Resilience-Based Rehabilitation Planning For Water Distribution Systems in Seismic Zones

Zeinab Farahmandfar, Clemson University


ABSTRACT Water supply systems (WSSs) are one of the most essential infrastructures in modern societies and their disruption due to adversities such as natural and anthropogenic hazards may cause severe inconvenience to consumers. Their continuous functioning is paramount in the case of earthquakes for fighting fires that usually transpire following an earthquake. Therefore, it is important for WSSs to continue satisfying demands without major degradation after an earthquake event. This ability is often referred as resilience of WSS. On the other hand, aging and deterioration of pipelines make water infrastructure more vulnerable to external hazards such as earthquakes. Although several previous research studies assessed seismic resilience of WSSs, there are limited demonstrations made on large-scale WSSs. Furthermore, limited approaches were formulated to consider multiple hazards in an integrated manner to determine capital improvement rehabilitation actions. Consequently, this research study developed a quantitative rehabilitation planning approach that can be used by utility managers and system engineers to enhance seismic resilience of large WSSs. The proposed rehabilitation framework is capable of assessing seismic hazards and inherent systemic uncertainties in an integrated manner. This study proposed and demonstrated two novel seismic resilience metrics – topology based and flow based - and comparatively evaluated them to choose the better metric to be used in the rehabilitation planning framework. A reasonably large WSS operating in the earthquake-prone Charleston region of South Carolina is used as a test bed for demonstration purposes in this study. The comparative analysis of the two resilience metrics revealed the trade-off between the computational benefit and resilience performance benefits. The better performing resilience metric is subsequently used to determine optimal rehabilitation actions by simultaneously considering seismic hazards and deterioration-related uncertainties. Rise in water demand has also been considered as part of the rehabilitation planning keeping in mind the population growth and changing life styles. Multiple evolutionary optimization algorithms are tested to determine the best suitable option to be used in the rehabilitation planning framework. The proposed resilience metrics, developed rehabilitation framework, demonstration approach and the obtained results present the design engineer or the utility owner an insight for formulating effective rehabilitation planning schemes.