SUMMARY

MISHA, a cooperative project between GRS and IKE, aims to establish a calculation chain for innovative MMR designs using the GRS developed simulation codes ATHLET and FENNECS. The Los Alamos National Laboratory’s Special Purpose Reactor was chosen as a reference design. To perform such a simulation, an initial model in the Monte Carlo code Serpent is required for the creation of a macroscopic cross-section library and as a reference to validate the future FENNECS model.

The Serpent model is presented and calculated core eigenvalues for different core configuration are compared to earlier work by the Idaho National Laboratory. The model shows good results for the excess reactivity without absorbers, reasonable shutdown margins and reaches the critcal state in a configuration similar to the reference.

KEYWORDS

SMR, MMR, Serpent, MISHA, SPR

INTRODUCTION

Micro Modular Reactors (MMRs) are a subcategory of Small Modular Reactors with an electrical power lower than 10 MW. In recent years they garnered research interest as potential power sources for remote or mobile applications, featuring a high transportability and little effort of deployment at location. These MMRs often feature innovative designs which differ significantly from the established reactor types.

One of these designs is the Special Purpose Reactor (SPR) designed by Los Alamos National Laboratory, a reactor concept for a 5 MWth, 2 MWel heatpipe-cooled fast reactor with a solid, monolithic core, also called Megawatt in earlier design stages [1]. The design was chosen as a reference as detailled information as well as thermal and neutronic analysis by the Idaho National Laboratory (INL) is publically available [2], [3]. It also features several similarities to the commercial eVinci® design by Westinghouse [4], [5], which has recently received a Front-End Engineering and Experiment Design contract by the Department of Energy to support the planned test reactor at INL [6].

In detail, the SPR features a hexagonal steel monolith core with an active length of 1.5 m and a diameter of 1.012 m flat-to-flat, which is divided azimuthally into six segments, each containing 352 UO2 fuel elements and 204 potassium heat pipes. A total mass of 5.22 t UO2 with 19.75 wt% U-235 enrichment is contained within. In the center of the core is a control void where a solid and an annular control rod

can be inserted. The core is surrounded by an Al2O3 reflector, in which 12 control drums are located. The control drums are rotatable with a crescent of absorber material on one side and function as the primary reactivity control. The core and radial reflector are contained within another steel vessel with additional neutron shielding. The absorber material in the drums, the control rods and the shielding is B4C with 90% Boron-10. A recuperated brayton cycle is used for power conversion. [2]

The MISHA project, a cooperation between GRS and the IKE of the University of Stuttgart, aims to esta- blish a simulation model for such MMRs as a coupled simulation between the thermal-hydraulic system code ATHLET [7] and the neutron diffusion code FENNECS [8], which are both part of the GRS code package AC2 [9]. To perform the FENNECS calculation, pre-computed macroscopic cross-sections are required. These can be calculated using the Monte Carlo code Serpent® [10] developed at VTT. The Serpent results also function as reference for later FENNECS calculations.

MODELLING

Based on the neutronic analysis and the MCNP model by INL [2], [3], a detailed model of the SPR was created in Serpent. The purpose of this study is to compare the INL results with Serpent calculations, which will in the future provide macroscopic cross-section data for the dynamic simulations with FENNECS.

While precise information on some of the details of the model like temperature of the different materials or the exact location and orientation of the control drums and the absorber crescent within, is not explicitly given, the Serpent model recreates the geometry and material data as closely as possible. It features, from the centre outwards, a hexagonal central channel for the central control rod and control annulus, the hexagonal steel monolith with the 2112 fuel elements and 1224 heat pipes, the radial reflector with 12 control drums and an additional steel vessel as well as a neutron shield. The control drums are distributed around the core in groups of two along the outer edge of the core segments, drums within a group have the same rotational position and each group is rotated by 60° when compared to its neighbours. Above and below the active core region are neutron reflectors made from steel with additional BeO elements in the fuel channels.

Core configurations

The core configurations ‘all poisons out’ and ‘all poisons in’ are pictured in Figure 1 and Figure 2 respectively. The first configuration has the absorber crescents of the control drums rotated out of the core as far as possible and has removed the two central control elements from the active core region. The second configuration has the control drums rotated in and both central control elements in the active core, reducing reactivity as much as possible. In the figures, yellow represents potassium heat pipes, red the UO2 fuel elements, the steel monolith is black and the Al2O3 reflectors are grey, lighter in the radial reflector and darker in the control drums. The absorber material B4C in the control drums, the central rods and the outer shielding is colored blue. The darker turqoise tone above and below the fuel elements is BeO in the bottom and top reflectors. The empty spaces in white are filled with Helium.

The other core configurations investigated are ‘all control drums in’, which rotates the drums into the core like in Figure 2, but does not apply the two central control elements, and ‘annular rod in’ as well as ‘solid rod in’, which apply either of the central elements with the control drums rotated out of the core. Finally, the ‘criticality’ configuration of the core represents operational state, which rotates all control drums (CD) by the same angle, so that the reactivity of the core is closest to 1. INL reaches this state with a rotation of 48°, while the Serpent model requires a rotation by 52°.