Comparison of Optimization Formulations for Waste‐Load Allocations

Abstract

A novel approach for including uncertainty in the water‐quality‐management process is presented. The approach is based on the use of multiple scenarios that reflect possible combinations of hydrologic, meteorologic, and pollutant loading design conditions. Each scenario, which represents possible “future management design conditions,” defines a set of transfer coefficients that describes the pollutant transport process under these design conditions. A general waste‐load‐allocation model is developed for which all sets of transfer coefficients are simultaneously incorporated in the water‐quality constraint set. Several optimization formulations of this general model are examined to evaluate efficient management solutions that account for a range of possible future conditions. The approach is applied to the management of BOD from 10 dischargers on the Willamette River in Oregon. The results indicate that model formulations based on minimizing a measure of total deviations from a target DO level identify efficient solutions for the waste‐load allocation.