A new method for determining the pore water pressure around twin shallow circular tunnels Authors: Tao Zhan, Shengbiao Shan, Chao Wang, Peiyuan He, Jinfeng Zou Published: 2025 ABSTRACT To obtain the distribution rules of pore water pressure for twin shallow circular tunnels, the shallow circular tunnels were regarded as fully saturated, homogeneous, isotropic structures in a semi-infinite space. Coordinate transformations were utilized to obtain the water pressure control equations in a bipolar coordinate system. Through the Schwarz alternating method, the calculation method for the pore water pressure around a single tunnel was generalized to twin tunnels, and a new method for determining the pore water pressure around twin shallow circular tunnels in a semi-infinite space was proposed for the first time by using the bipolar coordinate system method and the Schwarz alternating method. Solutions for the pore water pressure were obtained via multiple iterations. The calculation results obtained via the proposed method were compared with the other theoretical calculation results and numerical simulation results, and we further utilized the proposed method to analyse the influencing factors of the pore water pressure. The results show that the maximum error between the theoretical calculation results and the numerical simulation results was only 2.15%; thus, the rationality of this new method was effectively verified. The relative error between the results obtained by the proposed method and the results obtained by the existing method was only 1.0%, and both very well matched the numerical simulation results; thus, the high accuracy of the proposed method was confirmed. The number of tunnels and the centre distance between the twin tunnels are the main influencing factors of the pore water pressure. According to their influences, the distribution pattern of the pore water pressure around the twin shallow circular tunnels can be preliminarily determined. Therefore, the method proposed in this study provides a theoretical guidance for the development of design and construction programs for water-rich tunnel engineering. Introduction Underground works are invariably constructed with a strong emphasis on the impact of groundwater. Especially in the process of water-rich tunnel excavation, groundwater can easily flow into the tunnel due to the seepage effect, generating pore water pressure in the surrounding rock, which will negatively affect the effective stresses in the tunnel surrounding rock and cause the tunnel structure to fail. Therefore, to guarantee the safety of the tunnel structure during construction, the distribution pattern of pore water pressure around the tunnel under the influence of seepage should be analysed to maintain the stability of the tunnel structure in groundwater-rich areas1–3. To date, several studies have been conducted on the pore water pressure around tunnels. For a single tunnel, Shin et al.4 studied the potential effect of the pore water pressure on the tunnel lining via the finite element method and proposed a method for estimating the design curve of the pore water pressure for the secondary lining. Pan et al.5 analysed the effect of pore water pressure on the stability of the palm face via fluid mechanics and numerical simulation. Park et al.6 derived an analytical solution for the seepage force around a circular tunnel via the complex variable method. Zhang et al.7 used complex variable theory and the conformal mapping method to determine the half-plane excess pore pressure around an unlined tunnel caused by grouting pressure and further obtained an analytical solution for soil consolidation based on the conformal transformation method and Terzaghi–Rendulic consolidation theory. They also analysed the influencing factors of the pore water pressure around a tunnel and ground consolidation settlement. Yang et al.8 considered the influence of cutter head rotation and slurry specific gravity on pressure boundary conditions and proposed a multifield coupling model that can be used to analyse the distribution pattern of the excess pore water pressure. Lee et al.9 proposed a calculation approach for groundwater seepage in surrounding rock by using the limit analysis upper limit method. Yang et al.10 proposed stability solutions that consider the water seepage of shallow tunnels on the basis of strength reduction theory. For twin tunnels, Zhang et al.11 derived analytical solutions of the non-Darcy seepage field in shallow single-hole and double-hole circular tunnels via the principle of the mirror method. Zhu et al.12 derived the stable seepage field of twin parallel tunnels via the complex variable method. Zhang13 derived an analytical solution for the stable seepage field of twin parallel tunnels in a semi-infinite plane via the principle of the mirror method. Wan et al.14 used fully grouted vibrating-wire piezometers to measure rapid pore water pressure changes around tunnels, and they identified five distinct immediate pore water pressure responses. Guo et al.15 proposed a new alternating analytical approach to establish a new calculation method for the steady-state seepage field around underwater twin tunnels in semi-infinite saturated aquifers. Yang et al.16 used the nonlinear failure criterion and modified tangential technique to propose upper-bound solutions for the critical supporting pressure on a deep tunnel face under the conditions of pore water pressure. Hu et al.17 used Terzaghi–Rendulic theory to establish the governing equations for the dissipation of excess pore water pressure in the surrounding soil mass of a tunnel; thus, they verified the rationality of the novel analytical solution for the consolidation behaviour of viscoelastic saturated soft soil subjected to large-scale ground loading. According to our literature review, theoretical research on the pore water pressure around tunnels has focused mostly on single tunnels, and research on the pore water pressure around twin tunnels has been relatively limited, especially research on the change rule of the pore water pressure around twin shallow tunnels. In addition, although the change rule of pore water pressure can be described by analytical equations, it is often affected by complex boundary conditions, so these analytical methods usually lack practical applicability and have not attracted substantial interest in engineering. To facilitate the development of design and construction programs for water-rich shallow buried tunnel projects, it is necessary to study the theoretical calculation method for the pore water pressure around twin shallow circular tunnels to explore the distribution range of the pore water pressure and its change rule. The form of the surrounding pore water pressure distribution due to tunnel excavation is often complex, and twin shallow circular tunnels are represented as plane strain problems in semi-infinite spaces. Considering the difficulty of solving for the displacement and stress of the surrounding rock and the variation in pore water pressure in a semi-infinite space under the Cartesian coordinate system or polar coordinate system, according to Jeffery18, the bipolar coordinate system method is introduced to solve the elasticity problem in semi-infinite spaces. As a special curvilinear coordinate system, the bipolar coordinate system can greatly simplify the calculation flow when specific conditions such as perimeter rock displacement19 and pore water pressure are considered. Moreover, since twin shallow circular tunnels are symmetric about the vertical direction (y-axis), according to Hu et al.17, Bi et al.20 and Zhang et al.21, the Schwarz alternating method is often used to obtain analytical solutions for the mechanics of symmetric tunnels. Therefore, the bipolar coordinate system method and Schwarz alternating method are used in this study to solve for the pore water pressure around twin shallow circular tunnels via several iterative calculations. According to the results of the above literature review, the Schwarz alternating method has not yet been adopted for the theoretical study of pore water pressure around tunnels. Therefore, the number of iterations and the calculation accuracy of the pore water pressure around tunnels are controlled by setting the average pore water pressure tolerance value [p] in this study to obtain an accurate analytical solution for the pore water pressure around the twin shallow circular tunnels, thus effectively compensating for the deficiencies in the theoretical calculation of the pore water pressure around tunnels. In this study, shallow circular tunnels in a fully saturated, homogeneous, isotropic semi-infinite space are used as an example. We use the bipolar coordinate system method to derive a calculation method for the pore water pressure around a shallow circular tunnel. On this basis, we further use the Schwarz alternating method and set the allowable value of the pore water pressure to control the number of iterations and the computational accuracy of the output results to establish a new method for determining the pore water pressure around twin shallow circular tunnels. By comparing the results obtained via the proposed method with the numerical simulation results and the theoretical calculation results obtained via the existing method, the rationality of the method proposed in this study is verified, and its calculation accuracy is further improved. The method is further utilized to theoretically analyse the main influencing factors of pore water pressure around tunnels, and a basic law of seepage in shallow tunnels is proposed accordingly, which is highly important in guiding the safe construction of water-rich shallow tunnels. Problem description and basic assumptions In Fig. 1, twin shallow circular tunnels are arranged horizontally, and the centre depth of the left tunnel is the same as that of the right tunnel. The radius, depth, centre distance, and ground surface hydraulic head of the twin shallow circular tunnels are r, d, L, and hw, respectively. The left and right tunnels correspond to Tunnel I and Tunnel II, respectively, and the ratio of the centre depth to the radius of each of the left and right tunnels does not exceed 7; thus, the twin tunnels in Fig. 1 can be considered shallow tunnels.Fig. 1Twin shallow circular tunnels in a semi-infinite space. Twin shallow circular tunnels in a semi-infinite space. In Fig. 1, we take the straight line where the ground surface is located as the x-axis and the straight line above the centre of the left circular tunnel and perpendicular to the x-axis as the y1-axis. The intersection of the two is the coordinate origin O1 of the local Cartesian coordinate system x1O1y1. Similarly, we take the straight line over the centre of the right circular tunnel and perpendicular to the x-axis as the y2-axis, and the intersection of the two is the coordinate origin O2 of the local Cartesian coordinate system x2O2y2. The above two local Cartesian coordinate systems correspond to the left tunnel (Tunnel I) and the right tunnel (Tunnel II), respectively, and both belong to the global Cartesian coordinate system xOy, as shown in Fig. 1. Therefore, if the length of each tunnel is sufficiently long, the three-dimensional space problem can be simplified to a two-dimensional plane problem. The hydraulic head of the ground surface is greater than that of the ground, so the soil is fully saturated. The permeability coefficient of the soil is too small to keep the ground surface hydraulic head stable. Therefore, the shallow twin circular tunnels are buried in a fully saturated, homogeneous and isotropous semi-infinite space. In addition, owing to the mutual symmetry of the twin shallow circular tunnels parallel to each other in the semi-infinite plane, the centre distance between the twin tunnels is assumed to be smaller. When the twin shallow circular tunnels in a semi-infinite plane are regarded as the object of this study, it is very difficult to solve for the pore water pressure around the twin shallow circular tunnels directly. Therefore, we use the Schwarz alternating method to regard the problem of twin shallow circular tunnels in the semi-infinite plane as a superposition problem after solving the problem for the left and right tunnels in several iterations to obtain the distribution pattern of the pore water pressure around the twin shallow circular tunnels via the Schwarz alternating method. Derivation of the pore water pressure around a shallow circular tunnel According to Zou et al.22, with the constant hydraulic head seepage mode, the method for determining the pore water pressure around a shallow circular tunnel is derived in the bipolar coordinate system to provide a theoretical basis for solving the problem for the twin shallow circular tunnels. Stable seepage equations Because of the complexity of soil pores, the direction of water seepage is often random. Therefore, seepage is usually analysed macroscopically in engineering. Only the main direction of seepage can be considered in the analysis, and the influence of the soil can be ignored. Before continuous medium theory was proposed, Bernoulli’s equation was commonly used in hydraulics and is expressed as follows:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$u + \frac{1}{2}\rho v^{2} + \rho gh = C$$\end{document}where u is the fluid pressure, v is the fluid flow rate, ρ is the fluid density, g is the acceleration of gravity, h is the hydraulic head height, and C is a constant. Because Eq. (1) is derived from the law of conservation of energy, it is applicable only to ideal fluids that are non-viscous or less viscous and non-compressible. The pressure of the hydraulic head can be expressed as the total hydraulic head of the fluid, as shown in Eq. (2).2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h = \frac{u}{{\gamma_{{\mathrm{w}}} }} + z + \frac{{\overline{v}^{2} }}{2g}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{v}$$\end{document} is the permeation rate of water, γw is the gravity of water, and z is the elevation head, which indicates the distance in the positive direction of the reference plane. If z lies below the x-axis, its value is negative. Figure 2 shows that water flows from the high head to the low head through a connecting pore channel when there is a total head difference in the soil. The total hydraulic head difference (△h) can be expressed as follows:3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta h{ = }h_{A} - h_{B}$$\end{document}with4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h_{A} = \frac{{u_{A} }}{{\gamma_{{\mathrm{w}}} }} + z_{A} + \frac{{\overline{{v_{A} }}^{2} }}{2g}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h_{B} = \frac{{u_{B} }}{{\gamma_{{\mathrm{w}}} }} + z_{B} + \frac{{\overline{{v_{B} }}^{2} }}{2g}$$\end{document}Fig. 2Bernoulli’s principle. Bernoulli’s principle. The head loss of the unit process is the hydraulic gradient, which is expressed in Eq. (6).6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$J{ = }\frac{\Delta h}{{\Delta L}}$$\end{document}where J is the hydraulic gradient. Since seepage, which is a type of laminar flow in the soil, tends to flow slowly with small hydraulic gradients, the momentum changes can be ignored, and the flow rate can be considered zero. For the seepage of laminar flow, the hydraulic head at any position in the whole seepage space can be expressed as follows:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h = \frac{u}{{\gamma_{{\mathrm{w}}} }} + z$$\end{document} Darcy’s law reflects the linear relationship between the seepage rate of water in the soil and the head loss. When water flows in the soil in a laminar state, the seepage of water follows Darcy’s law and can be expressed as follows:8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Q = kAJ$$\end{document}9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$v = kJ$$\end{document} Darcy’s law indicates that the rate of passage is proportional to the magnitude of the hydraulic gradient and the permeability of the porous medium when water passes through the porous medium. However, the seepage rate of the fluid remains nonlinearly related to the hydraulic gradient when the flow rate of the fluid is too fast. Therefore, Darcy’s law is not applicable to turbulence. On this basis, the fluid of interest exhibits laminar flow, with a constant flow rate per unit time. In the Cartesian coordinate system, the water discharge velocity in the x-axis direction is μ, and the water discharge velocity in the y-axis direction is η. The water flow continuity equation is as follows:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{\partial \mu }{{\partial x}} + \frac{\partial \eta }{{\partial y}} = 0$$\end{document} The velocity potential energy equation can be expressed as follows:11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi (x,y) = - k\left( {\frac{u}{{\gamma_{{\mathrm{w}}} }} + z} \right) = - kh$$\end{document} According to Eq. (11), the water discharge velocity in the x-axis direction (μ) and the water discharge velocity in the y-axis direction (η) can be obtained by12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mu = \frac{\partial \varphi }{{\partial x}} = - k\frac{\partial h}{{\partial x}}$$\end{document}13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\eta = \frac{\partial \varphi }{{\partial y}} = - k\frac{\partial h}{{\partial y}}$$\end{document} The Laplace equation for two-dimensional stable seepage23 can be obtained by combining Eqs. (12) and (13) with Eq. (10), as expressed in Eq. (14).14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\nabla^{2} \varphi = \frac{{\partial^{2} \varphi }}{{\partial x^{2} }} + \frac{{\partial^{2} \varphi }}{{\partial y^{2} }} = 0$$\end{document} In the bipolar coordinate system, Eqs. (11) and (14) can be expressed as follows via the coordinate transformation technique:15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi (\alpha ,\beta ) = - kh = - k\left( {J\sinh \alpha + \frac{u}{\gamma }} \right)$$\end{document}16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\nabla^{2} \varphi = \frac{1}{{J^{2} }}\left[ {\frac{\partial }{\partial \alpha }(\frac{\partial \varphi }{{\partial \alpha }}) + \frac{\partial }{\partial \beta }(\frac{\partial \varphi }{{\partial \beta }})} \right]$$\end{document} Analytical solutions for determining the pore water pressure around a shallow circular tunnel We suppose that the permeability coefficient on the outside of the lining structure is relatively high; then, the head at the tunnel boundary is constant, and the hydraulic head at infinity is related only to the tunnel depth. In a semi-infinite space, the Laplace equation for the hydraulic head can be expressed as follows via the coordinate transformation technique:17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\nabla^{2} h = \frac{1}{{J^{2} }}\left( {\frac{{\partial^{2} h}}{{\partial \alpha^{2} }} + \frac{{\partial^{2} h}}{{\partial \beta^{2} }}} \right) = 0$$\end{document} According to the boundary conditions of stable seepage, the head of pore water in the surrounding rock can be expressed as18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h = h_{0} + \frac{{h_{{a_{i} }} - h_{0} }}{{\alpha_{i} }}\alpha$$\end{document}where ha is the hydraulic head at the tunnel boundary and where h0 is the hydraulic head at semi-infinite space infinity. Substituting Eq. (18) into Eq. (11), the pore water pressure in the surrounding rock under constant hydraulic head conditions can be obtained as follows:19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p = h_{0} \gamma_{{\mathrm{w}}} + \frac{{h_{{a_{i} }} - h_{0} }}{{\alpha_{i} }}\gamma_{{\mathrm{w}}} \alpha - J\sinh \alpha \gamma_{{\mathrm{w}}}$$\end{document} Establishment of a method for calculating the pore water pressure around twin shallow circular tunnels Calculation model As shown in Fig. 3, the left tunnel is Tunnel I, and the right tunnel is Tunnel II. Both Tunnels I and II are located below the ground surface hydraulic head. The head of the surface water on the tunnel is hw, and the hydraulic head in the surrounding rock is h1. In the bipolar coordinate system, the coordinates of any point of the tunnel are (α, β). The coordinates of the tunnel wall are denoted by αi. The coordinate of the ground is zero. In a coordinate system with the y-axis oriented vertically downwards in the positive direction, α is always greater than zero, and β ranges from −π to π. Supposing that the lining structures do not have infiltration capacity, the boundary conditions correspond to the constant hydraulic head seepage mode. Because twin tunnels correspond to the multiconnected domain problem24–26, the pore water pressure around twin shallow circular tunnels can be calculated via the Schwarz alternating method to obtain solutions for the pore water pressure around each shallow circular tunnel over multiple iterations.Fig. 3Calculation model for the pore water pressure around the two shallow circular tunnels. Calculation model for the pore water pressure around the two shallow circular tunnels. According to the calculation model for the pore water pressure around the twin shallow tunnels in Fig. 3, to ensure that the bipolar coordinate system method and the Schwarz alternating method can be applied to solve for the pore water pressure at the same time, we propose the following basic assumptions for this calculation model:The surrounding rock is uniform, continuous, and isotropic.The interaction relationship between the twin tunnels meets the calculation requirements of the Schwarz alternating method, and the twin tunnels can be regarded as two single tunnels for iterative solution.Considering the effects of physical strength, the weight of the soil and the weight of the groundwater in the surrounding rock must not be neglected.The state of seepage in specific soil pores is ignored, and the state of seepage in the flow space is analysed macroscopically. Only the main direction of seepage is considered, and the influence of soil particles within the seepage range is ignored; thus, it is assumed that seepage fills the entire flow space, including soil particles and circulation pores.The groundwater in the soil is assumed to be an ideal fluid that is non-viscous or has very low viscosity and no compressibility, and this fluid is in a stable seepage field, which is laminar, i.e., a two-dimensional stable fluid. The seepage process satisfies Darcy’s law, and Bernoulli’s principle can be used to describe the change in pore water pressure in the surrounding rock. The surrounding rock is uniform, continuous, and isotropic. The surrounding rock is uniform, continuous, and isotropic. The interaction relationship between the twin tunnels meets the calculation requirements of the Schwarz alternating method, and the twin tunnels can be regarded as two single tunnels for iterative solution. The interaction relationship between the twin tunnels meets the calculation requirements of the Schwarz alternating method, and the twin tunnels can be regarded as two single tunnels for iterative solution. Considering the effects of physical strength, the weight of the soil and the weight of the groundwater in the surrounding rock must not be neglected. Considering the effects of physical strength, the weight of the soil and the weight of the groundwater in the surrounding rock must not be neglected. The state of seepage in specific soil pores is ignored, and the state of seepage in the flow space is analysed macroscopically. Only the main direction of seepage is considered, and the influence of soil particles within the seepage range is ignored; thus, it is assumed that seepage fills the entire flow space, including soil particles and circulation pores. The state of seepage in specific soil pores is ignored, and the state of seepage in the flow space is analysed macroscopically. Only the main direction of seepage is considered, and the influence of soil particles within the seepage range is ignored; thus, it is assumed that seepage fills the entire flow space, including soil particles and circulation pores. The groundwater in the soil is assumed to be an ideal fluid that is non-viscous or has very low viscosity and no compressibility, and this fluid is in a stable seepage field, which is laminar, i.e., a two-dimensional stable fluid. The seepage process satisfies Darcy’s law, and Bernoulli’s principle can be used to describe the change in pore water pressure in the surrounding rock. The groundwater in the soil is assumed to be an ideal fluid that is non-viscous or has very low viscosity and no compressibility, and this fluid is in a stable seepage field, which is laminar, i.e., a two-dimensional stable fluid. The seepage process satisfies Darcy’s law, and Bernoulli’s principle can be used to describe the change in pore water pressure in the surrounding rock. Bipolar coordinate system method The bipolar coordinate system is a curvilinear coordinate system that was first introduced by Jeffery18 to solve semi-infinite spatial elasticity problems and simplify the related calculations. In Fig. 4, O1(0,a) and O2(0, −a) are the 2 poles of the bipolar coordinate system and are symmetric about the x-axis. Point P is the origin of the Cartesian coordinate system, and a represents the distance from pole O1 or O2 to point P. If any points M(x1,y1) and N(x2,y2) in the Cartesian coordinate system can be denoted as M(α1,β1) and N(α2, β2) in the bipolar coordinate system, then α1 = ln(l2/l1) and β1 = θ2−θ1 , where l1 and l2 are the distances from point M to points O1 and O2, respectively, and θ1 and θ2 are the angles of MO1 and MO2 with the x-axis, respectively. Therefore, the different α values in the bipolar coordinate system can be plotted as a series of coaxial eccentric circular curves called α curves with O1 and O2 as the poles, radius acsch(α), and centre coordinates (0,acoth(α)). Different β values can be plotted as circular curves from the 2 poles and orthogonal to each of the α circles, which are called β curves.Fig. 4Bipolar coordinate system. Bipolar coordinate system. We let the left and lower sides be the positive directions of the x- and y-axes of the Cartesian coordinate system, respectively; then, α is positive below the x-axis, and β is positive to the left of the y-axis. The coordinates (α, β) of any point in the bipolar coordinate system correspond to the intersection of different α and β curves. According to the rules specified above for the coordinates, the interval of α is (−∞, + ∞), and α = 0 when the coordinate point is located on the x-axis. When the coordinate point is located at either of the 2 poles O1 and O2, α is + ∞ and − ∞, respectively. The interval of values of β is (−π,π), and β is π and −π when the coordinate points are located in the O1P and O2P segments on the y-axis, respectively, and β is 0 when the coordinate points are located elsewhere on the y-axis. The larger the radius of the curve is, the smaller the absolute values of α and β are, as shown by the coordinate relationship between points M and N in Fig. 4. The conformal mapping relationship between the bipolar coordinates (α, β) of any point and its corresponding Cartesian coordinates (x, y) is as follows:20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x + {\mathrm{i}}y = {\mathrm{i}}a\coth \left( {\frac{{\alpha + {\mathrm{i}}\beta }}{2}} \right)$$\end{document} Solving for the Cartesian coordinates x and y in Eq. (20) yields21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\{ {\begin{array}{*{20}c} {x = \frac{a\sin \left( \beta \right)}{{\cosh \left( \alpha \right) - \cos \left( \beta \right)}}} \\ {y = \frac{a\sinh \left( \alpha \right)}{{\cosh \left( \alpha \right) - \cos \left( \beta \right)}}} \\ \end{array} } \right.$$\end{document} The angle of intersection of the tangents to the ϕ and β curves with the x-axis can be used to define the stretch ratio of the bipolar coordinate system, which is given by22\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\tan \left( \phi \right) = \frac{\cosh \left( \alpha \right)\cos \left( \beta \right) - 1}{{\sinh \left( \alpha \right)\sin \left( \beta \right)}}$$\end{document} On the basis of the above rules, it is possible to carry out mutual transformations between the Cartesian coordinates and the bipolar coordinates for calculations related to shallow tunnels. Schwarz alternating method The Schwarz alternating method was proposed for solving the Dirichlet problem, and its convergence in two and three dimensions was subsequently demonstrated by Mikhlin27 and Soboldff28. Therefore, for the analysis of the elastic zone of the surrounding rock in a multiconnected domain, the sequence of loading and geometry changes has no effect on the stresses in the surrounding rock. Therefore, the Schwarz alternating method can transform multiconnected domain problems into single connected domain problems29–31 to simplify the solutions for the pore water pressure around twin shallow circular tunnels in a semi-infinite space. The solution for the pore water pressure around a shallow circular tunnel in the bipolar coordinate system is extended to the twin tunnels via the Schwarz alternating method. (1) Supposing that Tunnel II does not exist in the fully saturated, homogeneous and isotropous semi-infinite space, the pore water pressure around only Tunnel I in the bipolar coordinate system can be obtained, which is expressed as follows:23\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{{\mathrm{w11}}}} = h_{0} \gamma_{{\mathrm{w}}} + \frac{{h_{{a_{i} }} - h_{0} }}{{\alpha_{i} }}\gamma_{{\mathrm{w}}} \alpha - J\sinh \alpha \gamma_{{\mathrm{w}}}$$\end{document}where pw11 is the pore water pressure around Tunnel I. (2) Supposing that Tunnel I does not exist in the fully saturated, homogeneous and isotropous semi-infinite space, a hydraulic head (hw21) is applied around Tunnel II to counteract the pore water pressure around Tunnel II that is generated by Tunnel I. The pore water pressure around Tunnel I that is generated by Tunnel II can be expressed as follows:24\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{{\mathrm{w21}}}} = h_{0} \gamma_{{\mathrm{w}}} + \frac{{h_{{a_{i} }} + h_{{{\mathrm{w21}}}} - h_{0} }}{{\alpha_{i} }}\gamma_{{\mathrm{w}}} \alpha - J\sinh \alpha \gamma_{{\mathrm{w}}}$$\end{document}where pw21 is the pore water pressure around Tunnel I that is generated by Tunnel II. (3) Similarly, supposing that Tunnel II does not exist in the fully saturated, homogeneous and isotropous semi-infinite space, a hydraulic head (hw12) is applied around Tunnel I to counteract the pore water pressure around Tunnel I that is generated by Tunnel II. The pore water pressure around Tunnel II that is generated by Tunnel I can be expressed as follows:25\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{{\mathrm{w12}}}} = h_{0} \gamma_{{\mathrm{w}}} + \frac{{h_{{a_{i} }} + h_{{{\mathrm{w12}}}} - h_{0} }}{{\alpha_{i} }}\gamma_{{\mathrm{w}}} \alpha - J\sinh \alpha \gamma_{{\mathrm{w}}}$$\end{document} (4) When the value of pw12 is small enough or equal to zero and when pw12 < [p], the interaction can be stopped. Otherwise, the interaction will continue until pw satisfies these conditions. An analytical solution for the pore water pressure around the twin shallow circular tunnels can be obtained by superimposing the results of each iteration. Analytical solutions for determining the pore water pressure around two shallow circular tunnels In the bipolar coordinate system, when the pore water pressure is determined, the parameters of the twin shallow circular tunnels in the semi-infinite space are shown in Table 1.Table 1Parameters of the twin shallow circular tunnels.ParameterValueGround surface hydraulic head hw (m)30Hydraulic head in the surrounding rock h1 (m)0Tunnel centre depth h (m)20Tunnel radius r (m)5Centre distance between the twin tunnels L (m)20Underground water gravity g (kN/m3)9.8 Table 1Parameters of the twin shallow circular tunnels.ParameterValueGround surface hydraulic head hw (m)30Hydraulic head in the surrounding rock h1 (m)0Tunnel centre depth h (m)20Tunnel radius r (m)5Centre distance between the twin tunnels L (m)20Underground water gravity g (kN/m3)9.8 Parameters of the twin shallow circular tunnels. Since ground surface water and groundwater are used twice during the calculation of the Schwarz alternating method, hw1 = 0.5 and hw = 15 m are set. The effect of groundwater should be ignored in the second iteration, and only the interaction between the tunnels is considered. When the interaction between the tunnels is small enough, an accurate analytical solution for the pore water pressure can be obtained. According to Wei et al.32 and Guo et al.33, the threshold value ([p]) can be set to 10–4 MPa, which controls the termination of the iterations, and the iterative computational flow of the Schwarz alternating method for calculating the pore water pressure around tunnels is shown in Fig. 5.Fig. 5Computational flow for determining the pore water pressure. Computational flow for determining the pore water pressure. In Fig. 6, the value range of the x-axis is set to [-100,100], and the value range of the y-axis is set to [0,100], from which the two cross sections at y = 10 m and y = 30 m in the semi-infinite space are taken as the study area to observe the pore water pressure distribution pattern at the same depth in the surrounding rock of the twin shallow circular tunnels. Sections 1–2, 3–4, 5–6, and 7–8 are selected as the monitoring points of the pore water pressure around Tunnel I.Fig. 6Monitoring points of pore water pressure around the tunnels. Monitoring points of pore water pressure around the tunnels. First iteration of the loop In the bipolar coordinate system, the pore water pressure around Tunnel I is solved for when Tunnel II is assumed not to be in the system. The pore water pressure in the Cartesian coordinate system can be expressed via the coordinate transformation technique. The pore water pressure around Tunnel I is that of a shallow circular tunnel. The calculation results for the pore water pressure around Tunnel I in the above sections are shown in Fig. 7.Fig. 7Pore water pressure around Tunnel I in the first iteration of the loop. (a) Sections I and Section II, (b) Section 1– 2, (c) Section 3– 4, (d) Section 5–6, and (e) Section 7–8. Pore water pressure around Tunnel I in the first iteration of the loop. (a) Sections I and Section II, (b) Section 1– 2, (c) Section 3– 4, (d) Section 5–6, and (e) Section 7–8. When the influence of the pore water pressure around Tunnel I on Tunnel II is analysed, the pore water pressure acting around Tunnel II should be the average value, and a hydraulic head (hw21) should be applied around Tunnel II to counteract the pore water pressure around Tunnel II. In the same way, the pore water pressure around Tunnel II is determined when Tunnel I is assumed to not be in the bipolar coordinate system. The results for the pore water pressure around Tunnel II should be superimposed with the pore water pressure around Tunnel I, as shown in Fig. 8.Fig. 8Pore water pressure around the twin shallow circular tunnels in the first iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Pore water pressure around the twin shallow circular tunnels in the first iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Second iteration of the loop In the first iteration of the loop, the pore water pressure around Tunnel II has a greater impact on Tunnel I. The pore water pressure acting around Tunnel I should be the average value, and a hydraulic head (hw12) should be applied around Tunnel I to counteract the pore water pressure around Tunnel I. In the Cartesian coordinate system, the pore water pressure can be expressed via the coordinate transformation technique. When the pore water pressure around Tunnel II is determined, the iteration should be stopped until the pore water pressure is less than [p] or equal to zero. Because the average pore water pressure is 1.703 × 104 Pa, which does not satisfy the criterion, a hydraulic head (h'w21) should be applied around Tunnel II to counteract the pore water pressure. The pore water pressure around Tunnel II can then be obtained and superimposed with the pore water pressure around Tunnel I, as shown in Fig. 9.Fig. 9Pore water pressure around the twin shallow circular tunnels in the second iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Pore water pressure around the twin shallow circular tunnels in the second iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Third iteration of the loop In the second iteration of the loop, the average value of the pore water pressure acting around Tunnel I is 6.518 × 103 Pa, which is greater than [p], after which the iteration process continues. A hydraulic head (h'w12) should be applied around Tunnel I to counteract the pore water pressure, and the pore water pressure around Tunnel I can be determined. The average value of the pore water pressure around Tunnel II is 2.494 × 103 Pa, which does not satisfy the criterion; thus, the iteration process is continued. The pore water pressure around Tunnel II can be obtained and superimposed with the pore water pressure around Tunnel I, as shown in Fig. 10.Fig. 10Pore water pressure around the twin shallow circular tunnels in the third iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Pore water pressure around the twin shallow circular tunnels in the third iteration of the loop. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Pore water pressure solutions for twin shallow circular tunnels After the sixth iteration of the loop, the average value of the pore water pressure around the tunnels is 20.463 Pa, which satisfies the iterative stopping condition. The pore water pressure around the twin shallow circular tunnels can be obtained by superimposing the results of the first six iterations, as shown in Fig. 11. In the first three iterations, the results of pore water pressure superimposed on the same horizontal plane tended to be symmetrically distributed. After the accuracy satisfied the criterion, the distribution of the pore water pressure also confirmed this finding. In the surrounding rock between Tunnel I and Tunnel II, the pore water pressure within the plane from the ground to the centres of the twin tunnels fluctuates due to the positions of the twin tunnels, whereas the surrounding rock area below the centres of the twin tunnels is consistent with the pore water pressure distribution around Tunnel I in Fig. 7a. Therefore, the distribution pattern of the pore water pressure around the twin shallow circular tunnels is similar to that around a single shallow circular tunnel when the tunnel centre depth is large enough.Fig. 11Pore water pressure around the twin shallow circular tunnels in Sections I and II. Pore water pressure around the twin shallow circular tunnels in Sections I and II. Figure 12a–c shows that the pore water pressure around Tunnel I after the sixth iteration of the loop is similar to the pore water pressure in the first iteration of the loop. Figure 12d shows that the pore water pressure is symmetrically distributed on the centreline of the twin shallow circular tunnels.Fig. 12Pore water pressure around the twin shallow circular tunnels after the sixth iteration of the loop. (a) Sections 1–2, (b) Section 3–4, (c) Section 5–6, and (d) Section 5–6. Pore water pressure around the twin shallow circular tunnels after the sixth iteration of the loop. (a) Sections 1–2, (b) Section 3–4, (c) Section 5–6, and (d) Section 5–6. Validation Comparison with the numerical simulation results and verification of the rationality of the method According to the above derivation of the calculation method for the pore water pressure around the twin shallow circular tunnels proposed in this study, to validate this method, we utilize FLAC3D finite difference software to establish a numerical model for seepage in the two-dimensional twin shallow circular tunnels19–21, which is shown in Fig. 13. The basic assumptions of this model are consistent with the basic assumptions of the calculation model proposed in this study. In addition, considering the characteristics of shallow tunnels in a semi-infinite domain, we set the dimensions of the model to 100 m × 80 m (X × Z) and set the tunnel centre depth, tunnel radius, and centre distance of the twin tunnels to 20 m, 5 m, and 20 m, respectively. Therefore, the size of the model is much larger than the size of the tunnels. On this basis, we mesh the model. To ensure the accuracy of the numerical simulation results, we grid the area near the twin shallow circular tunnels more densely and grid the area farther away from the twin shallow circular tunnels more sparsely. The type of unit is set to a free quadrilateral mesh; thus, the numbers of units and nodes of this model are 12,535 and 15,312, respectively.Fig. 13Numerical model for seepage in the twin shallow circular tunnels. Numerical model for seepage in the twin shallow circular tunnels. When establishing the numerical model, we consider the influence of seepage on underground water and set the ground surface hydraulic head, permeability coefficient, porosity, fluid modulus and fluid density to 30 m, 2 × 10–10 m/s, 0.5, 2 × 109 Pa, and 9.81 kg/m3, respectively, and the other parameters can be found in Table 1. On this basis, we set the number of computational steps to 150,000. In addition, the pore water pressure at the boundary of the model is constant because sufficient underground water is present, and the pore water pressure is evenly distributed with depth. Therefore, we use this numerical model to analyse the pore water pressure around twin shallow circular tunnels. Under the condition of the same hydraulic head, the theoretical calculation results for the pore water pressure at the above four sections (Sections 1–2, 3–4, 5–6, and 7–8) in Fig. 6 are compared with their corresponding numerical simulation results, which are shown in Fig. 14.Fig. 14Comparison of the numerical simulation and theoretical calculation results for the pore water pressure around the twin shallow circular tunnels. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Comparison of the numerical simulation and theoretical calculation results for the pore water pressure around the twin shallow circular tunnels. (a) Sections I and Section II, (b) Section 1–2, (c) Section 3–4, (d) Section 5–6, and (e) Section 7–8. Figure 14a shows that the numerical simulation results are more consistent with the theoretical calculation results when the pore water pressure is near the tunnel and the ground, but there is still a deviation in the results between the tunnel and the ground. The maximum deviation occurs at a depth of 11.5 m. The result of the theoretical calculation is 145.51 kPa, but the result of the numerical simulation is 147.65 kPa, which shows that the maximum error is approximately 1.45%. However, with increasing depth, the error decreases. Figure 14b shows that the results of the theoretical calculation and the results of the numerical simulation are very close in the area where the depth is less than 65 m, and the maximum error is only 0.398%. However, when the depth is greater than 65 m, the error between the theoretical calculation and the numerical simulation results gradually increases. The error is 2.15% when the depth is 80 m, which continues to increase with depth. Figure 14c also shows that the error between the theoretical calculation and numerical simulation results is smaller within 25 m from the left end of Tunnel I. However, the error increases as the distance from Tunnel I increases. When x = −50 m, the error between the theoretical calculation and numerical simulation results is 0.365%. In addition, Fig. 14d shows that the theoretical calculation and numerical simulation results are very close in the area close to the twin shallow circular tunnels or the midpoint of the centreline. When x = −4 m or x = 4 m, the error is only 0.795%, which is the maximum error. Because of the finite boundary in FLAC3D modelling, the calculation in the numerical simulation reaches the steady state faster than the theoretical calculation. During the theoretical derivation, the average value of the pore water pressure around the tunnel is used in the iterations by the Schwarz alternating method, which can also cause errors in the results of the theoretical calculation. However, the theoretical calculation results of pore water pressure obtained by the proposed method are analytical solutions, not numerical solutions. In addition, according to the characteristics of FLAC3D software that uses command flow to implement modeling, the proposed method can substitute the analytical solution directly into the numerical model as the initial state and boundary constraints. The mathematical properties of analytical solutions are used to improve the efficiency of numerical analysis and enhance the computational rate of numerical methods. For this study, the analytical solution obtained by this study can specify the control equations, boundary conditions and initial conditions. By solving the analytical solutions for the local simplified part of the complex problem of pore water pressure around the twin shallow circular tunnels, the results can be used as the input of the global numerical model to accelerate the numerical convergence and improve the computational efficiency of numerical simulation. The rationality and calculation accuracy of the analytical solutions of pore water pressure obtained by the proposed method can be verified by FLAC3D numerical simulation software. Therefore, based on the results of the above comparative analysis, the error between the theoretical calculation and numerical simulation results is very small in overall. Therefore, the calculation method for the pore water pressure around the twin shallow circular tunnels is verified to be reasonable. Comparison with existing methods To verify the accuracy of the proposed calculation method, we compare this method with the method of Guo et al.33. Considering that the method proposed by Guo et al.33 is derived by the complex variable method and Schwarz alternating method to solve the pore water pressure around the twin shallow circular tunnels, while the method proposed in this study is derived by the bipolar coordinate system method and Schwarz alternating method to solve the pore water pressure around the twin shallow circular tunnels, so the rationality of the proposed method can be verified by comparing the calculation results of the two methods. Although the bipolar coordinate system method and the complex variable method are two important methods for solving two-dimensional field problems, there are some differences between them. The bipolar coordinate system method is based on orthogonal coordinate system transformations and is designed for bifocal symmetry problems, which are solved by transforming the physical problem into the bipolar coordinate system and using the separated variable method combined with special functions. While the complex variable method relies on the theory of complex analysis, and maps complex geometric boundaries to simple regions by constructing analytic functions or conformal transformations, so as to satisfy the Laplace equation by directly utilizing the properties of analytic functions. The core difference between the two is that the bipolar coordinate system method relies on symmetry-adapted coordinate systems and is physically intuitive, while the complex variable method maps the geometric complexity through complex plane, which is more flexible and universal. Accordingly, in this study, the calculation case in Guo et al.’s study is used to analyse the distribution of the hydraulic head under the influence of the pore water pressure around the twin shallow circular tunnels at different locations, and the corresponding location division is shown in Fig. 15. We adopt the proposed method to determine the pore water pressure and corresponding hydraulic head in this case and compare the results obtained via the proposed method with those obtained via Guo et al.’s method and numerical simulation, as shown in Fig. 15, and the effective verification of the accuracy of the calculation method proposed in this study is achieved.Fig. 15Comparison of the results for the hydraulic head around the twin shallow circular tunnels at different locations. (a) Line 1; (b) Line 2; (c) Line 3; and (d) Line 4. Comparison of the results for the hydraulic head around the twin shallow circular tunnels at different locations. (a) Line 1; (b) Line 2; (c) Line 3; and (d) Line 4. Figure 15 shows that the distribution pattern of the hydraulic head around the twin shallow circular tunnels is different at different locations. The theoretical calculation results obtained by the proposed method fit well with the theoretical calculation results obtained by Guo et al.’s method, and both are similar to the variation trend of the numerical simulation results. However, the theoretical values obtained by the proposed method are slightly larger than the theoretical values obtained by Guo et al.’s method, and the theoretical calculation results obtained by Guo et al.’s method better fit the numerical simulation results. The average error between the theoretical calculation results obtained via the proposed method and the numerical simulation results is 1.5%, and the average error between the theoretical calculation results obtained via Guo et al.’s method and the numerical simulation results is only 0.5%. Therefore, the calculation accuracy of Guo et al.’s method is greater than that of the calculation method proposed in this study. However, the overall variation trend of the theoretical calculation results obtained via the proposed method is the same as those of the theoretical calculation results obtained via Guo et al.’s method and the numerical simulation results, and the average error of the theoretical calculation results obtained via the proposed method is only 1.5%. The results obtained via the proposed method are very close to the results obtained via Guo et al.’s method; thus, the calculation accuracy of the method proposed in this study is also high, and this method can be used to accurately calculate the pore water pressure around twin shallow circular tunnels and provide reliable ranges of parameter values for the design and construction of similar projects. In addition, considering that the solution for the pore water pressure around the twin shallow circular tunnels belongs to the problem of bifocal symmetry, compared with the complex variable method introduced in the method proposed by Guo et al.33, the bipolar coordinate system method involved in the proposed method has the advantage of high efficiency and intuition in solving this type of problem. Therefore, for the semi-infinite plane problem of the twin shallow circular tunnels, which has a relatively simple geometric boundary in this study, the Schwarz alternating method is adopted to solve the distribution pattern of pore water pressure around the tunnels in the bipolar coordinate system, which provides a reliable theoretical guidance for the parameter selection in the design and construction of water-rich tunnel engineering. Discussion about the main influencing factors of the pore water pressure The distribution pattern of the pore water pressure around twin shallow circular tunnels is different from that around a single shallow circular tunnel because of the interaction between the tunnels. Therefore, the number of tunnels and the centre distance between twin tunnels are important factors affecting the pore water pressure around shallow circular tunnels. As a result, we adopt the control variable method to analyse the distribution pattern of the pore water pressure around shallow circular tunnels under the influence of the number of tunnels and the centre distance between twin tunnels. The arrangement of the calculation positions is shown in Fig. 16, where Tunnel I represents the left tunnel, and Tunnel II represents the right tunnel.Fig. 16Arrangement of the calculation positions of the pore water pressure around single and twin shallow tunnels. Arrangement of the calculation positions of the pore water pressure around single and twin shallow tunnels. Influence of the number of tunnels According to the calculation method proposed in this study, the number of tunnels is one of the main influencing factors of the pore water pressure. Therefore, we calculate the pore water pressure around single and twin shallow circular tunnels, which are common in shallow buried tunnelling projects, sequentially, and the other parameters are kept constant, as shown in Table 2. According to Fig. 16, the distribution pattern of the pore water pressure at the corresponding four calculation positions, which are shown in Fig. 16, is obtained, as shown in Fig. 17.Table 2Parameters of the case study.ParameterValueGround surface hydraulic head hw (m)30Hydraulic head in the surrounding rock h1 (m)0Tunnel centre depth h (m)20Tunnel radius r (m)5Centre distance between the twin tunnels L (m)20Underground water gravity g (kN/m3)9.8Fig. 17Comparison of pore water pressure under various numbers of tunnels. (a) Pore water pressure at position 1, (b) pore water pressure at position 2, (c) pore water pressure at position 3, and (d) pore water pressure at position 4. Table 2Parameters of the case study.ParameterValueGround surface hydraulic head hw (m)30Hydraulic head in the surrounding rock h1 (m)0Tunnel centre depth h (m)20Tunnel radius r (m)5Centre distance between the twin tunnels L (m)20Underground water gravity g (kN/m3)9.8 Parameters of the case study. Comparison of pore water pressure under various numbers of tunnels. (a) Pore water pressure at position 1, (b) pore water pressure at position 2, (c) pore water pressure at position 3, and (d) pore water pressure at position 4. Figure 17 shows that the distribution pattern of the pore water pressure at position 1 when Tunnel I and Tunnel II are both present is significantly different from that when only Tunnel I is present. This fully reflects the influence of the interaction between the twin tunnels on the distribution pattern of the pore water pressure. When only Tunnel I is present, the pore water pressure at position 1 increases gradually as it moves away from the centre of the tunnel, but the increasing trend gradually slows. For both Tunnels I and II, the pore water pressure at position 1 tends to be symmetrical about the mid-penumbra of the line connecting the centres of Tunnel I and Tunnel II, which is consistent with the symmetry of the shallow buried bilinear circular tunnels with respect to the vertical direction. At positions 2, 3 and 4, the distribution patterns of the pore water pressure are generally consistent when only Tunnel I is present and when both Tunnel I and Tunnel II are present. At position 2, the pore water pressure decreases as the centre of the tunnel is approached, and the decreasing trend gradually increases. At position 3, the pore water pressure gradually increases with proximity to the surface, and the increasing trend gradually decreases. At position 4, the pore water pressure gradually increases with burial depth, but the growth trend also gradually flattens. Therefore, the surrounding pore water pressure is negatively correlated with the spacing at the centre of the tunnel for both single and twin shallow circular tunnels. However, the pore water pressure tends to decrease and level off as it moves away from the centre of the tunnel. In the upper tunnel region, the pore water pressure around the twin tunnels is slightly greater than that around a single tunnel. In addition, the pore water pressure around the twin tunnels is lower than the pore water pressure around the single tunnel at other computational locations, and this conclusion can especially be seen in Fig. 17c. In the region below the tunnels, the pore water pressures around the single and twin tunnels gradually converge with depth. Therefore, the distribution pattern of the pore water pressure around a single tunnel is different from that around twin tunnels, and the distribution curves of the pore water pressure for both tunnels cannot achieve a good fit at positions 1–4. Influence of the centre distance between the twin tunnels According to the calculation method proposed in this study, the centre distance between the twin tunnels is one of the main influencing factors of pore water pressure. Therefore, we set the centre distance L between the twin tunnels to 15 m, 20 m, 25 m, and 35 m in turn while keeping the other parameters constant. There are four groups of working conditions, and the remaining parameters are shown in Table 2. According to Fig. 16, the pore water pressure around the twin shallow circular tunnels under the influence of the centre distance between the twin tunnels is calculated, and the distribution patterns of the pore water pressure at the corresponding four calculation positions under different working conditions are shown in Fig. 18.Fig. 18Comparison of pore water pressure under the influence of the centre distance between twin tunnels. (a) Pore water pressure at position 1, (b) pore water pressure at position 2, (c) pore water pressure at position 3, and (d) pore water pressure at position 4. Comparison of pore water pressure under the influence of the centre distance between twin tunnels. (a) Pore water pressure at position 1, (b) pore water pressure at position 2, (c) pore water pressure at position 3, and (d) pore water pressure at position 4. As shown in Fig. 18, the overall trend of the pore water pressure around the twin shallow circular tunnels at different calculation positions is that the pore water pressure gradually increases with increasing distance from the centre of the tunnel, but this trend gradually slows. However, when the centre distance L between the twin tunnels increases from 15 to 35 m, the change rules of the pore water pressure at different calculation positions differ. At position 1, the pore water pressure around the twin shallow circular tunnels gradually increases with increasing centre distance L in the twin tunnels, and the two are positively correlated. At positions 2–4, the pore water pressure around the twin shallow circular tunnels decreases gradually with increasing centre distance L in the twin tunnels, and the two are negatively correlated. In addition, a comparison of four different centre distances in the twin tunnels reveals that the pore water pressure gradually increases with increasing centre distance between the twin tunnels at the calculation position 1. The maximum value of the pore water pressure is at the midpoint of the centreline of the twin tunnels. At calculation position 2, the pore water pressure gradually decreases as the centre distance between the twin tunnels increases. The pore water pressure around the twin tunnels at different centre distances gradually converges with increasing centre distance. Therefore, at different locations, the centre distance between the twin tunnels has different effects on the pore water pressure, but the change rule of the pore water pressure at each location is approximately the same. Conclusions In this study, the bipolar coordinate system method and the Schwarz alternating method are used to calculate the pore water pressure around twin shallow circular tunnels in a semi-infinite space, and a new method for determining the pore water pressure around twin shallow circular tunnels is proposed. We further use this method to analyse the main influencing factors of pore water pressure. The main conclusions of this study are as follows:A method for calculating the pore water pressure around twin shallow circular tunnels in a semi-infinite space is proposed for the first time by using the bipolar coordinate system method and the Schwarz alternating method. By counteracting the interaction between the twin tunnels through a number of iterative cycles, the pore water pressure around the twin shallow circular tunnels can be obtained, which provides a new approach for solving for the pore water pressure under steady seepage conditions.The error of the Schwarz alternating method can be controlled by setting the calculation accuracy so that the effective solutions for the design and construction parameters in the corresponding cases can be realized according to the engineering error requirements. The calculation accuracy of the proposed method is set to 1.0 × 10–3 MPa to ensure the theoretical calculation results fitted well with the numerical simulation results, and the maximum error between the two is only 2.15%, which verifies the good rationality and high accuracy of the method proposed in this study.Compared with that of an existing method, the efficiency of the proposed calculation method is greater. When the calculation accuracy is set to 1.0 × 10–3 MPa, the relative error between the calculation results obtained by the proposed method and the existing method is only 1.0%, which meets the error requirements specified by engineering experience. The theoretical calculation results obtained via the proposed method are highly consistent with the numerical simulation results and the theoretical calculation results obtained via the existing method, which reflects the greater accuracy of the proposed method, and can provide an accurate range of values for the parameters of actual engineering design and construction.The pore water pressure around a single shallow tunnel increases gradually with the distance from the centre of the tunnel increasing. In the intermediate region of the twin shallow circular tunnels, the pore water pressure shows a symmetrical distribution about the middle vertical line connecting the centres of the twin tunnels. In other regions, the influence of the number of tunnels on the pore water pressure is small. The change curves of the pore water pressure under different numbers of tunnels are well fitted, and all of them gradually increase with increasing distance from the tunnel centres, but the growth trend gradually decreases. In addition, the pore water pressure is positively correlated with the centre distance between the twin tunnels in the intermediate region of the twin shallow circular tunnels. The pore water pressure gradually increases with increasing distance between the centres of the twin tunnels, but the growth trend gradually decreases. While the two are negatively correlated in other regions. A method for calculating the pore water pressure around twin shallow circular tunnels in a semi-infinite space is proposed for the first time by using the bipolar coordinate system method and the Schwarz alternating method. By counteracting the interaction between the twin tunnels through a number of iterative cycles, the pore water pressure around the twin shallow circular tunnels can be obtained, which provides a new approach for solving for the pore water pressure under steady seepage conditions. A method for calculating the pore water pressure around twin shallow circular tunnels in a semi-infinite space is proposed for the first time by using the bipolar coordinate system method and the Schwarz alternating method. By counteracting the interaction between the twin tunnels through a number of iterative cycles, the pore water pressure around the twin shallow circular tunnels can be obtained, which provides a new approach for solving for the pore water pressure under steady seepage conditions. The error of the Schwarz alternating method can be controlled by setting the calculation accuracy so that the effective solutions for the design and construction parameters in the corresponding cases can be realized according to the engineering error requirements. The calculation accuracy of the proposed method is set to 1.0 × 10–3 MPa to ensure the theoretical calculation results fitted well with the numerical simulation results, and the maximum error between the two is only 2.15%, which verifies the good rationality and high accuracy of the method proposed in this study. The error of the Schwarz alternating method can be controlled by setting the calculation accuracy so that the effective solutions for the design and construction parameters in the corresponding cases can be realized according to the engineering error requirements. The calculation accuracy of the proposed method is set to 1.0 × 10–3 MPa to ensure the theoretical calculation results fitted well with the numerical simulation results, and the maximum error between the two is only 2.15%, which verifies the good rationality and high accuracy of the method proposed in this study. Compared with that of an existing method, the efficiency of the proposed calculation method is greater. When the calculation accuracy is set to 1.0 × 10–3 MPa, the relative error between the calculation results obtained by the proposed method and the existing method is only 1.0%, which meets the error requirements specified by engineering experience. The theoretical calculation results obtained via the proposed method are highly consistent with the numerical simulation results and the theoretical calculation results obtained via the existing method, which reflects the greater accuracy of the proposed method, and can provide an accurate range of values for the parameters of actual engineering design and construction. Compared with that of an existing method, the efficiency of the proposed calculation method is greater. When the calculation accuracy is set to 1.0 × 10–3 MPa, the relative error between the calculation results obtained by the proposed method and the existing method is only 1.0%, which meets the error requirements specified by engineering experience. The theoretical calculation results obtained via the proposed method are highly consistent with the numerical simulation results and the theoretical calculation results obtained via the existing method, which reflects the greater accuracy of the proposed method, and can provide an accurate range of values for the parameters of actual engineering design and construction. The pore water pressure around a single shallow tunnel increases gradually with the distance from the centre of the tunnel increasing. In the intermediate region of the twin shallow circular tunnels, the pore water pressure shows a symmetrical distribution about the middle vertical line connecting the centres of the twin tunnels. In other regions, the influence of the number of tunnels on the pore water pressure is small. The change curves of the pore water pressure under different numbers of tunnels are well fitted, and all of them gradually increase with increasing distance from the tunnel centres, but the growth trend gradually decreases. In addition, the pore water pressure is positively correlated with the centre distance between the twin tunnels in the intermediate region of the twin shallow circular tunnels. The pore water pressure gradually increases with increasing distance between the centres of the twin tunnels, but the growth trend gradually decreases. While the two are negatively correlated in other regions. The pore water pressure around a single shallow tunnel increases gradually with the distance from the centre of the tunnel increasing. In the intermediate region of the twin shallow circular tunnels, the pore water pressure shows a symmetrical distribution about the middle vertical line connecting the centres of the twin tunnels. In other regions, the influence of the number of tunnels on the pore water pressure is small. The change curves of the pore water pressure under different numbers of tunnels are well fitted, and all of them gradually increase with increasing distance from the tunnel centres, but the growth trend gradually decreases. In addition, the pore water pressure is positively correlated with the centre distance between the twin tunnels in the intermediate region of the twin shallow circular tunnels. The pore water pressure gradually increases with increasing distance between the centres of the twin tunnels, but the growth trend gradually decreases. While the two are negatively correlated in other regions.