Présentation EPFL-Public | 01.01.2013 1
Geomechanical investigation of CO 2 sequestration
R. Makhnenko, C. Li & L. Laloui
September 1, 2014 “Kraftwerk 2020”
Laboratoire de Mécanique des Sols
(LMS EPFL)
Présentation EPFL-Public | 01.01.2013 2
CO
2storage, no thank you
Motivation CCS
Présentation EPFL-Public | 01.01.2013 3
Carbon capture and storage
Motivation CCS
Switzerland:
decommission of nuclear power plants
combined-cycle gas-fired power plants might be temporatily used (each produces 0.7 Mt CO2/year)
Kyoto protocol (1997): reduction of CO2 emissions
www.avenirelectricite.ch
Problem:
Concentrations of the greenhouse gas carbon dioxide in the global atmosphere are approaching 400 parts per million (ppm) for the first time in human history
- Temperature raise ( anomaly: 0.56 ºC, 2 ºC by 2100 ) - Ocean level raise ( + 3 mm/year)
- Human health issues cdiac.ornl.gov
The Keeling Curve (University of San Diego)
Présentation EPFL-Public | 01.01.2013 4
Geologic sequestration
Motivation Geologic sequestration
CO2 can be sequestrated in one of the following three geological formations, widely spread, available and safe:
• abandoned oil and gas reservoirs,
• unmineable coal seams and
• deep saline aquifers:
highly permeable and porous rocks,
> 800m depth and saturated with undrinkable water
widespread and available practically anywhere
deep saline aquifer
abandoned oil/gas reservoirs (675 – 900 Gt C)
(1,000 – 10,000 Gt C)
unmineable coal seams (3 – 200 Gt C)
Présentation EPFL-Public | 01.01.2013 5
CO
2sequestration – worldwide
Motivation IPCC 2005
Depth: 800 - 2500 m
Overburden pressure: 20 – 100 MPa Water pressure: 7 – 40 MPa
Temperature range: 25 – 125 ºC
Geologic sequestration
250 km offshore, 800 m under sea floor Injection: 2.5 kt CO2/day (since 1996), 0.03 GT CO2 have been sequestered so far
Snohvit
2600m below the sea floor
Présentation EPFL-Public | 01.01.2013 6
CO
2sequestration – Switzerland
Motivation
Diamond et al., 2010
Geologic sequestration
PSI ETHZ
Univ. Bern EPFL
Pilot for demonstrating CO2 capture on gas-fired power plant
Pilot for assessing onshore CO2 storage in Switzerland Proposed sites for
gas-fired power plant
Switzerland (total):
- 2.7 Gt of CO2 can be stored
- current annual emission 11.3 Mt
-capacity of saline aquifers is sufficient for > 200 years
Upper Muschelkalk:
65 m thickness, Dolomite
8.7 % of interconnected porosity 0.7 Gt CO2 can be stored
Malm-Lower Cretaceous:
50-1200 m thickness, Limestone 5 % of interconnected porosity 1.5 Gt CO2 can be stored
Aquifers:
Chevalier et al., 2010 800 - 2500 meters
deep aquifers
Présentation EPFL-Public | 01.01.2013 7
Reservoir materials
Motivation CO2 sequestration
Sandstone
Limestone Shale
Mainly quartz and feldspar Porosity: 3 - 30%
Intrinsic permeability: 10-7 – 10-2 cm/sec Stiffness: 1’000 – 20’000 MPa
Dominant pore size: dozens of µm
Caprock
Aquifer
Under relevant CO2 geological storage conditions, limestone suffers from potential
alteration through chemical reactions with CO2
saturated water.
On the contrary, sandstone remains intact during the injection period.
Candidate for sealing material
Candidates for host rock material
Mainly clay minerals and tiny fragments Porosity: 5 - 30%
Intrinsic permeability: 10-10 – 10-7 cm/s Stiffness: 1’000 – 70’000 MPa
Dominant pore size: dozens of nm
Mainly calcite Porosity: 5 - 35%
Intrinsic permeability: 10-8 – 10-3 cm/sec Stiffness: 1’500 – 55’000 MPa
Dominant pore size: dozens of µm
Présentation EPFL-Public | 01.01.2013 8
Injection of supercritical CO
2Motivation CO2 sequestration CO2CRC
Beneath 800m underground CO2 exists in the supercritical state:
temperature > 31.1 ºC, pressure >
7.4 MPa, density > 600 kg/m3 Injection of CO2
Nova Sterilis, 2014.
CO2CRC
Présentation EPFL-Public | 01.01.2013 9
Trapping mechanisms
Motivation CO2 sequestration
Trapping processes take place over many years at different rates from days to thousands of years.
In general, CO2 becomes more securely trapped with geological time.
THM coupling behaviour during the injection phase is crucial to secure the CO2 storage.
Présentation EPFL-Public | 01.01.2013 10
Chair « Gaz Naturel »
Motivation
- understanding and prediction of the effects of surrounding environment, of mechanical and chemical changes as well as heat effect during CO2 injection and storage
- experimental and numerical interdisciplinary research on the interplay between transport, reaction and mechanics
- advance scientific knowledge and provide reliable solutions to the industry.
Assessment of CCS
Présentation EPFL-Public | 01.01.2013 11 Objective
Objectives of laboratory research
Geological sequestration
sandstone limestone shale
precipitation, cooling effect, chemical
degradation, suction effect, permeability mechanical integrity Characterization of thermo-hydro-mechanical
behavior of possible host and cap rocks in contact with water, brine, supercritical and liquid CO2
Présentation EPFL-Public | 01.01.2013 12
Chemical reactions Dissolution
Reaction with carbonates (days/weeks)
Reaction with silicates (years)
Host rocks: issues
Theory CO2 effect
Oye et al., 2012 Ciantia & Hueckel, 2013
- Change in poroelastic response due to chemical effect caused by CO2 injection - Change in inelastic parameters and failure characteristics
Présentation EPFL-Public | 01.01.2013 13
Poroelastic regimes
Theory Poroelasticity
Drained Undrained Unjacketed
=0
∆ ∆
= ∆
V p
V P K
“drained” ∆p= 0 “undrained” ∆mf = 0
=0
∆ ∆
= ∆
mf
u V
V P K
=0
∆ ∆
= ∆
mf
P B p
=0
∆ ∆
= ∆
mf
f f
f V
V p K
Skempton’s coefficient:
bulk modulus of pore fluid:
“unjacketed” ∆P = ∆p
P p
s V
V p K
∆
=
∆ ∆
= ∆ '
unjacketed pore bulk modulus:
P p
s V
V p K
∆
=
∆ ∆
= ∆
φ
" φ
Présentation EPFL-Public | 01.01.2013 14
Slightly anisotropic (5% difference for ultrasonic velocities and 7-8% in UCS)
Porosity = 23%, density = 2100 kg/m3 ,
UCS= 41-43 MPa, E = 13-15 GPa, and ν = 0.31 Permeability k = 40 mD (at 5 MPa mean stress)
Diffusivity
sec 2 m . ) 0
1 ( ) 2 1 (
) )(
1 (
2
22
2
=
−
−
−
= −
u
kG
uc µα ν ν
ν ν
ν
Berchenko et al., 2004
sec 01 . 4 0
2
=
≈ c
t L
- time to equilibrate ∆p due to ∆P = 1 MPaHost rock: sandstone
Lab testing Sandstones
Mineralogical composition:
Quartz ∼ 90%
Feldspar ∼ 7%
Calcite ∼ 1%
Clay – traces
Quartz grain size ∼ 0.2 mm
Berea sandstone (Ohio):
Makhnenko, 2013
Présentation EPFL-Public | 01.01.2013 15
Poroelastic response: sandstone
Lab testing Sandstones
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 1 2 3 4 5 6
Back pressure [MPa]
B cor
Bmax = 0.58
P′ = 5 MPa
f L u
cor
K K V V p
B
α σ
σ
ν −
∆
∆ +
∆
= +
3
) )(
1 (
1
3 1
3.10 3.15 3.20 3.25 3.30
0 1 2 3 4 5 6
Back pressure [MPa]
P-wave velocity [km/s]
P' = 5 MPa
−
+
=
"
1 1
s
f K
K K B
φ α
α
Makhnenko and Labuz, 2013
Présentation EPFL-Public | 01.01.2013 16
Constitutive response: sandstone
Lab testing Sandstones
0 5 10 15 20 25 30 35
0 2 4 6 8
γ [10-3]
τ [MPa]
Triaxial, undrained Plane strain,
undrained
Triaxial, drained
Model, undrained
Initial conditions:
P = 20 MPa po = 3 MPa Drained:
∆p = 0 Undrained:
∆mf = 0
Undrained:
(
K K)
KKeff =υ 1 f −1 s" +α 1
G K
H
K H
d d
eff eff
) (
1 µβ
µβ γ
τ
+ +
= + G
H H d
d
= + γ 1 τ
Drained:
Rice 1975, Rudnicki 1985
0.0 0.2 0.4 0.6 0.8 1.0
0.0 1.5 3.0 4.5 6.0 7.5 9.0
Plastic shear strain [10-3]
µ
0.0 0.2 0.4 0.6 0.8 1.0
β
µ
β
Makhnenko and Labuz, JGR-2014
Présentation EPFL-Public | 01.01.2013 17
Host rock: limestone
Lab testing Limestone
Close to be isotropic
(2% difference in ultrasonic velocities) Porosity = 33%, density = 1400 kg/m3 , UCS= 15 MPa, E = 7.3 GPa, and ν = 0.25 Permeability k = 3-5 mD (at P′ = 5 MPa)
Mineralogical composition:
Calcite ∼ 98%
Traces of other minerals Grain size = 0.05 - 3 mm
Calcarenite (Apulian limestone):
Makhnenko and Labuz, 2014
Présentation EPFL-Public | 01.01.2013 18
Poroelastic response: limestone
Lab testing Limestone
GPa 1
.
= 5 K
GPa 7
. 42 ' = K
s88 .
= 0 α
0 5 10 15 20 25 30 35
0 1 2 3 4 5 6
Volume strain [10-3]
Hydrostatic pressure [MPa]
K = 5.1 GPa Ks' = 42.7 GPa
0 5 10 15 20 25 30 35 40 45 50
0 2 4 6 8 10 12
Hydrostatic pressure [MPa]
Strain [10-3]
1.1 1.3 1.2
Ks′ is significally smaller than Kcalcite – a lot of very small and non-connected pores
Makhnenko and Labuz, 2014
Présentation EPFL-Public | 01.01.2013 19
Characterization of dissolution
Lab testing Limestone
Viscoporoelastic formulation for undrained constant mean stress response:
( ) (
P p)
dt dp K
K dt
d
s
−
−
− −
=
ηφ
φ
φ 1
' 1
1 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
1.5 2.0 2.5 3.0 3.5 4.0 4.5
Pore pressure [MPa]
B
P' = 5 MPa Bmax = 0.70
Connolly and Podladchikov 1998
( )(
P p)
dt dp
BK −
= −
φ η
α
φ 1 1
ηφ - matrix bulk viscosity
Makhnenko and Podladchikov, 2015
Présentation EPFL-Public | 01.01.2013 20
Scanning Electron Microscopy (SEM) X-ray CT scanning Mercury Intrusion
Porosimetry (MIP)
Non-destructive
Qualitative Destructive
Quantitative Destructive/non destructive Qualitative
- microstructures
- pore space morphology and porosity
- fluid saturation
- mineralogical composition
- pore size ditribution - relation between Hg pressure and volume of intruded pores.
- surface morphology and topography
8 μm/pixel 100-0.003 μm 4 nm/pixel
Romero et al, 2008
Characterization of chemical effect
Lab testing Change in composition
Présentation EPFL-Public | 01.01.2013 21
Not treated
1 MPa water
Complexity of pore morphology
Scanning Electron Microscopy (SEM)
Water injection (1 MPa for 4 days) Goal: observe change in porosity and pore morphology
Lab testing Change in composition
Présentation EPFL-Public | 01.01.2013 22
- Cumulative volume of pores with radius larger than 5 µm increases from 3 to 10%
- Increase in total porosity is about 2-3%
Dry specimen Treated specimen
Mercury Intrusion Porosimetry (MIP)
Lab testing Change in composition
Treated and dry specimens of similar size were tested with MIP
0 10 20 30 40 50 60 70 80 90 100
0.001 0.010 0.100 1.000 10.000 100.000 1000.000 10000.000
Cum. Volume [%]
Pore Radius [µm]
MIP results: dry specimen
0 10 20 30 40 50 60 70 80 90 100
0.001 0.010
0.100 1.000
10.000 100.000 1000.000 10000.000
Cum. Volume [%]
Pore Radius [µm]
MIP results: treated specimen
Présentation EPFL-Public | 01.01.2013 23
MatLab©
- Pixel size 8.41 μm - Grey-scale 0-255
- Each mineral has the same density - Grey intensity is
in proportion to pore size
- Normal distribution MIP ϕ = 0.385
MatLab© ϕ = 0.375
Lab testing Change in composition
X-ray CT scanning
Présentation EPFL-Public | 01.01.2013 24
CO
2injection project at Mont Terri
Lab testing Shales
• Main Objectives
– Build well elements
– Measure of the flow inside and and and outside the casing
-> sealing changes
– Sample fluid across time -> fluid changes
– Take samples of the different elements (overcoring)
-> mineralog. changes
Présentation EPFL-Public | 01.01.2013 25
Caprock - issues
Lab testing Shales
~ 1nm
Wellbore
- Seal permeability
(w/ respect to H2O and CO2) - Seal capacity
(CO2 retention properties) - Seal integrity (propensity for
brittle or ductile behavior) - Pressure build-up due to
injection of CO2
- Geomechanical/failure
characteristics (effect of in- situ stress variations)
- Change in mineralogy/
porosity/ permeability due to the chemical effect
mont-terri.ch
Présentation EPFL-Public | 01.01.2013 26
Geomechanical testing of shales
Lab testing Shales
Opalinus clay behavior at different mean stresses and temperatures:
HIGH-PRESSURE OEDOMETER
Water retention curve of Swiss shale
SORBTION BENCH
Oedometric curve of a Swiss shale [Ferrari, Manca, Laloui]
[Ferrari, Manca, Laloui]
Présentation EPFL-Public | 01.01.2013 27
Caprock: failure
Lab testing Shales
Mechanical weakness:
the interface between the caprock and the aquifer:
• Primary barrier to prevent CO2 from leakage
• Failure potentials to be evaluated
Φ
cohesion
Friction angle
σ'= −σ
= c c+ w w
f
f
p S p S p p
Shear stress Shear stress
Laloui and Li, 2014
Présentation EPFL-Public | 01.01.2013 28
Capillary effects
Lab testing Shales
2
4 cos 2
w CO
T T
p p
d R
− = − θ = −
Capillary stress : Non-wetting phase : CO2
Wetting phase : water
Présentation EPFL-Public | 01.01.2013 29
«Capillary» failure
Lab testing Shales
• CO2 pressure increase / capillary stress
increase / effective stress increase mass shrinkage (free shrinkage)
• If shrinkage is constrained, reaction forces arise.
• Three main causes of shrinkage constraint:
(1) Boundary restraint
(2) Moisture gradients inside the body (3) Internal structure
tensile stresses are built up, tensile strength is reached cracks appear and propagate.
Laloui and Li, 2014
Présentation EPFL-Public | 01.01.2013 30
CO
2retention behavior
Lab testing Shales
Water retention curve with CO2 at 20 ˚C and pressure of 8 MPa : σ = 0.030 [N/m]
θ = 20 [˚]
(Espinoza and Santamarina, 2010) Water retention curve with air at room temperature
and atmospheric pressure : σ = 0.073 [N/m]
θ = 0 [˚]
Water retention curve of Opalinus Clay:
Ferrari et al., 2014
Reduction of gas entry value from 13 to 8 [MPa]
Laloui and Li, 2014
Présentation EPFL-Public | 01.01.2013 31
Host rock and caprock microcracking
Lab testing Microcracking
undrained compression
0.0 0.5 1.0 1.5 2.0 2.5 3.0
0 5 10 15 20
Shear strain [10-3] Volume strain [10-3 ]
0 300 600 900 1200 1500
AE events
peak load inelastic deformation
drained compression
0.0 0.3 0.6 0.9 1.2
0.0 2.5 5.0 7.5 10.0 12.5
Shear strain [10-3] Volume strain [10-3 ]
0 200 400 600 800
AE events
inelastic deformation peak load
transducer microcrack
elastic wave
t V
Inelastic response (yielding) of rock is associated with microcracks, which generate elastic waves called acoustic emission, AE.
Makhnenko and Labuz, JGR-2014
Présentation EPFL-Public | 01.01.2013 32
Host rock: AE events locations
Lab testing Sandstones
0-1 1-2 2-3 3-4
0 50 100 150 200 250 300 350 400 450
0 0.05 0.1 0.15 0.2 0.25 0.3
Lateral displacement [mm]
Load [kN]
0 500 1000 1500 2000 2500
AE events
1 2 3 4 all events fractured specimen
Makhnenko, Ge and Labuz, 2015
Présentation EPFL-Public | 01.01.2013 33
Summary
Group business
Gaz Naturel Projects
- Geologic sequestration of carbon dioxide is promising option for reducing greenhouse gas emissions
- Thermo-hydro-mechanical processes occur during CO2 storage:
deformation and failure potentials triggered by injection-induced overpressure and cooling are the key issues to be addressed
- Laboratory testing is needed to characterize different aspects of rock-water- CO2 interactions and an advanced equipment has to be used to get accurate results and make reliable predictions
- Geomechanics will play a key role in seeking a balance between injectivity and integrity/safety of host and caprocks