Conference Agenda

Small Modular Reactors (SMRs) rely on passive safety systems, which use gravity-driven natural circulation to transfer heat from the core to a large reservoir without operator intervention for extended periods of time. Some SMR designs can reach Rayleigh numbers around 10^15, but the limited experimental data at this scale highlights the need for further validation of predictive thermal hydraulic codes.

This study presents preliminary Computational Fluid Dynamics (CFD) calculations to support the design of an experimental rig at the PANDA facility at PSI, Switzerland, under the OECD/NEA PANDA project. Using a 2D axisymmetric CFD model created with EDF's open-source Code_Saturne software, parametric studies were conducted to investigate the heat transfer mechanisms. Key variables, such as geometrical configuration and dimensions, operating conditions, and numerical options, were examined. Initial results indicate that achieving a Rayleigh number of 10^15 is feasible under the constraints of the facility, with favorable alignment to existing Nusselt-Rayleigh experimental correlations at lower Rayleigh numbers. The computational results allowed the sizing of the components and fixing the operating conditions. These results also underscore the importance of a detailed temperature measurement strategy, particularly near the wall, to be used for the validation and verification of the models.

Further studies explore the influence of different turbulence models on the results, revealing notable differences in predicting the laminar-to-turbulent transition zone and the peak wall temperature. These findings are useful for ensuring the validity of Nusselt-Rayleigh correlations over a wider range of applicability and allowing for accurate modeling of large-scale passive safety systems.

The conventional containment vessel of nuclear power plant acts as a barrier against radiological impacts and missile threats, independent of the reactor’s cooling. However, as the size becomes more compact, the containment vessel of an SMR roles as the final passive cooling system. The cooling capability around the dome must be ensured during severe accidents, such as Loss of Coolant Accident, to minimize coolant loss through condensation on the dome. Most condensation occurs in the upper dome of the containment vessel, however, studies on the external cooling capability of the dome reflecting the large Ra of SMR have not been conducted. Therefore, this study aims to analyze the heat transfer of domes under high Ra conditions, considering their geometric characteristics and flow behavior. Experimental range corresponds to 10⁹≤Ra_Db≤10¹³. The base diameter of dome, corresponding to the diameter of the cylindrical lower structure, is fixed, but its height can vary. These characteristics of the dome can be defined by a truncation angle, as the dome represents a segment of a sphere. The smaller the truncation angle, the closer the dome resembles a flat plate, while a truncation angle of 90° represents a hemispherical shape. Depending on the truncation angle, the slope of the dome surface varies, leading to differences in flow behavior. Plume flow enhances heat transfer and the separation point, where the point plume flow develops, is varied by the surface slope. Smaller truncation angles result in an earlier development of the separation point and further enhance heat transfer.

Due to the nonlinearity of the Natural Circulation (NC) process, flow and temperature perturbations in an NC loop result in a feedback mechanism. Under certain conditions, the prevailing circulatory flow can become unstable and ultimately stall, leading to insufficient heat transfer. This scenario is of particular interest to PWRs, since inadequate passive heat removal could result in initiation of postulated accident scenarios. Therefore, accurate predictive capability for stall in perturbed NC scenarios is of great interest to this work.

Performing full-scale high-fidelity simulations of the plant during perturbed scenarios is computationally prohibitive from a design basis analysis perspective. However, the phenomena of interest during perturbed scenarios are highly complex and multiscale; current 1D System Codes are not capable of adequately capturing the prevailing 3D flow phenomena. The current work considers a more computationally feasible approach. We explore surrogate modelling via Gaussian Process Regression as a method for accurately predicting bulk flow and therefore NC stall. The surrogate model is trained by considering Unsteady Reynolds Averaged Navier-Stokes (URANS) Computational Fluid Dynamics (CFD) simulations of NC in a simple experimental loop geometry containing an n-bend and multiple heat sinks. Various perturbed NC simulations are presented, in which we consider multiple transient definitions which are specified by an injection flow rate profile at a specified location, heater input power, and n-bend height. We then assess the surrogate modelling approach’s ability to capture the key features of quantities of interest by benchmarking the results against equivalent URANS CFD simulations.

As the primary mechanism of passive heat removal in nuclear reactors, understanding natural circulation is paramount to their safe operation and shutdown conditions. Natural circulation in molten salt reactors is of particular relevance due to the fluids high Prandtl number as well as the phase of the fissile material in liquid-fueled reactors. Natural circulation loops are employed to study the thermal hydraulic behavior of fluids when subject to thermal gradients and small flow disturbances. The objective of this work is to introduce and analyze the effects of local heating conditions on the velocity profile and near wall behavior (such as boundary layer thickness) of molten salts in a heated transparent test section. Particle image velocimetry (PIV) was performed, and the boundary layer was analyzed for three different heating conditions. Those conditions were applied to the transparent test section: a cooling test condition, a thermally isolated test condition, and a heated test condition. In addition, the other thermal conditions of the loop were held constant. As the test section is heated, the peak velocity and slope of the velocity profile increase with test section heater power. Additionally, preliminary transient thermal analyses are presented in this work.

The passive molten salt fast reactor (PMFR), under development in Korea, employs a helium bubbling system to remove insoluble fission products (IFPs). Notably, the helium injection significantly changes entire fluidic performance within the primary system. These changes, in turn, influence heat transfer efficiency. Accordingly, a lab-scale experiment needs to be performed to evaluate the impact of helium injection under PMFR conditions.

To this end, a two-phase natural circulation molten salt loop at a reduced scale was designed based on scaling laws to simulate the helium bubbling effect in the PMFR. New governing equations for the prototype (PMFR) and reduced model (molten salt loop) were established based on the drift-flux model. Similarity criteria were derived through the nondimensionalization of the new governing equations. These criteria were employed in the design of the reduced model. Furthermore, an enlarged model, whose geometrical shape is similar to the reduced model, was also designed to evaluate the effect of friction number.

Subsequently, the distortion among prototype and two models was evaluated using multiphaseEulerFOAM solver in OpenFOAM. The distortion analysis revealed significant discrepancies in void fraction and fluid velocity between the prototype and the reduced model. A major reason for the distortion is attributed to geometrical differences. However, the distortion between the reduced model and the enlarged model was relatively minor due to geometric similarity and friction number scaling. This study will contribute to developing advanced scaling laws applicable to molten salt reactors (MSRs), although there are still rooms for improvement.

The steady-state behaviour of two-phase natural circulation systems (TPNCs) is critical for their efficient and safe operation across various applications, including nuclear reactors, thermal power plants, and advanced passive cooling systems. In TPNCs, steady-state conditions represent a balance between the buoyancy driving force and the opposing frictional force. Understanding and predicting the steady-state characteristics are vital for optimizing system performance, particularly in Boiling Water Reactors (BWRs) and natural circulation boilers (NCBs), where high flow rates are desirable for enhancing the heat transport capability.

This paper provides an in-depth analysis of the steady-state behaviour in TPNCs, focusing on key factors such as loop flow regimes, the effects of system geometry and heat input on circulation patterns. The study examines the influence of system parameters like loop diameter, heat flux and pressure on steady state flow. In addition, the effect of loop inventory on the steady state performance has been studied. The predictions cover the inventory at which peak TPNC flow rate, breakdown of TPNC flow and heat-up of the heated surface occurs. Insights gained from this analysis are crucial for designing TPNC systems to maximize the heat transport capability for critical applications such as nuclear reactors and thermal power plants.