Five Hundred Tests on the Dielectric Strength of Oil

Abstract

While developing a dependable method of taking experimental data with a sphere gap in transil oil, considerable study was made of the behavior of oil under disruptive dielectric stress. The following notes on the dielectric strength of oil are offered as evidence that the nature and character of the dielectric breakdown of oil may be entirely different from that of air. Five hundred successive breakdowns were taken on a sphere-gap in oil at the same gap setting. Because of the well known inconsistency of breakdowns in oil these observations showed wide variation. A curve was plotted to show the relation between the breakdown voltage and the number of breakdowns at each voltage. If the disruptive breakdown of oil is due to the voltage exceeding the dielectric strength of the oil, as is the case with air, it should be possible to represent such a curve of probable error, or ``probability curve'' as it is usually called, by an exponential equation. In the following paper this is seen to be impossible, the most representative exponential curve being higher than the observations at higher voltages. The explanation is offered that this discrepancy is caused by foreign particles of low dielectric strength beinig drawn into the gap and that therefore the dielectric strength of oil differs from that of air in that it does not represent the true breakdowrn value of the oil but is instead a measure of the presence of foreign particles in the oil.