Abstract: The current criterion for the configuration of operating reserve capacity is based on the load percentage and maximum single-unit capacity, which can result in insufficient or excessively abundant reserve capacity in cases with a high proportion of renewable energy access. To address the challenge of accurately assessing the reserve capacity demand for high proportions of renewable energy grid operation during current dispatching operations, this study proposes a probabilistic dynamic assessment method for operating reserve requirements based on kernel density estimation. Furthermore, a dynamic confidence selection strategy considering the renewable energy penetration rate and evaluation period is proposed, which can facilitate the rolling differential evaluation of the up-and down-regulation of reserve capacity demand. Based on this proposal, a strategy for formulating conventional power supply startup capacity that considers dynamic reserve capacity demands was developed. This strategy maximizes the space for renewable energy consumption while ensuring sufficient operation reserve capacity. An application test was conducted using a provincial power grid with a high proportion of renewable energy sources, and the effectiveness of the proposed method was verified.