Identification of Critical Pipes for Proactive Resource-Constrained Seismic Rehabilitation of Water Pipe Networks

Abstract

Utility managers in charge of water pipe networks that are exposed to high seismicity make difficult decisions regarding the allocation of a limited rehabilitation budget to be most effective in enhancing a network’s postearthquake serviceability. Seismic vulnerability models are typically integrated with simple prioritization methods to identify the critical pipes subjected to earthquakes. These methods do not distribute resources at the system level and may not provide an economical solution. The objective of this paper is to develop an approach to identify the critical pipes for proactive seismic rehabilitation that will enhance a network’s postearthquake serviceability when only a finite length of pipes can be rehabilitated. To achieve this objective, a proper stochastic combinatorial optimization was formulated and then solved, using a genetic algorithm that was integrated with a network-level seismic vulnerability model. The approach was implemented to identify critical links for proactive seismic rehabilitation of two benchmark networks. The results showed that this approach outperforms the simple length-based prioritization methods used by the utilities, as well as the latest proposed methodology in the literature, in identifying the critical pipes in a water pipe network subjected to an earthquake.