Performance optimization of water distribution network using meta-heuristic algorithms from the perspective of leakage control and resiliency factor (case study: Tehran water distribution network, Iran)

Abstract

In the context of sustainable development and considering water distribution networks (WDNs) as vital infrastructure systems, designing a resilient and efficient network to deliver water demand to consumption nodes while adhering to engineering standards is of utmost importance. This study specifically focused on the complex structure of the north-west Tehran's WDN, encompassing 1124 pipes totaling 92552 m in length, along with four gravity reservoirs and 988 nodes. Genetic Algorithm (GA) and Nonlinear Programming (NLP) were employed as optimization techniques to enhance the WDN by minimizing leakage and improving its resilience. The study involved determining leakage coefficients for nodes using measured data and GA. Subsequently, the WDN was optimized by defining an objective function, constraints, and decision variables using both GA and NLP. The results demonstrated the superiority of GA in terms of pressure reduction, achieving a significant decrease of 23.7%. Additionally, GA outperformed NLP in enhancing the resiliency index, underscoring its effectiveness in optimizing the network's performance and ensuring its robustness against potential disruptions.