Power grid risk identification is a crucial aspect of power grid dispatch and operation. With the increasing complexity of new power system operation modes and the coexistence of diverse grid structures

potential risks in grid configurations are continuously growing. Traditional manual risk identification methods are prone to omissions and misjudgments

suffer from low computational efficiency

and fail to meet the practical needs of real-world dispatch applications. This paper proposes a graph-computing-based method for rapid risk identification in grid operations

leveraging graph theory and the topological characteristics of power equipment failures. First

a hybrid topological graph model tailored for risk identification is constructed

significantly reducing the scale of the graph model while ensuring accuracy

thereby improving computational efficiency. Second

a topological analysis algorithm based on graph computing for risk identification is introduced

along with a detailed explanation of its parallel implementation. Finally

the effectiveness of the proposed method is validated using real-world data from a provincial power grid. The results demonstrate that the proposed risk identification method achieves high accuracy

with analysis speed increased by dozens of times

significantly enhancing the efficiency of topological analysis and meeting the computational efficiency requirements for risk identification in maintenance planning across multiple time scales in large-scale power grids.