Detection of water leaks in metal pipes by electromagnetic induction and electrical resistivity tomography time-lapse

Abstract

This master thesis aims to define the feasibility of detecting water leaks in metal pipes using geophysical prospecting methods based on electromagnetic induction (EMI), with temporal monitoring. The work is carried out in partnership with the Société Wallonne des Eaux (SWDE), the project applicant. A preliminary characterization of the site using different geophysical methods is undertaken in order to validate the location of the site. Following this, a controlled infiltration of water is conducted in order to simulate a leak and three cases are studied. The first one consists in laying the foundations of the feasibility of water detection using EM methods by detecting the infiltration in the absence of any metallic disturbance. This case gives encouraging results, although some results are peculiar with a decrease in conductivity instead of an increase. The second case involves the detection of a variation in saturation in the presence of a metallic artefact, in this case the injection valve, located on the surface. After 4 days and an injected volume of 2 m³, the response of the saturated zone exceeds the response of the metallic element, both in amplitude and in spatial extension, meaning that the electromagnetic induction allows the detection of an infiltration despite a major disturbance. However, the results are limited due to the difficulty of saturating the soil and the majority of gravity flows. The third case is the closest to reality, with infiltration at the pipe. Unfortunately, this case did not give any results because the data are noisy due to the presence of an electric cable, discovered during the experiment. A section is dedicated to the recommendations to be followed in order to be able to renew the experiment under more favourable conditions. The data are presented in 1D form with the use of a single line of measurements, despite the 2D coverage of the area. This is justified by the very low injection rates and therefore a small saturated zone. The EM measurements are coupled with tomographic measurements of electrical resistivity in order to avoid the signature of the injection valve and to obtain a robust result regarding infiltration.