April 2005

William Austen Bradbury

Institute of Irrigation and Development Studies, School of Civil Engineering and the Environment, University of Southampton, Southampton, SO17 1BJ, United Kingdom.

ABSTRACT

The use of solar energy for water supply systems is not a new concept, but a design solution that has had little experience in a developing world context. This article explores the outcome of a study of 3 such projects carried out in southern Honduras, Central America, in 2000 and 2001. It examines the diverse aspects of solar-powered rural water pumping systems in relation with their sustainability and suggests a series of recommendations for the implementation of future projects, focusing on the areas found most important: community administration; technical assistance; suitability of the project; design of the water system; and training and institutional support.

Evidently there is more work that needs to be done on ensuring the sustainability of projects of this type; photovoltaic water pumping can be a viable solution to a community’s needs, but the requirements of the technology must be understood and fulfilled.

INTRODUCTION

Approximately half of the world’s population live in rural areas of developing countries (Omer, 2001) where poverty is often most severe. For communities away from the electricity grid, solar power has the potential to solve many problems by providing clean, renewable and reliable energy. Owing to the high initial capital costs of solar panels and their control devices, solar energy systems are usually used for communal purposes in development. Improving the water supply to a poor community provides a string of benefits and the use of solar energy is often the most economic solution in remote areas with no electricity (FAO, 2000).

Solar-powered water pumping systems have been in existence for many years and have been installed in many countries throughout the world, yet development in the field has been slow. In 1978, the World Bank set a target to install 10 million photovoltaic pumping (PVP) systems by the year 2000 (Barlow et al, 1993); however, by 1998 only 60 000 units were estimated to have been sold (Short et al, 2002a). It is estimated that millions of water pumps are still needed in areas far from the electric grid (Posorski, 1996).

This paper looks at the current situation in the field of solar-powered water pumping and discusses the outcome of a study carried out between July and September 2003 of three such projects implemented by the humanitarian aid organisation Action Against Hunger.

TECHNICAL SOLUTIONS TO RURAL WATER SUPPLY

A community located beneath a mountain spring or stream may be able to obtain its water supply from a gravity-flow distribution system, the solution which is most sought after as it requires no additional energy supply and so often relatively cheap to operate and maintain. Similarly, in areas with high rainfall, an energy-free system can be used to harvest rainwater.

Communities close to the electricity grid, with a water table at an acceptable depth, may use an electric pump to supply water. For communities away from the electricity grid, a wind-powered pump may be a viable option if the average wind speed is above 3 m/s (Omer, 2001); or, if insolation is above 15 MJ /m² / day, a solar pump could be considered. Where the climatic conditions are not appropriate, biomass or diesel pumps can be used, as long as there is an adequate fuel supply – if not, hand pumps may be the only option.

SOLAR-POWERED WATER PUMPING OVERVIEW

Solar technology solves the problem of providing energy to remote rural areas where the electricity grid does not reach. Provided that there is adequate insolation, solar panels can be used to pump groundwater to a reservoir which feeds domestic outlets. Compared to the use of a diesel-powered pump, the initial capital cost of a solar-powered system is high (in Honduras, typically around $40 000 (Verani, 2000)) but the running and maintenance costs are low (there is no fuel to pay for). PVP systems for water supply are generally designed for a lifetime of at least 10 years (Enersol, 2002), if not 15 or 20 (SNL, 2001), over which period the system can become very cost-effective.

Fig. 1 is a simple diagram of a PVP system, showing its main components. The reservoir is used to provide water when there is no sunshine to power the system and should have the capacity to provide 2-5 days’ water consumption, depending on climatic conditions. Batteries can be used instead to continue pumping water without sunshine, but a reservoir is more reliable and much more economical.

Fig. 1: A typical solar water pumping system

The controller is used to increase system efficiency by 10-15%. If the pump motor is AC, this must include a transformer to convert the DC signal from the panels to AC. If the pump motor is DC, the controller is optional.

The energy needed to pump the water, and hence the size of the system, will depend upon: the design flow; the depth of the water source below ground (which will vary); the height of the storage tank; and the frictional losses in the pipe-work. System efficiency is generally low (40-50%) because the energy goes through several different stages from its solar input to the water output (Short et al, 2002b). The various components have to be carefully matched for optimum efficiency. Choosing which components to include in the design is a matter of balancing cost, service, efficiency and reliability. There is much talk of the need for ‘matched systems’ but manufactures have not followed this up (ibid.).

Routine maintenance costs of PVP systems are very small – the PV panels require little or no maintenance and the controller simply needs to be maintained well-sealed to avoid dust, water and insects (SNL, 2001). Pumps should be cleaned regularly to avoid blockages. Centrifugal pumps require little additional maintenance, but positive displace-ment pumps need to have their diaphragms replaced every 2-3 years and their piston seals every 3-5 years. Pump motors with brushes must have them replaced after a few years, depending on their use, but this is usually a simple operation. Brushless motors do not require maintenance and may last 10-20 years (SNL, 2001).

Other routine maintenance required is similar for almost all other water systems: cleaning of the reservoir, cleaning of private water tanks, releasing of excess air in pipes, etc.

TABLE 1: ADVANTAGES AND DISADVANTAGES OF SOLAR WATER PUMPING SYSTEMS

Advantages	Disadvantages
Renewable energy – environmentally friendly Generally very reliable with long design life Little or no operating costs (no fuel needed) Few maintenance costs	Reliant on the weather (much insolation needed) High capital costs Complex electronics, technical support required Risk of theft Repairs may be expensive Little experience

RESEARCH REVIEW OF PHOTOVOLTAIC PUMPING SYSTEMS

Previous research on the sustainability of PVP systems is limited, but certain affirmations exist and were taken into account when designing the study and analysing its results.

Choice of Community

The choice of solar power as the design solution for a particular community must only be made after other solutions have been evaluated and eliminated (Enersol, 2002) – it is the water that is required, not the power source (Short et al, 2002b). If PVP is applied where alternatives are not considered, it may result in an unsustainable solution and a waste of money (ibid.).

Enersol Associates (2002) state that communities selected for PVP projects should be at least 1 km from the electricity grid, as nearer than this it is usually cheaper to extend the electricity grid and buy a transformer. They also state that it is economically feasible for communities of up to 200 families (~1000 people) whereas Münger (2000) advocates from 300 to 2000 people. Other prerequisites are that the community should have difficult access to water (far away or poor quality) and have access to a solar technician (Verani, 2000). Villagers must agree to conserve water and to pay a monthly flat fee for the service (Verani, 2000).

Water Tariff

The existence of a system for water management, including fee collection, is a condition for a PVP project (Kaunmuang et al, 2001), the development of which is a challenge for both project workers and community leaders (Verani, 2000). A fund is needed to pay for maintenance costs and to replace and repair equipment. Verani (2000) suggests that there is a willingness to pay for water service far above conventional notions and existing tariffs, but Posorski (1996) reports that revenues actually collected from tariffs rarely cover operating costs because poor service and lack of response from water authorities deters users from paying.

Community Development and Institutional Support

The social and institutional interaction necessary for a PVP system may be more important for the success of a project than technological factors (Omer, 2001). Local tradition, personnel, requirements and skills must be understood so that villagers can be involved in the planning and constr-uction of the system, with the aid of outside professional support (ibid.).

A local management committee is required to decide how to share out water equitably, to handle finances and to manage routine maintenance (Omer, 2001). This committee should be made up of democratically elected individuals committed to voluntary community service, whereas routine functions should be performed by separate individuals who are paid for their service (Johnson, 2003).

The involvement of local private businesses in a PVP project helps towards sustainability (Verani, 2000) and post-sale and post-installation support is a requirement for such systems (Ley, 2003). In Mauritania, 5-year maint-enance contracts with a private Mauritanian company comprised of a full guarantee for solar installation facilities including repairs and a yearly visit to the system for between US$ 260 and US$ 900 per year (Münger, 2000).

Technology

The solar technology involved in PVP systems is advanced and includes complex electronic systems that cannot be repaired at village level. The solar array is the most reliable component of the system and may have a guarantee of up to 25 years. Panel failure is rare – Münger (2000) reports that, over 4 years, there were no panel failures amongst 63 systems studied, but the rate of pump failures was 1.5% per year. Guarantees for pumps usually last for just 1 or 2 years but most pump failures are normally due to lack of simple maintenance rather than defects in the pump itself.

A study in Thailand of almost 500 PVP systems found that 32% of breakdowns were due to pump motor failures and 19% by inverter failures (all systems used AC pumps). Failures of pump motors were related to sediment build-up which could have been rectified by the villagers. The controller failures, however, occurred from lightning shocks, inadequate protection from moisture and over-heating (Kaunmuang et al, 2001). Short et al (2001) report on two more studies, both of which found inverter failures of around 20% and Posorski (1996) declares that the inverter is “the most sensitive system component” after finding similar results in 89 PVP systems. Posorski claims the controller’s maximum power point tracking is not reliable and Short et al even suggest one option as to do away with the controller altogether in order to increase reliability.

Security

Security is an important issue in solar systems because solar panels are valuable items and can too easily be stolen. Bannister (2000) reports on a variety of security methods: electric fences; alarm systems; locating panels in a community member’s yard; using latest designs of panels which are robust and hard to steal. An alternative is to employ a vigilance system whereby community members take turns in guarding the equipment, as has been adopted in Honduras and the Dominican Republic (Enersol, 2002).

PROJECTS STUDIED

The 3 projects studied were located in the department of Choluteca in the south of Honduras. Table 2 presents some characteristics of the climate in Choluteca (Clarke et al, 1998 and Smith, 1993). These conditions are perfect for solar energy projects.

TABLE 2: METEOROLOGICAL DATA OF CHOLUTECA, HONDURAS

Annual average rainfall	1000 mm
Average max. temperature	34.3 ºC
Average min. temperature	22.9 ºC
Average number of sunshine hours per day	6.9 (Sept) – 9.8 (Feb)
Solar radiation (insolation)	~20 MJ/m²/día (throughout the year )

The projects studied were located in the communities of El Fortín, El Naranjal and Los Achiotes. Each system consisted of: photovoltaic array (mounted on the roof of a hut); controller (inside the hut); electric pump; handpump; hand-dug well or borehole; reservoir; distribution network (with outlets in the grounds of each house). Table 3 above shows the main characteristics of each project.

Each community had a Water Committee (WC) formed as part of the project, to administer the running of the water system. A Plumber, paid by the WC, was responsible for operation and maintenance.

The sustainability of a rural water system is dependent on a large number of factors and hence is a complex issue. It can be split up into different aspects, all interlinked: administrative (or institutional), social, environmental, economical (or financial) and technical (Abrams, 2000). The study used various techniques to obtain information on all these aspects: informal interviews and observation checklists; focus groups (or small meetings) in which a range of community participation tools were employed (see Geilfus (1997 and Wespi et al (2001)); interviews with key members of the communities; semi-structured dialogues and questionnaires with beneficiaries of the water systems.

ANALYSIS AND DISCUSSION OF RESULTS

The results of the study are analysed by the various aspects of the sustainability of the water projects.

Administrative Aspects

The Water Committees were not holding regular meetings, yet these are essential to enable communication with the community and discuss problems. The committee members were generally unclear on their duties and the existence of dual roles and relatives within a committee no doubt negatively affected the group’s efficiency.

The Plumber tended to be the most active person with respect to the water system and was seen as a member of the Water Committee (WC), rather than an employee. He could run the system as he liked, which may not necessarily have been the best way for the beneficiaries. The Treasurer was also able to act freely, resulting in serious problems with accountability. Problems with payment of the tariff were severe in two communities where the Treasurer failed in keeping payments up-to-date and in complying with the rules on cutting off water connections to non-payers.

Due to the absence of a contract or fixed contact for technical assistance, communities experienced longs delays while trying to arrange repairs after breakdowns in the system.

Social Aspects

The effective running of the water system is dependent on the participation of both the WC and the community. There was a lack of interest in the water system and a severe lack of communication; villagers knew little about the manage-ment of the system but could identify a string of problems in the focus groups, some of which they had no basis for.

One of the risks that the communities did not experience problems with was that of theft and vandalism. Despite the inconvenience of spending a whole day every few weeks guarding the system, the beneficiaries of El Fortín and Los Achiotes did not complain and seemed to appreciate the value of their equipment. In El Naranjal, however, there was no security system in place, yet the equipment had suffered no problems. Some inhabitants of this village did not appreciate their system nor the service it provided and even chose to revert to their old water sources instead of paying for the new system.

Social exclusion occurred in the communities with respect to the poorest people and newcomers to the village. Some elderly people and single mothers were excluded from the projects owing to their inability to pay the monthly tariff or contribute to the construction of the system, and newcomers were not invited to connect to it.

Environmental Aspects

The study found no significant environmental impacts of the projects. However, one issue that the projects had not regarded was the handling of wastewater. With the new water supply, villagers used more water than before and hence the amount of wastewater had also risen. If not managed properly, this can cause environmental health problems such as an increase in breeding sites for mosquitoes and hence disease.

Economic Aspects

The values of the water tariffs in the communities studied (40-45 lps) were substantial for poor people who would earn only 50 lps (US $3) a day when they could find work. However, the economic problems encountered were largely linked with administrative and social issues rather than solely with the value of the tariff. In two communities, there were severe problems with payments. A host of factors affected villagers’ opinions of the tariff: level of service; accountability of money paid; knowing cost of repairs; etc. People were willing to pay for their water as long as they were happy with the service they received.

Lack of payments and lack of accountability severely jeopardised the projects’ sustainability, as any fault in the system would prove critical without the economic resources to repair it. The existence of a communal bank account proved to aid accountability considerably, giving the Treasurer the duty to deposit money and the ability to withdraw funds only with the signature of another member of the committee. However, without supervision or meetings to discuss funds, the Treasurer still had the opportunity to manipulate finances.

Technical Aspects

All three systems had weaknesses which meant that the villagers’ water requirements could not be fully met. Water was rationed by the Plumber by filling the reservoir and then distributing the water to sectors of the communities for certain durations. This resulted in a very unequal water distribution; the amount of water a beneficiary received depended on the amount of time the water was distributed to the house, the water pressure at the tap and the capacity for storing water (tanks, barrels or buckets). This had a lot of knock-on effects on the sustainability of the projects, to the extent that some villagers opted out of the system and reverted to retrieving water themselves. Many villagers also used the rain as a supplementary source of water.

Two of the systems had had several problems with their equipment since it had been installed, mainly with the pumps. Technical assistance was expensive (and possibly overcharged) and the communities had to wait long periods in which the handpumps proved indispensable. In Los Achiotes, when visited, the system was not working and reparation had been abandoned by the Water Committee.

Water was chlorinated using a ‘hypochlorinator’, a device which constantly drips a chlorine solution into the reservoir when applied. As the water entering the tank varies according to the sunshine, it is not chlorinated consistently. Nevertheless, the chlorinators were hardly used due to the additional work involved and the cost of chlorine powder.

PRINCIPAL RECOMMENDATIONS

Community Administration

PVP projects need a fully functioning and effective community management system to ensure sustainability. Water Committees should be set up before the PVP project and given duties such as promoting hygiene, saving water or dealing with wastewater. This can be used to assess the motivation and organisation within a community – if proved successful, the PVP project may follow.

In administrating PVP systems, it is accountability and communication that are most important. Training is needed for both the WC and the community on the importance of meetings and the requirements of committee members. With a PVP system, villagers may not understand why they should pay such a high water tariff when the operational costs are so low, so the WC must explain where the money goes and account for it. Targets should be set as to future upgrading of the system (e.g. purchasing additional panels) so that the community continues paying the tariff if there are no major costs (such as repairs) with the system. So that payments are met, the WC should employ a Service Fee Collector to collect the money at monthly meetings and discuss finances with the Treasurer and the community. If payments are not met, the administration must be effective in taking action.

It should be made clear that both the Service Fee Collector and the Plumber are employees of the WC and not part of it. The Plumber must report to the WC and not be allowed to manipulate the system as he or she likes.

Technical Assistance

The availability of technical assistance is a critical factor in the sustainability of solar projects. Taking into account that target populations of solar projects are usually situated in places with limited access and communications and that PVP systems are much rarer than other types of water system, the best way to ensure the technical assistance required is to implement a program of various projects and establish maintenance contracts with a private company, as with the projects in Mauritania (see 3.3). The community thus pays a fixed regular payment and the responsibility for maintaining and repairing the system lies with the outsiders. It should be ensured that the community has a means of communication with the company in case of breakdown.

There is a limited amount of maintenance work that can be carried out by the community itself, such as dismantling the pump to clean it and remove blockages, or doing simple performance tests. This should be promoted and included in training programmes,

PVP as a Solution for the Community

Two crucial and interdependent factors in the implement-ation of projects of this type are the suitability of the project for the community and the community’s acceptance of it. The characteristics of the community itself, rather than just the technical aspects of the design, are very important and will determine the sustainability of the project. The community’s enthusiasm and necessity for a water project must be assessed. Before deciding to use PVP, all other appropriate types of solution must be considered and what a PVP project will offer and what the community will need to contribute must be fully explained in advance.

The criteria for the distance from the electricity grid should not only be defined in terms of the cost of extending the grid to the community, but must consider 10-15 years into the future whereby the electricity grid may have expanded anyway.

Careful calculations of projected finances must be made, considering the costs of running the system, the cost of a maintenance contract (if possible) and the savings needed to replace equipment when guarantees run out. It can then be determined if a community is eligible or not, according to its size and the monthly tariff it is willing and able to pay. Ideally, a community should have the same standard of living throughout so all households are able to pay the same tariff and income must be regular and stable. These requirements are not so important if water meters are used (see 7.4, below).

Design of the Water System

The high capital costs of solar equipment mean that a PVP system will almost always be designed to provide a limited amount of water per person per day. Providing for all the beneficiaries’ wishes (watering the garden, bathing animals, washing the patio, etc) is simply too expensive. Water must thus be conserved, but the Honduran method of rationing it is not equitable.

Water meters connected at the outlet of each beneficiary allow the users to be charged according to the amount of water they use. Tariffs can be set so that a certain ration per month has a fixed, low price, but anything above this limit is charged at a higher rate. In this way, users will be less likely to waste water or use large quantities. The additional cost of the meters and their installation can be set back against the savings made on other system components (Johnson, 2002) such as fewer solar panels. The Service Fee Collector, as employee of the WC, would be responsible for recording meters every month and calculating fees. This method is particularly suitable for economically heterogeneous communities.

Even with water meters, there will almost always be a time when the system does not fulfil the needs of the users – if the sun does not shine for several days, the reservoir will dry up unless it is very large. To avoid the cost of a large reservoir, a PVP system can be complemented with rainwater harvesting. In periods with little sun, it is often likely, in tropical areas, that there will be intensive rain (as in Honduras). If the beneficiaries already have recipients to store water, the only additional capital costs involved are drainpipes to channel water from the roof of each house. Training sessions will be needed to teach the community how to install and how to use their rainwater system.

A method of disinfecting water that is more reliable and effective than the hypochlorinator is needed. Chlorinating water at a household level may prove better as each beneficiary would be responsible for disinfecting his or her own water, removing the dependency on the Plumber. This method would also be more resourceful, as only the small proportion of water that needs to be disinfected (that used for consumption and personal hygiene) would be treated.

Another recommended alternative is to use solar water disinfection. It is a very simple practice and can kill 99.9% of microorganisms if the water is heated to 50-60°C for one hour (EAWAG, 2002). This is also done at a household level, using transparent plastic bottles and avoids the cost of buying chlorine.

Training and Institutional Support

During the implementation of a project, the training given to a community should serve to ensure the project’s sustainability. Training is needed in many areas and should be carried out in a way that actively involves the community. A select group of people, including, but not limited to, the members of the WC, should be trained by the institution so that they can then relay what they have learnt to other members of the community in further sessions (with the supervision of the institution).

However, training only at the beginning of the project is evidently not sufficient, as people forget what they have learnt or committee members change and knowledge is not passed on to new members. Project sustainability will be considerably increased by follow-up assessment and training from institutions at intervals (e.g. once a year) after the project is completed. This could be achieved by coordinating projects with other organisations concerned with water supply or solar energy, including those concerned with local or national government strategies.

If more than one organisation is involved in a project, the roles of each organisation must be clearly defined so that the community knows what to expect from each one and who to contact if necessary.

ACKNOWLEDGMENTS

The author would like to thank the following for enabling and assisting in this successful research project:

Pamela Ortega; Luis Hebrero; Ben Fawcett; Catherine Wrigley; Eric Johnson; Kristin Bradbury; Action Against Hunger; and the communities of El Fortín, El Naranjal and Los Achiotes.

Special thanks to the following organisations for their financial support:

Charles Rayner Educational Trust; Douglas Bomford Trust; The Royal Academy of Engineering (UK).

REFERENCES

1. Abrams L. 2000. Understanding the Sustainability of Local Water Services. Water Policy International.

http:/www.africanwater.org/sustainability.htm

1. Action Against Hunger, 2000-1. Various archive project documents.

2. Bannister M. 2000. Solar power for community water supply. Proc. of 26th WEDC Conf.–Water, Sanitation and Hygiene: Challenges of the Millennium, Dhaka, Bangladesh, pp. 311–314. http://www.lboro.ac.uk/wedc/
   papers/26/W%20-%20Water%20Provision/bannister.pdf

3. Barlow R, B McNelis, A Derrick. 1993. Solar Pumping: An Introduction and Update on the Technology, Performance, Costs and Economics. World Bank Technical Paper No. 168. Intermediate Technology Publications Ltd., London. http://www-wds.worldbank.org/servlet/
   WDSContentServer/WDSP/IB/1999/04/28/000009265_3970311124155/Rendered/PDF/multi_page.pdf

4. Clarke D, Smith M and El-Askari K. New software for Crop Water requirements and Irrigation Scheduling, J Int Commission Irrigation and Drainage Vol 47 No 2. 1998.

5. Davila A. 2003. Water and Sanitation Promoter, Action Against Hunger (Choluteca). Personal communication..

6. EAWAG (Swiss Federal Institute for Environmental Science and Technology). 2002. Solar Water Disinfection – A Guide for the Application of SODIS.
7. Enersol Associates. 2002. Recommendations and Guidelines for Photovoltaic Powered Community Water Systems Based on Experiences in Honduras and the Dominican Republic.
8. FAO (Food and Agricultural Organisation of the United Nations). 2000. FAO : el potencial de los sistemas de energía solar en las zonas rurales a menudo sigue sin explotar. Waternunc.com.
9. Geilfus F. 1997. 80 herramientas para el desarrollo participativo: diagnóstico, planificación, monitoreo, evaluación. IICA-GTZ, San Salvador, El Salvador.
10. Johnson E. 2002. A Measured Step Toward Sustainability For Rural Community Water Supply: One Metering Strategy That Works.
11. Johnson E. 2003. Advisor for Community-Oriented Basic Service Delivery. Personal communication.

12. Kaunmuang P, K Kirtikara, R Songprakorb, S Thepa, T Suwannakum. 2001. Assessment of Photovoltaic Pumping Systems in Thailand – One Decade Experience. Solar Energy Materials & Solar Cells, Vol 67, pp. 529–534. http://www.sciencedirect.com/

13. Ley Lum D. 2003. Office of Renewable Energy, Sandia Laboratories. Proc. of Photovoltaic Systems