Preemptive Medium-Low Voltage Arc Flash Detection with Geometric Distribution Analysis on Magnetic Field

As power distribution systems become increasingly interconnected and concentrated, the risk of arc flash events in low magnitude arc faults are attracting more attentions. Since such event can occur without causing sufficiently high current due to high arcing resistance and weak system source, the fault is difficult to detect by over-current protection. Therefore, the early stage low-magnitude arc fault detection and potential hazard level evaluation are essentially to be investigated. To overcome the limitations of traditional arc fault detection method, such as low sensitivity, delayed detection, complex structure, susceptible to interference of other sources, etc., meanwhile improves response time and minimizes false alarms, this paper explores a novel arc flash detection mechanism, which utilized the magnetic sensor array based spatial magnetic field analysis, to be able to precisely locate arcing position, evaluate the arcing current with associated potential hazard level and take action to prevent faults in the apparatus. With a series actual cases based simulations are performed, the proposed method shows a better ability to detecting the dynamic behaviors of arc transients, especially on the forming stage of arc flash, and easy to be used in estimation of incident energy level and the possibly flow direction. These results offer valuable information to develop reliable system and techniques to evaluate and mitigate the the potential hazards during an arc flash.