Numerical Simulation of Gas Production Behavior Using Stepwise Depressurization with a Vertical Well in the Shenhu Sea Area Hydrate Reservoir of the South China Sea

Abstract

Stepwise depressurization is an important depressurization strategy in the development of natural gas hydrates. This work numerically analyzes the effects of different depressurization gradients and constant pressure durations on gas and water production during stepwise depressurization extraction with a vertical well in the Shenhu Sea area hydrate reservoir of the South China Sea. The results indicate that stepwise depressurization can reduce water production and raise the gas-to-water ratio in the early stages of production while ensuring cumulative gas output. When the vertical well is deployed at the model’s center with a completion length of 70 m and a constant pressure duration of 10 days, a depressurization gradient of 0.5 MPa, stepwise depressurization by 6 MPa, and continuous production for one year is achieved. Compared with direct depressurization, its cumulative gas production is 2.966 × 10⁶ ST m³, which only decreases by 2.94%. However, it maintains a higher gas-to-water ratio in the early stages of production. Considering factors such as engineering operability, cumulative gas output, and gas-to-water ratio, it is recommended to use a small pressure gradient and a medium constant pressure stabilization time for stepwise depressurization Stepwise depressurization can maintain a high gas-to-water ratio while ensuring gas production and reducing water production can alleviate sand production problems and improve economic efficiency. The understanding gained from this work has reference value for the development of similar hydrate reservoirs worldwide.

1. Introduction

Sustainable development of the global economy and technology relies on a stable energy supply. Currently, going green and low-carbon is the main direction for global energy structure adjustment, and there is an urgent need to develop clean energy. Natural gas hydrates (NGHs) are widely distributed in the permafrost and marine continental shelf. Its industrial exploitation may effectively promote global energy structure adjustment [1,2,3,4,5,6,7]. Japan and China’s offshore NGH testing production projects confirm the technological superiority of the depressurization method for NGH extraction [8,9,10,11]. However, the average daily gas output remains far below the commercial development standard of 50 × 10⁴ m³/d [2]. Therefore, many scholars focus on how to significantly increase production capacity and achieve effective recovery of NGHs, such as Too et al., 2018, who confirmed the possibility of forming artificial fractures in synthetic methane hydrate-bearing sand, which provides an opportunity for the efficient development of NGH [12]. Nair et al., 2018 conducted experimental studies on the dissociation behavior of hydrates in clay hydrate reservoirs using depressurization, thermal stimulation, and combination methods, and the results showed that the combination method had the best effect [13]. Aghajari et al., 2019 found that methane production during NGH exploitation is highly correlated with reservoir porosity and permeability, while the dissociation rate is positively correlated with depressurization and inversely correlated with reservoir temperature [14]. Vedachalam et al., 2020 believe that synchronous depressurization of 40 wells can achieve commercial exploitation of NGH in the Krishna Godavari Basin of India [15]. Terzariol et al., 2021 found that there is a synergistic depressurization effect between wells when multiple wells are synchronously depressurized, and its efficiency is higher than that of an equal number of independent working wells [16]. Wu et al., 2021 reviewed the relevant work and concluded that adopting different production well designs—such as horizontal wells, complex structure wells, etc., combined with varying strategies of depressurization and using different stimulation methods such as in situ heating or hot water injection, reservoir hydraulic fracturing, or near wellbore reservoir reconstruction—can significantly improve production capacity [17]. Optimizing the depressurization method is a noteworthy approach [18].

In recent years, efforts have been put into research on NGH development with the stepwise depressurization strategy. Phillips et al., 2019 experimentally studied the dissociation characteristics of hydrates with stepwise depressurization. In reservoirs with high salinity and saturation, salt diffusion and heat transfer slow down the dissociation of hydrates [19]. Yin et al., 2020 conducted a depressurization experiment in hydrate sediments and found that the water production rate grew linearly with the depressurization rate [20]. Zhao et al., 2020 found that dividing direct depressurization into several stages can accelerate hydrate decomposition [21]. Guo et al., 2020 conducted a set of experiments and found that stepwise depressurization helps alleviate initial water production; the fine pressure gradient during the high-yield water stage can increase the gas production rate to 31%, and the cumulative water production depends on the degree of depressurization [22]. Ravesh et al., 2019 studied the dissociation behavior of methane hydrates in porous media using single-step and multi-stage depressurization in a 25 L reactor and found that multi-stage depressurization can improve the gas recovery rate [23]. Yoon et al., 2021 conducted a numerical study on the geomechanical response of gas hydrate deposits in the Ulleung Basin of the Korea East Sea with various depressurization schemes and found that the productivity is similar among different depressurization schemes, with periodic depressurization schemes having relatively small subsidence [24]. Li et al., 2021 found that stepwise depressurization helps to increase the integrated gas-to-water ratio and avoid a significant decrease in reservoir temperature [25]. Yin et al., 2022 established a multi-field coupling model and investigated horizontal well depressurization’s gas production behavior. The results show that stepwise and cyclic depressurization can improve production capacity to varying degrees [26]. Wang et al., 2022 proposed an optimized simulation model based on the reservoir characteristics of the Shenhu Sea area and analyzed the impact of seepage characteristics caused by stepwise depressurization on hydrate extraction. The results indicate that relative gas permeability and porosity are key parameters affecting hydrate production [27]. Yin et al., 2022 found that the optimal duration of constant pressure for single-stage depressurization is 5–10 days. The optimal depressurization gradient for long-term stable production is 0.2 MPa/d [28]. Xue et al., 2023 conducted a numerical analysis of the gas production behavior of vertical wells with stepwise depressurization. The results show that the gas production increased by more than 10% [29]. Ge et al., 2023 proposed combining the hydrate dissociation model with a genetic algorithm to optimize the stepwise depressurization strategy. Research has found that increasing depressurization steps is detrimental to exploitation efficiency [30]. Wang et al., 2023 numerically analyzed the production behavior of stepwise depressurization extraction of NGH with a horizontal well. The results showed a significant decrease in water production [31]. Various studies in recent years likewise contribute relevant information and context on gas hydrate stability, destabilization, and changes to temperature and/or pressure influencing this hydrate stability (e.g., Biastoch et al., 2011; Burton et al., 2020; Kim and Zhang, 2022; Burton & Dafov, 2023) [32,33,34,35].

The insights gained from the above research on the stepwise depressurization strategy provide valuable theoretical references for practical application. Multi-well or group well exploitation is an important development direction for the commercial development of natural gas hydrates, and the vertical well is one of the important basic well types [17]. Currently, there is limited research on the stepwise depressurization of NGH reservoirs in the Shenhu Sea area with vertical wells. In this work, we established an ideal model of NGH reservoirs based on China’s first offshore NGH testing data, numerically studied the gas production behavior and reservoir physical characteristics’ evolution of vertical well stepwise depressurization, and focused on the effects of depressurization gradients and constant pressure durations on gas production behavior.

2. Methodology

2.1. Simulation Code

TOUGH+HYDRATE V1.0 can simulate various exploitation methods such as depressurization, thermal stimulation, and inhibitors, focusing on well-scale or reservoir-scale simulation (instead of basin-scale simulation, e.g., Piñero et al., 2016; Burton, 2022) [36,37,38]. The accuracy of the code has been verified through laboratory and field testing [39,40,41,42,43]. Specifically, the equilibrium model was adopted to simulate the stepwise depressurization by a vertical well [44]. This work assumes that sand production is controllable, Darcy’s law is effective, and the geological media does not deform. The main control equations for the simulation code are as follows [36]:

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

(1)

Here, $M^{\kappa }$ is the mass accumulation term, $F^{\kappa }$ is the Darcy flux vector, and $q^{\kappa }$ is the source/sink term. $V_n$, ${\Gamma }_n$, κ, and $t$ represent the grid volume (m³), grid surface area (m²), components, and time (s), respectively, whereas V and Γ are volume (m³) and surface area (m²), respectively.

2.2. Model Construction

The SHSC4 well in the Shenhu Sea area was selected for the simulations (Figure 1) [45]. The NGH reservoir at this site belongs to Class 1-type, which consists of three parts: the gas hydrate-bearing layer (GHBL, contains water and hydrates, with thicknesses of 35 m); the three-phase layer (TPL, contains water, free gas and hydrates, with thicknesses of 15 m); and the free gas layer (FGL, contains water and free gas, with thicknesses of 27 m) [10,46,47]. A multi-layer reservoir model with a size of 500 × 500 × 137 m was constructed under the x-y-z coordinate system, as shown in Figure 2a. Ten simulation cases were simulated to assess the stepwise depressurization with different depressurization parameters, including a 70 m long vertical wellbore deployed at the center of the model (model’s −21 and −91 m).

Table 1. The detailed settings of simulation cases.

Scenarios	Production Pressure Difference Regulation Process (MPa)	Depressurization Gradient (MPa)	Constant Pressure Duration (Day)
Case 01-DD-Ref	6.0	None	None
Case 02-SD1-20d	3.0→3.5→4.0→4.5→5.0→5.5→6.0	0.5	20
Case 03-SD1-10d	3.0→3.5→4.0→4.5→5.0→5.5→6.0	0.5	10
Case 04-SD1-05d	3.0→3.5→4.0→4.5→5.0→5.5→6.0	0.5	5
Case 05-SD2-20d	3.0→4.0→5.0→6.0	1.0	20
Case 06-SD2-10d	3.0→4.0→5.0→6.0	1.0	10
Case 07-SD2-05d	3.0→4.0→5.0→6.0	1.0	5
Case 08-SD3-20d	3.0→4.5→6.0	1.5	20
Case 09-SD3-10d	3.0→4.5→6.0	1.5	10
Case 10-SD3-05d	3.0→4.5→6.0	1.5	5

Note: DD is direct depressurization; SD is stepwise depressurization.

gives the detailed parameters for these cases.

The model is discretized into 23 × 23 × 81 grids in (x, y, z), with a total of 42,849 grids (Figure 2b). Specifically, corresponding to the radius of 0.1 m of the wellbore, the grid size of the wellbore in the x-y plane is 0.2 m × 0.2 m. To properly capture heat conduction, hydrate dissociation, and fluid flow, the vertical grid sizes of the GHBL, TPL, and FGL in the model were set to be 1.0 m.

2.3. Model Initialization and Boundary Conditions

The model is divided into three subdomains: the GHBL, TPL, and FGL. The top and bottom of the TPL are considered as the boundaries of the hydrate interface (i.e., the three-phase conditions for the coexistence of hydrates, gas, and water). Each subdomain is simulated separately to determine the initial pressure, temperature, and saturation steady-state conditions; during the process, the heat flux between each subdomain is adjusted by fine-tuning the geothermal gradient. When the heat flux of the contact surface between the three subdomains remains consistent, the combination is performed to complete model initialization, as shown in Figure 3 [48,49,50,51,52,53].

During the simulation process, the wellbore is considered the internal boundary, and fixed or stepwise production pressure differences are given to the wellbore grids according to different simulation cases. The wellbore grid is treated as a pseudo-porous medium with a porosity and permeability of 1 and 10,000 D, respectively [54]. The parameters for the reservoir and model are shown in

Table 2. Parameters of the reservoir.

Layer	Parameter	Value and Unit
OB [55,56,57,58,59]	Thickness	30 m
	Porosity	0.30
	Initial permeability	2.0 mD
GHBL [10,56,57,58,59]	Thickness	35 m
	Porosity	0.35
	Initial permeability	2.9 mD
	Initial hydrate saturation	Extracted from logging curve
TPL [10,56,57,58,59]	Thickness	15 m
	Porosity	0.33
	Initial permeability	1.5 mD
	Initial hydrate saturation	Extracted from logging curve
FGL [10,56,57,58,59]	Thickness	27 m
	Porosity	0.32
	Initial permeability	7.4 mD
	Initial free gas saturation	Extracted from logging curve
UB [55,56,57,58,59]	Thickness	30 m
	Porosity	0.30
	Initial permeability	2.0 mD

and

Table 3. Parameters of the model.

Parameter	Value and Unit
Wellbore radius [56,57,58,59]	0.1 m
Salinity [56,57,58,59]	3.5%
Grain density [56,57,58,59]	2600 kg/m3
Geothermal gradient [56,57,58,59]	43.653 °C/km
Gas composition [56,57,58,59]	100% CH4
Grain specific heat [56,57,58,59]	1000 J·kg−1·K−1
Dry thermal conductivity [56,57,58,59]	1.0 W·m−1·K−1
Wet thermal conductivity [56,57,58,59]	3.1 W·m−1·K−1
Capillary pressure model [56,57,58,59]	$P_{cap}=-P_0{\left[{\left(S^{*}\right)}^{-1/\mathsf{\lambda }}-1\right]}^{1-\mathsf{\lambda }}$, $S^{*}=\frac{\left(S_A-S_{irA}\right)}{\left(S_{mxA}-S_{irA}\right)}$
SmxA (maximum water saturation)	1
λ (porosity distribution index)	0.45
P0 (initial capillary pressure)	104 Pa
Relative permeability model [56,57,58,59]	KrA = [(SA − SirA)/(1 − SirA)]nA, KrG = [(SG − SirG)/(1 − SirA)]nG
nA (aqueous phase permeability reduction index)	3.5
nG (gas phase permeability reduction index)	2.5
SirG (residual gas saturation)	0.03

, respectively.

2.4. Model Validation

Based on China’s first offshore NGH testing data, a vertical wellbore was deployed at the model’s center with a completion length of 70 m [10,56,57,58,59]. Under the production pressure difference of 3 MPa, a 60-day gas production fitting was conducted to verify the model’s accuracy [10,56,57,58,59]. As shown in Figure 4, the fitting results are within the acceptable range.

3. Results and Analysis

3.1. Gas and Water Production

Two simulation cases were set up to compare the evolution characteristics of gas and water production behavior and reservoir physical properties between direct and stepwise depressurization: Case 01, which has direct depressurization by 6 MPa and continuous production for one year; Case 03, which has a constant pressure duration of 10 days, a depressurization gradient of 0.5 MPa, stepwise depressurization by 6 MPa, and continuous production for one year. From the gas production rate (Q_g) curves and cumulative gas production (V_g) curves in Figure 5a,b, it can be seen that Q_g of Case 01 reached its maximum value at the beginning, then slowly decreased to about 0.7 × 10⁴ ST m³/d, and the V_g reached 3.056 × 10⁶ ST m³. Direct and significant depressurization can quickly reduce reservoir pore pressure, promote rapid dissociation of hydrates, and rapidly decrease reservoir temperature. The huge pressure difference can cause the fluid to flow too quickly to the wellbore, which can easily lead to sand production problems. Usually, the seepage and heat transfer processes in the reservoir are slower than pressure propagation; so, significant depressurization cannot effectively exert the depressurization effect. The Q_g curve of Case 03 showed a stepwise increase followed by a gradual decrease within the initial 70 days of production and tended to be consistent with Case 01 around 135 days. The V_g reached 2.966 × 10⁶ ST m³, only a decrease of 2.94% compared to Case 01. Stepwise depressurization divides the significant depressurization into several small cycles with a small depressurization gradient, and the peak value of Q_g is much smaller than the direct depressurization, and a small depressurization gradient is beneficial for sand production control. Different from the above results, Xue et al., 2023 numerically studied natural gas hydrate production in the Nankai Trough of Japan with a vertical well and found that the stepwise depressurization resulted in a 10% increase in gas production compared to direct depressurization, which is due to the differences in hydrate reservoir types [29]. From Figure 5c,d, it can be seen that the water production rate (Q_w) curves of Case 03 in the early stages of production also show a stepwise increase followed by a gradual decrease trend and tends to be consistent with Case 01 around 135 days. Compared to Case 01, Case 03 can reduce the Q_w, this is similar to the research results of Wang et al., 2023 where stepwise depressurization can significantly reduce water production [31]. While lower water production is beneficial for sand production control, throughout the entire production cycle, Case 03 has better gas-to-water ratio (R_gw) performance compared to Case 01, especially during the initial 120 days of production, and a higher R_gw means better exploitation efficiency. The analysis results indicate that stepwise depressurization has a more significant impact on water production. This phenomenon may be related to the “gas slippage effect” [60,61,62].

3.2. Physical Properties

3.2.1. Pressure

From Figure 6, it can be observed that Case 01 and Case 03 have limited pressure propagation in the reservoir around the TPL. This is because the permeability of the reservoir around the TPL decreases with the formation of secondary hydrates, and the pressure propagation range of Case 03 is smaller in the early production stage. The same phenomenon can be observed in Case 01 and Case 03. Compared to the reservoir located in the GHBL, the pressure propagation range of the reservoir located in the bottom free gas layer is significantly smaller, which is caused by the gas expansion effect.

3.2.2. Temperature

From Figure 7, it can be observed that both Case 01 and Case 03 have formed low-temperature zones in the reservoir around the wellbore. This is due to the dual effects of hydrate dissociation in the GHBL and the Joule–Thomson effect. The Joule–Thomson effect caused by the significant direct depressurization in Case 01 during the initial production stage is stronger, resulting in a larger range of temperature drops in the reservoir around the wellbore. However, stepwise depressurization avoids a rapid drop in reservoir temperature, providing favorable conditions for later production. This phenomenon is similar to the research results of Li et al., 2021 where stepwise depressurization helps to avoid a significant decrease in reservoir temperature [25].

3.2.3. Hydrate and Gas Saturation

From Figure 8, it can be observed that during the initial stage of production within 60 days, Case 01 has a larger range of hydrate dissociation in the GHBL. After one year of production, Case 01 and Case 03 both have a hydrate dissociation radius of about 2 m at the GHBL. Affected by the strong Joule–Thomson effect, secondary hydrates were formed in the reservoir around the wellbore at the TPL in Case 01 and Case 03. As production progressed, the amount of the secondary hydrates gradually increased. At the end of production, the saturation and amount of the secondary hydrates in Case 01 and Case 03 tended to be consistent. It is worth noting that the amount of secondary hydrates formed in the TPL of Case 03 is relatively small, which also means that the stepwise depressurization method can help alleviate the formation of secondary hydrates in the early stages of production.

From Figure 9, it can be observed that as the hydrate dissociation in the GHBL and the free gas entered the wellbore in the TPL, both Case 01 and Case 03 formed gas at a relatively high saturation zone and gas at a relatively low saturation zone in the GHBL and TPL, respectively. In both cases, it can be observed that free gas moves upwards and downwards along the wellbore. The closer the free gas is to the wellbore, the faster it migrates, forming conical and inverted conical gas-bearing zones, respectively. After 120 days of production, the dissociation of hydrates located in the upper part of the TPL was observed in both Case 01 and Case 03, resulting in local high saturation gas accumulation around the wellbore.

4. Discussion

4.1. Effects of Depressurization Gradient

As shown in Figure 10, Case 04, Case 07, and Case 10 have the same constant pressure duration of 5 days, with depressurization gradients of 0.5 MPa, 1.0 MPa, and 1.5 MPa, respectively. From Figure 10b, it can be seen that under the premise of a fixed constant pressure duration, the V_g increases with the increase in the depressurization gradient. The V_g after one year of production is 3.019, 3.039, and 3.045 × 10⁶ ST m³, respectively. Compared with direct depressurization, it decreases by 1.21%, 0.55%, and 0.35%, respectively, and there is no significant difference between them. As shown in Figure 10d, the R_gw decreases with the increase in the depressurization gradient. Throughout the entire production cycle, Case 04 has the best R_gw performance. Under the premise of fixed constant pressure duration and considering engineering operability, V_g, and R_gw, Case 04 is the optimal choice among them.

4.2. Effects of Constant Pressure Duration

As shown in Figure 11, the depressurization gradients for Case 02, Case 03, and Case 04 are all set to 0.5 MPa, and the constant pressure durations are set to 20 days, 10 days, and 5 days, respectively. From Figure 11b, it can be seen that under the premise of a fixed depressurization gradient, the V_g decreases with the increase in constant pressure duration. The V_g after one year of production is 2.864, 2.966, and 3.019 × 10⁶ ST m³, respectively. Compared with direct depressurization, it decreased by 6.28%, 2.94%, and 1.21%, respectively. As shown in Figure 11d, the R_gw increases with the increase in constant pressure duration. Throughout the entire production cycle, Case 02 has the best R_gw performance. Under the premise of fixed depressurization gradient and considering the V_g and R_gw, Case 03 is the optimal choice among them.

4.3. Comparison of Production Performance

Figure 12 and Figure 13 show the R_gw and V_g for all simulated cases, respectively. Case 02 has the best R_gw performance, but its V_g of 2.864 × 10⁶ ST m³ is the lowest among all stepwise depressurization cases, a decrease of 6.28% compared to Case 01. The V_g of 3.045 × 10⁶ ST m³ in Case 10 is the highest among all stepwise depressurization cases, with only a decrease of 0.35% compared to Case 01, but its R_gw is the worst among all stepwise depressurization cases. The V_g of Case 03 is 2.966 × 10⁶ ST m³, which is only 2.94% lower than Case 01. Its R_gw ranks among the top in all stepwise depressurization cases, and it is also easy to achieve stepwise depressurization with small depressurization gradients in engineering, making it the best choice in all cases.

4.4. Implications and Future Recommendations

Multi-well or group well exploitation is an important development direction for the commercial development of NGHs, and the vertical well is one of the important basic well types. However, there is limited research on the stepwise depressurization of NGH reservoirs in the Shenhu Sea area with vertical wells. This work is based on SHSC4 well-logging data to establish an ideal numerical model and study the impact of different depressurization methods on gas and water production in Shenhu Sea area’s NGH reservoirs with a vertical well. The results show that compared to direct depressurization, stepwise depressurization can increase the gas-to-water ratio while ensuring gas production. Maintaining a high gas-to-water ratio during natural gas hydrate exploitation has the following positive effects: 1. A decrease in water production can reduce the energy consumed for lifting or pumping water in engineering, which is beneficial for improving economic efficiency; 2. Related studies have shown that water production is an important factor leading to sand production in reservoirs [63,64,65,66]. A decrease in water production can alleviate sand production in reservoirs and avoid a series of engineering problems caused by sand production, such as wear of wellbore pipelines due to sand production and failure of sand control due to excessive sand production. This is similar to the numerical analysis results of Yin et al., 2022 using horizontal wells to stepwise-depressurize natural gas hydrates in the Shenhu Sea area [28]. It is recommended to use a smaller pressure gradient and a medium constant pressure duration for stepwise depressurization in Class 1-type hydrate reservoirs with a vertical well. This work helps us better understand the Class 1-type hydrate reservoirs and has reference value for the development of similar hydrate reservoirs worldwide. Furthermore, this work contributes to gas hydrate, natural gas, and petroleum system studies broadly speaking (e.g., Magoon & Dow, 1994; Burton et al., 2018, 2019; Jang et al., 2020; Almashwali et al., 2022) [67,68,69,70,71].

5. Conclusions

Based on the NGH testing data in the Shenhu Sea area, an ideal NGH reservoir model of the SHSC4 well was established, and the vertical well depressurization production simulation was conducted. A comprehensive analysis was made between direct and stepwise depressurization, including gas and water production behavior, as well as the evolution characteristics of reservoir physical properties. The sensitivity of the depressurization gradient and constant pressure duration were analyzed, and the following conclusions can be drawn:

(1) Stepwise depressurization can divide direct depressurization into several small cycles with small amplitude depressurization. While ensuring V_g, it can reduce water production and increase R_gw in the early stages of production, which is beneficial for sand production control; in the early stage of production, stepwise depressurization can avoid a rapid decrease in reservoir temperature and alleviate the formation of secondary hydrates. Compared to the depressurization gradient, the effect of constant pressure duration on V_g and R_gw is more significant.

(2) When the vertical well is deployed at the model’s center with a completion length of 70 m and a constant pressure duration of 10 days, a depressurization gradient of 0.5 MPa is used for depressurization by 6 MPa for continuous production for one year. Compared to direct depressurization, its V_g is 2.966 × 10⁶ ST m³, a decrease of only 2.94%, but it maintains a higher R_gw in the early stages of production. It is recommended to use a small pressure gradient and a medium constant pressure stabilization time for stepwise depressurization exploitation, which can effectively improve production efficiency.

(3) Stepwise depressurization can maintain a high gas-to-water ratio while ensuring gas production, improving economic efficiency, and reducing water production can alleviate sand production problems. These understandings have reference value for the development of similar hydrate reservoirs worldwide. In the future, we will combine auxiliary heating, near-wellbore reservoir reconstruction, hydraulic fracturing, and other methods to further study the impact of stepwise depressurization on gas and water production in Class 1-type hydrate reservoirs.