than 500 inhabitants per square kilometer. This is also a region where the main grid is already in place, which should help to increase the electrification rate relatively quickly at a relatively low cost by densifying the existing distribution grid. While detailed demand survey is lacking and should certainly be done as a necessary preliminary step to prepare any ambitious rural electrification program, rough preliminary estimates indicate that the additional capacity required to serve these one million additional rural customers plus some productive uses could be around 70 MW. The need for a new model to scale up access However, past performance of the current institutional and financial arrangement for rural electrification over the last three decades indicates that such an ambitious goal would be out of reach without reshaping it deeply: As of end FY2004/2005, the cumulative number of new customers connected by the Rural Electrification program established in 1973is only 101,793. The limitations of current RE institutional model have been pointed out by several studies and include institutional and management constraints, depletion of the Rural Electrification Fund by operating losses worsened by the allocation formula and by inadequate tariffs, high unitary investment costs, and a restrictive connection policy that limits the integration of domestic customers. Besides the RE program, KPLC is also connecting urban and rural customers using its own resources. However, the current pace of connection, either by the RE program or by KPLC on its own, and either urban or rural, is around 50,000 new customers per year. On average, the rate of rural electrification has increased by 0.3% per year. Even assuming significant improvement of KPLC implementation capacity, which may concentrate on its core urban market, the ambitious objective set by the Sessional Paper No4 to support the implementation of the ERSWEC cannot be attained without leveraging additional resources and implementation capacity and without improving the efficiency of the use of the public resources that can be allocated for rural electrification. The GOK has already taken a number of steps to address some of the issues identified. In particular, it has prepared a draft Energy Bill which will establish a Rural Electrification Authority and a Rural Electrification Fund, and which will remove KPLC monopoly in power distribution activities by permitting the award of distribution licenses to third parties. Several studies have also been initiated to facilitate implementation of the 6/81 decisions already taken in the Sessional Paper on Energy regarding: (i) the institutional set-up for RE, (ii) the need for cost recovery tariffs; and (iii) a new connection policy for connecting customers to facilitate the achievement of access targets. Key recommendations Endorsing the Government’s decision to establish a specific institutional and financial mechanism for rural access expansion separate from the current set-up for the main system, and on the basis of lessons from international experience in this area, the study is proposing key principles that should structure this new arrangement, which are the following: Establish a new tariff structure for rural electrification that takes into account operational costs and new customers’ willingness to pay. On one hand current tariffs do not even cover O&M cost, and on the other hand, results from the few households surveys and field visit observations are supporting the evidence commonly faced in other African countries that substitutable energy expenses are significantly higher than the bill that would be charged by applying the existing tariff corresponding to these categories of customers Adopt a new connection policy that maximizes both the number of beneficiaries and revenues. A reshaping of the actual connection policy is required to fix both the non- desired side effects of the current one and to overcome the bottleneck of the high direct and indirect connection costs. Develop Private Public Partnerships to leverage additional implementation capacity. On one hand, even if additional funding like the support provided by the French and the Spanish cooperation allows KPLC to increase its contribution to electrification, the objective of 40% rural access by 2020 seems to be far beyond even improved KPLC capacity. On the other hand, experience of former projects, like PVMTI, consultation of different Kenyan stakeholders and experience of on-going public private partnership in other countries in the region, shows that there is a number of national and international stakeholders that are interested, and sometimes even already involved (agriculture cooperatives, PV dealers, SACOS, hire purchase companies, banks, contractors, etc.), that could play a more active role to implement the ambitious access target set up by the Set up a new Financing Mechanism. It is acknowledged that scaling up access would require setting up of an appropriate financing mechanism to channel public resources that are necessary to fill the gap between the initial investment cost and the resources that can be brought upfront by the customers and the private sector. According to the lessons from the PVMTI project, the appetite of the Kenya financial market to finance the share of investment cost that can be borne by the service provider, either KPLC or other private operators, may not be enough. As a consequence, there may also be a need for setting up a guarantee fund. Complete the regulatory framework to provide for integration of new technical solutions. Besides the conventional grid, new and more cost-effective technologies are available. On one hand, low cost grid technologies like single wire return by earth (SWARE) have been tested in different countries that reduced significantly the investment cost required to serve low and dispersed loads. On the other hand, over the last decades, the technological progress has benefited more to the new decentralized electrification systems than to the already very mature grid technology. It is therefore 7/81 recommended to review the technical specifications currently enforced by the ERB and to provide an adapted regulatory framework for decentralized solutions. Set up a strong institutional champion for expanding access. One of the key lessons from other rural electrification programs in other countries is that it is essential to set up institutionally a strong and autonomous champion whose unique commitment is to expanding access. The reason for recommending such a separate arrangement is to avoid new access expansion efforts ending up being high- jacked by inefficiencies, lack of specific capacity – especially regarding innovative solutions - and different ranking of priorities (urban growth, industrial demand development) that generally rule the power sector. The new Bill prepared by the Government will establish a Rural Electrification Authority (REA). It is recommended that the composition of its board should incorporate representatives of the civil society in a proportion that guarantees its autonomy. Organize multi-Sectoral partnerships for maximizing ratchet effect on rural development. The development of social and productive uses of electricity requires coordinated multi-sectoral actions that seldom occur timely and spontaneously. Investigations made during this study have revealed that there is a considerable potential for increasing impacts of projects and programs executed on the same territory by other sectors. The reason is that these projects and programs are facing technical and financial limitations to provide adequate energy infrastructure, especially because they cannot account with economy of scale to get access to cost effective and reliable energy equipment and O&M services. Options for a new model based on Public Private Partnership The participation of private companies in developing access to electricity inside or outside the areas already served by KPLC can be sought under a range of legal arrangements which vary according to the degree of autonomy and protection awarded by the sector authorities to the participating private companies. This ranges from simple works contracts, as done currently under the RE agreement, to fully integrated licenses awarded through competitive bidding, including a number of intermediary options of which advantages and limitations can be examined in the specific geographical and legal context of Kenya. Three potential options for Public Private Partnership to increase access rate outside “KPLC domain� have been determined by the study, which are the following: Option a: Turnkey contracts for clusters of RE schemes: This is the model currently being implemented under the Spanish and French funded programs. Instead of bidding limited number of schemes to each contractor, a large number of schemes (dozens) are bid at the same time. Advantages: This option has already been road tested through the Spanish and French funded programs, and as such will benefit from the lessons learned when implementing these programs. Limits This option allows neither the leveraging of additional private human and financial resources nor the bringing in of innovation related to decentralized solutions. Option b: Distribution licenses for large compact rural areas 8/81 Distribution licenses for compact rural areas are awarded through a competitive bidding process to the candidate that would commit for the higher number of connections against the pre-determined envelope of subsidies. The committed amount of connections is to be (i) built under a limited time and (ii) operated and maintained during the whole license period. Revenues are collected directly by the local rural electrification company that the winner will be requested to create in the license area. The choice of the technology is of the responsibility of the licensee, to the extent that each technology used complies with technical specifications or standards to be attached to the tender document. Advantages: The size of the area bid and the perspective of a series of tendering, the protection offered by the license, which ensures that revenue from customers can be secured over a period long enough to recover the investment, and the economic viability facilitated by the first cost subsidy can attract large private companies that are reliable over the long term, able to bring in significant amount of financing and able to achieve quickly ambitious targets. The competitive process will ensure that the candidates will optimize the use of the public subsidy. To win, international companies tend naturally to team up with local ones to reduce costs, as observed for instance in Senegal. Limitations and risks: This model supposes to elaborate quite a complex procurement process and to prepare a number of documents that are part of the tender package. This assumes that the REA receives technical assistance to prepare the first bidding package. Detailed preliminary market surveys for each license areas are important to provide enough information to candidates to prepare reliable proposals. Credibility, especially regarding financial and regulatory commitments from government, is essential. It should be clear that expected cash flow is adequate to cover more than O&M cost. It is worth note that all these conditions have been met by the Rural Electrification Agency of Senegal, where economic power of rural seems to be less than in Kenya. Option c: Build Operate Transfer (BOT) contracts for electrifying large compact areas: This option is similar to the previous one but differs with it to the extent that only functions are awarded but not licenses; KPLC remains accountable to the regulator. After several years of operation during which the winner of the bid collects the bills and maintains the equipment he has committed to install, the assets and the activity are transferred to KPLC. Advantage: No license needs to be issued, which simplifies the process. The choice of technology and the marketing can be innovative and more cost effective, and thus more customers are served than is the case in a scheme where the number of customers does not result from competition. As for the previous option, the long period of the contract generates a cash- flow that allows it to leverage private financial resources to be brought by bidders to maximize the number of connections. 9/81 Limitations and risks: This model assumes that the duration of the contract is properly determined ex ante on the basis of business plan simulations to ensure a reasonable financial return to potential bidders and that KPLC provides enough energy to the licensee’s grid for him to serve his final customers. Since the selected firm is a subcontractor and not a licensee, it has no formal link with the regulator and its contract with KPLC is not directly regulated. Private companies will commit themselves and engage money only to the extent they trust KPLC’s capacity to honor the contract over such a long period. The risk is that KPLC would interfere in their relation with the clients and thus no firm wants to engage money ; as a result the leverage effect would be nil. The risk exists when assets are transferred to KPLC when it lacks experience in managing a very different scheme. This option is very similar to the arrangement that has been implemented in Morocco by the national public utility, ONE. However, KPLC doesn’t have the same cumulated 20 years experience on solar PV systems that ONE had before setting up such an arrangement. Taking into account the specifics of Kenya and the possibility of benefiting from technical assistance based on previous experience, the study concludes that option b would be the most efficient and less risky option. 10/81 I. Introduction: Rural Electrification and the ERSWEC In the past two decades, Kenya has experienced a slowdown in growth levels across all economic sectors, and increased poverty from 48 percent of the population in 1990 to about 56 percent in 2003. During that period, agriculture grew by just about 1 percent. Manufacturing sector faced stagnating investment and negative productivity, and the quality of delivery of energy and other infrastructure services declined. With a relatively small urban population (around 25% only), Kenya remains mainly a rural country, and agriculture continues to be the mainstay of Kenya’s economy, accounting of 24% of the GDP in 2005. In order to turn economic performance around, and to improve the well- being of the population, the new Government that came into office in December 2002, has formulated an Economic Recovery Strategy for Wealth and Employment Creation (ERSWEC) 1. ERSWEC key policy objectives are: accelerated economic growth and employment creation; increased productivity across all sectors; equitable distribution of national income; reduction of poverty through provision of basic services to the population; and improved rural / urban balance. The ERSWEC reiterates that the achievement of the objectives is dependent on adequate and reliable least-cost energy. This implies that growth in energy supply to various sectors has to be at the minimum, commensurate with the rate of expansion of the economy. Otherwise it would become a constraint to economic growth and the achievement of the ERS objectives. The recent recovery of agriculture is acknowledged to be one of key elements, which allowed the Kenyan GDP to grow at an accelerated pace. The 2005 Economic Survey registers a 5.8 per cent growth, up from 4.9 per cent in the previous year; the survey shows that in spite of the worst drought, farmers raised their production by 6.7 per cent, reflecting increased activities in the agricultural sector, and improved investment, not only in the sector, but also in other sectors in the rural areas. As a consequence, ensuring adequate energy supply for the growth of activity in the rural areas is an important component to achieve the objectives of the ERSWEC. This is confirmed by the investigations made by this study regarding the specific energy needs for the different sectors of activity in the rural areas. Main conclusions, which are detailed in section II.3 below, are the following:  Agriculture, Livestock and Fishery Sector: While more analytical work would be needed to get a more precise picture, consultation of experts in the Ministry of Agriculture led to the conclusion that electricity is required to improve productivity for: (i) food processing (including post harvest processing); (ii) irrigation; (iii) food preservation (in particular refrigeration and cooling); and (iv) horticulture. Electricity is also required for fish preservation –cooling, drying, and processing. Just as the fish industry has been growing steadily, so has the requirement for energy in this sub-sector. The lack of cooling services, has, for 1 The ERSWEC is the Kenya’s variant of the Poverty Reduction Strategy Paper (PRSP) 11/81 example, resulted in the loss of many catches with attendant loss in incomes, profits and livelihoods. The poultry sub-sector is also a fast growing user of energy, although data on the energy requirements in the sub-sector is not readily available  Cooperatives Sector. There are two main types of cooperatives in Kenya: (i) those for production and marketing and (ii) those for saving and credit also known as (SACOS). There are hundreds of production cooperatives operating in different parts of the country, covering such areas as coffee (900 coops), cotton, dairy (400 coops), horticulture, pyrethrum, etc. Most of these cooperatives would like to apply electrical processing equipment that adds value to their products: pulping for coffee factories, ginneries for cotton, cooling tanks for dairies, etc. For example, substitution of diesel by electricity for the pulping process has been for years identified as a key productivity gain for Kenyan coffee producers.  Education Sector : the Education Sector has witnessed tremendous growth since independence; the number of students enrolled at various levels of education has substantially increased. Enrolment in formal public primary schools grew from 891,533 pupils in 1963 to 7.2 million pupils in 2004 due primarily to the introduction of Free Primary Education. The number of primary schools has increased over the years to stand at 19,587 in 2004. Boarding schools, in particular, require electricity for lighting, heating and cooking, computer applications and laboratories. Only 20 percent of the schools are connected to the grid, the remaining 80 percent rely on generators. Currently less than 5 percent of the primary schools have electricity, the schools in rural areas rely on generators, paraffin oil, and wood.  Health Sector. There are eight provincial hospitals, 75 district hospitals, 460 health centers and 1,600 dispensaries in Kenya, most of which are located in rural areas. Only about 5% of dispensaries have access to electricity and about 50% of the health centers are connected to electrical supply. Of the district hospitals about 90% are connected to electricity while about 90% have stand by generators.  Water Sector. Electricity is required for pumping water from the bore holes which is currently done by diesel operated water pumps. Many boreholes have been dug in Arid and Semi Arid Lands (ALAS) but have had to be abandoned for lack of pumps. This is one more example of unmet demand.  Telecommunication Sector. The communications sector is a major consumer of electricity in the rural areas, most of which are supplied by the service providers, who put a high premium on reliability and quality of their services to the customers. They have hundreds of repeater stations in the rural areas throughout the country. Because of the need to cool some of the equipment at the stations, electricity requirements at the stations are relatively high. Consequently the service providers have to make major investments to meet the electricity demand for their stations, but without benefiting from the economy of scale that could generate a large electrification program. Communications service providers are keen to work out modalities for KPLC to connect many of their rural stations, bearing in mind their stringent reliability and service quality requirements. 12/81 The objective of this study is then to try to quantify the challenge that the ambitious target of 20% access rate in rural areas by 2010 would mean, and, considering the institutional and financial arrangement currently in place for rural electrification, to explore possible ways to face such a challenge. II. Demand Analysis 1. Current level of access to electricity in Kenya This section presents the figures regarding direct access to electricity in rural and urban areas estimated from the different sources currently available in Kenya. Access is defined here as an individual access at household level to a grid, a mini-grid or any individual systems generating electricity (individual genset, individual solar PV system, etc.). KPLC statistics provide monthly updated figures regarding number of customers served by RES schemes or directly by KPLC and distinguishes between rural and urban.. These statistics are organized in 4 main regions: Nairobi Region, coast, West Region and Mount Kenya Region. The four regions are subdivided into eleven sub-regions and detailed according to 59 smaller areas (see Annex 1 - KPLC Statistics - Number of customers per region and Sub-region, April 2006). On the other hands, CBS statistics provides the distribution of rural households and urban households by districts and provinces in 1999 and demographic projections until 2010 (see Annex 2 - Projection of Population per District 2000-2006, Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistic). On that basis, it is possible to estimate electrification rates at two different point of time, e.g. 1999 and 2006. While the focus of this study is more on rural access, it has been possible to calculate a separate estimate of the electrification rate for rural and urban areas for the year 2006. It should be noted that is has not been possible to check precisely if the definitions of rural areas / population used by CBS and by KPLC were identical or not2. As a result the split in these two categories is to be used with caution3. For that reason, the global urban+rural figures are always presented together with the disaggregated ones. Perhaps more important, it has also been possible to estimate the number of households that are not yet connected to the grid, both in the urban and rural areas. Both connection rate and unconnected households are detailed for the 7 provinces. This required a cautious work of re-aggregation of detailed KPLC data per administrative province. So 2 For instance, all KPLC customers of the 3 districts of the North Eastern Province (Mandera, Wajir and Garissa) are counted as rural in KPLC statistics. 3 Over the past three decades during which RE has been undertaken in Kenya, some locations that were originally “rural� have developed and become “urban�. As the same time, there are rural schemes that have been undertaken by KPLC as normal expansion of its network in response to customers demand. Such connections are not captured in the statistics because the data collection and retrieval system only recognizes “rural� to mean Government’s RE Program. 13/81 far it was not possible to get disaggregated data at the districts level, since it has not been possible to re-aggregate the detailed KPLC data per districts4. However, since the electrification rate remains very low in most of the districts, the simple demographic data is also useful to get a sense of the distribution of non connected households at the district level. This information is also presented in annex 2. It should also be noted that the KPLC figures do not capture illegal connections or secondary connections, which would lead to higher figures, nor other means of access to electricity (private mini-grids, individual gensets and PV). However, a specific CBS Report on Housing Conditions and Households Amenities dated 1999 provides the Distribution of households by main type of lighting (see annex 3), which include electricity independently of the origin (grid, local mini-grid or individual genset) and individual solar PV systems. While this data is not very recent, it gives a sense of the relative importance of other means to access electricity services. 1.1.Electrification rate and Geographical distribution of connected customers From 2000 to 2006, the total number of KPLC customers has increased from around 500,000 to around 800,000 (see table 1 below). This includes all categories of customers, including small, medium and large commercial customers, industrial customers and street lighting. Domestic customers (urban+ rural) correspond only to tariff categories A0 and A0&D0. These two categories, which together represent 81.5% of the total number of KPLC customers, determine the number of households that are getting access to electricity through KPLC grid (Rural Electrification Schemes + KPLC connections). This number has increased from around 400,000 to 650,000 during the period 2000-2006. Annual growth of domestic connections varied between 6 to 8% per year during the period with a peak of 12% in 2002, which means around 50,000 new domestic connections every year5. Table 1: Evolution of Domestic Customers – 2000-2006 Tariff Category(KPLC+RES) 2000 2001 2002 2003 2004 2005 2006 AO only (Domestic) 356,541 378,751 430,580 468,848 504,535 548,251 594,691** AO & DO (Domestic) 44,581 47,279 46,424 49,582 50,609 51,168 51,540** Total Domestic 401,122 426,030 477,004 518,430 555,144 599,419 646,231** Annual growth 6.2% 12.0% 8.7% 7.1% 8.0% 7.8% A0 as % of total 70.5% 70.5% 72.5% 72.9% 73.5% 74.6% 75.0%* A0&D0 as % of total 8.8% 8.7% 7.8% 7.7% 7.3% 6.9% 6.5%* A0 + A0&D0 as % of total 79.3% 79.3% 80.4% 80.6% 80.9% 81.5% 81.5% Total customers all categories 505,951 537,079 593,621 643,274 686,195 735,144 792,922 * estimated; ** calculated as a percentage of the total Sources: KPLC, Annual Report and Accounts 2004-2005, and KPLC database for April 2006 total figure. See also detail of KPLC statistics for all categories of customers in Annex 1. 4 KPLC data follows a specific geographical division, which cannot easiliy compared with CBS district based data. 5 The 12% peak of growth between 2002 to 2003 corresponded to 49,076 new connections, the 7.8% growth between 2005 to2006 corresponded to 46,812 new connections. 14/81 Not surprisingly, KPLC domestic customers (all categories) concentrate principally in the Nairobi Region (41%). While also quite concentrated, the distribution of rural customers is significantly different, since 30% of them are located in the Rift Valley Region and 23% in the Central Region (see table 2). The comparison of these figures with CBS statistics, especially population projections for the year 2006, gives an estimate of the access rate nationwide and for each province, both for rural and urban population. However, two caveats need to be stressed with respect to the population projections results. First, the further we move away from 1999 into the future, the less reliable the projections' results become. And, second, internal migration plays a major role on sub- national projection outcomes. This is so because, obviously, migration patterns will directly influence sub-national populations but also because migration rates applied to sub-national populations bring a bias of their own in the calculations. Indeed, if one apply a positive migration rate to a sub-national entity, its population will grow more rapidly than in reality because you apply a positive rate to a growing population - in fact, it generates a snowball effect ; the opposite is true with negative migration rates and the population decrease faster than in reality. But since the latter situation is less frequent in the case of Kenya (only 3 districts out of 70 are deemed to have negative rates), the use of these projections may induce a slight over-estimate of the global population projections that I used, and thus a slight underestimate of the rate of access. Therefore, these results should be used with caution, and it is recommended that, when large projects will be prepared, the monitoring and evaluation plan integrates mini-census of selected areas to improve the quality of the access rate determination and the measurement of it evolution along the project. Table 2 : Regional distribution of domestic customers (A0 and A0&D0 categories only) - April 2006) Urban connected Rural connected Total connected % of total % of total No. of urban No of rural No of % of total Provinces households households households households households households non Nairobi 263,656 50% 0 applicable 263,656 41% Central 49,878 10% 27,543 23% 77,421 12% Coast 74,912 14% 13,752 11% 88,663 14% Eastern 28,189 5% 21,778 18% 49,968 8% North non eastern 0 applicable 5,025 4% 5,025 1% Nyanza 25,049 5% 10,964 9% 36,013 6% Rift Valley 70,489 13% 36,572 30% 107,062 17% Western 12,067 2% 6,356 5% 18,423 3% Kenya 524,240 100% 121,991 100% 646,231 100% 15/81 Source: KPLC database note: in KPLC statistics, all domestic customers in Nairobi (city) are considered urban, and all domestic customers located in the North Eastern province (Garissa, Mandera, Wajir) are registered as rural. Considering that the number of connected households is accurately measured by the number of A0 and A0+D0 KPLC customers, this gives an estimate of the current national electrification rate of 8.65% in April 2006. This is significantly below the figure of 15% frequently quoted in different texts6. However, it should also be noted that the KPLC figures do not capture illegal and/or secondary connections, which would lead to higher figures of households effectively connected, nor other means of access to electricity (individual gensets and PV). A specific CBS Report on Housing Conditions and Households Amenities dated 1999 provides the distribution of households by main type of lighting (see annex 3), which includes electricity independently of the origin (grid – legal or illegal connections - , local mini-grid or individual genset) and individual solar PV systems. While this data is not very recent, it gives a sense of the relative importance of other means to access electricity services. According to this survey, the proportion of households having access to some electricity services - either as a registered KPLC customer or through any other mean – was 14% in 1999. This means that only 61% of people accessing electricity are getting it officially through KPLC. This also means that the rest about 39%, are getting access to electricity under non regulated conditions without any regulatory protection regarding quality standards, safety standards or price paid. While no precise figure exists regarding the number of solar PV systems effectively working as of today, it should be noted that the number of such systems sold in the country has significantly increased (see section IV Solar Electrification). While the CBS survey counted only 28.531 households using such systems in 1999, it is estimated that the current number of such systems could be around 200,000. Such an increase represents an additional 0.5% in terms of households having access to electricity. 6 This figure of 15% seems to have been calculated by multiplying the number of customers by a too high estimate of the average number of persons per households and comparing the result it to an estimate of the total population. When using such a method, any variation of the number of persons per household – also called sometime family size – affects strongly the result. For instance, applying a family size of 8 persons per households on the total number of domestic customers (646,231) and comparing it to the population in 2006 (34,045,843) would give a rate of 15.2%, when in fact according to the CBS census of 1999, the average number of person per household is only around 4.50 (total number of households in 1999 was 6,371,370 ; total population in 1999 was 28,686,607; average number of persons per household calculated by dividing population by number of household is 4.50). Because of the high level of uncertainty attached to such a method calculating the population corresponding to customers, it is preferable to compare directly registered customers to number of households effectively counted by the demographic census. In the latter method, both the figures compared are counted, avoiding thus uncertainty attached to calculation based on a hypothesis regarding family size. 16/81 Urban electrification rate is of course higher (26.4%) peaking at 31.1% in Nairobi. Rural electrification rate is lower (2.2%) peaking at 4.21% in the Coast Province (see table 3 below). According to KPLC data for the year 2000, the number of domestic customers was 401,122, which represented a rate of electrification of 6.4% nationwide. This means that although the population has grown 12.7% in the period (see annex 2), the electrification rate – measured on the basis of registered domestic KPLC customers - has increased by 2.05%, which is 0.33% per year. 17/81 Table 3: Access to electricity in April 2006 (KPLC grid) Households connected (2) Number of Household (2006)(1) (KPLC customers statistics 2006, A0 & A0+D0 only) elec elec elec Number Number Number Rural Urban Total rate rate rate Rural Urban Total rural urban total non Nairobi 0 848,232 848,232 0 applic. 263,656 31.08% 263,656 31.08% Central 822,846 171,760 994,606 27,543 3.35% 49,878 29.04% 77,421 7.78% Coast 326,469 298,768 625,237 13,752 4.21% 74,912 25.07% 88,663 14.18% Eastern 970,210 89,595 1,059,805 21,778 2.24% 28,189 31.46% 49,968 4.71% North 195,673 35,255 230,928 5,025 2.57% 0 0.00% 5,025 2.18% eastern Nyanza 954,698 128,027 1,082,724 10,964 1.15% 25,049 19.57% 36,013 3.33% Rift 1,483,383 326,654 1,810,037 36572.31 2.47% 70489.35 21.58% 107061.66 5.91% Valley Western 734,994 84,660 819,654 6,356 0.86% 12,067 14.25% 18,423 2.25% Kenya 5,488,272 1,982,950 7,471,223 121,991 2.22% 524,240 26.44% 646,231 8.65% Source: (1) Analytical Report on Housing Conditions and Households Amenities, Vol X, Kenya 1999 Population and Housing Census, Central Bureau of Statistics, and Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistics. Number of rural and urban households in 2006 have been estimated by applying the projection of population growth from 1999 to 2006 to the number of rural and urban households established by the 1999 census. (2) KPLC, Annual Report 2004-2005. See also details regarding number of KPLC customers of all categories in Annex 1.. Table 4: Access to electricity in 2000 : comparison between connection to KPLC grid only and access to electricity including all possible means (grid, mini-grid, individual genset, PV) Access to Electricity Access to KPLC grid only (grid, mini-grid, genset, PV) (based on KPLC data, 2000) (based on CBS survey, 1999) Nb Number of Number Number Domestic Access Households Connection Regions Household 1999 electricity users customers KPLC rate % 2000 rate % (Census 1999) 1999 2000 (proj CBS) Nairobi 638,928 336,640 52.4 % 682,689 194,068 28.4% Central 912,760 149,691 16.3 % Coast 521,047 102,758 19.6 % Eastern 950,565 64,512 6.7 % North 146,176 8,072 5.5 % eastern Nyanza 956,979 49,089 5.1 % Rift 1,478,354 146,901 9.9 % Valley Western 699,016 23,859 3.4 % Total 6,301,825 881,522 14.0 % 6,636,122 401,122 6.4% Kenya 18/81 1.2.Geographical distribution of non connected households By subtracting the number of rural and urban domestic customers from CBS households, it is possible to get an estimate of the remaining population which doesn’t have access to the public service of electricity. As of today, it can be estimated that around 6.8 out of 7.5 million households (e.g. around 91.3%) have no access to the public service. Table 5 below gives the split in rural and urban population. These figures do not take into account illegal connections and sub-connections, and for that reason may be slightly overestimated. Table 5 gives also the regional distribution of this population. While the non-connected urban population concentrates principally in Nairobi (40% of total), the non-connected rural population is slightly more evenly distributed among provinces, with the exception of the provinces with smaller populations (Coast and North Eastern). As a matter of fact, these figures largely reflect how the provinces have been administratively delimitated, some being very small, others quite large. As a result, it is more appropriate to examine the settlement of these populations to better perceive where it concentrates and where the major efforts should be done to increase future rates. Table 5:Geographical distribution of households without access to the public service of electricity Households not connected to KPLC grid Regions Number Rural % rural Number Urban % urban Number total % total Nairobi 0 584,576 40.1% 584,576 8.6% - Central 795,302 14.8% 121,882 8.4% 917,184 13.4% Coast 312,717 5.8% 223,856 15.3% 536,574 7.9% Eastern 948,432 17.7% 61,406 4.2% 1,009,838 14.8% North eastern 190,648 3.6% 35,255 2.4% 225,902 3.3% Nyanza 943,733 17.6% 102,978 7.1% 1,046,711 15.3% Rift Valley 1,446,811 27.0% 256,165 17.6% 1,702,976 25.0% Western 728,638 13.6% 72,593 5.0% 801,231 11.7% Kenya 5,366,281 100% 1,458,711 100% 6,824,992 100% Source: (1) Analytical Report on Housing Conditions and Households Amenities, Vol X, Kenya 1999 Population and Housing Census, Central Bureau of Statistics, and Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistics. (2) KPLC, Annual Report 2004-2005 2) KPLC, Annual Report 2004-2005. Indeed, the geographical distribution of population in Kenya is very specific since the population settlement is highly contrasted in Kenya: while large regions remain very low densely populated – mainly the Northern part of the country – the highlands and the borders of the Victoria Lake are impressively highly populated (Western Region, Nyanza Region, the Central Region and the southern part of the Eastern Region). This can be easily observed in the maps on the next page (Fig 2 and 3). 19/81 Since the rural electrification rate is still very low nationwide (less than 4 percent), the areas where the rural population concentrates are the same where the non-connected rural households also concentrate. As a consequence, these maps give a realistic view of where rural electrification programs should focus to increase rate of electrification. The rural density is so high in these areas – it can be above 500 persons per square km - that it should be technically possible to increase very quickly and at a relatively low cost the number of grid connections. However, observation of the past trends has shown that the pace of access expansion through connections to KPLC grid has remained very low (+0.33% per year), meaning that it would require around 35 years (e.g. to year 2041) to increase the electrification rate from the actual level of 8.65% to the target of 20%. Such observations call for new modes of intervention in order to significantly accelerate this pace. 20/81 Fig. 1: KPLC interconnected grid, 2005 Fig.2: Geographical Distribution of Kenya Population Fig. 3: Density of Population in Kenya Towns (thousands of inhab) Most of the Kenyan population is concentrated in less than one third of the national territory, as the result of very high density of population in the western part of the country. This density is not the result of urbanization but is mainly due to the historically high concentration of population in rural areas of the highlands facilitated by the exceptional climatic conditions and quality of soils. As a consequence, the development of urban centers has principally occurred in that part of the country, which also triggered the concentration of Population the grid infrastructure in the same region. The Density in 1989 (inhab./ sq km) main grid being already in place, this should help to increase the electrification rate relatively quickly at a relatively low cost by densifying the existing distribution grid. (Source: Les Afriques au Sud du Sahara, Geographie Universelle, Belin- Reclus, p.331, 1994) 21/81 2. Energy and capacity requirement to achieve access goal This section presents an estimate of the minimal additional demand that would have to be served to reach the objective stated in the Sessional Paper N0.4 of 2004 of serving 20% of the rural population by 20107. For this purpose, the following steps have been followed: 1. Estimate of the number of additional rural households to be served until 2010. 2. Estimate of the demand for electricity uses: - domestic needs for electricity services of rural households - additional productive and social uses that would be triggered by the corresponding extension of access to electricity. 3. Calculation of corresponding energy to be served and additional capacity required 2.1.Estimate of the number of additional rural households to be served until 2010 On the basis of population projection data established for 2010 by CBS, it is estimated that the additional number of households that would have to be served would be around 1 million rural households. While the focus of this study is more on rural access, a similar calculation has been done for urban access, which gives an estimate of around 525,000 additional urban households that would have to be connected to reach an urban electrification rate of 50% nationwide by 2010. These results are presented in the table 6 below. Table 6: estimate of number of additional households to served to reach the goal of 20% of rural access and 50% of urban access by 2010. additional estimated nb estimated nb of targeted nb of households households to be of households households in connection already connected in connected by in 2006 2010 rate 2010 2006 2010 rural 5,488,272 5,885,219 20% 121,991 1,055,053 urban 1,982,950 2,126,370 50% 524,240 538,945 total 7,471,223 8,011,589 28% 646,231 1,593,998 Source: calculated from : (1) Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistics. (2) KPLC data 2006. 2.2.Estimate of the domestic needs for electricity services of rural households and additional productive and social uses that would be triggered by the corresponding extension of access to electricity. Precise results would require detailed surveys of domestic needs and productive and social uses in rural areas, including the determination of the statistical distribution of rural households according to different level of needs and willingness to pay; such detailed 7 The Sessional Paper No4 also states a goal of at least 40% of rural access to electricity by 2020. However, current data available, especially regarding population growth, are not enough to estimate the corresponding effort to be done. 22/81 demand analysis should be undertaken later, for instance when preparing a new access expansion project. For the time being, data available are very limited. As a consequence estimated average consumption patterns per household have been established on the basis of the limited data available, field trip observations and patterns observed in other countries in the region. This average pattern is presented in tables 7 and 8 below. Since the objective of this section is to determine the minimum additional demand to be served to reach the objective stated in the Sessional Paper N04 of 2004, it has been assumed that all new connected households would be equipped with efficient Compact Fluorescent Lamps, which is the most cost effective option8. In case CFLs are not widely distributed in future RE projects, the estimate would have to be considerably increased. Table 7: estimated average pattern for basic domestic demand per type of use in rural areas (energy and additional capacity) Nb average average energie energy max coeff of additional wattage daily use per day per year demand simultaneity capacity (W) (Wh) (kWh) (W) required (W) lamps 5 10 3 hours 150 55 50 0.3 15 (all CFLs) Radio 1 15 5 hours 75 27 15 0.3 4.5 Table 8: estimated average pattern for basic domestic demand per household in rural areas (energy and additional capacity) energie energy max additional per day per year demand capacity (Wh) (kWh) (W) required (W) total basic 225 82 65 19.5 demand 2.3.Calculation of corresponding energy to be served and additional capacity required in rural areas to reach the goal of 20% of access in 2010. One the basis of the above estimates, additional energy and installed capacity required to reach the goal of 20% of rural access to electricity in 2010 can be calculated and are estimated at around 86,646 MWh/year, and 21 MW respectively. Of course, it is important to stress that these estimate do not include the productive and commercial/uses that would also develop when access rate increases. Additional data and investigations 8 On the basis of international experience, it is far more cost effective to provide CFLs to customers to reduce demand than to build corresponding additional generation capacity; rough estimate shows that capacity savings through CFLs could cost less than $ 200,000 per MW. 23 23/81 will be required when preparing a project to calculate upstream impact of the new connections on the transmission and the generation. Table 9: estimate of additional energy and installed capacity required to achieve 20 % access rate in rural areas additional rural households to be additional energy to be served per year for additional installed connected until 2010 domestic, productive and social uses capacity required 1,055,053 86,646 MWh/y 21 MW 3. Electricity needs in productive and social sectors in rural areas 3.1.Methodology Very little data exists on electricity requirements of the different sectors of activity in the Kenyan rural areas. To get quantitative figures of the electricity demand related to productive and social uses in these areas would require a detailed survey which was not possible to undertake in the context of this study. Such a survey should be undertaken when preparing the next large rural electrification project. In the context f this study, the following methodological steps have been followed: (i) Review of current and planned sector programs and development reports to assess their requirements for electricity; (ii) Discussion with officials of various ministries, notably ministries of Energy and of Planning and National Development; (iii) Interviews with experts in the ministries of Agriculture, Health, Education, and Water and Irrigation; (iv) Discussion with officials of the Central Bureau of Statistics (CBS), the Kenya Power and Lighting Company (KPLC), the Communications Commission of Kenya and telephone network providers – Telecom Kenya, Celtel and Safaricom; (v) Review of data provided by the respective ministries - Energy, Planning and National Development, Education, Health- by various bodies such as CBS, KPLC, CCK, Telecom Kenya, Celtel and Safaricom; (vi) Field visits to selected trading centers to better ascertain current sources of energy for households, businesses, their expenditures on energy and the impact of readily available electricity on livelihoods, and o business and commercial activities; a summary of the main observations made during the field visits is presented in annex 4. 3.2.Sector programs Review was undertaken of current and planned development programs of various sectors covering Agriculture and Rural Development, Education, Health, General Economic Services (Trade and Industry; Tourism and Wildlife; Labour and Human Resources Development; Gender, Sports and Social Services; and Youth Affairs), Information Communication Technology, Physical Infrastructures, Public Administration, and Water and Irrigation. 24 24/81 This was supplemented by discussions with officials of a number of ministries e.g. Energy, Planning and National Development, Agriculture Health, Education, and Water and Irrigation all of which also provided additional data on activities in the rural areas. Discussions were similarly held with government organizations such as KPLC, the Communications Commission of Kenya (CCK), Telkom Kenya, the Central Bureau of Statistics (CBS), National Irrigation Board, National Water Conservation and Pipeline Corporation; these too provided valuable information and data. Furthermore, discussions were held with a private sector cell phone service provider, Safaricom, which also shared with us information and data on their electricity requirements in the rural areas. A selected number of sectors, including, agriculture, health, education, administration, information communication technologies, water and irrigation and co-operatives are examined below as part of an attempt to highlight the high level of unmet or under- served demand for electricity in the rural areas; also examined are the micro and small- scale enterprises many of which are located in rural areas. A limiting factor in this effort is the lack or adequacy of data on the demand for these activities, current or planned. 3.2.a. Agriculture Agriculture plays a major role in Kenyan economy: about 80 per cent of the population depends on agricultural activities; moreover, agriculture directly contributes 26 per cent of the country’s GDP and a further 27 per cent indirectly through linkages with manufacturing, distribution and other sectors. The agricultural sector cuts across different levels of electricity use, ranging from households; small and medium scale enterprises, including cottage industry; and agro- based industries, such as dairying, tea, coffee and sugar production. Electricity is also used other activities such as irrigation, crop drying, food preservation (in particular refrigeration and cooling), general food processing (including post harvest processing), irrigation, food storage, industrial processing and horticulture. Agricultural activities would expand significantly and productivity improved greatly if electricity were available for these activities To be sure, different levels require different types of energy depending on availability, accessibility, affordability, sustainability and socio-cultural factors among others. A selected number of sub-sectors are highlighted below, fishing, livestock, and poultry. Fish industry Just as the fish industry has been growing steadily, so has the requirement for energy in this sub-sector. But this requirement has not been fully met, far from it. The fish industry is large and employs thousands of people and with an annual turnover of KShs 17 billion. A strong case can be made for supplying selected beaches with adequate electricity for fish storage and preservation (cooling and drying); suitable areas within the vicinity of the beaches could also be supplied for processing the fish. No firm data is available for electricity requirements but they are quite large. The inadequacy, and in many cases, complete lack of cooling services, has, for example, resulted in the loss of many catches with attendant loss in incomes, profits and livelihoods. While no figures are available, the 25 25/81 unmet as well as underserved demand is high. The on-going development of the sub- sector policy will go a long way in addressing the electricity supply challenges facing the industry. Livestock Significant progress has been made in enhancing local and international market access for beef. This should be reflected in increased sales. Nevertheless, the unavailability or inadequacy of energy, in particular electricity, constitutes a major constraint in the development of livestock based industries in the rural areas. Electricity is, for example, required for operating abattoirs and for dairying. In the dairy industry, the implementation of reforms that includes the regularization of the informal milk sub-sector, revival of the KCC and improved extension services has led to significant growth in milk output (3.5 per cent in 2005). Moreover, the Government has initiated the review and formulation of various policies aimed at streamlining the dairy industry; this will result in expansion of activities and increased productivity in the industries and attendant increase in demand. Milk is pasteurized or converted to yogurt or to mala. Electricity is the main source of energy; some plants use diesel for normal operation or as standby, while others use firewood or charcoal. The total electricity consumption, estimated at 3,015,600kWh (in the year 2000), is thus expected to increase significantly in the light of these new developments. A Dairy Commercialization Project, to be funded by I FAD, is planned for 2006/2007 will enhance dairy production and add value to the milk. This too will result in increased demand for electricity. The poultry sub-sector is also a fast growing user of energy – for providing controlled heat and light incubation. Most of poultry breeding activities are located in rural and peri- urban areas. Data on the energy requirements in the sub-sector is not readily available but we believe it is substantial and, as noted, fast growing. 3.2.b. Education The education sector has witnessed tremendous growth since independence; the number of students enrolled at various levels of education has substantially increased: enrolment in formal public primary schools grew from 891,533 pupils in 1963 to 7.2 million pupils in 2004, due primarily to the introduction of Free Primary Education in 2003. The net national enrolment rate in public primary schools rose from 67.8 per cent in 2002 to 82,2 per cent in 2004. The number of primary schools has increased over the years to stand at 19,587 in 2004; of these, 17,678 are public schools and 1,839 are private ones. There are 1,160 day/boarding and 18,427 day primary schools. Table 10 shows the number of primary schools, public and private, in the period between 1999 and 2004, broken down by province 26 26/81 . Table 10: Number of Primary Schools by Province, 1999 -2004 Province 1999 2000 2001 2002 2003 Public Private Public Private Public Private Public Private Public Private Public Coast 1 031 82 1 019 143 1 007 131 1 004 143 1 034 120 1 039 Central 1 752 84 1 842 152 1 751 216 1 763 253 1 766 471 1 774 Eastern 4 071 96 4 115 123 4 044 130 4 048 134 4 098 285 4 112 Nairobi 188 58 183 185 164 178 188 182 188 94 189 Rift Valley 4 663 106 4 515 170 4 583 206 4 554 225 4 768 464 4 763 Western 1 926 16 1 923 29 1 915 31 1 895 30 1 956 92 1 985 Nyanza 3 648 127 3 677 429 3 509 460 3 411 469 3 575 316 3 604 North Eastern 181 - 179 5 183 5 194 5 209 15 212 Total 17 460 569 17 453 1 236 17 156 1 357 17 057 1 441 17 594 1 857 17 678 TOTAL 18 029 18 689 18 513 18 498 19 451 Source: Ministry of Education, Kenya 27/81 At the secondary level, enrolment grew to 862,908 students in 2003 (415,246 girls and 447,662 boys). The number of secondary schools, both public and private, has increased from 151 at independence to 4,100 in 2004. Of these 860 are boarding, 920 are boarding/day and 2,320 are day schools. Enrolment in secondary schools has increased from 738,085 in 2000 to 923,134 in 2004, an increase of 25.1 per cent. In spite of increased enrolment in secondary schools, the sub-sector is still faced with issues of access, equity and quality, mainly because growth of educational facilities at this level has not kept pace with growth in demand. Table 11 shows a brake down of public and private secondary schools for the period 1999 to 2004. The education sector is a large consumer of energy given that learning/teaching institutions rely on energy for lighting, heating and cooking. Of Kenya’s 20,000 educational institutions, about 90 per cent use wood fuel to prepare meals. Firewood collection leads to the destruction of trees that could absorb carbon dioxide emissions and degrades local ecosystems. Also, in some schools children use significant amounts of time searching for fuel wood – time which could otherwise be spent on learning. Currently, less than 20 per cent of all the secondary schools have access to grid connection while 30 per cent use stand alone generators and other forms of energy. Less than 5 per cent of the primary schools have electricity and most of the schools still rely on standalone generators, paraffin and wood fuel. The demand for electricity in rural schools, especially primary schools, while relatively low at the present time, has the potential of growing substantially if certain actions are implemented. Among these is the expansion of the current Rural Electrification Program so that as many primary schools as possible are connected; those far away from the grid should be provided with stand alone generators or supplied from other decentralized systems such as solar photovoltaic (PV), wind small scale hydro or a combination of these; indeed there is an ongoing Government PV solar electrification project to electrify some public secondary schools in Arid and Semi-Arid districts of North Eastern and Rift Valley Provinces. It has also, for example, been suggested that the Government in its endeavor to increase the stock of human skills should put in place policy measures to address the low transition rates from primary to secondary schools, currently estimated at below 50 per cent as a way of supporting the country’s overall development goals. Provision of conventional energy in secondary schools is one of the strategies to help the Government realize this noble goal. All in all there is a need to increase availability of electricity to secondary schools to 40 per cent by 2008, 50 per cent by 2010 and 80 per cent by 2015. Provision of use of generator should also be increased from the current 30 per cent to 50 per cent in 2010 and 70 per cent in 2015. The school mapping project, started in 2004, is not yet complete. It is undertaking a census and providing geographic locations of all schools; in the process the schools connected to the grid, as well as those using stand alone generators or other energy sources will be identified and mapped. When complete the initiative would provide useful information for assessing energy demand, including that for electricity, met and unmet, served and underserved, in schools in the country, both rural and urban, and for preparing the necessary strategies for meeting such demand. 28/81 Table 11: Number of Secondary Schools by Province, 1999 - 2004 1999 2000 2001 2002 2003 2004 Province Public Private Public Private Public Private Public Private Public Private Public Private Coast 150 16 146 17 153 31 150 38 143 43 142 43 Central 605 36 514 43 629 36 602 48 661 65 683 65 Eastern 576 88 645 100 611 73 634 86 706 90 713 90 Nairobi 48 47 48 47 48 47 48 47 48 47 48 47 Rift Valley 600 128 527 139 666 160 674 152 813 164 818 164 Western 406 16 356 13 449 14 448 14 453 17 452 17 Nyanza 581 47 531 44 663 52 669 52 737 62 740 62 North 22 3 24 3 22 3 22 - 22 - 25 2 Eastern Total 2,988 381 2,791 406 3,241 416 3,247 437 3,583 488 3,621 490 TOTAL 3,369 3,197 3,657 3,684 4,071 4,112 Source: Ministry of Education, Kenya 29/81 3.2.c. Health Among the public hospitals, there are two national referral hospitals, eight provincial hospitals, 75 district hospitals, 460 health centers and 1,628 dispensaries in Kenya. Table 12 shows the public health facilities broken down by province. Unfortunately, no information has been provided on private health facilities of which there are many and whose number is growing in the rural areas. Table 12: Government Health Facilities * Province Hospital Health Centers Dispensaries Total % GoK Facilities/ 100,000 population Central 15 56 234 305 13.8 7.6 Nairobi 7 33 59 99 4.5 3.7 Coast 16 45 185 246 11.2 8.6 Eastern 18 59 307 384 17.4 7.6 North Eastern 3 9 38 50 2.3 3.7 Nyanza 15 79 187 281 12.8 5.8 Rift Valley 33 114 554 701 31.8 8.6 Western 8 65 64 137 6.2 3.6 Kenya 115 460 1628 2203 100.0 6.7 Source: Ministry of Health, Kenya Typical equipment which requires energy in a dispensary includes refrigerators for vaccines, sterilization equipment, examination lamps, microscopes and centrifuges. In the health centre the equipment which requires energy to operate includes examination lamp and general lighting, refrigerators, sterilization equipment, heaters, suction machines, microscopes, centrifuges, transfer incubators, kitchen cooking equipment, ironing equipment and incinerators. In the district hospitals the number of equipment requiring energy to operate is even larger and includes:- examination lamp, general lighting, refrigerators, suction machines, dental equipment, heaters, kitchen cooking equipment, laundry equipment, theatre and sterilization equipment, mortuary plants, physiotherapy equipment, maternity and nursery equipment, ophthalmic equipment, pharmacy equipment, stand by generators, and incinerators. All referral and provincial hospitals are connected to the grid; so are about most the district hospitals. The number of dispensaries has been growing steadily in the rural areas, especially so in recent years, in response to increased population and ever mounting multiplicity of health problems. At present, dispensaries rely more on gas and only about 5 per cent of them have access to electricity. About 50 per cent of the health centers are connected to electrical supply and about 20 per cent have stand by generators; 5 per cent of health centers uses solar power; charcoal and firewood are used mainly for cooking. Of the 30/81 district hospitals about 90 per cent are connected to electricity while about 90 per cent have stand by generators. Here again one can appreciate, even without complete data, the fact that there is large unmet demand in the rural areas. It is the health centers and dispensaries most of which are located in the rural areas that will drive of demand for electricity by the health facilities in these areas. 3.2.d. Information and Communications The communications sector is a major consumer of electricity in the rural areas, most of itself supplied by the communications service providers, which put a high premium on reliability and quality of their services to the customers. They have a large and fast growing number of repeater stations in the rural areas throughout the country. For each station, the following are required: two generators, batteries and a rectifier. Consequently the service providers have to make major investments to meet the electricity demand for their stations. This demand is huge, and yet it is not reflected in the current electricity demand data in the country. Because of the need to cool some of the equipment at the stations, electricity requirements at each station are relatively high. If this demand were to be reflected it would constitute a significant portion of the total rural electricity requirements. Table 13 provides a sampling of the communications stations not connected to the grid installed by the Telkom Kenya Corporation, a major provider of line telephone network in Kenya. This is but a small portion if the total number of stations that are self supplied by Telcom Kenya. The supply of electricity to such stations is an expensive undertaking, one that has undoubtedly resulted in a lower rate of growth in meeting communications needs of the rural population, and that would, furthermore, greatly benefit from connection to the grid where possible. Electricity supply constraints may have, directly or indirectly affected the quality of communications services provided to the rural areas.. Table 13: Sample of Telkom Kenya sites not connected to the grid Type of power Exchange Make Region District KPLC DEG Solar standby Dual Bura Tana LCON Eastern Tana Riv --- 2x8KVA Cheptiret MB R.Valley -- 1kVA Dadaab RA/TER Eastern -- 2x8KVA Eshibuye LCON Western Kakamega --- 2KW Faza ‘’ Coast Lamu --- 2kW Gabra Tula ‘’ Eastern Isiolo --- 2x8KVA 2kW Habsweni PCM/TER Nothern? --- 2x8KVA Ianzoni LCON Eastern Machakos --- 2kW Kadel ‘’ Nyanza Migori 2kW KenduBay ‘’ Nyanza Rachuonyo --- 2kW Manga ‘’ Nyanza --- 20kW Migwani ‘’ Eastern Kitui --- 2kW 31 31/81 Modogashe PCM/TER Eastern --- 2x8KVA Mpeketoni LCON Coast Lamu --- 2kW Mtomo ‘’ Eastern Kitui --- 2kW Namanga ‘’ R.Valley -- 2kW Ndalu ‘’ R.Valley --- 2kW Ndaragwa ‘’ Central Nyandarwa --- 2kW Okia ‘’ Southern? --- 2kW Plateau ‘’ R.Valley Ouasingishu --- 2kW Rhamu ‘’ N. Eastern Mandera --- 2x8KVA SKinangop ‘’ Central Nyandarwa --- 2kW Songhor ‘’ Nyanza Nyando --- 2kW Tausa ‘’ Coast Voi --- 2kW Tawa ‘’ Eastern Mkuweni --- 2kW Wenje MB Eastern 2x8KVA Source: Telkom Kenya Safaricom is the main provider of cell phone services in Kenya. Table 14 too, gives a breakdown of the Safaricom stations throughout the country - 34 in all. Of these the Safaricom has applied to be connected by KPLC in 12 stations. A three phase 10kVA supply is needed at each station so a total 340kVA is currently self supplied. But this figure is bound to rise rapidly in the immediate future in the face of unprecedented growth in cell phone services. It has, for example, been reported that Safiricom is signing up on average 100, 000 new users each month and targets 5.5 million customers by 2007; this represents an annual rate of growth of 25 per cent. Communications service providers are keen to work out modalities for KPLC to connect many of their rural stations, bearing in mind their stringent reliability and service quality requirements. A Rural Telecommunications Project, to be funded by China is being planned; this would involve installing repeater stations at locations throughout the country within the vicinity of each Divisional Headquarters. The demand for electricity to supply the stations is expected to be enormous. The service provider-Telkom Kenya- would like to co-ordinate it effort with KPLC toward meeting the electricity requirements for the proposed project. The cell phone industry, spearheaded by Safaricom and Celtel companies, has come out as a major contributor to economic growth in the country. The number of cell phone users has grew by a staggering 56.9 per cent in the past year; the number of cell phone users is currently estimated at 5.6 million, a sizeable portion of which lives in the rural areas; this has brought with it a fast growing demand in the rural areas for charging cell phones. 32 32/81 Table 14: Safaricom sites without power S.No Re Location Site Site Name Site Co-ordinates KPLC Permit Remarks Code Latitude Longitude App Number 1 CE Embu 01181 Mbeere Kianjiru Hill 000° 40' 44.0 " S 037° 36' 44.8 " E E25202003100132 Remote power solution 2 CO Msa Rd 01855 Kyulu Hill Top 002° 55' 58.1 " S 038° 24' 30.7 " E Remote sites power solution 3 CO Voi 01905 Taru 003° 46' 19.5 " 039° 07' 09.3 " Remote sites power solution from Winafrique 4 CO Garsen 62325 Garsen Dual genset to be employed 5 CO Garissa Rd 71242 Katumba Remote power solution 6 CO Garissa 71243 Madogo Remote power solution 7 ET Makueni 01820 Lurmuti 002° 10' 59.4 " S 037° 38' 41.6 " E Remote sites power solution 8 ET EmbuMeru 01247 Kunati 000° 05' 54.1 " N 037° 54' 18.1 " E RPS employed. Dual gensets used. 9 ET Machakos/Kitui 01141 Mbubuni 001° 37' 15.30" S 037° 33' 36.40" E Remote Power solution to be employed 10 ET Machakos/Kitui 01142 Kikoko Hill 001° 48' 05.70" S 037° 23' 42.50" E RPS employed. 11 ET Machakos/Kitui 01143 Ikalaasa 001° 31' 23.0 " S 037° 39' 14.3 " E Remote Power Solution to employed. 12 ET Machakos/Kitui 01144 Kalawa 001° 39' 13.7 " S 037° 42' 02.3 " E Remote Power Solution to be employed 13 ET Machakos/Kitui 01169 Zombe 001° 26' 25.10" S 038° 14' 41.40" E Remote Power Solution to employed. 14 ET Machakos/Kitui 01171 Maviani 001° 22' 25.90" S 037° 49' 28.60" E E25802005110011 Remote power Solution Needed 15 ET Kitui Out 01165 Kisasi 001° 31' 23.90" S 038° 01' 18.20" E E21202006020037 Remote power solution to be employed 16 ET Kitui Out 01167 Kilonzo 001° 24' 11.70" S 038° 03' 20.70" E E21202006040033 Hub Site -Remote power solution to be employed 17 ET Kitui Out 01168 Mutomo 001° 50' 23.50" S 038° 12' 39.10" E Remote power solution employed 18 ET Mwingi 71241 Mbuvu Remote power solution 19 NY Migori 01581 Mwachi 001° 00' 54.50" S 034° 24' 43.20" E Remote power solution to be employed 20 NY Gwasi 01647 Nyenga 000° 43' 03.70" S 034° 07' 28.20" E Remote power solution to be employed 21 NY Kisii 01648 Nyangweta 000° 54' 26.40" S 034° 40' 55.30" E Remote Power Solution to be employed 22 NY Kuria 01649 Ntimaru 001° 20' 30.90" S 034° 41' 15.20" E Remote power solution to be employed 23 NY Kedu Bay 01562 Ndiru Hill 000° 22' 58.7 " S 034° 40' 04.2 " E E24102005050489 Siemens power solution employed 24 NY Kedu Bay 01563 Kanyamfwa 000° 23' 05.3 " S 034° 32' 02.7 " E E24102005050488 Remote power solution to be employed 25 RV Marakwet 01625 Chesongoch 001° 07' 52.2 " N 035° 38' 36.80" E Remote power solution to be employed 26 RV Marakwet 01626 Tot 001° 12' 17.40" N 035° 39' 20.80" E Remote power solution to be employed 27 RV Kitale 01738 Mutua 001° 09' 37.80" N 035° 01' 40.90" E E21202006020011 RPS to be considered 28 RV Laikipia 01189 Colobus Farm 000° 28' 55.59" N 036° 15' 24.59" E E21202006020032 Remote power solution to be employed 29 RV Namanga_Taveta Rd 01415 Namanga Town 002° 32' 56.30" S 036° 47' 45.20" E RPS:-Twin gensets employed 30 RV Namanga_Taveta Rd 01409 Maparasha Hill 002° 11' 07.2 " S 036° 48' 47.6 " E Remote Power Solution-Two gensets 31 CE Nyeri-Nyahururu 1477 Segera Ranch 000° 06' 49.90" N 036° 49' 30.90" E E21202006020009 RPS to be considewred 32 RV Namanga_Taveta Rd 01419 Emali Makutano 002° 32' 04.0 " S 037° 29' 35. 7" E E21132006010160 RPS to be considered. 33 WT Kericho 01727 Nasianda 000° 45' 27.5 " N 034° 54' 28.6 " E E27102004090189 Remote power sol. to be employed. Genset power available 34 WT Maragoli 1 01664 Misango Hill 000° 09' 06.2 " N 034° 39' 47.3 " E E24102005050499 Remote power solution to be employed Source: Safaricom 33/81 3.2.e. District development plans The District Development Committees (DDCs) were established as part of Government policy of decentralizing decision-making. The committees deliberate on projects at the district level. They examine on-going projects in the various sectors to ascertain constraints and challenges facing their implementation. They also make proposals on new projects; in the process they determine priorities for the project and programs, down to the divisional or lower levels. The priority projects and programs are then presented to line ministries for funding and implementation. In the rural electrification sub-sector, the DDCs presented to the Ministry of Energy a list of rural electrification project priorities for 2004/05 & 2005/06, broken down by constituency. There 210 constituencies, most of them in the rural areas. The priority projects which run into thousands cover a wide range of activities from schools, health facilities, water pumping from boreholes, irrigation projects; they also relate to connecting markets/trading centers, administrative centers, as well as such community projects as orphanages and children’s homes; others are concerned with supplies to productive activities including coffee factories. The estimated cost for the projects the two years alone is about Kshs8.9 billion (see Table 15). Even if only a small fraction of the list were to be implemented, the demand for electricity in the rural areas would increase manifold. 3.2.f. Water and Irrigation Apart from the huge requirement for electricity in the water sector for domestic purposes, there is a significant and growing demand for irrigation, pumping from boreholes and other activities the rural areas. The National Irrigation Board which is responsible for public irrigation projects in the country has provided valuable information on its irrigation activities. In Western Province, for example, it is involved in a project at Bunyala, currently covering 200ha of rice. Diesel generators are used to supply the electricity for irrigation. A major drawback is the inadequacy or unreliability of power supplies. Plans are already in place to increase the area under cultivation to 500ha in 2007, and to 1000ha by 2008. More farmers would like to be included in the project, but there are inadequate supplies: at present four old diesel generators supply a total of 160kVA; with the planned expansion, 3MVA would be required by 2009. The Ahero irrigation project in Nyanza Province is bigger: 1,000ha are under cultivation at the present time, planned to increase to 1,200ha in the year 2007 and then to 1,500ha in 2008, with attendant increase in the requirements for electricity for irrigation. With increased activities, more villages will come up, and the size of existing ones grow substantially, all of which will result in increased demand for electricity – for irrigation, milling, other productive activities, and to meet lighting requirements in the growing number and size of the villages/ trading centers. Close by Ahero is the West Kano scheme of 1,000ha, with water being pumped from Lake Victoria. Plans are under way to expand this project, too. This will undoubtedly result in increased demand for electricity, for pumping, milling and to meet the expected growth in the needs of the villages/ trading centers. 34/81 Table 15: Rural electrification project priorities for 2004/05 & 2005/06, broken down by constituency; also given estimated cost in million KShs Nairobi Coast cont. Eastern cont. Nyanza RValley Western Constituency Cost Constituency Cost Constituency Cost Constituency Cost Constituency Cost Constituency Cost Dagoreti 4.2 Msambweni 54.0 Kilome 16.5 Rongo 27.0 Baringo C. 84.0 Kimilili 48.5 Mathare 2.0 Kinango 30.0 Manyatta 17.87 Nyatike 84.0 Baringo E. 73.5 Bumula 46.0 Total 6.2 Mwatate 15.5 Runyenjes 25.5 Uriri 36.5 Baringo N. 102.5 Webuye 112.5 Central Voi 21.5 Tigania W. 13.3 Migori 38.0 Keiyo N. 38.5 Sirisia 49.0 Limuru 3.0 Wundanyi 24.0 Igembe 2.5 Bomachoge 28.3 Keiyo S. 13.5 Kanduyi 25.0 Kabete 38.2 Taveta 32.5 Ntonyiri 16.0 Bobasi 51.0 Saboti 30.5 Mt. Elgon 52.5 Lari 59.5 Magarini 167.0 Tigania E 13.0 S Mugirango 13.5 Kwanza 49.6 Lurambi 33.5 Githunguri 25.0 Malindi 11.5 Gachoka 28.0 Kisumu Rur 41.5 Cherangani 21.0 Shinyalu 34.5 Kiambaa 38.5 Bahari 18.0 Siakago 118.5 KisumuT W. 4.5 Buret 16.0 Ikolomani 15.5 Kinangop 30.0 Kaloleni 15.0 Tharaka 108.0 Nyando 18.5 Konoin 9.0 Malava 20.0 Ol Kalau 30.7 Ganze 57.5 Moyale Dist 28.0 Kisumu T. E 9.5 Samburu W. 4.0 Budalangi 31.0 Ndaragwa 41.0 Galole --- S. Imenti 66.5 Kasipul Kab 39.0 Samburu E. --- Butula 92.5 Kipipiri 94..5 Bura 4.0 N. Imenti 74.9 Karachuonyo 169.0 Eldama R 18.5 Nambale 18.5 Mwea 21.0 Garsen 160.0 C. Imenti 36.46 Gwassi 32.7 Mogotio 44.2 Funyula 4.5 Ndia 34.5 Lamu E. 36.0 Mwingi N. 64.0 Mbita 12.5 Laikipia E. 126.1 Amagoro 83.5 Kerugoya 24.2 Lamu W. 24.0 Mwingi S. 19.5 W.Mugirango 67.7 Laikipia W. 136.1 Butere 22.0 Gichugu 11.0 Total 829.3 Total 1,364.23 Kitutu Masab 149.5 Belgut 36.0 Mumias 45.7 Mathioya 48.4 Eastern N Eastern N.Mugirango 75.5 Kipkelion 132.0 Matungu 37.5 Kiharu 30.4 Kangundo 38.0 Wajir S. --- NyaribariChache 68.0 Ainamoi 4.0 Khwisero 15.5 Kangema 66.1 Machakos T. 37.5 Wajir N. --- Kitutu Chache 101.5 Marakwet E 174.0 Lugari 53.0 Kigumo 71.4 Yatta 13.0 Wajir East --- NyaribariMasaba 165.5 Marakwet W 109.5 Malava 20.5 Kandara 34.7 Mwala 131.5 Wajir W. --- Bonchari 32.0 Bomet 36.8 Sabatia 20.0 Maragwa 36.0 Kathiani 11.5 Dujis --- Total 1,869.2 Chepalungu 33.0 Hamisi 29.0 Juja 28.0 Masinga 30.0 Ladgera --- Rift Valley Sotik 11.5 Emuhaya 33.0 Gatundu S. 19.5 Saku 5.0 Fafi --- Kapenguria/Sigo 40.5 Subukia 6.0 Vihiga 17.5 Gatundu N. 21.05 Laisamis --- Total --- Kapenguria/Kac 57.5 Naivasha 49.5 Total 960.7 Gatanga 40.8 Kitui S. 149.0 Nyanza Sigor 71.5 Nakuru T 10,0 Othaya 62.05 Mutito 77.0 Ndhiwa 59.0 Kacheliiba 135.5 Rongai 21.0 Gran Total 8,884.63 Kieni 14.7 Kitui Central 53.4 Rangwe 89.5 Turkana C. --- Molo 26.0 Tetu 41.0 Kitui West 92.8 Gem 90.5 Kajiado South 104.1 Kuresoi 19.5 Mathira 24.9 Nithi 44.0 Ugenya 18.5 Kajiado C. 71.5 Tinderet 22.0 Mukurweini 49.0 Isiolo S. 33.0 Alego Uson 31.0 Kajiado North 3.0 Aldai 56.0 NyeriTown 12..5 Isiolo N. --- Rarieda 46.8 Eldoret North 41.5 Narok N 242..5 Total 1,051.6 Makueni 5.5 Bondo 21.5 Eldoret South 50.0 Narok S 212.5 Coast Kibwezi 69.0 Kuria 206 Eldoret East 43.0 Kilgoris 112.0 Kisauni 9..5 Mbooni 79.0 Nyando 11.5 Mosop 52.5 Total 2,803.40 Likoni 14..5 Kaiti 34.5 Nyakach 16.7 Emgwen 52.0 Changamwe -- Muhoroni 13.0 Matuga 134.8 Source: Ministry of Energy, Kenya 35/81 Within the Rift Valley Province, in Baringo’s Marigot area, there is an irrigation project for maize at three different locations, covering, respectively, 500ha at Pakeria, 250ha at El Dume, and 250ha at Sendai. There are plans to expend the project that will result in increased demand for electricity for milling and other productive activities associated with the expected growth in settlements around the project areas. The Mwea scheme in Central Province is a major rice irrigation project that has grown steadily over the years; 6,000ha are under cultivation, planned to increase to 10,000ha. Irrigation is by gravity; nevertheless, large community settlements have grown in step with the expansion of the scheme and are expected to get bigger; some of the settlements are connected to the grid while others are not. There will be need for more electricity, first and foremost for increased rice milling, and also for SMEs that are springing up at a fast rate around the scheme, as well as for lighting. In the Coast Province the main irrigation projects are at Bura and Hola. The Bura scheme is for cotton farming and horticulture; pumping for irrigation is supplied by a diesel plant rated at 3MVA. A total of 600ha are currently under cultivation, planned to increase to 2,500ha in 2007 and then to 6,000ha. Electricity requirements will increase accordingly – for irrigation pumping, milling, preservation, SMEs, and general lighting in the expanded settlements. There is also a good potential for using solar PV to meet household requirements at the two locations. The Hola undertaking is also for cotton farming and horticulture. Phase I has involved the cultivation of 900ha pumped by a diesel plant, rated at 480kVA; the plant is being rehabilitated; and arrangements are in place, during phase II, to expand land cultivated to 2,500ha. Here again, the demand for electricity will increase by a large margin to meet requirements similar to those of Bura. There are also other smaller projects throughout the country, for example mango in Nyanza, Sigor in the Rift Valley and Kilifi in the Coast, all of which are planned for expansion with attendant increase in electricity requirements. The National Water Conservation and Pipeline Corporation is charged with, among others, the task of drilling and equipping boreholes throughout the country, based on the priority listing by the District Development Committees. No firm data is available on the number of boreholes currently in use but attempts are under way to collect and collate the data. Most of the boreholes are located in rural areas. Energy is required for pumping water from boreholes; only a few of these are connected to the grid, with the rest relying almost exclusively on generators for water pumping, an undertaking which has turned out to be extremely expensive, especially so in Arid and Semi-arid Lands (ASALs) - in the light of exacting logistics and the long distances involved in transporting the fuel, thus adding to the already high prices. Water so pumped would be available, not only for drinking, but also for other productive purposes such as irrigation, livestock use, agriculture and the like. 36 36/81 The National Water Conservation and Pipeline Corporation has an on-going program for drilling boreholes – about 188 in 2006 and estimated at 250 in the year 2007. The depth varies considerably from 60m to well over 300m, the latter mainly in arid areas; the rated output for the generators range from 5kVA to 11kVA, with some as high as 22kVA. Even without firm data on the number of boreholes, the ballpark demand for electricity to meet pumping requirements for the ever growing number of boreholes throughout the country is large. 3.2.g. Cooperatives There are two main types of cooperatives in Kenya: (i) those for production and marketing and (ii) those for savings and credit also known as (SACOS). Hundreds of cooperatives of the first category are operating in different parts of the country, covering such areas as coffee (900), cotton, dairy (400), horticulture, pyrethrum, etc. Most of the production and marketing cooperatives would like to use processing equipment that adds value to their products: pulping for coffee factories, ginneries for cotton, cooling tanks for dairies, etc. For example, diesel generators are widely used in the pulping process in the coffee factories. Substitution of diesel by electricity has for years been identified as a key productivity gain for Kenyan coffee producers. An IDA loan (Second Coffee Improvement Project Cr 20620, MUS 46.8) and the first phase of the European Union both sponsored the Coffee Factories Rural Electrification Project (COFREP I) targeted the electrification of 168 factories out of 896. It turned out that of the 168 factories, 59 had already been connected; the remaining 109 factories were connected after long delays, arising largely from KPLC’s limitation to implement the project and poor management of the program, including coordination problems between Ministry of Energy and KPLC. The European Community has provided support for survey and design for an additional 177 factories under COREFII; these preparatory works were completed in December 2004. However, no financing has been secured so far to proceed with COFREP II. The above examples point to the importance of electricity and the appetite of cooperatives to be supplied with electricity in order to dramatically reduce their energy cost and increase their productivity; they also highlight the difficulties that rural stakeholders face to be connected to the grid. Agro-industries are expected to be the back bone of agricultural development on Kenya that would also ensure a wide range of agricultural related economic and industrial activities in the country with concomitant increase the demand for electricity. 3.2.h. Micro and small scale enterprises Micro and small scale enterprises (SMEs) are businesses employing up to 50 workers. Cottage industries are a component of SMEs and refer to bakeries, brick making, fish preservation (cooling, drying and smoking), milk processing, posho mills, small restaurants and tobacco curing; most of these require electricity input. Other SMEs cover a wide range of activities such as manufacturing, trade, restaurants and hotels, service 37 37/81 providers and construction, The majority of these are – over 66 per cent - are located in rural areas. Posho mills These are mills for crop grinding (mainly maize, millet or sorghum); most posho mills (about 88 per cent according to an estimate reported in 2000) with access to the grid use electricity because the latter is cleaner and more convenient (than diesel); furthermore it is cheaper to produce maize flour from electricity driven mills. The average consumption per mill in the year 2000 was estimated at about 3.1kWh per annum, or a total of 73,656kWh. With improved economic situation in the rural areas, the number of posho mills is expected to increase accordingly; this would result in increased demand for electricity, the preferred energy source. Bakeries Most bakeries use electricity, except for those without access to electricity; it was estimated that about 183 small scale bakeries in the rural areas consume on average 8,580kWh per year. The number of bakeries has since risen substantially as has their demand for electricity. Hotels and restaurants Many of these are to be found all over the country, including in trading centers in the rural areas; mostly the electricity is used for lighting, entertainment, refrigerators Total consumption of electricity by the cottage industries was estimated at 343,947,816kWh for the year 2000. Unfortunately there is no breakdown for the hotels and restaurants in the rural areas and for those in urban areas; this notwithstanding, it is reasonable to assume that, with the noted increase in economic and other activities in the rural areas, the number of hotels and restaurants has grown substantially in these areas, and that their demand for electricity has also by a large margin. Table 16 shows annual energy consumption by type and sector in the year 2,000. It shows that the share by the cottage industry for that year was 30 per cent and that for rural households was 8 per cent. Unfortunately the data not disaggregated the share of the cottage industries located the rural areas and which in urban. Other pieces of information provided are incomplete at best; for example, in a table on total electricity consumption by sector, the consumption of electricity in the agricultural sector is shown as zero; we believe, however, that a significant amount of electricity is consumed in the sector. Table 16: Annual energy consumption by type and sector Type of energy % share Sector Firewood Charcoal Wood Farm Electricity Kerosene LPG Total (43.8%) (46.0% wastes residue demand (0.6%) (6.4%) (0.7%) (2.2%) (0.2%) (100%) Rural 89.4% 46.2% 61.9% 99.5% 8% 53.1% 5.6% 80% household Urban 2.3% 36,5% 38.1% 0.5% 61.8% 46.3% 66.7% 13% household Cottage 8,3% 17,3% 0% 0% 30.2% 0.79% 27.7% 7% 38 38/81 industry Total 100% 100% 100% 100% 100% 100% 100% 100% Source: Study on Kenya’s Energy Demand, Supply and Policy Strategy for Households and Small-scale Industries and Service Establishments 3.2.i. Jua Kali sub-sector The Jua Kali is a fast growing informal sub-sector in both urban and rural areas and includes artisans with a wide variety of skills such as welding, machining, carpentry, shoe making, motor and bicycle repair etc. The sub-sector depends on electricity for electric machines, welding, battery charging, spraying machines, electric jacks, motor rewinding, sewing machine repair, wood work, band saws, computer repair, chemical mixers, boilers, hatcheries, electroplating chrome plating and driers, among others. A couple of examples can be seen in the field visits report presented in annex 4. Experience has shown that, for those in the rural areas, productivity and quality of work are improved by a large margin and business expanded substantially, following their connection to the grid. No information is available on the level of demand for the sub-sector, but it is believed that the demand is substantial and growing rapidly, and that much of the demand is unmet or under-served III. Structure, Organization and Performance of the Rural Electrification Program 1. Current structure of the Power Sector The power sector has been undergoing restructuring since the mid-90s. The GoK, through the Ministry of Energy (MoE), is responsible for policy formulation whilst the Electricity Regulatory Board (ERB), established in 1998, is responsible for regulation of the power sector. To date, the structural reforms developed under the just concluded multi-donor supported Energy Sector Reform and Power Development Project and the ongoing Energy Sector Recovery Project, have focused on creating and enhancing commercial relationships between the main players, eliminating government subsidies to the sector (except for the rural electrification program), introducing an autonomous regulatory system and securing private sector participation in power generation. Since 1998, the following main developments have taken place in the power sector; i) Unbundling of the power sector into two entities; KPLC and KenGen as on 1 st September, 1997. ii) Transfer of assets as of June 1999; iii) Implementation of an Interim Power Purchase Agreement (IPPA) between KenGen and KPLC. The IPPA was intended to run for a period of 2 years from 1st August 1999; but has been extended to 30th June, 2006. iv) Adjustment of retail tariffs that took effect from 1st August 1999 responding to exchange rate fluctuations and pass though policy for fuel prices ; 39 39/81 v) Publication of ERB’s Retail Electricity Tariffs Review Policy Framework in 2005 Today, publicly owned generation assets are owned, managed and operated by Kenya Electricity Generating Company (KenGen) since 1999. The KenGen supplies about 80 per cent of the total generated energy while the private sector provides the balance. KenGen is a listed company in the Nairobi Stock Exchange; until recently, its shareholding currently was 70 per cent Government and 30 per cent private. In 2006, the Government, as part of its policy to divest its ownership in public enterprises, undertook to sell its shareholding in KenGen to the public. The 30% sale was to raise about Kshs. 7.85 billion (US$ 109 million equivalent). At the closing date of the Initial Public Offer, the total subscription was about Kshs. 26.4 billion (US$ 366.4 million). This implies an oversubscription of Kshs. 18.5 billion (US$257.4 million). The overwhelming favorable outcome of the IPO demonstrates the confidence the public has in KenGen. All publicly owned transmission and distribution assets are managed and operated by KPLC, which is responsible for transmission and distribution functions throughout the country. KPLC is a listed company in the Nairobi Stock Exchange; until recently its shareholding was 52 per cent Government and 48 per cent private. In 2006, the National Social Security Fund (NSSF) has disposed part of its shareholding in KPLC to private shareholders thereby reducing the total public sector shareholding (NSSF and Ministry of Finance) from 52% to 48%. Thus, the private sector now holds the majority of shares in KPLC. 2. Institutional Setting for Rural electrification Kenya’s Rural Electrification (RE) program dates back to 1967 when the East African Power and Lighting Company (the predecessor to KPLC) began setting aside 1 per cent of its total yearly gross sales revenue for the electrification of rural areas. Subsequently in 1973 GoK, through grant funding from the Government of Sweden, established the Rural Electrification Fund [REF]. KPLC was then mandated by the GoK to carry out execution, and operation and management of rural electrification schemes, including maintenance of accounts and records of such schemes. An RE unit was established in the MoE with responsibilities for planning and evaluating the economics of schemes proposed by the District Development Committees [DDCs] and for monitoring implementation, by KPLC, of rural electrification schemes. A Committee was set up for approving RE schemes submitted by the DDCs which were the designated entities for originating proposals for RE schemes. The RE was to support sub-economic schemes whose selection was to be guided by the criteria set out in the rural electrification master plan. The Government was to bear all the capital expenditures as well as any shortfalls on operation and maintenance costs. The current RE institutional model with its dependence on KPLC for program implementation and operation management is limited in that KPLC lacks adequate management, capacity, technical skills, and human and financial resources to scale up the program to the levels required to achieve rapid access expansion. Field visits and conversations with several stakeholders suggested weaknesses in KPLC capacity to market its services, or respond to business opportunities from customers’ requesting 40 40/81 connections. Given its limitations and the challenges it faces in improving its operational and financial performance in the areas it currently serves, KPLC is not positioned to introduce new decentralized technologies which are necessary to expand access to remote areas from the grid. During interviews KPLC informed the Bank of the recent initiatives it has taken to enhance its management, and technical capacity to address the rural electrification agenda. However, in view of the modest pace of electrification observed in the past a huge challenge still remains for KPLC to expand access within the urban and peri-urban areas. Leveraging KPLC efforts by bringing in other players, financing mechanisms and technologies, amongst other measures, will be necessary for the achievement of the GoK objectives of increasing access to electricity to 40% by 2020 in the rural areas. Also noted was the inadequacy of planning capacity within GoK exemplified by lack of a systematic aggregated data base of sectoral demand for electrification; and the appearance that the criteria of sub-economic schemes that is needed for eligibility to REF resources (under the 1973 GoK/KPLC Agreement) is not always adhered to, but instead schemes may be funded under the RE due, amongst other things, to the inability of sponsors to raise connection costs. 3. Expenditures and Sources of Financing The data on capital expenditures and financing sources, numbers of customers connected, connection costs per customer, operating costs etc; since the inception of the RE program show the limited contribution that the program has made to the rural electricity expansion. Table 17 summarizes the expenditures and financing sources between 1973 and 2001/02 in nominal Kenya shillings. The figures will be expressed in constant US$ terms in the REES, but are given here to give a broad indication of the composition and trends of expenditures as well as of the sources of financing. Table 17: RE Expenditures and Financing Sources9 1973-1985 Ave. Ann. 1986-1998 Ave. Ann. 1999-2005 Ave. Ann Total % Increase Increase Increase Expenditures Investment 184.0 15.3 3495.0 268.8 5038.0 719.7 8717.0 57 Operating Losses 36.0 3.0 1047.0 80.5 5578.0 796.9 6661.0 43 Total Exp. 220.0 18.3 4542.0 349.3 10,616.0 1516.6 15378.0 100 Financing Sources REF 190.0 15.8 2387 183.6 9747.0 1392.4 12324.0 78 Consumers 19.0 1.6 299.0 23.0 219.0 31.3 537.0 3 Donors 54.0 4.5 1096.0 84.3 1926.6 275.2 3076.6 19 Total Financing 263.0 21.9 3782.0 290.9 11892.6 1666.0 15937.6 100 9 The years for which data is aggregated have been selected to align with changes in the REF fund levies 41 41/81 Surplus/Deficit 559.6 Composition of RE Expenditures. Over the period 1973-2002, operating costs constituted about 43 percent of the total expenditures while investment expenditures made up the balance of 57%. Under the 1973 GoK/KPLC Agreement KPLC recovers its operating costs for RE schemes, in the first instance through revenues from electricity sales to RE customers, and then covers any resulting deficit from the REF. Calculation of Operating Costs. The methodology for calculating operating costs was prescribed in the 1973 Agreement with the GoK and was subsequently detailed in Minute No. 216/77 of the Electricity Development Committee Meeting of May 15, 1977. In the methodology overheads10 are apportioned between RE and other KPLC customers on the basis of the number of customers, irrespective of the size of customers. The resultant high allocation of overheads to RE customers contributes to the overall losses from this group of customers, losses which KPLC recovers from the REF. The fact that the operating costs assigned to RE customers exceed revenues from this category of customers has limited the amount of the REF levy that can be used for investment and expansion of electricity to rural consumers. A revision of the allocation formula based on the marginal costs would release additional funds for investments. However this measure alone is unlikely to cause the RE program to break even financially, since tariffs do not fully cover O&M costs. Inadequacy of Tariffs to Cover O&M Costs. Nearly 40% of current KPLC customers belong to the A0 category, which tariff is 1.55 Ksh/kWh for energy, which corresponds to 1.9 USc/kWh. Considering that the least cost generation options are around 5 USc/kWh (Olkaria II extension cost is estimated at 4.9 USc/kWh), it means that current tariff charged doesn’t even cover half of the generation cost. As a consequence, without tariff adjustment, there is no way for rural electrification revenues to be sustainable, even if investment cost is highly subsidized, since revenues do not even cover O&M costs. However, capacity to pay is significantly higher, even in poor non connected rural areas, demonstrating that there is room to adjust tariffs. As a result, it would be necessary to achieve full recovery of O&M costs and recovery of capital investment, through adjustment of tariffs, albeit on a phased approach for the KPLC system with the exception of the isolated systems where the O&M subsidy might remain for a longer period as a matter of social policy. These issues are expected to be further examined by an ongoing tariff study. High Investment Costs. A comparison of investment costs per customer for Kenya’s RE program with other countries suggests that there is scope for cost reduction. These figures are as follows11: Kenya-US$1700, South Africa-US$1130, Bolivia-US$750, and 10 Overhead costs as defined in the agreement cover costs for administration, accounting, billing, metering, and general management. 11 Source: A study by EdF on Kenya ‘s Rural Energy Institutional Framework and financed by the AFD, 2005. 42 42/81 LDC average-US$1380. This very high figure seems to be the direct reflection of the combination of (i) the policy choice to give priority to productive and commercial loads, which does not create incentive for KPLC to maximize connections and (ii) high up-front connection fee /internal wiring that make it difficult for residential customers to connect. As a result the number of connections is relatively low compared to investment cost, leading to very high average connection cost. This is confirmed by the recent figures of the RE project being funded by the French cooperation and implemented by KPLC: the total investment is estimated at Ksh 2,5 billion (USD 35.6 million) for a total number of 21,740 connections12. As a consequence, such a high average connection cost is not representative of the unitary connection cost that should be possible to get in the highly populated Kenyan rural areas, with a different policy and different financial arrangement. The combination of less costly standards and arrangements to maximize the number of connection would very probably allow for significant reduction in the average connection cost and increase in the number of customers connected for a given amount of investment resources. At present KPLC charges for costs of supervision amount to about 17.5 %, an amount that is excessive and could be reduced to about 10%, thus helping to reduce the investment costs per customer connected. Table 18: RE Percentage Financing by Source 1973-1985 % 1986-1998 % 1999-2005 % Total % Financing Sources REF 190.0 72 2387.0 63 9747.0 82 12432.0. 78 Consumers 19.0 7 299.0 8 219.0 2 537.0 3 Donors 54.0 21 1096.0 29 1926.6 16.0 3076.6 19 Total Financing 263.0 100 3782.0 100 11892.6 100 15937.6 Relative Contribution of Different Sources of Financing to RE Expenditures. The REF contributed 59% to the total financing of the RE over the 30-year period to 2002 while donors financed 19 %, consumers 5% and KPLC 18%13. Apart from the initial Swedish grant, the bulk of the donor funding occurred during the period 1987-1991, but declined substantially between 1992 and 1997 as donors shunned Kenya due to disagreements over political liberalization. Donor interest was revived in the late 1990s when the EU provided funding for the electrification of coffee factories (Ksh 375 million) and the Spanish Government provided about US$10.6 million in 1998 followed by another US$10.3 million in 2001 towards rural electrification. Subsequently the French also provided Euro 9.2 million for rural electrification in 2000 followed another Euro 30 million in 2006. While the ratio for calculating the KPLC’s contribution was increased from 1% out of its gross sales revenues to 2% in 1985, the relative contribution of the company actually declined during the period 1986-1998 due to the higher donor contribution. The substantial increase in the relative contribution of the REF in recent years reflects that a 5 % levy payable by consumers on the basis of electrical energy consumed was introduced 12 Financial assessment of RE activity within KPLC, by EDF for the Ministry of Energy, February 2005, p.26 13 The reason and accuracy of a large KPLC contribution in the 1986-1998 period to be verified 43 43/81 in 1998 through Legal Notice No.96 of 1998 pursuant to the provisions of the Electric Power Act, No.11 of 1997. Private Sector Participation in RE. Thus far there has been no private sector financing of rural electrification except in the context of auto-generation and informal distributors. Informal distributors are not legal given KPLC’s monopoly distribution rights and hence there has been no systematic effort to collect data on the performance of these activities, and yet they do play an important role in electricity in some places such as Namanga on the border with Tanzania. 4. Outcomes of the RE Program With the indicated levels of expenditures and financing, since 1973 the RE had connected 101,793 customers by the end FY2004/05. The cumulative numbers of connected customers for selected years since 1985 is shown in table 19 below: Table 19: evolution of the number of customers connected under the Rural Electrification Program Fiscal Year No of customers 1984/85 6,069 1989/90 19,067 1994/95 43,718 2004/05 101,793 These are clearly modest achievements over a 30-year period and in relation to the Government’s development objectives for rural electricity expansion. The issues described above (institutional and management constraints, depletion of the REF by operating losses worsened by the allocation formula and by inadequate tariffs, high investment costs) have contributed to this modest progress in the quest for rural electricity expansion. In addition, another serious constraint is the recent GoK policy change to confine RE program to financing public and commercial services (administrative and health centers, schools, commercial undertakings and productive uses) and the high costs of connecting customers. With this change, RE customers are required to pay a flat connection fee of Ksh 15,000 for single phase connections and Ksh 40,000 for three-phase connections, plus taxes. The current policy of the Ministry is to confine the RE support to institutional and commercial customers. In practice, this means that efforts are concentrated on trade centers where public and commercial services (administration, health centers, schools, shops, jua kali (artisans) other productive uses, etc.) are generally located. As a result of this preference for public and commercial services, domestic customers are considered not eligible as such for connections financed under these RE schemes. It is then for KPLC to undertake the additional enrollment campaign and works to connect these domestic customers under the condition that these customers accept to pay the connection fee charged by KPLC. If the potential domestic customers are less than 600 meters from existing transformer, they are required to pay a flat connection fee of KShs31,500 for 44 44/81 single phase connections and KShs 43,000 for three-phase connections. Customers located more than 600 m from existing transformers are required to pay full connection costs based on a quotation by KPLC. The absence of a financing mechanism for customers to spread payments over time (there is currently a possibility of splitting the cost in three payments only, which is not widely implemented as per observations during field visits), such as through the monthly bill is a serious constraint as many potential customers cannot afford the high connection costs. The policy of excluding households from support under the RE program combined with KPLC’s high connection costs results in cases where, households in close proximity to RE and with ability to pay monthly electricity bills, remain without electricity because they cannot afford to pay connection costs. On the ground, this policy appears also to be difficult to apply without generating problematic side-effects.  A significant share of potential new beneficiaries is lost for a while: To the extent that both the processes of financing and enrolling customers are different in the respective RE and KPLC schemes, it appears to be generally impossible for KPLC to piggy-back on RE schemes and connect the domestic customers in the particular area. On top of that, the connection fees charged by KPLC are perceived as a major bottleneck by the customers, even if some limited flexibility regarding payment is theoretically proposed but not always applied in practice. As a consequence, it may take years before these customers are connected; the potential for increasing access to the RE schemes will thus remain marginally realized.  Evolution of access rate becomes more difficult to measure: It has been observed in the field that in order to overcome these barriers, potential customers organize themselves in a way that only one customer formally applies for connection and the others – either shops or households - get secondary connections from him. The result is a growing disconnect between the number of officially registered domestic customers and the number of households actually getting access to electricity. This may make it more difficult for the GOK to measure progresses toward the targeted 40% access rate. IV. Solar Photovoltaic Electrification While there are no reliable statistics available on individual solar systems in Kenya, different sources agree around the rough figure of at least 200,000 such systems installed in the last decade. Since this is a private market and retail is split in a number of retailers, it difficult to get precise data to check this estimate. Nevertheless, this estimate seems to be consistent with the annual volume of PV modules imported, which according to members of the Kenyan PV Association (name to be checked), is around 30,000 solar modules imported in 2005, some of which were re-exported to other countries in the sub- region. In regard to global capacity that these sales may represent, a study done in 2001 indicated a 20% annual average growth over the period 1995-2000. . A more recent study done in 2004 indicated that imports were around 800 kWp in 2004 (see Table 20)14. 14 K. Magambo, Rencon Associates, for Electric Link Ltd, April 2004. 45 45/81 Table 20: PV Imports/Sales into Kenya, 1999 - 2003 Year 1999 2000 2001 2002 2003 2004 Total kWp 430 510 670 750 730 800 Source: 1) Kenya Revenue Authority Statistics, cited by Magambo, 2004 2) Electrik Link Ltd, Magambo, April 2004 Besides the 9 main importers/distributors15, the total number of “PV dealers� is estimated around 150, most of which are small retailers. As a result, it can be said that the PV industry is well established in Kenya and has developed far beyond the early stage generally observed in other Sub-Saharan countries. When compared to the smaller figure of 101,000 rural customers connected to the grid by the Rural Electrification Program since 1973, the contribution from the fully private based PV industry to the increase of rural access to electricity service in the last decade is far from negligible. 1. The market is still driven by low cost small and low quality products The PV sector is facing a number of problems that considerably limits the effectiveness of its contribution. One of these is the low quality of the cheapest and most sold products. These cheap products are generally installed by the users themselves without a proper selection of the battery size or installation of a load controller. As a matter of fact, the very small size of systems bought in Kenya is one of the parameters that limits its growth, since such modules are not the priority of the main international manufacturers. As a result, this demand has to be served as a secondary market of the marginal production that does not comply with any specification of the main market for large systems located in industrialized countries. {to be reviewed} As a consequence, while there is no real survey of which proportion of the approximately 200,000 systems are actually working today, it can be estimated that more than 60% of the systems may not be working properly or even working at all today. Two different segments can be roughly distinguished: - a large segment of the very small systems (10-20 Wp), consisting mainly in amorphous silicon solar modules bought in small retailers shops by customers looking for the cheapest price, some of them of bad quality and installed without load controllers. Typically, modest customers will buy a cheap, low quality amorphous panel for 3,000 ksh, a car battery for 2,500 ksh and 2 to3 twelve volts incandescent bulbs for 200 ksh, for a total cost of 5,700 ksh (i.e. 83 USD). Today this segment represents probably more than 80% of total number of systems sold and 50% in kWp. - a limited segment above 45 Wp16, consisting mainly crystalline solar modules, which is subdivided by richer individuals that can afford them and in institutional 15 The main importers/distributors are: Kenital Ltd, Chloride Exide Ltd, Solagen Ltd, ASP Solar Ltd, Telesales, Solar Innovations, Electric Link, Sangyung Enterprises and Davis and Shirliff. 16 Magambo distinguishes between two sub-segments: (i) 20-60Wp, representing about 35% in kWp and (ii) >60Wp, representing the remaining 15%. 46 46/81 programs17. Most of these systems benefit from proper design regarding adequacy of size to needs and to battery and include a charge controller. According to observations made in Nairobi, the unitary retail price of a complete good quality 50 Wp system (50 Wp crystalline panel, one charge controller, 8 lamps 11W, one 100 Ah solar battery) is around 34,000 ksh (i.e. 496 USD). The main issue is the payment capacity of the majority of customers, who end up buying very small systems (20 Wp and below) of bad quality. It should be noted that several brands of amorphous silicon solar modules have improved the quality of their products in the last years. However, according to a recent study, there are still at least two widely available brands that perform well below their advertised levels. A couple of programs have tried to address this issue as a means of improving the quality and the size of systems. According to available surveys, there is a contrast between the very limited savings capacity and the quite significant capacity to pay on a monthly basis for electricity services. As a consequence, financing the purchase of the systems should allow customers to get access to better and larger systems. PVMTI, a program co-funded by IFC and GEF initiated in 1999, is the most important program that tried to enhance the financing capacity of the sector and increase the share of good quality products sold in the Kenyan market, mainly by offering loans to intermediaries that would in return offer loans or lease to individual customers. However, one of the main lessons of the PVMTI program is that the appetite for loans of both the financial intermediaries and final customers is surprisingly low. While there may be some need for financing from part of the intermediaries, the main need expressed by them seems to be more for risk sharing and guarantee. Regarding the final customers, the debt capacity is generally limited to one single loan at the same time. This is typically the condition imposed by micro-finance institutions and saving and credit cooperatives (SACOS). As a result, individuals prefer not to preempt the possibility to take a loan for more urgent priority, especially if the loan for PV has to be several years long to make installments bearable. As a consequence, the PVMTI programs faced major difficulties to commit its 5 million dollars credit line, and even after finally managing to commit it, it couldn’t achieve significant results in terms of number of PV systems actually purchased by final customers (less than 200). The Ministry of Energy is preparing a program intended to increase the installed generation capacity to 12 MW by the year 2008/9, by developing the institutional market to electrify 120 schools and to increase PV connected consumers from current estimated 200,000 to 350,000. While institutional programs are important to ensure a core business where quality is ensured through technical specifications imposed in tender documents, the issue remains how to ensure that modest rural household can get access to better quality and larger PV systems that would fulfill their basic electricity needs (principally lighting, radio and/or 17 For instance, ongoing Government electrification project for public boarding secondary schools in Arid and Semi-Arid Districts of Northern, Eastern and Rift Valley Provinces. 47 47/81 TV) for which they are spending relatively large amounts of money for bad quality products (paraffin, battery charging). The economic calculation below illustrates that overcoming the barrier of the initial cost can contribute significantly to address this challenge. 2. Fee-for-service arrangements as a way to make quality and size affordable Field observations have shown that when there is no grid, current expenditures for a single household can range around 300 to 700 ksh A particular household using a genset indicated that it spent up to 300 ksh per day on gasoline; one small mini-grid is charging 500 ksh per month for one single bulb 6 hours per day. It is possible to estimate what could be the value of a solar system that an Energy Service Company (ESCO) could install in rural household if charging a similar monthly bill. Fee- for-service means charging a monthly bill against the delivery of a service, which is typically what electricity distribution company does when charging a tariff to its clients. This calculation has be done for the case of a customer who is currently paying 500 ksh per month either for paraffin oil or for getting a limited service from a private mini-grid and would pay the same amount to an ESCO for a better service. Assuming that the ESCO will deliver the service to that customer for 20 years, and assuming a remuneration above the pure-cost recovery of 14% a year (i.e. a discount rate of 14%), the net value of the revenue that the ESCO will get from this customer is more than 42,000 ksh (i.e. around 620 USD). This has to be compared to the retail price of the good quality 50 W solar system indicated above (34,000 ksh, i.e. USD 496). Considering the economy of scale from which an ESCO will benefit, wholesale price of a similar system will be significantly lower, and even considering that the ESCO will have to replace the battery and fixed operational costs, this calculation shows the breadth of the improvement that a well designed fee-for-service program could bring to this customer. (to be reviewed) The result compared with the value that a micro-finance institution could finance against the same monthly installment of 500 Ksh over 12 months at a 14% interest rate. A similar calculation shows that the loan would not be more than 6,000 ksh (i.e.82 USD), which is more or less the value of a low quality 15Wp system without load controller. Table 21: economic calculations based on current energy expenditure that can be displaced by individual solar systems Ksh USD Monthly energy expenditure that can be substituted by a 500 ksh $ 7.3 solar system Value of 1 year loan that can be financed 5,593 ksh $ 82.0 Net present value of ESCO revenue for fee-for-service over 42,370 ksh $ 619.0 20 years 48 48/81 3. Main lessons of the Kenya solar experience: The main lessons derived from this analysis are the following: - there is a real demand and acceptability of PV systems on the Kenyan market; - the market is well developed in regard to import and retail network; - there is a significant capacity already built regarding designing, installing and maintaining PV systems; - there may be a potential for significant growth of sales of larger and better products if the initial investment cost can be overcome; - financing individual purchases through individual loans is not an appropriate answer; - fee-for-service model is worth to be explored as a means to provide good quality electricity services to individual households through individual photovoltaic solar systems. 4. Integrating photovoltaic systems and grid extension in large rural electrification projects It is important to stress that a fee-for-service model does not necessarily restrict the choice of the technology to serve customers to solar systems. A conventional power distribution company is a de facto a “fee-for-service� electricity service provider, in which the fee is determined by the tariff and the service measured in kWh. To the extent that level of services are similar – which is a reasonable assumption for modest rural households requiring only lighting and electricity to power radio/TV - a distribution company can either provide the service through a connection to the grid or through the installation of an individual solar system. The parameter to decide which technology should be chosen is the cost of each solution. Studies have compared the cost of delivering electricity services through the grid or through individual solar systems. Regarding the cost of the grid solution, key parameters are the number of customers to be connected at the same time and the length of the grid to be built to connect them to the existing grid. Such comparison is not easy to do because it requires a detailed data on number of rural electrification projects, and also a reliable estimate of cost associated to the installation and operation of solar systems. While such a comparison could be done using Kenyan figures, the exercise couldn’t be realized in the context of this study because of lack of sufficient data. However, for indicative purposes, figure 4 below presents the results of the comparison made in the context of poor peasant regions in the northeastern part of Brazil. This figure shows that for small groups of less 25 households (around 150 persons), individual photovoltaic systems (PV systems) are cheaper than the grid extension if these groups are located at more than 6 kms from the existing grid, assuming a cost of 700 USD per individual PV system. 49 49/81 Fig.4 Kenya is characterized by very high regional contrast in terms of demographic densities (see maps in Fig 2 and 3). According to statistics published by the Central Bureau of Statistics of Kenya, on the one hand, 11 out of 70 districts had a density lower than 19 persons per square kilometers. These 11 districts, located in low lands in the north, the east and the south of the country represent more than 60% of the national territory. On the other hand, exceptionally high rural densities are observed in the high and fertile lands of the western region, as high as 500 persons per square kilometers. As a result it is expected that the role the solar photovoltaic technology could play in achieving the ambitious target of 40% rural access at the national level would differ from one region to the other. V. Analysis of Options for Faster Access Expansions 1. Global Institutional Framework under the National Energy Policy For many years Kenya did not have an energy sector policy and strategy for guiding expenditure priorities in the sector. Sessional Paper No. 4 on Energy, which is consistent with the ERS objectives, charts the way forward and provides key policy directives that will govern and shape the implementation of various activities particularly the Rural Electrification Program in the foreseeable future. Some of the policy strategies include: 50 50/81 (a) Establishment of a Rural Electrification Authority that provides an institutional framework for rural electrification and strengthens the Rural Electrification Fund (REF). This is to promote both grid and off-grid rural electrification with a view to increasing rural access to power from the current 4 percent to at least 40 per cent by 2020; (b) Revising the legal framework for the energy sector to: (i) create more space for private sector participation through the creation of a common access power transmission system (a separate transmission company) and removing the Kenya Power and Lighting Company’s (KPLC) monopoly for distributing power; and (ii) strengthen the regulatory framework by harmonizing the institutional set-up in the energy sector for the petroleum and electricity sub-sectors by bringing all regulatory functions of these sub-sectors under one entity, the Energy Regulatory Commission; (c) Developing a plan to increase utilization of renewable energy particularly plans to harness more solar (through increasing installed PV capacity by 10% annually), wind, and small-scale hydros, as alternatives to grid extension for the provision, of electricity for telecommunications repeater facilities, water heating, crops drying, refrigeration, and water pumping; and (d) Promoting public private partnerships in the expansion of the rural electrification program. Therefore, Sessional Paper No. 4 of 2004 opens a wide range of new options and arrangements to implement the objective of scaling access to electricity. 2. Recommendations and Options for a New Institutional and Financial framework This section explores the different institutional arrangements that could be considered to face the challenge of scaling up access to electricity services in Kenya in order to achieve the goal set by the National Policy. The different key principles/recommendations that should structure this new arrangement according to the diagnosis of the current situation and to international experience are first presented in sub-sections 1 to 6. Section 7 lists different forms of public private partnerships that could constitute a mean to unleash the implementation capacity of access expansion programs to be funded by both the international donor community and the government of Kenya. 2.1.Revised modalities for customer financial participation: A basic condition to ensure that electrification is viable, is to ensure that the customer’s financial contribution will at least cover the O&M costs plus amortization. If not, the activity will generate a structural deficit and will not be able to survive on a standalone basis. This is supposed to ensure that potential beneficiaries effectively become customers and to define the appropriate level of customer financial participation. 51 51/81 2.1.a. A new tariff structure that takes into account operational costs and new customers willingness to pay Almost 40% of current KPLC customers belong to the A0 category and it can be taken for certain that the proportion of similar modest customers will be even higher in rural areas when getting access to electricity. The current A0 tariff is 1.55 Ksh/kWh for energy, which corresponds to 1.9 USc/kWh. Considering that the least cost generation options are around 5 USc/kWh (Olkaria II extension cost is estimated at 4.9 USc/kWh), this means that current tariff does not cover half of the generation cost. Without tariff increase, rural electrification will not be sustainable, even if investment cost is highly subsidized, since revenues do not even cover O&M costs. However, the ability to pay is significantly higher, even in poor non-connected rural areas, demonstrating that there is room to adjust tariffs. Results from few households surveyed during field visit, indicated that current energy expenditures for lighting and radio/TV in non-connected rural areas is above 300 ksh/month, which is far more than the average bill charged to A0 rural customers. The value of 300 Ksh/month corresponds to a monthly consumption of more than 200 kWh/month, when average monthly consumption of A0 clients is around 20 kWh/month. Such substitutable energy expenditures are generally considered as representing a conservative estimate of what people would accept to pay to get far better lighting and electricity for TV and radio. As a result, a new tariff policy is required as a pre-condition for sustainability large scale access expansion programs. Moreover, since metering small and distant consumption is very expensive and may increase significantly O&M costs, it is recommended that flat tariff fee option be developed for modest consumers. The government has already hired a consultant to revise the tariff structure. Depending on the outcome of this global study, additional tariff refinements may be required to ensure cost recovery under specific regional circumstances. This would be of the responsibility of the regulatory commission. 2.1.b. A new connection policy that maximizes both the number of beneficiaries and revenues Adjusting the tariff structure is not enough. If too many targeted customers do not meet the conditions to be connected, then there will not be enough revenue and again the system will become unsustainable. In fact, to the extent that the tariff issue is addressed, the more clients served, either domestic, service or productive, the bigger the revenue and the more sustainable will be the corresponding electrification project. Today, the current connection policy, especially as it is applied on the ground, does not create the enabling environment for maximizing the number of customers. Regarding the use of RE Fund resources, the current policy of the Ministry is to concentrate on RE schemes serving public and commercial services (administration, health centers, schools, shops, joacalis, other productive uses, etc.). In practical, this means generally the concentration of efforts on trade centers where these services are generally located. Whenever KPLC is asked by the Ministry to electrify a trade centre, a load survey and the design of the scheme are done by KPLC. KPLC also does the tendering of the works to be executed and the selection of a contractor (to be reviewed). 52 52/81 Customers are asked to pay flat connection fees of, respectively, 15,000 Ksh for single phase connections and 40,000 Ksh for tri-phase connections, plus taxes. Usually, the localities to be electrified are first proposed to Ministry by District Development Councils (DDC). As a result of this preference for public and commercial services, domestic customers are considered not eligible as such for connections financed under these RE schemes. It is then left to KPLC to connect these domestic customers under the condition that the required financial contribution from the customers is secured. If the potential domestic customers who apply for a connection are within 600 meters from existing grid, then KPLC charges them a flat connection fee, which is 34,000 Ksh for single phase connections and 42,000 Ksh for tri-phase connections, plus taxes. If they are located beyond 600 m from the grid, they receive from KPLC a quotation of the full cost of the works that are necessary, which determines how much these customers have to pay to be connected. In both cases, there is some flexibility for the customers to pay the connection cost in up to three installments; however, on the ground observations have revealed that customers are frequently not aware of such an option. The same policy applies to projects forwarded to KPLC by Constituency Development Funds (CDFs). In a situation where the demand of a customer changes, i.e. from pure domestic to productive use, which necessitates a need for the reinforcement of the grid upstream, the customer has to reapply under a new category and to receive a new quotation to determine how much he/she has to pay for the reinforcement of the grid. On the ground, this policy appears to be difficult to implement without generating at least two problematic side-effects. The first problem is that a significant share of potential new beneficiaries is lost for a while. To the extent that both the processes of financing and enrolling customers are different in the respective RE and KPLC schemes, it appears to be generally impossible for KPLC to piggy-back on the RE schemes and connect the domestic customers of the area. On top of that, the connection fees charged by KPLC are perceived as a major bottleneck by the customers, especially when it appears that connecting them would require an increase in the capacity of the grid, which initially meant for only the priority customers (public/commercial services); in that case, the full cost of the works are charged to them by KPLC. As a result, it may take years before these customers are connected. Because of these constrains, the potential for increasing access is only marginally realized, and the revenues obtained from the restricted number of customers connected remains limited, thus reducing the return that could be expected from the infrastructure assets that have been built. The second problem is that the results of the program in term of access expansion become more difficult to measure. It has been observed in the field that, especially in densely populated areas where potential customers are close form each others, households spontaneously develop strategies to benefit unofficially from the grid extension. As result these households are not accounted as beneficiaries in official KPLC 53 53/81 statistics. To overcome these barriers, potential customers organize themselves in a way that only one customer formally applies for connection and the others get secondary connections from such a customer. This occurs for instance in trading centers where shops officially connected offer secondary connections to other shops or to domestic tenants. This also occurs in highly densely populated peasant areas where one member of a community formally applies and secondary connections serve several neighbors. This results in a misrepresentative between official KPLC figures regarding the number of domestic customers and the number of households actually getting access to electricity. However, as mentioned above, it seems that a number of potential customers can afford to pay the monthly bills but cannot pay upfront the connection and internal wiring costs. To overcome this bottleneck, which seems to prevent a significant increase of the access rate, it is suggested to create the possibility to spread these costs in the bill over several years, as an additional separate component in the bill besides the consumption charged according to tariff. This would constitute in fact an extension of the basic principle that underlies the calculation of tariffs, which is to spread generation, transmission and distribution cost along the economic lifetime of the equipment and then charged it to customers through the tariff. Such a measure, which has already been adopted in other countries (i.e. Senegal), may allow for a far greater proportion of households located in area already electrified to apply for connection. The government is currently initiating a study to revise the current connection policy. Depending on the outcome of this study, which may not consider the new institutional arrangements proposed in the next section, additional specific modalities may need to be developed and enforced by the regulatory commission. 2.2.A new Financing Mechanism: Assuming that both the tariff structure and the connection policy have been adjusted to solve the above mentioned issues, and that on the basis of what has widely been observed in other countries, one can expect that the cash flow expected from the sales of electricity services will both cover the O&M costs of appropriate low cost delivery arrangements and serve a certain amount of debt or capital, there remains generally a gap to fully finance rural electrification investments. This gap defines the amount of first cost subsidies that need to be financed under an appropriate financing mechanism. This was one of the objectives of the original Rural Electrification Fund, created in 1973. However, as explained in above section, the fund was also used to cover ex-post calculated operational deficit of the RE schemes, which is a function which should not be required anymore, except for the existing 5 large diesel based isolated grids. To the extent that these systems are integrated to KPLC assets, such operational unbalance should be addressed as a specific issue – which is not related to objective of access expansion - by both the government and KPLC. Further studies when preparing specific access expansion projects could determine precisely what the O&M costs are associated to these projects what are the expected revenues from the customers, what share of the investment cost can be covered by these 54 54/81 revenues after discounting O&M costs, and thus what is the gap to be fulfilled by the financing mechanism. As detailed below, the amount of public financing to fill the gap can be optimized by a competitive process. It is acknowledged that scaling up access require to set up an appropriate financing mechanism to channel the investment cost subsidies and financing that are necessary to fill the gap between the resources that can be brought by the customers and the service provider. According to the lessons from the PVMTI project, the capability of the Kenya financial market to finance the share of investment cost that can be borne by the service provider, either KPLC or other private operators, may not be enough. As a consequence, there may also be a need for setting up a guarantee fund. Possible sources of funding for the REF are:  Levies charged on electricity customers paid through their bills;  International financings from donors  Financial contribution from the Government budget;  Revenues from REF activity, i.e. reimbursement of loans made by the REF to access providers if any. The proposed Energy Bill will provide for the establishment of a new Rural Electrification Fund, which will substitute the current one. Detailed modalities for this fund need to be further developed. 2.3.Private Public Partnerships to leverage additional implementation capacity While additional funding like the support provided by the French and the Spanish cooperation agencies allows KPLC to increase its contribution to electrification, the objective of 40% access by 2020 seems to be far beyond even improved KPLC capacity: preliminary figures shows that this objective will require around 2 million additional connections in less than 14 years, when current total number of KPLC is only 787,804. According to its statistics, KPLC has connected only 144,565 rural customers ( 85,714 rural customers under RE schemes plus 58,851 rural customers connected out of RE schemes) in the last 33 years (since 1973). And, while precise statistics are not available, in most localities electrified by KPLC, performance regarding rate of connection appears to be still rather low, generally less than 50%, even in urban areas. As a consequence, the achievement of the ambitious national objective will require unprecedented efforts both to densify connection in areas already covered by KPLC and to accelerate the implementation capacity in the vaster remaining areas. Since KPLC human resources that can be allocated to expanding access are limited, it is, therefore, recommended that KPLC’s mission should be reviewed so that it is made responsible for access expansion in the areas where it already has MV and LV lines and those it has plans for extending access to in the next 2-3 years. It is also recommended that these areas should be listed as to constitute the territorial basis of the KPLC distribution license. 55 55/81 Outside this “KPLC domain�, legal and regulatory conditions should be created for Public Private Partnerships, as a way to attract new players and increase the implementation capacity. The national policy has already stated this possibility as an option to be promoted and the proposed bill has made provisions to ensure a legal basis for such private participation by allowing the issuance of different types of licenses and permits. Lessons from previous projects, such as the PVMTI, consultation with different Kenyan stakeholders and on-going public private partnership in other countries in the region show that there is a number of national and international stakeholders that are interested, and sometimes even already involved (agriculture cooperatives, PV dealers, SACOS, hire purchase companies, banks, contractors, etc.), that could play a more active role to implement the ambitious access target set up by the Nation Energy Policy. Box 1: Rural Cooperatives and Rural electrification: The experience of previous rural electrification programs implemented with cooperatives, including the Cofferr Factory electrification project financed by the European Union and the PVMTI, illustrates at the same time (i) the importance of electricity and the willingness of cooperative to get electricity in order to reduce dramatically their energy cost and increase their productivity and (ii) the difficulty for rural stakeholders that are not from the sector to ensure their own electrification. By releasing such constrains, a rural electrification program can thus leverage very important impacts on the activities of these cooperatives. These cooperatives are also used to providing inputs to their members. Generally members have an account were the revenue of the sale of the production is transferred and from which payments for goods purchased by the member are retained at the source. SACOS cooperatives provide loans against savings accumulated by the members (the ratio is generally around 3 to 1). Collection of bills and other O&M activities that require direct interaction with customers (bill payment collections, enrollment of new customers, reporting of incidents at customer level, etc.) are very costly for utilities in rural areas. The reason is that such activities require long trips that are staff time consuming and generate high transportation costs. Securing payment is also difficult because the cost of enforcing default payers disconnection is also high and often difficult to manage socially. The presence of a cooperative in the area offers great opportunities to delegate efficiently such functions at a far lower cost: easy communication channels with potential customers are already in place, payments can be retained at the source, and other functions can also be delegated (see box 2). As a consequence, the advanced development of rural cooperatives in Kenya is a trump card for potential providers of electricity services in rural areas, to reduce the cost and improve the reliability of the commercial development and management of customers and revenues. Production and marketing cooperatives have already developed and centralized the demand for significant productive uses that will ensure immediate load and SACOS can provide members very competitive financing for the additional equipments that can be ran after electricity is there. Box 2: Functions that can be fulfilled by rural grassroots organizations 1. Enroll applicants, collect down payments, identify bad debtors beforehand. 2. Assist applicants to choose the level of service most appropriate for their needs and payment abilities, primarily during the period that new users are brought in (this ability to pay is easily ascertainable for 56 56/81 local rural organizations since all members have similar activities and are well aware of their neighbors’ level of business). 3. Run the payment-collection service on behalf of both the operator and the user. This service makes the payment period flexible and makes it possible to spread out payments to the association since it is always accessible to payers. In the case of individual payments at the market or at a banking institution, this flexibility is limited. Above all, this service eliminates the cost of travel that would be paid by the user in the case of individual payments at the market or at a banking institution. This cost can represent a significant increase to the base rate. 4. Guarantee the continuity of payments from customers that are temporarily experiencing financial difficulties (sickness, accident, etc.) by making their payments for two or three months, in accordance with current practices of solidarity in the region. 5. Pass along to the operator the users’ requests for upgrade/ system changes. 6. Facilitate technical communications between users and the operator by identifying and resolving minor problems and by correctly recording more complex problems. 7. The cooperative/ local organization can better manage the social cost of the decision to take a system back from a bad debtor or to exchange it with a less expensive system. 8. Be involved in managing a local “guarantee fund", the surplus from which will be returned to village associations that are project partners. 9. Foster an attitude of individual and collective responsibility concerning the equipment and the proper functioning of the project. The users and the associations see that the project will provide them with a benefit that they really want, so they are inclined to contribute actively to the project’s success. This includes carrying out certain tasks as a form of collective contribution in addition to the individual payments. 10. Pass along to the operator the requests to buy household electrical appliances that would work with the systems, assuming the operator develops a commercial business to distribute the most appropriate equipment. 11. Help to train new users to decentralized electrification systems (PV, windmills) during the first few months of operation. 12. Manage and paying monthly bills for collective equipment such as: lighting for a community room and public lighting, church, video devices and parabolic antenna/ decoder for community use, collective pump system with a fire hydrant, school, computer, internet modem, etc. 13. Manage and paying monthly bills for productive equipment used collectively such as equipment for processing crops. 2.4.Ensure that the regulatory framework provides for integration of new technical solutions Besides the conventional grid, new and more cost-effective technologies are available. On one hand, low cost grid technologies like single wire return by earth (SWARE) have been tested in the past by KPLC and implemented with success in different countries that reduced significantly the investment cost required to serve low and dispersed loads. It is therefore recommended that the technical specifications currently enforced by the ERB be revised and adapted grid standards be provided for rural areas. On the other hand, over the last decades, technological progress has benefited more to the new decentralized electrification systems than to the already very mature grid technology. This applies to solar photovoltaic individual systems, to windmills and to hybrid systems. As a result, the maximum distance of competitive expansion of the grid from existing one against decentralized solutions has been considerably reduced, especially when it comes 57 57/81 to electrifying small rural communities (see economic comparison between grid and PV in section above). On top of that, the lack of an appropriate electronic database and hardware/software for modeling MV networks doesn’t allow either for anticipating upstream constrains that may result from the connection of new grid schemes and customers, or for estimating the cost and planning appropriate reinforcement investments. As a result, future beneficiaries may not get the full benefits from the new assets created under national and international funding. On the contrary, already connected customers may suffer a degradation of the service. Efficient use of public resources requires that the most cost-effective technical solution be selected. The regulatory framework for electricity services delivery should be completed to ensure that new cost effective technologies are not discriminated against, including when it comes to the technology to be used by distribution licensees to comply with their obligations. Since the nominal installed capacity of decentralized individual systems is generally very low, for instance far below the 100 kW threshold below which no permit is required, it is recommended to address regulatory issues of such systems in the contracts under which public resources would be injected in Public Private Partnerships (regulation by contract). These regulatory clauses should, among other issues, address minimum technical standards applicable to components and minimum service level associated with regulated tariffs (especially for flat fee tariff if any). In case grid extension is considered the most cost effective solution, assessment of the impact on upstream grid and provision for necessary reinforcement should be mandatory. 2.5.Set up a strong institutional champion for expanding access Developing and multiplying Public Private Partnerships presupposes an important work of promotion, technical preparation, procurement procedures, financial resource management and monitoring that require the set up of a specific institutional capacity. Since this is related to implementation of the energy policy and not to its elaboration, it is recommended that this capacity be located outside the Ministry. It is also recommended that a specific institution be set up for access expansion rather than merging it in the current power sector entities, i.e. KPLC. The reason for recommending such a separate arrangement is to avoid new access expansion efforts ending up being high jacked by inefficiencies, lack of specific capacity – especially regarding innovative solutions - and different ranking of priorities (urban growth, industrial demand development) that generally rule the power sector. One of the key lessons from rural electrification programs in other countries is that it is essential to set up institutionally a strong and autonomous champion with a commitment for expanding access to electricity. As a matter of fact, as noted above, the proposed Energy Bill will establish a Rural Electrification Authority (REA). It is recommended that the composition of the board should be balanced, that is representatives of the civil society be integrated in a proportion that guarantees the 58 58/81 board’s autonomy. It is recommended that at least five members be from non- governmental institutions and be selected on the basis of competence in the following respective areas: - economy - finance - electrical engineering, including experience in - legal - rural development It is suggested that this REA be in charge of the allocation of the resources of the Rural Electrification Fund to be created by the new law. The REA should also be responsible for the preparation, the tendering process and the awarding of the contracts implementing the options to be retained among those proposed in the following section. It is also suggested that the regulatory commission should delegate certain functions to the REA, for instance, the elaboration of technical specifications (that are different from standards, which may be referred to in technical specifications), the selection process of licensees or permit holders for rural electrification, etc. 2.6.Multi-Sectoral Partnerships for maximizing ratchet effect on rural development Working on the premise that the advent of electricity in an un-served area does not always spontaneously induce expected use of electricity for social or productive uses, the REA should also be in charge of specific activities to ensure that these highly beneficial uses of electricity in terms of poverty alleviation take place and be maximized. Since the main barrier is that the development of social and productive uses of electricity requires coordinated multi-sectoral actions that seldom occur simultaneously, the REA should develop multi-sector energy sub-programs (MSEPs) together with other entities, agencies and ministry departments in charge of programs in other sectors. As a matter of investigations made during this study and presented above that there is a considerable potential for increasing impacts of projects and programs executed on the same territory by other sectors. The reason is that these projects and programs are facing technical and financial constraints to provide adequate energy infrastructure, especially because they cannot provide for economies of scale to get access to cost effective and reliable energy equipment and O&M services. MSEPs should provide the specific coordination needs, the technical assistance, and - when necessary - last resort financing for purchasing and installing energy equipment to ensure full effectiveness of multi-sectoral interfaces between access expansion projects supported by the REA and other sectoral programs. 2.7.Options for Public Private Partnership to scale up access to electricity The participation of private companies in developing access to electricity inside or outside the areas already served by KPLC can be sought under a range of legal arrangements which vary according to the degree of autonomy and protection awarded by the sector authorities to the participating private companies. This ranges from simple works contracts, as done currently under the RE agreement, to fully integrated licenses awarded, to competitive bidding, including a number of intermediary options of which 59 59/81 advantages and limitations can be examined in the specific geographical and legal context of Kenya. 2.7.a. Options to increase access rate inside “KPLC domain� as defined by the French consultant proposal: Option a: Inject more financing into current KPLC electrification procedure: Under this scenario, distribution infrastructure and connection investment financing is boosted but no change is introduced regarding the procedures under which KPLC connects new customers located in sub-locations where its MV grid is already present18. KPLC first register applications from new customers, then makes quotation and design of works to be done, and, accordingly to its procedures, awards a contract to a selected contractor to do the works on a turnkey basis. Only the financing constrain is released by allocating earmarked loans to KPLC from a revolving fund. Option b: Improve the efficiency and the pace by preparing larger volumes of projects awarded to same contractor in the same area: KPLC creates an internal commercial and technical taskforce specialized in electrification supported by an international consultant to bring in international experience, and which moves from one district to another to boost the local commercial and technical KPLC capacity to design and implement either urban of rural electrification schemes, including reinforcement of MV lines when necessary. As a result the applications are boosted and a large number of “densification projects� are elaborated quickly in same district and are awarded all together to contractor that should meet higher implementation capacity requirement. This option is quite similar to the arrangements adopted to implement the electrification program financed by the French cooperation, but adding to it the diagnosis and works of the reinforcement of upstream MV lines when necessary. Option c: Delegate to larger private sector entities the densification of whole compact areas Areas where statistics show that there is a significant potential for densification, are delineated and bid to private firms for densification. The contract includes both the commercial part and the works, including the technical design. Assets are transferred to KPLC for operation when commissioned. Preliminary studies should determine a reasonable volume of financing required to reach at least 40+% electrification rate in the considered area. The winner would be the candidate that would commit for the higher number of connections. The payment would be done partially or totally on an OBA basis. OBA, i.e. output based aid, means that the payment is made progressively after commissioning effective connections. It can be added as an additional specification of the tender that all social and commercial services that apply for connection in that area shall be included in bidders proposals. A variation of the above principle is to allow the winner to continue to connect additional customers 18 While procedures are not changed, a revolving fund will be established to increase connectivity. 60 60/81 beyond the initial number committed against a fix average payment per connection to make it attractive to increase the connection volume. To a certain extent, this is similar to the Revenue Management Contracts experimented by NEPA in Nigeria to reduce high non commercial losses which were able to attract 9 international firms. This model could be applied either in rural or urban areas and could be tested in the Kibera slung area. 2.7.b. Options to increase access rate outside “KPLC domain� as defined by the French consultant proposal: Option a: Turnkey contracts for clusters of RE schemes: This is the model currently being implemented under the Spanish and French funded programs. On the basis of a revised master plan, RE schemes are identified and pre- designed by KPLC. Instead of bidding limited number of schemes to each contractor, a large number of schemes (dozens) are bid at the same time. Advantages: This option is already road tested through the Spanish and French funded programs, and as such will benefit from the lessons learned when implementing these programs. Limits This option does allow neither to leverage additional private human and financial resources nor to bring in innovation related to decentralized solutions. It does not increase the management capacity of rural electrification beyond the current limited KPLC capacity ( e.g. does not improve operational efficiency or address marketing problems after initial connections). Resulting cost per customers are also considered by the GoK to be relatively expensive19. Option b: Distribution licenses for large rural areas Because of the insufficient pace of electrification in the past, we recommend to develop Public-Private Partnerships to attract new players in the area of rural electrification and thus leverage additional resources and implementation capacity. Based on international experiences where specific rural distribution companies or cooperatives have been established in both developing countries and in industrialized countries (Bangladesh, Morocco, South Africa, Senegal, USA, France, etc.), an alternative that can be envisioned in Kenya is to award rural distribution licenses to the private sector through a competitive bidding process., jointly with generation licenses where interconnection with KPLC is not possible. The master plan and other preliminary studies would delineate the adequate size perimeters of supply areas that can be made viable and attractive to new players, the volume of investment required to achieve a minimum target in terms of access rate in that area, and an estimate of the amount of first-cost subsidy that would required to make the activity sustainable and attractive. 19 Source: “Comments on the Rural Electrification Access Expansion Study Draft Report�, Ministry of Energy of Kenya, September 4, 2006. 61 61/81 Each corresponding rural electrification distribution license will then be awarded to the candidate that would commit for the higher number of connections against the pre- determined envelope of subsidies, which would be disbursed on an OBA basis. In order to avoid the bottleneck of the high cost of the connection fee and internal wiring, the contractual commitment of the licensee should cover connection cost, internal wiring and energy conservation lamps. The committed amount of connections would have to be (i) built under a limited time (for instance 2-3 years) and (ii) operated and maintained20 during the whole license period. The tariff would be determined by the Rural Electrification Authority on the basis of surveyed expenditures for lighting (paraffin, etc.) and other substitutable uses (i.e. battery charging for TV) and would be part of the tender. Revenues will be collected directly by the local rural electrification company that the winner will be requested to create in the license area. It is suggested that the choice of the technology should be of the responsibility of the licensee, to the extent that each technology used complies with technical specifications or standards to be attached to the tender document21. The master plan becomes only indicative and no more detailed design of each individual scheme by KPLC is required, since everything is included in the contractual commitment signed with the licensee. KPLC commercial campaigns are no more necessary either, since the licensee will be free to elaborate its own innovative commercial strategy. Advantages: The size of the area bided and the perspective of a series of tendering (which justifies the effort to prepare a bid since incremental cost for the next will be low and the probability to get at least one success is higher), the protection offered by the license (which ensure that revenue from customers can be secured over a long period), and the economic viability (which results from the first cost subsidy) can attract large private companies that are reliable over the long term, able to bring in significant amount of financing and able to achieve quickly ambitious targets. The competitive process will ensure that the candidates will optimize both the investments and the O&M cost, and will create an incentive to maximizing the financing (equity + debt) brought by the candidates besides the subsidy, as a way to increase the proposed number of connections. To win, the candidates will naturally tend to form groupings to aggregate different expertise and comparative advantages. Typically, electrical works firms mastering the grid technology may team up with PV industry firms to get the best technology mix, as observed for instance in Senegal. International companies may team up with local ones to reduce costs. Limitations and risks: 20 Maintenance could/should cover the replacement of components of internal installation, including energy conservation lamps, against return of defective elements and, possibly, a small penalty to avoid free-rider effect. 21 Document norms and standards for different grid technologies should be attached to the tender, i.e. tri- phase, double phase, single phase, earth return, etc. , for solar systems and for windmills, i.e. quality standards for components, minimum quantity of guaranteed energy or lighting hours, etc… 62 62/81 This model supposes to elaborate a quite complex procurement process and to prepare a number of documents that are part of the tender package. This supposes that the REA receives technical assistance to prepare these documents and to become familiar with the whole process. Good preliminary market surveys for each license areas are important to provide enough information to candidates to prepare reliable proposals. Credibility especially regarding financial and regulatory commitments from government is essential. It should be clear that expected cash flow is enough to cover more than O&M cost. All these conditions have been met by the Rural Electrification Agency of Senegal, where economic power of rural seems to be less than in Kenya. Finally, if the whole design is not enough credible, the main risk is the REA not be able to attract enough candidates to award the number of licenses that are necessary to achieve the target. Option c: Built Operate Transfer (BOT) contracts for electrifying large compact areas: This option is similar to the previous but differs to the extent that only functions are awarded but not licenses. The main objective and the process remains the same (bid compact areas for the maximum number of connections against a pre-determined amount of public financing). After several years of operation during which the winner of the bid collects the bills and maintains the equipment he has committed to install, the assets and the activity are transferred to KPLC. The period of operation has to be 15 years at least to ensure coverage of private financing brought by bidders. From the legal perspective, the selected firm works formally as a subcontractor of the licensee, i.e. KPLC. This is the model being implemented in Morocco. The model can be technology specific (for instance only individual PV systems in areas identified by the master plan) or technology neutral. A variation of this option would be to offer the possibility to the customers to buy PV systems through installments along a similar period. Such variation could even be implemented independently, as it is being currently successfully implemented in Bolivia. Advantage: No license need to be issued, which simplifies the process. The choice of technology and the marketing are innovative and more cost effective, and thus more customers are served than in a scheme where the number of customers does not result from competition. The long period of connection generates a cash-flow that allows leverage to private financial resources to be brought by bidders to maximize the number of connections. Limitations: Since the selected firm is a subcontractor, it has no formal link with the regulator and its contract with KPLC is not regulated. Private companies will commit themselves and engage money only to the extent they trust enough in KPLC capacity to honor the contract over such a long period. The risk is that no firm wants to engage money and then the leverage effect is nil. They may not want to take the risk that KPLC, which remains the licensee, interferes in their relation with the clients and put at risk the cash-flow based revenue. There is then no difference with the option c below, except that the contractor can choose the technology (i.e. individual PV systems) and the commercial relation with the clients, including intermediaries like local organizations or cooperatives. But then the risk exists when assets are transferred to KPLC that KPLC lacks experience in managing a very different scheme, which supposes to maintain a large number of decentralized PV 63 63/81 systems and collect bills from customers with no meter and that are not connected to the grid. This option is very similar to the arrangement that has been implemented in Morocco by the national public utility ONE to expand solar photovoltaic based rural electrification for scattered rural population. However, KPLC doesn’t have the same cumulated experience than the one the ONE of Morocco has acquired on solar since 1985 before deciding on its own to elaborate BOT type tenders. Would KPLC be able or even willing to lead a process of offering BOT contracts with solar on its own license area? So far the Bank would consider this path to uncertain regarding effective implementation and absorption capacity by KPLC and would not advise the GOK to follow it. VI. Conclusions While specific surveys would be necessary to get a more precise and quantified picture of the situation, consultations and data gathered by the study confirm that lack of access to electricity constrains rural development in Kenya. There is evidence of unmet electricity needs in productive activities and social services, namely agriculture, livestock, fishery, tea and coffee cooperatives, telecommunications, water pumping, health and education services. This supports the recognition by the ERSWEC and the Sessional Paper N0.4 of 2004 on Energy that adequate access to electricity is condition to achieve achieving the objectives of the ERSWEC, namely the strengthening of the economic growth – which still depends largely on rural activities – and enhancing equity and reducing poverty. On the basis of existing data issued by the Central Bureau of Statistics (CBS) on demography and households amenities surveys and the most recent KPLC statistics on domestic customers, it has been estimated that the current rate of access to any kind of electricity services would be around 14.5 %, of which about8.6% is provided by the KPLC grid. The gap between these two figures is due to (i) illegal and/or secondary connections, (ii) connection to private mini-grids, (iii) individual gensets and (iv) use of individual solar PV systems. Although these households manage to get some electricity services, this also means that this access is not regulated, that is they do not benefit from any regulatory protection against poor safety standards, bad quality and discontinuity of the supply, and very high price charged to them for these services. However, this also demonstrates the willingness to pay from these populations, which is illustrated by the impressive development of the small and low quality PV solar market. On the basis of population projections for 2010 made by CBS, the number of additional rural customers that should be connected by this date to reach the target of 20 % access rate in rural is around one million. While these figures seem impressive, the characteristics of the Kenya population settlement may help a lot to achieve large volumes of connections. The Kenyan population is concentrated in less than one third of the national territory, as result of very high rural density of population in the western and center parts of the country. High density populated regions are also the ones where the main grid is already in place, which should help to increase the electrification rate 64 64/81 relatively quickly at a relatively low cost, mainly by densifying the existing distribution grid. However, the past performance of the current institutional and financial arrangement for rural electrification over the last three decades indicates that such an ambitious goal would be out of reach without reshaping it deeply: the current pace of connection, either by the RE program or by KPLC on its own, either urban or rural is around 50,000 new customers per year. On average, the rate of rural electrification has increased 0.3% per year during the period 2000-2006. The limitations of current RE institutional model have been pointed out by several studies and include institutional and management constraints, depletion of the Rural Electrification Fund by operating losses worsened by the allocation formula and by inadequate tariffs, high unitary investment costs, and a restrictive connection policy that limits the integration of domestic customers Even assuming significant improvement of KPLC implementation capacity, which may concentrate on its core urban market, the ambitious objective set by the Sessional Paper No4 to support the implementation of the ERSWEC cannot be attained without leveraging additional resources and implementation capacity and without improving the efficiency of the use of the public resources that can be allocated for rural electrification. The GOK has already taken a number of steps to address some of the issues identified. In particular, it has prepared a draft Energy Bill which will establish a Rural Electrification Authority and a Rural Electrification Fund, and which will remove KPLC monopoly in power distribution activities by permitting the award of distribution licenses to third parties. Several studies have also been initiated to facilitate implementation of the decisions already taken in the Sessional Paper on Energy regarding: (i) the institutional set-up for RE, (ii) the need for cost recovery tariffs; and (iii) a new connection policy for connecting customers to facilitate the achievement of access targets. Endorsing the Government’s decision to establish a specific institutional and financial mechanism for rural access expansion, separate from the current set-up for the main system, and on the basis of lessons from international experience in this area, the study is making the following key recommendation to structure this new arrangement:  Set up a strong Institutional Champion committed exclusively to expand access. This would require a detailed description of the roles and responsibilities of the future Rural Electrification Authority, and of the corresponding human and financial means to be committed by the government.  Complete the regulatory framework so it provides for integration of new technical solutions and partners, including the dynamic solar PV industry. This would require additional work on the basis of other international experience but taking into account the specifics of the Kenya legal background;.  Develop Private Public Partnerships to leverage additional resources and implementation capacity. This would require several additional studies, especially market analysis, feasibility studies and promotion of the concept.  Establish a new tariff structure for rural electrification that takes into account operational costs and rural customers’ willingness to pay. A tariff study is currently 65 65/81 being done, which may need to be completed in case the specific case of rural electrification is not covered.  Adopt a new connection policy that maximizes both the number of beneficiaries and revenues. A study is in progress which would provide proposals to solve current issues. Specific provisions may be required in the case of rural areas.  Set up a new Financing Mechanism to provide and optimize first cost public subsidy. According to the option(s) retained, the appropriate mechanism(s) would still need to be designed.  Organize Multi-Sectoral Partnerships for maximizing ratchet effect on rural development. This would require specific studies to identify more precisely, in each areas considered to be electrified, what are the other projects going on in the same area and what additional provisions should be taken, beside the basic electricity infrastructure to work out efficiently the interface with the future electrification program. Participation of private companies in developing access to electricity inside or outside the areas already served by KPLC can be sought under a range of legal arrangements which vary according to the degree of autonomy and protection awarded by the sector authorities to the participating private companies. This ranges from simple works contracts, as done currently under the RE agreement, to fully integrated licenses awarded through competitive bidding, including a number of intermediary options of which advantages and limitations can be examined in the specific geographical and legal context of Kenya. Three potential options for Public Private Partnership to increase access rate outside “KPLC domain� have been determined by the study, which are the following : Option a: Turnkey contracts for clusters of RE schemes: This is the model currently being implemented under the Spanish and French funded programs. Instead of bidding limited number of schemes to each contractor, a large number of schemes (dozens) are bid at the same time. Option b: Distribution licenses for large compact rural areas Distribution licenses for compact rural areas are awarded through a competitive bidding process to the candidate that would commit for higher number of connections against the pre-determined envelope of subsidies. The committed amount of connections is to be (i) built under a limited time and (ii) operated and maintained during the whole license period. Revenues are collected directly by the local rural electrification company that the winner will be requested to create in the license area. The choice of the technology is the responsibility of the licensee, to the extent that each technology used complies with technical specifications or standards to be attached to the tender document. This is the model being implemented in Senegal, and, with some variations, in Argentina and South Africa. Option c: Built Operate Transfer (BOT) contracts for electrifying large compact areas: This option is similar to the previous but differs to the extent that only functions are awarded but not licenses. After a several years of operation during which the winner of 66 66/81 the bid collect the bills and maintain the equipment he has committed to install, the assets and the activity are transferred to KPLC. This is the model adopted in Morocco, and, with some variations, in Bolivia. Option (a) allows neither for the leverage of additional private human and financial resources nor for bringing in innovation related to decentralized solutions. It has also been noted by GoK that option (a) resulted in relatively high cost per customer. Since KPLC doesn’t have the same records as ONE and does not have such an cumulated experience of 20 years on solar PV systems, option (c) does not seem to be appropriate in the context of Kenya to attract the strong private partners required to face the ambitious target. Option (b), e.g. awarding distribution licenses for large rural areas, is recommended as the preferable option to be further explored on the basis of previous international experiences. The size of the area bid and the perspective of a series of tendering, the protection offered by the license, which ensures that revenue from customers can be secured over a period long enough to recover the investment, and the economic viability facilitated by the first cost subsidy can attract large private companies that are reliable over the long term, able to bring in significant amount of financing and able to achieve quickly ambitious targets. The competitive process will ensure that the candidates will optimize the use of the public subsidy. To win, international companies tend naturally to team up with local ones to reduce costs, as observed for instance in Senegal. While this model is supposed to elaborate quite a complex procurement process, the future Rural Electrification Agency can receive technical assistance to prepare the first bidding package and benefit procedures road tested in other countries in the region. However, it should be stressed that significant preparatory work would be required to pave the way for an ambitious implementation of this option, among others detailed domestic and productive demand surveys, indicative feasibility studies and business plans and preparation of legal documents, and the Rural Electrification Authority should be operational. 67 67/81 Bibliography and Data Sources Data and statistics 1. Annual Report KPLC, 2004-2005. 2. Number of Metered Customers RE per sub-region and tariff, March 2006. 3. Population Distribution by Administrative Areas and Urban Centers, 1999 Population and Housing Census, Vol 1, Central Bureau of Statistics. 4. Atlas of Population and Development Indicators, Kenya 1999 Population and Housing Census, Central Bureau of Statistics, October 2003. 5. Analytical Report on Housing Conditions and Households Amenities, Vol X, Kenya 1999 Population and Housing Census, Central Bureau of Statistics. 6. Analytical Report on Migration and Urbanism, Vol VI, Kenya 1999 Population and Housing Census, Central Bureau of Statistics. 7. Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistics. Legal and Policy Documents: 1. The Energy Bill, Bill for Introduction into the National Assembly, July 2004. 2. Sessional Paper N0.4 of 2004 on Energy, Ministry of Energy 3. Task Force 4 Report on Rural Energy, March 2003, Ministry of Energy 4. Roadmap for Reform of the Rural Electrification Program, Ministry of Energy, October 2005. Institutional and Financial Arrangements 1. Report by EDF to Ministry of Kenya on Rural Electrification Institutional Reform, November 2004. 2. Financial Assessment of RE activity within KPLC, report by EDF to Ministry of Energy financed by AFD. 3. Review of the Existing GOK/KPLC Agreement on the Rural Electrification Programme Fund, Discussion Paper by ISG, December 1999. Studies on Rural Electrification in Kenya: Supply options and Demand Assessment 1. Study on Kenya’s energy demand, supply and policy strategy for households, small scale industries and service establishments, August 2001, Kamfor Company. 2. Study of non-electrified Households in electrified Areas, Study by EDF for Ministry of Energy financed by AFD., Feb 2004. 3. Rural Electrification Master Plan, Tractebel and Kaburu Okelo &Partners, 1997 4. Development of MV and LV networks in Rural Zones, report by EDF to Ministry of Energy financed by AFD, November 2004. Documents on PV in Kenya 1. An overview of the Kenyan photovoltaic industry market, Study Report, Electrik Link Ltd, by Kiremu Magambo, April 2004. 2. Renewable Energy Department Activities and set up. Extract from Strategic Plan, Ministry of Energy, 2006. 3. Study for Solar Quality and Specifications and Market Penetration, BCEOM, FONDEM, Energy Alternatives Africa Ltd, August 2001. 68 68/81 4. “Field Performance Evaluation of Amorphous Silicon (a-Si) Photovoltaic Systems in Kenya: Methods and Measurements in Support of a Sustainable Commercial Solar Energy Industry�, Energy and Resources Group (ERG), University of California, Berkeley and Science, Technology, and Environmental Policy (STEP) Program, Princeton University, 2000. 5. “Photovoltaic module quality in the Kenyan solar home systems market�, R. D. Duke, A. Jacobson, D. M. Kammen 6. Energy and Resources Group (ERG), University of California, Berkeley and Science, Technology, and Environmental Policy (STEP) Program, Princeton University, in Energy Policy 30 (2002) 477–499, 2000. 7. “Putting Soar Home System programs into perspective. What lessons are relevant?� F. van der Vleuten, N. Sta. R. van der Plas, final draft for Energy Policy, 2006. 8. “Solar electricity in Africa: a reality�, R. van der Plas, M. Hankins, in Energy Policy 26 (1998) 295–305. 9. PVMTI News, No5, March 2004 10. “The value of vigilance: Evaluating Product quality in the Kenya Solar Photovoltaic Industry�, A. Jacobson, , Umboldt State University & D. Kammen, University of Califormia, Berkeley, June 2005. 11. “An Overview of the Kenyan Photovoltaic Industry Market Study Report�, K. Magambo, Rencon Associates, for Electric Link Ltd, April 2004. Documents related to current or future Donors Programs 1. Support to Rural Electrification Programme in Kenya, Final Intervention Document, Ministry of Energy of Kenya, in collaboration with Ministry for Foreign Affairs of Finland, May 2005. 2. ESMAP – Energy SMEs Program, Aide-mémoire of June 2005 mission. Documents related to Multisectoral Aspects: 1. Draft national Policy for the Sustaimable Development of Arid and Semi Arid lands of Kenya, December 2004. 2. Education Sector Report, February 2006. 3. Health Sector Working Group, February 2006. 4. Physical Infrastructure Sector MTEF Report, February 2006. 5. Agriculture and Rural Development Sector Report, February 2006. 6. Public administration sector Report, February 2006. 7. Kisima (Newsletter published by the Ministry of Water and Irrigaton and the Water and Sanitation Program-Africa), issue 2, July 2005 Maps No.1 Interconnected and Isolated grids No 2 Current Levels of connectivity for households and Market Centers by District No 3 Current Levels of Connectivity for Schools and Health Facilities by District No.4 Distribution of Telecommunication Facilities 69 69/81 VII. Annex 1 - KPLC Statistics - Number of customers (all categories) per region and Sub- region, April 2006 KPLC RES Urban (KPLC+RES) Rural (KPLC+RES) Total (KPLC+RES) % of total % of total Sub- urban rural % of Region Major Towns Urban Rural Urban Rural connected custom Connected custom Connected totalcustom Nairobi South Nairobi City 102,425 - - - 102,425 15.9% - 0.0% 102,425 12.9% Machakos 2,627 1,202 1,024 2,099 3,651 0.6% 3,301 2.2% 6,952 0.9% Kajaido 5,963 702 608 524 6,571 1.0% 1,226 0.8% 7,797 1.0% Makueni - - 986 4,260 986 0.2% 4,260 2.8% 5,246 0.7% Kibwezi 596 493 346 896 942 0.1% 1,389 0.9% 2,331 0.3% Garissa - 3,661 - - - 0.0% 3,661 2.4% 3,661 0.5% Wajir - - - 1,311 - 0.0% 1,311 0.9% 1,311 0.2% Moyale - - - 1,173 - 0.0% 1,173 0.8% 1,173 0.1% Mandera - - - 1,194 - 0.0% 1,194 0.8% 1,194 0.2% Marsabit - - - 926 - 0.0% 926 0.6% 926 0.1% Total Nairobi South 111,611 6,058 2,964 12,383 114,575 17.8% 18,441 12.3% 133,016 16.8% Nairobi West Nairobi City 98,615 - - - 98,615 15.3% - 0.0% 98,615 12.4% Kikuyu 4,407 1,800 1,204 1,696 5,611 0.9% 3,496 2.3% 9,107 1.1% Limuru 4,266 5,167 1,140 1,362 5,406 0.8% 6,529 4.4% 11,935 1.5% Ngong 6,700 4,018 986 1,953 7,686 1.2% 5,971 4.0% 13,657 1.7% Total Nairobi West 113,988 10,984 3,330 5,011 117,318 18.2% 15,995 10.7% 133,313 16.8% Nairobi North Nairobi City 122,464 - - - 122,464 19.0% - 0.0% 122,464 15.4% Kiambu 3,665 2,799 1,452 1,836 5,117 0.8% 4,635 3.1% 9,752 1.2% Ruiru 2,794 1,596 942 3,142 3,736 0.6% 4,738 3.2% 8,474 1.1% Total Nairobi North 128,923 4,395 2,394 4,978 131,317 20.4% 9,373 6.3% 140,690 17.7% 70/81 70 Total Nairobi Region 354,522 21,436 8,688 22,372 363,210 56.5% 43,808 29.3% 407,018 51.3% South Mombasa Coast City 33,558 2,130 - - 33,558 5.2% 2,130 1.4% 35,688 4.5% Kwale 1,415 1,176 254 996 1,669 0.3% 2,172 1.5% 3,841 0.5% Ukunda 2,991 513 136 924 3,127 0.5% 1,437 1.0% 4,564 0.6% Voi 3,260 533 123 1,026 3,383 0.5% 1,559 1.0% 4,942 0.6% Loitoktok - 620 - 462 - 0.0% 1,082 0.7% 1,082 0.1% Taveta - 1,215 117 125 117 0.0% 1,340 0.9% 1,457 0.2% Total South Coast 41,224 6,187 630 3,533 41,854 6.5% 9,720 6.5% 51,574 6.5% North Coast Malindi 13,948 1,231 720 424 14,668 2.3% 1,655 1.1% 16,323 2.1% Lamu 2,556 782 652 296 3,208 0.5% 1,078 0.7% 4,286 0.5% Nyali 24,645 2,526 524 845 25,169 3.9% 3,371 2.3% 28,540 3.6% Kilifi 6,521 587 496 464 7,017 1.1% 1,051 0.7% 8,068 1.0% Total North Coast 47,670 5,125 2,392 2,029 50,062 7.8% 7,154 4.8% 57,216 7.2% Total Coast 88,894 11,312 3,022 5,562 91,916 14.3% 16,874 11.3% 108,790 13.7% Central Rift Nakuru 33,222 5,116 867 2,678 34,089 5.3% 7,794 5.2% 41,883 5.3% Naivasha 5,275 1,325 523 1,532 5,798 0.9% 2,857 1.9% 8,655 1.1% Narok 1,032 483 867 1,303 1,899 0.3% 1,786 1.2% 3,685 0.5% Nyahururu 4,473 632 939 1,542 5,412 0.8% 2,174 1.5% 7,586 1.0% Molo 3,064 615 539 1,685 3,603 0.6% 2,300 1.5% 5,903 0.7% Salgaa - 507 248 352 248 0.0% 859 0.6% 1,107 0.1% Eldama Ravine - 265 42 126 42 0.0% 391 0.3% 433 0.1% Maralal - - - 739 - 0.0% 739 0.5% 739 0.1% Total Central Rift 47,066 8,942 4,025 9,957 51,091 7.9% 18,899 12.6% 69,990 8.8% West Kenya Kisumu 19,861 3,353 1,230 4,864 21,091 3.3% 8,217 5.5% 29,308 3.7% Kakamega 5,752 1,226 852 2,652 6,604 1.0% 3,878 2.6% 10,482 1.3% 71/81 71 Bungoma 4,518 512 557 1,020 5,075 0.8% 1,532 1.0% 6,607 0.8% Kisii 7,454 517 632 2,225 8,086 1.3% 2,742 1.8% 10,828 1.4% Migori 492 431 746 1,147 1,238 0.2% 1,578 1.1% 2,816 0.4% Busia 2,366 825 761 1,564 3,127 0.5% 2,389 1.6% 5,516 0.7% Bondo - 491 320 425 320 0.0% 916 0.6% 1,236 0.2% Kericho 2,034 262 243 1,453 2,277 0.4% 1,715 1.1% 3,992 0.5% Sotik - 253 760 820 760 0.1% 1,073 0.7% 1,833 0.2% Total West Kenya 42,477 7,870 6,101 16,170 48,578 7.6% 24,040 16.1% 72,618 9.2% North Rift Eldoret 9,361 2,150 1,563 5,863 10,924 1.7% 8,013 5.4% 18,937 2.4% Kitale 3,409 810 563 1,803 3,972 0.6% 2,613 1.7% 6,585 0.8% Kapsabet 1,923 516 900 1,252 2,823 0.4% 1,768 1.2% 4,591 0.6% Kabarnet - 548 235 2,744 235 0.0% 3,292 2.2% 3,527 0.4% Iten - 52 151 254 151 0.0% 306 0.2% 457 0.1% Total North Rift 14,693 4,075 3,412 11,916 18,105 2.8% 15,991 10.7% 34,096 4.3% Total West Region 104,236 20,887 13,538 38,043 117,774 18.3% 58,930 39.4% 176,704 22.3% Mount Kenya North Nyeri 19,764 1,788 708 3,252 20,472 3.2% 5,040 3.4% 25,512 3.2% Meru 4,251 2,516 966 2,362 5,217 0.8% 4,878 3.3% 10,095 1.3% Nanyuki 2,230 488 1,241 1,562 3,471 0.5% 2,050 1.4% 5,521 0.7% Isiolo 809 243 113 215 922 0.1% 458 0.3% 1,380 0.2% Embu 9,110 1,032 644 2,496 9,754 1.5% 3,528 2.4% 13,282 1.7% Kerugoya 5,698 700 623 2,275 6,321 1.0% 2,975 2.0% 9,296 1.2% Total Mount Kenya North 41,862 6,766 4,295 12,162 46,157 7.2% 18,928 12.6% 65,085 8.2% Mount Kenya South Thika 14,230 957 2,528 3,369 16,758 2.6% 4,326 2.9% 21,084 2.7% Kitui 2,755 531 569 1,254 3,324 0.5% 1,785 1.2% 5,109 0.6% Muranga 3,204 2,081 896 2,952 4,100 0.6% 5,033 3.4% 9,133 1.2% Total Mount Kenya South 20,189 3,569 3,993 7,575 24,182 3.8% 11,144 7.4% 35,326 4.5% 72/81 72 Total Mount Kenya Region 62,051 10,335 8,288 19,737 70,339 10.9% 30,072 20.1% 100,411 12.7% Totals Country Wide 643,239 100% 149,683 100.0% 792,922 100% KPLC RES Total Total Urban Rural Urban Rural Urban Rural Total Gen Totals Country Wide 609,703 63,969 33,536 85,714 643,239 149,683 792,922 Grand Total All Customers 673,672 119,250 73/81 73 Annex 2 – Projection of Population per District 2000-2006 Source: Analytical Report on Population Projections, Vol VIII Kenya 1999 Population and Housing Census, Central Bureau of Statistics. Evolution District 2000 2001 2002 2003 2004 2005 2006 2000- 2006 Nairobi 2,290,049 2,379,741 2,470,850 2,563,297 2,656,997 2,751,860 2,845,353 24.25% Kiambu 775,548 782,823 789,602 795,866 801,601 806,790 810,724 4.54% Kirinyaga 475,535 478,868 481,877 484,554 486,891 488,880 490,096 3.06% Murang'a 356,939 352,981 348,667 344,004 339,001 333,664 327,724 -8.18% Nyandarua 506,171 517,986 529,598 540,984 552,118 562,975 573,038 13.21% Nyeri 683,490 683,115 682,185 680,698 678,654 676,053 672,321 -1.63% Thika 667,525 667,159 666,251 664,799 662,802 660,262 656,617 -1.63% Maragua 400,945 400,570 399,866 398,833 397,471 395,781 393,426 -1.88% Centrai 3,882,021 3,918,438 3,952,369 3,983,728 4,012,433 4,038,407 4,058,098 4.54% Kilifi 573,962 587,377 600,763 614,106 627,388 640,593 653,144 13.80% Kwale 520,960 530,467 539,827 549,027 558,051 566,886 575,026 10.38% Lamu 76,243 77,536 78,804 80,043 81,253 82,431 83,503 9.52% Mombasa 704,571 725,046 745,670 766,422 787,280 808,221 828,513 17.59% Taita Taveta 256,972 259,183 261,208 263,041 264,676 266,108 267,101 3.94% Tana River 191,318 196,467 201,636 206,822 212,018 217,219 222,228 16.16% Malindi 297,765 305,778 313,824 321,895 329,983 338,077 345,872 16.16% Coast 2,622,794 2,684,095 2,745,267 2,806,237 2,866,931 2,927,273 2,984,625 13.80% Embu 289,965 292,672 295,210 297,574 299,759 301,761 303,313 4.60% Isiolo 106,719 109,625 112,514 115,381 118,222 121,032 123,699 15.91% Kitui 539,443 547,138 554,582 561,762 568,663 575,273 581,080 7.72% Makueni 811,035 826,824 842,338 857,548 872,427 886,947 900,310 11.01% Machakos 953,049 971,603 989,832 1,007,705 1,025,190 1,042,252 1,057,955 11.01% Marsabit 127,696 130,182 132,624 135,019 137,362 139,648 141,752 11.01% Mbeere 179,075 181,814 184,474 187,051 189,539 191,934 194,065 8.37% Meru Central 522,581 530,575 538,339 545,858 553,119 560,108 566,325 8.37% Mayale 56,020 56,877 57,709 58,515 59,293 60,042 60,709 8.37% Mwingi 318,325 323,268 328,073 332,731 337,233 341,571 345,440 8.52% Meru North 636,040 649,698 663,170 676,433 689,463 702,236 714,115 12.28% Tharaka 106,340 108,624 110,876 113,094 115,272 117,408 119,394 12.28% Nithi (Meru S.) 213,012 213,644 214,121 214,439 214,597 214,595 214,247 0.58% Eastern 4,840,947 4,902,006 4,960,626 5,016,693 5,070,098 5,120,733 5,164,069 6.67% Garissa 420,025 436,167 451,384 465,590 478,699 490,624 500,850 19.24% Mandera 267,923 278,220 287,926 296,988 305,350 312,956 319,480 19.24% Wajir 350,074 374,438 399,481 425,189 451,543 478,523 505,674 44.45% North Eastern 1,054,665 1,127,638 1,202,603 1,279,509 1,358,301 1,438,916 1,519,985 44.12% Gucha (Kisii S.) 484,356 493,601 502,704 511,652 520,429 529,022 536,955 10.86% Noma Bay 303,323 309,261 315,116 320,876 326,534 332,079 337,214 11.17% Kisii Centrai 513,812 520,087 526,112 531,877 537,372 542,585 547,039 6.47% Kisumu 526,948 533,383 539,563 545,476 551,110 556,457 561,025 6.47% Kuria 161,018 165,782 170,546 175,303 180,047 184,771 189,303 17.57% Migori 542,708 555,042 567,272 579,381 591,350 603,159 614,262 13.18% 74 74/81 Nyamira (Kisii 520,705 527,417 533,887 540,102 546,051 551,725 556,635 6.90% N.) Rachuonyo 321,063 325,202 329,191 333,023 336,691 340,190 343,217 6.90% Siaya 496,498 496,326 495,736 494,728 493,298 491,448 488,758 -1.56% Suba 163,910 167,439 170,933 174,384 177,787 181,138 184,272 12.42% Bondo 248,980 251,426 253,735 255,901 257,922 259,792 261,283 4.94% Nyando 316,950 325,128 333,274 341,375 349,419 357,393 364,972 15.15% Nyanza 4,598,485 4,666,140 4,731,887 4,795,614 4,857,210 4,916,569 4,969,323 8.06% Baringo 278,517 283,926 289,257 294,501 299,649 304,694 309,361 11.07% Barnet 402,353 410,167 417,868 425,443 432,880 440,168 446,911 11.07% Keiyo 151,378 154,516 157,621 160,690 163,717 166,697 169,482 11.96% Kajiado 433,139 449,279 465,697 482,379 499,314 516,487 533,427 23.15% Kericho 491,147 499,119 506,873 514,392 521,663 528,669 534,938 8.92% Koibatek 146,640 151,217 155,842 160,510 165,218 169,961 174,584 19.06% Laikipia 341,955 352,628 363,413 374,299 385,278 396,338 407,119 19.06% Marakwet 147,666 150,353 152,989 155,569 158,090 160,546 162,794 10.24% Nakuru 1,254,551 1,287,281 1,320,089 1,352,940 1,385,794 1,418,614 1,450,116 15.59% Nandi 609,203 622,108 634,897 647,552 660,054 672,384 683,938 12.27% Narok 386,496 396,512 406,549 416,595 426,638 436,668 446,289 15.47% Samburu 150,932 153,926 156,879 159,789 162,651 165,460 168,066 11.35% Trans-Mara 178,513 181,011 183,409 185,702 187,882 189,947 191,725 7.40% Trans-Nzoia 610,674 629,359 648,226 667,256 686,430 705,730 724,515 18.64% Turkana 472,727 480,481 488,029 495,355 502,445 509,286 515,420 9.03% Uasin Gishu 658,108 675,261 692,454 709,669 726,885 744,083 760,588 15.57% West Pokot 324,971 332,677 340,363 348,021 355,639 363,208 370,401 13.98% Buret 332,907 339,169 345,327 351,371 357,290 363,074 368,399 10.66% Rift Valley 7,386,459 7,581,637 7,777,407 7,973,552 8,169,849 8,366,071 8,554,647 15.82% Bu ngo ma 932,817 964,910 997,357 1,030,126 1,063,180 1,096,486 1,129,037 21.04% Busia 391,668 401,859 412,069 422,286 432,498 442,692 452,468 15.52% Mt.Elgon 141,399 143,509 145,564 147,561 149,496 151,366 153,036 8.23% Kakamega 641,539 662,824 684,322 706,009 727,863 749,861 771,315 20.23% Lugari 229,559 237,176 244,868 252,628 260,448 268,320 275,996 20.23% Teso 191,008 195,023 199,009 202,961 206,874 210,743 214,379 12.24% Vihiga 527,135 540,735 554,355 567,980 581,594 595,180 608,199 15.38% Butere/Murrias 500,076 508,347 516,454 524,385 532,128 539,671 546,534 9.29% Western 3,532,944 3,604,850 3,676,133 3,746,697 3,816,448 3,885,290 3,949,742 11.80% KENYA 30,208,365 30,864,544 31,517,142 32,165,328 32,808,269 33,445,119 34,045,843 12.70% 75 75/81 Annex 3 - Distribution of households by main type of lighting Source: Analytical Report on Housing Conditions and Households Amenities, Vol X, Kenya 1999 Population and Housing Census, Central Bureau of Statistics. Table A8 Electricity Pressure Lantern Tin Lamp Fuel Wood Solar Other Total Lamp Number % Number % Number % Number % Number % Number % Number % Nairobi 336,299 52.3 12,476 1.9 136,469 21.2 140,334 22 698 ` 0.1 341 0.1 16,289 2.5 642,906 Central 143,619 15.6 14,178 1.5 408,077 44.4 324,020 35.3 4,682 0.5 6,072 0.7 17,827 1.9 918,475 Coast 101,713 19.4 6,229 1.2 120,903 23.1 275,260 52.5 6,638 1.3 1,045 0.2 12,092 2.3 523,880 Eastern 55,779 5.8 9,334 1.0 424,473 44.5 388,746 40.7 48,800 5.1 8,733 0.9 18,336 1.9 954,201 North eastern 7,825 5 354 0 42,600 29 23,506 16 58,744 40 247 0 13,806 9 147,082 Nyanza 45,883 5 9,970 1 240,971 25 639,398 66 4,275 0 3,206 0 19,166 2 962,869 Rift Valley 139,381 9 14,614 1 578,618 39 521,094 35 186,393 13 7,520 1 39,439 3 1,487,059 Western 22,492 3.2 5,481 0.8 184,775 26.4 474,173 67.9 1,956 0.3 1,367 0.2 8,579 1.2 698,823 Kenya 852,991 13.5 72,636 1.1 2,136,886 33.7 2,786,531 44.0 312,186 4.9 28,531 0.5 145,534 2.3 6,335,295 share of each technology 13% 1% 34% 44% 5% 0% 2% 100% 76/81 76 Annex 4 – Field Visits : Main Findings Visits were made to selected trading/market centres - Namanga, Bissil, Fly-over and Njambini - some of which had been connected to the grid and others which had not been connected. During the visits families in the vicinity of and far from the market/ trading centres were met. The visits provided ample evidence of high levels of unmet, and based on expenditures on other energy forms, a clear indication of the willingness and ability of those interviewed to pay for electricity at current prices; the indications are exemplified by the examples given below. Businesses have installed own generators to supply individual homes or shops with electricity and charge Kshs500 per bulb; and there are many takers. The demand for such services is growing; so is the demand for electricity for entertainment and for productive purposes such as carpentry, welding and machining. Those with stand alone generators showed keen interest in getting connected to the grid in order to benefit from lower cost and more reliable supply. Meanwhile artisans – welders, carpenters, etc – expressed interest in being supplied from the grid as this, they affirmed, would significantly increase their productivity; with the growing demand for the better quality services and products, their businesses and incomes would increase accordingly. Fig. 1 A carpenter at Namanga; expressed interest in being connected as this would greatly increase the productivity of his business At the centres that had been connected to the grid (Bissil and Fly-over) there was evidence of business expansion and increased activities e.g. maize milling, welding, general machining, bars and restaurants. Business opportunities and the quality of facilities were greatly enhanced. More artisans were moving in; the artisans interviewed welcomed competition while the 77/81 customers relished the improved quality of services and products at competitive prices. None of the facilities which had been connected has been disconnected – a clear indication of willingness and ability to pay for the services rendered. Fig.2: Artisans at work at Bissil – machining motor parts; quality of work improved as a result of electricity connection and customers satisfied Visits to unconnected rural homes also revealed interest in electricity including those supplied by decentralised systems. A household was, for instance, using PV panels for lighting and entertainment. Another one went some length to install a standby generator for lighting, colour TV and computer at very high cost. Both expressed an interest in being connected to the grid and willingness to pay for the services. Other observations At Namanga, plans and arrangements are advanced for connection to the grid; a total of 238 customers are to be connected initially and it is envisaged that about 1,000 will be connected by 2009. The main customers will be hotels, shops, bars, petrol stations, government facilities such as police stations; and social facilities e.g. schools, dispensaries and health centres. The main stumbling block is the high cost of connection: currently going at Kshs 46,000 for a 3-phase system and Kshs 15,000 for a single phase one. More people expressed interest in participating in the scheme if only the connection fees could be spread out, say for at least three instalments. Another factor limiting the increase in demand is that households are not included in the current programme; yet there are many that would benefit; indeed some have expressed willingness to pay the connection fees. Their participation would increase the demand considerably. A sure benefit would be the extension of electricity across the boarder to Tanzania to meet the demand; this would be a good example of co-operation in the region At Bissil the bars have increased in number and quality of facilities as well as customer services enhanced; where before there used to be only two small refrigerators, now there are at least 20 large ones. More customers, households as well as businesses, want to be connected; 78/81 the demand for electricity is, therefore, likely to rise substantially to a much higher level than is currently the case. At Flyover still more customers want to be connected; the level of connection could be increased to optimise the use of electricity for expanded business activities; new business are coming in as are more tenants. There is also scope for greater use of solar panels, but given the climatic conditions at the area, there is inadequate sunshine; the solar panels, therefore, need to be augmented by electricity or better still be used as back up to the grid connected supply. At Njambini a new approach is being implemented:Rather than concentrate on one centre, this time emphasis was on meeting the requirements of several customers simultaneously: Njamini township, four secondary schools, Kanyenyaine trading centre and Munyaka. The new approach will go a long way towards increasing connections. A total of 7 transformers will be installed – two 200kVA, two 100 kVA, two 50kVA and one 25kVA. There will be 18.1 km of 11kV lines and11km of 220V lines. Our attention was in particular caught by the experience of a rural homestead in Kinangop. The owners of the homestead were willing to overlook the high cost (Kshs300 a day in fuel) of running a small 1,1kVA diesel driven generator for no more than 2 hours a day, for lighting, colour TV and computer; this is over and above that, also uses 5 litres of kerosene a month for lighting; uses gas for cooking. The experience of the household is representative of many in the rural areas in Kenya which point to the huge unmet demand in those areas. There is also room for more widespread use of decentralised systems to supply electricity to meet household demand for electricity in the rural areas, especially those far removed from the grid. Many households are willing to pay the high cost to be supplied with electricity for lighting and entertainment; for even the alternative source – kerosene - is becoming increasingly expensive in spite of being more inconvenient to use. With better information and guidance even more would opt for the use of decentralised systems to meet their electricity needs, thus increasing the demand substantially. Within the areas in the immediate vicinity of the grid where many are already enjoying the benefits of being connected, a large number of people patiently are still waiting to be connected to meet household electricity needs or to be upgraded to cater for productive activities such as bakery or irrigation for agriculture; but they have been informed that their needs cannot be met for the time being because of unavailability of suitable facilities and equipment. This, I am sure is, the case for many parts of rural (and urban) Kenya pointing the high levels of unmet or underserved demand for throughout the country. 79/81 Fig 3: Area already connected (at Lari); productive activities: compound houses a bakery and dairy cows; customer would like to apply for an up-grade from single to three-phase connection to further expand his business and for irrigation to boost agricultural productivity What is lacking and must be addressed as a matter of urgency is awareness on the part of the public at large about the programme and the modalities of as well as the benefits of being connected. Adequate information should be provided on (i) connection policy, (ii) connection charges, (iii) whether the connection charges can be paid over a period of time say in three instalments; and (iv) what step to take to apply for an up-grade, from single to three phase connection. 80/81 Annex 4 – Map of the Kenya Electricity Grid Source: Annual Report KPLC, 2004-2005 81/81