The standard 1.2/50 lightning impulse test has been important for factory testing equipment and evaluating their performance in electrical systems. In practice, however, such equipment is subject to non-standard waveform (NSW) of high frequencies, which may not be covered by standard tests. To overcome this difficulty, the present article outlines a method that simulates NSW in laboratory tests. Based on multiresolution wavelet analysis, the method considers the correlation between amplitude and entropy in the decomposition levels between waveforms. To that end, changes are proposed to the following parameters of standard impulse tests: amplitude, cut-off instant, and front time. The results show that, mostly, the combination of a lightning impulse waveform (LIW) with impulse waves cut at different times, sufficiently represent the NSW. Only when the NSW's density falls predominantly in a frequency range above 3.125 MHz, will there be need for a complementary LIW with a steep front time. The method's application allows manufacturers and concessionaires to subject the equipment to conditions that properly represent the NSW effects, through lightning impulse tests carried out at the factory.
Read full abstract