The high randomness and volatility of wind power pose new challenges to the operational reliability of power systems. Traditional reliability assessment methods struggle to balance economic efficiency and robustness when considering renewable energy outputs, leading to overly optimistic or overly conservative evaluations under high wind power penetration. To address this issue, a modeling and assessment approach for the operational reliability of power systems with high wind power penetration is proposed. First, a two-stage data-driven distributionally robust minimum-cost model is developed, where the first stage determines the optimal dispatch scheme, and the second stage determines the minimum load shedding and wind curtailment. Conditional value-at-risk is introduced to characterize and constrain the operational risk of system branches. Second, uncertainty sets based on the 1-norm and ∞-norm are constructed using historical data of available wind power output, and the column-and-constraint generation algorithm is utilized to solve the model. Finally, simulations on the IEEE-RTS79 bus system are conducted to analyze the effectiveness and robustness of the proposed method in different wind power penetration scenarios.