Review of Heat Transfer Characteristics of Natural Gas Hydrate

Abstract

As a typical unconventional energy reservoir, natural gas hydrate is believed to be the most promising alternative for conventional resources in future energy patterns. The exploitation process of natural gas hydrate comprises a hydrate phase state, heat and mass transfer, and multi-phase seepage. Therefore, the study of heat transfer characteristics of gas hydrate is of great significance for an efficient exploitation of gas hydrate. In this paper, the research methods and research progress of gas hydrate heat transfer are reviewed from four aspects: measurement methods of heat transfer characteristics, influencing factors of heat transfer in a hydrate system and hydrate-containing porous media systems, predictive models for effective thermal conductivity, and heat transfer mechanisms of hydrate. Advanced measurement techniques and theoretical methods that can be adopted for the heat transfer characteristics of gas hydrate in the future are discussed.

1. Introduction

Gas hydrates are compounds formed by the interaction of gas molecules and water molecules under conditions of low temperature and high pressure, resembling a crystalline lattice structure with cage-like formations [1,2]. In this structure, water molecules serve as the “host”, while gas molecules act as the “guest”, and based on the characteristics of the water molecule distribution, gas hydrates can be classified into three types: sI, sII, and sH. In natural gas hydrates, the guest molecules are low molecular weight hydrocarbon gases such as methane. They are widely found in the permafrost regions of continents and deep-sea sediment layers, but also in some ancient mountain ranges [3,4]. Under standard temperature–pressure conditions, 1 m³ of hydrate can release 0.8 m³ of water and about 164 m³ of natural gas. Currently, there is a growing global demand for natural gas resources. By 2040, the world demand for natural gas is projected to grow by a significant 44.0 percent (1.7 percent per annum), reaching a quarter of the primary energy mix (i.e., oil, natural gas, coal, nuclear, and renewables) [5]. Therefore, natural gas hydrates have attracted more and more attention worldwide due to the shortage of conventional resources and the daily increasing expectations for alternative energy [4,6]. And gas hydrates have significant applications in sustainable technologies including gas and energy storage, gas separation, and water desalination [7].

A natural gas hydrate reservoir is a multi-component complex sediment system composed of natural gas, water, hydrate, sediment, and so on. Natural gas exploitation in reservoirs involves a hydrate phase change, multi-phase seepage, heat and mass transfer, and other processes. Thus, the exploitation of natural gas hydrate is a complex and difficult process to control. As we all know, the decomposition of natural gas hydrate is a heat absorption process that can cause a decrease in the local temperature of the reservoir; if the environmental heat exchange efficiency is low, it will lead to the slowing down of hydrate decomposition near the decomposition front, the cessation of hydrate decomposition, or the secondary formation of hydrate, which seriously restricts the efficient development of natural gas hydrate [8]. Thus, the study of heat transfer characteristics of natural gas hydrate is of great significance for efficient exploitation. The current experimental methods of gas hydrate exploitation are depressurization, CO₂-CH₄ displacement, thermal exploitation, injection of chemical inhibitors, and the solid exploitation method. Considering the features of high yield, low environmental impact, and relatively mature technology of depressurization method, since 2002, some countries such as China, the United States, Canada, and Japan have primarily employed depressurization as the principal means of exploiting hydrate from sandy reservoirs and conducted production trials [6]. When the heat transfer is insufficient, the temperature at and around the hydrate decomposition interface will be lower than the corresponding phase equilibrium temperature, and the hydrate decomposition rate will be significantly reduced or even stopped, thus inhibiting gas hydrate decomposition. Therefore, heat transfer affects hydrate decomposition and even the efficiency of gas hydrate extraction [9]. In this method, temperature control is the key factor, so we need to understand the heat transfer characteristics of natural gas hydrates, make more use of the sensible heat of the reservoir, and enhance the heat transfer to the environment to improve the production efficiency of hydrates [10,11]. On the other hand, natural gas hydrates are stable only in a specific temperature and pressure range, elucidating the danger of climate change. For plausible changes in sea level and temperature over the next few centuries [7], the effect of rising temperatures on the gas hydrate reservoir will likely accelerate the decomposition of gas hydrates by speeding up heat exchange between the external environment and the gas hydrate reservoirs. This will reduce the rigidity of permafrost and cause geological disasters [12,13], such as sediment deformation and landslides, and seafloor slope failures on continental margins. Because of the widespread distribution of permafrost and the uncertainty of its gas hydrate content, CH₄ leakage from hydrate decomposition caused by permafrost can seriously pollute the atmosphere and accelerate the global greenhouse effect [7,14].

At present, experimental tests and theoretical simulations have focused on the heat transfer process of natural gas hydrate formation and decomposition, but there are still some deviations between theoretical and experimental data, and there is no unified view on the heat transfer mechanism in the process of hydrate decomposition/formation. This is mainly because the gas hydrate reservoir is relatively complex, and the reservoir sediments are mainly porous media. Its heat transfer is affected by many factors, including pore size distribution, pore size, and heat transfer performance of the media. On the other hand, there are many heat transfer modes such as convective heat transfer, phase change heat transfer, and radiative heat transfer in hydrate formation and decomposition [15]. This paper focuses on the heat transfer characteristics of natural gas hydrate formation and decomposition processes, summarizes the measurement methods of heat transfer characteristics, and analyzes the influencing factors of heat transfer in hydrate systems and theoretical methods that can be adopted for the heat transfer characteristics of natural gas hydrate in the future.

2. Heat Transfer Measurement Methods

Heat transfer is the process of transferring heat energy due to spatial temperature differences. The material transfers heat through the movement and interaction of internal microscopic particles, and the measurement method of the heat is controlled by the space and scale of transmission. The traditional heat transfer measurement methods include the hot plate method, heat flow meter method, and thermocouple method. The hot plate method can detect the stability of the test material, which is often used to measure the thermal conductivity of the thermal insulation material. The heat flow method is used to measure thermal insulation, foaming, multiple layers, electronic materials, etc. The thermocouple method can be used as a temperature sensor due to its intuitive temperature feedback. All of the above methods involve applying a heat source to the system and using the temperature change to obtain thermal conductivity. There are also methods that allow heat transfer to be measured without affecting the internal conditions of the system. A natural gas hydrate is an ice crystal compound formed by gas and water molecules. At present, its heat flow density measurement methods mainly include heat flux measurement, specific heat capacity measurement, and effective thermal conductivity ETC measurement.

2.1. Heat Flow Density and Specific Heat Capacity Measurement

The measurement of heat flux density can be reflected by temperature changes in different positions or at different times. The heat transfer in the vertical or horizontal direction of the hydrate formation/decomposition process can be analyzed and the heat loss can be calculated. The heat flux measurement method requires installing temperature sensors at different positions to record the temperature changes, realizing the visualization of the heat transfer process, and analyzing the specific situation of heat transfer by moving the interface. Li et al. [16] analyzed the influence of reservoir heat transfer factors on hydrate decomposition through the change in temperature distribution in the process of natural gas hydrate formation and decomposition. The distribution of temperature sensors in the experimental device and reactor is shown in Figure 1. The formation and decomposition process of natural gas hydrate can be analyzed by comparing the temperature distribution at different times at each level. The temperature distribution analysis method of ultrasonic tomography is also applied to the measurement of heat flux density [17,18,19]; in this method, the pulses generated by the transducer pass through the measured structure and return to the echo characteristics. The optical path difference between continuous echoes can be calculated. The temperature of different segments can be obtained, and the temperature distribution of the entire path can be obtained along the ultrasound. In addition, the temperature distribution can be measured by adding thermocouples on both sides of the measured object. The temperature is converted into an electrical signal for detection based on the thermoelectric effect as a way to obtain an overall real-time temperature distribution [20].

Specific heat capacity is the amount of heat required to raise a unit mass of substance by 1 K under certain conditions, and it is a physical quantity used to measure the thermal properties of substances. Since the hydrate sediments of natural gas need to consume their own sensible heat during decomposition to compensate for the temperature required for decomposition heat, the determination of the specific heat capacity of hydrate sediments is an important basic parameter for establishing corresponding exploitation technology [21]. Handa et al. [22] measured the specific heat and decomposition heat of different hydrates by using an improved Tian–Calvet calorimeter, and the pressure of the modified calorimeter was much greater than the phase equilibrium pressure under the measurement conditions of the hydrate, which could then ensure that factors such as the decomposition heat of the hydrate were avoided. Based on experimental values, Handa proposed empirical relationships for the molar heat capacity applicable to muffle, stone, and pyre firings:

$$C_m=a_0+b_{0}T+c_0{T}^2+d_0{T}^3$$

(1)

In Equation (1), C_m is specific heat capacity (J/(mol·K)) and T is temperature (K). a₀, b₀, c₀, and d₀ are empirical polynomial coefficients.

2.2. Effective Thermal Conductivity Measurement

In the study of thermal conductivity of materials, effective thermal conductivity (ETC) is widely used as a key parameter to evaluate the heat transfer process [23]. ETC can measure the heat transfer performance of hydrates at various stages, and its methods can be roughly divided into two categories: steady-state method and unsteady-state method [24]. In the steady-state method, the temperature at any position does not change with time under the same boundary conditions. The measurement principle of ETC obeys Fourier’s Law. However, due to the long experimental time required by the steady-state method, the heat conduction and the temperature of several points need to be measured, and the hydrate needs to be preserved and stabilized under high-pressure and low-temperature conditions. Therefore, this method is less applied in the field of hydrates. The unsteady-state method has the advantages of short measurement time, high precision, and simple environmental conditions, so it is widely used. The unsteady-state method mainly includes the hot wire method, hot strip method, probe method, and transient plane heat source method. Each method has its own characteristics and application scope [25], as shown in

Table 1. Effective thermal conductivity measurement.

Method	Diagram	Formula	Literatures
Hot wire		$\lambda =\frac{q}{4\pi }\times \frac{dT}{d\ln \tau }$	From ref. [26] Copyright (1996), with permission from Elsevier.
Hot strip		$\lambda =\frac{q}{4\pi }\times \frac{dT}{d\ln \tau }$	From ref. [27] Copyright (2016), with permission from Elsevier.
Probe		$\lambda =\frac{q}{4T\pi }\ln \frac{4\alpha t}{B{a}^2}$	From ref. [29] Copyright 2002 by the American Geophysical Union.
TPHS		$R(T)=R_0(1+k\Delta T)$ $\Delta T(\tau )=\frac{P_0}{\lambda r{\pi }^{\frac{3}{2}}}D(\tau )$	From ref. [30] Copyright © 2011, Science China Press and Springer-Verlag Berlin Heidelberg.

.

The hot wire method is widely used to measure the ETC of granular materials [26]. The heat source of the test system is a wire with a uniform initial temperature distribution, which not only provides an internal heat source but also acts as a sensor to measure the temperature change. The thermal conductivity is derived by calculating the temperature versus time. The heat strip method is based on the hot wire method whereby a linear heat source is flattened into a strip in order to expand the contact area between the heat source and the medium, and the temperature response curve in the area of the metal strip is recorded [27]. The probe method was first proposed by Herzen and Maxwell [28]. The single-ring heating line is wrapped with a subcutaneous needle, and there is a temperature-induced thermistor in the middle of the subcutaneous needle. The transient plane heat source method improves upon the hot wire method, which is one of the most convenient and accurate methods to study the heat transfer characteristics of materials. This method uses an instantaneous thermal plane probe and uses the principle of linear relationship between temperature and probe resistance. In a hydrate system, considering the transferability of the hydrate phase, the unsteady transient plane heat source method is more commonly used to measure thermal conductivity and its change.

In Table 1, q is the intensity of the heat source (J/(m³·s)), τ is the time (s), T is the temperature (K), α is thermal diffusion coefficient of sediment sample (m²/s), a is probe radius (mm), B is a constant whose value is 1.7811, R₀ is the initial resistance of the probe (Ω), k is the resistance temperature coefficient (1/K), and ΔT is the average temperature rise (K) in the probe surface.

2.3. Improvement in Measurement Method for Thermal Conductivity of Natural Gas Hydrate

Different measurement methods have their limitations. When using the transient plane heat source method, the probe will be severely squeezed and twisted due to the randomness of the natural gas hydrate crystallization process and the change in density after hydrate formation. Although the hot wire method has been successfully used to measure the ETC of gases, liquids, and solids at atmospheric pressure, this method is not suitable for high-pressure conditions in porous media due to the deformation and fracture of the wire. To overcome these limitations, Stoll et al. [21] measured the thermal conductivity of propane hydrate at 275.15 K and 1.0 MPa and the thermal conductivity of methane hydrate at 275.15 K and 10 MPa by using the thermal pulse probe technique. Shi et al. [31] measured the thermal conductivity of R141b (CH₃CCl₂F) hydrate and R113 (CH₃CCl₂F) hydrate by using the plate steady-state method. When the sample reaches thermal equilibrium, the thermal conductivity can be obtained by measuring the heat flux and temperature gradient per unit area of the sample. Yang et al. [32] adopted the point heat source method based on a thermistor, in which the heat-affected zone is a small sphere, rather than a large cylinder area as in the plate heat source method and the transient hot wire method. Yang used this method to study the ETC of tetrahydrofuran (THF)-containing hydrate samples. The results obtained represent the effective properties of smaller regions and can better reflect the local properties of hydrates. Zhao et al. [33] proposed a new device composed of a reaction cell, a thermistor probe, a high-precision injection pump, and a data acquisition system for the easy measuring of the hot wire under high pressure. The device can directly synthesize the natural gas hydrate and measure its ETC in situ with a thermistor, which can reduce the influence of the heater itself on the reaction results.

3. Study of Heat Transfer Characteristics of Gas Hydrate

Natural gas hydrate reservoirs contain a complex system of porous media, hydrates, water, gases, and other multi-phase and multi-component systems. The forms of heat transfer between the components include heat conduction, heat convection, and heat radiation. Therefore, the heat transfer characteristics of natural gas hydrate are affected by porous media, hydrate, water, gas, and other multi-phase structural properties, physical properties, and other factors.

3.1. Effect of Porous Media Material on Heat Transfer Characteristics of Hydrate

Li et al. [34] carried out methane hydrate decomposition experiments by using the depressurization method in porous media with three different thermal conductivities: quartz sand, white corundum, and silicon. The results showed that with the increase in ETC in porous media, the heat transfer rate increased, and the temperature rise rate of methane hydrate also increased. Wang et al. [35] studied the heat transfer effect of porous media materials by using four different porous media with different heat transfer properties: quartz sand, natural sand, corundum, and silicon carbide. The results showed that the porous medium silicon carbide with a larger ETC can improve the ETC of the whole multi-phase sediment to a greater extent, promote heat transfer in the gas production process, and eliminate the hydrate reforming and freezing caused by depressurization effectively. Yang et al. [32] measured the ETC of THF hydrate-bearing sediments using silica sands, natural sands, corundum, and silicon carbide with different ETCs. The study found that porous media with a larger ETC will improve the heat transfer performance of hydrate-bearing sediments. Subsequently, Yang et al. [36] studied the ETC of methane hydrate formation in these four porous media, obtained the same conclusion, and predicted that the ETC of porous media materials has a linear relationship with the ETC of hydrate-bearing sediments. A two-dimensional axisymmetric model developed by Zhao et al. [37] studied the effect of heat transfer on the decomposition of methane hydrate by using the heat shock method. It was found that porous media with high specific heat capacity can provide more sensible heat in a short time in the early stage of hydrate decomposition and can accelerate the rate. However, the decrease in hydrate decomposition caused by high specific heat capacity reduces the gas production rate in this process. Therefore, from the perspective of heat transfer, the reason why the decomposition rate of methane hydrate increases with the increase in ETC in porous media is that the heat of methane hydrate decomposition is provided by the sensible heat of porous media. This heat, as an in situ heat, can be immediately used for hydrate decomposition, so the decomposition rate of methane hydrate is also faster.

3.2. Effect of Porosity of Porous Media on Heat Transfer Characteristics of Hydrate

During the process of hydrate formation in porous media, fluid in the pore space of porous media contributes to heat conduction inside the pores and establishes heat transfer channels between particles for particle–fluid–particle heat transfer [38]. Priest et al. [39,40] and Best et al. [41] considered the change in heat transfer characteristics by changing the pore conditions of porous media and adjusting the distribution of hydrates. Those studies found that if the hydrate is suspended in the pore fluid, which is of the pore-filling type, there are some migration channels between the hydrate and the sediment so that the pores are filled with pore water, dissolved gas, or free gas; when the pore water is too much, it is easy to form the contact type. Tupsakhare et al. [42] studied the decomposition process of methane hydrate in porous media by using the thermal extraction method and found that free water in the pores of porous media played an important role in the process of transferring heat to hydrates far away. Dai et al. [23] emphasized the importance of non-gas phase distribution in the pore space to the thermal conductivity of unsaturated sediments. The increase in thermal conductivity may be caused by the redistribution and migration of water to neighboring contacts. When hydrate formation is associated with volume expansion, the contact area increases and water may extend into the hydrate grains and neighboring contacts, thus reducing porosity and increasing thermal conductivity. Yang et al. [32] found that the effective thermal conductivity of THF hydrate deposits was negatively correlated with porosity through experiments due to the fact that the thermal conductivity of porous media in the pore space is lower than that of hydrate crystals. Wei et al. [43] analyzed the ETC changes in hydrate-bearing sediments and THF hydrates in the South China Sea. Compaction operations reduce porosity and expel air, leading to a significant increase in the ETC of sediment-containing hydrates. The study showed that the ETC of hydrate-bearing sediments is negatively correlated with porosity. It is suggested that pore expansion may lead to a significant decrease in ETC, and pore space reconstruction may lead to heat transfer barriers. The low thermal conductivity of hydrate-bearing sediment samples in the South China Sea is attributed to the high porosity and irregular shape of the sediments. Wu et al. [44] proposed a new model for the ETC rate of gas hydrate-bearing sediments that integrates hydrate saturation and pore morphology evolution. Their findings suggest that when the intensity of pore morphological changes correlates with the pore structure and hydrate formation conditions, the ETC escalates in tandem with the increase in pore change intensity, which is contrary to the results of Yang’s [32] and Wei’s [43] study. As the hydrate pore morphology alters, given a constant water content and pore volume, the correlation between porosity and ETC is predominantly influenced by alterations in pore composition. As hydrate saturation increases, free gas is supplanted, leading to a greater hydration space occupation within the pores. Both an increase in water content within the pores and a decrease in gas content contribute to an elevated ETC in multi-phase samples. Similarly, Li et al. [30,45] evaluated the ETC of quartz sand samples containing methane hydrate formed in quartz sand with different porosities. The study found that porosity has a very obvious effect on thermal conductivity, and the ETC increases with the increase in porosity; they also found that the ETC of hydrate-containing porous media depends on the location of hydrate formation in the pore space, and the pore space also depends on the formation environment.

3.3. Effect of Particle Size of Porous Media on Heat Transfer Characteristics

Farahani et al. [46] in their study on soil thermal conductivity proposed that heat transfer in porous materials is controlled by the following particle-level mechanisms (as shown in Figure 2): (1) conduction within the particle, (2) particle–particle conduction through contact area, (3) particle–fluid–particle conduction, (4) radiation at the contact between particles, (5) conduction in pore fluids, (6) convection in pore fluids, (7) radiation media from the particle surface to the surrounding environment, and (8) radiation into the surrounding medium. Heat transfer in porous media hydrates is similar to that in soil particles, and the main factor determining the overall ETC comes from the porous media particle-level transport processes. Cortes et al. [38] studied the heat transfer characteristics in THF hydrate-bearing sediments. The particle-level heat transfer process in porous media is preferentially conducted through minerals, followed by contact conduction between particles, and finally the adjacent particle–fluid–particle contact conduction in the fluid, as well as conduction through the pore space. Jaeger et al. [47] and Nasirian et al. [48] pointed out that the contact between particles plays a key role in the heat transfer of granular media. The heat transfer between particles is proportional to the particle radius [49]. Midttomme et al. [50] posited that the rate of heat transfer in sediment increases with an increase in the particle size of the porous medium. According to the theory of heat conduction, the law is to preferentially find the path with the smallest thermal resistance. If a large contact thermal resistance exists between the porous media particles, this resistance attenuates the ETC. Li et al. [16] generated hydrates and decomposed them in sediments with four particle sizes of 14–20 mesh, 35–60 mesh, 80–120 mesh, and 400–500 mesh. From the analysis of temperature changes, it is concluded that the heat conduction rate in porous media with smaller particle sizes is also lower, which is consistent with the proportional relationship obtained by Midttomme et al. [50]. In addition, Zhan et al. [51] studied the hydrate dissociation behavior in sediments with different particle sizes, it was shown that the heat transfer of sediments increased with the increase in particle size. However, it is not accurate to judge ETC only according to the particle size. Ahn et al. [52] discovered that when the proportion of fine particles in the mixed particles is high, ETC will also increase. Wang et al. [53] employed the thermal extraction method to decompose hydrates and examined its correlation with sediment particle size. Their findings suggest that both the heat transfer rate and energy efficiency diminish as sediment particle size decreases. Larger particle sizes corresponded to enhanced permeability, potentially leading to a more expansive fluid flow sweep area, which is conducive to the heat supply of hydrate decomposition. However, in the actual hydrate reservoir, the pore and particle size of the sediment may be much larger than the artificial medium simulation used in the laboratory. The particle size distribution is shown in

Table 2. Particle size distribution of porous media.

Category	Porous Media Materials	Normal Particle Size
Natural sediments	Silicon dioxide	25–58 μm
	Grit	1.0–2.0 mm
	Montmorillonite	0.5–25 μm
	Clay	0.25–0.425 mm
Artificial sediments	Silica sand	300–450 μm
	Activated carbon	2.19 nm
	Quartz sand	13–27 nm
	Glass bead	0.105–1.397 mm
	Silica gel	0.105–0.150 mm

, and the effect of this difference on hydrate decomposition needs further study.

3.4. Effect of Hydrate Saturation and Water Saturation on Heat Transfer Characteristics

Yang et al. [32] obtained the THF hydrate ETC at 274 K by using THF and deionized water to form THF hydrates with different saturations in silica sand at different concentrations. When the porosity is constant, the lower the THF hydrate saturation, the higher the water saturation in the pores, and the higher the ETC, which helps to improve the ETC of hydrates. Dai et al. [23] measured the thermal conductivity of samples with different hydrate saturations and found that gas saturation was the dominant saturation component in unsaturated samples. The ETC of unsaturated hydrate-bearing sediments is largely affected by the volume fraction, pore size distribution of the hydrate, and the aqueous phase. When the porosity remains constant, the lower the saturation of the THF hydrate and the higher the water saturation in the pores, resulting in a higher ETC [32]. Dai et al. [23] studied the change in ETC during the formation of methane hydrate and found that ETC increased after hydrate formation when the hydrate saturation was less than 20%, while ETC decreased after hydrate formation when the hydrate saturation was 30~40%. The analysis showed that the increase was due to the increase in contact area caused by volume expansion during hydrate formation, while the decrease was due to the expansion of the sedimentary skeleton or contact separation of particles during hydrate formation, and the contact mode may change from particle–particle to particle–hydrate–particle or even particle–gas. Wang et al. [35] interpreted this relationship as follows: with increasing hydrate saturation, free gas is replaced and more pore space is occupied by the hydrate; the ETC of the hydrate-bearing sediments significantly changes with increasing hydrate volume in pore space. The effect of water saturation is due to the fact that water acts as a relatively large heat transfer bridge to promote heat transfer between particles [54]. Wei et al. [55] attributed this relationship to the correlation of cementation and heat transfer between particles in porous media and the easy phase transition property of the low-saturation sample skeleton. Subsequently, they studied the effect of hydrate distribution on ETC in the process of hydrate formation in quartz sand. With the increase in water saturation, the water distribution gradually expanded, and some heat transfer paths changed from ‘particle–water–particle’ to ‘particle–particle’, making ETC increase [56].

In view of the effect of water/hydrate saturation on ETC in different actual situations, some research has established models based on theory to analyze the causes of the influence. Sun et al. [25] established a model for the ETC of hydrate-bearing sediments. After simulation, it was concluded that the relationship between ETC and methane hydrate saturation is saturated > supersaturated > unsaturated. Therefore, there is a critical saturation of hydrate; when hydrate decomposition makes most of the cementation between porous media disappear, the cementation rate decreases rapidly, which is the critical saturation. Li et al. [57] pointed out that when the saturation of the hydrate decreases to the critical hydration saturation, the decomposition of the hydrate leads to a weakening of cementation between porous media and the heat transfer rate will decrease rapidly. Wu et al. [44] established a model for the ETC of natural gas-containing hydrate sediments. It was found that under the critical saturation of the hydrate, when the pore morphology of the hydrate changed from pore filling to particle coating, ETC decreased significantly but was not obvious. However, in an excess gas environment, ETC was significantly enhanced at critical hydrate saturation. Therefore, the ETC of hydrate-bearing sediments is very sensitive to hydrate saturation near the critical hydrate saturation threshold.

A gas hydrate reservoir is a multi-component complex sediment system, and the thermal conductivity of these components is different by one order of magnitude [38,58]. Therefore, the ETC of the gas hydrate reservoir is more complex, which is not only related to porous media material, pore structure, and particle size distribution, but also related to water content, hydrate saturation, and distribution. In a gas-saturated environment, ETC is mainly affected by hydrate saturation because the thermal conductivity of the hydrate is higher than that of gas. Whereas, in a water-saturated environment, ETC is mainly affected by water saturation [59].

3.5. Effect of the Hydrate Dissociation Process on Heat Transfer

Hydrate decomposition is a solid–gas–liquid phase transition process. In their study on propane hydrate decomposition, Kamath et al. [60] divided the process of hydrate decomposition into three stages, solid stage, liquid stage, and gas stage, emphasized the influence of the phase transition process on heat transfer, and pointed out that the bubble generated by the rapid release of gas from hydrate decomposition dominates the heat transfer process. Later, Kamath et al. [61] further proposed that hydrate decomposition is in turn limited by the heat transfer process in their study on methane hydrate decomposition. The water produced by hydrate decomposition continuously forms a thin liquid film on the surface of the undecomposed hydrate, thereby generating resistance to heat transfer. Zhao et al. [62] divided the decomposition region into three regions, the decomposed region, the decomposing region, and the hydrate region, when analyzing hydrate decomposition in porous media. The heat transfer path between regions is mainly transferred from the decomposing zone containing the hydrate to the decomposing zone containing gas and water. Their study found that the convection caused by the gas produced by the decomposition reduced the ETC and heat capacity of the hydrate. This confirms the conclusion previously mentioned by Pooladi-Darvish et al. [63] that gas convection leads to a decrease in heat capacity and gas production rate. Kou et al. [64] refined the heat transfer effect of the hydrate dissociation stage (as shown in Figure 3): stage 1 is the depressurization stage, where the decomposition heat source is mainly provided by the heat conduction from the environment; stage 2 is the constant pressure stage, where the heat source is mainly provided by the sensible heat, and the partially decomposed hydrate structure provides a new dissipative channel for heat transfer. When the sensible heat of hydrate-bearing sediments is consumed, the heat conduction from the environment becomes the main driving force of decomposition. Zhang et al. [65] established a hydrate dynamic decomposition model to simulate the decomposition process of hydrate particles in flowing water. When hydrate particles move in water, the heat transfer mode is heat conduction inside the particles and heat convection between hydrate particles and water. In the early stages of hydrate particle decomposition, the heat transfer on the surface of hydrate particles is dominant. When the surface temperature of hydrate particles gradually stabilizes, the internal heat transfer of the hydrate plays an important role, and the heat transfer rate of hydrate particles cannot be ignored. When considering the heat transfer of hydrate decomposition, Hao et al. [66] proposed that the two heat transfer modes of heat conduction and heat convection should be considered at the same time, as heat convection is the key factor of natural gas hydrate decomposition. Different from the influence of gas or water disturbance on heat transfer in the process of water and matter decomposition mentioned above, some scholars have analyzed the influence of hydrate heat transfer through the change in hydrate structure in the process of hydrate decomposition. In the work of Kanda et al. [67], they used a high-speed phase-shifting interferometer to observe the interfacial heat transfer phenomenon of methane hydrate and calculated the apparent density difference distribution near the interface. Wei et al. [43] used this idea to observe the changes in hydrate density and density distribution during the phase transition of hydrate decomposition by comparing the ETC changes in the THF hydrate during the decomposition process. It was concluded that the hydrate decomposition process destroyed the heat transfer path in the original hydrate system, which reduced the interface thermal resistance, in turn leading to an increase in the ETC of THF hydrate. He et al. [59] proposed an explanation for the change in the specific heat transfer path in the decomposition stage of methane hydrate. In the early stages of hydrate decomposition, the ETC increased due to the expansion of the water distribution range. In the later stages, the heat transfer path changed from ‘particle–water–particle’ to ‘particle–gas–particle’, and the contact mass between particles decreased, resulting in a downward trend of ETC.

3.6. Effect of Environmental Heat Conduction on Hydrate Dissociation Process

Li et al. [68] observed the evolution of the hydrate temperature gradient and the increase in temperature in the decomposition area of the hydrate boundary area in the experimental and simulation study of the hydrate depressurization decomposition process in sediments. In their study, the energy transferred from the surrounding environment was used as the latent heat in the process of hydrate decomposition equilibrium, and then, it was concluded that the heat transfer around the reservoir has an effect on the depressurization decomposition process. Li et al. [69] further analyzed the heat transfer characteristics of depressurized decomposition of methane hydrate near the freezing point and found that heat conduction from the environment is mainly the sensible heat of the reservoir, the boundary conduction heat, and the transition latent heat of ice. Wan et al. [70] found that under thermal stimulation, heat conduction plays a role in the early stages of hydrate decomposition, while hydrate decomposition mainly transfers heat from the environment under depressurization conditions. Oyama et al. [71] pointed out that the dissociation process of the hydrate is dominated by the heat transfer process around the hydrate core by analyzing the temperature distribution changes in the core center and the wall surface during the hydrate dissociation process. Heat transfer affects the temperature distribution in the hydrate, thereby controlling the hydrate decomposition process. Chen et al. [72] found that the combination of pressure and heat transfer in the hydrate decomposition process is the main driving force. In the early stages of decomposition, the driving force comes from the pressure and sensible heat of the sediment hydrate. When the temperature decreases to the phase equilibrium temperature, the decomposition enters the later stage. The driving force of this process comes from the pressure of the surrounding environment and the external heat transfer. In the process of decomposition, free water will migrate to the upper part of the system and affect the heat transfer. Robert et al. [73] and Mehran et al. [74] used numerical simulation to study the heat transfer process of hydrate depressurization decomposition with phase changes in porous media and proved that heat transfer and temperature gradient distribution inside the reservoir are the main driving forces of the depressurization decomposition process. It can be seen that heat transfer, as the basic control mechanism of the hydrate decomposition process, directly affects and controls the hydrate decomposition rate and decomposition efficiency, and has a crucial impact on the efficiency of natural gas hydrate decomposition.

4. Predictive Model for ETC

A natural gas hydrate reservoir is a multi-phase and multi-component system; its heat transfer process includes the heat conduction of multi-phase components such as porous media of sedimentary layers, hydrate, salt ions in solution, gas, ice, and water. At the same time, these multi-phase components migrate during the hydrate exploitation process, and the heat will be transferred with the seepage processes such as convective mass transfer and molecular diffusion mass transfer. Therefore, multi-phase heat transfer and multi-phase seepage in the process of hydrate exploitation are coupled with each other. It is of great scientific significance to clarify the variation law of multi-phase and multi-component ETC for the study of the coupling control mechanism between the hydrate decomposition reaction, heat, and mass transfer and multi-phase seepage in the process of hydrate exploitation.

The commonly used models for the effective thermal conductivity of the local heat balance condition are the parallel model, continuous model, discrete model, and square root model, as shown in

Table 3. Commonly used effective thermal conductivity models.

Model	Formula
Parallel model	${\lambda }_e=({\lambda }_H{S}_H+{\lambda }_G{S}_G+{\lambda }_W{S}_W)\varphi +{\lambda }_s(1-\varphi )$
Continuous model	${\lambda }_e={\left[(\frac{S_H}{{\lambda }_H}+\frac{S_W}{{\lambda }_W}+\frac{S_G}{{\lambda }_G})\varphi +\frac{(1-\varphi )}{{\lambda }_S}\right]}^{-1}$
Discrete model	${\lambda }_e={\lambda }_H^{\varphi S_H}\cdot {\lambda }_W^{\varphi S_W}\cdot {\lambda }_G^{\varphi S_G}\cdot {\lambda }_S^{(1-\varphi )}$
Square root model	${\lambda }_e={\lambda }_H\sqrt{\varphi S_H}+{\lambda }_W\sqrt{\varphi S_W}+{\lambda }_G\sqrt{\varphi S_G}+{\lambda }_S\sqrt{1-\varphi }$

[75]. In Table 3, λ_e is ETC, and λ_H, λ_W, and λ_G are the pure thermal conductivity of hydrate, water, and gas.

But in hydrate porous media, it is not only these three conditions that are considered above. Sun and Mohanty proposed the following model [76] under local thermal equilibrium conditions:

$$\lambda =\varphi {\displaystyle {\sum }_{i=W,G,H,I,S}{S}_i{\lambda }_i}+(1-\varphi ){\lambda }_R$$

(2)

In Equation (2), λ_R and λ_S are the ETC of porous media and salt ions, and their values are constants, according to the types of porous media and salt ions; λ_H is the ETC of hydrates, and 0.49 W/(m·K) is generally used for sI methane hydrates [75,76]; λ_G, λ_I, and λ_W are the ETC for the gas, ice, and water phase, and their value is a function of temperature, evaluated according to the empirical correlations described by Sun and Mohanty [77].

5. Study of Heat Transfer Mechanism of Hydrate

The study of heat transfer mechanisms will provide an explanation for ETC research under complex systems with different components and different conditions, and provide a basic theory for the heat transfer process in the process of hydrate exploitation.

5.1. Methods for Studying Heat Transfer Mechanisms on a Molecular Scale

Due to the complex composition and structure of natural gas hydrate, it is a great challenge to study the heat transfer behavior of natural gas hydrate through laboratory observation. Molecular dynamics simulation is an important research method to study the structure and properties of molecules or molecular systems through computer simulation, which can effectively analyze the microscopic reaction process on a molecular scale and has been widely used in the study of hydrate formation and decomposition mechanisms [78,79,80]. Using molecular dynamics simulation to analyze hydrate heat transfer can separate various promotion or inhibition effects on the heat transfer process from complex systems [81] and analyze the thermal transport performance of the system on a molecular scale. The application of molecular dynamics simulation in the study of the heat transfer process can be divided into equilibrium molecular dynamics simulation (EMD) and non-equilibrium molecular dynamics simulation (NEMD) according to the equilibrium state of the system [82]. EMD is a method based on the Green–Kubo relation to calculate the dynamic behavior of molecular systems in equilibrium, which is mainly based on the fluctuation–dissipation theory and the linear correspondence theory. In the linear response range, the ETC given by the Green–Kubo relationship is an integral of the heat flux vector autocorrelation function (HACF) of the system, and the equation is as follows [83]:

$$\lambda =\frac{1}{3k_B\mathrm{V}{T}^2}{\displaystyle {\int }_0^{\infty }<\stackrel{\rightharpoonup }{q}(t)\cdot \stackrel{\rightharpoonup }{q}(0)>} dt$$

(3)

In Equation (3), k_B is the Boltzmann constant; its value is 1.38066 × 10⁻²³ J/K.

For the two-body Lennard–Jones (L-J) potential, the effective heat flux q(t) is expressed as follows:

$$q(t)={\displaystyle \sum _i{e}_i{\overrightarrow{v}}_i+\frac{1}{2}{\displaystyle \sum _{i\ne j}({\overrightarrow{f}}_{ij}\cdot {\overrightarrow{v}}_j){\overrightarrow{r}}_{ij}}}$$

(4)

In Equation (4), ${\overrightarrow{v}}_i$ is the velocity of particle i, ${\overrightarrow{r}}_{ij}$ is the distance between particle i and j (${\overrightarrow{r}}_{ij}={\overrightarrow{r}}_i-{\overrightarrow{r}}_j$), and ${\overrightarrow{f}}_{ij}$ is the force of particle j on i; $e_i=\frac{1}{2}{m{v}_i}^2+\frac{1}{2}{\displaystyle \sum \varphi (r_{ij})}$ and $e_i=\frac{1}{2}{m{v}_i}^2+\frac{1}{2}{\displaystyle \sum \varphi (r_{ij})}$.

NEMD is a method based on the calculation of the dynamic behavior of molecular systems under non-equilibrium conditions. NEMD can be used in systems with irreversible heat flow and temperature gradient. The NEMD method can be divided into two methods: the isotropic NEMD method and the anisotropic NEMD method. The isotropic NEMD method ensures that the momentum of the whole system remains constant by adding a heat flux perturbation term that depends on the simulation time to the regular description of the equilibrium system [82]. The anisotropic NEMD method is used to simulate the non-periodic boundary and anisotropic heat transfer problems in the system space. When using the NEMD method to simulate heat transfer, the non-equilibrium heat transfer process is established by applying disturbance to the system, and then, the ETC of the system is calculated according to the Fourier heat conduction law. The equation is as follows:

$$\overrightarrow{J}=-\lambda \nabla T$$

(5)

5.2. Host Molecular Heat Transfer Mechanism

The main water molecules in the hydrate form cages through hydrogen bonding, and the interaction between the host water molecules dominates the hydrate crystals. The study of its thermal conductivity mechanism will help to analyze the heat transfer characteristics of hydrates. Phonon is the main carrier in the heat transfer process of hydrates, and it is the quantized form of lattice vibration [84]. The phonon heat transfer mechanism is shown in Figure 4 [85]. Usually, the phonon mode is fitted into two modes: long-range phonons and short-range phonons [86]. Liu et al. [87] analyzed the heat current autocorrelation function (HCACF) from the perspective of phonons using Equation (4). The relaxation time of phonons can be obtained by decomposing HCACF, and then, the influence of various modes of phonons on heat transfer can be analyzed. It is found that the interaction between host molecules is dominant in the heat transfer of methane hydrate. When the interaction between host and guest molecules is enhanced, it is helpful to improve the thermal conductivity of methane hydrate.

$$HCACF(T)={\displaystyle \sum _{i=1}^{n_{ac}}{A}_i\exp (-\frac{t}{{\tau }_i})}+{\displaystyle \sum _{j=1}^{n_{opt}}\left[{\displaystyle \sum _{k=1}^{n_{0,j}}{B}_{jk}\exp (-\frac{t}{{\tau }_{jk}})}\right]\cos {\omega }_{0,j}t}+{\displaystyle \sum _{j=1}^{n_{opt}}{C}_j{\omega }_{0,j}t}$$

(6)

English [88] used EMD to predict the ETC of sI-type methane hydrates and high-pressure sII-type and sH-type polymorphs in a wide temperature range. Comparing the contribution of water–water, water–methane, and methane–methane to heat transfer by analyzing the heat flux correlation function between water and methane, it is shown that the main water skeleton plays a leading role in the heat transfer process. Zhou et al. [89] used the mode-dependent thermal conductivity method to measure the contribution of each phonon to the total thermal conductivity. Based on the single-mode relaxation time approximation, the thermal conductivity of the system is described by the Boltzmann transport equation. By comparing the thermal conductivity values obtained by the phonon spectral energy density function and the EMD method, it is concluded that the difference between the two is because there are still non-propagating vibrations in the conduction process, and these vibrations will diffuse heat. Similarly, English et al. [90] simulated the thermal conductivity of sII methane hydrate. From the analysis of heat flux vector data, the slow decrease in long-range acoustic relaxation time shows that the energy transfer at this stage is mainly dominated by the host lattice behavior. The heat flux vector damping is due to the tight object–host contact in the double-occupied small cavity, and the damping leads to the amorphous behavior of heat transfer characteristics. By analyzing the radial distribution function, Li et al. [91] first explained that the hydrate structure will not change due to the lack of guest molecules, but the lack of host water molecules will lead to a large number of phonons scattering during heat transfer, which greatly affects the heat transfer characteristics. Yuan et al. [92] used NEMD simulation to study the thermal conductivity of the THF hydrate. The simulation results have the same low thermal conductivity at low temperature and non-negligible temperature dependence as the experiment. The projected density of state (PDOS) data indicate that the vibration of THF molecules exists in the low-frequency acoustic region of the main crystal lattice. On the other hand, from the velocity cross-correlation function (VCCF) between THF and water molecules, the correlation between water and THF molecular motion is proved. The two molecular vibrations are similar, so the influence of the object on the thermal conductivity of the hydrate itself can be ignored. Finally, it is concluded that the ultra-low thermal conductivity of the THF hydrate should be attributed to its complex large cell and distorted water cage crystal structure. Wan et al. [93] used EMD to simulate the thermal conductivity of sI-type methane hydrate with different crystal cavity occupancy at 263.15 K. The calculation shows that the poor heat transfer characteristics of methane hydrate are determined by the cage structure constructed by the main molecule.

5.3. Guest Molecular Heat Transfer Mechanism

The guest molecules in the hydrate occupy the cage constructed by the host water molecules, which has an important influence on the stability of the hydrate and the crystal structure of the hydrate. In general, the combination of non-harmonic interactions between single-crystal atoms and phonon scattering due to crystal defects results in a non-monotonic trend in thermal conductivity with temperature. However, as a crystal, natural gas hydrate shows the characteristic that the thermal conductivity of glass-like hydrates increases monotonously with temperature. Anderssson and Ross [94] studied hydrates containing 1,3-dioxolane, cyclobutanone, and THF as guests, respectively; the temperature-dependent characteristics of the thermal conductivity of all three cage hydrates clearly show glassy properties and differ due to the variation in guest species. Zakrzewski et al. [95] found that the addition of guest molecules makes the hydrate lattice become anharmonic and reduces the thermal conductivity of hydrates by comparing the thermal conductivity of empty hydrates and hydrates containing guest molecules. Zakrzewski thought that this feature is similar to glass-like materials. In the analysis of this situation, Zakrzewski only vaguely described its properties as ‘similar‘. However, in fact, in the previous work of Ross et al. [96], it was found that hydrates are abnormally dependent on temperature and pressure. Although hydrates are regarded as crystal-like substances, hydrates have some disordered structures. In the temperature range of 100–300 K, the guest molecules in hydrates will be redirected constantly, which provides a new scattering method. The disorder of guest molecules greatly increases the thermal resistance of hydrates and further affects heat transfer. Later, Inoue et al. [97] further analyzed the glass-like properties of hydrates and found that hydrates are only similar to the heat transfer behavior of glass at lower temperatures. The abnormal phenomenon that the thermal resistance of hydrates is not proportional to temperature is due to the translational motion of guest molecules. Different from the way in which the host molecule directly affects the heat transfer characteristics of hydrates through the water skeleton, the guest molecule affects the heat transfer process of hydrates by changing the vibration of the host molecule or by enhancing the phonon scattering. Most of the above studies used infrared or inelastic neutron scattering methods to detect the phonons at the center or boundary of the hydrate region. Tse et al. [98] chose to analyze the internal phonons in the Brillouin zone. They used a synchrotron radiation light source to excite nuclear resonance to study hydrates with Kr atoms as guest molecules. This method can characterize the local vibration of Kr atoms as guests effectively. It is observed that the vibration mode of guest molecules is a very obvious anharmonic motion, and this anharmonic motion is the cause of the very low thermal conductivity of hydrates; this is also an important manifestation of the host–guest coupling vibration. Therefore, the slight change in the guest molecule does not affect the glass-like heat transfer behavior of the hydrate. Wang et al. [99] attributed the heat transfer behavior of this type of glass to the fact that the guest molecules become more dispersed as the temperature increases, and the contribution of the guest molecules to the energy transfer process is also increasing. Rosenbaum et al. [100] obtained the ETC of methane hydrate by experiment and EMD simulation. In the simulation process, three water models and five rigid methane models were used to study the thermal conductivity of methane hydrate with a guest molecular occupancy of 80–100% in the crystal cavity. The simulation results also show that the ETC of the hydrate increases with the decrease in crystal cavity occupancy; they analyzed this phenomenon from the perspective of intermolecular forces and suggested this phenomenon is due to the fact that the hollow crystal cavity not occupied by guest molecules weakens the hydrogen bond stability between guest molecules and water molecules. This weakening effect is more conducive to the stability of the hydrate system and thus improves ETC. Through the comprehensive analysis of the heat flux correlation function, English et al. [81] clarified the effect of guest molecules on heat transfer dissipation in sI hydrates at a temperature greater than 100 K, and found that they had little effect on heat transfer dissipation in sII and sH hydrates at a temperature greater than 200 K. This is because the cage of sI hydrates is smaller than that of sII or sH hydrates, which makes the guest methane molecules and water molecules move closer together and allows the energy transfer between the host and the guest to occur more easily.

6. Development and Discussion of Heat Transfer in a Multi-phase Complex System

Hydrate-bearing sediments are a multi-phase complex system. It is a great challenge to accurately measure the ETC of hydrate-bearing sediments under the phase equilibrium conditions of gas hydrate. In the ETC measurement of hydrate-bearing sediments, the contact temperature sensors are usually inserted into the sediment and in full contact with the measured sample. Thermometers and temperature sensors can only measure discrete temperatures at independent single points, and continuous temperature field data cannot be obtained directly. In addition, the used temperature sensors may affect the properties and structure of the aqueous sediment, as well as the gas–water migration behavior in porous sediment samples. More importantly, most of these temperature sensors used in the literature can obtain data from discontinuous temperature measurement points only. Li et al. [101] applied infrared thermography to measure heat transfer in sediment hydrates; this method receives the infrared radiation of the hydrate surface through an infrared detector, generates images of continuous temperature field distribution, and provides a wide detection temperature range without touching and destroying the hydrate system. Its experimental setup is shown in Figure 5. In addition, in the transient liquid crystal measurement method, heat transfer measurements are accomplished by coating the surface of an object with a thermochromic liquid crystal [102], and in the liquid crystal thermal imaging method, heat transfer measurements are performed by comparing the color change in the thermochromic liquid crystal before and after the measurement [103]. The laser flash method is usually used to determine the thermal diffusivity of solids, where the absorption of pulsed energy causes the temperature of the test material to change and the thermal diffusivity properties of the material are calculated [104]. For micro/nano-scale substances, the measurement can be conducted by methods such as Raman thermometry [105].

These methods can study the heat transfer of the system more comprehensively without affecting the temperature, pressure, and other conditions of the original system, which also provides a new idea for the study of heat transfer characteristics of natural gas hydrates in the future. But the current in situ measurement of hydrate heat transfer is only small-scale testing; how to achieve in situ measuring and tracking on a larger scale, or even actual exploration, will be a big challenge in the future.

The objects faced by the hydrate exploitation process are mostly hydrate-bearing sediments, where multi-component thermal conductivity coefficients all have an impact on the system.

Table 4. Thermal conductivity of porous media.

Porous Media	Particle Size/μm	Porosity/%	Thermal Conductivity /(W·m−1K−1)	Studies
Quartz sand	20–63	41	0.45	[50]
Quartz sand	200–500	38	0.83	[50]
Quartz sand	600–830	35	1.03	[50]
Quartz sand	362	42	0.275	[110]
Silica sand	431	41	0.35	[111]
Silica sand	330	37	0.73	[52]
Silica sand	840	37	0.55	[52]
Silica sand	74	40.5	0.174	[112]
Silica sand	150	40.5	0.234	[112]
Silica sand	420	42.5	0.245	[112]
Silica sand	590	37.5	0.267	[112]
Silica sand	1680	39.0	0.402	[112]
Kaolin (300 °C)	-	40.0	0.2	[113]
Sea mud (Joetsu Basin, eastern Japan Sea)	-	70	0.90	[114]
Sea mud (Nankai Trough, Japan)	8.9	36	1.05	[115]
Sea mud (Ariake Sea, Japan) (undried)	-	-	0.7	[116]
Sea mud (Ariake Sea, Japan) (undried)	-	-	0.09	[116]

and

Table 5. Thermal conductivity of solutions and gas.

Gas/Liquid	Temperature/K	Thermal Conductivity/(W·m−1K−1)	Studies
Water	298.78	0.609	[117]
Ice	263.15	2.28	[118]
Sea water	273.15	0.581	[119]
Air	273.15	0.024	[120]
Methane	260.05	0.086	[121]
Methane (gas)	300.06	0.114	[121]
Propane (liquid)	261.72	0.123	[122]
Propane (gas)	300.35	0.018	[122]
NaCl solution	298.15	0.594	[123]
NaCl solution	298.15	0.583	[123]
NaCl solution	298.15	0.574	[123]
CaCl2 solution	293.15	0.55	[124]
CaCl2 solution	293.15	0.602	[124]

list the inherent ETC of porous media, gas, and solutions in the multi-phase components of natural gas hydrate. From Table 5 and

Table 6. Thermal conductivity of the hydrate in porous media.

Porous Media	Guest Molecules	Particle Size/μm	Porosity/%	Thermal Conductivity /(W·m−1K−1)	Studies
Quartz sand	CH4	50	39	0.421	[16]
Quartz sand	CH4	100	32	1.15	[125]
Quartz sand	CH4	150	40	0.457	[16]
Quartz sand	THF	125–250	42.4	2.58	[126]
Quartz sand	CH4	300–125	47	0.999–1.024	[127]
Quartz sand	CH4 + C2H6 + C3H8 + THF	300–125	47	1.238–1.294	[128]
Quartz sand	THF	300–125	47	1.879–1.967	[127]
Quartz sand	CH4	250–425	41	1.568–1.493	[45]
Quartz sand	CH4	250–425	44	1.32–1.5	[45]
Quartz sand	THF	250–500	40.5	2.65	[126]
Quartz sand	CH4	500	40.5	0.470	[16]
Quartz sand	THF	500–1000	40.5	2.66	[126]
Quartz sand	CH4	1150	42.5	0.46	[16]
Silica sand	THF	430	37.5	0.725	[32]
Silica sand	THF	800	39.0	0.7	[32]
Silica sand	THF	1190	40.0	0.675	[32]
Silica sand	THF	2000	40.5	0.66	[32]
Silica sand	THF	4330	42.5	0.649	[32]
Sea mud (Nankai Trough, Japan)	CH4	87	46.7	1.51	[115]
Sea mud (Nankai Trough, Japan)	CH4	140	46.7	1.65	[115]
Sea mud (Nankai Trough, Japan)	CH4	215	41.0	1.55	[115]
Sea mud (northern South China Sea)	THF	<60 (90%) + 60–140 (10%)	42.5	0.61	[43]
Sea mud (offshore Qingdao, China)	CH4	63–250 (22%) + 250–500 (78%)	40	1.134–1.270	[25]

, it can be seen that the order of the inherent ETC of multi-phase components is approximately ice > water > salt ions > hydrate > porous media > gas. The thermal conductivity of each phase and component is not the same, and if the proportion of each component is different, it will have a different contribution to the heat conduction coefficient of the whole system. Among these conditions, water saturation and salt ion content in the natural gas hydrate reservoir are the key factors affecting the thermal conductivity of the system. Water acts as a relatively great bridge of thermal conductivity, promoting heat transfer between grains. Therefore, the effective thermal conductivity of partially saturated porous media can increase with the degree of water saturation. As for salt ions, according to Duan et al.’s [106] study, salt ions reduce the potential of the porous media particles by the “compressed double layer” principle, thus increasing the contact between the particle surfaces. This principle is shown in Figure 6.

The previously mentioned Si and φ in Equation (2) can be calculated by experiment, but the distribution of multi-components and the relationship between components (such as the dissolution of gas in solution, etc.) will be the key factors affecting ETC under local thermal equilibrium conditions, which increase the accuracy and repeatability of ETC test. Table 6 lists the ETC of hydrate-bearing porous media tested under different conditions. It can be seen from this table that the ETC under different conditions is quite different. Under the coupling of multi-phase heat transfer and multi-phase seepage, it is difficult to obtain a unified law due to the influence of multiple factors. Therefore, analyzing the influence of various key factors on ETC will be an important step to accurately obtain the ETC of natural gas hydrate reservoirs. Most of the current methods of in situ measurement of natural gas hydrates are still through the direct insertion of heat source and sensing in the system, but this method will definitely change the original system. Therefore, how to measure the hydrate system in situ in a comprehensive way without affecting the original conditions will be an important method to understand the heat transfer characteristics of natural gas hydrates. At present, the rapid development and wide application of artificial intelligence (AI) technology provide a convenient way to study the physical properties of multi-component complex systems. Machine learning is an important branch of AI technology. It mainly uses algorithms and statistical models to complete complex tasks through big data statistics and predict macroscopic physical properties. It has been applied to the study of natural gas hydrate formation and exploitation [107,108,109]. Therefore, in the future, machine learning and other means can be used in the study of multi-phase heat transfer and multi-phase seepage in the hydrate mining process. Through rapid analysis and various prediction models, more efficient and stable data prediction can be achieved, for example, to achieve ETC prediction of multi-parameter variations under the coupling of multi-phase heat transfer and seepage. The application of AI provides efficient and accurate basic data for heat transfer research in the natural gas hydrate mining process.

7. Conclusions

The heat transfer characteristics of natural gas hydrate reservoirs are one of the important control factors in the gas production stage of hydrate exploitation which directly affect the input efficiency of external heat, thus controlling the subsequent reaction process. In this paper, the research methods and research progress of natural gas hydrate heat transfer are reviewed from four aspects: the measurement method of heat transfer characteristics, the influencing factors of heat transfer in a hydrate system and hydrate-containing porous media system, the predictive models for ETC, and the heat transfer mechanism of hydrate.

(a)

The heat transfer measurement methods of natural gas hydrate mainly include heat flux measurement and ETC measurement. The heat flux measurement method realizes the visualization of the heat transfer process by measuring the temperature changes at different positions and qualitatively analyzes the heat transfer path of the system; the ETC measurement method obtains the ETC of the system by calculating the temperature response between the measuring points and quantitatively analyzes the heat transfer characteristics of the system.

(b)

The phase change in the hydrate in the process of hydrate decomposition causes the change in heat conduction and heat convection. The effective heat transfer of hydrate plays an important role in the control of the hydrate decomposition rate.

(c)

The natural gas hydrate reservoir is a multi-component complex sediment system. The thermal conductivity of these components is different, and the decomposition of the hydrate itself and the environmental heat transfer of the reservoir in which the hydrate is located have an impact on the natural gas hydrate and the reservoir. Therefore, the heat transfer characteristics of natural gas hydrate reservoirs are not only related to the material of porous media, pore structure, and particle size distribution, but also to the water content of the system, hydrate saturation and distribution, hydrate decomposition, and reservoir heat conduction.

(d)

The study of the heat transfer mechanism of natural gas hydrates found that the interaction between host molecules is dominant in the process of heat transfer. The vibration of the hydrate cage lattice will diffuse heat. The local vibration of guest molecules affects the vibration of host water molecules. The coupling vibration of guest molecules determines the transfer intensity of heat energy. Therefore, the enhancement in the interaction between host and guest molecules will help to improve the thermal conductivity of hydrates.

At present, due to the limitations of different measurement methods, the measurement accuracy and repeatability of natural gas hydrate heat transfer research are low. With the continuous improvement in scientific research equipment, microscopic characterization techniques such as infrared and Raman thermal measurements can be used to study the heat transfer of hydrates, using a combination of quantitative and qualitative methods for analysis. The application of AI technology will realize ETC prediction of multi-parameter changes under the coupling of multi-phase heat transfer and multi-phase seepage, and provide efficient and accurate basic data for heat transfer research in the natural gas hydrate exploitation process. At the same time, we can apply EMD and NEMD to the study of the heat transfer mechanism of the hydrate-bearing porous media system, and analyze the heat transfer mechanism in the process of hydrate phase change in sediments. The combination of theory and experimental research will help to analyze the heat transfer path and influencing factors in the process of natural gas hydrate exploitation, and provide basic data and theoretical support for the study of natural gas hydrate exploitation. Although the exploitation of natural gas hydrate is still in its infancy, the in-depth study of the heat transfer characteristics of natural gas hydrate has made important progress. We predict that in the near future, the application of natural gas hydrate will actively promote the transformation of the modern economic development mode to low-carbon sustainable development. Therefore, we hope that this review will be helpful in the application and development of natural gas hydrate in sustainable industries.