Measurement of Time Domain Parameters for Series Arc Fault Detection. Sensitivity Analysis in the Presence of Noise

This paper presents a sensitivity analysis concerning the measurement of some time-domain parameters previously proposed by authors for series arc fault detection in electrical circuits. The parameters under study were introduced in an earlier paper; their formulation is based on current signal acquisition and processing, by means of cross-correlation analysis between subsequent cycles of the current waveform derivative. In this paper the effectiveness of the aforesaid parameters is investigated, considering the issues related to their measurement and thresholding in the presence of noise. The analysis is carried out by evaluating the success rate of each parameter in detecting the fault, for different thresholds values and in the presence of different noise levels.

I. Introduction

Series arc faults (SAFs) in electrical circuits are a prominent cause of fire risk in civil and industrial buildings and plants, due to insulation defects in cables and sockets, bad terminal connections, damaged or aged appliances and power systems components and so on. To protect against such fire risk, different international standards and codes currently require the use of arc fault protection devices or interrupters (AFDDs/AFCIs) [1]–[4].