Abstract: Risk early warning is the key to system security. To develop an analysis and calculation system is an urgent task to meet the requirements of current large-scale risk early warning. This paper proposes a strategy of power grid real-time risk early warning system based on failure rate and cloud platform, which improves the accuracy and real-time of large-scale risk early warning from the perspective of software and hardware. First, the real-time failure rate under the condition of equipment condition change is analyzed. The N–1 security checks for high-risk equipment are calculated one by one. Then a real-time contingency set are constructed. The power grid risk warning is realized through the analysis of contingency. And it is implemented in the cloud platform based on Docker technology. Finally, an example is given to verify the effectiveness and rapidity of the strategy.