Scholarly article on topic 'Numerical Study on Field-scale Behavior of Carbon in CO2 Micro Bubble Storage (CMS)'

Numerical Study on Field-scale Behavior of Carbon in CO2 Micro Bubble Storage (CMS) Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
Keywords
{"Carbon dioxide" / "Green house gas" / CCS / Micro-bubble / CMS / "Numerical simulation"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Satoru Miyoshi, Takashi Hitomi, Hideaki Miida, Hiroshi Wada, Kaoru Inaba, et al.

Abstract CO2 micro bubble storage (CMS) is proposed as one of the technical alternatives for CCS that micro bubble of CO2 is dissolved in groundwater in a borehole and the water saturated by CO2 is injected into aquifer. In this study, the preliminary study was done that CO2 behavior in CMS was numerically simulated under the operational condition that CO2 micro bubble is injected into one well and the groundwater flow around the well is controlled using four withdrawing wells one-hundred-meter apart from the injection well. The results show that CO2 migrates in the aquifer, the excess pore pressure generated by CO2-dissolved water injection is so small that rock around the injection well can be mechanically stable and CO2 independent phase is not generated.

Academic research paper on topic "Numerical Study on Field-scale Behavior of Carbon in CO2 Micro Bubble Storage (CMS)"

Available online at www.sciencedirect.com

SciVerse ScienceDirect

Energy Procedia 37 (2013) 5978- 5985

GHGT-11

Numerical Study on Field-scale Behavior of Carbon in CO2 Micro Bubble Storage (CMS)

Satoru Miyoshia, Takashi Hitomia, Hideaki Miidab, Hiroshi Wadab, Kaoru Inabac, Masayuki Yamaurad*

aObayashi Corporation, Shimo-kiyoto 4-640, Kiyose city, Tokyo 204-8558, Japan bEngineering Advancement Association of Japan (ENAA), Toranomon 3-18-19, Minato-ku, Tokyo 105-0001, Japan cTakenaka Corporation, Otsuka 1-5-1, Inzai citi, Chiba 270-1395, Japan dDia Consultants, Yoshino-cho 2-272-3, Saitama city, Saitama 331-0811, Tokyo

Abstract

CO2 micro bubble storage (CMS) is proposed as one of the technical alternatives for CCS that micro bubble of CO2 is dissolved in groundwater in a borehole and the water saturated by CO2 is injected into aquifer. In this study, the preliminary study was done that CO2 behavior in CMS was numerically simulated under the operational condition that CO2 micro bubble is injected into one well and the groundwater flow around the well is controlled using four withdrawing wells one-hundred-meter apart from the injection well. The results show that CO2 migrates in the aquifer, the excess pore pressure generated by CO2-dissolved water injection is so small that rock around the injection well can be mechanically stable and CO2 independent phase is not generated.

© 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT

Key Words : carbon dioxide; green house gas; CCS, micro-bubble; CMS; numerical simulation

1. Introduction

Carbon Capture and Storage (CCS) is one the technologies that is expected to contribute the decrease of Green House Gas (GHG). Among the technical alternatives of CCS, Koide and Xue[1] proposed CMS

* Corresponding author. Tel.: +81-424-95-1031; fax: +81-424-95-1261. E-mail address: miyoshi.satoru@obayashi.co.jp

1876-6102 © 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT doi: 10.1016/j .egypro .2013.06.525

that is economical and does not generate CO2 leakage. In CMS, CO2 is injected into shallow geology as microbubbles, while CO2 is injected into geology as supercritical phase in the general concept of CCS. Suzuki et al.[2] studied on the feasibility of CMS from the view points of technical, economic, and legal aspects. They proposed the operational condition that CO2 microbubbles are mixed with water in a vertical injection well and CO2-saturated water is injected. Groundwater is withdrawn through four vertical wells located around the injection well to control the groundwater flow around the wells. They indicated that CO2 can be economically and safely stored in an aquifer, a high-permeable sand rock layer that is present in the depth from 300 meters to 500 meters under the ground surface.

In this study, the field-scale behavior of CO2 in CMS is numerically simulated using TOUGH2 with ECO2N module (Pruess and Garcia[3], Pruess[4]) to understand the effect of various operational conditions such as the existence of an aquitard, a low-permeable layer as a cap layer over the aquifer, the vertical position of injection in the aquifer, and the distance between the injection well and each withdrawing well. In the numerical study, the multiphase multi-component fluid simulator TOUGH2 with the ECO2N module developed by Laurence Berkeley National Laboratory of United States is used because it should be confirmed if the dissolved CO2 is vaporized and free gas phase is generated.

2. Methods

2.1. Geological Settings and Basic Operation of CMS

The aquifer is present at the depth from 300 meters to 500 meters under the ground surface. There is low-permeable mud rock layer over and below the aquifer. Groundwater is withdrawn through four vertical wells located around the injection well to control the groundwater flow around the wells so that dissolved CO2 can be dispersed in the aquifer. The flow rate of the injection and the withdrawal is balanced. The injection rate is 7 kg/sec as the CO2 solution and groundwater is withdrawn through the withdrawing wells.

2.2. Evaluation Points

The results of numerical simulation of each case are evaluated from the three points: the distribution of CO2 concentration, pore pressure, and water saturation.

Through the contours of dissolved CO2 concentration, it is confirmed if injected CO2 is dispersed widely in the aquifer surrounded by the wells.

If the pore pressure around the injection well increases drastically, the stress field is drastically changed and rock could be destructed. It is not easy to set the threshold level of permissible pore pressure because it depends on various factors. It can be assumed that rock is not destructed if the change of pore pressure is below the value corresponding to the casual change of groundwater level such as several tens of meters. The distribution of pore pressure is confirmed through contours.

The permeability of gas phase is as about one thousand times as liquid phase. Buoyancy acts on gas phase according to the difference of density. As the result, the driving force to move CO2 upward increases if the gas phase of CO2 is generated. In that case, the capability of CO2 storage depends on the performance of the mud rock layer over the aquifer as the capillary barrier and the risk of CO2 leakage could increase. Because the density of CO2-saturated groundwater is a little greater than the normal groundwater, no upward flux of CO2 is generated in normal geological settings. Considering from the above, it is important in CMS to limit the CO2 injection rate within the level that no gas phase CO2 is generated. The distribution of independent phase saturation is confirmed through contours.

2.3. Numerical Simulator TOUGH2

The multiphase multi-component fluid simulator TOUGH2 with the ECO2N module developed by Laurence Berkeley National Laboratory of United States is used because it should be confirmed if the dissolved CO2 is vaporized and free gas phase is generated.

TOUGH2 is a multiphase multi-components fluid simulator. It can solve the behavior of the combination of various fluids and solutes. The combination is given as EOS modules that implement multi-phase fluid flow equations and the state equations of fluids and solutes. ECO2N, one of the EOS modules, implements those equations of CO2, water, and brine under the thermodynamic state that temperature is 10 to 100oC, the pore pressure is 0.1to 60MPa, and the concentration of sodium chloride is less than saturated solubility. It is sometimes used for the simulation of CCS [3].

Phase partitioning characteristics and thermodynamic properties of the mixture of water, CO2, and Sodium Chloride are implemented in ECO2N as follows.

CO2 is present as solute in water, independent phase, and both of them. Gas phase is the mixture of vapor and CO2. The mixing ratio depends on the temperature and the pressure. Sodium Chloride is present only in liquid phase as electrolytes. Because the concentration of sodium chloride affects the free energy of water and CO2, the mixing ratio in gas and liquid phase depends on it. Summarizing the data of laboratory experiments on the phase partitioning characteristics of water and CO2 under various conditions of the temperature, the pressure, and the sodium chloride concentration, Spycher[5] and Spycher and Pruess[6] proposed a numerical thermodynamic model, which is implemented in ECO2N.

Because the mixing ratio of water in gas phase is less than 1% at most, the error is negligible if the thermodynamic properties of gas phase are approximated as those of the independent phase of CO2. In ECO2N, the density and the viscosity of independent CO2 phase given by Altunin[7] is used as those of gas phase. The specific enthalpy of gas phase is given by Garcia[8]. The density, the viscosity, and the specific enthalpy of liquid phase are given by Garcia[8], Phillips[9], and Lorentz et al.[10] , respectively

2.4. Assumptions for Numerical Study

Because the aquifer is shallow, the groundwater in it originates from precipitation and the electrolytes concentration is negligibly low.

The excess CO2 more than solubility is returned to the ground surface through the double tubes in the injection well when CO2 micro bubble is injected. That means that the concentration of CO2 is the saturated solubility according to the temperature and the pressure. No independent phase CO2 is injected to the aquifer.

2.5. Settings of Numerical Simulation

The discrete model for the simulation is shown in Fig.1.

The three dimensional area of horizontal (x-y plane) 500m x 500m by vertical (z axis) 900m that involves the injection well and the withdrawing wells is divided to 10m x 10m x 10m grids. The dividing width is 100m within 100m from the sides, 200m from the surface, and 300m from the bottom. The injection well is given as the source to the grids that place horizontally at the center and vertically at the depth from 350 to 450 meters. Each withdrawing well is given as the sink from horizontally continuous two grids that place at the depth from 300 to 500 meters. The wells and the vicinity are not modeled precisely.

There is low-permeable cap layer at the depth from 250 to 300 meters and high-permeable aquifer at the depth from 300 meters to 500 meters.

For the boundary conditions, the pressure is the atmospheric pressure (100kPa) and 15oC at the ground surface. The sides and the bottom are no-flux boundary. The initial conditions are set using the geothermal gradient of 2.5oC/100m and the hydrostatic pressure gradient 1MPa/100m. The initial mass fraction of CO2 in groundwater is 0% and that of the injected water is 4.0% less than the saturated solubility under the CO2 partial pressure 4MPa , 4.4%[5J. 4MPa is approximately the hydrostatic pore pressure at the depth of 400 meters.

Fig. 1. The discrete model for numerical simulation

The permeability and the porosity of the sand rock and the mud rock are shown in Table 1.

Table 1 Permeability and porosity of sand rock and mud rock

sand rock mud rock

permeability ( m2) 1.0x10-13 1.0x10-16

porosity (-) 0.35 0.30

The characteristics of two phase flow are shown in the Table 2[3J. Here the phase dominated by water is called water phase and the phase dominated by CO2 is called gas phase. The same value of capillary pressure is used for the both layers. That setting is conservative for the performance of mud rock as a cap layer. Fig. 2 and Fig. 3 show the plot of relative permeability and capillary pressure, respectively.

The simulation cases are set as shown in Table 2. In the cases 1, 2, and 3, the distance between the injection well and each withdrawing well is 100 meters. The injcetion depth in the case 1 is 350 to 450 meters from the surface that is the middle 100 meters of the aquifer and the injection depth in the case 2 is 400 to 500 meters from the surface that is the lower 100 meters. In the case 3, the permeability of mud rock is the same as that of sand rock, that is to say, there is no cap layer. In the case 4, the distance between the injection well and each withdrawing well is 200 meters. The area for the simulation is devided into 20m x 20m x 20m grids.

Satoru Miyoshi et al. /Energy Procedia 37 (2013) 5978- 5985 Table 2 Two phase flow characteristics of sand rock and mud rock

relative permeability capillary pressure

van Genuchten function van Genuchten-Mualem model

(van Genuchten11-1) (Mualem12); van Genuchten11-1)

sand rock clay rock sand rock clay rock

X 0.457 0.457 X 0.457 0.457

Slr 0.30 0.30 Slr 0.0 0.0

Sls 1.0 1.0 1/P0 5.1x10-5 5.1x10-5

Sgr 0.05 0.05 Pmax 1.0x107 1.0x107

Sls 0.999 0.999

ft <5 >§0.

■-a t3

0 0.5 1

water saturation / dimensionless

1.E+06

1.E+04

1.E+02

2 Plot of water saturation - relative permeability

Table 3 simulation cases

1.E+00

0 0.5 1

^ water saturation / dimensionless Fig. 3 Plot of water saturation - capillary pressure

Injection depth Cap layer

Horizontal distance between injection well and withdrawing wells

Case 1 Case 2 Case 3 Case 4

350 - 450m 400 - 500m 350 - 450m 350 - 450m

Present Present

No Present

100m 100m 100m 200m

3. Results and Discussion

3.1. Case 1

Dissolved CO2 reached the withdrawing wells four years after starting the injection. The contour map

of CO2 mass fraction in pore water at that time is shown in Fig. 4. The isosurface of 4% is also shown in

it. The volume where the CO2 mass fraction is the same as the injected solution (4%) spherically dispersed in the aquifer surrounded by the withdrawing wells.

The pore pressure distribution at the same time on the horizontal cross section at the middle depth of the aquifer is shown in Fig. 5. The dimension of the number in the legend is Pa. Because the initial pressure at the depth of 400 meters is 4.1MPa, the increment is approximately 600kPa. There was no grid where the water saturation is less than 1.

3.2. Case 2

The isosurface of 4% CO2 mass fraction in pore water four years after starting the injection is shown in Fig. 6. The contour map at the depth of 300 meters and 500 meters are also shown in it. CO2 migrated horizontally in the aquifer and dispersed upward to some extent.

The pore pressure distribution at the same time on the horizontal cross section at the depth of 480 meters is shown in Fig. 7. Because the initial pressure at the depth is 4.9MPa, the increment of the pore pressure of the grids around the injection well is approximately 200kPa. There was no grid where the water saturation is less than 1.

3.3. Case 3

The isosurface of 4% CO2 mass fraction in pore water four years after starting the injection is shown in Fig. 8. The contour map at the depth of 300 meters and 500 meters are also shown in it. The distribution of CO2 mass fraction is almost the same as case 1. Even though there is no cap layer over the aquifer, remarkable upward flow is not generated.

The pore pressure distribution at the same time on the horizontal cross section at the depth of 400 meters is shown in Fig. 9. Because the initial pressure at the depth is 4.9MPa, the increment of the pore pressure of the grids around the injection well is approximately 250kPa. The reason why the increment of pore pressure is less than that in case 1 is that the pore pressure can propagate upward to some extent. There was no grid where the water saturation is less than 1.

3.4. Case 4

Dissolved CO2 reached the withdrawing wells six years after starting the injection. The contour map of CO2 mass fraction in pore water at that time is shown in Fig. 10. The isosurface of 4% is also shown in it. The volume where the CO2 mass fraction is the same as the injected solution (4%) spherically dispersed in the aquifer surrounded by the withdrawing wells.

The pore pressure distribution at the same time on the horizontal cross section at the depth of 400 meters is shown in Fig. 11. The increment of the pore pressure of the grids around the injection well is approximately 200kPa. There was no grid where the water saturation is less than 1.

4. Conclusions

From the above case studies, it is concluded that injected CO2 is migrated in the aquifer and no independent phase of CO2 is generated under the operational conditions given in this study. In addition, the increment of pore pressure generated by CO2-dissolved water injection is less than several tens of meters by piezometric head at most, which is within the casual change of groundwater level.

In the case 3 where there is no cap layer, there was upward flux greater than other cases to some extent. It can be pointed out that the upward flux of CO2-dissolved groundwater has to be monitored and controlled for CMS operation under the geological setting where there is no effective cap layer.

Fig. 4 Contour of dissolved CO2 mass fraction and isosurface of 4% CO2 mass fraction after four years in case1

Fig. 6 Contour of dissolved CO2 mass fraction and isosurface of 4% CO2 mass fraction after four years in case2

XC02o o.o-i

Fig. 5 Contour of pore pressure after four years in case1

S 1E+06 SOeE+CB 5.OS E+06 5.04E+06 5 02E+06

4.90E+06

4.9e£+O0

4.94S+06 4 92E+06 4.9E+06

Fig. 7 Contour of pore pressure after four years in case2

Fig. 8 Contour of dissolved CO2 mass fraction and isosurface of 4% CO2 mass fraction after four years in case3

Fig. 9 Contour of pore pressure after four years in case3

Fig. 10 Contour of dissolved CO2 mass fraction and isosurface Fig. 11 Contour of pore pressure after four years in case4 of 4% CO2 mass fraction after four years in case4

Acknowledgements

Our deepest appreciation goes to JKA (Rokuban-cho 4-6, Chiyoda-ku, Tokyo 102-0085, Japan) that financially assisted this study from the fiscal year 2010 to 2011.

References

[1] Koide, H. and Xue Z. Carbon microbubbles sequestration: A novel technology for stable underground emplacement of greenhouse gases into wide variety of saline aquifers, fractured rocks and tight reservoirs. Energy Procedia 2009; 1, 3655-3662.

[2] Suzuki et al. Feasibility Study on CO2 Micro Bubble Storage (CMS), Proceeding of the 11th International Conference on Greenhouse Gas Control Technologies (GHGT-11) 2012; in printing.

[3] Pruess, K. and Garcia, J Multiphase flow dynamics during CO2 disposal into saline aquifers, Environmental Geology 2002; 42, 282-295.

[4] Pruess, K. ECO2N: A TOUGH2 Fluid Property Module for Mixtures of Water, NaCl, and CO2, Lawrence Berkeley National Laboratory Report LBNL-57952, Berkeley, CA; 2005.

[5] Spycher, N. et al. CO2-H2O Mixtures in the Geological Sequestration of CO2. I. Assessment and Calculation of Mutual Solubilities from 12 to 100 °C and up to 600 bar, Geochim. Cosmochim. Acta 2003; 67, 3015 -3031.

[6] Spycher, N. and K. Pruess. CO2-H2O Mixtures in the Geological Sequestration of CO2. II. Partitioning in Chloride Brines at 12-100 °C and up to 600 bar, Geochim. Cosmochim. Acta 2005; 69, 3309-3320.

[7] Altunin, V.V. Thermophysical Properties of Carbon Dioxide, Publishing House of Standards, Moscow (in Russian)

[8] García, J.E. Density of Aqueous Solutions of CO2, Lawrence Berkeley National Laboratory Report LBNL-49023, Berkeley, CA; 2001

[9] Phillips, S.L., et al. A Technical Databook for Geothermal Energy Utilization, Lawrence Berkeley National Laboratory Report LBL-12810, Berkeley, CA; 1981

[10] Lorenz, S., D. Et al. Eine analytische Funktion zur Bestimmung der Enthalpie wässriger NaCl Lösungen, draft report, Institut für Sicherheitstechnologie, Köln, Germany; 2000