Dynamic fault location method for distribution networks with distributed generation

Abstract

This paper presents a new analytical time-based fault location algorithm for radial distribution networks with synchronous generators used for distributed generation (DG). The algorithm uses the voltage and current samples recorded at the main feeder head along with the scheduled injected active and reactive powers by the network synchronous generators to estimate the fault location. A full order synchronous machine model has been used to consider dynamic behaviour of the DG plants during fault transients. The algorithm, therefore, uses time domain analysis and is dynamic in the sense that it uses the updated synchronous machines and network data at each time step. The algorithm is implemented in two sequential steps. In the initialisation step, by using the measured pre-fault voltage at the main feeder head and scheduled power injections by the DG plants, a pre-fault load flow calculation is performed to evaluate different parameters of the synchronous generators, initial values of the network node voltages, line section currents, loads and capacitive charging currents. In the second step, the numerical fault location algorithm is performed by moving the data window from the pre-fault to a post-fault cycle. The accuracy of the proposed algorithm has been validated by several simulated fault studies carried out on a 205-node 20 kV practical radial distribution feeder. The results have shown remarkable accuracy in fault location when considering only the current and voltage measurements at the main feeder head have been used.