Feasibility Evaluation of I–Shaped Horizontal Salt Cavern for Underground Natural Gas Storage

Abstract

Underground salt cavern gas storage has been widely applied due to its numerous advantages. Most of China’s salt resources are derived from lacustrine deposits. As high–quality resources in the central sedimentary area are gradually exploited, exploring the utilization of thin salt layers at the edges of sedimentary centers is the future development trend. However, the use of thin salt layers faces challenges such as low resource utilization, small cavern volumes, and poor economic feasibility, which limit its engineering applications. Therefore, a comprehensive assessment of constructing gas storage in thin salt layers is necessary. This paper first analyzes the necessity of building gas storage in thin salt layers and surveys cavern construction methods and their applicability. Based on geological seismic data, the feasibility of constructing gas storage in the Pingdingshan thin salt layer is proposed. A novel I–shaped cavern design is introduced, which, according to engineering economic evaluations, reduces investment by 9.6% compared to traditional single–well vertical cavern construction methods. Finally, rock mechanics tests were conducted to study the impact of mudstone interlayers and cyclic operation modes on the stability of the I–shaped cavern under three different injection and production conditions. The analysis shows that multi–cycle injection and production can effectively suppress cavern shrinkage and the development of the rock–relative plastic zone. The safety factor (SF) for different conditions is greater than 1, indicating that the I–shaped cavern has good stability and can adapt to various operational conditions. This study provides valuable insights into the geological conditions and rock mechanics characteristics for the future construction of gas storage in thin salt layers in China.

1. Introduction

As is well known, natural gas can effectively reduce the emissions of SO₂ and NO_x compared to traditional energy sources. China has significantly increased its natural gas usage, with consumption nearing 4000 × 10⁸ m³ in 2023. To ensure a stable and reliable gas supply, it is essential to develop a large number of storage facilities [1,2]. Currently, the storage capacity has reached 381 × 10⁸ m³, accounting for only 9.7% of consumption, which is much lower than the international standard of 15–18% [3]. Therefore, China faces a serious shortfall in its natural gas supply, particularly in winter or during emergencies, highlighting the urgent need to accelerate the development of the storage capacity [4].

UGS can be divided into depleted gas–oil reservoirs, aquifers, and underground salt caverns [5]. Among these, salt caverns are favored due to several key advantages: (i) Rock salt has extremely low permeability and porosity, along with excellent plasticity and self–healing properties. (ii) Rock salt (NaCl > 90%) is water–soluble, which facilitates the easy and cost–effective creation of large caverns. (iii) Salt cavern injection and withdrawal rates are the highest [6], resulting in a minimal cushion gas stock [7]. And (iv) salt deposits are widely distributed across the world. Consequently, underground salt caverns are extensively utilized for oil and natural gas storage [8,9] and gradually applied to compressed air energy storage [10,11] and even hydrogen storage [12,13,14]. China is rich in salt mineral resources, with approximately 21 large salt mines, all of which belong to lacustrine deposits [15]. The thickness of the salt layers in the sedimentary centers can reach 100–600 m, as shown in Figure 1, which can be used for large–scale underground energy storage.

UGS and compressed air energy storage stations have been successively deployed in most of China’s medium–thickness salt layers. However, areas with thinner salt layers have not been effectively utilized, as shown in Figure 2, necessitating the further development of methods for constructing gas storage in thin salt layers. Research on the construction of gas storage in thin salt layers has primarily focused on vertical caverns, addressing issues related to the stability, safety, and sealing of caverns within layered salt rock. Most studies have utilized laboratory experiments and numerical computation methods. For example, Zhang et al. studied the mechanisms of roof collapse and leakage based on the physical properties of salt and non–salt layers [16]. Zhang et al. revealed the failure mechanisms of the rock surrounding caverns containing different types of interfaces. Xiong et al. revealed gas seepage surrounding gas storage salt caverns [17]. Liu et al. proposed a horizontal cavern design scheme involving dual vertical wells and suggested a reasonable operating pressure [6]. Jiang et al. optimized the horizontal spacing between the dual vertical wells and conducted a stability assessment [18]. Jing et al. and Cheng et al. forecasted the evolution of ground subsidence above the gas storage cavern [19,20].

These studies provide important references for the safe operation in thin salt rock layers. However, creating caverns in thin salt layers quickly and economically is challenging and has rarely been reported. Based on the construction experience of the Jintan gas storage, an effective volume of 20 × 10⁴ m³ can be achieved using a single–well–vertical cavern, requiring a continuous salt rock layer thickness of 160–170 m, with an insoluble content not exceeding 20% [21,22,23]. The thin salt layer makes it impractical to directly apply the existing single–well–vertical cavern construction method used in Jintan. Considering China’s significant demand for natural gas storage and the requirements for balancing renewable energy storage facilities [24], constructing horizontal caverns for energy storage is a promising option [25].

Researching storage technologies for thinly bedded salt rock can expand the range of salt cavern UGS and enhance the overall utilization of salt mines. To provide reliable evaluations of thinly bedded salt rock in China, this paper focuses on the thinly bedded salt rock area in southern Pingdingshan. It conducts the following studies: (i) An applicability analysis of various construction methods under thinly bedded salt rock conditions. (ii) Rock mechanics testing to support stability evaluation. And (iii) the proposal of a novel I–shaped cavern design, analyzing the feasibility of constructing UGS in thinly bedded salt rock from geological, cavern construction, economic, and stability evaluation. This research not only provides a comprehensive evaluation of the feasibility of constructing storage facilities in thin salt layers, but also offers important information for the development of solution–mined caverns.

2. Construction Methods and Applicability

2.1. Cavern Construction Method

Salt cavern gas storage is constructed using the water–soluble cavern formation method, which can be categorized into various approaches based on the configuration of the tubular string. This study surveys and compares different cavern construction methods used globally.

2.1.1. Single–Well–Vertical (SWV) Cavern

The SWV cavern method is the primary cavern construction technique used around the world. This process involves vertical drilling from the surface to reach the target salt layer. Production casing, outer tubing, and inner tubing are sequentially lowered into the borehole. Then, freshwater or low–concentration brine is injected into the cavern and the brine formed by dissolving salt rock is discharged to the surface, as shown in Figure 3.

Based on a wellbore structure, cavern construction methods can be divided into conventional borehole caverns and large borehole caverns. Conventional borehole caverns utilize 9 5/8” production casing, with a typical brine injection rate of 100 m³/h. In contrast, large borehole caverns utilize 13 3/8” production casing, allowing for brine injection rates of up to 300 m³/h, which offers advantages such as high flow rates and lower friction energy consumption.

The main advantages include the easier control of the cavern shape and the mature process, and the cavern can be measured by sonar. However, there are drawbacks, such as slower cavern construction speeds; for instance, constructing a 20 × 10⁴ m³ cavern at the Jintan salt cavern gas storage takes about 4 to 5 years. Additionally, it can be challenging to form large–volume storage in salt formations with a high interbedded content and thin layers.

2.1.2. Two–Well–Directional (TWD) Cavern

China’s salt mines are characterized by multi–layered, low–grade geological formations. During cavern construction, insoluble materials undergo fragmentation and expansion, leading to significant debris accumulation in the cavern’s bottom space, which ultimately results in a limited upper cavern volume. The TWD cavern is a technique for expanding gas storage by utilizing voids in the sediment. This approach consists of one vertical well and one directional well. During the cavern construction phase, a SWV cavern method is utilized, and during the gas injection and brine discharge, the directional well facilitates brine removal from the sediment bottom. This enables the efficient utilization of the sediment voids at the cavern’s bottom. However, it does not increase the usable underground volume compared to the SWV cavern method, as shown in Figure 4. This cavern construction method has undergone preliminary field trials in the Shandong and Hubei provinces in China, pending the verification of directional well drilling, cementing, and completion techniques in the sediment.

2.1.3. Small–Spacing Two–Well (SSTW) Cavern

The SSTW cavern method involves drilling two wells into the target salt layer at a close distance, ranging from 6–30 m apart. After the two wells are connected, fresh water is injected into one well while brine is extracted from the other to dissolve the salt layer, as shown in Figure 5. This technology has been implemented in the US Strategic Petroleum Reserve (SPR) in the Gulf of Mexico, USA, and in the Manosque of France. In Manosque, the salt layer reaches a thickness of 800 m, and the caverns constructed using SSTW caverns are cylindrical in shape, featuring a significant height and a total volume of approximately 50 × 10⁴ m³.

The main advantages of the SSTW cavern technology are (i) the low energy consumption and significant water injection rate, reaching up to 300 m³/h, which notably enhances the cavern formation speed; (ii) creating larger caverns, increasing the storage volume, and improving the efficiency of salt layer utilization; (iii) reducing water hammer and vibration phenomena; and (iv) increasing the gas injection and production rates. However, this cavern construction process is still underdeveloped, particularly regarding the determination of well spacing, optimal flow rates, and cavern shape control, which require further practical exploration.

2.1.4. Two–Well–Horizontal Saddle–Shaped (TWHS) Cavern

The TWHS cavern involves drilling two wells from the surface to reach the salt layer, with a distance of approximately 150–300 m between them. One well is vertical and the other is horizontal open–hole completion, allowing for a connection between the two wells to create a horizontal exposed borehole. Production casings are then installed in both wells, and an alternating water injection is used to dissolve the salt rock and discharge brine, as shown in Figure 6.

Since the construction of the TWHS cavern at the Xiangheng Salt Mine in Hunan in 1991, it has been in use in domestic salt mines for nearly 30 years. It has been widely applied to salt deposits with burial depths of 300–4000 m, salt layer thicknesses of 1.5–200 m, and NaCl grades of 40% to 90%.

The main advantages of the TWHS cavern are (i) high cavern formation efficiency and high brine concentration, with maximum discharge rates reaching 300 m³/h. (ii) This will allow for the more effective utilization of salt resources, which is well–suited to the characteristics of multi–layered salt rock formations in China.

However, there are notable drawbacks: (i) The absence of dissolution inhibitors makes it difficult to control the cavern morphology during formation, leading to a saddle–shaped cavern and potential roof collapse due to roof penetration. (ii) Salt rock near the horizontal channel may not be adequately dissolved, especially if the distance between the wells is too large, resulting in significant unrecoverable salt rock. (iii) Sonar measurement technology faces challenges in accurately assessing the morphology of horizontal well caverns, and the evaluation criteria for horizontal well caverns as gas storage remains insufficiently developed [26].

2.1.5. Two–Well Retreating Horizontal (TWRH) Cavern

The TWHS cavern formation technology for horizontal wells has evolved from the TWHS cavern. As illustrated in Figure 7, two wells are drilled from the surface to reach the salt strata, with a distance of 250–450 m between them. One well is vertical, while the other is horizontal. Using directional drilling into the target salt layer, the drilling trajectories are connected to form a horizontal exposed channel. Subsequently, the casing is lowered into both wells. During cavern formation, the water injection point is gradually retracted from the vicinity of the vertical well toward the inclined well, allowing for the phased dissolution of the salt layer and ultimately creating a tunnel–shaped cavern [27].

From 1964 to 1965, a horizontal cavern with a volume of 20 × 10⁴ m³ was constructed in Russia’s Podzemgasprom. During the cavern formation, the prolonged contact time between freshwater and salt rock allows for high brine concentrations early in the operation. However, it is challenging to maintain a uniform cross–sectional shape along the entire horizontal segment, and incomplete brine discharge during a gas injection leads to the significant volume loss of the cavern. In 1993, France began cavern formation experiments in a 25 m thick salt layer at the Mulhouse salt mine, resulting in a small cavern with an approximate volume of 750 m³ [28,29,30].

The main advantages of this method are (i) the high utilization of salt layers, allowing for the construction of large caverns within limited–thickness salt layers, thus maximizing the use of thin salt formations for storage; and (ii) it enables the avoidance of thick interlayers by selecting pure salt sections between layers for cavern formation. Many salt mines in China feature complex interlayers, and crossing thick interlayers during cavern formation can lead to difficulties in controlling the cavern and the potential collapse of the interlayer, which may damage the drilling column. Avoiding thick interlayers can effectively increase the cavern formation rate.

2.2. Cavern Construction Economic and Technological Comparison

Currently, in China, the development of salt cavern UGS is gradually shifting towards thin salt layers and complex low–grade salt mines. The dual–well single–cavern technology offers advantages such as reduced friction, diversified cavern formation methods, and convenience for a gas injection and brine discharge without applying pressure to the wells, making it a trend for future storage development. However, as cavern formation becomes more complex, drilling and completion costs are rising, necessitating a comprehensive evaluation of its economic feasibility. Under geological conditions of thin salt layers, various cavern formation methods are illustrated in Figure 8. The advantages and disadvantages of each method are summarized in

Table 1. Comparison of cavern formation method for thin–layer salt mines.

Method	Advantages	Disadvantages	Applicability	Drilling and Completion Cost
SWV cavern	(i) Mature technology and wide applicability	(i) Slow cavern formation rate (ii) High tubing friction	Wide applicability	CNY 14 million
TWD cavern	(i) Sediment voids utilized (ii) Increased gas storage capacity	(i) Risk of directional good blockage	Low–grade salt rock with a thick salt layer	CNY 28 million
SSTW cavern	(i) Low cavern formation cost (ii) Fast cavern formation rate (iii) High salt utilization rate	(i) Immature technology	The thickness of the salt layer is relatively large	CNY 24 million
TWHS cavern	(i) Low cavern formation cost (ii) Fast cavern formation rate (iii) High salt utilization rate	(i) Risk for roof collapse (ii) Difficult to control cavern shape	Big, thick salt layer	CNY 42 million
TWRH cavern	(i) Low cavern formation cost (ii) Fast cavern formation rate (iii) High salt utilization rate	(i) Difficult to control cavern shape	Thin salt layer without thick interlayer	CNY 32 million

. After comparison, it is believed that the methods SWV cavern and TWRH cavern are suitable for constructing a thin salt layer for gas storage.

2.3. The Experiment and Evaluation Method

Experimental and evaluation methods are important ways to analyze the feasibility of constructing thin–salt–layer UGS. This article proposes a comprehensive feasibility evaluation method based on experiments, feasibility evaluation theory, and numerical simulations.

2.3.1. The Experiment Site

The Pingdingshan area is rich in salt mineral resources and the construction of gas storage facilities began at the center of salt mineral sedimentation. According to the plan of PipeChina to build large storage facilities and stations in the Pingdingshan area, this article chose Pingdingshan City in the Henan Province as the experimental site. The reason for choosing this is that there is a large demand for natural gas consumption in central and southern China, and the Henan Province has good pipeline trunk lines, as shown in Figure 9. At the same time, PipeChina conducted a lot of preliminary work in the Pingdingshan area, so we chose the salt rock in that area to conduct mechanical tests.

2.3.2. The Experiment Methods

To ensure the feasibility of constructing UGS in thin salt layers, an analysis was conducted, considering factors such as rock mechanics tests, geological conditions, and cavern formation methods. As shown in Figure 10, rock mechanics tests were first performed to obtain key mechanical parameters. Subsequently, geological, cavern, economic, and stability models were analyzed.

In the geological model, various geological parameters were selected based on the geological information from Pingdingshan to establish the model. The cavern model optimized suitable methods for constructing storage in thin salt layers, proposing a novel I–shaped cavern based on the current cavern formation practices. The economic evaluation model assessed the feasibility of different cavern formation methods by considering drilling, completion, and cavern formation costs. In the stability evaluation model, a three–dimensional geological model was established to analyze the stability under various operating conditions. Through the analysis of these models, a comprehensive evaluation of the feasibility of constructing storage in thin salt layers was achieved.

3. Salt Rock Laboratory Tests

Rock laboratory testing is a fundamental condition for evaluating the feasibility of gas storage stability and an important way to obtain key mechanical parameters. This article introduces four types of rocks that have been tested in the laboratory to obtain the tensile strength, uniaxial compressive strength, elastic modulus, Poisson’s ratio, shear force, cohesion, internal friction angle, and creep parameters. We conducted the following rock mechanics tests at room temperature using the MTS815 rock mechanics testing machine, which had a maximum axial loading capacity of 2800 kN and a maximum confining pressure of 80 MPa.

3.1. Uniaxial Compression Tests

Uniaxial compression tests are conducted to obtain the uniaxial compression strength, elastic modulus, and Poisson’s ratio of rock samples. The five tests of the salt rock and one test of mudstone are conducted, as shown in Figure 11. The uniaxial compression strength, elastic modulus, and Poisson’s ratio are calculated by the following equation [31].

$${\sigma }_C={\displaystyle \frac{P_{max}}{A}}$$

$$E={\displaystyle \frac{∆\sigma }{\Delta {\epsilon }_1}}$$

$$v={\displaystyle \frac{\mathsf{\Delta }{\epsilon }_3}{\mathsf{\Delta }{\epsilon }_1}}$$

where ${\sigma }_c$ represents the uniaxial compressive strength (MPa), $P_{max}$ denotes the peak load (kN), $A$ is the cross–sectional area of the sample (m²), $E$ indicates the elastic modulus (GPa), $\nu$ is Poisson’s ratio, and $\mathsf{\Delta }{\epsilon }_1$ and $\Delta {\epsilon }_3$ represent the strain differences in the axial and circumferential directions corresponding to the elastic deformation section of the stress–strain curve.

The parameters obtained from the tests are shown in

Table 2. Uniaxial compression test results of rock samples.

Lithology	Specimen ID	Uniaxial Compression Strength (MPa)	Elastic Modulus (GPa)	Poisson’s Ratio
Salt rock	U–2–12	23.35	8.49	0.21
	U–3–14	25.22	5.9	0.23
	U–3–13	23.64	6.98	0.28
	U–3–12	15.56	4.36	0.31
	U–3–15	22.57	8.19	0.25
Mudstone	U–4–1	25.52	11.2	0.24

. The elastic modulus of mudstone is higher than that of salt rock, indicating that mudstone serves as a hard interlayer. Additionally, Poisson’s ratio of salt rock exceeds that of mudstone, suggesting that under the same stress conditions, both axial and circumferential strains in salt rock are greater than those in mudstone. This superior deformation behavior of salt rock is beneficial for the construction of UGS.

3.2. Triaxial Compression Tests

Under confining the pressures of 15 MPa, 20 MPa, and 30 MPa, triaxial compression tests were conducted to obtain the triaxial compressive strength, cohesion, and internal friction angle of rock samples. Five sets of salt rock and one set of mudstone were tested, with the triaxial compressive strength, elastic modulus, and Poisson’s ratio derived from the following equation:

$${\sigma }_1=m{\sigma }_3+b$$

$$m={\displaystyle \frac{1+\sin \phi }{1-\sin \phi }}$$

$$b={\displaystyle \frac{2c\cos \phi }{1-\sin \phi }}$$

where ${\sigma }_1$ represents the ultimate axial stress (MPa), ${\sigma }_3$ denotes the confining pressure (MPa), $C$ indicates cohesion (MPa), and $\phi$ is the internal friction angle. Using ${\sigma }_1$ and ${\sigma }_3$, Mohr’s limit stress circle is plotted on a coordinate system, and the strength coefficient under triaxial stress conditions is determined based on Coulomb–Mohr strength theory.

Table 3. Three–axis compression test results of rock samples.

Lithology	Specimen ID	Confining Pressure (MPa)	Axial Compression (MPa)			Cohesive Force (MPa)	Internal Friction Angle
			Maximum	Strain 1%	Strain 2%
Salt rock	T–2–11	15	74.58	25.851	−15.395	5.77	34
	T–2–12	15	45.24	2.427	−0.902
	T–2–13	20	58.47	4.832	−1.8
	T–2–14	20	47.1	2.766	−2.028
	T–2–16	30	73.99	9.104	−4.596
Mudstone	T–3–1	15	64.039	5.017	−4.804	7.74	26.5
	T–3–2	20	59.975	2.5	−2.92
	T–3–11	30	89.212	2.417	−0.539
	T–4–1	15	55.044	3.48	−1.462
	T–4–2	20	57.345	3.604	−3.267
	T–5–1	30	78.886	2.653	−0.387

lists the strength, axial strain, and radial strain of salt rock and mudstone under peak conditions, along with the cohesion and internal friction angle for dark gray salt rock. The cohesion of salt rock is 5.77 MPa, and its internal friction angle is 34°. The cohesion of mudstone is 7.74 MPa with an internal friction angle of 26.5°.

3.3. Brazilian Disk Splitting Test

To obtain the tensile strength of the rock samples, the Brazilian disk splitting tests were conducted. The tests were performed under displacement control with a rate of 0.002 mm/s. The tensile strength is calculated by the following equation [32]:

$${\sigma }_t=-{\displaystyle \frac{2P}{\pi GL}}$$

where $P$ represents the fracture load of the rock sample at the point of splitting (kN), and $G$ and $L$ are the diameter and length of the rock sample (mm), respectively. ${\sigma }_t$ denotes the tensile strength of the rock sample (MPa). A total of 21 splitting tests were conducted on two types of rock samples: 9 tests on salt rock samples and 12 tests on mudstone samples, as shown in

Table 4. Brazilian disk splitting test results of rock samples.

Lithology	Specimen ID	Breaking Load (kN)	Tensile Strength (MPa)	Average Value (MPa)
Salt rock	B–2–11	2.54	1.35	1.34
	B–2–12	2.59	1.33
	B–2–13	3.17	1.582
	B–2–21	2.14	1.23
	B–2–22	2.48	1.4
	B–2–23	2.8	1.525
	B–2–31	3.17	1.596
	B–2–32	2.27	1.119
	B–2–33	1.74	0.892
Mudstone	B–3–16	10.93	5.579	3.22
	B–3–17	3.62	1.863
	B–3–18	7.34	3.83
	B–3–19	11.78	6.453
	B–3–21	4.89	2.6
	B–3–31	6.1	3.154
	B–3–32	5.86	2.918
	B–3–33	4.38	2.311
	B–3–34	5.52	2.861
	B–3–35	6.21	3.283
	B–3–36	3.62	1.877
	B–3–37	3.7	1.97

. The average tensile strengths for mudstone and salt rock samples were 1.34 MPa and 3.22 MPa, respectively. After the test shown in Figure 12, an open tensile crack appeared near the center of the sample surface.

3.4. Triaxial Compression Creep Tests

The creep behavior of salt rock is critical to the long–term stability of salt cavern underground gas storage facilities. The creep test plays a key role in analyzing the mechanics of salt caverns. Many constitutive relationships have been used to describe the rheology of salt rock. Detailed experiments were conducted and the final results were adopted, as shown in

Table 5. Triaxial compression creep test results of rock samples.

Lithology	Specimen ID	Confining Pressure (MPa)	Deviatoric Stress (MPa)	Steady–State Creep Rate (h−1)	Time (h)
Salt rock	C–1–2	30	15	6.0 × 10−5	67
		30	20	1.3 × 10−4	98
		30	25	2.4 × 10−4	96
	C–1–3	30	15	5.0 × 10−5	21
		30	20	1.0 × 10−4	25
		30	25	4.0 × 10−4	57
	C–1–4	10	30	1.87 × 10−4	48
		10	45	7.19 × 10−4	510
	C–1–5	10	55	5.68 × 10−5	35
	C–1–6	10	15	3.29 × 10−5	48
		10	20	6.51 × 10−5	48
		10	25	2.95 × 10−4	110
		10	30	6.72 × 10−4	75
		10	40	2.4 × 10−3	30
Mudstone	C–2–2	20	10	1.02 × 10−6	56
			20	2.87 × 10−6	64
			30	5.9 × 10−6	60

.

The parameters in the Norton power law are calculated by the equation [33]:

$$\epsilon =A{\left(\overline{\sigma }\right)}^n$$

where ε represents the steady–state creep rate; $\overline{\sigma }$ represents the deviatoric stress ($\overline{\sigma }={\sigma }_1-{\sigma }_3$); ${\sigma }_1$ and ${\sigma }_3$ represent the axial stress and confining pressure, respectively; and A and n represent the creep parameters; the creep parameters were calculated. The average creep parameters for salt rock were A = 1.02 × 10⁻⁷, n = 3, and for mudstone, A = 2.40 × 10⁻⁸, n = 1.7. These parameters can be used in the stability analysis.

4. Feasibility Analysis of Construction

4.1. Geological Analysis

The selection of geological judgments is a fundamental condition for the construction of gas storage, as different locations have different geological characteristics. First, we need to determine the geological parameters based on the geological features of the experiment site. This paper determines the geological parameters based on the 3D seismic exploration and well data collected by PipeChina in the Pingdingshan area.

The Pingdingshan salt mine is located in the flatlands of Yexian and Wuyang County, within the Wuyang depression, which is part of the western area of the Zhoukou sag in the North China basin. It exhibits a structural pattern characterized as “deeper in the north and shallower in the south” with “two depressions and one uplift.” The salt–bearing strata belong to the Paleogene Eocene Niuhuangyuan Formation, specifically the first and second members, with the primary salt layer found in the first member. The lithology mainly consists of salt rock interbedded with gypsum–rich mudstone and mudstone.

The surface above the sedimentation center of the Pingdingshan salt mine is urbanized, making it unsuitable for constructing salt cavern UGS. However, thicker salt layers surrounding the urban area have already been designated for storage, as shown in Figure 2. In response to the national strategy for large gas storage, three–dimensional seismic inversion data were analyzed for the southern experimental area. The seismic data indicate that the salt layer thickness for groups 1–7 is 55–75 m, for groups 8–13 is 30–50 m, and for groups 14–20 is 90–120 m. Groups 14–20 were selected as the optimal thin salt layer for the construction of the gas storage. The top of the storage layer is approximately 1310 m deep, with an average salt layer thickness of about 120 m. Based on the seismic inversion data and well–log data from the surrounding areas, the basic parameters for the construction of the gas storage are shown in

Table 6. Basic geological parameters.

Geologic Parameter	Value
Top depth of salt cavern construction salt layer (m)	1290–1410
Salt layer thickness (m)	120
Depth of casing shoe (m)	1300
Depth of cavern top (m)	1315
Comprehensive insoluble content ($c$)(%)	30

, and the stratigraphic model is shown in Figure 13.

4.2. Cavern Construction

The horizontal cavern is a suitable gas storage solution for thin salt layers, such as the TWRH cavern. Its horizontal formation allows for a larger cavern volume within the layered salt rock, resulting in a higher utilization of the salt layers and better adaptability for constructing gas storage in thin salt formations. However, the geological conditions of thinly bedded salt rock in China differ significantly from those in foreign TWRH cavern construction projects. Due to the inability to move the cavern formation tubing like in the vertical cavern formation process, it is impossible to control the collapse of the interlayer. Foreign construction is carried out in high–purity salt domes, and there are no cases of TWRH cavern construction through thick interlayers, and the horizontal cavern layer should avoid these.

This paper proposes a gas storage solution suitable for thinly bedded salt rock, designed as a combination of one horizontal cavern and two vertical caverns using two vertical wells and one horizontal well, forming an I–shaped cavern structure. The cavern construction can be divided into three stages, as shown in Figure 14. Based on the experience of the Jintan gas storage cavern project, the radius of the dissolution cavern is typically no more than 40 m [34,35]. Considering that the salt layer is relatively thin, the dissolution cavern radius is set to 35 m. Based on the surface conditions in the deep salt layer block of Pingdingshan, a horizontal cavern length of 400 m is recommended. According to the seismic data on the thickness of the salt rock layers in the target storage area, the height of the horizontal cavern should not exceed 42 m, determining the dimensions of the horizontal cavern. Due to the presence of interbedded mudstone and salt rock with a thickness of no less than 30 m at both the top and bottom of the construction salt layer, the thickness of the top and bottom protective shelf complies with the recommended SMRI guidelines. The cavern size parameters are shown in

Table 7. Investment in different cavern formations for thinly bedded salt rock.

	SWV Cavern	TWRH Cavern	I–Shaped Cavern
Neck height (m)	15	15	15
Vertical cavern radius (m)	35	0	35
Vertical cavern height (m)	80	0	80
Horizontal cavern height (m)	/	42	42
Horizontal cavern length (m)	/	400	400
Total volume ($V$) (104 m3)	20	48	69.5
Effective volume ($V_{eff}$) (104 m3)	10.4	24.9	36.14
Drilling and completion investment	CNY 14 Million	CNY 32 Million	CNY 40 Million
Cavern formation investment	CNY 12 Million	CNY 30 Million	CNY 41.7 Million
Total investment cost per m3	CNY 250	CNY 243	CNY 226

, and different types of caverns are illustrated in Figure 15.

The total volume ($V$) of the salt cavern includes both soluble salts and insoluble materials. The effective volume refers to the space formed in the salt cavern after considering the accumulation of insoluble residue and expansion. Due to the high content of insoluble materials in the mining area, it is necessary to account for the accumulation and expansion of insoluble sludge at the bottom of the cavern. Based on the experience from the Jintan gas storage project, and considering the expansion coefficient of the insoluble sludge in the saturated brine as $u=1.6,$ the effective volume $V_{eff}$ is given by the formula $V_{eff}=V(1-uc)$, where $C$ is the comprehensive insoluble content [36].

4.3. Economic Evaluation

The economic analysis compares construction costs in the Henan region with underground engineering investment expenses, including drilling, completion, cavern excavation water, electricity, and downhole operation expenses. Drilling and completion investments are referenced from the “Market–Oriented Quotas for Engineering Technical Services in Huabei China Oilfield (2023 Edition),” while cavern formation costs are based on the calculations for the Jintan salt cavern gas storage project. Investment calculations are detailed in Table 7. Diagrams of the SWV cavern, TWRH cavern, and I–shaped cavern are shown in Figure 15. After optimization, the investment for the horizontal cavern construction is reduced by 9.6% compared to the SWV cavern method and by 7.3% compared to the TWRH cavern, demonstrating significant economic benefits.

4.4. Analysis of Cavern Stability

To analyze the stability of the cavern, a 3D geological model was established using FLAC^3D software based on the geological conditions of Pingdingshan, and mechanical calculations were performed. Four evaluation criteria were proposed: the displacement, plastic zone, volume shrinkage rate, and safety factor. Additionally, the stability of the I–shaped cavern was analyzed under various injection and extraction operating conditions over a 30–year period [37,38,39].

4.4.1. Three–Dimensional Stability Analysis Model

The overall stratigraphic calculation model measures 1000 m × 250 m × 510 m. The actual burial depth of the model’s top layer is 1100 m. Vertical simply supported constraints are applied at the bottom surface of the model, while simply supported constraints in the normal direction are applied at the four side surfaces to prevent normal displacement, as shown in Figure 16.

4.4.2. Calculation Parameters

This elastoplastic calculation employs the classical Mohr–Coulomb strength theory, applicable to both plastic rocks and brittle salt rock shear failures. Assuming the compressive stress is negative, the yield function for the Mohr–Coulomb strength condition is calculated using the following equation:

$$f_s={\sigma }_1-{\sigma }_3{N}_{\varphi }+2c\sqrt{N_{\varnothing }}$$

$$N_{\varnothing }={\displaystyle \frac{1+\sin \varnothing }{1-\sin \varnothing }}$$

Rock salt is a typical visco–plastic material, and its creep is very apparent once deviatoric stress exists. Step–creep experiments were performed to investigate the creep behavior by using the rock salt sample from a pilot well in the Pingdingshan Salt Mine. The transient creep of rock salt is completed in a short period of time, and constructing a cavern takes about 2–3 years, so transient creep can be ignored. Therefore, during operation, we only consider the steady–state creep state of salt cavern energy storage. The Norton–Power model is used to describe the stable creep behavior of rock salt as follows.

$$\epsilon \left(t\right)=A{q}^n$$

$$\epsilon \left(t\right)=D\exp \left(-{\displaystyle \frac{Q}{RT}}\right){\sigma }^m$$

$$q=\sqrt{3J_2}$$

In the equation, $J_2$ is the second invariant of the stress deviator, while $A$, $n$, D, and $m$ are experimental constants for salt rock materials. $Q$ is the activation energy, $R$ is the universal gas constant (8.3143 KJ/mol), $T$ is the absolute temperature in Kelvin, and $\sigma$ is the stress deviator. When temperature changes are not considered, the values of $A$ and $D$ can be interconverted. The calculation parameters are shown in

Table 8. Basic parameters table for numerical simulation.

Rock Type	Elastic Modulus (GPa)	Poisson’s Ratio	Cohesion (MPa)	Internal Friction Angle (°)	Tensile Strength (MPa)	A [(MPa)h−1]	n
Salt Rock	6.8	0.26	5.77	34	1.34	1.02 × 10−7	3
Mudstone	11.2	0.24	7.74	26.5	3.22	2.4 × 10−8	1.7

.

4.4.3. Work Conditions

Due to the seasonal fluctuations in natural gas consumption, it is necessary to cycle the injection and production of natural gas within a year to meet the market demand, while ensuring the stability of the cavern. This results in varying internal pressures within the cavern. Three operational conditions have been designed based on different injection and production frequencies, as shown in Figure 17: Condition 1 involves one injection and production cycle per year, Condition 2 involves two cycles per year, and Condition 3 involves three cycles per year.

According to the operating principles of underground salt cavern gas storage, the operation can generally be divided into four stages: (i) low–pressure maintenance, (ii) pressurization for gas storage, (iii) high–pressure storage, and (iv) depressurization for gas production. Experience indicates that during the gas production phase, the pressure within the cavern decreases, increasing the deviatoric stress on the rock salt and accelerating the creep rate. Therefore, throughout the entire operation, the cavern experiences maximum contraction and deformation during the gas extraction phase, followed by the low–pressure maintenance phase.

4.4.4. Deformation Analysis of Surrounding Rock

After 30 years of injection and production operation, the surrounding rock shows an overall trend of inward contraction under different operational conditions. The displacement of the cavern decreases with increased injection and production cycles, indicating that multiple operations effectively suppress cavern shrinkage. The displacement distribution cloud maps for 5, 10, 20, and 30 years are shown in Figure 18, Figure 19 and Figure 20. The deformation of the upper cavern roof is significantly greater compared to that of the bottom slab, with the maximum displacement located at the center of the horizontal cavern. The maximum displacements for different conditions are presented in Figure 21. The vertical cavern design on both sides adopts a “pear” shape, with the roof deformation also being greater than that of the bottom slab. The maximum displacement occurs at the shoulder of the horizontal cavern, and the difference between the two increases over time. As the number of injections and production increases, the displacement around the cavern decreases, but the maximum displacement difference between Conditions 1 and 3 is less than 6%.

4.4.5. Volume Shrinkage

By calculating the displacement of the cavern under different injection and production conditions, the volume shrinkage rate of the cavern at different times can be calculated. The changes in the volume shrinkage rates of the cavern over 30 years under the three operational conditions are similar. The overall volume of the cavern decreases continuously over time, with the volume shrinkage rate reducing as the number of injection and production cycles increases. The first five years exhibit a rapid volume shrinkage rate, which then gradually slows down, showing a linear increase over time, as illustrated in Figure 22. In Condition 1, the final volume shrinkage rate after 30 years is 27.3%, averaging 0.91% per year. In Condition 2, the final volume shrinkage rate is 18.6%, averaging 0.62% per year. In Condition 3, the final volume shrinkage rate is 15.7%, averaging 0.52% per year. These results meet the requirement that the volume shrinkage rate over 30 years does not exceed 30% [6]. Through analysis, we see that the volume shrinkage mainly increases during the low–pressure stage. Therefore, in order to reduce the volume shrinkage, the low–pressure stage should be shortened.

4.4.6. Relative Plastic Zones

The cyclic modes have significant influences on the relative plastic zones (RPZ). The RPZ of the two cyclic operating modes are shown in Figure 23. RPZs increase with oscillations. Within each cycle, RPZ has a peak and a bottom plateau. The RPZs of the different modes have a similar increasing trend with time; that is, RPZs increase quickly in the initial years and slow down in the later periods. ‘One cycle/year’ has a larger RPZ than the others. This indicates that the longer the period of each cycle, the higher the amplitude of the variation of the RPZs, which is adverse to the safety of the UGS cavern. Through analysis, we see that the RPZ is mainly increased during the ‘Low–pressure stage’; thus, to improve the safety margin of the cavern, the ‘Low–pressure stage’ should be shortened as much as possible.

4.4.7. Safety Factor

During the injection and production operations of the gas storage, internal fractures in the rock develop and connect under stress, leading to a phenomenon of volume expansion. This results in a significant increase in the permeability of the surrounding rock, which can easily cause a gas leakage from the storage facility. The safety factor is calculated using the following equation [40,41]:

$$\sqrt{J_2}={\displaystyle \frac{D_1{\left(\frac{I_1}{\mathrm{sgn}\left(I_1\right){\sigma }_0}\right)}^n+T_0}{\sqrt{3}\cos \theta -D_2\sin \theta }}$$

$$SF={\displaystyle \frac{-0.27I_1}{\sqrt{J_2}}}$$

where $D_1$ and $D_2$ are material parameters; $I_1$ is the first stress invariant, defined as $I_1={\sigma }_1+{\sigma }_2+{\sigma }_3$; ${\sigma }_0$ is a dimensional constant; $\theta$ is the angle of repose; $T_0$ is the uniaxial tensile strength; SF is the volume expansion safety factor; and $J_2$ is the second stress invariant, calculated as $J_2={\displaystyle \frac{1}{6}}\left[{\left({\sigma }_2-{\sigma }_1\right)}^2+{\left({\sigma }_3-{\sigma }_2\right)}^2+{\left({\sigma }_3-{\sigma }_1\right)}^2\right]$. Figure 24 shows the distribution map of the safety factor in the surrounding rock.

After 30 years of cavern operation, the safety factor of the surrounding rock is greater than 1 under all three working conditions, and the overall distribution is consistent. The areas with lower safety factors are located at the junction of the horizontal section and the vertical cavern (SF = 1.03) and at the interlayer (SF = 1.6). The lower safety factor at the junction between the horizontal section and the vertical cavern is caused by the irregular geometry of the model and the finer mesh division, which does not reflect the actual conditions; as such, irregular geometric shapes do not occur during the cavern construction process. The safety factor of the surrounding salt rock remains greater than 1.8. According to the safety factor evaluation standards, regions with a safety factor below 1.5 indicate rock damage, below 1.0 indicate rock failure, and below 0.6 indicates rock collapse. The overall safety factor of the salt cavern is relatively high, indicating a stable condition.

5. Conclusions and Prospect

A new method for constructing gas storage caverns in thin salt layers has been proposed based on the construction needs in China, with a series of studies conducted in the geological conditions of southern Pingdingshan. The main conclusions and recommendations are as follows:

(i)

A systematic comparison of various cavern construction methods was performed, analyzing their economic feasibility in thin salt layers. The horizontal multi–step method was identified as the optimal choice. The primary issue with conventional vertical cavern construction is the small volume and poor economic returns. Although the horizontal multi–step method remains the best option for constructing thin salt layers, it requires a specific salt quality and interlayer thickness, making it unsuitable for most of China’s multi–layered thin salt mines.

(ii)

A series of rock mechanics tests (unconfined compression, triaxial compression, Brazilian splitting, and triaxial creep) were conducted. Test results indicate that the experimental area has potential for constructing salt cavern gas storage.

(iii)

Based on 3D seismic data analysis, optimal locations within the 90–120 m range were identified.

(iv)

A new I–shaped cavern design, suitable for thin salt layers with multiple interlayers and low cavern formation rates, was proposed. The optimized horizontal cavern increases the effective volume by approximately 39% compared to the SWV cavern. Comparatively, the investment for the optimized horizontal cavern construction is reduced by 9.6% compared to conventional single–well, single–cavern methods and by 7.3% compared to the horizontal multi–step construction method.

(v)

An analysis of three injection and production scenarios was conducted for thin–salt–layer constructions, employing criteria such as the displacement, volume shrinkage rate, plastic zone, and safety factor for stability evaluation. The results demonstrate that the I–shaped cavern exhibits good stability, ensuring the stability of the surrounding rock under gas storage operation conditions.

Most importantly, the new I–shaped cavern shows ideal feasibility for natural gas storage in thin–salt–rock formations. As high–quality thick–salt–layer resources become increasingly scarce, further research and the practical application of the new I–shaped cavern should be prioritized to enhance China’s gas storage and peak–shaving capabilities.