UC Berkeley

The concept of a molten salt reactor (MSR) originated at Oak Ridge National Laboratory (ORNL), under the direction of Alvin Weinberg. In the 1960s, the Molten Salt Reactor Experiment (MSRE) operated successfully for five years and demonstrated the viability and safety of such concept. Recently, MSRs have attracted world's attention again as part of the six reactor technologies selected for further research and development by the Generation IV International Forum (GIF).

Modeling of liquid-fueled molten salt reactors involves the simulation of peculiar phenomena whose treatment is not readily available in common reactor physics tools as these are typically developed with stationary solid fuel in mind. In particular, the effect related to delayed neutron precursor circulation, the impact of compressibility on pressure wave propagation in liquid salt, and the consequence of radiative heat transfer (RHT) within liquid salt need specific modeling and simulation capabilities.

The purpose of this work is to develop multi-physics high-fidelity MSR models which are able to analyze phenomena peculiar to the liquid-fueled MSR. Since no MSR has been built yet, with the exception of the MSRE, high-fidelity models can help to better understand the significance and impact of these phenomena on the performance and safety of MSRs. Furthermore, high-fidelity tools and models can provide reference data for low-fidelity (less computational demanding) models/tools verification if experimental data are not available.

Two MSR models, based on the Molten Salt Reactor Experiment and the Molten Salt Fast Reactor (MSFR), respectively, were developed using the Monte Carlo particle transport code Serpent 2 and the multi-physics code GeN-Foam. The MSRE GeN-Foam model was used to study the effect of delayed neutron precursor drift. When fuel moves rapidly through the core, delayed neutron precursors decay in a different location as compared to the one they are generated. This leads to delayed neutrons being emitted in lower importance regions of the core or even outside of the core, leading to a lower neutron multiplication factor and a lower effective delayed neutron fraction. A multi-physics model of the MSRE that couples fuel flow and neutronics, as well as an adjoint solver to calculate the effective delayed neutron fraction, was developed and the results showed close agreement with the experimental values from the MSRE. The MSFR GeN-Foam model was used to study the effects of fuel salt compressibility and radiative heat transfer in MSRs. During a reactivity-initiated accident, negative reactivity is obtained from the Doppler effect but also from the salt expanding and exiting the core region. The latter effect is unique to liquid fuel, but failing to recognize the impact of the salt compressibility alters its nature. The assumption of incompressible salt, although usually acceptable in core simulations, makes the reactivity feedback prompt. In reality, this feedback is delayed and depends on the speed that the pressure waves propagate through salt. The effect of the salt compressibility was studied in the MSFR using either fluoride salt or chloride salt. It was found that a larger power excursion occurs when salt compressibility is properly accounted for and this is more severe in chloride salt due to its harder neutron spectrum and shorter neutron prompt lifetime. Finally, preliminary results show that the impact of radiative heat transfer is not significant during a loss of flow accident in the MSFR.