Optimal Groundwater Remediation System Design Via Phased Response Surface Optimization
Mathematically optimizing well placement and pumping rates to simultaneously clean-up and contain two multi-layered plumes is computationally intensive, especially if plume boundaries overlap yet differ from each other. The Wurtsmith AFB Mission Street TCE and DCE plumes are an example of such a situation. A phased approach was used to address this task. First, management goals and constraints were specified through negotiation and by evaluating preliminary simulations. The goal of maximizing mass removal was selected to most robustly address six-year cleanup goals. Of twenty potential well locations the S/0 model selected five as optimal for cleanup. Then additional potential wells were considered to achieve containment goals. The S/0 model selected two additional wells and increased pumping at some of the five initially selected wells. The developed optimal pumping strategy should reduce TCE and DCE concentrations to target levels within six years and satisfy plume containment goals.
Techniques employed here have been or are currently being included in the REMAX software. REMAX can address groundwater management problems ranging from regional sustained yield planning; to conjunctive water use (coordinated use of groundwater and surface waters); to sophisticated contaminant plume management. To achieve this functional range, REMAX employs a suite of simulation and optimization approaches. It can efficiently address large linear optimization problems having thousands of decision and state variables, as well as complex nonlinear plume management problems having relatively few variables.
Peralta, R. C., "Optimal Groundwater Remediation System Design Via Phased Response Surface Optimization" (1988). Civil and Environmental Engineering Faculty Publications. Paper 1162.