Locating power transformer fault based on sweep frequency response measurement by a novel multistage approach. Issue 8 (1st November 2018)

• Record Type: Journal Article
• Authors: Jahan, Mohammad Sadegh; Keypour, Reza; Izadfar, Hamid Reza; Keshavarzi, Mohammad Taghi
• Is Part Of: IET science, measurement & technology. Volume 12:Issue 8(2018)
• Page Start: 949
• Page End: 957
• Publication Date: 2018-11-01
• Subjects: power transformers -- Fourier series -- transformer windings -- fault diagnosis -- optimisation -- matrix algebra; magnetic optimisation algorithm -- power transformer fault -- sweep frequency response measurement -- disk space variation -- mechanical damage -- two‐level method -- sweep frequency response analysis -- multistage method -- Fourier series coefficients -- electromagnetic forces -- data analysis -- winding movement model -- fault detector algorithm -- coupled network matrices -- voltage 64 kV; Measurement -- Periodicals; Electrical engineering -- Periodicals; Electronics -- Periodicals; Nanotechnology -- Periodicals; Electromagnetism -- Periodicals; Medical instruments and apparatus -- Periodicals; 621.3
• DOI: 10.1049/iet-smt.2018.0003
• Languages: English
• ISSNs: 1751-8822

Abstract

Abstract : The windings of the power transformers are prone to deformation or displacement as a result of huge electromagnetic forces. Regular monitoring of transformers for detecting minor winding movements is crucial for preventing them from developing into major faults and eventually, causing an outage. In this study, a novel procedure is proposed to determine the location and severity of the disk space variation, a frequently occurring mechanical damage in transformers, by developing a two‐level method. At the first level, an accurate model of the winding is determined from sweep frequency response analysis measured data corresponding to the healthy status. Analysis and interpretation of the obtained results by the proposed fault detector algorithm lead to locating the position and severity of the fault, in the next level. The process of determining the winding model from the frequency response analysis data is achieved by a multistage method utilising an enhanced version of magnetic optimisation algorithm, which comprises of utilising Fourier series coefficients and phase functions, and fast calculation of coupled network matrices. Experimental results of two 64 kV transformer windings verify the effectiveness of the proposed method in detecting the winding faults.