Multiphysics study of sodium-cooled fast neutron reactor FFTF

The next generation of nuclear reactors (Gen IV), is expected to provide answers to many of the problems limiting public acceptability of earlier commercial nuclear power plants. Amongst the proposed designs, the sodium-cooled fast reactor (SFR) is the most mature, and has been extensively studied at CEA, the French Atomic Energy and Alternative Energy Commission. One of the most dangerous accidental transient in a SFR is the Loss Of Flow WithOut Scram (LOFWOS). To study how to limit its consequences, several experiments were carried out in the 1980s on an early research fast reactor in the USA, named FFTF. A benchmark based on one of these tests (LOFWOS Test #13) has been lately proposed by the International Atomic Energy Agency (IAEA) to improve the simulation capabilities and member states, and CEA is among the partecipating organizations. The traditional way to treat the neutronic and thermal-hydraulic transient problems in a nuclear reactor is the chaining methodology, which consists in solving the two problems one after the other. Coupling consists instead in solving the two problems in parallel, allowing in principle greater accuracy in taking into account the reciprocal influence of each physics on the other. The scope of this work is to couple neutronics and thermal-hydraulics for the simulation of FFTF LOFWOS Test #13 in the framework of the IAEA benchmark. The work will be presented according to the following structure: First, an introduction to the host organization, context, SFRs in general and the FFTF reactor in particular, as well as the LOFWOS transient will be presented. Secondly, the neutronic problem will be addressed, starting from the physical models and numerical methods, passing through the description of its FFTF application and its numerical resolution through deterministic codes. The thermal-hydraulic problem will then be described in a less detailed way: in the course of the internship, it has not been changed much from the version previously developed at SESI, thus only the principles, the set of equations solved and the assumptions made will be presented. The coupling methodology will then be addressed, presenting the coupling platform and explaining the choices made concerning the coupling algorithm. Subsequently, the main results will be presented and commented, with some comparisons of the coupled and chained simulations with the results measured and/or calculated by ANL. Finally, some conclusions will be drawn from the results, the relative importance of coupling and correctly modeling the single thermally-driven feedbacks in the thermal-hydraulic model will be assessed and some ideas for possible perspective work will be presented.