Experimental Investigation of Supercritical CO2-Rock-Water Interactions in a Tight Formation with the Pore Scale during CO2-EOR and Sequestration.

In recent years, gas injection, especially CO2 injection, has been acknowledged as a promising approach for enhanced oil recovery (EOR) and CO2 capture and storage (CCS), especially for tight reservoirs. However, when CO2 is injected into the oil reservoirs, it can disturb the equilibrium of the system and lead to chemical reactions between CO2, formation water, and reservoir rocks. The reactions will alter some geochemical and physicochemical characteristics of the target reservoirs. However, the reactions still lack quantitative characterization at the pore scale, especially under reservoir conditions. Herein, we conducted an experimental study of the interactions between CO2, brine, and rocks in the Mahu oilfield at 20 MPa and 70 °C. The low-field nuclear magnetic resonance (LF-NMR) measurements showed that the incremental amplitude for tight cores of CO2-rock-water tests was larger than that for CO2-rock tests, and the amplitude alteration presented significant differences corresponding to different types of minerals and pores. Furthermore, the interplanar spacing of the core samples was increased with the increase of reaction time in the CO2-rock experiments but still lower than that in CO2-rock-water tests. This research demonstrated evident changes in the geochemistry in tight reservoirs caused by CO2, brine, and rock reactions. The results of this study may provide a significant reference for the exploration of similar reservoirs in the field of CO2-EOR and CO2 sequestration.

Introduction

Unconventional reservoirs including tight oil and shale oil have been drawing increasing attention owing to the increasing energy demand. According to the estimation, approximately 30 billion barrels of unconventional oil are distributed worldwide in 24 oil reservoirs.[1−3] The recoveries of these reservoirs (Bakken oilfield, Eagle Ford oilfield, and Vaca Muerta oilfield) are believed to be less than 10% even after fracture, owing to their low porosity and ultralow permeability.[4,5] Therefore, enhanced oil recovery (EOR) approaches should be applied to disclose the locking.[6,7] The commonly used water flooding is not suitable for tight reservoirs (Bakken oilfield and Eagle Ford oilfield) owing to the extra high injection pressure.[8] CO2 flooding has been proven to be useful and had the potential to enhance the recovery of tight reservoirs (Bakken oilfield, Eagle Ford oilfield, and Changqing oilfield) among all of the effective EOR methods.[9−11] In addition, reducing CO2 emissions has become an urgent worldwide problem. Thus, CO2–EOR associated with CO2 sequestration illustrated an excellent development potential in the future.[12]

It was concluded that the equilibrium of the natural condition would be broken owing to the injection of CO2 into the subterranean layer. Subsequently, the alteration of geochemical and physicochemical features of the target reservoirs would change due to the chemical reaction among reservoir rocks and formation brine and consequently affect the behavior of CO2–EOR.[13−15] Therefore, it is of vital importance to have a comprehensive understanding of the reactions triggered by the injection of CO2. Lots of efforts have been devoted to studying the interactions of CO2–water–rock in recent years.

Zhang et al. investigated the interactions of reservoir rocks (Lucaogou formation of Jimsar sag, Junggar Basin), formation brine, and supercritical CO2 under reservoir conditions. They found that the dissolving of supercritical CO2 in the formation water would generate an acidic condition, which would cause the dissolving of minerals and their subsequent precipitation. Furthermore, the rock surface after exposure to CO2 was changed to be hydrophilic owing to mineral dissolving, kaolinite formation, and surface corrosion.[13] Abedini et al. claimed that the chemical interactions might lead to dissolution and precipitation of certain minerals and alter the geophysical properties, including porosity and permeability of reservoir rocks.[16] Yu et al. claimed that mineral wettability, composition, and oil saturation were the main controls on the exposed surface area of grains, and mineral wettability, in particular, led to selective dissolution.[17] Fuchs et al. evaluated the effects of geochemical reactions on the geomechanical integrity of representative siliciclastic reservoir samples. The fracture toughness results demonstrated that carbon storage reservoirs might undergo geomechanical weakening with CO2 injection, which could lead to redistribution of stresses that are able to induce fracture slippage and trigger microseismic events.[18] Zou et al. found that mineral dissolutions caused numerous large etched pores, which eventually resulted in a significant increase in porosity and permeability in their experiment.[19] Zhang et al. utilized computed tomography (CT) scanning-discrete element method (DEM) combined approach to explore the alterations of limestone rock mechanical properties during CO2 injection.[20] Wei et al. investigated the interaction dynamics between CO2, water, and rock minerals under realistic reservoir conditions. The results indicated that CO2-triggered reactions increased the permeability of the tight core, leading to the consumption of injected CO2.[8] Tang et al. explored the mechanism that alters the characteristics of the reservoir for the CO2–brine–rock reaction during CO2 injection and storage in gas reservoirs. The results showed that the interaction resulted in the alteration of petrophysical properties that core permeability reduced as the porosity increased. Therefore, the dry CO2 ought to be injected into the water area to decrease the side effect of CO2–brine–rock interactions and guarantee the practical implementation of CO2 capture and storage (CCS) projects in gas reservoirs with the aquifer.[21] However, few researchers have investigated the effects of different minerals on CO2–water–rock interactions at the pore scale in a realistic reservoir environment. Actually, it is vital to investigate the mineral types suitable for CO2 storage.

Owing to the extremely low permeability of the tight reservoir, little efforts have been devoted to exploring the reactions between CO2 and rock minerals at the pore scale. Moreover, most of the existing studies failed to quantify the impacts of the interactions between CO2, water, and rock minerals. Therefore, this work focuses on quantifying the alterations caused by the interactions between CO2, tight core, and water of the Mahu tight conglomerate reservoir in the Junggar Basin, northwest China, under reservoir conditions (20 MPa, 70 °C) at the pore scale. X-ray diffraction (XRD), scanning electron microscopy (SEM), and low-field nuclear magnetic resonance (LF-NMR) spectroscopy were applied to characterize the reaction process in this study. The LF-NMR measurement results indicated that the increase of amplitude for tight cores of CO2–rock–water tests was mostly higher than that of CO2–rock tests. The increase of amplitude for big pores was higher than that for small pores in the case of CO2–rock trials, while the opposite results were observed in CO2–rock–water tests. Furthermore, the amplitude alteration presented great differences corresponding to different types of minerals and pores. Notable alteration of the mineral surface could be observed in SEM analysis owing to the interaction between CO2, water, and rock. The XRD measurements of the cores also indicated that the saturated CO2 could further expand the pore size. The research of this paper offered a further explanation of the CO2–EOR and CCS in tight reservoirs.

Experimental Methods

Materials

Synthetic core samples with pure minerals (calcite, feldspar, illite, kaolinite) were prepared before the tests. Then, they were dried in an oven at 100 °C for more than 90 h to diminish the influence of water before gas permeability and porosity measurements (LkiQR-168 Jiangsu Haian Petroleum Scientific Research Instrument Co., Ltd., China). The petrophysical characteristics of the used core are shown in Table . The deionized water was used in tests 5–8, and the CO2 was sourced from CO2 cylinders with a purity of 99.99%.

Table 1

Petrophysical Properties of the Core Samples

scenarios	tests	length (cm)	diameter (cm)	permeability, Kair (mD)	porosity, Φ (%)
CO2–core	1 (calcite)	6.57	2.54	0.09	8.11
	2 (kaolinite)	6.41	2.54	0.12	9.26
	3 (illite)	6.26	2.53	0.07	8.04
	4 (feldspar)	6.39	2.54	0.08	8.21
CO2–water–core	5 (calcite)	6.52	2.51	0.06	8.07
	6 (kaolinite)	6.29	2.54	0.10	8.96
	7 (illite)	6.28	2.54	0.07	8.13
	8 (feldspar)	6.40	2.54	0.09	8.19

Reaction between CO2 and Cores

The experimental process is described as follows: (1) vacuuming of the system and (2) injection of CO2 into the high temperature–high pressure (HT–HP) (Hastelloy, Haian Petroleum Technology Co.) cell at 20 MPa and 70 °C until equilibrium. The schematic diagram of the process is shown in Figure .

Figure 1

Schematic of the experimental setup for CO2 and cores experiments.

Reaction between CO2, Cores, and Deionized Water

The core samples were first soaked in deionized water for 72 h, as shown in Figure . Then, they were placed in the HT–HP cell, as illustrated in Figure . The experimental procedures are briefly described as follows: (1) the HT–HP cell was filled with deionized water; (2) the temperature was increased to 70 °C and air was vacuumed from the whole system, and then CO2 was pumped into the reference cell and pressurized to 5.0 MPa until the pressure was stable for 1 h; (3) the valve was opened and CO2 was injected into the HT–HP cell to a pressure of 20 MPa for 10 days; and (4) the mass loss of the core before and after the reaction was calculated. The percentage of mass loss was calculated as followswhere Wt stands for the percentage of mass loss of the rock, Wb is the weight of the rock after reaction (g), and Wb represents the weight of the rock before the reaction (g).

Figure 2

Schematic illustration of the experimental setup for the vacuum saturation device.

Figure 3

Schematic illustration of the experimental setup for different tight cores soaked in supercritical carbon dioxide (SC-CO2) under deionized water.

Scanning Electron Microscopy (SEM) Analysis

A core piece was cut from the end face of the tight core before and after being soaked in the supercritical carbon dioxide (SC-CO2). The morphology of core tablets was observed by scanning electron microscopy (SEM) with a FEI Quanta 450.

Low-Field Nuclear Magnetic Resonance (LF-NMR) Test

The core plugs were subjected to vacuum at 10–1 MPa for one week using a vacuum pressurization saturation device (KDZB-II, Kedi, China) and then pressurized to 30 MPa to saturate the core with brine (room temperature). Then, the cores were subjected to LF-NMR spectrometry (AniMR-150, Shanghai Niumag Electronic Technology Co., Ltd., China) to conduct the measurements of the nuclear magnetic resonance (NMR) T2 spectrum. The permanent magnet of the NMR spectrometer is 0.23 ± 0.03 T with a resonance frequency of 12 MHz. The echo and scanning numbers were 18 000 and 64, respectively. All of the measurements were performed at room temperature and atmospheric pressure.

Results and Discussion

LF-NMR Analysis

To comprehensively describe the CO2–rock–water interactions, low-field nuclear magnetic resonance (LF-NMR) was applied to illustrate the reaction process. The NMR transverse relaxation time (T2) spectra of the original, CO2–rock, CO2–rock–water for cores with different mineral types are presented in Figure . Based on the bimodal T2 curve, the pores of the tight cores can be divided into two regimes (small pores and large pores).[22−24] The increase of amplitude for tight cores of CO2–rock–water tests was mostly higher than that for CO2–rock tests owing to the dissolution reaction.[25]Figure simply presents the mechanisms of altering the pore size before and after the test. In the case of CO2–rock reaction tests, CO2 was trapped in porous media primarily through the adsorption process, and it could slightly enlarge the pore size of tight cores. However, this effect was not as strong as expected even though the CO2 was under a supercritical state. By contrast, the significant increase of pore size caused by the dissolution reaction is much more obvious than that by CO2–rock tests. Furthermore, it was found that the increase of amplitude for big pores was much higher than that for small pores in CO2–rock–water tests, as shown in Figure . This might be due to the fact that there was a large amount of micropores in tight cores, which were helpful to the dissolution interactions between CO2, rock, and water. The dispersed micropores were well connected after being exposed to saturated CO2. It should be noted that the alteration of amplitude presented significant differences corresponding to different types of minerals and pores. The highest increase of amplitude of total pores and big pores occurred in feldspar during the CO2–rock test.[26] But in the case of CO2–water–rock measurements, the highest increase of amplitude of total pores and big pores occurred in kaolinite. The mass loss and permeability/porosity alterations of tight cores after CO2–water–rock tests also present the same results, as shown in Figure (Figures , 6, and 7).

Figure 5

NMR T2 spectra of different cores ((a) calcite, (b) kaolinite, (c) illite, (d) feldspar) during CO2–rock and CO2–rock–water measurements.

Figure 7

Schematic of the alteration mechanisms of pore size during CO2–rock and CO2–rock–water measurements.

Figure 6

Increments of amplitude with different cores during CO2–rock and CO2–rock–water measurements in different pore intervals.

Figure 4

Porosity/permeability increment (a) and mass loss (b) of the core sample after the interaction between CO2, water, and rock.

Mineral Surfaces

Figure shows the mineral surface morphology of the rock disks before and after the CO2–water–rock reaction by SEM. It can be seen that the rock consisted of fine cemented minerals, as shown in Figure a. After being exposed to saturated CO2, the dissolved pores and pits were clearly observed, which could notably increase the connectivity of the tight reservoir rocks, as shown in Figure b.[27] Furthermore, the SEM images presented conclusive evidence of feldspar dissolution (Figure b) and kaolinization (Figure d).

Figure 8

Mineral surface morphology of the rock disks before (a) and after the CO2–water rock reaction (b–d).

XRD Analysis

It can be seen from Figure that the intensities of the peaks become notably weak with the increase of reaction time, which corresponds to the expansion of the mineral during the injection of CO2. New characteristic peaks were not observed during the experiments, indicating that only physical interaction was triggered between CO2 and tight core. On the contrary, the new characteristic peaks occurred in the case of CO2–water–rock experiments (as shown in Figure ), which suggested that there existed a chemical reaction during the injection of saturated CO2. Furthermore, the interplanar spacing D, which can comprehensively reflect the alteration of the pore size of the cores before and after the injection of CO2/saturated CO2 was calculated by the following equationwhere K is the Scherrer constant, λ denotes the wavelength of X-ray, B represents the half-width of the peak for the core samples, and θ is the diffraction angle.

Figure 9

XRD spectra of different cores during CO2–rock experiments.

Figure 10

XRD spectra of different cores during CO2–rock–water experiments.

It can be seen from Tables –5 that the interplanar spacing of the core samples was increased with the increase of reaction time in the CO2–rock experiments but still lower than that in CO2–rock–water tests. This may be attributed to the expansion of pore size being limited since only physical interactions occurred during the injection of CO2. In contrast, the chemical reaction caused by the injection of saturated CO2 could further expand the pore size.

Table 2

Parameters of XRD during Calcite–CO2 and Calcite–CO2–Water Tests

calcite–CO2	D (interplanar spacing, nm)	B (half-width of peak)	2θ
origin	0.2987	0.204	29.5904
30 h	0.3001	0.179	29.7604
50 h	0.3016	0.147	29.8964
calcite–CO2–water	0.3026	0.154	29.4954

Table 5

Parameters of XRD during Illite–CO2 and Illite–CO2–Water Tests

illite–CO2	D (interplanar spacing, nm)	B (half-width of peak)	2θ
origin	1.0107	0.148	8.6966
30 h	1.1343	0.133	8.7096
50 h	1.1661	0.132	8.7266
illite–CO2–water	1.1674	0.129	8.6836

Conclusions

To clarify the interaction process between CO2, water, and rock at the pore level during the CO2–EOR operations, we systematically presented an experimental investigation of characterizing the reaction mechanism in tight cores under reservoir conditions using LF-NMR, XRD, and SEM methods. Based on the experimental data, the following conclusions can be generally drawn:

The low-field NMR tests indicated that the increase of amplitude for CO2–rock–water tests was larger than that for CO2–rock tests due to the dissolution reaction.

The amplitude alteration presented great differences corresponding to different types of minerals and pores.

The interplanar spacing of the core samples was increased with the reaction time in the CO2–rock experiments but was still lower than that in CO2–rock–water tests.

Although limited works have been conducted in this paper, it provides some insights into the study of CO2–EOR and CCS in tight reservoirs. For example, core samples with a single mineral in this study were not representative to reflect the actual reservoir condition. Therefore, the natural cores will be used to characterize the CO2–EOR process at the pore scale in our future works.

Table 3

Parameters of XRD during Feldspar–CO2 and Feldspar–CO2–Water Tests

feldspar–CO2	D (interplanar spacing, nm)	B (half-width of peak)	2θ
origin	0.3138	0.492	28.114
30 h	0.3152	0.262	28.128
50 h	0.3168	0.205	29.264
feldspar–CO2–water	0.3234	0.155	28.051

Table 4

Parameters of XRD during Kaolinite–CO2 and Kaolinite–CO2–Water Tests

kaolinite–CO2	D (interplanar spacing, nm)	B (half-width of peak)	2θ
origin	0.3599	0.204	26.7854
30 h	0.3619	0.164	26.8346
50 h	0.3628	0.162	27.0064
kaolinite–CO2–Water	0.3657	0.17	26.7584