A spatial assessment of the likelihood of failure of water mains, a case study of NWSC Kansanga branch.

Abstract

NWSC has aided in achieving SDG 6 (clean water and sanitation for all) through its mandate of providing water and sewerage services to areas entrusted to it and currently has a total piped water and sewer network of 20,513 Km and 714 Km respectively (NWSC Corporate plan,2021-2024). However, more focus is put on water extensions in comparison to network rehabilitation as these pipes tend to fail as they age. Water pipe failures are majorly associated with pipe characteristics, material properties and environmental conditions (Hossein Rezaei, Bernadette Ryan & Ivan Stoianov, 2015). These have caused supply unreliability, increased NRW, traffic inconveniences, wastage of resources spent on reinstatement among others. Recent research has focused on statistical and machine learning models to predict pipe failures which are complex and require expert knowledge. Furthermore, their probability of failure was not mapped to the geographic location of the pipe allowing creation of a map of the network with the failure state attached. This project uses Weibull probability distribution to determine survival probabilities, AHP multi criteria evaluation techniques to spatially assess the likelihood of failure of water mains before they break focusing on pipe age, pipe failures, size, material and its survival probability. The model developed resulted in a map showing location of pipes with their likelihood of failure. The results also showed that sub branches of Bukasa and Kansanga had pipes with the highest likelihood of failure majorly occurring in pipes with diameters between 50mm_250mm for steel and HDPE materials thus aiding NWSC management to come up with a pipe replacement schedule plan to replace most of the metallic pipes in the identified locations thereby improving supply reliability, reduction in water losses and O and M costs.