Friday, February 22, 2002
11:30 am, Science Building, Small Auditorium

Karen Ricciardi

University of Vermont

Optimal Groundwater Remediation Design Subject to Uncertainty

Abstract: The main objective of this research is to include the uncertainty of the hydraulic conductivity into a cost effective pump and treat groundwater remediation design subject to gradient and concentration constraints. The method used to include this uncertainty, called robust optimization, is one that examines many different realizations of the uncertain parameter simultaneously. Violations of the constraints are permitted, however, a cost is added to the objective to account for violations of the constraints.

The cost associated with violations of the constraints becomes a part of the objective function in the robust optimization setup, and so the characteristics of the objective function take on the characteristics of the constraints. The cost per unit of pumping and the hydraulic head response to pumping is linear, however, the concentration of contaminant in the groundwater has a nonlinear response to pumping. Because the gradient response to pumping is linear and the concentration response is nonlinear, the robust optimization problem for gradient constraint problems is linear and is a much more simple problem to solve than one subject to concentration constraints. These problems have led to the development of a new method to solve numeric optimization problems.