Determination of the optimal location of wells in aquifers with an accurate simulation-optimization model based on the meshless local Petrov-Galerkin

Abstract

In this research, a simulation-optimization model is utilized to determine the optimum location of wells on two unconfined aquifers (steady condition and transient state) with different boundary conditions. Meshless local Petrov-Galerkin (MLPG) and particle swarm optimization (PSO) are employed for the simulation model and optimization algorithm, respectively. MLPG model is first verified by comparing results with analytical and finite difference solutions. The obtained results fromMLPG showmore accuracy than FDM. In PSO, the objective function is defined as the summation of the absolute difference between water table before and after extraction in each node. In the first aquifer, results show that the content of objective function before using simulation-optimization model is 61 m; however, it reaches 17 m when the optimal location is selected. The value of objective function after using simulation-optimization model has 27% reduction in value. The second aquifer has ten extraction wells and investigated in two cases with different boundary conditions. In both cases, the value of objective function is decreased significantly after applying SO model.