Multi-Objective Optimization of Groundwater Remediation System Design

In recent years, the use of multi-objective optimization techniques in designing groundwater remediation systems has increased because of their computational efficiency for solving problems with two or more objectives, such as minimizing the remediation costs while maximizing the design reliability. For any given multi-objective problem, there exists a set of Pareto-optimal solutions or nondominated solutions. The purpose of this study is to identify those sets of nondominated solutions using a multi-objective optimization technique, Non-dominated Sorting Genetic Algorithm (NSGA). NSGA is coupled in this study with the commonly used groundwater flow and transport models MODFLOW and MT3DMS, to design pump-and-treat systems that achieve the dual objectives of minimum cost and minimum mass remaining. The coupled code is demonstrated to work well for both 2-D hypothetical test problems and 3-D real-world design problems, leading to efficient identification of the Pareto optimal solutions (tradeoff curves).