Novel Resilience Assessment Methodology for Water Distribution Systems

Availability of safe drinking water is at the core of economic welfare: Water Distribution Systems are ubiquitous yet silent, until they fail. These systems face diverse threats: natural disruptive events like hurricanes, tornadoes, and earthquakes, and man-made disruptors such as deliberate attack, vandalism, or accidents. Even slow-developing processes like drought or ageing are a possible threat. For these reasons, infrastructure risk previously focused on critical infrastructure protection is shifting towards a dual approach including system resilience: The ability to recover efficiently after failure. This paper presents a novel methodology for estimating the probability of water infrastructure resilience. Existing approaches to measuring system "resilience" are frequently narrow and may ignore essential features of resilient systems. In this methodology, resilience is defined in terms of performance loss, recovery time, and recovery cost. A real-world resilient system should be able to withstand damage and recover within realistic time and cost constraints. A Stochastic Simulation approach is applied on a network model subject to a set of failure scenarios. Each scenario is developed randomly using asset failure patterns according to the type of disruption and specific probability distributions. Hydraulic analysis of the damaged network provides performance values that can be compared with baseline performance (undisturbed network). Finally, the damaged network is assumed to recover using available resources, which are also probabilistic. Applying this sampling process on many repetitions it is possible to estimate the posterior probability that the infrastructure system assessed will be resilient in case of disruption.