Relationship between environmental factors and water pipe failure: an open access data study

The present study provides a proof-of-concept data collection methodology for water pipes failure from publicly available (open) data sources and identification of the common causes of water pipes failure in a water supply network, which may be automated with artificial intelligent (AI) methodology. Using Adelaide (South Australia) as the location of our case study, open source databases, such as pipe failures of SA Water (Twitter feeds), data from the Bureau of Meteorology, Adelaide Metropolitan soil data reports, etc., have been used to identify pipe failure locations, which was then correlated with the environmental factors e.g. soil type, monthly rainfall pattern, etc. The results were compared with the SA Water’s renewal program to validate the methodology introduced by this research. The results confirmed SA Water’s asset management system is very effective in managing its linear assets with the areas focused on the water main replacement program matched well with the pipe failure locations. This not only validated the methodology introduced by this research is capable to identify the high-risk pipe failure areas but also facilitated linkage between government and public agencies externally via their open access data which may complement their core services.