Imaging Wellbore Cement Degradation by Carbon Dioxide under

Jul 23, 2012 - Characterization of reactive flow-induced evolution of carbonate rocks using digital core analysis- part 1: Assessment of pore-scale mi...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/est

Imaging Wellbore Cement Degradation by Carbon Dioxide under Geologic Sequestration Conditions Using X‑ray Computed Microtomography Hun Bok Jung,† Danielle Jansik,† and Wooyong Um*,† †

Pacific Northwest National Laboratory, Richland, Washington 99354, United States S Supporting Information *

ABSTRACT: X-ray microtomography (XMT), a nondestructive three-dimensional imaging technique, was applied to demonstrate its capability to visualize the mineralogical alteration and microstructure changes in hydrated Portland cement exposed to carbon dioxide under geologic sequestration conditions. Steel coupons and basalt fragments were added to the cement paste in order to simulate cement-steel and cement-rock interfaces. XMT image analysis showed the changes of material density and porosity in the degradation front (density: 1.98 g/cm3, porosity: 40%) and the carbonated zone (density: 2.27 g/cm3, porosity: 23%) after reaction with CO2saturated water for 5 months compared to unaltered cement (density: 2.15 g/cm3, porosity: 30%). Three-dimensional XMT imaging was capable of displaying spatially heterogeneous alteration in cement pores, calcium carbonate precipitation in cement cracks, and preferential cement alteration along the cement-steel and cement-rock interfaces. This result also indicates that the interface between cement and host rock or steel casing is likely more vulnerable to a CO2 attack than the cement matrix in a wellbore environment. It is shown here that XMT imaging can potentially provide a new insight into the physical and chemical degradation of wellbore cement by CO2 leakage.

■

INTRODUCTION Portland cement is a common component consisting of a sealing material for wellbores for geological carbon sequestration to prevent vertical fluid migration and provide mechanical support.1 Hydrated products formed by mixing Portland cement with water include a semiamorphous gel-like calciumsilicate-hydrate (C−S−H) and a crystalline portlandite [Ca(OH)2(s)].2,3 The leakage of CO2 stored in a deep geological carbon sequestration site may potentially occur at the interface between the steel casing and cement, cement and host rock, or through the cement pore space and cracks.4 Hydrated cement is a nonhomogeneous material with macroscopic properties that depend on its local microstructural characteristics, which highlights the importance of the understanding of the threedimensional (3-D) microgeometry. A number of analytical methods, such as mercury intrusion porosimetry (MIP), nitrogen gas adsorption, and electron microscopy have been applied to characterize microstructure change of cement-based materials5−8 and to evaluate the degradation of cement by CO2 attack under geologic sequestration conditions.9−14 However, most of these methods assume the pore geometry in the interpretation of results and provide microstructure distribution information only in two-dimensions at the specimen surface.15 In addition, these methods require destructive sample preparations such as cutting, polishing, drying, and carbon coating that could cause the change of the original properties. Preserving microstructures is essential when the physical properties of the materials are dependent on the dynamic © 2012 American Chemical Society

evolution of the chemical composition at the micrometer-scale. For example, Fabbri et al. (2009)16 used a triaxial apparatus to measure the hydromechanical properties of Portland cement at in situ pressure conditions. X-ray microtomography (XMT) is a noninvasive and nondestructive 3-D imaging technique that uses a series of radiographic images to reconstruct a map of an object’s X-ray absorption, which can provide information on spatial distribution of microstructure of a material in three dimensions with minimal sample preparation. Because X-ray absorption is a function of the elemental composition and density of the object, X-ray imaging can be related to the microstructure and mineralogical composition of the material. In greyscale XMT images, brightness is proportional to X-ray absorption, when a material is imaged near or above 100 KeV, with dark and light regions corresponding to low- and highdensity phases, respectively. Despite numerous applications of XMT for geological materials17−20 and cement-based materials,21−26 there is little literature on the application of the XMT technique to investigate the degradation of wellbore cement material by CO2 under carbon sequestration condition. Recently, Kutchko et al. (2009)1 showed X-ray computer Special Issue: Carbon Sequestration Received: Revised: Accepted: Published: 283

April 3, 2012 July 18, 2012 July 23, 2012 July 23, 2012 dx.doi.org/10.1021/es3012707 | Environ. Sci. Technol. 2013, 47, 283−289

Environmental Science & Technology

Article

Table 1. Summary of Cement Columns Reacted with Wet Supercritical CO2 or CO2-Saturated Water for 2 to 5 Monthsa vessel no.

sample ID

1

NRAP-4 NRAP-8 NRAP-10 NRAP-16 NRAP-24 NRAP-25 NRAP-32 NRAP-33

2

3 4

cement type type type type type type type type type

II-V II-V II-V II-V II-V I-II I-II I-II

sample size 14 14 14 14 14 13 13 13

mm mm mm mm mm mm mm mm

D D D D D D D D

and and and and and and and and

90 90 90 90 90 25 25 25

water-to-cement ratio mm mm mm mm mm mm mm mm

L L L L L L L L

0.33 0.33 0.33 0.33 0.33 0.38 0.38 0.38

reaction period 5 months 5 months

2 months 3.5 months 3.5 months

degradation avg. depth 3.53 3.11 4.23 2.98 3.17 1.44 1.52

mm mm mm mm mm mm mm

comments cement-steel cement-steel layered cracks cement only cement-basalt cement only cement only cement only

a

For NRAP-4, -8, -10, -16, and -24, the top half of each sample was exposed to wet supercritical CO2, while the bottom half of the sample was submerged into CO2-saturated synthetic groundwater. For NRAP-25 and -33, the whole cement column was submerged into the water, while the whole cement column of NRAP-32 was exposed to only wet supercritical CO2. The average depth of cement degradation by CO2-saturated water under carbon sequestration condition (50 °C and 10 MPa) is estimated from the middle vertical slice in XMT images.

atm) before the CO2 reaction. Immediately after curing (without drying process), the hardened cement columns were placed in Parr pressure vessels (300 mL size with 64 mm diameter × 102 mm depth) containing 130 mL of synthetic groundwater to simulate the groundwater composition of Wallula Basalt Carbon Dioxide Sequestration Pilot Site (Table S1). For the cement columns with 14 mm diameter and 90 mm length, the bottom half of each sample was submerged in synthetic groundwater saturated with CO2 at a temperature of 50 °C and pressure of 10 MPa, while the top half was exposed to wet supercritical CO2. The cement columns with 13 mm diameter and 25 mm length were either completely submerged in synthetic groundwater (130 mL) saturated with CO2 or exposed to wet supercritical CO2 at 50 °C and 10 MPa. After the reaction for 2−5 months, cement samples were depressurized slowly for 24 h in order to minimize microstructural modifications upon decompression.16 Cement samples were characterized by XMT, as well as scanning electron microscopy-energy dispersive spectroscopy (SEMEDS) and X-ray diffraction (XRD) for duplicate samples (see the Supporting Information). Porosity and air permeability of a duplicate cement column (type I-II; 25 mm diameter × 38 mm length; water-to-cement ratio = 0.38) before and after the 2month reaction with CO2-saturated water were also measured by Core Laboratories using a CMS-300 Automated Permeameter (see the Supporting Information). X-ray Microtomography Measurement. The XMT analysis was conducted using an NSI X-View Digital X-ray Imaging and Microfocus Computed Tomography (XMCT) system manufactured by North Star Imaging, Inc. (Rogers, Minnesota) (Figure S1). X-rays are generated by a microfocus X-ray source (Comet Feinfocus model 160.48 160 kV), and images are collected by a PaxScan 2520 V flat panel digital Xray detector (pixel pitch of 127 μm and a total active imaging area of 8 × 10 in.). Data acquisition and image reconstruction were achieved using commercially available X-View IW or efXct image reconstruction software. Columns were imaged with one image per 0.5 degree of rotation at 96 KeV and 527 μA for cement columns with 13 mm diameter and 25 mm length (image resolution = 15 μm), while a rotation at 98 KeV and 536 μA was used for cement columns with 14 mm diameter and 90 mm length (image resolution = 29 μm). Image segmentation and advanced visualization were performed using AVIZO Fire 7.0 image processing software. Initial density calibration curves were generated by using the point measurement tool to quantify the image intensity of density-known

tomography images indicating a density change in altered region of pozzolan-cement blend after exposure to CO2saturated brine at 50 °C and 15 MPa, while Laudet et al. (2011)27 provided X-ray tomography images exhibiting a progression of the carbonation front in oil well cement exposed to CO2-saturated water at 90 or 140 °C and 8 MPa. The objective of this study was to demonstrate and validate the capability of XMT to visualize and quantify the chemical degradation and physical pore structure changes of hydrated Portland cement exposed to carbon dioxide under geologic sequestration conditions. An industrial microfocus X-ray computed microtomography with 15−29 μm resolution was utilized to image the alteration of cement columns through cement pores and along the cement-steel and cement-rock interfaces after reaction with wet supercritical CO2 and CO2saturated water at 50 °C and 10 MPa over a time scale of 2−5 months.

■

MATERIALS AND METHODS

Portland Cement Preparation and Carbonation at High Pressure−Temperature Conditions. Cement slurry was prepared by mixing Portland cement with water at a waterto-cement ratio of 0.33 (Ash Grove, type II−V; comparable to class G cement) for cylindrical cement columns with 14 mm diameter and 90 mm length or at a water-to-cement ratio of 0.38 (Lafarge North America, type I-II; comparable to class A cement) for cylindrical cement columns with 13 mm diameter and 25 mm length (Table 1). See the Supporting Information (SI) for chemical composition of cement. Cement samples with 14 mm diameter and 90 mm length (water-to-cement ratio = 0.33) were cast in the form of cylinders by pouring the slurry into a 15-mL centrifuge tube containing steel coupons (NRAP4 and -8; C = 0.28%, Mn = 1.16%, Fe = 98.5% by weight) or basalt fragments (NRAP-24) from the Wallula CO2 sequestration pilot site in Washington State. Cement columns including steel or basalt were prepared to test the capability of XMT to evaluate the potential alteration of cement along the interfaces with steel casing or host rock in the wellbore environment. The NRAP-16 cement column was prepared without additives, while the NRAP-10 cement column was prepared by assembling several hydrated cement fragments using cement slurry (water-to-cement ratio = 0.33) to form layered cracks. The NRAP-25, -32, and -33 cement samples with 13 mm diameter and 25 mm length (water-to-cement ratio = 0.38) were placed into a plastic mold without additives. All cement specimens were sealed to prevent the evaporation of moisture and cured for 28 days under ambient condition (20 °C and 1 284

dx.doi.org/10.1021/es3012707 | Environ. Sci. Technol. 2013, 47, 283−289

Environmental Science & Technology

Article

Figure 1. A: A schematic diagram adopted from Kutchko et al. (2007) of the degradation processes of a hydrated Portland cement by CO2-saturated water across the horizontal cross section. B: SEM image and EDS data (atom %) for NRAP-15 (duplicate for NRAP-16) reacted with CO2-saturated water for 5 months (spectrum 1: 16% Ca and 5% C, spectrum 2: 12% Ca and 15% C, spectrum 3: 6% Ca and 20% C, spectrum 4: 5% Ca and 26% C). C-E: Greyscale XMT images of vertical and horizontal slices after reaction of CO2-saturated water with NRAP-25 (13 mm diameter; w/c = 0.38) for 2 months, NRAP-33 (13 mm diameter; w/c = 0.38) for 3.5 months, and NRAP-16 (14 mm diameter; w/c = 0.33) for 5 months, respectively. Due to slight differences in reconstruction settings for cement columns reacted for 2−3.5 months (C and D) and 5 months (E), the comparison between these images describes differences in the density of the sample only qualitatively. Dark spots correspond to air-filled voids, and the regions with lighter and darker gray color indicate higher-density carbonated zone and lower-density Ca(OH)2 depleted degradation front, respectively. Ring artifacts observed in the XMT images are generated by differences in sample centering during the XMT scan.

while light regions correspond to high-density phases. XMT imaging of a cement column reacted with CO2 for 2, 3.5, and 5 months displays obvious changes in greyscale color as a result of cement degradation by CO2-saturated water, contrary to relatively constant greyscale color in the XMT image of an unaltered cement column (Figure 1C-1E). The degradation front, which corresponds to the Ca(OH)2 depleted zone, is clearly shown by a dark, thin layer with a thickness less than ∼0.5 mm between the inner unaltered cement matrix and the outer carbonated zone (Figure 1C-1E). Darker greyscale color of the degradation front than inner unaltered cement matrix is attributed to relatively lower bulk density in the degradation front due to higher porosity caused by depletion of Ca(OH)2(s) in the cement matrix. The degradation front is not clearly displayed by backscattered electron (BSE) due to its narrow width and the small size of the leached grains of Ca(OH)2(s).10 However, XMT has a great capability of visualizing the degradation front because of a significant density contrast between the unaltered cement matrix (2.15 g/cm3) and the degradation front region, which can facilitate the estimation of the spatial extent of cement degradation. The outer carbonation zone is evidently shown with lighter color because of precipitation of CaCO3 (mainly calcite by XRD; see Figure S4B) with a typical density of 2.71 g/cm3 (Figure 1). This carbonation zone seems to correspond to the zone with similar Ca and C atom % as an evidence of CaCO3 precipitation by SEM-EDS analysis (Figure 1B, Figures S2 and S3). The outermost zone of porous silica enrichment due to further dissolution of CaCO3 and C−S−H(s)10 was not obviously distinguishable from the outer carbonated zone in the XMT

materials within the sample. See the Supporting Information for detailed information.

■

RESULTS AND DISCUSSION Visualization of Cement Degradation and Carbonation. As Portland cement comes in contact with supercritical CO2 dissolved in the aqueous phase, carbonic acid attacks the cement matrix. Portlandite [Ca(OH)2(s)] is dissolved by carbonic acid as the carbonated water diffuses into the cement matrix, resulting in the increased leaching of Ca2+ out of the cement matrix and an increase in the porosity.10,11 As the cement degradation proceeds, CaCO3 begins to precipitate as a result of the reaction between Ca2+ diffused out of the cement and dissolved carbonate ion diffused into the cement (Figure 1A). The reaction of CO2 with C−S−H phase forms CaCO3 and amorphous silica gel. The filling of the cement pore space with CaCO3 precipitates creates a denser and less permeable zone.10 Further cement degradation by carbonic acid enhances the depletion of Ca(OH)2(s) from the cement matrix and also causes dissolution of CaCO3. With the dissolution of CaCO3 by continuous CO2 reaction, only amorphous silica gel with reduced mechanical stability is remained in cement paste.10 SEM-EDS analyses of duplicate cement columns show a systematic decrease of Ca and increase of C atom % from unaltered inner zone (average 20% Ca and 7% C) to degraded/ carbonated outer zone (Ca < ∼15% and C > ∼15%) as a result of cement carbonation (Figure 1B, Figure S2, and Figure S3). In a slice image of XMT (imaged and reconstructed under the same settings) brightness is proportional to X-ray absorption; thus dark regions correspond to low-density phases, 285

dx.doi.org/10.1021/es3012707 | Environ. Sci. Technol. 2013, 47, 283−289

Environmental Science & Technology

Article

image probably because of the similar density between calcite (2.71 g/cm3) and silica (2.65 g/cm3) and the narrow thickness of the porous silica zone.27 The outermost porous silica zone appears to correspond to the zone with much lower Ca/Si ratio of ∼1 and Ca/C ratio of 0.2−0.5 in atom % compared to unaltered cement matrix (Ca/Si = ∼2.5 and Ca/C = ∼3; Figure 1B, Figures S2 and S3). Estimation of the Extent of Cement Alteration. Twodimensional or 3-D XMT imaging comprehensively visualizes the extent of cement degradation, which is the depth to the degradation front (Figures 1, 2 and Figure S5). The cement degradation depth for the bottom half of the cement columns (water-to-cement ratio = 0.33) after the reaction with CO2saturated water for 5 months are heterogeneous ranging from nearly 0 mm−7 mm. The varying cement degradation depths are attributed to heterogeneous micropore structure of the hydrated cement that develops during hydration of the anhydrous cement grains reacting with water. The average degradation depth of cement columns (NRAP-25 and -33; the water-to-cement ratio = 0.38) without additive (steel or basalt) after the reaction with CO2-saturated groundwater for 2 and 3.5 months was 1.44 mm and 1.52 mm, respectively, while the degradation depth was 2.98 mm on average for the cement column NRAP-16 (the water-to-cement ratio = 0.33) after 5 months reaction (Table 1 and Figure 1). Although the cement degradation process from the XMT imaging is generally in agreement with previous studies using backscattered electron (BSE) image,10,11,13 the cement degradation rate from our study could be different from that in the real wellbore environment because cement columns were cured under ambient condition for this study. Kutchko et al. (2007)10 reported that cement degradation depth was shallower when cement was cured at high pressure−temperature condition than ambient condition because of an increase in hydration and the change in the microstructure of Ca(OH)2(s). The extent of cement degradation is shown to be greater with the presence of additives such as steel coupons (NRAP-4 and -8) and basalt fragments (NRAP-24), as well as layered cracks (NRAP-10), than the pure cement column (NRAP-16) without additives. The average cement degradation depths for NRAP-4, -8, -10, and -24 were 3.53, 3.11, 4.23, and 3.17 mm for 5 months, respectively (Table 1 and Figure 2). The cracks or the interfaces between the cement matrix and steel coupons or basalt fragments could provide preferential pathways for carbonic acid, leading to enhanced cement degradation compared to the pure cement micropores. This suggests that the interface of wellbore environment between cement and reservoir rock or steel casing could be most vulnerable to the degradation by CO2.9,28 The degradation depth is usually negligible for the top half of cement columns exposed to wet supercritical CO2 compared to the bottom half of the cement columns exposed to CO2-saturated water,11 although a random and spatially isolated degradation zone with a depth of ∼3 mm is observed (Figure 2), which is attributed to condensed droplets of surface moisture on the outside surface.11 XMT images of cement columns prepared with the water-to-cement ratio of 0.38 consistently indicate negligible cement degradation by wet supercritical CO2 for 3.5 months, except for a random and spatially isolated degradation zone likely caused by condensed droplets of surface moisture (Figure S6). XRD and SEM-EDS data also show that portlandite is the dominant mineral phase with a minor calcite in the cement matrix exposed to wet supercritical CO2 for 5 months (Figure S4A),

Figure 2. Greyscale XMT images of vertical slices for top and bottom half of cement columns NRAP-4 and NRAP-8 with steel coupons, NRAP-10 with layered cracks, NRAP-16 (pure cement column), and NRAP-24 with Wallula basalt fragments. The top half of the cement columns was exposed to wet supercritical CO2 for 5 months, while the bottom half was reacted with CO2-saturated water for 5 months at 50 °C and 10 MPa. Ring artifacts observed in the XMT images are generated by differences in sample centering during the XMT scan.

and the changes of Ca atom % (16−20%) and C atom % (5− 9%) from inner zone to outer zone is insignificant (Figure S2). In contrast, calcite is the dominant mineral phase with no remaining portlandite in the cement matrix carbonated by CO2saturated water for 5 months (Figure S4B). Quantification of Cement Alteration and Pore Structure Change. It has been shown that a linear relationship between the X-ray attenuation coefficient and the mineral density exists when the column is imaged at energy near or above 100 KeV, and the gray levels of the reconstructed XMT slices correspond to a map of attenuation coefficients within a sample.20,22 The XMT produces images where the voxel intensity is proportional to the density of the material when the columns are scanned and reconstructed under identical conditions. Despite the potential limitations of XMT application because of artifactssuch as beam hardening effects and ring artifacts (see the Supporting Information)as well as instability of the source, X-ray tomography has been used to measure mineral or material density in combination with specific image processing methods29 and to evaluate the density variation in microcrystalline cellulose compacts obtained at different mean porosity.30 After wet supercritical CO2 reaction at 50 °C and 10 MPa for 5 months, the spatial variation of image intensity that is related to material density change is minor with ∼125 along the horizontal cross section for the top half of cement columns, suggesting insignificant cement degradation by wet supercritical CO2 (Figure S7). In contrast, the variation of image intensity is more pronounced for the bottom half of cement columns, with lower intensity 286

dx.doi.org/10.1021/es3012707 | Environ. Sci. Technol. 2013, 47, 283−289

Environmental Science & Technology

Article

Table 2. Average X-ray Image Intensity, Average ln (I/I0), Known and Average Estimated Densities (g/cm3), and Average Estimated Porosity of Regions from 5 Cement Columns Reacted with CO2-Saturated Water for 5 Monthsa region

number of analysis

intensity

ln (I/I0)

known density

background unaltered cement basalt degradation front carbonated zone calcium carbonate air-filled void

16 15 1 9 11 3 4

39 124 152 113 132 150 57

0.004 0.504 0.582 0.488 0.527 0.588 0.148

0.00 2.15 2.70

estimated density

estimated porosity

1.98 2.27 2.54 0.65

40% 23%

a

Densities of the degradation front, the carbonated zone, air-filled voids, and calcium carbonate were estimated based on the linear relationship between ln (I/I0) and known material densities (background air, unaltered cement, and basalt) for each individual cement column (Figures S9 and S10). Porosity is estimated based on material density assuming that porosity of unaltered cement is 30% and all pores are initially filled with water.

Figure 3. Greyscale XMT images showing the precipitation of CaCO3 near the interface between cement matrix and steel coupon (A and B: NRAP4 and C and D: NRAP-8) or between cement matrix and basalt (E and F: NRAP-24), as well as CaCO3 precipitation along the crack (G and H: NRAP-10). A, C, E, and G: the top half of the cement columns exposed to wet supercritical CO2 for 5 months; B, D, F, and H: the bottom half of the cement columns reacted with CO2-saturated water for 5 months. I and J: XMT visualization of horizontal slices of the NRAP-10 cement column with layered cracks showing the calcium carbonate precipitation filling a crack during carbonation by wet supercritical CO2 (I) and CO2-saturated water (J) for 5 months. Light blue and dark blue colors correspond to low- and high-density regions, respectively. Ring artifacts observed in the XMT images are generated by differences in sample centering during the XMT scan.

cement matrix (n = 15) to 113 ± 4 and 0.488 ± 0.033 for the degradation front (n = 9), 132 ± 5 and 0.527 ± 0.027 for the carbonated zone (n = 11), and 150 ± 7 and 0.588 ± 0.037 for CaCO3 minerals (n = 3; mainly calcite with minor aragonite by XRD, Figure S11) formed on the outside surface of cement columns, respectively (Table 2). Based on the linear relationship between ln (I/I0) and known material density derived from XMT image analysis of individual samples (Figure S9), the averaged bulk density of the degradation front (dark, thin inner layer), the carbonated zone (light outer zone), and CaCO3 mineral is estimated to be 1.98 ± 0.07 g/cm3 (1.89−2.09 g/ cm3; n = 9), 2.27 ± 0.08 g/cm3 (2.18−2.45 g/cm3; n = 11), and 2.54 ± 0.07 g/cm3 (2.48−2.62 g/cm3; n = 3), respectively (Table 2). Such a wide range of estimated density for the degradation front and the carbonated zone indicates a variable degree of cement degradation within a cement column and between cement columns because of intrinsic heterogeneity in cement pore structure. The estimated density of large air-filled voids varies from 0.26 to 1.32 g/cm3 (Table 2), which is higher than that (∼0.001 g/cm3) of air in background. The reason is uncertain, but it could be attributed to the presence of high salt fluid or precipitates (e.g., CaCO3) or it could be a phase contrast artifact. The lower average density of 1.98 g/cm3 for

(∼110) in the degradation front and higher intensity (∼130) in the carbonated zone compared to unaltered cement (∼125) due to higher extent of cement degradation by CO2-saturated water (Figure S8). For each of the five cement columns (NRAP-4, -8, -10, -16, and -24) that has been degraded by CO2-saturated water for 5 months, ln (I/I0) shows a good linear relationship (R2 > 0.99) with the known density values of the materials, which are air in background (∼0 g/cm3), unaltered cement matrix (2.15 g/cm3), and basalt fragment (2.70 g/cm3) (Figures S9 and S10). The bulk densities of unaltered cement (water-to-cement ratio = 0.33) and basalt fragment were directly calculated after measuring the mass and the volume of samples. The linear relationship was generally similar for all cement columns with the slope that varies from 4.0−4.5. Slight differences in the slope are attributed to minor differences in column position and variation in the X-ray tube power. For NRAP-4 and NRAP-8 cement columns containing steel coupons, the density estimation was performed on the regions away from steel coupon because of an interference of X-ray image intensity resulting from high X-ray absorption of steel materials. The average image intensity and ln (I/I0) in all five cement columns varies from 124 ± 3 and 0.504 ± 0.038 for unaltered 287

dx.doi.org/10.1021/es3012707 | Environ. Sci. Technol. 2013, 47, 283−289

Environmental Science & Technology

Article

filled voids were filled with CaCO3 formed during cement alteration by both wet supercritical CO2 and CO2-saturated water (Figure 3G-3J). It is interesting to note that CaCO3 filling of cracks or cement pores along the interface between cement and basalt occurred by both wet supercritical CO2 and CO2-saturated water, whereas CaCO3 precipitates on the outside surface of the cement column were formed only by CO2-saturated water. Condensed water drops could have played a role in increasing the diffusion of Ca and carbonate ions in the top half of cement column exposed to wet supercritical CO2 to form CaCO3 in the cracks (Figure 3G and 3I) or cement pores along the interface with basalt (Figure 3E). Application of XMT for Wellbore Integrity Evaluation. The present study demonstrates that X-ray computed tomography is capable of nondestructively visualizing and quantifying the spatially heterogeneous degradation of wellbore cement material, including density and pore structure changes due to CO2 attack under geologic sequestration conditions. The XMT imaging of hydrated Portland cement reacted with CO2 under geologic sequestration conditions obviously visualized the degradation front with low density and high porosity as well as the carbonated zone with high density and low porosity in three dimensions. XMT can be utilized to image and quantify the time-dependent degradation process of cement materials during exposure to supercritical CO2 in laboratory. Whereas laboratory experiments can investigate the cement degradation only during a limited period of time (