Management of water resources has become more complex in recent years as a result of changing attitudes towards sustainability and the attribution of greater attention to environmental issues, especially under a scenario of water scarcity risk introduced by climate changes and anthropogenic pressures. This study addresses the optimal short-term operation of a multi-purpose hydropower system under an environment where objectives are conflicting. New optimization models using mixed integer nonlinear programming (MINLP) with binary variables adopted for incorporating unit commitment constraints and adaptive real-time operations are developed and applied to a real life hydropower reservoir in Brazil, utilizing evolutionary algorithms. These formulations address water quality concerns downstream of the reservoir and optimal operations for power generation in an integrated manner and deal with uncertain future flows due to climate change. Results obtained using genetic algorithm (GA) solvers were superior to gradient based methods, converging to superior optimal solutions especially due to computational intractability problems associated with combinatorial domain of integer variables in the unit commitment formulation. The adaptive operation formulation in conjunction with the solution of turbine unit commitment problem yielded more reliable solutions, reducing forecasting uncertainty and providing more flexible operational rules.

[1] Optimal operation models for a hydropower system using new fuzzy multiobjective mathematical programming models are developed and evaluated in this study. The models use (i) mixed integer nonlinear programming (MINLP) with binary variables and (ii) integrate a new turbine unit commitment formulation along with water quality constraints used for evaluation of reservoir downstream impairment. Reardon method used in solution of genetic algorithm optimization problems forms the basis for development of a new fuzzy multiobjective hydropower system optimization model with creation of Reardon type fuzzy membership functions. The models are applied to a real-life hydropower reservoir system in Brazil. Genetic Algorithms (GAs) are used to (i) solve the optimization formulations to avoid computational intractability and combinatorial problems associated with binary variables in unit commitment, (ii) efficiently address Reardon method formulations, and (iii) deal with local optimal solutions obtained from the use of traditional gradient-based solvers. Decision maker's preferences are incorporated within fuzzy mathematical programming formulations to obtain compromise operating rules for a multiobjective reservoir operation problem dominated by conflicting goals of energy production, water quality and conservation releases. Results provide insight into compromise operation rules obtained using the new Reardon fuzzy multiobjective optimization framework and confirm its applicability to a variety of multiobjective water resources problems.Citation: Teegavarapu, R. S. V., A. R. Ferreira, and S. P. Simonovic (2013), Fuzzy multiobjective models for optimal operation of a hydropower system, Water Resour. Res., 49,[3180][3181][3182][3183][3184][3185][3186][3187][3188][3189][3190][3191][3192][3193]

Optimal operation models for a hydropower system using partial constraint satisfaction (PCS) approaches are proposed and developed in this study. The models use mixed integer nonlinear programming (MINLP) formulations with binary variables. The models also integrate a turbine unit commitment formulation along with water quality constraints used for evaluation of reservoir downstream water quality impairment. New PCS-based models for hydropower optimization formulations are developed using binary and continuous evaluator functions to maximize the constraint satisfaction. The models are applied to a reallife hydropower reservoir system in Brazil. Genetic Algorithms (GAs) are used to solve the optimization formulations. Decision maker's preferences towards power production targets and water quality improvements are incorporated using partial satisfaction constraints to obtain compromise operating rules for a multi-objective reservoir operation problem dominated by conflicting goals of energy production, water quality and consumptive water uses.

No abstract

No abstract