Two-Phase Flow and Heat Transfer in Highly Permeable Porous...

36
Oct 7, 2015 Transport in Porous Media 1/36 M. Quintard D.R. CNRS, Institut de Mécanique des Fluides, Allée Prof. C. Soula, 31400 Toulouse cedex – France [email protected] http://mquintard.free.fr Two-Phase Flow and Heat Two-Phase Flow and Heat Transfer in Highly Permeable Transfer in Highly Permeable Porous Media Porous Media Michel Quintard

Transcript of Two-Phase Flow and Heat Transfer in Highly Permeable Porous...

  • Oct 7, 2015

    Transport in Porous Media1/36

    M. Quintard

    D.R. CNRS, Institut de Mécanique des Fluides, Allée Prof. C. Soula, 31400 Toulouse cedex – France

    [email protected] http://mquintard.free.fr

    Two-Phase Flow and Heat Two-Phase Flow and Heat Transfer in Highly Permeable Transfer in Highly Permeable

    Porous MediaPorous MediaMichel Quintard

    mailto:[email protected]

  • Transport in Porous Media 2/36M. Quintard

    OutlineOutline

    Background One-phase flow at high Re number Two-phase flow: quasi-static and dynamic models Hybrid Model: pore-scale / network-scale Macro-Scale Model with phase splitting for

    structured porous media Boiling in porous media Conclusions

  • Transport in Porous Media 3/36M. Quintard

    BackgroundBackground

    Nuclear Engng, Nuclear Safety Chemical Engng: distillation, catalytic columns Flow (Oil,...) in gravel, blasted rocks, … High T geothermy

  • Transport in Porous Media 4/36M. Quintard

    Upscaling One-Phase FlowUpscaling One-Phase Flow

    Pore-scale equation

    Upscaling

  • Transport in Porous Media 5/36M. Quintard

    Closure: Stokes problemClosure: Stokes problem

    PDEs for deviations

    Re~0

    (Whitaker, 1986)

  • Transport in Porous Media 6/36M. Quintard

    Closure and Macro-Scale Closure and Macro-Scale EquationEquation

    See Sanchez-Palencia (homog.), Whitaker, ...

  • Transport in Porous Media 7/36M. Quintard

    Non Darcean regimesNon Darcean regimes

    Heuristic: Forchheimer, Ergun, …

    Upscaling? Darcy Weak Inertial Strong InertialWeak

    Turbulent

    ~10 ~30 ~1000

    ~Re~Re3

    ~Re2

    ~Re2

    ~Re

    Re

    (passability)

  • Transport in Porous Media 8/36M. Quintard

    Non Darcean regimesNon Darcean regimes

    Laminar inertia effects → ~generalized Forchheimer equation

    – Re → 0: Darcy– Re ~ 0: weak inertia, F.〈vβ ~ v〉 〈 β〉3 (Levy, Mei

    & Auriault, ...)– Re > 0: strong inertia, F.〈vβ ~ v〉 〈 β〉2 (Whitaker,

    1996; Lasseux et al., 2011;...)

  • Transport in Porous Media 9/36M. Quintard

    Experimental Evidence: Darcy Experimental Evidence: Darcy regimeregime

    4×4 mm prisms - Air flow

    8.00E-09

    8.50E-09

    9.00E-09

    9.50E-09

    1.00E-08

    1.05E-08

    1.10E-08

    1.15E-08

    1.20E-08

    1.25E-08

    0 1 2 3 4 5

    Re

    µV/ (

    -∂P

    / ∂z

    -ρg

    ) (m

    ²)

    PermeabilityDarcyNon Darcy

    Clavier et al., 2014

  • Transport in Porous Media 10/36M. Quintard

    Experimental Evidence: inertia Experimental Evidence: inertia regimeregime

    Clavier et al., 2014

    4×4 mm prisms - Water flow

    y = 0.0002x2.3831y = 0.0064x1.1916

    0.1%

    1.0%

    10.0%

    100.0%

    Re

    (-∂P

    /∂z

    -ρg

    -µU

    /K)

    / (µ

    U/K

    )

    Weak Inertial

    Transition

    1 10 100

    Non-spherical particle beds - Air flow

    0.0E+00

    1.0E-04

    2.0E-04

    3.0E-04

    4.0E-04

    5.0E-04

    6.0E-04

    7.0E-04

    0 500 1000 1500 2000

    Re

    ρU²

    / (-∂

    P/∂z

    -ρg

    -µU

    /K)

    (m) c5x5

    c5x8c8x12p4x4p6x6

    Weak inertia Strong inertia

  • Transport in Porous Media 11/36M. Quintard

    Turbulent flows in porous mediaTurbulent flows in porous media

    Turbulence: time and spatial averaging (see book De Lemos, 2006; ...)– time and spatial averaging commute!– However: not necessarily the same result if sequential closure!?

    • for one-phase flow: scheme “II” seems preferable –contrary to Antohe & Lage (1997), Getachew et al. (2000)–see discussion: Nakayama & Kuwahara (1999), Pedras and de Lemos (2001), etc...

    • for multiphase flow?• Legitimacy of seq. Closure? → Simultaneous closure over [R3 R✕ t]? More

    complex sequential closures (t → x → t → ...) depending on the hierarchy of scales?

  • Transport in Porous Media 12/36M. Quintard

    Turbulent flows in porous media (continued)Turbulent flows in porous media (continued)

    Localized Turbulence: i.e., nearly periodic (see Jin et al., 2015 for DNS results)– Spatial averaging of RANS models → generalized

    Forchheimer equation, F not necessarily ~ v〈 β〉 or v〈 β〉2

    Porous media turbulence models (i.e., modified k-ε, k-ω, etc...)?– Pedras & De Lemos, Nakayama & Kuwahara (1999), ...– note: useful for fluid/porous medium interface (D'Hueppe et

    al., 2012)

    Soulaine and Quintard, 2014

    Example: structured packings

  • Transport in Porous Media 13/36M. Quintard

    Two-Phase FlowTwo-Phase Flow

    Pore-scale

    L

    β-phase

    averaging volume V

    l γ

    γ-phase

    σ-phase

  • Transport in Porous Media 14/36M. Quintard

    Upscaling, quasi-static theoryUpscaling, quasi-static theory

    Case of B.C. 4Whitaker, 1986; Auriault, 1987; Lasseux et al., 1996; ...

    + Re numbers + Dynamic Bond numberif ≈0

    ⇒ ?

  • Transfert de chaleur 15/36M. Quintard

    Macro-Scale Models: quasi-Macro-Scale Models: quasi-staticstatic Heuristic (Muskat)

    Upscalingimbibition

    w = wetting phase

    drainage

    Pc

    1- Sor 1Swi

    Sw Sw

    kr

    1- Sor 1Swi

    Phase interaction

  • Transfert de chaleur 16/36M. Quintard

    Inertia EffectsInertia Effects

    Ergun (Heuristic)

    Schulenberg and Muler (1987) (Heuristic and ⛐)

    Upscaling (Lasseux et al., 2008)

  • Transfert de chaleur 17/36M. Quintard

    Importance of Cross-Terms, and Importance of Cross-Terms, and Non-Linear EffectsNon-Linear Effects

    from Clavier et al. (2015)

    see also Taherzadeh & Saidi (2015)

    Case Vβ=0 :

  • Transfert de chaleur 18/36M. Quintard

    Comparison various models: dPComparison various models: dP

    0

    0.2

    0.4

    0.6

    0.8

    1

    1.2

    0 0.2 0.4 0.6 0.8

    4 mm beads 〈 vl =0〉

    Exp (Clavier et al.)

    Clavier et al.

    Lipinski

    Reed

    Hu&Theofanous

    Schulenberg

    Tung&Dhir

    〈 vg (m/s)〉

    Models without cross-terms

  • Transfert de chaleur 19/36M. Quintard

    Comparison various models: Comparison various models: SaturationSaturation

    0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

    1

    0 0.2 0.4 0.6 0.8

    Exp (Clavier et al.)

    Clavier et al.

    Lipinski

    Reed

    Hu&Theofanous

    Schulenberg

    Tung&Dhir

    4 mm beads 〈 vl =0〉

    〈 vg (m/s)〉

  • Transfert de chaleur 20/36M. Quintard

    More: Dynamic ModelsMore: Dynamic Models

    impact of ∂S/∂t, Vα, av...:– Pseudo-functions in Pet. Engng– Quintard & Whitaker (1990, from large-scale

    heterogeneity effects and multi-zone)– Hilfer (1998, multi-zone)– Panfilov & Panfilova (2005, meniscus)– Hassanizadeh and Gray (Irr. Therm., av as state

    variable, 1993), also Kalaydjian (1987)...– ...

  • Transfert de chaleur 21/36M. Quintard

    Examples of dynamic equationsExamples of dynamic equations

    Quintard & Whitaker, 1990

    ...see also Petroleum Engng literatureon pseudo-functions!

    ω

    η

    η

    ω

    0 0.5 1

    Ωβ=0 Pa/m

    10-1

    10-2

    10-3

    10-4

    10-5

    Ωβ=10-4Pa/m

    Ωβ=10+4Pa/m

  • Transfert de chaleur 22/36M. Quintard

    Mixed or Hybrid ModelsMixed or Hybrid Models Motivation: highly

    permeable media, trickle beds

    Example of challenging problem: jet dispersion

    Trickle Bed (X-ray, IFP)

    Horgue et al., 2013

  • Transfert de chaleur 23/36M. Quintard

    Mixed or Hybrid ModelsMixed or Hybrid Models

    Network model Dynamic rules (may come from local VOF

    simulations)

    Melli & Scriven, 1991; results from Horgue et al. (PhD CIFRE/IFP/IMFT), 2012

    VOF

    Dynamic Network Simulation

  • Transfert de chaleur 24/36M. Quintard

    Application to tomographic Application to tomographic imagesimages

  • Transfert de chaleur 25/36M. Quintard

    Macro-Scale Models with Phase Macro-Scale Models with Phase “Splitting”“Splitting” Example: Flow through Structured

    Media

    Upscaling with phase splitting (Soulaine et al., 2014): role of momentum exchange term

    Mahr and Mewes (2007) Alekseenko (2008)

    Model with liquid phase splitting

  • Transfert de chaleur 26/36M. Quintard

    Macro-Scale Models with Phase Macro-Scale Models with Phase “Splitting”“Splitting” Example: Soulaine et al. Exp.

    Comparison with Fourati et al. (2012) experiments

    Model with liquid phase splitting

    1. Identification on 1st stack2. Application to several stacks

  • Transport in Porous Media 27/36M. Quintard

    Coupling with Heat Transfer: Coupling with Heat Transfer: boiling in porous mediaboiling in porous media

    3-Temperature model:– decoupled 2-phase flow, quasi-steady →

    Generalized Darcy-Forchheimer? (time-space ergodicity?)

    – 3-T model, extension of 2-T model: Berthoud & Valette (1994); Petit et al. (1999); Duval et al. (2004);...based on quasi-steady approx.

    – Heuristic: time averaging of averaged equations → porous media Nukiyama curves (see Sapin et al., 2014)Note: Highly open problem!

    App. Nuclear safety

  • Transport in Porous Media 28/36M. Quintard

    3T-Model3T-Model

    withBoiling rate

    Note: two-phase flow model (inertia+cross terms) +

  • Transport in Porous Media 29/36M. Quintard

    Heat Exchange Terms: Impact of Heat Exchange Terms: Impact of phase configuration!phase configuration! Quasi-static two-phase flow theory

    0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

    0.25

    0.5

    0.75

    1

    1.25

    1.5

    1.75

    2

    10−

    6(W

    m−

    3K

    −1 ) Chang SLG

    Chang SGLStaggered SLGStaggered SGL

    ⟨ Tl⟩l: exch. ⟨ Tl⟩l- T sat

    Sl

    s

    g

    l

    R

    εℓ = 0.28, εg = 0.42, Sℓ = 0.4Phase repartition:

    VOF or Cahn-Hilliard Duval et al., 2004, ...

  • Transport in Porous Media 30/36M. Quintard

    Pore-Scale ExperimentationsPore-Scale Experimentations

    Sapin et al., 2014

    Nucleate Boiling Film Boiling Intense Boiling

    Pt wiring R0=100Ω

    ceramic

    ceramic coating

    Remark Need time-averaging!

    Slow motion!

  • Transport in Porous Media 31/36M. Quintard

    Porous Media Nukiyama CurvesPorous Media Nukiyama Curves

    0

    5

    10

    15

    20

    25

    -20 30 80 130 180

    Wall fluxqps (W/cm²)

    Tsat = Tp - Tsat (°C)

    qchf

    qmin

    TLeidenfrost

    Forced Convection

    boiling crisis

    Vapor film collapse

    Impact on non-linear properties: kr, h, D, etc...?!

  • Transport in Porous Media 32/36M. Quintard

    Critical FluxCritical Flux

    0

    5

    10

    15

    20

    25

    30

    -20 -10 0 10 20 30 40 50 60 70∆Tsat = Tp - Tsat (°C)

    Qmp = 20mWQmp = 50 mWQmp = 100 mWQmp = 150 mW

    q ps (

    W/c

    m2 )

    Surrounding heating changes critical flux!...but the behavior up to CHF seems to be unaffected (contrary to modeling guess)?!

  • Transport in Porous Media 33/36M. Quintard

    Macro-Scale Model?Macro-Scale Model?Bachrata et al., 2013 (model implemented in CATHARE Safety code): work confirms main required features (inertia, NLE,…)

    → ∃ Macro-scale behavior: confirmed by experiments but heuristic time averaging or time-space ergodicity

    → Need for adapted correlations for phase permeabilities and cross terms, inertia terms, heat transfer coefficients, etc...

    Example for Solid-fluid exchange: IRSN Experiments:PRELUDE, SYLPHIDE, ...

  • Transport in Porous Media 34/36M. Quintard

    Macro-Scale Model: reflooding Macro-Scale Model: reflooding simulationsimulation

    0

    0.1

    0.2

    0.3

    0.4

    0.5

    0.6

    0.7

    0.8

    0.9

    1

    1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 3998000

    99000

    100000

    101000

    102000

    103000

    104000PRELUDE

    mesh number

    void fractiontotal heat flux [W]

    wall T [°C]

    gas vel. [m/s]

    liq. vel. [m/s]pressure [Pa]

    Qmax = 0.8 10+5

    W/m2

    Tmax = 414 °CVg,max= 3.9 m/sVl,max= 9.4 10

    -3 m/s

    dp = 2 mm

  • Transport in Porous Media 35/36M. Quintard

    Example of Macro-Scale Example of Macro-Scale Simulations (Quenching)Simulations (Quenching)

    15s

    25s

    15s

    25s

    Temperature field

    550K

    750K

    steam

    Water and steam velocities

    4 m

    1.2 m

    QTregoures et al. (2003)

  • Transport in Porous Media 36/36M. Quintard

    Conclusions (high permeable Conclusions (high permeable media)media) Various types of models Momentum equations:

    – Generalized Forchheimer equation ~OK up to some high Re numbers, including localized turbulence

    – Two-phase flow: need for cross-terms Boiling:

    Need for: inertia, non-equilibrium, ... Problem: coupling and time averaging?

    Effective properties Experimental procedures and interpretation? DNS?

    Numerical aspects

    Slide 1Slide 2Slide 3Slide 4Slide 5Slide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Slide 16Slide 17Slide 18Slide 19Slide 20Slide 21Slide 22Slide 23Slide 24Slide 25Slide 26Slide 27Slide 28Slide 29Slide 30Slide 31Slide 32Slide 33Slide 34Slide 35Slide 36