Series DC Arc Fault Detection Algorithm for DC Microgrids Using Relative Magnitude Comparison

This paper describes a detection algorithm for series dc arc faults to improve safety issues in dc microgrids. Series arc faults typically occur when gaps form in a current-flowing wire or contactor. Arc faults can create serious fire hazards if not detected and removed in a short time. To identify series arc faults accurately and quickly in dc microgrids, a series dc arc fault detection algorithm using relative magnitude comparison is proposed. The proposed algorithm detects the time of occurrence of an arc fault based on the magnitudes of the load current in both the time and frequency domains simultaneously. The algorithm accurately works in a dc microgrid with multiple switching converters, as it relies only on relative current information. The proposed detection algorithm is experimentally verified on a 5-kW 380 V dc microgrid prototype with an arc fault generation unit.