Table of Contents

GEO Report No. 298

Table of Contents

Ground Control for

EPB TBM Tunnelling

HKIE Geotechnical Division

Working Group on Cavern and Tunnel Engineering
and Geotechnical Engineering Office
Table of Contents
Table of Contents

Geotechnical Engineering Office

Civil Engineering and Development Department
The Government of the Hong Kong
Special Administrative Region

Table of Contents

Ground Control for

EPB TBM Tunnelling

Table of Contents

GEO Report No. 298

Table of Contents

HKIE Geotechnical Division

Working Group on Cavern and Tunnel Engineering
and Geotechnical Engineering Office

Table of Contents

The Government of the Hong Kong Special Administrative Region

Prepared by:
Geotechnical Engineering Office,
Civil Engineering and Development Department,
Civil Engineering and Development Building,
101 Princess Margaret Road,
Homantin, Kowloon,
Hong Kong.

Table of Contents

First published, July 2014

Table of Contents
Table of Contents
Table of Contents

Table of Contents

Preface

Table of Contents

In keeping with our policy of releasing information

which may be of general interest to the geotechnical
profession and the public, we make available selected internal
reports in a series of publications termed the GEO Report
series. The GEO Reports can be downloaded from the
website of the Civil Engineering and Development Department
(http://www.cedd.gov.hk) on the Internet. Printed copies are
also available for some GEO Reports. For printed copies, a
charge is made to cover the cost of printing.
The Geotechnical Engineering Office also produces
documents specifically for publication in print. These include
guidance documents and results of comprehensive reviews.
They can also be downloaded from the above website.

Table of Contents

The publications and the printed GEO Reports may be

obtained from the Governments Information Services
Department. Information on how to purchase these documents
is given on the second last page of this report.

H.N. Wong
Head, Geotechnical Engineering Office
July 2014
Table of Contents

Foreword

Table of Contents

This report should be used as a guide only. The content of this report may require
further updating as more experience in EPB TBM tunnelling and performance data in local
ground conditions become available. Users applying the guidance in this report should take
into account the actual ground conditions, the monitoring and controls available for ground
control in the TBM chosen for the project, and the specific risks that the TBM will introduce,
in preparing the design calculations, risk registers and work procedures, and in planning the
risk mitigation measures and contingency plans for ground control. In all cases, the risk
owner and key personnel assigned to control ground risk posed by the TBM operation must be
clearly identified, and experienced tunnelling and geotechnical professionals must be
employed to carry out and take responsibility for the design, site supervision and risk
management.

Table of Contents

The HKIE Geotechnical Division's Working Group on Cavern and Tunnel Engineering
undertook to produce this guidance document using voluntary input from its members, with
input and advice from local and international professionals in the tunnelling industry. The
aim of the report is to provide guidelines for the auditing of the design calculations and work
procedures relating to ground control during EPB TBM tunnelling in Hong Kong. There
have been a number of major reports on pressurised TBM tunnelling in the last ten years
including BTS/ICE (2005) and publications by international and national tunnelling
associations (ITA WG-14 (2007)). It is not intended to duplicate the general advice given in
these documents, but to provide more specific guidelines on issues relating to ground control
for EPB TBM tunnelling in Hong Kong. The focus of the report is on key issues that affect
the magnitude of ground movement due to tunnelling, and therefore the potential impact on
third parties.

Table of Contents

In December 2009, the Geotechnical Engineering Office published GEO Report

No. 249, Ground Control for Slurry TBM Tunnelling, based on a report prepared by Golder
Associates (HK) Ltd under an expert advice consultancy. Consultation had been made
widely with the tunnelling and geotechnical industries before publication. One of the most
frequent comments received was the need for an equivalent set of guidelines for Earth
Pressure Balance Tunnel Boring Machine (EPB TBM) tunnelling.

As most of the guidelines given in GEO Report No. 249 are also relevant to ground
control of EPB TBM tunnelling, the relevant parts of the GEO Report, with suitable
amendments where needed, are reproduced in this report. The agreement by Golder
Associates (HK) Ltd for the extracts from the GEO Report No. 249 to be reproduced in this
report is gratefully acknowledged.
Table of Contents

A draft of this report was circulated to members of the HKIE Geotechnical Division's
Working Group on Cavern and Tunnel Engineering, as well as to the Association of
Geotechnical & Geoenvironmental Specialists (Hong Kong), the Hong Kong Contractors
Association, International Society for Soil Mechanics and Geotechnical Engineering TC204,
Lars Babendererde of Babendererde Engineers, Alastair Biggart of Alastair Biggart
Tunnelling LLP, Richard Lewis of YL Associates Ltd, Nick Shirlaw of Golder Associates
(HK) Ltd, Roger Storry and Bruno Combe of Bouygues Travaux Publics, Andy Raine of

Table of Contents

P.L.R. Pang
Deputy Head (Mainland)
Geotechnical Engineering Office
July 2014

Table of Contents

Dragages Hong Kong Ltd, Piers Verman of Leighton Contractors (Asia) Limited, Darren Page
of OTB Engineering UK LLP, Shinichi Konda and Lok Home of the Robbins Company,
Thomas Camus of NFM Technologies and a number of prominent individuals in the TBM
tunnelling field in Hong Kong and overseas. Useful and constructive comments and
suggestions for improvement were received. H.P. Lo, Kenny Kam, Ivan Chan and Patrick
Chau, under the direction of N.F. Chan, coordinated the GEO input into this document,
including the finalisation and production of the document after consultation with the
profession and individual experts. All contributions are gratefully acknowledged. Special
thanks are given to David Salisbury of MTRCL for his dedication to drafting the document,
coordinating and resolving comments from various parties, as well as sourcing figures and
references from tunnel practitioners.

Table of Contents
Table of Contents

Contents
Page
No.
Title Page

Preface

Foreword

Contents

List of Tables

List of Figures

10

Scope and Objective

11

Glossary of Terms and Symbols

12

Face Pressure Assessment and Control

12

3.1

Factors of Safety

15

3.1.1

Partial Factors on Shear Strength Parameters

15

3.1.2

Partial Factors on Surcharge

15

3.1.3

Partial Factor on Water Pressure

15

3.2

3.3

Effective Stress Calculations of Minimum Face Pressure

16

3.2.1

ULS Calculation - Method Based on Anagnostou &

Kovari (1996)

20

3.2.2

SLS Calculation - Method Based on Proctor &

White (1977)

22

3.2.3

Assessing Pressures for Tunnels in Rock Using

Effective Stress Methods

23

Total Stress Calculations of Minimum Face Pressure

24

3.3.1

Total Stress ULS Calculations

24

3.3.2

Total Stress SLS Calculations

25

3.4

The Effect of Interfaces

26

3.5

Dealing with Highly Variable Ground Conditions

27

3.6

Assessing the Maximum Acceptable Face Pressure

28

3.6.1

29

Maximum Pressure in Intact Ground

7
Page
No.
3.6.2

Pressures for Using Compressed Air for Head Access

30

3.8

Use of Compressed Air or Semi-EPB Mode

36

3.8.1

Ground Conditions Unsuitable for Semi-EPB Mode

of Operation

37

3.8.2

Pressure Distribution Over the Face under Semi-EPB

Mode of Operation

38

3.8.3

Risk of Short-term Heave

38

3.8.4

De-oxygenated Air

40

3.8.5

Risk of Encountering an Open Path

40

3.10

30

3.7

3.9

Maximum Pressure with an Open Path to the Ground

Surface

Adjustment to the Target Face Pressures Based on Observations

40

3.9.1

Adjustment of Target Face Pressures

41

3.9.2

Adjustment of Actual Face Pressures

41

Presentation and Communication of Target Face Pressures

42

3.10.1

Presentation of Face Pressure Calculations Prior to

Start of Tunnelling

42

3.10.2

Presentation of Target Face Pressures for Regular

Review during Tunnelling

43

3.10.3

Communication of Target Face Pressures to the

Operator and Other Tunnel Staff

44

3.11

Why the Target Face Pressure Might Not be Applied in Practice

44

3.12

Some Key Issues for Designers and Design Checkers

46

Screw Conveyor

48

4.1

Types of Screw Conveyor

50

4.1.1

Centre Shaft Screw Conveyor

50

4.1.2

Ribbon Screw Conveyor

50

4.1.3

Twin Screw Conveyor

51

4.1.4

Articulated Screw Conveyor

52

4.1.5

Two-stage Screw Conveyor

52

Conditioning Agents

54

5.1

55

Foaming Agents

8
Page
No.

5.2

Polymers

56

5.3

Anti-clogging Agents

56

5.4

Abrasion Reducing Agents

56

Tail Void and Shield Skin Grouting, Limiting Pressures and Volumes

58

6.1

Types of Grout

60

6.1.1

Single Component

60

6.1.2

Two or More Components

61

6.2
7

Shield Skin Grouting

61

Excavation Management Control System

61

7.1

Accuracy of the Volume and Weight Measurements

62

7.1.1

Volume of Additives

62

7.1.2

Belt Weighers and Belt Scanners

62

7.2

Assessment of the Volume and Weight of the Material

Excavated

62

7.3

Use of the Information from the Excavation Management

Control System

63

High Risk Activities (Break-in, Break-out, Interventions, Interfaces,

Wear, Flushing)

64

References

73

Glossary of Terms

77

Glossary of Symbols

80

List of Tables
Page
No.

8.1

High Risk Activities and Some Mitigation Measures That

Can Be Considered for Reducing the Likelihood of Loss
of Ground during EPB TBM Tunnelling

65

8.2

Some Mitigation Measures That May Be Considered for

Controlling the Consequences of Excessive Loss of
Ground during EPB TBM Tunnelling

72

Table of Contents

Table
No.

Table of Contents
Table of Contents
Table of Contents

10

List of Figures
Page
No.

3.1

Variation of Pressure over the Face

17

3.2

Dimensions

19

3.3

Fluctuation of Face Pressure

21

3.4

Compressed Air Pressure (Balancing Water Pressure 1 m

above the Base of the Exposed Face)

32

3.5

Effect of Use of Compressed Air in a Lens or Pocket of

Sand

33

3.6

Idealised Pressure Distribution in Semi-EPB Mode, Based

on the Spoil Level at Tunnel Axis Level

39

3.7

Section View of a Tunnel Showing Ring Numbers

47

4.1

Centre Shaft Screw Conveyor (Courtesy of Herrenknecht

AG)

50

4.2

Ribbon Screw Conveyor (Courtesy of Richard Lewis)

51

4.3

Twin Screw Conveyor (Courtesy of NFM Technologies)

51

4.4

Articulated Screw Conveyor (Courtesy of The Robbins

Company)

52

4.5

Two-stage Screw Conveyor (Courtesy of Herrenknecht

AG)

53

4.6

Example of Excessive Wear on Screw Conveyor

54

8.1

Typical Launch Seal (Courtesy of Richard Lewis)

69

8.2

Example of Rock Fragment Trapped in the Cutterhead

69

8.3

Example of Rock Fragment Trapped in the Opening of the

Cutterhead

70

Table of Contents

11

Table of Contents

Schematic of an EPB TBM Showing Some of the Major

Features (Courtesy of Herrenknecht AG)

Table of Contents

1.1

Table of Contents

Figure
No.

11
1 Scope and Objective

(a) determination of the appropriate face pressure to support the

ground,

Table of Contents

The objectives of this report are to provide background, calculation methods,

references and other information about ground control during Earth Pressure Balance Tunnel
Boring Machine (EPB TBM) (Figure 1.1) tunnelling. The scope of the report includes:

(b) controlling target face pressures,

(c) modifying excavated material parameters using additives,
(d) establishing limiting pressures and volumes for tail void
grouting,

(f) high risk activities (break-in, break-out, head access,

interfaces, wear, flushing).

Excavation Chamber

Table of Contents

(e) the use of Excavation Management Control (EMC) system,

and

Thrust Cylinder
Segmental Lining

Cutterhead

Table of Contents

Screw Conveyor

Segment Erector

The report concerns ground control during tunnelling using an EPB TBM in Hong
Kong, in the superficial deposits and in weathered granitic and volcanic rocks. For
tunnelling in other strata, such as in marble, additional considerations will apply, due to
particular geological features. Site investigation is mentioned but not covered in detail.
Readers should refer to GEO TGN 24 (GEO, 2009) and AGS (HK) (2004a, 2004b & 2005)

Table of Contents

Figure 1.1 Schematic of an EPB TBM Showing Some of the Major Features (Courtesy
of Herrenknecht AG)

12

This report provides guidance on the review of design calculations, drawings which
show the planned face pressures, and documented work procedures. To help to minimise the
risk of unacceptable ground movements, it is required that these documents and drawings are
prepared prior to tunnelling. While the preparation of appropriate design calculations,
drawings and work procedures is a necessary precursor to tunnelling, such preparation is only
part of the process and will not control all of the risks associated with the tunnelling. Other
factors can lead to excessive ground movements. These factors would include:

Table of Contents

for reference. Also, risk assessments are referred to, but not discussed in detail. Readers
should refer to other references on these subjects (e.g. Chiriotti et al, 2010; Kovari & Ramoni,
2006; Shirlaw et al, 2000).

(a) poor decision making by the tunnelling staff,

(c) mechanical breakdown, and

(d) unexpected ground behaviour.
The risk of an incident due to such factors can be reduced, and/or the consequences
controlled, by employing sufficient qualified and experienced staff, allowing sufficient time in
the construction programme for preparing and reviewing the design calculations and drawings,
and developing suitable risk mitigation measures, contingency plans (including for dealing with
mechanical breakdowns) and work procedures before tunnelling commences.
Most
importantly, management of the tunnelling by experienced tunnelling staff and geotechnical
professionals is essential to manage the risk associated with all types of TBM tunnelling.

The papers referred to in this report often use different terms/symbols to refer to the
same parameters, such as the depth to the axis level of the tunnel. In order to provide a
consistent document, tunnelling related terms/symbols and a brief are given in the Glossary of
Terms and Glossary of Symbols. Terms/symbols for which there are standards, such as for
soil strength, are not provided in the list of terms/symbols.

Table of Contents

2 Glossary of Terms and Symbols

Table of Contents

(b) priority given to cost and programme control over ground

control,

3 Face Pressure Assessment and Control

Table of Contents

Establishing, and then maintaining, the correct face support pressure (face pressure)
for the ground and groundwater conditions is critical to the safe operation of an EPB TBM.
If face pressure is inadequate, this will lead to excessive ground movement, and may result in
collapse of the tunnel face. Instability leading to major loss of ground or collapse at the face
of the tunnel is an Ultimate Limit State (ULS). Tunnelling in Hong Kong will generally
require stringent control of the ground movements due to tunnelling, to minimise the effect on
overlying and nearby buildings, structures and utilities. In this case, the Serviceability Limit

13

Generally, EPB TBMs are selected for use in clays, silts and sands. The minimum
face pressure should be calculated based on the effective stress method with an additional
check based on total stress method for cohesive materials, such as clays and silts, and other
coarse or granular materials embedded within a cohesive soil mass, to determine which
governs the designed face pressure. The maximum acceptable face pressure to prevent
heave must also be calculated.

(a) Open mode No active face support is provided under this

mode which will be applied for tunnels in stable rock
(Section 3.2.3) or where the ground is stable and without
any water inflow, groundwater drawdown and settlement
problems.

Table of Contents

(b) Semi-closed, semi-EPB or compressed air mode The

excavation chamber is approximately half full of spoil, with
the spoil fully covering the inlet to the screw conveyor and
the upper half of the chamber supported by compressed air.
Typical EPB conditioning agents, e.g. foam and/or bentonite
slurry, are used to reduce air loss. This mode may not be
considered as a traditional EPB mode and is only used
where traditional EPB operation is either not possible or
extremely slow, or causes excessive damage to the TBM
(see Section 3.8).
This area of technology is still
developing, with the appearance of hybrid EPB/Slurry
machines now being produced for pilot use in appropriate

Table of Contents

An EPB TBM can be operated in three basic modes (Babendererde et al, 2005):

Table of Contents

EPB machines (unlike slurry machines) are reactive rather than proactive. Face
pressure of an EPB TBM is maintained by a combination of propulsion thrust and removal of
excavated material at the correct rate to match the rate of advance, in order to minimise the
ground settlement and heave. If face pressure is lost, it can be maintained only by advancing
the TBM. If the TBM cannot move forward, the face pressure can be maintained by
pumping conditioners into the excavation chamber. An EPB TBM can be used in any soft or
loose material, cohesive or non-cohesive, but is more effective and efficient in soft clays, silts
and sands. In granular material there should be sufficient fine material to reduce the
permeability of the excavated ground to minimise pressure loss along the screw. Reference
to the grading curves interpreted from the ground investigation data is critical to the decision
on TBM selection and its design parameters (Meritt et al, 2013). The approximate limit of
permeability for the suitability of EPB machines is considered to be 10-5 m/s. By increasing
the percentage of conditioning agent, or by adding a thickener, such as pulverised limestone,
an EPB machine can be used in soil with a permeability greater than 10-5 m/s.

Table of Contents

State (SLS) has to be considered as well as the ULS. For intact clays and clayey silts the
face pressure is commonly calculated using total stress methods. Although effective stress
methods can be used for intact clays and silts, this requires an advanced soil model that
correctly models the stress path to failure. For sands and silty sands, effective stress
methods are used.

14
ground conditions.

In assessing the face support pressure for the full EPB mode discussed throughout in
Section 3, it is assumed that the ground is fully saturated and that the excavation chamber is
completely filled with excavated material, commonly referred to as excavation paste, with or
without the addition of soil conditioning additives (Section 5).

Table of Contents

(c) Full EPB mode or closed EPB mode The excavated

chamber is completely filled and the face support pressure is
provided through the control of the rate of TBM advance
and the rate of removal of the excavated spoil.

The target face pressure should be calculated at intervals along the tunnel alignment.
The intervals chosen will depend upon the variability of the predicted geology, the ground
cover to the tunnel and the sensitivity of adjacent structures.

(a) factors of safety,

(b) effective stress calculations for the minimum face pressure,

Table of Contents

The following are discussed below:

(c) total stress calculations for the minimum face pressure,

(d) the effect of interfaces,
(e) dealing with highly variable ground conditions (e.g. sand
lenses or interface contact zones),
Table of Contents

(f) assessing the maximum acceptable face pressure,

(g) pressures for using compressed air for head access,
(h) use of compressed air or semi-EPB mode,
(i) adjustment to
observations,

the target

face

pressures

based

on

(j) presentation and communication of target face pressures,

(l) some key issues for designers and design checkers.

Table of Contents

(k) why the target face pressure might not be applied in practice,
and

15
3.1 Factors of Safety
Table of Contents

Rather than using a global Factor of Safety, partial factors are applied to the
parameters used in design, including the shear strength parameters of the ground (soil or rock),
and the imposed loads, such as surcharge. Where reference is made to parameters, such as
tan ', c', su, or to the surcharge, q, in the rest of this report, the equations refer to factored
values of the parameters/surcharge.

3.1.1 Partial Factors on Shear Strength Parameters

In deriving the factored shear strength parameters for ULS design calculations, the
following partial factors are applied:
(a) on tan ': divide by 1.2.

(c) on su: divide by 1.5.

(d) on unit weight: multiply by 1.0.

Table of Contents

(b) on c': divide by 1.2 (noting that c' is generally taken as zero
for soil in these calculations).

For SLS calculations, a partial factor of 1.0 is applied to all of the soil parameters.

3.1.2 Partial Factors on Surcharge

(a) When the surcharge is unfavourable (i.e. for inward yielding

of the face): multiply by 1.5.
(b) When the surcharge is favourable (i.e. for heave of the
ground surface): multiply by 0.

Table of Contents

For ULS calculations:

For SLS calculations, a partial factor of 1.0 is applied to the surcharge loads.

3.1.3 Partial Factor on Water Pressure

Where the groundwater level is more than about one tunnel radius above the tunnel

Table of Contents

The water pressure to be used for the calculations should be the most onerous likely
pressure at the level of the tunnel, based on a critical assessment by the geotechnical engineer
of the available piezometric data and the groundwater flow regime interpreted from the
piezometric measurements. Allowance should be made for seasonal or tidal variation, where
appropriate. On this basis, a partial factor of 1.0 is applied to the water pressure.

16

Table of Contents

crown, the water pressure will generally be the dominant factor in the target face pressure
calculated based on effective stresses. In these conditions, the practical effect of the partial
factors applied to the shear strength parameters is usually very small, as a proportion of the
total face pressure. The use of the partial factors will not compensate for poor assessment of
the water pressure. It is essential to have as accurate an assessment of the water pressure at
tunnel level as possible. Sufficient piezometers should be placed along the whole tunnel
alignment, at appropriate elevations, to establish the piezometric level at tunnel level with
reasonable accuracy.
It is quite common to define the water pressure in relation to ground level. If this
approach is adopted, care must be taken in areas where the ground level changes suddenly, for
example at a cutting. Water pressure does not change as abruptly as ground levels can,
particularly where the natural terrain has been modified.

Table of Contents

Currently, there is limited information on the pore pressure changes near the face of an
EPB TBM tunnel. In granular soils, there is typically some degree of drainage out of the
face and through the screw conveyor, reducing the pore pressure but creating seepage
pressure. This effect can be minimised by the correct use of conditioning agents. The use
of foam, as an additive, may cause some transient increase in pore pressures just ahead of the
face, and may partially counteract the effects of the drainage. Generally, it is assumed that
the pre-tunnelling groundwater pressure represents an upper bound pore pressure for EPB
tunnelling. However, this assumption should be reviewed for each specific case.

Table of Contents

One of the effects of closed face tunnelling is that an excess pore pressure can develop
ahead of the face, provided that there is no inward seepage. This excess pore pressure
reduces effective stresses in the area of the face, and increases the support pressure required
(Broere, 2003). The net effect on the face pressure required is small unless the tunnel is in a
confined or semi-confined aquifer with a permeability of between 10-3 and 10-5 m/s. These
conditions are most likely to occur in Hong Kong when tunnelling through beach or marine
sands. Where these conditions are identified, reference can be made to Broere (2003) for the
influence of this effect. Where there is inward seepage, this will counteract against the
excess pore water pressure ahead of the face. This effect can migrate a substantial distance
from the excavation face.

3.2 Effective Stress Calculations of Minimum Face Pressure

In general terms, the target (minimum) face pressure (PEt) is calculated with an
allowance for variation, using the following equation:

The water pressure varies from crown to invert (Figure 3.1), so the minimum face
pressure required to balance the water pressure also varies. The pressure calculation
therefore needs to be specific to a particular level. Typically, calculations are for a particular
pressure sensor at a known level below the crown of the TBM or tunnel axis.

Table of Contents

PEt = water pressure + pressure to balance effective pressure

from soil and surcharge + allowance for variation in pressure ............ (3.1)

Table of Contents

Surcharge (q)
Ground Level

Table of Contents

Pressure to Balance
Water Pressure

17

Water Pressure

Table of Contents

Excavation
Paste Pressure

TBM
Please refer to Glossary for the definitions of symbols.

Figure 3.1 Variation of Pressure over the Face

Table of Contents

Note:

18

The equations provided in this report are for the calculation of the pressure at sensor 1
at the crown (Figure 3.2). Substituting the appropriate terms for the respective sensor into
the equations can derive the pressure at sensors 2 and 3. In order to carry out the
calculations, it is necessary to know the location of the sensors, relative to the axis or crown
of the TBM, with reasonable accuracy (200 mm). The position of the sensors should be
shown, with dimensions, on the drawings giving the internal arrangement and key dimensions
of the TBM. These drawings should be included as part of the design document setting out
the basis for the calculation of the face pressures.

(a) cement grouts used for ground treatment, such as for jet
grouting or deep soil mixing,

Table of Contents

It should be noted that the properties of the excavation paste can be affected adversely
by contact with saline groundwater, cement or other chemicals. Examples would include:

Table of Contents

For soil pressure, two particular methods of calculating minimum face pressure using
effective stress methods are summarised here: The method based on Anagnostou & Kovari
(1996) and the method based on Proctor & White (1977). In this report, both of the methods
in the original publications have been adopted for use in EPB TBM tunnelling. Only a brief
summary of the application of these methods will be given here. The original publications
should be consulted for the basis of the calculations.

(b) chemical grout used for ground treatment, see Jefferis (2003)
for an example,
(c) tail void grout, where this runs forward over the shield skin,
and

Table of Contents

For guidance on the choice, mixing, quality control testing and treatment of the
excavation paste by additives, reference should be made to manufacturers product data and
laboratory testing of soil samples prior to the commencement of the tunnel excavation. In
order to achieve effective support during EPB TBM tunnelling, it is necessary to ensure that
appropriate additive materials are used and to carry out detailed laboratory and site trials well
before commencement of tunnelling, as the effect of additives used singularly or in
combination varies. The trials will inform the requirements for regular quality control for
the properties of the excavation paste. Tests on slump value, bleeding ratio and filtered
water can be considered as quality control tests to verify the properties of the excavation paste.
A detailed testing schedule and acceptable ranges for test results should be included in the
work procedures. The site supervisory staff should check compliance with the requirements.

Table of Contents

(d) diaphragm walls or piles, where the machine has to break

through them.

Table of Contents

Ground Level

Piezometric
Level
ZW

Table of Contents

Z0

ZS1

19

Sensor 1

D0

Sensor 2

Sensor 3
TBM

Please refer to Glossary for the definitions of symbols.

Figure 3.2 Dimensions

D1
Table of Contents

Note:

Table of Contents

Piezometer

20

The target face pressure at the crown of the tunnel (PEt(Crown)) is calculated, using the
membrane model, from:
PEt(Crown) = pressure due to water + presure due to soil + pressure
due to surcharge + allowance for variation in pressure ............. (3.2)

Table of Contents

3.2.1 ULS Calculation - Method Based on Anagnostou & Kovari (1996)

Table of Contents

In slurry TBMs, the slurry forms a filter cake in advance of tunnelling and there is no
significant penetration or loss of slurry into the ground. Hence, the membrane model given
in Anagnostou & Kovari (1996) applies. In EPB TBMs, Anagnostou & Kovari (1996)
suggested use of a seepage model where face pressure significantly lower than water pressure
can be used provided that the additional effective stresses due to inward seepage is considered.
However, the seepage model is not reliable because it is difficult to control the amount of
inward seepage, which can induce consolidation settlement and may not be tolerated. Also,
it may not be possible to maintain inward seepage/drainage during the build cycle and if
significant water and earth pressures build up the paste in the screw conveyor will be blown
out when the discharge gate is opened, leading to loss of ground. Therefore, it is common
practice to carry out calculations for EPB TBMs based on the membrane model, which gives
an upper bound value for the required total face pressure. For critical cases such as shallow
ground cover, highly variable ground conditions, sensitive structures in the vicinity of the
tunnel, etc., it may be appropriate to use numerical analysis to provide additional justification
for the target face pressure.

Table of Contents

The contractor should also provide, as part of the work procedures, a contingency plan
for identifying and responding to any previously unidentified soft/loose soil or highly
permeable zones encountered during the tunnelling. Where such risks are considered likely,
which could result in large volume losses (Shirlaw, 2002), the machine specification, design
and work procedures should contain adequate provisions to mitigate the risks. For example,
the TBM specification and design may include a bentonite slurry injection system, sometimes
known as Auxiliary Face Support (AFS) system, which can automatically inject slurry into
the excavation chamber when pressure at the tunnel face drops below the agreed minimum
target face pressure. This system should have an interlock mechanism to prevent further
excavation advance and to close the screw discharge gate. The work procedures should
include the maintenance of an adequate store of bentonite slurry.

The basis of the Anagnostou & Kovari method is a simple limit equilibrium
calculation. The result of the equation provides the minimum pressure required to avoid
face collapse (ULS condition).

The pressure due to water at the crown = (C ZW )W ...................... (3.3)

The pressure due to the soil = F0 ' D - F1 c', where F0 and F1 are factors derived from
the charts in Anagnostou & Kovari, and depend on ' and the ratio C/D. 1 This equation is a
1

In the paper H is used as the symbol for cover over the tunnel, where C is used here.

Table of Contents

The terms in Equation 3.2 can be evaluated as follows:

21
truncated version of the equation in their paper, as the membrane model is assumed for
slurry/EPB tunnelling.
Table of Contents

The pressure due to surcharge at or near to the ground surface can be assessed using
the charts in Atkinson & Mair (1981). However, for critical locations, such as tunnelling
under buildings supported by piles, the pressure needed to control the foundation movement
should be assessed by more detailed calculation. It may be necessary to use numerical
analysis to take into account the interaction between the tunnelling and the piles.
The surcharge used as a basis for the calculation should be a realistic assessment of the
actual ground surface load applied during tunnelling; it is not intended that design loads such
as prescribed nominal HA or HB traffic loadings should be used in the calculation of target face
pressures during tunnelling, as these are unlikely to reflect the actual loading condition at the
time of tunnelling.

Target
Pressure

Pressure

-ve

Table of Contents

Time

Figure 3.3 Fluctuation of Face Pressure

Table of Contents

+ve

Table of Contents

A typical allowance for the variation in pressure, v, may be 0.2 bars (20 kPa), i.e. the
actual face pressure may be higher or lower than the target face pressure applied by 0.2 bars
(Figure 3.3). Consistently achieving a target face pressure within the tolerance requires a
skilled operator, the correct machine configuration, and appropriate additives for the ground
conditions. The value for the variation in pressure varies with the density of the ground and
will usually be greater for mixed ground than for a tunnel in a single material. The value for
the variation in pressure to be used in the calculations should be based on the ground
conditions, its variation and documented experience for the type of TBM proposed. It
should be regularly reviewed during tunnelling as it may require the face pressure calculations
to be revised.

22

PEt(S1) = PEt(Crown) + ( ZS1 C )E .......................................... (3.4)

As the unit weight of the excavation paste is more than that of water, the critical point
is at the crown of the tunnel. Hence the initial calculation is for the tunnel crown, and the
target face pressure at the sensor required to achieve the calculated pressure at the crown is
then assessed.
The equation given above is for the target face pressure.
pressure should not be allowed to fall below PE, where,

Table of Contents

The pressure applied varies depending on where on the face the pressure is measured,
which depends on the vertical distance below the crown adjusted for the unit weight of the
excavation paste. In order to achieve the required pressure at the crown of the tunnel, the
target face pressure at sensor 1, PEt(S1), can be derived from:

During tunnelling, the actual

One assumption used by Anagnostou & Kovari in deriving their charts is that

d/' = 1.6. Since d/' = Gs/(Gs - 1), the assumption is that Gs = 2.67. This assumption is
reasonable, within practical limits, for most soils in Hong Kong.

Table of Contents

PE(S1) = PEt(S1) - v .................................................... (3.5)

The charts in Anagnostou & Kovari are derived from calculations using a limit
equilibrium model proposed by Horn (1961). An alternative to using the charts is to derive
the necessary pressure directly, by carrying out calculations specific to the project. Broere
(2001) provides guidance on such calculations.

3.2.2 SLS Calculation - Method Based on Proctor & White (1977)

Table of Contents

There is limited basis to relate ground movements to face pressures in Hong Kong. A
simple method is to adopt the analytical methods proposed by Proctor & White (1977).
These were developed to derive the pressure against the walls of supported shafts or tunnels
in various soils, and, by analogy, can be used to assess the pressure exerted by the soil at the
face of the tunnel if only limited movement is allowed.
The calculation for the total face pressure is the same as that given in Equation 3.1,
except that the excavation chamber pressure to balance the soil pressure is given as 0.2 'D for
dense sand to 0.6 'D for loose sand.

(a) for saprolite with SPT-N > 30, use 0.25 'D,
(b) for saprolite or residual soil with SPT-N < 30 but > 10, use
0.4 'D, and

Table of Contents

There is limited information on which to base the application of the method to soils in
Hong Kong. The following are suggested, subject to review as experience is gained:

23
(c) for granular superficial deposits with SPT-N < 10, use
0.55 'D.

The face pressure is not the only factor that influences the volume loss over the tunnel;
other factors include the size of the overcut gap and tail void and the time taken to install the
permanent stiff support, i.e. the tunnel lining, the quality of which also requires effective
grouting of the tail void (Section 6). The actual minimum face pressure should be reviewed
regularly and adjusted based on observation.

3.2.3 Assessing Pressures for Tunnels in Rock Using Effective Stress Methods

Where the top of the stable rock is well established and there is low risk of a change in
the geology along the tunnel drive, the EPB TBM can be designed to be modified from closed
EPB mode or semi-EPB mode to work in open mode by removing the screw conveyor and
using a belt conveyor instead. This approach is particularly recommended if the length in
stable rock is considerable. Additional spoil handling plates are required in the excavation
chamber during open mode operation to lift and discharge the excavated rock onto the
discharge conveyor boot end.

Table of Contents

If the rock is jointed and the TBM operating pressure is less than the water pressure,
then seepage will develop towards the TBM. The effect of this seepage on pore pressures in
compressible soils above the rock and any resulting consolidation settlement should be
considered in setting the target face pressure. Additional effects on the annulus grout
injection and groundwater seepage affecting the grout should be considered if planning to
operate an EPB in open mode. This should only be considered in stable ground with no
hydrostatic head, low permeability soil or in an area where the effects of lowering the
hydrostatic head are not a concern.

Table of Contents

It is common for tunnels to be driven through stable rock without a face pressure.
Where part of an EPB TBM drive is in a full face of stable rock, then it may be possible for
the machine to operate in open mode, even to the extent of replacing the screw conveyor
with a belt conveyor for the majority of the rock sections of the tunnel, as face pressure is not
required for stability. If the screw is retained, but the face is stable with or without the use
of compressed air, this is more accurately referred to as the semi-EPB mode (Section 3).
There still needs to be sufficient material in the excavation chamber to ensure that the screw
conveyor entrance is covered so that the screw can work at the designed capacity. This can
increase efficiency and reduce the wear or risk of blockage on the screw conveyor.

Table of Contents

In determining the target face pressures, consideration needs to be given to the

anticipated ground and groundwater conditions along the tunnel alignment. Particular
concerns would be areas of high permeability, mixed ground conditions especially at rock/soil
interfaces, which could lead to differential pressures across the tunnel face, or soils with a low
shear strength (such as loose sand and soils rich in mica) and variations in the depth of cover.

Table of Contents

Generally, the method based on Proctor & White gives a minimum face pressure
slightly higher than that from Anagnostou & Kovari (1996) for dense saprolite, and
significantly higher for loose sand.

24

Zones of highly fractured rock may be treated as a granular soil for the purposes of
assessing the face pressure.

Table of Contents

Where there is a risk of encountering highly fractured rock, or the consequence of

encountering such a feature causing partial face collapse is high, consideration should be
given to operating in semi-EPB mode with the excavation chamber only partially filled.
This does provide the advantage over open mode in that it allows the excavation chamber to
be promptly filled and pressurised with excavated material, or with a bentonite slurry from the
reserve tank within only several minutes, thus minimising ground or excessive water loss
from the excavation face and its impact. The criteria on when the excavation chamber
should be filled should be assessed in design and communicated to the TBM operator for him
to follow. The semi-EPB mode without compressed air is however not recommended and
should be used only where a risk assessment indicates the consequences of substantial face
loss and settlement due to groundwater inflow and drawdown are acceptable.

Table of Contents

In generally stable ground, there is a risk of encountering local features that require
face pressure for stability. An example would be a fault zone in an otherwise stable rock
mass. The zone around the fault may be further weakened by weathering extending from the
fault. The contractors work procedures should include methods for identifying and
responding to the presence of such local features (Section 3.5).

3.3 Total Stress Calculations of Minimum Face Pressure

Table of Contents

For TBM tunnelling, effective stress calculations generally govern the minimum face
pressure in coarse materials. However, for EPB TBM tunnelling, where the excavation is
often in clay, total stress calculations may govern. In Hong Kong, this would typically apply
in cohesive marine deposits, and may apply in residual soils, and clayey colluvial or alluvial
deposits. However, it should be noted that unstable granular soils could exist as beds or
lenses within a relatively stable cohesive mass (i.e. Young & Dean, 2010). Where it is
uncertain as to whether effective or total stress calculations are appropriate, the minimum face
pressure required can be checked using both types of calculation, and the more onerous used
for design. The basis for assessing the minimum face pressure, and settlement due to ground
movement at the face of the tunnel, using total stress calculations, is given below.

3.3.1 Total Stress ULS Calculations

The minimum face pressure required to avoid collapse of the face of the tunnel can be
calculated from:

NTC is the stability number at collapse, and can be assessed from charts in Kimura &
Mair (1981) (reproduced in OReilly (1988)). The value for NTC varies based on the ratios
C/D and P/D, P being the length of the unsupported heading as defined in Kimura & Mair.
For the ULS case, P can be taken as zero. If there is sufficient movement, the ground will
close around the TBM skin, so that the ground will be fully supported except at the face.

Table of Contents

PE = ( Z0 + q) - ( su NTC ) ............................................. (3.6)

25
The target face pressure is then:

This target face pressure is the average pressure over the face, and can be taken as the
pressure at the spring line of the tunnel. The target face pressure at sensor 1 is then:
PEt(S1) = PEt - ( Z0 - ZS1)E .............................................. (3.8)

Table of Contents

PEt = PE + v ....................................................... (3.7)

3.3.2 Total Stress SLS Calculations

LF = N NTC ...................................................... (3.9)

where

= ( Z0 + q PE) / su

Table of Contents

The potential volume loss at the face and along the body of an EPB shield can be
evaluated using the results of the model tunnel tests presented in a chart in
Kimura & Mair (1981). The chart relates volume loss to the Load Factor (LF), the ratio of
the Face Stability Index (N) and the limiting stability index at the point of total collapse (NTC):

In using the chart, the value of P should be taken as 0 and L (the length of the tunnel
shield) for an EPB TBM without slurry injection and an EPB machine with slurry injection
around the shield skin respectively.

While an unsupported overcut may stay open long enough in stiff cohesive soils for
grouting to be implemented, unsupported overcut will close in unstable ground. Unless
there are ports in the shield for injecting bentonite to support the surrounding ground, the
volume loss due to the overcut should be evaluated in the SLS calculations. It is
recommended that all EPB TBMs should be designed with ports in the shield skin for
injecting bentonite into the shield annulus to control the ground movement into the overcut
gap or to reduce the thrust during TBM advancing.

PE = ( Z0 + q ) ( su NTC LF ) ........................................ (3.10)

Table of Contents

It is unlikely that the ULS calculations will satisfy SLS requirements for an urban area,
and, in consequence, a lower LF is likely to be needed. The allowable LF can be evaluated
from the allowable volume loss: after subtracting an allowance for the volume loss at the tail
void, the residual allowable volume loss due to movement at the face and along the shield
skin can be established. Then, the allowable LF can be read off the chart in
Kimura & Mair (1981) or calculated from the equation in Dimmock & Mair (2007), and the
minimum face pressure required to achieve the allowable LF can be calculated from:

Table of Contents

The volume loss derived from the charts represents only the volume loss due to
movement at the face and along the shield skin. Any volume loss at the tail of the shield (at
the tail void) will be additive to the loss due to movement at the face and along the shield skin.
The volume loss at the tail void will depend on the effectiveness of the grouting of the tail
void around the tunnel lining (Shirlaw et al, 2003).

26
For the SLS case in total stress analysis, it is not necessary to allow for the variation in
the face pressure, provided the variation is kept with a limited range. Then:

Equation 3.8 applies unchanged.

Table of Contents

Equation 3.7 changes to: PEt = PE

3.4 The Effect of Interfaces

The tunnel may pass through one or more interfaces between soils or rocks of very
different nature. Examples are:
(a) a palaeo-channel eroded into weathered granite, with the
palaeo-channel infilled with superficial deposits,

(c) in superficial deposits, there may be several beds in the face,

with each bed having different geotechnical properties.

Table of Contents

(b) the boundary between the rock and soil grades of weathered
granite (e.g. a soil/rock interface or a corestone-bearing
layer), and

The equations given in Section 3.2 are for homogeneous ground conditions. The
application of the equations in practice has to consider the effect of heterogeneous conditions.
There are two simple ways of doing this:

In the second of these approaches, it is implicitly assumed that the minimum face
pressure is controlled by the pressure required to support the weakest of the units in or close
to the face, and that the pressure required to support that unit in mixed conditions is not

Table of Contents

(b) Where an assessment is carried out for each of the units

present in or close to the face of the tunnel, the analysis is
carried out as if the tunnel were in homogeneous conditions
based on that unit, except for the overburden pressure,
which is based on the assessed unit weight and thickness of
each of the units over the tunnel. The face pressure
selected is that required to control the weakest unit (the one
requiring the highest face pressure). In practice, it is
generally simpler to identify which of the units present will
be critical, and carry out calculations only for that unit.

Table of Contents

(a) Where the tunnel is being driven in clay, and there are
granular soils or significantly weaker clay above the tunnel,
the cover over the tunnel (C) is taken as the cover of the
stiffer clay. In this case the weaker soils are treated simply
as a load (see Dimmock & Mair (2007) for an application of
this in London Clay).

27

The effect of contrasting permeability between units on the minimum face pressure has
to be considered in addition to the effects of differing strength. A confined or semi-confined
aquifer in the face will require a higher support pressure than required in homogenous
conditions, as discussed in Section 3.1.3, and the pressures calculated in the second approach
would need to be adjusted for this.

(a) Carrying out calculations based on the model proposed by

Horn (1961). The charts in Anagnostou & Kovari (1996)
are for homogeneous conditions, using the model proposed
by Horn (1961).
Heterogeneous conditions can be
assessed by calculation based on the original model.
However, the Horn (1961) model is only appropriate for
sands; for clay, the failure surface above the tunnel proposed
by Horn (1961) is inappropriate.

3.5 Dealing with Highly Variable Ground Conditions

In order to plan the target face pressures to be used during tunnelling, it is essential

Table of Contents

Where the minimum face pressure has to be higher in one unit than another along the
tunnel drive, it is too late to wait until the weaker unit is encountered to raise the face pressure.
The face pressure needs to be adjusted in advance of the interface, if the pressure needs to be
raised, or after passing through the interface, if it can be reduced. The pressure changes need
to be managed such that the higher pressure required to support the weaker unit is applied
while there is still a sufficient cover of the stronger unit over the crown of the tunnel. As a
rule of thumb, the cover should be at least one tunnel radius. Detailed calculations using
numerical methods or the results of centrifuge modelling can be used to assess the minimum
cover required in critical locations. Any uncertainty in the location of such interfaces needs
to be considered in the planning of the target face pressures, and a conservative assumption
made where there is uncertainty.

Table of Contents

(b) By numerical analysis. Finite element or finite difference

analysis can be used to assess the face pressure in
heterogeneous conditions, and can be used to calibrate the
simpler analyses discussed above.

Table of Contents

The methods outlined above for assessing tunnelling in heterogeneous ground

conditions are highly simplified. Depending on the complexity of the ground conditions, the
size and depth of the tunnel, and the environment over the tunnel, it may be appropriate to
refine the initial calculations by more detailed analysis. This can be done by the following
two methods:

Table of Contents

greater than if the conditions were homogeneous. Sample calculations for two layers of
contrasting strength in the tunnel face were presented by Broere (1998). The results of these
calculations show that for those cases where the calculated minimum pressure with two units
in the face is greater than for homogeneous conditions, the difference is small and can be
ignored for practical purposes.

28

The location of interfaces between units requiring a significantly different target face
pressure must be assessed conservatively, recognising the likely variation in ground
conditions. The lower face pressure required in the stronger unit should only be applied
where there is a high degree of confidence that the tunnel will be in that unit, with sufficient
cover to the weaker unit.

An EPB TBM may be excavating in generally stable ground conditions, but with local
features that require face pressure and operation in a suitable mode for stability. An example
would be a TBM operating in rock with local faults or deeply weathered seams. The
possible presence of the faults or seams of soil grades may be inferred, but their location and
the groundwater pressure may not have been identified in the ground investigation. In this
case, it may be appropriate to select the minimum face pressure based on the worst expected
ground and groundwater conditions, i.e. for the faulted or weathered zone(s) or soil.

3.6 Assessing the Maximum Acceptable Face Pressure

It is necessary to consider the maximum acceptable face pressure, as well as the
minimum. Excessive face pressure is caused by advancing the machine with inadequate

Table of Contents

The selection of an option to give the highest minimum face pressures might appear to
be the safest option. However, the wear on TBM cutting tools tends to increase with
increasing face pressure. Increasing wear results in more frequent stoppages for head access
to change cutting tools and repair the machine. Head access is one of the activities
associated with a higher than normal risk of loss of ground, as well as adverse health and
safety implications to the workforce where more frequent and higher pressure interventions
are required. Excessive wear may also result in other parts of the machine malfunctioning
such as excessive wear on the screw conveyor, shield skin and cutter housings. This may
lead to an inability to exert the desired face pressure, when weaker soils are encountered.
Thus, excavating through rock or strong soils with an unnecessarily high face pressure may
not reduce the overall level of risk.

Table of Contents

It may be possible to make provision for identifying the location of the weaker zones
along the tunnel by methods such as directional drilling prior to the commencement of
tunnelling or probing from the TBM while it is still in good rock or strong soil, and adjust the
minimum face pressure depending on the revealed conditions.

Table of Contents

The soil shear strength parameters that are used in the face pressure calculations
should be the lowest credible parameters for a particular section of tunnel, to cover the likely
range of conditions that will be encountered. How far to divide the tunnel into different
sections for the purpose of selecting the design parameters depends on the nature of the
ground and the groundwater conditions, the level of investigation and judgment.

Table of Contents

that sufficient site investigation be carried out to identify the ground and groundwater
conditions along the tunnel route. However, even a comprehensive site investigation will
not provide complete information on the conditions to be encountered during tunnelling.
Therefore, in assessing the minimum face pressure, consideration has to be given to the fullest
reasonable variation in ground and groundwater conditions given the anticipated geology.

29

Table of Contents

muck removal through the screw conveyor. This can be due to careless or poor operation of
the TBM. It can cause damage to the TBM (main seals and tail seals) and may result in
heave at the ground surface (less common than in Slurry TBMs). If an EPB TBM
encounters an open path to the surface (such as an old borehole) there is likely to be an
immediate loss of face pressure. This can lead to an eruption of foam, other conditioning
agents and/or grout to the surface, but does not necessarily lead to a subsequent face collapse.
However, when operating under similar conditions, if there is a significant loss of compressed
air during an intervention, this can result in instability of the face and endanger the workers in
the chamber through decompression or face collapse.

3.6.1 Maximum Pressure in Intact Ground

The pressure required to cause unacceptable ground heave is effectively the pressure
required to lift the block of ground over the tunnel - the reverse of the problem analysed by
Anagnostou & Kovari (1996).
This pressure is equal to the insitu stress at tunnel level plus a value based on the
strength of the ground mass. If the maximum face pressure is kept at or below the total
vertical overburden pressure, there should be no risk of excessive heave in intact ground.
For the pressure at sensor 1, this can be achieved if the target face pressure PEt(S1) is checked
for:

Table of Contents

It is not unusual to have a small (< 5 mm) heave at the ground surface as the TBM face
passes under a monitoring point; this level of heave is generally not a major concern. The
concern is with heave that is sufficient to cause damage, or to rupture the overburden.

PEt(S1) < C + ( ZS1 C )E - v ......................................... (3.11)

The measured pressure, PE(S1) , at sensor 1, any time, should not exceed:

Table of Contents

C is the total overburden pressure at tunnel crown. (ZS1 C) E is the difference in

pressure between the crown and the sensor, and v is the variation in pressure. In the
equation, v is subtracted to derive the maximum target face pressure.

PE(S1) < C + ( ZS1 C )E ............................................ (3.12)

Table of Contents

The check against total overburden pressure is simple, and in most cases will be
sufficient to confirm that there will not be unacceptable heave at the maximum design face
pressure. However, in some cases, such as relatively shallow tunnels in the superficial
deposits, it may be found that the maximum design face pressure (maximum target face
pressure + v) exceeds the overburden pressure. In this event, a more detailed calculation can
be carried out, allowing for the resistance provided by the shear strength of the ground.
Where the shear strength is taken into account, the appropriate partial factors for shear
strength parameters and surcharge should be used, as this represents an ULS.

30
3.6.2 Maximum Pressure with an Open Path to the Ground Surface
Table of Contents

As discussed above, where there is an open path, foam, other conditioning agents
and/or grout can erupt at the surface. The presence of the material at the surface can cause
problems for third parties, such as when it appears on a road, pavement, in a basement or
flows into a drain. Loss of tail void grout to the surface can also cause problems in the
tunnel, due to the inadequate tail void grouting that can result from the loss. It is unlikely
that an EPB TBM operated under the full EPB mode will cause any substantial eruption at the
surface, because as soon as an open path is encountered the pressure will dissipate rapidly.
The potential impact on the tunnel and third parties has to be managed as part of the
risk management for the tunnel. Common mitigation measures include:
(a) identifying any likely open paths and grouting them in
advance of tunnelling,

(c) testing whether the compressed air pressure can be

maintained before allowing access into the chamber.

Table of Contents

(b) maintaining continuous surface watch during TBM

tunnelling, and implementing control measures if a loss of
material is observed, and

All of the above measures are normally implemented in urban areas as it is unlikely
that all of the open paths can be identified from existing records.

Table of Contents

Because the face pressure in a slurry machine is pro-active whereas in an EPB machine
it is reactive, if there is an open path which allows the excavation paste to move up it, face
pressure in an EPB machine will be immediately lost until the shield is pushed forward.
Unlike slurry it will not continue to flow up the path. If pressure is regained by moving the
shield forward the flow along the path will be restricted by the high viscosity of the paste as
well as by its density. Hence, it is often preferable to continue excavating forward with an
EPB TBM until the location of the pressure loss is cleared, after which it can be grouted from
within the tunnel or from the surface if access is possible and working space is sufficient.
Selecting a face pressure that is appropriate to the ground and groundwater conditions,
but is not conservatively high, will help to limit the consequences of any material loss.

3.7 Pressures for Using Compressed Air for Head Access

Table of Contents

Compressed air can be used to allow access into the excavation chamber in unstable
ground or where water seepage is unacceptably high, for inspection and maintenance.
Compressed air was used to allow open face tunnelling in Hong Kong in the 1970s and 1980s,
particularly for tunnelling through saprolite and superficial deposits.

31

PCA = [ Z0 + ( D/2) ( ZW + 1)]W ...................................... (3.13)

In coarse soil with a low fines content, such as in the beach and marine deposits in
Hong Kong, there are a number of issues with the use of compressed air.

Table of Contents

This results in the air pressure in the upper section of the face being higher than the
water pressure, tending to dry out the soil in the upper section. CDG, when dry, will tend to
ravel, and for long stoppages additional face support needs to be provided using timbering or
sprayed concrete. At the base of the exposed face the air pressure is usually slightly less
than the groundwater pressure, leading to some seepage into the tunnel. CDG is generally
able to tolerate a very small seepage head, over a limited period, due to the residual
cementation of the soil. The effect of the air pressure can be observed in the face, and the
pressure is adjusted as necessary, based on those observations.

Table of Contents

The pressure exerted by compressed air is constant over the exposed face (Figure 3.4).
With an exposed face of granular soil, the compressed air will penetrate into the soil pores.
The pressure of the compressed air minimises groundwater flow towards the face, but does
not provide support to the effective stresses in the soil skeleton. For open face compressed
air tunnelling in saprolite, the air pressure would be typically set to balance the water pressure
at a level about 1 m above the base of the exposed face. For a full face this would mean that:

(a) Instability of the dried out soil: the sand near the crown is
dried out by the effect of the compressed air and becomes
running ground.

Table of Contents

(b) Instability of the soil at the base of the exposed face:

superficial sand deposits are less stable under seepage forces
than saprolite. Even the minimal net water pressure can
cause some erosion of the sand at the base of the exposed
face. This can be exacerbated if the sand is a discrete lens.
In this case the water cannot be pushed away from the face
by the compressed air.
Instead, the compressed air
pressure raises the water pressure, causing increased flow
and erosion at the base of the lens (Figure 3.5).
(c) Instability following a prolonged stoppage: the material in
the face dries out and collapses following a re-start.

Table of Contents

Table of Contents

Surcharge (q)
Ground Level

Table of Contents

Pressure to Balance
Water Pressure

32

Water Pressure

Table of Contents

Air Pressure
(Full Face Exposed)

TBM

Please refer to Glossary for the definitions of symbols.

Figure 3.4 Compressed Air Pressure (Balancing Water Pressure 1 m above the Base of the Exposed Face)

Table of Contents

Note:

Table of Contents

Ground Level

Table of Contents

Air Enters
Clay
TBM
33

Clay

Sand Lens

Water is driven out of base of sand lens, as

there is no other escape route.

Table of Contents

Figure 3.5 Effect of Use of Compressed Air in a Lens or Pocket of Sand

Table of Contents

Clay

34

(b) a layer of the sludge is left on the face which acts as a filter
cake, and
(c) the screw conveyor is left full of softer material which
makes it easier to start after a longer stoppage.

Table of Contents

(a) this material is easier to draw down with the screw

conveyor,

Table of Contents

The experience from the 1980s was for the use of compressed air in conjunction with
open face shields or sprayed concrete lining. For slurry TBM tunnelling, the stability of the
face can be improved by ensuring that a good filter cake is formed at the face prior to
intervention. For EPB tunnelling a filter cake is not normally formed during the process of
tunnelling. Where possible, during manned interventions, it is preferable to only remove the
spoil within the excavation chamber down to axis level. This will minimize the size of the
compressed air volume and therefore reduce the overpressure at the crown of the tunnel and
minimize the risk of face instability as far as practicable. However, prior to a compressed air
intervention a partial filter cake can be formed by replacing the muck with clean slurry for the
300 - 500 mm of excavation before the intervention. Any foam soil conditioning is stopped
and replaced with reasonably large amounts of bentonite, so that, for this final section of
driving, a very sloppy mixture of bentonite and excavated material fills the face. Results of
this are:

Where an emergency face pressure support system is incorporated into the TBM and
back-up, this can provide the means to fill the excavation chamber with bentonite slurry. A
filter cake will help by:

(b) reducing the amount of air lost through the face, and
(c) penetrating coarse sand (such as beach sand), thus creating a
zone of bentonite impregnated sand at the face.

The filter cake can be maintained during interventions by spraying the face with fresh
bentonite.
If an effective filter cake is formed, the compressed air pressure can be increased

Table of Contents

The adoption of a bentonite replacement system for interventions on an EPB TBM

should be considered in terms of the ground risk, risk to adjacent structures, risk to the
intervention team, available space within the machine, the additional cost of the system, the
additional complexity of the processes involved, and the time/cost of the replacement process
at interventions. It is becoming more common to incorporate an emergency face support
system into the larger diameter EPB TBMs, as it provides an additional option for reducing
the risk related to excessive settlement due to EPB tunnelling and interventions during the
tunnelling.

Table of Contents

(a) allowing the compressed air to support the soil particles, and
not just push the water ahead of the face,

35

(a) shorten the working time in the pressurised chamber; this

can increase the total time for the intervention. The longer
the intervention the greater the risk of a loss of ground,

Table of Contents

above the value given in Equation 3.13. After allowing for the difference in the distribution
of face pressure between EPB mode and compressed air mode, the total face support pressure
can be increased to match that required in EPB mode, provided that this does not result in an
unacceptable factor of safety against blow-out. While increasing the compressed air
pressure will provide increased support, there are disadvantages in increasing the pressure.
In particular, an increase in pressure will:

(b) increase the rate with which the filter cake dries out and
starts to peel off the face,

(d) increase the health and safety risks to the intervention team.
Setting the compressed air pressure involves a delicate balance between providing
sufficient support, but minimising the adverse effects listed above.

For planning purposes, it is suggested that:

In CDG, Equation 3.13 is used to assess the face pressure for full emptying of the
excavation chamber, while Equation 3.14 (below) can be used for drawdown to axis level.

Table of Contents

The compressed air pressure needed will depend on the level to which the excavated
material is drawn down and the nature of the ground conditions in and over the face. For
most interventions, the material level in the excavation chamber is not fully drawn down, but
is maintained at about axis level. This level of material allows the cutting tools to be
changed, as the head can be rotated to allow access to all of the tools. However, for major
maintenance it can be necessary to completely empty the excavation chamber. This should
only be undertaken after careful assessment and planning as it adds greatly to the risk if
personnel are working below axis within the excavation chamber.

Table of Contents

(c) increase the overpressure at the crown of the tunnel,

reducing the factor of safety against a blow-out, and

PCA = [ Z0 ( ZW + 1)]W ............................................. (3.14)

In superficial deposits, the compressed air pressure should be at least as high as for
CDG. Where a slurry filter cake is formed, the compressed air should be checked for:

The higher of the face pressures from Equations 3.14 and 3.15 should be used in
superficial deposits. In Equation 3.15, PEt is calculated on the basis given in Sections 3.2
and 3.3 to satisfy ULS, using a pressure variation (v) of 0.1 bars. The exposed face refers
to the area of the face exposed to compressed air. In settlement sensitive areas, the need to

Table of Contents

Average PCA over the area of the exposed face

Average PEt over the area of the exposed face ........................ (3.15)

36
satisfy SLS should be considered, but the benefit of increasing the air pressure, in terms of
limiting settlement, may be outweighed by the risk factors discussed above.

The air pressure is constant over the exposed height of the face, creating a significant
overpressure at the crown of the tunnel. It is necessary to check that the pressure at the
crown does not exceed the total overburden pressure. Otherwise, the ground may rupture,
leading to a blow-out of the compressed air. This is a severe risk to the safety of those
working in the compressed air, and to the public at the surface. In heterogeneous soils, the
tunnel may be driven in a permeable soil (an aquifer) with a lower permeability soil (an
aquitard) providing a cap above the tunnel. In this case the compressed air pressure can
develop at the base of the aquitard, i.e. at a level higher than the crown of the tunnel.

For smaller machines, airlocks can be provided further back in the tunnel, so allowing
part, or all, of the tunnel to be pressurised. In this case, the full length of the pressurised
tunnel should be checked to ensure that the weight of the overburden at all points is greater
than the compressed air pressure, to avoid the risk of a blow-out. Calculations to
demonstrate that there is factor of safety of at least 1.1 against blow-out should be carried out
at all locations where compressed air may be applied.

Table of Contents

During a long intervention, the compressed air will dry out the exposed face or, where
used, the filter cake. A dried filter cake will start to peel off the face. Depending on the
nature of the ground, the exposed ground can then ravel or run. For such situations,
provision should be made to replace the filter cake, or alternative materials/products such as
sprayed polymers may be used to provide a seal at the face to prevent the compressed air from
drying out the ground. Additional measures should be considered prior to re-starting
excavation to avoid dried out ground ahead of the excavation face ravelling immediately after
the re-start.

Table of Contents

For closed face TBM tunnelling, compressed air is typically applied only to the
forward chambers, to allow head access. This is done where the machines are large enough
to accommodate airlocks within the TBM, in which case the TBM pressure bulkhead is also
used as the compressed air bulkhead.

Table of Contents

The equations given above are for the planning of face pressures during interventions.
The actual pressure used should be adjusted based on regular inspections, the behaviour
observed during those inspections and monitoring data.

Interventions continue to be a major source of ground loss during closed face TBM
tunnelling. The risks associated with interventions, and some possible control measures, are
discussed in Section 8 below.

Semi-EPB mode has been used in mixed faces of rock and saprolite, particularly where
there is a high percentage of rock in the face. There are, however, a number of limitations
and risks in the use of this mode, which need to be managed. In particular:

Table of Contents

3.8 Use of Compressed Air or Semi-EPB Mode

37
(a) Some ground conditions are unsuitable for this mode.

(c) There is a risk of short-term heave during tunnelling; the

heave may initially mask the longer term settlement over the
tunnel.

(e) There is a risk of sudden loss of pressure if an old borehole,

well or other open path to the surface is encountered.

Table of Contents

(d) The compressed air can lose oxygen as it passes through the
ground; there is therefore a risk of de-oxygenated air
entering confined spaces, such as other tunnels in the
vicinity. In addition the de-oxygenated air can leak back
into the excavation chamber and air locks. Therefore,
these chambers should be purged before the next manned
entry.

Table of Contents

(b) The pressure distribution over the face and the target face
pressure required will be different to the closed EPB mode,
and the risk of a blow-out has to be checked.

3.8.1 Ground Conditions Unsuitable for Semi-EPB Mode of Operation

Table of Contents

The permeability of ground to air is much higher than to water. The compressed air
pressure can therefore readily penetrate into the pores in the soil or discontinuities in fractured
rock. As a result, an excess compressed air pressure (over the water pressure) applied to the
face of the tunnel does not support the soil skeleton or fractured rock, unless there is a low
permeability membrane covering the face or the soil mass is of very low permeability and has
a low air entry value. In semi-EPB mode, even when bentonite slurry is also injected, any
filter cake is constantly being destroyed by the action of the cutterhead. In order to operate
in soil in semi-EPB mode, the soil has to have sufficient cohesion for the face to be stable
under compressed air. Semi-EPB mode is therefore only possible where the ground is stable
under compressed air alone.
There has been extensive experience of compressed air tunnelling in Hong Kong.
General observations from this experience are that:

Table of Contents

(a) Superficial sand layers, such as alluvial and beach sands, are
unstable under compressed air. The use of compressed air
dries out the sand, which then runs. Venkta et al (2008)
recorded face stability problems when trying to operate in
semi-EPB mode in a mixed face of rock and sand in
Singapore. If the sand bed is a confined aquifer, the
compressed air pressure will tend to enter the top of the bed,
pushing the pore water out of the base of the layer into the
face. The flow of water at the base will erode the bed of
sand.

38
(b) Compressed air provides an effective support pressure in
clays and other very low permeability soils.
Table of Contents

(c) Many saprolites are initially stable under compressed air.

Over time, the drying effect of the compressed air will cause
the saprolite to ravel. This needs to be considered,
particularly if the TBM is stopped for a long period of time,
as semi-EPB mode may not be suitable during the period of
the stoppage.

3.8.2 Pressure Distribution Over the Face under Semi-EPB Mode of Operation

Face pressure calculations for closed EPB mode are not appropriate for semi-EPB
mode. New calculations should be carried out considering the different face pressure
distribution, the reduced variability in the face pressure, and to ensure that there is an
adequate factor of safety against blow-out. Typically, the face pressure required in
semi-EPB mode will be significantly lower than in closed EPB mode at the level of the tunnel
axis, but higher at tunnel crown. Setting the target face pressure to the minimum needed for
face stability will maximise the factor of safety against blow-out, and minimise the loss of
compressed air and the rate of heave.

Some loss of compressed air during semi-EPB mode is inevitable, although the rate of
loss depends on the permeability of the ground in the face, the area of the face exposed and
the magnitude of the excess pressure (over water pressure) at tunnel crown. The lost
compressed air will migrate from the tunnel; during compressed air tunnelling it is common,
after heavy rainfall, to see a mass of bubbles in the ponded rainwater due to the escaping air.

The heave at the ground surface will dissipate eventually, as the compressed air
escapes, and the underlying settlement due to the tunnelling becomes evident. Until this
occurs, the heave will mask the magnitude of the underlying settlement.

Table of Contents

If the compressed air encounters a low permeability soil above the tunnel, the
compressed air will build up under the cap of the low permeability soil. Ultimately, the
pressure under the capping layer can build up to the pressure applied at tunnel level. This
can lead to heave, or rupture, of the capping layer. The rate at which the heave develops
depends on many factors, including the relative permeability between the soil at the face and
the capping layer.

Table of Contents

3.8.3 Risk of Short-term Heave

Table of Contents

Compressed air pressure is constant over the height of the face that is exposed to the
air. The pressure distribution is therefore different to that in closed EPB mode, where the
pressure generally increases with depth due to the self-weight of the spoil. Idealised
pressure distribution in semi-EPB mode is shown in Figure 3.6. In semi-EPB mode, there is
an excess pressure at the crown of the tunnel. This excess pressure creates the risk of a
blow-out, particularly if the tunnel is shallow relative to the tunnel diameter.

Table of Contents

Excess Compressed Air Pressure CA

Water Pressure at Crown

Table of Contents

Tunnel Radius

39

E
Water Pressure at Invert
TBM

Please refer to Glossary for the definitions of symbols.

Figure 3.6 Idealised Pressure Distribution in Semi-EPB Mode, Based on the Spoil Level at Tunnel Axis Level

Table of Contents

Note:

Table of Contents

40
3.8.4 De-oxygenated Air
Table of Contents

As the compressed air passes through the ground it can become de-oxygenated. This
is a safety risk if the compressed air enters a confined space, such as another tunnel or a
basement. It is important to check the oxygen levels of such confined spaces before entry
after tunnelling.

3.8.5 Risk of Encountering an Open Path

Table of Contents

If the EPB TBM operates in semi-EPB mode, and an open path is encountered, the
compressed air will escape much more rapidly than normal, in the same manner as an airlift.
If the rate of loss of air is greater than the available rate of supply, this will lead to a loss of
face pressure and, potentially, a major loss of ground. Most open paths are man-made, such
as unsealed boreholes, instrumentation (such as standpipe piezometers or inclinometers),
wells, some types of pile, or where piles or sheetpiles have been removed. It is important,
before starting tunnelling, to carry out a detailed desk study and site reconnaissance of the
locations of such items on or near the tunnel route, and to take measures to seal the paths
before the TBM approaches. It is also important to remove any of the installations that
could obstruct the tunnelling.

3.9 Adjustment to the Target Face Pressures Based on Observations

An example of a TBM target face pressure review process is given below for reference,
but this needs to be adapted to the particular circumstances on each project.

Table of Contents

In order to review the performance of the TBM, it is essential that the key operating
parameters and the monitoring data are recorded, stored and available to those who need to
access it. The key TBM operating parameters should be recorded in the on-board computer
in real time. The work procedures should identify how the TBM operating data and the
monitoring data are stored, and how access to the data is provided. All this data should be
reviewed in conjunction with the corresponding instrumentation and monitoring data from the
surrounding ground, surface and structures.

Table of Contents

The minimum and maximum design face pressures and the target face pressures should
be reviewed regularly by the designer and the design checker, and where necessary they
should be adjusted, as part of the overall management of the tunnelling. The work
procedures developed by the contractor should include a clear process for the review and
adjustment of the target face pressures, and the updating of documentation and staff
instructions. The key personnel to be involved in this process will depend on the
contractors organisation. The work procedures should identify the minimum frequency of
review, and identify the geotechnical and tunnelling professionals with the authority to adjust
the minimum/maximum design face pressures and to make changes to the target face
pressures. Based on this data the target face pressures for the following week can be issued.

41
3.9.1 Adjustment of Target Face Pressures

Unless the designer and design checker attend the daily meeting, it is useful to hold a
weekly review meeting to assess whether the basis for the target face pressures needs to be
adjusted. This meeting would typically be held every week, and confirm the target face
pressures for the next one to two weeks. The review should consider any information on,
inter alia:

Table of Contents

A regular review of the target face pressures should be carried out daily. The site and
supervisory staff directly involved in the tunnelling should meet to review the progress,
performance and monitoring data from the previous day, and to confirm the planned activities
for that day.

(a) deviations from the target face pressures,

(c) any deviation from the planned ranges of grout volume and
pressure,
(d) any new information on ground or groundwater conditions,

Table of Contents

(b) any evidence of loss of ground, from the EMC system

records, monitoring or observation,

(e) construction or excavation works in the vicinity of the

tunnel alignment,
(f) instrumentation readings,

(h) measured ground movements and the settlement of

buildings, structures and utilities, and
(i) the actual variations in the pressure applied.

Table of Contents

(g) observed ground behaviour in the face,

Based on this review, it may be necessary to adjust the target face pressures.

3.9.2 Adjustment of Actual Face Pressures

(a) evidence of excessive excavation, based on the EMC

system,

Table of Contents

The actual face pressure used should generally be within the target face pressure range.
However, the operator will have to respond to observations during tunnelling, which may
result in or require the use of a pressure outside of the range set. Examples of observations
that could lead to such an adjustment include:

42
(b) loss of face pressure during ring building, when the TBM is
stationary, and

3.10 Presentation and Communication of Target Face Pressures

The target face pressures should be communicated to various parties involved,
including the designer and the design checker, the TBM operator and other tunnel staff. The
means by which this is done should form part of the contractors work procedures for the
tunnelling. Examples of work procedures are given below for reference, but they need to be
adapted to the particular circumstances on each project.

The calculation of the planned minimum/maximum design face pressures and the
target face pressures is carried out before the start of tunnelling. The calculations should be
documented in a design report by the geotechnical professional (the designer) and checked by
the design checker. The basis of the calculations should be provided, as well as the detailed
calculations and assessed pressures. The methods outlined above can readily be set up on a
spreadsheet. Typically calculations are carried out at intervals of 10 to 50 m along the

Table of Contents

3.10.1 Presentation of Face Pressure Calculations Prior to Start of Tunnelling

Table of Contents

Equipment designed to inject bentonite slurry from a reserve into the crown of the
cutterhead, if face pressure falls below a set value, is commonly specified. Concurrent with
this, the discharge gate closes and the advance stops - all to maintain face stability and
prevent blow outs or sink holes. This equipment is also used for supplying bentonite during
a TBM stoppage. During weekend stoppages (if any), foam is degrading and bentonite
slurry has to be injected into the chamber to compensate properly for the drop in level and
maintain the proper face pressure.

Table of Contents

The face pressure will need to be adjusted if there is evidence of face instability,
over-excavation, piezometric pressures outside the range assumed, excessive settlement/heave
or increasing/reducing pressure in the excavation chamber. The necessary action will
depend on the particular circumstances and will need to be implemented quickly, so the
decision has to be devolved down to the shift engineer or TBM operator and the key site
supervisory staff responsible for ensuring public safety. The key TBM management staff
(such as the tunnel manager) and the key site supervisory staff at management level (such as the
geotechnical professional responsible for risk management) should be available at all times, to
review the actions needed and the adequacy of the actions taken in response to such
observations, where there is significant risk to public and worker safety. The authorised
persons for modifying the adjustment of the confinement pressure must be clearly defined in the
work procedures for TBM operation. The work procedures should identify critical
observations and the decision making process for responding to those observations. If an AFS
system is installed, its use and trigger pressures should be included in the work procedures.

Table of Contents

(c) adverse monitoring data, such as significant change in

piezometric levels near the areas of tunnelling and excessive
ground settlement.

43

(a) where the tunnel is close to or at a major interface between

different geological units,
(b) where there is shallow ground cover,

Table of Contents

tunnel, depending on the variability of the ground and groundwater conditions, and the rate of
change in the depth of the tunnel. More closely spaced calculations or interpolation may be
deemed necessary:

(c) where the ground level changes suddenly or where the

piezometric levels fluctuate significantly,
(d) where there are major underground structures (i.e. bored
tunnels, underpasses or culverts) over or under the tunnel,

(f) where there is a significant (e.g. > 0.1 bar) difference in the
calculated pressure between adjacent sets of calculation,
(g) at the break-in and break-out, and

Table of Contents

(e) where there are piles over or adjacent to the tunnel,

(h) at planned intervention locations.

The minimum/maximum design face pressures and the target face pressures at
intervals along the tunnel alignment should be summarised on drawings, in combination with
the anticipated ground conditions and piezometric levels, the location of key facilities affected,
instrumentation, and anticipated obstructions or open paths to the ground surface.

Table of Contents

The focus of the regular design and management review of target face pressures is
typically on the next one to two weeks of production. The operators and other staff in the
tunnel are typically provided with this information along with a daily briefing for at least the
next one to two days of production noting any adjustments required following the data from
the recent excavation. The tunnel manager, or a key staff member designated by the tunnel
manager, provides to them a simple summary of the target face pressures together with the
other key target operating parameters, such as the screw operating speed, pressure for the
sensors along the screw conveyor casing, maximum torque, grout volume and pressure, etc.
Detailed calculations are typically too large to be readily used for this purpose. It is useful to
provide a simple summary sheet for the next 100 to 200 m of tunnelling, depending on the
tunnelling rate and variability in ground conditions. Target face pressures, as well as the
compressed air pressures to be used at intervention locations should be included in the daily
briefing sheets as well as the ring by ring target data for the TBM operator to work to, based
on the results of the daily meetings reviewing the TBM performance and settlement figures,
as well as the expected ground conditions ahead. This daily information should include as a
minimum:

Table of Contents

3.10.2 Presentation of Target Face Pressures for Regular Review during Tunnelling

44
(a) anticipated ground conditions, including piezometric levels,

(c) target face pressures,

(d) minimum face pressures,

Table of Contents

(b) any other expected obstacle or structures,

(e) proposed intervention locations and pressures,

(f) conditioning details,
(g) grouting pressures,
(h) grouting volumes,

(j) anticipated location of key interfaces,

(k) the location of buildings, structures, tunnels, roads and
utilities over or close to the tunnel,

Table of Contents

(i) other key TBM operating parameters,

(l) the location of subsurface instrumentation and known

boreholes and piezometers, and
(m) the location of piles, known/suspected obstructions or wells
on or close to the line of the tunnel.
Table of Contents

The information can provide a ready reference for planning of the work and when
assessing the appropriate response to any problems during the tunnelling. The target face
pressures summarised in the sheet should be reviewed (Section 3.9.1), and adjusted as
necessary, based on the experience gained during tunnelling.

3.10.3 Communication of Target Face Pressures to the Operator and Other Tunnel Staff

3.11 Why the Target Face Pressure Might Not be Applied in Practice
Even though the target face pressures are reasonably calculated and communicated to
the tunnel crew, they may not always be correctly implemented in practice. This may arise

Table of Contents

The operators and other staff in the tunnel are typically provided with information for
the next one to two days of production. The tunnel manager, or a key staff member
designated by the tunnel manager, provides to them a simple summary of the target face
pressures together with other key operating parameters, pressure sensors along the screw,
maximum torque, grout volume and pressure, etc.

45
from human errors or mechanical problems.

(a) the target face pressures are not communicated to the TBM
operator,
(b) the target face pressures are misread,

Table of Contents

Human error could result in inadequate face pressure if, for example:

(c) the operator deliberately reduces the pressure below the

target face pressure, for example, to improve the speed of
tunnel advance,

(e) the pressure sensors are not correctly set or maintained

(sensors must be checked and cleaned during each
intervention).

Table of Contents

(d) the operator is not sufficiently experienced to maintain the

face pressure within the minimum/maximum planned
pressure range (the target face pressure plus or minus the
variation in pressure used in the design calculations), and

Mechanical and other problems could result in inadequate (or excessive)

face/compressed air pressure, for example:

(b) There is a failure/breach of the main pressure bulkhead due

to excessive abrasion.

Table of Contents

(a) The discharge gate should close automatically if the face

pressure drops below a preset figure. This should be built
into the programmable logic controller (plc) controls and
logic of the TBM controls. If there is a boulder or such
preventing closure, the gate will need to be momentarily
released to clear the block and then closed again
immediately.

(c) There is an excessive wear on the screw conveyor.

(e) The pressure sensors are malfunctioning, leading to

incorrect information on the face pressure. Most machines
are fitted with multiple sensors and hence a single sensor
failure should be easily identified and corrected.

Table of Contents

(d) There is a power failure on the machine not allowing the

discharge gate to close. It is essential to specify that
discharge gates are fitted with a fail-safe closure mechanism
(accumulator) to prevent this.

46

Table of Contents

The risk that the target face pressures are not achieved in practice is something that has
to be considered in the risk assessment, and appropriate risk control measures should be
established. The contractors work procedures, organisation, supervision and maintenance
regime should be developed to reduce the risk to an acceptable level, reflecting the results of
the risk assessment. The action parties for implementing the risk control measures should be
clearly identified in the drawings or work procedures, and adequate training and audits should
be provided.

3.12 Some Key Issues for Designers and Design Checkers

The equations given above are simple, and a spreadsheet can be set up quickly to
derive appropriate limiting values for the face pressure. Carrying out a check using the
methods given here is therefore simple. Providing the original calculations give reasonably
similar values to the check calculations, there is no need to debate the respective merits of the
basis of those calculations.

Table of Contents

In practice, the face pressure is defined with reference to the ring number. It is
essential to correctly relate the location of the face to the ring number being used as the basis
for monitoring progress in the tunnel. The advance of a TBM tunnel is defined in terms of
the ring number that is built at the end of that shove. The ring is built in the tailskin at the
back of the shield, so the face during that shove is in advance, by slightly less than one TBM
length, of the plan location of the ring being used as a reference. It is important to
understand this, as the number of the ring to be built is also the basis for defining the target
face pressure to the TBM operator. If the operator is told to use 2 bars of pressure during the
shove for ring 100, the face at this point will not be at the location of ring 100, but at the
location where, say, ring 105 to 108 (depending on the relative length of the ring and the
TBM) will later be built (Figure 3.7). It is common to calculate the face pressures in relation
to length along the tunnel drive or to a chainage based on a project datum, initially, and then
translate this into ring numbers for use by the TBM operator. This translation is a common
source of error, and needs to be checked carefully.

Table of Contents

The number and locations of pressure sensors in the excavation chamber and the screw
conveyor need to be considered and agreed during the design stage of the TBM. A
minimum number should be specified by the client. Consideration should be given to their
positions to monitor the pressure distribution of the excavation paste at the face and along the
extrusion path, bearing in mind the possible effect of heterogeneity of the paste on the face
pressure applied and the need to have sufficient reliable data for the control of the risk against
variable face pressure. A level of redundancy and the ability to replace these sensors should
be considered in the TBM specification and design.

Table of Contents

Outlined above are some simple methods for establishing the minimum/maximum
design face pressures and the target face pressures for EPB TBM tunnelling. The methods
given are not the only ones available. There are viable, alternative approaches; the
alternative approaches should provide similar, but not identical, values. Broere (2001)
includes a summary of various methods that have been developed to assess the minimum face
pressure, including those by Jancsecz & Steiner (1994) and Anagnostou & Kovari (1996).
The important result is the proposed face pressure, not the means of establishing that value.

Table of Contents

TBM

Table of Contents

Rings Built
Direction of
Tunnelling
47

Ring Being Built

Ring 100
Ring 108

Table of Contents

Figure 3.7 Section View of a Tunnel Showing Ring Numbers

Table of Contents

Rings Built

48

(a) Is the geological interpretation accurate and has there been

enough desk study and ground investigation to define the
ground conditions along the tunnel alignment, especially the
location of interfaces and the potential presence of boulders
or corestones?

Table of Contents

The assumptions that are used in deriving the face pressures need to be reviewed in
detail, and be regularly revisited during tunnelling. Some questions that should be posed
are:

(b) Is there enough piezometric data to allow the design

piezometric level at tunnel level to be defined along the
alignment? Is there enough permeability test wells carried
out to allow the assessment of groundwater flow?
Table of Contents

(c) Has a review of records and site history been carried out for
the presence of old boreholes, instrumentation, underground
structures, utilities (including abandoned utilities) or
weakened/open zones (such as those due to the removal of
temporary works along the tunnel alignment in past
projects)?
(d) Is the assessed variation (v) in the face pressure achievable
in practice?

(f) Is there any performance data on EPB TBM tunnelling in

similar ground and groundwater conditions, and is this
consistent with what is being proposed?

Table of Contents

(e) Is the behaviour of the ground and the facilities affected

(from instrumentation and monitoring) consistent with the
calculations? For example, if the tunnel is driven close to
a vibrating wire piezometer in soil, do the piezometric
pressures rise as the face is driven past, reflecting the higher
than hydrostatic pressures calculated for the target face
pressures?

(g) Is the quantity of excavated material calculated by the EMC

system consistent with those for the face pressures used and
settlements recorded?

4 Screw Conveyor
The screw conveyor of an EPB TBM is the key component to regulate and maintain
the confinement pressure and to achieve the controlled pressure drop along its length. It

Table of Contents

(h) Are the results of the settlement monitoring above the tunnel
within the predicted range?

49

The discharge gate is a guillotine gate, the opening of which can be varied for
achieving the required face pressure during excavation. Discharge gates must be fitted with
a fail safe automatic closure mechanism in case of a power failure or a blow-out through the
screw conveyor which should also be designed to close automatically when the face pressure
drops below a pre-set value which is changed as the tunnel drive progresses according to face
pressure requirements. The use of a double discharge gate reduces the risk of it failing to
close due to being blocked and is an option which should be considered in the TBM
specification and design.

By varying the rotational speed of the screw conveyor the face pressure can be
maintained. A number of variations of the screw conveyor have been developed. These
are described below and several of them are shown in Figures 4.1 to 4.5.
The face pressure that a particular TBM can exert depends largely on the design of the
TBM, although the nature of the (conditioned) spoil also has an effect. The practical design
of screw conveyors is a key component in determining the limiting pressure under which EPB
TBMs can operate. Other components of the TBM have to be designed to ensure that the
machine can be operated at the maximum planned face pressure, plus a margin for
safety/unforeseen conditions, including:

(b) the seals for the main bearing,

(c) the soil conditioning system, and

Table of Contents

(a) the pressure bulkhead,

Table of Contents

It is common to utilise wear protection plates on the screw conveyor. These can
either be welded into place or of a bolt-on type to allow them to be replaced more easily.

Table of Contents

controls the discharge of the excavated material and uses the shear resistance of the excavated
material to counter the earth pressure at the bottom of the screw. The excavated material
then drops onto a belt conveyor running the length of the TBM back-up where it is removed
by further conveyors or muck wagons.

(d) the thrust rams, for which the face pressure is one of several
resistances that have to be overcome.

Table of Contents

The screw conveyor(s) have to be designed to maintain the difference between the
pressure in the excavation chamber, at the inlet to the conveyor, and the pressure at the outlet
from the system (the discharge gate). Generally, the earth pressure reduces along the flight
of the screw conveyor until it effectively balances with atmospheric pressure at the discharge
gate. However, a few machines have incorporated systems which allow controlled discharge
at a pressure greater than atmospheric, thus reducing the difference in pressure. The
pressure drop that can be achieved by the screw conveyor is largely determined by the number
of flights along the screw, which in turn is approximately related to the length of the screw
conveyor. However, the pressure difference is also affected by the type of screw (centre
shaft or ribbon), the number and arrangement of the screw conveyors used, whether there is
excessive wear, etc.

50
4.1 Types of Screw Conveyor

This is the most common type of screw conveyor, used on the majority of EPB TBM.
It is effectively an Archimedean screw inside a tube with a drive at the upper end. This type
of screw is the preferred design for all soils except where the size of boulders (or rock
fragments) anticipated is too large to pass up the flights (Figure 4.1).

Table of Contents

4.1.1 Centre Shaft Screw Conveyor

Table of Contents
Table of Contents

Figure 4.1 Centre Shaft Screw Conveyor (Courtesy of Herrenknecht AG)

4.1.2 Ribbon Screw Conveyor

Table of Contents

Where there is a risk of obstructions such as cobbles and boulders that would not be
able to pass up the flights of a centre shaft screw conveyor, a ribbon type screw can be
adopted. The larger openings in the ribbon screw allow boulders up to 60% of the screw
diameter to pass through the conveyor. However, ribbon screws have limitations in the
amount of torque which can be applied to them. They require a peripheral drive and can
be susceptible to blow-out through the centre of the screw. Ribbon screw are less effective
at dissipating the face pressure along the length of the screw and so are less suitable where
there is a high hydrostatic/piezometric head. They have torque limitations due to the
torsional strength of the ribbon (Figure 4.2). To cope with large boulders which may
arrive at the discharge gate, the gate has to have sufficiently large openings to discharge the
boulders.

51

Table of Contents

4.1.3 Twin Screw Conveyor

Table of Contents

Figure 4.2 Ribbon Screw Conveyor (Courtesy of Richard Lewis)

On some larger diameter EPB TBMs a pair of side by side screw conveyors have been
used to effectively double the excavation rate while maintaining the same face pressure
control. Such options are only suitable in ground conditions where large boulders or other
obstructions are unlikely to be encountered. It is more common to increase the diameter of a
screw conveyor to suit the ground conditions as the control systems of twin screw conveyors
are complicated (Figure 4.3).
Table of Contents
Table of Contents

Figure 4.3 Twin Screw Conveyor (Courtesy of NFM Technologies)

52
4.1.4 Articulated Screw Conveyor
Table of Contents
Table of Contents

Some machines are fitted with a centre shaft or ribbon screw conveyor inclined up to
the crown of the tunnel, which allow boulders to travel up the shaft, where they are disposed
of through a boulder collecting gate (removing boulders without losing pressure is not simple
and requires enough space). This is followed by an articulation joint beyond which a second
centre shaft screw conveyor is mounted horizontally in the crown of the tunnel. This second
screw can be of any length required, allowing greater earth pressure to be dissipated smoothly
(Figure 4.4). Articulated screw conveyors that have both lengths as centre shaft screws are
more common. These are used to increase the length of the screw conveyor for pressure
reasons, or to lift the screw conveyor higher through the build area. Forward and rear
sections of the auger can be joined by a universal joint to allow a single drive, or they can
have separate drives in which case they become a two-stage screw conveyor. Some screw
conveyors are articulated at the forward end to allow the rear discharge points to be
maintained directly above the belt conveyor on the first gantry. This feature is common
where the TBM has to negotiate tight curves.

Table of Contents

Figure 4.4 Articulated Screw Conveyor (Courtesy of The Robbins Company)

4.1.5 Two-stage Screw Conveyor

Table of Contents

A number of different configurations of two-stage screw conveyor systems have been

used to provide greater resistance to earth pressure and greater control over the effective face
pressure by combining a second, shorter, counter-rotating screw in series to the primary screw.
Peripheral drives are required for the forward screw. Japanese manufacturers have
experimented with two-stage screw conveyors which vary the rotation rate automatically with
the variation in face pressure. This type of screw conveyor has been used successfully for
both soft cohesive clays, in sands and gravels, and for deeper tunnels in cohesive soils

53

Table of Contents

(Figure 4.5). One of the advantages of a two-stage screw conveyor is that by varying the
speed of the second screw compared to the first screw, a sand plug of compressed material
can be created between the two screws, thus improving the overall resistance of the double
screw to pressure.

Table of Contents

Figure 4.5 Two-stage Screw Conveyor (Courtesy of Herrenknecht AG)

Excessive wear of the screw conveyor will reduce the effectiveness of the screw
conveyor as a critical part of the face pressure control system (Figure 4.6). The design of the
screw conveyor needs to give due consideration to the abrasiveness of the ground. Items to
be considered include, inter alia:

Table of Contents

All screw conveyors should include access ports for unblocking, ports for injection of
conditioning agents at various points along the length, pressure sensors at each end of each
stage as a minimum, and preferably in the centre as well. There should be wear protection
along the length, the first 5 - 6 flights being the most important.

(a) additional wear protection,

(b) facility for inspecting the screw conveyor for wear without
removing it from the casing, and

The ability to withdraw a screw conveyor and close a guillotine gate at the entrance to
the screw in the excavation chamber bulkhead should be included in the TBM specification
and design to allow maintenance of the screw conveyor while maintaining face support
pressure. Careful consideration should be given to the maintenance method if this feature is
not included in the TBM design.

Table of Contents

(c) facility to remove and maintain/replace the screw conveyor

without risk of loss of ground.

54

Table of Contents

5 Conditioning Agents

(a) to enable the

cutterhead to
(homogeneous,
allow it to pass
conveyor,

excavated material retained within the

be workable by having properties
viscous, flow, permeability) that would
easily through the cutterhead to the screw

Table of Contents

In EPB TBM technology, the correct use of soil conditioning agents is essential for a
successful and efficient TBM drive. It is important to form a paste of a suitable and
consistent density and viscosity. Jancsecz et al (1999) and Thewes (2007) outline the range
of soil grading that can be conditioned for EPB TBM tunnelling. With the future
development of EPB TBM technology and conditioning agents, the soil grading range of
application of EPB TBM tunnelling will not be limited to soft clays, clayey-silts and
silty-sands. The aims of providing conditioning agents are:

Table of Contents

Figure 4.6 Example of Excessive Wear on Screw Conveyor

(b) to support the excavated soil and any rock fragments in the
paste mix without segregation and settlement of the larger
rocks to the bottom of the chamber,

(d) to reduce friction and therefore reduce the required torque

and wear on the cutterhead and the screw conveyor, and
(e) to make the paste able to transmit the pressure from the
bulkhead onto the face to support it.

Table of Contents

(c) to make the paste effectively impermeable when it enters the

screw conveyor,

55
There are many products that have been used and are still useful.

Conditioning agents can be broken down into a number of major types, as described in
the following subsections.

5.1 Foaming Agents

(a) fluidising effect on soils because of the decrease in surface

tension, and as such the soil particles are no longer bound to
each other by linked water, and

Table of Contents

The main demand of foam as a conditioning additive is to create a paste, to build up

and maintain the necessary face support pressure in the excavation chamber, and to minimise
pressure variations. It creates small bubbles (foam) in the excavation paste, and is also used
to obtain a suitable rheology. The reduction in torque and abrasion is a very important
additional effect. Foam can be created out of a turbulent mixing of a surfactant solution and
air. The main surfactant properties are:

Table of Contents

Polymer is added to foaming solutions to give strength and stability to the bubbles in
the foam, i.e. as a stabiliser. Polymer is also useful to reduce the friction or adhesion
between the paste and the side walls of the excavation chamber, especially if there is a lot of
clay in the spoil.

Table of Contents

Plain water is commonly used. However, for many soil conditions, a pure bentonite
slurry is used. These were the only additives in the early days of EPB machines, following
which polymers were used. The main disadvantage of water-based fluids is that if the
quality control is not maintained properly the material can become too fluid when it emerges
from the screw conveyor. It can be too wet for handling easily on the belts and for disposal.
Foams were developed because they impart fluidity to the material, without making it wet.
Thus, as the bubbles collapse on the belt conveyor and the air escapes from the spoil, the
remaining material is relatively dry. However, there are still applications where water,
bentonite or polymer is more effective than foam.

(b) electrostatic repulsion effect which can separate two

particles attracting each other by electrostatic forces.

(a) concentration of surfactant agent in the foaming solution,

(b) ratio of mixing foaming solution with air (to create a foam),
and
(c) ratio of mixing foam with soil.

Table of Contents

Surfactants are a combination of a hydrophobic chain and a hydrophilic head. Both

parameters can be varied. Each soil type, from stiff clay to sandy gravel, requires its own
type of foam to reach its best effectiveness. The type of surfactant which should be used for
a special site has to be determined by laboratory tests with the insitu soil. Three important
foam parameters are:

56
5.2 Polymers

(a) water binding polymers to dry out (liquid) soils, and

(b) soil structuring polymers which are useful in loose, coarse
soils to change the excavation paste properties and which
prevent sedimentation.

5.3 Anti-clogging Agents

Table of Contents

Some polymer developments are based on hydrocarbon chains and are produced by
bacterial fermentation. These polymers are water soluble, biodegradable and compatible
with the foam surfactants. They are safe for the foaming generator. In consequence, they
can be mixed with the foaming solution and passed through the foam generator. Polymers
can induce a more stable support pressure in the excavation chamber during the tunnel drive
and when stopping the machine for a short time. All polymers should preferably be in liquid
form to avoid dosing problems.

Table of Contents

In addition to their foam stabilising effect, there are two main functional types of
polymers:

EPB TBMs were originally designed for soils with at least 30% of fines (Thewes,
2007). However, many clays have a tendency to stick to the cutterhead and affect the cutting
efficiency of the TBM. Anti-clay polymers carry a high charge density which separates the
soil particles, and creates an electro-chemical barrier which avoids binding/sticking effects.
Surfactants can also fulfil these functions.

Anti-abrasion additives are used in highly abrasive soils or rock excavation. They are
designed to protect the cutterhead, the tools mounted on the cutterhead and the screw
conveyor. Anti-abrasion additives are generally introduced directly into the cutterhead, in
the mixing chamber or in the screw conveyor.

Table of Contents

5.4 Abrasion Reducing Agents

Other conditioning agents, such as high-density slurries, have been used to deal with
specific ground conditions. Specialist advice and local geotechnical knowledge should be
sought when planning a project to ensure that the design is as comprehensive as possible.
Table of Contents

Conditioning agents from different suppliers should be tested for each project, as the
effectiveness of the chemical will be affected and altered in each situation by the natural
minerals occurring in the ground and in the local groundwater. Site and laboratory testing is
essential to ensure that the desired paste properties can be achieved and therefore should be
specified. Confinement pressure can have a significant effect on the performance of
foaming agent and therefore insitu results should be compared to the site and laboratory
testing at an early stage of the excavation, at any changes in geology, and periodically along
the drive to ensure the expected performance is being achieved.

57
Important parameters and considerations include:

(b) the dilution ratio with water to achieve the desired quantity
of usable chemical,
(c) the injection ratio as a percentage of solid ground excavated,
for foams, this is referred to as the FIR (Foam Injection
Ratio),

Table of Contents

(a) the concentration of the conditioning agent as purchased,

(d) for foams, the FER (Foam Expansion Ratio), and

(e) some polymers are also anti-clogging agents, which are
useful when sticky clays are excavated.

Although primarily a means of adjusting the properties of the excavation paste to make
it suitable for EPB tunnelling, the use of soil conditioning also has other beneficial effects.
These include a reduction of torque to both the cutterhead drive and the screw conveyor drive,
and a reduction on abrasive wear from the material to the steel components in contact with it
(cutterhead, shield, cutter tools and screw).

Table of Contents

The best place for injection of foam, polymer or other conditioning agents is through
nozzles in the cutterhead face fed through the cutterhead arms. This enables the
conditioning agent to be mixed with the excavated soil from the very moment it is excavated
by the cutterhead. Injection ports are also fitted to the pressure bulkhead but these are less
effective. It is also important to have some injection ports on the screw conveyor so that if
the screw torque increases the material can be softened. In this respect, it is better to
introduce the conditioning agent near the entrance to the screw conveyor.

Table of Contents

The FIR/FER will need to be estimated very early in the project, as these will
determine the design capacity of the air compressors and receivers required on the TBM
backup.

Nozzles on the cutterhead should be spaced so that each nozzle covers an

approximately equal swept area, thus radially the spacing is slightly closer together near the
outside of the face. However, the central area also needs an extra injection port as the
mixing action is reduced toward the centre of the head and clogging is more likely here.
Spacing of nozzles on the cutterhead should be approximately 1 m apart radially.
Table of Contents

Injecting agents through the rotating cutterhead arms requires a rotary joint on the
TBM. The port through this joint should be large enough to allow passage of the foam.
Each port going to each injection port should have its own pump. If nozzles are piped
together and fed from a single pump all the foam or polymer will pass through the port with
the least resistance and the other ports will receive nothing. Moreover, they will become
blocked very quickly. It is important to have a rotary joint with several injection lines in
order to have the ability to control the injection conditioning differently for each port and to
provide redundancy should one port/line become blocked or damaged.

58

6 Tail Void and Shield Skin Grouting, Limiting Pressures and Volumes
Tail void grouting is essential, both to minimise ground settlement due to tunnelling
and to ensure the load on the tunnel lining is reasonably uniform. If the tail void grouting is
not carried out properly, the load on the tunnel lining may be highly uneven. This can lead
to distortion of the lining, cracking and, ultimately, collapse of the tunnel. The tail void
grout must also quickly develop sufficient internal shear strength to stop the ring from
floating upwards within the grout.

(a) the use of gauge cutters (in rock) or hydraulically

extendable overcutters (in soil),
(b) any vertical or horizontal curves that have to be negotiated,

Table of Contents

The purpose of the grouting is to completely fill the annulus around the lining with
grout. The annulus is caused by the shield machine cutting a larger hole than the extrados of
the lining. The size of the annulus may be further increased by, inter alia:

Table of Contents

The injection of the conditioning agent should be monitored carefully and the volume
injected has to be considered when the excavated volumes and weights are being measured
for comparison with theoretical and for assessment of over excavation.

(c) the inclination of the TBM relative to the direction of

tunnelling, and
(d) any loss of ground at the tunnel face.

It is normal to define, for the tail void grouting: a minimum volume, a minimum
pressure and a maximum pressure. The work procedures should also state the means of
injection, the grout mix, testing to confirm the properties of the grout and methods of
checking that the grout is spreading around the ring and filling the full tail void. Examples
of grout mixes, testing and injection methods are discussed in Shirlaw et al (2004).
The theoretical minimum void can be calculated. This is typically increased by at

Table of Contents

The effective volume of the grout injected behind the lining is affected by losses into
the surrounding ground by permeation or over excavation and by bleeding of the grout. It is
therefore normal to inject more grout than the minimum theoretical volume of the tail void
(annulus). Typically the injected grout volume will be at least 20% above the theoretical
annulus volume.

Table of Contents

Where the ground is unstable without a support pressure, the tail void will close
rapidly unless a support pressure is provided. EPB TBM tunnelling will generally involve
such ground conditions. In order to provide a continuous support pressure at the tail void, it
is necessary to grout continuously during the advance of the TBM. Providing grout pipes,
which are laid within the tail skin and behind the tail seals, can do this. Such grout pipes
allow the injection of the grout simultaneous with TBM advance, commonly known as
simultaneous tail void grouting.

59
least 20% to arrive at the minimum volume to be injected.

The maximum grout pressure should be calculated, on the basis of avoiding excessive
heave of the ground surface. The pressure to heave the ground surface is the full overburden
pressure plus the pressure needed to overcome the shear strength of the ground. For shallow
tunnels, or for tunnels in soft clay, the minimum pressure to ensure effective grouting may be
greater than the pressure required to heave the ground surface. Where the two criteria are in
conflict, the critical criterion is to apply sufficient pressure to ensure effective grouting of the
ring. Uneven loading due to inadequate grouting of the tail void risks the stability of the
tunnel. The consequences of tunnel instability are generally much more severe than a
limited amount of heave, but this needs to be reviewed on a case-by-case basis.

Table of Contents

Tail seals should be incorporated into the tail skin, to prevent grout, water or soil
particles from entering the TBM via the tail void. A typical tail seal consists of multiple
rows of wire brushes, with grease injected under pressure into the gaps between the wire
brushes. The pressure of the tail seal grease needs to be kept at or above the grout pressure.
A lower pressure risks damage to the tail seals, which in turn risks loss of ground through the
damaged tail seals. Tail seal grease is generally measured by volume used as the TBM
progresses. The pumps inject shots on a timed basis. Pressure is measured in the pump
line, not at the tail seal. Whilst the pressure in the grease should be above the grout pressure,
it is not something which can practically be adjusted. Measured pressures (in the supply line)
typically vary throughout a shove between 2 bars and 40 bars depending on the timing of the
shots and the pressure measurement. The main objective is to keep the tail seal full of grease,
as a thin layer, typically 0.8 - 1.0 mm thick depending on the roughness of the segment
extrados, deposited on the outside of the lining as the TBM advances.

Table of Contents

Effective tail void grouting is a key factor in ensuring that settlement due to tunnelling
is minimised. Other potential sources of settlement include movement at the face (discussed
in Section 3), movement into the gap around the machine caused by overcutting and
negotiating curves, and consolidation of compressible soils. Effective tail void grouting is
also essential to maintain the build quality of the completed segmental lining which also
reduces long term settlement due to squatting of the rings.

Table of Contents

For shallow tunnels, or for tunnels in soft clay, this may equal or exceed the
overburden pressure at the crown. In addition, for shallow tunnels (with a ground cover less
than two tunnel diameters), the pressure should be assessed very carefully and in most cases
should be much less than 1 bar over the target face pressure.

Table of Contents

The injection pressure to be used has to be greater than the groundwater pressure, and
has to be sufficient to move the grout around the annulus around the ring. The effective over
pressure will also be affected by the number of grout ports available around the shield
perimeter. It is preferable for the injection pressure to be slightly greater than the face
pressure in the head of the machine, as the excavated material can connect to the tail void by
passing around the machine through the overcut. This effect is less likely in EPB machines
as the overcut tends to be smaller and the excavated material is less liquid than in a slurry
TBM. Typically, the minimum grout pressure is set at 1 to 2 bars over the target face
pressure, due to the location of the grout pressure sensors. It is useful to carry out a trial for
measuring the pressure loss along the grout lines. This can also be checked with an
instrumented section (trial plot).

60

Reverse spring plates on the extrados of the tail shield are recommended to prevent
grout migration along the extrados towards the cutterhead.

Table of Contents

Consolidation settlements can be caused by drainage of compressible soils, or by the

dissipation of excess pore pressures generated during tunnelling. Positive excess pore
pressures can be caused by excessive grouting, or by grouting into the tail void after the
ground has already moved down onto the ring. Poorly planned or executed grouting is a key
factor in the development of consolidation settlements over the tunnel (Shirlaw et al, 1994).

6.1 Types of Grout

Types of grout can be sub-divided into three categories:

(b) Semi-inert - Mainly composed of the same elements as

found in the Inert type with the addition of a small
percentage of a material that will cause some degree of
hydration to occur, e.g. < 100 kg/m3 cement. The grout
may take a considerable period of time to harden although
stiffening will occur more quickly.

Table of Contents

(c) Inert - The system contains no Portland cement, mainly

hydraulic lime and/or flyash and therefore only stiffens
slightly over time. It provides initial stiffness due to a
mechanical lock up of the mix. It is important to maintain
the mix and avoid any segregation during handling, as this
can lead to no mechanical lock up and float of the tunnel.
The rheology of this grout (shear strength and yield value)
has to be defined in order for the grout to be able to support
the encountered load.

Table of Contents

(a) Active - Grouts containing high percentages of Portland

cement causing full hydration of the grout material.

The selection of the type of grout will depend on the particular tunnel requirements,
such as the risk of settlement, the allowance for future deformation, effective down time of
the machine (dayshift only operation) and the soil conditions.
A second sub-division of grouting types is single or two component grouts.

A single component grout is a mixture of sand, a binder (usually cement or cement and
lime), water, and usually a workability additive. The material must be capable of being
workable for a long enough period from mixing to injection and be capable of being pumped
into the annulus while at the same time providing properties to avoid the segmental lining
moving, distorting or floating as the TBM advances. This can be difficult to achieve and
depends heavily on the types of sand available locally to the project.

Table of Contents

6.1.1 Single Component

61
6.1.2 Two or More Components

Two component grouting can give better results so far as filling of the annular void is
concerned, but requires more sophisticated equipment. Pipe blockages can occur at the
injection nozzles if the equipment is not operated and maintained correctly. The risk of
blocked grout pipes can result in inadequate grouting, which can lead to increased settlement,
ring deformation and cracking due to unbalanced loading. Site trials should be undertaken
to demonstrate that the mixed two-part grout can penetrate throughout the tail void before
gelling occurs.

To control ground movements into the overcut gap around an EPB shield, in particular
for EPB TBM with a step shield configuration where the volume of the annular gap increases
as the TBM advances, a system to allow injection of bentonite around the shield skin should
be provided. Injection of bentonite may not be required in loose or soft ground if the paste
formed from the excavated material can continuously fill the annular gap during the advance
of the TBM. However, for EPB TBM tunnelling in granular or unstable ground, the
excavation paste may not continuously fill the annular gap. Therefore, injection of bentonite
may be required in granular or unstable ground to minimise the ground movement.

Table of Contents

6.2 Shield Skin Grouting

Table of Contents

Two component grouts are generally described as A/B grouts where Component A is a
stabilised mortar filler and Component B is an activating liquid (typically sodium silicate)
which is mixed with Component A at the point of injection.

7 Excavation Management Control System

Table of Contents
Table of Contents

The net volume of the ground removed during excavation can be calculated by
measuring the volume of excavated material mucked out, and subtracting from it the volume
due to bulking and the volume of soil conditioning agents added. This can be compared
with the theoretical excavated volume calculated from the TBM cutterhead diameter and the
stroke length. The net weight of the ground removed can be assessed by measuring the
weight of excavated material mucked out and subtracting from it the weight of soil
conditioning agents added. This can be compared with the theoretical weight of excavated
material calculated from the theoretical excavated volume and the insitu bulk density of the
excavated material. Such calculations and comparisons provide the basis for assessing
whether there has been significant over-excavation at the face of the machine. While, in
principle, this is simple, there are a number of practical issues that must be considered,
including the accuracy of the volume or weight measurement and the accuracy of the bulking
factor and insitu bulk density of the excavated material used. Due to variability of materials
along the tunnel drive, the effects of bulking of the material excavated and the difficulty in
accurately measuring the insitu density of the excavated material, it is difficult to measure the
actual volume or weight of material excavated to within 10% - 15%.

62
7.1 Accuracy of the Volume and Weight Measurements

Even with regular calibration, the accuracy of the calculations can only be within about
10% - 15%. This can introduce significant variation in the measured net volume or net
weight of material mucked out. The longer the excavation cycle, the greater the potential
error becomes.

7.1.1 Volume of Additives

It is essential when calculating the volume of material discharged that account is taken
of the effects of any additive used. The potential loss of additives into the surrounding
ground which does not come back through the screw conveyor should also be considered in
the overall volumetric or weight calculations.

Table of Contents

The measurement of volume or weight can also be affected by gains and losses to the
system. Gains can come from groundwater entering the excavation chamber (i.e. inward
seepage). This should only happen if the face pressure is lower than the groundwater
pressure. Errors can also occur due to loss of additives into the ground or to the surface, and
during maintenance and head access where bentonite slurry is used to prepare the excavation
face for the maintenance interventions.

Table of Contents

Laser scanners are the direct way of continuous monitoring of the excavated volume
while belt weighers are the direct way of continuous monitoring of the excavated weight.

7.1.2 Belt Weighers and Belt Scanners

Belt weighers can suddenly malfunction, giving a spurious reading and causing
unnecessary concern over apparent over-excavation. It is becoming more common to
specify that two belt weighers are used simultaneously, on the basis that it is unlikely that
both will malfunction at the same time.

Table of Contents

Belt weighers and laser scanners/profilers can provide continuous measurement of

spoil quantity (BTS/ICE, 2005).

The use of belt weighers to measure the weight of material excavated and laser
scanners to measure the volume and hence deduce the density of material on the belt conveyor
is the best method of monitoring the spoil quantity. Regular cross-checking via sampling
and calibration of the belt weighers and laser scanners should be undertaken.

There is a significant difference in the weight of 1 m3 of soil and 1 m3 of excavated

rock. This is a function of the way in which the TBM cutterhead has been tooled to excavate
the ground and an accurate assessment is essential for the design of the tunnel mucking
system. It introduces a particular problem for tunnelling through weathered rock. In
partially weathered, heterogeneous rock masses, the tunnel will pass through a highly variable

Table of Contents

7.2 Assessment of the Volume and Weight of the Material Excavated

63

Table of Contents

mixture of rock and soil grades of weathered rock. Nakano et al (2007) discuss this issue for
tunnelling through weathered granite in Singapore. They give the example of the dry weight
of the solids from Grade III granite being nearly 50% higher, for each cubic metre excavated,
than Grade VI granite (Residual Soil). During tunnelling in heterogeneous weathered rock
masses, the relative proportions of rock and soil grades in the face and the moisture content
will be constantly varying, and so the insitu bulk density of the excavated material per ring
will also vary significantly.

7.3 Use of the Information from the Excavation Management Control System

Table of Contents

The information from the Excavation Management Control (EMC) system is important
in assessing the degree of ground control being achieved in EPB TBM tunnelling. The
guidelines for best practice with closed face tunnelling machines (BTS/ICE, 2005) list the use
of the EMC system in the conclusions and recommendations section, while also noting some
of the potential sources of inaccuracy in measurement. Despite the limitations given above,
the EMC system is a primary warning that significant over-excavation may be occurring, and
is an essential part of EPB TBM tunnelling, particularly in an urban environment. By
plotting the cumulative net volume (or net weight) of material excavated against the
theoretical volume (or weight), during the excavation of each ring, any trend of significant
deviation from the theoretical value can be identified. Where possible significant
over-excavation is observed, control measures should be implemented. Control measures
typically include:
(a) raising the face pressure (but not such as to induce excessive
heave),

(c) implementing the appropriate contingency measures, as

identified in the Risk Management Plan, if a potential void
(due to over-excavation) is suspected. An example of such
contingency measures would be to temporarily close the
part of a road over the suspected cavity, for safety of the
road users.

Table of Contents

(b) assessing whether it is necessary to grout, to fill the

potential void due to over-excavation, and

Because of the potential limitations of the system:

(b) the results of the EMC system are considered in conjunction

with the actual face pressures used, the variations in the face
pressure, the geological and hydrogeological conditions, the
specific observations on the nature of materials discharged

Table of Contents

(a) the results are typically considered on a rolling average over

5 or 10 rings, as well as for each individual ring; a deviation
of less than 20% in excavated volume for a single ring may
be due to errors in measurement, or over-excavation, or a
combination of the two, and

64

Systems are available for real-time remote measurement of the TBM parameters such
as torque, thrust, RPM, advance rate, face pressures and grout volumes. Continual
assessment of these parameters, together with the plots of deviations from the theoretical
volume (or weight) of material excavated, can indicate when over-excavation is occurring,
and allow the appropriate control measures to be taken expeditiously.

Table of Contents

The results of the settlement monitoring on the surface are important and should be
used in the daily and weekly meetings referred to in Section 3.9 when assessing TBM
performance and face pressures. Many urban projects require the use of real-time
monitoring using remote sensing instrumentation to provide continuous data which is
processed instantly and provide very timely and useful data to detect trends or provide early
warnings of unpredicted ground movement. With traditional manual survey monitoring, by
the time the recorded data has been processed, there is often little or no time to take action to
mitigate the effects. Major over-excavation typically appears at the surface as a local
sinkhole, and is unlikely to be captured by the settlement monitoring, as any instrument
fortuitously located at the sinkhole will become unreadable.

Table of Contents

from the screw conveyor, results of cross checking using the

belt weigher and laser scanner data, and the instrumentation
and monitoring results (including measured volume losses)
and other observations, before deciding whether significant
over-excavation has in fact taken place.

8 High Risk Activities (Break-in, Break-out, Interventions, Interfaces, Wear, Flushing)

Based on experience with EPB tunnelling in Singapore, Shirlaw et al (2003) identified
the following activities where there was a higher than normal risk of loss of ground:

(b) Break-out from the tunnel to the TBM recovery shaft

(recovery of the shield).

Table of Contents

(a) Break-in to the tunnel from the TBM access shaft (launch of
the shield).

(c) Interfaces between different strata.

(d) Mixed face conditions, consisting of rock and soil (or rock
and soil grades of weathered rock).

(f) Long drives in abrasive ground, leading to wear of the

machine.
The appropriate mitigation measures for these activities should be identified in the risk
assessment carried out for the project. As an aid to reviewing the risk assessment, reference
can be made to Table 8.1, which lists the high risk activities and a number of possible
mitigation measures. The table is based on Shirlaw et al (2005). The mitigation measures

Table of Contents

(e) Head access for maintenance.

65

Table 8.1 High Risk Activities and Some Mitigation Measures That Can Be Considered
for Reducing the Likelihood of Loss of Ground during EPB TBM Tunnelling
(Sheet 1 of 4)
High Risk Activity

Table of Contents

listed in Table 8.1 are not intended to be applied without a critical consideration of the
site-specific circumstances. There are alternative methods for mitigating risk that may be
applicable, depending on the circumstances. The actual mitigation measures to be applied
should be developed on a project specific basis.

Mitigation Measures

Table of Contents
Table of Contents

Shield recovery

Where possible, select site location with low risk for the
recovery.
Provide minimum length of ground treatment - typically length
of shield plus the length of two rings. This measure does not
need to be implemented if it can be demonstrated by
investigation, probing and calculation that the ground for this
section will be stable during tunnelling without a support
pressure.
Provide a seal between the shield and the wall of recovery shaft.

Table of Contents

Shield launch

Where possible, select site location with low risk for the launch.
Provide a seal between shield and wall of launch shaft
(Figure 8.1).
Test the seal pressure prior to launch beyond stable ground.
Provide minimum length of ground treatment - typically length
of shield plus the length of at least two rings. This measure
does not need to be implemented if it can be demonstrated by
investigation, probing and calculation that the ground for this
section will be stable during tunnelling without a support
pressure.
Provide detailed plan for how face pressures will be built up to
general operating level.
When launching through a soft tunnel eye created using steel
fibres or carbon fibre reinforcement, provide a contingency plan
for blockages in the screw conveyor while the head is still
within the treated zone.
Where Tubes-a-Manchette (TaMs) are used for grouting, ensure
that the material used for the tubes is brittle and breaks up when
cut by the head, to reduce the risk of blockage of the head or
screw conveyor. TaM tubes should be fully grouted to avoid
creating a flow path to the surface.

66

High Risk Activity

Mitigation Measures

Table of Contents

Mixed face conditions

Provide detailed planning of face pressures approaching,

through and after soil/rock interfaces, where identified.
Where necessary, grout the ground around the tunnel at the
interface.
Assess the interface conservatively. There has to be a high
level of confidence that the tunnel is in, and has sufficient cover
of rock, to operate the TBM at a face pressure lower than that
required for the overlying soil. See comments above on the
use of investigation, directional drilling or probing to provide
additional information.
Reduce the rate of advance, and/or speed of cutterhead or screw
rotation. This can reduce impact damage to the cutting tools
and the risk of plucking boulders/rock fragments from the face.
However, this needs to be balanced against the increased risk,
with a slower advance rate, that any overbreak will migrate
before it can be filled by the tail void grout.

Table of Contents

Interfaces between
stable & unstable
geological units

Provide detailed planning of face pressures approaching,

through and after the interface, where identified.
Where necessary, grout the ground around the tunnel at the
interface. While grouting is an option, this has to be
considered cautiously, as the effect of the grouting may be to
just move the location of the interface from one place to
another.
Reductions in face pressure should only be made where there is
a high degree of confidence that there is sufficient cover of
stable ground. Additional investigation, directional drilling or
probing from the tunnel can be used to provide more detailed
information on the location of the interface. However, probing
carries a risk of making a connection between the TBM and
permeable strata, so it is generally preferred to use a
conservative face pressure (i.e. higher than the hydrostatic
pressure) where there is a risk of such an interface close to the
tunnel, rather than probing.

Table of Contents

Shield recovery
(Contd)

Provide detailed plan for how face pressures will be reduced

from general operating level to zero.
Where TaMs are used for grouting, ensure that the material used
for the tubes is brittle and breaks up when cut by the head, to
reduce the risk of blockage of the head or screw conveyor.
TaM tubes should be fully grouted to avoid creating a flow path
to the surface.

Table of Contents

Table 8.1 High Risk Activities and Some Mitigation Measures That Can Be Considered
for Reducing the Likelihood of Loss of Ground during EPB TBM Tunnelling
(Sheet 2 of 4)

67

Mechanical problems

Regularly inspect and maintain the mechanical parts.

Provide hardening/wear plates for exposed areas of machine and
other key mechanical parts, and have ready essential equipment
for repair/replacement.
Provide discharge gates from the screw conveyor with a
fail-safe closure mechanism that operates automatically in the
event of a power failure.
Provide inspection ports along the casing of the screw conveyor,
so that the screw can be inspected without removing it.
Provide facility so that the screw conveyor can be partially
retracted, then hydraulically operated gates closed and a
blanking flange inserted to seal off the opening in the main
pressure bulkhead, so that the screw conveyor can be safely
removed and replaced, if necessary.
Provide a wear detection system that includes selected: cutting
tools, muck collecting buckets and the main cutterhead.
Mixing arms should be of substantial construction and provided
with wear protection.

Chamber/head access
(intervention)

Minimise the need for interventions in critical areas, such as

under buildings. Inspect and maintain TBM as necessary just
before entering critical areas.
Form safe havens, by grouting the ground around the tunnel,
to allow regular inspection/maintenance in critical areas.
Define level of authorisation required before intervention in a
critical area.
Give detailed work procedures for access, including details of
how the face pressure will be replaced by compressed air
pressure, and vice versa.
Provide compressed air locks (double air locks where the size of
the TBM allows). Compressed air access to be available
within 72 hours at all times.

Table of Contents

Mixed face conditions

(Contd)

Regularly inspect and maintain the cutting tools. This is

normal good practice, but the frequency of inspection should be
much higher in mixed soil/rock conditions than in uniform
conditions, due to the higher rate of damage.
Provide shrouds to protect the bearings of the discs from impact
damage or clogging with soil.

Table of Contents

Mitigation Measures

Table of Contents

High Risk Activity

Table of Contents

Table 8.1 High Risk Activities and Some Mitigation Measures That Can Be Considered
for Reducing the Likelihood of Loss of Ground during EPB TBM Tunnelling
(Sheet 3 of 4)

68

High Risk Activity

Table of Contents

During tunnelling through mixed ground conditions, boulders or rock fragments may
be removed from the ground. If these are too large to pass through the cutterhead openings,
they will be rotated with the TBM cutterhead until broken up. This may cause overbreak.
If a boulder/rock fragment is trapped in the cutterhead (Figures 8.2 and 8.3), the part
projecting ahead of the TBM will disturb the excavated face. Whether a boulder or rock
fragment is plucked from the face by the cutting tools depends on a number of factors,
including the size of the boulder or the joint spacing in the rock mass, the strength of the
matrix soil or rock joints, the forces applied by the TBM and its rate of advance, and the type
and condition of the cutting tools. Applying the correct chamber pressure will help to

Table of Contents

Provide remote camera facilities in the cutterhead to allow the

face stability to be checked before entering the excavation
chamber, and also to monitor the team during the intervention.
Prior to start of tunnelling, provide planned compressed air
pressure along the full tunnel alignment, in case of need for
emergency access.
Provide service penetrations through pressure bulkhead to avoid
running lines (i.e. compressed air for hand tools, water,
electricity, drainage, communications) through air locks.
Ensure any soil adhering to door through main pressure
bulkhead is cleaned off at the start of intervention - so that it can
be closed rapidly and securely in the event of instability.
Provide external seal around tailskin (between skin and
rock/soil) and/or inject polyurethane grout around skin prior to
entry, to minimise water flow along shield towards chamber.
Provide means to close head openings during intervention to
minimise water inflow and provide ground support (not possible
on most TBMs).
Carry out inspection of the exposed face at the start of the
intervention, and at regular intervals during the intervention.
Minimise the need for long interventions. Generally, it is
better to have more numerous, short, interventions to inspect
and replace (as necessary) the cutting tools than to risk the much
longer intervention(s) than may be needed if there is major
damage to the cutterhead due to operating with worn or
damaged cutting tools. For long drives in abrasive ground, a
combination of relatively frequent interventions to
inspect/replace cutting tools with occasional long interventions
at intermediate shafts or at safe havens (see above) can reduce
the risk associated with the interventions.

Table of Contents

Chamber/head access
(intervention)
(Contd)

Mitigation Measures

Table of Contents

Table 8.1 High Risk Activities and Some Mitigation Measures That Can Be Considered
for Reducing the Likelihood of Loss of Ground during EPB TBM Tunnelling
(Sheet 4 of 4)

69

Table of Contents

minimise the likelihood of this occurring, by preventing seepage flow towards the face.
Seepage will tend to reduce the strength of the soil matrix. Interventions may be necessary
to remove trapped boulders/rock fragments if these are causing significant damage to the
cutting tools or overbreak. Regular interventions for the inspection and maintenance of the
cutting tools will help to ensure that the boulders/rock fragments are broken up into small
pieces and extracted by the TBM rather than plucked out of the face.

Table of Contents

Figure 8.1 Typical Launch Seal (Courtesy of Richard Lewis)

Table of Contents
Table of Contents

Figure 8.2 Example of Rock Fragment Trapped in the Cutterhead

70

Table of Contents

Prior to a compressed air intervention, part (usually) or all (occasionally) of the paste
in the excavation chamber has to be removed, and the support pressure replaced, where
necessary, by compressed air pressure. When it is necessary or possible to form a filter cake
on the face there are intermediate steps of introducing the bentonite slurry, over-pressurising
the slurry, and then replacing it with compressed air. At the end of the intervention the
chamber has to be refilled, ready for the start of tunnelling in EPB pressure. The excavation
paste is only formed as the excavation re-commences. The face pressure has to be

Table of Contents

As discussed in Section 3.7, access to the head of the machine is a risk for both loss of
ground and the safety of those accessing the head. One method of reducing the risk to safety
of personnel is to temporarily close some or all of the openings in the cutterhead.
Temporary closure of the openings may also help to control loss of ground during the
intervention. However, if the cutterhead has been retracted to allow the tools to be changed,
significant volume loss can still occur if the ground collapses onto the cutterhead. This type
of machine with closure doors or plates is not common due to the constraints such systems
place on the overall design of the cutterhead.

Table of Contents

Similarly, rock fragments, which are too large to pass through the EPB screw conveyor
will get trapped and/or block the conveyor. This may stall the screw conveyor, lead to
excessive wear, damage or create a void in the screw conveyor, which results in deterioration
of the pressure plug formed by the screw conveyor. Ultimately this can lead to a blow-out
through the screw conveyor. To reduce the possibility of large fragments entering the screw
conveyor, bars can be installed across the cutterhead openings to limit the size of fragments
passing through into the excavation chamber. Screw conveyors should be able to allow the
operator to reverse their normal rotation to assist in freeing any blockage. Work procedures
to undertake this safely must be provided.

Table of Contents

Figure 8.3 Example of Rock Fragment Trapped in the Opening of the Cutterhead

71
maintained above the minimum pressure required for stability, at all times, during these stages;
this requires skill and experience.

Table of Contents

A limit is specified under the Factories and Industrial Undertakings (Work in Compressed Air) Regulation 12
that no person shall be employed in compressed air at a pressure exceeding 50 pounds per square inch, which
is equivalent to 3.45 bars, without permission from the Commissioner for Labour except in the case of an
emergency.

Table of Contents

Table of Contents

At the beginning of the intervention, the excavated face should be inspected, to check
that the ground and other conditions are safe for intervention using the planned compressed
air pressure (if any) and work procedures. Face pressures and intervention pressures should
be calculated for every ring based on the daily reviews and included in the daily excavation
instruction sheet, sometimes called a permit to excavate. Safe havens should also be
considered where intervention pressures higher than 3.45 bars may be required. 2 However,
even if safe havens are prepared and used, there should always be a contingency plan for
interventions in the case of emergency situations outside the safe haven. The contractors
work procedures should identify the key staff or grade of staff, who must be suitably
experienced to undertake this inspection. Unless the face is in stable rock or treated ground,
the inspection should be made in compressed air, and the air pressure should only be reduced
if it is confirmed from the inspection, by a competent professional who is qualified and

Table of Contents

The risk posed by interventions can be reduced by minimising the number of

interventions necessary in particularly settlement sensitive areas, such as when tunnelling
directly under buildings or other tunnels. The cutterhead should be inspected and all
necessary tool changes and maintenance carried out just before entering the settlement
sensitive area. This precaution should reduce the need for interventions in the settlement
sensitive areas. The locations for precautionary interventions can be planned in advance of
tunnelling, investigated for and reviewed and, where necessary, adjusted during tunnelling.
Other than the precautionary interventions, it is not possible to plan the location of most
interventions in advance, as the tunnel crew will have to respond to any indications that
maintenance is required due to wear or damage. However, general planning for face
interventions should be carried out, as far as practicable, prior to the start of tunnelling,
including face pressures, work procedures, precautionary measures and contingency plans.
The face pressure, if an intervention is required, can be calculated at the same intervals as the
face pressure for tunnelling, so that there is a default value available if an intervention is
required. However, assumptions regarding pore pressures should be reassessed prior to
entry if other works have affected them (e.g. dewatering), emphasising the requirement for
sufficient piezometers and on-going pore pressure measurement. The contractors work
procedures should identify the level of authorisation required in the contractors organisation,
and where necessary, by the Engineers site supervisory staff, before undertaking each
intervention. In settlement sensitive areas this authorisation should come from the senior
management personnel responsible for risk management. Where tunnelling is continuously
in settlement sensitive areas for long distances, risk can be reduced by preparing safe havens,
to allow major maintenance and repair of the TBM. The safe havens typically consist of a
block of grouted or otherwise treated ground, providing improved face stability during head
access. The head of the TBM is driven into the safe haven for a major intervention. As a
general rule in abrasive ground, manned interventions should be carried out at fairly frequent
intervals such as every 24 or 48 hours. This is to confirm that no undue wear or breakage is
taking place with the cuttertools.

72

Table 8.1 provides examples of possible mitigation measures, which can be considered,
to reduce the likelihood of loss of ground during TBM tunnelling. Another means to reduce
risk is to mitigate the consequences in the event of a loss of ground. In Table 8.2, a number
of possible mitigation measures for the consequences of ground loss are listed. Most of
these measures have been used in practice, either on their own or in combination with the
measures given in Table 8.1. The measures in Table 8.2 should only be applied where
appropriate, based on the risk assessment for the particular project and in consultation with
the stakeholders. The list in Table 8.2 is not exhaustive, and other mitigation measures not
listed in the table may be appropriate in particular circumstances.

Purpose

Table of Contents

Table 8.2 Some Mitigation Measures That May Be Considered for Controlling the
Consequences of Excessive Loss of Ground during EPB TBM Tunnelling
(Sheet 1 of 2)

Table of Contents

experienced to assess the risk of ground instability, that it is safe to do so. The initial
inspection should be made before entering the excavation chamber, by looking through the
access door, so that the door can be closed quickly if the conditions are unsafe. After the
initial assessment, a more detailed assessment can be made from within the excavation
chamber.

Mitigation Measures

Mitigating the effects

- buildings

Temporary evacuation as the TBM passes beneath the building.

Temporary propping of parts of the building.
Strengthening the foundations by underpinning or forming a raft
under the building or by compensation grouting.
Strengthening the building to span over a cavity.

Mitigating the effects

- roads

Temporary, localised, road closure above the tunnel.

Construct thickened road slab to bridge over any cavity that
might develop; ducts should be provided through the thickened
road slab to allow any cavities that develop to be filled. When
a thickened road slab has been constructed, take positive steps
to look for cavities below the road slab, which may not be
noticed for some time after the TBM has passed.

Table of Contents

Grouting the cavity caused by the ground loss before it reaches

the surface; this will work only where it is possible to detect the
cavity quickly and there is enough time for the grouting to be
carried out before collapse of the cavity. This measure is not
suitable for soft clay or granular soils under the water table,
where the cavity will likely migrate to the surface within a very
short time.
Purging the trapped air with a dedicated system to prevent
possible small collapses of ground at the crown caused by foam
degradation (which could result in formation of air bubbles
trapped on top of the excavation chamber).

Table of Contents

Intercepting the
ground loss

73

Purpose

Mitigation Measures

Mitigating the effects

- utilities

For water, sewage and gas pipes, ensure that there are valves to
close off the utility if necessary, and that this can be done at
short notice.
Shut down the utility temporarily as the TBM passes beneath it.
Construct a utility bridge to support the utility to cope with the
event of a cavity forming.
Grout the ground beneath the utility.

AGS (HK) (2004a).

Ground Investigation Guidelines 04.3 - Deep Excavations.
Association of Geotechnical and Geoenvironmental Specialists (Hong Kong), 2 p.
AGS (HK) (2004b). Ground Investigation Guidelines 04.4 - Rock Tunnels.
Geotechnical and Geoenvironmental Specialists (Hong Kong), 2 p.

Table of Contents

9 References

Table of Contents

Table 8.2 Some Mitigation Measures That May Be Considered for Controlling the
Consequences of Excessive Loss of Ground during EPB TBM Tunnelling
(Sheet 2 of 2)

Association of

AGS (HK) (2005). Ground Investigation Guidelines 04.6 - Soft Ground Tunnelling.
Association of Geotechnical and Geoenvironmental Specialists (Hong Kong), 2 p.

Atkinson, J.H. & Mair, R.J. (1981). Soil mechanics aspects of soft ground tunnelling.
Ground Engineering, UK, pp 20-26.

Table of Contents

Anagnostou, G. & Kovari, K. (1996). Face stability in slurry and EPB shield tunnelling.
Proceedings of the Symposium on Geotechnical Aspects of Underground Construction
in Soft Ground, London, pp 379-384.

Babendererde, S., Hoek, E., Marinos, P.G. & Cardoso, A.S. (2005). EPB-TBM face support
control in the Metro do Porto Project, Portugal. Proceedings of the 2005 Rapid
Excavation & Tunneling Conference, Seattle, pp 1-12.

Broere, W. (1998). Face Stability Calculation for a Slurry Shield in Heterogeneous Soft
Soils. Tunnels and Metropolises, Sao Paolo, Brazil, 1998, pp 215-218.
Broere, W. (2001). Tunnel Face Stability & New CPT Applications.
University of Delft, 194 p.

PhD thesis, Technical

Table of Contents

Bezuijin, A. & Talmon, A.M. (2006). Grout properties and their influence on back fill
grouting. Proceedings of the Symposium on Geotechnical Aspects of Underground
Construction in Soft Ground, Amsterdam, pp 187-195.

74

BTS/ICE (2005). Closed-face Tunnelling Machines and Ground Stability: A Guideline for
Best Practice. British Tunnelling Society with the Institution of Civil Engineers, UK,
77 p.

Table of Contents

Broere, W. (2003). Influence of excess pore pressures on the stability of the tunnel face.
(Re)Claiming the Underground Space, International Tunnelling and Underground
Space Association, Amsterdam, pp 759-765.

Chiriotti, E., Jackson, P. & Taylor, S. (2010). Earth Pressure Balance tunnel boring
machines, experience in mixed face conditions. Proceedings of the ITA-AITES World
Tunnel Congress entitled "Tunnel Vision Towards 2020", Vancouver.
Dimmock, P.S. & Mair, R.J. (2007). Estimating volume loss for open-face tunnels in
London Clay. Proceedings of the Institution of Civil Engineers, Geotechnical
Engineering, UK, vol. 160, Issue GE1, pp 13-22.

EFNARC (2005). Specification and Guidelines for the Use of Specialist Products for
Mechanised Tunnelling (TBM) in Soft Ground and Hard Rock. Experts for
Specialised Construction and Concrete Systems, 45 p.

Table of Contents

Dobashi, I., Sakae, T., Hiroshige, N. & Kitayama, H. (1999). Test of twin-screw earth
pressure-balanced shield system at great depths. Proceedings of the 54th Annual
Conference of the Japan Society of Civil Engineers, Hiroshima University, Session 6,
pp 152-153.

GEO (2009). GEO Technical Guidance Note No. 24, Site Investigation for Tunnel Works.
Geotechnical Engineering Office, Hong Kong, 9 p.

ITA WG-14 (2007). Standards: Manuals of Mechanized Tunnelling.

Tunnelling and Underground Space Association, 2 p.

International

Table of Contents

Horn, N. (1961). Horizontaler Erddruck auf senkrechte Abschlussflchen von Tunnelrhren.

Landeskonferenz der Ungarishchen Tiefbauindustrie, Budapest, pp 7-16.

Jancsecz, S. & Steiner, W. (1994). Face support for a large Mix-Shield in heterogeneous
ground conditions. Tunnelling 94, London, pp 531-550.
Jancsecz, S., Krause, R. & Langmaack, L. (1999). Advantages of soil conditioning in shield
tunneling: experience of LRTS Izmir. Proceedings of International Congress on
Challenges for the 21st Century, Rotterdam, pp 865-875.

Kimura, T. & Mair, R.J. (1981). Centrifugal testing of model tunnels in soft clay.
Proceedings of the 10th International Conference on Soil Mechanics and Foundation
Engineering, Stockholm, Volume 1, pp 319-322.

Table of Contents

Jefferis, S.A. (2003). Long term performance of grouts and the effects of grout by-products.
Vol. 2. Proceedings of the Third International Conference on Grouting and Ground
Treatment, New Orleans, February Issue, pp 1141-1152.

75

Leblais, Y., Andre, D., Chapeau, C., Dubois, P., Gigan, J.P., Guillaume, J., Leca, E., Pantet, A.
& Riondy, G. (1999). Settlements induced by tunnelling. English Text of AFTES
Recommendations Originally Published (in French) in 1995, Tunnels et Ouverages
Souterrains, pp 129-151.

Table of Contents

Kovari, K. & Ramoni, M. (2006). Urban tunnelling in soft ground using TBMs.
Proceedings of the International Conference on Tunnelling and Trenchless Technology,
Subang, Selangor, Malaysia, pp 17-31.

Merritt, A., Jefferis, S., Storry, R. & Brais, L. (2013). Soil conditioning laboratory trials for
the Port of Miami Tunnel, Miami, Florida, USA. Proceedings of the ITA-AITES
World Tunnel Congress entitled "Underground the way to the future", Geneva.

OReilly, M.P. (1988). Evaluating and predicting ground settlements caused by tunnelling in
London Clay. Tunnelling 88, London, The Institution of Mining and Metallurgy,
pp 231-241.
Proctor, R.V. & White, T.L. (1977). Earth Tunnelling with Steel Supports.
Shearing and Stamping Co, Youngstown, 247 p.

Commercial

Table of Contents

Nakano, A., Sahabdeen, M.M., Kulaindran, A. & Seah, T.P. (2007). Excavation management
for slurry TBM tunnelling under residential houses at C853 (CCL3) Project.
Proceedings of the Underground Singapore 2007, Singapore, pp 38-45.

Shirlaw, J.N., Busbridge, J.R. & Yi, X. (1994). Consolidation settlements over tunnels, a
review. Proceedings of the 1994 TAC Annual Conference, Canada, pp 253-265.

Shirlaw, J.N. (2002). Controlling the risk of excessive settlement during EPB tunnelling.
Keynote Lecture. Proceedings of the Conference on Case Studies in Geotechnical
Engineering, Singapore, pp 147-174.

Table of Contents

Shirlaw, J.N., Hencher, S.R. & Zhoa, J. (2000). Design and construction issues for
excavation and tunnelling in some tropically weathered rocks and soils. Proceedings
of GeoEng2000, Melbourne, pp 1-44.

Shirlaw, J.N., Ong, J.C.W., Rosser, H.B., Tan, C.G., Osborne, N.H. & Heslop, P.J.E. (2003).
Local settlements and sinkholes due to EPB tunnelling. Proceedings of the Institution
of Civil Engineers, Geotechnical Engineering, UK, Vol. 156, Issue GE4, pp 193-211.

Shirlaw, J.N., Westland, J. & Doran, S.R. (2005). Planning for mechanised tunnelling in soft
and mixed ground. Proceedings of the International Congres Les Tunnels, Cl d'une
Europe Durable (Tunnelling for a Sustainable Europe), Chambery, pp 177-182.

Table of Contents

Shirlaw, J.N., Richards, D.P., Raymond, P. & Longchamp, P. (2004). Recent experience in
automatic tail void grouting with soft ground tunnel boring machines. Proceedings of
the ITA-AITES World Tunnel Congress, Singapore, Vol. 19, July-September Issue.

76

Venkta, R., Hoblyn, S., Mahatma, S. & Lim, H.C. (2008). EPB tunnelling under 2-storey
shophouses in mixed face conditions. Proceedings of the International Conference
on Deep Excavations (ICDE 2008), Singapore.

Table of Contents

Thewes, M. (2007). TBM tunnelling challenges - redefining the state of the art. Collection
of Keynote Lectures, Proceedings of the ITA-AITES World Tunnel Congress, Prague,
pp 13-21.

Young, D.J. & Dean, A. (2010). Estimating face pressures for closed face tunnelling in soft
ground. Proceedings of the ITA-AITES World Tunnel Congress, Vancouver.

Table of Contents
Table of Contents
Table of Contents

77

Table of Contents

Table of Contents

Glossary of Terms

Table of Contents

Table of Contents

78

Glossary of Terms

Extrados The outside face of a structural element, e.g. tunnel lining extrados.
Invert The lowest surface of a tunnel.

Table of Contents

(A) General terms related to tunnelling, used in this report:

Overcut gap The annulus around the shield skin created by the difference between the
excavated diameter and the outside diameter of the shield skin. The size of the
annulus varies due to different diameters of different parts of the shield skin. It is
also affected by steering of the shield, as well as by the degree of overcutting.

Tail void The annulus around the extrados of the tunnel, due to the difference between the
outside diameter of the lining and the excavated diameter. Also known as the tail gap
or annular space.

Table of Contents

Spring line In a circular tunnel, the spring lines are at the opposite ends of the horizontal
centreline. For a circular tunnel the spring line is also known as the axis level.

(B) Terms relating to EPB TBM tunnelling, used in this report:

Additive Additives are materials that are added at the cutterhead, into the excavation
chamber and/or the screw conveyor to improve the properties of the excavated material.
The purpose is to alter the properties of the excavated material to form a suitable paste
for EPB operation.
Table of Contents

Auxiliary Face Support (AFS) system A system based around a tank of bentonite slurry
within the TBM backup. The tank is linked to the excavation chamber, such that if
the face pressure drops below a pre-set value, the slurry is automatically injected into
the chamber to maintain the face pressure.
Bentonite annular injection TBM-mounted system to inject bentonite slurry around the
shield in squeezing ground. Also to maintain the support pressure between the face
and the annular grouting zone.
Conditioning agent Conditioning agents include foams, polymers, binders and bentonite
slurry, either alone or in combination. Synonymous with additive.

Earth Pressure Balance (EPB) TBM An EPB TBM is a closed face tunnel boring machine.
This type of TBM was originally developed primarily for cohesive soils, but has since
been adapted for use in a wide range of soil and weathered rock. The excavated
material is mixed, usually with injected additive(s), in the excavation chamber to form

Table of Contents

Discharge gate The opening at the upper end of the screw conveyor where the excavated
material discharges at atmospheric pressure.

79

Table of Contents

a paste, which must completely fill the excavation chamber and is then discharged via
a screw conveyor. The excavation paste is pressurised against the pressure bulkhead
by the propulsion thrust exerted by the TBM thus providing active face support. The
face support pressure can be altered by changing the relative rate at which material
enters the excavation chamber (defined by the rate of TBM advance) compared with
the rate at which excavation paste is removed (through the screw conveyor). To
maintain this pressure, the paste must form a plug in the screw conveyor that supports
the pressure difference between the inlet and discharge ends of the screw conveyor
(Figure 1.1).
Foaming agent Surfactant chemical used as an additive to aerate the excavation paste and
provide confinement, aid excavation, modify the density of the paste and reduce wear
of the TBM components.
Intervention Access into the excavation chamber of the machine.

Pressure bulkhead The pressure bulkhead separates the excavation chamber of the TBM
from the rest of the TBM, separating the pressurised area from the non-pressurised
area (Figure 1.1).

Table of Contents

Polymer Additive often used to enhance the viscosity of the excavation paste or to stabilise
the foam bubbles. Usually applied to homogenise excavated granular materials. In
clay soils, some polymers act as a dispersant to reduce stickiness.

Ribbon conveyor A type of screw conveyor fabricated without a central shaft to allow larger
solid objects to pass along the screw.

Shove Advance of the machine to allow one ring to be installed.

Table of Contents

Screw conveyor A screw conveyor is used to transport excavation paste from the excavation
chamber, and also forms part of the regulation system for the face pressure. The
screw conveyor is largely housed in a casing, except for the section forward of the
pressure bulkhead. Refer to Figure 4.1 for a screw conveyor and discharge point.

Soil conditioning Soil conditioning techniques are used to alter the properties of the
excavated materials to make them more suitable for excavation by the TBM. This
technology can also expand the range of soil conditions suitable for EPB TBM
tunnelling.

Table of Contents

80

Table of Contents

Table of Contents

Glossary of Symbols

Table of Contents

Table of Contents

81

Glossary of Symbols

Ground cover over the tunnel (depth from ground level to the highest excavated
point of the tunnel)

Excavated diameter of the tunnel

D1

Internal diameter of the lining

D0

External diameter of the lining

Unsupported length of the tunnel heading

Z0

Depth from the ground surface to the axis level of the tunnel

ZS1

Depth from the ground surface to pressure sensor 1

ZS2

Depth from the ground surface to pressure sensor 2

ZS3

Depth from the ground surface to pressure sensor 3

ZW

Depth from ground surface to the water table, or to the piezometric level at the
tunnel, as appropriate 1

Bulk unit weight of the ground (soil or rock, as appropriate)

'

Submerged unit weight of the ground (soil or rock, as appropriate) 2

Dry unit weight of the ground (soil or rock, as appropriate)

Unit weight of the excavated material (including additives)

Unit weight of water

This report assumes a simple hydrostatic distribution down to the level of the tunnel. The equations provided
may not be appropriate for more complex groundwater regimes, the influence of which should be assessed
carefully in the design.

This report is for tunnelling through saturated soils below the water table.

Table of Contents

Table of Contents

(B) Unit weight

Table of Contents

Table of Contents

(A) Dimensions

82
(C) Pressure and external loads

PE

Pressure in the excavation chamber

PE(S1)

Pressure measured at sensor 1 (or (S2) for sensor 2, etc.)

PEt

Target face pressure in the excavation chamber

PEt(Crown)

Target face pressure at the crown of the tunnel

PEt(S1)

Target face pressure at sensor 1 (or (S2) for sensor 2, etc.)

Average surcharge pressure at the ground surface

Maximum variation in applied face pressure, due to control accuracy, with v being
the change in pressure above OR below the target face pressure (Figure 3.3)

Table of Contents

Compressed air pressure

Table of Contents

PCA

Table of Contents
Table of Contents

GEO PUBLICATIONS AND ORDERING INFORMATION

Copies of GEO publications (except geological maps and other

publications which are free of charge) can be purchased either
by:

(
):

Writing to
Publications Sales Unit,
Information Services Department,
Room 626, 6th Floor,
North Point Government Offices,
333 Java Road, North Point, Hong Kong.

333
6626

or

Calling the Publications Sales Section of Information Services

Department (ISD) at (852) 2537 1910
Visiting the online Government Bookstore at
http:// www.bookstore.gov.hk
Downloading the order form from the ISD website at
http://www.isd.gov.hk and submitting the order online or by
fax to (852) 2523 7195
Placing order with ISD by e-mail at [email protected]

((852) 2537 1910)

http://www.bookstore.gov.hk
(http://www.isd.gov.hk)
((852) 2523
7195)
([email protected])

Map Publications Centre/HK,

Survey & Mapping Office, Lands Department,
23th Floor, North Point Government Offices,
333 Java Road, North Point, Hong Kong.
Tel: (852) 2231 3187
Fax: (852) 2116 0774

333
23

: (852) 2231 3187

: (852) 2116 0774

Requests for copies of Geological Survey Sheet Reports and

other publications which are free of charge should be directed to:

For Geological Survey Sheet Reports which are free of charge:

Chief Geotechnical Engineer/Planning,
(Attn: Hong Kong Geological Survey Section)
Geotechnical Engineering Office,
Civil Engineering and Development Department,
Civil Engineering and Development Building,
101 Princess Margaret Road,
Homantin, Kowloon, Hong Kong.
Tel: (852) 2762 5380
Fax: (852) 2714 0247
E-mail: [email protected]

:
101

(:)
: (852) 2762 5380
: (852) 2714 0247
: [email protected]

For other publications which are free of charge:

Chief Geotechnical Engineer/Standards and Testing,
Geotechnical Engineering Office,
Civil Engineering and Development Department,
Civil Engineering and Development Building,
101 Princess Margaret Road,
Homantin, Kowloon, Hong Kong.
Tel: (852) 2762 5346
Fax: (852) 2714 0275
E-mail: [email protected]

:
101

: (852) 2762 5346

: (852) 2714 0275
: [email protected]

Table of Contents

1:100 0001:20 0001:5 000

Table of Contents

1:100 000, 1:20 000 and 1:5 000 geological maps can be
purchased from:

Table of Contents

http://www.cedd.gov.hk

Table of Contents

A selected list of major GEO publications is given in the next

page. An up-to-date full list of GEO publications can be found at
the CEDD Website http://www.cedd.gov.hk on the Internet under
Publications. Abstracts for the documents can also be found at
the same website. Technical Guidance Notes are published on
the CEDD Website from time to time to provide updates to GEO
publications prior to their next revision.

MAJOR GEOTECHNICAL ENGINEERING OFFICE PUBLICATIONS

Geotechnical Manual for Slopes, 2nd Edition (1984), 302 p. (English Version), (Reprinted, 2011).
(1998)308(1984)
Highway Slope Manual (2000), 114 p.

Table of Contents

GEOTECHNICAL MANUALS

GEOGUIDES
Guide to Retaining Wall Design, 2nd Edition (1993), 258 p. (Reprinted, 2007).

Geoguide 2

Guide to Site Investigation (1987), 359 p. (Reprinted, 2000).

Geoguide 3

Guide to Rock and Soil Descriptions (1988), 186 p. (Reprinted, 2000).

Geoguide 4

Guide to Cavern Engineering (1992), 148 p. (Reprinted, 1998).

Geoguide 5

Guide to Slope Maintenance, 3rd Edition (2003), 132 p. (English Version).

(2003)120()

Geoguide 6

Guide to Reinforced Fill Structure and Slope Design (2002), 236 p.

Geoguide 7

Guide to Soil Nail Design and Construction (2008), 97 p.

Table of Contents

Geoguide 1

GEOSPECS
Geospec 1

Model Specification for Prestressed Ground Anchors, 2nd Edition (1989), 164 p. (Reprinted,
1997).

Geospec 3

Model Specification for Soil Testing (2001), 340 p.

GEO PUBLICATIONS
Review of Design Methods for Excavations (1990), 187 p. (Reprinted, 2002).

GEO Publication
No. 1/93

Review of Granular and Geotextile Filters (1993), 141 p.

GEO Publication
No. 1/2006

Foundation Design and Construction (2006), 376 p.

GEO Publication
No. 1/2007

Engineering Geological Practice in Hong Kong (2007), 278 p.

GEO Publication
No. 1/2009

Prescriptive Measures for Man-Made Slopes and Retaining Walls (2009), 76 p.

GEO Publication
No. 1/2011

Technical Guidelines on Landscape Treatment for Slopes (2011), 217 p.

The Quaternary Geology of Hong Kong, by J.A. Fyfe, R. Shaw, S.D.G. Campbell, K.W. Lai & P.A. Kirk (2000),
210 p. plus 6 maps.
The Pre-Quaternary Geology of Hong Kong, by R.J. Sewell, S.D.G. Campbell, C.J.N. Fletcher, K.W. Lai & P.A.
Kirk (2000), 181 p. plus 4 maps.

TECHNICAL GUIDANCE NOTES

TGN 1

Technical Guidance Documents

Table of Contents

GEOLOGICAL PUBLICATIONS

Table of Contents

GCO Publication
No. 1/90