Investigation and Understanding the Conditions of Power Transformer Internal Faults using On-line Technique

Abstract

Loss of a power transformer in a utility, generation plant or process can cost many millions of pounds, depending on how long it is out-of-service. Internal faults are said to be the most likely cause of disruption in transformer performance and interruption of power supply. Improving understanding of the relationship between types of problem in transformers and their indicators will help to identify internal faults and their locations. Developing an on-line method to monitor and investigate the health conditions of the transformer will help asset managers assess plant while it is operational, leading to reduced running costs and increased life. The proposed method is based on measuring electrical parameters at both sides of a transformer to differentiate between healthy and faulty conditions while it is still in service. In this study, two windings of five sections each have been simulated. The values for each section’s resistance and impedance are calculated based on copper windings, interleaved construction and insulation between sections. The simulated power transformer is connected to an inductive load. Comparison of simulated input and output voltages and currents has been conducted to identify indicators of the transformer’s health status. Developed faults at different locations in the transformer windings are used to study the transformer performance and to recognise the fault indicators. The simulation results show very clearly that there are trends in the measured parameters that are attributed to the type of fault within the transformer. Hence a simple logic comparator can be easily deployed to identify the fault location in relation to the transformer health status.