A probabilistic framework for water network resilience by integrating pressure indicator information and hydraulic simulations

In an era marked by population growth, urbanization, climate change, and aging infrastructure, water networks face increasing pressures threatening their reliability and efficiency. Timely response to incidents and prioritizing critical pipes for intervention are key aspects of ensuring network resilience. Traditionally, pipe criticality ranking has relied on population density on the network, pipe size, and replacement cost. While these factors are valuable, pressure indicators offer an additional layer of insight, which take account of fluctuation with demand, operational changes, and network conditions. Typically, network characteristics, such as robustness, redundancy, and other topological aspects, have been used to estimate network resilience, relying on deterministic methods based on graph theory. This paper proposes a probabilistic approach for modeling resilient water distribution networks and offers an alternative method to deal with real-world uncertainties. Pressure information after a failure is used for identifying critical links that are most important in enhancing network resilience. An application of the proposed methodology to an example network demonstrates that incorporating pressure indicator information can improve the system recovery time by 13%, also providing an opportunity to the infrastructure owner to allocate resources more effectively, prioritize replacement works, and proactively address disruptions. Including information from pressure indicators and probabilistic modeling of responses to disruption has a potential to enable water companies to respond swiftly to incidents, reduce service disruptions, and ensure the continuous delivery of safe and reliable water services. In addition, it also provides valuable insights into a holistic approach to enhancing network resilience, contributing to improved sustainability and reliability of water infrastructure systems.