Data-Driven Characterization and Forecasting of Metal-Oxide Varistor Degradation in DC Circuit Breakers

Energy absorption devices play a critical role in semiconductor-based dc circuit breakers (DCCBs), where they generate a counter-voltage to reverse fault current direction and dissipate fault energy during the interruption process. The metal-oxide varistor (MOV) is the mainstream candidate for this role. However, since MOVs are primarily designed for electrical surge protection, their degradation patterns under dc fault interruption conditions, characterized by exposure to unique DCCB voltage and current waveforms, have not been thoroughly studied. Understanding and predicting MOV degradation (an indicator of MOV lifetime) is crucial for ensuring the reliability and longevity of DCCBs. Therefore, there is a need to (i) systematically test and characterize the degradation of MOVs during dc fault interruptions and (ii) develop forecasting methods to estimate their long-term degradation. In this work, MOV degradation data is collected using an experimental DCCB testbed designed for accelerated degradation testing. Based on this data, statistical analysis and machine learning techniques are then employed to analyze the data, providing insight into DCCB MOV degradation trends and enabling the forecasting of their long-term performance. This work lays the foundation for enhancing the operational reliability and life-cycle management of DCCBs.