Buried pipe leak detection and localization via ground microphone and GPR

Buried-pipe leakage is a common issue in urban water distribution systems worldwide.  Apart from environmental problems such as water waste and pollution, leakage can lead to serious soil erosion and, in some cases, road collapse if not detected and repaired promptly. To date, various approaches for leakage detection and localization have been developed, including ground microphone, district metered areas, closed-circuit television, infrared thermal imaging, and Ground Penetrating Radar (GPR). Among these approaches, ground microphone uses listening devices to detect and localize the pipeline leakage by directly tracing the sound emitted at the point of leakage. However, the accuracy of ground microphone method heavily relies on the experience of the operator, as the sound signal is often contaminated by ambient noise. To address this issue, we combine the noise-resistant GPR method with the ground microphone method. With advantage of high resolution and efficiency, GPR has been widely applied in the localization of buried pipelines and thus has great potential in the detection of leaks.

This paper proposes a combined approach to detect and localize the leaks of buried pipelines. Firstly, the ground microphone method is used to collect acoustic data above the buried pipelines. During this step, the acoustic signals are processed to improve the signal-to-noise radio by using wavelet analysis [1] and loudness units referenced to digital full scale. Then, the relationship between the accuracy and the recognition precision of ground microphone data is analyzed. In the next step, a machine learning-based classifier [2] is established based on the features of acoustic data of buried pipelines, enabling automatic recognition of leaks. Finally, 3D GPR investigation is performed and a relative wavelet entropy (RWE) [3] method is introduced to localize the leakage point.

A laboratory and two filed experiments were carried out to validate the proposed approaches. In the laboratory experiment, we tested the RWE method, and the results show that the method can accurately localize the leaky point from 3D GPR data. Then, the results of two filed tests indicated that the combined approach effectively combines the advantages of ground microphone and GPR, which can efficiently and accurately detect and localize the buried pipeline leaks. The proposed approaches can benefit the health operation of water distribution system in urban cities.

References:

[1] G. Mallat, "A theory for multiresolution signal decomposition: the wavelet representation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 11, no. 7, pp. 674-693, 1989, doi:10.1109/34.192463.

[2] Qu, H. Feng, Z. Zeng, J. Zhuge and S. Jin, "A SVM-based pipeline leakage detection and pre-warning system," Measurement, vol. 43, no. 4, pp. 513-519, 2010, doi:10.1016/j.measurement.2009.12.022.

[3] O. A. Rosso, S. Blanco, J. Yordanova, V. Kolev, A. Figliola, M. Schürmann and E. Başar, "Wavelet entropy: a new tool for analysis of short duration brain electrical signals," Journal of Neuroscience Methods, vol. 105, no. 1, pp. 65-75, 2001, doi:10.1016/s0165-0270(00)00356-3.