Leak detection in water carrying pipe using a peak detector

This study is a continuation of previous work carried out here in the Chemical Engineering at the University of Bahrain. In the previous work, we first implemented what we called polarity correlation to determine the location where the leak occurs. However, the problem was the large amount of computation. The technique was modified to a more efficient one by implementing what was called zero crossing correlation, where computation was reduced extensively. Though both techniques contribute well in estimating the leak, it was thought that the computation could be reduced further by completely, eliminating the convolution. The idea behind the new technique is to place two hydrophones, one at each end of the pipe. When a leak occurs, it generates two signals, one upstream and one downstream. Depending on how far the leak from the transmitters is, there will be a delay between the two signals. Those signals generate a high peak with small secondary insignificant peaks. The time delay is generated as follows. A very fast 32-bit timer is set as a freerunning timer. When the first peak occurs on the first transmitter, the processor records the time. When the delayed peak occurs on the second transmitter, the processor records the second time. It then calculates the difference between the two. This difference represents the delay. Knowing the length of the pipe, as well as the geometry of the pipe and the flow velocity, the distance where the leak has occurred, is easily estimated.