Vulnerability Assessment of Water Distribution Networks Considering Performance Data

Abstract

A methodology is presented for the reliability assessment of urban water distribution networks (UWDN) on the basis of component analysis, network topology, and—most importantly—survival analysis to include the effects of a network’s past performance on its seismic and/or nonseismic reliability assessment. This study investigates the effects of a network’s historical performance on seismic vulnerability through the introduction of the number of observed previous breaks (NOPB) risk factor. This study discusses how the American Lifeline Alliance (ALA) guideline’s repair rate (RR) metric underestimates the seismic effects on the vulnerability of a network and recommends how the damaged (prior performance) and undamaged network states are included in the calculation of a pipe’s probability of failure. In essence, the proposed methodology acknowledges that the repair rate and the resulting probability-of-failure metric, proposed by the ALA guideline, utilize statistical observations from the performance of a generic UWDN whose historical nonseismic performance is unknown and improves on these metrics by introducing the utilization of localized nonseismic performance data through survival analysis. A spatial analysis is then employed to demonstrate the effects of such nonseismic performance data on the seismic network reliability.