SANDY application for the analysis of nuclear data covariance

Nuclear data form the basis of nuclear science and technology since they represent human knowledge of physical phenomena. These data are used for a wide range of applications: from personal dosimetry to nuclear power plants or radioactive waste management, to give a few examples. With the increasing computational power, simulations are becoming more and more accurate for more complex problems, but the basis remains the same: the numerical values of the nuclear data. Therefore, the updating and development of these data, as well as of codes that allow their optimization and analysis, is a priority for the nuclear industry. One of the current issues for the nuclear industry is radioactive waste management. After a reactor stops operating, high-level radioactive waste needs to be cooled for several years before long or intermediate storage. The spent fuel is a source of heat and if it is not properly removed, could melt the fuel, so its characterization and evolution of the spent nuclear fuel play a fundamental role in today's nuclear industry. In Europe, the work package WP8 of the EURAD project is trying to improve the spent fuel characterization and evolution until disposal. To achieve this, an upgrade in nuclear data is inevitable. One of the aims of this thesis is to contribute to the work package WP8 of the EURAD project by updating the fission yield data of the contributors to the 148Nd for the 235U thermal fission reaction, i.e., update the probability per isotope of contributing to the formation of 148Nd after thermal fission of 235U. The choice of the 148Nd isotope is because its production in the reactor is proportional to the fission rate, making it very important for spent fuel characterization. To achieve this goal, the SANDY code developed at SCK-CEN was used. SANDY is a code written in python that can read, write and perform a set of operations on nuclear data files. Its primary objective is to produce samples to propagate uncertainties, i.e., to quantify the variability of the output due to the variability of the input parameters, in this case, the variability of nuclear data file information. In addition, it is an open-source code available on GitHub, so this thesis will contribute to the scientific development of the SANDY code. Therefore, this thesis contributes to two major projects: the work package WP8 of the EURAD project and the scientific development of SANDY. In turn, the results of the thesis can also be divided depending on which project(s) it contributes. The most important results of the thesis consist of: • Check the consistency between different physical models and nuclear data related to fission yields.(EURAD and SANDY) • Apply the statistical procedure called General Least Squares (GLS) method to update the fission yield data. (EURAD and SANDY) • Demonstrate visually that the set of samples generated with SANDY keeps the information of the original file, together with a check of the sampling distribution and statistical convergence. (SANDY)