Modelling of Stochastically Dependent Forecast Errors for Load Flow Simulations in the Transmission Grid Using Convolution

The increasing share of renewable infeed results in a more volatile infeed situation. Especially in the grid security assessment, for example in transmission grid operational planning, the prognosis of the infeed from renewable energy systems becomes highly relevant. Uncertainties of this prognosis and possible dependencies have to be taken into account to ensure a stable system operation. Monte Carlo simulation is a common approach used for probabilistic load flow simulation. Due to the long computation time, this article proposes a probabilistic load flow approach using convolution to consider these uncertainties in simulations for large-scale transmission grids. Computation time plays an important role when uncertainties and their effects like cascading failures are to be considered. The methodology is validated using a comparison to deterministic approaches but has to be improved in dealing with correlations. In general, the methodology shows a good performance in simulating cascades.