Mixed-Integer Chance-Constrained Models for Ground-Water Remediation

Abstract

Ground-water remediation optimization models were formulated using a statistical optimization methodology, chance-constrained programming (CCP), to account for uncertainty in the coefficients of the models. Several models were formulated that depended on which set of coefficients were considered uncertain. Such models were either mixed-integer linear programming models or mixed-integer nonlinear programming models. The CCP method transformed the probabilistic models to deterministic models. The deterministic models are easier to solve and use less computer memory and less storage space than probabilistic models. Results are presented that demonstrate the models formulated. The results showed that incorporating uncertainty into a ground-water optimization model using CCP could be a practical method for making decisions on well locations and pumping rates in ground-water remediation.