Leak and blockage detection in pipelines

The detection of both leaks and blockages in pipelines is crucial to save money and avoid environmental adverse effects. Studies for leak and blockage detection in pipelines using unsteady flow that occurs during normal pipeline operations are reported herein. A new technique is developed utilizing the system pressure frequency response to locate and quantify leaks and blockages. The pressure frequency response is calculated at the downstream valve by the transfer matrix method. It is found that the presence of leaks and blockages alters the shape of the system pressure frequency response, the amplitude of pressure oscillations at even harmonics is increased and the amplitude at the odd harmonics is decreased. The increase in amplitude at the even harmonics is used to detect the location of leaks or blockages based on the leak and blockage frequency. For a simple pipeline, the technique successfully detected blockages up to 30% in size and leaks up to 2% of the mean discharge. Moreover, the computed results are verified by comparing with experimental measurements and those obtained by using the method of characteristics.; The concept of "wavelet multiresolution analysis" is utilized with the possibility of extracting details from the transient signal due to leaks and blockages. The transient flow equations are initially solved directly for the expansion coefficients at a certain level of resolution using the Wavelet-Galerkin method. Two wavelet bases, Haar and Daubechies Db3 are used in the Galerkin scheme. The expansion coefficients are used to compute the pressure variation at the downstream valve and the computed results are in excellent agreement with those obtained by the method of characteristics. Moreover, the calculated expansion coefficients are used to initiate the wavelet multilevel decomposition. It is found that transient features can be extracted in the decomposition levels at various levels of resolutions and various scales. Work is extended to wavelet multiresolution decomposition of transient laboratory experiments by using several compactly supported wavelets. The computed results show that the features due to blockages and leaks can be successfully extracted from the decomposition process with the best performance by Haar wavelets. Moreover, the extracted detail coefficients are used to differentiate between leak and blockage in the wavelet decomposition hierarchy.