An optimisation approach for real time water allocation in complex irrigation distribution systems has been developed. Quadratic programming provides an efficient solution to the optimisation problem defined. Through simulation experiments on an irrigation system in Bali, Indonesia, it has been demonstrated that the optimisation approach is very effective, and can significantly improve equity in water allocations.

The optimisation approach has been integrated with a soil moisture balance model to permit effective evaluation of drainage returns. The soil moisture balance routines may also be used to evaluate irrigation requirements in real time. For the Lower Ayung system on which the models have been evaluated, there was no benefit in computing irrigation demands in real time. On large systems with little rice cultivation, real time demand calculation may be more valuable, and remains an option in the model.

The use of an objective function based on equitable yield response does not produce equitable financial returns. With complex cropping patterns in which the financial values of crops varies significantly, an equitable yield response function is inappropriate. Introduction of a financial factor to the objective function still does not achieve the desired result, and further work is required to develop a function that will result in an equitable financial return per unit of water diverted.

At this stage, pilot testing and demonstration of the effectiveness of the equitable water allocation function in practice would be more appropriate than further development of yield based or financial return based objective functions. Operating a system with objective functions based on yield or financial returns would be difficult to monitor and explain to farmers. A simpler system based on equitable water allocation, with cropping decisions left to farmers would be more appropriate.

A user interface has been prepared to assist in the creation of network files for the model, and an outline given of how other aspects of a prototype system should be developed.

The optimisation approach could form the core of a decision support system for large-scale irrigation management. The approach can be used to assist in planning maintenance schedules, as well as in planning rehabilitation and improvements. The required components of a decision support system may vary significantly between irrigation systems. A toolbox of component routines will assist in improving the efficiency with which support systems can be put together. The optimisation routines developed here, along with the network definition software, provide important components of such a toolbox.

1. Wardlaw, R. Computer optimisation for better water allocation. Agricultural Water Management, accepted, September 1998.

2. Wardlaw, R.B., and Barnes, J.M. Evaluating the potential of optimisation in real time irrigation management. ASCE Irrigation Division, submitted July 1998, accepted March 1999.

3. Wardlaw, R.B., Moore, D.N., and J.M.Barnes, 1997. An assessment of the potential of optimisation in real time irrigation management. International Commission on Irrigation & Drainage, Proceedings of the 18th European Regional Conference "Water - An economic good ?". E & FN Spon, September 1997. ISBN 0 419 218408

4. Wardlaw, R.B., and Barnes, J.M., 1996. Real time operation and management of irrigation systems. International conference on new challenges for civil engineers of developing countries in the 21st century. Indian Society of Environmental Management, New Delhi, March 1996.

5. Wardlaw, R.B., 1998. Improved irrigation system planning and management: optimal allocation of irrigation water supplies. ODA TDR Research Project No 6261, Phase III report, June 1998.

6. Wardlaw, R.B., and Barnes, J.M., 1997. Improved irrigation system planning and management: optimal allocation of irrigation water supplies. ODA TDR Research Project No 6261, Phase II report, June 1997.

7. Wardlaw, R.B., & Barnes, J.M., 1996. Improved irrigation system planning and management: optimal allocation of irrigation water supplies. ODA TDR Research Project No 6261, Phase I report, April 1996.

There are two main areas in which the outcomes of this research can be progressed:

i) through a pilot application of the approach to an irrigation system;
ii) through further research and development of objective functions based on crop yield and financial returns.

The research carried out has demonstrated that the techniques developed have significant potential in improving water management in irrigation systems with complex distribution networks. A pilot application, through which real benefits could be demonstrated, and the wider requirements of decision support considered more fully would be very valuable. The demonstration of a real application might also lead to a wider uptake of the techniques developed. Implementation of the approach on the South Lombok irrigation system would be possible through the Water Operations Centre, which is already in existence there. It would be possible to set up a collaborative project with participation from DGWRD and DPMA in Indonesia. A proposal for such pilot testing has been submitted to DFID under the KAR programme.

Tel: +44 131 650 5944
Fax: +44 131 650 7167
Email: Robin.Wardlaw@ed.ac.uk