VeHIF current arcing, volatility and vegetation ignition development: An EMD based volatility analysis for earth-fault classification

Abstract

Vegetation High Impedance Fault (VeHIF) events pose severe risks to public safety, wildlife, and forests. Overcurrent protection schemes generally cannot clear VeHIF events, which have previously resulted in bushfires. This research presents a VeHIF classification method based on the statistical variance analysis of volatility in the empirical mode decomposed feeder current. The author applied the proposed scheme to a dataset of 130 earth-faults with 100 % success rate and a mean classification time of 10 s (standard deviation of 14.32 s). While faster VeHIF fault classification should ideally be achieved, research findings show that faster classification speeds may not always be feasible or necessary. This finding relates to the fact that bushfire risk increases, when charring starts to spread over a branch with a breakout of flames. The present work further validates a positive pairwise correlation between a fault current's volatility and arcing, demonstrating the efficacy of High Frequency (HF) signals as reliable VeHIF volatility signatures.