Evaluation of the Gas Production Enhancement Effect of Boundary Sealing and Near-Wellbore Stimulation for Class 1 Hydrate Reservoir Step-Wise Depressurization with a Horizontal Well

Abstract

Natural gas hydrates (NGHs) currently lack economic feasibility; efficient exploitation methods must be continuously explored to increase their production capacity. Drawing on field data from China’s first offshore NGH trial production, a numerical simulation method was used to evaluate a comprehensive development strategy that combines a horizontal well with boundary sealing, near-wellbore stimulation, and step-wise depressurization to improve the recovery of Class 1 NGH reservoirs. The results indicated that boundary sealing has a strong enhancement effect: it inhibits water invasion and thus concentrates the energy for hydrate dissociation. The use of high-pressure water jets for near-wellbore stimulation generates highly permeable channels, greatly accelerating hydrate dissociation and gas flow; step-wise depressurization optimizes the production behavior by controlling water production. The combined application of these technologies significantly improves development performance, with cumulative gas production (V_g) increasing to 220.1% and the gas-to-water ratio (R_gw) increasing to 102.6% compared to the base case, providing an effective strategy for the development of Class 1 NGH reservoirs.

1. Introduction

Natural gas hydrates (NGHs) are cage-shaped crystalline compounds formed with water and gas molecules (mainly methane) under low-temperature, high-pressure environments, widely occurring in deep-sea sediments and terrestrial permafrost [1,2,3,4]. Their global resource potential is enormous, with different studies evaluating methane resources ranging from approximately 8.2 × 10¹² to 5 × 10¹⁵ m³, making them widely recognized as a high-potential clean energy source and expected to become an important supplement to fossil fuels [5]. According to the classification system proposed by Moridis et al., natural gas hydrate reservoirs are categorized into three classes (Class 1, 2, and 3) based on their thermodynamic and phase conditions [6]. Among these, Class 1 NGH reservoirs with typical structures are composed of overlying gas hydrate layers (GHBLs), middle three-phase coexisting layers (TPLs), and bottom free gas layers (FGLs), which are the key targets of current exploration and development. The development of NGHs involves changing the temperature and pressure conditions of the reservoir through physical or chemical methods, causing it to deviate from the equilibrium state of hydrates, thereby promoting its dissociation into natural gas and water. The main methods include thermal stimulation, inhibitor injection, CO₂ replacement, depressurization, and their combined applications. Among them, the thermal stimulation method generally faces problems of heat loss and low energy efficiency; the inhibitor injection method has limitations such as high cost, slow action, and potential environmental risks; the CO₂ replacement method is limited by slow reaction kinetics and difficult gas separation, resulting in low efficiency. In contrast, the depressurization method has become the dominant technology in current field testing and application due to its high energy efficiency and engineering feasibility [7,8,9,10]. China successfully carried out NGH trial production in the Shenhu Sea area in 2017 and 2020, using vertical and horizontal wells with the depressurization method, respectively [7,8]. Although the horizontal well has significantly increased the daily gas production to 2.87 × 10⁴ m³, it is still far below the commercial mining threshold of 5 × 10⁵ m³/day [9,10]. This gap reflects the limitations of a single depressurization method, and there is an urgent need to improve production capacity through the combination of multiple technologies [9,10].

The commonly used methods to increase production capacity currently include the following: using complex structured wells to expand the drainage area; applying thermal stimulation to improve the hydrate dissociation rate; and implementing reservoir stimulation, such as hydraulic fracturing, high-pressure water jetting, and boundary sealing, to change the local or overall permeability of the reservoir, etc. In various reservoir stimulation technologies, boundary sealing is achieved by constructing low-permeability artificial barriers, forming a vertical closed structure to optimize the pressure propagation within the reservoir and promote the efficient dissociation of hydrates [9,10]. In recent years, numerical simulation studies have gradually revealed the efficiency-enhancing mechanism and application potential of this technology [11,12,13,14,15,16,17,18,19,20]. For example, Lv et al. (2022) [18] evaluated the adaptability of artificial barriers for vertical well depressurization, and the results indicated a 20.88% reduction in water production alongside an improved gas-to-water ratio. Simulation studies by Nie et al. (2024) [19] revealed that the co-implementation of boundary sealing and hot water injection yields a synergistic outcome. Notably, adopting a five-spot well pattern shortens the production cycle by 680 days and raises the gas–water ratio by 17% relative to the three-spot pattern. Qin et al. (2025) [20] further elucidated the impact of sealing layer performance on mining efficiency and proposed optimized sealing parameters for vertical wells. These studies have promoted the development of boundary sealing from mechanism exploration to engineering parameter optimization, providing a theoretical reference for practical applications [11,12,13,14,15,16,17,18,19,20].

Although boundary sealing has shown promising application prospects, its potential for synergistic step-wise depressurization and near-wellbore stimulation remains insufficiently investigated, largely due to the significant modeling and operational complexities associated with concurrently integrating multiple enhancement technologies in numerical simulations. Moreover, reservoir heterogeneity presents a common challenge for gas hydrate production across various geological settings. Studies such as Bazaluk et al. [21] (2021) highlight similar difficulties in achieving uniform dissociation and stable recovery in heterogeneous deposits. The integrated strategy proposed in this work, which combines boundary sealing, near-wellbore stimulation, and step-wise depressurization, addresses these widespread challenges by engineering more favorable flow conditions, suggesting its broader applicability to non-uniform reservoirs. Accordingly, a 3D geological model of a horizontal well was constructed in this work to evaluate the enhancement effects of three key technologies for developing Class 1 hydrate reservoirs: boundary sealing (at both the top of the GHBL and the bottom of the FGL), high-pressure rotating water jets (constructing high-permeability channels near-wellbore), and step-wise depressurization (regulating gas and water production behavior by stepped depressurization). The findings can provide new insights for the efficient development of Class 1 NGH reservoirs.

2. Methodology

2.1. Geological Backround

The SHSC4 well is located in the northern South China Sea (Figure 1) [7,22]. The water depth is approximately 1266 m. The seabed temperature is around 3 °C and it has a geothermal gradient of 43.65 °C/km, which offers conducive geological and thermodynamic conditions for the formation and distribution of NGHs [22]. Integrated logging and drilling data identify the reservoir as a Class 1 NGH reservoir. Its vertical succession is subdivided into three distinct reservoir units with different physical properties: GHBL occurring between 201 and 236 mbsf (meters below seafloor) with a thickness of 35 m. This interval is characterized by high-saturation hydrates coexisting with pore water. TPL is found within the 236–251 mbsf depth range and measures 15 m thick; this zone represents a transitional facies where gas hydrate, high-saturation free gas, and pore water coexist. FGL is located from 251 to 278 mbsf with a thickness of 27 m; this unit is primarily composed of low-saturation free gas and pore water [7,22].

2.2. Simulation Code

TOUGH-HYDRATE V1.0 is a numerical simulator designed for coupled multi-physical modeling of NGH production processes [24]. It incorporates various dissociation methods, including depressurization, inhibitor injection, thermal stimulation, and their combinations, and includes both equilibrium and kinetic models to capture hydrate formation and dissociation behavior. This code includes a multi-component (κ: methane-m, water-w, salt-i, hydrate-h) and multiphase (β: hydrate-H, aqueous-A, gas-G, ice-I) framework, enabling comprehensive representation of phase transitions and compositional dynamics under different production scenarios [24]. It has been extensively validated against experimental and field tests [25,26]. This work employed its parallel version, pT+H v1.0, and an equilibrium model, under the assumptions of Darcy-flow behavior and negligible geomechanical responses. This simplification enables research to focus on the coupled fluid flow and heat transfer processes that dominate during simulated production. However, the potential geomechanical effects, such as wellbore stability and sand production issues caused by geological deformation, may be significant over longer time scales or in specific geological environments. Therefore, the results and insights of this work should be explained within this simplified framework [27,28]. The main governing equations are as follows [24]:

Mass Conservation Equation:

$${\displaystyle \frac{d}{\mathit{dt}}}{\int }_{V_n}{M}^{\mathsf{\kappa }}\mathit{dV}={\int }_{{\mathsf{\Gamma }}_n}{F}^{\mathsf{\kappa }}\cdot \mathrm{n}d\mathsf{\Gamma }+{\int }_{V_n}{q}^{\mathsf{\kappa }}\mathit{dV}$$

(1)

where M^κ is mass accumulation, F^κ denotes mass flux, and q^κ represents source/sink terms.

Energy Conservation Equation:

$${\displaystyle \frac{d}{\mathit{dt}}}{\int }_{V_n}{M}^{\mathit{\theta }}\mathit{dV}={\int }_{{\mathsf{\Gamma }}_n}{F}^{\mathit{\theta }}\cdot \mathrm{n}d\mathsf{\Gamma }+{\int }_{V_n}{q}^{\theta }\mathit{dV}$$

(2)

where θ is heat component, is heat flux, and M^θ, F^θ, and q^θ are the heat accumulation, flux, and source/sink ratio, respectively.

2.3. Model and Cases Design

The schematic diagram is shown in Figure 2a. The model domain was discretized using a structured grid system. The X–Y plane of the model consists of 6303 grids (Figure 2b), while the Z-direction is divided into 95 layers (Figure 2c). The entire model has a total of 598,785 grids, which can precisely capture the dynamic variations in physical fields (e.g., multiphase flow and heat transfer) during hydrate exploitation, ensuring the reliability of numerical simulation results.

Four simulation cases, summarized in

Table 1. Simulation cases.

Cases	Boundary Sealing	Near-Wellbore Stimulation	Depressurization Mode
Case 1	-	-	Direct depressurization
Case 2	Yes	-	Direct depressurization
Case 3	Yes	Yes	Direct depressurization
Case 4	Yes	Yes	Step-wise depressurization

, were designed following a progressive structure (baseline → single technology → dual technology → triple technology), which is sufficient to systematically reveal the individual contributions and synergistic effects of each technology. Case 1 serves as the base case, without boundary sealing and near-wellbore stimulation, and adopts direct depressurization for exploitation. Case 2 was established to examine the impact of sealing layers on gas recovery. Case 3 activates the near-wellbore stimulation based on Case 2. Case 4 maintains the settings of sealing layers and near-wellbore stimulation consistent with Case 3, and adjusts the depressurization mode to step-wise depressurization, focusing on studying the response of reservoir gas production under different depressurization modes.

2.4. Initial and Boundary Conditions

Since pore water is connected to seawater, pore pressure can be regarded as hydrostatic pressure, which can be calculated by the following formula [29,30,31]:

P_pw = P_atm + ρ_swg (H + Z) × 10⁻⁶

(3)

This formula includes the following key parameters: porewater pressure (P_pw, MPa), atmospheric pressure (P_atm, MPa), seawater density (ρ_sw, kg/m³), and gravitational acceleration (g, m/s²). In addition, H represents water depth (m), and Z represents depth of seabed sediment (m). Figure 3 [32] illustrates the initial conditions, which were established via a self-balancing simulation performed with the TOUGH+HYDRATE simulator.

In this model, a first-type boundary condition [33] was applied, and the production wellbore was defined as an internal boundary. Critical multiphase flow parameters followed previous research. The van Genuchten model was employed for capillary pressure characterization, with the following parameter settings: S_mxA = 1, S_irA = 0.30, λ = 0.45, and P₀ = 1 × 10⁴ Pa. Similarly, the Stone model was utilized for relative permeability calculation, with the following parameter settings: S_irA = 0.30, S_irG = 0.03, n_A = 3.5, and n_G = 2.5. These selected parameter values are consistent with and supported by findings from earlier research [34,35,36].

Table 2. Model’s physical properties.

Parameter Type	Parameters	Value and Unit
Formation Thickness	OB and UB	30 m
	GHBL	35 m
	TPL	15 m
	FGL	27 m
Permeability	OB and UB	2.0 mD
	GHBL	2.9 mD
	TPL	1.5 mD
	FGL	7.4 mD
Porosity	OB and UB	0.30
	GHBL	0.35
	TPL	0.33
	FGL	0.32
Saturation	GHBL and TPL hydrate saturation	Obtained from the logging curve
	FGL gas saturation (free gas)	Obtained from the logging curve
Sealing Layer Properties	Permeability of sealing layers	0.0001 mD
	Thickness of sealing layers	1 m
	Location of sealing layers	Top of GHBL and bottom of FGL
Wellbore	Wellbore radius	0.1 m
Depressurization Strategy	Direct depressurization	6.5 MPa
	Step-wise depressurization (10 days per step)	3.5 → 4.0 → 4.5 → 5.0 → 5.5 → 6.0 → 6.5 MPa
Simulation Zone Settings	Number, width, radius, spacing, permeability, porosity	25, 0.5 m, 5.5 m, 11.5 m, 5 D, 0.7
Multiphase flow	van Genuchten mode for capillary pressure characterization	$P_{cap}=-P_0{\left[{\left(S^{*}\right)}^{-1/\mathsf{\lambda }}-1\right]}^{1-\mathsf{\lambda }}$, $S^{*}={\displaystyle \frac{\left(S_A-S_{irA}\right)}{\left(S_{mxA}-S_{irA}\right)}}$
	The maximum aqueous saturation (abbreviated as SmxA)	1
	The capillary pressure exponent (abbreviated as λ)	0.45
	The capillary pressure reference value (abbreviated as P0)	104 Pa
	Stone mode for relative permeability calculation	KrA = [(SA − SirA)/(1 − SirA)]nA, KrG = [(SG − SirG)/(1 − SirA)]nG
	The aqueous relative permeability exponent (abbreviated as nA)	3.5
	The gas relative permeability exponent (abbreviated as nG)	2.5
	The gas irreducible saturation (abbreviated as SirG)	0.03
	The aqueous irreducible saturation (abbreviated as SirG)	0.30
Others	Gas composition: 100% CH4; geothermal gradient: 43.653 °C/km; grain density: 2600 kg/m3; salinity: 3.5%; grain specific heat: 1000 J·kg−1·K−1; thermal conductivity (dry and wet): 1.0 W·m−1·K−1 and 3.1 W·m−1·K−1

details the model parameters. The basic structure consists of overburden (OB)/underburden (UB) at 30 m each, GHBL at 35 m, TPL at 15 m, and FGL at 27 m. Their permeability values are 2.0, 2.9, 1.5, and 7.4 mD, respectively; the porosities are 0.30, 0.35, 0.33, and 0.32, respectively [6]. The model includes 1 m thick sealing layers located on the GHBL top and FGL bottom (permeability 0.0001 mD). Depressurization strategies include direct depressurization (6.5 MPa) and step-wise depressurization (3.5 → 4.0 → 4.5 → 5.0 → 5.5 → 6.0 → 6.5 MPa).

2.5. Model Validation

The validation process involved comparing the model with field data from the 2017 Shenhu nature gas hydrate trial production. The simulation configured a vertical well at the center, with a completion interval from −201 to −271 m below seafloor (mbsf) and a constant pressure difference of 3 MPa [37]. As shown in Figure 4, the simulated 60-day cumulative gas production is approximately 3.6 × 10⁵ m³, while the field measurement is about 3.1 × 10⁵ m³, yielding a relative error of +16.1%, which demonstrates the model’s capability to replicate field production behavior and supports its utility for subsequent studies.

3. Results and Discussion

3.1. Production Performance of Gas and Water

Under a constant production pressure difference of 6.5 MPa, the 360-day evolution of gas production rate (Q_g) and cumulative gas production (V_g) for all cases is shown in Figure 5a,b. The pressure difference drives the rapid dissociation of hydrates near the wellbore, and high-saturation free gas from TPL flows into the wellbore rapidly. Case 1 exhibits an initial peak in its Q_g curve, and as production proceeds, the bottom water gradually inflows into the wellbore, weakens the driving force, slows down hydrate dissociation, and thus leads to a gradual decrease in Q_g. Compared to Case 1, Case 2 added the use of boundary sealing, which effectively reduces the invasion of bottom water and greatly reduces water production. The energy of the pressure difference is mainly used to drive the dissociation of hydrates. Therefore, the overall trend of its Q_g curve is higher than that of Case 1, and the rate of decline in the later stage is slower than that of Case 1. Compared to Case 2, Case 3 added the use of near-wellbore stimulation. Its Q_g reached a peak value larger than that of Case 2 at the beginning, and exhibited a slow downward trend in the later period as a result of the weakened driving force, but overall maintained a much higher gas production rate than Case 2. Compared to Case 3, Case 4 added the use of step-wise depressurization. The entire trend of its Q_g curve matches that of Case 3. The difference is that step-wise depressurization divides direct depressurization into several small cycles, and a minor depressurization gradient is conducive to inhibiting water production. Although the gas production is reduced, it can improve the overall gas-to-water ratio. When production reached 360 days, the V_g of the four cases were 886.1, 1079.5, 2005.5, and 1949.9 × 10⁴ m³, respectively. Compared with Case 1, Cases 2, 3, and 4 increased by 121.8%, 226.3%, and 220.1%, respectively.

Although this work is based on simplified assumptions and numerical simulations, the average daily gas production of Case 4 is about 5.41 × 10⁴ m³/d, which exceeded the average daily gas production reported in most previous international field tests, such as the 2008 Mallik vertical well depressurization trial production in Canada (0.21 × 10⁴ m³/d), the 2017 Nankai Trough vertical well depressurization trial production in Japan (0.92 × 10⁴ m³/d), and the 2020 South China Sea horizontal well depressurization trial production in China (2.87 × 10⁴ m³/d) [8,38]. This result indicates that the simultaneous adoption of multiple stimulation methods (boundary sealing, near-wellbore stimulation, and step-wise depressurization) in a horizontal well can significantly improve the overall system’s productivity. Production is increased by applying boundary sealing to suppress bottom water invasion, thereby achieving a broader and more uniform pressure distribution throughout the reservoir; near-wellbore stimulation effectively forms the high-permeability channel, which makes the dissociation of hydrates more intense; step-wise depressurization inhibits water production and enhances the overall gas-to-water ratio.

The evolution of the water production rate (Q_w) and gas-to-water ratio (R_gw) over 360 days is shown in Figure 6a,b. The Q_w curve of Case 1 rises sharply at the initial stage as hydrates around the wellbore decompose rapidly. As production proceeds, bottom water gradually invades the wellbore, and the overall water production rate Q_w shows a slow upward trend. Similarly to Case 1, the Q_w curve of Case 2 increases rapidly in the initial stage, but due to the boundary sealing, its Q_w curve shows a slow downward trend after 60 days, as the driving force is weakened and the decelerated hydrates dissociate. Compared to Case 2, Case 3 added the use of near-wellbore stimulation, which not only increases natural gas production but also promotes water flow. Therefore, its Q_w rises sharply at the early stage, with the initial water production rate being much higher compared to Cases 1 and 2. Subsequently, as the driving force is weakened and the decelerated hydrates dissociate, its Q_w curve shows a slow downward trend, but its overall water production curve Q_w is much higher than that of Cases 1 and 2. Compared to Case 3, Case 4 added the use of step-wise depressurization. Within 60 days of step-wise depressurization, the dissociation of hydrates is relatively less severe, and the water production rate Q_w curve shows a step-wise increase. Subsequently, as the driving force weakens, the dissociation of hydrates slows down and gradually decreases.

After 360 days of production, the R_gw values for the four cases were measured at 368.1, 635.6, 373.7, and 377.7. The R_gw (ST m³ of CH₄/ST m³ of water) serves as a key indicator of production efficiency, where a higher value denotes better economic performance. Relative to Case 1, Cases 2, 3, and 4 showed increases of 172.6%, 101.5%, and 102.6%, respectively. R_gw is influenced by both V_g and V_w, and the application of near-wellbore stimulation results in a lower R_gw for Case 3, but within an acceptable range. The application of step-wise depressurization in Case 4, while slightly sacrificing gas production, resulted in better R_gw performance throughout the entire production cycle (especially within the first 120 days) compared to Case 3. The overall assessment indicates that Case 4 demonstrated superior performance in terms of boundary sealing, near-wellbore stimulation, and step-wise depressurization, demonstrating the best performance and achieving more robust and sustainable efficient production.

3.2. Spatial Variation in Reservoir Properties

3.2.1. Spatial Variation in Pore Pressure

Figure 7 tracks the changes in pore pressure over 360 days for the four cases. It can be observed that after 60 days, there is no obvious change in Case 1, and the pressure drop is strictly limited to the near-wellbore area. This is due to the inflow of bottom water, which consumes the driving energy of the pressure drop. The direct consequence is that the effective pressure drop used to drive hydrate dissociation rapidly decreases, corresponding to the rapid decrease in the gas production rate in Figure 5a. Cases 2, 3, and 4 adopted boundary sealing, effectively suppressing the inflow of bottom water and concentrating the depressurization energy on the dissociation of hydrates, rather than consuming it on the produced water. It can be observed that as production progresses, the wellbore pressure in the reservoir becomes wider and more uniform, which is the key to maintaining a long-term stable driving force, directly supporting its better gas production performance. Cases 3 and 4 added the use of near-wellbore stimulation, where a large low-pressure area around the wellbore was observed. This is because the near-wellbore stimulation formed high-permeability channels and significantly enlarged the drainage area of the near-wellbore reservoir, which directly enhances the fluid flow capacity towards the wellbore, thereby increasing the initial gas production rate. Case 4 also added the use of step-wise depressurization, and the pressure gradient around the wellbore in the first 60 days is significantly smaller than that in Case 3. This smoother initial depressurization helps to suppress the dominant flow of water in the reservoir, which is the intrinsic reason for improving the early gas–water ratio. After 60 days, the pressure propagation range and pressure drop gradient of the two cases tend to be consistent.

3.2.2. Spatial Variation in Reservoir Temperature

In Figure 8, it can be observed that under the influence of the Joule-Thomson effect, a low-temperature area was observed within the near-wellbore reservoir in Case 1. After 60 days, as production advanced, the temperature of the near-wellbore reservoir gradually recovered as the hydrate dissociation slowed down, and remained basically unchanged thereafter. Case 2 similarly exhibited a near-wellbore low-temperature area; in comparison to Case 1, the boundary sealing concentrated the pressure difference-driving energy for hydrate dissociation. After 60 days, as production progressed, the dissociation of hydrates promoted the expansion of the low-temperature area around the wellbore. Based on boundary sealing, Cases 3 and 4 also adopted near-wellbore stimulation, which brought more gas production through high-permeability channels, resulting in a stronger Joule-Thomson effect. A clear low-temperature area was generated around the wellbore, and as production progressed, the affected area around the wellbore also widened. This drastic initial temperature drop is direct evidence of higher gas production, but also suggests potential risks of secondary hydrate formation. In the later stage (240 to 360 days), it can be seen that the temperature in the stimulation zone at the bottom of the wellbore has increased, resulting from the pressure difference-driven upward migration of bottom warm water. During the initial 60 days of production, it can be observed that the range and gradient of the low-temperature area around the wellbore in Case 4 are smaller compared to Case 3. This is because Case 4 employed step-wise depressurization in the early production stage, which prevented a rapid decline in reservoir temperature and provided favorable conditions for subsequent production. This milder thermal disturbance is a positive consequence of the step-wise depressurization strategy. This phenomenon is consistent with the research findings of Li et al. (2021) [39].

3.2.3. Spatial Variation in Hydrate Saturation

As shown in Figure 9, for the 360-day evolution of hydrate saturation, in all cases, a small amount of secondary hydrates were formed near the root and toe of the wellbore. This is because the influx of high-saturation free gas caused a strong Joule-Thomson cooling effect. As production proceeds, the secondary hydrates in the reservoir near the wellbore root and toe gradually dissociate. In addition, it can be observed that due to the use of near-wellbore stimulation in Cases 3 and 4, the hydrate saturation near the stimulation area is significantly lower than that in the same area in Cases 1 and 2. This demonstrates the crucial role of high-permeability channels in expanding effective dissociation, and the direct consequence is the significant increase in gas production.

3.2.4. Spatial Variation in Gas Saturation

Driven by pressure difference, in Cases 1 and 2, high-saturation free gas migrated from TPL into the wellbore, resulting in a continuous decrease in gas saturation around the wellbore over 360 days, as shown in Figure 10. Based on boundary sealing, Cases 3 and 4 also adopted near-wellbore stimulation to accelerate the dissociation of hydrates near the stimulation area, leading to a low-saturation gas accumulation near the stimulation area. This low-saturation gas accumulation zone is a manifestation of the efficient dissociation of hydrates and rapid gas extraction. As production proceeds, hydrate dissociation near the stimulation area is more intensive with a wider range, and the low-saturation gas accumulation zone near this area is also growing steadily. The evolution pattern of gas saturation is highly consistent with the region of decreasing hydrate saturation in Figure 9, indicating that near-wellbore reservoir stimulation has created high-permeability channels, achieving more efficient utilization of the reservoir.

4. Conclusions

Based on field data from China’s first offshore NGH trial production, this work employed numerical simulation to analyze how boundary sealing, near-wellbore stimulation, and step-wise depressurization enhance the development of Class 1 NGH reservoirs via horizontal wells. The results provide a theoretical reference for exploiting such reservoirs in various regions. The main conclusions are as follows:

The comprehensive development strategy integrating boundary sealing, near-wellbore stimulation, and step-wise depressurization can significantly enhance the recovery rate compared to depressurization using horizontal wells alone. In this work, when using boundary sealing and near-wellbore stimulation, combined with step-wise depressurization, the V_g and R_gw increased to 220.1% and 102.6%, respectively, compared to the base case. The boundary sealing enables the pressure to be transmitted more widely and uniformly in the reservoir, effectively preventing the invasion of bottom water and concentrating pressure energy to drive the dissociation of hydrates and the recovery of free gas. The near-wellbore stimulation has formed high-permeability channels around the wellbore, enhanced pressure propagation, accelerated hydrate dissociation, and significantly increased gas production. Step-wise depressurization effectively suppresses water production and improves the overall gas-to-water ratio. The combined use of multiple methods can effectively promote the more complete and uniform dissociation of NGH reservoirs, significantly improve recovery efficiency, and achieve more robust and sustainable efficient production.

The conclusion of this article is based on numerical simulation research, and practical applications still face several challenges, including the precise construction of sealing layers in deep-sea environments, engineering control of near-wellbore stimulation, and potential geological deformation and sand production risks that may occur in long-term production. These engineering and geomechanical issues need to be fully evaluated in subsequent research and on-site testing.