Permeability and Mineral Composition Evolution of Primary Seal and Reservoir Rocks in Geologic Carbon Storage Conditions

Abstract

It has been reported that deep saline aquifers represent the largest geologic CO₂ storage resource. To better predict containment effectiveness and long-term reservoir behavior of these formations, it is important to understand the potential geochemically induced changes to the porosity and permeability of both the primary sealing formation and CO₂ storage formation rocks. To investigate these potential changes, an experimental study to probe the geochemical interactions of CO₂/brine/rock system under geologic CO₂ storage conditions was conducted in a static reaction system. Marine shale (primary sealing formation) and Lower Tuscaloosa sandstone (CO₂ storage formation) core samples taken from the Plant Daniel CO₂ storage test site (Jackson County, Mississippi) were exposed to CO₂-saturated brine in a batch reactor at relevant geologic storage conditions (85°C and 23.8 MPa CO₂ pressure) for 6 months. X-ray diffraction, scanning electron microscopy, computed tomography, and brine chemistry analyses were performed before and after the exposure. Permeability measurements from the marine shale and sandstone core samples before and after CO₂/brine exposure indicated a significant effective permeability change. Sealing marine shale permeability increased following exposure while the permeability of the sandstone from the storage formation was observed to decrease. Analysis results of the primary sealing formation sample (marine shale) at the Plant Daniel CO₂ storage test site have not been reported before. The permeability decrease of the Lower Tuscaloosa sandstone sample reported in this study verifies the results reported in a previous study. These results have implications for the integrity of the primary seal in a CO₂ storage setting.

Introduction

Many options have been proposed to reduce or avoid emissions of anthropogenic CO₂ to the atmosphere to ameliorate the impacts of greenhouse gas effects on global climate change. One proposed technology option involves the capture of CO₂ from the gaseous by-product stream of large point sources (e.g., fossil fuel-fired power plants), compression, and transport of that separated CO₂ in a dense state to sites where it can be injected into deep geologic formations for long-term storage. The Intergovernmental Panel on Climate Change has reported that injection of CO₂ into confined geological formations represents one of the most promising solutions for mitigating global climate change (Pachauri and Reisinger, 2007). Candidate geological formations for possible CO₂ storage include depleted oil and gas fields, unmineable coal seams, basalt formations, shale intervals, and deep saline aquifers (Carbon Storage Atlas, 2015).

Recent studies have been reported regarding the sequestration of CO₂ in geological formations, including a storage capacity estimate for saline aquifers, effects on coal, coupling with enhanced oil recovery, and a modeling study of a CO₂ storage project in a reservoir in Cranfield, Mississippi (Haljasmaa et al., 2011; Rodosta et al., 2011; Delshad et al., 2013; Middleton, 2013; Ranjith et al., 2013; Jia et al., 2016). Deep saline aquifers represent the largest volume class of CO₂ storage targets (Carbon Storage Atlas, 2015) due to the large pore volume in sedimentary basins and the high density and solubility of CO₂ at the pressure and temperatures that exist in saline aquifers (Bachu and Adams, 2003).

Oil and natural gas bearing formations are also recognized as having significant CO₂ storage capacity, and CO₂ injection into those formations has the added advantage of enhancing the recoverable hydrocarbons, allowing a field to be more profitable. Demonstration scale injections of CO₂ storage in saline aquifers and hydrocarbon-bearing formations have been performed, and these experiences generate critical information to build confidence in the viability of those options.

The Southeast Regional Carbon Sequestration Partnership (SECARB) has identified the Lower Tuscaloosa formation in the Southeast region of the United States as a viable injection horizon for large-scale CO₂ storage demonstrations (Rodosta et al., 2011). One SECARB project targeting that formation is a small-scale injection conducted at a coal-fired power plant in Escatawpa called Plant Daniel in Jackson County, Mississippi (Rodosta et al., 2011); another is a larger scale injection in the Cranfield oil field near Natchez, Mississippi (Rodosta et al., 2011). For both injection locations the storage reservoir is the Lower Tuscaloosa sandstone formation, and the Tuscaloosa marine shale is the primary confining unit, with the Selma chalk group and Midway shale serving as secondary and tertiary containment units, respectively (SECARB, 2016).

Lower Tuscaloosa sandstone is about 36.5 m thick with an average porosity of 24% and permeability ranging from about 800 to 1,500 mD. The Lower Tuscaloosa sandstone, in southeast Mississippi, is composed of over 90% fine to medium grained sandstone. It is described as white and finely micaceous with a fine clay matrix interbedded with mudstone and siderite nodules (Griffith et al., 2011). The marine Tuscaloosa shale is a fissile, fine-grained, dark-gray micaceous shale with minor interbeds of calcareous, glauconitic sandstone with some argillaceous limestone lentils, laminated claystone, mudstones, and calcareous siltstone. The marine Tuscaloosa shale is predominantly composed of quartz, kaolinite, illite, hematite, and siderite based on point counts of thin section (Griffith et al., 2011). The average thicknesses of the confining systems are 152, 107, and 396 m for marine Tuscaloosa shale, Midway shale, and Selma chalk, respectively (Griffith et al., 2011).

As part of the SECARB projects, CO₂ has been injected into the Lower Tuscaloosa sandstone at both sites. About 4,300,000 metric tons of natural and anthropogenic CO₂ have been injected at Cranfield at depths from 2,591 to 3,124 m, and about 2,490 metric tons of CO₂ have been injected at Plant Daniel at a depth of 2,621 m. As CO₂ is injected into saline aquifers, it will dissolve into the formation fluids, form carbonic acid, lower solution pH, and possibly react with minerals in both CO₂ storage interval and the primary seal above the interval. The acidification sequences will impact the rock's properties (e.g., porosity, permeability, as well as reactive surface area) (Tutolo et al., 2015; Li et al., 2016; Pan et al., 2016). Better understanding of how rock/brine/CO₂ interactions affect injectivity and containment effectiveness will build confidence in the successful operation of subsurface CO₂ injection and storage (Delshad et al., 2013).

Investigations into host rock properties using both experimental and modeling studies have been conducted (Morse and Arvidson, 2002; Luquot and Gouze, 2009; Soong et al., 2014b; Zhang et al., 2015). Luquot and Gouze (2009) experimented with flowing CO₂-enriched aqueous phase through a carbonate core resulting in porosity and permeability changes from dissolution and/or precipitation at in situ geologic carbon storage (GCS) conditions of 100°C and pore fluid pressure of 12 MPa. Specifically, precipitation of Mg-rich calcite resulted in decreased permeability and porosity. Gouze and Luquot (2011) also conducted an X-ray microtomographic characterization of porosity, permeability, and reactive surface changes during the dissolution of pure calcite in a brine/CO₂ mixture and showed that an increase in permeability was due to decreasing the tortuosity for homogenous dissolution. Also, they concluded that extensive GCS in deep saline aquifers that contain carbonate minerals could result in dissolution of these minerals, which could in turn affect permeability (Gouze and Luquot, 2011).

Soong et al. (2014b) found that mineral dissolution/precipitation could occur under GCS conditions for samples obtained from the Mount Simon formation (Indiana). In addition, Zhang et al. (2015) conducted core-scale numerical simulations of Mount Simon sandstone porosity and permeability evolution under GCS conditions. The simulation predicts dissolution of feldspar and quartz with subsequent formation of kaolinite causing a decrease in permeability. A study conducted by Lu et al. (2012) on CO₂/rock/brine interaction in the Lower Tuscaloosa formation showed that mineral reactions in the Lower Tuscaloosa reservoir were minor during CO₂ injection and brine chemistry remained largely unchanged. The reservoir is composed mainly of minerals with low reactivity (e.g., quartz), which are relatively unaffected by CO₂.

Li et al. (2016) conducted a 10-day batch reaction study on rock/water/CO₂ interaction at conditions relevant to GCS (T = 100°C, P = 10 MPa) with rock obtained from the Guantao formation of Bohai Bay Basin, China. The major mineral composition of the rock is quartz balanced with K-feldspar, albite, and clay minerals. They reported K-feldspar and albite dissolution as well as the formation of kaolinite, chlorite, and Fe-bearing minerals of pyrite and hematite postreaction. In addition, HCO₃, Ca, Mg, and K and trace elements of Al, Cr, Mn, Ni, and Zn in water showed a significant increase postreaction. Tutolo et al. (2015) conducted studies in K-feldspar-rich sandstone from the Eau Claire formation (Indiana) under GCS conditions (150°C, and 20 MPa) using a flow-through apparatus. The results indicated dissolution of feldspar and precipitation of secondary aluminum minerals, such as boehmite. They also reported the decrease in bulk permeability and lower porosity in the postexperiment core samples.

Although laboratory studies have been conducted to investigate the interaction between CO₂ and the minerals comprising the storage interval rocks, few studies have been conducted to investigate the interaction between CO₂ and the minerals comprising the rock of the primary seal above the CO₂ storage interval (Miller et al., 2016; Mouzakis et al., 2016). In fact, dissolution of carbonates and silicates in the primary seal is possible when the seal is in contact with CO₂-saturated brine (Miller et al., 2016). The rate of dissolution is dependent on pH decreases in the brine from CO₂ dissolution, reactive area/mass transfer properties, rock/water ratio, and rock composition (Morse and Arvidson, 2002; Luquot and Gouze, 2009; Xiao et al., 2017a). Dissolution of carbonates and silicates can result in changes in permeability, porosity, and reactive surface area. Cation concentration increases can supersaturate pore fluids with carbonates, with subsequent precipitation that can further alter porosity and permeability of the primary seal.

To fill this knowledge gap, the present study considers the effects of GCS on a sample from the primary confining unit above the Lower Tuscaloosa sandstone CO₂ storage interval (marine shale) through laboratory experiments exposing the sample to CO₂-saturated brine for 6 months at in situ conditions. Mineralogical, textural, and geochemical changes as a result of the exposure were determined using X-ray diffraction (XRD), computed tomography (CT) scanning, optical microscopy, scanning electron microscopy (SEM), and brine chemistry analyses. A Lower Tuscaloosa sandstone sample was also exposed to CO₂-saturated brine for 6 months and analyzed by XRD, CT, SEM, and so on to verify property changes of another sample taken from the same interval reported in Soong et al. (2016).

Experimental

Core samples and brine

Two core samples from the Plant Daniel test site, ∼2.54 cm in diameter by 5.08 cm in length, were used in this study. One was a marine shale core from a depth of 2,418 m and the other was a Lower Tuscaloosa sandstone core from a depth of 2,611 m. Marine shale is the primary seal of Lower Tuscaloosa sandstone, which serves as the CO₂ storage reservoir. The approximate major mineral compositions of the Lower Tuscaloosa sandstone sample and the marine shale sample were determined by semiquantitative bulk XRD analysis (Table 1). The chemical composition of the synthetic brine (Table 2) was designed based on the fluid chemistry of Lower Tuscaloosa brine. The initial pH of the synthetic brine was 5.4. Since the Lower Tuscaloosa brine is in contact with rocks in Lower Tuscaloosa formation for a very long time, the equilibrium condition is expected to have been reached. Therefore, no reactions are expected to take place between the synthetic brine and rock samples before CO₂ is added. Before any measurements, the samples spent at least 3 days in a nitrogen-purged desiccator at relative humidity of 8–10%.

Table 1.

X-Ray Diffraction Analysis of Lower Tuscaloosa Sandstone and Marine Shale

	Lower Tuscaloosa sandstone (2,611 m) (%)	Marine shale (2,418 m) (%)
Quartz	77	60
Chlorite	6	9
Kaolinite	3	5
Illite	3	5
K-feldspar	1	2
Calcite	1	5
Plagioclase	7	12
Pyrite	2	2

Analysis was performed by Weatherford Laboratories.

Table 2.

Analysis of Synthetic Brine and Brine Interacted with CO₂/Sandstone for 6 Months Under CO₂ Sequestration Conditions

	Synthetic Lower Tuscaloosa brine, batch # 4 Ave. (ppm)	Brine interacted with marine shale Ave. (ppm)	Synthetic Lower Tuscaloosa, brine batch # 3 Ave. (ppm)	Brine interacted with Lower Tuscaloosa massive sand Ave. (ppm)
Al	1.13	∼	∼	2.5
Ba	9.48	2	9.3	6.9
Ca	11,030	11,392	10,983	10,743
Cr	0.13	1	0.13	0.6
Cu	∼	∼	∼	3.2
Fe	128	205	128	176
K	373	340	374	372
Mg	1,009	1,081	999	1,033
Mn	∼	7	∼	2.7
Na	42,063	43,073	41,122	42,555
Ni	0.59	8	0.6	3.4
Si	∼	35	∼	30.2
Sr	661	565	651	663
Chloride	87,524	87,008	87,412	88,230
Bromide	459	459	458	458
Sulfate	235	389	236	239

Reactors and experimental procedures

A set of 1.8 liter (N4683-T-HC-7500 C276) high-pressure vessels (9.53 cm I.D. by 26.67 cm depth) manufactured by Parr Instrument Company, Erie, PA, was used for this study. An open Teflon cell (7.62 cm I.D. × 10.16 cm length) and a smaller Teflon reaction cell (6.35 cm I.D. × 9.52 cm length) were used inside the pressure vessel. An ISCO 260D syringe pump was used to charge each vessel with CO₂ and to maintain the pressure in the reactor. Temperature was maintained and controlled within ±1°C. A sketch of the reaction system can be found in Fig. 1.

FIG. 1.

Sketch of rock—brine—CO₂ reaction system. Sample to brine volumetric ratio in Teflon cell #1 is 1:5.

The standard procedure for conducting the experiments was as follows:

1. The Teflon container holding the samples (core and two small chips) and brine was loaded into another large Teflon container in the pressure vessel.

2. The gap between the sample-holding container and the large container was filled with brine. Ensuring the CO₂ phase remains saturated with water vapor without substantially altering the salt concentration in the sample-holding container.

3. The pressure vessel was closed and purged with CO₂ three times to remove air.

4. The pressure vessel was charged with CO₂ to 4.08 MPa and heated to 85°C.

5. CO₂ was added to a pressure of 23.8 MPa and held at these conditions for 6 months.

6. On termination of the experiment, the temperature and pressure were slowly returned to ambient, and the samples were removed.

7. The solid samples were rinsed with deionized water multiple times to remove residual NaCl and dried in a desiccator under N₂ flow.

8. Liquid from the sample-bearing container was recovered for analysis. The reacted brine analysis is given in Table 2.

Measurements and analysis methods

Before and after the 6-month test, the solids were analyzed by CT scanning, XRD, SEM, and optical reflected light microscopy. Inductively coupled plasma–optical emission spectroscopy (ICP-OES) was used to determine metal concentrations in the liquid samples, but due to high salt concentrations, sample dilution with distilled/deionized water was required. Solutions were analyzed for the full range of metals quantifiable by ICP-OES, including Al, Si, and Fe. The details of cation and anion analyses, XRD, and SEM-energy dispersive X-ray analyses were reported elsewhere (Soong et al., 2014a).

Porosity and permeability of the samples were measured both before and after the exposure. The porosity was measured using a helium porosimeter (HP-401; TEMCO, Inc., Tulsa, OK) at 0.7 MPa at ambient temperature. For the high-permeability sample (sandstone), a TEMCO Ultraperm 500 “flow-through” permeameter was used. The low–medium permeability core (marine shale) was measured using the pressure transient method (details of the method were reported in Siriwardane et al., 2009). The permeability of the core samples measured at the confining pressure close to the depth of the core was collected with various effective pressures to obtain the permeability profile of the core. The initial porosities were 26.5% for the sandstone core and 8.65% for the marine shale core. The initial permeability was 2,955 mD for the sandstone core and 47 μD for the marine shale core.

All the cores were scanned in the North Star Imaging M-5000 industrial CT scanner. Each of the reconstructed pre- and postimages of the 2.54-cm-diameter Lower Tuscaloosa sandstone core has a voxel resolution between 21 and 22 μm. The marine shale pre-exposure core was scanned at a resolution of 32.7 μm and the postexposure core was scanned at 21.2 μm. All resolutions mentioned are not fine enough to resolve the micro- to nano-scale porosity present in a sandstone or shale rock. The undetectable pores may connect larger voids that facilitate mass transport. Instrument settings, operating conditions, parameters are described elsewhere (Soong et al., 2016). The reconstruction properties of the images were selected to provide the highest fidelity rendition of the scanned sample and to construct a 3D volumetric representation of the object. The data were exported as a series of 16-bit grayscale cross-sectional TIFF images of the core. Image postprocessing was performed using the open-source software ImageJ (Rasband, 2014) and PerGeos^®.

Results and Discussion

Exposure response of storage formation rock (Lower Tuscaloosa sandstone)

Figure 2 shows the permeability of the Lower Tuscaloosa sandstone measured as a function of effective pressure. The postexposure permeability of 2,400 mD was about 17% lower than the measured prereaction permeability of 2,900 mD in the fresh core at the effective pressure of 35.7 MPa. This result suggests that the CO₂/brine exposure alters the pore characteristics by restricting flow and resulting in permeability reductions, and this result is consistent with permeability reduction of another Lower Tuscaloosa core after a 6-month exposure to CO₂-saturated brine, as reported in Soong et al. (2016). The initial porosity of the Lower Tuscaloosa sandstone was 26.5%, and the postreaction porosity of the rinsed and dried Lower Tuscaloosa sample was 24.8%, which shows alteration of pore characteristics.

FIG. 2.

Permeability of Lower Tuscaloosa sandstone as a function of effective pressure.

Figure 3 shows CT analysis results of the sandstone before and after the 6-month exposure to the CO₂/brine environment. The pre-exposure scans (Fig. 3A) of the sandstone core were scanned at a 21.2 μm resolution, and the postexposure core was scanned with a resolution of 21.8 μm (Fig. 3B). The pre- and postexposure scans are similar in location for each image pair. Both pore space (dark zones) and higher attenuating minerals within the core (bright zones) can be observed in Fig. 3A and B. Figure 3C (pre-exposure) and D (postexposure) is zoomed-in images from the red square highlighted regions presented in the lower left portion of Fig. 3A and B. The yellow circled region in Fig. 3C shows a gray void structure present in the pre-exposure scan that is no longer visible in the postexposure scan in Fig. 3D. Moreover, higher attenuating minerals (bright spots in Fig. 3C) appear to have been altered after 6 months of exposure to CO₂/brine (Fig. 3D). This level of detail was only discernable in the zoomed-in images. The void was located ∼2 mm from the core edge and may indicate that mineral precipitation closed the void after the core was exposed to CO₂-saturated brine.

FIG. 3.

Lower Tuscaloosa (A) pre-exposure slice and (B) postexposure slice, each with the identical square ROI highlighted in red. (C) Cropped pre-exposure and (D) cropped postexposure zoomed-in images of (A, B) located ∼2 mm from the edge of core. The yellow circle highlights the pore space change from exposure. ROI, region of interest.

To ensure the observed precipitation was not a result of slight changes in resolution or an error in 2D slice matching from pre- to postexposure, the cropped 2D stacks that included the slices presented in Fig. 3C and D were examined in 3D using PerGeos (Fig. 4A, B). The red 3D pore space disappears from pre-exposure (Fig. 4A) to postexposure (Fig. 4B) further confirming that the pore space closed after 6 months of brine/CO₂ exposure. The resolutions of these two scans are not fine enough to resolve the microporosity present in a sandstone core. Microporosity is known to connect larger scale voids and contribute to flow. However, the observed precipitation occurred within a pore ∼200 μm in diameter, well above the resolution detectable within the scans reviewed. Undetectable microporosity may have influenced the observed reductions in porosity and permeability. However, the pores detectable at this resolution would contribute more to mass transport and, therefore, would contribute more to a reduction in mass transport if closed.

FIG. 4.

Lower Tuscaloosa sandstone (A) pre-exposure, (B) postexposure red pores (in red) and purple dense materials. The yellow circle highlights the region where precipitation occurred from (A) to (B). Cropped region located ∼2 mm from the edge of core.

Based on this evidence, it is theorized that the observed 17% permeability decrease and 6.4% porosity decrease result from mineral precipitation occurring in the sample pores with a net effect of restricting flow. Mineral precipitation may also have occurred within pore throats in the porous medium, thus contributing to net permeability decreases, as illustrated in these CT scans. Studies of acidizing hydrocarbon reservoirs for enhanced oil recovery such as Smith et al. (2013) have identified similar permeability decreases as well as a similar study conducted with CO₂/brine/Mount Simon sandstone (Soong et al., 2014b).

Comparing the analyses of the postreacted brine composition with that of fresh brine (Table 2) showed that after the 6-month exposure period, the concentrations of Al, Fe, Mg, Na, and Si had increased. This change in solution chemistry suggested that CO₂-acidizied brine dissolved some of the nonquartz minerals.

Figure 5A–F shows selected secondary electron SEM images of the Lower Tuscaloosa sandstone sample. The specific site selected is a feldspar grain (identified based on energy dispersive X-ray spectroscopy), which is expected to be more susceptible to chemical change during exposure. The fresh sample is shown in Fig. 5A and D. Figure 5B and E shows the sample images after 6 months of CO₂/brine exposure. Figure 5C and F further illustrate the sample after additional rinsing with deionized water to remove the NaCl deposited on the surface. The NaCl surface deposit was probably formed due to slow diffusion of residual brine out of pore structure to the surface during drying, which is commonly referred to as a “salting out” effect. In the top center of the fresh (Fig. 5D) and exposed (Fig. 5F) circled images, possible evidence of mineral dissolution was observed. Possible mineral dissolution and precipitation were also observed after the 6-month CO₂/brine exposure (by comparing middle and bottom circled areas in Fig. 5D [fresh] to those circled areas in Fig. 5F [exposed]). The precipitation of secondary minerals (possibly amorphous SiO₂ [am] and kaolinite, as reported in Zhang et al., in press) is possibly the primary contributor to a decrease in bulk permeability. When CO₂ is injected into a saline formation, the brine in the reservoir becomes more acidic. The increase in acidity can lead to some dissolution of minor minerals in the sandstone such as feldspar and chlorite that are stable in the original conditions (Fu et al., 2009; Zhang et al., 2016). The increase in cation concentration produced by mineral dissolution can supersaturate the pore solution with respect to other minerals resulting in precipitation of secondary minerals such as SiO₂ (am) and kaolinite. Zheng et al. (2015) conducted an experimental investigation of quartz–feldspar–detrital sandstone properties under CO₂/NaCl solution/rock interaction. Their results show that the addition of CO₂ enhanced the partial mineral dissolution and the compressibility for sandstone, and a permeability reduction is observed. The injected CO₂ could be dissolved in pore fluid and precipitated with the calcium, magnesium, and ferrous ions in pore solution from the sandstone skeleton and thus resulted in the permeability reduction (Zheng et al., 2015).

FIG. 5.

SEM images of the Lower Tuscaloosa sandstone fresh (A, D) versus exposed (B, E) to CO₂/brine for 6 months. (C, F) Images of (B, C) rinsed with deionized water to remove residual NaCl. SEM, scanning electron microscopy. Circles show mineral dissolution/precipitation.

Exposure response of primary sealing formation rock (marine shale)

Characterization of the marine shale sample was done using the same procedure described for the Lower Tuscaloosa sandstone sample. The permeability that was measured at in situ conditions corresponded to a depth of ∼2,418 m (effective pressure of 33.5 MPa). The Autolab unit (NER, Inc., White River Junction, VT) with argon was used for more precise measurements of the marine shale permeability for both fresh and exposed samples due to the low permeability of shale. Figure 6 shows permeability as a function of different effective pressures. Here, a significant increase in permeability (170 μD of the core when exposed for 6 months vs. 30 μD of the fresh core at the effective pressure of 33.5 MPa) was observed. Mineral dissolution could increase the average size of pathways, which could lead to increased permeability. The porosity of the fresh sample was 8.65%, and the porosity measured after the 6-month CO₂/brine exposure was only slightly lower at 8.56%.

FIG. 6.

Permeability of marine shale as a function of effective pressure.

CT analysis results of the marine shale core before and after the 6-month exposure to the CO₂/brine environment are shown in Fig. 7. Figure 7 displays approximately the same slice from pre- (7A) and postexposure (7B). One main fracture spanning the diameter of the core in Fig. 7A appeared to have decreased in size after the 6-month exposure in CO₂/brine environment (Fig. 7B). On the contrary, two adjacent fractures were widened in the postexposure image (Fig. 7B). Furthermore, dense materials (bright spots) were formed within the left-most fracture following exposure to CO₂-saturated brine (Fig. 7B circled areas).

FIG. 7.

(A) Marine shale pre-exposure slice and (B) marine shale postexposure slice. Aperture changes of fractures are indicated and precipitated dense materials are marked in circles.

Figure 8 contains resliced images in the YZ-plane for pre- (Fig. 8A) and postexposure (Fig. 8B) samples. Two regions of interest are identified: a red rectangle and an orange oval. The red rectangle focuses on a fracture that became more connected after 6-month exposure. A widened and more connected fracture detectable at the scan resolution would contribute to mass fluid transport. In addition, the orange oval highlights a fracture where some dense mineral growth is apparent after the 6-month exposure (Fig. 8A, B). The combination of both may result in the increase of permeability of the sample after the 6-month exposure. Precipitation caused the main fracture to close in diameter, while two large fractures widened over the 6-month exposure. The additional stresses within the precipitated fracture may have caused the adjacent fractures to expand.

FIG. 8.

Marine shale resliced images in the YZ-plane. (A) Pre-exposure and (B) postexposure. Two ROIs are highlighted, one rectangle showing dissolution and one oval showing precipitation.

Both fracture connection and mineral precipitation in the fracture after the CO₂/brine exposure are indicated by CT images, and permeability results show that the combination of both causes an increase of permeability of the sample after the 6-month exposure (Fig. 6). To better interpret those results, the nature of these fractures, their relevance to in situ behavior of formation rock, and implications for GCS system performance need to be considered. The evolution of fracture features in the marine shale cores may be related to a geochemical alteration of the sample along a bedding plane due to a preferential geochemical alteration related to compositional differences (e.g., presence of more easily weathered mineral(s) along specific planes). It may also be due to delamination of the sample along the bedding plane after the sample was removed from the in situ confining stress. This decompression and mechanical delamination would be augmented by sample swelling with exposure to brine.

A modified experimental approach would need to be pursued to resolve the extent to which each of these factors contributed to fracture evolution in the marine shale core sample under experimental conditions. Also, because the direction of observed shale fracturing/delamination is parallel to the bedding plane, and nominally orthogonal to the path of leakage through the seal, the change in effective permeability in the X-Y plane, associated with shale delamination that could potentially occur in the marine shale, would not significantly contribute to the change in effective permeability of the seal in the Z direction. A large permeability increase in the Z direction due to delamination cracks would depend somewhat on the degree of interconnection of the cracks. As such, the overall impact to containment effectiveness of the marine shale seal may be negligible in the Lower Tuscaloosa GCS system. Further experimental simulation and modeling are warranted to further explore this idea.

The postreaction brine was analyzed following contact with the marine shale and CO₂ (Table 2). Comparing the reacted brine with the fresh brine, the following trends were observed. The concentration of Ca, Fe, Mg, Na, and Si increased after the 6-month exposure period. Except for K, the increases in reacted brine concentration for the marine shale were more than those observed for the Lower Tuscaloosa sandstone core exposure experiment. This observation was not surprising since the shale contains a higher content of minerals (i.e., chlorite, plagioclase, etc.) that are more susceptible to CO₂/brine attack. Only a slight increase in the concentration of Ca was observed even though 5% of the marine shale is calcite. Preferred dissolution of silicates such as chlorite and plagioclase over calcite is possibly the reason for slow increase of Ca in solution.

Chemical reactions between reservoir rock and fluids involve two types of minerals: the first type are the relatively fast-reacting minerals, including Ca, Mg, or Fe-rich frame work minerals such as feldspars, clays, micas, and the second type includes minerals such as Fe oxides and quartz, which interact slowly with CO₂ (Zheng et al., 2015). The addition of CO₂ induced the dissolution of feldspathic minerals, which increased the concentration of metal ions such as K and Na and may decrease the concentration of calcium ions precipitated by carbonic acid. Dissolution of feldspar and other silicates can also lead to supersaturation of SiO₂ and precipitation of SiO₂ (Miller et al., 2016). In the NaCl/CO₂/rock environment, the feldspar reacts with water and generates mica, kaolinite, and so on, while the dissolved CO₂ can result in formation of carbonate minerals such as dolomite, calcite, and siderite (Zheng et al., 2015).

Studies of CO₂ injection into depleted oil and gas reservoirs have shown that the initial chemical equilibrium between the saline formation fluid and reservoir rock may be disturbed resulting in new chemical reactions (Fischer et al., 2010). Such interactions can result in changes in the physical and chemical properties of the reservoir system (Wigand et al., 2008; Fischer et al., 2010; Soong et al., 2014b; Zhang et al., 2015). The marine shale located above the potential storage formation (Lower Tuscaloosa sandstone) is considered as the primary seal. The observed increase in permeability after the CO₂/brine exposure from this study is likely due to the heterogeneity of the core sample and the experimental setup, which resulted in core sample cracking during the exposure period. Those results may not represent actual in situ behavior.

It also can be seen from Table 1 that the major mineral composition is not significantly different between sandstone and marine shale. In addition, marine shale does have more reactive feldspar and calcite than sandstone. Dissolution of reactive K-feldspar and calcite could have created additional connective pathways thus contributing to an increase in permeability after CO₂/brine exposure. This might impact the integrity of the long-term storage of CO₂. However, the very low permeability of the shale relative to the sandstone and capillary pressure barrier effect would suppress CO₂ to enter the pore space of the caprock, which limits bulk rock dissolution and reduces the risk to seal rock formation (Tian et al., 2014; Xiao et al., 2017b). Furthermore, secondary cap rock formation, Selma chalk, above marine shale was found not to be reactive with CO₂/brine in our previous study (Soong et al., 2016).

In summary, while potential geochemical alteration of sealing intervals in GCS scenarios could significantly affect local seal behavior, the presence of a relatively thick primary caprock, and secondary low permeability sealing intervals will likely minimize upward CO₂ and brine migration. To be specific, even though the primary confining layer to the Lower Tuscaloosa formation (marine shale) shows some evidence of reactivity and fracturing in experimental studies, the presence of Selma chalk as a secondary containment unit gives confidence that effective containment will be ensured.

For the Lower Tuscaloosa sandstone sample, the combined information from XRD, SEM, CT scanning, and brine analysis indicates dissolution of minerals, most likely feldspar, kaolinite, and chlorite in flow pathways along with precipitation of minerals in flow pathways of the Lower Tuscaloosa sandstone sample. For the sandstone sample, mineral dissolution, most likely feldspar, and secondary mineral precipitation in flow pathways had the net effect on decreasing permeability. These results suggest that the injection of CO₂ into a deep saline aquifer can affect porosity as a result of both mineral dissolution and precipitation in the formation. The change of porosity due to mineral reactions can adversely influence permeability. This permeability change depends not only on porosity but also on the details of pore space geometry and the distribution of precipitates within the pore space.

In future work, there will be two focuses: (1) to determine if the permeability increase of the marine shale sample results in an impact to containment effectiveness of the marine shale seal in the Lower Tuscaloosa GCS system and (2) to use multiple marine shale samples from different depths to verify the results reported in this study. To address the first, a reservoir-scale model will be developed to investigate how thick the zone will be that has a permeability increase in the marine shale. To address the second, multiple marine shale samples will be obtained from the Plant Daniel CO₂ storage site to conduct duplicate experiments.

Conclusions

In this study, chemical interactions between a primary seal sample (marine shale core sample) and CO₂-saturated brine were investigated under in situ GCS conditions, as well as a CO₂ storage reservoir sample (Lower Tuscaloosa sandstone core sample). The Lower Tuscaloosa CO₂ storage reservoir was immediately below marine shale primary seal. The interaction between core samples and CO₂-saturated brine was investigated using a 6-month duration static system. CT, XRD, SEM, petrography, brine composition, porosity, and permeability analyses were performed on the samples before and after exposure to CO₂-saturated brine under GCS conditions. For the marine shale, the significant change in permeability is due to the combined effects of reactive mineral composition, sample heterogeneity, and sample delamination, although the permeability of the altered shale was observed to be 1,000 times less than the permeability of sandstone. Moreover, after the Marine shale sample was retrieved from deep depth below the surface, the drop in confining pressure might lead to formation of secondary fractures due to decompression and subsequent delamination. The formation of these secondary fractures could lead to overprediction of the permeability increase. For the Lower Tuscaloosa sandstone, there was a 6.4% decrease in porosity and a 17% decrease in permeability found in the sandstone core after it was exposed to CO₂-saturated brine for 6 months. The decrease in porosity and permeability was due to feldspar dissolution, migration, and secondary mineral precipitation altering the sandstone pore/crack structure. The low content of active mineral phases in the Lower Tuscaloosa sandstone sample and the high initial porosity of the sample lead to a smaller permeability change for the sample after the exposure. Our previous study suggests that even with an increase in marine shale permeability, the existence of secondary and tertiary seals is likely to provide a stable caprock seal in the presence of CO₂.

Footnotes

Acknowledgments

We thank Bryan Tennant for providing the CT images of the studied rock samples. We are also indebted to Dr. T.T. Phan for preparing the brine and to AECOM for conducting the brine sample analysis.

Disclaimer

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

Author Disclosure Statement

No competing financial interests exist.

References

Bachu

, and Adams

J.J.

(2003). Sequestration of CO2 in geological media in response to climate change: Capacity of deep saline aquifers to sequester CO2 in solution. Energy Convers. Manag., 44, 3151.

Carbon Storage Atlas. (2015). Carbon Storage Atlas-Fifth Edition (Atlas V). Available at: www.netl.doe.gov/research/coal/carbon-storage/atlassv (accessed April 2017 ).

Delshad

, Kong

, Tavakoli

, Hosseini

S.A.

, and Wheeler

M.F.

(2013). Modeling and simulation of carbon sequestration at Cranfield incorporating new physical model. Int. J. Greenh. Gas Con., 18, 463.

Fischer

, Liebscher

, and Wandrey

(2010). CO2-brine-rock interaction—First results of long-term exposure experiments at in situ P-T conditions of the Ketzin CO2 reservoir. Chem. Erde, 70, 155.

, Lu

, Konishi

, Dilmore

, Xu

, Seyfried

W.E.

, and Zhu

(2009). Coupled alkali-feldspar dissolution and secondary mineral precipitation in batch systems: New experiments at 200°C and 300 bars. Chem. Geol., 258, 125.

Gouze

, and Luquot

(2011). X-ray microtomography characterization of porosity, permeability and reactive surface changes during dissolution. J. Contam. Hydrol., 120, 45.

Griffith

C.A.

, Dzombak

D.A.

, and Lowry

G.V.

(2011). Physical and chemical characteristics of potential seal strata in regions considered for demonstrating geological saline CO2 sequestration. Environ. Earth Sci., 64, 925.

Haljasmaa

I.V.

, McLendon

T.R.

, Jikich

S.A.

, Goodman

, Siriwardane

, Soong

, McIntyre

D.L.

, Bromha

G.S.

, Irdi

G.A.

, and Dobroskok

(2011). North Dakota lignite and Pittsburgh bituminous coal: A comparative analysis in application to CO2 sequestration. Int. J. Oil Gas Coal Technol., 4, 264.

Jia

, McPherson

B.J.

, Pan

, Xiao

, and Bromhal

(2016). Probabilistic analysis of CO₂ storage mechanisms in a CO2-EOR field using polynomial chaos expansion. Int. J. Greenh. Gas Con., 51, 218.

10.

, Pang

, Yang

, and Jin

(2016). Geochemical responses of a saline aquifer to CO2 injection: Experimental study on Guantao formation of Bohai Bay Basin, East China. Greenh. Gas Sci. Technol., 6, 1125.

11.

, Kharaka

Y.K.

, Thordsen

J.J.

, Horita

, Karamalidis

, Griffith

C.A.

, Hakala

J.A.

, Ambats

, Cole

D.R.

, Phelps

T.J.

, Manning

M.A.

, Cook

P.J.

, and Hovorka

S.D.

(2012). CO2-rock-brine interactions in Lower Tuscaloosa formation at Cranfield CO2 sequestration site, Mississippi, U.S.A. Chem. Geol., 291, 269.

12.

Luquot

, and Gouze

(2009). Experimental determination of porosity and permeability changes induced by injection of CO2 into carbonate rocks. Chem. Geol., 265, 148.

13.

Middleton

R.S.

(2013). A new optimization approach to energy network modeling: Anthropogenic CO2 capture coupled with enhanced oil recovery. Int. J. Energy Res., 37, 1794.

14.

Miller

Q.R.

, Wang

, Kaszuba

J.P.

, Mouzakis

K.M.

, Navarre-Sitchler

A.K.

, Alvarado

, McCray

J.E.

, Rother

, Bañuelos

J.L.

, and Heath

J.E.

(2016). Experimental study of porosity changes in shale caprocks exposed to carbon dioxide-saturated brine II: Insights from aqueous geochemistry. Environ. Eng. Sci., 33, 736.

15.

Morse

J.W.

, and Arvidson

R.S.

(2002). The dissolution kinetics of major sedimentary carbonate minerals. Earth Sci. Rev., 58, 51.

16.

Mouzakis

K.M.

, Navarre-Sitchler

A.K.

, Rother

, Bañuelos

J.L.

, Wang

, Kaszuba

J.P.

, Heath

J.E.

, Miller

Q.R.

, Alvarado

, and McCray

J.E.

(2016). Experimental study of porosity changes in shale caprocks exposed to CO₂-saturated brines I: Evolution of mineralogy, pore connectivity, pore size distribution, and surface area. Environ. Eng. Sci., 33, 725.

17.

Pachauri

R.K.

, and Reisinger

(Eds) (2007). IPCC fourth assessment report on climate change; synthesis report. In Prepared by Working Group I, II and III of the Intergovernmental Panel on Climate Change. Geneva, Switzerland, p. 104..

18.

Pan

, McPherson

, Esser

, Xiao

, Appold

, Jia

, and Moodie

(2016). Forecasting evolution of formation water chemistry and long-term mineral alteration for GCS in a typical clastic reservoir of the Southwestern United States. Int. J. Greenh. Gas Con., 54, 524.

19.

Ranjith

P.G.

, Perera

M.S.A.

, and Khan

E.A.

(2013). Study of safe CO2 storage capacity in saline aquifers: A numerical study. Int. J. Energy Res., 37, 189.

20.

Rasband

W.S.

(2014). ImageJ. Bethesda, MD: U.S. National Institutes of Health, 1997. Available at: http://imagej.nih.gov/ij/ (accessed April 2017 ).

21.

Rodosta

, Litynski

, Plasynski

, Spangler

, Finley

, Steadman

, Ball

, Hill

, McPherson

, Burton

, and Vikara

(2011). U.S. Department of Energy's Regional Carbon Sequestration Partnership Initiative: Update on validation and development phases. Energy Procedia, 4, 3457.

22.

SECARB. (2016). Saline Reservoir Field Test: Mississippi Test Site Fact Sheet. Available at: www.secarbon.org/files/mississippi-test-site.pdf (accessed January 2017 ).

23.

Siriwardane

, Haljasmaa

, McLendon

T.R.

, Irdi

, Soong

, and Bromhal

(2009). Influence of CO2 on coal permeability determined by pressure transient methods. Int. J. Coal Geol., 77, 109.

24.

Smith

M.M.

, Sholokhova

, Hao

, and Carroll

S.A.

(2013). Evaporite caprock integrity: An experimental study of reactive mineralogy and pore-scale heterogeneity during brine-CO2 exposure. Environ. Sci. Technol., 47, 262.

25.

Soong

, Hedges

S.W.

, Howard

B.H.

, Dilmore

R.M.

, and Allen

D.E.

(2014a). Effect of contaminants from flue gas on CO2 sequestration in saline formation. Int. J. Energy Res., 38, 1224.

26.

Soong

, Howard

B.H.

, Dilmore

R.M.

, Haljasmaa

, Crandall

, Zhang

, Irdi

, Romanov

, Lin

, and Mclendon

T.R.

(2016). CO2/brine/rock interactions in Lower Tuscaloosa formation. Greenh. Gases Sci. Technol., 6, 824.

27.

Soong

, Howard

B.H.

, Hedges

S.W.

, Haljasmaa

, Warzinski

R.P.

, Irdi

, and McLendon

T.R.

(2014b). CO2 sequestration in saline formation. Aerosol Air Qual. Res., 14, 522.

28.

Tian

, Pan

, Xu

, McPherson

, Yue

, and Mandalaparty

(2014). Impacts of hydrological heterogeneities on caprock mineral alteration and containment of CO₂ in geological storage sites. Int. J. Greenh. Gas Con., 24, 30.

29.

Tutolo

B.M.

, Luhmann

A.J.

, Kong

, Saar

M.O.

, and Seyfried

W.E.

(2015). CO2 sequestration in feldspar-rich sandstone: Coupled evolution of fluid chemistry, mineral reaction rates and hydrogeochemical properties. Geochim. Cosmochim. Acta, 160, 132.

30.

Wigand

, Carey

J.W.

, Schutt

, Spangenberg

, and Erzinger

(2008). Geochemical effects of CO2 sequestration in sandstone under simulated in situ conditions of deep saline aquifers. Appl. Geochem., 28, 2735.

31.

Xiao

, Dai

, Viswanathan

, Hakala

, Cather

, Jia

, Zhang

, and McPherson

(2017a). Arsenic mobilization in shallow aquifers due to CO₂ and brine intrusion from storage reservoirs. Sci. Rep., 7, 2763.

32.

Xiao

, McPherson

, Bordelon

, Viswanathan

, Dai

, Tian

, Esser

, Jia

, and Carey

(2017b). Quantification of CO₂-cement-rock interactions at the well-caprock-reservoir interface and implications for geological CO₂ storage. Int. J. Greenh. Gas Con., 63, 126.

33.

Zhang

, Soong

, and Dilmore

(2016). Investigation on porosity and permeability change of Mount Simon sandstone (Knox County, IN, USA) under geological CO2 sequestration conditions: A numerical simulation approach. Greenh. Gas Sci. Technol., 5, 1.

34.

Zhang

, Soong

, and Dilmore

(2017). Numerical investigation of Lower Tuscaloosa sandstone and Selma Chalk caprock under geological CO₂ sequestration conditions: Mineral precipitation and permeability evolution. Greenh. Gas Sci. Technol. [Epub ahead of print]; DOI: 10.1002/ghg.1703.

35.

Zhang

, Soong

, Dilmore

R.M.

, and Lopano

(2015). Numerical simulation of porosity and permeability evolution of Mount Simon sandstone under geological carbon sequestration. Chem. Geol., 403, 1.

36.

Zheng

, Feng

, and Pan

(2015). Experimental investigation of sandstone properties under CO2-NaCl solution-rock interactions. Int. J. Greenh. Gas Con., 37, 451.