Risks Posed by Brines Containing Dissolved CO 2 Ron Falta 1 , Larry - PDF document

Risks Posed by Brines Containing Dissolved CO 2 Ron Falta 1 , Larry Murdoch 1 , and Sally Benson 2 Catherine Rupecht 1 , Lin Zuo 2 , Kirk Ellison 1 , Chris Patterson 1 , Shuangshuang Xie 1 , Miles Atkinson 1 , Laura Daniels 1 , Qi Zheng 1 1 Clemson University 2 Stanford University January 7, 2013 834383 1

CO 2 Density and Solubility with Depth CO2 Phase Density CO2 Solubility 900.00 60.00 800.00 50.00 700.00 600.00 40.00 Solubility, g/l density, g/l 500.00 30.00 400.00 20.00 300.00 200.00 10.00 100.00 0.00 0.00 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Depth, feet Depth, feet Calculated using TOUGH2-ECO2N assuming 35 o C and 10,000 mg/l NaCl Risks Posed by Brines Containing Dissolved CO2 2 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

The high CO 2 solubility is significant  At 3000 ft depth, we get ~50 g/l (50 times more CO 2 than beer!) Brine Density  When CO 2 dissolves, the 1020.000 aqueous phase becomes more 1015.000 dense (about 1% here) brine density, g/l brine density with dissolved CO2  Upward flow would require a 1010.000 brine density without dissolved CO2 caprock defect, and an upward 1005.000 hydraulic gradient > density 1000.000 difference 0 1000 2000 3000 4000 5000 depth, feet Calculated using TOUGH2-ECO2N Risks Posed by Brines Containing Dissolved CO2 3 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012 3

The he Di Disso ssolved C CO 2 is is Se Secu cure -- -- Or Is r Is It It?  Solubility trapping – CO 2 dissolves in pore water (up to 60 g/l)  Density increase favors downward flow of CO 2 saturated brine  Upward flow would require a caprock defect, and an upward hydraulic gradient > 1%  However, if a CO 2 saturated brine moved upward, the CO 2 would come out of solution (exsolve), leading to a potentially mobile gas phase IPCC, 2005 4 Risks Posed by Brines Containing Dissolved CO2 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Outline  Experiments  Pore  Core  Relative permeability  Modeling  Fault  Wells  Dissolved and supercritical injection  Outcrop Risks Posed by Brines Containing Dissolved CO2 5 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Laboratory Micromodel Study (Zuo, Zhang, Falta, and Benson, AWR, 2013) Binary Thin section image used micrograph for of Mt. micromodel Simon sandstone Micromodel: 530 mD; PV=1.35 uL Risks Posed by Brines Containing Dissolved CO2 6 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

a Micromodel  Initially fill micromodel with water saturated with dissolved CO 2 at 90 bars, 45 o C  Depressurize at a rate of 10 b bars/hr  Images taken at 1 second CO 2 intervals after onset of exsolution at 31 bars  CO 2 first starts to flow out at c 23.5 bars, with a CO 2 phase saturation of 56% CO 2 100um Risks Posed by Brines Containing Dissolved CO2 7 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Comparison of Exsolution and Supercritical CO 2 Injection exsolution , 31 bar exsolution , 25 bar exsolution , 18 bar CO 2 injection , 45 bar Risks Posed by Brines Containing Dissolved CO2 8 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Core Scale Experimental Setup Core Holder CT Scanner Dual-pump System Risks Posed by Brines Containing Dissolved CO2 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Core Experiments Mobility of exsolved gas porosity a ( Zuo, Krevor, Falta, and Benson , TIMP, 2012)  Fill core with CO 2 saturated water at 124 bar, 50 o C  Depressurize to 27 bar at a rate of 12 bars/hr  CO 2 phase saturation reaches CO 2 saturation >40%, but very low mobility c  No gravity redistribution after 11days.  CO 2 is mobile at 3% gas saturation during flood of the same core Risks Posed by Brines Containing Dissolved CO2 10 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Core Experiments Relative permeability Mt. Simon Sandstone (15.7 mD, 23.9 % porosity) Core Flood Relative Permeabilities Exsolution Relative Permeabilities 1 1 0.9 Krw 0.8 0.1 Relative Permeability Relative Permeability Krg 0.7 krw model krw_new 0.6 0.01 krg model krg_new 0.5 krw model 0.4 0.001 krg model 0.3 0.2 0.0001 0.1 0 0.00001 0.5 0.6 0.7 0.8 0.9 1 0.5 0.6 0.7 0.8 0.9 1 Water Saturation Water Saturation CO 2 phase injection CO 2 exsolution from brine Risks Posed by Brines Containing Dissolved CO2 11 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Core Experiments Hysteretic CO 2 phase trapping  Core flood experiments where CO 2 saturation was cyclically increased and decreased to Linear Trapping Model measure trapping 1.0  CO 2 saturation was measured 0.9 0.8 by CT scan Residual S CO2 0.7 0.6 0.5  Trapped CO 2 is a linear 0.4 0.3 function of maximum CO 2 y = 0.5x 0.2 saturation 0.1 0.0 0.00 0.20 0.40 0.60 0.80 1.00 Maximum S CO2 Risks Posed by Brines Containing Dissolved CO2 12 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Core Experiments New relative perm model for hysteretic CO 2 phase trapping Hysteric CO2 Relative Permeability 0.10 Berea Sandstone  Simple approach: residual saturation a function of maximum k rCO2 saturation 0.05  Continuously update the max residual saturation  Allows use of existing relative permeability 0.0 models 0.5 0.6 0.7 0.8 0.9 1.0 ( ) 2 m S w 1 = − ˆ − ˆ k k 1 S 1 S m rg rg max w w − S S = − ˆ w wr S − w 1 S S wr gr Risks Posed by Brines Containing Dissolved CO2 13 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Modeling Open fault model using TOUGH2-ECO2N Ground Surface +200 30m Confining Layer No flow 0 k x = 10 -11 m 2 k z = 10 -12 m 2 Drinking Water Aquifer -100 Open Fault k z = 10 -11 m 2 , 50m wide Confining Layer k x = 10 -15 m 2 k z = 10 -16 m 2 - 700 k x = 10 -11 m 2 k z = 10 -12 m 2 Saline Formation with Dissolved CO2 -800 50.7 g/L CO 2 Model extends 5,000m in either direction Risks Posed by Brines Containing Dissolved CO2 14 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Model using regular core flood relative permeabilities. Time is 30 years. Dissolved CO 2 mass fraction Gas saturation 400 400 XCO2a: 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 SG: 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 200 200 Y Y 0 0 -200 -200 -400 -400 -600 -600 2000 4000 6000 8000 2000 4000 6000 8000 X X Dissolved salt mass fraction 400 Gas phase CO 2 reaches the XNACL: 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 200 DWA, and spreads to the Y boundaries at 5000m within 30 0 years if the drawdown is -200 maintained. -400 -600 2000 4000 6000 8000 Risks Posed by Brines Containing Dissolved CO2 15 X Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Model using exsolution relative permeabilities. Time is 30 years. Dissolved CO 2 mass fraction Gas saturation 400 400 XCO2a: 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 SG: 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 200 200 Y 0 Y 0 -200 -200 -400 -400 -600 -600 2000 4000 6000 8000 2000 4000 6000 8000 X X Dissolved salt mass fraction • Leakage much less using 400 exsolution relative permeability XNACL: 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 200 Y 0 • Related simulations for wells similar -200 • In all cases, CO 2 migration stops -400 when head imbalance is corrected, -600 no runaway effect 2000 4000 6000 8000 X Risks Posed by Brines Containing Dissolved CO2 16 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Modeling CO 2 injection as dissolved or supercritical Q Properties: Formation: 300m thick, 20km x 20 km Typical of deep sandstone Slope: 0.008, 8m/1km Stochastic distribution Hysteretic capillary and rel. Injection rate: 10 kg CO 2 /s for 20 years Monitoring period: 30 years perm functions Risks Posed by Brines Containing Dissolved CO2 17 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Modeling results CO 2 injection as dissolved or supercritical Supercritical Dissolved 1 km Similar areal footprints after injection ~10 km 2 • Supercritical CO 2 moves after injection, increasing area by 50% (14.9 km 2 ) • Dissolved CO 2 sinks after injection, decreasing area contacting caprock (8.9 km 2 ) • Risks Posed by Brines Containing Dissolved CO2 18 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012

Conclusions  Brine containing dissolved CO 2 can be mobilized upward by modest hydraulic gradients  As the carbonated brine is depressurized, the CO 2 comes out of solution (exsolves) throughout the pore space  The exsolved CO 2 phase has a very low relative permeability, even at high phase saturations. Exsolution relative permeability function  Hysteric relative permeability represented by updating residual saturation in standard models. Simple, fits data well.  Upward flow of brines containing dissolved CO 2 stops when the external driving force is removed, no runaway instability seen.  Injection of CO 2 as a dissolved phase is likely to have a similar “footprint” to supercritical CO 2 injection, less mobile after injection. Risks Posed by Brines Containing Dissolved CO2 19 Falta, Murdoch, Benson, USEPA STAR Progress Review 7 Jan 2012