Multi-objective conflict resolution optimization model for reservoir's selective depth water withdrawal considering water quality.

This paper develops a multi-objective conflict resolution simulation-optimization model based on a leader-follower game to resolve conflicts between different water users while optimizing water quality in the river through selective depth water withdrawal from the reservoir. Iran Water Resources Management Company (IWRMC), given the nature of the power distribution in this region, is selected as leader, and agricultural, domestic, and industrial water users are selected as followers. Nash-Harsanyi bargaining theory is used as a nested model in this general framework to model competition between followers. The proposed selective withdrawal approach considers four reservoir outlets, located at 120, 145, 163, and 181 m above sea level. Water withdrawal from multiple outlets addresses reservoir thermal stratification and water quality. Temperature and water quality are simulated based on different possible scenarios of reservoir inflow and release using a calibrated CE-QUAL-W2 model. Five artificial neural network (ANN) surrogate/meta models are then trained and validated based on CE-QUAL-W2 model results for each water quality variable. Subsequently, these validated surrogate models are coupled with the NSGA-II optimization model, which along with the utility functions of different stakeholders, constitute the building blocks of our conflict resolution multi-objective optimization model. Finally, three decision-making methods, namely AHP, PROMETHEE, and TOPSIS, are utilized to choose the superior compromise solution. Our results show that water withdrawal from multiple reservoir outlets ensures optimal water allocation to different stakeholders while satisfying the desired water quality criteria. In this study, the top outlet (181 m) has desirable quality, and the IRWQISC water quality criterion at the top and deepest outlets are highest and lowest, respectively.