Correlation between degree of polymerization and data from transformer oil analysis
Kjani Guri
The aging of transformers increases the possibility of failures in the insulation system and structural parts. In this case, the insulation system is the most attacked, because the main component is cellulose. The condition of the insulating paper can be identified based on the degree of polymerization (DP) measurement. In some cases, when the degree of polymerization measurement is difficult, the degree of polymerization estimation can be done by collecting information on some of the test parameters, such as oil acidity, moisture content, and furan concentration in the specific 2-furfuraldehyde. In this paper, to build a mathematical model and correlation between oil analysis, the multiple linear regression method was used, where 50 samples of transformers were taken in the study, to create a relationship that connects the degree of polymerization with by-products during the paper degradation process. The empirical formula for calculating the degree of polymerization based on the concentration of 2-furaldehyde was developed and gave promising results to calculate the degree of polymerization and the correlations that exist between these two components. The results show an estimation error of less than 4.06 % for the proposed logarithmic equation with a high correlation coefficient of 0.9995 and an estimation error of 7.21 % for the proposed polynomial equation, against the measured DP with a coefficient excellent correlation of 0.983, excluding oil, acidity, and moisture content.
Cite this article as:
Guri K., Correlation between degree of polymerization and data from transformer oil analysis. Electrotechnica & Electronica (Е+Е), Vol. 59 (1), 2024, pp.24-32, ISSN: 0861-4717 (Print), 2603-5421 (Online)
