Transformer Diagnosis Using Frequency Response Analysis

  • Mozon Shadid

Student thesis: Master's Thesis

Abstract

Power transformers are important components of electric networks that plays a vital role in the stability of electrical power systems. Furthermore, power transformer failure can lead to power outage and interruption of energy supply, resulting in significant expenses owing to loss of operation and consequent maintenance costs. Consequently, it may have an impact on the transformer's capacity to tolerate further mechanical or electrical faults. Therefore, it is essential to diagnose faults in transformers rapidly and precisely. Condition monitoring techniques have been developed and implemented to monitor faults in transformers, reduce failure rates and extend the life of transformers. Frequency response analysis is a widely used tool to detect power transformer winding deformation. FRA has been used as an offline diagnostic tool over the past couple of decades. However, with the recent development in smart grid systems there is now a growing interest in the development of on-line monitoring techniques. Therefore, switching FRA assessment from offline to online approaches has attracted attention and interest to reduce the maintenance costs as well as minimizing the probability of unexpected outages, or potentially to prevent disastrous failures in power grids. This study contributes into transformer winding turn-to-turn short circuit diagnosis using online IFRA measurement. The proposed method is investigated using two identical transformers. Transient impulses are used as an input signal injected to the transformer winding to emulate the naturally occurring transients in power systems. The frequency response attained has been analysed graphically and by using various statistical tools for different frequency bands to interpret the changes in the frequency response due to turn-to-turn short circuit. Moreover, the proposed method has been validated by comparing the obtained signature with the standard SFRA measurements. Additionally, a sensitivity analysis of the proposed method for two turn-to-turn short circuit severity is performed using EMTP simulation based on a high frequency model of the transformer winding.
Date of AwardJul 2022
Original languageAmerican English

Keywords

  • Transformers
  • fault diagnosis
  • SFRA
  • IFRA
  • frequency response
  • transient impulses.

Cite this

'