A CDORR-Based Reliability-Oriented Method for Spinning Reserve Allocation Under Multiple Sources of Uncertainty

With the increasing penetration of renewable energy and the intensification of load fluctuations, power electrical equipment frequently operates under complex conditions such as heavy loads. Traditional spinning reserve allocation methods based on fixed outage rates are inadequate for accurately characterizing the dynamic operational risks of the system. To address this issue, this paper proposes an optimal spinning reserve allocation method for power systems considering operational reliability. First, by introducing the condition-dependent outage replacement rate (CDORR), a reliability model under multi-uncertainties is established, which considers load forecast errors, wind power randomness, and equipment operating conditions. Second, to overcome the “curse of dimensionality” caused by alternating iterative solving in complex scenarios, analytical expected energy not served (EENS) formulations based on the sensitivity method are derived to rapidly quantify the power flow variations between pre-contingency and post-contingency states. Finally, aimed at the non-convexity introduced by bilinear terms in the model, an improved segmented McCormick envelope relaxation algorithm is designed, which is combined with the Tangent Plane Cut Collection (TPCC) strategy to effectively eliminate redundant envelope regions. Case studies based on the IEEE 30-bus system demonstrate that, while ensuring evaluation accuracy, the proposed method accelerates computational speed by nearly 20 times compared to traditional nonlinear solvers. Moreover, it effectively mitigates potential operational risks under complex conditions, achieving an optimal balance between system reliability and economics.