Seismic resilience assessment of water distribution systems considering improved fragility relations and CIPP rehabilitation

Water distribution systems (WDSs) in seismic regions are highly vulnerable to earthquake-induced ground deformation, yet network-scale assessments often rely on generic fragility curves that overlook variability in pipe diameter, soil conditions, and rehabilitation strategies. Despite its widespread use, the effectiveness of cured-in-place pipe (CIPP) lining as a network-level seismic mitigation strategy remains understudied. This study evaluates the seismic resilience of WDSs primarily composed of cast iron (CI) pipelines and explores the effectiveness of CIPP rehabilitation as a resilience improvement strategy. Diameter- and soil-specific fragility relations for CI pipelines, with and without CIPP lining, are used to demonstrate the proposed resilience evaluation and improvement schemes, considering two benchmark WDSs under 10,000 randomly generated earthquake scenarios. Specifically, this study aims to (1) assess network-level seismic resilience using diameter- and soil-specific fragility relations, (2) quantify the resilience improvement achieved through budget-constrained CIPP rehabilitation, and (3) distinguish between scenario-specific rehabilitation priorities and systemwide pipeline importance identified across earthquake scenarios. Seismic resilience is quantified using the Topologic Resilience Metric (TRM), which integrates pipeline failure probabilities, network connectivity, and nodal demand. A one-at-a-time (OAT) sensitivity analysis evaluates the effects of hazard and system parameters on resilience. Budget-constrained optimization using a genetic algorithm (GA) is employed to identify CIPP rehabilitation strategies. In addition, the systemwide importance of individual pipelines is quantified based on their mean relative resilience improvement across the simulated earthquake scenarios. The discussion section interprets results in terms of network characteristics, rehabilitation prioritization, and resilience-based pipeline management. Results show that resilience varies across networks, and hazard-related parameters primarily govern resilience variability. Moreover, across six independent GA runs, optimized CIPP rehabilitation increases mean resilience under the selected governing seismic scenario from 61.9 % to 69.2 % in D-Town (7.3 percentage points; 11.8 % relative) and from 53.1 % to 57.6 % in Exnet (4.5 percentage points; 8.5 % relative). These findings highlight how trenchless rehabilitation can be integrated into seismic resilience-based planning for aging water distribution systems.