Review of statistical water main break prediction models

This paper aims to provide a critical review of statistical water main break forecasting prediction models recently published in the past ten years (2002–2012). The statistical characterization of the break rate of water mains by means of historical failure data are the key similarity between the models reviewed. Recent models reviewed were categorized into deterministic, probabilistic, and soft computing methods. The applicability of a model and its data requirements are of critical importance, and were emphasized as such. The review discusses each model’s research contribution to the field of main break prediction. Based on the review, areas of future research were recommended.