A field implementation of linear prediction for leak-monitoring in water distribution networks

• Published in: Advanced engineering informatics 2020-08, Vol.45, p.101103, Article 101103
• Main Authors: Cody, Roya A., Narasimhan, Sriram
• Format: Article
• Language: English
• Subjects: Acoustic signals; Leak detection; Linear prediction; Water distribution networks
• ISSN: 1474-0346; 1873-5320
• DOI: 10.1016/j.aei.2020.101103

Water distribution networks (WDNs) are complex systems that are subjected to stresses due to a number of hydraulic and environmental loads. As a result, system leaks remain an unavoidable reality. Leaks which are not large enough to become visible at the street level can often go undetected for prolonged periods of time; the presence of smaller leaks can be concealed in system variability. The current paper addresses the problem of leak-detection and localization in WDNs, using a data-driven methodology which utilizes linear prediction (LP) theory. LP has a relatively simple mathematical formulation and has been shown in laboratory studies to effectively capture leak-induced signatures in fluid-filled pipes. In this paper, the performance of LP for leak-detection is verified, using field data in an operational WDN. In addition, a two-part localization approach is proposed which utilizes LP pre-processed data, in tandem with the traditional cross-correlation approach. Results of the field study show that the proposed method is able to perform both leak-detection and localization in full-scale systems using relatively short time signal lengths. This is advantageous in continuous monitoring situations as this minimizes data transmission requirements, which are one of the main impediments to full-scale implementation and deployment of leak-detection technology. In addition to the analysis results, a novel hydrant-mounted data-acquisition system is proposed, along with its unique hardware and software capabilities.