Statistical analysis of the AC breakdown voltages of ester based transformer oils

D. Martin, Z. D. Wang

    Research output: Contribution to journalArticlepeer-review

    Abstract

    Recent decades have seen the rapid and successful deployment of ester oils as transformer dielectrics. When evaluating the compatibility of ester oils with existing transformer insulation designs, it is essential to not only compare the mean breakdown voltages of the oils, but also to consider the role of dispersion and thus the withstand voltage levels (with the acceptable breakdown likelihood). Insulation designers sometimes estimate the withstand voltage of mineral oil from the dispersion of the data, assuming that the breakdown voltages of the oil follow a parametric distribution. However, there is a lack of discussion as to whether this method would be suitable for estimating the withstand voltages of ester oils. This paper uses samples of breakdown voltages of a synthetic ester, a natural ester and a mineral oil to analyze their distributions, and discusses the applicability of using Weibull and Gaussian distributions to estimate the withstand voltages of the esters. It is found that the distributions of ester breakdown voltages, in particular the lowest breakdown voltage, are similar to those of mineral oil. Consequently there is evidence that from the ac withstand voltage point of view the tested esters are compatible for the use in power transformers. © 2008 IEEE.
    Original languageEnglish
    Article number4591226
    Pages (from-to)1044-1050
    Number of pages6
    JournalIEEE Transactions on Dielectrics and Electrical Insulation
    Volume15
    Issue number4
    DOIs
    Publication statusPublished - Aug 2008

    Keywords

    • Dielectric breakdown
    • Oil insulation
    • Power transformers
    • Statistics

    Fingerprint

    Dive into the research topics of 'Statistical analysis of the AC breakdown voltages of ester based transformer oils'. Together they form a unique fingerprint.

    Cite this