Maintenance of irrigation systems in the developing world is often poorly carried out and is generally under-funded. Governments in Asia typically allocated the equivalent of $6-12/ha at 1992 prices, whereas recommendations made in various countries suggest that $13-40/ha would have been needed to stem progressive deterioration of the systems. The Note demonstrates the positive impact of maintenance on project economics and lifetime. Many institutional and policy issues affecting maintenance were also identified.

1 The analyses showed that in all cases good maintenance cost less than the combined costs of poor maintenance and early rehabilitation whilst at the same time safeguarding output which would be lost under a poor maintenance regime. A discount rate of 10% was used in the base case scenarii. The outcomes remained valid under a series of sensitivity tests.

The annual net benefits resulting from good maintenance were in the range $50-$100/ha, depending on the cropping pattern and scheme, under the base case conditions. On a medium-sized project of 20,000 ha, the annual benefits would therefore have been $1-$2 million (1992 prices).

On the other hand, poor maintenance would have required extra investment by government of up to $175/ha discounted over the project lifetime, to cover rehabilitation. The savings resulting from good maintenance on the medium-sized project could have been used to provide a safe supply of drinking water to over 20,000 people. (World Bank estimates of supply costs: $150 per person.).

2 Deterioration of the system may have other serious consequences for the farmer and for the nation. Typically, the water supply and/or the drainage networks become progressively more unreliable, and the distribution of available water increasingly inequitable, as channels fill with sediment and/or weeds. Farmers in disavantaged locations, typically at the tail of a system, may be forced out of agriculture, increasing unwelcome migration to the cities. Environmental problems may also affect human health. Such outcomes are real, but they are intangibles for any given project, so they were not included as disbenefits in the economic analyses.

3 It might seem economically efficient to fund only such maintenance as is necessary to ensure that the design lifetime is achieved. In the present state of knowledge, it is only possible to predict that a given low level of maintenance is inadequate to stem progressive deterioration. It is currently unrealistic to attempt to select a level of spending which will allow 'managed' decline.

Recommendations:

The consequences for national economies of the prevailing 'Build-Neglect-Rebuild' policies need to be more widely recognized and publicized.

1. Firm information on maintenance spending and its impact on economic performance and project sustainability is needed to guide government policies on water charges, turnover and target-setting.

2 Better methods of identifying maintenance works and putting priorities on those with principal impact on performance are needed, so as to make best use of available resources.

3. Many governments aim to turn over the O&M of parts of systems to farmers, without knowing the consequences for the system. Guidelines identifying the circumstances under which turnover is likely to have positive, or negative impact on the infrastructure, and the support services which farmers will require, are badly needed.

4. There is no guidance available to governments on the best way to restructure inefficient institutional arrangements so as to improve maintenance. Turnover policies which leave O&M of parts of the system to unreconstructed agencies are unlikely to be effective.

5. New or rehabilitated systems are still generally designed as if O&M will remain the responsibility of governments. Low maintenance designs, suited to O&M by farmers, are increasingly needed.

There is a move throughout the world to turn over the operation and maintenance of major parts of public irrigation systems to farmers. There are some success stories, like Mexico, where maintenance has improved. However, Mexico is unrepresentative of the large, Asian smallholder schemes growing low value cereal crops. There are many uncertainties attached to turnover in such circumstances. There is a danger that the cycles of 'Build- neglect-rehabilitate', which have become typical of agency-managed irrigation schemes, will become even shorter under farmer-management.

Informed policies to tackle the maintenance deficit are badly needed.

J.C. Skutsch (1998) Maintaining the value of irrigation and drainage projects. OD/TN 90. HR Report.

J.C. Skutsch (1999) Realising the value of irrigation and drainage maintenance. IPTRID Issues paper no2. March 1999.

A synthesis and elaboration of the content of OD/TN 90 has recently been prepared for IPTRID with the aim of extending exposure to the issues in the irrigating and funding communities.

Tel: +44 1491 835381
Fax: +44 1491 826352
Email: http://www.dfid-kar-water.netdfid-kar-water@hrwallingford.co.uk