Redefining water networks efficiency: a synergy of hydraulic modeling and NSGA-II optimization framework for leakage, cost, and pressure

Reducing background leakage in water distribution networks (WDNs) is critical for decreasing energy consumption and promoting sustainable development. Recent advancements in computer modeling enable the design of networks that actively minimize leakage from the outset. This study employs hydraulic analysis using the gradient algorithm method, incorporating pressure-dependent leakage calculations to enhance accuracy. By integrating pressure-sensitive leakage estimation directly into the design process, the work offers novel insights into holistic network planning and sustainable water resource management. The nondominated sorting genetic algorithm II multi-objective optimization technique was applied, targeting three key objectives: minimizing costs (F1), reducing allowable total pressure deficiency (F2), and lowering total background leakage (F3). A case study on a real urban WDN demonstrated that the three-objective optimization approach effectively reduces both background leakage and construction costs by 24.47 and 11.48%, respectively. Specifically, the average background leakage flow decreased from 26.58 l/s under existing conditions to 24.76 l/s with the two-objective design, and further to 23.85 l/s with the three-objective optimal design. This level of leakage reduction could potentially serve approximately 1,024 people, underscoring the approach's practical significance for sustainable water management.