*3.1. Liquid CO2 Injection Process*

When the CO2 liquid was injected from the bottom of the cell, a CO2 liquid plume moved upward because of a buoyancy force. Eventually, the CO2 liquid front reached the CO2 hydrate stability zone, and then CO2 hydrates formed. The CO2 hydrate-bearing sediment layer then obstructed the upward flow of the CO2 liquid. While water and CO2 were consumed in the CO2 hydrate formation process in the hydrate stability zone, the CO2 hydrate-bearing sediment layer prevented CO2 supply. Thus, a pressure difference between the upper and lower part of the cell appeared. The P-wave velocity and electrical resistance monitoring results indicated the formation of CO2 hydrates and the blockage of the CO2 flow. Detailed experimental results are shown in the following sections.

#### 3.1.1. Temperature and Pressure

Figure 4 shows the pressure of each layer in the cell over time. The CO2 liquid was injected at 1300 min after data logging started. The pressure in each layer was scattered until 5000 min because of a difference in pressure between the injection pressure and backpressure (i.e., ~0.6 MPa, refer to Section 2.3.2), and volume change of pore fluids due to CO2 dissolution into pore water. The pressure in each layer was very similar to one another because the pore space was well connected throughout the sample. CO2 hydrates started to form when the injected CO2 reached the CO2 hydrate stability zone (herein, between layers A3 and A4; Figure 3). Then, the pressure in layers A1, A2, and A3 rapidly dropped to 3.3 MPa after 5000 min while the pressure of layers A4 and A5 were nearly constant and identical to the injection pressure. The difference in pressure between the upper and lower part of the cell was induced by the sealing (i.e., pore clogging) effect of the CO2 hydrate bearing-sediments layer.

**Figure 4.** Pressure of the cell with lapsed time during CO2 injection.

The sealing capacity of the CO2 hydrate-bearing sediment layer gradually increased during the growth of CO2 hydrates in the pore space of the sample. Meanwhile, water and CO2 were consumed during CO2 hydrate formation. For a constant-volume process, the consumption of CO2 and water during CO2 hydrate formation leads to a pressure decrease because the molar volume of the CO2 hydrates is smaller than the original molar volume of the consumed fluids. For the lower part of the cell (represented by layers A4 and A5) the pressure was preserved because CO2 was supplied continuously from the bottom of the cell during the experiment. For the upper part of the cell (represented by layers A1, A2, and A3), however, the pressure decreased because the CO2 hydrate-bearing sediment layer prevented the CO2 to be supplied from the lower part of the cell.

Figure 5 shows the pressure–temperature evolution during the CO2 injection test. Note that layers A2 and A3 were in the CO2 hydrate stable condition while the others were not. The pressure of the upper part of the cell (i.e., layers A1, A2, and A3) decreased when the sealing capacity of the CO2 hydrate-bearing sediment layer increased to a level that prevented flow. Note that there were no CO2 hydrates in layer A2 because the injected CO2 did not reach it, even though this layer is in the CO2 hydrate stability zone. Meanwhile, the lower part of the cell (i.e., layers A4 and A5) maintained a constant pressure level. Then, the pressure of the upper part of the cell gradually increased. There are two mechanisms for the pressure recovery of the upper part of the cell: (1) The uppermost CO2 hydrates dissociated with the pressure decrease. Therefore, pressure was recovered restrictively via emitted CO2 and water from the CO2 hydrates, and (2) CO2 hydrate saturation is limited by capillary pressure, which is determined by the pore size [23].

**Figure 5.** The pressure–temperature relationship during CO2 injection. (**a**) 4680–8100 min, (**b**) 8100– 25,300 min.

In the central part of the CO2 hydrate-bearing sediment layer, CO2 hydrates grew until the CO2 hydrate saturation reached maximum CO2 hydrate saturation. Thus, the consumption of CO2 and water diminished because any additional formation of CO2 hydrates was restricted. Finally, the pressure of the upper part of the cell increased to 5.2–5.5 MPa. However, the pressure of the upper part of the cell was still lower than that of the lower part of the cell. The repetitive ascending and descending pressure could be due to the continuous repetition of the CO2 formation and dissociation process at the CO2 hydrates front.

#### 3.1.2. P-Wave Velocity

Figure 6 shows the results of the P-wave velocity measurements during the CO2 injection process. Before the water injection process, the cell was partially saturated, and the P-wave velocity was about 900 m/s. When the sediment sample was saturated by distilled water, the P-wave velocity of all the layers was about 1600 m/s. Then, when the CO2 liquid was injected, the P-wave velocities of the lower part of the cell (i.e., layers A4 and A5) decreased because the bulk modulus of the CO2 liquid was much lower than that of the water [24,25]. Meanwhile, the P-wave velocity of layer A3 gradually increased because of the stiffening effect induced by CO2 hydrate formation. This P-wave velocity increase in layer A3 indicates that the CO2 hydrate bearing-sediment layer is between layers A3 and A4. On the other hand, the P-wave velocity of layer A1 and A2 did not change during CO2 injection. This is evidence that the CO2 hydrate-bearing sediment layer prevents any upward flow of the CO2 liquid.

**Figure 6.** P-wave velocity of the unconsolidated sediment sample during CO2 injection. (**a**) layer A1, (**b**) layer A2, (**c**) layer A3, (**d**) layer A4, (**e**) layer A5. P-wave velocity of layers A1 and A2 did not change during CO2 injection because the CO2 hydrate-bearing sediment layer prevented the upward flow of CO2 liquid.

#### 3.1.3. Electrical Resistance

Figure 7 shows the normalized electrical resistance (*R*/*R0*) during the CO2 injection process, where *R* is the measured electrical resistance and *R0* is the initial electrical resistance of the distilled water-saturated sample at each layer (i.e., layers B1–B4). For the lower part of the cell (i.e., layers B3 and B4), the electrical resistance decreased with CO2 injection because of the dissolution of CO2.

**Figure 7.** Normalized electrical resistance *R*/*R0* of the unconsolidated sediment sample during CO2 injection. (**a**) layer B1, (**b**) layer B2, (**c**) layer B3, (**d**) layer B4.

Electrical resistance increased in-situ in the marine sediments because the conductive pore water (i.e., brine) was replaced by CO2, which is a nonpolar molecule. This electrical resistance increase induced by the CO2 replacement was weakened by the dissolution of CO2 and the surface effect of the mineral grains [26]. For typical brine, the effect of CO2 dissolution on electrical resistance is negligible because the concentration of salt (i.e., NaCl) is much larger than the ionic concentration increased by CO2 dissolution [27]. However, if the salt concentration is low (e.g., onshore sediments), electrical resistance of in-situ sediments can decrease during the CO2 permeation [28]. In this experiment, the effect of dissolved CO2 was dominant on electrical resistance because the pore water was distilled.

Meanwhile, the electrical resistance of the upper part of the cell (i.e., layers B1 and B2) showed minor changes during CO2 injection. This is additional evidence demonstrating that CO2 liquid did not reach the upper part of the cell. Based on the change of the pressure, P-wave velocity and electrical resistance, we can presume that the CO2 hydrate formation front is located between layers A3 and B3. Meanwhile, CO2 hydrates formation was not observed in the electrical resistance data. The electrical resistance of in-situ water-saturated sediments increased when CO2 hydrates formed

because the electrical resistance of CO2 hydrates is higher than pore water [29,30]. In this study, however, the change in electrical resistance was insignificant in spite of the presence of CO2 hydrates, because distilled water was used as pore water in this experiment. This is the one of the limitations of this experiment.

#### *3.2. Depressurization Process*

#### 3.2.1. Temperature and Pressure

Figure 8 shows the pressure of each layer in the cell over time. The pressure of the cell was reduced step-wise using a back-pressure regulator. Pressure differences between the upper and lower part of the cell remained, even though CO2 was vaporized. This pressure discrepancy indicated that the sealing capacity of the CO2 hydrate-bearing sediment layer was preserved. When the CO2 hydrate dissociated completely, the pressure of each layer became equal.

**Figure 8.** Pressure of the cell with lapsed time during depressurization.

Figure 9 shows the pressure–temperature relationship during the depressurization test. The pressure of the cell dropped with the pressure release using the back-pressure regulator. When the pressure of layers A4 and A5 reached the CO2 vapor pressure, their pressure and temperature relation moved along the CO2 vapor pressure (Figure 9a). The path of the pressure and temperature relationship of layers A4 and A5 was similar to that of the isometric process because the flow of fluids was obstructed by the remaining CO2 hydrate-bearing sediment layer (Figure 9b). Then, the pressure and temperature relationship of layer A2 moved along the CO2 hydrate equilibrium line (Figure 9c). This is evidence that the CO2 hydrates re-formed and dissociated in layer A2. In the previous CO2 injection process, CO2 liquid did not reach layer A2 because of the sealing effect of the CO2 hydrate-bearing sediment layer, which was located between layers A3 and B3. However, CO2 was supplied to layer A2 when the existing CO2 hydrates partially dissociated by depressurization. CO2 hydrates then re-formed because layer A3 is in the CO2 hydrate stability zone. Then, CO2 hydrates in layer A3 dissociated by additional depressurization. During the dissociation of CO2 hydrates in layer A3, a self-preservation effect was observed in the pressure and temperature relationship as seen in previous experimental studies on CO2 hydrate dissociation [12,31]. Finally, CO2 hydrates completely dissociated with step-wise depressurization (Figure 9d).

**Figure 9.** The pressure–temperature relationship during depressurization. (**a**) 0–980 min, (**b**) 980–1770 min, (**c**) 1770–3300 min, (**d**) 3300–8700 min.

#### 3.2.2. P-Wave Velocity

Figure 10 shows the results of the P-wave velocity measurements taken during the depressurization process. The sealing effect of the original CO2 hydrate-bearing sediment layer reduced because some portion of the original CO2 hydrates dissociated. Thus, the P-wave velocity of layer A1 decreased because CO2 intruded the upper part of the cell. On the other hand, the P-wave velocity of layers A2 and A3 suddenly increased because CO2 hydrates formed using the CO2 supply from the lower part of the cell. Note that layers A2 and A3 were in the CO2 hydrate stability zone until 3300 minutes (refer to Figure 9). Meanwhile, the P-wave velocity of layers A4 and A5 decreased because CO2 vaporized during depressurization. When the pressure was lower than the equilibrium pressure of the CO2 hydrates (i.e., 4000–6000 min, refer to Figure 9d), the P-wave velocity of layers A2 and A3 suddenly decreased because reformed CO2 hydrates in these layers dissociated.

**Figure 10.** P-wave velocity of the unconsolidated sediment sample during depressurization. (**a**) layer A1, (**b**) layer A2, (**c**) layer A3, (**d**) layer A4, (**e**) layer A5.
