Gas hydrate in-situ formation and dissociation in clayey-silt sediments: An investigation by low-field NMR

Abstract

The hydrate reservoir in the Shenhu Area of the South China Sea is a typical clayey-silt porous media with high clay mineral content and poor cementation, in which gas hydrate formation and dissociation characteristics are unclear. In this study, the CO₂ hydrate saturation, growth rate, and permeability were studied in sandstone, artificial samples, and clayey-silt sediments using a custom-built measurement apparatus based on the low-field NMR technique. Results show that the T₂ spectra amplitudes decrease with the hydrate formation and increase with the dissociation process. For the artificial samples and Shenhu sediments, the CO₂ hydrate occupies larger pores first and the homogeneity of the sandstone pores becomes poor. Meanwhile, compared with the clayey-silt sediments, CO₂ hydrate is easier to form and with higher hydrate saturation for the sandstone and artificial samples. In hydrate dissociation process, there exists a protection mechanism, i.e. the dissociation near the center of hydrates grain is suppressed when gas pressure drops suddenly and quickly. For permeability of those samples, it decreased with hydrate forms, and increases with hydrate dissociation. Meanwhile, with the same hydrate saturation, permeability is higher in hydrate formation than in dissociation.

Keywords

Gas hydrate clayey-silt sediments NMR formation and dissociation permeability

Introduction

Natural gas hydrates (NGHs), composed of water and natural gas, are solid clathrate compounds found under relatively high-pressure and low-temperature conditions, such as sub-marine sediments and underneath permafrost ( Klauda and Sandler, 2005; Makogon et al., 2007; Sun et al., 2018 ). The estimated global in place (GIP) of natural gas hydrate gas is on the order of 10¹⁵ to 10¹⁸ m³, which is almost twice as much as the world’s proven conventional fossil energy reserves (Lerche and Bagirov, 2004; Li et al., 2016; Sloan and Koh, 2007). Therefore, NGHs are believed to be a strategic energy resource with the potential of replacing conventional natural gas in the 21st century.

The China Geological Survey (CGS) performed the first production test on an NGH-bearing silty clay formation in the target Shenhu Area of South China Sea (SCS) in 2017 (Li et al., 2018). Compared to the NGHs in sandstone reservoirs with relatively high permeability (more than 100 mD) and saturation (about 70%) in Japan (generally coarse sand reservoir) and the USA (conglomerate reservoir), the mean median grain size of the sediments is about 12 μm in the NGH reservoir of the Shenhu Area, which leads to low permeability (2–5 mD) and saturation (about 33%) (Li et al., 2018; Ye et al., 2018). Moreover, the high clay mineral content (26%–30%), which is made up mostly of montmorillonite and illite, results in high bound water concentration of more than 65%. To summarize, the hydrate reservoir in the Shenhu Area of the SCS is mainly composed of fine-grained clay silt with low permeability and low NGH saturation, making exploitation and development difficult.

Numerous laboratory NGH formation and dissociation experiments have been conducted to study the permeability (Johnson et al., 2011; Komai et al., 2009; Kumar et al., 2010), the triaxial compressive properties (Miyazaki et al., 2010), and the electrical properties (Li et al., 2012a) of gas hydrate-bearing sediments. However, these experiments are commonly carried out using glass beads, silica sand, sandstone, or various clay minerals (Sun et al., 2015; Yan et al., 2019; Zhou et al., 2008), which are far from natural sediment properties. According to Hyndman and Davis (1992), gas hydrates beneath the seafloor display a wide range of growth habits and distribution within the host sediment (Booth et al., 1996; Ginsburg and Soloviev, 1994). Therefore, it is necessary to perform hydrate formation and dissociation experiments in silt-clay sediments of the Shenhu Area.

In addition, temperature, pressure, and resistivity measurements have been taken to detect the transition of water and methane to the solid phase in order to monitor and model the formation and dissociation processes of NGHs. (Li et al., 2012b; Linga et al., 2009). A number of kinetic studies have been conducted on NGHs where the gas pressure is measured in high-pressure hermetic vessels during hydrate formation and dissociation. (Englezos and Bishnoi, 1988; Katipot et al., 2019; Mekala et al., 2014). Moreover, during the process of gas hydrate formation or dissociation, the resistivity was measured providing insights into conductance changes (Buffett and Zatsepina, 2000). However, sensors installed in the samples causes heterogeneity in the sediments. The intrusion of sensors might lead to uneven nucleation resulting from irregular cooling at the metal boundary of sensors with high thermal conductivity (Linga et al., 2009). Therefore, non-intrusive methods for detecting and modeling NGH characteristics should be encouraged. In recent years, X-ray Computed Tomography (X-CT) has provided a visual technique to depict the spatial distribution of different phases of gas, water, and hydrate phases with different densities (Ai et al., 2017; Kneafsey and Moridis, 2014; Seol and Kneafsey, 2009). However, current X-CT technology is not suitable for rapid measurements in the initial periods of hydrate formation as it requires several hours to obtain high-resolution X-CT images.

Low field nuclear magnetic resonance (NMR) provides a fast, easy, and non-destructive way of detecting hydrogen-bearing fluids. The low field NMR method was used to determine petrophysical characteristics of conventional and unconventional porous media in oil and gas fields, such as porosity, pore geometry, pore connectivity, permeability, and even the saturation of fluids in reservoirs (Liu et al., 2019; Sun et al., 2016; Timur, 1969, 1972). With this in mind, the NMR technique was applied to monitor the dynamic behavior of CO₂ hydrate formation and dissociation as well as permeability changes in silt-clay sediments with a series of transverse relaxation time (T₂) measurements. In this study, gas hydrate formation and dissociation processes as well as reservoir properties were studied through time using the configurations of T₂ spectra distribution changes of the water in the sediments. This process allows for the quantitative evaluation of the gas formation rate, saturation, and permeability at different hydrate saturations, which is applicable for modeling formation and dissociation characteristics of NGHs and successful production operation in the SCS.

Theory of NMR measurement

The NMR used in the petroleum and natural gas industry is tuned to describe the response of the protons within the hydrogen nucleus to an applied magnetic field. Hence, the NMR measurements provide a unique look at the interactions of hydrogen bearing fluids and the characterization of the pores within reservoir rock (Howard et al., 1993; Kenyon et al.,1988). T₂ acts as a sensitive indicator to proton response to the applied magnetic field, which is defined by three relaxation mechanisms in rocks: (1) molecular motion in fluids (bulk relaxation), (2) surface relaxivity at the pore wall (surface relaxation), and (3) molecular diffusion in the internal gradients (diffusion relaxation). This can be described as

\frac{1}{T_{2}} = \frac{1}{T_{2 B}} + \frac{1}{T_{2 S}} + \frac{1}{T_{2 D}}

(1)

where T_2B presents bulk relaxation, T_2S presents surface relaxation, and T_2D is diffusion relaxation. The diffusion relaxation is the pore fluid relaxation induced by diffusion in the gradient magnetic field, while in this study the magnetic is a homogeneous internal field, so that the relaxation caused by gas diffusion can be neglected. The surface relaxation time shows a linear relationship with the pore size that corresponds to the surface-to-volume ratio (S/V). In this case, the characteristic of proton transverse relaxation time is as follows

\frac{1}{T_{2}} = \frac{1}{T_{2 B}} + \frac{1}{T_{2 S}} = \frac{1}{T_{2 B}} + ρ_{2} \frac{S}{V}

(2)

where ρ₂ is the surface relaxivity, and S/V is the specific surface area of pores. The bulk relaxation time, which represents the intrinsic properties of the fluid (oil, gas, and water), is long and a result of slow relaxation. Surface relaxation, which is affected by the interactions of hydrogen-bearing fluid and the surface, occurs rapidly due to a short relaxation time (Sun et al., 2016; Zheng et al., 2020).

In addition, NMR measurement responses are associated with not only the status of the fluid (oil, gas, and water) but also the hydrogen content. The quantity of ¹H atoms detected by NMR is proportional to the total amplitude of T₂ spectra, thus the water content within pores can be obtained from T₂ spectra. In this study, the ¹H atoms in water molecules are the measurement target. As solid and liquid phase are two different forms of water, the response of the NMR measurements is different as well. After the water molecules freeze, the relaxation speed of ¹H becomes so fast that it cannot be detected by the low-field NMR instrument. Thus, the principle for ¹H atom in hydrate showing no signal is the same as that of ice. In this study, the total amplitude of T₂ was used for quantitative analysis of the water during the formation and dissociation processes of the hydrate.

Equipment and methods

Experimental equipment and procedure

Figure 1 shows a schematic of the design setup used to conduct the in-situ hydrate formation and dissociation experiments. The experimental apparatus system comprises an NMR spectrometer, a core holder, an upstream gas supply system, an upstream water supply system, circulation and refrigeration confining systems, a back-pressure system, and a data acquisition system (Figure 1). The main unit of the apparatus system is the high-pressure thermostatic core holder, which is designed for holding hydrate-containing specimens or sample cell. The core holder is specially designed using nonmetallic and nonmagnetic material. The liquid and gas supply systems pump water and gas to the samples at a constant velocity and pressure, respectively. The confining systems consist of a cooling bath and high-pressure circulating pump for cooling and circulation of fluorinated liquid (confining liquid). The back-pressure system connected to the core holder outlet includes a pressure transducer, a regulator controlled by a pump, and a wet gas flowmeter, which is used to obtain the gas flow rate during permeability measurements.

Figure 1.

Schematic diagram of the experimental setup.

Sample preparation and experimental procedure

In this study, three kinds of porous media with different pore size distributions and pore structure were used to study their T₂ spectra responses in the gas hydrate formation and dissociation processes. The first sample is a well-consolidated sandstone with a length of 5 cm, a diameter of 2.54 cm, and a porosity of 22.6%. The second sample is an unconsolidated artificial sample consisting of fine sands and clays with grain diameters ranging from 70 to 500 μm. The third sample is a fresh unconsolidated clayey-silt sediment collected from the SCS.

The sandstone sample was polished and cleaned to remove reservoir fluids and drilling mud and then dried for 24 hours at a temperature of 80 °C. For the artificial sample, the sample was dried in a vacuum-oven at 80 °C for six hours then packed into a sample cell. Using the dried sample weight and density, the porosity was quantified and found to be 30%. The sandstone and artificial samples were saturated with distilled water. The clayey-silt sediments were obtained from the Shenhu Area. To prevent the evaporation of water, the fresh samples were carefully packaged in plastic sheeting during transportation. With the T₂ spectra measurement, the porosity of the clayey-silt sediments was found to be 38%. For the unconsolidated sediments, they were put into a special made poly-Teflon sample cell for the NMR measurement.

Using the specially designed apparatus, we performed hydrate formation and dissociation experiments with parallel, real-time T₂ measurements during each process. In the experimental procedure, the samples were placed into a thermoplastic sleeve and inserted into the holding chamber. Then, fluorinated liquid was injected into the space between the sleeve and the chamber to compress the sample. When the confining pressure stabilized, the cooling bath was switched on to decrease and keep the experiment temperature at 4 °C. The samples were subjected to vacuum to -0.1 MPa, and then the CO₂ was injected with low pressure to keep water from being displaced. The T₂ spectra were measured from the start of CO₂ gas injection until the end of the formation process, when the changes of two adjacent T₂ were negligible. Meanwhile, the visual images of samples were obtained using an MRI technique. The NMR spectrometer, manufactured by Niumag Corporation Ltd., China, has a 60 mm diameter magnet coil and a frequency of 23.15 MHz.

Results and discussion

The formation of hydrate

The T₂ spectra of samples were measured before the CO₂ hydrate formation. As shown in Figure 2, there are three distinct peaks for the samples: the left peak (P1) centered at 0.1–10 ms, the middle peak (P2) at approximately 10–100 ms, and the right peak (P3) at > 100 ms. The configurations of the T₂ spectra distributions represent the pore size distribution, with smaller pores having a shorter relaxation time, and a longer relaxation time for larger pores (Yao et al., 2019). Generally, the P3 peaks represent the bulk water content in the corner and surface of the samples. Thus, the “bulk water” peak is unimportant in our research. For the artificial sample that consists of fine sand and clay, it shows three unique peaks resulting from the large pores between the fine sand and the micropores mainly distributed in the clay. For the clayey-silt sediments, the micropores are uniform in size; thus, the T₂ spectrum shows a narrow P1 peak centering at small relaxation time. Meanwhile, the P2 peak near 100 ms corresponds to the large pores in the clayey-silt sediments, which is a small proportion of the total pore space.

Figure 2.

T₂ spectra of different water-saturated samples.

The CO₂ hydrate formation experiment results show that there is an observable decrease in the amplitude of T₂ spectra corresponding to the formation process, as shown in Figure 3. For the sandstone sample, there is an obvious decrease in P2, whereas there is only a slight decrease in the P1 peak. Moreover, the P2 peaks move towards the left (indicating a faster relaxation time) and then move to the right slightly with the further growth of the gas hydrate. The P2 peaks moving toward the left at the first 200 minutes suggest the growth of hydrate near the pore walls decreases the pore size leading to an increase in the surface-to-volume ratio. With hydrate continuing to grow, some of the water is surrounded by the hydrate resulting in the isolation from the free CO₂ gas. It is difficult for the residual water to form hydrate as it is difficult to make contact with the free CO₂ gas due to pore plugging. Therefore, the residual water in hydrate with long bulk water relaxation causes the leftward movement of the P2 peak. It is observed that the P2 peak became wider, which corresponds to a wider range of pore diameters. Therefore, the various pore diameters indicate the homogeneity of the sandstone pores becomes poor. The reason for this is the increased complication of pore structure with hydrate growth leading to multiscale pores. For the artificial samples, the P1 and P2 peaks decrease and shift leftward. In addition, for the clayey-silt sediments, the P1 peak decreases with CO₂ hydrate formation, and the decreased area is mostly on the right side corresponding to the relatively large pores. For the artificial sample and the Shenhu sediments, the peaks move to the left as the hydrate grows. This indicates that the CO₂ hydrate occupies the large pores, and the decrease in porosity leads to the stronger surface relaxation of the small pores. For the artificial sample, the position changes are more obvious than in the Shenhu sediments, which is a result of the multi-scale pore distribution.

Figure 3.

T₂ spectra during CO₂ hydrate formation experiments for the different samples.

In addition, compared to the Shenhu sediments, the CO₂ hydrate is easier to form in the sandstone and artificial samples. For hydrate in the Shenhu clayey-silt sediments with high clay content and small pores, it took four times as long to reach the equilibrium, as shown in Figure 4. As the gas hydrate is undetectable by the low-field NMR and the sample is fully saturated, CO₂ exists as a dissolved gas. Thus, in this study, $S_{h} + S_{w} = 1$ ; meanwhile, $S_{h}$ is obtained by the area changes between the water-saturated sample spectra ( $A_{initial}$ ) and the measured spectra at any time ( $A_{t}$ ), which is expressed as

S_{h} = 1 - S_{w} = \frac{A_{initial} - A_{t}}{A_{initial}}

(3)

where

S_{h}

is hydrate saturation,

S_{w}

is water saturation,

A_{initial}

is the area of water-saturated sample spectra, and

A_{t}

is the area of sample measured at any time.

Figure 4.

The CO₂ hydrate saturation in the formation process for different samples.

The hydrate saturation in clayey-silt sediments is low (about 36.7%), whereas the hydrate saturation is 77.4% in the sandstone sample and 83.8% in the artificial sample. Therefore, it is easier for the gas hydrate to reach an equilibrium state for samples with large pore diameters than those with small pore diameters. The CO₂ hydrate formation process in clayey-silt sediments showed a slow formation process and low hydrate saturation as a result of smaller pores. Meanwhile, the hydrate formation rate is limited by CO₂ diffusion through water, which is determined by gas pressure, temperature, and the pore structure in water-saturated porous media. For clayey-silt sediments with small grain sizes and pores, the CO₂ diffusion is slower than that in the sandstone and artificial samples (Kim et al., 2018).

Meanwhile, the experimental results show that the P2 peak of the T₂ spectra with longer relaxation time decreased with the formation time, whereas the P1 peak decreased slowly, indicating the gas hydrate forms in larger pores preferentially for all samples, as shown in Figure 5. For the sandstone, 81% of water combines with the CO₂ molecules to form hydrate in the larger pores, whereas only 45% of water forms hydrates with CO₂ in smaller pores_. For the artificial samples, the percentage is 91% in larger pores and 83% in smaller pores. In the clayey-silt sediments, the percentage is about 100% in larger pores and 34% in smaller pores.

Figure 5.

The changes of P1 and P2 amplitude for different samples.

The dissociation of hydrates

In this study, the multi-stage depressurization mode (pressure decreased step by step) was used, during which the T₂ spectra were measured at every pressure stage. The results show that the T₂ spectra peaks increase with hydrate dissociation. At the first stage, where gas pressure dropped to 3 MPa, there is little change in the T₂ spectra for the three different samples. As the pressure is above the thermodynamic equilibrium curve, the gas hydrate is in a stable condition. With gas pressure dropping to 2 MPa, the hydrates begin to dissociate. As shown in Figure 6, the area of the T₂ spectra begins to increase; however, the hydrates do not dissociate completely, although the pressure is below the thermodynamic equilibrium curve (the equilibrium pressure is about 2.2 MPa under 4 °C in pure water for CO₂ hydrate) (Ohgaki et al., 1993). When the pressure dropped to 1.0 MPa, massive hydrate dissociation occurs. Therefore, there is a pressure threshold for abrupt changes in the hydrate dissociation process. However, some hydrate remains, which dissociates with further drops in pressure. The experiment result suggests that although gas pressure drops below the thermodynamic equilibrium curve, hydrate dissociation is incomplete, which is in agreement with the results in previous studies (Kneafsey and Nakagawa, 2011). A sudden and quick depressurization brings hydrate out of the stable P-T zone, which results in a rapid dissociation from the hydrate-gas interface. Meanwhile, hydrate dissociation exhibits a self-protection mechanism in that as temperature behind the interface begins to decrease, hydrate dissociation in the center of the pores is suppressed. For the artificial sample, the P1 peak increases, and then the P2 peaks were observed during hydrate dissociation. For clayey-silt sediments, the P1 peak begins to increase in the left portion of the peak, which is the opposite of the formation process. The peak position changes of these two samples indicate that hydrate dissociation is initiated in the small pores and then in the macropores.

Figure 6.

T₂ spectra at different pressures for different samples CO₂ hydrate dissociation.

However, there is a significant difference between the T₂ spectra distribution after hydrate dissociation and from before hydrate formation for artificial sample, where the P1 peak decreases and P2 peak increases. We assume that the redistribution of peaks is a result of the water molecules moving from small to large pores to form hydrate, causing the pore structure to change. Therefore, it is suggested that the formation and dissociation of hydrate changes the pore structure and water distribution of unconsolidated samples. Tian et al. (2019) also observed pore structure changes during hydrate dissociation. Figure 7 shows that the water signal disappears during CO₂ hydrate formation and reappears after hydrate dissociation in the MRI images. It also shows that the water in samples is redistributed after CO₂ hydrate formation and dissociation. Noticeably, the MRI image or T₂ spectra collected by low-field NMR is the composition of the sample along the magnetic field direction. The MRI image is a composite two-dimensional image. Therefore, the distribution of gas hydrate or porosity is an average but not the real three-dimensional distribution.

Figure 7.

Nuclear magnetic resonance image during CO₂ hydrate formation and dissociation in the artificial sample (a→d, the process of hydrate formation, and d→e, the process of hydrate dissociation).

The changes in permeability

The Shenhu sediments are a poorly cemented porous media, thus it is difficult to obtain permeability using conventional methods (Lei et al., 2020). This is especially true as hydrate grows in the sediments under high pressure and low temperature conditions with water or gas flowing through the sample. Therefore, quantifying permeability changes during hydrate growth in clayey-silt sediments can provide unique insights. There are two primary methods for estimating permeability with NMR results, one is the SDR model based on the estimation of an average pore size, and the other is the Coates model based on irreducible water saturation (Machado et al., 2008). The Coates equation uses the porosity and free-fluid to bound-fluid ratio determined from the T₂ distribution to predict permeability (Coates et al., 1991). As we cannot obtain the free-fluid to bound-fluid ratio in the hydrate formation and dissociation processes, the Coates method is not applicable. The SDR method is based on the T₂ distribution being a pseudo-pore size distribution of the formation. In this study, the main fluids are water and CO₂, thus we can use SDR to measure the permeability. The SDR model is

K = a T_{2 g m}^{2} \emptyset^{4}

(4)

where K is the permeability,

T_{2 g m}

is the logarithmic mean of the T₂ distribution (mD), and

\emptyset

is the porosity (%). The parameters a in the model is determined from NMR measurements on samples with a known porosity and permeability. As we cannot obtain the original permeability of fresh Shenhu sediments, the parameter a cannot be obtained. Thus, we calculate the ratio of permeability (

\frac{K}{K_{o}}

) at the initial condition (with the subscript o) and any condition during hydrate formation, as shown in Figure 8. It shows that for all samples the permeability decreases with hydrate formation; meanwhile, the permeability increases with hydrate dissociation. The results show that the permeability during the dissociation process is higher compared to during the formation process at the same hydrate saturation. We fit the permeability to hydrate saturation with a power exponential function, which fit the data well:

\frac{K}{K_{0}} = a {(1 - S_{h})}^{n}

(5)

Figure 8.

The permeability in hydrate formation and dissociation for different samples.

As mentioned above, since the sample is fully saturated CO₂ exists as dissolved gases. Thus, in this study, $S_{h} + S_{w} = 1$ . However, $S_{w}$ can be represented as the ratio of porosities ( $\frac{\emptyset}{\emptyset_{0}}$ ) obtained from NMR spectra:

S_{h} = 1 - \frac{\emptyset}{\emptyset_{0}}

(6)

Thus, there is a relationship between permeability and porosity:

\frac{K}{K_{0}} = a {(\frac{\emptyset}{\emptyset_{0}})}^{n}

(7)

where the parameter a and n are empirical values. Civan (2002) described the Kozeny-Carman equation application for the permeability reduction of porous media with fine particles deposition (Sharma et al., 2000) as

\frac{K}{K_{0}} = f_{F τ} f_{Σ} f_{ϕ}

(8)

where the factors

f_{F τ}

f_{Σ}

, and

f_{ϕ}

denote the variations of the tortuosity and the shape factor product, pore surface area, and porosity by deposition of fines in porous media, respectively. Thus, the empirical value a is related to the geometric structure of pore spatial distribution, which is affected by the tortuosity, the pore throat, and the pore surface area. In addition, compared to the hydrate formation process, the parameter a increased during the hydrate dissociation process, which indicates that with the same porosity (i.e., the same hydrate saturation), the permeability is higher during the dissociation process. We assume that the formation of gas hydrate is initiated at regions of pore throats where the pore size is small. The hydrate plugs the throat and rapidly decreases the permeability. Meanwhile, the dissociation of gas hydrate starts around regions of the pore throat and expands into the pore body. Wang et al. (2020) studied the hydrate occupancy and microstructure evolution with hydrate dissociation on the basis of processing and statistical analysis of X-ray CT images. They also found that the interpore hydrate frameworks interconnected through the pore throats tend to be dissociated in priority in the fully brine-saturated sediment. Therefore, the pore connectivity is better, and the permeability is higher with the same hydrate saturation.

For the artificial sample whose structure is not compact, the permeability after complete hydrate dissociation is higher than that of the original condition. Besides the effect of hydrate saturation on the pore structure, the sample’s own structure changes with hydrate formation and dissociation. However, for the unconsolidated clayey-silt sediments sample, the changes of the pore structure are very small. The main reason is that the content of clay is high (about 26%–30%), and the cementation is strong.

When CO₂ hydrate growth reaches equilibrium, we tried to measure the gas permeability by decreasing the backpressure to 3.7 MPa. It is worth mentioning that we assumed that there was no hydrate formation during permeability measurements as the hydrate formation has reached equilibrium. However, the experiment results show that there is no gas detected at the outlet of backpressure regulator. This indicates that the porosity and permeability decreased dramatically with the formation of hydrate. The permeability of dry sandstone is 645.6 mD, and the permeability found using NMR is 2 mD after hydrate formation, which is in good agreement with the permeability measurement results. This is just a preliminary experiment for permeability changes, and a series of successor experiments will be performed to study permeability under different hydrate saturations.

Conclusions

In this study, NMR was used to evaluate the CO₂ hydrate saturation and permeability changes in sandstone, artificial samples, and clayey-silt sediments during hydrate formation and dissociation. The following conclusions were made.

The T₂ spectra representing the water in samples decreases with hydrate formation and increases with hydrate dissociation. The difference in the T₂ spectra was used to calculate the hydrate saturation. With hydrate growth, the homogeneity of pores becomes poor, and the CO₂ hydrate occupies larger pores first, leading to the leftward movement of the T₂ spectra.

Compared to the sandstone and artificial sample, the hydrate saturation in clayey-silt sediments is low (37%), and it takes four times as long to reach hydrate growth equilibrium. Moreover, gas hydrate forms preferentially in larger pores for all samples.

There is self-protection mechanism in the process of hydrate dissociation in which the complete dissociation is not achieved under the dissociation condition, and the hydrate dissociation initiates in the small pores first and then in macropores.

The permeability of the porous media decreases with hydrate formation and increases with hydrate dissociation. With the same hydrate saturation, the permeability during hydrate dissociation is higher than that in hydrate formation.

Footnotes

Acknowledgments

The authors are grateful to Donghui Xing, Lantao Geng, and Chenlu Xu for their participation in performing experiments.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the National Natural Science Foundation of China (51991365), the GH Programs of China Geological Survey (No. DD20190232) and the Guangzhou Science and Technology Project (No. 201909010002).

ORCID iD

Xiaoxiao Sun

References

Zhao

Wang

, et al. (2017) Analyzing permeability of the irregular porous media containing methane hydrate using pore network model combined with CT. Energy Procedia 105: 4802–4807.

Booth

Rowe

Fischer

(1996) Offshore gas hydrate sample database with an overview and preliminary analysis. U.S. Geology Survey Open File Reports 28: 96–272.

Buffett

Zatsepina

(2000) Formation of gas hydrate from dissolved gas in natural porous media. Marine Geology 164(1–2): 69–77.

Civan

(2002) Multiphase model for evaluation of filtration during drilling and fracturing of wells. In: Proceedings of the SPE annual technical conference and exhibition, San Antonio, TX, USA, 29 September–2 October.

Coates

Miler

Gillen

, et al. (1991). The MRIL in Conoco 33-1: an investigation of a new magnetic resonance imaging log. In: Proceedings of the annual logging symposium of the SPWLA, Midland, TX, USA, 16–19 June.

Englezos

Bishnoi

(1988) Prediction of gas hydrate formation conditions in aqueous electrolyte solutions. AIChE Journal 34(10): 1718–1721.

Ginsburg

Soloviev

(1994) Mud volcano gas hydrates in the Caspian Sea. Bulletin of the Geological Society of Denmark 41: 95–100.

Howard

Kenyon

Straley

(1993) Proton-magnetic resonance and pore-size variations in reservoir sandstones. SPE Formation Evaluation 8(03): 194–200.

Hyndman

Davis

(1992) A mechanism for the formation of methane hydrate and seafloor bottom-simulating reflectors by vertical fluid expulsion. Journal of Geophysical Research 97(B5): 7025–7041.

10.

Johnson

Patil

Dandekar

(2011) Experimental investigation of gas-water relative permeability for gas-hydrate bearing sediments from the Mount Elbert Gas Hydrate Stratigraphic Test Well, Alaska North Slope. Marine and Petroleum Geology 28(2): 419–426.

11.

Katipot

Pramoch

Santi

, et al. (2019) Effects of temperature and pressure on the methane hydrate formation with the presence of tetrahydrofuran (THF) as a promoter in an unstirred tank reactor. Fuel 255: 115705.

12.

Kenyon

Day

Straley

, et al. (1988) A three-part study of NMR longitudinal relaxation properties of water saturated sandstones. SPE Formation Evaluation 3(03): 622–636.

13.

Kim

Yun

(2018) Visualization of gaseous and dissolved CO₂ migration in porous media. Environmental Earth Sciences 77(8): 301.

14.

Klauda

Sandler

(2005) Global distribution of methane hydrate in ocean sediment. Energy & Fuels 19: 459–470.

15.

Kneafsey

Moridis

(2014) X-Ray computed tomography examination and comparison of gas hydrate dissociation in NGHP-01 expedition (India) and Mount Elbert (Alaska) sediment cores: Experimental observations and numerical modeling. Marine and Petroleum Geology 58: 526–539.

16.

Kneafsey

Nakagawa

(2011) Repeated methane hydrate formation and dissociation in a partially water saturated san: Impact on hydrate heterogeneity and sonic-frequency seismic properties. In: Proceedings of the 7th International Conference on Gas Hydrates, Edinburgh, Scotland, UK, 17–21 July.

17.

Komai

Sakamoto

Nakashima

, et al. (2009) NMR measurement and analysis of porosity and permeability for hydrate-bearing sediments. In: Proceedings of the eighth ISOPE ocean mining symposium, Chennai, India, 20–24 September.

18.

Kumar

Maini

Bishnoi

, et al. (2010) Experimental determination of permeability in the presence of hydrates and its effect on the dissociation characteristics of gas hydrates in porous media. Journal of Petroleum Science and Engineering 70: 114–122.

19.

Lei

Liao

Chen

, et al. (2020) Stress dependent gas-water relative permeability in gas hydrates: A theoretical model. Advances in Geo-Energy Research 4(3): 326–338.

20.

Lerche

Bagirov

(2004) World estimates of hydrate resources, basic properties of hydrates, and Azerbaijan hydrates. Energy Exploration & Exploitation 22(1): 3–56.

21.

Sun

, et al. (2012a) Experimental studies on the evolvement of electrical resistivity during methane hydrate formation in sediments. Energy & Fuels 26(10): 6210–6217.

22.

Qin

, et al. (2018) The first offshore natural gas hydrate production test in South China sea. China Geology 1(1): 5–16.

23.

Zhang

, et al. (2016) Investigation into gas production from natural gas hydrate: A review. Applied Energy 172: 286–322.

24.

Yang

Zhang

, et al. (2012b) Experimental investigation into gas production from methane hydrate in sediment by depressurization in a novel pilot-scale hydrate simulator. Applied Energy 93: 722–732.

25.

Linga

Haligva

Nam

, et al. (2009) Gas hydrate formation in a variable volume bed of silica sand particles. Energy & Fuels 23(11): 5496–5507.

26.

Liu

Xie

Yao

, et al. (2019) Preliminary study of influence factors and estimation model of the enhanced gas recovery stimulated by carbon dioxide utilization in shale. ACS Sustainable Chemistry & Engineering 7(24): 20114–20125.

27.

Machado

Azeredo

RBV

Ramos

PFO

, et al. (2008) A NMR permeability model based on partial least square (PLS) regression for a low permeability gas sand in Santos basin. In: Proceedings of the SCA Annual Symposium, Abu Dhabi, UAE, 29 October–2 November.

28.

Makogon

Holditch

Makogon

(2007) Natural gas hydrates – A potential energy source for the 21st century. Journal of Petroleum Science & Engineering 56(1-3): 14–31.

29.

Mekala

Babu

Sangwai

, et al. (2014) Formation and dissociation kinetics of methane hydrates in seawater and silica sand. Energy & Fuels 28(4): 2708–2716.

30.

Miyazaki

Masui

Tenma

, et al. (2010) Study on mechanical behavior for methane hydrate sediment based on constant strain‐rate test and unloading‐reloading test under triaxial compression. International Journal of Offshore and Polar Engineering 20: 61–67.

31.

Ohgaki

Makihara

Takano

(1993) Formation of CO₂ hydrate in pure and sea waters. Journal of Chemical Engineering of Japan 26(5): 558–564.

32.

Seol

Kneafsey

(2009) X-ray computed-tomography observations of water flow through anisotropic methane hydrate-bearing sand. Journal of Petroleum Science Engineering 66(3-4): 121–132.

33.

Sharma

Pang

Wennberg

, et al. (2000) Injectivity decline in water-injection wells: An offshore Gulf of Mexico case study. SPE Production & Facilities 15(01): 6–13.

34.

Sloan

Koh

(2007) Clathrate Hydrates of Natural Gases. 3rd ed. Boca Raton: CRC Press.

35.

Sun

Yao

Liu

, et al. (2016) Interactions and exchange of CO₂ and H₂O in coals: an investigation by low-field NMR relaxation. Scientific Reports 6: 19919.

36.

Sun

, et al. (2018) Hydrate dissociation induced by gas diffusion from pore water to drilling fluid in a cold wellbore. Advances in Geo-Energy Research 2(4): 410–417.

37.

Sun

Zhang

Guo

, et al. (2015) A study of the formation morphology and phase equilibrium of fractured methane hydrates. Energy Exploration & Exploitation 33(5): 745–754.

38.

Tian

Wei

Yan

, et al. (2019) A NMR-based analysis of carbon dioxide hydrate dissociation process in silt (in Chinese). Scientia Sinica Physica, Mechanica & Astronomica 49: 034615.

39.

Timur

(1969) Pulsed nuclear magnetic resonance studies of porosity, movable fluid, and permeability of sandstones. Journal of Petroleum Technology 21(06): 775–786.

40.

Timur

(1972) Nuclear magnetic resonance study of carbonates rocks. Log Analysts 13: 3–11.

41.

Wang

Liu

, et al. (2020) Study of hydrate occupancy, morphology and microstructure evolution with hydrate dissociation in sediment matrices using X-ray micro-CT. Marine and Petroleum Geology 113: 104138.

42.

Yan

Chen

, et al. (2019) Methane hydrate formation and dissociation behaviors in montmorillonite. Chinese Journal of Chemical Engineering 27(5): 1212–1218.

43.

Yao

Liu

, et al. (2019) A new application of NMR in characterization of multiphase methane and adsorption capacity of shale. International Journal of Coal Geology 201: 76–85.

44.

Qin

Qiu

, et al. (2018) Data report: Molecular and isotopic formations of the extracted gas from China’s first offshore natural gas hydrate production test in South China Sea. Energies 11(10): 2793.

45.

Zheng

Yao

Elsworth

, et al. (2020) Dynamic ﬂuid interactions during CO₂-ECBM and CO₂ sequestration in coal seams. Part 2: CO₂-H₂O wettability. Fuels 279: 118560.

46.

Zhou

Tao

Liang

, et al. (2008) Simulation of the decomposition of natural gas hydrate in porous media by hot water injection. Energy Exploration & Exploitation 26(4): 267–279.