Data-Driven Distributionally Robust Planning of Energy Hub Coupled with Carbon Capture Utilization and Storage Facilities

Toward carbon peaking and carbon neutrality goals in China, a transition from traditional fossil power to renewable energy becomes an imperative need. This paper aims to address the planning problem of a highly renewable-energy penetrated energy hub system, coupled with carbon capture facilities and downstream production processes of hydrogen and methanol, under the uncertainty of renewable power output. To this end, we propose a two-stage distributionally robust optimal planning model, where investment decisions are the first-stage variables and operation variables are determined at the second stage after the realization of uncertainty. In the proposed model, we utilize historical data samples to construct a strengthened ambiguity set that leverages both Wasserstein distance and moment information for a better characterization of renewable generation uncertainty. Moreover, to efficiently solve the proposed planning problem, we develop a simplified affine decision rule method in a general setting with both moment and Wasserstein ambiguity, and then apply it to reformulate the distributionally robust planning problem. The derived reformulation possesses the favorable property that its scale is free from the sample size, thus greatly reducing the computational complexity. Computational results validate the effectiveness of the proposed planning model and the computational efficiency of the solution methodology.