Clay Particle Migration and Associated Permeability Damage in Natural Gas Hydrate-Bearing Clayey-Silty Sediments: A Review

Abstract

Natural gas hydrate (NGH) is a highly promising alternative energy source for the future, which is widely distributed in marine clayey-silty sediments. Permeability is the key factor determining the efficiency of NGH exploitation. However, clay particles can migrate and clog the pores, leading to a decrease in reservoir permeability during the development of NGH. This review summarizes the permeability damage law during the NGH production from clayey-silty sediments, with a focus on the influence of clay particle migration. For the scientific problem of clay particle migration, the governing equation of clay particle migration was first clarified through force balance analysis. Then, the influencing factors and laws of clay particle migration were systematically summarized from two aspects: internal factors such as clay type, content, particle size, reservoir heterogeneity, and external conditions such as salinity, flow rate, temperature, pH, and stress field. The detachment, migration, aggregation and clogging characteristics of clay particles in porous media were observed and outlined based on microscopic visualization technology. Thirdly, the numerical simulation methods of particle migration were summarized, and the permeability damage laws and its influence mechanism were analyzed. Finally, the limitations on clay particle migration and permeability damage in the current research were discussed, and corresponding suggestions were given to promote the efficient development of NGH.

1. Introduction

Natural gas hydrate-bearing clayey-silty sediments (NGHBS) refer to fine-grained sediment deposits containing natural gas hydrates within their pore spaces [1,2]. This type of reservoir makes up over 90% of the global total [3]. With the gradual depletion of conventional energy resources such as petroleum and coal, natural gas hydrates have gained widespread recognition as a critical strategic alternative energy source for the future [4,5].

Permeability is a crucial parameter for gas hydrate extraction, playing a decisive role in the recovery efficiency of methane and other natural gases [6,7]. In theory, when natural gas hydrates dissociate from solid to fluid, they release pore space in methane hydrate-bearing clayey-silty sediments, during which permeability should gradually increase. However, an anomalous phenomenon of permeability damage has been increasingly observed in laboratory experiments [8,9,10]. Research suggests that this phenomenon may be attributed to the following: the dissociation of gas hydrates produces fresh water and methane, leading to a rapid decrease in reservoir salinity. Given the swelling properties and salinity sensitivity of clay minerals, hydrate dissociation readily triggers clay swelling and migration. These particles can clog the pore throats of methane hydrate-bearing clayey-silty sediments [11,12], significantly altering the reservoir pore structure [13,14], and ultimately resulting in permeability damage.

Scholars have extensively investigated permeability damage [15,16,17] through theoretical, experimental, and numerical approaches, examining aspects [18,19,20,21,22], such as: permeability evolution patterns during gas hydrate dissociation, clay particle migration behavior, mathematical models of particle migration, internal/external conditions of particle migration and permeability damage patterns, mechanisms of particle clogging in pore throats, and particle migration path tracing [23,24,25,26,27,28]. Therefore, this paper reviews the literature on clay particle migration in methane hydrate-bearing clayey-silty sediments, with a focus on its impact on permeability. The review is structured as follows: studies on particle migration and permeability damage; mathematical models; influencing factors; and numerical simulations. Its novelty lies in systematically summarizing the multi-factorial controls on migration patterns and damage extent, as well as the mechanisms triggered by hydrate dissociation. Based on the identified gaps, future research directions are proposed.

2. Clay Particle Migration and Permeability Damage in NGHBS

Due to the challenges associated with simulating in-situ hydrate extraction in natural reservoirs, most research has been conducted in laboratory settings [29,30]. The initial phase of these experiments involves the preparation of HBS.

Table 1. Preparation methods of HBS in experimental studies.

Experimental Materials	Hydrate Formation Methods and Conditions	Reference
Kaolinite mixed with ice powder, compacted at −10 °C	Excess gas method at 1 °C and 5.5 MPa	Liu et al. [31]
Montmorillonite mixed with ice powder, compacted at −8 °C	Excess gas method at 1 °C and 6 MPa	Wu et al. [32]
Core samples from China’s Qilian Mountain hydrate reservoir, saturated with 5% KCl solution	Excess gas method at 0.85 °C and 5 MPa	Wang et al. [33]
Quartz sand packing, deionized water saturation	Excess gas method	Shen et al. [34]
Kaolinite, montmorillonite, and illite mixed with ice powder, compacted at −10 °C	Excess gas method at 1 °C and 6 MPa	Wu et al. [35]
Standard sand and montmorillonite, refrigerated for 24 h	Excess gas method at 1 °C and 5 MPa	Zhao. [36]
Unconsolidated quartz sand, compacted and water-saturated	Excess gas method at 8.5 °C and 15 MPa	Li et al. [37]
A mixture of quartz sand and sodium bentonite with 3.5% NaCl solution at −1.95 °C, compacted	Excess gas method at −1.95 °C and 10 MPa	Sun et al. [38]
Quartz sand and silty clay mixture	Excess gas method at 8.5 °C and 20 MPa	Han et al. [39]
Kaolinite and montmorillonite	Excess gas method at 2.8 °C and 7.5 MPa	Wang et al. [40]
Quartz sand and illite mixture (wet compaction method)	Excess gas method at 2 °C and 8 MPa	Lei et al. [41]

summarizes the methods utilized in the laboratory preparation of HBS.

2.1. Permeability Damage During Hydrate Dissociation

In laboratory settings, hydrate dissociation is frequently induced by depressurization, a method whose operating principle is to depress the system pressure below the phase equilibrium pressure of the hydrates to force their decomposition. Wu et al. [32,35] observed that hydrate dissociation caused methane gas permeability damage in montmorillonite, illite, and kaolinite, with severity correlating to their swelling capacity (Figure 1). Maximum damage occurred at 15.15% hydrate saturation, declining beyond 15.5% as expanding clay aggregates created new flow pathways. The mechanism involves clay–water electrostatic interactions forming bound water, with particle swelling and hydration clogging pore throats.

Wang et al. [33] investigated naturally fractured hydrate-bearing cores from the Qilian Mountain permafrost. XRD analysis showed a composition dominated by quartz and kaolinite, classifying the reservoir as clayey-silty sediments. As shown in Figure 2, water permeability damage was observed after both hydrate formation and dissociation. Post-formation damage was attributed to pore clogging by hydrates. Experiments revealed that dissociation-induced permeability damage was primarily caused by fine particle mobilization due to salinity reduction from released freshwater, consistent with findings by Bedrikovetsky et al. [42].

Table 2. Summary of potential factors causing permeability damage during hydrate dissociation.

Factors	Reference
Clay hydration and swelling lead to pore-throat clogging	Qi et al. [43]
Reduced pore water salinity triggers particle migration and subsequent pore-throat clogging	Jang et al. [44] and Feng et al. [45]
Flow pathways and hydrate distribution patterns	Hou et al. [46]
Stress and confining pressure	Zhao et al. [47]; Zhao et al. [48] and Zhao et al. [49]

summarizes the primary causes of permeability damage during hydrate dissociation in HBS. Key mechanisms include clay particle migration and expansion, driven by salinity sensitivity and hydration–induced swelling. Other influencing factors, such as hydrate distribution, stress state, and confining pressure, have also been identified.

2.2. Behavior of Clay Particles During Hydrate Dissociation

The freshwater generated by dissociation can be trapped as bound water, which, given the salinity sensitivity and swelling nature of clays, often triggers clay migration and swelling. These processes alter sediment pore structure, reducing porosity and causing permeability damage, with the influencing factors summarized in

Table 3. Factors potentially influencing clay particle migration and swelling behaviors during hydrate dissociation.

Factors	Reference
Clay particles aggregate due to flocculation effects, while pore water salinity and vertical pressure gradients.	Sun et al. [38]
Initial hydrate saturation, depressurization rate and magnitude, and whether the generated gas can continuously exist	Li et al. [50]
The generated gas exists in the form of bubbles, which continuously accumulate and expand, while clay particles aggregate at the water-gas interface.	Jung et al. [51]
Water content (saturation) and initial hydrate saturation	Guan et al. [52]

.

Sun et al. [38] simulated depressurization-induced hydrate dissociation under confined and unconfined conditions. In confined settings, minimal water/sand production and negligible permeability damage occurred, as dissociation-generated water was constrained by capillary forces or clay absorption. Pre-existing clay aggregates from hydrate formation limited further permeability damage from particle swelling and migration. In unconfined settings with seawater intrusion, vertical pressure gradients and salinity changes intensified clay swelling and migration, causing upper particle compaction, continuous pressure buildup, and significant permeability damage. As shown in Figure 3, depressurization yielded distinct responses. Confined environments exhibited immobilized pore water, minimal particle activity, and unimpeded gas flow, while unconfined conditions featured pronounced downward clay migration, compaction, and swelling.

Furthermore, Feng et al. [45] investigated the influence of clay particles (e.g., montmorillonite and illite) on methane hydrate-bearing clayey-silty sediments. As shown in Figure 4, hydrate dissociation triggered clay migration and swelling, which induced particle bridging and pore-throat clogging. Based on cavity formation in Ottawa sand (Figure 5), Jung et al. [51] proposed that gas bubbles generated during hydrate dissociation drive particle migration. Clay accumulates at gas–water interfaces, exacerbating pore-throat clogging and ultimately forming cavities.

Ultimately, permeability damage is primarily caused by clay particle migration and swelling, which are instigated by dissociation-induced salinity changes and two-phase flow. To elucidate this key mechanism, the following reviews progress in modeling, influencing factors, and simulation of particle migration.

3. Mathematical Models of Particle Migration

Theoretical models aim to describe particle detachment, migration, and clogging. Macroscopic approaches, based on mass conservation (Bedrikovetsky et al. [42]), fail to capture behavior under abrupt salinity or flow changes, or reveal underlying mechanisms (Miri et al. [53]). Microscopic models analyze force equilibrium between migration and adsorption torques but require extensive experimental calibration (Bedrikovetsky et al. [42]).

3.1. Macroscopic Perspective

Particle transport is typically modeled using a mass conservation equation coupled with a kinetic equation for particle clogging. A widely adopted formulation, which distinguishes between suspended and retained particles, integrates processes of detachment, migration, and clogging as follows:

$$\begin{array}{c}{\displaystyle \frac{\partial }{\partial t}}\left(\varphi c+\sigma \right) + U{\displaystyle \frac{\partial c}{\partial x}} ={ D}_0{\displaystyle \frac{{\partial }^{2}c}{\partial x^2}}\end{array}$$

(1)

$$\begin{array}{c}{\displaystyle \frac{\partial \sigma }{\partial t}}=\lambda \left(\sigma \right)cU-k_{\det }\sigma \end{array}$$

(2)

where c and σ are the suspended and retained particle concentrations, respectively, U is the Darcy flow velocity, D₀ is the initial erosion front velocity, λ is the filtration coefficient, and k_det is the detachment coefficient [42,54,55].

Another approach classifies retained particles into two types: surface-deposited and pore-throat clogging, leading to the following transport equations:

$$\begin{array}{c}{\displaystyle \frac{\partial }{\partial t}}\left(\varphi c + {\sigma }_{\mathrm{a}} + {\sigma }_{\mathrm{s}}\right) + U{\displaystyle \frac{\partial c}{\partial x}} = 0\end{array}$$

(3)

$$\begin{array}{c}{\displaystyle \frac{\partial {\mathsf{\sigma }}_{\mathrm{a}}}{\partial t}}={ \lambda }_{\mathrm{a}}cU -{ k}_{\det }{\sigma }_{\mathrm{a}}\end{array}$$

(4)

where σ_s and σ_a are the particle concentrations deposited on pore surfaces and clogging pore throats, respectively, λ_s and λ_a are the filtration coefficients for deposition on pore surfaces and clogging at pore throats, respectively [56,57], the other parameters are the same as above.

The simulation results of particle migration based on the microscopic theoretical approach are presented in Figure 6. While Runkel et al. [58] performs well under steady flow, but it shows significant errors in transient conditions, an issue partially addressed by Saaltink et al. [59] via concentration-dependent flow terms. Furthermore, such equations lack mechanistic insight [42] necessitating integration with microscopic theories (e.g., maximum retained concentration) for fundamental improvement.

3.2. Microscopic Perspective

Firstly, force analysis is conducted on the particles. According to Saffman’s lift theory [61] and colloidal stability (DLVO theory), the forces acting on the particles are determined as follows: gravitational force F_g, electrostatic attraction F_q from matrix particles, lift force F_l, and drag force F_d from fluid flow [42,62,63]. The schematic diagram of force analysis is shown in Figure 7. In Figure 7a, the light blue circle at the top represents the study object, while the dashed gray circle at the bottom represents a virtual matrix particle. Figure 7b illustrates the behaviors of particles in porous reservoir media under environmental changes, including detachment and migration, bridging at pore throats, and clogging. In terms of force effects, both gravitational force and electrostatic attraction from the matrix exhibit adsorption effects on particles, while fluid forces promote particle migration.

As shown in Figure 8, considering the moment balance of particles at the intersection point O of the retention layer, they remain in a static state, i.e.:

$$\begin{array}{c}{F}_{\mathrm{d}}{l}_{\mathrm{d}} + F_{\mathrm{l}}{l}_{\mathrm{n}} = \left(F_{\mathrm{q}} +{ F}_{\mathrm{g}}\right)l_{\mathrm{n}}\end{array}$$

(5)

where ln is the horizontal lever arm, and l_d is the vertical lever arm.

In practice, based on the above force analysis, the determination method for particle motion states shown in Figure 8 is obtained. Particles remain adsorbed if the retention torque dominates; otherwise, they migrate. Migrating particles re-adsorb or clog at throats upon reaching torque equilibrium or are produced if the migration torque prevails.

The expression for the gravitational force on the particle is:

$$\begin{array}{c}{F}_{\mathrm{g}}={\displaystyle \frac{4}{3}}\pi r_{\mathrm{s}}^3∆\rho g\end{array}$$

(6)

where r_s is the particle radius, Δρ is the density difference between the fluid and particle, g is the gravitational acceleration, taken as 9.8 m/s².

The electrostatic attraction is determined based on the colloidal stability (DLVO) theory and the diffuse double-layer theory. Numerically, the electrostatic attraction equals the negative partial derivative of the three short–range interaction potential energies V between the matrix particles and migrating particles with respect to their separation distance h, i.e.:

$$\begin{array}{c}\left\{\begin{array}{c}{F}_{\mathrm{q}} = -{\displaystyle \frac{\partial V}{\partial h}}\\ V ={ V}_{\mathrm{LVA}} +{ V}_{\mathrm{DLR}} +{ V}_{\mathrm{BR}}\end{array}\right.\end{array}$$

(7)

where V_LVA, V_DLR, and V_BR represent the van der Waals potential energy, double-layer repulsive potential energy, and born repulsive potential energy.

Van der Waals forces [64], which vary with interparticle distance, comprise induction, orientation, and dispersion forces. Experimental data [62] demonstrate that dispersion forces dominate in most colloidal systems and are primarily responsible for colloidal stability. Consequently, dispersion forces are commonly used to quantify van der Waals interactions, with their potential energy expressed as:

$$\begin{array}{c}\left\{\begin{array}{c}{V}_{\mathrm{LVA} }= -{\displaystyle \frac{A_{132}}{6}}\left[{\displaystyle \frac{2\left(1+\mathrm{Z}\right)}{\mathrm{Z}\left(2+\mathrm{Z}\right)}}+\ln \left({\displaystyle \frac{\mathrm{Z}}{2+\mathrm{Z}}}\right)\right]\\ \mathrm{Z} = {\displaystyle \frac{h}{r_{\mathrm{s}}}}\end{array}\right.\end{array}$$

(8)

where A₁₃₂ denotes the Hamaker constant. Reference values of the Hamaker constant for different medium systems can be found in [65].

According to DLVO and diffuse double-layer theory, charged fluids generate an electric double layer at their surfaces, consisting of a charged layer and a diffuse layer. Simultaneously, the charged fluid induces electrostatic repulsion (i.e., double-layer repulsion) between particles and the attached rock matrix. The characteristics of the double layer can be described by the Poisson–Boltzmann theory [66].

The expression for the double-layer repulsive potential energy [42] is as follows:

$$\begin{array}{c}\left\{\begin{array}{c}{V}_{\mathrm{DLR} }= {\displaystyle \frac{{\epsilon }_0{D}_{\mathrm{e}}{r}_{\mathrm{s}}}{4}}\left[2{\phi }_1{\phi }_2\ln \left({\displaystyle \frac{1 + {\mathrm{e}}^{-k_{\mathrm{D}}h}}{1 -{ e}^{-{\mathrm{k}}_{\mathrm{D}}\mathrm{h}}}}\right) - \left({\phi }_1^2 +{ \phi }_2^2\right)\ln \left(1 {- e}^{-2k_{\mathrm{D}}h}\right)\right]\\ k_{\mathrm{D} }= \left\{\begin{array}{c}\sqrt{{\displaystyle \frac{e^{2}c}{{\epsilon }_{0}D{k}_{\mathrm{B}}T}}} \left(\mathrm{i}\right)\\ \sqrt{{\displaystyle \frac{e^2\sum _{i=0}^n{v}_i{z}_i^2}{{\epsilon }_{0}D{k}_{\mathrm{B}}T}}}\left(\mathrm{i}\mathrm{i}\right)\end{array}\right.\\ \end{array}\right.\end{array}$$

(9)

where ε₀ is the permittivity of free space, taken as 8.854×10⁻¹²C⁻²J⁻¹ m⁻¹, D_e is the relative permittivity, taken as 78, φ₁ and φ₂ are the zeta potentials of the particle and rock matrix, respectively, k_D is the Debye length, whose reciprocal characterizes the charge effectiveness of charged particles (i.e., double-layer thickness), for monovalent salt solutions (e.g., NaCl, KCl), use Equation (1); otherwise, apply Equation (2), c is the salinity, k_B is the Boltzmann constant, e is the electron charge, taken as 1.6 × 10⁻¹⁹ C, v_i is the ion concentration, z_i is the ion valence, other parameters are the same as above.

Born repulsion [65] is a short-range repulsive force that can be derived from the Lennard–Jones potential σ_LJ:

$$\begin{array}{c}{V}_{\mathrm{BR} }= {\displaystyle \frac{A_{132}}{7560}}{\left({\displaystyle \frac{{\sigma }_{\mathrm{LJ}}}{r_{\mathrm{s}}}}\right)}^6\left[{\displaystyle \frac{\left(8+\mathrm{Z}\right)}{{\left(2+\mathrm{Z}\right)}^7}} + {\displaystyle \frac{6-\mathrm{Z}}{{\mathrm{Z}}^7}}\right]\end{array}$$

(10)

Combining Equations (1)–(7), the electrostatic attraction F_q can be expressed as a function of salinity.

The expression for the hydrodynamic drag force, based on the c [53], is given as follows:

$$\begin{array}{c}{F}_{\mathrm{d}} = {\displaystyle \frac{\omega \pi \mu r_{\mathrm{s}}^{2}u}{H-2h_{\mathrm{c}}}}\end{array}$$

(11)

where ω is the correction factor for adsorption force, with a value range of 10~60, μ is the fluid viscosity, r_s is the radius of clay particles, u is the fluid velocity, H is the pore throat height, h_c is the clay layer thickness.

The expression for the hydrodynamic lift force is derived based on Saffman’s lift theory:

$$\begin{array}{c}{F}_1 = \chi r_{\mathrm{s}}^3\sqrt{{\displaystyle \frac{\rho \mu u^3}{{\left(H-2h_{\mathrm{c}}\right)}^3}}}\end{array}$$

(12)

where χ is the fluid lift force constant, taken as 1190, ρ is the density of clay particles.

Substituting Equations (6), (11), and (12) into Equation (5) and approximating yields:

$$\begin{array}{c}{\displaystyle \frac{F_{\mathrm{d}}}{F_{\mathrm{q}}+F_{\mathrm{g}}-F_{\mathrm{l}}}} = {\displaystyle \frac{l_{\mathrm{n}}}{l_{\mathrm{d}}}}\approx {\displaystyle \frac{1}{\sqrt{3}}}\end{array}$$

(13)

$$\begin{array}{c}{F}_{\mathrm{q}}\left(c\right)+{\displaystyle \frac{4}{3}}\pi r_{\mathrm{s}}^3∆\rho g-\chi r_{\mathrm{s}}^3\sqrt{{\displaystyle \frac{\rho \mu u^3}{{\left(H-2h_{\mathrm{c}}\right)}^{3 }}}}={\displaystyle \frac{\sqrt{3}\mathsf{\omega }\mathsf{\pi }\mathsf{\mu }{\mathrm{r}}_{\mathrm{s}}^2\mathrm{u}}{\mathrm{H}-2{\mathrm{h}}_{\mathrm{c}}}}\end{array}$$

(14)

then a quantitative relationship is established between the maximum retained particle concentration and salinity, seepage velocity, viscosity, and particle radius. The traditional expression for maximum retained particle concentration is:

$$\begin{array}{c}{\sigma }_{\mathrm{cr}} = \left[1-{\left(1-{\displaystyle \frac{2h_{\mathrm{c}}}{H}}\right)}^2\right]\left(1-{\varphi }_{\mathrm{c}}\right)\varphi \end{array}$$

(15)

where σ_cr is the maximum retained concentration, h_c is the clay layer thickness, H is the pore throat height, ϕ_c is the clay layer porosity, and ϕ is the rock matrix porosity.

Bedrikovetsky et al. [42] proposed an approach to process the moment balance equation (Equation (14)) by quantifying the clay layer thickness, thereby establishing a quantitative relationship between the parameters and σ_cr:

$$\begin{array}{c}U = \varphi u\end{array}$$

(16)

$$\begin{array}{c}1+{\displaystyle \frac{4}{3F_{\mathrm{q}}\left(\mathrm{c}\right)}}\pi r_{\mathrm{s}}^3∆\rho g-\chi {\displaystyle \frac{\sqrt{\rho F_{\mathrm{q}}\left(\mathrm{c}\right)}}{\mu }}\sqrt{{\displaystyle \frac{{\mu }^3{r_{\mathrm{s}}^{6}U}^3}{{{\varphi }^3\left(H-2h_{\mathrm{c}}\right)}^3{\left[F_{\mathrm{q}}\left(c\right)\right]}^3}}}=\sqrt{3}\omega \pi {\displaystyle \frac{\mu r_{\mathrm{s}}^{2}U}{\varphi \left(H-2h_{\mathrm{c}}\right)F_{\mathrm{q}}\left(c\right)}}\end{array}$$

(17)

where U is the Darcy seepage velocity.

Performing the following equivalent substitutions:

$$\begin{array}{c}x =\sqrt{{\displaystyle \frac{\mu r_{\mathrm{s}}^{2}U}{\varphi H\left(1-\frac{2h_{\mathrm{c}}}{H}\right)F_{\mathrm{q}}\left(c\right)}}}\end{array}$$

(18)

Then, Equations (15) and (17) are transformed into Equations (19) and (20):

$$\begin{array}{c}1+{\displaystyle \frac{4}{3F_{\mathrm{q}}\left(\mathrm{c}\right)}}\pi r_{\mathrm{s}}^3∆\rho g-{\displaystyle \frac{\chi \sqrt{\rho F_{\mathrm{q}}\left(c\right)}}{\mu }}{\mathrm{x}}^3 = \sqrt{3}\omega \pi x^2\end{array}$$

(19)

$$\begin{array}{c}{\sigma }_{\mathrm{cr}}\left(c,U,\mu ,r_{\mathrm{s}}\right)=\left[1-{\displaystyle \frac{{\mu }^2{U}^2{r}_{\mathrm{s}}^4}{{\varphi }^2{H}^2{\left[F_{\mathrm{q}}\left(c\right)\right]}^2{x}^4}}\right]\left(1-{\varphi }_{\mathrm{c}}\right)\varphi \end{array}$$

(20)

Through Equation (20) above, a quantitative relationship is established between the maximum retained concentration and salinity, seepage velocity, viscosity, and particle radius. This allows revealing the particle migration mechanism based on the maximum retained concentration.

Figure 9a illustrates particle detachment and migration during gas-water flow. Calculations in Figure 9b, c for salinity increasing from 0.1 to 105 mg/L show that elevated salinity enhances electrostatic attraction but reduces particle migration, as retention torque increases while migration torque remains constant, thereby inhibiting particle mobilization.

4. Study on Influencing Factors of Particle Migration

Table 4. Factors influencing particle migration behavior and resulting in permeability damage.

Categories	Factors	Reference
Internal Factors	Particle content	(Russell et al. [68] and Mohan et al. [69])
	Particle type	(Yang et al. [63] and Lei et al. [70])
	Particle swelling	(Anderson et al. [71]; Low rt al. [72]; Hensen et al. [73]; Sameni et al. [74] and Karpiński et al. [75])
	Particle size	(Yuan et al. [76])
	Reservoir heterogeneity	(Yang et al. [77])
External Factors	Seepage velocity	(Bedrikovetsky et al. [78] and Marquez et al. [79])
	Salinity	(Mahani et al. [80]; You et al. [81])
	pH	(Vaidya. [82] and Tang et al. [83])
	Fluid type	(Bai et al. [84] and Song et al. [85])
	Temperature	(You et al. [86])
	Stress field	(Ma et al. [87,88])

summarizes potential influencing factors on reservoir particle migration behavior and associated permeability damage reported in the literature.

4.1. Internal Factors of Particle Migration

The structure and properties of the reservoir itself influence the particle migration behavior within it. In most cases, the particles that migrate and cause permeability damage in reservoirs are clay particles [89]. Therefore, reservoir properties such as clay content, clay type, and particle size are related to the extent of particle migration. Generally, reservoirs with high clay content have a greater potential risk of internal particle migration, but this is not absolute, as clay type plays a more significant role. For instance, Yang et al. [63] found that under the same conditions, reservoirs with high clay content may exhibit less particle migration than those with low clay content.

For reservoirs with the same clay type, Russell et al. [68] identified a threshold for kaolinite content. Once this threshold is exceeded, further increases in kaolinite content do not significantly elevate the overall degree of permeability damage. Lei et al. [70] observed that compared to illite, montmorillonite exhibits stronger swelling upon contact with pore water, leading to more pronounced pore clogging after migration. Song et al. [90] found that kaolinite particles tend to aggregate and clog pore throats along the main flow direction, while expandable montmorillonite particles have a higher probability of clogging pore throats. As shown in Figure 10, clay particle swelling exacerbates the degree of clogging at pore throats. Yuan et al. [76] noted that larger particles can directly clog small-sized pore throats, whereas smaller particles may form bridging clogging in larger pore throats.

Additionally, Yang et al. [77] highlighted that when reservoirs exhibit nanoscale heterogeneity, surface charge distribution also becomes heterogeneous. In such cases, the presence of energy barriers complicates the interactions between surfaces and particles, rendering DLVO theory inapplicable and thereby affecting the determination of electrostatic attraction.

4.2. External Conditions for Particle Migration

External environmental factors critically influence particle migration within reservoirs, including fluid seepage velocity, salinity, pH, fluid type (viscosity), temperature, and stress field.

4.2.1. Fluid Seepage Velocity

The maximum retention concentration theory (Equations (11) and (12)) indicates that an increase in seepage velocity mobilizes particles by enhancing lift and drag forces beyond constant adsorption forces, a phenomenon supported by experiments. Bedrikovetsky et al. [78] showed that a shift from low to high rate prevents permeability damage by mobilizing particles, while the reverse shift induces damage. This observation confirms that higher flow velocities enhance particle mobilization. Marquez et al. [79] reported that although high- and low-rate injections initially caused similar permeability damage (~50%), their effects diverged markedly beyond 30,000 pore volumes, with damage worsening only at high rates, demonstrating that elevated flow velocities intensify particle migration.

4.2.2. Salinity

Based on Equations (7)–(10) and (20), salinity reduction expands the electric double layer, which weakens electrostatic attraction and ultimately reverses the retention layer by increasing the detachment torque over adsorption torque for surface particles. This phenomenon manifests as an increase in particle detachment and migration. Yang et al. [63] observed through core flooding and SEM-EDX-XRD analysis that while injections of 0.4% KCl caused negligible permeability reduction across all core types, significant impairment occurred only with deionized water, indicating that substantial salinity reduction is required to promote particle migration in certain reservoirs. For Birkhead sandstone, Mahani et al. [80] found that stepwise salinity reduction gradually decreases in permeability, which plunges below 0.05 mD with DIW injection, confirming salinity-dependent mobilization. You et al. [81] experimentally studied permeability damage caused by particle migration in geothermal wells. They highlighted the acute sensitivity of clay minerals (kaolinite, illite, chlorite) to salinity changes, where clay detachment and migration occurred instantaneously upon salinity reduction.

4.2.3. pH

The pH value critically governs the zeta potential and surface potential of reservoir particles, both of which directly influence electrostatic attraction forces. Vaidya et al. [82] reported that in Berea sandstone, pH 11 triggers permeability loss by increasing zeta potential and EDL repulsion to detach particles, though this mechanism is mineral-dependent. For instance, Tang et al. [83] demonstrated that CO₂-induced acidic conditions trigger particle migration through rock dissolution, a mechanism contrasting with alkaline-driven processes, with damage severity dependent on core permeability and saturation.

4.2.4. Fluid Type

The complexity of particle migration in multiphase flow can be partially explained by fluid viscosity effects. As described by Equations (11) and (12), viscosity differences under constant flow velocity lead to variations in the migration torque exerted on particles, resulting in differing degrees of mobilization. Bai et al. [84] found that coal fines production is proportional to the cube of fluid velocity once flow channels form, with negligible migration during single-phase gas flow but significant permeability damage during gas–water flow due to water’s superior transport capacity. The study further noted that more complex pore structures and poorer connectivity exacerbate this migration. Zhou et al. [91] found that low-viscosity fluids facilitate particle migration during water inrush by lowering the critical hydraulic gradient. Although higher viscosity increases the driving force, it also induces greater fluid resistance that reduces flow velocity, ultimately suppressing particle migration.

4.2.5. Temperature

Temperature affects particle migration by influencing the electric double layer (EDL) thickness. As shown in Equations (7)–(10), increasing temperature leads to greater EDL thickness and enhanced EDL repulsion, resulting in reduced electrostatic attraction. You et al. [86] found that elevated temperature intensifies particle migration in geothermal reservoirs despite reduced water viscosity. This occurs because the electrostatic attraction and drag force, which are 2–3 orders of magnitude stronger than gravity and lift, are significantly reduced by elevated temperature, thereby overwhelmingly controlling the migration behavior.

4.2.6. Stress Field Mutation

Ma et al. [87,88] observed that abrupt stress field changes during coal mining destabilized reservoir stresses, activating faults and generating massive particles. These particles migrated with the water flow, triggering water–sand inrush geohazards. Additionally, Kozhevnikov et al. [92] noted that the pore pressure drop during hydrate dissociation leads to an increase in effective stress, which can readily cause reservoir deformation and instability. This, in turn, induces significant damage to reservoir permeability.

In summary, particle migration, activated by external factors, leads to permeability damage through pore-throat blockage and particle deactivation, with the specific impairment pattern being case-dependent.

4.3. Visualization Studies on Particle Migration-Clogging

Focusing on hydrate dissociation in fine-grained sediments, this subsection employs microfluidic models to directly visualize particle dynamics, offering key mechanistic insights. These laser-etched microchannels, which replicate reservoir pore networks, represent a well-established tool for research on particle migration and see page [93,94,95,96,97].

The key factors controlling pore-throat clogging in micro-visualization models—particle concentration, geometric ratios (D/ds, O/ds), and pore fluid parameters—are illustrated by the physical photographs and partial plans of common models shown in Figure 11a,b. Existing results [98,99,100,101] demonstrate that particle clogging and migration are primarily controlled by the throat-to-particle size ratio (O/ds), with a secondary effect from the pore-to-particle ratio (D/ds). The studies delineate specific threshold ratios, clogging at O/ds < 1.67, bridging at O/ds = 1.67~10, pore-filling at O/ds = 10~100, and free migration at O/ds > 100 (with corresponding D/ds thresholds shown in Figure 11c–e). This hierarchy confirms the dominant role of O/ds compared to D/ds. Importantly, Cao et al. [102] demonstrated that clogging behavior is material-specific and chemically controlled. While quartz and bentonite clogging are inhibited in DIW, kaolin is promoted, and mica remains unaffected, showing that chemical properties can override geometric effects. As shown in Figure 12, the distribution of particle suspensions (ds = 19.2 μm) in microfluidic chips with varying throat diameters reveals an increase in clogging risk with decreasing throat size.

Additionally, Jung et al. [103,104] identified key clogging mechanisms where particles accumulate at water–gas interfaces to form bridges, and higher flow rates enhance clay detachment yet can destabilize bridges through pressure fluctuations, creating a complex dependence of clogging severity on both fluid type and flow dynamics. Bate et al. [105] identified O/ds as a critical clogging determinant, noting non-clogging triangular deposits forming upstream of cylindrical obstacles that interact with flow fields; enhanced bentonite deposition with increasing salinity. Jarrar et al. [106] discovered that even at O/ds = 12~30, high particle concentrations could still cause throat clogging, emphasizing clay content as another crucial influencing factor.

5. Numerical Simulation Studies of Particle Migration

Numerical simulation methods comprehensively solve particle tracing, fluid flow fields, velocity fields, pressure fields, etc. [107,108,109,110,111], providing insights into the impact of particle migration on both fluid and solid phases. These methods are widely applied in particle migration research [112,113,114]. Compared to experimental approaches [115,116,117,118,119], numerical simulations significantly reduce economic costs and are not constrained by experimental conditions [120,121,122].

Table 5. Summary of numerical methods for simulating particle migration.

Methods	Main Equations	Core Concept
CFD–DEM	$\begin{array}{c}\rho {\alpha }_{\mathrm{g}}\left({\displaystyle \frac{\partial \overrightarrow{v}}{\partial t}}+\overrightarrow{v}·\nabla \overrightarrow{v}\right)=-\nabla {\alpha }_{\mathrm{g}}p+\mu {{\alpha }_{\mathrm{g}}\nabla }^2\overrightarrow{v}+{\alpha }_{\mathrm{g}}\rho g-\overrightarrow{S}\end{array}$$\begin{array}{c}{m}_{\mathrm{s}}{\displaystyle \frac{\mathrm{d}\overrightarrow{v_{\mathrm{s}}}}{\mathrm{d}t}}={\overrightarrow{F}}_{\mathrm{g}}+\sum {\overrightarrow{F}}_{\mathrm{D}}+\sum {\overrightarrow{F}}_{\mathrm{n}}+\sum {\overrightarrow{F}}_{\mathrm{c}}\end{array}$	Fluid phase treated as continuum, solid phase as discrete
LBM–DEM	$\begin{array}{c}{f}_{\mathsf{\alpha }}\left(\overrightarrow{\mathrm{x}}+{\overrightarrow{\mathrm{e}}}_{\mathsf{\alpha }}{\delta }_{\mathrm{t}}, t+{\delta }_{\mathrm{t}}\right)-f_{\mathsf{\alpha }}\left(\overrightarrow{\mathrm{x}}, t\right)=\\ -{\displaystyle \frac{1}{\tau }}\left[f_{\mathsf{\alpha }}\left(\overrightarrow{\mathrm{x}}, t\right)-f_{\mathsf{\alpha }}^{\left(\mathrm{eq}\right)}\left(\overrightarrow{\mathrm{x}}, t\right)\right]+F_{\mathsf{\alpha }}\left(\overrightarrow{\mathrm{x}}, t\right){\delta }_{\mathrm{t}}\\ \end{array}$
TFM	$\begin{array}{c}{\displaystyle \frac{\partial \left({\epsilon }_{\mathrm{g}}{\rho }_{\mathrm{g}}\right)}{\partial \mathrm{t}}}+\nabla ·\left({\epsilon }_{\mathrm{g}}{\rho }_{\mathrm{g}}{\overrightarrow{u}}_{\mathrm{g}}\right)=0\end{array}$$\begin{array}{c}{\displaystyle \frac{\partial \left({\epsilon }_{\mathrm{m}}{\rho }_{\mathrm{m}}\right)}{\partial t}}+\nabla ·\left({\epsilon }_{\mathrm{m}}{\rho }_{\mathrm{m}}{\overrightarrow{u}}_{\mathrm{m}}\right)=0\end{array}$	Both fluid and solid phases treated as continuum

Summary of numerical simulation methods for particle migration studies: core concepts, governing equations. The limitations of CFD–DEM and LBM–DEM method approaches include, the current model has not yet accurately captured fluid turbulence effects or accounted for irregular particle morphologies [123]. Additionally, the TFM method is unable to characterize particle packing, migration, and clogging behaviors [112]. The CFD–DEM approach provides detailed results on particle positions, temperature profiles, migration velocities, and size distributions within porous media.

Among these methods, the CFD–DEM approach has been widely adopted for particle migration studies. This method first models the fluid phase, with its mass conservation equation given by:

$$\begin{array}{c}{\displaystyle \frac{\partial {\alpha }_{\mathrm{g}}\rho }{\partial t}} + \nabla ·\left({\alpha }_{\mathrm{g}}\rho \overrightarrow{v}\right) = 0\end{array}$$

(21)

where α_g is the volume fraction of fluid in the grid cell, ρ is the fluid phase density, and v is the fluid velocity vector.

The momentum conservation equation for the fluid phase, i.e., the Navier–Stokes (N–S) equations, is given by:

$$\begin{array}{c}\rho {\alpha }_{\mathrm{g}}\left({\displaystyle \frac{\partial \overrightarrow{v}}{\partial t}}+\overrightarrow{v}·\nabla \overrightarrow{v}\right) = -\nabla {\alpha }_{\mathrm{g}}p+\mu {{\alpha }_{\mathrm{g}}\nabla }^2\overrightarrow{v}+{\alpha }_{\mathrm{g}}\rho g-\overrightarrow{S}\end{array}$$

(22)

where p, μ, g, and S represent pressure, fluid viscosity, gravitational acceleration, and the momentum form of fluid–solid interaction forces.

The calculation formula for the momentum form of fluid–solid interaction forces is given by:

$$\begin{array}{c}\overrightarrow{S}= {\displaystyle \frac{\sum _{i=1}^n{\overrightarrow{F}}_{\mathrm{D}}}{V}}\end{array}$$

(23)

where F_D is the resultant force exerted by the fluid.

For the solid phase (i.e., particles), the discrete element method (DEM) approach is adopted. The particle motion is governed by substrate-interaction forces, fluid forces, interparticle collision forces, non-contact forces, and gravity. According to Newton’s second law, the momentum conservation and rotational equations for particles are given by:

$$\begin{array}{c}{m}_{\mathrm{s}}{\displaystyle \frac{\mathrm{d}\overrightarrow{v_{\mathrm{s}}}}{\mathrm{d}t}} ={\overrightarrow{ F}}_{\mathrm{g}}+\sum {\overrightarrow{F}}_{\mathrm{D}}+\sum {\overrightarrow{F}}_{\mathrm{n}}+\sum {\overrightarrow{F}}_{\mathrm{c}}\end{array}$$

(24)

$$\begin{array}{c}{I}_{\mathrm{s}}{\displaystyle \frac{\mathrm{d}\overrightarrow{{\omega }_{\mathrm{s}}}}{\mathrm{d}t}}=\sum \overrightarrow{M}\end{array}$$

(25)

where m_s, v_s, F_n, F_c, I_s, ω_s, M represent particle mass, velocity, wall interaction forces (including friction), non-contact forces (e.g., electrostatic forces), moment of inertia, angular velocity, and resultant torque on particles.

The fluid–solid interaction coupling is then performed using methods such as the fictitious domain method and immersed boundary method. The fictitious domain method solves the fluid-solid interaction forces as follows:

$$\begin{array}{c}{\overrightarrow{F}}_{\mathrm{D}} = {\displaystyle \frac{\rho }{{\rho }_{\mathrm{s}}}}{m}_{\mathrm{s}}{\displaystyle \frac{\mathrm{d}\overrightarrow{v_{\mathrm{s}}}}{\mathrm{d}t}} - {\int }_{{\Omega }_{\mathrm{s}}}\overrightarrow{\lambda }\mathrm{d}\overrightarrow{x}\end{array}$$

(26)

where λ is the fictitious force (λ ≡ 0 in particle-free regions), other parameters are defined above.

Immersed boundary method (IBM):

$$\begin{array}{c}\sum {\overrightarrow{F}}_{\mathrm{D}}={\displaystyle \frac{1}{V}}{\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}}{\displaystyle \frac{\beta V_{\mathrm{si}}}{1-{\alpha }_{\mathrm{g}}}}({\overrightarrow{\mu }}_{\mathrm{g}}-{\overrightarrow{v}}_{\mathrm{si}})\delta \left(\overrightarrow{r}-{\overrightarrow{r}}_{\mathrm{i}}\right)\end{array}$$

(27)

where V, β, V_s, δ (r−r_i) represent total grid volume, coupling coefficient, solid volume occupied in grid, and the direct delta function.

The above methods primarily address laminar flow problems. For transitional flow (Re > 2000) and turbulent flow, further modifications are required. The Reynolds number is calculated as:

$$\begin{array}{c}{R}_{\mathrm{e}} = {\displaystyle \frac{\rho uL}{\mu }}\end{array}$$

(28)

In clayey-silty sediment, clay and other particles are mostly irregular in shape. However, the modeling methods for non-spherical particles—including their contact forms, collision models, and contact force models—are highly complex [124], making it difficult to accurately describe particle migration behaviors. To simplify research, these particles are typically approximated as regular spherical particles.

Through numerical modeling, scholars have revealed valuable, practically significant patterns. Bate et al. [105] established that larger micropillars generate larger triangular low-velocity zones, thereby controlling particle paths and promoting deposition through dominant hydrodynamic forces. Through CFD–DEM simulations, Fang et al. [125] found that matching LCM density to water is critical, allowing 75% of particles to bypass pores, seal fractures, and achieve controlled injection. Li et al. [126] demonstrated that applied pressure gradients can remobilize clogged particles to restore permeability, a key insight gained from their LBM–DEM–IBM simulations of particle migration and force interactions in porous media. Feng et al. [127] demonstrated through CFD simulations that particle radius and concentration non-linearly affect migration and permeability, with smaller radii causing instantaneous clogging at pore throats and altering flow paths. Using the LBM–DEM method, Fan et al. [128] demonstrated that while increasing effective stress reduces conductivity by compacting proppants, employing diverse proppant diameter combinations can create more complex pore spaces, thereby enhancing particle migration and mitigating clogging. By reconstructing pore structures via micro-CT and employing a base-point increment method, Su et al. [129] discovered that particles interact via fluid–mediated forces without direct contact, with larger particles exhibiting reduced flow capacity and greater dispersion. Sanematsu et al. [130] demonstrated that attractive rock surfaces promote the transport of larger, faster-moving nanoparticles, whereas repulsive surfaces enhance the production of smaller particles.

Based on a comprehensive review and analysis of theoretical, experimental, and simulation studies, we propose a mechanism for clay-particle-migration-induced permeability damage triggered by methane hydrate dissociation. As illustrated in Figure 13, the permeability damage from hydrate dissociation is attributed to a three-stage process: salinity reduction mobilizes particles, two-phase flow amplifies their migration, and large-scale movement causes formation damage.

6. Problems and Challenges

This review summarizes clay particle migration and the resulting permeability damage during NGH production from clay-silty sediments. Focusing on the theme of particle migration, we have systematically sorted out the following aspects: mathematical models, controlling factors, micro-visualization experiments, and numerical simulations. Further, clay particle migration characteristics and their impact on permeability were discussed. However, the following critical challenges remain unresolved:

(1)

Multiphase-multifield coupling. Methane hydrate-bearing clayey-silty sediments involve multiphase systems (free methane (gas), pore water (liquid), and mobile clay particles (solid)), along with coupled physical fields including gas-liquid-solid flow, endothermic dissociation, chemical progress, and stress changes. Clay particle migration is thus governed by complex interactions among multiple physical mechanisms, and influenced jointly by intrinsic reservoir properties and external fluid conditions. Future research should focus on elucidating the dynamics of particle behavior under these coupled environmental conditions.

(2)

The complexity of clay particle migration. The behavior of clay particles in porous media involves a series of continuous and complex processes, including detachment, migration, and clogging, which are interconnected and mutually influential. Clay particle migration occurs only after detachment from the pore surfaces, while the subsequent clogging induced during migration directly impacts the permeability of the porous medium. Therefore, future research should adopt a systematic approach to analyze clay particle migration.

(3)

Quantitative characterization of permeability damage. It has been demonstrated that the migration-clogging of clay particles is the main cause of permeability damage in clay-silty sediment. However, a clear correlation between the microscopic behavior of clay particles and the macroscopic permeability variation has not yet been established. The permeability damage caused by clay particle migration remains challenging. How to apply the behavior of a clay particle to the permeability damage calculation is the focus for future research.

7. Conclusions

This paper reviews the relevant literature from the following four perspectives: clay particle migration and permeability damage during NGH production from clayey-silty sediments; mathematical models and influencing factors of clay particle migration; clay particle migration-clogging characteristics; and the application of numerical simulation methods, such as CFD–DEM, in particle migration research. The following conclusions are drawn:

(1)

In NGH-bearing clayey-silty sediment, an abnormal phenomenon of "permeability decrease instead of increase" is observed during methane hydrate dissociation. Clay particle migration and the resulting pore clogging are recognized as the primary causes. The dissociation of NGH leads to a reduction of pore-water salinity, which triggers clay particle migration. Meanwhile, the two-phase flow of methane and water further promotes this process. As hydrate saturation increases, the permeability damage degree in clayey-silty sediments first increases and then decreases.

(2)

The behaviors of clay particles and other fine particles in sediment include detachment, migration, collision, deposition, and clogging, which are governed by hydrodynamic forces, porous matrix forces, interparticle contact forces, and gravitational forces. Among these, hydrodynamic forces generally facilitate particle migration, while the other forces tend to inhibit particle mobilization and migration. The specific behavior is determined by the torque exerted by these forces on the particle.

(3)

The migration and clogging behavior of clay particles are jointly influenced by both intrinsic reservoir properties and external fluid conditions. Among the intrinsic factors, the type and content of clay particles play a dominant role. With respect to external factors, pore-water salinity is particularly critical. Furthermore, the ratio of throat size to particle diameter plays a critical role in determining whether clogging occurs at the throat. When the ratio falls below 10, clay particle clogging becomes highly prone to occur.