Comparison of generator circuit-breaker stresses in test laboratory and real service condition

The paper shows that there may be substantial discrepancies regarding delayed current zeros and circuit-breaker arc energies when exposing a generator circuit-breaker to a short-circuit in a real system or in a test lab. Depending on the pre-loading of the generator in the real system the degree of asymmetry may be far higher than that being adjustable in a test lab. Based on realistic generator data the courses of the short-circuit currents have been simulated for various cases also taking into account the arc voltages of different circuit-breaker types. For the same breaking current zero crossings may be delayed for several periods and arc energies may be substantially different in the real system compared to the test lab. Improved generator modeling is indispensable for correct simulation of such phenomena and explains more over "voltage depression".

I. Introduction

This present treatise endeavors to show that the stresses applied to a generator circuit-breaker in practical applications are not always correctly represented by short-circuit breaking tests performed with a short-circuit generator. The reason for this discrepancy lies in the asymmetrical nature of the short-circuit currents. The degree of asymmetry, and hence the occurrence of delayed current zeros, is not only dependent on the generator characteristics but also greatly influenced by its loading conditions. The existence of this dependency is generally recognized [1] and is treated by the relevant Standards. However the analogies drawn in [2] and the prevailing ideas of the effect of the initial loading are, at least in part, incorrect.