Influence of voltage application history on insulation test with one-minute step-up method

Abstract

For the development of test conditions for a short-duration power-frequency withstand voltage test, which is one of the insulation specifications of substation equipment, there is a precise method of evaluating reliability using a Weibull distribution function. For this method, "independence method" and "accumulation method" have been proposed in order to handle the coexistence of multiple voltage levels. In this paper, an insulation test based on the one-minute step-up method as condition for the coexistence of multiple voltage levels is used to calculate and evaluate the test results using these two methods. Consequently, the insulation characteristics obtained from the one-minute step-up test are found to have differences from the true values. If the independence method is adopted, there is a certain deviation regardless of the Weibull time shape parameter "a-value" and, the smaller the Weibull voltage shape parameter "m-value", the larger the deviation. If the accumulation method is adopted, on the other hand, the larger the "a-value" or the smaller the "m-value", the larger the deviation. The 50% breakdown voltage value, the standard deviation "σ-value", and the "m-value" that can be obtained from the insulation test based on the one-minute step-up method were evaluated for their deviations from the true values. The results suggest that, when a test is conducted on an oil-filled transformer, the design may turn out to be on the dangerous side due to overestimation of the dielectric strength by around 8%.