Water Distribution Network Reliability: are Surrogate Measures Reliable?

Water distribution networks (WDNs) must be reliable infrastructures since they provide an essential service to society. Reliability assessment is a complex task and involves various aspects: mechanical, hydraulic, water quality, water safety, among others. This paper focus is on the hydraulic reliability. Hydraulic reliability is computationally hard to measure directly, therefore researchers came up with surrogate measures, like the resilience index, the modified resilience index, the flow entropy or the diameter-sensitive flow entropy, that are simple and fast to compute. But, are these surrogate measures reliable to be used in the design of WDNs? This paper proposes a new reliability index based on the surplus flow available on each node to mitigate the effects of a pipe failure. To illustrate the applicability of this new index, a WDN example is optimally designed (simulated annealing algorithm embedded in WaterNetGen) in order to maximize some reliability index subject to a given budget. Results show that the solutions based on the flow entropy or the proposed index are more reliable than the others, and, also, the maximization of the other reliability indexes gives only a residual contribution to the global reliability (or even no contribution at all).