Robust Optimal Design for Wastewater Treatment. Part I: General Approach

Abstract

One important problem with using mathematical optimization models for design of wastewater treatment systems is that parameter values, and thus, model results, are often uncertain. A general approach, called robust optimal design, is developed for extending nonlinear optimization models to include first-order sensitivity-based measures of system robustness. Robustness is defined narrowly as the ability of the system to maintain a level of performance even if the actual parameter values are different from the assumed values; thus less sensitive designs should be more robust. A solution procedure based on nonlinear optimization and system sensitivity analysis techniques is discussed. The approach can be used to generate alternative designs that recognize traditional modeled objectives such as cost and effluent water quality measures, and that also reflect concerns about uncertainty. The approach is applied in a companion paper to a complex activated sludge treatment system with 55 uncertain parameter values.