Coherenceogram for leak detection in water pipes
Files
(Accepted version)
Date
2022
Authors
Zeng, W.
Cazzolato, B.
Lambert, M.
Stephens, M.
Gong, J.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Journal article
Citation
Journal of Sound and Vibration, 2022; 530:1-23
Statement of Responsibility
Wei Zeng, Benjamin Cazzolato, Martin Lambert, Mark Stephens, Jinzhe Gong
Conference Name
Abstract
A widely used technique for locating leaks in buried water distribution pipes is the cross- correlation between two measured acoustic signals at either side of a leak. The technique is not always effective in identifying leaks with high certainty due to the dispersion of the hydro- acoustic waves, unknown frequency characteristics of the leak signal and the potential pres- ence of strong ambient noise. In this paper, a new method termed “coherenceogram” has been proposed for leak detection in water pipes. The frequency-dependent time delay of the leak signal at two measurement stations is estimated based on the principle that the coherence of the two signals will maximize when they are synchronized to contain the same segment of the leak signal. The effect of temporal windowing is investigated, along with the removal of uncertainty, to enhance the efficacy of the coherenceogram for leak detection. Field validations were then carried out on a water pipe (a part of a water distribution system in a busy city area) and a trunk main. The results of the field experiments in comparison with other methods (cross correlations and cross time-frequency spectrum) illustrate that: 1) the coherenceogram does not need any signal filtering and is easy to implement; 2) it can guide the selection of the signal filter for cross- correlation techniques; 3) it accommodates the frequency-dependent characteristics of the leak signals and can be also used in many other fields for time-delay estimation, such as processing acoustic waves and seismic waves which possess strong wave dispersion.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
© 2022 Elsevier Ltd. All rights reserved.