Uncertainty and sensitivity analyses for steady-state thermal–hydraulics of research reactors

Abstract

The assessment of uncertainty is a desirable capability for thermal–hydraulic codes allowing the achievement of uncertainty associated with any code calculation results. In this paper a methodology, based on multivariable Taylor series expansion, suitable for introducing such a capability into a thermal–hydraulic code, method of evaluation of thermal–hydraulic parameters and the main results obtained from its application are discussed and presented. The computer code, THYC, was developed using standard correlations adopted by International Atomic Energy Agency (IAEA) for the analyses of steady-state thermal–hydraulics of research reactors. For the purpose of accuracy evaluation and showing the applicability of the proposed code, the steady-state operating conditions of the IAEA nuclear research reactor were assessed at a power level of 2 MW. The steady-state thermal–hydraulic results obtained show close agreement between the results of the present code and those of benchmark calculations. The purpose of benchmark calculations was to check how reliable the results obtained are when the code is applied to identical cases. The sensitivity study for 31 main input variables identified the most important design parameters affecting the peak clad temperature, ONB and DNB heat fluxes for plate-type research reactors. Also, the analysis used for assessing sensitivity in steady-state thermal–hydraulics of research reactors provides a means of determining which design variables are of (a) dominant, (b) significant, and (c) minor importance in severe accident decision-making.