Abstract
Consequence analysis, including the modelling of corrective actions, is an important component when performing power system reliability analyses. Using an integrated methodology for power system reliability analysis, we investigate the impact of different modelling choices for the consequence analysis on estimates for the energy not supplied. These investigations corroborate the large impact modelling assumptions for corrective actions have on the resulting reliability indices. We have also identified other features of the consequence analysis, such as islanding and distributed slack, that can be important to take into account. The findings and the underlying structured approach contribute to improving the accuracy of power system reliability analyses.