Statistical inference for breakdown voltage in SF6 GIS from first breakdown data

The statistical distributions of first ac breakdown field data were shown to be represented by the three-parameter Weibull distribution. Many experimental data compiled from the literature in various gap configurations and in sphere-plane gap system in the present work, obtained in SF6 at pressures of 0.2, 0.4 and 0.6 MPa were analyzed. By making use of the maximum likelihood method, the algorithm suggested provides parameter estimates with no computational difficulties. The compiled field data produces lower parameter estimates compared to those evaluated for individual data groups at the same pressure, and although the limiting field strengths tend to decrease as gas pressure increases, the slopes of the distributions are almost unaffected.