Wave velocity selection for leakage localization of water pipeline by Variational Mode Decomposition

International Conference on Sustainable Energy and Green Technology 2018

能源学;生态环境科学

Lee, W.T.^1 ; Tan, A.C.^1 ; Tey, J.Y.^1 ; Rahman, F.A.^1

Universiti Tunku Abdul Rahman, Bandar Sungai Long, Kajang

43000, Malaysia^1

关键词: Acoustic emission sensors;    Cross correlation techniques;    Leak localization;    Leakage localization;    Localization errors;    Mode decomposition;    Time delay of arrival;    Wave dispersion;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/268/1/012054/pdf
DOI  :  10.1088/1755-1315/268/1/012054

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Water pipelines are susceptible to leakage due to deterioration over time, pipe corrosion, water hammer and soil movement. Water leakage results in losses of revenue and can cause devastating consequence to the environment. Water leakage often happens below ground and remains undetected until a sizable crack has been developed on the pipeline. In the event of a leak, water from the pressurized pipeline expels through the crack at high velocity and generates acoustic waves which propagate along the pipeline. These waves can be picked up by acoustic emission (AE) sensors attached on the surface of the pipe without interfering with the operation of the pipeline. To pinpoint the location of a leak, it requires at least two sensors to be placed on each side of the leak along the pipe. The location of the leak can be located by knowing the time delay of arrival (TDOA) of the acoustic waves at both sensors by employing cross-correlation technique. Leak location is commonly calculated using a constant wave velocity which corresponds to the pipe material. However, the localization error is often large because acoustic wave does not travel along the pipeline at a fixed velocity due to wave dispersion phenomena. The velocity of the acoustic wave is dependent on the frequency and mode of the wave. This paper proposes a leak localization method with an innovative wave velocity selection method based on mode separation by Variational Mode Decomposition (VMD). The proposed method is validated by a leak simulation experiment and it achieves an accuracy of within 7.72% in leak localization.