Series DC Arc Fault Detection Method for PV Systems Employing Differential Power Processing Structure

Abstract

Arc fault detection is an important process for ensuring the safety of PV and grid-connected inverters and is essential for producing PV systems in real applications. However, it is difficult to detect series dc arc faults using conventional fuses because these faults produce only small current variations. In addition, the arc fault detection devices have limited performance, because they focused on arc fault detection between PV arrays and inverter. Moreover, they require the additional current and voltage sensors to implement the arc fault detection algorithm. This article proposes an arc fault detection algorithm employing differential power processing (DPP) structure. The proposed algorithm only uses intrinsic voltage sensors of DPP and inverter, which can improve the cost-effectiveness of PV systems. In addition, the proposed algorithm can integrate the functionality of maximum power processing for each PV panel and arc fault detection. The voltage relationship between DPP and inverter is analyzed to detect the arc fault condition according to the DPP voltage sensing type. The performance of DPP units and its voltage relationship are verified with the control hardware-in-the-loop system. The arc fault detection performance is verified with the prototype PV system.