Carbon Capture, Utilization, and Storage (CCUS) processes have gained significant momentum in recent years for reducing carbon dioxide (CO₂) content in the Earth’s atmosphere. Carbon storage is realized by injecting CO₂ into geological structures, mainly depleted oil reservoirs [1]. This process takes advantage of using the pre-existing wells in Enhanced Oil Recovery (EOR) projects to inject CO₂ for cost reduction. Another advantage of using depleted oil reservoirs to store CO₂ is the high injectivity of CO₂ into oil reservoirs. A critical concern of using EOR wells as CO₂ receivers is the risk of CO₂ leakage through EOR wellbores due to the attack of CO₂ on the well cement sheath [2][3]. The leakage is promoted by the high mobility (low viscosity) of supercritical CO₂ in high-temperature conditions that prevail in EOR wells. It has been practiced to inject CO₂ into sub-sea water zones where CO₂ should exist in liquid form in the low-temperature environment. This should significantly increase CO₂ viscosity and reduce the risk of CO₂ leakage [4]. Field-scale CO₂ injection projects include the Sleipnir project in the Norwegian Sea [5], the Hokkaido project in Japan [6][7] and the Gulf Coast project in the United States [8]. Simulation studies have shown that the possibility of CO₂ leakage from the sub-sea water zones is very low [9]. However, CO₂ injection into the sub-sea water zones has a common issue of low injectivity due to the low-permeability nature of the water zones. To further reduce the risk of CO₂ leakage from its storage reservoirs, CO₂ storage in its hydrate form has been considered.

Small-molecule gases (e.g., methane, ethane, propane, iso-butane, CO₂, and H₂S) are trapped inside the cages of water molecules, forming gas hydrates under high-pressure and low-temperature conditions. The molecular size of these hydrophobic gases and volatile organic compounds is 3.8 Ȧ to 9.2 Ȧ. In appearance, the hydrates resemble snow or loose ice [10].

Injecting CO₂ into permeability zones where the in-situ temperatures are lower than the temperature required for hydrate formation allows for CO₂ storage in hydrate form, eliminating the possibility of CO₂ leaking through wellbores. This idea has been investigated in previous studies [11][12]. However, it takes a lengthy time for CO₂ molecules to fully contact water molecules to form CO₂ hydrates due to the limited rate of mass transfer. In fact, this has been done in methane hydrate reservoirs for more than a decade in the CH₄-CO₂ swapping processes. The unsolved issue is that the CO₂ swapping rate is limited by the heat-transfer efficiency. Using geothermal energy to improve heat transfer efficiency may be a viable solution [13]. However, this idea is still in the stage of theoretical investigations.

The first common disadvantage of CO₂ carbon storage in geological structures, either depleted oil reservoirs or low-temperature water zones, is that wellbores are required for CO₂ injection, which increases the cost of operations. The second common disadvantage is the low capacity of storage due to the limited reservoir pore volumes. These issues have brought back the idea of depositing CO₂ directly onto sea/ocean floors [14]. If CO₂ is mixed with seawater near the sea floor in conditions that are favorable for hydrate formation, CO₂-hydrates should form and theoretically sink to the bottom of the sea/ocean by virtue of the relatively high density of the hydrate.

Guo et al. [15] presented an analytical method for determining the minimum seawater depth required to dispose of CO₂ in hydrate form on the seafloor. Amponsah [16] determined the minimum CO₂-storage seawater depths in the major seas around the world and the Arctic Ocean. His work shows that the minimum required seawater depth varies greatly from 200 m to 650 m. The Persian Gulf is too shallow for CO₂-storage in hydrate form due to its high temperature. The minimum CO₂-storage seawater depth in the Arctic Ocean is about 120 m.

The CO₂ storage in hydrate form on the seafloor requires that the density of hydrates be greater than that of seawater. It is known that 1 m³ of pure hydrates contains approximately 170 sm³ of CO₂ depending on the formation process. This makes CO₂-hydrates heavier than seawater. However, free CO₂ (bubbles) can be trapped inside the bulk hydrates due to the excessive proportion of CO₂ during the chemical reaction in non-equilibrium conditions, making the density of bulk hydrates less than that of seawater [17]. Guo and Mahmood [18] proposed a jet-cooling method to accelerate the formation of CO₂-hydrates. Although the jet-cooling technique can generate CO₂ hydrates quickly, the generated bulk hydrates may trap CO₂ bubbles, making the bulk hydrates float and unable to settle down to the seafloor. On the other hand, free water may be trapped inside the bulk hydrates due to the excessive proportion of water during the chemical reaction in non-equilibrium conditions, making the density of bulk hydrates greater than that of seawater. Therefore, it is necessary to study the hydrate-forming processes to find a reliable and reproducible method to produce heavier-than-water CO₂ hydrates for CO₂ deposition onto seafloors.

CO₂-hydrate formation under various conditions was observed through a windowed reactor shown in Figure 1.The reactor has see-through windows made of sapphire glass at both ends. The glass has a design strength of 400 MPa at 20˚C and 275 MPa at 500˚C, with a Young’s modulus of 345 GPa and Poisson’s ratio of 0.27. Its flexural strength (modulus of rupture) is 0.9 GPa to 1.9 GPa, with a Knoop hardness of 1525 - 2000. The reactor was designed to handle a stress (safe working stress) up to 150 MPa (22,000 psi) with a safety margin. The reactor’s 3-inch diameter, 0.29 inch (7.4 mm) thick, sapphire glass plate can hold up to 2000 psi internal pressure. The reactor was water-pressure-tested up to 2000 psi for 24 hours with no leaks identified.

The whole observation system is illustrated in Figure 2 with the windowed reactor drawn at the center. The ISCO metering pump provides water to the reactor through ports either at the top or bottom. The CO₂ bottle feeds the Accumulator with CO₂ gas up to 1000 psi. The ISCO metering pump squeezes water to the Accumulator so that CO₂ is displaced through a CO₂ Condenser into the reactor through ports either at the top or bottom. The Cooler with Pump circulates coolant through the coils around the reactor covered by an insulation layer. The backpressure regulator controls the pressure in the reactor. Pressure and temperature inside the reactor are measured by the pressure gauge (PG), pressure transducer (PT), and temperature transducer (TT). These parameter values are stored and analyzed in the Computer.

Figure 1. Image of a reactor with see-through windows at both ends.

Figure 2. Schematic of the observation system.

CO₂ hydrate formation was observed in two mixing procedures: 1) CO₂-bubbling into the water phase and 2) water-dropping into the CO₂ phase. The procedure for CO₂-bubbling into the water phase is outlined as follows:

1) Fill the reactor with water and pressurize it to the desired pressure level using the ISCO metering pump.

2) Fill the upper section of the Accumulator with CO₂ using the CO₂ bottle.

3) Lower the reactor temperature using the Cooler with Pump until it reaches the desired temperature level.

4) Use the ISCO Metering Pumper to displace the CO₂ in the Accumulator through the CO₂ Condenser into the reactor through a 1/16 bottom port at 5 ml/min.

5) Observe CO₂-hydrate formation at different CO₂ injection rates through the reactor windows.

The procedure for water-dropping into the CO₂ phase is outlined as follows:

1) Fill the upper section of the Accumulator with CO₂ using the CO₂ bottle.

2) Use the ISCO Metering Pumper to displace the CO₂ in the Accumulator through the CO₂ Condenser into the reactor through a bottom port until it reaches the desired pressure level.

3) Lower the reactor temperature using the Cooler with Pump until it reaches the desired temperature level.

4) Inject water at a 5 ml/min flow rate into the reactor through a 1/16 top port using the ISCO metering pump.

5) Observe CO₂-hydrate formation at different water injection rates through the reactor windows.

Amponsah [16] determined the minimum water depths required for CO₂-storage in hydrates in 6 seas around the world and the Arctic Ocean. His results show that the minimum required water depth varies from 120 m in the Arctic Ocean to 650 m in the Mediterranean Sea. The temperatures at the minimum required water depth range from 0˚C in the Arctic Ocean to 11˚C in the Gulf of Mexico.

Figure 3 reproduces the static conditions for forming CO₂ hydrates [18]. It shows that, in the temperature range between 273.15 K (0˚C) and 284.15 K (11˚C) and the pressure range between 3.3 MPa (485 psi) and 5.4 MPa (800 psi), pure CO₂ exists in liquid form and liquid CO₂ can form hydrates when water is present. CO₂ hydrate formation was investigated in the temperature range from 0˚C to 3˚C and in the pressure range from 500 psi to 800 psi. Therefore, the observed CO₂ in this study is CO₂ liquid.

Figure 3. Static conditions for forming CO₂ hydrates [18].

Figure 4 shows two images of CO₂ hydrate formation during CO₂-bubbling at 0.4˚C and 800 psi. The left image was captured when CO₂ entered the reactor in the form of bubbles at the bottom of the water phase. The right image was captured when hydrates formed around the CO₂ bubbles before the hydrate-shelled bubbles floated up to the top of the water phase. The CO₂ bubbles with hydrate-shells accumulated on the top side of the water phase, indicating that the bulk hydrates (pack of hydrate-shelled CO₂ bubbles) have a bulk density less than the water density. This is because the bulk hydrates contain liquid CO₂ as an excess phase.

Figure 4. Images of CO₂ hydrate formation in CO₂-bubbling at 0.4˚C and 800 psi: (a) CO₂ enters the reactor in the form of bubbles, (b) hydrates form around the CO₂ bubbles before floating up.

Figure 5 shows two images of CO₂ hydrate formation during CO₂-bubbling at 2˚C and 800 psi. The left image was captured when a CO₂ bubble rose in the water phase. The right image demonstrates accumulations of CO₂ bubbles and hydrate-shelled CO₂ bubbles at the top of the water phase. The accumulation of hydrate-shelled CO₂ bubbles on the top side of the water phase implies that the bulk hydrates (a pack of hydrate-shelled CO₂ bubbles) have a bulk density less than the water density. Again, this is because the bulk hydrates contain liquid CO₂ as an excess phase.

Figure 5. Images of CO₂ hydrate formation in CO₂-bubbling at 2˚C and 800 psi: (a) a CO₂ bubble floating up in the water phase, (b) accumulation of CO₂ bubbles and hydrate-shelled CO₂ bubbles.

Figure 6 presents two images of CO₂ hydrate decomposition when the reactor pressure was reduced from 800 psi to 500 psi at 2˚C (the equilibrium pressure is 515 psi). The left image shows the collapse of hydrate-shelled CO₂ bubbles. The right image shows collapsed hydrate shells floating at the interface between liquid CO₂ and water phases, suggesting that the residual hydrates still trapped liquid CO₂.

Figure 6. Images of CO₂ hydrate decomposition with a pressure reduction from 800 psi to 500 psi at 2˚C: (a) collapse of hydrate-shelled CO₂ bubbles, (b) collapsed hydrate shells.

Figure 7 presents two images of CO₂ hydrate formation during continuous CO₂ injection at 2˚C and 800 psi. The image on the left shows a hydrate-coated CO₂ channel at a 5 mL/min CO₂ flow rate. The image on the right shows a hydrate-coated CO₂ channel at a 10 mL/min CO₂ flow rate. It is seen that a straighter CO₂ channel formed at a higher CO₂ flow rate.

Figure 7. Images of CO₂ hydrate formation during CO₂-injection at 2^°C and 800 psi: (a) hydrate-coated CO₂ channel at 5 mL/min CO₂ flow rate, (b) hydrate-coated CO₂ channel at 10 mL/min CO₂ flow rate.

Figure 8 shows two images of CO₂ hydrate formation during water dropping at 0˚C and 500 psi. The image on the left indicates that a clear water pocket remained at the bottom initially. The image on the right shows CO₂-hydrate crystals at the bottom of the reactor after 1 hour of growth, indicating that the hydrate crystals have a density greater than that of water.

Figure 8. Images of CO₂ hydrate formation during water dropping at 0˚C and 500 psi: (a) the water pocket stayed at the bottom of the CO₂ liquid at the beginning, (b) CO₂-hydrate crystals at the bottom side of the reactor after 1 hour.

Figure 9 shows two images of CO₂ hydrate formation after water dropping at 0˚C and 500 psi. The image on the left indicates CO₂-hydrate crystals grown from the water pocket after 4 hours. The image on the right shows CO₂-hydrate crystals everywhere in the reactor after 24 hours.

Figure 9. Images of CO₂ hydrate formation from water dropping at 0˚C and 500 psi: (a) after 4 hours, (b) after 24 hours.

Figure 10 presents two images of CO₂ hydrate formation from water dropping at 2˚C and 500 psi. The image on the left indicates the initial water pocket at the bottom. The image on the right shows CO₂-hydrate crystals remaining at the bottom after 3.45 hours.

Figure 11 shows two images of CO₂ hydrate formation from water dropping at 3˚C and 500 psi. The image on the left indicates the initial water pocket at the bottom. The image on the right shows a hydrate-shelled water pocket staying at the bottom after 12 hours. The slow formation of CO₂-hydrate is because this condition is very close to the equilibrium condition indicated by Figure 3. However, as shown in Figure 12, a large amount of hydrate crystals was found after 34 hours.

In summary, although the density of pure CO₂-hydrate is theoretically greater than that of water, the observed floating behavior of the bulk hydrates formed during CO₂-bubbling suggests that the density of the bulk hydrates is less than that of water. This is because free CO₂ is trapped inside the bulk hydrates and outside the hydrate crystal structure. On the other hand, the observed settling behavior of the bulk hydrates formed during water-dropping suggests that the density of the bulk hydrates is greater than that of water. This is because no free CO₂ is trapped inside the bulk hydrates and outside the hydrate crystal structure. Either some free water or no water is trapped inside the bulk hydrates, which can cause the settling behavior.

Figure 10. Images of CO₂ hydrate formation during water dropping at 2˚C and 500 psi: (a) water pocket stayed at the bottom of CO₂ liquid at the beginning, (b) CO₂-hydrate crystals at the bottom side of the reactor after 3.45 hours.

Figure 11. Images of CO₂ hydrate formation during water dropping at 3˚C and 500 psi (near the equilibrium point): (a) a water pocket stayed at the bottom of CO₂ liquid at the beginning, (b) a hydrate-shelled water pocket stayed at the bottom side of the reactor after 12 hours.

Figure 12. Images of CO₂ hydrate formation from water dropping at 3˚C and 500 psi (near the equilibrium point), CO₂-hydrate crystals on the bottom side of the reactor after 34 hours.

Discharging carbon dioxide (CO₂) to seafloors in the form of CO₂-hydrates can potentially store CO₂ in virtually unlimited quantities. Realizing the process requires a thorough understanding of CO₂-hydrate forming behavior in seawater conditions. One of the major concerns is the gravitational stability of the produced bulk CO₂-hydrates. The objective of this study was to seek the optimum procedure for producing bulk CO₂-hydrates that would stay at the bottom of the water due to gravity. This objective was achieved by direct visualization of CO₂-hydrates through the glass windows of a reactor under various pressure and temperature conditions with mixing methods of CO₂-bubbling and water-dropping. The following conclusions are drawn.

1) CO₂ bubbling into the water phase initiates CO₂ hydrates at the surfaces of bubbles, creating bubble shells. The bubble shells reduce further contact between CO₂ and water, hindering the growth of hydrates. The hydrate-shelled CO₂ bubbles move upward due to buoyancy.

2) Reducing pressure causes the collapse of bubble shells. However, the collapsed shells still stay at the interface of water and CO₂, indicating that the collapsed shells are CO₂-rich, i.e., there are still free CO₂ molecules outside of the hydrate structures.

3) Continuous injection of CO₂ to the bottom of the water phase can create hydrate-coated CO₂ channels. The tortuosity of the channels decreases as the CO₂ flow rate increases. Further studies are needed to determine the critical condition for the change from CO₂ bubbling to CO₂ channeling.

4) Water-dropping into CO₂ liquid initiates CO₂-hydrates at the surfaces of water droplets, creating hydrate-coated water droplets. The hydrate crystals grow outside the hydrate coatings. The bulk hydrates stay at the bottom of the water phase, indicating the gravitational stability of the bulk hydrate.

5) To produce gravitationally stable CO₂-hydrates for deposition on seafloors, CO₂-hydrates should be generated using the process of water spraying into the CO₂ phase, not the process of CO₂-injection into the water phase.

The authors are grateful to the administration of the Energy Institute of Louisiana at the University of Louisiana at Lafayette for its administrative assistance throughout this research.

This research was supported by the Louisiana Board of Regents Support Fund (BoRSF), Grant No. LEQSF (2024-27)-RD-B-04.

Muhammad Towhidul Islam—Experimental investigations. Boyun Guo—Manuscript writing and resources.

The data supporting the findings of this study are included within the article.