This project investigates the problem of determining the exact location of power system faults based on the recorded fault transients. The potential of using traveling waves created by power system faults, as a means of determining the location of faults has long been recognized. However, lack of suitable technological tools for high bandwidth monitoring of transients as well as the limitations of existing Fourier based signal processing methods in localizing abrupt changes in signals, have discouraged the efforts of developing comprehensive traveling wave based fault location methods so far. This project focuses on the use of the Discrete Wavelet Transform in processing the traveling waves along the faulted transmission lines. The ability of the Wavelet Transform to capture the fault transients with a high time resolution in lower scales, is exploited in order to develop a fault location algorithm. Essential idea behind the algorithm is to capture the travel time of the modal components of the three phase transient signals during the fault, as they travel between the fault point and the line terminal at which the recording is done. This information can then be translated into an equivalent distance, based on the knowledge of the travel velocities of various modes for the monitored line. The project studies the behavior of the traveling wave based methods' performance under difficult yet common network topologies, such as mutually coupled lines, lines with series compensation capacitors, parallel lines with loop flows, etc. The effectiveness and computational efficiency of the Wavelet Transform is also carefully examined in terms of the optimal choice of the mother wavelet, number of scales, and the modal signals to be processed. Developed methods and the related algorithms are critically evaluated based on digital simulation of actual fault transients using the detailed models of power system components. The project's results are expected to initiate a new way of processing fault transients, using the information at the higher frequency end of their spectrum, as opposed to most of the techniques currently in use today. These results can potentially be used in related applications such as harmonics monitoring, detection and characterization of disturbances, distribution system fault monitoring and incipient failure detection of power transformers.
