Page 46 - Thermal Hydraulics Aspects of Liquid Metal Cooled Nuclear Reactors

Thermal-hydraulic challenges in liquid-metal-cooled reactors 21

produced at the free surface of the pool by bubble detachment from vortices or by the nucle-
ation of dissolved gas in the cold part of the pool can be carried by the primary flow toward
the inlet plenum of the core (the diagrid). If this is the case, a gas pocket may form in the low
flow zones of the diagrid, and the destabilization of this pocket (due to pump flow-rate
changes, for instance) could lead to the passage of a large gas bubble in the core, leading
to an overpower transient. In order to assess this risk, the phenomenology of gas bubble
transport from their sources to the core inlet plenum must be analyzed.
State of the art
Transport of bubbles in liquid metals was touched upon by Arien et al. (2004). Simulation
methods, both in system thermal-hydraulic codes and in CFD, still need a lot of develop-
ment. Typically state-of-the-art Euler-Lagrangian CFD methods can only deal with the
transport of small spherical bubbles.
Development needs
If bubbles get larger, Eulerian-Eulerian or volume-of-fluid methods are probably required,
which are computationally very costly and probably only allow small geometries. This topic
will become important in core voiding scenarios and certain beyond design basis accidents.
Therefore, basic fundamental experiments and experiments directed toward application will
be required together with model development. State-of-the-art developments in measure-
ment techniques for bubbles in liquid metal are described by Andruszkiewicz et al.
(2013) and applied by Vogt et al. (2015) for basic gas entrainment experiments.
l Particle transport
Challenge
Liquid-metal-cooled reactors are designed to operate in a single-phase flow regime, and the
presence of a secondary phase of solid particles may have damaging effects. Among the
threats are core damage, criticality effects related to fuel accumulation and coolability
effects due to potential blockage and local heat flux increase. Various sources of particles
can be assumed, ranging from particles formed through chemical reaction (IRSN, 2015)
between coolant and impurities, air or water, or through corrosion of structural material
to particles from damaged or molten fuel. The accurate prediction of transport and possible
accumulation of particles with diverse properties is challenging.
State of the art
The Euler-Lagrange particle-tracking approach has been used as general modeling method-
ology to investigate dispersion of small particles in the large plena of a pool-type liquid-
metal reactor by Buckingham et al. (2015). This method assumes the particles as “point
masses” within the computational domain, and only particles below the mesh size can be
simulated. It is used for very dilute systems, where particle interactions can be neglected.
In those cases where the particle volume must be considered when modeling hydrody-
namics and wall effects (e.g., for flow blockage studies), an alternative method within
the Euler-Lagrange framework is the new macroscopic particle model (Agrawal et al.,
2004). For a dense particulate system, particle-particle interactions dominate the particle
dynamics.
This can be the situation when dealing with fuel particle sedimentation/accumulation
after the release caused by a severe accident. Coupled CFD-DEM methods offer a solution
where the motion of the particle phase is obtained accurately by the discrete element method
(DEM) and interactions can be described through interaction forces (Guo et al., 2014). To
reduce the computational time of DEM, particle-particle interactions can be modeled by
fluid properties, such as granular pressure and viscosity, in a so-called Euler-granular model.
The latter is also used in combination with population balance method to study solid particles