Water Network Assessment and Reliability Analysis by Use of Survival Analysis

A holistic and sustainable strategy for the management of urban water distribution networks should be composed of two equally important pillars: (1) efficient methods for monitoring, repairing or replacing aging infrastructure, and (2) effective tools for modelling the deterioration in the network and for proactively assessing the risk of failure of its components so as to devise preventive measures for avoiding such failures. The paper presents a framework for devising such a proactive risk-based integrity-monitoring strategy for the management of urban water distribution networks. The framework presented is based on a combination of artificial neural network, parametric and nonparametric survival analysis and it is utilized in the estimation of time-to-failure metrics for pipe networks.