Thermal-hydraulic reliability (TH-R) analysis for passive system requires repeated runs of thermal-hydraulic models or codes to get enough input/output data. Thermal-hydraulic models or codes of passive system are costly in terms of running time which results in slow computational speed and low estimation efficiency. To improve the estimation efficiency, surrogate models (or response surface) can be used to replace the real thermal-hydraulic model. During response surface construction, there are many input parameters, some undetected interactions or nonlinear effects involved in thermal-hydraulic model, which debase the accuracy and validity of the response surface. As a global sensitivity analysis method, Morris method is to determine factors which are important, nonlinear and non-additive, or involved in interactions with other factors. In this paper, screening key parameters for passive system is studied based on Morris method. The results of Morris method are insightful to select a reasonable surrogate model according to the interactions or non-linear characteristics of the model in the process of response surface construction. As a result, estimation efficiency of TH-R analysis for passive system can be improved greatly while ensuring the response surface’s accuracy.

This content is only available via PDF.
You do not currently have access to this content.