Development of a general coupling interface for the fuel performance code TRANSURANUS tested with the reactor dynamics code DYN3D

A general interface is presented for coupling the TRANSURANUS fuel performance code with thermal hydraulics system codes, sub-channel codes or reactor dynamics codes. Beside its generality, other main characteristics of this interface are the calculation at either fuel assembly or fuel rod level, one-way or two-way coupling, automatic switch from steady to transient conditions in TRANSURANUS (including update of the material properties etc.), writing of all TRANSURANUS output files and manual pre- and post-calculations with TRANSURANUS in standalone mode. The TRANSURANUS code can be used in combination with this coupling interface in various scenarios: different fuel compositions in the reactor types BWR, PWR, VVER, HWR and FBR, time scales from milliseconds (i.e. RIA) over seconds/ minutes (i.e. LOCA) to years (i.e. normal operation) and thence different reactor states.

As first application of the interface the reactor dynamics code DYN3D was coupled in order to analyze the impact of a more detailed description of the fuel rod behavior during system transients. More precisely, the influence of the high burn-up structure formation, geometry changes and fission gas release are included. In the coupling, DYN3D provides only the time-dependent rod power and thermal hydraulics conditions to TRANSURANUS, which in turn transfers parameters like fuel temperature and cladding temperature back to DYN3D. Results of the coupled code system are presented for a control rod ejection transient in a German PWR, along with a sensitivity study for the full core. The results reveal that the detailed fuel rod behavior modeling influences the neutron kinetics in the core due to the Doppler reactivity effect of the fuel temperature. In particular it appears that for high burn-up fuel DYN3D-TRANSURANUS systematically calculates higher value for the node centerline fuel temperature compared to DYN3D standalone. The main reasons of the differences seem to be the UO2 material properties (e.g. thermal conductivity), and the radial power density profile over the fuel pellet radius.

Furthermore results of the DYN3D-TRANSURANUS code system are shown for the planed RIA experiment CIP3-1 in the CABRI water loop facility in France. The experimental data including time-dependent rod power was taken from the recent RIA fuel codes benchmark organized by the OECD/NEA. DNB is predicted by calculations under the typically PWR coolant conditions in CIP3-1.

No convergence problems occurred for DYN3D-TRANSURANUS. The coupled code system can improve the assessment of safety criteria, at a reasonable computational cost since it rises on average only by a factor 4 on the same workstation for RIA (determined from reaching the initial state of the transient), when compared to the DYN3D code standalone.