Modeling the stochastic dependencies in a probabilistic load flow including wind generation

Nowadays, the implementation of environmentally-friendly and uncontrollable primary energy sources in the electrical power system production is increasing, introducing abatement in generation dispatchability. The incorporation of high levels of small-scale non-dispatchable distributed generation leads to the transition from the traditional 'vertical' power system structure to a 'horizontally-operated' power system, where the distribution networks contain both stochastic generation and load. This leads to a large number of stochastic system inputs and complex dependencies between them being involved in the analysis. In this paper, a Monte Carlo Simulation approach is used for the power system multivariate uncertainty analysis problem. The Stochastic Bounds Methodology is applied in order to tackle this problem through the definition of extreme dependence structures that can bound all real cases. On the other hand, the Joint Normal Transform models the exact dependence structure between the clusters based on the mutual correlations obtained by data analysis and provide more accurate results.