Understanding failure rates in cast iron pipes using temporal stratification

Abstract

The need for proactive replacement schedules in the management of water supply infrastructure is highlighted by the ever-increasing cost of reactive repairs. Proactive replacement schedules require models of asset performance, and, in turn, these models require appropriate stratifications of the data in order to produce subgroups that are expected to behave uniformly over time. This paper outlines a new approach to determine relevant temporal stratifications through an examination of historical performance. This represents a novel alternative to current practice where stratification regimes are identified using a priori knowledge of criteria deemed relevant. A key advantage of using the actual break histories to identify appropriate stratifications lies in the ability to identify changes in behaviour that may not be reflected in a priori criteria. In an example using Hunter Water Corporation break data a significant change is discovered and found not to correlate with previously hypothesised partition criteria. A possible cause for the change in behaviour is suggested.

Pipe Break Rate Assessment While Considering Physical and Operational Factors: A Methodology based on Global Positioning System and Data-Driven Techniques

“…Ultimately, soft computing techniques especially a variety of Artificial Intelligence (AI) models, were generally flexible to combine with Evolutionary Algorithms (EAs) for various purposes such as optimizing time scheduling, costs on rehabilitation, replacement, repair, pressure fluctuations, and break rates (Clark et al,1982;Kettler and Goulter, 1985;Goulter et al, 1993;Silinis and Franks, 2007;Berardi et al, 2008;Clair and Sinha, 2012;Tang et al, 2019;Robles-Velasco et al, 2020).…”