In order to accurately analyze the running features of load nodes with large-scale new energy sources access to the grid, according to the Thevenin equivalent theory, the electric network system including large-scale new energy sources can be equivalent to 3 nodes. On the basis of the equivalent system, the state vulnerability assessment method and index based on short-circuit capability for load nodes are built up, which makes the system of state vulnerability assessment more perfect. Based on the results of state vulnerability index, the PAM clustering algorithm and Homogeneity-Separation (HS) index are used to find out the nodes which need more attention. In addition, the improved IEEE-39 node system is simulated and analyzed. The result is compared with that of traditional voltage margin index assessment, simulation shows the effectiveness of this state vulnerability assessment method based on short-circuit capability and the rationality of using the PAM clustering algorithm to identify weak nodes, which have the application of on-line prospect. This work is supported by National Natural Science Foundation of China (No. 51177111).