| description abstract | Actions taken to restore water supply service after a disruptive event can greatly influence the ultimate duration of outages, and thus the impacts on households and businesses. It is important, therefore, to design postevent restoration processes to be as effective and efficient as possible. In this paper, we compare and evaluate 15 possible water supply restoration plans to examine the effects of alternative crew management approaches, task prioritization methods, immediate damage discovery, and rerouting efforts. The analysis was conducted using the Restoration of Water after Event Tool (REWET), a newly developed model that simulates the restoration of a damaged water supply system. By allowing a detailed, realistic representation of the water supply system, system functioning, and restoration process, REWET enables a notably nuanced understanding of the behavior of the water supply systems, and the interactions among the system configuration, damage pattern, and restoration plan. Two water supply systems are analyzed—one very small (the Anytown hypothetical network) and one very large (Los Angeles)—for multiple disruptive events—earthquakes of varying intensity. Comparing across systems and events offers insight into the generalizability of the restoration strategy comparison results. Findings indicate that restoration activities result in a substantial reduction in customer-days without service, with an average of 77% of the expected amount avoided in Anytown and 67% in Los Angeles. The analyses also highlight the potentially substantial effect of simplifying the restoration process in a way that omits key tasks, like discovering or rerouting around damage. Finally, they show that the best task prioritization approaches depend on the network size and layout, and the extent and pattern of the damage. | |
