A method is developed for performing statistical analysis of the uncertainty of the thermophysical calculations. This method makes it possible to construct with a prescribed reliability the confidence interval for the estimated parameter, determine the factors which have a large effect on the computational result, and estimate how close the variant calculations are to the base (undeviated) calculation. The methodology has been approved, within the framework of a deeper substantiation of the safety of the No. 1 unit of the Kursk nuclear power plant, for analysis of two types of accidents, using the RELAP5/MOD3.2 computer code, with rupture of the group distribution collector.

No abstract

A library of electronic databases for thermophysical data is developed using Microsoft Access. This library includes 13 databases containing experimental data. Existing experimental data and their reduction to a unified electronic form for maintaining a large volume of data in their real state were collected and systematized to create the library. The proposed organization of the information greatly facilitates statistical analysis of the experimental data and the computational uncertainty for assessing the safety of operating nuclear power plants and verifying computer codes. Data on the thermohydraulics of RBMK fuel assemblies, critical efflux, and transient regimes in operating power-generating units and data for three-dimensional thermohydraulic calculations of rod assemblies have been placed into the library.A library of electronic databases for thermophysical data has been under construction for the last several years at the Research and Design Institute of the Electric Technology (NIKIÉT) as part of an agreement with Rosénergoatom. The team doing this work has set as its goal collecting and systematizing experimental data and reducing them to a unified electronic form, since many scientific archives exist only on paper (in the form of reports which are difficult to access).A common deficiency of the experimental data which are ordinarily available to anyone doing computations is the lack of exhaustive information on the experimental setup, the measurement apparatus (measurement errors), and the initial and boundary conditions for performing an experiment. To eliminate these deficiencies, we have drawn upon the most diverse sources (documents and reports) including direct contact with the experimenters (oral information).Another goal of this work was unification of the experimental data and maintaining a large volume of data in their real form. The fact that different sets of experimental data are used in computational modeling makes it more difficult to compare the computational results with one another, since the investigators performing an analysis of experiments exclude data which show the largest discrepancies between calculations and experiments or because of disagreement with a proposed expression. The goal of the library is to preserve the primary experimental information, which will not only make it possible to unify the computational modeling of each separate experiment but also facilitate statistical analysis of the data and analysis of the computational uncertainties [1, 2] when making a safety assessment of operating nuclear power plants and verifying computational codes, which answers the requirements of departmental documents [3].First of all, the data used for additional verification of domestic and foreign computational codes in application to RBMK reactors are placed in the library. The computational modeling in this case is concentrated on transient processes. The following phenomena are characteristic for such regimes:• outflow from a rupture in the circulation loop, including critical o...

No abstract

No abstract