Dynamic Groundwater Remediation Design with Genetic Algorithmsby Amy Chan Hilton,
Document Type: Proceeding Paper
Part of: WRPMD'99: Preparing for the 21st Century
In groundwater remediation design it has been shown that dynamic policies, in which the pumping strategies may vary throughout the remediation period, can result in more cost-effective remediation strategies than static policies. The dynamic approach uses management periods, which are portions of the total remediation period during which the remediation policy is unchanged. This work applies genetic algorithms to solve several dynamic formulations of a granular activated carbon pump-and-treat design example and an enhanced in situ ioremediation design example. The genetic algorithm is a search method based on concepts from natural selection. The simultaneous dynamic and the sequential dynamic formulations are compared. In the simultaneous dynamic formulation, all the decision variables for all management periods are optimized at once. Thus decisions made for the first management period impact all future periods. In the sequential dynamic formulation, the decision variables of each period are optimized independent of the other management periods, with the ending conditions of one period being the starting conditions for the next period. Furthermore, the length of each management period is handled as a decision variable. The results from these applications are analyzed, with attention to the optimal costs and policies as well as the computational effort required to solve these examples.
Services: Buy this book/Buy this article
Return to search