An Integrated Framework for Enhancing Water Distribution Network Performance Using Complex Network Theory

Abstract

Water Distribution Networks (WDNs) are critical infrastructure systems that could ensure the equitable and reliable supply of potable water to both urban and rural populations. With the increasing challenges of skewed urban expansion, fluctuating water demand, and climate variability, managing WDNs effectively has become more complex. Key issues such as leakages leading to pressure and water losses, degradation in water quality due to contamination and poor circulation, and inefficient distribution significantly impact both water security and operational costs. This necessitates robust strategies for efficient network operation and maintenance. Conventional WDN management approaches rely heavily on hydraulic model simulations, real-time sensor observations, and extensive field data collection, which are often impractical for large-scale networks. In recent years, Complex Network Theory (CNT) has emerged as a powerful tool for WDN analysis to device strategies for improved/efficient maintenance and operation. CNT provides topology-based insights that complement hydraulic information for more effective network analysis. This thesis explores the potential of CNT-based approaches in three critical/key aspects of WDN management: nodal performance assessment, sectorization of the network into District Metered Areas (DMAs), and optimal sensor placement for effective leak localization. Novel methodologies are proposed for facilitating investigations towards each of these aspects, aiming to enhance the reliability, efficiency, and scalability of WDN management strategies. The first major contribution of the thesis is the development of two novel nodal performance indices, the Integrated Sustainability Index (ISI) and the related Criticality Index (CI). These indices integrate hydraulic indicators with a topological measure to provide a more holistic evaluation of nodal performance. The hydraulic indicators focus on demand satisfaction with desired pressure at nodes, whereas the topologic measure accounts for the interdependencies and interconnectedness of the network components. The ISI is demonstrated to be more effective than the conventional hydraulic indicators-based sustainability index in differentiating the relative performance of nodes, even when they appear similar in hydraulic or topological characteristics. Furthermore, the CI outperforms five existing centrality measures in identifying the most critical nodes whose failure could lead to severe service disruptions. The ability to rank nodes based on their relative importance enables utilities to prioritize maintenance for the smooth operation of the network. This will ensure proactive decision-making for network rehabilitation, risk mitigation, and overall infrastructure resilience. The subsequent chapter of the thesis focuses on the sectorization of a WDN into DMAs. Effective sectorization facilitates controlled management of the desired quantity and quality of water and increases the system’s ability to tackle any unforeseen conditions (e.g., pipe failure, and contaminant breakout). While the conventional sectorization approach offers enhanced network control, it simultaneously makes the network susceptible to failure during emergency situations. To overcome this limitation, a novel greedy algorithm-based WDN sectorization approach is presented which accounts for the multiple associations of nodes across various communities/sectors/DMAs and domain knowledge (e.g., hydraulic and topologic properties) of the network. For use with the approach, a fuzzy variant of the complex network theory-based modularity index is developed. The sectorization approach eliminates the need to prespecify the number of DMAs and effectively partitions a WDN into compact and dynamic DMAs. The results demonstrate that the proposed approach yields partitions with fewer intercommunity/boundary pipes, thereby improving system redundancy and network management while lowering sectorization costs. Additionally, it provides multiple sectorization options, allowing utilities to select the most optimal partitioning based on trade-offs between hydraulic efficiency, resilience, water quality requirements, and operational constraints.