Experimental Study on Seawater Intrusion Law and Countermeasures within Island Underground Water-Sealed Oil Storage Caverns

Abstract

Underground water-sealed oil storage caverns constructed in island environments is a promising approach for expanding oil storage caverns. However, few researchers have studied the risks of seawater intrusion and the distribution characteristics of intrusion interfaces in large underground water-sealed oil storage caverns in island environments. In this paper, we established a visualized physical simulation platform to investigate the characteristics and control measures of seawater intrusion in single fracture of rock masses within island underground water-sealed oil storage caverns. In addition, the effects of the excavation of caverns, the distance between the cavern and coast, fracture width, seawater level, oil storage stage, water curtain, and water injection pressure were evaluated. The results show that excavation of oil storage caverns carries the risk of seawater intrusion. Specifically, reducing the distance between the oil storage caverns and the coast, increasing the fracture width, and raising the seawater level all contribute to accelerated seawater intrusion into the caverns. However, the vertical water curtain can effectively inhibit seawater intrusion, and increasing the injection pressure of vertical water curtain can avoid the risk of seawater intrusion into the caverns. The research results provide an experimental basis for the study of seawater intrusion in underground oil storage caverns in island environments.

1. Introduction

Underground water-sealed oil storage caverns play a crucial role in the energy reserve systems of countries worldwide due to their high safety and economic benefits [1,2,3,4,5]. These facilities are commonly established in inland regions characterized by stable surrounding rock and abundant groundwater resources [6,7,8,9,10,11,12]. However, due to the stringent geological conditions required, the availability of suitable land-based reservoir sites is decreasing, posing challenges to fulfilling China’s oil reserve needs.

In recent years, in order to increase oil reserves and the utilization rate of island resources, it has become a new trend to construct underground water-sealed oil storage caverns in island environments [13]. Along the southeast coast of China, there are a considerable number of granitic islands with stable surrounding rock, abundant groundwater, and excellent hydraulic conditions, which are suitable for the construction of large-scale underground water-sealed oil storage caverns. Moreover, the location of many islands, either being already developed or situated near oil terminals, offers convenient access for oil import and export operations. Consequently, the utilization of islands for the establishment of large-scale underground oil storage facilities carries significant importance in expanding the national petroleum reserves. Currently, there are limited instances of constructing underground water-sealed oil storage caverns in island environments [14,15,16,17].

Compared to traditional land-based environments, the seepage field operation law of the underground oil depot system in island environments is significantly more complex, requiring careful consideration of issues such as seawater intrusion, hos in hydraulic conductivity, and tidal dynamics [18,19,20,21]. In addition, seawater intrusion will not only corrode the structure of the underground caverns (steel, concrete, etc.) but also affect the quality of the crude oil [22]. Therefore, it is crucial to pay more attention to seawater intrusion during the construction and operation of underground water-sealed oil storage caverns in island environments.

Research on seawater intrusion primarily emphasizes analytical modeling and laboratory simulation [23,24,25,26,27]. According to the compatibility between seawater and freshwater, the theoretical model of seawater intrusion is divided into sharp interface model and density-dependent model. The sharp interface model disregards the hydrodynamic dispersion between saltwater and freshwater. It assumes the immiscibility of seawater and freshwater, and there is a distinct interface between them in a state of static equilibrium. The hydraulic heads of saltwater and freshwater along this interface exhibit continuous distribution [28,29,30,31]. Despite the fact that the sharp interface approach has been commonly adopted for modeling saltwater intrusion, it may not be valid for situations where transition from the saltwater to freshwater region is not abrupt and the width of the transition zone is greatly affected by hydrodynamic dispersion [32]. For such situations, the density-dependent model was proposed by researchers. Henry regarded saltwater and freshwater as homogeneous fluids and provided an analytical solution for the distribution of salinity in a two-dimensional confined aquifer under the influence of convection and diffusion [33]. In order to enhance the applicability and stability of the analytical solution for the Henry problem, numerous scholars have derived both analytical and numerical solutions by considering different boundary conditions, different parameters, and the dispersion effect [34,35,36,37]. These theoretical advancements were built upon the typical Henry problem, enabling the resolution of more complex seawater intrusion issues.

Physical model tests have become one of the effective methods for studying seawater intrusion. Sand tank experiments are a powerful tool for the investigation and visualization of seawater intrusion, which can help researchers to identify general patterns and features [38]. Numerous experimental studies on seawater intrusion have been conducted by scholars [39,40,41]. These experimental studies have examined various factors such as the influence of density and water head differences between saltwater and freshwater on interface migration, changes in the salt–fresh water interface during tidal fluctuations, and the impact of different water-bearing media on seawater intrusion [42,43,44]. However, there is no corresponding experimental study on the problem of seawater intrusion in the construction of groundwater-sealed oil storage caverns.

The control countermeasures of seawater intrusion are mainly as follows: remediation of underground aquifer, rational exploitation of groundwater, artificial recharge of groundwater, and set up barrier wells along the coast [45,46,47,48]. Barrier wells located along the coast can effectively decrease seawater intrusion into the underground caverns [16]. Therefore, the effect of water curtain on inhibiting seawater intrusion is worth studying.

In this paper, a visualized physical simulation platform of seawater intrusion of the island underground water-sealed oil storage caverns was established. The fracture media model was adopted, and the experimental model was designed based on the hydrogeological condition and spatial structure of the oil storage caverns. Figure 1 shows the shows the process of the experimental study. The distribution characteristics of seawater intrusion interfaces before and after cavern excavation were explored by model test. This study investigated the effects of cavern excavation, distance from the coast, fracture width, seawater level, and oil storage on seawater intrusion interfaces. Additionally, the effect of a vertical water curtain on inhibiting seawater intrusion was verified. The study fills an experimental gap in the field of seawater intrusion within island underground water-sealed oil storage caverns, and the results provide a theoretical basis for the construction of underground water-sealed oil storage caverns in island environments in the future.

2. Experimental Materials and Methods

2.1. Theory and Ratio of Similarity

The relationship between the site and the model is established through dimensional analysis. The relevant parameters affecting seawater intrusion are shown in

Table 1. Influence factors of similar model test.

Control Parameters	Symbol	Standard Unit	Dimension
Structural dimension	D	[m]	L
Space coordinate	X	[m]	L
Space coordinate	Y	[m]	L
Hydraulic path	ΔL	[m]	L
Water level	h	[m]	L
Time	t	[s]	T
Density	ρ	[kg/m3]	ML−3
Pressure	ΔP	[kg/m·s2]	ML−1T−2
Velocity	V	[m/s]	LT−1
Kinematic viscosity coefficient	νw	[m2/s]	L2T−1
Permeability coefficient	K	[m/s]	LT−1

.

In this paper, the dimensional matrix method was used to establish the corresponding similarity model test. The similarity criterion solution process is as follows:

(1) According to the model similarity theory, the function of the influencing factors is:

$$f(D,X,Y,\rho ,V,{\nu }_w,\Delta P,K,\Delta L,h,t)=0$$

(1)

In this paper, according to the principle of independence between the three physical quantities in the dimension selection, the structural size D was selected in the geometric quantity, the intrusion time t was selected in the kinematic quantity, and the pressure ΔP was selected in the kinetic quantity as the basic physical quantity containing the reference physical dimension.

(2) The selected basic physical quantities and other remaining physical quantities were combined into dimensionless quantities, and the relationship (2) expressed by eight dimensionless combination quantities π and the dimensionless power index expression (3) are written as follows:

$$F({\pi }_1,{\pi }_2,{\pi }_3,{\pi }_4,{\pi }_5,{\pi }_6,{\pi }_7,{\pi }_8)=0$$

(2)

$$\begin{array}{l}{\pi }_1=X\times {(D)}^{a_1}\times {(T)}^{b_1}\times {(\Delta P)}^{c_1}\\ {\pi }_2=Y\times {(D)}^{a_2}\times {(T)}^{b_2}\times {(\Delta P)}^{c_2}\\ {\pi }_3=\rho \times {(D)}^{a_3}\times {(T)}^{b_3}\times {(\Delta P)}^{c_3}\\ {\pi }_4=V\times {(D)}^{a_4}\times {(T)}^{b_4}\times {(\Delta P)}^{c_4}\\ {\pi }_5={\nu }_w\times {(D)}^{a_5}\times {(T)}^{b_5}\times {(\Delta P)}^{c_5}\\ {\pi }_6=K\times {(D)}^{a_6}\times {(T)}^{b_6}\times {(\Delta P)}^{c_6}\\ {\pi }_7=\Delta L\times {(D)}^{a_7}\times {(T)}^{b_7}\times {(\Delta P)}^{c_7}\\ {\pi }_8=h\times {(D)}^{a_8}\times {(T)}^{b_8}\times {(\Delta P)}^{c_8}\end{array}$$

(3)

(3) According to the dimension concordant principle, the similarity criterion is solved, and the non-dimensional expression of similarity model is shown in

Table 2. The non-dimensional expression of similarity model.

Similarity Criterion	π1	π2	π3	π4	π5	π6	π7	π8
Expressions	$\frac{X}{D}$	$\frac{Y}{D}$	$\frac{\rho }{D^{-2}{T}^2\Delta P}$	$\frac{V}{D^2{T}^{-1}}$	$\frac{{\nu }_w}{D^2{T}^{-1}}$	$\frac{K}{D{T}^{-1}}$	$\frac{\Delta L}{D}$	$\frac{h}{D}$

.

According to the similarity relationship between the practical project and the test model, the similarity index can be obtained to meet the model test design:

$$\left.\begin{array}{l}\frac{C_X}{C_D}=1\\ \frac{C_Y}{C_D}=1\\ \frac{C_{\rho }}{C_D^{-2}{C}_T^2{C}_{\Delta P}}=1\\ \frac{C_V}{C_D^2{C}_T^{-1}}=1\\ \frac{C_{{\upsilon }_w}}{C_D^2{C}_T^{{}^{-1}}}=1\\ \frac{C_K}{C_D{C}_T^{-1}}=1\\ \frac{C_{\Delta L}}{C_D}=1\\ \frac{C_h}{C_D}=1\end{array}\right\}$$

(4)

By comprehensively considering the relevant factors of the test, the similarity ratio of the model size is finally determined to be 1:500. Therefore, according to the similarity criterion, the similarity ratio of space coordinate is 1/500, the similarity ratio of hydraulic path is 1/500, the similarity ratio of water level is 1/500, the similarity ratio of time is 1/500², and the similarity ratio of pressure is 1/500.

2.2. Experimental Setup

Figure 2 illustrates a visualized physical simulation platform for studying the law of seawater intrusion in a single fracture of rock masses in island groundwater-sealed oil storage caverns.

The device has dimensions of 160 cm (length) × 2.7 cm (width) × 30.5 cm (height) and is divided into three compartments: the seawater and freshwater chambers located at the left and right ends, and a central flow chamber that contains the fractured medium. The seawater and freshwater chambers represent the landward and seaward boundary, respectively, and the dimensions of both chambers are 15 cm (length) × 15 cm (width) × 60 cm (height). There are overflow outlets on the outside of the two chambers. The hose is connected to the overflow outlet, allowing accurate adjustment of the water level using the U-tube principle.

To guarantee both the structural integrity and visual clarity of the experiment, two 15 mm thick organic glass plates were used to simulate a single fracture surface of granite. The back plate is fixed, while the front plate is adjustable. Gaskets, with thicknesses ranging from 0.2 mm to 0.5 mm, were placed at both the lower and upper boundaries of the seepage zone. An elastic deformation strip was inserted into grooves located in front of the adjustable plate. The two plates were tightly joined together using deformation extrusion, followed by reinforcement of the upper section of the model using a tooling clamp. Consequently, the distance between the two plates represents the width of the fracture width. Moreover, the device is equipped with two rotatable supports at its base, facilitating adjustments in orientation as required.

The size of the caverns in the model is reduced proportionally to 1/500 based on the actual section size of the oil storage caverns. The width of the caverns in the test model is 50 mm, the height is 60 mm, and the distance between cavern and coast is 60 mm. The main caverns are 280 mm away from the bottom of the model, and the nearest distance from the seawater chamber is 500 mm. The horizontal water curtain is positioned 50 mm from the top of the cavern, and the vertical water curtain is 50 mm away from the nearest cavern.

Seawater was prepared by dissolving 33.4 g of salt and 1 g of carmine pigment per liter of tap water. The density was measured using a densitometer to ensure a density of 1.025 g/L. For observation purposes, the blue pigment was added to the tap water to simulate freshwater. The solute transport rates of food pigments and NaCl were verified to be consistent, and food pigments have been successfully used in image analysis experiments [39,40].

A coordinate grid with centimeter accuracy was affixed to the outer surface of the plate for reading the test data. The experimental process was recorded by a camera at ten-second intervals.

2.3. Cavern Excavation Experiment

In order to research the influence of underground oil storage cavern construction on the dynamic evolution of the interface of seawater intrusion, the excavation experiment of underground caverns was carried out. Before excavation, the drainage outlets of the caverns were closed, and the fracture width was adjusted to 0.2 mm, the preset fresh water level height was 42 cm, and the seawater level height was 44 cm. Freshwater was injected into the freshwater chamber to form a stable flow from freshwater to seawater, and then seawater was injected into the seawater chamber. Upon filling the seawater chamber, the seawater intrusion process commenced. The distribution characteristics of seawater intrusion interface under natural conditions were clarified, and this experiment served as the control group for subsequent experiments.

In the excavation experiment, the seawater intrusion experiment under natural conditions served as the basis. The drainage outlets were opened to simulate the excavation of the caverns, and the dynamic change in the seawater intrusion interface under the excavation of the caverns was recorded using a video camera.

2.4. Influential Experiments on Relevant Factors

To investigate the influence of various factors on the dynamic evolution of the seawater intrusion interface in single fracture, relevant parameters of the experimental device were modified to research the dynamic evolution.

In order to study the impact of the distance between caverns and coast on the degree of seawater intrusion, three separate experiments were conducted after the seawater intrusion reached a steady state under natural conditions. These experiments involved individually opening the drainage outlets of caverns 1#, 2#, and 3#, respectively, to observe and record the dynamic changes in seawater intrusion process.

The seawater level was adjusted to 42.5 cm, 43 cm, and 43.5 cm, respectively, to investigate its influence on the degree of seawater intrusion. Subsequently, seawater intrusion experiments were conducted sequentially to observe the changes in the seawater intrusion interface before and after excavation of the cavern.

To investigate the influence of fracture width on the distribution characteristics of seawater intrusion interface, the fracture widths were adjusted to 0.2, 0.3, 0.4, and 0.5 mm by replacing the internal gasket in the device. For experimental convenience, after reaching a steady state of seawater intrusion under natural conditions, only chamber 1 was opened. Subsequently, the dynamic change process of the seawater intrusion interface was observed, and the time of the seawater intrusion into cavern under different fracture widths was recorded.

To determine whether underground oil storage caverns will be susceptible to seawater intrusion during the operational period after storing oil, the simulated industrial oil was stored inside the cavern 1#. Seawater intrusion experiment was conducted to observe dynamic changes in the interface of seawater intrusion after oil storage in the cavern.

2.5. Water Curtain Experiment

During the construction and operation stages, underground oil storage caverns may be at risk of seawater intrusion. Therefore, it is particularly important to explore methods to inhibit seawater intrusion. In this study, a specially designed device was used to examine the inhibitory effect of different water curtains on seawater intrusion during the excavation phase of a cavern. A horizontal water curtain was installed 5 cm above the cavern, and a vertical water curtain was installed 5 cm outside the cavern on the side facing the sea. Before excavation, water was injected into the horizontal water curtain and the vertical water curtain, respectively, to simulate the water replenishment effect of the water curtain. Subsequently, the drainage outlet of cavern 1# was opened to investigate the inhibitory effect of different types of water curtains.

The water injection pressure of the water curtains is crucial during operation. In order to investigate the impact of changes in water injection pressure before and after cavern excavation on seawater intrusion, the pressure of the water curtains was adjusted using water pumps and pressure meters.

3. Results and Discussion

3.1. Dynamic Variation Characteristics of Seawater Intrusion

3.1.1. Natural Conditions

Figure 3 shows the dynamic variation characteristics of the seawater intrusion interface under natural conditions. When the seawater completely filled the seawater chamber and reached the preset concentration, the seawater intrusion formally began. As time progressed, the seawater intrusion interface gradually became clearer, transitioning from an initial straight-line shape to a downward concave parabolic shape. The toe shape became convex upward near the toe of the saltwater wedge as a consequence of the dispersion. This effect was observed by Kohout and explained by Henry [49]. Simultaneously, the seawater wedge (SW) continued to intrude inland, and the toe of the seawater wedge (TOSW) increased from the initial 30 cm to 42 cm after 90 min. The results show that seawater will intrude inland over time under the natural conditions, but the intrusion rate gradually slows down until it reaches equilibrium.

3.1.2. After Excavation of Oil Storage Caverns

The excavation of the underground caverns disrupted the initial equilibrium state between seawater and freshwater under natural conditions, leading to changes in the seawater intrusion interface. Specifically, after the excavation of the underground cavern, the upper edge of the seawater wedge (UE) gradually ascended, while TOSW extended towards the bottom of the cavern, gradually forming an intrusion pathway (Figure 4). Subsequently, seawater intruded into the cavern along the bottom, gradually invaded to caverns 1#, 2#, and 3#. The results show that the excavation of the caverns will aggravate the degree of seawater intrusion. When the seawater invades inland, the TOSW extends to the cavern, gradually forming an intrusion path extending from TOSW to the bottom of the caverns, and then the seawater invades the interior of the cavern along the intrusion path.

3.2. Influence of Different Factors on Seawater Intrusion Interface

3.2.1. Influence of Distance between Excavated Caverns and the Coast

The distribution of the seawater intrusion interface was similar when the cavern was excavated at different distances from coast (Figure 5); seawater invaded the interior of the excavated cavern along the intrusion path. However, as the distance between the cavern and the coast increased, the extent of the seawater intrusion inland expanded, and the time required for it to intrude into the cavern also increased (Figure 6). During the excavation of the cavern 1#, 2#, and 3#, when the seawater intrusion reached the equilibrium state, the displacement of SW remained constant, and the values were 61 cm, 69 cm, 80 cm, respectively. In addition, according to the Figure 6, it can be seen that the time for TOSW to reach the same displacement point decreases with the decrease in the distance between the cavern and the coast.

The results demonstrate that increasing the distance between the underground cavern and the coast effectively delays the advancement of seawater intrusion into the cavern. Therefore, in practical engineering, the process of seawater intrusion can be delayed by appropriately increasing the distance between the excavated cavern and the coast.

3.2.2. Influence of Fracture Width

As the fracture width varied, the characteristic of the seawater intrusion into caverns 1#, 2#, and 3# remained consistent and unaltered. As shown in Figure 7, before the excavation of the cavern, the time taken for seawater to intrude to the observation point increased as the fracture width decreased, which indicates a positive linear correlation between the rate of seawater intrusion and the fracture width. This finding is also confirmed in Figure 8. Both the formation time of the seawater intrusion path and the time taken to intrude into each cavern increase with the decrease in the width of the fracture. However, it is worth noting that in Figure 7, for different fracture widths, the displacement of TOSW was 42 cm when seawater intrusion reached a stable state. This verifies that the permeability coefficient of the aquifer only affects the degree of seawater intrusion when it falls below the threshold of 46 m/d [50]. Notably, the calculated permeability coefficient based on the fracture width employed in our study exceeded this threshold.

The result indicates that the influence of fracture width on the degree of seawater intrusion can be ignored. However, according to the linear regression analysis, there is a positive linear correlation between the rate of seawater intrusion and the fracture width. In other words, the intrusion rate is proportional to the square of the fracture width.

$$\frac{t_a}{t_b}={(\frac{D_b}{D_a})}^2$$

(5)

$$\frac{v_a}{v_b}={(\frac{D_a}{D_b})}^2$$

(6)

where $t$ is the time of seawater intrusion, $v$ is the velocity of seawater intrusion, the subscripts a and b represent two different fracture widths.

3.2.3. Influence of Seawater Level

Figure 9 presents the impact of different seawater levels on the dynamic change characteristics of the seawater intrusion interface. At seawater levels of 42.5 cm, 43 cm, and 43.5 cm, the seawater intrusion displacement to the equilibrium state before cavern excavation measured 46 cm, 50 cm, and 55 cm, respectively. For identical intrusion durations, the intrusion distance exhibited an upward trend as the seawater level increased until reaching the equilibrium state (Figure 10).

The results show that as seawater levels rise, the seawater intrusion distance increases, while the time required for seawater to intrude into the cavern decreases. In engineering practice, fluctuations in seawater levels induced by tides, typhoons, and other environmental factors can exacerbate the intrusion of seawater into caverns. Therefore, according to the influence of seawater level on the degree of intrusion, reducing the buried depth of caverns can alleviate the influence of seawater level change.

3.2.4. Influence of Oil Storage

Figure 11 illustrates the similar distribution characteristics of the seawater intrusion interface under conditions of oil storage and excavation. Ultimately, seawater intruded into the cavern from the bottom. However, there is a discrepancy in the amount of seawater intrusion at the same time. The amount of seawater intrusion during the oil storage stage of the cavern is significantly less than that during the excavation stage. The color of the seawater wedge obviously fades. The discrepancy arises from the fact that during excavation, the cavern possesses zero internal pressure, whereas during oil storage, the cavern experiences internal pressure. As a result, the pressure differential between the seawater boundary and the interior of the cavern is diminished, which in turn reduces the force of intrusion. Nevertheless, the risk of seawater intrusion persists even after oil storage within the cavern. Therefore, relevant countermeasures need to be taken to deal with the risk of seawater intrusion during cavern excavation and oil storage.

3.3. Inhibition of Water Curtain on Seawater Intrusion

3.3.1. Influence of Different Water Curtains on Seawater Intrusion

The above studies have revealed that constructing underground water-sealed oil storage caverns in island environments carries the risk of seawater intrusion. Therefore, the effect of water curtain holes on inhibiting seawater intrusion in island storage caverns was tested. Figure 12a illustrates that the seawater intruded into the caverns 1#, 2#, and 3# without water curtain, and the SW spread inland. Upon installing horizontal water curtain holes, the effect was limited to impeding the UE from advancing inland, as depicted in Figure 12b. However, the extent of seawater intrusion into the excavated caverns was not significantly reduced, and seawater was still able to intrude into the caverns.

As shown in Figure 12c–h, after the vertical water curtain was put into operation, the initially clear interface between seawater and freshwater gradually became blurred. Before the excavation of the cavern, the area of SW continued to decrease. Specifically, the TOSW retreated from the initial 42 cm to 36 cm, and the UE also reduced from the initial 33 cm to 20 cm. After excavation of the cavern, the SW area rapidly increased, and displacements of TOSW and UE rapidly increased (Figure 13). The width of the seawater intrusion path significantly decreased due to the influence of the vertical water curtain holes. The seawater intruded inland along the bottom of the device, finally intruded into the cavern, and the amount of intruded seawater also decreased significantly. Additionally, the change curve indicates that under the effect of the vertical water curtain, whether before or after excavation, the displacement rates of TOSW and UE exhibit an initial rapid increase followed by a slower increase.

The results show that the horizontal water curtain cannot effectively prevent the intrusion of seawater into the excavated cavern. However, the vertical water curtain can form a water pressure wall on both sides of the underground caverns, which effectively reduces the thickness of the seawater wedge. Therefore, setting up vertical curtains is an effective way to prevent seawater intrusion, and is worthy of further study.

3.3.2. Influence of Water Injection Pressure on Seawater Intrusion

The above studies have found that a vertical water curtain hole can effectively inhibit seawater intrusion. Therefore, a study on the correlation between the injection pressure of vertical water curtain holes and their effectiveness in inhibiting seawater intrusion was conducted.

As shown in Figure 14, with a water injection pressure of 1 kPa, the seepage area affected by the vertical water curtain was limited, resulting in minimal effectiveness in inhibiting seawater intrusion. A comparison of Figure 14a,b reveals that at lower water curtain pressures, the vertical water curtain influenced seawater inflow on the left side, creating a concave seawater barrier zone at the bottom of the excavated cavern. This led to the narrowing of the seawater intrusion path and reduction in the amount of seawater intruding into the cavern. Figure 14c demonstrates that the vertical water curtain exhibited a substantial inhibitory effect on seawater intrusion when the injection pressure was raised to 2 kPa. The influence of the vertical water curtain on the seepage area expanded as the injection pressure increased. This led to the formation of a larger seawater barrier zone at the bottom of the caverns, resulting in a significant reduction in the width of the TOSW and a substantial narrowing of the intrusion path. Consequently, the seawater had to circumvent the barrier zone by intruding into the cavern from the device’s bottom. When the water injection pressure of the vertical water curtain reached 3 kPa, the affected seepage area and formed seawater barrier zone expanded even further. Even after the excavation of the cavern, the TOSW was unable to penetrate the seawater barrier formed by the vertical water curtain. As a result, the formation of an intrusion path was successfully prevented, thereby averting seawater intrusion into the cavern. The results show that increasing the water injection pressure of the vertical water curtain can form an effective seawater barrier zone between the cavern and seawater, thereby inhibiting seawater intrusion into the cavern.

4. Conclusions

In this study, a visualized physical simulation platform to investigate seawater intrusion in a single fracture of rock masses within an underground island oil storage cavern was established, and model experiments were conducted to investigate the dynamic changes in the seawater intrusion interface during the excavation of the caverns and the inhibitory effect of water curtain holes on seawater intrusion. It obtained the following conclusions.

Under natural conditions, seawater intrusion progresses slowly and the degree of intrusion is limited. After excavation of the caverns, seawater further intrudes inland and intrudes into the interior of the caverns along the bottom. The distance between the excavated cavern and the coast directly impacts the formation rate of the intrusion path and the speed of seawater intrusion. In the experimental range of fracture widths, the influence of fracture width on the degree of seawater intrusion before and after excavation is negligible. However, the speed of seawater intrusion demonstrates a proportional relationship with the square of the fracture width, indicating an increase in intrusion rate as the fracture width expands. Additionally, as the seawater level rises, the degree of seawater intruding inland intensifies, resulting in an accelerated formation of the intrusion path and an increased speed of seawater infiltration into the caverns. Even during the operation of the oil storage caverns, the risk of seawater intrusion persists. The distribution characteristics of the intrusion interface resemble those observed during excavation conditions, although the amount of intruded seawater is reduced. Therefore, island underground water-sealed oil storage caverns will face the problem of seawater intrusion during the excavation and operation, and some countermeasures need to be taken to deal with the risk in the actual project.

The horizontal water curtain has limited inhibitory effects on seawater intrusion, while the vertical water curtain exhibits a significant inhibitory effect. With the increase in the water injection pressure of the vertical water curtain, the convection effect of freshwater on seawater is obviously strengthened. Additionally, the impact range of the vertical water curtain expands, effectively blocking the hydraulic connection between the coastline and the underground oil storage cavern, thus preventing the risk of seawater intrusion. Therefore, it is recommended to set a vertical water curtain on the side of the cavern adjacent to the coast to deal with the problem of seawater intrusion in the actual project of the island groundwater-sealed oil storage caverns.