Optimal design of district metered areas based on multiple hydraulic performance indicators for water distribution systems

The design of district metered areas (DMAs) requires considering multiple factors to achieve effective partitioning. This study proposes a two-step method for DMA design. In the first step, during the clustering phase, an improved multilevel graph partitioning algorithm is utilized. It incorporates hydraulic characteristic indicators to enhance node weighting, considering constraints on the partition size and the number of boundary pipes. Multiple hydraulic characteristic indicators, such as pipe length and flow rate, can be simultaneously considered. The second step involves the dividing phase. A particle swarm optimization algorithm is employed to optimize the placement of flow meters and valves based on three objective functions. The feasibility of the proposed method is evaluated based on Exnet water distribution networks (WDNs) and a real WDNs. The results demonstrate that the proposed method can be applied to medium-sized and larger networks, providing efficient partitioning and an ample number of feasible DMA design solutions. Specifically, the proposed method significantly improves the balance of demand, pipe length, and pressure among the partitions. It effectively improves the quality of partitioning, aligning with the practical requirements of WDNs.