Evaluation of overshoot rate of lightning impulse withstand voltage test waveform based on new base curve fitting methods

The k-factor, which evaluates an overshoot of the withstand voltage test waveform in a lightning impulse withstand voltage test, is about to be adopted in IEC 60060-1 that specifies high-voltage test techniques for electric power equipment. In its procedure, it is defined that the recorded waveforms are fitted with a double exponential function to derive a base curve. For some waveforms, however, an extracted base curve sometimes deviates upward from the central lines of the recorded waveforms in the wavefront area. In particular, the degree of such dissociation is significant in waveforms with a high overshoot rate and a low frequency, and further it can not be said that the base curve is appropriate in terms of insulating properties. Accordingly, such significant dissociation may result in an irrational calculation of the overshoot rate. To resolve these problems, a numerical equation resulting from a relationship of a solution between an equivalent electric circuit and dominant equation was used, and fitting methods were overviewed in order to extract a more reasonable base curve in previous studies. In this paper, various options from these fitting methods were narrowed down to several options, which were then applied to simulated recorded-waveform; for which the overshoot rate and frequency of superimposed oscillating served as parameters, and a fitting method for extracting a more reasonable base curve was examined. Consequently, the calculation results of the overshoot rate were brought closer to their actual values than with the existing method by using a new base curve extraction method by which the fitting can be performed with removing the oscillatory part from the recorded waveform. While there is significant dissociation in the existing method, especially in the waveform on which a relatively low frequency oscillatory wave is superimposed, evaluation results of the overshoot rate were significantly improved as a result of the use of the new base curve extraction method. In the meantime, it should be noted that the k-factor filtering scheme was implemented after the application of various fitting methods resulted in the shape parameters of the test voltage waveforms, such as crest value, wavefront and wavetail durations, emerging as virtually identical, alongside confirmation of the minor effect of the fitting methods on the test voltage waveform.