Distributionally Robust Generation Expansion Planning Considering Large-scale Displacement of Fossil Fuel-Fired Units

With the large-scale displacement of fossil fuel-tired units due to the increasing penetration of renewable generation, the generation expansion planning (GEP) problem has been facing a major challenge to deal with the renewable power uncertainty. To handle this issue, this paper introduces a distributionally robust GEP model which explicitly considering the large-scale displacement of fossil fuel-fired units. In the first stage, the long-period generation expansion planning takes into account the expectation of the worst case; While, in the second stage, the short-term operation is modeled based on the distributionally robust optimization (DRO). The wind power uncertainties are treated as ambiguity sets using the Wasserstein distance technique. Aiming to ease the computational complexity of the proposed model, Benders decomposition with a reformulation technique to deal with the worst-case constraints is introduced. Case studies on a modified IEEE RTS-79 system verify the effectiveness of the proposed approach.