Influence of test methods on dc flashover performance of ice-covered composite insulators and statistical analysis

Operational statistics and studies have revealed that, the reduction of insulation strength of transmission line insulators caused by icing is one of the critical reasons of the power grid faults. This paper compares of various test methods for evaluating the flashover characteristic of ice-covered composite insulators. Two types of composite insulators, i.e. FXBW-±800/530 short sample and FXBW-110/100 are investigated. The influences of various test methods on flashover performance of ice-covered insulators are studied in an artificial climate chamber. The test results revealed the influences are significant. The average flashover voltage (U_av) using even raising method is highest, followed by the 50% withstand voltage (U_50%) which is corresponding to up and down method, the lowest flashover voltage (U_f.min) using Ucurve method is the third, and the maximum withstand voltage (U_ws) using the maximum withstand voltage method is the lowest. Through the statistical analysis, the U_50% can be estimated by U_av and σ_av though the estimated value U_e (U_e=(U_av-k₁σ_av)/( 1+k₂δ_e)). The estimated errors (E_e) are less than 1.80% in this paper. Therefore, there is an intrinsic relationship among the test results obtained by using various test methods. This simplifies selection of the appropriate test method according to the testing specific circumstances.