Assessment of Criticality in Water Distribution Networks via Complex Network Theory

Water distribution networks (WDNs), which are responsible for delivering water of adequate quantity and quality, are vulnerable to threats such as leaks, pipe breaks, and contaminant intrusions. Hence, it is important to identify critical network elements to develop more assertive maintenance strategies for water systems. This paper aims to perform a risk assessment on leaks and pipe breaks to support the identification of critical elements in water supply systems. To this end, complex network theory (CNT) is applied as an alternative to conventional approaches that rely on multiple hydraulic simulations. Metrics such as robustness, redundancy, centrality, and connectivity are used to analyze graphs representing WDNs. Failures are modeled using hydraulic simulations to evaluate their impact on parameters such as pressure and flow. CNT metrics are then applied, including shortest path calculations between water sources and demand vertices to assess pipe importance, and vertex centrality metrics to evaluate node influence on the network. The results of the hydraulic simulations are compared with the outcomes of CNT-based analyses. Multi-criteria analysis is then employed to determine the asset maintenance priority, considering multiple failures and the associated impacts on the system. The results highlight a novel approach that shifts the focus from hydraulic state-based assessments to topology-driven analysis, reducing the influence of uncertainties inherent in water distribution network models.