Optimizing Pump-and-Treat Method by Using Optimization-Simulation Models

Abstract

The goals of this research include investigating the efficiency of the finite element method and its combination with meta-heuristic algorithms to solve the optimization problem of the pump and treat (PAT) system. In this research, the hybrid optimization-simulation models were developed to determine the optimal groundwater remediation strategy using the pump and treat (PAT) system. The results indicated that when we consider minimizing the contaminant in groundwater at the end of the remediation period as the objective function, locating the pumping wells in the path of the contaminant flow and close to the contaminant source. In a single objective problem, the GA-FEM model with an average value of 0.0005036 in five runs of the model had the best performance among other models. The results of the two-objective problem indicated that MOMVO-FEM, despite a few solutions in optimal Pareto-front, could find a better location for pumping wells. Finally, it can be said that among factors such as the location of pumping wells and pumping rate, the most influential factor in choosing the right pumping and treatment policy is the proper location of pumping wells. Also, the location of contamination pumping wells does not necessarily correspond to the location of the contamination seepage.

Keywords

• drawdown of groundwater head
• finite element method
• health-risk assessment
• meta-heuristic algorithms