Challenges of Tunnel Support in Low Overburden Zones in Urban Areas—Case Study

Abstract

This paper systematically analyzes the challenges of stabilizing tunnel excavations in zones with low overburden in urban environments, through an engineering-validated case study of the Kobilja Glava Tunnel. A combined approach involving the New Austrian Tunneling Method (NATM) and the pre-installation of steel pipe umbrellas was applied as the primary pre-support measure under complex geotechnical conditions. The design, drilling, grouting, and formation of the temporary support arch were thoroughly documented, along with the implementation of shotcrete, lattice girders, self-drilling anchors, and reinforcement meshes. A numerical analysis was performed using the PLAXIS 2D software package, encompassing the modeling of deformations, shear forces, axial forces, and bending moments, with precisely defined support parameters. Geodetic monitoring recorded maximum surface settlements of up to 70 mm at an overburden of less than 3 m, while deformations were reduced to 28 mm at an overburden of 20 m. The numerical model confirmed soil plasticization within a 3 m wide zone, with maximum displacements reaching 6.3 cm, consistent with field measurements. Calculated tensile strain and angular distortion were classified according to established building damage criteria, confirming minimal structural impact on adjacent buildings. The applied combination of the NATM and the pipe umbrella pre-support system proved to be an effective and reliable solution for controlling deformations and ensuring excavation stability under conditions of limited rock cover and dense urban development. The obtained results provide a verified framework and practical recommendations for future tunneling projects in similar geotechnical and urban conditions, aiming to enhance safety, stability, and cost-effectiveness.

1. Introduction

The rapid urban growth in modern times has increased the demand for the efficient utilization of underground spaces [1]. Tunnel construction is a complex and interdisciplinary process [2]. Modern tunnel design and construction require the use of appropriate techniques and technologies at all stages of execution [3]. Tunnel excavation is considered the most critical phase due to its potential consequences [4]. The construction of road tunnels ranks among the most demanding engineering tasks, requiring careful planning, advanced excavation methods, and the application of cutting-edge technological solutions to ensure structural stability and safety [5].

Tunnel excavation under shallow overburden conditions inherently induces surface settlement, particularly when conducted in soft soil strata or geologically unfavorable rock masses. Such settlements become critical when tunnels are constructed in densely populated urban environments, directly beneath residential buildings [6,7,8,9]. The support measures applied during excavation play a key role in controlling deformations and increasing the safety of tunnel works. Properly selected methods significantly contribute to the stability of underground structures [1,10]. Excavation methods such as open-cut excavations in combination with quasi-rectangular jacking boxes showed good adaptability to complex surrounding environments, providing the good ground control in soft working environments needed for excavation [11]. Implementation of the pipe umbrella method as a pre-support system in the construction of shallow tunnels contributes to the control of the subsidence and groundwater barrier, ensuring safe excavation underneath. It is important to control pipe installation deviations and to appropriately design spacing between the pipes of a sufficient diameter [12]. Knowledge transfer from different mining and tunneling projects is of great importance for guiding engineers both in the design and construction phases. It contributes to the appropriate selection of the excavation method, pre-support and support constructions, monitoring techniques, prediction models for measurable parameters during excavation, etc. [13,14,15,16].

This paper provides an overview of the support measures implemented during the excavation of the left tunnel tube in the low overburden section (chainage 4 + 132 to 4 + 043), as part of the construction of the Kobilja Glava Tunnel—a segment of the First Transversal, a strategic road connecting the center of Sarajevo with Vogošća, intended to alleviate city traffic congestion. Located in a densely populated urban zone with approximately 500 residential buildings, this twin-tube tunnel passes through the hill of the same name, facing complex geotechnical challenges due to the extremely shallow overburden above the structure [17]. The total length of the right tunnel tube is approximately 635 m, while the left tube is about 639 m.

Tunnel construction in urban areas, especially in close proximity to residential and infrastructure facilities, requires careful planning and strict control to prevent unwanted ground settlements and damage to nearby structures [18]. Any excavation in rock that creates a void disturbs the initial stress state of the rock mass, causing a redistribution of forces and the development of a secondary stress field around the tunnel gallery [19]. These changes lead to the propagation of deformations beyond the immediate excavation area, potentially causing displacements and stresses in surrounding structures. Depending on the geotechnical characteristics of the rock mass, such effects can compromise the stability of adjacent buildings, making the timely and adequate application of support measures crucial for the safety of the entire system [20]. The geographical location of the Kobilja Glava Tunnel on the Sarajevo–Vogošća route is shown in Figure 1. The coordinates (WGS84) of the western tunnel portals area are 43.886089 N and 18.383092 E.

2. Engineering–Geological Characteristics

The geological profile along the Kobilja Glava Tunnel alignment is relatively simple, consisting exclusively of Neogene sediments and Quaternary deposits. The Neogene sequence is represented by Upper Miocene (1M₃) sediments of the Sarajevo–Zenica Basin, also known as the Koševo Series.

The excavation zone of the left tunnel tube with shallow overburden (H < 2D), as shown in the geological profile (Figure 2), is characterized by a distinctly stratified structure of soil and rock. The upper horizons are dominated by weakly consolidated clayey and sandy sediments, while the deeper parts of the profile consist of more solid marl and sandstone layers.

The total thickness of the overburden above the tunnel axis is relatively small, classifying this zone as a low overburden area. This geometry creates challenging conditions for construction, as it reduces the natural geostatic pressure that contributes to the stability of the overlying material [21].

In Figure 3, the cross-section of the exploratory borehole B-5 is presented.

According to UCS Classification, marls of horizon 5a belong to groups of lean clay (CL-1) and fatty clay (CH-1). The natural volumetric weight is at an average of 21.0 kN/m³. Natural water content is on average 16%. The plasticity index is on average 20.2%. The consistency index is on average 0.99, while the monoaxial strength value is on average 2250 kPa. The test of shear resistance characteristics was performed in a direct shear apparatus, on samples with dimension of 6.05 cm × 6.05 cm and the average internal friction angle was 24^° while average cohesion was 32 kPa.

3. Case Study

The construction of the Kobilja Glava Tunnel, as defined by the main project [22], is to be carried out using the New Austrian Tunneling Method (NATM). This method is based on the principle of maximizing the load-bearing capacity of the rock mass or soil, with minimal primary support applied immediately after excavation.

A key concept of the NATM is the continuous observation of rock mass behavior and the flexible adaptation of support measures according to actual field conditions [23]. Fundamentally, the method involves the use of shotcrete, lattice girders, steel mesh, and self-drilling anchors, combined with geotechnical monitoring that enables deformation control and optimization of stabilization measures [24].

The NATM has proven effective in complex geological conditions, especially in zones with variable lithological composition and minimal overburden, as is the case along the Kobilja Glava Tunnel route. By applying flexible support measures and continuously monitoring rock mass behavior, the method maintains ground equilibrium, controls stress redistribution, and reduces settlement risk. This adaptability is crucial for the safe and efficient construction of tunnels in urban environments with complex geotechnical conditions [25,26,27,28]

3.1. Pipe Umbrella as Method of Roof Support Prior to Excavation

A critical factor contributing to the enhanced adaptability of the New Austrian Tunneling Method (NATM) in soft ground and minimal overburden conditions is the systematic implementation of pipe umbrella systems for pre-support during excavation [29]. This technique provides a viable and efficient solution to the associated geotechnical challenges, as it avoids surface excavation and minimizes interference with surface infrastructure and traffic flow [12].

The pipe umbrella method involves installing a series of steel pipes to form a temporary roof that supports the underground space beneath it. This technique is widely used for tunnel excavation under shallow soil cover [30]. The pipe umbrella is constructed in phases by installing steel pipes through boreholes ahead of the tunnel face, creating a protective arch just above the tunnel axis. This configuration enables load redistribution, reduces stresses in the tunnel crown, and stabilizes the excavation face—factors that are particularly important under shallow overburden conditions [31,32,33].

Figure 4 illustrates the roof support method using the pipe umbrella prior to tunnel excavation in the low overburden zone, while Figure 5 shows the flowchart summarizing the sequential steps in the pipe umbrella installation process used for pre-support in low overburden tunnel excavation.

As shown in Figure 5, the flowchart for the sequential steps consists of the following steps. After evaluation of ground conditions and identification of zones with low overburden or soft soil, a proper design of the pipe umbrella system is required (determining pipe length, spacing, inclination, and installation sequence based on geotechnical data). When previous conditions are met, the drilling of a series of boreholes along the designed alignment is performed using a drilling jumbo. When all boreholes are drilled, steel pipes are installed in them and grouting around the pipes to fill voids is applied. The next step is the formation of a temporary roof (umbrella arch) from preinstalled pipes which help in the redistribution of loads and the stabilization of the excavation zone beneath the pipe umbrella.

3.2. Support Measures During the Excavation of the Left Tunnel Tube from the Portal

The excavation of the Kobilja Glava Tunnel required the implementation of advanced and carefully designed support measures to ensure the stability of the tunnel profile and minimize deformations within the rock mass. One of the key solutions was the installation of steel pipes with a diameter of Ø114 mm and a length of 21 m, arranged in two rows with an overlap of 5 m (Figure 6). The boreholes were drilled with an enlarged diameter to allow the formation of a thicker grout layer, which increased the efficiency of cutting potential slip planes and further reduced the load on the left tunnel face slope [34].

The stabilization of the tunnel excavation involved the use of several support elements: shotcrete, self-drilling anchors (IBOs), reinforcing mesh, and lattice girders. These components proved highly effective in preserving the integrity of the tunnel subgrade. Particularly noteworthy was the use of Ø51 mm IBO anchors with a capacity of 530 kN, installed in the tunnel roof and bench according to a precisely defined layout. These anchors significantly contributed to stabilizing the excavation face and reducing the risk of uncontrolled rock mass movements.

Shotcrete of class C25/30 was applied in layers of varying thickness—5 cm for the initial stabilization of the excavation face and 30 cm for the roof, bench, and invert—achieving effective deformation control and increasing the load-bearing capacity of the subgrade. For additional protection of the excavation face, Q257 reinforcing mesh was installed at every third excavation step, while Ø32 mm IBO anchors, 12 m long, were used to further reinforce the face zone.

For the validation of the adopted support construction, numerical analysis has been conducted using PLAXIS 2D software package. Modeling parameters for shotcrete were selected as follows: thickness d = 0.3 m, axial stiffness EA = 4.5 × 10⁶ kN/m, flexure stiffness EI = 15.0 × 10⁶ kNm²/m, and Poisson’s ratio ν = 0.2. Modeling parameters for anchors were selected as follows: diameter d = 0.051 m, bolt spacing L = 0.8 m, Young’s modulus E = 72.44 × 10⁶ kN/m², and unit weight γ = 37.23 kN/m³. Deformations, shear forces, axial forces, and bending moments for the chosen support construction are presented in Figure 7.

To limit deformation of the surrounding material and the primary support, it was essential to strictly adhere to the technological sequence of the works, ensuring that the distance between the excavation face and the temporary invert did not exceed 2 m.

Compared with practices in other tunnels constructed under similar geological conditions, the applied support measures proved to be an optimal solution for excavation stabilization and effective deformation control. Figure 8 provides an overview of the support measures used for roof stabilization in the low overburden zone of the left tunnel tube.

The cross-section of the tunnel with the support type is presented in Figure 9, while the scheme of the anchors is presented in Figure 10.

The longitudinal section of the excavating steps is shown in Figure 11.

3.3. Methods of Predicting Surface Deformations

Considerable research efforts have focused on improving the prediction of soil behavior in response to stress changes induced by tunneling, as well as on developing technically demanding mitigation strategies to address these challenges [35]. Owing to the inherent characteristics of the tunneling process, the actual excavated profile typically exceeds the theoretical design, resulting in a series of ground displacements directed toward the excavated spaces. This phenomenon is commonly referred to as ground loss or, more precisely, volume loss [36]. According to Peck [36], surface settlement arises from both radial displacements within the tunnel cross-section and longitudinal deformations along the tunnel axis (Figure 12) [7].

A range of methodologies has been developed for the accurate prediction and management of maximum surface settlement, including empirical, analytical, numerical, and artificial intelligence-based approaches [37]. Conventional methods typically rely on empirical correlations or analytical formulations derived from engineering experience and theoretical assumptions. One of the earliest investigations into the surface settlement profile above underground tunnel structures was conducted by Martos [38], who analyzed the initial shape and distribution of the settlement.

Martos [38] proposed that the surface settlement profile above a tunnel can be effectively represented by a normal (Gaussian) distribution curve, as illustrated in Figure 10. Building upon this concept, Peck [36] extended this research by demonstrating the suitability of the Gaussian curve for modeling ground settlement induced by tunnel construction. Through the analysis of settlement data from numerous underground construction projects, he established that the settlement profile is approximately symmetrical about the vertical axis of the tunnel. Based on these findings, a semi-empirical approach was developed to calculate ground settlement, expressed by the following expression:

$${\mathrm{S}}_{\mathrm{V}}={\mathrm{S}}_{\mathrm{m}\mathrm{a}\mathrm{x}}\mathrm{e}\mathrm{x}\mathrm{p}\left(-{\displaystyle \frac{{\mathrm{x}}^2}{{2\mathrm{i}}^2}}\right),$$

(1)

where S_V represents subsidence value, S_max represents theoretical maximum subsidence along the tunnel centerline, x represents lateral distance from the tunnel centerline, and i represents lateral distance from the tunnel centerline to the inflection point in the Gaussian distribution curve.

Empirical equations for estimating surface settlement induced by tunnel construction remain useful for preliminary assessments [39]. However, in complex geological or structural conditions, advanced numerical analyses—particularly finite element analysis (FEA) and soil–structure interaction studies—are recommended to obtain more accurate results [40]. Numerical simulation techniques, such as the finite element method (FEM) and the finite difference method (FDM), provide a more flexible and sophisticated modeling framework. These methods can accommodate a wide range of geotechnical parameters and effectively capture the complex interactions between soil strata and tunnel linings [7]. Owing to their theoretical rigor and adaptability, numerical approaches are widely applied in geotechnical engineering and are regarded as a more realistic methodology for predicting surface settlements [41].

The modeling parameters of the soil used for numerical analysis in Plaxis software are presented in

Table 1. Modeling parameters of soil.

ID	Material	Model	Drainage	Unit Weight γ (kN/m3)	Oedometer Modulus Eoed (kN/m3)	Unloading/Reloading Stiffness Eur (kN/m3)	Cohesion c (kN/m3)	Internal Friction Angel φ (°)
1	El-dl	Hardening soil	Drained	19	30,000	120,000	9.50	23
2	Marl	Hardening soil	Drained	22	40,000	120,000	30	55

.

4. Results and Discussion

Numerical analyses were conducted using the finite element method with Plaxis software. The geometry of the model is presented in Figure 13, while the geometry of the excavated model is shown in Figure 14.

The Plaxis package indicated expected displacements of the left tunnel tube of up to 6.30 cm (Figure 15), with the zone of plastified material around the tunnel confined within a width of 3.0 m (Figure 16). These results were crucial for validating the stability of the tunnel subgrade and for optimizing subsequent excavation phases.

For broader validation of the numerical results, a comparison was conducted between the measured surface settlements and the empirical model proposed by Peck [36], which approximates the settlement profile above a tunnel using a Gaussian distribution, as expressed by Equation (1). For the purposes of this analysis, the maximum settlement was taken as S_max = 70 mm, while the parameter representing the width of the zone of influence was set to i = 8 m, corresponding to the geotechnical conditions in an area of shallow overburden. The calculated settlement values for distances up to ±10 m from the tunnel axis are presented in

Table 2. Calculated settlement values according to Peck model (S_max = 70 mm and i = 8 m).

Distance From the Tunnel Axis, x (m)	Calculated Settlement, S(x) (mm)
–10	43.94
–9	48.94
–8	53.59
–7	57.73
–6	61.30
–5	64.26
–4	66.56
–3	68.17
–2	69.06
–1	69.41
0	70.00
+1	69.41
+2	69.06
+3	68.17
+4	66.56
+5	64.26
+6	61.30
+7	57.73
+8	53.59
+9	48.94
+10	43.94

.

The diagram presented below (Figure 17) illustrates the distribution of surface settlements above the tunnel according to Peck’s empirical model. The diagram shows the maximum settlement occurring along the tunnel axis. This visual confirmation further supports the numerical results and demonstrates the reliability of the applied stabilization measures under conditions of shallow overburden.

Before the start of excavation, a geodetic monitoring system was established on the surface above the left tunnel tube to track ground settlement. Survey profiles were installed at intervals of approximately 10 m, with three markers placed on each profile—one directly above the tunnel axis and two laterally, to the left and right, at a distance of about 10 m from the axis. Monitoring was carried out using standard measuring equipment, specifically a total station.

Simultaneously with numerical modeling, deformations of the primary tunnel lining were monitored under real conditions. The monitoring system included a network of measurement-control profiles and reference points on the surface (Figure 18), enabling the reliable recording of settlements and timely assessment of the excavation’s impact on surface deformations.

Figure 19 shows the measured maximum settlements on the ground surface above the left tunnel tube in the low overburden zone, covering the range from the start of excavation at chainage 4 + 133 to chainage 4 + 043. The horizontal axis represents the chainages, while the vertical axis indicates the settlement values in millimeters.

The greatest settlements were recorded at the start of excavation (chainage 4 + 133), where the overburden height was less than 3 m, with settlement values reaching approximately 70 mm. Moving toward lower chainages, as the overburden thickness increased, a gradual decrease in deformations was observed. At chainage 4 + 043, where the overburden reached about 20 m, the settlement was approximately 28 mm.

By comparing the measured values along the alignment (e.g., 70 mm at station 4 + 130, 52 mm at station 4 + 120, etc.), it is confirmed that the actual data align well with the theoretical distribution. This correlation further validates the reliability of the numerical model and the effectiveness of the applied stabilization measures in the shallow overburden zone.

The measured settlements on the site closely matched the limits predicted by numerical modeling, with an accuracy of approximately 90%, confirming the reliability of the computational approach and the validity of the defined input parameters. The reduction in settlement with increasing overburden thickness can be explained by the redistribution of stresses through the surrounding rock mass. When the overburden height exceeds twice the tunnel diameter (H > 2D), the rock mass bears a significant portion of the load, acting as a natural supporting element. This mechanism, known as the arching effect, reduces the load on the tunnel lining and limits deformations [35,42,43,44].

For a precise assessment of the impact of tunnel excavation on surrounding structures, a settlement of 70 mm was further analyzed through calculations of maximum tensile strain (ε_t) and angular distortion (θ), with the results then correlated to damage classification schemes according to Boscardin and Cording [45] and Burland [46].

At the site corresponding to chainage 4 + 080, the first structure, located directly above the axis of the left tunnel tube, with a length of L = 10 m and a height of 10 m, was subjected to a deformation assessment based on the measured differential settlement of $∆ =$20 mm (0.02 m). The structure is a residential, masonry construction with reinforced foundations.

Angular distortion (θ) is defined as the ratio of the differential settlement of the structure $∆$ to its length L:

$$\mathsf{\theta }={\displaystyle \frac{∆}{\mathrm{L}}},$$

(2)

In our case, it is as follows:

$$\mathsf{\theta }={\displaystyle \frac{20}{\mathrm{10,000}}}=0.002={\displaystyle \frac{1}{500}}$$

(3)

According to the damage classification presented in

Table 3. Damage classification based on angular distortion (θ) [46].

Damage Category	Range of θ	Damage Description
Negligible	≤1/500	No or very fine cracks, deformations almost imperceptible
Minor	1/500–1/300	Superficial cracks, not significant for structural stability
Moderate	1/300–1/150	Visible cracks in walls
Severe	1/150–1/50	Structural cracks, repair required
Very severe/Critical	>1/50	Significant damage, risk to structural stability

, θ = 1/500 places the structure in the “Negligible/Minor” category, indicating minimal or superficial cracks with no significant impact on the structural stability.

Field observations and the calculated results confirmed that the risk of damage to the structure is minimal. The measured values of angular distortion (θ) and maximum tensile strain (ε_t) place the structure in the “Negligible/Minor” categories, which is consistent with the visual assessment. Damage is limited to very fine cracks, primarily on external surfaces such as sidewalks, with no impact on the load-bearing elements of the structure.

5. Conclusions

The construction of a tunnel under shallow overburden in a densely populated urban area, as in the case of the left tunnel tube of the Kobilja Glava Tunnel, represents a significant technical and safety challenge due to the limited thickness of the rock mass above the excavation, variable geomechanical properties, and the proximity of existing structures. This study demonstrates that careful selection of the excavation method and support measures, combined with continuous geotechnical monitoring, is crucial for controlling deformations and ensuring structural stability.

The application of the New Austrian Tunneling Method (NATM), complemented by additional roof support using a pipe umbrella, allowed for gradual advancement and adaptive adjustment to in situ conditions. Systematic pre-support using steel pipes with a diameter of Ø114 mm, in combination with shotcrete, self-drilling anchors, and reinforcement meshes, proved effective in reducing stresses in the tunnel crown, stabilizing the excavation face, and limiting surface settlements.

Numerical modeling conducted in Plaxis predicted displacements of up to 6.30 cm and plasticization of the soil within a zone approximately 3 m wide, while field measurements recorded maximum settlements of approximately 70 mm in areas of minimal overburden (<3 m). As the overburden increased toward chainage 4 + 043 (overburden approximately 20 m), settlements decreased to about 28 mm. The close agreement between modeling results and field monitoring confirms the reliability of the computational parameters and the effectiveness of the applied engineering solutions.

The observed trend of decreasing settlement with increasing overburden confirms the theoretically known arching effect, whereby for H > 2D, the rock mass carries a significant portion of the load, relieving the tunnel lining. This mechanism substantially contributes to stability and deformation control, which is particularly important in urban environments with strict requirements to limit the impact on surface structures.

In conclusion, the combination of the NATM with the targeted use of the pipe umbrella and precise settlement monitoring ensured the stability of the tunnel under extremely complex shallow overburden conditions. This integrated approach can serve as a reliable model for the design and construction of tunnels in similar geotechnical and urban settings.