Abstract
The residual energy stored in DC line inductance is dissipated when a switch is opened because there is no natural zero crossing that allows for circuit interruption during periods when no current flows. To prevent switch damage due to line inductance energy induced DC switching surges, a parallel connected metal oxide varistor (MOV) has proven to be more viable in larger DC systems than using a freewheeling protection diode. The MOV device clamps the line inductance energy by absorbing and diverting accumulated surge charge away from an opening switch. Switch-off operations occur occasionally and therefore, sufficient time is available for absorbed MOV material surge energy to be dissipated as heat to ambiance. Under occasional DC switching surge conditions, a ZnO based MOV still degrades and its energy absorption capability reduces over time. However, the MOV should not fail before the expected DC switch lifetime operation. Available literature suggests that it is still not well understood how to represent and benchmark MOV resilience or durability for this specific energy handling capability to occasional DC switching surge degradation. In this study, the durability model estimate proposed is the average of the total accumulated conduction charge transfer through each MOV of a statistical significant sample that has exceeded a degradation level of 10 % change in 𝑉1𝑚𝐴 due to occasional DC switching surge degradation. To confirm across MOV manufactures, a total number of 270 low-voltage MOV devices from three different manufacturers with similar size and electrical characteristics, are individually degraded by 5000 generated DC switching surges. Three samples of 30 MOV devices for each manufacturer is degraded by different surge charge content sizes, where the DC switching surge amplitude remains constant but the surge duration of the three samples are set at 2 ms, 3 ms, and 4 ms respectively. The findings of this thesis are that the estimated average of the total accumulated charge transfer of the three samples analysed, are determined to be statistically equivalent even if the charge content size per DC switching surge varies. The statistical results obtained from confidence interval comparisons, hypothesis pairwise testing and single factor analysis of variance, indicates that the proposed durability model could not be rejected. This proposed durability model may be used by designers to estimate the number of DC switching surges and expected MOV lifetime protection. In addition, this model provides a more reliable way to compare durability capability between different MOV manufacturers with similar device specifications, for this specific type of degradation.