Coupled Hydraulic-Genetic-Algorithm Optimization for the Joint Minimization of Water and Energy Losses in Urban Water Distribution Networks: A Case Study of an Arid-Region City

Abstract

Aging urban water distribution networks in arid regions lose a substantial fraction of the water they transport, and because every cubic metre of water must be pumped, each physical leak simultaneously wastes the electrical energy used to lift and pressurize it. Excess and spatially uneven pressure is the common root cause of both losses: it drives leakage through pressure-dependent orifices and forces pumping stations to expend energy generating head that is not required for service. This paper develops and demonstrates a coupled optimization framework that treats water losses and pumping energy as a single, jointly minimized system rather than as two independent problems. A pressure-dependent (FAVAD-type) leakage formulation is embedded in a steady-state hydraulic model built on the global-gradient method, and pumping energy is expressed directly through the hydraulic-regime variables. A single economic objective function-the combined annual cost of lost water and wasted energy-is minimized under service-pressure and equipment constraints using a genetic algorithm coupled to the hydraulic solver, with pressure-reducing-valve set-points and variable-frequency-drive pump speeds as decision variables. The framework is applied to a representative model of the Karshi city network (Kashkadarya region, Uzbekistan). For the studied configuration, the optimized pressure regime lowered the network-average pressure from about 58 m to about 33 m while maintaining the minimum service head, reducing modeled physical water losses by roughly 31 % and pumping energy by roughly 27 %, with an estimated simple payback of about two years. The results confirm that district-metered-area zoning combined with active pressure control delivers water and energy savings that neither measure achieves alone.