Safety Analyses of Sodium Fast Reactors

SAS-SFR and SAS-LFR are designed to perform deterministic analyses of beyond design accidents in sodium (SFRs) or lead/lead-bismuth (LFRs) cooled fast reactors with mixed oxide fuels, which are foreseen for transmutation in fast systems. Detailed mechanistic models describe both the steady state operation and accident conditions caused e.g. by protected and/or unprotected loss of coolant flow accidents or reactivity insertion accidents. The initiation phase of an accident is modelled, including coolant heating and boiling, fuel heat-up, melting and pre-failure in-pin fuel motion. Consequences of clad failure or melting and subsequent core materials relocation (fuel, clad, fission gases, coolant) in the coolant channel and, after loss of integrity of the fuel pin, in a broken-up configuration are simulated. The calculations with SAS-SFR are stopped, when a long-term sub-critical core state is reached or when the subassembly shroud (hexcan) loses its integrity.



For simulation of a reactor core, a single channel represents the behaviour of one or more subassemblies simulating a representative fuel pin, its associated coolant inventory, and the respective hexcan structure. For an adequate representation of a whole core, a large number of channels must be employed. The heat transfer in pin and structure is performed with a quasi two-dimensional heat conduction model, where the radial heat conduction in pin and structure is coupled with the one-dimensional axial coolant flow. Reactivity feedback effects associated with the behaviour of the different core materials are taken into account by first order perturbation theory (point kinetics), including fuel and clad heat up (axial expansion and Doppler effect), coolant heating and boiling, and clad and fuel motion. 



SAS-SFR has been extensively qualified with a large variety of results from various experiments.

The CABRI FAST LOF-TOP experiment with EFM1 pin from VIGGEN4 demonstrates the quality of the validation.




Figure on the right shows CABRI FAST LOF-TOP experiment with EFM1 pin: VIGGEN4 with 12 at% burn-up; Loss of Flow halving time: 6.5 – 8.3 seconds. Structured power pulse triggered 8.03 seconds after bulk boiling onset; half width of main pulse ≈100 milliseconds.




 The code system SAS-SFR is the result of long-term cooperation between scientists from the KIT/INR (Germany), CEA, IRSN (France) and JAEA (Japan). The development started in the early eighties, taking the SAS4A code developed at ANL (USA) as a basis, but it became actually a new code frame due to a large number of improvements. Numerous coding errors and inconsistencies, found in the original code version, were corrected, existing models were extended and new ones were introduced, and the experimental database for qualification of the updated and the newly developed models was broadened. Due to this extensive work, requiring much manpower over the last three decades, progress has been reached in two aspects: the new code could be applied with sufficient reliability for the simulation of rather different core and plant designs, and it had a stable performance irrespective of the application.. Today SAS-SFR still represent the reference code for the assessment of initiation phase of accidental conditions.

Presently, SAS-SFR is used at KIT in the framework of research and development activities, investigating safety issues related to innovative sodium cooled critical systems. These activities are mostly imbedded in research programmes supported by the European Commission in the Horizon 2020 Programme. Presently, SAS-SFR is used at KIT for theoretical support of experiments with sodium boiling foreseen in the KASOLA facility. SAS-SFR has also supported the development of other system codes such as ASTEC-Na (JASMIN Project).

SIM-SFR (being one of SIM-‘‘family” codes: SIM-ADS, SIM-LWR, SIM-SFR, SIM-LFR, SIM-GFR, SIM-MSR) is a PC-based; multi node point kinetic (PK) model that describes the nuclear and thermal-hydraulic characteristics of both critical and sub-critical reactor cores. Two distinctly separate neutronic core models are run concurrently, namely a single node point kinetic (PK) model that acts as the driver for a multi-node  thermo-hydraulic model, and a multi-node neutronic model  allowing changes in the axial power profile during the transient according to nodal temperature and nodal reactivity variations (such as changes in control rod positions or local coolant void formation). In the single node PK neutronic model, the axial power profile is assumed to remain unchanged during a transient. Both neutronic models should yield identical results in core averaged parameters for ‘‘simple” transients, thereby serving as a continuous calculational cross-check to each other. The neutron kinetics of each nuclear model is described by 6 or 8-delayed neutron precursor groups on the nodal level, and a complete set of nodal dependent reactivity feedback coefficients such as Doppler, coolant and fuel expansion, diagrid expansion, void, control rod position, etc. The nodal time characteristics of the decay heat after reactor shutdown are described by up to 10 decay heat groups.

The family code system SIM-SFR, SIM-GFR, SIM-ADS, SIM-MS, etc. has been developed and used extensively over the last fifteen years to perform transient analyses for various European projects such as both LBE-cooled and He-cooled ADS designs during the PDS-XADS project, the EFIT-Pb and EFIT-He facilities, and XT-ADS experimental ADS during the EUROTRANS project, the ETDR experimental reactor during the GCFR project, the AMSTER molten salt reactor during the MOST project, and the CDT, ESFR, LEADER, and GoFastR projects as one of the system codes. SIM-SFR has been furthermore validated against available Superphénix transient data, and has been tested and validated extensively against actual LWR transient plant data for both PWRs and BWRs. The SIM-family has also been tested and benchmarked to other large code systems such as RELAP, TRACE, CATHARE, and SAS4A during the various projects listed above. The SIM-family codes can describe the nuclear core behavior during the initial accident phase before the geometric integrity of the core is compromised.



Perez-Martin S, Pfrang W, Girault N, et al. Development and assessment of ASTEC-Na fuel pin thermo-mechanical models performed in the European JASMIN project. Annals of Nuclear Energy (2018) 119 454-473

Bubelis E., Tosello A., Pfrang W., Schikorr M., et al. System codes benchmarking on a low sodium void effect SFR heterogeneous core under ULOF conditions. Nuclear Engineering and Design (2017) 320  325-345

Schikorr, M., Bubelis, al. Assessment of SFR reactor safety issues. Part I: Analysis of the unprotected ULOF, ULOHS and UTOP transients for the SFR(v2b-ST) reactor design and assessment of the efficiency of a passive safety system for prevention of severe accidents. Nuclear Engineering and Design (2015) 285, 249-262

Kruessmann, R., Ponomarev, A., Pfrang, W., Struwe, D., Champigny, J., Carluec, B., Schmitt, D., Verwaerde, D. Assessment of SFR reactor safety issues. Part II: Analysis results of ULOF transients imposed on a variety of different innovative core designs. Nuclear Engineering and Design (2015) 285 263-283

Sato I, Lemoine F, Struwe D. Transient fuel behavior and failure condition in the CABRI-2 experiments. Nuclear Technology (2004) 145(1) 115-137

Charpenel J, Lemoine F, Sato I,  Struwe D et al. Fuel pin behavior under the slow power ramp transients in the CABRI-2 experiments. Nuclear Technology (2000) 130(3) 252-270

Royl P, Pfrang W, Struwe D. Reactivity feedback evaluation of material relocations in the Cabri-1 experiments with fuel worth distributions from SNR-300. Nuclear Engineering and Design (1994) 147(1) 85-91

Imke, U., Struwe, D., Niwa, H., Sato, I., Camous, F., Moxon, D. Status of the SAS4A-code development for consequence analysis of core disruptive accidents. In: Proceedings of an International Topical Meeting. Sodium Cooled Fast Reactor Safety, Obninsk, Russia, October 1994, Paper 2-232.