Verification of the Aging Impact of MOV Ceramics by Recording Wide-Range V/I Characteristics

Abstract

Nowadays, varistors are widely used because of their energy absorption capability (EAC) with concomitant economy and efficiency. Surges and overvoltage transients unnoticeably degrade Metal Oxide Varistors (MOV) over time. A newly developed condition monitoring system is based on a time-resolved measurement of the discharge current. In order to be able to compare and interpret acquired data sets from the monitoring system with relevant aging models, extensive investigations of the measurable influences are required. Therefore, an experimental setup has been developed for determining different aging characteristics and process influences. A precise continuous curve recording over nine current decades from the ohmic leakage current area to the limiting resistance region up to several thousand Amperes allows studying aging characteristics of differently stressed MOVs. This method combines common spot-measurements from the curves like mA-Point, leakage current and the nonlinearity exponent alpha. In this paper, a fully automated test bench is presented. The performance of the system and its advantages are compared to manually operated tests and different test equipment. First experimental results show that single point measurements, depending on the type of stress, are insufficient to detect degradation characteristics. Especially in the transition area from the ohmic linear range to the non-linear range, irregularities terms of degradation are ascertained. These studies are the basis for building up relevant aging models depending on the process measurement variables. This results in more accurate predictions for the state of health of varistor ceramics by using recorded data sets as the time-resolved measurement of the discharge current. Finally, it helps to understand the change of characterization by different impacts. In addition, the system is scalable for the study of layered surge arresters and thus applicable to all network levels.