High impedance fault detection and localization in 11kV distribution system / Mohd Syukri Ali

Mohd Syukri , Ali (2018) High impedance fault detection and localization in 11kV distribution system / Mohd Syukri Ali. PhD thesis, University of Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (1638Kb)
	PDF (Thesis PhD) Download (3293Kb)

Abstract

High impedance fault (HIF) is a type of fault, where the fault current flow is restricted by the high impedance element. The insufficient fault current will not trigger the fuse or conventional overcurrent protection relay. As such, any undetected HIF will lead to potential hazards to human life and equipment damages. On the long run, it can affect power system quality in terms of service continuity, disturbance propagation and energy losses if HIF left untreated. In order to overcome the stated problems, it is imperative for the utilities to detect the HIF and its location in distribution system quickly. However, the detection of HIF is challenging as the resultant distortion on the voltage profile and increment in harmonic content are similar to the non-HIF events such as load switching, motor starting and capacitor switching. Moreover, locating the HIF is also a complicated task especially for a distribution network. This is due to the complexity of the network such as the presence of non-homogeneous line, lateral branches and load variations. Considering the importance of detecting and locating HIF in the distribution system, this study proposes a combined technique comprising of two independent techniques utilizing single measurement point. The first technique, based on Phase Displacement Computation (PDC) is proposed to detect the occurrence of HIF utilizing the three-phase voltage waveforms. Concurrently, the proposed technique is able to discriminate the HIF from non-HIF events. The second technique, based on intelligent approach is proposed to identify the location of HIF. For this purpose, the measured three-phase voltage and current waveforms are analyzed using the wavelet transform (WT). Then, the extracted approximation features from WT are fed into Artificial Neural Network (ANN) to be trained. In this process, the grey wolf optimization (GWO) technique is adopted in order to provide the optimal value of ANN parameters. Fault simulations were carried out using iv the PSCAD/EMTDC software to obtain the distorted voltage and current waveforms during the HIF events. Subsequent algorithm simulation results show that the proposed methods are able to detect and subsequently locate the faults with high accuracies. In addition, the proposed methods are found to be effective and economical since the methods require only single measurement of voltage and current waveforms.

Actions (For repository staff only : Login required)