Stanford UniversityGlobal Climate & Energy Project

Integrated Experimental and Modeling Studies of CO2 Sequestration in Coal

Mark ZobackDepartment of Geophysics

Stanford University

GCEP SymposiumSeptember 20, 2006

Sequestration in Coalbeds

– CO2 is sequestered as an adsorbed phase.– CO2 injection can result in enhanced coalbed methane (ECBM) recovery,

off-setting costs for sequestration.– Potential to store large volumes of CO2 (U.S./Canada/China/India/Russia) – Critically important in rapidly growing economies of China and India

http://www.ieagreen.org.uk/8.pdf

Transport Modeling of CO2 Storage and ECBMCarolyn Seto, Kristian Jessen and Lynn Orr

Sorption and Transport in Coalbeds: A Laboratory and Simulation Investigation

T. Chaturvedi, G.-Q. Tang, W. Lin, K. Jessen and Tony Kovscek

Seismic Monitoring of CO2 Storage in Coal BedsChuntang Xu, Jolene Robin-McCaskill, Youli Quan and Jerry Harris

Simulation Studies of CO2 Sequestration and ECBM in Coalbeds of the Powder River Basin

Hannah Ross, Paul Hagin and Mark Zoback

_______________

Using CO2 Injection to Quench Coal Bed Fires – Lynn Orr

GCEP-Supported Coal Sequestration Project Studies at Stanford

Motivating Questions

1. Is it feasible to sequester CO2 in unmineable coalbeds? Are there potential show-stoppers such as a severe reduction of perm with CO2 saturation or the plasticization of coal in the presence of CO2?

2. If so, what volumes of CO2 can be sequestered? Are we accurately simulating injection into a heterogeneous, dual-porosity medium in which transport is controlled by complex physical and chemical processes?

3. To what degree could CO2 injection result in significant improvements in CH4 recovery (ECBM)? Can we can we accurately simulate cost recovery in commercial applications?

4. How do we effectively monitor CO2 storage in coal? Can we distinguish among gases present in situ (CO2, CH4, N2, etc.)?

Transport In Coal

increasing size

increasing sizeincreasing

sizeincreasing

size

adsorption on internal coal surfaces

diffusion through matrix and micropores

bulk flow in the fracture network

If diffusion times are small, matrix is in equilibrium with fracture system

Asorption/Desorption During Transport

Kovscek and Tang, 2004

Fluid Flow SimulationCO2 Sequestration and ECBM

5-spot, 80-acre well spacing using the simulator GEM.

1. Primary production 5 years:– 4 production wells

2. Base case: – One injector and 4 producers – Years 6-11.– Injector BHP constraint of 4 MPa (~600 psi).

3. Hydraulic fracture case: – Hydraulic fracture placed at base of injection well

• 60m in radius, permeability of 1000 md, and porosity of 30%.

Powder River Basin

Currently ~15,000 active CBM wells.

~50,000 more to be drilled in next decade.

History-Matching to Constrain Model Parameters

0.050.011-0.1Matrix porosity

0.5 mD0.04-0.7 mDMatrix permeability

100 mD10-160 mDVertical face cleat permeability

100 mD10-160 mDHorizontal butt cleat permeability

300 mD100-500 mDHorizontal face cleat permeability

ModeMinimum and Maximum ValueProperty

930937Well 531513153Well 416951696Well 328542844Well 217621768Well 1

History-matched Average Water Production per

Month (bbl/month)

Average Water Production per Month (bbl/month)

(WOGCC, 2006)Production Wells

History-matching water production

Geostatistics

3D Model and Cleat Permeabilities

Depth to the top: 315-361 m (1030-1180 ft)Pressure grad: 7.25 kPa/m (0.315 psi/ft)

Horizontal face cleat permeability

Horizontal butt cleat permeability

Vertical permeability

CO2 and CH4 Adsorption/DesorptionAfter 11 Years

CO2 Adsorption CH4 Desorption

Mole/m3 Mole/m3

Volumes of CO2 Injected and CH4 Produced after 6 Years of Injection

• One injection well can inject ~23 kt of CO2 per year.

• Sequester ~95% of total CO2 injected.

• With ECBM, CH4production increased by ~7 fold.

• To first order, need ~2,300 injection wells to sequester the annual CO2 emissions for the state of Wyoming.

• Hydraulic fracture increased injection by ~30%.

Importance of CH4 Adsorption

CO2, CH4, N2 Adsorption/Desorption

• Pure components are well fit by Langmuir isotherms

• CO2 adsorbs preferentially to CH4 and N2

• Adsorption hysteresis for all gases (small for N2)

Powder River Basin (WY) Coal

Kovscek and Tang, 2004

Kovscek and Tang, 2004

Total Volumes of CO2 Injected and CH4 Produced after 6 Years of Injection

• One injection well can inject ~23 kt of CO2 a year.

• Sequester ~95% of total CO2 injected.

• With ECBM, CH4production increased by ~7 fold.

• To first order, need ~2,300 injection wells to sequester the annual CO2 emissions for the state of Wyoming.

• Hydraulic fracture increased injection by ~30%.

Coal Matrix Shrinkage and Swelling

• Shrinkage of coal matrix• Increase in permeability with CH4 desorption.

• Swelling of coal matrix• Decrease in permeability with CO2 adsorption

• Stress dependent permeability• Decrease in permeability with decrease in pore

pressure.

Apparatus for sorption/k-reduction/displacement

• Pore pressure: 60~1100 psi• Gas mixtures made in the lab by weight

Permeability Reduction with CO2 AdsorptionCrushed Coal Pack

Severe decrease in permeability due to CO2 adsorption (matrix swelling).

How do changes in porosity and permeability with adsorption or desorption affect flow processes? Overburden Pressure=400 psi

N2

CO2

50% CO2, 50% N2

25% CO2, 75% N2

85% CO2, 15% N2

75% CO2, 25% N2

0

2

4

6

8

10

12

14

16

18

20

0 100 200 300 400 500 600 700 800 900 1000Pore Pressure (psi)

Per

mea

bilit

y (m

D)

5x

Constant Confining Pressure of 400 psi

100x

Net Permeability ReductionPure and Mixed Gases (Crushed Coal Pack)

• Small fraction of N2prevents severe perm reduction

• Overall perm reduction may be smaller at elevated temperature

• Overall perm reduction may be smaller in whole cores

Pure CO2Mixed Gases

Simulation of CO2 Selectivity During Flowco

al &

CH

4

CH4 + CO2 + N2

gas analyzer

CO2 /N2

p = 600 psiSw = 0

Flue Gas

Post Separation

CH4

DataSimulation

N2

85/15 CO2 /N2

CO2

24/76 CO2 /N2

N2

CH4

CO2

Sensitivity to Hysteresis

CH4 + N2 + CO2

54%N2+46% CO2

• Case 1: Base case• Case 2: Ignoring hysteresis of N2 adsorp/desorp

N2

CH4CO2

12

Advanced Modeling Questions

Analytical Solutions for Simple Flow Paths • Can multiphase, multicomponent flow in coal be

described with analytical 1D solutions?

Streamlines for Computational Efficiency• Can 1D analytical solutions be used for rapid

and efficient streamline calculations or are numerical solutions along streamlines required?

Analytical Solutions by the Method of Characteristics (and Finite Differences)

0

0.2

0.4

0.6

0.8

1

1.2

0 0.5 1 1.5 2

Sg, u

D

SgSg (fd 1000)uD

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2

zCH

4 , zC

O2 ,

zH2O

zCH4zCO2zH2O

λ

I

A B

CC

O

CH4

CO2H2O

IA

C

BO

• Compositional mechanisms are revealed by the analytical theory.

• Numerical solutions will likely be required for streamline simulations with gravity.

• 1D solutions will be useful for evaluation of adsorption models and comparison with experiments.

Large CH4 bank propagates through displacement

Single PhaseThree Component

Dimensionless Flow Velocityinlet outlet

Analytic/Finite Difference

Extension of Analytic Solutionsto 4 Component/2 Phase Flow

H2O

CH4

N2 CO2

crossover tieline

AB

O

CD

AFE

I

0

0.4

0.8

1.2

0 0.5 1 1.5 2 2.5 3

zN2,

z CH

4, z C

O2,

z H2O

zN2zCH4zCO2zH2O

0

0.4

0.8

1.2

1.6

0 0.5 1 1.5 2 2.5 3

Sg, u

D

SgSg (fd 1000)uD

λ

A

OB

CCF

D D

E

I

Relatively diffuse displacement front

Analytic/Finite Difference

Dimensionless Flow Velocity

Streamlines for 3D Coal Bed Simulation?

• Streamlines can capture flow in heterogeneous systems.

• If they evolve slowly, then 1D solutions along streamlines can be used to model the combination of compositional effects and flow.

• Streamline computations are fast and affected less by numerical dispersion.

• Compare with conventional coal bed simulation approaches.

Adaptive pressure solveAdaptive pressure solve Move compositionsMove compositionsMOC (analytic) or MOC (analytic) or

HO (numerical) upwindHO (numerical) upwind

Accurate streamline tracingAccurate streamline tracing

Gas injection: streamlines for transport

Improved mapping to streamlinesImproved mapping to streamlinesOr, Improved mapping to pressure gridOr, Improved mapping to pressure grid

Stop

Seismic Monitoring of CO2 Storage in Coal Beds

• Acoustic measurements of saturated coals- Differential Acoustical Resonance Spectroscopy- Changes in compressibility and attenuation with saturation

• Seismic monitoring of CO2 Storage- Time-lapse simulation of seismic attributes- Integration of active and passive monitoring strategies

Time Lapse:AcousticalProperties

vs.Saturation:

CH4, CO2, H2O

DARSAcoustical

Spectroscopy

Low FrequencyCompressibilityAttenuation

3-D Seismic Simulation

Active/Passive SeismicSeismic attributesDynamic aperture geometryTime regularized inversion

DARS: Differential Acoustical Resonance Spectroscopy

Phase-lockAmplifier

Power Amp

SignalGenerator

Charge Amp

Detector

Xducer

Tank

Cavity

PC ControllerSamplePositioner

Resonance Profile

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 12000

0.05

0.1

0.15

0.2

0.25

Frequency (Hz)

Acoustic Resonance Spectrocopy

d31:c15.85Td31:c46.85T

Boise SSEmpty

Q= f / W

ws

AAs

w

f fs

Ener

gy D

ensi

ty

Frequency (Hz)

Posi

tion

AcousticVelocity(1st mode)

Acousticpressure

DARS Calibration with Gas Permeability

k - gas injection (mDarcy)

κe = κ∞ +

2ϕκ f

l0

Diω

1− exp(− iω / D l0

2)

⎡

⎣ ⎢

⎤

⎦ ⎥

D = k /ϕη(κ f + κ p ) l0 = samplelength

Dynamic Diffusion

Permeability (gas injection), mD

Per

mea

bilit

y (d

iffus

ion

mod

el),

mD

Resonance Profiles

DARS Measurements of Coal Samples

2.42+/-0.3240*87+/-0.5Coal

Sample 5

2.31+/-0.2770*12.111+/-0.5Coal

Sample 4

2.51+/-0.2325*13.915+/-0.5Coal

Sample 3

2.62+/-0.15218*7.1418+/-0.5Coal

Sample 2

2.73+/-0.150.05 – 0.11.919+/-0.5Coal

Sample 1

DARS Bulk

Modulus (GPa)

EstimatedPermeability

(md)

MeasuredPorosity

( %)

Measured Volume

(cc)

Coal Samples

* Dynamic diffusion estimates of permeability

High Interest: Attenuation

Berea-1 Berea -2 Boise Chalk CoalDensity (g/cc) 2.1 2.1 2.3 1.8 1.1Porosity (%) 22.0 19.2 12.0 34.0 1.9Permeability (mD) 367.7 127.1 0.9 2.1 0.1V dry (km/s) 2.5 2.8 2.6 3.0 2.2

3.00

3.20

3.40

3.60

3.80

4.00

4.20

4.40

1076 1078 1080 1082 1084 1086 1088 1090 1092

Frequency (Hz

Berea-1

Berea-2Chalk

BoiseCoal

Line

wid

th(H

z)

Frequency (Hz)

Q = f /W

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 12000

0.05

0.1

0.15

0.2

0.25

Frequency (Hz)

Acoustic Resonance Spectrocopy

d31:c15.85Td31:c46.85T

Boise SSEmpty

Q= f / W

ws

A

As

w

f fs

Ener

gy D

ensi

ty

Frequency (Hz)

DARS: Flow-induced Attenuation(Drained vs Undrained Sample)

Berea Sandstone Closed

Open

250

260

270

280

290

300

310

320

330

340

350

-50 -40 -30 -20 -10 0 10 20 30 40 50

Sample positioSample position

Qsys

Seismic Monitoring of CO2 Storage in Coal Beds

• Acoustic measurements of saturated coals- Differential Acoustical Resonance Spectroscopy- Changes in compressibility and attenuation with saturation

• Seismic monitoring of CO2 Storage- Time-lapse simulation of seismic attributes- Integration of active and passive monitoring strategies

Time Lapse:AcousticalProperties

vs.Saturation:

CH4, CO2, H2O

DARSAcoustical

Spectroscopy

Low FrequencyCompressibilityAttenuation

3-D Seismic Simulation

Active/Passive SeismicSeismic attributesDynamic aperture geometryTime regularized inversion

Triaxial Deformation Experiments

• Permeability as a function of effective stress, differential stress, CO2 saturation and temperature

• Static and dynamic elastic moduli

• Coal plasticization

Conventional Triaxial PressMax. Confining Pressure = 100 MPa

Pressure Vesselwith Heater Coil

CoreholderSample assembly

Modification of Triaxial Testing Systemfor Coal

Pore Pressure Inlet

Axial and Radial Strain Gauges

Axial Load Cell

Pore Pressure Outlet

Ultrasonic P/S2Transducers

Core Holder Assembly

Permeability - Effective Stress

0.1

1

10

100

1000

0 200 400 600 800 1000 1200 1400 1600

Effective Stress (psi)

Pe

rmeab

ilit

y (

mD

)

Field Measurements(San Juan Basin)

Lab Measurements(Appalachian Basin)

McKee et al., 1988

Permeability and Effective Stress

Laboratory Samples(Appalachian Basin)

Field Measurements(San Juan Basin)

Per

mea

bilit

y, m

D

Effective Stress (psi)

Effect of CO2 on Coal Plasticity

0

50

100

150

200

250

300

350

400

450

0 1 2 3 4 5 6 7

Effective Stress (MPa)

So

fte

nin

g T

em

pe

ratu

re (

Ce

lsiu

s)

CO2Helium

Khan and Jenkins (1985)

Effect of CO2 on Coal Plasticity

Pla

stic

izat

ion

Tem

pera

ture

of B

itum

inou

s C

oal,

ºC

400

300

200

100

0

Effective Stress (MPa)0 1 2 3 4 5 6 7

CO2

He

Transport Modeling of CO2 Storage and ECBMCarolyn Seto, Kristian Jessen and Lynn Orr

Sorption and Transport in Coalbeds: A Laboratory and Simulation Investigation

T. Chaturvedi, G.-Q. Tang, W. Lin, K. Jessen and Tony Kovscek

Seismic Monitoring of CO2 Storage in Coal BedsChuntang Xu, Jolene Robin-McCaskill, Youli Quan and Jerry Harris

Simulation Studies of CO2 Sequestration and ECBM in Coalbeds of the Powder River Basin

Hannah Ross, Paul Hagin and Mark Zoback

_______________

For more information, see 5 posters this afternoon

GCEP-Supported Coal Sequestration Project Studies at Stanford

Permeability ReductionDue to Surface Coverage?

Pure Gases

N2

CO2

Crushed Coal Pack

Sensitivity Analyses