Renewable energy connecting to the grid increases the uncertainty of power system. In order to realize the risk assessment of power system, a novel method of line overload safety risk assessment under typical operating scenarios for wind power grid-connected system is proposed in this paper. Firstly, K-means algorithm is used to construct typical system scenarios. Furthermore, the risk indicator of line overload is evaluated combined with risk theory, which characterizes the fluctuation of the line power flow and the change of the system state. Finally, a simulation is carried out to calculate the risk indicators under typical scenarios. The results show that the changes in typical system scenarios directly affect the operation of power system and risk indicators. Among them, in scenarios where the wind power level is higher and the load level is lower, the risk indicators are higher.