Target accuracy requirements for MYRRHA

Research in nuclear physics and the applications derived from it require reliable data. Basically, two types of data are needed, those describing the interaction of particles with the nuclei conforming the system; and those describing the nuclear changes in the components of the system, as a result of fission, activation or decay. All these data are considered nuclear data. There is currently an increasing demand for nuclear data not only from scientific institutions, but also from industry and regulators, as they are essential for the design and safety analysis of advanced fast reactors and for the development of strategies for the best management of spent nuclear fuel.
The MYRRHA reactor has been conceived as an ADS but can also operate on its critical configuration when decoupled from the accelerator and adding a certain number of fuel assemblies. In this thesis, the main objetive is to perform a Target Accuracy Requirements analysis (TAR) for MYRRHA, that is the inverse problem of the uncertainty quantification. For this purpose, the project has been divided into 4 phases. First, a homogenized and a RZ model of the critical configuration have been developed. These models have been verified by comparing their main reactor parameters and the netruon flux spectrum with those of the heterogenous model. Afterwards, a sensitivity analysis of keff and βeff has also been carried out obtaining the list of the ten most important neutron induced reactions and nuclides. Morevover a sensitivity analysis of some reactivity coefficients such as void, Doppler and the Power Peaking Factors is also presented. Thirdly, uncertainty in all these parameters and coefficients have been quantified. This allowed to justify the need of the TAR analysis for keff. The analysis showed the importance of selecting an appropriate energy grid sctructure to get accurate results. Finally, a method developed in Python to perform TAR analysis is presented.
The implemented method is very innovative because, unlike the codes used in previous studies, it is very easy to use and it is open and free for any potential user. It also presents new features that previous programs were not capable of performing, such as analysing several parameters of the same reactor at the same time, analysing several reactors at the same time or modifying the values of the cost factors for each energy group.