Transient wave-leak interaction analysis in water supply pipelines

Abstract

In recent years, transient wave-based leakage detection methods (TWLDMs) have received increasing attention in detecting pipes' defects for several advantages, such as high efficiency, low cost, and non-intrusive characteristics. Among different transient-based leak detection methods, the transient wave reflection-based method (TWRM) has been proven to be convenient and efficient for applications. Previously, the TWRM has demonstrated its applicability for a rough estimation of leak size, they either consider a frictionless case, or use a weak steady friction assumption in the elastic pipelines, or elastic pipe-wall assumption through testing viscoelastic pipes. Consequently, the amplitude damping of transient waves in elastic pipes and the additional phase shift effects of transient waves in viscoelastic pipes cannot be properly accounted for by the current TWRM. Although acceptable results were reported for some specific case studies in the literature, still the inclusion of these two effects in the derivation of the reflection coefficient helps enhance the accuracy and applicability of this type of TWRM for practical applications. To this end, it is necessary to perform a systematic analysis for the leak-induced reflection wave and its propagation considering both friction and viscoelasticity effects as well as the measurement distance influence for improving the accuracy of leakage detection, which is the main motivation and scope of this thesis research. The analysis indicates that, the inclusion of these damping-related terms (friction and viscoelasticity of pipe wall) enhances the accuracy of TWRM for leakage detection. It is also noted that despite numerous publications and research projects to address the accuracy and efficiency of the transient wave-based leakage detection methods (TWLDMs), still the order of dominance of various parameters is not well understood, demanding a systematic investigation of the role of various parameters in leakage detection, which helps to evaluate and characterize the TWLDMs. To achieve this, this study adopts dominance analysis to rank the relative importance of the critical parameters involved in the detection process. The preceding studies only investigated specific factors that influence leak detectability through leak-reflection signals. In contrast, this study seeks to examine their relative importance, in general. Through extensive numerical tests and dominance analysis, it is found that, among the dimensionless leak-related influential quantities, leak ratio is the most dominant factor which plays the most important role in the leak-induced reflection coefficient at the measurement point whereas the measurement distance is the least important. Last, the validity of the leak law in transient modeling which is treated as a discrete point in almost all TWLDMs may be the source of minor errors in the magnitude of the pressure peaks and frequency shift of the corresponding spectrum. Meanwhile, no solid evidence has yet to be shown to illustrate the rationality of this simplification. As an extension, this research is one of the first attempts to thoroughly examine the most-common longitudinal crack-induced leakage and develops the corresponding analytical solution based on the Fourier series in typical reservoir-pipeline-reservoir systems to capture the time-domain pressure signals in transient operation.