Numerical simulation tools are widely used in the nuclear field to evaluate the behavior of nuclear power plants during hypothetical severe accidents (SA), and the development and optimization of accident management measures aimed at preventing and mitigating the consequences of SA rely on the numerical simulation results using severe accident analysis programs. As SA tools can predict important outcomes such as the failure time of safety barriers and potential radiation source terms released into the environment, it is crucial to understand the accuracy of these quality factors and explain the inherited phenomenological uncertainty in modeling. This study takes the AP1000 reactor as an example, selects an accident sequence, uses the developed uncertainty calculation program to sample the severe accident uncertainty input parameters, calculates the input files generated by the sampling, and uses the uncertainty visualization display program to perform statistical analysis on the output parameters to obtain the impact of severe accident uncertainty on the accident process. The results of the analysis can provide reference for predicting the accident process and making decisions on mitigation measures in the event of a severe accident in a nuclear power plant.