Physical Properties of Gas Hydrate-Bearing Pressure Core Sediments in the South China Sea

Gas hydrates are a potential future energy resource and are widely distributed in marine sediments and permafrost areas. The physical properties and mechanical behavior of gas hydrate-bearing sediments are of great significance to seafloor stability and platform safety. In 2013, a large number of pressure cores were recovered during China’s second gas hydrate drilling expedition in the South China Sea. In this study, we determined the gas hydrate distribution, saturation, physical properties, and mechanical behavior of the gas hydrate-bearing sediments by conducting Multi-Sensor Core Logger measurements and triaxial and permeability tests. Disseminated gas hydrates, gas hydrate veins, and gas hydrate slabs were observed in the sediments. The gas hydrate distribution and saturation are spatially heterogeneous, with gas hydrate saturations of 0%–55.3%. The peak deviatoric stress of the gas hydrate-bearing sediments is 0.14–1.62MPa under a 0.15–2.3MPa effective confining stress. The permeability is 0.006–0:095 × 10−3 μm2, and it decreases with increasing gas hydrate saturation and burial depth.


Introduction
Natural gas hydrates, which are widely distributed in continental margins and permafrost areas, have been suggested to be a potential renewable energy resource in the future. Understanding the physical properties of gas hydratebearing sediments is important for gas hydrate production from marine sediments [1][2][3][4][5]. Depending on the upward methane flux, temperature, pressure, and pore water salinity, the burial depth and distribution of the gas hydrates vary significantly from the seafloor to the depths of several hundred meters below the seafloor [6][7][8][9]. Previous lab tests and numerical simulations have revealed that gas hydrate formation and decomposition may change the sediments' properties, such as their permeability and strength [10][11][12][13][14][15]. Strength decreases resulting from gas hydrate dissociation may lead to geohazards, such as continental landslides [16,17] and seafloor subsidence.
Lab experiments are the main method used to determine the physical properties of gas hydrate-bearing sediments, such as their permeability and shear strength. Key parameters, such as the permeability and shear strength, needed for numerical simulations can be obtained through lab experiments and field testing [18]. The gas hydrates used for lab experiment are either synthetic [19][20][21] or natural samples [22][23][24]. In order to prevent gas hydrates from dissociating during core retrieval from marine sediments [23,25,26], various pressure coring tools have been developed in recent years [3,[27][28][29]; however, only a few studies have been conducted on pressure core gas hydrate-bearing sediments [22,23,30].
Six gas hydrate drilling expeditions [31][32][33][34][35][36][37] and two production tests were successfully conducted by the Guangzhou Marine Geological Survey between 2007 and 2019 on the northern continental slope of the South China Sea [38]; however, the geomechanical properties and permeability of the gas hydrate-bearing sediments have not been publically reported. In this study, we determined the physical properties of the hydrate-bearing pressure core sediments obtained from the Pearl River Mouth Basin in the South China Sea in 2013. X-ray imaging was conducted and the P-wave velocity and gamma density were measured using a Multi-Sensor Core Logger (MSCL). The gas hydrate saturations were estimated via depressurization. The physical properties of gas hydrate-bearing sediments, including the permeability and strength, were measured using a Pressure Core Analysis and Transfer System (PCATS) developed by the GeoTek Company. By comparing our results with those of previous studies, we investigated the influences of the gas hydrate distribution, saturation, effective confining strength, and burial depth on the permeability and strength of the sediments.

Materials and Methods
The sediment cores were retrieved using pressure coring tools during the GMGS2 in the Pearl River Mouth Basin on the northern continental slope of the South China Sea ( Figure 1). The PCATS has been used for gas hydrate exploration for a long time [22]. It can keep the gas hydratebearing sediments under the original in situ pressure (high pressure) during the entire coring and processing procedure, which enables further onshore analysis of the gas hydrates. By using the PCATS, the in situ pressure was maintained and the cores were transferred from the autoclaves into a logging chamber where the gamma density and P-wave velocity were measured and the X-ray imaging was conducted. Furthermore, using the PCATS, the cores were cut into subsamples under the in situ pressure (Table 1) for further tests, including triaxial and permeability tests. Some subsamples were depressurized to estimate the gas hydrate saturation. The temperature and pressure of the cores were recorded throughout the entire process ( Figure 2).

PCATS MSCL Tests.
The gamma density was measured using a 137 Cs source and a sodium iodide detector, with an error of ±2%. The P-wave velocity was measured using the pulse transmission technique at 250 kHz, with an error of ±1%. The X-ray images were collected using a microfocus source and an image intensifier with a pixel resolution of 120 microns. The voltage and current of the polychromatic X-ray source were tuned to provide the best quality. All of the datasets were collected at a spacing of 0.5 cm.

PCATS Triaxial Strength and Permeability Tests
2.2.1. Preparation. Based on the MSCL logging data, specific intervals were cut from the cores for the triaxial tests. Subsamples were transferred to a triaxial transfer vessel under in situ pressure and a temperature of 8°C-10°C to prevent the gas hydrates from dissociating. The subsamples (length: 115 mm, diameter: 57 mm) were transferred to the PCATS triaxial apparatus [22].
Consolidation was followed by the continuous application of effective stresses of up to the calculated in situ values ( Table 1). The pore pressure was reduced while maintaining the confining pressure in order to apply an effective stress on the samples. The in situ effective stress was estimated using the logging-while-drilling density log [39]. A positive axial load, which was achieved by incrementally moving the top cap down as the sample consolidated, was also maintained to ensure contact between the top cap and the sample.

Permeability
Test. The permeability was measured by subjecting the samples to a hydraulic head and measuring the flow rate through the samples. The head pressure was adjusted to create sufficient flow through the samples on a reasonable time scale to accomplish the permeability measurements (up to 12 hours). The pressure and fluid volume were recorded over time, and based on this, the hydraulic conductivity (K) was calculated and converted to the permeability (Ƙ) as follows: where Q is the water flow rate, A is the cross-sectional area, h is the hydraulic head, L is the sample length, μ is the fluid dynamic viscosity, ρ is the fluid density, and g is the gravitational acceleration. The dynamic viscosity of the water was assumed to be 1:298 × 10 −3 Pa•s at 5°C [40].

Triaxial Test.
Following the permeability test, the undrained triaxial test was performed using a constant axial strain rate (2% of the sample's height per hour to record the sample changes). The tests were conducted until the sample failed or the torque limits of the servo drives delivering the load to the samples were reached. The axial load, axial strain, cell pressure, and pore pressure were recorded during the test.

Depressurization and Gas Hydrate Quantification.
The subsamples were depressurized, and all of the released water and gases were collected. During depressurization, the effective confining stresses were greater than the pore pressure in order to avoid fluid leakage. The total gas volume was calculated based on the gases released during the depressurization, which were collected at 8°C and atmospheric pressure in a water-filled 2000 mL-measuring cylinder. The temperature and pressure of the sediments measured by the Fugro probes were used to calculate the in situ methane saturation [41]. When the total gas volume was greater than the methane from the saturated pore waters, the extra methane was assumed to be from the methane hydrates.

Results
Three pressure cores were retrieved from boreholes GMGS2-05B, GMGS2-08C, and GMGS2-16D. The basic information for the subsamples is presented in Table 1.

P-T Conditions of the Pressure Cores.
The pressure and temperature of the cores were recorded by the sensors in the corer (Figure 2). During the deployment, the temperature 2 Geofluids decreased and the pressure increased with increasing water depth. At a water depth of~600 m, the coring tool entered the gas hydrate stability zone (Figure 2(a)). After the sediments were sealed in the coring tool, the pressure was main-tained at an almost constant value, but the temperature increased slowly due to warming of the seawater and air during the core retrieval. During the triaxial and permeability tests, the pressure varied between 11 and 19 MPa due to the  3 Geofluids core operations. The temperature was less than 10°C during most of the coring operations (Figure 2(b)). The permeability test was followed by degassing. Although the pressure and temperature varied during the coring operations, based on the phase diagram (Figure 2(a)), the gas hydrates did not decompose.

MSCL Measurements and Gas Hydrate Saturation.
The depressurization of subsample 05B-2P-1b shows that the methane concentration did not exceed the in situ methane gas saturation in the pore water, indicating that there were no gas hydrates in this subsample. The X-ray image of 05B-2P-1b shows a homogenous lithology, which is consistent with the observed greyish clay in the core description ( Figure 3). The average gamma density was 1.9 g/cm 3 , with a variation of less than 0.1 g/cm 3 , and the average P-wave velocity (Vp) was 1505 m/s, with a variation of less than 15 m/s. The X-ray image of subsample 08C-4P-1b shows many 1-5 mm gas hydrate layers and veins (Figure 3). The depressurization of 08C-4P-1a5 and 08C-4P-1b revealed that their gas hydrate saturations were 9.7% and 14.4% of the pore volume, respectively. Their gamma density and Vp values were almost constant, with average values of 1.8 g/cm 3 and 1610 m/s, respectively.
Four subsamples of 16D-13A were depressurized. Except for subsample 16D-13A-1a2, which lost methane during  4 Geofluids depressurization, the gas hydrate saturations of the other three samples were 44%-55% of the pore volume. The Xray image of 16D-13A-1a2 did not show a distinct gas hydrate distribution (e.g., layers and veins); therefore, it is speculated that the gas hydrates may have been finely distributed in the pore space ( Figure 3). This core consists of homogenous silty clay with biscuit intervals. The density did not vary significantly; however, the Vp reached high values of 1700-2050 m/s.

Strength of Gas Hydrate-Bearing Sediments.
Triaxial compression strength tests were conducted on each subsample under different effective confining pressures. The peak deviatoric stress was defined as the deviatoric stress at an axial stain of 10% if the peak value was not reached by the end of the test. Figure 5 shows the stress-strain curves of the subsamples. Except for 05B-2P-1b, which exhibits slight strain softening, the other six samples exhibit strain hardening, especially subsamples of 16D. Based on the slopes of the initial strain-stress curves, the stiffness of 16D-13A is much higher than those of 05B-2P and 08C-4P. The peak deviatoric stress of 05B-2P-1b was 0. It should be noted that some of the curves exhibit sudden decreases in the deviatoric stress; for instance, the deviatoric stress of sample 16D-13A-1b at an axial strain of 7% decreased incrementally from 1.40 MPa to 1.05 MPa, but it recovered in a short time.

Gas Hydrate Distribution and Burial
Depth. The gas hydrate distribution is controlled by the burial depth [42] and the fluid migration patterns [25,[43][44][45][46]. The X-ray images show that the gas hydrates are in the form of veins within the shallow sediments at~17 m (08C-4P-1a5 and 08C-4P-1b). In the deep sediments, the gas hydrates are   13A-1b, depth~200 m). This observation is consistent with the speculation of Dai et al. [43] that the gas hydrate distribution is influenced by the effective stress and the growing space. Within the shallow sediments, gas hydrate nucleation and crystallization can overcome the lithostatic stress and form gas hydrate veins and layers. While in the deep sediments, the gas hydrate growth cannot overcome the lithostatic stress, and thus, they can only form in the pore space and preexisting faults and fractures.

Factors Controlling the Physical Properties of Gas
Hydrate-Bearing Sediments. It has been suggested that the physical properties of gas hydrate-bearing sediments are closely related to the gas hydrate saturation and effective confining stress [24,26]. In addition, gas hydrate layers and veins have been discovered in marine sediments in marine sea areas, which can change the marine sediments from isotropic to anisotropic materials [47,48]. This inhomogeneity of the material properties increases the difficulty of predicting the stability of the seabed sediments during gas hydrate production [38].

Impact of Gas Hydrate Saturation on the Sediment
Strength and Permeability. The permeability of the sediments is influenced by the gas hydrate saturation (Figure 6). Our results show a negative relationship between the permeability 6 Geofluids and hydrate saturation ( Figure 6). When the gas hydrate content increases in the sediments, the pore space is potentially occupied by gas hydrate grains, leading to decrease in the sediment's permeability [10,14,49]. Previous research has revealed that the stiffness and strength of coarse-grained sediments increase with increasing gas hydrate saturation [21,50,51]. Waite et al. [52] pointed out that the impact of gas hydrates on the peak deviatoric stress is quite limited when the saturation is less than 30%; however, the impact increases dramatically when the saturation is greater than 40%. Hyodo et al. [11] experimentally determined the exponential relationship between the peak deviatoric stress and gas hydrate saturation. The shear wave velocity did not distinctly increase before the gas hydrate saturation reached 40%, indicating that the effect of gas hydrates on the strength and stiffness of the sediments is not significant below this point. Yoneda et al. [24] pointed out that the effects of gas hydrate saturation are analogies for natural gas hydrates and lab synthetic specimens. In reality, the lab synthetic specimens are more likely to be homogeneous for the gas hydrates formed in the pore spaces of the grains, while gas hydrates in nature are usually more heterogeneous in both pore spaces and the fractures. Therefore, this should be taken into consideration when comparing the effects of the gas hydrate saturation of lab synthetic and natural gas hydrate-bearing sediments. Figure 7 shows the relationship between gas hydrate saturation and sediment strength. Since subsamples 05B-2P, 08C-2P, and 16D-13A were subjected to different effective stresses, the influences of the gas hydrate saturation on the strengths of these samples cannot be compared. The eclipses in Figure 7 illustrate three pairs of subsamples, which are from the same burial depth and were subjected to the same effective confining stress during the triaxial tests. The result shows that the stiffness and strength of the sediments increased as the gas hydrate saturation increased. This increase in stiffness is inferred to be caused by the gas hydrates occupying the pore spaces. The gas hydrate cementation between the sediment 7 Geofluids grains resists the rotation, slip, and rearrangement of the grains, which consequently increases the density and strength of the sediments [17,24]. This cementation effect increases with increasing gas hydrate saturation.

Impact of Effective Confining Stress on Sediment
Strength. The sediment stiffness and strength increase as the effective confining stress increases [42,50,53]. As is shown in Table 1, the in situ effective stress of subsample 05B-2P-1b (0.44 MPa) is three times higher than the initial effective stress (0.15 MPa), leading to an overconsolidation ratio of~3, which is speculated to result in the smallobserved post peak softening behavior during shearing ( Figure 5). A highly effective confining stress will increase the effective friction between the sediment grains, and more energy is needed to overcome this friction [50]. The strainstress curve changes from strain softening to strain hardening as the effective confining pressure increases. Figure 8 illustrates the relationship between the effective confining stress and the strength. It is inferred that the increase in the effective confining stress leads to an increase in the interparticle coordination and the friction between grains, which consequently increase the stiffness and strength of the gas hydrate-bearing sediments.

Conclusions
In this study, the physical properties of the hydrate-bearing pressure core sediments recovered during the GMGS2 in the South China Sea were analyzed and reported for the first time. The conclusions of this study are as follows: (1) The gas hydrates within the pressure cores exhibit different morphologies, such as slabs, pore filling, and veins (2) The gas hydrate distribution is very heterogeneous in the sediments with hydrate saturations of 0%-55.3% (3) The maximum deviatoric strength of the subsamples was 0.135-1.619 MPa. The stiffness and strength of the gas hydrate-bearing sediments increase as the gas hydrate saturation and effective confining stress increase (4) The permeability of the subsamples ranged from 0:006 × 10 −3 μm 2 to 0:095 × 10 −3 μm 2 , and the permeability is negatively correlated with hydrate saturation

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 that they have no conflicts of interest.  Figure 8: Relationship between the effective confining stress and the peak deviatoric stress. Red star: this study. Black box: previous study [17]. The numbers are the gas hydrate saturation within the sediments. 8 Geofluids