Formation of design and operational databases on the reliability of equipment at Nuclear Fuel Cycle facilities

Abstract

The uniqueness of the nuclear fuel cycle facilities and their equipment, which must be adjusted to the specific conditions of manufacturing and (or) reprocessing the fuel of a particular reactor type in each case, complicates the application of traditional methods for assessing reliability and safety, thus relying on the availability of a significant operating experience. As a result, at the design stage, the reliability level of such equipment is determined by the available data on the reliability of analog elements. Following completion of commissioning of a nuclear fuel cycle facility, the regulatory requirements demand the updating of reliability and safety calculations to bring them in line with the actual state of the facility based on the results of the construction, commissioning, as well as pilot- and pilot-industrial operation. The article presents a universal method for accounting for the statistics of equipment operation at a high level of reliability in the absence of representative samples during the first years of its operation, which is compensated by the use of prior information within the framework of the Bayesian approach. The simulation of the time between failures is demonstrated using the synthesis of the prior information and the results of prototype bench tests for 5 years.