Effect of Fluid/Rock Interactions on Physical Character of Tight Sandstone Reservoirs during CO2 Flooding

Structures of pore-throat and permeability alteration caused by precipitation and the dissolution of rock matrix are serious problems during CO2 flooding into reservoirs for enhanced oil recovery (EOR). Experiments were conducted under pressure boost and reduction conditions, which simulate CO2-brine scaling in different parts of the reservoir during CO2 flooding. And experiments on the dissolution and scaling of CO2-brine-rock were carried out. The results show that the pH of brine with CO2 under high pressure is small, and no precipitation is formed, so there is no precipitation generated near the gas injection well. Pressure drops sharply near the production well, CO2 dissolved in the formation fluid escapes in large quantities, pH increases, carbonate precipitates are generated, so inorganic scale is formed near the production well. The increase of permeability of core saturated high scale-forming ions is smaller than that of saturated no scale-forming ions brine after CO2 flooding. The accumulation and attachment of salt crystals were found in some large pores of the core with scale-forming ions water after CO2 flooding. The ratio of medium size pores decreased, while that of large and small pores increases, and the pore radius distribution differentiates toward polarization.


Introduction
Change in injectivity is a problem in CO 2 injection wells, in enhanced oil recovery and sequestration projects, because of reactions between CO 2 , brine, and rock [1]. For tight Sandstone Reservoirs, the pores and throats of formation are tiny (the pore-throat radius is about 2~0.03 μm) and the water saturation is high, and the reservoirs are easily damaged. So protecting the tight reservoirs is an important issue.
During CO 2 huff and puff or CO 2 flooding, crude oil, rock, and brine in the reservoir will contact with CO 2 and undergo physical-chemical reactions [1,2]. Because CO 2 is dissolved in the brine to form carbonic acid, carbonate brine can dissolve the rock matrix in the reservoir. Carbonate minerals and silicate minerals are the main ones dissolved in the matrix, which results in forming new circulation channels and thus increasing the permeability of the stratum [2][3][4][5]. It has also been found that the carbonate ion in the formation water and metal ion (Ca 2+ , Mg 2+ ) produced precipitation which will be deposited on the surface of the rock. Mean-while, the generated kaolinite, intermediate products, and clay particles dissolved by carbonate cement migrated to pore and throats, and some smaller pores and throats would be blocked, leading to permeability reduction [6][7][8][9][10]. A carbonated water injection field experiment was conducted in Daqing oilfield, and the oil recovery increased by 10.5%. However, due to the scaling and corrosion of injection wells and pipelines were very serious; the field test was finally terminated [11].
The variation trend of rock properties during CO 2 flooding is related to the distribution of pores and throats, brine composition, and thermodynamic conditions, which depends on time and space. So some research results show that CO 2 flooding improved the reservoir, while some considered this technology damages the reservoir. Ross et al. [12] conducted CO 2 flooding experiments on cores of oolitic limestone, calcite, and dolomite, respectively, and the experimental results showed that CO 2 flooding increased core permeability. Sayegh [13] conducted similar experiments, selecting sandstone cores for CO 2 flooding. Permeability first increase and then increases. Omole et al. [14] studied the interaction between CO 2 and dolomite, and the results showed that the permeability near the injection well increased while that near the production well decreased. They believed that this was due to the massive dissolution of carbonate minerals around the production well and the new secondary minerals forming near the production well. But most of the previous studies focused on the influence of the interaction between carbonate water and rock on the reservoir [15][16][17][18][19][20][21] and, rarely, considered the effect of the reaction between CO 2 and the brine with lots of scale-forming ions in the reservoir. Especially for tight sandstone with formation water containing high concentrations of Ca 2+ and Mg 2+ , the pore-throat structures are very complex, so it is necessary to study the interaction between formation water and CO 2 .
In this work, taking the rocks and formation water of reservoirs in Changqing oilfield as the research objects, experiments of reaction between brine and CO 2 under pressure boost and reduction condition and CO 2 flooding were carried out to study the precipitation formation of CO 2 -formation water in the process of pressure rising and falling. The permeability and porosity of cores before and after CO 2 flooding were tested. The rock mineral composition, micromorphology, composition change, and rock pores and throats radius distribution change were analyzed before and after the CO 2 flooding by EDS analysis, SEM electron microscope analysis, and NMR.

Experiments
2.1. Materials. According to the composition of Triassic produced water in Block X of Ordos Basin, the composition of brine was tested and shown in Table 1. The purity of CO 2 used in the experiment is 99.99%. Simulated formation water was prepared, and the scale-forming ions (Ba 2+, Sr 2+, Ca 2+, Mg 2+ ) in the formation water are replaced with K + and Na + ; the TDS (Total dissolved solids) is still 92455 mg/L.
Cores: the natural core of Triassic X reservoir in Block X of Ordos Basin is adopted. After the cores were cleaned and dried, they are vacuumed for 48 hours and, then, saturated with clean formation water, and the porosity of the cores is obtained. Then, the permeability of the cores is measured by injecting formation water into the cores. As shown in Table 2, the porosity measured ranged from 9.56% to 13.09%, and the absolute permeability ranged from 0.14 to 0:4 × 10 −3 μm 2 . Mineral composition of the cores is shown in Table 3.

Experimental Equipment
2.2.1. Experimental Equipment. The structure of the high temperature and high pressure (HTHP) visual reaction vessel is shown in Figure 1. The internal volume of the reaction vessel is 2000 mL, which can withstand the maximum pressure of 70 MPa. The pressure is controlled by the ISCO pump, and the solution changes can be observed through the sapphire visual window. There is a heating device inside the container; the temperature range is 20~180°C. X-ray photoelectron spectrometer (XPS): Thermo Fisher K-Alpha model; the pH meter (BPH-7200) accuracy was 0:01 ± 0:0005, and the standard buffer solution with pH = 4, 7, 9 was used for calibration before and after the experiment. CO 2 was injected into the vessel with formation water and inserted the pH meter. After preheating the solution to the    Figure 1: HTHP visual reaction vessel.
2 Geofluids specified temperature, gradually increase the pressure and record the balance reading displayed on the pH meter at each pressure point. A schematic diagram of the high-pressure core-flooding apparatus used for CO 2 flooding experiments is shown in Figure 2. Formation water, simulated formation water (water with no scale-forming ions), and CO 2 were contained separately in three high-pressure cylinders. A dual ISCO syringe pump was used to inject fluid from the high-pressure cylinders to the core holder (P max = 80 MPa; T max = 130°C). A pump was used to maintain confining pressure, and another pump and a back-pressure valve were used together to regulate and maintain the back-pressure. The cylinders and core holder were placed in a constant-temperature oven (T max = 100:0°C). The produced brine, oil, and gas were collected and quantified by a gas−liquid cyclone separator and a mass flow meter. Flow data and pressure were logged automatically by a computer during the experiments.

Experimental Methods and Procedures
2.3.1. Experiments of Reaction between CO 2 and Brine. 250 mL of brine was injected into the HTHP reaction vessel. CO 2 was injected into the reaction vessel through the cylinder. As the pressure changed, precipitation was observed in the solution through the visual equipment. The pH of brine dissolved with CO 2 under different pressures was measured by the pH meter. The precipitate was filtered, dried, pressed, and then measured by X-ray electron spectroscopy (XPS) to determine the elemental composition of the precipitate.

CO 2 -Water-Rock Interaction Core Flooding
Experiment. The schematic diagram of the core displacement experiment is shown in Figure 3. Before the experiment, the cores were put into the extraction device and cleaned with toluene and methanol to remove oil, salt, and cement. Oil, water, and CO 2 are injected into the cores.
Each group of CO 2 flooding experiment was carried out according to the following steps. First, slices A and D with thickness of 2 mm were cut at the injection end and the outlet end of the core, respectively, as analysis samples. The distribution of pore radius of cores saturated with simulated formation water (water with no scale-forming ions) was determined by NMR. The cores were saturated with formation water and then the permeability was measured. CO 2 and water were injected alternately at experimental temperature and pressure. First, 5 PV (pores volume) CO 2 was injected, and then, 5 PV water was injected. Three cycles of injection were performed. After CO 2 flooding, the cores were cleaned and dried, and pore volume and permeability were tested. And then, the MRI was tested and, then, cut core pieces B and C with a thickness of 2 mm at the injection end and the outlet end of the core, respectively, clean and dry them. Scanning electron micrographs (SEM), surface element, and rock mineral analysis were carried out together with the core sections before the flooding. Replace the core and repeat the above procedure under different conditions. The porosity change rate is defined by Eq. (1). The permeability change rate is defined by the following Eq. (2).
where η φ is the porosity change rate, φ A:0 is before the reaction, and φ A is the porosity after the reaction.

Geofluids
where R is the permeability change rate, K d is the permeability before the reaction, and K o is the permeability after the reaction.

Results and Discussion
3.1. CO 2 -Formation Water Interaction. During CO 2 injection, the pressure around the borehole of the gas injection well is high. So a large amount of CO 2 will be dissolved in the formation water in this area. The pressurization can simulate the brine-CO 2 reaction around the injection well. However, the pressure funnel is obvious near the production well, and the sharp pressure drop inevitably causes a large amount of CO 2 dissolved in the formation fluid to escape. To simulate the reaction between formation water and CO 2 when the pressure near the well drops sharply, a pressure-reduced formation water-CO 2 reaction experiment was used. There is an excess of CO 2 throughout the reaction.
To determine whether the brine will precipitate at the formation temperature, pour the formation water into the reaction vessel, keep the temperature at 50°C for 48 h, and then observe through the visual window, no precipitation is observed, as shown in Figure 4(a). CO 2 was injected, increase the CO 2 pressure to 20 MPa, and observe for 12 days, there was no precipitation, as shown in Figure 4(b). Reduce the pressure drops to atmospheric pressure, and precipitation was observed after 2 days, as shown in Figure 4(c). Precipitation is not only deposited on the bottom of the vessel but also a large amount of precipitation is attached to the inner wall of the reaction vessel. CO 2 in the system is in a supercritical fluid state. At the same time, small bubbles containing salt are dispersed in the CO 2 supercritical fluid. When the system is depressurizing, scale precipitates in small water droplets and adheres to the inner wall of the container. Due to the upward flow of CO 2 during pressure relief, the deposition of scale is mainly distributed at the mouth of the container.
The result shows that as the pressure increases, the dissolution of CO 2 in the formation water increases, but there was no precipitation formed. During the depressurization process, precipitation was observed. In response to the above phenomenon, a high-pressure pH test device is used to measure the change in the pH of the CO 2 -formation water system during the pressure change under constant temperature conditions. The change in pH with pressure is shown in Figure 5.
The pH of the CO 2 -formation water decreases with the increase of pressure as CO 2 is continuously injected. When the pressure is between 0 and 15 MPa, the pH of the solution is between 6.5 and 3.1. This is due to the increase in pressure and the solubility of CO 2 in water, which increases the amount of CO 2 dissolved in water, and more CO 2 reacts with brine to form carbonic acid, thus resulting in the decrease of pH of the solution with the increase of pressure. During the experiment, no precipitation was found in the solution. Although the concentration of Ca 2+ , Mg 2+ , and Ba 2+ divalent ions is relatively high, formation water is acidic because of the injection of CO 2 , and CO 2 mainly exists in the form of HCO 3 and H 2 CO 3 (as Eq. (3) 4 Geofluids the high-pressure near the gas injection well, CO 2 will not react with CaCl 2 water-type formation water to produce precipitation. As time goes by, the amount of CO 2 dissolved in the formation water at different positions in the deep part of the reservoir gradually increases, and the pH of the formation water gradually decreases. CaCl 2 formation water in the deep part of the reservoir is the same as the near-well area of the gas injection well, and no inorganic precipitation was generated.
In the near-well zone of the production well, due to the sharp pressure drop, a large amount of CO 2 escapes. When the partial pressure of CO 2 in the solution decreases, the reaction proceeds to the left, and the originally dissolved Ca(HCO 3 ) 2 decomposes to form the CaCO 3 scale (as Eq. (6)-(10)). Therefore, during the CO 2 injection, CaCl 2 water-type formation water will lead to the formation of inorganic scales near the production well.
Some researchers have found that only when the downhole pressure is lower than the bubble point pressure, the CaCO 3 scale was found in the produced water of the oilfield [12]. Vetter (1980 [22,23] believes that this is because the pressure drops, and CO 2 is flash vaporized from the system, causing the pH value of the system to rise, and the CO 3 2concentration in the system rises accordingly; thus, the system is in a state of supersaturation and begins to precipitate carbonate [8]. Dawe and Zhang [24] believed that as the pressure drops and the gas was released from the liquid, the gasliquid interface appeared, which promoted the generation of scale. In their experiments, they found that the presence of small bubbles in the experimental system helped the crystallization of CaCO 3 and believed that the reason might be that the surface of the small bubbles had a high Gibbs free energy, which can act as a catalyst for the crystallization process [24]. To study the mechanism of CO 2 -formation water forming precipitation, starting from checking the Scanning electron microscope (SEM) image of the precipitation, the element and chemical composition of the sediment were analyzed. There are three common crystal phases of calcium carbonate crystals: calcite, vaterite, and aragonite [25]. Some researchers believed that under low-temperature conditions (below 80°C), C a CO 3 will precipitate out of the water to form the thermodynamically unstable vaterite [26,27]. It can be seen from the SEM pictures ( Figure 5) that the solution produced vaterite crystals which are loose and soft.
The precipitate was measured by X-ray electron spectroscopy (XPS), and the measurement results are shown in Table 4. The main elements in the precipitate are C and O, followed by metal elements, mainly Ca, Sr, Mg, and Ba, as shown in Figure 6. According to the ion composition measured by the formation water and the ionization reaction of CO2 dissolving into the formation water to produce CO 3 2and HCO 3 -, it can be concluded that the precipitates mainly consist of CaCO 3 , BaCO 3 , SrCO 3 , MgCO 3 , CaMg(CO 3 ) 2 , etc.
3.2. Effects of CO 2 Flooding on Properties of Rock. Previous scholars conducted high-pressure reaction experiments in cylinders of CO 2 -formation water-rock interaction and drew some conclusions about the influence of CO 2 -formation water-rock interaction on reservoir physical properties [6,15,16]. However, the reactor is a closed system, which cannot reflect the structural characteristics of porous media. Experiments have also been carried out with artificial cores [28][29][30]. Although there is no water-rock reaction in the artificial tight sandstone core, the pore structure of the artificial low-permeability cores is relatively simple, which cannot simulate the complexity of the pore structure of the tight sandstone cores. Therefore, it is necessary to carry out the formation water-CO 2 flooding experiment in the natural tight sandstone cores. By CO 2 flooding cores saturated with no scale-forming ions water, precipitation generated by brine and CO 2 reaction can be excluded, and the influence of CO 2 -water-rock reaction on the physical properties of rock can be obtained. By comparing the cores saturated with no scale-forming ions water and high concentration scale-      7 Geofluids forming ions water flooding experiments, the influence of the precipitation generated by CO 2 and formation water in the reservoir on the physical properties of rock is obtained.

Changes in Core Permeability and Porosity after CO 2
Flooding. The changes in core permeability and porosity after CO 2 flooding were shown in Figures 7 and 8. The permeability and porosity of cores saturated with simulated formation water (no scale-forming ions water) and cores saturated with formation water increased after CO 2 flooding under different pressures. Under the pressure of 15 MPa, the pH of the solution is about 3.1 (as Figure 9), and the permeability of the cores saturated with simulated formation water (no scaleforming ions water) increased by 2 times, and that of the core saturated with formation water increased by 1.5 times, and the higher the flooding pressure was, the more it increased. It can be seen that the effect of mineral dissolution on expanding and connecting pores is greater than the effect of particle migration and sedimentation to plug pores. CO 2 flooding can improve the permeability of the reservoir. And according to Figure 5, the higher the pressure, the greater the solubility of CO 2 in water, the smaller the pH of the solution, and the stronger the CO 2 -water-rock interaction. However, the increased permeability of cores saturated with simulated formation water (no scale-forming ions water) is larger than that of the core saturated with formation water, which may be caused by precipitation caused by formation water-CO 2 interaction.     Figure 9: pH of formation water at different pressures during CO2 injection (T = 50°C).

Geofluids
As shown in Figure 8, the porosity of cores saturated with simulated formation water increased by 6% after CO 2 flooding, with a very small amplitude, while that of core saturated with formation water increased by 2%. These changes are all caused by rock-CO 2 -formation water interaction and fluid flooding. Reactions in sandstone rocks generally include dissolution of carbonate minerals, alkaline feldspar, of which mass fraction was 19.7% totally, as shown in Table 3.

Geofluids
3.2.2. The Changes in Pore-Throat Structure of Cores after CO 2 Flooding. Permeability and porosity are the macroscopic parameters of rock physical properties, and the changes in the overall pore radius distribution of the cores caused by changes in the microscopic morphological structure of pores are the essential factors for changes in rock permeability. To study the effect of interaction between CO 2 and formation water on pore structure, NMR measurements of cores before and after flooding were carried out.
The NMR mechanism shows that the relaxation time (T 2 ) is proportional to the pore radius (r). Therefore, relaxation time is converted to the pore radius according to Eq. (12) [31].
where r is the pore radius, m; T 2 is NMR relaxation time, ms; and C is the conversion factor, whose value is 20.23 ms/m. The pore radius distribution curve of T 2 spectrum transformation was well fitted with the conventional mercury injection curve, and the correlation was high.
NMR transverse relaxation time of pores is less than 1 ms defined as small pores. 1-10 ms corresponds to medium pores, and greater than 10 ms corresponds to large pores. The corresponding pore-throat radius is classified as less than 0.36 μm, 0.36~1.82 μm, and greater than 1.82 μm. Figures 10-12 show the NMR spectra of cores saturated with formation water under different pressures before and after CO 2 flooding. It was found that the pore proportion between 0.1-1 ms and 10-1000 ms increased, and the pores proportion between 1 and 10 ms decreased, that is, the probability of large and small pores distribution increased, and the probability of medium pores distribution decreased. For tight reservoirs, on the one hand, macropores are the main flow channels, and the permeability is also determined by the proportion of macropores. The injected CO 2 mainly enters into macro-pores, where the flow rate of CO 2 is relatively high, making it easier for CO 2 to be replenished and updated in time, leading to mineral dissolution, particle release, and pore space enlargement. Meanwhile, there may also be particle migration and pore blockage (the pore space decreases), but generally, the precipitation is vaterite whose particles soft scale and loose, as shown in Figure 5 which are not enough to block the macropores and have a small probability of blocking the macropores. For medium pores, the reaction degree is bounded between large and small pores. Due to the different connection structures of the medium size pore-throat, the medium size pore-throat eventually differentiates into both ends of large pores and small pores. If the middle pores are connected with more throat channels with a smaller radius, the dissolved small particles in the reaction cannot be migrated well, which will plug up the pores and turn the middle pores into small pores. If the middle pores and macropores are connected, the small particles dissolved because of the reaction can be well transported out, and the middle pores before the reaction will be transformed into macropores. For small pores, on the one hand, CO 2 is a nonwet phase, and the resistance of the Jamin effect should not be ignored when it enters small pores. CO 2 mainly enters small pores through diffusion, and the amount is relatively small, so the change is not obvious. On the other hand, as the minerals in cores react with CO 2 , secondary micropores are created. In tight sandstone reservoirs, permeability is mainly contributed by macropores, which are consistent with the previous conclusion that cores permeability increases greatly but porosity changes little after CO 2 flooding.    Figure 12 shows the variation of the distribution of small, medium, and large pores of cores saturated with formation water before and after CO 2 flooding under different pressures. The proportion of small-medium pores increases, while the proportion of larger medium pores decreases, while the proportion of large medium pores increases, and the proportion of small pores increases. As Figure 12 shown, the probability of small pores distribution increases with the increase of pressure drop, which is caused by the increase of precipitation caused by formation water-CO 2 interaction. The probability amplitude of macropore distribution increases first and then decreases with pressure drop. The decreasing range of porosity distribution probability first increases and then decreases with increasing pressure drop.
Comparing Figure 11 with Figure 13, it can be seen that the reduction of the medium size pore-throat of cores saturated with formation water is larger than that of cores saturated with no scale ionized water after CO 2 flooding, but the increase in the amount of large and small pores is smaller than that after CO 2 flooding with no scale-forming ions water, which is due to metal scaling in the formation water. The reaction between ions and carbonate can form a precipitate in all pores, which results in a decrease in the volume of all types of pores. Figure 14 shows the dissolution of minerals under different flooding pressures because of the interaction of CO 2 , water with no scale-forming ions, and rock. As shown in Figure 14(a), before the reaction the crystal shape of the rock is complete and there is a little mineral debris on the quartz surface. After CO 2 flooding, carbonate on the rock surface dissolves, the detrital minerals at the throats in the rock dissolve, the detrital minerals on the rock surface decrease at the back-pressure of 10 MPa, and karst pits appear, as shown in Figure 14(b). With the increase of CO 2 pressure, the dissolution intensifies. After flooding at the back-pressure of 15 MPa, feldspar is more severely dissolved, and the shape of feldspar turns into filamentous, as shown in Figure 14(c). When the flooding pressure increases to  11 Geofluids 20 MPa, carbonate and feldspar are more severely eroded, resulting in some secondary pores in some areas, as shown in Figure 14(d).

Core Surface Morphology Changes after CO 2 -Water-Rock Interactions.
The permeability of the reservoir can be improved by increasing the connectivity between pores. The changes of pore morphology caused by the above reasons will seriously affect the permeability of the ultralow-permeability rock, but the porosity will be less affected. In particular, macropore throats, which have a greater contribution to rock permeability, are the main flow channels of fluids and the main sites for particle migration in the process of flooding. Therefore, SEM photographs show that the microstructure of macropores changes significantly, and the distribution of pore radius reflects that proportion of macropores increased.
The rock surface morphology changes after the CO 2 -formation-rock interaction (Figure 15(a)-15(c)). SEM showed that some of the precipitates were salt crystal fragments, and these minerals were loosely arranged and did not occupy too much pore space to block the large pore throats (Figures 15(b) and 15(c)). There is good connectivity between the throat and the throat, and this type of pore morphology has little effect on the permeability of the rock. In summary, after CO 2 flooding, the size and morphology of some pores in the core have undergone major changes. The distribution of core pore radius has changed on the entire core scale.

Conclusion
The reaction of CO 2 with high mineralization formation water during a pressure-boosting and pressure-reducing were conducted, and the results showed that in the CO 2 injection process, the near-well-bore area of the gas injection well and the deep reservoir because of the pressure rising, the dissolution of CO 2 in the formation water quantity rising, pH of the solution, no precipitate is formed in CaCl 2 formation water. However, in the near-well area of the oil well, with the sharp drop of pressure, a large amount of CO 2 dissolved in the formation water will escape, and the pH of the solution will rise, resulting in the formation of inorganic salt precipitation. According to the analysis of X-ray photoelectron spectroscopy, the main components of the precipitation are found to be CaCO 3 , BaCO 3 , SrCO 3 , MgCO 3 , etc., which is the loose and soft vaterite crystals.
The permeability, porosity, pore throat distribution, and surface morphology changes of the cores saturated formation water and cores of saturated no scale-forming ions water  12 Geofluids before and after CO 2 flooding were compared. The results showed that, in Changqing oilfield, the porosity, permeability of cores saturated no scale-forming water, and formation water increased after CO 2 flooding, according to the SEM found that the effect of mineral dissolution pore expansion. Unicom is greater than the effect of particle migration and precipitation of jam pore, even with relatively high-cost scale ion formation, CO 2 injection can still improve the permeability of the reservoir, but the medium size of pore throat turned into large pores and throats, which will aggravate the heterogeneity of strata. For highly mineralized formation water, carbonate precipitation will generate in the core after CO 2 injection, and the change amplitude of permeability, porosity, and core mass is smaller than that after deionized water displacement, carbonate salt crystals are precipitated in the pores.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare no conflict of interest.