Use of SALQR Optimization in Large Aquifer Cleanup

by Christopher M. Mansfield, (S.M.ASCE),
Christine A. Shoemaker, (M.ASCE),

Document Type: Proceeding Paper

Part of: North American Water and Environment Congress & Destructive Water


A modified form of Differential Dynamic Programming (DDP), called Suc cessive Approximation to a Linear Quadratic Regulator (SALQR), is used to compute optimal policies for remediation of contaminated groundwater. The results presented here differ from previous work in this area in that the finite element simulation model employed uses the equations for Row and transport in an unconfined aquifer, and that policies are computed for a large finite element mesh using data from a specific field site. The policies computed meet specified clean-up or containment goals for the contaminant considered within a specified time horizon, using pump and treat techniques. Data used in the computer simulation are taken from field investigations at the study site, which were performed by US DOE consultants. Simulation and optimization of the study site required a mesh with over 1500 state variables. The contaminants examined in the optimization study were two of the contaminants of concern at the site. Time-varying optimal policies for the removal or the containment of the contaminants are computed and compared.

Subject Headings: Pollutants | Finite element method | Computer programming | Groundwater management | Field tests | Computer models | Water treatment | United States

Services: Buy this book/Buy this article


Return to search