چكيده لاتين
Critical infrastructures are considered one of the most important assets of any country, where their optimal design, improved operation, and performance contribute to the advancement and development of nations. One such vital infrastructure is urban water distribution networks. These networks require optimal management due to their aging conditions and the need to maximize customer satisfaction. The network must also be capable of providing adequate service during critical scenarios of minimum and maximum network pressure (such as the lowest and highest water withdrawal levels). To assess the quantitative and qualitative parameters of the network, two types of analyses are performed: demand-driven analysis and pressure-driven analysis. The results of the pressure-driven analysis are closer to the actual conditions of the network and provide better outcomes. Additionally, with technological advancements, more physical, chemical, and biological pollutants are entered into water sources, emphasizing the importance of improving the quality of water delivered to consumers. One of the most important quality parameters is the residual chlorine concentration, which must remain within allowable limits to ensure the water is delivered with minimal health risks. If the residual chlorine concentration falls below the allowable limit, the water becomes an environment conducive to the growth of hazardous microorganisms. Conversely, if the concentration exceeds the limit, carcinogenic compounds like trihalomethanes may form. Therefore, in addition to hydraulic pressure management in the network, controlling residual chlorine concentration also becomes a key focus. Furthermore, significant amounts of water are lost annually within the network, and measures need to be taken to reduce these losses. Since there is a direct relationship between pressure levels and leakage, pressure management can help control leakage in the network. In this study, the urban water distribution network of Najafabad, located in Isfahan Province, was used as a real-world case study, along with the Polakis network as another water distribution network. The Najafabad network was modeled using a mathematical graph, and based on graph theory principles, the Girvan-Newman algorithm was used for network partitioning, with acceptable configurations selected from the partitions. Subsequently, the connecting pipes between the zones were optimized using a genetic algorithm based on distinct objective functions, including the maximization of hydraulic reliabilities, water quality reliabilities, combined reliabilities, and the minimization of average network pressure. Since merely opening the connecting pipes between the zones may not bring pressure levels within the desired range, it is necessary to install pressure-reducing valves (PRVs) on the connecting pipes to regulate pressure. To this end, using the NSGA-II and NSGA-III algorithms, the optimal regulatory pressure values for each case are determined. By applying the proposed method for partitioning the network into district metered areas, the average network pressure and leakage rates are reduced, while residual chlorine concentration is increased. A comparison of results reveals that combined reliability levels in the Poulakis and Najafabad networks have improved by 13.7% and 7.3%, respectively, using the proposed method compared to the Girvan-Newman approach. Consequently, the combined reliability values for these networks are 46.8% and 59.4%, respectively. Additionally, in the Najafabad network, leakage rates using the proposed method have decreased by 13.2% and 4.9% for warm and cold days of the year, respectively.
