Leak detection in gas pipelines using wavelet-based filtering

Abstract

The article presents the application of the wavelet filter for detection of leaks in gas pipelines. The idea behind the method is based on the detection of echoes reflected from turbulences induced by a leak. Such echoes are often very hard to detect, mainly due to high noise levels. The common way of enhancing the analysis is to apply a filter. Different types of filters can be used in practice. The selection of the appropriate filter type and its parameters is the major difficulty. The wavelet-based filter, selected for leakage detection, is optimally tuned using maximum values of kurtosis. The proposed method is verified using simulated and experimental signals with high noise levels. The results demonstrate that for the conditions used, the approach is superior over existing signal processing techniques. The method allows not only for leakage detection but also for leakage location and estimation of its severity.