|FHWA > Bridge > Tunnels > Technical Manual for Design and Construction of Road Tunnels - Civil Elements|
Technical Manual for Design and Construction of Road Tunnels - Civil Elements
Chapter 7 - Soft Ground Tunneling
7.4 Ground Loads and Ground-Support Interaction
The main objectives of tunnel support system are to (1) stabilize the tunnel heading, (2) minimize ground movements, and (3) permit the tunnel to operate over the design life. In general, the first two functions are provided by an initial support system, whereas the third function is preserved with a final lining.
The loading on the support system and its required capacity is dependent on when and how it is installed and on the loadings that will occur after it is installed. If the final lining is installed after the tunnel has been stabilized by initial support, the final lining will undergo very little additional loadings such as contact grouting pressures, thermal stresses, groundwater pressure, and/or time dependent loading (creep).
Generally, two types of loading have been considered to generate analytical solutions in tunneling in soil - overpressure loading and excavation loading. If a ground is assumed to be isolated and a pressure is applied to the upper surface, it is considered to be overpressure loading, where the support system is placed in the ground when it was unstressed and the lining and the ground is normally handled by applying lateral pressure to the ground.
Practically the support system is never placed in an unstressed ground, instead it is placed in the opening after the initial deformation has occurred, and before any additional deformation occurs and the additional deformation induces loading into the support system. This induced loading is called excavation loading.
The load developed on the support system (initial support and final lining) is a function of relative stiffness of the lining with respect to the soil (ground-lining interaction). Both analytical solutions and numerical methods have been commonly used by design engineers to evaluate the effect of the relative lining stiffness on the displacement, thrust, moments in the lining for various loading configurations. The available methods are summarized in this section. The reader should refer Chapter 10 for the final lining design practice.
7.4.2 Loads for Initial Tunnel Supports
This section presents a simplified system of determining the load on the initial support for circular and horseshoe tunnels in soft ground. These presented loads are patterned after Terzaghi's original recommendations (1950) but have been simplified. In all cases, it is important that the experience and judgment of the engineer also be applied to the load selection. Table 7-6 shows the loads recommended for design of initial tunnel supports in soft ground.
The vast majority of tunnels in soft ground are driven with modern tunneling machines and are, therefore, circular. However, some tunnels are still driven by hand and are often horseshoe or modified horseshoe in shape, for example, pump stations or cross passages between transit tunnels. Therefore the table also provides initial support recommendations for horseshoe tunnels.
The term tunnel liner actually should be broken into two concepts that have historically had distinct but related functions. Initial support is that support needed to make the soft ground tunnel opening stable and safe during the complete construction operation. It includes the gamut of support measures from reinforcement to grouting to freezing to shotcrete to ribs and boards to precast concrete segments and everything in between.
Final lining is the concrete or other lining placed to make the tunnel acceptable aesthetically and functionally, e.g., smooth to air or water flow, and to make the tunnel permanently stable and safe for its design life of 100 years or more.
While technically this distinction should still be made, with the advent of tunnel boring machines and high quality precast concrete lining systems (which are needed to propel the machines) this distinction is becoming blurred. For most modern tunnels a single lining of precast concrete segments is typically installed as the tunnel is advanced and used for both functions.
7.4.3 Analytical Solutions for Ground-Support Interaction
The state of stress due to tunnel excavation and interaction between rock support system and supporting ground were previously discussed in Chapter 6. The elastic formulations and interaction diagram discussed in Section 6.6.2 are also valid for a tunnel in soft ground.
Analytical solutions for ground-support interaction for a tunnel in soil are available in the literature. The solutions are based on two dimensional, plane strain, linear elasticity assumptions in which the lining is assumed to be placed deep and in contact with the ground (no gap), i.e., the solutions do not allow for a gap to occur between the support system and ground.
Early analytical solutions by Burns and Richard (1964), Dar and Bates (1974), and Hoeg (1968) were derived for the overpressure loading, while solutions by Morgan (1961), Muir Wood (1975), Curtis (1976), Rankin, Ghaboussi and Hendron (1978), and Einstein et. al. (1980) were for excavation loading. Solutions are available for the full slip and no slip conditions at the ground-lining interface. Appendix E present the available published analytical solutions in Table E-2, as well as the background (excerpt from FHWA Tunnel Design Guidelines published in 2004). Appendix E also presents a sample analysis is in Table E-3, for a 22ft diameter circular tunnel with 1.5 ft thick concrete lining. The tunnel is located at 105 ft deep from the ground surface to springline and groundwater table is located 10 ft below the ground surface. Details of input parameters are shown in Table E-3a. The calculated lining loads from various analytical solutions are presented in Table E-3b. The result of finite element analysis is shown in Figure 7-8.
7.4.4 Numerical Methods
Application of the analytical solutions is restricted when the variation of stress magnitude is significant with depth from the tunnel crown to the invert, such that assumptions made in the analytical solutions are not valid. Then, numerical method can be used to simulate support-ground interactions.
Numerical modeling has been driven by a perceived need from the tunneling industry in recent times. It has led to large, clumsy and complex numerical models. Properly performed numerical modeling will lead engineers to think about why they are building it - why build one model rather than another - and how the design can be improved and performed effectively.
An outline of the steps recommended for performing a numerical analysis for tunneling is as follow:
For the analysis of tunneling in soil, continuum analysis is generally accepted, where the domain can reasonably be assumed to be a homogeneous media. The continuum analysis includes Finite Element Method (FEM), Finite Difference Method (FDM), and Boundary Element Method (BEM). The details of numerical analysis softwares are discussed in Section 6.6.3. Sample loads on the concrete lining calculated by Finite Element analysis on a tunnel (Appendix E) are shown in Figure 7-8 . Figure 7-8 illustrates loads on the concrete liner including axial force, bending moment, and shear force calculated from two-dimensional, plain strain analysis.
Figure 7-8 Loads on a Concrete Lining Calculated by Finite Element Analysis: (a) Axial Force, (b) Bending Moment, (c) Shear Force
7.5 Tunneling Induced Settlement
Ground settlement is of greater concern for soft ground tunnels than for rock for two reasons:
With modern means and methods, both the designer and the contractor are now better equipped to minimize settlements and, hence, their impact on other facilities.
7.5.2 Sources of Settlement
Although there are a large number of sources or causes of settlement, they can be conveniently lumped into two broad categories: those caused by ground water depression and those caused by lost ground.
Groundwater Depression Groundwater depression may be caused by intentional lowering of the water during construction or by the tunnel itself (or other construction) acting as a drain. When either of these occurs the effective stress in the ground increases. Basic soils mechanics can then be applied to estimate the resulting settlement. For tunnels in granular soil the settlement due to this increase in effective stress is usually reflected as an elastic phenomenon requiring knowledge of the low stress modulus of the ground and calculation of the change in effective stress. Unless the soil contains silt or very fine sand, this elastic settlement will typically represent the majority of the total but its absolute value will also be relatively small.
For fine grained soils, the situation is a bit more challenging but certainly manageable using normal soil mechanics approaches. With fine-grained soils, the conditions are reversed. In most instances, the settlement is mostly due to consolidation brought on by the changes in effective stress and hence is analyzed by the usual soil mechanics consolidation theories. In some instances, primarily if lenses of sands are contained in the soil, there may also be a relatively small contribution by elastic compression. In comparison to the settlement of granular soils, consolidation can lead to several inches of settlement when the consolidating soils are thick and the change in effective stress is significant.
Lost Ground Lost ground has a number of root causes (at least nine) and is usually responsible for the settlements that make the headlines. By definition, lost ground refers to the act of taking (or losing) more ground into the tunneling operation than is represented by the volume of the tunnel. Thus it is highly reflective of construction means and methods. As will be discussed, modern machines can be a great help in controlling lost ground but in the end it usually comes down to quality of workmanship.
For the purposes of this manual, the causes of lost ground are lumped into three groups: face losses, shield losses and tail losses.
7.5.3 Settlement Calculations
Estimates of settlement in soft ground tunneling are just that, estimates. The vagaries of nature and of construction are such that settlements cannot be estimated in soft ground tunnels to the same level of confidence as, say, the settlement of a loaded beam. In tunneling we rely heavily on our experience with some assistance from analysis. Thus, there are two related methods to attack the problem: experience and empirical data.
Experience can be used where a history of tunneling and of taking measurements exists. An example of this is Washington, D.C., where soft ground tunnels have been constructed in well-defined geology for over 40 years. During that time the industry has progressed from basic Brunel shields to the most current closed-face tunneling machines. For this case it would be anticipated that an experienced contractor would achieve between 0.5 and 1.0 percent ground loss (see Table 7-7 ). An inexperienced contractor would attain 1.0 to 2.0 percent loss.
When there is no record to rely upon, the design would have to be based strictly on empirical data and an engineering assessment of what the contractor could be expected to achieve with no track record to rely upon. In that case the above evaluations might be bumped up one-half percentage point each as an insurance measure
State-of-the-art pressurized-face tunnel boring machines (TBM) such as EPB and SFM as discussed in Section 7.3.2 minimize the magnitude of ground losses. These machines control face stability by applying active pressure to the tunnel face, minimizing the amount of overcut, and utilizing automatic tail void grouting to reduce shield losses. Typically, ground loss during soft ground tunnel excavation using this technology limits ground loss to 1.0 percent or less assuming excellent tunneling practice (adequate pressure applied to the face and effective and timely tail void grouting).
The volume of ground loss experienced during tunneling can be related to the volume of settlement expected at the ground surface (Peck, 1969). For a single tunnel in soft ground conditions, it is typically assumed the volume of surface settlement is equal to the volume of lost ground. However, the relationship between volume of lost ground and volume of surface settlement is complex. Volume change due to bulking or compression is typically not estimated or included in the calculations. Ground loss will produce a settlement trough at the ground surface where it can potentially impact the settlement behavior of any overlying or adjacent bridge foundations, building structures, or buried utilities transverse or parallel to the alignment of the proposed tunnel excavation. Empirical data suggests the shape of the settlement trough typically approximates the shape of an inverse Gaussian curve (Figure 7-9 ).
The shape and magnitude of the settlement trough is a function of excavation techniques, tunnel depth, tunnel diameter, and soil conditions. In the case of parallel adjacent tunnels, surface settlement is generally assumed to be additive. The shape of the curve can be expressed by the following mathematical relationships (Schmidt, 1974).
The settlement trough distance, i is defined as:
The maximum settlement, wmax is defined as:
Table 7-7 summarizes likely volumes of lost ground as a percentage of the excavated volume and a function of combined construction practice and ground conditions.
Figure 7-9 Typical Settlement Profile for a Soft Ground Tunneling
For geometrics other than a single tunnel, adjustments of the types given below should be made to obtain settlement estimates:
As shown in Figure 7-9 the width of the settlement trough is measured by an i value, which is theoretically the horizontal distance from the location of maximum settlement to the point of inflection of the settlement curve. The maximum value of the surface settlement is theoretically equal to the volume of surface settlement divided by 2.5 i. Figure 7-10 illustrates assumptions for i values (over tunnel radius R) for calculating settlement trough width in various ground conditions.
Figure 7-10 Assumptions for width of settlement trough (adapted from Peck, 1969)
The ground settlement also can be predicted by numerical methods. The numerical method is extremely useful when the tunnel geometry is not a circular or horse-shoe shape since analytical/empirical method is not directly applicable. A sample finite element settlement analysis is shown in Figure 7-11.
Figure 7-11 Example of Finite Element Settlement Analysis for Twin Circular Tunnels under Pile Foundations
7.6 Impact on and Protection of Surface Facilities
7.6.1 Evaluation of Structure Tolerance to Settlement
Evaluation of structural tolerance to settlement requires definition of the possible damage that a structure might experience. Boscardin and Cording (1989) introduced three damage definitions for surface structures due to tunneling induced settlement (where settlement is calculated per Section 7.5):
A number of methods for evaluating the impact of settlements on building or other facilities have been proposed and used. In 1981, Wahls collected and studied data from other investigators (e.g., Skimpton and MacDonald, 1956: Grant, Christian, and Vanmarked; Polshin and Tokar) plus his own observations (totaling more than 193 cases). From that study Wahls proposed the correlation of angular distortion (the relative settlement between columns or measurement points) and building damage as shown in Table 7-8 .
As an alternative initial screening method, Rankin (1988) proposed a damage risk assessment chart based on maximum building slope and settlement as shown Table 7-9 .
a Safe limit includes a factor of safety.
7.6.2 Mitigating Settlement
Where the settlement is or would be caused by groundwater lowering the first, and usually the simplest, approach is simply to reduce or eliminate the conditions causing or allowing dewatering. This could include, for example:
Where the settlement is or could be caused by lost ground in the tunneling operation that settlement can nearly always be mitigated with proper construction means and methods. For example consider:
7.6.3 Structure Protection
The concept of and methods for structure protection are already woven into earlier paragraphs. First and foremost are the tunneling procedures of maintaining face pressure (control) and immediately grouting to fill the annular (or any other) void.
The next step is ground improvement either by consolidation or jet grouting and, closely related compensation or compaction grouting.
As a last resort, to be applied when all else appears to be unsuccessful and or unworkable, is underpinning. Like the use of compressed air, this method is now seldom used because modern tunneling techniques make it unnecessary. At times, as with the Pershing Square garage in Los Angeles it is still applicable, but most of the time practitioners believe it to have the possibility to do more damage than to be beneficial. Typical steps of underpinning method are summarized as follow:
Instrumentation and monitoring for the existing structures are discussed in Chapter 15 Geotechnical and Structural Instrumentation.
7.7 Soil Stabilization and Improvement
Until fairly recently essentially all the design effort for tunnels in soft ground was to provide a support system or systems that would stabilize the existing ground during construction and then, perhaps with some modification, would permanently support the ground and provide an opening suitable for the long term mission of the tunnel. In the last two or three decades, however, the situation has changed such that in some applications a dual approach is taken. First, the characteristics of the ground are modified by stabilization and/or improvement to make that ground contribute more to its own stability. Then, secondly a supplementary but less costly support/lining system is installed to make the tunnel perform for its full lifetime. In this section the various methods of soil stabilization and improvement are summarized. References with more details on these methods are also given.
7.7.2 Typical Applications
The decision to use soil stabilization or improvement must be made on each individual case. This decision may sometimes be easy with there being no other way to construct the tunnel. More often, the decision comes down to a trade off among treating the ground, using high-tech machines, and/or a combination of the two. With all of the possibilities it can be said that there are now no unacceptable construction sites. Table 7-10 summarizes the challenging ground sites and corresponding treatment methods.
It is to be noted that the boundaries between both ground conditions and treatment methods are not fixed. Also, the use of vibrocompaction techniques or dynamic compaction is typically applicable at or near the tunnel portals as these techniques are applied to the ground surface and are not effective beyond about 100 ft depth for vibro compaction and 35 ft depth for dynamic compaction. Both are generally effective only in granular soils.
Readers are referred to the Ground Improvement Methods Reference Manual (FHWA, 2004) for more detailed discussion for the soil stabilization and improvement techniques presented below.
7.7.3 Reinforcement Methods
Soil Nails Soil nails may be used to stabilize a tunnel face in soil during construction. Steel or fiberglass rods or nails are installed in the face and the resulting reinforced block(s) are analyzed for stability much as for usual slope stability analyses. Several methods (e.g., Davis, Modified Davis, German, French, Kinematrical, Golder, and Caltrans) are used for these analyses. Walkinshaw (1992) has studied these methods and concluded that all had some level of inconsistencies, such as:
For more discussion readers are referred to GEC No.7 Soil Nail Walls (FHWA, 2003), which also recommends that the Caltrans SNAIL program be used because it will handle both nails and tiebacks. However, it must be recognized that application of that or any other program must be tempered with appropriate judgment, measurements and case history experience.
Soil Doweling Soil doweling entails the installation of larger reinforcement members than does nailing. These dowels act in tension like soil nails but are large enough in cross section that they also develop some shearing resistance where they pass through the sliding surfaces.
As they are applied to tunneling, micropiles are essentially the same as soil dowels. These are typically drilled piles two to six inches in diameter that contain a large reinforcing bar centered in the hole and the hole backfilled with concrete. As opposed to pin piles that are typically installed at the surface (and that act in compression), the pin piles placed in tunnels typically act in tension and shear across the sliding surfaces.
Soil nails, soil dowels, and pin piles are typically installed at the face of the tunnel to stabilize that face for construction. Thus, they are continually being installed and mined out of the face. For ease in this mining operation, fiberglass bars (rods) are typically used in these applications because they are much easier to mine out and cut. In contrast, spiling tends to look out around the perimeter of the tunnel, thus steel is more likely to be used for spiling bars or plates.
Readers are also referred to "Micropile Design and Construction Reference Manual" (FHWA, 2005f) for more details.
7.7.5 Grouting Methods
All grouting involves the drilling of holes into the ground, the insertion of grout pipes in the holes, and the injection of pressurized grout into the ground from those pipes. The details of the operations, however, are distinctly different. Readers are referred to the Ground Improvement Methods Reference Manual (FHWA, 2004) for more detailed discussion for the grouting techniques discussed hereafter.
Permeation Grouting Permeation grouting involves the filling of pore spaces between soil grains (perhaps displacing water). The grout may be one of a number of chemicals (but is usually sodium silicate or polyurethane) or neat cement using regular, micro- or ultra- fine cement, along with chemicals and other additives. Once injected into the pore spaces, the grout sets and converts the soil into a stable, weak sandstone material. Permeation grouting usually involves grout holes at three to four feet centers with enough secondary holes at split spacing to verify that all the ground is grouted. If necessary to get full coverage all of the split spacing holes may have to be grouted and verification performed by the tertiary holes.
Compaction Grouting Compaction grouting uses a stiffer grout than does permeation grouting. In compaction grouting the goal is to form a series of grout bulbs or zones four to six feet above and around the tunnel crown. By pumping the stiff grout in under pressure these bulbs compress (densify) the ground above the tunnel and between the tunnel and overlying facilities.
The pipes for compaction grouting are pre-positioned and drilled into place and all the grouting pumps, hoses, header pipes, instrumentation and the like are in place before the tunnel drive begins. Instrumentation is read as the tunnel approaches and passes a facility and the grouting operation is adjusted real time in response to the movement readings. Actually, in most applications it is possible to either pre-heave the ground or to jack it back up (at least partially) by pumping more grout at higher pressures.
Compensation Grouting Compensation grouting is, in some ways, similar to compaction grouting. The goal is to monitor ground movements, primarily between the tunnel and any overlying facility. When it is apparent that ground is being lost in the tunneling operation, a grout, typically slightly more liquid than the compaction grout mix, is injected to replace (compensate for) the lost ground. As indicated the differences between these two schemes are relatively minor - compaction grouting seeks to recompact the ground by forming grout bulbs, compensation grouting seeks to refill voids created by the tunneling operations.
Jet Grouting Jet grouting is the newest of the grouting methods and is rapidly becoming the most widely used. Jet grouting uses high pressure jets to break up the soils and replace them with a mixture of excavated soils and cement, typically referred to as "soilcrete". There are a number of variations of jet grouting depending on the details of the application and on the experience and expertise of both the designer and the contractor.
The design of a jet-grouted column is influenced by a number of interdependent variables related to in situ soil conditions, materials used, and operating parameters. Table 7-11 presents a summary of the principal variables of the jet grouting system and their potential impact on the three basic design aspects of the jet-grouted wall: column diameter, strength and permeability. Table 7-11 gives typical ranges of operating parameters and results achieved by the three basic injection systems of jet grouting. It should be noted however, that the grout pressures indicated in this table are based on certain equipment and can vary. This table can be used in feasibility studies and preliminary design of jet-grouted wall systems. The actual operating parameters used in production are usually determined from initial field trials performed at the beginning of construction.
Jet grouting is frequently used as a ground control measure in conjunction with tunneling in soft ground using Sequential Excavation Method (Chapter 9).
7.7.6 Ground Freezing
As with much of tunneling technology, ground freezing was developed first in the mining industry and was probably first used in sinking mine shafts. For a mine the shaft (and the mine) is located where the ore is. Thus, means of obtaining access in unfavorable ground conditions, of providing emergency support in unstable ground below the water table, and of maintaining stability of working faces below the water table, such as freezing, often had their roots in the mining industry.
In its simplest form, ground freezing involves the extraction of heat from the ground until the groundwater is frozen. Thus converting the groundwater into a cementing agent and the ground into a "frozen sandstone". The heat is extracted by circulating a cooling liquid, usually brine, in an array of pipes. Each pipe is actually two nested pipes, with the liquid flowing down the center pipe and back out through the annulus between the pipes. When the pipes are close enough and the time long enough, the cylinders of frozen soil formed at each pipe eventually coalesce into one solid frozen mass. This mass may be a ring or donut as needed to support a shaft or a solid block of whatever shape necessary to stabilize the working face or heading.
Because of the dearth of engineering data on the properties of frozen ground (especially clays) it is recommended that two steps be taken early in any design of ground freezing:
However, a few general guidelines can be stated as follows (after Xanthakos, 1994).
Groundwater flow across the site requires special considerations closer pipe spacing, multiple rows of pipes and the like. Groundwater flow velocities approximately >2 m/day may impede or prevent freezing. A number of special challenges associated with ground freezing should be considered in both the design and construction stage. Those are creep of frozen ground, sensitivity of frozen ground properties to loading condition, ground heave or settlement, and others.
Readers are referred to the discussions and details of ground freezing application in Chapter 12.