Due to great deal of operation experience and technology accumulation, sodium cooled fast reactor (SFR) is the most promising among the six Generation IV reactors, which has advantages of breeding nuclear fuel, transmuting long-lived actinides and good safety characteristics. Thermal-hydraulic computer codes will have to be developed, verified, and validated to support the conceptual and final designs of new SFRs. However, work on developing thermal hydraulic analysis code for SFR is very limited in China, while the common software RELAP5 MOD3 is unable to analyze liquid metal systems. So the modified RELAP5 MOD3.2 is being considered as the thermal-hydraulic system code to support the development of the SFRs.
The thermodynamic and transport properties of sodium liquid and vapor have been implemented into the RELAP5 MOD3.2 code, as well as the specific heat transfer correlations for liquid metal. The sodium liquid properties use polynomial equations based on data obtained from Argonne National Laboratory, and the vapor is assumed to be perfect gas. The property equations are acceptably accurate for analysis of SFR, especially for single-phase liquid. New files are added to the fluids directory to generate property tables for new working fluid, which are similar to the table interpolation subroutines for light and heavy water in the original file directory. The method of code modifications are universal for other working fluids and will not affect the code original performance. Some basic verification work for the modified code are carried out. The steam generator of CEFR is analyzed to verify the modified code. The calculated results show that all the water will boil off in the evaporator and the calculated results are in good agreement with the design values. By using modified RELAP5 to model the primary loop of EBR-II fast reactor, the SHRT-17 PLOF test was analyzed. The results show that the natural circulation can be established in the EBR-II primary system after main pumps off to remove the core decay residual heat effectively, and the peak temperature under the safety limits. Moreover, the results computed in this work compared well with the test experimental data for the steady state condition. During the transients, the changing trends of temperature and pressure are similar to experimental data. The discrepancies between calculation and experiment are considered acceptably which need to be improved in the future work. Our work could demonstrate the capability and reliability of the modified RELAP5 for the analysis of SFRs further.