Abstract
Water Distribution Systems resilience indicators often describe network topology at a fixed condition or measure hydraulic surplus without explicitly accounting for component aging. Resilience informed management of these systems requires metrics that capture how pipe deterioration affects water supply over time. This study develops the Weighted Independent Pathways metric, WIPW(τ), a time-dependent resilience metric linking pipe deterioration, reliability, hydraulic performance, and network redundancy. The framework combines a geoadditive Bayesian deterioration model developed from 37,484 georeferenced observed watermain failures recorded between 1956 and 2025 across a real-world network, material specific age dependent reliability functions, 24-hour pressure-driven hydraulic simulations and weighted independent pathways. WAIC, DIC, and spatially blocked cross-validation were used to select the best performing deterioration model between Weibull, log-linear non-homogeneous Poisson process (NHPP), and Cox-type models. Closed-form NHPP survival functions propagate pipe age and failure uncertainty to the system level, while pathway weights are determined using the CRITIC method and assessed through sensitivity analysis. The framework is demonstrated on the Anytown, C-Town, and Richmond benchmark networks and compared with the Todini resilience index and a time-dependent reliability index (TRI). Results show that WIPW(τ) captures aging-related losses in supply redundancy not reflected by energy-based indicators. Gradient-guided renewal recovers approximately half of the attainable resilience gain at 17–25% of the full renewal cost, supporting resilience based infrastructure renewal prioritization.

This publication has 24 references indexed in Scilit: