Multiobjective optimization of water-using networks with multiple contaminants

A systematic multiobjective optimization procedure is developed in this study to produce realistic multicontaminant water-using network designs. In every given design problem, the most appropriate structure is identified by solving three mathematical programming models (that is, one nonlinear program and two mixed-integer nonlinear programs) sequentially so as to satisfy different criteria in order of decreasing importance. Freshwater conservation is given the top priority in the present work because of its scarcity in environment and the pollution problems caused by wastewater effluents. Minimization of the total number of interconnections in the water network is treated as the next design goal since network complexity is directly related to operability, controllability and safety. The final step in the sequential procedure is to minimize the total throughput and, consequently, the operating and capital costs of all water using units as well. Five examples are presented in this paper to demonstrate the effectiveness of proposed procedure. When compared with the network designs obtained with available methods reported in literature, it can be observed that better solutions can usually be generated with our approach.