On the use of the Serpent Monte Carlo code for few-group cross section generation

Serpent is a recently developed 3D continuous-energy Monte Carlo (MC) reactor physics burnup calculation code. Serpent is specifically designed for lattice physics applications including generation of homogenized few-group constants for full-core core simulators.
Currently in Serpent the few-group constants are obtained from the infinite lattice calculations with zero neutron current at the outer boundary. In this study, in order to account for the non-physical infinite-lattice approximation, B1 methodology, routinely used by deterministic lattice transport codes, was considered for generation of leakage-corrected few-group cross sections in the Serpent code. A preliminary assessment of applicability of the B1 methodology for generation of few-group constants in the Serpent code was carried out according to the following steps. Initially, two-group constants generated by Serpent were compared with those calculated by Helios deterministic lattice transport code. Then, 3D analysis of a Pressurized Water Reactor (PWR) core was performed by a nodal diffusion code DYN3D employing two-group cross section sets generated by Serpent and Helios. At this stage thermal-hydraulic (T-H) feedback was neglected. The DYN3D results were compared with those obtained from the 3D full core Serpent MC calculations. Finally, the full core DYN3D calculations were repeated taking into account T-H feedback and assuming Hot Full Power (HFP) conditions.
B1 two-group cross sections and diffusion coefficients generated by the Serpent and Helios codes agree within less than 2.5%. The results of the DYN3D calculations with the Serpent B1 cross-section sets agree very well with those of the Serpent full core MC calculations. The relative difference in keff is below 300 pcm. The maximum and root mean square (RMS) difference in the radial power distribution is 2.7% and 1.1% respectively. The results of the DYN3D full core calculations with T-H feedback obtained with Helios and Serpent generated cross-section libraries show an excellent agreement as well. The estimated critical boron concentration agrees within 5 ppm. The maximum and RMS difference in the core radial power peaking factors is 0.8% and 0.4% respectively.
In this study a Matlab script was used for calculation of the leakage-corrected few-group cross sections, however the B1 methodology has recently been implemented directly in the Serpent code.