Studies on Sensitivity and Uncertainty Analyses for SCOPE and WAFT with Uncertainty Propagation Methods

摘要

The purpose of Steam condensation on cold plate experiment facility (SCOPE) and Water film test (WAFT) is to verify the steam condensation and water film evaporation correlation within the parameter variation range of CAP1400 passive containment cooling system. These correlations were used for containment response analysis. Uncertainty and sensitivity analyses were performed for SCOPE and WAFT tests in this paper. Sampling-based sensitivity analysis with uncertainty propagation is a new parameters sensitivity analysis method, and the importance of input parameters could be evaluated by calculating the correlation coefficients between input parameters and the output target parameter. This method was used to acquire the influence of the measured input parameters uncertainty on the output target parameter. The results show that air and steam flow rate, coolant flow rate, inlet and outlet water temperature are the main source of the uncertainty for SCOPE. Inlet film flow rate, inlet air flow velocity and plate surface temperatures are the main source of the uncertainty for WAFT. Sensitivity analysis results may provide support for experiment measurement system optimization to reduce the target parameter error range. Uncertainty analysis is one important aspect of test data analyses, which is meaningful to the assessment of test results. Conventionally the partial derivative with respect to the input parameters is used to transfer uncertainty from the input parameters to the output parameter. However, in this method the partial derivatives of the output parameter sub the input parameters must be calculated. For complex engineering problems, it is usually difficult to acquire theoretical correlations for the partial derivatives. WILKs formula is used to determine the parameter tolerance interval with certain probability content and confidence level. The tolerance interval is a good way to well describe the uncertainty of parameters. The nonparametric statistics with WILKS correlation were widely used in the best-estimate plus uncertainty (BEPU) accident analyses. However, little work has been conducted on the experiment results uncertainty analysis with that method. In this paper nonparametric statistics with WFLKS correlation was used to acquire key parameters uncertainty. And the results show that key output parameters uncertainty for SCOPE and WAFT are within the reasonable range. Uncertainty Propagation Methods were implied for test results Sensitivity and Uncertainty Analysis in the paper, which may be conveniently applied to the other experiment data analyses and also valuable to the engineering project.