PHYSOR 2014 – The Role of Reactor Physics Toward a Sustainable Future The Westin Miyako, Kyoto, Japan, September 28 – October 3, 2014, on CD-ROM (2014) DEVELOPMENT OF A HIGH-FIDELITY MONTE CARLO THERMAL-HYDRAULICS COUPLED CODE SYSTEM SERPENT/SUBCHANFLOW – FIRST RESULTS M. Daeubler, J. Jimenez and V. Sanchez Institute for Neutron Physics and Reactor Technology Karlsruhe Institute of Technology, Karlsruhe, Germany miriam.daeubler@kit.edu javier.jimenez@kit.edu victor.sanchez@kit.edu ABSTRACT Worldwide efforts to develop high-fidelity or ab-initio, high performance multi-physics tools are on-going due to the availability of relatively cheap, large-scale parallel comput- ers. In order to arrive at a coupled neutronics and thermal-hydraulics tool with a higher spatial resolution and accuracy than currently used Best-Estimate tools, an external cou- pling between the Monte Carlo Reactor Physics code Serpent and the sub-channel code SUBCHANFLOW has been developed. The coupled code system is intended to serve as reference for deterministic reactor dynamics code developments in future exploiting the fact that Serpent was conceived as a lattice code for such deterministic tools. The achieved coupling is based on Serpent’s recently introduced universal multi-physics in- terface. Enabling the interface, Serpent treats temperature dependence of nuclear data us- ing target motion sampling. Furthermore, variance reduction techniques and un- der-relaxation schemes were utilized. The development is verified by code-to-code com- parison with the deterministic tool DYNSUB. Simulation results of both code systems for a pressurized water reactor fuel assembly under hot full power conditions were found in good agreement. Finally, the impact of the thermal-hydraulic feedback on Serpent’s nu- merical performance was assessed. For a hot full power simulation, Serpent ran roughly ten times slower than with pre-broadened continuous energy cross sections for homoge- neous hot zero power thermal-hydraulic conditions. However, the developments present- ed here simply represent a first step towards Serpent/SUBCHANFLOW as a comprehen- sive reference for deterministic safety analysis tool development as many issues such as combining target motion sampling with bound atom scattering have yet to be resolved. Key Words: coupled codes, Serpent, SUBCHANFLOW, DYNSUB, safety 1. INTRODUCTION The availability of relatively cheap, large-scale parallel computers has triggered the development of high-fidelity reactor simulation tools for nuclear reactor design, optimization and safety anal- ysis in academia. These are intended to replace Best-Estimate (BE) methods in future. Recently, the large number of challenges one is facing developing such high-fidelity or ab-initio tools has been summarized by Smith and Forget [1]. While no truly high-fidelity multi-physics tool exists so far, most on-going efforts concentrate on increasing both the spatial resolution and on im- proving the accuracy of coupled neutron physics and thermal-hydraulic codes. As a consequence, for example, all fuel rods of a light water reactor (LWR) may be modelled explicitly without any significant geometric simplifications. These coupled code systems rely both on deterministic and Monte Carlo methods to solve the neutron transport problem. As an illustration, the whole core