Abstract: In order to reflect the impact of wind power fluctuation on the overall operation risk of the power system in the future time periods, this paper establishes the load shedding risk change indicators, the line overload risk change indicators and voltage violation risk change indictors, and designs a three-layer assessment indicators with a progressive relationship in the power system with large-scale wind power integration based on the value at risk theory. Firstly, the assessment indicators in the first layer such as the load shedding risk indicator, the line overload risk indicator and the voltage violation risk indicator are calculated based on the wind power prediction output sequence. Secondly, synthesizing the results of the first layer assessment indicators at all moments and using the logistic distribution and exponential distribution to describe the probability distribution of each indicator, the value at risk theory is introduced to calculate the second layer indicators that reflect the overall risk level of the system in the future time periods, and the comprehensive risk indicators that are used as the third layer indicators are introduced to comprehensively evaluate the power system operation risks. Finally, the rationality of the proposed model and method is verified with the analysis in the IEEE-RTS79 system. This paper further analyzes the impact of different wind farm access nodes or access capacities and wind farm centralized access or decentralized access on the system operation risks.