Adaptive robust AC optimal power flow considering intrahour uncertainties

Given the increasing share of variable renewable energy resources (VREs), power system operations need to account for the associated uncertainty with a fine resolution. This paper formulates an adaptive robust optimal power flow, which secures the hourly schedule against uncertain intrahour power injections. The uncertainty is characterized by spatially correlated polytopic sets. Second-order cone programming relaxation is employed to address the nonconvexity of power flow constraints. A sequential convex programming (SCP) procedure is developed to close the relaxation gaps. Due to convexity, the vertices fully represent the uncertainty sets, which alleviates the computational complexity stemming from full recourse. The effectiveness of the proposed solution framework is verified on 14-, 118-, and 588-bus systems with 80% VRE penetration and various uncertainty sizes. The SCP procedure recovers high-quality AC-feasible solutions in 3-17 iterations within 0.1%-41.4% of the planning horizon time span, which makes it suitable for practical use. The robust optimization can prevent load shedding and reduce operational costs by 2.0%-13.6%, while incurring 2.5%-5.0% reduction in VRE utilization.