Data-driven distributionally robust transmission expansion planning considering contingency-constrained generation reserve optimization

Releasing quick-response spinning-reserves yields a critical corrective implementation to deal with emergency conditions. Since the reserve capacity allocation is always omitted in transmission expansion planning problems, the robustness to uncertainty factors cannot be strictly guaranteed. In this context, a two-stage data-driven distributionally robust transmission expansion planning model is proposed to incorporate the pre- and postcontingency generation reserve optimization. For the sake of robustness enhancement, the bundled uncertainty of renewable output power, renewable probability distribution, and N-K security is simultaneously guarded against in the planning procedure. Specifically, the first-stage problem decides on transmission expansions and allocates the preventive reserve capacity. Afterwards, limited by the planning schemes and available reserves, the post-contingency operation schedules under bundled uncertainty are produced with optimal reserve utilization in the second-stage problem. Moreover, without any assumption of renewable probability distribution, a data-driven methodology based on the mixed confidence uncertainty set is employed to describe the probability fluctuations, which enables to improve the model conservativeness by inserting more historical data. Furthermore, a parallel column-and-constraint generation (C&CG) algorithm is presented to accelerate the solution process. At last, the proposed method is applied in both IEEE 24-bus and 118-bus test systems to derive the planning schemes, which strongly demonstrates the robustness and economic efficiency under bundled uncertainty.