Use of a Decision Tree to Define Transformer Oil Contamination from On-Load Tap-Changer Gases to Ensure Power Quality in the Grid
S. Bustamante(1), J. L. M. Lastra(2), M. Manana(1), and A. Arroyo(1) 1. Department of Electrical and Energy Engineering. Universidad de Cantabria. Santander, Spain. 2. Faculty of Engineering and Natural Sciences. Tampere University Tampere, Finland 2026-01-20

Abstract

Power transformers are one of the most importantand critical assets in the electricity distribution and transmission network. Power quality (PQ) can be disturbed when a power transformer is brought into or out of service, so it is very important to be sure of the reason for this action. Dissolved gas analysis (DGA) in oil can be used to diagnose the condition of transformer insulation. There may be situations where the DGA results indicate the presence of a serious fault which would lead to the transformer being taken out of service, when in fact the high gas concentrations are due to the leakage into the main oil tank of gases generated in the on-load tap-changer (OLTC) during normal operation. In previous work, using machine learning techniques and a distribution system operator's DGA database, a decision tree (DT) was developed to identify oil contamination from OLTC gases. In this work, the developed DT is applied to a new DGA database to identify contaminated transformers and test its accuracy. A total of 1161 DGA results from 95 transformers with OLTC were used, giving an initial DT accuracy of 83.13% when all samples were analysed and 85.26% when the last DGA result from each transformer was used.

Key words: Communicating OLTC, dissolved gasanalysis, maintenance management, oil insulation, power transformer.

References

[1] R. Kumar, B. Singh, R. Kumar, and S. Marwaha, “Recognition of underlying causes of power quality disturbances using stockwell transform,” IEEE Transactions on Instrumentation and Measurement (2020), vol. 69, no. 6, pp. 2798-2807.

[2] Y. Ma, X. Xiao, and Y. Wang, “Identifying the root cause of power system disturbances based on waveform templates,” Electric Power Systems Research (2020), vol. 180.

[3] R. Martinez, P. Castro, A. Arroyo, M. Manana, N. Galan, F.S. Moreno, S. Bustamante, and A. Laso, “Techniques to Locate the Origin of Power Quality Disturbances in a Power System: A Review,” Sustainability (2022), vol. 14, 7428.

[4] IEC 60599:2022 Mineral oil-filled electrical equipment in service. Guidance on the interpretation of dissolved and free gases analysis (2022).

[5] IEEE Std C57.104-2019 Guide for the interpretation of gases generated in mineral oil-immersed transformers (Revision of IEEE Std C57.104-2008) (2019).

[6] M. Duval and J. Buchacz, “Identification of Arcing Faults in Paper and Oil in Transformers—Part I: Using the Duval Pentagons,” IEEE Electrical Insulation Magazine (2022), vol. 38, no. 1, pp. 19-23.

[7] M. Duval and J. Buchacz, “Gas Formation from Arcing Faults in Transformers—Part II,” in IEEE Electrical Insulation Magazine (2022), vol. 38, no. 6, pp. 12-15.

[8] T. Breckenridge, S. Ryder, Overview of transformer and reactor procurement, CIGRE Green Books, Springer, Cham (2022).

[9] S. Bustamante, M. Manana, A. Arroyo, A. Laso, and R. Martinez, “Determination of Transformer Oil Contamination from the OLTC Gases in the Power Transformers of a Distribution System Operator,” Applied Sciences (2020), vol. 10, 8897.

[10] Advances in DGA Interpretation, JWG D1/A2.47, Technical Brochure No. 771, CIGRE (2019).