The Serpent Monte Carlo code and the Serpent-Ants two step calculation chain are used to model the hot zero power physics tests described in the BEAVRS benchmark. The predicted critical boron concentrations, control rod group worths and isothermal temperature coefficients are compared between Serpent and Serpent-Ants as well as against the experimental measurements. Furthermore, radial power distributions in the unrodded and rodded core configurations are compared between Serpent and Serpent-Ants. In addition to providing results using a best practices calculation chain, the effects of several simplifications or omissions in the group constant generation process on the results are estimated.
Both the direct and two-step neutronics solutions provide results close to the measured values. Comparison between the measured data and the direct Serpent Monte Carlo solution yields RMS differences of 12.1 mg/kg, 25.1 × 10-5 and 0.67 × 10-5 K-1 for boron, control rod worths and temperature coefficients respectively. The two-step Serpent-Ants solution reaches a similar level of accuracy with RMS differences of 17.4 mg/kg, 23.6 × 10-5 and 0.29 × 10-5 K-1.
The match in the radial power distribution between Serpent and Serpent-Ants was very good with the RMS and maximum for pin power errors being 1.31 % and 4.99 % respectively in the unrodded core and 1.67 %(RMS) and 8.39 % (MAX) in the rodded core.