Surrogate Based Simulation-Optimization Approach in Optimal Operation of Waterbody Considering Quality and Quantity Aspects

Abstract

In this research, deriving optimal reservoir operation in selective withdrawal scheme considering quality and quantity aspects has been studied. Surrogate based simulation-optimization approach (SBSOA) has been applied to improve downstream water supply and enhance reservoir outflow water quality. CE-QUAL-W2 as the hydrodynamic and water quality simulation model in the river-reservoir system and multi-objective particle swarm optimization (MOPSO) algorithm as an efficient tool have been applied in simulation-optimization (SO) approach. To overcome the computational burdens of multiple calls of CE-QUAL-W2, as a numerical high fidelity model, various surrogate models have been developed to simulate reservoir outflow water quality parameters (DO, NO3, PO4, BOD, and Fe). The developed surrogate models and mass balance model have been coupled with MOPSO algorithm in SBSOA. In this study, the water quality objective function is defined as water quality index (WQI), which integrates various water quality parameters, DO, NO3, PO4, BOD, and Fe. The proposed approach has been applied in Ilam river-reservoir system to derive optimal reservoir operating strategies in the selective withdrawal scheme. The results show suitable efficiency and accuracy of the developed surrogate models in approximation of various water quality parameters compared with CE-QUAL-W2 simulation results (the approximation error of DO, NO3, PO4, Fe, and BOD has been respectively 3%, 1%, 2%, 2%, 3%). The studies indicate enhancing reservoir outflow water quality is consistent with downstream water supply satisfaction. It means, the increasing of reservoir outflow rate leads to the reservoir detention time decreasing, pollutant settling rate reductions, and chemical/biological reaction attenuation.