Joint UNDP/World Bank Energy Sector Management Assistance Program Activity Completion Report No. 032/85 Country: THE GAmBIA Activity: PRE-INVEST.IENT REPORT ON SOI. PHOTOVOLTAIC APPLICATIONS IN THE HEALTH AND TELECOMMUNICATIONS SECTORS MARCH 1985 Report of the joint UNDPAVortd Bank Energy Sector Management Assistance Program This document has a restricted distribution. Its contents may not be disclosed without authorization from the Government, the UNDP or the World Bank. * *t~~~~~~~~' ** -. * - - -- - -- -- - --- - - - - . ENRY SECOR M UNAGEMENT ASSISTANCE PROGRAM The Joint UNDP/World Bank Energ- Sector Management Assistance ?rogram (ESMAP), started in April 1983, assists countri:es in implementing the main invest-ent and policy recommendations of the Energy Sector Assessment Reports produced under another Joint UNDP/World Bank Program. ESMAP provides staff and consultant assistance in formulating and justifying priority pre-investment and investment projects and i A providing management, institutional and policy support, The reports produced under this Program provide governments, donors and potenti.1 investors with the information needed to speed up project preparation and implementation. ESMAP activities can be classifted broadly into three groups: - Energy Assessment Status Reports: these evaluate achieve- ments- in the year following issuance of the original assessment report and point out where urgent action Is still needed; - Proje't Vormulation and Justification: work designed to accelerate the preparation and Implementatitu of investment projects; and - Institutional and Policy Support: this work also frequently leads to the identification of technical assistance packages. The Program aims to supplement, advance and strengthen the impact of bilateral and multilateral resources already available for technical assistance in the energy sector. Funding of the Program The Program is a major international effort and, while the core finance has been provided by the UNDP and the World Bank, important financial contributions to the Program have also been made by a number of bilateral agencies. Countries which have now made or pledged initial contributions to the programs through the UNDP Energy Account, or through other cost-sharing arrangements with UNDP, are the Netherlands, Sweden, Australia, Switzerland, Fialand, United Kingdom, Denmark, Norway, and New Zealand. Further Inforuation For further information on the Program or to obtain copies of completed ESMAP reports, which are listed at the end of this document, please contact: Division for Global and OR Energy Assessments Division Interregional Projects Energy Department United Nations Development World Bank Program 1818 H Street, N.W. One United Nations Plaza Washington, D.C. 20433 Neu York, N.Y. 10017 THE GAMBIA PRE-INVESTMENT REPORT ON SOLAR PHOTOVOLTAIC APPLICATIONS IN THE HEALTH AND TELECOMMUNICATIONS SECTORS MARCH 1985 i ABBREVIATIONS AND AC8ONYMS Ah Ampere-hours DEL Direct Exchange Line DOH Department of Health EPI Expanded Program for Immunization GPMB Gambia Produce Marketing Board CAMTF,. Gambia Telecommunications GOTS Government of The Gambia GTZ Cerman Technical Assistance Agency GUC Gambia Utilities Corporation ITU International Telecommunications Union IBRD World Bank km kilometers kVA kilovolt-amperes kWh Kilowatt hours m meters MCH Maternal and Child Health Care Program MEPID Ministry of Economic Planning and Industrial Oevelopment MHz Megahertz mm Millimeters PANAFTEL Pan-African Telecommunications PV Photovoltaic UHF Ultra High Frequency UN United Nations UK-ODA United Kingdom Overseas Development Administration VHF Very High Frequency CURREC 3Q0ALIS Dalasis 4.20 = US $1.00 (Decembert 1984) D 3.75 = US $1.00 (April, 1984) 0 1.68 = US $1.00 (1981) CCOVU3SION FACTORS British thermal unit (Btu) - 1.054 kJ 1.054 x 101 erg = 1054 J 778 ft-lb - 252 cal 0 0.293 Wh Btu/ft2 - 11.35 kJ/m2 0.271 cal/m2 - 1.135 J/cm - 3.15 X 10-3 kWh/m2 cal/cm2 h = 11.63 W/m2 (langley/h) - 3.687 Btu/hft2 0 0.1667 cal/cm2/min - 41.8 kJ/4 m2 kw/m2 = 1000 W/m - 3600 kJ/h m2 - 317.2 Btu/hf 2 - 1.432 cal/cm/min kWh/m2 - 3600 kJ/m2 - 317 Btu/f 2 - 86 cal/cm' TABLE OF CONTENTS Page SUMMARY 4 ** ~~~~~~~~~~i-xii Background to the Rep o r t 1 Objectives of the Report ................ . 1 Viable Solar Photovoltaic AppLio dtio... 2 The Project.3^..................... , 3 rI. HEALTH SECTOR/A.... 0 0 Rural Health Sector anrd the Immunization Programa...... 4 Structure of the Service and its Organization 4 The Expanded Program for Immunization (EPI) 6 Energy-Related Prbb l e m s 8 GUC-Powered Health Centers and Dispensaries 8 Other Health Centers and Dispensaries ******. O10 Solar Photovoltaic Vaccine Refrigerators*.......... ;0 12 Well-defined Needs: Refrigeration, Lighting and Water Heating ................. .0 *o......*....12 Less-clear Needs - Telecommunitations and Water Supplies ... ................ *.* }16 Recommendations for the Rural Health Sector *o..* r,,**** 17 Solar Photovoltaic Vaccine Refrigerator and Lighting ................... ?7 Solar Water Heating ..........0000..... , 18 Solar Telecommunication Systems and Water Suppilies ............... l;. III. TELECOMMUNICATIONS SECTOR ..... ............... O..... 20 The Rural Telecommunications Network................... 20 The Multi-channel Trunk System*..,...... ... ..... 22 Provincial Telephone Exchanges on Trunk Network ....... 23 Single-Channel UHF Links to the Villages.............. 24 Structure of Telecommunications Operation. ..............* 28 Gamtel Finances ...... ..... ,o..0 . 28 Maintenance Problems .......................... 30 Power Supply Problems in the Telecommunications UHF Repeater Stations ..*.............. .. . *..... 31 Provincial Terminal Exchanges......................... 32 Single-Channel Rural VHF Links .................... 0*0* 32 Recommendations for the Telecommunications Sector...... 33 UHF Repeater Stations *. 33 Provinciat Exh sa n g e s 35 Single-Channel Rural VHF Links..i... 37 IY. SOLAR WATER PUMPING...... *................................ 42 Water Puaping Ueeds in the Health Sector ............... , 42 Relationship with the Activities of the Department of Wt.ter Resources ....................... 43 Irrigation Pumping and the Use of Solar Pumps in McCarthy Island Division......................... 43 Experience with Solar Pumps in The Gambia............. 44 Comparison of Solar Pumps with Wind and Hand Pumps.... 45 Recouuendations ............................................ 46 Budgetary Requirements ....... 44.4.... ..e*e**.*o.o**** 47 V* THE PROJEC. .............................................. 49 Project Description ........ ..... .... ....... ......... .... 49 Health Sector ........; 49 Telecommunications Sector ............................ 50 Implementation s.......o........ 4 50 Budget .. 4.44......... 444444444 444 4444444444444444444444 54 ANNEXES 1. Comparative Analysis of Kerosene and Solar Vaccine Refrigerators 2. Technical Description of Solar Vaccine Refrigerators and Lighting Systems 3. Photovoltaic Power System Sizing and Cost Analysis 4. Analysis of Repeater Station Power Systems 5. Replacement Battery Requirements for Provincial Exchanges 6. TechnicaL and Cost Analysis of VHF Links 7. Solar Photovoltaic Water Pumps TECHNICAL SUPPLEMENT A. Preliminary Draft Tender Document B. Climatic Data for The Gambia MAPS Rural Health Facilities IBRD 18624 Gamtel Long Distance Network IBRD 15991x1 SUMUY Background and Objectives The Country Background 1. The Republic of The Gambia, located on the west coast of Africa at about latitude 13' north, measures about 3iJO km from east to west and is narrow, averaging about 20-30 km in width from north to south. Banjul, the capital, lies on the western-most end of the country and faces the Atlantic Ocean. The population of about 600,000 is over three quarters rural, with the urban population concentrated in and around Banjul. 2. The Gambia is one of the poorer countries of Africa, with an estimated GNP per capita of US$300. Its economy is based on non-energy- intensive agriculture, the main cash-crop and export commodity being groundnuts. The country has suffered from serious economic problems in recent years, resulting from a combination of increased import prices, particularly oil prices, and declining export earnings. This decline was targely t,ue to reduced groundnut production brought on by a decade of lower than average rainfall. Oil imports consumed only 9% of commodity export eavnings in 1974-75, but in 1980-81, which was a parcicularly bad year for agricultural production and when rainfall was especially poor, they absorbed 70% of commodity export earnings. The government is cur- rently proceeding with the Second Development Plan, which is premised on an average of 5% annual growth in GDP between 1980 and 1986. This rate of growth is based on the assumption that groundnut production will greatly increase. 3. Living conditions for the majority of the popuLation are poor and malnutrition is combined with a high incidence of disease. Much of the disease is due to contaminated water supplies or epidemics that could be controlled through mass immunization. Infant mortality is particu- larly high in the rural areas where, in some districts, more than one third of the children die before the age of two. It is a government priority to combat these problems. Major programs have been initiated, with international support, to introduce a nation-wide immunization campaign as a key element in this effort. The improvement of village water wells is also a high priority item in the government's program. Another sector of the rural economy which has hitherto been neglected, and is somewhat in a state of disarray, is telecommunications. The rural telecommunications network has been affected by the fuel shortages, unreliable power supplies, and inadequate maintenance. Since 80% of the population is dependent on the rural network, the Government views its rehabilitation and improvement as important components in the national development program. - ii - Background to the Report 1/ 4. tn 1982, a mission under the auspices of the Joint UNDPI World Bank Energy Sector Assessment Program reviewed all aspects of the energy sector in The Gambia and prepared a report which inter alia evaluated some possibilities for substituting imported oil with indigenous energy resources, including solar energy. 2/ Subsequently, in April 1984, at the request of the Government, a Technical Assistance Identification mission 3/ under the Energy Sector Management Assistance Program ,ESMAP) identified the fuel substitution potential ol solar photovoltaic systems in two vital rural systems -- rural health care and telecommunications systems - and prepared detailed terms of reference for a pre-investment study. A follow-up ESMAP mission visited The Gambia in July, 1984, to prepare this report. Objectives of the Report S. The primary objective of the July 1984 mission was to produce a pre-investment study which would assist the GOTG in obtaining external *funding to replicate the use of solar photovoltaic power units in the rural health care and telecommunications systems. The specific objec- tives, as set out in the mission's terms of reference, were: (a) to establish the costs and benefits from the introduction of solar photovoltaic power units to supply some or all of the critical electricity needs of the existing network of rural health centers and dispensaries in The Gambia (for refrig- eration of vaccines, for lighting in-patient rboms, for hot water sterilization and, possibly, for pumping water from wells); (b) to establish basic specifications of solar photovoltaic power units to supply a combination of the above needs for which potential savings can be made, and to establish the level of investment required to supply these systems to various health centers and dispensaries; * 1/ This report is based on the findings of a mission comprising Messrs. Anwer Malik and Peter Fraenkel (Consultant) that visited the Gambia in July 1984. The report was written by Messrs. Fraenkel, Malik and Akanda. 2/ Joint UNDP/World Bank Energy Sector Assessment Program. The Gambia: Issues and Options in the Energy Sector, Report No. 4743-GM, November 1983. 3/ This Technical Assistance mission comprised Mr. Amarquaye Armar.- - iii - (c) to review the existing recurrent costs and the status of the use of solar photovoltaic power units at some of the rural single channel VHF stations in The Gambia's rural telecommuni- cations system, as a basis for establishing the level of in- vestment required to retrofit remaining rural single-channel VHF stations with solar photovoltaic power kits, the assessment of single channel VHF stations to be done with special refer- once to the needs of GOTG's Primary Health Care ProSram; (4) to assess the feasiLility and level of investment required to retrofit existing diesel power units at UHF repeater stations and provincial multi-channel VHF stations with solar photo- voltaic systems; (e) to prepare, on the basis of (a), (b), (c), and (d) above, a report specifying investment costs, benefits and an implemen- tation plan: (i) of an optimal program to equip health centers and dispensaries with the appropriately sized solar photo- voltaic power system for their essential electricity needs (inclusive or exclusive of water pumping); and (ii) to equip rural VHF stations with appropriately sized solar phocovoLtaic power systems; and (f) to briefly look into the feasibility of utilizing solar photo- voltaic powered pumps for small-holder irrigation in the McCarthy Island Division. The Mission and its Approach 6. After initial meetings with officials from the relevant govern- ment agencies, the mission visited a cross-section of health and telecom- munications facilities to identify potential solar energy applications in. these two sectors. A total of 9 out of 29 govRrnment health centers and disp4. saries, all 3 UHF repeater stations, 3 out of 7 rural telephone exchanges, and 7 out of the 13 remaining single channel VHF public tele- phone terminals were visited. The mission was accompanied in the field by personnel from the Department of Health (DOH) and from Gambia Telecom- munications (Gamtel) who concurred that the sample of installations visited was truly representative and gave a realistic impression of the existing situation. Findings and Recommendations of the Mission Rural Health Sector 7. The primary health care program in The Cambia includes the Expanded Program for Immunization (SPI). The main targets for the EPI are the below-23 months age group and pregnant women. The program is - iv - implemented through a network of 18 rural health centers, 13 dispen- saries, and two private missions which cooperate with the government program. The health centers range in size from having five or six staff and a few beds to being, in effect, small hospitals, with a dozen staff and up to 20 beds. Dispensaries usually have four para-medical or nursing staff working from a single building with minimal equipment. There is also a network of rural sub-dispensaries that do not have per- manent staffing but which are serviced from the dispensaries and health centers which offer clinics once or twice per week. 8. Four of the health centers have no power supply, and the rest receive it only intermittently. The dispensaries have no power at all. At present, all health centers and dispensaries are equipped with refrig- erators for storing vaccines. Due to the general unreliability or non- availability of mains electrical power supplies in The Gambia, cnly two health centers have mains electric refrigerators, two have experimental solar photovoltaic refrigerators, and the rest rely on kerosene-fueled absorption refrigerators. 9. The EPI attempts to provide a regular supply of kerosene to all health centers and dispensaries officially dependent on kerosene re- frigerators, and additional amounts of fuel are provided to alLow for lighting requirements, spillage, or other losses. The supply of fuel, however, is highly inadequate. The situation has been aggravated by inadequate equipment maintenance and fuel shortage problems which have delayed or prevented kerosene deliveries. These problems significantly reduce the number of successful immunizations achieved and result in a great wastage of vaccines. They also increase costs because of the extra travel required of health center/dispensary staff to collect additional vaccines. In addition to the effect on the immunization program, the absence of an adequate fuel supply has other implications. For example, significant hazards to life can occur from lack of adequate lighting when dealing with emergencies at night and also from inadequate sterilization resulting from lack of fuel to boil water. Hot water needs (e.g., bathing newborn babies, dressing wounds, etc.) also cannot be met because of this fuel shortage. 10. The mission concluded that the principal applications for solar energy in the rural health centers and dispensarIes were in: (a) power- ing vaccine refrigerators; (b) providing lighting; and (c) heating water for use at childbirth and other end uses and boiling water for steriliza- tion. Other applications where solar energy could make an effective con- tribution were in: (d) pumping water from existing open wells at five health centers and 10 dispensaries currently using sub-standard water from open wells; and (e) water pumping for irrigation on small farms in the McCarthy Island Division. 11. The mission recommends that an equipment package consisting of a solar-powered photovoltaic vaccine refrigerator, a pair of fluorescent Lights (to run off the same power supply as the refrigerator) and a small, integrated solar water heater and storage tank unit shouLd be ^v suPplied to 25 of the rural health centers and dispensaries involved in the EPI. The gross procurement, shiprina and installation cost for these 25 sets of equipment, plus an adequate inventory of spare parts, is estimated at US$161,840. 12. The quantifiable benefits of using solar refrigerators are: (a) substitution for kerosene supply and delivery costs: US$778 per year per refrigerator, or a total of US$19,450; (b) in terms of reduced vac- cine losses, an estimated US$4,605 worth of vaccines will be saved per year through reduced refrigerator down-time; and (c) improvement in vac- cine availability due to an anticipated reduction in refrigerator down- time. It is expected that, with refrigerator availability, the producti- vity of the immunization program will improve by over 20%. The annual- ized life-cycle costs using solar refrigerators instead of kerosene are expected to be reduced from US$1,018 to an average of US$762. The cost per dose will drop from US$1.12 to approximately US$0.91 -- a saving of 191. 13. Other benefits that cannot easily be quantified include: (a) reduced maintenance workload for technicians and medical personnel (i.e., no more kerosene burners to be cleaned and adjusted); (b) reduced need to transport vaccines to an alternative refrigerator many kilometers away when kerosene shortages force a shutdown; and (c) reduced incidence of disease if a more productive immunization program is achieved. 14. The provision of lighting does not replace any comparable existing technology, nor does it significantly affect DOH's recurrent costs. The main, justification is that a certain minimum level of lighting is necessary to operate an effective rural health service, and that the cost of providing the lighting "on the back" of the solar vac- cine refrigerator is small. Substituting photovoltaic lighting will result in a nominal annual saving of about US$650 (US$26 per center) of kerosene. If, at each center, the same degree of illumination were to be provided by kerosene lamps as would be obtained by the proposed fluor- escent lights, six hurricane lamps would be required and the fuel cost would be about US$250 per year for each center. The unquantifiable benefits of providing lighting include: reduced risk for patients; im- proved staff and patient morale; and elimination of the frequent need for the staff .to pay for candles and flashlights at their own expense. 15. Similarly, the provision of a small solar water heater at each center would meet an obvious need for hot water. This will considerably reduce the occasional use of kerosene or the frequent effort required at present to collect firewood for water heating. It would cost about US$60 a year with a kerosene heater to provide the equivalent energy of a solar water heater at each center. Other benefits of providing solar water heaters are: improved hygiene (hence, reduced health risk); reduced work for staff; improved staff morale. 16. The mission concluded that the need for solar water pumps at health centers and dispensaries with sub-standard water supplies was not - vi - clear-cut. Despite the technical feasibility of solar pumps, a generally applicable "package" could not be recommevted because the water require- ments were different at each site. Moreover, the economic feasibility of solar pumps compared with wind, hand, and diesel pumps had not been established. Similarly, while the technical feasibility of using solar pumps for irrigation has been estabiished, a detailed economic evaluation has not been done. Consequently, the mission recommends that a separate consultant study be undertaken to assess the role of solar water pumps for the health facilities and irrigation in the McCarthy Island Divi- sion. The study will cost between US$15,800 and US$23,800 to complete. The Rural Telecommunications Network 17. Background. In The Gambia, the telephone network is operated by Gamtel, an autonomous corporation formed in 1984 out of the former government Department of Telecommunications (DOT). The Gambia has approximately 3,400 telephones, of which fewer than 300 are distributed throughout the countryside where 80a of the population lives. The present quality of the domestic service provided by Gamtel is generally poor, particularly in the rural network. While plans are being made to rebuild the peri-urban Banjul area network with French bilateral aid, there is no immediate provision for overhauling the rural network. Major technical problems cause a high proportion of the rural telephones to be out of service for extended periods, in some cases almost permanently. This results in high costs to maintain the rural telecommunications sector, a low quality of service and reduced revenues for Gamtel. 18. The rural telecommunications network is based on radio telecom- munications using a single UHF multi-channel trunk route running the length of the country linking Banjul to Georgetown at the eastern end. A VHF extension runs eastwards from Georgetown to Basse via Bansang. Various single channel VHF links connect the main exchanges at provincial towns on the UHF trunk to surrounding villages and towns. 19. Power and maintenance problems, which are often interlinked, *are the key factors affecting the performance of the rural telecommunica- tions sector. The mission observed that any faults other than the most commonplace and trivial need to be investigated and rectified by cen- trally-based staff, which is a time-consuming operation. Ironically, the repair teams are often incommunicado for many hours due to the very faults they are trying to correct, and therefore cannot be notified from headquartars of further problems that may have arisen in their area because they do not have two-way radio/telephones in their vehicles. The vehicles are often ill-equipped, and major components frequently have to be taken back hundreds of kilometers to Banjul for minor repairs because adequate mobile facilities for testing are not available. Similarly, there are to spare batteries, so that flat batteries have to be taken away to be recharged and returned later at great expense in travel costs and resulting in unnecessary extra down-time. - vii - 20. The importance of adequate telecommunications for economic development is difficult to establish, but examples of tangible benefits exceeding costs by several huuidred percent have been found in case studies relating to the development of rural economies in many developing countries, including some in West Africa. 4/ Therefore, there are good reasons for believing that improvements to the sparse and inadequate Gambian rural telecommunications network would have a net positive impact on the economic development of the country. a 21. Three main areas with maintenance and power supply problems which can be effectively addressed by solar-related equipment were iden- tified by the mission. These are: (a) the national* trunk system, (b) the provincial exchanges, and (c) the single channel vHF lines between the village terminals and the provincial exchanges. 22. The National Trunk System. The UHF system is equipped with 48 channels. Beyond Georgetown, the trunk backbone continues as a UHF nomi- nal 24-channel link to Bansang and then 16 channels continue to Basse. There are also UHF 8-channel branches from Banjul to Barre, irom M4ansa Konko to Farafenni and from Georgetown to .untaur. The syscem uses drop repeater stations at the provincial exchanges, but three repeater relay stations are necessary to bridge gaps between provincial towns. rhese repeater relay stations, at Jelekato, Jattaba and Makagui, ace identi- cally equipped with twin 3.4 kW diesel generating sets, owned and main- tained by Gamtel to provide their power requirements via a bank of bat- teries which are charged on a 12-hour per day duty cycle. Power supply problems with these repeaters have been responsible for a number of sys- tem failures in recent months and require frequent travel by engineers' and technicians from Banjul to implement repairs. Fuel shorLages com- pound the problems. The repeater stations are vital to the nationai telephone network. In fact, two of them carry all trafEi_ becw:t-=n BAiLjuL and the central and eastern part Ot tho ccrutry. Thus. cape.arer SLat,.on failures can have major straregic conseqv-inces since hey Tav cause a complete break in communications between onjui. and the ptovirnces. AlsX, breakdowns result in revenue losses which are probably substantial but cannot readily be quantified from available records. Therefore, it is imperative that reliable power supplies be- ptovided at rho repoat,r stations. 23. The mission recommends that the power supplies ai all three repeater relay stations on the UHF trunk route be converted to solar power. This will involve the provision of solar photovoltaic arrays and new batteries. A system sized to deliver sower with 99% availabiiity in the least sunny month will cost about US$20,500, giving a Sross procure- ment and installation cost for the chree stations of UJS$61,5OO. 4/ For example, tTU-DECD Project Document, TeLecommunication for Devel- opment, International Telecommunications Union (ITU), reneva, 1983. - viii - 24. A conservative analysis of cost data indicates that the recom- mended solar powered systems will produce electricity at about 53% of the cost of the diesel generated electricity. The savings in direct recur- rent diesel engine O&M costs for Gamtel resulting from the use of solar power for the repeater stations will be approximately US$11,400 per year. Other important benefits include: reduced demand on the limited mainten- ance resources of Gamtel; reduced system down-time (and hence increased traffic revenue and a more secure network); the release of six generating sets with about three to four years useful life each as extra standby plant for use at telephone exchanges; and the release of the rooms which currently house diesel engines for use as maintenance workshops. 25. Provincial Telephone Exchanges. There are seven exchanges 5/ associated with the trunk network, plus a small manual exchange at Kerewan to serve the town's 8 Direct Exchange Lines (DELs) and one re- maining (intermittently available) rural spur to Salikene. All the provincial exchanges and repeater stations are dependent either on CUC local mains electric supplies or on Gamtel diesel generating sets, or sometimes on both. Since neither the mains supply nor the generating sets generally can provide a 24-hour service, the telecommunications equipment runs from a bank of lead-acid batteries which are charged by the mains or the generator. Maintaining an adequate state of charge in the batteries is proving to be a general problem for many of these exchanges. Irregular charging has led to excessively deep cycling of the batteries which in many cases has caused them to deteriorate and further reduce their capacity. 26. The mission recommends that batteries with four times the capa- city of the e xiting equipment be provided for the inter-town radio network. The investment necessary to undertake these improvements at seven key rural exchanges will be approximately US$39,200 ($5,600 per exchange). It also should be noted that in any case most of the existing batteries are in urgent need of replacement and will cost Gamtel a mini- mum of US$1,450 per exchange (or US$10,150 altogether), so this invest- ment represents a marginal extra cost of around US$29,000, or US$4,285 per exchange. 27. The recommended changes will allow much longer periods of operation without the need for using standby generators (extended from 11 hours to 44 hours for 50% discharge) and will also allow recharge at a higher rate without damage; so a full recharge can be obtained in 7.7 hours instead of 12 hours. Moreover the depth of cycling of the bat- teries will be reduced so as to extend their useful lives from the current 5.5 years (at best) to more than 13 years. Regarding benefits, the cash savings in recurrent costs depend entirely on how much use of standby generators is avoided. No saving is obtained with 100% u-se of 5/ Farafenni, Mansa Konko, Kaur, Kuntaur, Georgetown, Bansang, and Bassev -ix GUC mains power, but US$3,800 per exchange can be saved when 100% use of standby generators is necessary. On the basis of past records, it seems likely that significant savings will accrue, probably around US$1,000 to US$2,000 per exchange, or US$7,000 to US$14,000 p.a. in total. 28. Although solar power can be cost-effective in situations where the only alternative is a small diesel generating set (as indicated for the case of the repeater stations), it cannot be cost-competitive with GUC mains power costing only about US$0.13/kWh, provided, of course, that such power is available. Therefore, the mission concluded that the prob- lem at rural exchanges relates to the intermittancy of the mains power supply rather than the high cost of providing a small diesel supply and that solar power would not be justified in this case. Hence, it was recommended to address this problem by re-equipping the rural exchanges with significantly larger battery banks to provide power for the crucial radio telephone network. 29. Single-Channel VHF Lines _o the Villages. These lines are be- set with serious reliability problems. The mission visited a number of VHSF single channel links and found that at least 14 out of 17 were not working and only two (solar-powered systems) were generally serviceable, - wfunctioning only when the sun shines strongly enough on the existing solar photovoltaic arrays. The equipment was originally intended to be powered from EMU type primary batteries; these, however, proved expensive and unreliable and were replaced by lead-acid secondary *torage batteries (standard car batteries were used) which have been recharged either by a local generating set or by removing them for recharging elsewhere and substituting a freshly charged unit at appropriate intervals. Charging of batteries in this way has proved to be problematic, since they do not readily survive periods of over-discharge and it has been impossible to arrange reliable recharging. Therefore, a UN/ITU expert seconded to the DOT (GamteL's predecessor) under a UNDP project improvised solar photo- voltaic power systems which were fitted to nine village terminals. The solar powered systems were relatively successful when first introduced but, due to errors in sizing of their batteries, alignment of their arrays and other design problems, most of their batteries appear to have failed after a year or two of use. Consequently, most of the VHF links no longer function. 30. The mission found other individually less serious problems with the VHF single channel link terminals, but which collectively can render the system unserviceable. These include: poorly oriented antennae masts; faults with transceivers housed in poorly ventilated and over-heated con- ditions; human problems such as lack of knowledge as to how to use the telephone, interference by children, and people leaving the hand-set "off the hook"; and indications of poor siting in locations remote from poten- tial users. The mission concluded that, for the VHF systems to operate satisfactorily, it was essential that they be properly designed and installed, and also supervised by trained attendants. After reviewing revenue receipts from various VHF installations, the mission concluded x that efficiently operating systems would restore public ccnfidence and markedly increase revenuqs. 31. The mission also found that all health center and dispensary staff interviewed rated the provision of a telephone as about their high- est priority. If working telephones were generally available at health centert and dispensaries, significant savings in travel and distribution costs could be expected for DOe as well as possible reduction in risk to life through being able to refer emergency patients to hospitals or larger centers with quaLified doctors. 32. The mission recommends that twenty solar-powered new VHF siaigle channel village telephone systems be provided to replace thirteen of the existing systems and also seven new links be provided for health centers and dispensaries without a telejhone nearby. It is recommended that new batteries be provided to-allow eight of the existing solar powered VHF systems to be rehabilitated and made to function effectively in addition to the twenty new systems. This will require a total investment of US$191,000. The mission also recommends that in order to meet Gamtel's need for supervision oif rural telephones and Doe's -need-for telephonesj insofar as possible, the rural telephones be located at or near DOH faci- lities and the DOH staff be given supervisory responsibility for the itelephone in such situations. A mutually agreeable arrangement bet,Ieen the two organizations could be worked out for this purpose. Investigations made by the mission indicated that such an agreement be ween the two organizations is possible. 33. The direct benefits to Gamtel accruing from the proposed new, well-designed, properly `.astalled and reliable solar-powered single- channel VHF links are difficult to determine with certainty, since they depend on the amount of revenue generated. Hlowever, it can be shown that on the basis of current tariff rates and average lengths and costs of rural calls, it will repuire 21 to 27 minutes per day of traffic per link to break even on a pure revenue/cost basis. Since this is only 3.3% of che average avaiLable dayLight hours when the system is most likely to be used, it seems reason.ble to expect this Level of usage not only to be attained, but, based Jr .analysis of revenue receipts, to be sig,iificantly exceeded, especially when public confidence in the telephone system is restored. The revenues that can be expected at each station from, say, 80-160 minutes of telephone calls, range from US$96,000 to US$192,000 a year. Also, significant benefits can be expected for the rural economy. The Project 34. To implement its recommendations for introducing solar photo- voltaic systems in the rural health and telecommunications sector, the mission proposes a project comprising the following elements: - Xi - (a) the procurement of equipment for the rural health and telecom- munications sectors, as discussed above; (b) a technical assistance program to assist with the installation, commissioning and initial period of operation of the equipment, and to help train Gambian personnel in using it; and (c) a consultant study to assess the potential of solar water pump- ing for water supply in the rural health sector, and for irri- gation in the McCarthy Island Division. 35. The invitation of tenders for supplying equipment and the re- sulting selection of equipment would be implemented by personnel from the relevant Gambian departments, assisted where necessary by advisers from associated donor agencies. As part of the technical assistance a short- term "Specialist Adviser" would be recruited for the initial phase which would include equipment approval, installation, and commissioning and subsequently would be avaiLable for two more brief visits to respond to any specific requests for advice which may arise during the course of the project. The Specialist Adviser would help select and train an expa- triate "Assistant Engineer" who would remain in the country for about 15 months, that is, until the equipment has been functioning satisfactorily for some months and is ready to be handed over to The Gambian counter- parts. The Specialist will be responsible for producing a final report on the project. This should be of interest not only to The Gambia but to other developing countries. 36. The Consultant study on water supplies for health facilities and for small-hoLder irrigation in McCarthy Island Division is also pro- posed as a separate activity. However, the are-investment study is included as an element of the technical assistance recommended for this project. rhe study would involve a mission to The Gambia to conduct the necessary investigations. The recommendations that result would require separate funding, but it is possible that they could be implemented in conjunction with this project. Schedule and Budget 37. -The total duration of the proposed project is about 27 months. If work commenced in June, 1985, as planned, equipment should begin to arrive in February, 1986 and installation should be completed by June, 1986. The final project report would be prepared at that stage for completion by about June, 1987. The tctal butdgetary requirement is estimated at US$612,000-678,000, consisting of US$500,000 for equipment procurement and US$112,000-178,000 for technical assistance. The actual cost is likely to be cLoser to the lower figure. r - xcii - Conclusions 38. The mission- recommends undertaking the proposed project as the majority of the energy problems relating to the rural health and telecom- munications sectors can be significantly reduced by introducing solar energy systems. For an estimated expenditure of US$612,000-US$678,000, W including about US$15,000 for the follow-up consultant study on solar water pumps, the project will yield gross quantifiable benefits of US$145,000-US$250,000 annually. This would give rates of return on investment of 24-41Z based on a 0S-year system life. The proposed iniestment will also yield a number of unquantifiable but nevertheless important benefits which have been discussed earlier. 39. Finally it is worth noting that this project will be unique in that it uses, for the first time, solar-powered systems to substitute for conventional systems throughout two sectors of a country. It is expected that, apart from its value to The Gambia, much can be learned from the proposed project that will be of considerable relevance to a number of other developing countries. I. DIUTODUCTION Background to the Report 1.1 The Gambia relies entirely on imported petroleum to meet all its commercial energy needs, including electricity. In recent years, the country has encountered energy problems which, in large part, can be attributed to difficulties in servicing the import bill for petroleum and subsequent interruptions in petroleum supply. The government has re- quested assistance from various donor agencies to develop a strategy to resolve these problems. 1.2 A mission under the Joint UNDP/World Bank Energy Sector Assess- ment Program visited The Gambia in August, 1983, and prepared a report 6/ to assist the GOTC transform its energy policy into action. This policy, outlined in broad terms, aims to: (a) secure adequate energy supplies to meet future requirements of the economy; and (b) minimize the cost of energy to the economy by improving the efficiency of energy use. The Energy Assessment Report reviewed all aspects of the energy sector in detail and identified solar energy as an indigenous energy resource which could substitute for imported oil in certain applications. Subsequently, in April, 1984, at the request of the COTG, a Technical Assistance Identification mission 7/ under the Energy Sector Management Assistance Program (ESMAP) identifted that solar photovoltaic applications had the potential to improve the reliability of energy supply for critical needs in the rural health and telecommunications sectors and prepared detailed terms of reference for a pre-invettment study. A follow-up ESMAP mission visited The Gambia in July, 1984, to prepare this report. Objectives of the Report 1.3. The primary objective of the July, 1984, mission was to produce a pre-investment study which would assist the GOTG obtain external funding to replicate the use of solar photovoltaic power units in the rural health care and telecommunications systems. The objectives as set out in the mission's terms of reference were: (a) to establish the costs and benefits from the introduction of solar photovoltaic power units to supply some or all of the 6/ Joint UNDP/World Bank Energy Sector Assessment Program. The Gambia: Issues and Options in the Energy Sector, Report No. 4743-CM, November 1983. 7/ This mission comprised Mr. Amarquaye Armar. critical electricity needs of the existing network of rural health centers and dispensaries in The Gambia (for refrigera- tion of vaccines, for lighting in-patient rooms; ror hot water sterilization; and, possibly, for pumping water from wells); (b) to establish basic specifications for sotar photovoltaic power units to supply a combination of the above needs for which potential savings can be made, and to establish the level of investment required to supply these systems to various health centers and dispensaries; (c) to review the existing recurrent costs and the status of the use of solar photo'oltaic power units at some of the rural single channel VHF stations in The Gambia's rural telecommuni- cations system, as a basis for establishing the level of investment required to retrofit remaining rural single-channel VHF stations with solar photovoltaic power bits; the assessment of single-channel VHF stations to be done with special refer- en;e to the needs of the Government of The Gambia's Primary Health Care Program; (d) to assess the feasibility and level of investment required to retrofit existing diesel power units at UHF repeater stations and provincial multi-channel VHF stations with solar photovol- taic systems; (e) on the basis of (a), (b), (c), and (d) above, to prepare a report specifying investment costs, benefits, and an implemen- tation plan: (i) of an optimal program to equip health centers and dispensaries with appropriately sized solar photovoltaic power systems for their essential electricity needs (including or excluding water pumping); and (ii) to equip rural VHF stations with appropriately sized solar photovoltaic power systems; and (f) to briefLy look into the feasibility of utilizing solar photo- voltaic powered pumps for smallholder irrigation in the McCarthy Island Divison. Viable Solar Energy Applications 1.4 In both the health and telecommunication sectors, technically reliable and cost-effective solar energy systems can help meet important energy needs. In the health sector, living conditions for the majority of the population are poor and there is malnutrition combined with a high incidence of disease. Much of the disease is due to contaminated water supplies or epidemics that could be controlled through mass immuniza- tion. Infant mortality is particularly high in the rural areas where in some districts more than one-third of the children die before the age of -3- two. A successful immunization program and improved village water wells are critical to the country's development and, as such, are items of high priority for the government. The ESMAP mission investigated ways of con- tributing to improvement efforts and found that, since fuel and power supplies are erratic and unreliable at rural health centers and dispen- saries, the provision of solar photovoltaic vaccine refrigerators and solar water heaters would go a long way towards alleviating the current energy supply problems. The systems would also help provide minimal but essential lighting at these centers/dispensaries. 1.5 In the telecommunications sector, the problem has been that of fuel shortages, unreliable power supply, and inadequate maintenance. It is clear that telecommunications is very important for economic develop- ment, and this is particularly true for The Gambia, where 80% of the population is dependent on a rural network which provides generally poor service. The Government of The Gambia (GOTG) views the rehabilitation and improvement of the telecommunications sector as important components in the national development program. The ESMAP mission identified three main areas with significant maintenance and power supply problems which can be effectively addressed by introducing solar photovoltaic powered systems. These areas are: (a) the national trunk system; (b) the pro- vincial exchanges; and (c) the single channel VHF telephone lines 'between the village terminals and the provincial exchanges. 1.6 The technical feasibility of solar water pumps is well-estab- lished for: (a) water supplies at health centers and dispensaries cur- rently using sub-st4ndard water from open wells; and (b) irrigation for small landholdings im the McCarthy Island Division. However, it will require extensive analysis to determine the economic viability of such applications. The scope of the investigation was too large to be carried out in this report and, as such, warrants a separate consultant study. The Project 1.7 The recommendations of the mission have led to the formulation of a project which is presented in this report. The project has three main elements: (a) procurement and installation of solar energy equip- ment; (b) technical assistance to help The Gambia's personnel procure and install the equipment, and also train them in its use; and (c) a consul- tant study to assess the potential of solar photovoltaic water pumps for use in the health sector and in irrigating small landholdings in the McCarthy Inland Division. The project will cost an estimated $612,000 - 670,000, require 27 months to implement, and has a rate of return of about 24-41% based on a 15 year system life. In addition to making an impact on two sectors of The Gambia's economy, the project will also serve as an important example for many other developing countries. -4- II. UEALT1 SECTOR Rural Health Sector and the Immunization Program Structure of the Service and its Organization 2.1 The primary health care program in The Gambia has two main com- ponents: the Expanded Program for Immunization (rI), and the Maternal and Child Health Program (MCH). The components are managed through a nettwork of 18 rural health centers, 13 dispensaries and 2 private sector missions which cooperate with the government program. The dispensaries are generally smaller, less well-equipped and have fewer staff than the health centers. Details of the primary health care network are shown in Map 18624 and Table 2.1. 2.2 It can be seen from Table 2.1 that the health centers range in size from 5 or 6 staff members and a few beds, to 12 staff members and up to 20 beds (becoming, in effect, small hospitals). Four of the health centers have no pow&- supply, and the rest have only intermittent ser- vice. In one or two cases, recently constructed centers have operating cheaters which have not yet become functional because of unreliable electricity supply. The dispensaries are very basic, usually employing four para-medical or nursing staff, working with minimal equipment from a single building. Most have one or two beds, usually intended for deliv- ery of babies; serious cases are generally referred to either the nearest health center or to a hospital. None of the dispensaries have any power supply. It is government policy to upgrade all dispensaries into health centers by the year 2000. There is also a network of rural sub-dispen- saries which do not have permanent staffing but which are serviced from the dispensaries and health centers usually once or twice per week. This work is carried out by a small team that visits the sub-dispensaries by Land-Rover to hold regular clinics. Vaccines are transported to these clinics in insulated cold-boxes. The sub-dispensaries, too, are without power supplies. 2.3 The Gambia also has a number of privately operated health centers or dispensaries run by religious missions. Most of these are outside the state health service and levy fees for their services, but two private missions offer free medical service similar to that of the government centers; these cooperate with the government program. The mission at Sibanor deserves special mention in that it is fully inte- grated into the government program. Consequently, on the recommendacion of the Medical Director for the Rural Health Service, it has been in- cluded in the proposed program discussed in this report. Table 2.1: COVERMIENT HEALTH CENTERS AND DISPENSARIES Location Staff Beds Patients a/ Immunizations a/ Health Centers .akau Basse b/ 30 45 3,000 1,000 + Brikama Bwiamb 6 6 1,200 300 Essau Farafenni Fatoto 6 8 - 800 Georgetown Gunjur b/ 6 2 Kaur 9 14 3,000 500 Kerewan - 4 Kiang Karentaba / 12 20 Kudang 12 20 5,000 400 Kuntaur 5 5 2,000 1,750 Mans4 Konko Serrekunda Sibanor Yorobawal - 8 Dispensaries Baja Kunda - 1 Brikamaba 4 1 1,000 1,800 Brufut Chamen Dankunku 4 2 3,000 400 Diabugu Basilla / Gambisarra 4 1 1,500 1,000 Kuntair Marakissa Medina Bafaloto Sami Karantaba Salikene - 4 Sukuta - 2 */ Estimates of monthly averages. b/ Health centers and dispensaries visited by the mission. c/ Private sector missions which are integrated with the government's EPI program. Source: DOH, The Gambia. -6- 2.4 Official statistics for the catchment size served by individual health centers and dispensaries do not exist, but the villages in which most are located typically comprise 4,000-10,000 people, with another 10,000 or more people in the surrounding hinterland. Several dispensers estimate their total catchment at about 20,000 people. The six health centers and four dispensaries visited by the mission dealt with about 1,000 to S,000 persons per month. 2.5 There are two full-scale hospitals in the country, one in Banjul$ the capital, and the other in Bansang, which serves the eastern half of the country. These two fall outside the scope of this report as they are not administered within the rural health program. The Expanded Program for Immunization (EPI) 2.6 The target population for the BPI is the below-23 months age group. They are being vaccinated against yellow fever, diptheria, tuber- culosis, polio and measles. Pregnant women are also vaccinated with tetanus toxoid as a precaution against infection at childbirth. 2.7 The RPI depends on the availability of reliable refrigeration capacity at health centers and dispensaries for storing vaccines. Most vaccines need to be stored at temperatures in the -10' to oeC range. Some, however, may be damaged by freezing and need to be stored in the +4" to -+8C range. Any deviation from these temperature ranges tend to render the vaccine ineffective; the longer the deviation and the further from the recommended storage temperature, the shorter the shelf-life of the the vaccine and greater the risk of spoilage. 2.8 At present, all health centers and dispensaries use refrigera- tors to store vaccines. Due to the unreliable mains electrical power supplies in The Gambia, only two health centers have mains electric re- frigerators, both located in places with better than average power supplies. Two other health centers have experimental solar photovoltaic- powered electrical refrigerators supplied by an American company, Solar Pawer Corporation; these were provided under a project funded by USAID (para. 2.15). The rest of the health centers and all the dispensaries rely on 154 liter-volume kerosene-fuelled absorption refrigerators, (model RAK660) manufactured by Electrolux, a Swedish company. The kero- sene refrigerators which normally operate with reasonable reliability are beset with major problems in The Gambia. Fuel supply and maintenance facilities for them are quite unreliable. With the consequent refriger- ator down-time and loss of vaccines this results in a serious drop in the rate of immunization. 2.9 Details of the costs of the MCH/EPI program in The Gambia for the fiscal years 1982-83 and 1983-84 are shown in Table 2.2, which re- flects a major increase in the cost of the program. This increase in costs, from Dalasis 1.4 to 2.4 million, is probably related to the devaluation of the Dalasi that has occurred over the past few years (from Dalasis 1.68 to US$1.00 in 1981, to Dalasis 3.75 for US$1.00 in 1984). It is difficult to separate BPI and MCH costs, since they share many staff and facilities, but the latest year for which a breakdown of the actual SPI program costs was available was 1980-81 and at that time the BPI budget accounted for about 701 of the total. Table 2.3 presents average monthly v&ccine disbursements recorded for the BPI for 1983, and approximate unit costs and total costs for different vaccine types. As shown, about 492,000 doses of vaccine are being distributed each year in The Gambia at a cost of some US$31,000. It should be noted that the vaccine budget is financed mostly through international grants as a part of the International CCCD Program, which is providing funds to health programs in Africa. The exact breakdown between local and aid funds in the procurement of vaccines is not known. Table 2.2: MCH AND EPI RECURRENT EXPENDITURE, 1982-84 (Dalasis) 1982/83 1983/4 MCR/EPI Unit 139,710 143,840 MC/REPI Salaries/Wages 67,110 69,340 Health Centers 878,890 1,737,770 Dispensaries/Sub-Disps. 314,860 463,080 Total 1,401,170 2,414,030 Source: Government of The Gambia "Estimates of Recurrent Revenue and Expenditure, 1983-84". Table 2.3: ANNUAL VACCINE DISBURSEMENTS, 1983 Vaccine Type Delivered Unit Cost Annual Cost a/ (doses/yeal (USU) f(uS) DPT 106,068 0.50/20 3,050 Polio 126,444 0.40/20 2,858 BCC 81,480 0.95/20 4,450 Measles 52,716 1.25/10 7,577 Tetanus Toxoid 89,880 0.45/20 2,325 Yellow Fever 379236 5.00/20 10,705 Total 491,664 30,965 a/ Total allows for 151 excess required to cover unavoidable waste in discarding used vaccines in vial at conclusion of clinic. Source: DOB estimates. 8 - Energy-Related Problems CUC-Powered Health Centers and Dispensaries 2.10 Twelve health centers in some of the larger provincial centers and villages are provided with electricity by The Gambia Utilities Corp- oration (CUC) electricity supply. Table 1.4 lists the health centers having power supplies and indicates the nature of those supplies. The CUC is also generally responsible for water supply to those centers, since the water is usually pumped from a borehole. 2.11 All mains electricity in the provinces is supplied from isolated networks energized by small diesel generating plant that are rated at tens or, in a few cases, hundreds of kilowatts. Although these isolated electricity systems originally were intended to provide 24-hour service per day, GUC policy since May 1984 has been to run the generating plants for no more than 12 hours in 24, usually in two 6-hour periods from 6:00 a.m. to 12:00 noon, and from 6:00 p.m. to midnight. GVC adopted this policy in response to the worsening fuel supply situation in The Gambia which reflects the shortage of foreign exchange with which to pay for petroleum products. In reality, however, electricity supply has virtually never been continuously available in most provincial centers, and even the current daily 12-hour service can hardly be maintained. While fuel shortage is the primary cause, there are also significant maintenance problems. Given the high cost of maintaining the isolated systems and the recurring Cael shortages, it seems highly unlikely that the degree of reliability offered by solar photovoltaic systems will ever be achieved by CUC in its provincial electricity supply services. 2.12 At some of the smaller centers, the local diesel generating plant provides power almost exclusively to the health centers, and is some cases managed by DOH. However, DOH has recently handed over the responsibility for operating all such diesel generating plants to the GUCC whose own staff operate and maintain the equipment. In cases where the equipment is small and used principally for supplying the needs of the health center, the GUC charges DOH on a full cost recovery basis, but regular GUC electricity tariffs apply when the supply comes from the local GUC station. In the situations where the CUC charge is on a full cost recovery basis and the local power plant is out of service for extended periods, DOH has to pay significant overheads for power (Tables 2.5 and 2.6). Where a CUC provincial grid is involved, the tariff varies from D 0.335/unit (for first 30 units) to D 0.51 (for over 1,000 units). However, as of January 1984, CUC has increased its tariff by an average of 30Z, which has in effect raised the price of electricity to some USCIO-16/kWh. Table 2.4: SERVICES AVAILABLE AT HEALTH CENTERS AND DISPENSARIES Power Water Location Supply Supply Telephone Refrigerator Health Centers Bakau mains piped yes kerosene Basse mains piped yes e/ kero & elec Brikama mains piped yes electric Bwiam diesel borehole yes e kerosene Essau none well nearby kerosene Farafenni mains piped nearby e kerosene Patoto diesel piped nearby kerosene Georgetown none a/ piped nearby kerosene Gunjur none well nearby solar PV Kaur none b well nearby e/ solar PV Kerewan none well nearby kerosene Kiang Karantaba diesel borehole yes e kerosene Kudang diesel borehole nearby kerosene Kuntaur none river nearby kerosene Mansa XConko mains piped yes electric Serrekunda mains piped nearby none Sibanor diesel borehole no kerosene Yorobawal diesel piped yes c/ kerosene Dispensaries Baja Kunda none well no kerosene Brikamaba none veil t no kerosene Brufut none borehole nearby kerosene Shamen none well no kerosene Dankunku none well nearby !/ kerosene Diabugu 8asilla none well nearby e kerosene Gambisarra none well nearby kerosene Kuntair none well no kerosene Medina Bafaloto none well no kerosene Sami Karamtaba none well no kerosene Salikene none well nearby e/ kerosene Sukuta none well nearby kerosene a/ Mains power in town, but health center not connected. b/ Health center about to move to new premises with diesel. c/ Telephone equipment removed after failure and not replaced. d/ Well frequently dries up leaving no water supply. eI Telephone believed to be generally not working. Source: DOH, The Gambia. 10 - 2.13 Some typical examples illustrating the power supply situation for those health centers having GUC connections are: (a) The large health center at Basse (30 staff and 45 beds) re- ceived no electrical power at all for five and a half months (January to June 1984) due to major maintenance and repair problems with the local GUC power plant. This failure caused great inconvenience to this health center since its water supply, and hence sanitation, lighting, and refrigeration were all dependent on regular electricity supply. Medical staff were reduced to driving in private cars to a public water point some distance away and bringing water back in buckets. (b) The large Kudang health center, only just completed with bila- teral assistance from the Peoples' Republic of China, is more like a small hospital than a health center as it even has a well-equipped theatre. However, it had no electrical power for over twelve weeks prior to the mission's visit due to a tech- nical problem in integrating the center's new diesel generating plant with the local GUC network. (e) The small (59 kVA) generating station at Bwiam is limiting its output to only 6 hours per day (6 p.m. to midnight) because of inadequate fuel supplies. (d) The Peoples' Republic of China has built two other health centers, also equipped with diesel generating plants, at Kaur and Kiang Karantaba; these centers have electric water pumps, numerous electric lights, refrigerators, and other electrical equipment. Although the generating plant is new and mechan- ically reliable, it is expected that restricted fuel supplies will limit the electricity supply even after regular operation of the generating plants has been established by GUC. (e) Although Georgetown, the provincial administrative center for McCarthy Island Division, has GUC power supply, the health center there is no longer connected to it. However, the GUC still provides the water supply. Other Healch Centers and Dispensaries 2.14 As mentioned before, most of the smaller health centers and all the dispensaries are without either electric power or piped water supply. They rely mostly on kerosene to power their vaccine refrigerators, and either kerosene or candles, and sometimes even privately owned flash- lights, for lighting. Hot and boiling water is produced by using kero- sene or, more often, wood. Drinking water from the wells is obtained through the use of buckets. Inadequate and unreliable kerosene fuel supplies is the major problem facing these health centers and dispen- saries. * ~~~~- 11 - Table 2.5s ELECTRICAL POWER PROCURED ON A FLAT-RATE COST BASIS FROM GUC BY THE DOH FOR HEALTH CENTERS WITH GUC-RUN CENERATING PLANT Cost in Dalasis per Quarter Health Center Ist/83 2nd/83 3rd/83 4th/83 lst/84 Kudang 20,475 19,866 25,467 25,644 29,463 Karantaba 6,553 2,581 (1,266) (l,266) (1,266) Patoto 13,169 9,349 8,485 9,121 10,399 Bwiam 9,687 9,211 11,114 12,312 17,534 Yorobawal 23,759 10,298 (4,367) (4,556) 10,069 Note: Figures in parentheses indicate periods when breakdowns prevented any power from being delivered. Variability of other figures indicates shorter breakdowns probably occurred and reduced operating costs as less fuel would have been consumed as a result. Source: GUC Accounts Department. Table 2.6: ELECTRICAL POWER PROCURED ON A METERED COST BASIS FROM GUC BY THE DOH FOR HEALTH CENTERS TAKING MAINS ELECTRICAL POWER Cost in Dalasis per Quarter Health Center 2nd/83 3rd/83 4th/83 1st/84 2nd/84 Mansa Konko n/a 760 1,067 981 306 Mansa Konko n/a n/a 4,080 852 50 Georgetown 0 0 0 0 272 Farafenni n/a n/a 177 228 245 Bansang hospital 462 575 535 237 1,245 Basse n/a 673 986 0 0 Note: Zero figures indicate periods when no power was de- livered. Variability of other figures indicates that shorter breakdowns probably occurred and reduced the power that could be consumed. Source: GUC Accounts Department. - 12 - Solar Photovoltaic Refrigerators for Vaccines 2.15 The dissatisfaction with kerosene refrigerators has spurred interest in solar refrigerators in The Gambia. Two solar photovoltaic powered (electric) vaccine refrigerators have therefore been introduced on an experimental basis in The Gambia at the health centers in Gunjur and Kaur. These are identical units manufactured by Solar Power Corp- oration Inc., a former subsidiary of the Exxon Oil Company which has recently been liquidated. Each unit consists of an electric compression refrigerator/freezer in a well insulated, top-opening casing. The unit is powered from a bank of six 105Ah 12V Delco model 2000 lead-acid bat- teries, which are charged by a photovoltaic array rated at approximately 315 i(p). 2.16 These units were supplied by the Center for Disease Control, (CDC) of Atlanta, Georgia (USA) working with the National Aeronautical and Space Administration (NASA), Lewis Research Center, and financed by USAID. The two solar refrigerators in The Gambia have been fitted with instrumentation to measure internal and ambient temperatures, plus other key operating parameters such as solar irradiance, operating hours, battery voltage, etc. The data are analyzed periodically by NASA-Lewis in the U.S. 2.17 The mission was told that the two solar refrigerators had performed quite satisfactorily since their installation, except for one occasion when the unit at Kaur had to be defrosted and disconnected for about three days. This was done on instructions from NASA-Lewis in order to fully recharge the batteries after the data had indicated that the batteries were not staying at an adequate voltage. There are two prob- able reasons for inadequate voLtage: first, there may only have been a minor fault either in the charging circuit or in one of the batteries since all the primary components appeared to be functioning properly; second, there may have been some abuse of the system, e.g., through overloading of the refrigerator by the operators with food or drinks in addition to the normal vaccines and ice-pacKs. The former could have been corrected easily if the attendant were properly trained, as is planned through the present project. The latter fault could have been prevented by equipping the refrigerators with-a padlock, as indeed the kerosene refrigerators are currently equipped. The experience with solar refrigerators in The Gambia can, therefore, be considered successful. Well-defined Needs: Refrigeration, Lighting and Water Heating 2.18 There are certain energy-related applications which are common to all The Gambian rural health centers and dispensaries where solar energy systems might be used quite effectively. For these applications the solar energy systems can form a standard "packaget which would be common for all intended locations. These are: refrigeration of vac- cines, lighting; and water heating for bathing newly delivered babies as well as for sterilization. - 13 - 2.19 Refrigerators for Vaccines. It has already been demonstrated in The Gambia and several other developing countries that the use of solar powered vaccine refrigerators is technically feasible. Further- more, there are now a number of solar vaccine refrigerators on the market. Therefore, it is feasible to invite bids for the supply of reliable solar powered vaccine refrigerators which could be used to substitute for all the existing kerosene refrigerators in rural health centers and clinics in The Gambia. Technically speaking, the solar refrigerators can be introduced into The Gambia to successfully overcome the problems currently being experienced due to kerosene shortages and distribution difficulties, provided only well-proven equipment is procured, local staff are trained, procedures to prevent misuse are implemented, and suitable technical assistance is provided in the initial stages of the project implementation. 2.20 It is important to ensure that the investment in solar photo- voltaic vaccine refrigerators represents a reasonably sound use of devel- opment funds when compared with the costs of the existing kerosene fuelled system. As such, this report has analyzed the economics of solar versus kerosene vaccine refrigerators. As the EPI is not an "economic activity" the only meaningful quantified results relate to the relative costs of the options and their likely influence on the achievement of EPI goals. This latter point is important since the overheads for the EPI are significant at about US$350,000 per year and if better results can be obtained through the introduction of improved technology, it can be argued that this represents a better return in relation to EPI general overbeads. The analysis is summarised in Table 2.7 and presented in detail in Annex 1. Table 2.7: COMPARATIVE ANALYSIS OF SOLAR AND KEROSENE REFRIGERATION (US$) Solar Solar Kerosene Low High Capital cost (installed) 400 4,680 6,569 Recurrent costs 853 60 120 Annualized life cycle costs 901 716 1,108 Reliability (assumed) (X) 80 100 90 Yearly Doses per refrigerator 13,373 16,716 15,044 Refrigerator direct cost/dose .07 .04 .07 EPI overhead per dose 1.05 .84 .93 Total cost per dose 1,12 .88 1.00 *4. 2.21 This analysis assumes two scenarios for solar refrigerators, designated as 'solar low' and 'solar high'. The 'solar low' scenario is - 14 - based on the lowes range of refrigerator prices available in the market plus a higher degree of system reliability (1001 - performs faultlessly throughout the year). The 'solar high' scenario, on the othe- hand, is marked by a higher range of refrigerator market prices and permissible system reliability (901). The two scenarios, therefore, provide the lower and upper bounds for evaluating the economics of solar refrigera- tors in The Gambia. As Table 2.7 shows, the effect of displacing kero- sene refrigerators by solar refrigerators would effectively reduce the cost per dose by about 12-22%. 2.22 The main benefits to be expected from the use of solar vaccine refrigerators include: (a) substitution for kerosene (worth US$19,450 p.a.); reduction in spoiled vaccines (worth US$4,605 p.a.); (b) reduced refrigerator down-time due to elimination of problems caused by kerosene supply problems; (c) increased immunization success rate due to both improved vaccine availability and better chance of being able to provide follow up "booster" shots; and (4) reduced maintenance work-load for skilled technicians and medical personnel, thus releasing them for other tasks. 2.23 Lighting. One of the most serious difficulties pointed out to the mission by medical personnel was the lack of proper lighting at the health centers and dispensaries. This problem applies even in centers nationally connected to a mains supply, since the supply is both inter- mittent and unreliable. Lack of adequate lighting can actually be a hazard to life in emergency situations, such as difficult deliveries of babies, and stitching of serious wounds. These are often conducted either by candlelight or the failing batteries of a small flashlight. Solar PV lights can effectively address this problem. 2.24 The economic benefits of providing lighting cannot readily be quantified, since there are no costs incurred by the DOH that will be displaced, other than a marginal excess of kerosene added to that supplied for the refrigerators at present (notionally 0.06 liter/day delivered, or about US$26.00 per annum). No doubt that small sum of money will also be saved by medical personnel who currently purchase candles and torch batteries from their own funds. Besides that, insuffi- cient information on the mix and usage patterns of different types of lighting sources prevents any meaningful analysis. In addition, none of the existing options can, in reality, be considered as real alternatives since the basic reason for considering solar powered lighting is that existing options are inadequate. However, if the same degree of illumi- nation were to be provided by kerosene lamps as that provided by two 20W solar fluorescent lamps, 6 kerosene hurricane lamps would probably have to be used, with a total recurrent cost of over US$250 per year. 2.25 An important factor to be considered is that, at present, while kerosene is notionally intended to be used for lighting, and a hurricane lamp is provided by DOH, the kerosene supplies in practice are quite inadequate. If the kerorene refrigerators are replaced by solar refri- gerators, then even the limited supply of kerosene that is at present - 15 - available will probably cease to be supplied. So it can be argued that some steps should be taken to provide an alternative to kerosene for lighting, or an already difficult situation could be made worse. 2.26 If solar powered vaccine refrigerators, equipped as they are (in the interest of extra reliability) with siseable battery-electric storage, are provided, then a low powered lighting package could easily be provided at a marginal extra cost in relation to the total budget for the proposed program. This will help provide the minimal lighting for night-time emergencies which should be an integral component of an effec- tive rural health service. 2.27 Energy demand for lighting, and hence costs, will depend on the average daily usage. Unless adequate solar array and battery storage capacity are made available, excessive usage could result in over-dis- charge of the batteries which in turn could introduce a hazard for the stored vaccines. Some measures proposed to guard against this include: (a) conservative sizing of the array; (b) stricter regulation by DOH; and (c) provision of "time clocks" on the lighting systems. 2.28 Water Heating. Hot water is needed for activities such as bathing newborn babies, dressing wounds, and staff hand washing. Boiling water is needed to sterilize instruments and syringes, especially as the high cost of disposable syringes causes them to be re-used, on average, seven times in the rural health sector. A supply of boiled water is also necessary for making up medications and for other applications requiring sterile water. Kerosene left over from the fuel supply for the vaccine refrigerators is expected to be used for heating and boiling water, but as already explained, it is not available in sufficient quantity, and wood is often used instead. It should be noted that often hot water, even when it is needed, is not used for want of fuel. 2.29 It would be a simple matter, technically, to provide a continu- ous small supply of hot water by providing a combined solar water heater and insulated storage tank and thereby help meet some of the basic hot water needs in the rural health sector. Obviously, the hot water would require subsequent boost in temperature (through the use of kerosene or wood) to bring it to boil and meet sterilization needs, but the fuel needs will be more than halved through the use of solar water heating. 2.30 There are a number of commercially available solar water heat- ers capable of fulfilling the needs of The Gambian health sector. Their costs are variable, but to produce about 30 gallons/day of water typi- cally at 60C would require a system costing approximately US$300. Each solar water heater would harvest about 8 million Btus per year. If this energy were to be provided by kerosene, it would cost about US$60. Therefore, for 25 centers the fuel savings would be about US$1,500. Additionally, the availability of hot water would have a favorable impact both on the hygienic conditions in the health centers and on staff morale. - 16 - Less Clear Needs - Telecommunications and Water Supplies 2.31 It is clear that additional serious difficulties are experi- enced in the rural health sector due to either lack of communications or unreliable communications, and also inadequate and often polluted water supplies. However, unlike the needs discus,ed in the previous section, these requirements are less easily solved through c simple technical "package". 2.32 Telecommunications. As indicated in Table 2.4, only 7 out of 30 centers have a telephone, which in many cases is quite unreliable, while one privately-run health center and six dispensaries lack any access to a nearby telephone. In reality, then, with the general unreli- ability of the rural' telecommunications system; the majority of the tele- phones that are installed or lie within an accessible distance are so unreliable that there is little expectation of ever being able to make a successful call, except at centers close to Banjul. Consequently, the health centers and dispensaries are generally not able to use telephone communication. Chapter III describes in more detail proposed steps for rehabilitating and extending the rural telephone system to provide a reasonably reliable telephone link at or near each health center and dispensary. Some of the benefits for the rural health sector 'expected from this effort include: (a) saving of unnecessary travel and distri- bution costs; (b) reduced risk in emergencies; and (c) improved morale for medical staff. 2.33 Without detailed microeconomic studies, the benefits listed above are difficult to quantify, although cash savings should result. For example, benefit (a) above should reduce the amount of travel that is needed by DOH personneL, whose only method of seeking information or advice, or obtaining urgently needed supplies is to travel to an appro- priate source of supply. Additionally, lack of communications prevents the efficient distribution and restocking of consumables, which inevit- ably involves unnecessary journeys, wastage through over-delivery, and inconvenience through under-delivery. 2.34 It is proposed to extend and upgrade the existing rural tele- communications system to improve the reliability of the existing system and to expand it so that all health centers and dispensaries either have a telephone at the center or one in close proximity. The justifications, details and costs of doing this are presented in Chapter III. 2.35 Water Supplies. The mission identified at least five health centers and ten dispensaries with sub-standard water supplies where measures need to be taken to upgrade and improve the existing facili- ties. These facilities are described in Table 2.4. The precise require- ments for each center will generally be different in terms of both water needs (quantities required) and physical pumping requirement (pumping head and distance to be delivered). It was beyond the scope of this mission to conduct a detailed survey for all centers, which would have been necessary to specify the equipment requirements and costs of -17 - providing an adequate and reliable water supply system. It is proposed that the next step should be a consultant study of the water needs of all health centers and dispensaries with the objective of defining in detail the optimum pumping systems to satisfy the requirements. This is dis- cassed in detail in Chapter IV. Recommendations for the Rural Health Sector Solar Photovoltaic Vaccine Refrigerator and Lighting-Package 2.36 It is recommended that: (a) all 25 kerosene-fuelled vaccine refrigerators listed in Table 2.4 be replaced by photovoltaic (PV) vac- cine refrigerators; and (b) each vaccine refrigerator should have, in- cluded as part of each package, two small fluorescent lights running off the same batteries. A total of 26 complete PV systems should be procured in order to provide one complete spare system, and that an additional system should also be procured -- minus the photovoltaic array -- to provide a second complete spare "balance of system" package for use as spares. This latter provision is necessary because arrays are less like- ly to develop faults than the remaining components, but they are the most expensive single element of each system. However, in practice, it is re- commended that when the government invites bids, each prospective sup- plier should be requested to incorporate an appropriate spare-parts in- ventory to accompany-the consignment offered in their respective bids. 8/ 2.37 All the systems to be procured should be identical and provided by one supplier. The provision of an adequate warranty is an important criterion for selecting the supplier. Annex 2 provides a general des- cription of vaccine refrigerators, while Annex 3 gives details of methods recommended for sizing and evaluating the economics of PV svstems. rhe Technical Supplement includes a draft tender document (Annex A) for inviting bids from suppliers which effectively details the procure-ment requirements. Also included as part of the draft tender document are technical specifications of the solar refrigerators to be procured for The Gambia. 2.38 The displaced kerosene refrigerators and kerosene lamps should be retained by the centers concerned and remain the property of DOH rather than being sold. It is recommended that these be offered for the use of staff (who would be responsible for providing kerosene). The displaced equipment, thus, will be available if the solar unit were to fail, which would be unlikely. Also, the availabi-lity of kerosene refrigerators for staff use (for drinks, etc.) would further reduce the possibility of abuse of solar refrigerators by the staff. 8/ The additional cost of two PV systems has been factored into the economics of solar refrigerators in Annex 1. - 18 - 2.39 It is recommended that specialized technical assistance be pro- vided as a component of the project to assist the Covernment of The Gambia in appraising the tenders received, selecting the equipment and then with installing and commissioning the delivered vaccine refrigera- tors. The details of how this should be implemented are given in Chapter V. 2.40 The cost of solar refrigerators and fluorescent lamps is shown below: Table 2.8: COST OF SOLAR REFRIGERATORS AND FLUORESCENT LIGHTS (a) Solar Refrigerators (US$) 26 systems 133,825 I spare subsystem 29425 (minus the array) Total cost of component: 136,250 (b) Fluorescent Lights 2 x 12V 20W fluorescent lighting units 100 680 Wh extra battery capacity to allow 4 h/night operation (allowing for battery efficiency at 70F and up to 33Z maximum discharge) 150 70 Wp extra photovoltaic modules to provide the power supply (2 modules) 420 Total cost of each set: 670 Total procurement of 27 sets (including 2 spares) 18,090 The cost of fluorescent lights can be reduced to one half (i.e., US$9,045) if only one fluorescent light is provided at each center. This, however, will not provide adequate lighting, and should be avoided. Solar Water Heating 2.41 It is recommended that 25 solar water heating systems be procured, one for each center. At a price of US$300 per system, the estimated cost for all 25 systems would be US$7,500. - 19 - Solar Telecoamunication Systems and Water Supolies 2.42 Since the telecoumunication application in the health care program is shared with that of the telecommunication sector overall, the recommendations and the budgetary requirements are discussed in the next chapter. Similarly, the consultant study on water supplies, which is linked with irrigation water pumping in the McCarthy Island Division, is covered in Chapter V. - 20 - III, TBLECONMUNICATIONS SECTOR The Rural Telecommunications Network 3.1 Geographically, The Gambia is a long and narrow country with the capital (Banjul) located at one end. As a result, the telecommunica- tions network is based on a single UHF trunk running the length of the country, linking Banjul to Georgetown via the provincial towns of Mansa Konko and Kaur. A VHF trunk extension runs eastwards from Georgetown to Basse via Bansang. Also, various single channel VHF links connect the main exchanges on the UHF trunk to surrounding villages and towns (IBRD Map 15991R1). 3.2 In July 1983 The Gambia had approximately 3,400 telephones, ot which 92% were located in and around Banjul, the only truly urban area in the country. In other words, less than 300 telephones are distributed throughout the rest of the country, where 80% of the population live. This gives The Gambia 0.55 telephones per 100 people, compared with the average for Africa of 0.66 per 100. The current UN Transport and Commnu- nication Decade in Africa has a target of achieving one telephone per 100 people by 1987. Currently The Gambia is clearly below even the relatively low standards of Africa and seems unlikely to achieve the Decade target on schedule. The actual number of Direct Eschange Lines (DELs) is 1,920 (June 1981, Table 3.1), of which 1,670 are located in Banjul and the adjoining Serrakunda and Yundum area, leaving just 250 "rural" DELs for the rest of the country. The situation is further aggravated by the fact that most rural telephones are very unreliable and many of them are not functioning. 3.3 The telephone network is operated by Gamtel, an autonomous corporation formed in 1984 from the former government Department of Tele- communications (DOT). Gamtel has also recently taken over responsibility for international telephone, telex and cables, which up to 1984 had been handled by Cable and Wireless PLC, a UK-based company, via a Standard B satellite earth station located near Banjul and inaugurated in 1978. There is also an international microwave link with Dakar in Senegal which comes under the Pan-African Telecommunications (PANAFTEL) scheme. The general quality of international connections is reportedly good and the capacity appears adequate for present international traffic. 3.4 By contrast, the present quality of the domestic service pro- vided by Gamtel is generally poor. The peri-urban network suffers from almost life expired equipment and lack of adequate maintenance. -9/ 9/ Gambia Telecommunications Sector Memorandum, The World Bank, Trans- portation, Water and Telecommunications Dept., Western Africa Regional Office, Report No. 3603-GM, December 23, 1981. - 21 - There is such severe congestion in the trunk network that many local and trunk calls cannot be completed, especially at peak periods. The national long distance network is mostly manually operated, and " ...the quality of service provided by the operators is unsatisfactory and there are frequent complaints". 10/ This situation refers to 1981, and the impression gained by this mission is that, if anything, the situation is worse in 1984, particularly in the rural network. In fac:, the UHF trunk backbone connects to the peri-urban network in Banjul through a manual switchboard of obsolete design which is usually manned by only one or two people. Table 3.1: GAMBIAN TELEPHONE EXCHANGES Exchange Capacity Type DELs a/ Banjul 1,400 s/s auto 1,027 Serrekunda 800 " 587 Yundum 100 " 56 Brikama 50 " 36 Barra 50 RAX 15 Farafenni 50 " 38 Kaur 50 " 34 Kuntaur 50 " 20 Bansang 50 " 20 Basse 50 " 47 Total auto s/s 2,650 l,880 Kerewan 50 Manual 8 Mansa Konko 50 " 17 Georgetown 50 is 15 Total manual 150 40 Total all types: 2,800 1,920 a/ Number of lines as of June 1981. Source: Gamtel. 3.5 Further problems occur in the UHF trunk backbone and in the rural network. Many of these problems stem from Lack of reliable power supplies for exchanges, repeater stations and single-channel VHF termi- nals, and from maintaining and operating life-expired and inadequate equipment with limited numbers of staff, insufficient spare parts, and a 10/ Ibid. - 22 - shortage of transport. These problems are detailed in the following sec- tions, and means to overcome them are partially addressed by the recom- mendations made in this report. 3.6 The peri-urban network is expected to be rebuilt and overhauled with modern automatic equipment provided as French bilateral aid, so it is expected that the constraints on the rural network caused by the in- adequacies of the peri-urban network will soon be largely removed. How- ever, there are no plans at present to overhaul the rural network, this despite the fact that 802 of the population lives in the rural areas and the economy of The Gambia is largely based on produce from the rural sec- tor. The improvement of the peri-urban system will release some mainten- ance resources which could be used by the rural networks. 3.7 The economic importance of telecommunications is difficult to establish, but there is considerable evidence of a close correlation between invescment in telecommunications and economic growth. A project document of the ITU 11/ argues, inter alia, that investment in tele- communications has often been a consequence of economic development rather than a prerequisite and therefore economic development has been inhibited. It shows also that, although the costs of rural telecom- munications are relatively high and no doubt inhibit such investment, particularly where benefits tend to be intangible, the benefits revealed by recent analytical studies are exceptionally high, especially in remote rural areas having poor transport facilities. Examples of tangible bene- fits exceeding costs by many hundreds of percent have been found in case studies relating to development of rural economies in many developing countries (cited in the above document). Unfortunately, considerable on- the-spot investigation and analysis would be needed to quantify benefits from improving and enlarging The Gambia's rural telephone network. Therefore, although this report looks at the direct benefits to Camtel of implementing the recommendations which follow, and justifies them mainly on that basis, much greater intangible benefits to The Gambia can be expected as well. The Multi-Channel Trunk System 3.8 The first trunk system in The Gambia was installed in 1961/62 by General Electric Company, Ltd. This was a VHF five channel backbone connecting Banjul to Basse via Kerewan, Hansa Konko, Kaur, Kuntaur, Georgetown and Bansang. By 1978 only two channels in the system were still working. Project UNDP/ITU CAM/78/003 was the mair element of technical assistance for the rural telephone network and as provided with the services of an ITU appointed expert, who implemented a number of developments specifically relating to single-channel VHF spurs and links, 11/ Project Document of ITU-vECD Project, Telecommunications for Devel- opment, International Telecommunication Union (ITU), Geneva, 1983. - 23 - which are detailed later. He also assisted the DOT in.acceptance trials of the UHF trunk backbone. 3.9 The UK government financed the construction of the replacement trunk system, this being the 470MHz UHF trunk backbone currentlj used; this project was executed by the Crown Agents, starting in 1977, and was implemented by the UK firm, Pye. After a number of delays caused by technical problems, it was finally commissioned and accepted in June 1980. A UK engineer has been seconded to Gamtel by the UK Overseas Development Administration (UK-ODA) to provide technical assistance in maintaining the UHF system. 3.10 The UHF system is generally described as a 60-channel system, which is the maximum capacity of the equipment, but the system in The Gambia, in reality, is only equipped with 48 channels. These 48 channels are dedicated in blocks of 12 to different destinations. Beyond Georgetown, the trunk backbone continues as a nominal 24-channel UHF link to Bansang, and thereafter 16 channels carry through to Basse. This is illustrated in Map 15991R and in Table 3.2. It can be seen that there are also 8-channel UHF branches from Banjul to Barre, from Mansa Konko to Farafenni and from Georgetown to Kuntawr. 3.11 The UHF/VHF system generally has a range of 50 km to 60 km between stations. The new system follows the new tarred road on the south Bank of The qambia River, rather than going along the shorter but more difficult north Bank route via Kerewan as before. Two repeater stations were consequently located between Banjul and Mania Konko (at Jelekato and Jattaba) and suffer from being at almost the maximum range. The Banjul to Jelekato link also has potential reception problems due to interference from the signal passing low over a substantial stretch of water. Mansa Konko is the first exchange out of Banjul and takes 12 channels to serve itself and surrounding towns. The UHF trunk route then switches to the north bank and passes via another exchange at Kaur and a repeater at Makagui to Georgetown, provincial headquarters of McCarthy Island Division, the province from which most of The Gambia's commercial agricultural produce is derived. The eastern end of the country is served from Basse exchange, which is reached via the repeater and exchange at Bansang. The Bansang-Basse 24-channel UHF link is again at the limit of its range and suffers at times from interference caused by two hilltops, Nella Kunda and Bagadagi, very close to the line-of-sight for signal transmission. Provincial Telephone Exchanges on Trunk Network 9. 3.12 There are seven exchanges associated with the trunk network, plus a small manual exchange at Kerewan to serve the town's 8 DELs and one remaining (intermittently available) rural spur to Salikene. The seven exchanges are: Farafenni, Mansa Konko, Kaur, Kuntaur, Georgetown, Bansang, and Basse. The first five exchanges in Table 3.1 are in effect part of the peri-urban system. - 24 - 3.13 There is no national electricity grid in The Gambia, so all the provincial exchanges and repeater stations are dependent either on small GUC local mains electric supplies, on Gamtel diesel generating sets, or sometimes both. Since neither the mains supply nor the generating sets generally can give a 24-hour service, the telecoemwnications equipment runs from a bank of lead-acid batteries which are charged by the mains or the generator. Table 3.3 summarises the power supplies and battery capa- city installed at the exchanges. 3.14 Maintaining an adequate state of charge in the batteries is proving to be a problem for many of these exchanges due to both deteri- orating batteries and difficulties in providing regular and adequate recharging. This is discussed further in (para. 3.29). Single-Channel VHF Links to the Villages 3.15 There are 17 single channel VHF links and spurs from the pro- vincial town exchanges to villages and small towns. These are all 148- 174 MHz VHF duplex two-frequency systems which terminate in the villages at a public pay phone, usually in a brick-built kiosk fitted with a 10 to 15m guyed mast and antenna. The standard equipment is Plessey (UK) model PRD707 1W or 8W transceivers of about 1978 vintage, but a few Philips (Australia) 1W or 15W FM880s of about 1980 vintage are also in use. Standard Associated Automation 2-button pay phones (UK) are connected to the transceivers. 3.16 The equipment was originally intended to be powered from EMU type primary batteries. These proved expensive and unreliable under tro- pical conditions and were replaced by lead-acid secondary storage bat- teries recharged, in a few cases, by a local generating set, or removed for recharging elsewhere and replaced by a freshly charged unit at appro- priate intervals. Standard car batteries of 12V 3OAh rating are used for this purpose as they are readily available in The Gambia. Charging of car batteries in this way has proved to be problematic, since they do not readily survive periods of over-discharge and it has been impossible to arrange reliable enough recharging facilities. Therefore a solar photo- voltaic power system using 20W Solarex (UK) solar panels and German sealed lead-acid "Dryfit" batteries was devised for this purpose. - 25 - Table 3.2: UHF AND VHF RURAL RADIO TELEPHONE LINKS (1984) (excluding Banjul semi-urban network) In From To Channels Distance b/ use? a/ (used nominal) (km) (YIN) A. 470 MHz UHF Trunk System Banjul Jelekato 48/60 41 Y Jelekato Jattaba 48/60 50 Y Jattaba Mansa Konko 48/60 38 . Y Mansa Konko Kaur 32/60 35 Y Kaur Makagui 32/60 46 Y Makagui Georgetown 32/60 28 Y B. Multi-channel 470 MHz UHF Trunk and Spurs Banjul Barra 8 5 Y Mansea Konko Farafenni 8 14 Y Georgetown Kuntaur 8 19 Y Georgetown Bansang 24 16 Y Ransang Basse 16 53 Y C. Single-channel VHF Links and Spurs Banjul Gunjur 1 33 Y Banjul 8wiam 1 58 N Banjul Ndungu Kebbe 1 30 Y Banjul Kerewan 1 55 Y Kerewan Jowara 1 8 N Kerewan Manduar 1 16 N Kerewan Salikene 1 14 N Mansa Konko Illiasa 1 25 N Mansa Konko Kwinella 1 30 N Kaur Dankunku 1 16 N Kaur Kudang (GPMB) 1 25 N Kuntaur Kudang (village) 1 18 N Bansang Sami Karantaba 1 16 N Basse Gambisarra 1 12 N Basse Diabugu 1 30 N Basse Patoto 1 30 ? Basse Sutukoba 1 30 N Basse Yarobawal 1 12 N a/ Denotes whether link was functioning as of July, 1984. b/ Line-of-sight distance. - 26 - 3.17 Table 3.4 indicates the power sources and status of all the VHF single-channel links, including nine converted during 1980-81 to solar power. The solar powered systems were relatively successful, especially when first introduced, but due to errors in sizing their batteries and aligning their arrays, most of their batteries appear to have failed after a year or two of use brought on by excessively deep discharge cycles and inadequate recharge. As a result, most of the solar powered systems now can only function when the solar irradiation is strong enough to produce sufficient power to run the transmitter directly from the solar array. 3.18 The problems are further compounded by improper alignment of antennas, incorrect tilt of PV arrays, weak supporting guy wires and, in general, poor maintenance. There are also human problems such as lack of knowledge on the use of the telephone, interference by children, and the habit of leaving the hand-sets "off the hook", which leaves the trans- mitter switched on and rapidly drains the battery. Fortunately, vandal- ism has not been a problem. Another important prcblem relates to siting, namely, a sufficiently unobstructed "line of sight" from the exchange antenna to the terminal antenna. The terminal antennae are generally quite low (10-15m) and are therefore prone to obstruction and interfer- ence by trees or low hilltops. The mission recommends that the adequacy of the line-of-sight should be re-checked before renovating of existing single channel lines and, if necessary, there should be re-siting of systems to gain a more favorable transmission path. Yet another major problem has been occasional indiscriminate siting of the systems in underpopulated regions. 3.19 The problems are exacerbated for people wishing to make trunk calls from single channel VHF links the further up-country from Banjul they are due to the many other weak links in the chain. In any case, more than 90% of rural calls from pay-phones are operator-assisted, and even when the system is technically functional, problems have arisen due to poor performance of operators. Similarly, the poor performance of the system dissuades people from even attempting to use the telephones. If public confidence in the reliability of these systems could be re-estab- lished, it would have a marked effect on the number of people attempting to use the telephone. 3.20 Tables 3.2 and 3.4 indicate that, as a result of these prob- lems, only about three, or possibly four, out of the seventeen original single-channel VHF units are functioning at this time. In fact the reli- ability of these systems has been so poor, due mainly to power supply problems, that a. number have been removed from their locations (Table 3.4) and returned to storage in Banjul; for example, there is no longer any equipment at Jowara, Kwinella, Manduar or Yarobawal. The two main exceptions, at Gunjur and Fatoto, appear to have functioned relatively reliably and have yieltded revenue consistently. This is, perhaps, due to the fact that they have properly aligned antennae and better than average solar arrays. The system at Fatoto is also unique in having a paid attendant who cleans the solar array, teports faults and explains its use - 27 - to potential customers, and this certainly has a positive influence on its reliability. The system at Gunjur is sufficiently close to Banjul to give a good signal strength even under marginal operational conditions and to be accessible for maintenance by headquarters technical staff. Table 3.3: POWER SOURCS ANO BATTERIES AT PPOVINCIAL EXCHANGES AND THE REPEATER STATIONS (In approx. order eastwards from Banjul) Location Average 9atotry Power Source and Telecom Battery Charger Type Charge Outy (Type) Load Capacity Rating e/ Rate Cycle Remarks &/ (amp) (/ lno)x(Ah) (amp) (amp) (h/day) Jelekato SA 12x2OOAh 30A 2,4kVA D IIA 12 24V MUR) Jattaba SA 12x200Ah 30A 2x4kVA D IIA 12 24V (U,R) Farafenni IA 2xWMAh 30A GUC & 17A 12 48V Exchange (V,T,X) SkVA 0 24V Radlo Mans" itonko 10-llA 12x200Ah 40A GMC & 22A 12- 24V Exchange (U,V,X-R,X) SkVA D 24V Radio Kaur 10-llA 12x200Ah 50A GUC & 22A 12- 48V Exchange (U,V,-R,X) lx8kVA 0 24V Radio IxSkVA D Makagul SA 12x200Ah 30A 2x4kVA D IIA 12 (U1,R) Kuntaur BA 12x200Ah 30A 3x5kVA D 17A 12+ 48V Exchange (Y,D-R,X) 24V Radio Geowgetown BA I2x200Ah 40A GUC & 20A 12- 24V Exchange (U,V,O-R,X) 5kVA 0 24V Radio Bansang OA 12x200Ah 30A GUC only 17A 24 48V Exchange (V,D-R,X) 24V Radio Basso 12A 12x200Ah 30A GUC & t7A 12 48V Exchange (V D-R,X) SkVA D 24V Radio I/ Legend: UUJF VuVHF D-Rudrop repeater RaRepeater TaTerminal X4Exchange - -b Load given as me current (at 24V In all cases), c/ Capacity at lOh discharge rate given (where known), d/ HIarmer & Simmons SCI/24 transductor units In all cases. D .1 SiC a mains power (usually suppiled 12h per day); 0 a Gamtel standby diesel; the duty cycle given Is -intended" rather than "actual", Source: GSate . - 28 - Table 3.4: POWER SOURCES AND BATTERIES AT RURAL SINGLE CHANNEL VHF TERMINALS Working? Reason for Location Power Source Battery a/ Failure Gunjur 2OWp solar 6Ah 12V Yes Ndungu Kebbe 2OWp solar 6Ah 12V Yes Bwiam diesel mains 3OAh 12V No flat battery Jowara none No removed Salikene diesel mains / 3OAh 12V No flat battery Manduar none No removed Illiasa none 3OAh 12V No flat battery Kwinella none, No removed Dankunku 2OWp solar 6Ah 12V No faulty installation Kudang (CPMB) diesel gen. missing No no battery Kudang village 2OWp solar 6Ah 12V No faulty installation Sami Karantaba 2OWp solar 6Ah 12V No faulty installation Cambissara 2OWp solar 6Ah 12V No faulty installation Diabugu 2OWp solar 6Ah 12V No faulty installation Fatoto 2OWp solar 6Ah 12V ? faulty installation Sutukoba 2OWp solar 6Ah 12V No faulty installation Yarobawal 2OWp solar 6Ah 12V No removed a/ Note that some of the faulty solar powered systems shown as not working do occasionally function, while those marked as working do function most days, but usually without a high level of reliability. b/ Systems with no power source originally used EMU cells; when these proved too troublesome, no alternative was available so the system was removed. c/ This system had a car battery to replace an EMU cell; this is supposed to be regularly taken away to be charged, but lack of transport prevents this. Source: Gamtel. Structure of Telecommunications Operation 3.21 Until recently, Gamtel was a government department integrated with the postal service. It is now a commercial corporation with a board of directors and a mandate to seek profits. Camtel Finances 3.22 On the financial side, a major problem for the old DOT was that around 50X of telephone traffic had been through government departmental telephones and it had not been reimbursed for this traffic. Consequent- ly, the DOT operated at a loss. For example, telecommunications - 29 - operating revenue for 1978-9 was approximately D 587,000 and expenditure was D 710,012, resulting in a loss of about D 123,000. According to the Government estimates, the expected expenditure for PY83-84 has risen to D 1,316,400, and an associated loss of over D 350,000. Since Gamtel as an autonomous corporation will now be able to bill government depart- ments, it is expected that a large proportion of the potential revenue from government traffic will be collected and that Gamtel will achieve a far better financial return than the DOT did. In addition, it is ex- pected that the imminent rehabilitation of the peri-urban system, which contains most of the current telephone DELs, will generate extra traffic and additional revenues. 3.23 Revenue from the rural network is naturally small at present due to the limited number of lines in use and the poor condition of the system. Nevertheless, Table 3.5 indicates that ten public call boxes yielded D 12,285 per quarter on average for the five quarters from January 1983 through March 1984. The data shows that the call box at Basse appears to function reliably and has yielded nearly three times the average revenue per call box; this may give some indication as to the possible revenue to be obtained from call boxes if they develop a reli- able reputation. The mean cost per call is also estimated in Table 3.5 and can be compared with the tariff rates given in Table 3.6. The table shows that local calls are not timed and are at a flat rate, but trunk calls are timed and have a minimum charge for three minutes. The high average cost of calls originating from the Banjul call box may be due to a number of international calls to Senegal or because a large proportion of calls are from people with families in the. prov£nces. For similar reasons the average cost of calls originating in Basse is high, as any calls from there to Banjul will be at the highest rate. The average cost per call for the ten call boxes is D 1.73 and for the nine call boxes (excluding Banjul) is still D 1.43. Table 3.5: USE OF PUBLIC CALL BOXES c-apital & main centers) a/ Location No. of Calls Gross Revenue Mean Cost/Call (per qtr) (Dalasis) (Dalasis) Banjul 828 1952 2.36 Brikama 924 730 .79 Kerewan 197 120 .61 Farafenni 567 984 1.73 Mansa Konko 364 610 1.68 Kaur 853 1038 1.22 Kuntaur 525 823 1.57 Georgetown 853 1478 1.73 Bansang 566 935 1.65 Basse 1860 3615 1.94 a/ Average figures for 1983 plus 1st quarter 1984. Source: Camtel. - 30 - Table 3.6t SUMMARY OF TELEPHONE CHARGES (1984) Type of Call Cost (Dalasis) Local call (same exchange or auto-dial area) private subscriber: 0.20 public call box : 0.25 Trunk calls per minute (minimum charge 3 minutes) same for both private subscriber and public call boxes: From Banjul up to Kerewan or Bwiam 0.25 From Banjul to Farafenni, Mansa Konko & Dankunku 0.50 From Banjul to Kuntaur, Georgetown and beyond 0.75 Source: Camtel. Maintenance Problems 3.24 The UWF trunk back-bone and the single-channel VHF links come under Gamtel's Department for Transmission of Radio and Telex, which has a limited number of centrally-based technical maintenance staff (three or four plus two vehicles). Any faults other than the most commonplace and trivial need to be investigated and rectified by centrally-based staff. Therefore, if a number of simultaneous faults develop in different parts of the rural network, it is a time-consuming operation to send out a team headed by a technically proficient person to correct the faults. Ironi- cally, the team is often incommunicado for many hours due to the very faults it is trying to correct. It often cannot be notified from head- quarters of further problems that may have arisen in its area, simply because there are no two way radio-telephones in team vehicles. In fact, the vehicles are ill-equipped and often major systems have to be taken back to Banjul for repair because not enough mobile facilities for test- ing are available. Similarly, there are no spare batteries, so that flat batteries have to be taken away to be recharged and returned later at great expense in travel costs and unnecessary extra down-time. Therefore a modest investment in spare batteries, mobile radio-telephones and mobile test-equipment could radically improve the effectiveness of the maintenance team. This need appears to be well known within Gamtel, but the resources to procure the necessary items do not exist at present. 3.25 Some proposals to remedy this problem are discussed later in this chapter. The general philosophy for the future specification of equipment should be to pay a premium, where necessary, on new equipment to minimize failure modes. For example, the provision of better battery charge regulators might reduce the incidence of battery problems and limit the transmission of excessive voltage spikes from the generating - 31 - sets, while the provision of better shaded and ventilated enclosures for the VHF equipment would no doubt reduce the incidence of electronic com- ponent failures. A poor country like The Gambia can ill-afford the use of cheap equipment. Systems procured for The Gambia need to be as failure-free as possible in order to allow technical staff to success- fully complete routine maintenance; otherwise considerable time will be spent rushing from one breakdown to the next, as appears to be the case at present. Power Supply Problems in the Telecommunications Sector 3.26 It is clear that power supply problems, including fuel short- ages, undersized and damaged batteries and inadequate megns to charge them, represent perhaps the biggest problem area for the rural telephone system. Power problems directly or indirectly cause significant but unquantifiable system down time, as illustrated in the following paragraphs. UHF Repeater Stations 3.27 One particular problem arises at the three UHF repeater sta- tions at Jelekato, Jattaba, and Makagui, which, being in remote areas, do not have a CUC supply but are provided with a pair of Gamtel diesel gen- erating sets. These stations have only one semi-skilled attendant and security guard who operates and maintains the generating plant and reports faults. In recent months, power supply problems with these repeaters have caused a number of system failures and a great deal of travel for engineers and technicians from Banjul to make repairs; fuel shortages have seriously aggravated the problem. The weakest points in the UHF links are recognized to be the UHF repeaters and their unreliable power supplies. 3.28 An indirect problem associated with the power supplies relates to voltage regulation. Even very brief "voltage excursions" can cause damage to batteries and to charging equipment, and can possibly be trans- mitted through the battery charging system and affect telecommunications equipment. Transient excess voltages can occur because the electrical load at the repeater stations is very small in relation to the generating plant capacity, and the diesel system is grossly derated. 12/ The gross derating also no doubt adversely affects both diesel engine efficiency and its maintenance, since continuous running of diesels at part load can 12/ The repeater stations use small Lister single cylinder air-cooled diesel single-phase generating sets rated at 3.8 kVA (3.4 kW). The normal (or intended) charging rate, however, is only 11A at 24V, or 264W. This requires about 35OW from the generator, which is only about IOZ of its rated output. - 32 - cause carbonization and hence require more frequent maintenance. Unfor- tunately, there is no solution to this problem while using diesel power, as the engines being used are about the smallest size available that have adequate durability and reliability. This is why the application in question is particularly well suited to conversion to solar photovoltaic power. Provincial Terminal Exchanges 3.29 The battery storage is generally adequate for a situation where a regular and reliable 12h/day charging current is available (except at Farafenni where two car batteries with inadequate capacities are in use). However, as explained in Chapter II, the GUC supplies at present tend to be 12h/day at best. In some cases there have been many weeks of GUC supply down time. This is indicated in Table 3.7 where, for example Gamtel paid only the standing charge of D 10.00 for Basse during the 1st quarter of 1984 because zero kilowatt hours were metered. Only the standing charge was paid because of the many months of down-time at the Basse GUC plant. Unlike DOH, Gamtel has had to use its own standby plant with diesel engines similar to those at UHF repeater stations as the prime power source. Thus, the Gamtel plants suffer from exactly the same problems as the repeater stations. 3.30 Unfortunately, a combination of excessively deep discharge cycles and excessive voltage excursions has led to evidence of serious deterioration of many batteries. Once a few cells in a series string begin to fail, the remaining cells receive an excessive charging voltage, which in turn shortens their lives. Lack of spare batteries has led to such desperate measures as the insertion of odd smaller 2V cells in a series of larger cells (witnessed by the mission at Basse) to keep the voltage per cell at an acceptable level; despite this "bodge", many of the cells are gassing badly and their electrodes are visibly breaking up. There is therefore a vicious circle of decreasing battery capacity due to deteriorating batteries and lengthy gaps between recharging due to a deteriorating electricity supply service. Even Gamtel's own standby plant, which appears to be relatively well maintained and functional and which, therefore, is being increasingly depended on, is threatened by nationwide fuel shortages. Single-Channel Aural VHF Links 3.31 Transport difficulties, due mainly to fuel shortages, have made the regular replacement of discharged lead-acid batteries with recharged ones impracticable, and these installations have generally been removed. Besides, even if transport were not a deterrent, the high cost of trans- porting batteries would render this option very questionable. 3.32 The solar power systems, for the VHF links use sealed lead acid "Dryfit" batteries of only 5.7 Ah depending on assumptions on the duty cycle (transmit time), are significantly below the necessary capacity for reliable operation. Actually, about 50 Ah would have been more suitable. - 33 - This undersizing of batteries has made the operation of solar powered VHF systems nearly impossible. Table 3.7: ELECTRICAL POWER PROCURED ON A METERED COST ASIS FROM GUC BY GAMTEL FOR EXCHANGES TAKING MAINS ELECTRICAL POWER Cost in Dalasis per Quarter Exchange 2nd 83 3rd 83 4th 83 1st 84 2nd 84 Mansa Konko 835 502 751 711 691 Georgetown T n/a n/a 1,523 1,439 1,981 Bansang VHF stn 1,314 1,573 1,797 1,943 1,874 Basse P & T 2,583 1,428 857 10 514 Note: Low and variable figures indicate breakdowns probably occurred and reduced power that could be consumed. D 10 is the quarterly standing charge, therefore no power was procured in Basse in 1st quarter 84. Source: GUC Accounts Department. 3.33 The single channel VHF units proved that they can work under favorable circumstances, but a combin4tion of errors in their implementa- tion marred the systems sufficiently to render the majority unserviceable by mid-1984. As mentioned earlier, two or three continue to function; nevertheless, it is probable that they, too, are prone to develop faults and that they could be improved by introducing appropriate corrective measures. Recommendations for the Telecommunications Sector 3.34 The objective of these recommendations is to eliminate or re- duce some of the aforementioned constraints to providing an adequate rural public telephone service in The Gambia. A number of further recom- mendations follow, aimed at further reducing some of the factors known to inhibit the operation of the rural telephone service. UHF Repeater Stations 3.35 It is recommended that the existing diesel generating sets, battery chargers and battery banks be completely replaced at all three UHF repeater stations (Jelakato, Jattaba and Makagui) with solar photo- voltaic powered systems. All three repeater stations are identically equipped and can, therefore, be identically re-equipped. It is assumed that this electricity demand is to be met by providing an appropriate new bank of lead acid batteries, plus, of course, a photovoltaic array and battery charge regulator. - 34 - 3.36 Results of Analysis and Justification. The costs and benefits to be expected from the implementation of ths recommendation are de- tailed in Annex 6 and summarized in Table 3.8. The financial benefits to be gained are primarily savings in both diesel fuel and maintenance of diesel generating plant, combined with a reduction in down-time gained from the use of an inherently more reliabLe power system. Additional benefits are: first, that three, and eventually all six of the diesel generators, plus the three Harmer and Simmons 30A battery chargers can be redeployed to telephone exchanges to provide extra stand-by capacity; and second, the space created in zhe engine house can be converted intu extra workshop space for overhauling telecommunications equipment. No finan- cial allowance for these latter benefits has made in the analysis, and consequently the benefits shown are conservative. Table 3.8s COST COMPARISON OF SOLAR AND DIESEL AT REPEATER STATIONS (Us$) Solar Low Solar High Diesel Installed Cost 18,238 27,220 6,463 Annual Recurrent Cost (O & M) 100 200 3,938 Annualized Life-Cycle Cost 2,617 3,869 4,914 Unit Cost/kWh 1.68 2.49 3.17 Cost as X of Diesel 53% 781 1002 3.37 The substitution of solar photovoltaic power systems will result in real cost savings, and the cost of power ($/kWh) will decrease by about 22-47%. The high cost of diesel-generated electricity is attri- butable to the fact that the .diesel engines are considerably derated and operate at about 10% of their rated capacity. 3.38 As far as Camtel is concerned, the financial benefits will be much greater if the solar powered systems are provided in the form of bilateral aid, as the recurrent costs of the solar system are negligible in relation to those of the existing diesel plant. Conversion of the repeater stations to solar power will reduce the annual running costs of each one by A minimum of US$3,838 ($11,514 for all three). Other bene- fits to be derived from the conversion to solar power include: (a) mini- mal maintenance requirements, which would release skilled technicians for maintenance of other weak links in the rural system; and (b) increased telephone revenue from reduced UHF system down-time. The increased reve- nues to be derived from improving system reliability could be signifi- cantly higher than might be assumed simply by proportioning it to down- time. This is because many potential users are no doubt deterred from using the telephone by poor expectations of obtaining a connection reasonably quickly. - 35 - Provincial Exchanges 3.39 Since the provincial exchanges are an essential link between the proposed rural VHP single channel terminals, it is important to en- sure that these are able to function properly. Therefore, it is recommended that all the batteries associated with the trunk network and the VHF single channel feeders at provincial exchanges be replaced and provided with new battery- chargers with a higher rating. It is, in any ease, essential to replace the existing batteries because most of them are near or beyond the end of their useful lives and there would be sig- nificant system down-time resulting from over-discharge of the batteries if no early replacements are provided. However, it is recommended that the existing worn out batteries not simply be replaced with batteries of a similar sized bank, but that the replacement batteries also offer a larger storage capacity. This would reduce the risks of over discharge in the future arsd allow more rapid charging during the generally shorter periods for which power is available from the GUC or, where stand-by generators are used, to allow charging at a power rate more compatible with the size of the generating set. 3.40 Results of Analysis and Justification. The technical and cost assumptions used for the analysis, and the results obtained are detailed in Annex S. A summary is provided in Table 3.9. The technical specifi- cations related to the choice of batteries are also discussed in Annex 5. The replacements will be necessary at seven telephone exchanges, namely, ?arafenni, Mansa Ronko, Kaur, Kuntaur, Georgetown, Bansang and Basse. The total cost to upgrade the battery storage as recommended will be about US$40,000. The main cost savings will be in reduced use of standby generators, as follows: cost of mains power (Table 3.6) average D 0.50 = US$0.13/kWh; and cost of 3.4 kV diesel per hour (Annex 5) US$1.10/h. It is clear that the present inadequately sized system is marginally less expensive in straight life-cycle cost terms than the pro- posed replacement, despite having an assumed life of 5.5 years instead of 13.7 years (based on 1 cycle/day). Table 3.9: COST SAVINGS FROM USING NEW BATTERIES Existing New Battery Oattery Difference (US$) (US$) (US$) Annual cost with 100% mains 351 351 Hours needed for 25% engine use 1,095h 703h Annual cost with 25% engine use 1,468 1,037 431 Hours needed for 50% engine use 2,190h 1,406h Annual cost with 50% engine use 2,584 1,722 862 Hours needed for 100% engine use 4,380h 2,810h Annual cost with 100Z engine use 4,818 3,091 1,727 - 36 - 3.41 The benefits to be gained by replacing the existing batteries with a larger bank of batteries are: first, greater certainty and pre- dictability of the actual life of the new system; second, reduced battery maintenance and risk of down-time caused by battery problems. This latter benefit is the primary justification for the recommendation, although there is a distinct possibility that implementation will also result in direct financial savings. The reason for this is the signif- icant, but not easily predictable, benefit co be gained through reducing the period of each battery recharge and through extending the allowable gap between recharging. This should significantly reduce the need to use standby generators. However, given the largely unpredictable mains supplies, the need to use the standby generators will continue. 3.42 The cost implications of this are indicatad at the bottom of Table 3.9, where the electricity costs with the existing system are about US$351/yr. with 100% mains use and increase to around US$4,818/yr. with 100% diesel generating set use. Increasing the size of the battery bank as proposed will have two effects: first, it will reduce the probability of having to use diesel generating sets, since only an average of 7.7h/day of mains will be needed, and the gap between recharges will be extended to perhaps 48h (compared with 12h/day and 22h at present); second, if lengthy periods of mains down-time occur, then the generating set only needs to be used for perhaps 7.7h in 24 instead of for 12h in 24. The financial benefit of this can range from zero in the unlikely case of 100% mains availability (for both options) to as much as US$1,727/yr. in the event that the system is run continuously on standby generators. Also, the actual efficiency of the standby generators is likely to increase, and their lives will be extended, through running them at a greater percentage of their rated power for fewer hours per year; this will actually reduce generator running costs, although no allowance for this has been made in the analysis. Hence, the probability is that the larger battery storage will allow engine running cost savings to compensate for its extra annualised life-cycle cost once the need for using the standby generators exceeds approximately 15% of charging time. 3.43 The proposed larger batteries will have significant advantages for Camtel in financial terms if the investment is provided as grant-aid, in that the recurrent costs and future battery replacement costs will be reduced. Assuming that the engines are used 50% of the time, the annual savings for seven provincial stations would equal about US$6,000. Other benefits which are difficult to quantify but which nevertheless could be significant are reduced system down-time, reduced demand on manpower and resources for maintenance, and perhaps increased telephone traffic. 3.44 The budgetary requirement for replacing the existing batteries with a larger bank, as recommended, of 800Ah capacity, will be US$5,600 per exchange including new battery chargers, or US$39,200 for all seven rural exchanges. The existing batteries are, in any case, urgently in need of replacement, and this could be done at a minimum cost of US$1,450 per exchange (if the old chargers are not replaced) or US$10,150 for all seven exchanges. However, the minimum recommended replacement should - 37 - actually be twice the present capacity and would thus cost about US$29,050 (without new chargers), and the marginal cost of the recommen- dation would then be US$29,050. Single-Channel Rural VHF Links 3.45 It is recommended that a total of 20 single channel VHF links terminating with public pay-phones (and one commercial private line to GPMB), be established or, in some cases, rehabilitated; Map 15991R and Table 3.10 illustrate their distribution. These are all to be solar powered, using a fully integrated and purpose-designed solar power system and purpose-designed ventilated enclosures to house the electronic systems. 3.46 Location of Telephones at Health Centers. It is recommended that the new VHF single channel terminals be located at or as near as possible to health centers and dispensaries. In fact, the choice of locations is intended to ensure that every location having a government health center or dispensary will also acquire a public telephone con- necting to the rest of the country. There are three main reasons for locating public telephones at or near health centers and dispensaries: (a) these centers are generally accessible to a sizeable catchment of people, and where hundreds of people already gather at clinics several times per week, the locations should be accessible to enough people to generate useful telephone traffic; (b) as explained in Chapter I, the provision of telecommunications for health centers and dispensaries will have major health-related benefits, and DOH staff accords a high priority to it; and (c) the DOH officer in charge could be given responsibility to assign one of his staff to act as caretaker of the public telephone. This appears to be a feasible proposition acceptable both to DOH and Gamtel, and would solve a number of the problems outlined below. 3.47 The mission believes that many of the operational problems experienced with the first generation of rural public telephones can be overcome when an attendant is provided to clean the solar array, and to assist people unfamiliar with the telephone in using it for the first time, and, of course, to report faults. The evidence for this is the successful installation at Fatoto which is unique in being about the only working rural single-channel link up-country, and also unique in having a paid caretaker. Gamtel cannot afford to employ caretakers for remote public telephones, but discussions held by the mission with senior Gamtel and DOH staff indicate that there are advantages in arranging for the health centers and dispensaries to be officially authorised by Gamtel to be responsible for the telephone. DOH would have to be compensated for this service. One possible option would be to allow the respective health centers to make a pre-specified number of calls each month for free, or alternatively receive every month a percentage of the proceeds from the coin boxes. 3.48 Rationale for Selecting the Locations. The first thirteen proposed VHF single-channel links listed in Table 3.10 consist of - 38 - existing or formerly existing VHF links, eleven of which already ter- minate at or near a health center or dispensary. They are marked on the table with "Tal", having been originally sited by former DOT and, there- fore, obviously selected to suit telephone department needs and, coin- cidentally, suiting DOH needs, too. The otherst at Illiasa and Kwinella (marked "T"), do not have a nearby health center and, while the location selected was to ser-'e DOT rather than DOH interests, they cater to suffi- ciently large catchments to warrant the telephone which they already have - except that it rarely works. Additionally, seven new single channel VHP links are proposed for all the locations having health centers or dispensaries that do not currently have a telephone nearby. 3.49 The telephone at the GPMB warehouse at Kudang also needs immediate rehabilitation because of important economic implications. Kundang collects agricultural produce for the whole region, but it is not part of the public telephone network; therefore, it is recommended that Camtel be provided with an extra solar power unit for this telephone installation, using the existing transceiver and telephone equipment since the operation of the CPHB diesel generator is not reliable enough to warrant its use. 3.50 Table 3.10 also indicates that three former links are not recommended for renews.1 because they do not serve any DOW interests and their siting is somewhat questionable. These are at Jowara (because it is quite near Kerewan and the reason for having a telephone there was not clear), at Manduar (because a better site is a few kilometers away at the large, newly built health center at Kiang Karantaba), and at Sutukoba where the original installation was badly sited and where a new unit. at the nearby dispensary at Baja Kunda can replace it. 3.51 Eight further centers that could benefit from a later phase or from an optional extension to the present project are also listed as "second priority" locations for single channel links. These eight cen- ters could be incorporated in the program as a result of rehabilitating and redeploying some of the existing failed solar powered systems. In this way sufficient equipment would become available to equip these eight additional locations immediately at marginal extra cost. The reason for selecting these extra eight was that se-ten of them, marked "T" were in the old DOT plan for extending the rural telephone network and should therefore be pursued, provided that Gamtel can demonstrate that the original reasoning for selecting them is still valid. The eighth is the one health center (at Essau) not given a telephone as a "first priority" on the grounds that Essau is very near to Barra, which has a telephone exchange and is just a ferry ride from Banjul; however, for completeness it seems proper that Essau should also eventually have a more convenient telephone at or near the health center too. - 39 - Table 3.10: SUMMARY OF RECOMMENDATIONS FOR LOCATIONS OF SOLAR POWERED SINGLE CHANNEL VHF TERMINALS exchange est. recrmondatton locatlon liak dist. proposer (acton) (km) T4Gamt.l (HC^Health Cr.) A. Existing or Former Stngle Channel Links (1) replace/rehabilitate 13 pro-exIsttng systems Gunjur Banjul 35 + T&H renew & resote at NC Ndungu Kobbe Bmnjul 30 + TUH renew & rolte at HC fviwm Banjul 58 + . TUH renew Salikene Kerewan 14 TaH renew & resite at HC IllIla Mansa K. 25 7 ronev 4 review s1te Kwlnella Manse K, 30 T renew & revlew site Dankunku Kaur 16 TiH renew & resite at NC Kudong village Kuntaur is TAH renew & rosute at MC Saml Karantabo Bonsang 16 TAH renew S resite at NC Gambissara Bsse 12 Taml renew l resite at HC Oiabugu Basse 30 + TAH renw resite at NC fatoto Basso 35 + T&H renew Yarobwal BEsso 12 TUH renew at HC (ii) rehabilittate single GPMB link gower system Kudang (GPMB) Kour 25 GPMB Gamtel to renew battery U provide solar power source (iii) scrap following pro-existing systems Jowora Kerewan a T do not renew (close enough to Kerewan) Manduar Kerewan 16 T do not renew (replace by new system at K'taba) Sutukoba Basse 30 T do not renew (replace by new syst. at sajakunda) S. Seven New Slngle Channel Links (rationale: to serve HCs and dispensaries currently not near to a telephone) Medinoa Beto Banjul 21 H site at dispensary Karantaba sanjul 50 + H&T site at HC (replaces Nanduar) Stbanor eanjul sO + H&T site at misston Kuntair BanJul 40 + H stte at dtspensary Chamen Banjul 20 H site at dtspensary Brikaumba G¢town 16 H site at dispensary - Baja Kunda Basso 25 + H site at dtspensary C. Eight Second Priorlty Single Channel Ltnks (ratlonale: part of old OaT plan and/or to serve HCsfdtspensertes not covered by A or 8 beause they have access to a nearby phone) Feraba Santa Banjul 21 T 0OT plan proposal Kafuta Bsanjul 30 + T DOT plan proposal Jutfure Banjul 25 T Jutfure and Albreda are or Albreda too close together to Justify separate lInks Essau BanJul 8 H site at HC (not far Barra, hence Cat.C) Manduar Kerewan 16 T defunct lInk (nr Koran- taba which Is Cat. B) Pakall Ba N'Konko . 32 + T DOT plan proposal Fulabontang G(town 6 T DOT plan proposal Bakadagy Bansang 30 + T DOT plan proposal a/ Note that llnks marked "+" need more powerful 5-lOW transcelver and a matching solar power systeom - 40 - Table 3.11: COST ANALYSIS OF SINGLE CHANNEL VHF RURAL LINKS Item Cost (US$) A. Total Procurement 11 High power (5-10W) systems 102,000 9 low power (1W) systems 75,000 Solar power at GPMB (kaur)? 1 W system 1,000 Spare batteries (for rehabiliation) 1,000 Cross-country vehicle 10,000 Two-way mobile radio-plus test equipment for vehic e 2,000 Total 191,000 B. Analysis per installation 21 Installations 29 Istallations Investment per installation 9,095 6,724 Annualized investment cost (10%, 15 yrs) 1,087 803 Annual recurrent costs 200 200 Annualized total cost 1,287 1,003 C. Analysis per call Mean cost per call 0.45 0.45 Average daily traffic per link to breakeven with mean duration of 7.83 calls 6.11 calls 3.5 minutes per call 27 minutes 21 minutes 3.52 Finally, it must be emphasized that these suggested locations must be treated as tentative since a proper survey (by Gamtel) of the terrain profile is vital before finalising a decision on locating any line-of-sight VHF equipment. In addition, it is important that the original justification for the sites should be reviewed and if on recon- sideration these sites do not seem viable enough to attract sufficient traffic, alternative sites be selected. However, the number and sizing of the solar PV system is expected to remain unchanged. 3.53 Results of Analysis and Justification. Analyses of the load, necessary solar power system size and cost data, together with an econo- mic analysis, are presented in Annex 6 and summarized in table 3.11. The complete installed cost of. each link, including the exchange equipment, the complete terminal and its solar power system, if "single-sourced", will be in the region of US$8,300 to 9,300. This estimate is based upon information obtained from potential suppliers and is felt to be conser- vative. If several specialized suppliers were involved rather than a single supplier, there might be lower costs, but this could result in higher risk and, as mentioned earlier, should be avoided. The total - 41 - budget installed cost for all the equipment necessary to implement the recommendations is US$191,000. If only 21 working links are installed then the average annualised cost per installation using a 10% discount rate over 15 years will be US$1,287, while if 29 installations are achieved (by rehabilitation of old equipment as well), then the average annualised cost will be US$1,003. 3.54 It is not feasible to predict the benefits in terms of actual revenue, but on the basis of experience with other call boxes, the traf- fic requirements to break even may be predicted. Table 3.5 indicates that a typical revenue yield per call from public call boxes is D 1.73; (if the -average call in the rural areas is at the intermediate D 0.50/ minute rate - see Table 3.6 - this would imply a mean call duration of approximately 3.5 minutes). Consequently, an average rate of about 8 calls/day would be needed to yield an annual break-even revenue for the 21 installation objective or about 6 calls if 28 installations are realized. In other words, if the assumption of 3.5 minutes per call is correct, it is necessary to achieve a traffic duration totalling about 27 minutes per day in the first case and 21 minutes in the second. Any traffic exceeding these requirements would bring in a profit. While it is difficult to predict with any certainty the traffic that may be gen- erated if the rural telephone system is perceived by most people as being likely to work most of the time, it seems reasonable to expect this modest usage necessary for break-even to be substantially exceeded. As shown in Table 3.12, even a modest 54-108 minutes/day of traffic can generate revenues of US$64,260-$128,520 a year. Moreover, as explained earlier, there are also significant intangible benefits for the rural areas and health centers and dispensaries. Table 3.12: ESTIMATED REVENUE FROM VHF LINKS No. Calls Total Average Average Daily No. Calls per Terminal Revenue Utilization per Day per Year a/ from 29 Links (minutes) (US$) (US$) 27 7.1 1,108 32,130 54 14.2 2,216 64,260 81 21.3 3,324 96,390 108 28.4 4,432 128,520 135 35.5 5,540 160,650 162 42.6 6,648 192,780 a/ Assumes 5% incentive payments to DOH for supervision of telephones. Source: Mission estimates. - 42 - IV. SOLAR WATER PUMPING Water Pumping Needs in the Health Sector 4.1 The government health centers and dispensaries have four pos- sible types of water supply: piped, borehole, open well, and, in one case, directly from the river. The safest of these is the piped supply, provided in most cases by the GUC, which meters the volume delivered and levies an appropriate water rate. Piped supplies usually originate from a borehole producing water which is generally unpolluted and clear. How- ever, it is sometimes beset with a problem in that the water is generally lifted by an electric pump to a storage tank, after which it is distri- buted by gravity flow through pipes; consequently, long-term power fail- ures can make it difficult to refill the main storage tank and render the water source inaccessible. On the other hand, intermittent power supply marked ',ith short to medium term power failures are of little consequence so long as the tank gets filled. Since the water supply situation with piped systems is not absolutely critical, the mission does not propose any changes to these nine health centers. 4.2 Four health centers and one dispensary are equipped with a borehole supply. This is similar to a piped supply, except that it generally has a single output rather than a piped distribution system. The borehole supplies at health centers usually also have electric pumps energised from small diesel generating sets; in these cases, so long as reasonable diesel fuel supplies can be guaranteed, water supply problems are not likely to be critical. The only dispensary with a borehole (Brufut) has no power, and a hand-pump is used to lift the water in- stead. As this is a new dispensary, it is too early to judge how effec- tive a handpump is in this context. Again, since the water supply problems are not absolutely critical there, the mission does not propose any water supply changes to health centers and the dispensary equipped with a borehole supply. 4.3 The remaining five health centers and the dispensaries all have sub-standard water supply from open wells which are subject to contamination and are a major source of disease both for DOS staff and patients. Excessive demand is also made of their water resources. In addition to supplying the usual needs of the staff, hundreds of patients and other local people frequently use the wells because others have been drying up in recent years due to the falling water table in The Gambia. Whereas the staff water needs would typically range from 100-1,000 litres/day, the additional demand often requires water supplies in the range of 2,000-10,000 litres/day. Moreover, the usage of wells by additional people multiplies the risk of contamination. 4.4 The mission recommends that initially, attention be focused on improving the water suppLy at the fifteen health centers and dispensaries with open wells. Except for the coastal locations, (where windpumps - 43 - could be more viable), solar water pumps could be expected to adequately meet their pumping needs at a cost ($/unit hydraulic energy) lower than would be incurred with diesel systems and often with even handpumps. However, owing to the variety of different water resources, varying water needs and pumping heads at these wells, it is not possible to recotpmend a standard "packaged" system which could be installed in each location. A consultant study is needed which would investigate each site in detail, and develop specific recommendations. The study is proposed as a part of a larger project which would also explore the use of solar water pumps for irrigation in the McCarthy Island Division. It is to be noted, however, that although windpumps on the coastal sites and solar pumps for the remainder of the country are expected to be the most viable options for mechanical water pumping, the only benefits to be derived from their use is better quality water, less contamination and improved public health. No direct energy savings would result. (Annex 7 describes solar water pumps in more detail.) Relationship with the Activities of the Department of Water Resources 4.5 About 10,000 hand dug wells serve 80X of the population in The Gambia. A major UNDP-sponsored program which is being implemented by the Department of Water Resources is seeking to improve water quality by con- structing concrete lined wells with adequate above ground protection. Many of these wells would probably not be covered because rope and bucket would probably remain the only water lifting device available. The chances of contamination, though reduced, would therefore still be high and the productivity of the wells would remain low. In the interests of cleaner water and higher productivity a number of handpumps are therefore being installed through support from some bilateral donors; also two experimental solar and two windpumps have been installed in recent months (para. 4.12). The consultant study of a water supply improvement program for the health centers and dispensaries should take into consideration these parallel activities of the Department of Water Resources as well as the data obtained from the experimental solar, wind and handpumps. Irrigation Pumping and the Use of Solar Pumps in McCarthy Island Division 4.6 Irrigation is not widely practised in The Gambia, even though it could conceivably have a significant impact on agricultural production owing to the lengthy dry season and often unreliable rains. The limited irrigation in use is at both ends of the scale of irrigation techniques in terms of cost and volume of water delivered, area of fields covered, etc. A few large commercial agricultural enterprises use standard com- mercial irrigation equipment with diesel pumps, and villagers also are known to use a small amount of well water for their vegetable gardens. 4.7 In general, the cost of delivering irrigation water is closely related to the vertical pumping head, because the energy requirements for a given output are directly related to head. Therefore, it is generally uneconomic to irrigate using well water lifted 10 m or more (as is neces- sary in The Gambia) unless there are no other water sources and energy is - 44 - cheap. Since The Gambia is a riverine country, the most useful source of irrigation water is The Gambia River, but only up-country, as the lower reaches (for a great distance inland) are partially tidal and fresh is intermixed with sea water, 4.8 The main area for agricultural production is Georgetown in McCarthy Island Division, the central province of the country. Hence Gambia Produce Marketing Board (GPMB) has its main warehouse and collec- tion point at Kaur on The Gambia River. The terms of reference for the mission included a to brief review of the feasibility of using solar pumps for small-holder irrigation in the McCarthy Island Di-vision. How- ever, in terms of its suitability for solar water pumping, this is a very different pumping requirement, both technically and in financial/economic terms, from pumping small amounts of well water for health center or vil- lage use. The volumes of water required are often tenfold (typically 50 to l00 m3/ha per day), and the pumping heads need to be much lower (maximum of 4m) if a solar pump is to be economically feasible. In addi- tion, the volume for irrigation purposes is not constant, but seasonal, and highly variable even in the dry season. Since the water source is almost certain to be surface water, it requires a different pump from the ones used for village water supply. The economic value of the delivered water, which determines the maximum acceptable pumping cost, will be a fraction of what is acceptable for drinking water suppiies. Irrigation water typically needs to be in the region of US$0.05/m , while drinking water can often cost much more and still be economically justified in extreme cases where no other water source is available nearby. 4.9 While there is a sound economic rationale for using solar pumps in village water supply, the case for their use in irrigation is much more marginal. Much depends on the crops to be grown, the efficiency of the farmers and the market value of the crops. It is, however, worth mentioning that solar pumps are at their most competitive with the chief alternative (namely i.c. engine pumps) in small scale applications such as irrigating small-holdings through a low head, where only a small power requirement (typically 100-200 W) is necessary. 4.10 The mission recommends that the most effective way to assess the role of solar water pumping in irrigation in the McCarthy Island Division is to commission a detailed consultant study. The study could be undertaken in tandem with that which investigates solar water pumping in the health sector. Experience with Solar Pumps in The Gambia 4.11 Some experience with solar pumps in The Gambia has been obtained through the Tanji and Sarakunda projects. A village solar pump - 45 - has also been installed at the village of Kaiaf (in May 1984). 13/ The system at Kaiaf was supplied by the Solarex Corporation (USAT using Solarex solar panels and a Grundfos (Denmark) m tor-pump unit. Field testing h s shown that itt c testin/ h e day i typo ca.ly delivers 20 D /day in an irradiation of IdMJ/m per day, up to 45 m/day in 23MJ/m. per day - these being typical "low" and "high" daily irradiation levels for The Gambia on a hazy/cloudy day and on a clear day, respectively. This particular system has a nominal power rating of 1.2 kW(p) and a delivered and installed (turnkey) cost of US$25,000. It is probable, therefore, that this system is sitnificantly larger than would generally be appropriate for rural health tenter and dispensary wells; for example, it may be expected that systems about one third of this rating (400Wp) would deliver at least 5 to 10 m3/day and would cost (installed) in the region of US$8,500 each. 4.12 One other solar pumping system has been installed in The Gambia at Jambanjali. This was financed by Saudi Arabian bilateral aid and installed under contract by the German bilateral agency GTZ. It is a German system with a Pleuger motor-pump unit and AEG-Telefunken solar panels, having a peak rating also of approximately 1.2 kW(p). This solar pump has been installed as one of the Saudi-assisted water supply proj- ects in West Africa. Being a demonstration unit in The Gambia, the solar pump has -been deliberately oversized and is consequently somewhat more expensive than the one at Kaiaf. Future units are expected to be cheaper. No performance data on the GTZ pump was available. Comparison of Solar Pumps with Wind and Hand Pumps 4.13 In areas without indigenous power supplies such as The Gambia, there are three potentially viable options that might be used for lifting water from closed 'wells or boreholes. They are, in order of increasing capital cost: handpumps, windpumps, and solar pumps. These are dis- cussed below. (a) Handpumps may in many cases be the most cost-effective solu- tion. However, many of them are unreliable and suffer from having a low output at heads in the 10-15m range which is common in The Gambia. Hence they may not be able to approach the output that would be obtained from several people using ropes and buckets simultaneously, as they do at present. This could cause queueing at the well and encourage people to seek other unhygienic water supplies instead. Also, unlike mech- anized methods of water lifting, handpumps cannot generally be used to pump water into a storage tank to feed a small dis- tribution network, so they tend to involve longer mean dis- tances for people to walk to collect water. Handpumps should, 13/ Peter L. Fraenkel, Supplementary Report on Solar Photovoltaic Water Pumps (I T Power Ltd, UK), for Republic of the Gambia Department of Water Resources and UNDTCD, May 1983. - 46 - however, always be the first possibility for consideration in public water supplies of this kind. (b) Given an adequate wind regime, windpu!ps are generally the least-cost mechanised method of water lifting for the quan- tities of water required in this case; however, The Gambia has only a limited wind resource. 14/ This factor limits the effective use of windpumps to the west of the country, espe- cially near the coast, and in sites clear of large trees where a windpump gains good exposure to the wind. However, as a result of the recommendations in the Fraenkel (1983) study cited above, two Tozzi & Bardi (Italy) windpumps with 6 m diameter rotors have been installed in The Gambian villages of Tanji, near the coast, and Sarakunda, 130 km inland. The total installed cost of each system, including a water storage tank, was US$9,400, which is low compared with most other mechanical water lifting possibilities. Water supplied by the windpump is generally well in excess of thi local water requirement, with daily outputs in the 10 - 30 m range, Even the unit located inland is performing better than anticipated and it appears that inland locations may prove adequately windy for a lightly- loaded windpump. Clearly, useful lessons relevant to the rural health sector may be learnt from these two installations, although it is unlikely that more than one or two health centers or dispensaries will have the appropriate needs and conditions to allow the use of a windpump. The consultants should at least consider the windpum) option for the health centers and dispensaries located near the coast. (c) Solar pumps are relatively capital intensive, but the solar regime in The Gambia is adequate to allow their use in almost any location in the country. Therefore, in situations where say 10 cubic meters per day or more of water are required, a solar pump may prove to be the most effective solution since a hand-pump will not be adequate to provide the required output and in most cases wind conditions will be inadequate for windpumps. Recommendations 4.14 Although the need for improved water supply is urgent in The Gambia, it is not possible to recommend immediate procurement and instal- lation of equipment without a more detailed review of needs specific to 14/ Peter L. Fraenkel, Report on a Consultancy Visit on Water Pumping Windmills, (IT Power Ltd, UK), for Republic of the Gambia Department of Water Resources and UNDTCD, May 1983. - 47 - selected locations. Therefore, it is recommended that a consultant with specialised knowledge of water supplies, irrigation and water lifting techniques be appointed to fulfill the following objectives: (a) DiscusS the water supply problem with officials in DOW, the Department of Water Resources, the UNDP and the Department of Agriculture, to establish in some detail their views on the problem, the priorities, and how any new project will fit with existing activities. (b) Visit all health centers and dispensaries with sub-standard water supplies in order to make a basic site survey and deter- mine the specification of suitable systems for each site. These systems could be solar pumps,, but a case for their use needs to be made in techno-economic terms against other options such as hand-pumps or wind-pumps. (c) Determine with further investigation whether there may be a case for initiating the use of solar pumps for small scale irrigation in McCarthy Island Division, i.e., if the market value of the crops, size of land-holding, typical pumping head and length of the irrigation season appear to offer some pros- pect for economic viability. This could be done using data collected in Banjul during the initial discussion referred to above. The consultant should then visit typical smallholdings in the relevant area in order to determine the specific tech- nical requirements, with a view to carrying out a techno-eco- nomic feasibility study. The study should seek to compare the use of solar irrigation pumps with alternative options, such as diesel, as well as demonstrating that solar power can deliver water at an economic cost. (d) Prepare a detailed report, justifying (wherever feasible) and specifying: (i) pumping equipment most suited for specific health centers and dispensaries (which could be a mixture of hand-pumps, solar pumps, etc.); and (ii) the case, if any, for introducing solar irrigation pumps in the McCarthy Island Division, or elsewhere in The Cambia, and preparation of a costed proposal for a possible project. Budgetary Requirement 4.15 It is estimated that a three to four week mission to The Gambia, followed by two man-weeks of investigation, analysis and report writing, will be necessary to implement the recommendations set out above. The tentative budget is as follows: - 48 - US$ International travel 1,800 Travel within The Gambia 200 Subsistence (26 days) 2,800 Consultancy (approx 44 man-days) 9,000 - 17,000 Miscellaneous (telephone, telexes, secretarial, etc.) 2,000 Totall 15,800 - 23,800 The lower consultant charges are for a free-lance consultant, while the higher figure is for a person hired through a company with associated overheads. The lower figure of US$15,800 is the more realistic, since qualified individuals should be available at that rate. - 49 - V. tHE PROPOSED SOUAR APPLICATIONS PROJECT Project Description 5.1 The project consists mainly of the procurement and installation of solar powered equipment for use in the Health and Telecommunications sectors of the rural areas of The Gambia. It also proviies for associated technical assistance and for the possible development of a consultant study on solar water pumping for use in the health sector and the McCarthy Island Division. 5.2 The initial call for tenders and procurement would -be issued from Banjul by DOH and Gamtel working in association with the Ministry of Economic Planning and Industrial Development (MEPID). If need be, they could seek technical advice from ESMAP. DOH and Camtel wotLd supervise the acceptance and installation of the equipment. They would be assisted in their task by a Specialist Adviser whose services would be paid for through the project. He would be responsible for technical aspects of the installation and work closely with Gambian counterparts and train them in the operation and maintenance of the new systems. The Specialist Adviser would also assist in the appointment of an Assistant Engiaeer whom he would partly triin. The Assistant Engineer would remain in the country for about a year after equipment installation is completed and correct any equipment problems that may ensue. The Specialist Adviser would return briefly, probably once in the middle of the first year of system operation and again about a year after commissioning the equipment, in order to prepare the Project Review Report. 5.3 The proposed consultant study on solar water pumping would be undertaken by MEPID in collaboration with Water Resources Board, the Department of Agriculture and other relevant agencies in accordance, with the outline presented in Chapter IV. Annex B of the Technical Supplement provides the climatic data which will be needed by bidder: to design their equipment. Health Sector 5.4 It is recommended that a total of 25 health centers and dispen- saries be each supplied with a standard package consisting of a photovoltaic powered vaccine refrigerator and two fluorescent lights powered from the same array and battery pack. In addition, a solar water heating unit should be provided for each center. It is also recommended to provide public telephone terminals at or near all the gove.rnment rural health centers and dispensaries as part of the telecommunications package discussed below (para. 5.5). Pinally, the preparation of a separate con- sultant study is recommended to define specific technical improvements for the water supplies at the five health centers and ten dispensaries not at present equipped with piped water. Although solar pumps have been discussed as a possible solution, the study would review all appropriate - 50 - options (solar, wind, handpumps, diesel etc.) with a view to recommending the most effective solution for each location. The study would also review a non-health-sector related activity, namely, the prospects for using solar water pumps for irrigation in the McCarthy Island Division. The objectiie of the study would be to prepare a separate pre-investment report which could be used to seek financing for the procurement and installation of suitable water pumping equipment. Telecommunications Sector 5.5 It is proposed that three primary elements of the rural telephone network be re-equipped in order to overcome power supply and fuel supply problems which currently plague the effective operation of the system specifically, recommendations are: . (a) conversion of the existing diesel-electric power supplies for the three key UHF trunk repeater stations to photovoltaic power; (b) replacement of life-expired and under-sized battery banks for powering the VHF links and spurs from seven provincial telephone exchanges; and (c) replacement of 19 existing (or formerly existing) single channel VHF public telephone links in the villages, and one private single channel VHF link at Kuntaur, with solar power units and new telecommunications systems. This is to be supple- mented, if feasible, by rehabilitation of existing solar powered VHF systems and creation of eight further VHF single channel terminals. Insofar as possible, the VHF terminals are to be located at or near the health centers or dispensaries. It is recommended that, wherever possible, DOH personnel be responsible for operating and supervising the public telephone booths. Implementation Plan 5.6 The proposed schedule of activities is given in Table 5.1. These are summarised below: (a) Approval by the Government of The Cambia. The project was approved by the Government in December 1984. (b) Choice of donors. The Government would identify prospective donor(s) to finance the project. In this task they could seek assistance from ESMAP - June 1985. (c) Invitation of tenders. Tender documents would be finalized and circulated to potential equipment suppliers. 15/ The call for tenders would be issued from appropriate government departments and technical assistance may be provided by ESMAP in preparing the equipment specifications -- August 1985. 15/ Draft documents are presented in Annex A of the Technical Supplement. * aF . A Tab$* 5.t UJECT fPt Trt OtnStr t 0 tf SCLM OOM SYSTtS 13 TME #OM4M TAND 06 S1 TI1IS SSC S Oif TM AMIA Ouplameatlt ---t9-- to- SOS - -two t Activll Aseaw tt m DMe t Ju r Aug Sao Oct Wte Dee JOe ftb Nr A NW Au Jul Ag So Oat WV Dot Sa F W WV Joe Jut ug I. Approval oy f*e gwvoremt at MSIWV A. Choieotet of amIg (EDS) *4 - J. lntitatnle aft 4at t1t *&lw T ~ 4. lolwe0ttee at IeUre wre tflW) S. Pl orders am * . Ill sy acept o *. tasStl atIn of OuSt_ SA& 00t *e e..m e. *. ON***$ ltsto at 46"0 E A a amfn a.Ieislst Iewoluarflo & trebRIn Al -bl SO GMc.me.ee.. SC. Projet SWWI £ Ret port SA so. _epaat' btua On Mntir S1paS ll Seem toat upasleIlt Aalse pree.t Sh tAlsteat faltt_ru t y SA * Seaolttt Advewr at A- * ltst talee - 52 - (d) Evaluation of tenders. Tenders should be received in Banjul in about 2 months and be reviewed by the appropriate government officials with possible assistance from ESMAP -- October 1985. (e) Placement of ordets. Orders would be placed through the Government's normal procurement channels of the Government November 1985. (f) Delivers and acceptance of equipment. It is expected that equipment would begin arriving at Banjul in February 1986. At this stage the Specialist Adviser would arrive to assist with the acceptance and installation of the equipment and to start training his Gambian counterparts. The Specialist Adviser would- ensure that the equipment supplied is acceptable and corresponds to specifications -- February-March 1986. (g) Installation of equipment. The Specialist Adviser will locate the exact sites and supervise site preparation and installation of equipment. The Government, with assistance from the Spe- cialist Adviser, would then appoint the Assistant Engineer, who should be scheduled to arrive in March 1986. The Assistant Engineer would support the Specialist Adviser with equipment installation. An important element of the work of the Special- ist Adviser during this period would be on-the-job training of Gambian counterparts in the correct installation and mainte- nance of. the equipment - February-June 1986. (h) Equipment evaluation and training. It is possible during the early months of system operation, that some typical shakedown problems might surface. The Assistant Engineer would be responsible -- with minimal guidance from Specialist Adviser -- for their resolution. The Gambian counterparts and the Assis- tant Engineer would establish a planned maintenance and moni- toring program in which all the installed equipment would be regularly visited. They would check that the equipment is pro- perly used and develop long-term maintenance plans and fault reporting procedures. The Specialist Adviser would be present during the first part of this phase and be available to respond to any questions that cannot be resolved by the Assistant Engineer and his colleagues. He might undertake a brief return mission, perhaps for two to three weeks around 1986 to visit the sites and resolve any problems that may have arisen and which may require his specialist expertise - July 1986-August 1987. (i) Completion. The projected completion date for the project is June 1987, after which the relevant Gambian institutions would take sole responsibility for the operation and maintenance of the equipment. At this point the Assistant Engineer's contract would terminate. - 53 - (j) Project Review and Report. Much will no doubt be learnt during the course of the various activities outlined above which will be of value for both future planning in The Gambia and else- where in the developing countries. Therefore, the Specialist Adviser would return to The Gambia to evaluate the project and prepare a review report for the use of the Government, donor agency(ies) and ESMAP. It would probably also be of serious interest to other developing countries -- August/September, 1987. (k) Consultant study on solar water pumping. It is recommended that a consultant be appointed to investigate and make recommendations for specific measures to be taken to improve the water supplies at a number of health centers and dispensaries. He would also explore the potential use of solar pumping systems for irrigation in the McCarthy Island Division. Since this activity is not linked to the mainstream project, its scheduling' is flexible and could tentatively be carried out during 1985. - 54 - Budget 5.7 The breakdown of the budget is given below: I. Equipment Procurement Uss (a) Equipment for Health Depsrtment (25 rural 1health centeirs and ipnais 26 solar vaccine refrigerators plus spare parts 136,250 16/ 27 pairs of fluorescent lights with extra PV modules and batteries 18,090 25 integral solar water heaters with storage tanks 7,C04 Subtotal 161 840 Unit cost (b) EKuipment for Camtel -iJ power suFplies for 3 UHF repeater stations Subtotal 61 '450 unit cost (i) new batteries for 7 provincial exchanges Subtotal 39,200 unit cOst (iii) 20 new VHF single channel links pilus i extra solar power system and 8 spare sets of batteries; including cross-country vehicle with two-way radio and test-equipment Subtotal 191 000 Unit co-st Total equipment procurement budget: 453,490 contingency fund at about IOX 46,5lO Total 500,000 II. Technical Assistance (a) Light vehicle (eg. ick-up truck) 7,000 Running costs for 28,ooo miles (15 months) 6,000 Subtotal 13,000 16/ Existing international prices for equipment were reviewed and the budgeting figures used here represent the lowest cost plus 25X of the differentiaL between highest and lowest price. - 55 - (b) Specialist Adviser: Fees: (8 man-months) 17/ 28,000 - 80,000 International travel T3x) 5,400 Subsistence (30 weeks at US$350) 10,500 Subtotal 43,900 - 95,900 (c) Assistant Engineer: Pay: (15 man months) 18/ 15,000 International Travel (2U) 3,600 Subsistence (65 weeks at US$175) 11,375 Subtotal 29,375 (d) Water supply consultancy: Fees: (approx 44 mod) 9,000 179000- International travel (lx) 1,800 Local travel 200 Subsistence (26 days) 2,800 Miscellaneous (telephone, telexes, secretarial, etc). 2,000 Subtotal 15,800 - 23,800 Total technical assistance budget: 102,075 - 162,075 contingency (approx. lOZ) 9,793 - 15,925 Total 112,000 - 178,000 Total budget requirement 1. Equipment procurement 500,000 2. Technical assistance 112,000 - 178,000 612,000 - 678,000 5.8 It is expected that an individual freelance consultant should be available to undertake the consultant study on solar water pumps (para. 4.14) as ;;ell as take on the Specialist Adviser's position. Therefore, the technical assistance budget should tend toward the lower figure. 17/ Reflects fee for consultant hired through company. 18/ Assumes junior/volunteer status with payment at US$10,000 p.a. plus 50% overhead to cover social costs and recruitment overheads. - 56 - Annex 1 Page 1 of 4 COMPARATIVE ANALYSIS OF KEROSENE AND SOLAR VACCINE REFRIGERATORS A. Kerosene Refrigerators Electrolux model RAK660 costs US$348 f.o.b. Sweden. Estimated 15X shipping gives installed price of US$400.. Fuel consumption 1.6 liter/day; 200 liter of kerosene delivered every four months (residue of 0.04 liter/day for lighting purposes). Price kerosene D 2.32/litre = US$0.77 per liter ex-Banjul. Delivery costs average US$175 per 400 liters or US$0.44 per liter (on basis of average of two drums per delivery average distance return from Banjul of 390 miles at 45C mile). Delivered price of kerosene is therefore US$1.21/litre. Spare parts recurrent costs for kerosene fridges as follows: wicks (size 10) 125 units at D 40.00 D 5000 glasses 150 units at D 50.00 D 7500 burners 2 units at D 200.00 D 400 hence average spares per unit is D 516 = US$136 hence average per unit: fuel US$453 fuel delivery US$264 spares US$136 average annual recurrent costs (total): US$853 annualised capital cost: (15 yr at 10) US$ 48 annualised life-cycle cost: US$901 number of doses per refrigerator-year: a/ 13,373 refrigerator cost (annualised) per dose: b/ 6.74C EPI annual overheads (est) US$350,000 ($14,000 per center) therefore EPI overhead per dose 3 104.69C Total cost per dose: 111.43C a/ Allowing for 152 loss to cover unavoidable wastage on basis of 1983 vaccine disbursement (Table 2.3) plus further 202 loss due to refrigerator down-time. b/ Using 1981 dollar total minus allowance for vaccines and for kerosene at reduced 1983 rate; see Table 2.3. - 57 - Annex 1 Page 2 of 4 BS Solar Vaccine Refrigerator As smot ions average electrical load 1 kWh per 24 hrs a/ battery 902 charged at end of sunnrest month (October) 99% availability required in least sunny month (December) using one yea-'le daily insolation data for the Gambia. Array size 280W(p), and Battery 8 kWh(e) b/ Cost functions: unit as sized ~~~ - ~~~~low high low high US$ per US$ 4 Lead-acid batteries 190 210 /kWh 1,520 1,680 PY modules 5 7 /Wp 1,400 1,960 Controls & ancilliaries 1 1.5 /Wp 280 420 Refrigerator cabinet 500 700 each 500 700 sub-total: s700 4;760J System integration 15 20 X 555 952 sub-total f.o.b. price: $ 4,255 5,712 Shipping and installation 10 15 X 425 857 Total installed cost: 4,680 6,569 N Note that the SPC PV refrigerator at Gunjur has maintained operating temperatures within WHO spec. at typical daily electrical loads of 840 to 1080 Wh/day. The system at Kaur had a higher load of 960 to 1440 Wh/day and frequently failed to keep within the spec. Actual measured results are most meaningful in this case since the primary variable affecting refrigeration load is operator activity in terms of opening cabinet and inserting warm items rather than simple heat transfer through the cabinet. b/ For comparison, the SPC units in use in the Cambia have a nominal 315 Wp array with 7.56 kWh(e) battery storage. 101 Discount Rate lo0w high (US$) Reliability 100% 90% Doses per refrigerator-year a/ 16,716.00 15,044.00 Annualised capital cost 656.00 898.00 Annual O& costs (est) 60.00 120.00 Annualised life cycle cost 716.00 1,018.00 EPI overhead per dose 0.84 - 0.93 Refrigerator cost per dose 0.04 - 0.07 Total cost per dose 0.88 - 1.00 a/ , Allowing for 15% loss to cover unavoidable wastage before deducting any losses due to refrigerator and on basis of 1983 vaccine disbursemenr (Table 2.3). - 58 - Annex 1 Page 3 of 4 C. Sumnary Comparison: Kerosene and Solar Refrigerators 10% Discount Rate Kerosene. Solar Solar low high CUSF) Reliability 801 100X 901 Number of doses per annum 13,373.00 16,716.00 15,044.00 Capital cost (installed) 400.00 4,680.00 6,569.00 Recurrent costs 853.00 60.00 120.00 Annualised life cycle cost 901.00 716.00 1,018.00 Refrigerator cost per dose 0.07 0.04 0.07 BPI overhead per dose 1.05 0.84 0.93 Total cost per dose 1.12 0.88 1.00 PW of capital invested in solar option [A] 5,497.00 7513.00 Net benefit of solar option in saved recurrent costs 793.00 733.00 PW of net benefit (B] 6,637.00 6,135.00 PW of recurrent costs 502.00 1,004.00 Net benefit/cost ratio [B/A+Cl 1.11 0.72 Gross benefit per dose 23.40 11.60 PW of gross benefit 32,743.00 14,606.00 PW of recurrent costs 502.00 1,004.00 Gross benefit/cost ratio [D/A+E] 5.46 1.71 PW a Present Worth (or Present Value) 1. The methodology used for the analysis consists of calculating life cycle costs for each option by taking the summed present values of their respective cash flows and annualizing these using a 10% discount rate. It is also assumed that the solar vaccine refrigerator will be 90- 1001 reliable, compared with 80X reliability of kerosene refrigerators, i.e. 90-100% of the vaccines stored (after allowing for 15% unavoidable losses from -he total disbursement due to packing) in solar refrigerators will be usa.te, but only 801 from kerosene units. In other words, it is expected that there will be an improvement in the number of successful immunizations by reducing the occurrence of vaccine unavailability due to down-time with kerosene refrigerators. Since there are no reliable statistics for actual down-time, the somewhat conservative figure of 20X losses from kerosene refrigerators has been used as it appears to accord with several independent estimates by field personnel. 2. It is important to note that many immunizations require a course of several vaccinations to be successful. Refrigerator down-time is, thus, more damaging in terms of wasted vaccinations than simply the loss on the days in question, since people who return for boosters, often after walking several kilometers, find they cannot have them and then rarely return on another occasion. Such incomplete and ineffective courses also reduce the general credibility of the DOH campaigns to en- courage the population to seek immunization. - 59 - Annex 1 Page 4 of 4 3. The current true overheads are not known with certainty, but it is probable that the BPI program total costs in terms of US dollars have remained approximately the same as they were in 1980-81 - US$400,000. Allowing for kerosene costs, plus kerosene distribution costs which are already incorporated in the refrigerator cost analysis, and for a rather higher vaccine budget in 1981 than in 1983 (the year for which immuniza- tion statistics are being used), the possible EPI overhead stands at about US$350,000, or US$14,000 per health center or dispensary. It has been assumed in the analysis that on the basis of 1983 figures, 13,373 doses per refrigerator year would be administered using kerosene refrig- erators, while 15,044 to 16,t716 could be administered from the same supply of vaccines using solar refrigerators. These figures reflect 20X, 102 and zero vaccine losses respectively due to refrigerator down-time. It can be seen from Section A that the EPI overhead per dose ranges from US$1.12 with a kerosene refrigerator down to US$1.00 to US$0.88 using more reliable solar units. This is not of course a cost-saving as such, but it does draw attention to the substantial overheads involved in giv- ing a vaccination over and above the costs of the refrigerator and its operation and maintenance. 4. The conclusions and assumptions of the economic analysis set out in Section A indicate that the annualized life cycLe cost of a kerosene refrigerator is US$901 and the cost per dose for this option is US$1.12. The solar low option has an annualized life-cycle cost of US$716, with i cost per dose of US$0.88, while the annualized cost of solar high is US$1,018 at US$1.00 per dose. 5. Finally, it is worth pointing out that the recurrent costs of solar refrigerators are much smalLer than those of kerosene refrigera- tors. Therefore, if the solar powered vaccine refrigerators are financed by grant aid, there will be a major impact in reducing the recurrent costs, and will release these limited funds for other purposes. The analysis estimates an annual recurrent cost of US$853 per kerosene refrigerator, ($21,325 for all 25 units) compared with a probable recur- rent cost from US$60 to US$120 for solar units. Therefore the net saving in recurrent costs for the EPI program will be in the range of US$18,325 to US$19,825 simply on refrigerator operation and maintenance. 6. The solar refrigerators are not significantly more expensive in life-cycle cost terms, and their advantages consists largely of signifi- cant savings in recurrent costs combined with an expected improvement in the efficiency with which vaccines are stored and used, which is after all the primary objective of the entire US$400,000 BPI program. Since the current problems in the Gambia with fuel shortages are likely to persist, it is possible that the advantages to be gained from solar vaccine refrigerators will increase in future. - 60 - Annex 2 Page 1 of 2 TECHNICAL DESCRIPTION OF REFRIGERATORS AND LIGHTING SYSTEMS 1. The general requirement is for a proven model of solar photo- voltaic vaccine refrigerator, preferably having World Health Organization (WHO) approval and supplied complete with an appropriate photovoltaic array and battery pack. The refrigerator must be capable of achieving the WHO specification of holding a freezer temperature between 0 and -15'C and a refrigerator temperature within the ran8e 4 to 8eC with a day time ambient temperature averaging 43C and a night time holdover at 35C average. A supplementary lighting package is also to be supplied, as detailed below, powered from the same array and battery pack. The sup- plier is required to provide evidence that their product has previously been used, successfully, in tropical locations. Under no circumstances should. a new and untried make or model of solar vaccine refrigerator be procured. A list of suppliers of tried and tested equipment is given in Annex AS. 2. The freezing compartment is required to hold a minimum of 4 ice-packs as well as vaccines. It is preferable to procure a unit cap- able of freezing up to 8 ice-packs. The ice-packs are plastic containers filled with water which held approximately 300gm of water and have dimen- sions of approximately 165 x 115 a 19 mm or (6.5 x 4.5 x 0.75 in). Therefore the ideal freezer size is in the region of 50 litres, (1.8 cu. ft.). The existing Solar Power solar vaccine refrigerators in use in the Gambia have a freezer of 34 litre capacity which is about the minimum acceptable volume. Refrigerator volume is less critical as only a small proportion of the vaccines are stored in that compartment. 3. The refrigeration load consists of refreezing ice-packs, as detailed above, twice per week. It can be assumed that the ice-packs will contain water at about 25*C when inserted into the refrigerator and that at least 48 hours will elapse before they ere removed and replaced. The vaccines are inserted into the refrigerator only once per month and in cold condition and therefore do not represent a significant Load on the system as such. The remaining load is due to opening and closing the door to remove vaccines and ice-packs, and due to heat gain through the cabinet. 4. The cabinet requires to be fitted with a lock to ensure that only authorised staff have access and that the door is less likely to be opened unnecessarily. It must be supplied with three keys, two for each centre and a spare to be retained at headquarters from which further spares can be cut. 5. The solar refrigerator must be supplied with a small lighting package consisting of two appropriate low voltage DC fluorescent lights in robust portable enclosures. These lights should be about 20 W each using transisterized high frequency inverters built to a tropical speci- lication and integral with the units. Each light needs to have suffi- cient heavy insulated cable to allow it to be positioned up to 10 m from - 61 - Annex 2 Page 2 of 2 the solar refrigerator. The lights should draw power from the same battery pack and PV array as the vaccine refrigerator, and it is essen- tial that the battery and photovoltaic array sizing must take account of the power needed to operate the lights on a 4-hour in 24-hour duty cycle. 6. The following minimum instrumentation is also to be provided to facilitate monitoring of the refrigerator performance (and to allow fault finding) to be carried out: (i) max-min thermometer in freezer; (ii) max-min thermometer in refrigerator; (iii) max-min ambient (exter- nal) thermometer battery voltage indicator; and (iv) array current (charge current) ammeter. It would in addition be desirable to have some or all of the following additional instrumentation, if they can be sup- plied within the budget: (i) freezer temperature continuous; (ii) re- frigerator temperature continuous; (iii) compressor running hours meter; and (iv) battery ampere-hour (or watt-hour) meter. A compromise measure worth considering would be to procure two or three sets of the second group of instruments for trouble-shooting purposes with units suspected of being faulty. . - 62 - Annex 3 Page 1lof 5 PHOTOVOLTAIC POWER SYSTEN SIZING AND COST ANALYSIS System Description 1. Photovoltaic systems are electrical generating systems based on photovoltaic, or solar cells (built into arrays) which convert incident sunlight directly into DC electricity. The amount of electric current generated by a solar cell array is priguarily dependent on the intensity of solar radiation striking its exposed area. The open circuit voltage produced, however, is primarily dependent on temperature of the cells. Voltage and cell temperature are inversely related (i-e., an increase in temperature lowers both the output voltage and the output power, thereby reducing the cell's efficiency). 2. A practical stand-alone photovoltaic power system typically requires several elements in addition to the array in order to satisfy the intended load. A typical configuration for. a DC load is 'shown in Figure (1). BLOCKING IDIODE LOLTAGE DC-DC DC VOLTAGE V ICONVERTER LOAD REGULATOR (OPTIONAL) 3. Battery storage, most commonly involving a lead acid battery in present applications, stores electrical energy produced by the solar array in daytime for use during the night or under cloudy conditions. To be considered practical for remote applications, a storage battery should have a long life, require low maintenance, and be able to survive a rnumber of deep discharge cycles with subsequent recharge by the array. 4 * Another important component of a photovoltaic power system using storage is a voltage regulator that controls the output voltage from the array when used to charge the battery. The regulator also limits the loss of water that could occur from gassing of the battery if it were permitted to become overcharged. At night or on cloudy days, a blocking diode (see Pigure 1) prevents the electrical energy stored within 'the battery from discharging through the voltage regulator or the - 63 - Annex 3 Page 2 of 5 array. A system serving a DC-load may require a DC-DC converter to match the system output voltage to the rated voltage of the load. 5. Some photovoltaic applications do not necessarily require battery storage. An example of this is water pumping, which nee4s only a photovoltaic array, a water pump, a water storage tank, and minimum power regulation. in this case, water is pumped into the storage tank during sunlight hours. The storage tank then acts as a reservoir, providing water during non-sunny periods. system Sizing Insolation Characteristics and Their Effect on System Sizing 6. Array and battery storage size requirements to adequately serve the load energy requirement ultimately depend on the amount of insolation at the location of the site. Since the amount of energy received from the sun will depend on location, season, weather, and array orientation, it is essential to account for these factors in sizing the systems. Furthemore, since the intended load may also vary seasonally, sizing will normally require a systematic accounting approach to ensure that sufficient solar energy will be captured by the array to achieve accept- able system reliability throughout the year. Insolation characteristics important for sizing photovoltaic system components include the follow- ing: diurnal variation, regional and seasonal variability, weather (cloud cover and its frequency) and array orientation relative to the direction of the sun. Introduction 7. The term "sizing" means estimating the required size o-r capa- city of all major photovoltaic system elements so that the system will be able to satisfactorily serve the intended load. The sizing methodology described in this section is based on a target level of photovoltaic system reliability that is equivalent to conventional diesel or battery alternatives. The size estimates obtained by this methodology are used to cost the system, and the cost estimates, in turn, are compared with estimates for other possible power generating options in order to deter- mine whether selection of a photovoltaic power system is justified on a comparative life-cycle cost basis. Sizing Methodology 8. The steps in the sizing methodology are summarized below: (a) Calculate the Load. The average daily energy load in kilowatt- hours is calculated for each month of the year. In the simpl- est case, a single load element draws constant power at all times. If several load elements are present, the individual - 64 - Annex 3 Page 3 of 5 elements must be summed. If the load totals vary from day to day, an average daily load over each mouth of the year will be required. (b) Determine the Local Insolation. The appropriate amount of input energy (i.e., insolation, I) to the photovoltaic system at the application site may be obtained from local monthly insolation data for various array tilt angles. (c) Calculate "Worst-Month" Insolation and Load Values. The sizing approach requires identification of the load and insolation values expected during the "worst month" of the year. This is done by constructing a table of average insolation and load values for each month of the year and then determining the month with the lowest ration of insolation to load. The inso- lation and load values for the selected month are used in subsequent steps to calculate required array size and battery storage capacity. (d) Determine Arry and Battery Storage Sizing Factors. The approach used in the methodology to size the array and storage is to apply previously determined "sizing factors" 1/ in array and storage calculations in order to scale the system to achieve a desired level of autonomy (i.e., availability). In some sizing application, the user may specify the level of system autonomy by placing a requirement on the number of sun- less days during which the system must be able to satisfy the intended load. (e) Calculate Array Power and Area. Calculating array size means calculating both its required peak power output in watts and its corresponding area in square meters. The array sizing cal- culations incorporate the worst-month load, array sizing factor based on the worst-month insolation, efficiencies of all major system elements, and a term to accoant for long-term array degradation. (f) Calculate Battery Storage Size. Calculating battery storage size requires scaling the storage to supply the daily energy load for a sufficient period of time to assure that the photo- voltaic system meets the load requirements at least 96 to 98X of the time (or equivalently, a 0.1 "loss of energy probabi- lity" (LOEP) during the woret month of the year). Alterna- tively, the storage can be sized to supply the load for a specified number of sunless days. In either case, the actual 1/ Macomber, H. L., Ruzek, J.8., Photovoltaic Stand-ALone Systems, DOES/NASA/0195-I, NASA CR-165352 M206, August, 1981. - 65 - Annex 3 Page 4 of 5 rated capacity is adjusted to ensure battery operation within acceptable depth of discharge limits. (s) Calculate the Voltage Regulator Size. The battery charging voltage regulator is sized to handle the maximum amount of array output power (DC) that is not being used to supply the load. For conservatism, this will be considered to be the full rated peak array output power to account for situations in which the load has been disconnected. System Cost Analysis 9. Photovoltaic power system life-cycle cost is estimated from the initial cost of the system installed at the site and the present = value of .l recurrent costs associated with system operation. Photovoltaic systems are typically capital intensive (i.e., they require a large initial capital expenditure), but have low operating costs (i.e, zero fuel cost and small expenditures for replacement, operation, and main- tenance). The sum of capital and recurrent expenditures represents the equivalent amount of money required at the time of system installation to completely cover all costs associated with the photovoltaic system, including a return on the investment, over its operating lifetime (typi- cailly.assumed to be 30 years). Life-cycle cost of the alternative to a photovoltaic system similarly combines the associated first cost and operating cost for comparison with the photovoltaic option. 10. The recurrent cost in photovoltaic systems are attributed primarily to battery replacement plus battery operation and maintenance. 11. Battery lifetimes are typically between 5 to 10 years depending on manufacturer-specific battery characteristics, number of discharge cycles, design depth of discharge, and operating temperature. In con- trast, the anticipated life of a photovoltaic array is estimated to be about 30 years. Battery replacement at regular intervals are, therefore, required throughout the operating lifetime of the photovoLtaic array. 12. T.?e present value of the sum of all battery replacements (RPV) over the photovoltaic system lifetime, escalated and discounted to account for the timing of the expenditure, is estimated as follows: RPV a [(BATC)(l-SV) + LREPJ x 1 - escb j x k * j~~~~~+l 1 +d dv where BATC delivered cost of batteries SV = fractional saliage value of batterieb at time of replacement - 66 - Annex 3 Page 5 of 5 LREP a labor cost of battery replacement (in base-year dollars) j a counter for number of battery replacements (1, nrep) k a battery lifetime (years) j x k = constrained to be strictly less than photovoltaic system lifetime or lifetime used in life-cycle cost analysis (years) 13. Regular operation and maintenance costs are estimated on an annual cost basis. These costs include expenditures for activities such as array, battery, and inverter maintenance; component replacements (other than batteries); and grounds, structural and electrical upkeep. Annual operation and maintenance costs (OM) can be estimated on the basis of the number of required visits to the site per year times the cost per trip in base-year dollars. A heuristic frequently used to estimate annual operation and maintenance expenditures is to assume that annual expenses are a fixed percentage of the initial cost of equipment. The present value of the cost of operation and maintenance procedure is the derived annual amount summed over the system lifetime, including any real escalation and discounting of expenditures over time. Annual expenditures are, therefore, growing in dollar amount at the constant rate of real escalation, if any. Present value of operation and main- tenance cost (OMPV) is presented as follows: fi1+ esc 1-1 esc Ni OMPV = OM x I om x om e O if dr # escom dr - escom 1 + dr or OMPV a OM x N, if dr -escom where OM a annual operation and maintenance cost in base year dollars escom a real (above inflation) annual escalation rate for operation and maintenance activities (fraction), typicaly 0% Photovoltaic system life-cycle cost (LCC) is then the sum of the initial installed system cost, and the present value of recurrent costs, Equation (1) plus Equation (2) above. - 67 - Annex 4 Page 1 of 2 ANALYSIS OF REPSATRR STATION POWER SYSTEMS A. Electrical Load (a) Telecommunications loads SA at 24V 24h/day 2.88kWh/day (b) Lighting toad: (i) Security lights for 12h/day 3 x 18W sodium (SOX) 12h/day 0.65kWh/day (ii) Transmitter room & caretaker's lights - 4s30W fluorescent 6h/day 0.72kWh/day Total electrical load: 4.25kWh/day B. Solar Powered System (a) Assumptions: Array tilt angle 15' facing due 1outh Least sunny month (December) 4.5 kWh/ml day (in the plane of the array) Sunniest month (October) 5.8 kWh/mZ day (in the plane of the array) Photovoltaic system life 1S years Battery life 7.5 years (b) Calculated results: Availability Array Rating Battery Capacity (tZ) W(p)/kWh W(p) -kM (days) 95 266 1,130 25.5 6 99 280 1,190 34 8 (c) Cost assumptions (for 99% available system): solar solar low high Solar PV modules at $5-7/W(p) 5,950 to 8,330 Battery at $170-210/kWh 5,780 to 7,140 Control equipment at $.75-1.25/kWh) 892 to 1,487 AncilLiaries at $0.5-1/w(p) 595 to 1,190 Component cost sub-total: 13,217 to 18,147 System integration at 202 of above 1,982 to 3,629 Bid price (foo) range 15,199 to 21,776 Shipping & installation range 3,039 to 5,444 Installed budget cost range $ 18,238 to 27,220 -68- Annex 4 Page 2 of 2 AlterAatively, if procured on a "turnkey" basis, the cost could be expected to be in the $20-25 per peak watt range, as follows: Turnkey total cost range $ 23,800 to 29,750 C. Diesel Generating Set SYstem (a) Assumptions: Existing generating sets & battery chargers have 50% of their useful lives available (total life 7.5 yrs) Existing batteries need immediate replacement; (when charged from an engine useful life 5 yrs) Capital cost of 2 x 3.4kW generating sets $ 3,000 Capital cost of 12 x 200Ah 2V batteries $ 1,010 Capital cost of H&S 30A battery charger $ 14170 Capital cost of control/switch geal. (10%) $ 440 Total procurement cost: $ 5,620 Installation & shipping cost (at 15%) $ 843 Tota. installed cost: $ 6,463 Fuel costs at Ds 2.32/liter $ 0.60/1 Fuel delivery 45c/mile x 180ml x 2000 1 $ 0.04/1 Total fuel cost (delivered) $/liter $ 0.64/1 Annual fuel consumption (at 16 liters/day) 5,840 Annual gross fuel cost $ 3,738 Engine O&M costs $ 200/yr Annualised capital cost @ 10% discount rate $ 976 Annualised life-cycle-cost Q 10% discount rate $ 4,914 D. Economic Cost-Benefit Comparison (a) Assumptions: costs are anmualised over 15 years (b) Calculated results: for 10% discount - solar low solar high diesel (in'US$) A. Installed cost 18,238 27,220 6,463 Recurrent costs (C&M) p.a. 100 200 3,938 Annualised life-cycle cost 2,617 3,869 4,914 Unit cost per applied kWh 1.68 2.49 3.17 Cost as percentage of diesel 53% 78% 100% Benefits of solar (saved recurrent costs) 3,838 3,738 - Simpte pay-back time (years) 4.75 7.28 - B. P.W. of benefits of solar 32,124 31,287 - Recurrent costs of solar p.a. 100 200 - C. P.W. of recurrent costs 837 1,674 - B Benefit/Cost ratio (---) 1.68 1.08 - A+C 69- Annex S Page 1 of 2 REPLACEMENT BATTERY REQUIREMENTS FOR PROVINCIAL EXCHANGES Assumptions and Results Telecommunications load: 7 to 9A at 24V 9 (cont.) Other loads: nil Average power output (taking 9A load) 216 W Assumed charging efficiency 70 X Energy input required per 24h 7.405 kWh System: existing new Time needed to fully charge batteries 12 h 7.7 h Mean power input during charge 617 W 1,234 W Charging current at 24V 26 A 40 A Battery capacity 200 Ah 800 Ah Recommended charge rate (Tungstone) 10 A 40 A Battery storage capacity 4.8 kWh , 19.2 kWh Period for 1002 discharge 22 h 88 h Depth of discharge for proposed cycle 45 X 91 2 Depth of discharge after 12h 45 Z 86 X Number of useful charging cycles/life 2,000 5,000. Life in years assuming 1 cycle/day 5.5yr 13.7yr4 Capital cost of replacement battery $1,300 $3,900 Capital cost of replacement charger $1,300 $1,500 Shipping and installation (est) $ 150 $ 200 Installed cost (est) $2,750 $5,600 Annual O&M costs $ 200 $ 100 Present worth over 15yr at 10% disc. $5,179 $7,292 Annualized capital costs $ 619 $ 871 Annualized life-cycle costs $ 819 $ 971 POWER SUPPLY FOR PROVINCIAL - TECHNICAL CONSIDERATIONS 1. Recognizing the unreliability of the power supplies and ' charging systems, it seems reasonable to design the battery storage to have several days' capacity without recharging, instead of about 22 hours and to be able to accept an adequate recharge in as short a time as possible. 2. The table above indicates that the existing system can only safely support a maximum of about Ilh between charges (for 50X depth of discharge) and even if recharged in 12 hours, as is intended but not always achieved at present, the necessary charging current (at 26A) considerably exceeds that recommended by the manufacturers for float charging (it should be IOA). The number of cycles a kattery can survive is closely related to the maximum depth of discharge as well as to -70 - Annex S - iTf 2 Page 2 of whether the charging current is within the limits recommended to avoid gassing. Typical life-time estimates for tubular plate cells (as are used at present) are: (i) 20% maximum discharge - 5,000 cycles; (ii) 50S maximum discharge - 5,000 cycles; and (iii) 801 maximum discharge -- 2,500 cycles. 3. The occasional 1001 (or nearly lOO) discharge, believed to occur with existing equipment, when combined with an excessive rate of charging, will in reality make the achievement of 2,500 cycles problematic. Therefore, for the purposes of analysis, a useful life of 2,000 cycles has been assumed for the existing batteries. This is probably generous, as the life becomes generally unpredictable, and can be radically reduced due to occasional periods of gross over-discharge combined with regular gassing due to the excessive rate of charge. 4. The type of cell normally used is the Tungstone 4TPG5O 200Ah 2V cel, which has tubular plates, making it resistant to damage, but which is limited to a recommended IOA float charge rate and is clearly significantly too small for the existing 12h charge rate of about 26A, let alone the proposed rate of 50A. The consequence is that the batteries have virtually been worn out in thrQe to five years instead of achieving a possible 10-15 year life. The retommended Tungstone replacement would therefore be the 8TPG100 with lOhr discharge capacity ratings of 800 Ah. This implies increasing the battery capacity by a factor of 4 compared with at present. 5. The sizing proposed for the replacement batteries of four times the present capacity was dictated primarily for reasons of allowing a high enough charging current so that only 7.7 hours per day -"e needed instead of 12 hours. In any case, if 12 hour recharging is planned, any replacement battery would need to be bigger by about 1001 so as to reduce the rate of charge as the voltage rises. Obviously, a compromise is possible in that a lesser capacity then recommended could be procured, although this should really exceed about 400 Ah if only to ensure the charging rate is acceptable. - 71 - Annex 6 fajllof 3 TECHNICAL AND COST ANALYSIS OF VHF LINKS A. Electrical Load Nominal transmitter rating 1W 8W Standby power need (idle) 35mA 35mA Transmit power need (at 12V) 0.8A 3.SA Energy requirement (6h/day output) 68Wh/day 2S2Wh/day B. Solar Power S8stem Assumptions exactly as set out in Table 3.11 (B) for 99X availability in least sunny month plus, duty cycle of 6h/day total transmit time: Nominal transmitter rating 1W 8W Precise array sizing l9W(p) 71W(p) Actual array sizing (std. modules) 20W(p) 70W(p) Battery sizing 0.54kWh 2.02kWh Battery rating 12V 45Ah 168Ah C. Cost Data Actual ccats will vary, depending on supplier, specification and technical features. These are guideline figures for a single VHF link gained as a result of making enquiries to several potential suppliers: Item Unit Cost Quantity Total Cost (US$) (per link) (US$) Transceiver unit 2,000 2 4,000 Yagi antenna 250 2 500 Mast and fittings 500 1 500 Solar power system a/ 1000-2,000 1 1000-2,000 Weatherproof housing 500 1 500 Telephone/coin collector 1,500 1 1,500 Telephone booth 500 1 500 Shipping and installation 800 1 80 Total per system: 8,300 to 9,300- a/ The lower price relates to a 2OWp array with a 45Ah (12V) battery to power a nominal 1W transceiver and the higher price relates to a 7OWp array with a l7OAh (12V) battery to power a nomi.aal 8W transceiver. - 72 - Annex 6 Page 2 of 3 D. Economic Analysis Total procurement proposed requires: USs (i) 11 high power (S-IOw) systems 102,000 (ii) 9 low power (1W) systems 75,000 (iii) Provision for solar power at GPMB at Kaur (assuming 1W system) 1,000 (iv) Provision for spare batteries to allow rehabilitation of old systems 1,000 (v) Cross-country vehicle -10,000 (vi) Two-way mobile VHF radio plus test equipment for vehicle 2,000 Total budget cost: 191,000 There are two options available; namely to implement 20 new VHF single-channel links plus the GPMB and to implement an additional 8-second priority ones b3 using rehabilitated existing equipment with new batteries at an estimated cost of US$500 per rehabilita- tion. Hence: Number of installations implemented 21 29 Investment per installation 9,095 6,724 Annualised investment cost (10X 15yrs) 1,087 803 Estimated annual recurrent costs 200 200 Annualised total cost 1,287 1,003 Mean cost per call: ($ 0.45) -(Table 3.7) approximate mean duration per call: 3.5 minutes Average traffic per day to break-even on above costs: (calls) 7.83 6.11 (time) 34 mins 26 mins - 73 - Annex 6 Page 3 of 3 Technical Considerations 1. Each installation requires the following components: exchange VHF transceiver; exchange feeder cable; exchange yagi directional anten- na; terminal yagi directional antenna; terminaL feeder cable; terminaL antenna/array support mast; terminal transceiver; terminal; telephone booth; terminal headset and coin collector; terminal solar array and battery charge regulator terminal; battery pack; and terminal ventilated equipment enclosure(s). It is strongly recommended that supliers be invited to bid to supply complete, fully integrated, sets of the above equiqment. Although the solar power components can be bought from sepa- rate sources, perhaps marginally at a lesser cost, the advantage of 'single-sourcing' the equipment is that there is then a clear rr0.3 >1.0 >10.0 January 1.1 0.9 0 0 0 February 0.8 0.6 0 0 0 March TR TR 0 0 0 April TR TR 0 0 0 May 4.7 4.1 0 0 0 June 73.0 28.4 6 5 3 July 269.4 69.9 16 15 8 August 439.5 92.0 22 19 12 September 295.6 68.0 18 16 9 October 97.4 40.0 8 7 3 November 7.2 5.2 1 0 0 December 0.8 0.5 0 0 0 Yearly total 1189.5 69 62 35 Source: Department of Water Resources, Government of The Gambia, Banjul, Gambia. .-RD 15991RI THE GAWtIA ,- - - GAMTEL NEGA LONG DISTANCE NETWORK … S .~~~~~~~~~~~~~~~~~~I/ I y / ts' _-. - - ----- N - G - - - - - - s m P..W- \"" *-~- 'U ac- ~ ~ ~ ~ ~ - ** .* a .$FmAk.Mt- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ j '- - ~.e,. o - j *.t U. 110 3'~C * -' t -------- _ ---- --_-._ 70 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ _ " THE GAMBIA 4.. -.--e - /. \RURAL HEALTH FACILITIES _ . \ - - o < = > ._ * _ A. ~~~~~~~~~w.S #sa**tt^\ ( - . A ~~~~~~ r~- I~~~~~~~~~~~~~ BA - - .. ~-- X X srS~~~~~~~~~~~~~~~~~~~~ - .4. . . . . , ....... , X 1E, * o e I~~~~woDo t4 T .r c e. OW _4 _- I y w u S E NwE G A 4~~ ~~~~~~~~~~~~ ~ t-_* /;,e.4l r- S E N E G ~~~~A I. Il. #AV* goM it"