Abstract
Transmission line’s fault location algorithms are designed for estimation of fault distance after line tripping but before the repairment of transmission line. Because fault current is contaminated with high frequency and decaying DC component, the current and voltage signals must be preprocessed for extraction of the fundamental components for calculation of the fault distance making use of line parameters. Recent fast development of global positioning system allows the synchronous sampling of current or voltage at both terminals of the faulted transmission line, yielding a higher fault location accuracy. This thesis uses the fundamental phasors of voltage and current to estimate fault distance. By integrating three existing fault location algorithms, and by adapting these existing algorithms to the application of synchronous sampling of two-terminal signals for both the single and double-circuit lines. The thesis contents are comprised of 4 parts. Part 1 contents refer to the single-circuit line, by integrating two existing algorithms. One uses the single-terminal signals; the other uses the two-terminal signals nonsynchronously sampled. The latter algorithm is further modified into a general formula for fault location applicable to 4 categories of total 10 types of fault, using the nonsynchronous or synchronous samples of two-terminal signals. Part 2 refers to the double-circuit line by rewriting the existing fault location formula to be the same format as those of Part 1. The formula is formula modified for use of the synchronous samples of two-terminal signals. Most importantly, the 3 algorithms in Part 1 and 2 algorithms in Part 2 have all been reconfigured so that the fault location processes are all based on the same general formula, which can thus save tremendous amount of computer implementation effort. Part 3 is concerned with the modification of signal preprocessing. In the modification, the fast Fourier transform (FFT) has been widely applied. The FFT is applied to the detection of fault inception point within the transient waveform, which replaces the traditional approach by detecting fault inception with only three samples. The FFT is also applied to the removal of decaying DC which replaces the conventional minic circuit method. Both have either vastly reduced the signal processing error or saved the computational time. In Part 4, the aforementioned 5 fault location algorithms, 3 for single-circuit line and 2 for double-circuit line, have been tested on the numerically simulated transmission lines. The simulation is done by using ElectroMagnetic Transient Program (EMTP) for simulation of Taipower’s 345kV and 161 kV lines. The 5 algorithms have been further tested on the transient data recorded during a 345 kV –line fault. Characteristics on the fault location accuracy have been compared and analyzed by both the direct fault location test and the sensitivity analysis tests.