Various existing opinions for evaluating the tunnel face stability in soft soil in the tunnelling industry were the inspiration for publication of this document. The recommendations aim to offer a guide to the available methods for tunnel face stability assessment in mechanised tunnelling and to provide a best practice guideline for face support pressure calculations.
Furthermore, the recommendations intend to help with the choice between available calculations methods depending on the expected ground conditions. In the recommendations, it is also specified the difference between two tasks, which are sometimes mixed.
These tasks are:
- Face stability calculation
- Analysis of machine ground interaction to evaluate surface settlements
The recommendations consist of seven chapters (excluding the introduction and reference chapters). The second chapter introduces the general aim of a tunnel face stability assessment and briefly describes soft ground mechanised tunnelling technology, with a focus on tunnel face stability. The German safety concept regarding the face stability assessment is outlined in the third chapter. Following this, the most important scientific approaches dealing with face stability calculations are discussed in chapter four focusing on support pressure due to earth pressure at the tunnel face and in chapter 5 summarising support pressure due to groundwater pressure. The most relevant calculation methods for shield tunnelling practice are presented in detail in chapter 6. Additional aspects regarding face stability are discussed in chapter 7. Two examples of face support pressure calculations are provided at the end of this document in chapter 8.
The recommendations can be downloaded in English through the following link: http://www.daub-ita.de/fileadmin/documents/daub/gtcrec1/ gtcrec10.pdf
Following abstracts are providing a short overview about the content of the published recommendations.
Viewpoints on Face Stability Calculations
The aim of a tunnel face stability assessment is to investigate groundwater pressure and earth pressure acting at the tunnel face, and to analyse the bearing capacity of the tunnel face. If the self-bearing capacity of the tunnel face is insufficient, tunnel face support has to be provided. For this, the support medium must counter the earth and groundwater pressures to stabilise the tunnel face.
Two fundamental points of view are called on for tunnel face support design. With the first viewpoint, the tunnel face pressure calculations deal solely with tunnel face stability. These types of calculations are the midpoint of the document. The calculations do not consider the development of ground deformations as a relevant criterion when calculated support pressure is applied to the tunnel face. This approach is called the “Ultimate Limit State Approach” since only a minimal required face support pressure is determined to avoid a tunnel face collapse.
The second point of view focuses on keeping the ground deformation below a pre-determined limit. Thus, this defines the support pressure (and consequently the tail void grouting pressure) based on the required ground deformation limit. This approach may be termed the “Serviceability Limit State Approach” since it considers ground deformation during excavation to be the main design criterion.
German Safety Concept in Face Stability Calculation
Two operational limits for the support pressure are defined in the German regulation ZTV-ING (2012): the lower and the upper limit. It is to note that RiL 853 (2013) references ZTVING regarding support pressure calculations. The lower support pressure limit (Fig. 1) has to ensure a minimal support force (Sci), which consists of two components and their respective safety coefficients (Eq. 1). The first component of the support force (Emax,ci) has to balance the earth pressure, and is calculated here based on active kinematical failure mechanism of the tunnel face. The second component of the support force (Wci) has to balance the groundwater pressure and is determined based on the elevation of groundwater level above the tunnel crown.
The upper support pressure limit (scrown,max) is defined as a limiting pressure to avoid a break-up of the overburden or blow-out of the support medium (Eq. 2). Therefore, the maximal support pressure has to be smaller than 90 per cent of the total vertical stress at the tunnel crown 
. It is necessary to point out, passive failure of the tunnel face due to high support pressure is unlikely to happen. Before reaching this limit, support medium will blow-out from the excavation chamber.
Break-up/blow out safety:
Overview of Calculation Methods
Various methods to determine the required support pressure due to the acting earth pressure can be found in the literature and are presented in the recommendations. All available approaches calculation approaches can be divided into four fundamental groups:
- Analytical methods
- Empirical methods
- Experimental methods
- Numerical methods
In this short overview, however, the focus will be given only to analytical and experimentally/empirical methods. The group of analytical methods includes limit equilibrium and limit state methods. These methods assume a possible failure mechanism of the tunnel face or a stress distribution in the ground respectively and from that determine a support pressure at collapse. Common features of most of the analytical methods are based on the adoption of two widely used laws of failure in soil mechanics.
On one hand, the Mohr-Coulomb law of failure is broadly adopted for frictional or frictional–cohesive materials where the associated flow rule is dominating among the formulations. On the other hand, Tresca law of failure (associated) is mostly applied for purely cohesive materials.
The limit equilibrium methods can be characterised by the required assumption of a kinematical failure mechanism of the tunnel face. The first limit equilibrium failure mechanism was suggested by Horn (1961) and assumes a sliding wedge in front of the tunnel face that is loaded by a rectangular prism stretching up to the terrain surface level (Figure 1). This sliding wedge mechanism for investigation of tunnel face stability was introduced to mechanised tunnelling by Anagnostou & Kovari (1994) and Jancsecz & Steiner (1994). Additional improvements of this method have been conducted by Anagnostou (2012) or Hu et al. (2012). On the sliding mechanism, the equilibrium condition for acting forces is formulated and required support force is determined.
Other solutions for the tunnel face stability have been formulated based on bound theorems within the plasticity theory. This group of approaches are often known as “limit state methods”. The solution for the tunnel face stability can be obtained by adopting the upper or the lower bound of the plasticity theory. The bound theorems were employed in methods suggested by Davis et al. (1980), Leca & Dormieux (1990) or Mollon et al. (2010).
The stability ratio method by Broms & Bennermark (1967) represents the most important calculation approach from experimentally/empirical methods. Broms & Bennermark (1967) conducted laboratory experiments that investigated the extrusion stability of a clayey soil through a vertical circular opening in a sheet pile wall. The stability ratio depends on the subtraction of the pressure supporting the opening from the vertical stress at the opening axis then divided by the undrained shear strength (cu) of the soil. Consequently, they suggested to apply the developed method for investigation of tunnel heading stability.
Depending on the soil type, the soil can show drained or undrained behaviour during excavation, stoppage or standstill.
Therefore, particular calculation methods are fitted to certain ground conditions. The theoretically calculated pressures at tunnel face collapse were mostly not validated in practice with real scale experiments. Only laboratory experiments were used to validate the particular calculation methods. For drained conditions, the best fit between theoretical calculations and experiments was found for limit equilibrium methods formulated by Anagnostou & Kovari (1994) or Jancsecz & Steiner (1994) and the limit state upper bound solution by Leca & Dormieux (1990). However, the upper bound calculations become complicated when the failure mechanism is close to the reality.
Thus, various formulations of limit equilibrium on the sliding wedge are commonly used in practice. Note that for drained conditions, the output of models represents only the required support force due to earth pressure (Eq. 1).
For undrained behaving soils, the lower bound limit state method assuming a cylindrical tunnel model was the best predictor of the support pressures at collapse (Davis et al., 1980). The authors determined required support pressures by using the stability ratio approach (Broms & Bennermark, 1967). This calculation method is very straightforward, and its concepts have become a standard for excavations in purely cohesive soils with undrained behaviour.
The Most Important Recommendations for Calculation Practice
Beside non-cohesive soils, the limit equilibrium method is employed in practice also typically when cohesive and non-cohesive soil layers are alternating at the tunnel face. The effective (drained) shear parameters of a soil are assumed in this case. It is generally not recommended to use the limit equilibrium approach for calculations with undrained shear soil parameters. Further, it is necessary to note that limit equilibrium approach delivers a minimum support pressure at which the ground theoretically becomes unstable. It is achieved only due to the application of safety factors that ground deformations are relatively acceptable when obtained support pressure is applied.
Since the adopted safety factors are generalized for every soil, the amount of obtained displacement is varying based on the actual stiffness of ground.
The minimal calculated support pressures exhibit relatively wide scattering in non-cohesive soils depending on the adopted formulation of the limit equilibrium approach. The scattering was highlighted by Vu et al. (2015) or Kirsch (2009). In cohesive frictional soils, the size of calculation scattering will decrease.
It is recommended to ensure that any assumptions used to calculate variables in limit equilibrium methods are consistent with regard to the adopted mobilisation of soil´s shear resistance.
For face stability calculations in undrained conditions using stability ratio method, the assumptions for critical stability ratio and amount of undrained soil cohesion are key factors that determine the results. The critical stability ratio should be adopted based on a case by case basis and local experience.
Furthermore, the amount of undrained cohesion should be conservatively assessed. Nevertheless, the critical stability ratio method often shows the acting groundwater pressure as the decisive factor for support pressure calculations.
When designing the support pressure for excavation in the particular ground conditions, it is necessary to consider also the extent of possible consequences such as failures or settlements, due to an inappropriate design. The extent of the worst case scenario determines how conservative the calculation approach should be. A general rule of thumb is that inappropriate support pressure may lead to immediate failure up to the surface for a combination of slurry shields and non-cohesive ground. With a combination of EPB shields and cohesive soil, it may lead “only” to extensive surface deformations without failure up to the surface. The height of the overburden plays a significant role for the extent of consequences. Thus, it is recommended for excavations in difficult ground conditions, e.g., in complicated weak ground or under surface constructions, that the analytical calculation of the support pressure should always be supplemented by the numerical analysis of machine – ground interaction.
Final important aspect of the calculation is the adopted support pressure deviations during excavation. Support pressure deviations must be considered in calculations, which are performed according to ZTV- ING (2012):
For the EPB shield, the range of deviation was defined to be larger because of a higher degree of uncertainty for the support pressure regulation. The deviations are added to the lower support pressure limit and subtracted from the upper pressure limit (compare Figure 1). However, the large range of deviations for an EPB shield may in some cases lead to a limited feasibility of EPB shield drives. Thus, EPB shield deviations may be reduced for special cases upon proper justification. Reduction of the deviations range should focus especially on the deviations at the upper pressure limit, since the risk of overburden breakup or support medium blow-out in the case of EPB is comparably lower. A good shield operation, process controlling and good design of excavation process (e.g., soil conditioning) are fundamentally required for such a reduction.
