An innovative simulation algorithm for optimising the cost of communal water supply systems using EPANET

Abstract

In this work, the EPANET software is used to parametrize the design, simulate and optimise the overall cost of building and running a reliable, sustainable and efficient communal Water Supply System. Considering a model community of forty houses with an average of six occupants per home, the developed optimization algorithm using the simulated annealing heuristics approach minimizes the total cost of continuous water distribution to the community through a system of pipe network in a way that satisfies the critical hydraulic rules and constraints for efficient water distribution. Thus, the controlled variables of the simulated system include volumetric flow, velocity and head loss. To realize the desired goal, the prototyped communal water distribution system was first simulated by water free fall i.e., only gravity feed under a peak demand flow condition. This facilitated the identification and localization of head-loss zones. Subsequently, virtual pump stations of various capacities were optimally introduced as boosters at identified nodal points to overcome critical head losses. The response of the system to varying capacities of the booster pumps was then used to analyse and determine the optimum capacity of the pump. Simulation outcome showed that the optimum least cost design for a sixtyyear system’s life cycle is achieved by using a combination of gravity and 5 hp (3.73 kW) pump with optimum pipe diameter ranging from 60 - 150mm.