Martin Blunt, Branko Bijeljic, Tara C LaForce, Stefan Iglauer, Ran Qi, Saleh Al- Mansoori, Chris Pentland and Erica Thompson Impact of capillary trapping.

Martin Blunt, Branko Bijeljic, Tara C LaForce, Stefan Iglauer, Ran Qi, Saleh Al-Mansoori, Chris Pentland and Erica Thompson

Impact of capillary trapping on geological CO2 storage

Outline

Field scale: Streamline Simulation

Core scale: Column Experiment

Pore scale: CT scan

Background

Long term fate, how can you be sure that the CO2 stays underground?

Field scale - The streamline method

Permeability field

Initial saturation

Pressure solve

SL tracing

Saturation along SL

Saturation for the next

time step

Streamline method for CO2 transport

Hydrocarbon phase Aqueous phase

Todd&Longstaff

Fingering model for CO2 in oil

Phases (3) Components (4)

Hydrocarbon

Aqueous

Solid

CO2

Oil

Water Salt

+

+

+

+

+

+

+

+

+

Streamline method for CO2 transport

Trapping model

Pore-scale model matches experimental data.• Kr is from Berea sandstone, which matches Oak (1990)’s

experiments.• CO2/water system is weakly water-wet (Chiquet et al., 2007)

contact angle (θ) = 65º.

New trapping model (Juanes et al., 2006)

2maxmaxgggt SSS

))((4

)1(1

2

1max

2

gggtggtggf

SSSSSSS

Design of carbon dioxide storage

The ratio of the mobility of injected brine and CO2 to the formation brine as a function of the injected CO2-phase volume fraction, fgi.

0.01

0.1

1

10

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

f ci

Mo

bil

ity r

ati

o

Mobility ratio between carbon dioxide/brine mixture and formation brine

Mobility ratio between chase brine and carbon dioxide/brine mixture during chase brine injection

Mobility ratio = 1.0

fgi

Design of carbon dioxide storage

1D analysis: Numerical simulation vs. analytical solution

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0 200 400 600 800 1000 1200 1400

Distance (m)

Sg

Simulation

Analyticalsolution

Trapped CO2 Mobile CO2

Dissolution front

Advancing CO 2 front

Chase brine front

fgi = 0.5

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 200 400 600 800 1000 1200 1400

Distance (m)

Sg

Simulation

Analytical solution

Trapped

CO2

Mobile CO2

Dissolution front

Advancing CO 2 front

Chasebrine front

fgi = 0.85

Design of carbon dioxide storage

Mobile CO2 saturation

Z

170m

X3200m

Y

2280m

Trapped CO2 saturation

X3200m

Y

2280m

Z

170m

Injector

Producer

SPE 10 reservoir model, 1,200,000 grid cells (60X220X85), 7.8 Mt CO2 injected.

Two years after chase water injection for fgi=0.85.

Design of carbon dioxide storage

3D simulation: Storage efficiency vs. trapping efficiency

Storage efficiency =

the fraction of the reservoir pore volume filled with CO2

Trapping efficiency =

the fraction of the injected mass of CO2 that is either trapped or dissolved

The storage efficiency is highest for fgi = 0.85, which also requires minimum mass of chase brine to trap 95% of CO2.

Design Criterion

• Inject CO2+brine where mobility ratio = 1.0

(fgi=0.85 in this example).

• Inject chase brine that is 25% of the initially injected CO2 mass.

• 90-95% of the CO2 is trapped.

Issues arising from field scale simulation

• Streamline-based simulator has been extended to model CO2 storage in aquifers and oil reservoir by incorporating a Todd-Longstaff model, equilibrium transfer between phases (dissolution) and rate-limited reaction;

• Trapping is an important mechanism to store CO2 as an immobile phase. Our study showed that WAG CO2 injection into aquifer can trap more than 90% of the CO2 injected;

• We have proposed a design strategy for CO2 storage in aquifers, in which CO2 and formation brine are injected simultaneously followed by chase brine.

• Streamline-based simulation combined with pore-scale network modeling can capture both the large-scale heterogeneity of the reservoir and the pore-scale effects of trapping.

Future work

Injection strategy design

• Require better experimental data, since the trapping model used has a significant impact on the results.

• Design of an injection strategy to maximize CO2 storage and oil recovery.

CT ScanningCT Scanning

A homogeneous sandpack was compressed A homogeneous sandpack was compressed and the porosity was determined via mass and the porosity was determined via mass balance (Φ = 38,93).balance (Φ = 38,93).n-Heptane was injected; when no more n-Heptane was injected; when no more brine was produced, another CT scan was brine was produced, another CT scan was performed at the irreducible water performed at the irreducible water saturation, Ssaturation, Swiwi..COCO22 was injected again. Gas injection was was injected again. Gas injection was stopped when no more liquid production stopped when no more liquid production was observed. Another CT scan was taken.was observed. Another CT scan was taken.30 pore volumes (PV) of brine were 30 pore volumes (PV) of brine were injected and a final CT scan was taken at injected and a final CT scan was taken at the residual gas saturation Sthe residual gas saturation Sgr gr .. resolution resolution 9 µm9 µm

Sandpack at irreducible water saturation

Brine – blueSand – redOil - orange

• Oil penetrates on average mainly into the larger pores as expected by capillary pressure considerations. • Thin water layer is visible on the rock surface as expected for quartz.• Oil has penetrated into the middle of some pores.

Sandpack at residual gas saturation

Brine –blueSand – redCO2 - yellow

• The largest CO2 ganglia is continuously spread over the largest available pore.• Though overall gas accumulates in the larger pores, a random distribution between large and medium size pores is observable.• Several tiny gas bubbles are randomly distributed in the pore volume. Though they might originate from the segmentation process, it is thought that they are real.

Vertical column experiments – Sor vs. Soi

•Sand-packed columns were oriented vertically.

•5 pore volumes of de-aired brine were injected to reach full saturation.

•Decane reservoir connected to top of columns and brine allowed to drain under gravity from the base. Decane enters the top of the column. No pumping.

•Equilibrium reached where both columns have a (theoretically) identical oil saturation profile versus height.

•One column removed for slicing and sampling – Soi.

•Second column has brine injected from the base, Brine sweeps oil leaving an Sor. Coulmn sliced and sampled.

brine flow

Oil

flow

Oil

flow

brine flow

COLUMN A - Soi COLUMN B - Sor

Vertical column experiments – Sor vs. Soi - results

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.0 0.2 0.4 0.6 0.8 1.0

Soi

So

r

Land - Experiment 2Experimental Data - Exp. 2Core Flood ResultsLand - Core Flood