Development of a coupled neutronic thermalhydraulic code system for accident analysis in sodium-cooled fast reactors

The aim of the doctoral thesis is the development of a coupled neutronic-thermal hydraulic tool for analysis of severe accidents in sodium-cooled fast reactors (SFRs). With the foundation of the Generation IV International Forum (GIF IV) this reactor technology gained new interest, as it fulfills the main goals defined by the experts of GIF IV for future nuclear power plants. Besides an excellent fuel utilization due to the possibility of breeding during operation, fast reactors can help to reduce long-lived radioactive waste (burner reactor design) and can be used for the realization of a closed fuel cycle management. However, one big safety issue is the fact that SFRs are not in their most reactive configuration under operation conditions. Unlike in Light Water Reactors (LWRs), a loss of coolant does not imply a nuclear shut down of the reactor. Instead the loss of sodium leads to a higher neutron flux in the system due to reduced parasitic absorption of neutrons in the sodium. Another consequence of the reduced sodium fraction is a harder neutron energy spectrum as the neutrons are no longer slowed down by scattering reactions. This effect implies a further increase of neutrons in the system, as the shifting of the neutron spectrum to higher energies is associated with a higher neutron yield in important isotopes (e.g. Pu239). Hence, the analysis of severe accidents in SFRs requires both, the consideration of the thermal hydraulic behavior of the core as well as the neutronic aspects. For the coupling a set of existing codes was chosen. The thermal hydraulic part is calculated by the ASTEC-Na Code developed by IRSN and GRS. The code is based on the well validated modular code ASTEC for LWRs. In the JASMIN project (2012-2016) the code was adapted for sodium and new models were implemented. The modern code architecture and the modular code structure enable an easy implementation of further developments. Simulations of CABRI experiments show, that the code is able to catch the main phenomena occuring during a transient perturbation of sodium-cooled fast systems. In the JASMIN project a point kinetic model was implemented to account for the neutronics. But as a point kinetic model cannot calculate changes in the flux shape, the idea was to couple the ASTEC-Na thermal hydraulics with a more detailed neutronic tool. For the neutronic part the code DYN3D from HZDR (Helmholtz-Zentrum Dresden/Rossendorf) is used. DYN3D is a nodal diffusion code with either square or hexagonal geometry options. The use of a nodal code requires preprocessing of problem specific homogenized, group-wise cross section data that covers the range of expected changes in important influence parameters during transient calculations. The cross section library is calculated with the Monte Carlo Code SERPENT2 (VTT,Finnland) for variations in the fuel temperature, the fuel density (to account for axial expansion) and the sodium density in 35 energy groups. The comparison of DYN3D calculations using the generated cross section data base with results obtained from detailed Monte Carlo simulations with SERPENT2 and MCNP6 confirmed the applicability of the used method. The development of suitable exchange scripts enables the data exchange between the two codes to get a full description with feedback of the system during transient conditions. For demonstration two kinds of accident scenarios are applied to a generic SFR core (definition from OECD/NEA Benchmark): an Unprotected Loss Of Flow (ULOF) and an Unprotected Transient OverPower accident (UTOP). The test cases show that the coupled code version is appropriate to simulate the early stage of severe accidents in sodium-cooled fast reactors.