Multi-objective Optimization Models for Distributing Biosolids to Reuse Fields: A Case Study for the Blue Plains Wastewater Treatment Plant

摘要

This paper presents multi-objective optimization models that simultaneously minimize biosolids odors as well as wastewater treatment and biosolids distribution costs and can be proactively used by biosolids managers at advanced wastewater treatment plants. The computational aspects of these models are daunting, typically involving over 200,000 constraints and about 145,000 variables and are thus difficult to solve. Also, some of the mathematical terms involve bilinear, nonconvex pieces which need to be approximated. This paper presents both a numerical approximation scheme to linearize these nonconvex terms using Schur's decomposition and SOS2 variables as well as a Dantzig-Wolfe decomposition method to improve the computational times. The computational results as well as a sensitivity analysis are performed using data from the Blue Plains advanced wastewater treatment plant in Washington, DC.