📝 SummarXiv

Abstract

Nuclear data libraries serve as the foundation for all calculations in the nuclear field. Their quality directly affects the accuracy of computations. When new nuclear data libraries are released, they must undergo validation through the use of integral experimental data. This work introduces an automated simulation pipeline that converts raw nuclear data into predictions for various reactors, significantly enhancing the validation process. This pipeline allows for the assessment of potential biases in integral measurement predictions for nuclear power plants before the release of a new nuclear data library. To achieve this, an implementation of history variables in DRAGON was developed to determine the macroscopic cross sections of the nodal code PARCS. This method has been verified for seven depletion cycles across three pressurised water reactors (PWRs) against publicly available data. The errors on the predictions are in the uncertainty range attributed to variations in nuclear data. Additionally, a methodology to assess the quality of novel nuclear data libraries against experimental measurements in PWRs is presented.