Accuracy comparison of partial discharge location algorithms in power cabl

Power cables are widely used in power transmission and distribution system. As the role of power cables is getting important, the condition monitoring of power cables is very essential. Partial-discharge measurements are a proven way to diagnose power-cables. Partial discharges (PDs): PD is defined as "a localized electrical discharge that only partially bridges the insulation between conductors and which may or may not occur adjacent to a conductor." PDs are generally produced when there are some defects on cables. These PDs will gradually degrade and erode the cable dielectric materials, eventually leading to insulation breakdown. Therefore, finding the location of PDs is a powerful and useful tool for the maintenance and operation of power cables to prevent unexpected discontinuities and maintain reliability of power transfer.;There are two major algorithms to find the location of PDs in power cables; the time-domain and phase-domain algorithms. The time-domain algorithms consist of Peak Detection, 50% Peak Detection, 20% Peak Detection, Akaike Information Criterion (AIC), Cross Correlation (CC), Energy Criterion (EC), and Gabor Criterion algorithms. The phase-domain algorithm has been introduced recently and it is based on the phase-difference method.;This thesis deals with the location accuracy comparison of the time-domain and phase-domain algorithms. The algorithms are implemented to the two-end and single-end PD measuring systems. Computer simulation and laboratory experiment are carried out to evaluate the performances of each algorithm and, then their statistical location errors are compared. The computer simulation is performed using the Electromagnetic Transient Program (EMTP) software, while the laboratory experiment is performed using the PD signal generator injected to a 3.3 kV, 50 m XLPE (Cross-linked Polyethylen) cable. The signal post-processing is carried out using the MATLAB. Furthermore, several factors which affect the location accuracy are discussed in detail and a suitable PD measuring system with a proper algorithm is proposed.;The best algorithm for two-end measuring system is trigger/threshold level algorithms (i.e., peak detection, 20% detection, and 50% detection). Meanwhile for single-end measuring system the best algorithm is Akaike Information Criterion (AIC) and Energy Criterion (EC) for simulation case and phase difference algorithm for laboratory experiment case. There are two main factors in the PD signal that affect the location accuracy: the rise time of PD pulse shape and the length of recorded signal..;In order to have good location accuracy, it is proposed that both two-end and single-end measuring systems are used in the measurement. This is because each measuring system has its own superiority and both can complement to each other. In the two-end measuring system, in general, the location error becomes higher when the defect site is further away from the middle of cable. In the single-end measurement method, the location error gradually improves as the defect site is further away from the measuring end.