Capability Of Actual CFD Codes to Support Fuel Rod Bundle Design

1
Capability Of Actual CFD Codes to Support Fuel
Rod Bundle Design

• Eckhard Krepper, Forschungszentrum Rossendorf,
  Germany
• Bostjan Koncar, Josef Stefan Institute
  Ljubljana, Slowenia
• Modeling and Measurements of Two-Phase Flows and
• Heat Transfer in Nuclear Fuel Assemblies
• October 10-11 2006, KTH, Stockholm, Sweden

2
Importance of Critical Heat Flux for Fuel
Assemblies

• Integrity of fuel rods has to be insured during
  operationand accidents ? the heat flux has to
  stay below thecritical heat flux
• The critical heat flux depends on
• fluid properties
• flow conditions ? geometry
• Advanced FA designs can allow higher permissible
  heat fluxes
• Verification of improvements in the critical heat
  flux require expensive measurements

3
Measurement of CHF by experiments

• Example Multifunction Thermal Hydraulic Test
  Loop KATHY, operated by AREVA in Karlstein
• Measurement of
• Pressure drop
• Critical heat flux
• Dryout tests for BWR (full 10x10 bundle)
• DNB tests for PWR (5x5 test bundle)
• Void distribution (BWR)
• ongoing tests carried out by FZ Rossendorf
• Transients (BWR)
• Stability (BWR)

4
Prediction of CHF

• Empirical correlations
• W-3 correlation Tong (1967)
• Correlation of Levitan (1975)
• etc.
• Valid only for a certain geometry in a narrow
  range of thermal hydraulic parameters
• Look-up tables
• 10000s of data points gained by experiments
• Interpolation
• Valid only for a certain geometry in a slightly
  extended range of thermal hydraulic parameters
• All these methods depend on geometry

5
Modeling of CHF independent on geometry

• Only possible by CFD methods
• Calculation of the phenomena on mesoscale
• Modeling of the phenomena on microscale (Closure
  relationships)

6
Flow pattern in a gas-liquid flow
Annular flow
Churn flow
Slug flow
Bubbly flow
7
Euler/Euler-approach

• For all phases the full set of conservation
  equations is solved
• Concerning the conservation equations the phases
  are considered interpenetrating each other (sum
  of volume fractions 1)
• Calculation of the closure relations based on
  microscaled models
• continuity equation ? mass transfer
• momentum equation ? drag, non drag e,g,
• energy equation ? heat transfer

8
? limitation on low gas volume fractions
Annular flow
Churn flow
Slug flow
Bubbly flow
9
Dependency of Tsat on the geodetic height at
different system pressures
10
CFX concept modelling subcooled boiling at a
heated wall

• Constant pressure ? given Tsat
• Overall heat flux Qw given
• Heat flux partition
• Qw Qf Qe Qq
• Qf - single phase convection
• Qe - evaporation
• Qq - quenching
• (departure of a bubble from the heated surface ?
  cooling of the surface by fresh water)
• Calculation of Qf, Qe, Qq and Twall by iteration

11
Validation Comparison to measured cross
sectional averaged Steam Volume Fractions
P4.5 MPa, D15.4 mm
5.7.105 W/m2
900 kg/(s.m2)
Bartolomej, G.G., Chanturiya, V.M., 1967, Thermal
Engineering Vol. 14, pp. 123-128
12
Temperatures
13
Heat flux partition
14
Bubble diametres
15
Mass transfer in the volume
Condensation rate
16
Investigation of the parameter range of
validityQ 1.0 MW/m2, G 1000 kg/(m2s)
17
P 7 MPa, G 1000 kg/(m2s)
18
Variation of pressure, heat flux and mass flow

• The larger the generated steam the larger the
  model deviations
• Acceptable agreement to the experiments at
• Pressure 3..11 MPa
• heat flux up to 1.2 MW/m2
• mass flow rate at about 1000 kg/(m2s)

19
Calculation of the flow conditions in a hot
channel of a fuel assembly

• consideration of the channel between 4 rods
• Calculation of a 0.5 m axial section between two
  spacers
• periodic boundary conditions
• Parameters
• pressure 15.7 MPa
• ? Tsat 619 K
• Qwall 1.0.106 W/m2
• Vin 5 m/s, Tsub 12 K

20
Simulation of the swirl by given inlet conditions
periodic
Rod4
Rod1
periodic
Rod2
Rod3
periodic
21
Influence of the swirl on the flow in the channel
water streamlines
vapour streamlines
22
Bubble forces
Air/Water at P105 Pa
Steam/Water
23
Bubble forces
24
Vapour Volume fraction distribution in the cross
section at z 0.475 m
without swirl
with swirl (VR 3 m/s)
25
Influence of the swirl on averaged values
vapour volume fraction
heat flux components
wall superheating
26
Wall superheating at the rod surface
Tsup Twall-Tsat K
VR 0 m/s
VR 1 m/s
27
Wall superheating at the rod surface
VR 2 m/s
VR 3 m/s
28
There exist still no CFD model describing CHF

• But Relative CHF performance of different spacer
  grids can be assessed by comparing CFD results
  like
• differential pressure
• swirl
• crossflow between channels
• concentrations of bubbles
• inhomogeneous distribution of wall superheating

29
Next steps

• Improvement of the actual approach
• Coupling of TW (result of wall boiling model)
  with heat conduction in the wall
• Modelling of bubble size at departure by
  mechanistic model
• Modelling of bubble size in the bulk by bubble
  size class model (inhomogeneous MUSIG)
• Modelling of turbulence
• Turbulence model of higher order
• Modulation of the turbulence of the liquid phase
  by gaseous bubbles
• Wall functions for two phase turbulent flow
• Extension of the actual approach
• Simulation of further mechanisms subcooled
  boiling towards CHF

30
For the validation of model extensions more
detailed experiments are necessary

• Desirable values
• wall temperatures
• vapor volume fraction distribution
• bubble size distributions
• liquid turbulent parameters
• Difficulties
• Narrow geometry of the channels (ca. 10 mm)
• bubble sizes expected 0.5...2 mm
• fluid temperature typical for nuclear reactor
  applications
• Solutions
• Using of model fluids ? up-scaling of geometry