A Multiscale Approach to Leak Detection and Localization in Water Pipeline Network

Abstract

Leak detection and localization in water pipeline networks is of paramount importance to industry, especially in regions where water is scarce. In this paper, we present a novel multi-modal and multi-scale approach for leak detection and localization in water pipeline networks, in which pressure measurements at various points on the network are used to localize the pipe segment in which the leak is occurring, and then the vibration sensors are used to localize the leak within this segment. In some situations where the complete pipeline model is not available, pressure data alone may not be effective in localizing the leak. However, in such a situation, by supplementing pressure data with vibration data, the leak can be localized, as these additional data are easier to acquire at arbitrary points, since vibration sensors are non-invasive. In order to validate the effectiveness of the approach that needs both pressure and vibration data, we simulate the pipeline model using EPANET that includes models for flow and pressure at various points on the pipeline, then integrate the vibration model with it in MATLAB, since EPNAET does not include models for vibration measurements. A case study of a pipeline network is considered, and the proposed scheme is used to detect and localize the leak. Extensive simulation results show the effectiveness of the proposed scheme in providing accurate leak detection and localization.