Analysis of Signal Processing Techniques for High Impedance Fault Detection in Distribution Systems

High Impedance Faults (HIFs) occur by the contact between an energized conductor and a high impedance surface. Due to the low fault current level, HIFs cannot be detected by conventional protection and there is no fully efficient solution to this problem. HIF detection methods often extract metrics using signal processing techniques, such as Fourier Transform, Wavelet Transform, Stockwell Transform, and Mathematical Morphology. However, these techniques are applied under specific conditions, which hinders comparative and critical analyses among them. Therefore, this paper presents a critical review of HIF detection methods based on the aforementioned techniques, and also shows a detailed investigation of the performance of the metrics commonly used with them. To do this efficiently, the authors proposed a set of assessment indices based on the ratio between the metrics' characteristics and another one based on the repeating of the metrics features. The proposed indices revealed that some of these metrics fail to distinguish HIF from other typical occurrences in power distribution systems, and their performances are negatively affected by the fault location and by the existence of noise in the measurements. Additionally, the results showed a need to specify system conditions in which any HIF detection technique is valid.

Automated summary

High Impedance Faults (HIFs) occur when an energized conductor comes into contact with a surface of high impedance, and conventional protection methods are unable to detect them due to low fault current levels. Various signal processing techniques, such as Fourier Transform and Wavelet Transform, are used for HIF detection, but their application under specific conditions hinders comparative analyses. The study proposes assessment indices based on metrics' characteristics and their features to evaluate the performance of commonly used HIF detection methods, revealing shortcomings in distinguishing HIF from other system occurrences. System conditions and noise levels significantly impact the effectiveness of these techniques.