Study on DC series arc fault in photovoltaic systems for condition monitoring purpose

Photovoltaic (PV) systems are increasingly being used. Due to ageing effects and also the trend toward higher DC voltage level, incidents of DC arc faults in PV systems are becoming more common, which have serious impacts on system stability and human safety. Parallel arcs draw high current compared to series arc faults and so detection of the latter is more challenging. This paper attempts to extract the features of DC series arc fault for condition monitoring purpose, achieved through studying the arc characteristics, carrying out simulation in Matlab/Simulink and laboratory experiments. Different arc models are investigated, and a set of parameters for the heuristic model is formulated for low current arcs. It is shown that the simulated arc by the heuristic model is consistent with experimental data. The arc noise features under electrodes' separation region and steady-arcing states with different gap width are investigated. It is found that the arc noise floor will increase after arc fault occurrence, especially for the frequency band below 50 kHz. This characteristic can be exploited for series DC arc fault detection. Moreover, frequency contents below 5 kHz are relatively more sensitive to change of air gap width.