Research on Leakage Mechanism of Underwater Shield Tunnels with Different Soil Layers during Operation Period

Abstract

Different soil layer properties have great influence on tunnel leakage. In this paper, the finite element software is used to analyze the homogeneous leakage model of different soil layers and different degrees of lining deterioration. The working conditions of different soil layer properties and different degrees of lining deterioration are established by means of soil layer permeability coefficient and lining permeability coefficient, and the law of tunnel leakage under each working condition is analyzed. This paper innovatively explores the relationship between the severity of tunnel seepage caused by different lining deterioration and the soil layer properties of the tunnel. Finally, by comparing the field survey data of Dinghuaimen with the leakage law and mechanism summarized by the model, the reliability and feasibility of the model method to explore the leakage mechanism are verified. The model test results are consistent with the project survey data. In this paper, the seepage law and mechanism of underwater shield tunnel are explored and the result shows: (1) The leakage laws and stress distribution laws of a single homogeneous soil layer with different strata properties are similar. The flow velocity in the lower half of the tunnel lining is greater than the upper half, and the maximum velocity occurs at the arch foot; the velocity inside the lining is much smaller than that inside the soil layer. The maximum absolute value of the axial stress of the lining occurs at the arch waist on the left and right sides, and the minimum values of the axial stress of absolute values all occur at the dome. (2) The leakage of underwater tunnels is related to the properties of soil layers. The larger the soil permeability coefficient is, the more serious the deterioration of the tunnel lining will be, and the more serious the tunnel leakage will be. (3) The leakage of complex strata follows the law of leakage in a single homogeneous stratum. The leakage of the tunnel is the most serious in the fine sand layer with the largest permeability coefficient. Moreover, when the seepage phenomenon occurs, the water flow tends to seep along the fine sand layer to the tunnel.

1. Introduction

After entering the 21st century, with the rapid development of infrastructure construction in China, a large number of underwater shield tunnel projects under construction and already under construction have appeared. China is known as the country with the largest number of underwater tunnel projects in the world [1]. Due to the harsh geological conditions [2,3], changes in water level [4], earth pressure [5] and other unfavorable factors, underwater shield tunnels are prone to leakage diseases. The structural safety and driving safety of tunnels are affected by leakage diseases.

Leakage disease is known as one of the main diseases of shield tunnels during operation. A large number of scholars have done research on tunnel leakage disease. The existing leakage-related researches are mostly aimed at the joints, including the waterproof performance of the joint sealing gasket [6,7,8], research on the joint structure and parameters [9,10] and the study of mechanical properties of joints [11]. The waterproof structure of joint sealing gasket has been improved by some scholars, for example, a new double sealing gasket waterproof system has been designed by Hongming Xie [12]. In addition, a series of new testing devices were proposed to explore the correlation between the waterproof performance and mechanical performance of joint gaskets [13,14]. The influence of strata on tunnel leakage has not been studied by researchers.

Zhang D.M. [15] used analytical methods and numerical simulation methods to deduce the settlement prediction caused by seepage water in saturated clay, considering the relative permeability of tunnel lining and soil. Shin [16] simulated the degree of joint leakage by changing the ratio of permeability coefficient, and explored the influence of the deterioration of hydraulic conditions in the formation on the pore water pressure, structural deformation and long-term leakage of segment joints. Zhang [17] studied the stratum and tunnel responses under three conditions of homogeneous leakage of the whole tunnel, local leakage on one side of the tunnel and local leakage on both sides of the tunnel, through numerical simulation based on the leakage statistics of the Shanghai Metro shield tunnel. Shin [4] studied the long-term effects of tunnel lining structure permeability and groundwater level changes on the tunnels in the clay of London, UK, and Seoul, South Korea. Wongsaroj [18] established a three-dimensional finite element model of soil–water coupling. The lining structure of the tunnel is regarded as uniform infiltration, and it is divided into two cases: drainable and non-drainable. The response of the tunnel under the long-term action of soil is considered. It can be seen that the scholars’ focus on the influence of stratum properties on tunnel leakage is relatively single, mainly for shield tunnels in soft soil layers. However, the stratum conditions of underwater tunnels are complex, and there are not only soft soil strata. At present, there are few studies on the complex geological conditions with high water pressure and strong water permeability for a long time.

Regarding the study of underwater shield tunnels, Shin [19] used a combination of analytical and numerical methods, and considered the relationship between seepage flow and pore pressure in the presence of linings. JOO [5] established the relationship between the pore water pressure in the underwater tunnel and the leakage flow rate of the tunnel by analytical method. Moon [2] studied the importance of the influence of high permeability stratum on the tunnel seepage through numerical simulation, and concluded that the high permeability stratum has an influence on the groundwater flow state and infiltration rate. Existing simulations for highly permeable strata in underwater tunnels have relatively simple settings for stratum conditions and leakage conditions, which cannot restore the influence of strata on tunnel leakage in real situations.

In this paper, finite element analysis software is used to establish a single homogeneous soil model considering different stratum properties and a numerical model based on the actual geological survey report at a section of the N-line of the Dinghuaimen Tunnel in Nanjing, China. Several common soil parameters of underwater tunnels are selected for a single homogeneous soil model, and the working conditions of different soil layers and different degrees of lining deterioration are realized. By setting different lining permeability coefficients, and analyzing the seepage field and seepage velocity of each working condition of the model, the seepage law and mechanism of the tunnel in a single homogeneous soil layer are obtained. The seepage field and seepage velocity law of underwater shield tunnel under complex formation conditions are explored and summarized. By analyzing the numerical model of the N-line project of the Dinghuaimen Tunnel in Nanjing, China, and comparing the numerical model of the project and the single homogeneous soil model under various working conditions, the seepage field and seepage velocity law of the underwater shield tunnel under complex stratum conditions can be explored and summarized. The prediction and treatment of seepage in the project has certain significance because the seepage mechanism and law in line with the actual working environment of the tunnel are summarized.

2. Numerical Computation Model

In order to study the long-term effect of tunnel lining leakage and analyze the mechanism of tunnel leakage under the action of different factors, a numerical model of a single homogeneous soil layer was established based on the finite element analysis software ABAQUS. Several common soil parameters of underwater tunnels are selected for a single homogeneous soil model, and by setting different lining permeability coefficients, the working conditions of different soil layers and different degrees of lining deterioration are realized. The seepage field and seepage velocity of each working condition of the model are analyzed to obtain the seepage law and mechanism of the tunnel in a single homogeneous soil layer.

2.1. Model Construction

In the tunnel water leakage analysis, Shin et al. [16,17,20] proved the reliability and rationality of the two-dimensional tunnel model in the analysis of deformation, settlement and stress. The two-dimensional tunnel model is used in this paper to study the influence of water leakage on the seepage field and seepage velocity of shield tunnels during operation, and to achieve the purpose of exploring the mechanism of tunnel leakage.

The tunnel center buried depth is h = 15 m, the outer diameter of the tunnel lining is 14.5 m, the inner diameter is 13.9 m, and the thickness of the concrete lining segment is 0.6 m. In order to minimize the influence of the setting of seepage boundary on the calculation results [21], the length of the tunnel model soil is 290 m and the height is 94.75 m. The soil and lining are all analyzed by CPE4P element for pore fluid/displacement coupling analysis. There is a “tie” constraint between the soil and the lining that does not vary with calculation time.

The accuracy of the calculation results is affected by the form of meshing. Through the author’s experiment, when the mesh element size is approximately 4 m, the calculation results cannot converge. In this paper, the mesh element size approximately 1~3 m is called mesh accuracy 1~3. In order to explore a reasonable grid division form, the four mesh partitioning methods selected in this paper are shown in Figure 1: all grids are uniformly set to grid accuracy 1, 2 and 3, or mesh accuracy 1 or 3 is selected according to the mesh position. (The mesh accuracy is 1 when the element is located near the tunnel. The mesh accuracy is 3 when the element is far away from the tunnel.)

As can be seen from Figure 2, four different meshing methods lead to different calculation results. By comparing the vertical stress on the outside of the lining, it can be seen that the calculation results of mesh accuracy 1 and mesh accuracy 2 are similar and the most accurate, but both of them take too long to calculate. The calculation result of mesh accuracy 3 is quite different from that of mesh accuracy 1 and mesh accuracy 2, so the calculation result of mesh accuracy 3 is not accurate enough. The calculation result of mesh accuracy 1&3 is accurate enough, and the calculation process is also time-consuming, so this form of meshing is selected in this paper. In this paper, mesh accuracy 1&3 is adopted. That is, the mesh size is approximately 1 m when the element is located near the tunnel. When the element is far away from the tunnel, the mesh size is approximately 3 m.

2.2. Research Scheme of Leakage Mechanism of Shield Tunnel under the Action of Different Factors

2.2.1. Research Scheme of Leakage Mechanism of Shield Tunnel under Different Formation Conditions

When exploring the mechanism of tunnel leakage, the research method that the shield tunnel lining is equivalent to a homogeneous permeable body is used by most scholars [4,18,20]. At present, the research on tunnel leakage in soft soil layer has been widely concerned by the academic circles [4,20,22,23]. In fact, the stratum that the underwater shield tunnel passes through includes not only soft soil, but also a variety of complex strata such as silty sand, pebble gravel and hard rock.

Several typical strata often traversed by underwater shield tunnels are selected as the analysis objects. The physical and mechanical parameters are shown in

Table 1. Soil parameters.

	Dry Density ρ (kg/m3)	Elastic Modulus E (MPa)	Poisson’s Ratio	Permeability Coefficient k (m/s)
Silty sand	1.94	17.030	0.3	5.78 × 10−6
Fine sand	2.02	19.480	0.26	2.315 × 10−4
Silty clay	1.86	5.699	0.39	2.31 × 10−7
lining		2.68 × 104	0.16	5 × 10−13

. Using the control variable method, several typical strata are set as a uniform single soil layer, and the tunnel lining model is simplified to uniform permeability (the permeability coefficient is 5 × 10⁻¹³ m/s) for seepage analysis. The leakage mechanism of underwater shield tunnels with different stratum conditions during the operation period was explored.

2.2.2. Research Scheme of Tunnel Leakage Mechanism under Different Degrees of Deterioration of Lining Waterproof Performance

Due to the long-term service of underwater shield tunnels in unfavorable environments such as high water pressure and running water erosion, the mechanical properties of the lining concrete are deteriorated by the influence of the working environment and the operating time of the tunnel. The impermeability of the lining is reduced due to the deterioration of the lining, and the permeability coefficient is increased. The variation range of the lining permeability coefficient is set as 5 × 10⁻¹³~5 × 10⁻⁹ m/s. As the permeability coefficient increases, it means that the deterioration degree of the tunnel lining is more serious.

As shown in

Table 2. Classification of working conditions for different lining deterioration degrees.

Working Condition	Soil	Soil Permeability Coefficient ks (m/s)	Lining Equivalent Permeability Coefficient k1 (m/s)
1	Silty sand	5.78 × 10−6	5 × 10−12
2			5 × 10−11
3			5 × 10−10
4			5 × 10−9
5	Fine sand	2.315 × 10−4	5 × 10−12
6			5 × 10−11
7			5 × 10−10
8			5 × 10−9
9	Silty clay	2.31 × 10−7	5 × 10−12
10			5 × 10−11
11			5 × 10−10
12			5 × 10−9

, various working conditions were established by setting different tunnel lining permeability coefficients in three soil layers. The severity of the tunnel leakage is divided by the flow velocity of the leaking water when the tunnel leaks in the homogeneous stratum. The seepage field and leakage law of each working condition were analyzed, and the leakage mechanism of the underwater shield tunnel under different degrees of deterioration of the waterproof performance of the lining during the operation period was explored.

2.3. Calculation Analysis Step and Setting of Boundary Conditions

This paper focuses on the leakage during tunnel operation, so the leakage analysis during tunnel excavation is not considered. The numerical model is mainly divided into three analysis steps to realize the calculation [24]: ① Gravity balance analysis step: The purpose is to simulate the situation that the surrounding soil does not seepage and consolidate in a very short period of time during tunnel excavation. The initial stress distribution of the tunnel under its own weight after excavation will be displayed. ② Load analysis step: To approximate the actual situation, the effect of the river pressure above needs to be considered. A uniform load corresponding to the depth of the river is set on the model surface. ③ Seepage consolidation analysis step: To simulate the actual situation, the inner boundary pore pressure of the tunnel lining is set to 0, and the seepage consolidation time of up to 10,000 d is set.

Mechanical boundary conditions: The entire model is subject to its own gravity, and a uniform pressure formed by a water depth of 41 m is distributed on the upper surface of the model. The left and right boundaries of the model are constrained to move in the horizontal direction. The bottom boundary is constrained to move horizontally and vertically. The upper surface of the model is a free boundary.

Seepage boundary conditions: According to the data of the initial hydrostatic pressure that can be formed at the 41-m-deep river water level, the initial pore pressure distributions on the upper surface and the left and right boundaries of the model are set. The underside of the model is set to be impermeable. The pore pressure at the inner boundary of the tunnel lining is set to 0 during operation. The section soil of the underwater shield tunnel is saturated soil, and the change of water level is not considered.

3. Results Analysis

3.1. Leakage Mechanism of Shield Tunnels under Different Stratum Conditions during the Operation Period

It can be seen from Figure 3 that the permeability coefficient of the lining is 5 × 10⁻¹³ m/s, that is to say, the self-waterproof performance of the lining is good. At this time, the seepage velocity inside the lining is much smaller than the velocity inside the soil layer. It can be seen that when the waterproof level of the lining of the underwater tunnel meets the requirements, the possibility of leakage of the lining is reduced to almost nothing. This is because the seepage field in the soil will be affected by the existence of the tunnel, and the water in the soil will have a seepage trend towards the tunnel. This tendency to seepage can be blocked by a lining with good waterproofing properties, thus preventing leakage diseases from occurring. For example, in the silty clay homogeneous soil layer, the maximum velocity of the soil at the arch foot is 7.043 × 10⁻¹⁰ m/s, and the maximum velocity of the lining with good waterproof performance is 5.933 × 10⁻¹¹ m/s. It can be seen that most of the seepage path of the water flow is blocked by the lining, and only a small part of the water flow can penetrate into the tunnel.

As shown in Figure 4, the angle between the leakage position on one side and the vertical diameter of the tunnel is set as the abscissa, and the seepage velocity in the soil is set as the ordinate. It can be seen from the figure that, when uniform leakage occurs in the underwater shield tunnel, basically the same leakage law is displayed in different homogeneous soils. The flow velocity in the lower half of the soil body where the tunnel is located is greater than that in the upper half, and the maximum velocity occurs at the arch foot. It is the result of the combined action of water pressure and gravity. In homogeneous soil, the closer to the bottom of the tunnel, the greater the soil water pressure. The hydraulic gradient in the lower half of the tunnel is larger than that in the upper half, resulting in a larger flow velocity in the lower half of the soil. Under the influence of gravity, the upward flow velocity of water seepage in the soil at the bottom of the tunnel arch is reduced, so the maximum flow velocity of seepage occurs at the arch foot in all underwater tunnels. In three different homogeneous soils, the seepage velocity of the tunnel in the soil increases with the increase of the soil permeability coefficient. Just as the distribution of permeability coefficient of homogeneous soil is fine sand > silty sand > silty clay, the maximum flow velocity distribution is fine sand > silty sand > silty clay.

When uniform leakage occurs in underwater shield tunnels, similar stress distribution laws in different homogeneous soils are shown. As shown in Figure 5, when the seepage occurs in the underwater shield tunnel, the maximum absolute value of the axial stress in the lining appears at the apron on the left and right sides, and the maximum axial stress of a single soil layer with different soil qualities is basically the same. For example, the maximum values of silty sand, fine sand and silty clay are 10.8439 kPa, 11.1789 kPa and 11.2365 kPa, respectively. When the seepage occurs in the underwater shield tunnel, the minimum value of the absolute value of the axial stress all appears at the vault, and the minimum value is greatly affected by different soil conditions. For example, the minimum values of silty sand, fine sand and silty clay are 6.4676 kPa, 6.0348 kPa and 7.8361 kPa, respectively. It can be seen that when the seepage occurs in the underwater shield tunnel in different homogeneous soil layers, the distribution positions of the maximum and minimum values of the axial stress of the lining are basically the same, but different soil bodies have an influence on the value of the axial stress.

3.2. Leakage Mechanism of Tunnel under Different Degrees of Deterioration of Lining Waterproof Performance

Due to the long-term service of underwater shield tunnels in unfavorable environments such as high water pressure and running water erosion, the mechanical properties of the concrete lining will be deteriorated by the influence of the working environment and the operating time of the tunnel. The simulation of lining deterioration is realized by increasing the equivalent permeability coefficient of the lining. It can be seen that in a single homogeneous soil layer, the seepage velocity of water flow will gradually increase with the gradual deterioration of the waterproof performance of the lining. As shown in Figure 6, cracks are more likely to form in the places where the water flow inside the tunnel is larger, which leads to an increase in the potential leakage points of the tunnel. In a single homogeneous layer, the water flow velocity at the arch foot is the largest, so the lining concrete material is more likely to form cracks and gradually develop at the arch foot under the scour of water flow.

As shown in Figure 7, the leakage severity of each working condition in Table 2 is divided according to the water flow velocity when leakage occurs in the soil. It can be seen from the figure that the properties of the soil layer and the degree of deterioration of the lining have an impact on the leakage. When comparing the same lining deterioration degree, the leakage flow velocity of the fine sand layer is the largest, and the leakage is the most serious. For example, the fine sand layers in each lining deterioration degree are seriously leaking. The silty sand and silty clay stratum will only appear serious leakage when the permeability coefficient of the lining is as high as 5 × 10⁻⁹ m/s. In the same homogeneous layer, the seepage velocity of water flow increases with the severity of lining deterioration. For example, in the silty clay layer, the maximum flow velocity at the arch foot increases gradually with the serious deterioration of the lining. They are 1.45697 × 10⁻⁹ m/s, 6.27083 × 10⁻⁹ m/s, 5.45356 × 10⁻⁸ m/s, and 5.14172 × 10⁻⁷ m/s, respectively.

4. Project Example Verification

Nanjing Dinghuaimen Yangtze River Tunnel is a tunnel in Nanjing City, Jiangsu Province, China. It consists of a southern line and a northern line. The total length of the northern line is 7014 m, including the underwater shield tunnel section of 4135 m. The total length of the southern line is 7363 m, including the underwater shield tunnel section of 3557 m. The tunnel is designed as a double-pipe double-deck eight-lane structure with a design speed of 80 km/h.

The geological conditions of the Dinghuaimen Tunnel in Nanjing are complex. The minimum radius of the horizontal curve of the shield section is 1000 m, the maximum longitudinal slope is 4.5%, and the minimum radius of the vertical curve is 4500 m for the convex shape and 2700 m for the concave shape. The maximum covering soil thickness is about 51 m, the maximum water depth above the vault of the tunnel in the middle section of the river is about 62.3 m, and the maximum water pressure at the bottom of the tunnel is about 0.765 MPa. Difficult formations such as soft soil, silt sand, gravel and hard rock are passed through by the tunnel.

As shown in Figure 8, section A located on the east side of the river of the Dinghuaimen Tunnel N-line with a larger water depth and a smaller covering thickness is selected in this paper. The stratum type of the tunnel section belongs to fully permeable stratum. The upper half of the stratum is the sandy permeable layer (silty sand, fine sand) and the lower half of the stratum is silty clay. The surface deformation of the stratum is large, the subsidence is large, and it is easy to flood water and sand. The typical section of this section passing through silty sand, fine sand and silty clay is selected to establish the tunnel leakage model of the composite soil. The leakage laws of the previously simulated tunnels under the influence of factors of different single homogeneous layers and different lining deterioration degrees are compared with the model results. The tunnel leakage mechanism of the actual working conditions of the underwater shield tunnel during the operation period is explored in the following.

4.1. Establishment of Dinghuaimen Engineering Case Model

The stratigraphic properties of typical sections of the Nanjing Dinghuaimen underwater tunnel are shown in

Table 3. Soil physical and mechanical parameters.

Code	Soil	Thickness (m)	Dry Density ρ (kg/m3)	Elastic Modulus E (MPa)	Poisson’s Ratio	Permeability Coefficient k (m/s)
①	Silty sand	12.5	1.94	17.030	0.3	5.78 × 10−6
②	Fine sand	6.5	2.02	19.480	0.26	2.315 × 10−4
③	Silty clay	11	1.86	5.699	0.39	2.31 × 10−7
④	Medium weathered siltstone	68.75	2.71	10	0.12	1.04 × 10−6
	Tunnel lining			2.68 × 104	0.16	5 × 10−13

. The thickness of the covering soil above the tunnel is 15 m. The river water level is 41 m according to the once-in-a-hundred-years flood level. The diameter of the tunnel is 14.5 m, the thickness of the lining is 0.6 m and 10 lining segments are installed on each ring of the tunnel. The strata distribution of the tunnel section is shown in Figure 9. The finite element analysis software ABAQUS and the previous model construction method were used to establish the leakage model of this section.

4.2. Leakage Analysis of Tunnel Example

According to the geological survey report of the Dinghuaimen Tunnel, the fine sand layer passed by the N-line is 1273.25 m long, accounting for 35.76% of the total length of the northern line. The silty clay layer accounted for 449.78 m, accounting for 12.63% of the total length. In addition, the upper half layer of 760.64 m tunnel is fine sand. The northern line has a total of 1782 ring lining. According to the field statistics since 2019, the number of leakage points and the percentage of section rings in the tunnel can be used to understand the severity of N-line leakage of the Dinghuaimen Tunnel under different geological conditions. The percentage of the number of seepage points and the number of cross section rings in the fine sand formation is as high as 51.52%. The silty sand layer is 41.76%, while silty clay is 28.23%. The tunnel severity data of different strata investigated in the field are highly consistent with the simulation conclusion in the third part of the paper, which shows the feasibility of the calculation model for the leakage simulation of the Dinghuaimen Tunnel.

As shown in Figure 10, the waterproof performance of the lining at this typical section is good, and the leakage degree of the tunnel lining at each part is related to the properties of the soil around the tunnel. In the composite soil of this section in Dinghuaimen, the maximum flow velocity of leakage also occurs in the silty sand layer. According to the previous simulation results of a single homogeneous soil layer, in the homogeneous fine sandy soil, the maximum velocity of the vault is 4.264 × 10⁻⁷ m/s, and the arch foot is 6.395 × 10⁻⁷ m/s. In the homogeneous silty clay soil, the maximum velocity of the vault is 1.174 × 10⁻¹⁰ m/s, and the maximum velocity of the arch is 7.043 × 10⁻¹⁰ m/s. In the composite soil, the fine sand layer is located in the upper half of the tunnel, and the silty clay layer is located in the lower half of the tunnel. The flow velocity in the lower half of the tunnel is slightly higher than the flow velocity in the upper half due to water pressure. The leakage flow velocity is more affected by the soil permeability coefficient, the maximum flow velocity in the soil appears in the fine sand layer.

In the composite soil, the gradual increase in the leakage sites of the fine sand layer (located on the vault) is caused by the deterioration of the waterproof performance of the tunnel lining. The flow trend of the water flow is to seepage along the silty sand layer to the tunnel lining. As shown in Figure 11, the top of the soil layer traversed by the tunnel in this section of Dinghuaimen is a silty sand layer, the soil permeability coefficient is small. The upper part is fine sand stratum with a large soil permeability, and the lower part is a silty clay stratum with the smallest soil permeability coefficient. The permeability around the tunnel is weak in the upper and lower parts and strong in the middle part, which is caused by the distribution of soil layers. The permeability of shield tunnel lining is increased from 5 × 10⁻¹³ m/s to 5 × 10⁻⁹ m/s, the water flow velocity of each part of the tunnel in the composite formation increases. The obvious seepage phenomenon appears in the fine sand layer on the upper part of the tunnel lining. The seepage parts of the fine sand layer of the vault increase significantly, and the seepage velocity increases. The main source of leakage water comes from this layer. The phenomenon of seepage occurs, and the water flow tends to leak along the fine sand layer to the tunnel lining. This phenomenon is caused by the feature that the permeability of the composite formation is weak in the upper and lower parts and strong in the middle part.

As shown in Figure 12, according to the survey site of the leakage situation of the section in the Dinghuaimen Tunnel, the leakage disease exists at the lining vault of the section. Wet stains appeared in some segments, and water droplets occasionally dripped from the joints of the vault segments. The leakage of the vault part of this section is more serious than other parts. According to the previous research on the leakage mechanism of the tunnel, the soil layer of the leakage part of the section is fine sand soil layer. The fine sand soil layer is the soil layer with the largest permeability coefficient in the composite soil layer of this section. From the previous conclusions, it can be inferred that the source of the seepage water is mainly provided by the silty sand layer.

5. Conclusions

In this paper, the finite element software is used to analyze the homogeneous leakage model of different soil layers and different degrees of lining deterioration. Through the analysis of tunnel seepage field and seepage velocity under different working conditions to judge the severity of tunnel leakage and explore the mechanism of leakage. This paper innovatively explores the relationship between the severity of tunnel leakage caused by different lining deterioration and the soil layer properties of the tunnel. The reliability and feasibility of using this model method to explore the mechanism of leakage are verified by the field survey data of the Dinghuaimen Tunnel. Finally, according to the actual working conditions of the Dinghuaimen project, the seepage law and mechanism of the Dinghuaimen Tunnel under complex stratum conditions are discussed. The conclusion is as follows:

(1)

Under the condition that the self-waterproof performance of the tunnel lining is good, the leakage law of the phenomenon that the uniform leakage of the underwater shield tunnel occurs in a single homogeneous soil layer is basically the same. The flow velocity in the lower half of the tunnel is greater than the flow velocity in the upper half, and the maximum velocity occurs at the arch foot. The velocity inside the lining is much smaller than the velocity inside the soil layer. For example, in the silty clay homogeneous soil layer, the maximum velocity of the soil at the arch foot is 7.043 × 10⁻¹⁰ m/s, and the maximum velocity of the lining with good waterproof performance is 5.933 × 10⁻¹¹ m/s.

(2)

The stress distribution law of tunnel leakage lining under different stratum conditions is similar. The maximum value of the absolute value of the axial stress in the lining occurs at the arch waist on the left and right sides of the tunnel. The minimum value of the absolute value of axial stress appears at the vault.

(3)

The leakage of underwater tunnels is related to the properties of soil layers. The different strata where the tunnel is located and the different degrees of deterioration of the lining have an impact on the leakage of the tunnel. The larger the soil permeability coefficient, the more serious the deterioration of the lining, and the serious leakage. For example, in the silty clay layer, the maximum flow velocity at the arch foot increases gradually with the serious deterioration of the lining. They are 1.45697 × 10⁻⁹ m/s, 6.27083 × 10⁻⁹ m/s, 5.45356 × 10⁻⁸ m/s, and 5.14172 × 10⁻⁷ m/s, respectively.

(4)

The method to study the leakage properties of tunnels in different soil layers with a single homogeneous soil layer is available. The leakage law of complex formations follows the analysis results of a single homogeneous formation. In the leakage analysis of several typical single homogeneous layers, the leakage of fine sand layers is serious. In the leakage analysis of the typical section of the Dinghuaimen underwater tunnel, the leakage of the silty sand layer on the vault is the most serious. The permeability of the composite formation is weak in the upper and lower parts and strong in the middle part. Seepage leads to the tendency of water flow to seep along the silty sand layer to the tunnel lining. The silty sandy soil layer is the soil layer with the largest permeability coefficient among the composite soil layers in this section of the Dinghuaimen Tunnel. The area that is most prone to leakage and the source of the leakage water is mainly provided by the silty sand layer.

The method adopted in this paper is reasonable, and the results reach the expectation, which proves the reliability and feasibility of this modeling method. In the next step, the author will use this method to further explore the leakage mechanism of different leakage forms under different strata, such as the influence of different stratum properties on annular seam leakage, longitudinal seam leakage and bolt hole leakage.