Numerical Simulations of Atmospheric Carbonation in Concrete Components of a Deep Geological Intermediate Low Level Waste Disposal

NUCPERF 2012

P. Thouvenot1, O. Bildstein1, S. Poyet2, I. Munier3 , B. Cochepin3 , X. Bourbon3 , E. Treille3

1 CEA (French Alternative Energies and Atomic Energy Commission), LMTE, Cadarache

2 CEA (French Alternative Energies and Atomic Energy Commission), LECBA, Saclay

3 ANDRA (French Radioactive Waste Management Agency)

FRENCH CONCEPT : RADWASTE REPOSITORY IN A CLAYSTONE FORMATION AT 500 M DEPTH

Current design of deep underground repository for high and intermediate level long-lived waste

FRENCH CONCEPT : RADWASTE REPOSITORY IN A CLAYSTONE FORMATION AT 500 M DEPTH

Atmospheric carbonation of overpack during the operating period

• Bitumized waste• Cemented waste• Compacted metallic waste• Organic waste

CARBONATION ISSUES FOR RADWASTE REPOSITORY

Ventilation (100 years)

DRYING AND CARBONATION PROCESSES IN ILLW OVERPACK

Dry air

(Rh = 40 %)

T = 25°C to 50°C

SlWater vapor diffusion

CO2 gas diffusion

T

Aqueous diffusion of reactants

Two phase water/air flow

Dissolution/precipitation : porosity reduction, permeability variations

Brine formation

CO2 gas dissolution

Dry air

(Rh = 40 %)

T = 25°C to 50°C

SlWater vapor diffusion

CO2 gas diffusion

T

Aqueous diffusion of reactants

Two phase water/air flow

Dissolution/precipitation : porosity reduction, permeability variations

Brine formation

CO2 gas dissolution

PHENOMENOLOGY: CAPILLARY FLOW

Flow law (generalized Darcy law):

Lowering of the dew point due to capillary effects

(Kelvin equation in EOS 4):

Water relative permeability (Van Genuchten):

Gas relative permeability (Corey):

Klinkenberg effect (gas flow at low pressure):

)(

gPkkF r

))(ln()( lrw

wrcap ShM

RThP

21

11)(

m

mrrrrl SSSk

lrls

lrlr SS

SSS

22 ˆ1ˆ1 SSk rg grlr

lrl

SS

SSS

1

ˆ

rgg kp

kk

1int

PHENOMENOLOGY: DIFFUSION

Air and water gases diffusion:

CO2 and other gases:

Aqueous diffusion:

Effective diffusion :

Tortuosity (Millington-Quirk):

15,273

15,273,, 000,,0,,0

T

P

PTPdTPd ii

M

RT

PNd

RTd i

8

23 2,,0

TR

Edd a

KOHi

1

15,298

1exp15,298,,,0 2

0,,0, ii dD

baS 0

PHENOMENOLOGY: DIFFUSION

Air and water gases diffusion:

CO2 and other gases:

Aqueous diffusion:

Effective diffusion :

Tortuosity (Millington-Quirk):

15,273

15,273,, 000,,0,,0

T

P

PTPdTPd ii

M

RT

PNd

RTd i

8

23 2,,0

TR

Edd a

KOHi

1

15,298

1exp15,298,,,0 2

0,,0, ii dD

baS 0

1st MAJOR COUPLING EFFECT!! Sliq

Di,g

DRYING PHENOMENON : PARAMETERS VALUES

3 different concrete materials:• High Performance Concrete (HPC)• Intermediate Performance Concrete (IPC)• Low Performance Concrete (LPC)

HPC IPC LPC

Porosity 0.08 0.12 0.16

Intrinsic permeability to liquid (m²) 1e-21 1e-19 1e-17

Intrinsic permeability to gas (m²) 1e-19 1e-17 1e-15

Relative permeability m – Slr – Sls – Sgr0.481 – 0.0 – 1.0 –

0.00.424 – 0.0 – 1.0 –

0.00.367 – 0.0 – 1.0 – 0.0

Capillarity pressure m – P0 (MPa) – Pmax (MPa) 0.481 – 45 - 1500 0.424 – 15 - 1500 0.367 – 5 - 1500

Molecular diffusion coefficient gaseous phase (m²/s) water

2.4e-05

Molecular diffusion coefficient gaseous phase (m²/s) CO2 1.6e-05

Molecular diffusion coefficient in aqueous phase (m²/s) 1.9e-09

Millington-Quirk a parameter 2

Millington-Quirk b parameter 4.2

Klinkenberg parameter (MPa) 0.45

SIMULATIONS CONFIGURATION

1D half section package container (section = 11 cm) Carbonation on both sides Ventilation air at 25°C and 40% relative humidity Initial liquid water saturation assumed to be 0.8

11 cm

25°C40%RH

25°C40%RH

DRYING RESULTS

TR EOS9 (Richards) and TR EOS4 (full multiphase) comparison

Drying process slows down when transport characteristics of concrete are enhanced.

Drying with Richards’ equation (EOS9 without gaseous diffusion) is slightly slower than with full multiphase model (EOS4).

NUMERICAL RESOURCES FOR CARBONATION SIMULATIONS

Simulations performed with Intermediate Performance Concrete

for aqueous species and mineral phases

full multiphase flow

CHEMICAL PARAMETERS

Primary phases

Secondary phases

Kinetics of dissolution / precipitation

nnnn Akr 1

15,298

11exp15,298 TR

EkTk a

n

Phase Volume %

Calcite 72.12

Portlandite 5.73

CSH 1.6 13.76

Monocarboaluminate 2.26

Ettringite 3.60

Hydrotalcite 0.39

Hydrogarnet-Fe (C3FH6) 2.05

Phase type Phases

Oxides Magnetite, Amorphous silica 

Hydroxides Brucite, Gibbsite, Fe(OH)3 

Sheet silicates Sepiolite

Other silicates CSH 1.2, CSH 0.8, Straetlingite, Katoite_Si

Sulfates, chlorides, other salts Gypsum, Anhydrite, Burkeite, Syngenite, Glaserite, Arcanite, Glauberite, Polyhalite

Carbonates Calcite, Nahcolite

Other Hydrotalcite-CO3, Ettringite, Dawsonite

CHEMICAL PARAMETERS

Primary phases

Secondary phases

Kinetics of dissolution / precipitation

nnnn Akr 1

15,298

11exp15,298 TR

EkTk a

n

Phase Volume %

Calcite 72.12

Portlandite 5.73

CSH 1.6 13.76

Monocarboaluminate 2.26

Ettringite 3.60

Hydrotalcite 0.39

Hydrogarnet-Fe (C3FH6) 2.05

Phase type Phases

Oxides Magnetite, Amorphous silica 

Hydroxides Brucite, Gibbsite, Fe(OH)3 

Sheet silicates Sepiolite

Other silicates CSH 1.2, CSH 0.8, Straetlingite, Katoite_Si

Sulfates, chlorides, other salts Gypsum, Anhydrite, Burkeite, Syngenite, Glaserite, Arcanite, Glauberite, Polyhalite

Carbonates Calcite, Nahcolite

Other Hydrotalcite-CO3, Ettringite, Dawsonite

amorphous CSH phases

CARBONATION RESULTS

pH decrease, portlandite dissolution and calcite precipitation over a thickness of about 2 cm after 100 years

CARBONATION RESULTS

Dissolution of CSH 1.6, ettringite, monocarboaluminate and hydrotalcite on 2 cm after 100 years

Precipitation of CSH 1.2, CSH 0.8, straetlingite, amorphous silica and gypsum on the same thickness

Precipitation of small amounts of sepiolite, gibbsite and katoïte-Si is also predicted

CARBONATION RESULTS

from a performance assessment point of view:

looking at the different concrete performance

similar paragenesis from 1 cm to 4 cm after 100 years

sensitivity calculations on diffusion properties a and b (tortuosity parameters)

less than 1 cm (a, b +50% with HPC) in 100 years complete carbonation - 5.5 cm - (a, b -50% with LPC) in

25 years

CARBONATION RESULTS: COMPARISON WITH EXPERIMENTAL RESULTS

Modeling results- alteration is complete (amorphous silica)- no residual primary phases (portlandite)

CARBONATION RESULTS: COMPARISON WITH EXPERIMENTAL RESULTS

Modeling results- alteration is complete (amorphous silica)- no residual primary phases (portlandite)

Experimental results- residual primary phases (portlandite) alteration is not complete

Distance (mm)

Calcite front

Portlandite front

Drouet, 2010

CARBONATION RESULTS

Effect of water content on reactivity (Bazant type function)

2nd MAJOR COUPLING EFFECT!!

Sliq

diffusionchemical reactivity

CARBONATION RESULTS

Effect of water content on reactivity (Bazant type function)

2nd MAJOR COUPLING EFFECT!!

Sliq

diffusionchemical reactivity

CARBONATION RESULTS

Significant reduction in the amplitude of carbonation (less dissolution of portlandite and CSH 1.6 and less precipitation of amorphous silica and other secondary CSH)

Lower reactivity accompanied by a greater penetration of carbonation front due to lower consumption of CO2 at the surface

Effect of water content on reactivity (Bazant type function)

CARBONATION RESULTS

Significant reduction in the amplitude of carbonation (less dissolution of portlandite and CSH 1.6 and less precipitation of amorphous silica and other secondary CSH)

Lower reactivity accompanied by a greater penetration of carbonation front due to lower consumption of CO2 at the surface

Effect of water content on reactivity (Bazant type function)

CARBONATION RESULTS

Significant reduction in the amplitude of carbonation (less dissolution of portlandite and CSH 1.6 and less precipitation of amorphous silica and other secondary CSH)

Lower reactivity accompanied by a greater penetration of carbonation front due to lower consumption of CO2 at the surface

Effect of water content on reactivity (Bazant type function)

CONCLUSIONS Drying process of 11 cm thick waste packages depends strongly on

the concrete nature and slightly on the flow model (Richards or full multiphase)

Considering full multiphase model, carbonated depth is about 2 cm after 100 years for the Intermediate Performance Concrete. degraded thickness is totally carbonated (total dissolution of primary mineral phases)

If we consider a chemical reactivity depending on the liquid saturation (Bazant type function), a considerable reduction in the amplitude of carbonation and a greater penetration of carbonation front are observed

calibration with accelerated carbonation experiments (Drouet, 2010)

also: need for improved knowledge on kinetics parameters and thermodynamic data, especially for CSH with low Ca/Si ratio

Other perspectives include:

• taking into account a protective effect of secondary minerals

• modeling corrosion of rebars