Optimizing Solution in Decision Supporting System for River Basin Management Consisting of a Reservoir System

Abstract

Decision support systems tackle problems and require systematic planning. They consider physical data, hydrological data, and sediment levels to achieve efficiency and adaptability in various situations. Therefore, this research aims to identify alternative engineering choices for the management of a river basin with a single reservoir system. Optimization techniques, including marine predator algorithm (MPA), genetic algorithm (GA), genetic programming (GP), tabu search (TS), and flower pollination algorithm (FPA), were applied to find the optimal reservoir rule curves using a reservoir simulation model. The study focused on the Ubolratana Reservoir in Thailand’s Khon Kaen Province, considering historic inflow data, water demand, hydrologic and physical data, and sedimentation volume. Four scenarios were considered: normal water scarcity, high water scarcity, normal excess water, and high excess water. The optimal rule curves derived from the reservoir simulation model, incorporating sedimentation and hedging rule (HR) criteria, were found to be the best engineering choices. In the normal and high water scarcity scenarios, they minimized the average water shortage to 95.558 MCM/year, with the lowest maximum water shortage 693.000 MCM/year. Similarly, in the normal and high excess water scenarios, the optimal rule curves minimized the average excess water, resulting in a minimum overflow of 1087.810 MCM/year and the lowest maximum overflow 4105.660 MCM/year. These findings highlight the effectiveness of integrating optimization techniques and a reservoir simulation model to obtain the optimal rule curves. By considering sedimentation and incorporating HR criteria, the selected engineering alternatives demonstrated their ability to minimize water shortage and excess water. This contributes to improved water resource management and decision-making in situations of scarcity and excess.