Effect of Residual Water in Sediments on the CO₂-CH₄ Replacement Process

Abstract

CO₂ replacement is a promising method of gas hydrate recovery. However, the influence of residual water in the replacement process and selections of a suitable mining area remain uncertain. To better understand this method, we examined the influence of the particle size and initial hydrate saturation on the replacement process while using the same amount of residual free water. The results showed that during the replacement process, two stages of rapid reaction and slow reaction occurred, which were manifested by the speed of pressure change in the reactor. The CO₂ sequestration ratio decreased with the increase in sediment particle size and increased with the increase in initial hydrate saturation. During the replacement process, two reactions occurred: CH₄ was replaced by CO₂ and CO₂ hydrate was formed, and the replacement amount and recovery efficiency of CH₄ increased with a decrease in sediment particle size. When the sediment particle size was less than 166 μm, the CH₄ recovery efficiency was significantly affected by the particle size. The replacement amount of CH₄ increased with the increase in initial hydrate saturation, and the recovery efficiency decreased. This study provides a basis for selecting suitable hydrate-accumulation areas for on-site mining.

1. Introduction

Natural gas hydrate is a crystalline compound similar to ice that is formed by small molecular gas and water molecules in high-pressure and low-temperature environments [1,2]. Since natural gas hydrate is found to have a huge reservoir in seabed and permafrost regions and the major component of natural gas hydrate is methane (CH₄), it is considered as an unconventional energy source and has received growing attention worldwide [3,4]. The common hydrate mining methods include depressurization, thermal stimulation, and chemical injection. All these methods recover natural gas by promoting hydrate dissociation [5,6]. In recent years, a new method for natural gas hydrate recovery has attracted the attention of researchers as a CO₂-CH₄ replacement. This method can replace the methane molecules in natural gas hydrates and capture CO₂ molecules in the seabed or permafrost in the form of hydrates, ultimately achieving a “win-win” situation for economic development and environmental protection [7].

Feasibility of the CO₂-CH₄ replacement method has been proved and the behaviors of CO₂-CH₄ replacement are now being intensively investigated. Ota et al. [8] studied the effects of temperatures from 272.15–275.15 K and found that under certain pressures, the replacement efficiency and rate increase with an increase in temperature. The replacement rate was relatively high in the initial 12 h, but reduced gradually after that. Li et al. [9] measured the gas composition in the CO₂-CH₄ replacement and noted that the amount of CH₄ replaced and the consumption of CO₂ had basically the same trend. Zhou et al. [10] studied CO₂-CH₄ replacement at a temperature of 274.15 K and a pressure range of 3.97–6.26 MPa. The results showed that the replacement rate and the amount of CH₄ replaced increased with an increase in the replacement pressure; however, when the pressure was close to the liquefaction pressure at the experimental temperature, the amount of CH₄ replaced decreased instead. Zhou et al. [11] used liquid CO₂ emulsions to enhance the efficiency of CO₂-CH₄ replacement and found that the amount of the replaced CH₄ was about 1.5 times higher than liquid CO₂. Lee et al. [12] studied the effect of the particle size of sediment on the replacement rate and noted that the CO₂-CH₄ replacements performed in small particles (75–90 μm) was 1.4 times higher than those in large particles (125–150 μm). A simulation study conducted by Bai et al. [13] revealed that the CO₂-CH₄ replacement would get promoted with an increase in CO₂ concentration. The formation of CO₂ hydrates was assumed to eliminate the mass transfer barrier of the amorphous CO₂ hydrate layer on CH₄ hydrate and improve the replacement efficiency. The effect of hydrate saturation on CO₂-CH₄ replacement were studied by Zhou et al. [14], and they noted the replacement efficiency would decrease with an increase in hydrate saturation. Wang et al. [15] found that the smaller the sediment particle size, the greater the replacement efficiency; however, the effect of the sediment particle size on the replacement efficiency was not obvious. As CH₄ hydrate is stored in complex marine sediments, it is affected by many factors in the actual replacement mining process, including initial hydrate saturation [16], temperature [17,18] pressure [8,19,20], salinity [21], the CO₂-injected state [11], particle size [22], and sediment type [23].

However, the CH₄-CO₂ replacement in hydrate-bearing sediments is still complicated to predict. In 2012, ConocoPhillips and the U.S. Department of Energy [20] carried out a field test of CH₄-CO₂ replacement in hydrate reservoir located in the Alaska area of the United States. A total volume of 6114 m³ of gas containing 22.5% CO₂ and 77.5 N₂ was injected into the reservoir where the pressure and temperature were controlled at around 9.8 MPa and 278 K, respectively. After 48 days, 1421 m³ N₂ and 826 m³ CO₂ remained in the reservoir, while a total of 24,410 m³ CH₄ was recovered. Although the investigation of the sequestrated gas indicated a preferential retention of CO₂ in the hydrate-bearing sediment, it was not clear if the sequestrated CO₂ was participated in the CH₄-CO₂ replacement. This could also be attributed to the adsorption of porous media or the formation of CO₂ hydrate from residual water in the sediment.

Generally, there are at least four phases in methane hydrate-bearing sediments: hydrate, gas, water, and grain. Free water not involved in hydrate formation is residual water, which exists together with methane hydrate in the sediment matrix, and this residual water plays an important role in the process of CO₂ injection into CH₄ hydrate-bearing sediments [24]. It has been documented that when CO₂ is injected, CO₂ first interacts with residual water at the gas-liquid interface to form a CO₂ hydrate film instead of participating in the reaction. The affinity of CO₂ to form a CO₂ hydrate film at the gas-liquid interface reduces the availability and selectivity of CO₂ to reach the surface of the CH₄ hydrate-bearing sediment [25], resulting in the reduction in CO₂ infiltrating into the CH₄ hydrate-bearing sediment and affecting the replacement effect [26]. In addition, there is still controversy regarding the displacement mechanism, especially regarding the microscopic mechanism. Furthermore, whether CH₄ hydrate dissociation occurs before the formation of CO₂ hydrate, or if CO₂ molecules directly replace the trapped CH₄ in the hydrate cages remains unclear.

To elucidate the influence of residual pore water in hydrate bearing sediment, CH₄-CO₂ replacements were carried out in synthesized hydrate-bearing sediments in 277.15 K, 3 MPa. The CO₂ captured by the hydrate bearing sediments and the CH₄ recovered were calculated by sampling the gas composition during CH₄-CO₂ replacements. The effect of initial hydrate saturation and particle size on replacement efficiencies were analyzed. The results revealed the advantages of residual water for CO₂ sequestration and the side effects on CH₄ recovery efficiency, which would provide a possible path to describing the CO₂ behavior in CH₄ hydrate reservoir and lay a theoretical basis for selecting suitable hydrate–bearing sediments.

2. Materials and Methods

2.1. Experimental Apparatus

A schematic of the experimental apparatus is shown in Figure 1. The main part of this apparatus is composed of a high-pressure reactor with an effective volume of 492 mL, an effective height of 150 mm, an inner diameter of 65 mm. The reactor is equipped with a buffer tank with an effective volume of 1117 mL. Their design pressures were 15 MPa and 20 MPa, respectively. Devices for measuring temperature and pressure were arranged in both the reaction kettle and the buffer tank. The thermometer was made of Pt100 platinum with an accuracy of ±0.1 K. The pressure transducer had a range of 0–20 MPa and an accuracy of 0.1% F.S. The constant-temperature water bath has an adjustment range of 270.15–303.15 K. The data were collected every 10 s using an Aigent34970 acquisition instrument. To simulate the environment of natural gas hydrate reservoirs, a cylindrical mold was used, and the wet sand was put into the mold for multiple beating and compaction times, and then placed in the refrigerator for 6 h to form a cylindrical sand column with a diameter of 50 mm, height of 65 mm, and effective volume of 127.6 cm³.

2.2. Experimental Materials

Methane and carbon dioxide with a mole fraction of 0.999 were supplied by the Foshan Huate Gas Industry Corporation, and the experimental distilled water was made in the laboratory with a resistance of 18 MΩ/m. Reservoir sediments were composed of four different median particle sizes of natural river sand, the properties of which are shown in

Table 1. Properties of natural river sand.

Number	Median Particle Size (μm)	Porosity (%)
1	101	44.68
2	166	43.08
3	279	42.01
4	377	41.38

.

Figure 2 shows scanning electron microscopy (SEM) images of natural sand at different magnifications. There are many voids and textures on the surface of the sand, which provide good reaction sites for the formation of hydrates [27].

2.3. Experimental Procedures

The experimental process consisted of two steps: the preparation of methane-containing hydrate sediment materials and the replacement reaction of CO₂-CH₄.

2.3.1. Preparation of Methane Hydrate in Porous Media

Before the experiment, 150 g of dried natural sand was mixed with 8.13, 12.12, 16.04, or 20.05 g of distilled water to prepare 20, 30, 40, or 50% of the initial water saturation of the wet sand. Then, the wet sand was placed into a mold, beat, and compacted. The mold was placed into a refrigerator to form a cylindrical sand column at 263.15 K for 6 h. After that, the frozen sand column was placed into the reactor. The reactor was set to a temperature of 277.15 K using the water bath. Finally, valve 11 was opened, and CH₄ was slowly fed into the reactor to form CH₄ hydrate. The reaction was stopped when the weight of residual water in the sediment was approximately 6.2 g (the calculation process is described in Section 2.4.).

Table 2. Hydrate formation at different particle sizes.

NO.	Median Particle Size (μm)	Water Saturation (%)	Initial Hydrate Saturation (%)	Residual Water Weight (g)	CH4 Consumption (mol)
Exp.1	101	30	16.2	6.25	0.0552
Exp.2	166	30	17.1	6.15	0.0581
Exp.3	279	30	16.0	6.23	0.0550
Exp.4	377	30	16.5	6.2	0.0556

and

Table 3. Hydrate formation at different initial water saturations.

NO.	Median Particle Size (μm)	Water Saturation (%)	Initial Hydrate Saturation (%)	Residual Water Weight (g)	CH4 Consumption (mol)
Exp.5	377	20	5.6	6.23	0.0192
Exp.6	377	30	16.5	6.20	0.0556
Exp.7	377	40	28.3	6.18	0.0976
Exp.8	377	50	39.2	6.24	0.1340

show the hydrate formation conditions for eight groups of experiments.

We selected sand with a median particle size of 101,166,279,377 μm and an initial water saturation of 30% to prepare sediments containing methane hydrate and to explore the effect of sediment size on replacement.

Similarly, we selected sand with initial water saturations of 20, 30, 40, and 50% and a median particle size of 377 μm to prepare methane hydrate-containing sediments and study the effect of the initial hydrate saturation on displacement.

2.3.2. Replacement Reaction Process

The replacement reaction process was divided into three steps:

• Gaseous CO₂ injection. The remaining CH₄ in the reactor was quickly discharged, valve 11 was opened, and the reactor was purged with CO₂ for 15 s to ensure that the CH₄ gas in the reactor was completely discharged to the emptying pressure, and then valve 12 was closed. Subsequently, gaseous CO₂ was injected for one minute. When the pressure was stable at 3 MPa, valve 11 was closed. This moment was recorded as time zero of the replacement reaction.
• Gas sample analysis. Gas samples were collected at time zero, and a gas chromatograph (GC9790, USA) was used to measure the gas sample components and content, record it as the content before the reaction, and then collect gas samples and measurements at time intervals. The volume of the single gas sample did not exceed 20 mL. The effects on temperature and pressure were negligible.
• Replacement reaction process. The replacement temperature was 277.15 K, the replacement pressure was 3 MPa, the replacement time was set to 4 days for each set of experiments, the temperature fluctuation during the reaction process did not exceed 0.1 K, and the pressure drop at the end of the reaction did not exceed 0.1 MPa.

2.4. Calculations

2.4.1. Calculation of Methane Hydrate Saturation

During the experiment, the skeleton structure of the sediment was almost incompressible; therefore, it can be assumed that the initial porosity in the natural sand remained unchanged, CH₄ gas was basically insoluble in water, and methane hydrate was formed uniformly in the sediment. Therefore, we believe that all the methane consumed is used to form CH₄ hydrates, and the molar ratio of methane hydrate in the sediment can be calculated by Equation (1) [28]:

$$∆n=V_{\mathrm{g}}\frac{P_0}{Z_0{R}_0{T}_0}-V_{\mathrm{g}}\frac{P_1}{Z_1{R}_1{T}_1},$$

(1)

where △n is the CH₄ gas consumption during the formation process (mol), R is the gas constant, 8.134 J·mol⁻¹·K, Z₀ and Z₁ are the compressibility factors of CH₄ gas under the corresponding temperature and pressure conditions, P₀ and P₁ are the initial and final pressures (MPa), T₀, and T₁ are the initial and final reactor temperatures (K), and V_g is the volume of the gas phase in the cell, which can be determined by the volume of the sediment layer, the volume of the sediment void, and the volume of the reactor. The volume expansion caused by the phase transitions is negligible. Therefore, it can be expressed by Equation (2):

$$V_{\mathrm{g}}=V_t-V_s-V_w+V_k,$$

(2)

where V_t is the effective volume of the reactor (mL), V_s is the volume of sand added to the reactor (mL), V_w is the volume occupied by the residual water (mL), and V_k is the deposition void volume (mL), which can be expressed by Equation (3):

$$V_k=V_s·\phi ,$$

(3)

where φ is the initial porosity of natural sand with the corresponding particle size (%). The saturation of methane hydrate can be described by Equation (4):

$$S_h=\frac{N_H{D}_n·M_w}{{\rho }_H·V_k},$$

(4)

where N_H is the hydration number of methane. In this study, we chose 6.0 as the hydration number of methane. M_w is the molar mass of water (g/mol), and ρ_H is the density of methane hydrate, 0.925 g/cm³.

2.4.2. Recovery Efficiency of CH₄

The recovery efficiency(η) of CH₄ is calculated using Equation (5):

$$\eta =(n_{CH}-n_{CH,0})/∆n,$$

(5)

where n_CH is the number of moles of methane in the gas phase in the reactor (mol), n_CH,0 is the number of moles of methane in the gas phase of the reactor at zero time (mol), and n_CH is determined by Equation (6):

$$n_{CH}=x_{CH}{\mathrm{Z}}_{CH}P{V}_g/\left(RT\right),$$

(6)

where x_CH is the mole fraction of methane in the gas phase in the reactor, which is determined after analyzing the gas sample by gas chromatography; P and T represent the temperature and pressure inside the reactor one second before the gas takes, respectively. Z_CH is the compression factor of CH₄. It can be calculated using the P-R equation.

2.4.3. Weight of Residual Water in the Sediment

The weight of residual water (w_r) in the sediment can be calculated using Equation (7):

$$w_r=w_0(1-C_w),$$

(7)

where w₀ is the weight of the water added to the sediment (g) and C_w is the water conversion ratio, which can be described by Equation (8):

$$C_w=\frac{N_H·∆n·M_w}{w_0}.$$

(8)

2.4.4. CO₂ Storage Efficiency

The moles of carbon dioxide in the gas phase of the reactor can be calculated using Equation (9):

$$n_{CO}=x_{CO}{Z}_{CO}P{V}_g/\left(RT\right),$$

(9)

where x_CO is the mole fraction of carbon dioxide in the gas phase in the reactor, which is determined after analyzing the gas sample by gas chromatography; P and T represent the temperature and pressure inside the reactor one second before the gas takes, respectively. Z_CO is the compression factor of CO₂. It can be calculated using the P-R equation.

The moles of carbon dioxide sequestered in sediments can be expressed by Equation (10):

$$∆n_{CO}=n_{CO,0}-n_{CO},$$

(10)

where n_CO,₀ is the number of moles of carbon dioxide gas in the reactor at time zero.

The CO₂ storage efficiency(ε) is defined as the ratio of the moles of CO₂ consumed to the total moles of CO₂ injected, which can be calculated using Equation (11):

$$\epsilon =∆n_{CO}/n_{CO}.$$

(11)

3. Results and Discussion

3.1. Variation in Pressure during the Replacement Process

The variation in pressure inside the reactor for different particle sizes with time during the replacement reaction is shown in Figure 3.

In the initial stage (the first 10 h), the pressure in the reactor drops rapidly in Experiments 2–4 because the rate of CO₂ hydrate formation between the injected CO₂ and the residual water in the sediment is much greater than the rate of replacement of CH₄; that is, the consumption of CO₂ is greater than the generation of CH₄, resulting in a decrease in the total pressure in the reactor. On the contrary, it also shows that in the replacement reaction, the priority of the formation of CO₂ hydrate in Experiments 2–4 is higher than that of CO₂ replacement of CH₄ hydrate [29].

The variation in pressure in Experiment 1 is opposite to that of Experiments 2–4. In the initial stage, the pressure in the reactor in Experiment 1 increased rapidly. This was because the amount of CH₄ gas replaced during the reaction in the first 10 h is greater than the sum of the CO₂ gas consumed in the replacement reaction and the CO₂ gas consumed in the generation of CO₂ hydrate with water, so the total pressure in the reactor tends to rise. The reason is that the specific surface area of the porous medium is larger when the particle size is smaller, and the area of the CO₂ replacement reaction surface is larger, which means the effective area of the replacement reaction is larger, so the amount of CH₄ replaced will be more. The particle sizes used in Experiment 1 was the smallest among the four experiments, which has the largest effective area of the replacement reaction [30]. When CO₂ is injected, the replacement reaction occurs rapidly, and the reaction speed is faster than that of CO₂ hydrate formation, which eventually leads to a pressure increase in the initial stage of the reaction.

In the later stage, the pressure of Experiments 1–4 showed a slow pressure drop, which can be explained as follows. First, during the replacement process, both the CO₂-CH₄ replacement reaction on the sediment surface, and the reaction of the residual water with CO₂ to form CO₂ hydrate, occurred at the same time [31]. Both reactions form a dense CO₂ hydrate film on the sediment surface [24]. As the thickness of the CO₂ hydrate layer gradually increased, the reaction became more difficult because of the increased diffusion barrier of CO₂ molecules into the CH₄ hydrate in the sediment and CH₄ molecules passing through the CO₂ hydrate layer. Second, as the replacement process proceeds, the fugacity difference between CO₂ and CH₄ in the gas and hydrate phases decreases [20], which means that the driving force of the replacement reaction decreases; thus, the reaction becomes slower. Thus, the diffusivity of gas molecules in the sediment plays a crucial role in the displacement reaction.

The pressure curves in the reactor over time at different initial hydrate saturations are shown in Figure 4.

Figure 4 shows that the pressure changes during the displacement process in Experiments 5–8 dropped sharply in the initial stage of the reaction and decreased slowly in the later stage. It also shows that the replacement reaction mainly occurred on the surface of the sediment. The CO₂ hydrate layer formed in the initial stage was thinner and the reaction speed was faster, resulting in a large pressure drop. However, as the reaction progressed, the CO₂ hydrate film formed around the sediment surrounded the entire sediment surface, and its thickness continued to increase, which eventually led to an increase in the channel resistance for CO₂ molecules to enter the sediment to participate in the reaction; thus, the pressure curve in the later period did not decrease significantly.

3.2. CO₂ Storage Efficiency

Figure 5 and Figure 6 show the storage efficiency of CO₂ in the replacement reactions of Experiments 1–4 and Experiments 5–8, respectively.

As the residual water quality was approximately equal in all experiments, it can be assumed that the amount of CO₂ dissolved in the residual water was the same. Therefore, the influence of residual water on the results was consistent. As shown in Figure 5, there is little difference in the storage efficiency of CO₂ for different particle sizes, and the smaller the particle size, the greater the storage efficiency. There are two possible explanations: one is that the smaller the particle size, the greater the porosity, and the relatively smaller the resistance for CO₂ molecules to enter the sediment to participate in the reaction [32]; second, the smaller the particle size, the larger the specific surface area, and with more CO₂ hydrates generated on the surface, more CO₂ needs to be consumed. As a result, the storage efficiency would increase.

Figure 6 shows that the CO₂ storage efficiency of Experiments 5–7 is not significantly different, because the formation of CH₄ hydrate in the sediment is grain-coating, and the CO₂ consumption at this time is only affected by the hydrate saturation. As the hydrate saturation decreases, the permeability of the sediment increases, which is conducive to the diffusion of CO₂ into the sediment. However, when the hydrate saturation is greater than 35%, CH₄ hydrate exists in the form of pore-filling [33,34], whose permeability is greatly increased compared with grain-coating; thus, the CO₂ storage rate in Experiment 8 is significantly increased. However, it also shows that the influence of hydrate morphology difference on gas permeability is greater than that of hydrate saturation on gas permeability.

3.3. Variation in CO₂ and CH₄ in the Gas Phase

Figure 7 and Figure 8 show the variation in the total gas content in the reactor and the gas phase, respectively, during the replacement process of Experiments 1–4.

Calculated by Equation (1), the reduction in gas in the reactors for Experiments 1–4 were 0.0118, 0.0286, 0.0268, and 0.0256 mol, respectively. These values represent the amount of gas consumed by the newly formed CO₂ hydrates during the replacement process. The variation in the total gas content in the reactor of Experiment 1 was the smallest within the same replacement time, but the content of the gas component CH₄ increased the most, reaching 0.042 mol. From this, it can be concluded that the cumulative replacement of CH₄ and the consumed CO₂ are not significantly different, which means that CO₂ alone is less involved in the process of forming CO₂ hydrate and mainly participates in the process of replacing CH₄, which is beneficial to replacement mining.

The gas content in Experiments 2–4 changed significantly with time, indicating that CO₂ was mainly involved in the formation of CO₂ hydrate, while the CO₂-CH₄ replacement reaction was not dominant, so the cumulative replacement amount of CH₄ was less than that in Experiment 1. The replaced CH₄ in Experiments 2–4 was 0.022, 0.017, and 0.015 mol, respectively. Under the same conditions, the replaced CH₄ increased with a decrease in the sediment particle size.

This can be explained as follows. First, owing to the formation of vein-like hydrates in fine sediments [35,36], the effective area of the interface between CO₂ gas and hydrates is large, while the hydrates formed by coarse sediments are pore-filled hydrates, and the effective area of the reaction is small. Second, with smaller particle size, there is greater permeability, making it easier for CO₂ to enter the sediment for replacement reactions.

Through the comparison and comprehensive analysis of Figure 7 and Figure 8, it can be concluded that the reaction shows a trend of decreasing total gas content as the reaction proceeds, during which CO₂ is decreasing, CH₄ is increasing, and the decrease in CO₂ is greater than the increase in CH₄. Therefore, in the replacement reaction, CO₂ does not simply replace CH₄ in equal amounts but also forms new CO₂ hydrates due to the combination of CO₂ and residual water.

Figure 9 shows the variation in the total gas content in the reactor over time during the replacement process of Experiments 5–8.

Calculated by Equation (1), the reductions in gas in the reactor for Experiments 5–8 are 0.038, 0.0256, 0.0183, and 0.0321 mol, respectively, which is the amount of gas consumed by the newly formed CO₂ hydrate during the replacement process. As shown in Figure 9, with an increase in the initial hydrate saturation, the reduction range of the total gas content in the reactor first decreased, and when the initial hydrate saturation reached a certain value, the reduction range of the total gas content increased. This can be explained by the fact that hydrate saturation and morphology control the CO₂ flow into the sediment and the effective reaction area. This is because they directly control the relative gas permeability in CH₄ hydrate-containing sediments. The difference in the hydrate morphology affects the relative permeability of the gas and the contact interface of the CO₂ gas hydrates. According to experimental records [34], when the initial hydrate saturation is less than 35%, CH₄ hydrate exists in the form of a grain-coating (mainly nucleated on the sediment surface; it grows along the sediment surface [37]). When the initial hydrate saturation is greater than 35%, and CH₄ hydrate exists in the form of pore-filling (mainly nucleating in the pores between the sediments [35,38], hydrate crystals are suspended in the pores and not in contact with the sediments, and there is no interaction). Based on this, it can be inferred that the hydrate formed in Experiments 5–7 is grain-coated, and the hydrate structure has the same effect on the permeability, so the permeability of the sediment will be determined by the hydrate saturation. In Experiment 5–7, the hydrate saturation gradually increases, so the permeability of the sediment is gradually decreasing [39]. This hinders the occurrence of the replacement reaction, and finally leads to a decrease in the reduction range of the total gas content in the reactor. Although the hydrate saturation in Experiment 8 was the highest, the gas relative permeability in the sediment was improved due to the formation of pore-filling hydrate, so the final gas reduction increased.

Figure 10 shows the variation in the gas component content in Experiments 5–8 with time.

Calculated from Equation (1), the accumulated amounts of CH₄ replaced in Experiments 5–8 were 0.007, 0.015, 0.024, and 0.028 mol, respectively. It can be concluded that under the same conditions, with an increase in the initial hydrate saturation, more CH₄ was replaced. This is because with higher hydrate saturation, there are more CH₄ hydrates in the sediment, which results in a larger interface area between CO₂ gas and the hydrate. This implies that there is a larger effective area for the reaction. The results demonstrate that in addition to the exchange of CO₂ and CH₄ molecules, the replacement process will also be accompanied by the process of CO₂ combined with residual water to form new CO₂ hydrates. Compared to the exchange of CH₄ molecules, CO₂ molecules form CO₂ hydrates more easily at the gas-liquid interface.

3.4. CH₄ Recovery Efficiency

The calculated recovery efficiency and replacement amount of CH₄ for Experiments 1–4 is shown in Figure 11.

The replacement amount and recovery efficiency of CH₄ increases with a decrease in particle size. However, the median particle sizes of 166, 279, and 377 μm had little effect on the recovery efficiency, whereas the recovery efficiency of the median particle size of 101 μm was much greater than that of other particle sizes. A possible reason for this is that according to the microscopic distribution of hydrate formation, CH₄ forms pore-filling hydrates in sediments with larger particle sizes, whereas it forms vein-like hydrates in sediments with smaller particle sizes. The surface area of vein-like hydrates was larger than that of pore-filling hydrates, and the interface area between CO₂ gas and CH₄ hydrates was larger during the replacement process. In addition, the porosity and permeability of the sediment increased as the particle size decreased, and the diffusion of CO₂ gas in the sediment improved accordingly. Therefore, the replacement and recovery efficiencies of CH₄ were higher.

The calculated recovery efficiency of CH₄ hydrate for Experiments 5–8 is shown in Figure 12.

The replacement of CH₄ increases with an increase in the initial hydrate saturation, whereas the recovery efficiency decreases with an increase in the initial water saturation. It is possible that the recovery efficiency is mainly controlled by CO₂ permeation into the interface between CH₄ and CO₂ hydrate. The CH₄ hydrate attached to the surface of the sediment increases with an increase in the initial hydrate saturation; therefore, the higher the hydrate saturation, the more CH₄ is replaced. However, as the reaction progresses, the hydrate film layer becomes thicker, and the resistance of CO₂ molecules to penetrate the sediment and participate in the replacement reaction increases, thereby reducing the recovery efficiency. When the hydrate saturation was greater than 35%, the amount of replaced CH₄ was more significantly affected by hydrate saturation. When the hydrate saturation was less than 35%, the amount of replaced CH₄ was less affected by hydrate saturation. This is because when the hydrate saturation is greater than 35%, CH₄ hydrate exists in the form of pore-filling, whereas when the hydrate saturation is less than 35%, it exists in the form of grain-coating. The permeability of pore-filling is known to be greater than that of grain-coating [35]. Therefore, it is easier for CO₂ to permeate the sediment and react to replace more CH₄.

4. Conclusions

Based on the consistency of the amount of residual water for all the experiments, the effects of sediment particle size and initial hydrate saturation on CO₂-CH₄ displacement were discussed using a self-built displacement experimental device. Eight experiments were conducted to examine the effect of displacement under different conditions. The experimental results show that during the replacement process, two stages of rapid reaction and slow reaction occurred, which were manifested by the speed of the pressure change in the reactor. The CO₂ sequestration ratio decreased with the increase in sediment particle size and increased with the increase in initial hydrate saturation. During the replacement process, two reactions occurred: CH₄ was replaced by CO₂ and CO₂ hydrate was formed, and the replacement amount and recovery efficiency of CH₄ increased with a decrease in sediment particle size. When the sediment particle size was less than 166 μm, the CH₄ recovery efficiency was significantly affected by the particle size. When the particle size was greater than 166 μm, the CH₄ recovery efficiency was not significantly different. The replacement amount of CH₄ increased with the increase in initial hydrate saturation, while the recovery efficiency decreased. When the initial water saturation in the sediment was greater than 35%, the recovery of CH₄ was significantly affected by the initial hydrate saturation. When the initial water saturation in the sediment was less than 35%, the CH₄ recovery was less strongly affected by the initial hydrate saturation.

In conclusion, in CO₂-CH₄ replacement production, selecting a storage area with high hydrate saturation, small particle size, and water saturation greater than 35%, will be beneficial to the CO₂ storage efficiency and the amount of CH₄ replaced.