Planning Models for Optimal Deployment of Water Leak Sensors and Their Effectiveness

Water pipelines, which are indispensable infrastructure for our society, have been buried in large quantities starting from the high economic growth period. After several decades now, the underground pipelines are undergoing aging problems (deterioration and damage) over time. In addition, the Japanese society is suffering from a shortage of engineers due to the decrease in population. Accordingly, in order to check the condition of the underground pipe, the evaluation performed by human auditory sense is replaced by a mechanical method through a leak sensor. However, leak sensors are offered at high unit prices. Furthermore, the area requiring diagnosis is wide, and the number of pipelines buried in large areas is even greater. Therefore in this study, with the focus on the optimal allocation of leak sensors in a water-supply pipeline network, we propose a planning model based on the “k-median problem”, which is a type of mathematical optimization problem, and verify the effectiveness of the planning model via case studies (application of a planning model to the pipeline network with 30 candidate sensor installation sites).