Evaluation of Stakeholder Utility Risk Caused by the Objective Functions in Multipurpose Multireservoir Systems

Abstract

Optimal operation of reservoir systems is a challenging task in water resource management, involving stakeholders with different utilities. These utilities are conflicting, especially in a multireservoir system when the number of stakeholders increases compared to a single-reservoir system. A reservoir system with hydropower energy generation and supplying downstream demands is an example of these challenges. An appropriate objective framework can improve mitigation of stakeholder conflicts. This paper considers a 1-year optimal operation of a multireservoir system, with two parallel upstream reservoirs and a downstream reservoir in series. System objectives are hydropower energy generation as well as supplying downstream demands. Existing conflicts of this system are determined in single-objective reservoir operation models. The mathematical form of objective functions can limit utility risk in multiobjective problems. To overcome impacts of existing conflicts, a simulation model has been coupled with the weighting method and Nash equilibrium in multiobjective models. Results show that although the calculated objectives using the Nash equilibrium are less (worse) than the maximum (best) values of the objective function with the largest weight for the first (highest) objective priority using the weighting method, in two parallel reservoirs the values of the objective function are larger (better), about 70.83 and 54.17%, respectively, for hydropower energy generation and supplying downstream demands objectives. Moreover, these objectives calculated by the Nash equilibrium are 58.33 and 50% better, respectively, than same values by the weighting method in series reservoirs. Thus, utilities were significantly balanced and increased from the worst (minimum) utility values by the Nash equilibrium. The Nash equilibrium increases considerably the reliability value for all stakeholders at the same time.