Abstract: The paper makes the study of the nuclear power plant main pipeline.The residual-life of the main pipeline that failed due to thermal aging has been studied by the use of performance degradation theory and Bayesian updating methods.Firstly,the thermal aging impact property degradation process of the main pipeline austenitic stainless steel has been analyzed by the accelerated thermal aging test data.Then,the thermal aging residual-life prediction model based on the impact property degradation data is built by Bayesian updating methods.Finally,these models are applied in practical situations.It is shown that the proposed methods are feasible and the prediction accuracy meets the needs of the project.Also,it provides a foundation for the scientific management of aging management of the main pipeline.