Effect of numerical integration on critical time evaluation in power system stability studies

The effect of numerical integration on time-domain power system simulations when evaluating the fault critical clearing time (CCT) is discussed in relation to stability and accuracy of integration methods. The specificity of CCT simulations is that the power system is close to its physical stability limits. While decreasing the time step is widely thought to be an effective measure to prevent numerical instability and ensure sufficient accuracy, it is shown that in real stability studies, numerical phenomena can be non-monotonic and cause misrepresentation of power system transient stability. For a given method, the effect depends on a combination of the loading of the generator whose CCT is evaluated and fault and network characteristics. The effect is illustrated with real-world simulation results. The paper makes recommendations on how the quality of CCT evaluation can be enhanced.