Optimizing Groundwater Remediation Design under Uncertainty

There always exists some degree of uncertainty associated with groundwater problems, often in determining the aquifer parameter values. Therefore finding an optimal remediation strategy based on a deterministic description of the system may not yield an optimal and feasible design. This work develops a genetic algorithm (GA) approach that takes into account the uncertainty of hydraulic conductivity values when determining the best remediation design possible. During the GA optimization, the heterogeneous hydraulic conductivity field realization varies between generations and on-going performance is measured. A policy's fitness is based on its performance over multiple generations. Therefore the most fit policy should provide a robust solution since this policy would be a good design over a range of aquifer realizations. Results of this approach applied to a hypothetical contaminated aquifer remediated by a pump-and-treat system indicate that a set of non-dominated policies can be generated by this modified GA. Additionally, this work shows that using a deterministic description of the aquifer, either homogeneous or heterogeneous, can result in significant under-design, with poor reliability.