Chapter 19. Photovoltaic power supply to a village in Upper Volta*

* Contributed by the ILO.

IN 1979, THE United States Agency for International Development (USAID), in close collaboration with the National Aeronautics and Space Administration (Photovoltaic Development and Support Service, Lewis Research Centre, NASA Le RC), embarked on this project in Tangaye, Upper Volta, as part of its programme entitled “Studies of Energy Needs in the Food System”. The objective of this programme was to improve the quality of life and productivity of small farmers in rural areas of developing countries. The project has as its immediate objective, “the demonstration of the potential for the use of solar cells as a power source for common village tasks with special emphasis on women’s tasks”¹. A village in the Sudano-Sahel zone of West Africa, Tangaye, (Latitude 13°N, Longitude O°) was selected for this pilot scheme (see Fig. 19.1). More recently, photovoltaic power supply to villages has also been provided on a pilot basis in Colombia, Mexico and Tunisia (see Annex).

Tangaye is located in the eastern region of Upper Volta, 190 kilometres east of Ouagadougu. Within the Sudano-Sahel zone there are two distinct seasons - the dry season from May to October inclusive and the rainy season from November to April inclusive. The dry season is characterised by high temperatures (40 degrees Centigrade), and low humidity. Hand-dug wells often dry up and water is scarce. In the rainy season, rainfall averages 100 centimetres annually. Mean daily (total horizontal) solar radiation ranges from 450 langleys in August to 560 langleys in March.

In 1978, Tangaye had a population of 2,172 people comprising 290 families. About half of the population is above 14 years, and of these 54 per cent are women. The village with a populated area of four square kilometres is divided into 8 quarters and 20 sub-quarters. Each quarter has a number of shallow wells dug by the villagers in addition to two large concrete wells built by the government, and a cement-lined one built by an individual. The hand-dug wells are 2 to 6 metres deep and tend to dry up a month or so after the end of the rainy season while the cement/concrete wells are about 10 metres deep and contain water throughout the year.

The people of Tangaye are dependent on subsistence agriculture, the main occupation being farming and cattle rearing. Farming is done mainly in the rainy season when millet, rice, corn, beans, peanuts, sesame, soybeans and cotton are grown. Villagers also raise cattle, sheep, goats, donkeys, horses, pigs and poultry. Men spend much of the time farming (although women help during the peak farming seasons) while women are responsible for all aspects of family care: these include fetching water and pounding or stone-grinding of cereals, which occupy two to three hours of work per day.

I. THE SYSTEM

In March 1979 when the system was installed, it consisted of the following:

- two solar arrays: one rated at 1.8 kWpk, 120 V providing power for all load components and another rated at 111W(pk) 12 V supplying the instruments and control panels;

- two battery subsystems: a 120 V battery for loads and a 12 V battery for the instruments and controls;

- three control subsystems: system voltage and battery charge regulation; over and under-voltage protection and pump and mill controls. Two sets of voltage control equipment (one electromechanical and the other electronic) are used to increase the reliability of the system. Over-voltage protection is required to disconnect the load in the event of failure of the voltage control equipment thus preventing damage to the load, while under-voltage protection is required to avoid excessive discharging of the batteries in periods when radiation and voltage output are low. The pump control consists of two water level sensors: one located in the water tank to stop and start the pump when required and the other in the well to switch off the pump when the water level falls below the pump intake. Mill control consists of interlock switches and a timer to limit daily operating time;

- instrumentation for measuring instantaneous values of system voltage and array, mill and pump currents, ampere-hour meters and run-time-meters;

- a 1 hp burr mill (initially) powered by a 120 V DC permanent magnet motor;

- a positive displacement lift pump powered by a 1/4 hp 120 V DC permanent magnet motor. The pump feeds water to a storage tank of 6 cubic metre capacity.

Due to excessive wear of the plates of the burr-mill, it was replaced by a 3 hp Bell hammer mill in September 1979 and by a Jacobson hammer mill in May 1981.

Thermal stresses led to the premature failure of the solar modules. By April 1981, 29 per cent of the cells had failed and in May 1981 the array was replaced by one with 3.6 kW peak loads and of a different make. The system as it now stands is shown diagramatically in Figure 19.1.

It is perhaps noteworthy to mention that all the components (including the PV cells) were off-the-shelf models.

Performance

A summary of the system’s operating data is shown in Table 19.1. For the first two years of operation, the system (and load bus) were on line 97 per cent of the time. Figs. 19.2 and 19.3 show the outputs of both mill and pump for the first two years after installation.

Figure 19.1. System diagram: photovoltaic-powered pumping/milling system, Tangaye, Upper Volta²

Table 19.1. Tangaye Data Summary March 1, 1979 to March 19, 1981

Burr Mill
Total running hours	522 hours
Rated current	7 amps
Actual average current	7.02 amps
Total amp-hrs used	3,665 amp-hrs
Total energy used	440 kilowatt-hrs
Hammermill
Total running hours	1,639 hours
Rated current	21 amps
Actual average current	14.43 amps
Total amp-hrs used	23,649 amp-hrs
Total energy used	2,838 kilowatt-hrs
Total grain ground	48,861 kilograms
Grain grinding rate (very fine)	30.2 kilograms/hour
	2.1 kilograms/amp-hr
Pump
Total running hours	4,530
Rated current	2.5 amps
Actual average current	1.05 amps
Total amp-hrs used	1.05 amps
Total energy used	4,756 amp-hours
Total water pumped	571 kilowatt-hrs
Water pumping rate	5,539 cu meters
	1.18 cu meters/hour
	1.12 cu meters/amp-hr
Total
Total amp-hrs used	32,070 amp-hrs
Total energy used	3,848 kilowatt-hrs

II. SOCIAL IMPACT OF THE TANGAYE VILLAGE POWER SUPPLY

A number of studies on the social impact of the Tangaye project were undertaken by Roberts.³ These studies examined the “impact of mechanising arduous, time-consuming village tasks on the life of women, with the power source viewed as a black-box”.⁴ They were studies of the effects of pumped water supply and a power-operated grain mill in a remote village. The results of the studies could be of potential interest to other countries, since these are major village activites and photovoltaic power generation might be the only viable method of providing electricity in some remote rural areas of Africa. They are summarised in this section.

Pumped Village Water Supply

The well with a solar-powered pump was found to attract many of the villagers in preference to existing hand-drawn wells from the same aquifer. 90 per cent of the users were to be found in the nearest two of Tangaye’s 16 hamlets, and in the dry season, this formed the major secondary source for three other villages. The average daily estimated per capita use of water in the dry season was 10 litres (9 litres annual average). Water drawn from the well was used for laundry (35 per cent), bathing (29 per cent), animal watering (24 per cent), gardening (5 per cent) and miscellaneous activities including construction and beer brewing (7 per cent). It was observed that people using other wells did not spend as large a proportion of water on laundry or bathing and that users of the well washed their clothes and bathed more often than non-users. The well had become a social focal point in the village.

Figure 19.2. Tangaye mill use

Source: J.E. Martz, A.F. Ratijczak, R. de Lombard, Operational performance of the photovoltaic powered grain mill and water pump at Tangaye, (Upper Volta), op.cit.

Figure 19.3. Tangaye water use

Source: J.E Martz, A.F. Ratijczak, R. de Lombard, Operational performance of the photovoltaic powered grain mill and water pump at Tangaye, (Upper Volta), op.cit.

Power-Driven Mill

Mill use was heavy in the dry season and low in the wet. This can be explained by the fact that in the dry season, farming activities are minimal and there are a number of festive activites which call for large amounts of flour for feeding guests or brewing beer. In addition, this is the season when disease is prevalent and women who are too sick or too busy nursing the sick, find it convenient to take their grains to the mill. On the other hand, the rainy season is the time when money from the sale of crops of the previous harvest is dwindling and when many of the villagers camp near their farms. Women used the mill with varying frequency. The blind, disabled, lepers, childless and elderly hardly ever used it while a definite correlation was found between household farm size and mill use. In addition, women who had acquired wealth through beer-brewing were found to use the mill frequently.

Station Management

In order to integrate the mill into everyday village life, a cooperative to manage the station was formed from representatives of 14 pre-existing agricultural cooperatives in the village. Although originally intended to have all female representation, the present cooperative has totally excluded women from everyday decision-making and work at the mill. The president of the cooperative liaises between the village chief and the cooperative; the millers were chosen among cooperative members. The station manager is a young man from the village.

Milling fees are determined by the cooperative in consultation with the chief. Salaries of the station manager and millers are paid from these fees and any surplus is paid into the cooperative’s bank account. Daily upkeep of the facilities is assumed by the station manager and millers.

It is interesting to note that the villagers themselves had taken initiatives to engage in activities not originally proposed in the project. Thus, a grass-root community forestry project has been embarked upon with the proceeds from milling. Adult literacy classes use PV-powered lights and a vegetable garden to utilise run-off water from the tank was planted.

The USAID proposes to hand over primary responsibility for the demonstration to the Department of Water and Rural Works (HER) of the Government of Upper Volta. HER will assume complete control, technically seconded by the staff of the Institute for Research on New Energy, IREN, (the Upper Volta representative to the Regional Centre for Solar Energy, CRES, being established in Bamako, Mali).

Villagers’ Input to the Project

The building to house the mill was constructed by the villagers using locally-made mud bricks, cement, lumber, steel rods, and sheet-metal roofing. System hardware installation was accomplished with considerable assistance from the men in the village, who prepared the trenches for underground wiring from the PV array to the controls, the well and the mill building. They helped in other tasks including the unloading of trucks and the putting of batteries in racks. The station keeper and millers were trained in system operation, maintenance and repair. This included: PV system safety procedures, description of the system, instructions on PV array and battery maintenance, troubleshooting and replacement/repair of components in the system, semi-automatic and manual operation of the PV system controls in the event of failure, routine maintenance and the operation of the mill. In three days, the trainees were able to operate the mill and develop their grain and flour-handling routines.

Minor problems which occurred in the system were resolved satisfactorily by local personnel and shut-down time was minimal. Advice on problem-solving was obtained with LeRC personnel through telephone calls, letters and cables.

III. SOME ECONOMIC CONSIDERATIONS

Cost-benefit analysis based on data collected from this project would prove meaningless for the following reasons. First, one of the activities (water pumping) was not intended to collect revenue. Second, as a pilot scheme many concessions were made which would not apply to real systems. Third, technical testing of the equipment was being undertaken simultaneously, which necessitated the installation of instruments and regular recording of data. Fourth, major changes had to be made to the system during the period under review.

In this section, an attempt is made to compare this photovoltaic system with alternative methods of providing electricity for the above village applications.⁵ These are:

- by extension of the main power grid to supply the village;
- by using decentralised diesel-generating sets.

These alternatives are compared assuming an annual electricity consumption of 3.6 kW for 8 hours a day, 200 days a year or 5.760 kWh.

Grid Extension

The cost of grid extension can be calculated from the formula:

... (1)

where

G = total cost per kWh consumed
U = change per unit (kWh) supplied by grid
K = length of transmission line
C = unit cost per km of transmission line
A = annual typical capital charge

12 per cent to cover depreciation and interest on capital

P = power consumption (in kWh per annum).

As can be seen from (1) G is directly proportional to the length of the extension k and inversely proportional to the annual power consumption P. Taking line costs C = US$4,000/km and unit electrical energy costs U = US$0.10 per kWh, the cost of electricity from grid extension from Ouagadougu (190 km away) to feed the water pump and mill at Tangaye⁶ would be approximately US$16.0 per kWh consumed.

Figure 19.4 shows the variation of cost of grid extension as a function of the distance from the nearest supply point for an annual consumption of 5,740 kWh.

Small-scale Decentralised Diesel Generation

Diesel sets incur high maintenance and fuel costs, and in remote rural areas, pose problems of fuel supply. The marginal (fuel) cost alone for generating power from diesel sets is approximately US$0.5 per kWh of electricity generated assuming the cost of diesel at US$0.44 per litre (cost in Upper Volta in 1979). In general, the fuel cost per kWh of power generated can be estimated as:

C_kWh = 0.23 C_d (approximately)
where C_d is the cost per litre of diesel in us dollars.
Fuel consumption is assumed as 4.5 litres per hour per 20 kW.

In 1979, capital cost of diesel-generating sets ranged between US$2,700 for a 3 kVA generator and US$5,500 for a 10 KVA generator, plus US$1,000 for installation. Cost per unit of electricity generated by diesel can be calculated as

where

C - total capital cost of equipment
t - annual depreciation and interest (30 per cent)
P - total annual consumption of electricity (kWh)
m - annual maintenance and repair costs taken as equal to the annual capital cost

Duplication of supply is also recommended by manufacturers⁷ of diesel-generating sets.

The cost per kWh of electricity generated is estimated at US$1.50 assuming duplication of the 5 KVA diesel generator, a five-year life of equipment, 10 per cent interest rate and a 3.6 kW load (5,760 kWh of electricity annually). Fig. 19.5 shows the cost/annual consumption curve of a 5 kVA diesel generator plant. While for the size of load initially installed in 1979 (1.8 kW) the cost of electricity generated by diesel sets was estimated to be US$2.50 per kWh,⁸ with the increased load demand of 3.6 kW, (due to the extension in 1981) the cost per kWh of alternative diesel-generated electricity dropped by about one dollar.

Photovoltaic Power Generation

Photovoltaic costs are divided into

(a) module cost

Figure 19.4. Cost of extension of grid supply to supply 5760 kWh annually

Figure 19.5. Comparison of costs of different methods of electrical power generation for varying annual loads

(b) balance of system costs (BOS)

The module cost is the cost of the set of environmentally-protected solar cells while balance of system costs include cost of the array structure, electrical wiring, control and projection circuits, instruments, enclosures and buildings, storage batteries, racks and maintenance equipment.⁹ In 1978 the module cost was estimated at US$14.70 to US$18.30 per Wp, while BOS costs were estimated to lie between 57 per cent and 46 per cent of total system cost.¹⁰ In 1978, the unit BOS cost was estimated between US$11 and US$17 per peak watt thus implying a total system cost ranging between US$25.70 and US$35.30.¹¹

In 1979, the Tangaye system cost US$50,400 for 1.8 kW peak (or US$28 per peak watt). The annual capital cost was calculated to be US$3.29 assuming a 10 per cent interest rate, and a 20-year life of the equipment. Assuming about 15 per cent of annual capital cost for maintenance and replacement, the annual maintenance and replacement cost was US$0.44 per peak watt. The annual energy output per peak watt (at 8-hour daily utilisation for 200 days a year) is 1.6 kWh. The cost per kWh of electricity generated was thus estimated at US$2.33.¹²

IV. CONCLUSIONS

The Tangaye project achieved its objective to demonstrate the potential for the use of solar cells as a power source for common village tasks in remote areas. Replicability of the system would however depend on a number of factors.

The cost of electricity generated by photovoltaic cells is by no means low. Although it can be argued that the benefits of rural electrification cannot be overestimated, one is forced to ask whether they warrant such high costs. Even if the price of photovoltaic cells drops to US$0.61 per peak watt the calculated cost per kWh of electricity would be US$0.47 at 1978 prices. Although this compares favourably with other methods of power generation it is still considerably high for poor subsistence villagers.¹³ Thus, solar and any other type of village electrification scheme would require some degree of subsidy to ensure their widespread use in the foreseeable future.

Where a rural electrification scheme is being planned it would be desirable to make comparisons between the different methods of providing power. In Figure 19.4 it is shown that for the present annual load in Tangaye (5,760 kWh) extension of the grid over a distance greater than 27 km would prove less economical than photovoltaic power generation within the village. Figure 19.5 compares the three methods for the provision of electrical power assuming Tangaye to be 50 kilometres from the nearest power supply point. From this figure, it can be seen that for the load at Tangaye it is less economical to extend the grid but more economical to use a 5 kVA (in duplicate) diesel generator. Breakeven point between extended grid supply and solar-generated electricity is 10,750 kWh annually and between diesel and solar-generated, 3,000 kWh.

The calculations done in this chapter assume 1978 prices for the equipment.¹⁴ It is obvious that a lot would depend on the cost of photovoltaic systems. In 1983, costs did not decrease as predicted in most of the literature, but as research efforts intensify and costs are lowered the breakeven point for solar versus other forms of power generation would become greater. This means that solar electricity would be the most viable (of the three compared here) for higher annual consumption.

The unpredicted open-circuit failure of the solar arrays due to thermal stresses points to the need for more vigorous field-testing of solar cells. It is only under field conditions that unforeseen problems can be detected and corrected.

The studies of the social impact of village power supply once more highlight the unquantifiable benefits of village power supply - better hygiene and health, reduced burden, and increased rural construction activities. They also demonstrate the willingness of inhabitants of this African rural village to participate actively in such a scheme.

At present, it would seem that low power applications could derive immediate benefit from solar electricity. Thus telecommunication receiver stations, educational TV sets and radios, refrigeration for drugs and vaccines in the village clinics, deserve top priority.

Solar electricity supply in remote villages would be particularly useful when diesel is scarce and essential services need to be powered by electricity. In these cases, solar electricity might be the only feasible method of power production.

NOTES AND REFERENCES

1. William J. Bifano, Anthony F. Ratijczak, and James E. Martz: “A photovoltaic power system in the remote African village of Tangaye, Upper Volta”, in NASA Technical Memorandum 79318, Washington, DC, 1979.

2. James E. Martz, Anthony F. Ratijczak and Richard De Lombard: “Operational performance of the photovoltaic-powered grain mill and water pump at Tangaye, Upper Volta”, in NASA Technical Memorandum 82767, Washington, DC, February, 1982.

3. Allen F. Roberts: A final evaluation of the solar impact of the Tangaye (Upper Volta) solar-energy demonstration, USAID Contract AID 686-089-80, 24 September, 1980; Social impact of the Tangaye (Upper Volta) solar energy demonstration: A summary report, USAID/Afr 0000-00-1045-00, 1981; and An update of the socio-economic impact research on the Tangaye (Upper Volta) solar energy demonstration, NASA Lewis Research Center, Cleveland, Ohio, 1982.

4. James E. Martz, et al., op. cit.

5. These do not include site-specific sources such as wind and hydro power.

6. If other villages are to be supplied on the route this cost would be reduced. Also the nearest utility supply point may not be as far as Ouagadougu.

7. Most single rural generating sets are only operated intermittently four hours a day.

8. William J. Biffano, et al., op. cit.

9. To these could be added the cost of inverters if system is to supply A.C. power.

10. Module peak power as determined for 60 degrees Centigrade cell temperature; 100 mW cm^-2 solar insolation, measured at 15.8 volts.

11. Louis Rosenblum, William J. Bifano, Ferald F. Hein and Anthony F. Ratijczak: “Photovoltaic power systems for rural areas of developing countries”, NASA Technical Memorandum 79097, revised, Washington, DC, February, 1979.

12. If this cost of the PV system (total array and BOS costs) falls to US$10.45 per peak watt then the cost of PV-generated electricity would drop to US$0.86.

13. Cost of electricity to urban dwellers was US$0.10 per kWh in 1978.

14. At current prices the cost per peak watt of photovoltaic systems (array, batteries and controls) is estimated at US$15.