An Approach to Assess the 500 kV Insulator String Condition Using Ultraviolet Imager

The condition of each cup-and-pin insulator that forms a string must be considered in order to obtain effective guidance for contamination flashover mitigation. This paper describes an approach to assess the string condition based on the ultraviolet (UV) emitted by each cup-and-pin insulator during partial discharge condition. The UV patterns on the string were observed visually in the laboratory test. The criteria of string condition level was built based on the patterns and applied to the strings installed in the field. The index value expressing the string condition level in the field (level status) was then compared with the UV quantitative parameter extracted from the UV video. At the final stage, a logarithmic regression classification model was built based on these two values. The laboratory test results showed that there were three levels of string condition, namely level one (safest), level two, and level three (most dangerous). Most strings in the field had category level two, while most strings with level three were of the suspension type. The UV emission area parameters expressed in pixels and percentages were able to represent the sporadic nature of the discharge phenomenon. The proposed approach provided a predictive model with a mean absolute error of 0.182.

Automated summary

This paper presents an approach to assess the condition of cup-and-pin insulators in a string by analyzing the ultraviolet (UV) emission during partial discharge. The UV patterns observed in laboratory tests were used to establish criteria for string condition levels, which were then applied to field installations. A regression classification model was built based on the comparison between the index value representing the field string condition level and the UV quantitative parameter extracted from the UV video. The proposed approach provided a predictive model with a mean absolute error of 0.182.