NEW FAULT LOCATION ALGORITHMS FOR HIGH VOLTAGE TRANSMISSION LINES

Abstract

The detailed developments of fault location algorithms for single-circuit and two individually transposed parallel transmission lines have been presented in this thesis. These algorithms are based on the widely used and time tested symmetrical components theory. Digital simulations and sensitivity analysis have also been conducted for the assessment of the proposed fault location algorithms. The simulation results indicate that the objectives of the research work have been successfully achieved. First, a fault location algorithm for a completely transposed single-circuit transmission line is presented. This algorithm is derived in terms of voltage and current phasers at both terminals of the transmission line. In the development of this algorithm, a single-circuit transmission line is represented by the lumped-parameter TI model. Voltage and current phasors acquired at both the terminals of the transmission line are synchronized with each other using the pre-fault data. Due to the utilization of two-terminal data, the current in the fault-path can be estimated. This makes the algorithm have a high fault location accuracy. However, this proposed algorithm requires data transfer between the two tenninals of the line and it is only suitable for short and medium transmission lines. Next, two algorithms of fault distance calculation for two parallel transmission lines, which are physically separated and individually transposed, have been developed. One of them carries out fault point determination by estimating source impedance of the remote system and the other performs fault location estimation by solving a non-linear algebraic equation. Both the algorithms only need the data recorded at one common terminal of the two lines. Finally, another fault location algorithm for two parallel long distance transmission lines is developed. This algorithm is based on voltage and current data available at one common tenninal of the two lines. Four first order differential equations are created to reflect the relationship among sequence voltages and currents at any point of the lines. With the solutions of these differential equations, the current fed from the remote terminal to the fault resistance can be estimated so that fault location accuracy of this algorithm is independent of fault resistance and source impedances. An iterative process is used for the implementation of fault location in this algorithm. Simulation studies for each fault location algorithm derived in this thesis have been carried out using the steady-state and transient data which are generated by the Electromagnetic Transients Program (EMTP). The steady-state data are used to examine accuracy of the fault location algorithm itself, while simulation results based on the transient data include all the errors of the algorithm, phasor calculation and EMTP modelling. It should be borne in mind that all the fault location algorithms presented in this thesis are used for post-fault analysis rather than for on-line distance protection. Therefore, efforts have been made to improve fault location accuracy of these algorithms. The main contribution of the thesis is that four fault location algorithms have been developed and they can be used in practical power systems. All fault location algorithms proposed in this thesis have the features that (1) the algorithms only need parameters of faulted lines and voltage and current data recorded by digital equipment, (2) the algorithms are not sensitive to fault resistance, source impedances and configurations of power grids and (3) the algorithms have a high accuracy of fault point determination (the maximum error is less than 1.7% under transient conditions).