Study on Physical Simulation of Shale Gas Dissipation Behavior: A Case Study for Northern Guizhou, China

Abstract

The Longmaxi from the Anchang Syncline in northern Guizhou exhibits a high degree of thermal evolution of organic matter and significant variation in gas content. Because the synclinal is narrow, steep, and internally faulted, the mechanisms controlling shale gas preservation and escape remain poorly understood, complicating development planning and engineering design. Research on oil and gas migration and accumulation mechanisms in synclinal structures is therefore essential. To address this issue, three proportionally scaled strata—pure shale, gray shale, and sandy shale—were fabricated, and faults and artificial fractures with different displacements and inclinations were introduced. The simulation system consisted of two glass tanks (No. 1 and No. 2). Each tank had three rows of eight transmitting electrodes on one side, and a row of eight receiving electrodes on the opposite side. Tank 1 remained fixed, while Tank 2 could be hydraulically tilted up to 65° to simulate air and water migration under varying formation inclinations. A gas-water injection device was connected at the base. Gas was first injected slowly into the model. After injecting a measured volume (recorded via the flowmeter), the system was allowed to rest for 24–48 h to ensure uniform gas distribution. Water was then injected to displace the gas. During displacement, Tank 1 remained horizontal, and Tank 2 was inclined at a preset angle. An embedded monitoring program automatically recorded resistivity data from the 48 electrodes, and water-driven gas migration was analyzed through resistivity changes. A gas escape rate parameter (G_d), based on differences in gas saturation, was developed to quantify escape velocity. The simulation results show that gas escape increased with formation inclination. Beyond a critical angle, the escape rate slowed and approached a maximum. Faults and fractures significantly enhanced gas escape.

1. Introduction

In 2020, the proven geological resources of shale gas in northern Guizhou reached 3380 billion cubic meters at burial depths of less than 3500 m, indicating strong development potential prospects. The Longmaxi formation occurs at depths of 2500–3000 m. During the late Ordovician to early Silurian, the Anchang Syncline was situated in a transitional setting from deep to shallow shelf, leading to the decomposition of the organic-rich black shales. However, because of the synclinal structure and subordinate features, such as faults and microfractures, the mechanisms controlling shale gas preservation and escape remain unclear [1,2,3,4]. To address this, the present study conducted a physical simulation to investigate the dissipation behavior of shale gas within the syncline.

Following the principle of proportional scaling and the key characteristics of the Anchang syncline strata, two glass tanks were constructed: Tank 1 was fixed, and Tank 2 could be tilted at various angles. One row of eight electrodes was installed on one side of each tank as transmitting electrodes, and three rows of eight electrodes each were placed on the opposite side as receiving electrodes. Three simulated strata with different lithologies—pure shale, calcareous shale, and sandy shale—were then prepared. By monitoring resistivity changes during injection and drainage, the study analyzed gas–water migration and the preservation or dissipation of shale gas, providing a basis for further research on evolution in this block. Research on the dissipation mechanisms and controlling factors of shale gas in the northern Guizhou syncline, combined with reservoir evolution simulations, clarified the evolution and late-stage dissipation characteristics of shale gas in the region. A reservoir evolution and late-stage dissipation model for syncline-type shale gas in the Wufeng–Longmaxi Formation of northern Guizhou was developed to support shale gas exploration and development in the area.

2. Geological Background and Reservoir Characteristics

2.1. Geological Background of Anchang Syncline

Northern Guizhou lies in the transition zone between the tank–horst structures at the edge of the Sichuan Basin, bounded by the central Sichuan and the central Guizhou uplifts. The Anchang Syncline, located on the outer margin of the Sichuan Basin, within the strongly deformed northern Guizhou fold belt, extends from Daozhen in the north to Zheng’an in the south and belongs to the Yangtze Platform. Multiple intense geological tectonic movements led to regional uplift, resulting in numerous large faults in the center and margins of the basin (Figure 1). These tectonic stages include Caledonian (late Early Silurian–late Devonian), wherein the Hanjiadian Formation shale was exposed and weathered, contributing to peneplanation; Hercynian (Early Carboniferous), a brief transgression Huanglong Formation limestone, followed by uplift, erosion, valley incision, and the formation of an initially weathering crust; and Late Carboniferous–Early Permian, a stable hot, humid climate improved ferrallitic weathering. Subsequent transgression–regression cycles and uplift drove karstification, forming features such as karren, solution grooves, and lake basins (e.g., the Daozhen–Zheng’an depression). In the Middle Permian (Liangshan Stage), weathering intensified, and aluminum-rich clastics accumulated in low-lying areas, forming bauxite layers. In the Mesozoic–Cenozoic period, repeated tectonic movements (e.g., Yanshanian, Himalayan) produced complex folds and faults, creating a heterogeneous in situ regime [5,6,7,8]. Multiple phases of deformation created numerous faults, especially in the southwest part of the syncline, with the eastern limb showing more complex faulting than the western limb. Rapid sedimentation combined with structural complexity produced distinctive shale gas reservoir conditions characterized by the following variables: lithostratigraphic assemblages, high organic matter maturity, heterogeneous gas-bearing properties, and abundant internal faults within the narrow, steep syncline (Figure 2) [9,10,11,12].

Under three directional compressive deformations, the Anchang Syncline evolved into a residual syncline. It is tightly closed in the northeast and opens toward the southwest, trending 20–30° NNE, approximately 30 km long and 5–12 km wide. The hinge shows wave-like undulations, and the axis curves slightly in an “S” shape [13,14,15]. The eastern limb dips 290–340° at 17–34°, while the western limb is steeper, dipping 135–150° at 16–75°, giving the structure a “gentle east–steep west, gentle south–steep north” geometry. Deposition occurred in a restricted shallow marine environment, preserving a thick, continuous sedimentary sequence. Eight potential shale or shale–marl formations are present: Niutitang, Baota, Wufeng-Longmaxi, Xintan, Shiniulan, Qixia, Maokou, and Longtan formations (Figure 3) [16,17,18].

2.2. Characteristics of the Longmaxi Formation in the Anchang Syncline

Drilling data from Wells Anye-1 to Anye-5 in the Anchang syncline of northern Guizhou show that the oldest encountered stratum is the Ordovician Baota Formation. From oldest to youngest, the stratigraphic sequence comprises the Baota, Wufeng, Guanyinqiao, Longmaxi, Xintan, Shiniulan, Hanjiadian, Liangshan, Qixia, Maokou, Heshan, Yelang, and Jialingjiang formations. The Longmaxi shale lies at a burial depth of 2200–2400 m and is 30–150 m thick. It is dominated by black carbonaceous shale composed mainly of quartz and clay minerals. In the Anchang syncline, it appears as grayish-black to black shale, with abundant graptolites and pyrite, characteristic of a deep-shelf sedimentary environment. For example, in Well AY-2, quartz content ranges from 50% to 75% (average 62.73%), and the clay content ranges from 14% to 36% (average 21.3%) [7]. The average formation pressure is 21.3 MPa, and the temperature is 70 °C. Core experiments show that the Longmaxi L1 formation has an average porosity of 7.48%, an average permeability of 0.0255 mD, and an average density of 2.692 g/cm³. The simulated core sample has an average porosity of 7.22%, an average permeability of 0.6689 mD, and an average density of 2.682 g/cm³. The Wufeng–Longmaxi interval generally has good sealing above and below and does not produce water. To accelerate gas migration and obtain results within a reasonable timeframe, N₂ was first used for air displacement. However, the resulting resistivity was so high that measurements exceeded the instrument’s upper limit, preventing detection of gas migration. Therefore, water displacement was used instead, although the resistivity of the injected water was much higher than that of the actual formation water.

3. Physical Simulation Research

3.1. Fabrication of Simulation Device and Simulated Strata

The syncline simulation device consists of three components:

(1)

Shale model containers: Two tempered-acrylic tanks (Tank 1 and Tank 2) used to hold the shale models. Tank 1 is fixed, while Tank 2 can be tilted. Each tank measures 0.8 m × 0.4 m × 0.3 m and has a wall thickness of 0.05 m.

(2)

Power system: Including a hydraulic drive unit and a fluid injection–drainage unit, the hydraulic system allows Tank 2 to tilt from 0° to 65°, while the injection–drainage unit is used for gas and water injection and withdrawal.

(3)

Measurement system: Used to measure the resistivity of the simulated strata, including an IM3750-type LCR meter and an Arduino-Mega2560 programmable module.

A motor drive is installed beneath Tank 2, and its tilt angle is adjusted using a lifting switch. To monitor fluid migration in real-time during injection and drainage, 24 circular copper electrodes (3 rows × 8 columns) were installed on one inner wall of each tank as receiving electrodes, and eight electrodes (1 row × 8 columns) were installed on the opposite wall as transmitting electrodes, 64 electrodes in total. Each electrode was wired, numbered, routed outside the tanks, and connected to the Mega2560 module. Automatic lattice measurement was controlled by embedded software. The simulation device is shown in Figure 4.

Approximately 2.5 tons of shale outcrops were collected from the study area. After crushing, the shale was mixed in prescribed proportions to prepare the simulated stratigraphic model. Based on the stratigraphy, lithology, fault types, fault displacements, and fractures of the Longmaxi Formation, the model was constructed following the principle of proportional scaling. The specific parameters are given in

Table 1. Parameters related to simulated strata.

Tank No.	Thick-ness/cm	Lithology	Electrode No.	Fault No.	Fault Throw/mm	Fault Dip Angle/°	Fracture Length/cm	Fracture Dip Angle/°	AZIM/°	DIP/°
1	13	Silty Shale	/	/	/	/	/	/	/	0
		Shale 70%, Silt 25% Limestone 5%
	9	Silty Shale		1	5	71	/	/	/	0
		Shale 80%, Silt 20%
	8	Pure Shale	3-3~3-6	1	5	71	21	30	110	0
2	13	Silty Shale	1-4~1-5	0	0	0	14	0	95	<65°
		Shale 70%, Silt 25% Limestone 5%
	9	Silty Shale	2-2~2-3	1	2	71	20	10	120	<65°
		Shale 80%, Silt 20%
	8	Pure Shale	3-2~3-3	1	1	71	30	20	110	<65°
			3-4~3-6	2	1	43	25	60	100	<65°

. The simulated strata consisted of three layers from bottom to top: pure shale, silty shale, and calcareous silty shale. Standard samples with the same composition were prepared for each layer to measure porosity and resistivity parameters. The average porosities of the three layers in Tank 1 were 5.82%, 8.42%, and 7.47%, respectively; in Tank 2, they were 5.63%, 8.51%, and 7.44%, respectively. The resistivity of the injected water was 36.841 Ω·m.

Based on the device parameters, the pore volume of each layer was calculated (as listed in

Table 2. Pore volume of each layer (mL).

Tank No.	Layer No.	Thickness /cm	Length /cm	Width /cm	Permeability /mD	Density /g/cm3	Porosity /%	Pore Volume	Total Pore Volume
1	1	13	80	40	0.0491	2.674	7.47	3107.52	7022.4
	2	9	80	40	0.0979	2.512	8.42	2424.96
	3	8	80	40	0.0617	2.817	5.82	1489.92
2	1	13	80	40	0.0526	2.688	7.44	3095.04	6987.2
	2	9	80	40	0.1063	2.839	8.51	2450.88
	3	8	80	40	0.0337	2.562	5.63	1441.28

).

3.2. Injection–Drainage Scheme

Based on the pore volume of each layer, water was injected in the following manner: 1000 mL of water was slowly injected from the bottom at each step. Injection stopped after 7000 mL had been injected into Tank 1 and 6960 mL into Tank 2.

Table 3. Pore volume and water saturation of each layer (%).

Tank No.	Layer No.	Porosity /%	Volume /mL	1000	2000	3000	4000	5000	6000	7000
				mL
1	1	7.47	3107.52	0.00	0.00	0.00	2.74	34.92	67.10	99.28
	2	8.42	2424.96	0.00	21.03	62.27	100.00	100.00	100.00	100.00
	3	5.82	1489.92	32.88	100.00	100.00	100.00	100.00	100.00	100.00
2	1	7.44	3095.04	0.00	0.00	0.00	3.48	35.79	68.10	99.12
	2	8.51	2450.88	0.00	22.80	63.60	100.00	100.00	100.00	100.00
	3	5.63	1441.28	30.62	100.00	100.00	100.00	100.00	100.00	100.00

presents the pore volume of each layer and the corresponding water saturation after injection. Note that the water saturation of Layer 2-1 in Tank 2 was calculated using the 6960 mL injection volume.

Injection–drainage procedure:

(1)

Gas injection: Connect the air pump and inject gas into the simulated formation at a constant pressure of 5 MPa, equivalent to hydrostatic pressure at ~500 m depth. When 7000 mL has been injected into Tank 1 and 6960 mL into Tank 2, gas escapes from the outlet, indicating full gas saturation. Close the gas inlet valve and stop injection.

(2)

Stabilization: Let the system stand for 48 h, periodically checking for any gas leakage at the outlet.

(3)

Water injection: Open the water inlet and slowly inject water from the bottom of the tank. The injected water has a resistivity of 36.841 Ω·m at 20 °C. Measure resistivity after each 1000 mL increment.

(4)

Equilibration: As water accumulates at the bottom, allow it to stand until it gradually seeps into the pore space before continuing injection. The full water injection phase lasts 69 h and 50 min.

(5)

Resistivity monitoring: Activate the automatic monitoring system to record resistivity in real time. Tank 1 remains horizontal, while Tank 2 is tilted upward in 10° increments, with resistivity recorded at each angle.

(6)

Data collection: Because Tank 1 remains stationary, only one dataset is collected for it. Tank 2 yields 8 × 8 = 64 datasets. The gas dissipation rate G_d is calculated, and resistivity maps are constructed to analyze gas–water distribution. Only key stages are presented due to the large volume of data.

3.3. Gas Escape Rate

The purpose of this simulation is to observe and analyze the behavior of gas dissipation in syncline strata during water-driven gas displacement. Gas dissipation can be approximated by monitoring changes in formation resistivity. During static displacement, the volumes of injected gas and water are measured with a flowmeter. With known porosity, the pore volumes occupied by water and gas (i.e., their saturations) can be calculated.

To quantify gas dissipation, the gas dissipation rate G_d is introduced. Assuming the strata contain only two fluid phases (water and gas), water saturation (S_w) and gas saturation (S_g) sum to 100%. According to Archie’s equation, $S_w=\sqrt[n]{abRw/Rt{\mathrm{\varnothing }}^m}$, the relationship between pore fluids and formation resistivity (Rt) is expressed as Equation (1):

$$S_w^n=K{\displaystyle \frac{1}{Rt}}$$

(1)

where K is a comprehensive formation factor reflecting lithology, pore structure, formation water, cementation type, and related properties. During water drainage, S_w increases while S_g decreases sequentially. Thus, the following sequence of water saturations is obtained:

$${S_w}_n>{S_w}_{n-1}>\cdots >{S_w}_3>{S_w}_2>{S_w}_1$$

(2)

The gas dissipation rate G_d is defined as the ratio of the change in gas saturation (∆S_g) between two adjacent stages to the gas saturation of the latter stage during the water-driven gas displacement. This ratio helps to eliminate the influence of factors such as lithology, porosity, cementation, and formation water. Because the simulated strata have relatively uniform lithology, simple structure and pore geometry, and high porosity, n ≈ 2. Thus, G_d can be expressed by Equation (3):

$$G_d={\displaystyle \frac{∆S_g}{{S_w}_2}}={\displaystyle \frac{{S_w}_2-{S_w}_1}{{S_w}_2}}=1-\sqrt{{\displaystyle \frac{{Rt}_2}{{Rt}_1}}}$$

(3)

In fact, the gas escape rate derived from resistivity changes after two consecutive water injections can be normalized through ratio calculations, making it independent of the specific model used. This is because the ratio remains consistently preserved across the Archie model, the dual-water model, and other alternatives, thereby leaving the value of G_d unchanged.

During water-driven gas displacement, S_w increases, and formation resistivity decreases (i.e., Rt₂ < Rt₁), resulting in G_d > 0, referred to as positive dissipation (water invasion and gas loss). During uplift, formation water moves downward, allowing for pores in the uplifted strata to fill with gas; in this case, Rt₂ > Rt₁ and G_d < 0, referred to as reverse dissipation (water withdrawal and gas filling).

It should be emphasized that the gas dissipation rate is not equivalent to the gas dissipation volume. G_d reflects the degree of dissipation; a larger value indicates a higher dissipation rate and easier gas loss. The next section analyzes the gas dissipation behavior using physical simulation and real-time resistivity monitoring.

4. Analysis of Physical Simulation Results

4.1. Analysis of Gas Dissipation Rate in Tank 1

After gas injection, the device was left to stand for 48 h before water injection began. After each 1000 mL increment of water injection, the system was allowed to stand, and resistivity was measured every 30 min until two consecutive readings were nearly identical, indicating fluid equilibrium in the formation. Tank 2 was then tilted, with resistivity measured at each 10° increase. The gas dissipation rate was calculated from resistivity changes to analyze gas diffusion. Throughout the tilting of Tank 2, Tank 1 remained horizontal and served as a control. The following section analyzes the resistivity and dissipation rate of Tank 1 throughout the injection–drainage process.

The gas dissipation behavior during water injection and tilting was evaluated using Equation (3).

Table A1. Resistivity of Tank 1 under different water injection volumes (Ω·m).

No.	0	1000 mL	2000 mL	3000 mL	4000 mL	5000 mL	6000 mL
1-1	840.65	840.32	840.88	840.68	837.3	820.87	802.12
1-2	839.77	840.45	840.66	840.51	837.52	821.29	803.90
1-3	840.16	841.72	840.82	842.21	837.90	819.77	803.55
1-4	840.22	841.31	841.72	841.12	837.88	822.07	802.61
1-5	839.81	840.77	843.19	840.95	837.22	820.71	802.70
1-6	840.32	839.45	838.75	839.95	838.54	821.38	802.57
1-7	840.31	839.75	840.51	839.92	837.56	821.30	804.06
1-8	840.44	840.34	840.19	841.73	839.32	820.67	801.69
2-1	613.14	612.91	598.03	572.8	549.78	549.64	549.59
2-2	613.15	612.96	598.91	572.63	549.82	549.61	549.87
2-3	612.44	612.02	598.76	567.13	550.20	550.01	550.04
2-4	612.77	614.34	598.57	572.52	545.73	549.63	549.48
2-5	612.65	613.27	598.29	572.41	550.31	547.12	545.84
2-6	613.22	614.27	598.27	573.02	550.45	550.40	550.33
2-7	612.86	612.06	598.81	572.93	549.68	550.39	550.10
2-8	613.11	612.97	597.99	573.27	549.99	550.27	549.9
3-1	560.83	559.4	530.33	530.17	530.07	530.38	529.63
3-2	560.79	559.36	530.29	530.80	530.65	529.90	529.09
3-3	561.13	556.70	504.40	504.21	503.42	503.38	503.24
3-4	560.54	556.22	503.88	503.16	502.47	502.14	501.92
3-5	560.98	556.58	504.27	504.11	503.23	503.73	503.45
3-6	560.73	556.38	504.05	504.12	503.99	503.56	503.70
3-7	561.29	559.84	530.77	530.61	530.61	530.61	530.61
3-8	561.04	559.6	530.53	530.24	530.24	530.24	530.24

in Appendix A presents the resistivity measurements for Tank 1 during the entire water injection stage.

First, the gas dissipation rate of Tank 1 after each water injection was calculated, and the results are listed in

Table A2. Gas dissipation rate of Tank 1 under different water injection volumes.

No.	Gd1	Gd2	Gd3	Gd4	Gd5	Gd6	Gd7
1-1	0.020	−0.033	0.012	0.201	0.986	1.149	0.978
1-2	−0.041	−0.012	0.009	0.178	0.974	1.065	0.994
1-3	−0.093	0.054	−0.082	0.256	1.088	0.994	1.027
1-4	−0.065	−0.024	0.036	0.193	0.948	1.190	1.039
1-5	−0.057	−0.144	0.133	0.222	0.991	1.103	1.060
1-6	0.052	0.042	−0.071	0.084	1.028	1.152	0.999
1-7	0.033	−0.045	0.035	0.141	0.975	1.055	1.010
1-8	0.006	0.009	−0.092	0.143	1.117	1.163	1.034
2-1	0.019	1.221	2.132	2.030	0.013	0.004	0.042
2-2	0.016	1.153	2.218	2.012	0.020	−0.024	−0.002
2-3	0.034	1.090	2.676	1.504	0.017	−0.002	0.041
2-4	−0.128	1.292	2.200	2.368	−0.357	0.014	0.014
2-5	−0.051	1.229	2.187	1.949	0.290	0.117	0.121
2-6	−0.086	1.311	2.133	1.989	0.004	0.007	0.010
2-7	0.065	1.089	2.184	2.050	−0.064	0.026	0.004
2-8	0.012	1.229	2.089	2.052	−0.025	0.033	0.016
3-1	0.127	2.633	0.015	0.010	−0.030	0.071	0.002
3-2	0.127	2.633	−0.048	0.014	0.071	0.076	0.073
3-3	0.395	4.813	0.019	0.078	0.004	0.014	−0.086
3-4	0.386	4.822	0.071	0.069	0.033	0.022	0.228
3-5	0.393	4.815	0.015	0.087	−0.049	0.027	0.059
3-6	0.389	4.819	−0.007	0.013	0.043	−0.014	0.068
3-7	0.130	2.631	0.015	0.000	0.000	0.000	0.000
3-8	0.128	2.632	0.027	0.000	0.000	0.000	0.000

in Appendix A.

The data above reveal the following patterns:

(1)

1000 mL water injection: In Layer 3, the water saturation was S_w = 32.88%. Resistivity at most electrodes changed only slightly due to (a) low formation porosity (5.82%); (b) water remaining mainly within the 0~3 cm zone, while electrodes were positioned at 5 cm; (c) uneven pore distribution, making 1000 mL insufficient to noticeably affect formation resistivity. However, resistivity decreased significantly at electrodes 3-3 to 3-6 (fracture zone). Using Equation (3), the average G_d at fractures was 0.39%, compared with ~0.12% in non-fracture areas. Layer 2 had an average G_d =−0.015%, and Layer 1 had −0.018%, indicating no water infiltration and only minimal gas displacement.

(2)

2000 mL water injection: Layer 3 became fully saturated, and G_d increased significantly to 4.96% (2.63% in non-fracture areas). Compared with the unsaturated state, fracture-zone G_d was 3.25 times the non-fracture value; after saturation, this ratio dropped to 1.8. For Layer 2, S_w = 22.8%. Resistivity decreased by 14.65 Ω·m relative to full gas saturation. At electrodes 2-3 to 2-6 (faults zone), the average G_d = 1.23%; in non-fault areas, it was G_d = 1.17%. Layer 1 showed G_d = −0.019%, indicating slight gas displacement.

(3)

3000 mL water injection: Layer 3 resistivity decreased by only 0.13 Ω·m, which is negligible. For Layer 2, S_w = 63.6%, G_d = 2.30% at faults and 2.17% in non-fault areas. Layer 1 showed G_d = −0.002%.

(4)

4000 mL water injection: Layer 2 became fully saturated with an average G_d = 1.99%. For Layer 1, S_w = 3.48% and G_d = 0.177%. Layer 3 showed a negligible average G_d = 0.036%.

(5)

5000 mL water injection: Layer 3 had G_d = −0.008%, and Layer 2 had G_d = −0.012%, both negligible. Layer 1 reached S_w = 34.92%, with an average G_d = 1.013%.

(6)

6000 mL water injection: Layer 1 reached S_w = 67.10%, with an average G_d = 1.109%.

(7)

7000 mL water injection: Layer 1 reached S_w = 99.28%, with G_d = 1.017%, completing the water injection process.

(8)

Overall Trend: Maximum gas dissipation in Layer 3 occurred at electrode 3-4, and in Layer 2 at electrode 2-3, confirming that fractures or faults act as primary channels for gas-flow pathways. Layer 3 resistivity ranges from 499.64 Ω·m to 561.29 Ω·m, giving G_d = 5.652%. Layer 2 resistivity ranges from 544.53 Ω·m to 613.22 Ω·m, giving G_d = 5.767%. The maximum G_d in Layer 1 was 3.354%. Overall, faults exert the strongest influence on gas dissipation.

A summary of the S_w and G_d results for Tank 1 is provided in

Table 4. Results of the S_w and G_d for Tank 1 under different water injection volumes.

Tank 1 Water Injection (mL)	Layer 1		Layer 2			Layer 3
	Sw (%)	Gd (%)	Sw (%)	Gd (%)		Sw (%)	Gd (%)
		Non-Faulted		Non-Faulted	Faulted		Non-Faulted	Faulted
1000	-	−0.018	-	−0.015	-	32.88	0.12	0.39
2000	-	−0.019	21.03	1.17	1.23	100	2.63	4.96
3000	-	−0.002	62.27	2.17	2.30	100	2.63	4.96
4000	2.74	0.177	100	1.99	-	100	0.036	-
5000	34.92	1.013	100	1.99	-	100	0.008	-
6000	67.10	1.109	100	-	-	100	-	-
7000	99.28	1.017	100	-	-	100	-	-
Overall	100	3.354	100	5.767	-	100	5.652	-

. For Tank 2 (horizontal state), the dissipation rate calculated followed similar trends and similar basic behaviors. The main difference was that after the first 1000 mL of water injection, the dissipation rate at Layer 3 fractures in Tank 2 was relatively high (average 2.963). Although Layer 2-3 porosity was lower than Layer 1-3, the longer fracture extension in this zone further confirmed that fractures dominate gas dissipation. In Layers 2 and 1 of Tank 2, dissipation rates increased sharply during early injection (before saturation) while non-fracture zones changed only slightly. Other patterns are consistent and are not repeated here.

4.2. Analysis of Gas Dissipation Rate When Tank 2 Was Horizontal

When 1000 mL of water was injected into Tank 2, it could only enter the pure shale of Layer 3 (pore volume: 1441.28 mL, see

Table A3. Resistivity of Tank 2 under different water injection volumes (Ω·m).

No.	0	1000 mL	2000 mL	3000 mL	4000 mL	5000 mL	6000 mL
1-1	841.13	842.05	841.86	841.6	834.73	816.84	798.95
1-2	839.77	838.89	840.17	839.77	833.47	815.58	797.69
1-3	841.84	841.84	841.84	842.08	835.39	817.50	799.61
1-4	841.29	842.22	841.87	841.72	809.26	791.94	774.63
1-5	839.93	839.93	839.59	839.93	808.04	790.72	773.41
1-6	840.42	840.73	840.44	840.28	834.07	816.18	798.29
1-7	840.03	840.03	840.03	839.63	833.71	815.82	797.93
1-8	840.17	839.42	840.17	840.42	833.84	815.95	798.06
2-1	612.54	612.34	612.42	586.58	563.55	563.75	563.17
2-2	613.45	613.62	594.76	569.75	547.44	547.80	546.76
2-3	612.23	612.23	593.69	568.68	546.37	546.89	545.71
2-4	612.18	612.15	612.10	586.25	563.22	563.17	563.58
2-5	613.25	613.64	613.07	587.23	564.20	563.91	563.91
2-6	612.53	611.67	612.41	586.57	563.54	562.75	562.86
2-7	612.90	612.90	612.75	586.91	563.88	563.53	563.90
2-8	612.29	612.29	612.19	586.35	563.32	563.68	562.55
3-1	559.95	559.56	530.5	530.5	530.46	529.31	529.22
3-2	557.93	557.66	528.59	528.18	528.09	528.03	527.55
3-3	558.38	525.64	513.22	513.39	512.17	514.94	514.07
3-4	551.95	519.76	507.34	507.34	507.00	506.31	506.34
3-5	552.21	519.99	507.58	507.58	507.57	507.58	508.14
3-6	550.16	518.13	505.71	505.71	505.23	506.92	505.36
3-7	555.81	555.66	526.59	526.09	526.14	525.99	525.17
3-8	560.41	560	530.94	530.55	530.12	529.99	528.19

in Appendix A). Because faults and fractures were present between electrodes 3-2 and 3-6, water infiltrated these zones more readily, producing larger resistivity changes than other electrodes. With 2000 mL of injected water, 558.8 mL entered the silty shale of Layer 2, and Layer 3 became fully saturated. At 3000 mL, the water saturation of Layer 2 reached 63.6%; at 4000 mL, Layer 2 became fully saturated, and 107.84 mL of water entered the calcareous silty shale of Layer 1. Water injection stopped at 6960 mL when Layer 1 reached 99.12% saturation. Figure 5 shows resistivity changes throughout the water-driven gas displacement process.

The figure clearly shows gas dissipation during water injection: starting at the bottom, resistivity in each layer decreased as water saturation increased. Fractures acted as preferential fluid channels, causing more pronounced resistivity reductions. Significant decreases at electrodes 3-2 to 3-6 (Layer 3), 2-2 to 2-3 (Layer 2), and 1-4 to 1-5 (Layer 1) indicated that: (a) fractures served as the dominant flow pathway, and (b) the fractures themselves became filled with water. A slight increase in resistivity at some electrodes not yet reached by water could be attributed to gas migration into those areas, which increased local gas saturation.

4.3. Analysis of Gas Emission Rate When Tank 2 Was Sloped

The variation patterns of Tank 2 during gradual tilting are summarized as follows:

(1)

After injecting 1000 mL of water: Tilting mainly affected Layers 3 and 2 (see

Table 5. Gas dissipation rate of Layers 2-3 in Tank 2 at 1000 mL water (θ = 10°~65°) (%).

No.	Gd1	Gd2	Gd3	Gd4	Gd5	Gd6	Gd7
2-1	−0.0021	0.0021	4.0823	−0.0835	0.0376	0.0458	0.0622
2-2	0.0412	0.0353	4.0094	0.0232	−0.0961	0.0671	−0.0622
2-3	−0.0621	0.0620	4.0823	0.0134	0.0283	−0.0911	−3.5101
2-4	0.0348	0.0030	−0.0379	0.0718	−0.0009	−0.1138	0.6633
2-5	−0.0348	0.0372	−0.0024	0.0806	−0.0807	−0.0635	0.0635
2-6	0.0095	−0.0095	0.0000	0.0000	0.0463	0.0113	−0.0576
2-7	0.0000	0.0408	−0.0822	0.0023	−0.0254	0.0644	0.0000
2-8	0.0363	−0.0834	0.0945	−0.0322	−0.0752	−0.0050	0.0648
3-1	0.0310	−0.0310	0.0000	−0.0341	−0.0259	0.0600	−0.0500
3-2	−0.0672	0.1460	−0.0469	0.0280	−0.0600	0.0583	0.0278
3-3	0.0559	−0.0272	−0.0372	−0.0218	0.0785	−0.0483	−0.0932
3-4	0.1017	−0.0430	−0.0588	0.0581	−0.0715	0.0133	0.0000
3-5	0.0267	−0.0579	−0.0092	0.0404	−0.0837	−1.3181	0.0000
3-6	0.0748	−0.1104	0.0355	−0.0279	−1.5569	0.0571	−0.0336
3-7	−0.0356	−0.0524	0.1153	−1.2804	0.0710	−0.0753	−0.0456
3-8	0.0526	0.0345	−1.3088	0.0860	0.0196	0.0172	−0.0113

). When the dip angle θ reached 30°, 40°, 50°, 60°, and 65°, sudden changes in gas dissipation rate occurred at electrodes 3-8, 3-7, 3-6, and 3-5, respectively. However, these changes reflected reverse dissipation, gas-driven water displacement, with G_d values of −1.31, −1.28, −1.56, and −1.32. As the tilt angle increased, water in distal pores migrated toward proximal areas, and pores in the uplifted part of the formation became refilled with gas. At 30°, sudden increases in G_d (>4.0) were observed at electrodes 2-1, 2-2, and 2-3 in Layer 2, indicating that water seeping from Layer 3 had reached these three proximal electrodes. In Layer 1, dissipation rates remained very small from 0° to 65°, showing that water had not reached this layer, and resistivity changed minimally. This process also suggests that formation uplift is not entirely destructive: as water drains from some pores, those pores may be refilled by gas migrating upward from below.

(2)

2000 mL water injection: In Layer 3, a sudden change occurred at electrode 3-6 when θ = 30°, but no abrupt changes appeared at electrodes 3-7 or 3-8, due to fractures in this zone. At 40°, a sudden change occurred at electrode 3-8. In Layer 2, dissipation rates at electrodes 2-1 to 2-3 were very large, and the rate at electrode 2-4 was also significantly higher than at other electrodes.

(3)

3000 mL water injection: In Layer 3, clear reverse dissipation occurred at electrodes 3-8 and 3-7 when θ = 50°. Reverse dissipation also appeared at electrode 2-8 in Layer 2 because it was a distal electrode that was uplifted first, causing water to migrate toward proximal electrodes. Strong positive dissipation occurred at electrode 2-4 due to fractures in that area. Layer 1 showed no notable dissipation.

(4)

4000 mL water injection: After injecting 4000 mL, Layers 3 and 2 showed no significant gas dissipation changes during formation tilting. Obvious reverse dissipation occurred at electrode 1-8 in Layer 1 when θ = 60°; resistivity and fluid content at other electrodes changed very little (see

Table 6. Gas dissipation rate of Tank 2 at 4000 mL of water (θ = 10~65°) (%).

No.	Gd1	Gd2	Gd3	Gd4	Gd5	Gd6	Gd7
1-1	−0.0572	0.0215	0.0741	−0.0458	0.0220	−0.0148	0.0312
1-2	0.0337	0.0036	−0.0057	0.0047	−0.0744	0.0141	0.0607
1-3	−0.0488	0.0559	0.0131	0.0000	0.0212	−0.0307	−0.0917
1-4	0.0000	0.0004	−0.0607	0.0000	−0.0087	−0.0317	0.0409
1-5	−0.0109	0.0041	−0.0446	0.0359	−0.0125	−0.0614	0.1333
1-6	0.0000	0.0000	−0.0399	0.0110	0.0436	−0.0571	0.0988
1-7	0.0162	0.0000	−0.0162	−0.0361	0.0662	0.0051	−0.4557
1-8	0.0560	−0.0629	−0.0415	0.0000	0.0000	−0.5492	0.0660
2-1	0.0851	0.0000	0.0000	−0.0525	0.0317	0.0799	−0.0860
2-2	0.0851	0.0315	0.0000	−0.0583	−0.0341	0.0727	−0.0119
2-3	0.0851	−0.0771	0.0093	0.0301	0.0521	−0.0144	0.0000
2-4	0.0851	0.0000	−0.0048	0.0048	0.0296	−0.0803	0.0506
2-5	0.0851	0.0000	−0.0218	0.0436	0.0367	−0.1327	0.0742
2-6	0.0851	0.0283	0.0000	0.0362	−0.0440	0.0321	−0.0754
2-7	0.0851	0.0130	0.0000	−0.0732	0.0647	−0.0867	0.1681
2-8	0.0851	0.0666	−0.0127	0.0000	−0.0170	−0.0512	−0.0543
3-1	0.0563	0.0364	−0.0081	0.0000	0.0801	−0.0283	−0.0066
3-2	0.0563	−0.0286	0.0884	0.0321	−0.0435	0.0609	0.0000
3-3	0.0563	0.0278	−0.0717	0.0468	−0.0479	−0.0534	0.0398
3-4	0.0563	0.0924	−0.0653	0.0000	0.0518	0.0498	−0.0140
3-5	0.0563	0.0357	−0.0948	−0.0246	−0.0792	0.0114	−0.0419
3-6	0.0563	0.0000	0.0000	0.0000	0.1113	0.0118	0.0043
3-7	0.0563	0.0541	0.0000	−0.0667	0.0484	0.0183	−0.0092
3-8	0.0563	0.0753	0.0503	0.0016	−0.0683	0.0801	−0.1570

and Figure 6). This indicated that gas dissipation might reach an equilibrium under the combined influence of porosity, saturation, fractures, and dip angle. Once a layer became fully saturated, its gas dissipation rate did not change within a certain tilt range; significant changes appeared only when the tilt exceeded a threshold. In this simulation, the threshold angle was approximately 60°, referred to as the critical angle for gas dissipation (θ*).

(5)

5000 mL water injection: Layers 3 and 2 showed almost no changes during uplift, indicating complete water saturation. Resistivity in Layer 1 generally decreased, reflecting increasing water saturation. Proximal electrodes showed little change at low angles, mostly positive dissipation, meaning water had entered but responded weakly to uplift. At θ = 60°, reverse dissipation at electrode 1-8 intensified; at 65°, electrode 1-7 showed reverse dissipation, with other electrodes only slightly.

(6)

6000 mL water injection: Layers 3 and 2 remained nearly unchanged. Resistivity in Layer 1 continued to decrease with rising water saturation. Proximal electrodes showed almost no dissipation changes; the most notable change still occurred at electrode 1-8.

(7)

6960 mL water injection: Resistivity changes in all three layers were minimal. During tilting, dissipation rates remained nearly constant, except for electrode 1-8.

A summary of the S_w and G_d results for Tank 2 is provided in

Table 7. Results of the S_w and G_d for Tank 2 under different water injection volumes.

Tank 1 Water Injection (in mL)	Layer 1		Layer 2			Layer 3
	Sw (%)	Gd (%)	Sw (%)	Gd (%)		Sw (%)	Gd (%)
		Non-Faulted		Non-Faulted	Faulted		Non-Faulted	Faulted
1000	-	0.044	-	0.663	−3.510	30.62	0.017	−1.318
2000	-	−0.063	22.8	0.663	−3.510	100	−0.075	−0.069
3000	-	0.674	63.6	−1.154	0.032	100	0.002	0.015
4000	3.48	−0.456	100	0.168	−0.012	100	−0.157	0.0398
5000	35.79	−0.805	100	0.047	0.125	100	0.009	0.089
6000	68.10	0.052	100	0.007	0.171	100	0.005	0.001
7000	99.12	−1.418	100	0.001	0.0-	100	0.002	0.001
Overall	100		100		-	100	5.652	-

Note: G_d values in this table are at formation dip 60°.

.

4.4. Validation of the Simulation

As shown in Figure 2, Wells X1, X2, and X3 are the three wells currently in production. Wells X1 and X3 are located on the left limb of the syncline. The target formation of Well X1 is the Ordovician Baota Formation, whereas Wells X2 and X3 target the Silurian Longmaxi Formation. The dip angle of the Longmaxi Formation is approximately 40.6° at Well X1, 29.5° at Well X3, and 58.8° at Well X2. Mud-logging data report total hydrocarbon ranges of 0.253–1.451 for Well X1, 0.834–85.930 for Well X2, and 0.019–0.246 for Well X3. To date, the cumulative gas production is 1.09 million cubic meters for Well X1, 4.7122 million cubic meters for Well X2, and 88,000 cubic meters for Well X3. The results of this simulation are fully consistent with this trend.

5. Discussion

This study successfully simulated gas escape from the Anchang anticline and clarified the main escape patterns within the anticline strata. However, the following limitations should be noted:

(1) Although the porosity, permeability, and density of the simulated strata are similar to those of the actual formation, the pore structure, capillary forces, and bound water characteristics may not fully match in situ conditions.

(2) A large-scale physical simulation device was used to investigate the overall gas escape behavior during development. No confining pressure was applied because increasing pressure to true formation levels could alter the pore structure, porosity, permeability, and fault conditions of the simulated strata. For this reason, a lower-pressure environment was adopted.

6. Conclusions

This study examines the design of a water-driven gas displacement simulation device for synclinal shale reservoirs and analyzes the gas dissipation behavior in low-porosity, low-permeability formations. The important findings are as follows:

(1)

Water-driven displacement: When the gas dissipation rate G_d > 0, gas is displaced by water. During the uplift, some pores exhibit reverse dissipation escape, with (G_d < 0), indicating water retreat and gas invasion. This shows that uplift does not entirely reduce gas content; the gas from lower parts of the formation can migrate into the uplifted zone.

(2)

Effect of dip angle: In an up-dip formation, G_d increases with the dip angle θ. When θ reaches a critical angle (θ*), G_d changes abruptly and reaches a maximum, after which it stabilizes or decreases. In this simulation, the critical angle is approximately 60°. Faults and fractures further enhance gas dissipation.

(3)

Role of fractures and faults: Fractures and faults form more effective gas loss pathways than pores. In this study, the gas dissipation rate in fractured or faulted zones is about three times higher than in purely porous formations.

In conclusion, the gas escape rate is controlled by multiple factors, including formation dip angle, faults, pore–fracture structures, and development practices, and does not increase monotonically with tilt angle. The simulation results clarify the fundamental gas-escape behavior in anticlinal formations and provide a scientific basis for the efficient and economical development of shale gas in the Anchang anticline.