Wavelets for the Analysis of Transient Pressure Signals for Leak Detection

Abstract

Transient tests can be reliably used for the diagnosis of pressurized pipe systems. As a matter of fact, the pressure signals acquired during these tests can reveal the presence of anomalies, e.g., leaks, since any irregularity in the pipe gives rise to reflected waves which in turn create discontinuities in the observed signal at the measurement section. In order to make the most of the interpretation of the pressure signals, as well as to improve the effectiveness of the location of leaks, wavelet analysis—a powerful tool within the realm of harmonic analysis—can be used. It aids in the diagnosis of pressure pipe systems by better exposing pressure signal discontinuities and precisely determining the arrival time of the pressure waves reflected by leaks, thus locating the leak itself. Since many wavelet transforms are available and the choice is driven by the specific application, both continuous and discrete wavelet transforms are considered in the paper, comparing different mother wavelets. Then the reliability of the technique with respect to the noise effects is tested on numerically simulated and experimental pressure signals. Specifically, it is shown that the wavelet analysis of numerical signals, with and without superimposed white noise, facilitates testing the ability of this tool to recognize small step variation—corresponding to small leaks—using low cost transducers.