Some of the factors affecting the dielectric strength of transformer oil are investigated. These factors include the stabilization phenomenon, oil and electrode pretreatment, the effect of oil velocity, the effect of a capacitance parallel to the test cell, and the effects of electrode area and gap spacing. The experimental apparatus and procedure are described, and the results are presented for each of the above factors. Suggestions for future research are offered. >

Breakdown of transformer oil

In order to examine the area and the volume effects on breakdown strength in liquid nitrogen (LN/sub 2/), we measured dc and ac breakdown voltages in LN/sub 2/ with sphere to plane and coaxial cylindrical electrode configurations We also carried out statistical analysis of the experimental results using the Weibull distribution and discussed with the statistical stressed electrode area (SSEA) and/or liquid volume (SSLV) The dc and ac breakdown strength in LN/sub 2/ decreased to 1/5 with increasing SSEA and SSLV over the wide range from 10/sup 0/to 10/sup 5/ mm/sup 2/ and from 10/sup -1/ to 10/sup 5/ mm/sup 3/, respectively The Weibull shape parameter m for the sphere to plane and the coaxial cylindrical electrodes was estimated to be 6 to 8 and 11 to 13, respectively; the breakdown strength in LN/sub 2/ saturated with increasing SSEA and SSLV From these results, it was verified that the area and the volume effects definitely determined the breakdown strength in LN/sub 2/ The results of the Weibull statistics were compared with the conventional concept of the stressed electrode area (SEA) and liquid volume (SLV) Consequently, SSEA or SSLV was found to be nearly equal to 80 to 85% of SEA or SLV >

Weibull statistical analysis of area and volume effects on the breakdown strength in liquid nitrogen

In an attempt to clarify the persisting controversy over the effect of the electrode area versus that of the stressed oil volume in large-oil-volume- breakdown, a study was carried out under well defined conditions of the oil quality, particularly with regard to the degree of particle contamination. The results indicate that both the electrode area and the stressed oil volume can affect the dielectric strength of transformer oil, with the stressed-oil-volume effect being most pronounced under particle contamination conditions. Test results with technically clean transformer oil as currently accepted for use in power apparatus indicate that the degree of particle contamination in these oils is sufficient to produce an observable stressed-oil-volume effect. Finally, it is demonstrated that the observed phenomenon can be interpreted in terms of an apparent effect of either the electrode area or the stressed oil volume. This has led to the development of a semi-empirical method of quantitatively assessing the breakdown phenomenon in large oil volumes with reasonable accuracy.

Statistical Dielectric Degradation of Large-Volume Oil-Insulation

The use of design criteria for assessing the stresses to which composite dielectric structures are subjected implies careful and informed application. The convention required in the application of these criteria is examined in the light of the physical processes expected in an oil/cellulose transformer structure and related to an equivalent stressed-oil-volume theory of breakdown. The magnitude of the stresses implied by such criteria has been gauged from an examination of a number of case studies in which finite-element analysis has been used to determine the cumulative stress characteristic. >

An assessment of the physical basis for the application of design criteria for dielectric structures

The efficiency of producing uniform fields in the test gap was evaluated for practical electrode configurations corresponding to the plane, the Rogowski and the Bruce profiles. The field intensity was calculated using the known charge simulation method and the calculated distributions of the field intensity at the surface of different electrode configurations were compared. The results obtained show a pronounced end effect at plane-profile electrodes. With the Rogowski profile, the effective uniform-field region is dependent on the gap spacing. The Bruce profile was found to be a good compromise, with a negligible end effect and a nearly constant uniform field region. Practical considerations are also presented together with the limitations of various profile configurations in the design of uniform field electrodes.

Electrode Design for Testing in Uniform Field Gaps

Breakdown of transformer oil is affected not only by electrostatic field configurations, but also by many other factors. The latters make it difficult to predict break- down voltage from electrostatic field, as made for air and S F6 gaps. Alternatively, many experimenters tried to express the breakdown stress of oil as a function of stressed oil volume. But, results are limited to their respective electrodes.

Y. Kawaguchi

Papers

Breakdown of Transformer Oil