Bayesian method for estimating water supply network repair time

The task of the collective water supply system (CWSS) is to supply residents with an appropriate amount of potable water, ensuring the required water pressure at all times, which requires a reliable water supply network (WSN). It is problematic to estimate the duration of repair of a water supply pipe, which is influenced by many factors, e.g. type of network, pipe diameter or pipe material and failure cause. Bayesian network was used to model the relationship between the time of repair of a water pipe and the identified factors. Repair time estimates were made for 3 case studies: a 40 mm diameter PVC house connection; a 150 mm diameter PE distribution network; and a 400 mm diameter cast iron main network. The shortest repair time was determined for the house connection (17.80 h), while the longest for the main network (38.18 h). The developed method allows for estimating the repair time of the water supply pipes based on operational data for any water supply system. It can be used by water supply companies, for supporting the water supply management, especially during a crisis situation.