Abstract
Abstract : In a power system, the function of a power transformer is to adjust the voltage to an appropriate value in different parts of the power grid by stepping it up or down. However, power transformer outages are a major source of unreliability of power supply and their outage directly affects the system. Transformer outages contribute significantly to financial loss due to power outages and the high cost of repair or replacement. Hence, power transformer outages constitute an important subject that needs to be investigated. A literature review on analysis of outages for power transformer population was done. Factors such as voltage levels, age and transformer ranges were considered in various studies. Several studies analysed different data lengths of transformers outages. Previous studies also addressed number of failures. Statistical methods and models such as Weibull distribution, regression model, Cox’s regression model, probability density function, were utilised in the analyses. However, no previous studies considered subtransmission transformer outages for primary voltages ranging between 44 – 132 kV. It was observed that the frequency of subtransmission transformer outages, in terms of various variables of events, had not been analysed previously. No studies had investigated the dependence of variables and the significance of that dependence between variables before. The duration of outages for distribution network components had been studied, but no studies particular to subtransmission transformer outages had been done. The aim of this study was to analyse the historical subtransmission transformer forced outage data for transformers with primary voltage of 44 kV, 66 kV, 88 kV and 132 kV based on 10-year data. The focus of the study was to analyse subtransmission transformer forced outages with respect to frequency and duration, utilising statistical techniques. Under the frequency of forced outages, the combination of actual number of outages and failure rates allowed the identification of critical levels of variables that needed prioritisation in order to improve frequency of forced outages. The bivariate bar graphs, combined with chi-squared test,...
M.Phil. (Electrical Engineering)