Security-Constrained Stochastic Optimal Power Flow For Hybrid AC/DC Grids Considering Non-Gaussian Uncertainty

The integration of renewable energy sources (RES) introduces uncertainty into power systems, which contributes to grid congestion by necessitating higher operational margins. System operators manage congestion in real-time through costly remedial actions, such as generation redispatch and RES curtailment, or alternatively, by leveraging non-costly transmission grid flexibility. Even though effective use of this flexibility can minimize the dependence on costly actions, it requires congestion management strategies supported by models that accurately represent uncertainty while ensuring the system remains N-1 secure. To overcome this challenge, this paper proposes a Security-Constrained Stochastic Optimal Power Flow model based on Polynomial Chaos Expansion. This chance-constrained model minimizes the expected redispatch costs and RES curtailment by utilizing HVDC transmission flexibility in hybrid AC/DC grids. The model enables system operators to optimize the HVDC converter set-points, RES curtailment, and generator redispatch while accounting for continuous non-Gaussian uncertainty in RES forecast errors to ensure a realistic representation of uncertainty. Two case studies are presented on a 5-bus AC/DC test system. The first case investigates the impact of HVDC flexibility on redispatch costs and RES curtailment, analyzing preventive and corrective actions for topological uncertainties, as well as here-and-now and wait-and-see decisions for operational uncertainties. The second case highlights the limitations of assuming Gaussian input uncertainty by benchmarking it against the proposed model.