Détection des fuites dans les réseaux de transport d'eau par radar hyperfréquence

De Coster, Albéric [UCL] Nottebaere, Martijn [ERM, SIC, Belgium] Neyt, Xavier [ERM, SIC, Belgium] Perez Medina, Jorge Luis [UCL] Vanderdonckt, Jean [UCL] Alkhalifeh, Khaldoun [UCL] Craeye, Christophe [UCL] Lambot, Sébastien [UCL]

In modern water supply systems, the amount of water lost through leaks can reach 20% to 30% of the total production. In Wallonia, leaks represent more than 20% of the produced drinkable water. The economic loss, which takes into account the cost of water production as well as damages engendered to buildings and infrastructures, is substantial and the environmental impacts are considerable. Water leak detection is therefore essential regarding to the distribution network management. The most common techniques used to detect water leaks are based on acoustic devices. However, all the current methods show limitations with respect to the background noise, required skills or the configuration underneath the road. Ground-penetrating radar (GPR) is a complementary technology which has raised a substantial interest regarding leak detection. GPR consists in emitting electromagnetic waves and receiving the backscattered fields in order to probe lossy dielectric materials. This technology can be used to detect structures and changes in material properties within the medium. Even if this non-destructive tool appears to be very promising in detecting pipes and leakages of water, some difficulties are often encountered when attempting to detect these ones in field conditions because environmental conditions can be intricate. In this context, the innovative SENSPORT project aims at developing the GPR capabilities to detect water leaks in water distribution networks and to provide detailed information about underground structures in a non-destructive way. The general objective can be divided into three specific parts. The first one consists in designing an ultra-wideband antenna in order to maximize the information contained in the radar data. The expected radar system will be based on the vector network analyzer (VNA) technology allowing us to perform accurate measurements in the frequency-domain. The second part deals with the signal processing and follows the adaptation and the integration of advanced radar imagery techniques. This step characterizes the qualitative aspect of GPR measurements. The quantitative aspect will be introduced through signal inversion which is used to get additional information about the location of pipe leaks. The last part sets the target of developing a visualization strategy for mobile and web applications which can be used in field conditions. It includes identifying the specific application needs and designing a human-machine interface prototype.