REGIONAL-SCALE GROUND-WATER QUALITY MONITORING VIA INTEGER PROGRAMMING

A network design approach developed in this paper identifies monitoring sites in multilayered, regional ground water flow systems at risk of contamination from waste storage facilities. Candidate locations are assigned weights that quantify monitoring value in terms of the prospect of plume detection and exposure hazard criteria. Detection criteria are based on the location of a site with respect to the source of contamination and potential contaminant plumes. Exposure criteria are based on the size of the population that consumes water from a supply well and the distances between the supply well and probable zones of contamination. The weights are used in a binary integer mathematical programming problem which selects the monitoring locations. On a 100-point rating scale developed to quantify composite plume detection and characterization efficiency, the network design model scored 87, compared with a score of 76 for a pre-existing monitoring network. (The model and pre-existing schemes involve the same number of wells.) The model selects well sites that are close together near the source of contamination, facilitating early detection of a contaminant release, and further apart downgradient, resulting in areal coverage for plume characterization.