Leakage from pipe has been a serious problem in recent years and its reduction has become a major priority for water suppliers and pipeline network owners throughout the world. Due to the increased pressure on organizations to prevent financial loss and environmental risk due to leakage, simple methods with high accuracy to overcome this problem are a major aim. Although the reasons for leaks are well-known (e.g. excessive pressure, bad installation, vibration, pipe age, etc.) some of the current methods are useless or imprecise. Transient analysis is often used nowadays due to its high potential to achieve this goal. The method that is used in this paper for leak detection is based on analysis of transient pressure signals using a Cepstrum method. After the analysis of pressure signal by wavelet and cepstrum methods there are some extra features in the processed results that identify the presence and location of leak in the pipeline network. It can also identify small leaks which are difficult to identify by other methods. The combination of orthonormal wavelet transformation (OWT) and cepstrum analysis is capable of filtering and analysing a transient signal from a fluid filled pipe network in order to extract features. This has now been shown to work for experimental data. The ends of the pipes, the junction joining them, and the existence of leaks have all been clearly identified and their position located. The larger the leak, the bigger the peak produced in the analysed result.
Leak detection in pipeline with use of Cepstrum method
2005
8 Seiten, 6 Bilder, 3 Tabellen, 12 Quellen
Aufsatz (Konferenz)
Englisch
Leak detection in pipelines using cepstrum analysis
Tema Archiv | 2006
|Kraftfahrwesen | 1981
|Canadian pipeline installs leak-detection system
Tema Archiv | 1988
|Model based pipeline monitoring with leak detection
Tema Archiv | 2009
|Study outlines pipeline leak-detection selection
Tema Archiv | 2012
|