Comparison of superimposed voltage tests with spark gap and coupling capacitor at a gas-insulated model arrangement

Superimposed voltage tests are widespread to qualify HVDC equipment. Two possible test circuits are used, which result in different voltage shapes: superimposition by a coupling capacitor or by a spark gap. So far, there is no data available, which compares the resulting breakdown voltages of both methods with each other. This paper closes this gap by determination of the breakdown voltage at a model arrangement using both methods. The investigation was focused on gas-insulated systems, which are a typical application in HVDC. The model arrangement was built such that it represents a typical gas-solid interface in a coaxial GIL assembly. HVDC effects were included by heating of the earthed electrode. The determination of reproducible data was challenging, but could be achieved by adaptions inside the test circuit, as well as a suitable test procedure. The collected data with the built test arrangement is presented and discussed further in this paper. Based on the results a first conclusion is given, whether the different voltage shapes resulting from the use of a coupling capacitor and a spark gap, respectively, also result in different breakdown voltages.