Water Leak Detection of Underground Pipes in Steelmaking Works Using Digital Signal Processing of Vibration Data

Abstract

   The cross-correlation method based on vibration measurement is widely used to detect and locate water leaks in underground pipes. However, it cannot be used in noisy condition such as steelmaking works because the leak signal is interfered by the noise. In order to solve this problem, a new leak detection and location technique using digital signal processing has been developed. This technique enables us to extract useful signal and cancel useless noise from measured vibration data. The validity of this technique has been verified by proof-of-concept experiments. Also a water leak in real underground pipes has been successfully detected and located in a steelmaking works.