Article ID: | iaor20022433 |
Country: | United Kingdom |
Volume: | 8 |
Issue: | 4 |
Start Page Number: | 439 |
End Page Number: | 464 |
Publication Date: | Jul 2001 |
Journal: | International Transactions in Operational Research |
Authors: | Vuuren J.H. van, Grndlingh W.R. |
Keywords: | artificial intelligence: decision support |
Decisions regarding good release strategies for large open air reservoirs may be surprisingly complex, even for experienced reservoir managers. This is due to the dual requirement that water levels should neither be kept too high (resulting in excessive evaporation) nor too low (resulting in an inability to supply water during particularly dry periods). Decisions do not only affect the reservoir's efficiency in the short term, but may also do so in the long run, and this significantly complicates matters. A new decision support system called ORMADSS (for Optimal Reservoir Management Active Decision Support System) is discussed here. This system is designed to aid medium-scale reservoir managers in their complicated decision-making processes. The visually attractive and informative computer-implemented system is based on determining an optimal release strategy for years of average climate and then attempting to steer release strategies for non-average years in such a manner that the reservoir content approaches the optimal level for average years, while still satisfying legal and environmental constraints, as well as demand set by reservoir users. The success of ORMADSS is discussed in a special case study regarding its implementation at Keerom Dam, the second-largest privately-owned open air reservoir in South Africa.