Detection of Patterns in Water Distribution Pipe Breakage Using Spatial Scan Statistics for Point Events in a Physical Network

摘要: Infrastructure systems of many U.S. cities are in poor condition, with assets reaching the end their service life and requiring significant capital investments. One primary requirement to optimize allocation investments such is an effective assessment physical condition assets. This paper addresses drinking water distribution by analyzing pipe breakage data as main source evidence about current pipes over space. From this spatial perspective, questions whether sets present unexpected clustering breaks, where those break clusters located if they do exist. presents a novel approach that aims detect locate points network. The proposed extends existing scan statistic approaches, which commonly used for detection disease outbreaks two-dimensional framework, collected from networked infrastructure systems. described tested set consists 491 breaks occurred six years 160-mi results presented indicate adapted applied networks able noncompact shapes, these significantly higher than expected rates even after accounting age diameter. Several possible hypotheses explored potential causes clusters.