A novel approach to account for uncertainty and correlations in probabilistic power flow

Abstract

Power system security assessment techniques can greatly benefit from the adoption of probabilistic approaches, as these can thoroughly describe the effect of uncertainties on the operating condition. The paper proposes a methodology to evaluate the probability of exceeding operational constraints under normal and contingency situations, accounting for uncertain, correlated power injections. In the proposed methodology, an analytical technique for probabilistic power flow is combined with copulas, the latter being introduced to deal with correlations. Copulas allow representing in an effective way the correlation information among variables, i.e. the multi-variate distributions of stochastic variables. A test system based on a realistic transmission network is used to exemplify the method.