Detection of DC Arc-Faults in Battery Energy Storage Systems

This paper proposes a new DC Arc-fault Detection method in battery modules using Decomposed Open-Close Alternating Sequence (DOCAS) based morphological filters. The proposed method relies on the State of health, state of charge and temperature measurements from battery management systems (BMS). The detailed electrochemical model of the battery is used, and the proposed DC arc-fault detector identifies the existence of an arcing fault by generating spikes at its output. The algorithm is enhanced by an adaptive threshold classifier for noise suppression. The proposed method is tested for different types of series and parallel DC arc-faults in Battery Energy storage systems. The simulation results demonstrate that effectiveness of the proposed method.