Rehabilitation Framework for Branched Water Distribution Systems

Abstract

Water distribution systems are essential for providing safe drinking water, yet their design presents complex challenges. These systems necessitate careful selection of pipes, pumps, and tanks, balancing service quality and cost while adhering to hydraulic equations and accounting for uncertainties. Various optimization techniques have been employed, including linear programming, nonlinear programming, and metaheuristics. This thesis develops an Integer Linear Programming (ILP) model for rehabilitating branched water distribution systems, taking into account future demand and generating optimal daily operational schedules accordingly. We apply our model to rural piped networks, focusing on multi-village redesign water networks in the Khardi and Karegaon villages of Maharashtra, India. The model is implemented in Julia and produces INP files of the generated networks that can be simulated and analyzed further using software like EPANET. It opens avenues for future research, including extending the model to looped systems, integrating with smart grids, and developing resilience to power outages in developing countries