On the probabilistic evaluation of indices for power system security assessment

In power system security assessment, security indices or approximate system performance (ASP) models are used to predict the system security. The authors develop performance curves based on probabilistic methods which indicate the effectiveness of the various security indices used in steady state security assessment. These curves give the probability of detecting insecurity as a function of false alarm probability, thus showing the tradeoff obtained when a particular security criterion is selected. These performance curves are used to illustrate the change in the effectiveness of the indices as the operating scenario and the uncertainties change. A 14 bus system is used to illustrate the probabilistic evaluation of the three scalar indices for predicting thermal limit violations