Non-intrusive ultrasonic sound detection of internal arc fault in medium voltage switchgear

Abstract

Arc fault protection is becoming ever more critical for all power systems to maintain the safety of personnel and reduce damage to expensive equipment. An internal arc fault is an unintentional event within an enclosed power system where the released energy from the arc causes a rapid rise in temperature and pressure. The fast-acting arc fault detection sensors used today require access to the switchgear compartment.

It has become a topic of interest to investigate alternate arc fault detection, which utilizes other means of sensor technology that may prove similar in performance but are non-intrusive. Since an internal electric arc will generate high temperature at a fast rate, and the speed of sound is affected by temperature, the objective of this thesis is to investigate the feasibility of using non-intrusive ultrasonic transducers to detect an internal arc fault within medium voltage switchgear.

This thesis gives a theoretical understanding of the generation of electric arc in the air, limiting systems used in medium voltage switchgear, and how the theory of ultrasound, the speed of sound, and Lamb waves can be used for non-intrusive ultrasonic sound detection of internal arc fault in medium voltage switchgear. Simplified models of the theory were developed and simulated through 5 test cases in COMSOL Multiphysics ® with different parameters and conditions. The results from the simulations showed potential in non-intrusive ultrasonic detection of arc fault using Lamb waves, but should be further investigated with more detailed modeling and actual physical experiments to draw a decisive conclusion.