The Development of a Method for Validation of the Uncertainties of the Models Used in Thermal-Hydraulic Codes

Abstract

The article deals with one of the most crucial and, at the same time, problematic stages of implementing the uncertainty analysis of thermal-hydraulic calculations by the GRS method—the validation of the statistical characteristics of model uncertainties—namely, of the circuit thermal-hydraulic block’s closing relations. The above relations developed, as a rule, on the semiempirical approach, are one of the main sources of uncertainty of the results calculated by the code. A task of developing the method proposed by the authors earlier that allows performing the validation procedure using verification technology and objective statistical criteria was set. The method in question allows going from an expert evaluation to the objective assessment of the model parameter uncertainties, thus enhancing the objectivity of the calculated results obtained within the framework of the GRS method. Particular problems are formulated the solution of which expands the capabilities of the method to use experimental information and to assess the uncertainties of the model parameters that cannot be directly measured experimentally. This, in turn, allows using the method for validation of the uncertainty of all most-critical model parameters of the design code. The stages of the proposed method are set forth sequentially followed by discussion of the topical problems that arise at every stage. Using practical examples, the ways of solving the formulated problems are demonstrated. The results obtained using the method for evaluation of uncertainties of a series of model parameters (closing relations) of the KORSAR thermal-hydraulic design code are provided. The results are recommended for use when analyzing uncertainties within the framework of the KORSAR code by the GRS method. Moreover, the proposed method can also be applied to other design codes.