Bi-objective optimization of water distribution networks

Water Distribution Networks (WDNs) are essential in urban centers and industrial processes, presenting high installation and operational costs that are influenced by commercial pipe diameters and pumping requirements. In this paper, a bi-objective optimization model with a MINLP formulation is presented, utilizing disjunctive programming. The model considers unknown flow directions and simultaneously minimizes total costs, including installation, operation, and pump costs, as well as CO ₂ emissions from pipe manufacturing and system operation. Implemented in GAMS, it was tested with three case studies from the literature. The ε-restricted method was employed to address conflicting objectives, yielding results that reconcile economic efficiency and environmental sustainability, thereby highlighting the importance of incorporating environmental impacts into WDN planning.