THE RISK OF VERY LARGE SETTLEMENTS DUE TO EPB TUNNELLING

Shirlaw, J.N.
Golder Associates, Singapore
Boone, S.
Golder Associates, Mississauga, Canada

ABSTRACT

Tunnelling using EPB shields generally allows good control of surface settlement. However,
there are reports from many projects of occasional, localised, sudden, and very large
settlements or sinkholes over EPB driven tunnels. There are a number of readily identifiable
risk factors associated with sudden, very large, localised settlements. The relative influence
of these risk factors is identified using detailed data from seven major EPB projects with
which the authors have been involved. The cases are discussed and illustrated, and high risk
areas are identified. Measures to minimise the risk of large settlements are outlined, and
issues related to risk assessment are discussed.

1 INTRODUCTION

The settlement caused by tunnelling can be a major issue, particularly in urban areas. One of
the major advantages of the Earth Pressure Balance (EPB) and Slurry tunnel boring machines
that have been introduced in the last few decades is that surface settlements can generally be
kept to a very small magnitude. However, locally large settlements or sinkholes have also
been recorded on projects using these types of machines. As a large local settlement under a
building, rail track, other settlement sensitive structure or utility would have serious
consequences, it is important to understand why such large settlements occur, and how they
can be avoided. In this paper the focus will be on tunnelling using EPB shields, although the
problems are not necessarily limited to EPB shields.

In principle, settlements during EPB tunnelling can be minimised simply by ensuring that:

• the face pressure is maintained at approximately the full overburden pressure

• the shield is kept on alignment, with as little ploughing and overcutting as possible,
and
• the tail void is filled immediately with a grout that does not shrink or bleed

In soft or loose soils only a pressure close to full overburden pressure will minimise
settlement, as shown by Shirlaw et al (1). Settlements increase rapidly with reducing face
pressure in these types of soil. Stronger soils are relatively less sensitive to reductions in face
pressure than soft or loose soils. While there will be some increase in settlement with
reducing face pressure, this increase is often small enough to be tolerable in all but the most
settlement sensitive areas. In soft clay, excessive face pressure will result in the development
of excess pore pressures that then dissipate, resulting in long term consolidation settlements
that have the same shape and width as the immediate settlements (Shirlaw (2)). It is therefore
unwise to exceed the full overburden pressure by more than 10% to 15%.

In order to fill the tail void as it is created, it is necessary to grout simultaneously with the
advance of the shield. To do this effectively requires grout pipes that are placed along the
tailskin and through the tail seals, as conventional grouting through the rings cannot be
carried out quickly enough in soft or granular soils to ensure that the grouting keeps the gap
filled at all times.

Essentially, minimising settlement requires that adequate support is applied to the ground at
all times. This appears to be both simple and reasonably achievable with EPB machines. It is
therefore worth studying the exceptional cases of large settlements and sinkholes, to
understand their causes and to learn how to avoid such instances in future.

2 CASES OF EXCEPTIONALLY LARGE SETTLEMENT

Exceptionally large settlements over tunnels can be divided between those that are not visible
to the naked eye, but can be measured by survey, and those that can be seen. Typically,
localised settlements need to be at least 150mm before they become obvious to the eye, and
then only if they occur on roads or other paved surfaces.

It is common to express settlement over tunnels in terms of volume loss, which is the unit
volume of the surface settlement trough divided by the unit volume of the excavated tunnel,
expressed as a percentage. For example, for a 6m diameter tunnel driven through clay at a
depth of 25m, a settlement of 150mm equates to about 16.5% volume loss. Design values for
volume loss over EPB driven tunnels are commonly taken at between 1% and 3%, depending
on the ground. The value taken would reflect the upper limit that the designer would expect,
based on good tunnelling practice. For the purpose of this paper, anything over 3% volume
loss will be considered an exceptional settlement.
100.0

90.0

80.0

70.0
Cases, percent

60.0

50.0

40.0

30.0

20.0

10.0

0.0
< 0.49 0.5 to 1 to 1.49 1.5 to 2 to 2.49 2.5 to 3 to 3.99 4 to 4.99 >5
0.99 1.99 2.99

V olu m e Loss, percent

Figure 1. A summary of the volume losses recorded at 485 monitored settlement points
over EPB tunnels, North East Line, Singapore
To study the risk of exceptional settlements occurring, it is necessary to take the data from a
large number of settlement points installed over long lengths of tunnelling. Shirlaw et al. (1)
did this for the construction of 20 kilometres of subway tunnelling in Singapore. Figure 1
summarises the overall results of the study. Of the 485 points included, 37 (7%) recorded
settlements equivalent to a volume loss of over 3%. The study also recorded 16 cases of very
large settlement, exceeding 150mm; many occurred as localised sinkholes. The subsurface
conditions in Singapore are severe for EPB tunnelling, with rapid changes, including many
stretches of mixed face conditions. However, the development of localised, large surface
settlements and sinkholes over EPB driven tunnels has been recorded over many tunnels and
in many countries, so this phenomenon should not be considered as being peculiar to
Singapore. Table 1 provides a summary of some of the published references to large local
settlements and sinkholes over EPB driven tunnels. Only a few of these references give the
number, size and causes of the settlements and sinkholes in any detail. Many owners and
contractors are unwilling to record such cases at all, and it is likely that the published record
is only a sample of what has really occurred.

Project Localised settlement problems Reference

documented
Changi Line, Singapore One case of settlement>150mm Raju et al (3)
Anacostia River, USA Settlements> 150mm and a Clough and Leca (4)
number of sinkholes
Taipei RTS, ROC Very large settlements and Moh et al. (5)
sinkholes
Sheppard Line, Canada Fifteen cases of large settlement Boone et al (6)
Storebaelt, Denmark Many large seabed depressions, Biggart and Sternath (7)
up to 2200m3
Railway tunnels, Japan Sinkhole in railway embankment Kimura (8)
Lille Metro, Belgium Several sinkholes Leblais et al (9)
Oporto Metro, Portugal Sinkholes, one causing a building Anon (10)
to collapse
Shenzen Metro, China Sinkhole in a golf course Wallis (11)
CTRL, England Sinkhole in the garden of a home Anon (12)
Table 1. Some published cases of large settlements and sinkholes over EPB driven tunnels.

The authors have summarised the number of cases of large, visible settlements and sinkholes
for seven major projects involving EPB tunnelling with which they have been directly
involved (Table 2). There were 57 incidents in 76.9km of tunnelling, an average rate of one
incident per 1.35km. This table only includes cases of large, visible settlements. Less visible
settlements, but which exceeded 3% volume loss, are not included.

Tunnel Year Length Incidents No./km

Singapore MRT Phase 2 1986-1987 1.6 0 0
St Clair River, Canada 1993-1994 1.9 3 1.58
Allen Sewer, Canada 1994 1.1 1 0.9
Sheppard Line, Canada 1997-1998 6.4 15 2.3
Changi Line, Singapore 1998-1999 7 1 0.1
North East Line, Singapore 1998-2000 20 16 0.8
Tunnels, Singapore 1999-2005 38.9 (to October 2003) 21 0.5
Table 2. Cases of large (>150mm) settlement or sinkholes over EPB driven tunnels on
seven major projects
3 THE CAUSES OF EXCEPTIONALLY LARGE SETTLEMENTS

As discussed in Shirlaw et al (1) the records of the settlement over the tunnels in Singapore
show that most of the cases of volume loss of over 3% were directly related to the use of an
inadequate face pressure. For the cases listed in Table 2, the settlements were often so large
that they can only be explained in terms of loss of ground at tunnel level, and must also be
due to an inadequate face pressure. All of the incidents were localised and isolated, so the
question is why, suddenly, the machine operator applied inadequate face pressure when, for
the rest of the tunnelling, sufficient pressure was applied. Reviewing both this specific data
and the more general cases summarised in Table 1, there appear to be a number of recurring
factors that are associated with a number of cases:

• Launching the shield

• Breaking into the shaft at the completion of the drive
• Driving from strong, stable ground into weaker, unstable ground, or vice versa
• A mixed face of rock or hard cohesive soil and potentially unstable soil, particularly
saturated granular soils
• Head access for maintenance or inspection
• Mechanical problems, particularly during long drives in abrasive ground

In Table 3 the 57 documented cases are divided by the major associated factor in each case,
except for four cases which do not fall readily into any of these categories.

Risk area Number

Launching the shield 10
Breaking into recovery shaft 4
Interfaces between stable and unstable soils 6
Mixed faces of rock, or hard cohesive soil, and granular soil 15
Head access for maintenance 13
Mechanical failures 5
Others 4
Total 57
Table 3. Large settlement/ sinkhole cases for seven major projects, divided by main risk
areas

It appears that the factors listed in Table 3 represent the locations where there is a significant
risk that the operator will not apply a face pressure adequate to control the ground. It is
important to emphasise that this is a risk, not a certainty. In the major projects reviewed, there
were many instances of successful EPB tunnelling through locations where one of the risk
factors were present, but where there was no excessive settlement.

During the launching of the shield the operator has to build up the face pressure from zero to
full operating pressure. He may also have to deal with the rapidly varying conditions
resulting from local grouting or from disturbance of the ground associated with the
construction of the launching shaft. During launching, there is an open annulus around the
machine leading back into the shaft which can act as a conduit for ground and water. All of
the factors, combined, result in the launching of the shield being one of the major risk areas
for excessive settlements. The break into the shaft at the end of the drive is basically the
reverse of the launch, with the same attendant risks.
A potentially unstable soil, such as a soft clay or uniform sand, requires a face pressure close
to full overburden pressure to minimise settlement. In contrast, tunnelling through a stronger
soil, such as hard clay, can require relatively less face pressure to adequately control
settlement. In a transition from one soil to the other it will be necessary to change the face
pressure to suit the changing ground conditions. However, it is not adequate to adjust the
pressure only as the different soil is encountered. In the transition from strong soil to weak
soil the face pressure must be raised in advance of the transition. In the reverse case, any
reduction in pressure can only be made once there is an adequate cover of the stronger soil
over the tunnel. Managing the face pressure in the transition therefore requires a detailed
knowledge of the ground conditions ahead of the shield and careful planning. A lack of
information about the location of the interface and/or a lack of planning will result in a high
risk of settlement problems developing locally.

A mixed face of rock or hard cohesive soil and granular soil presents two problems. Part of
the face consists of stable ground that takes time to cut; meanwhile the other part of the face
is in a highly mobile soil that moves readily into any gap. The vibration associated with
machine excavation exacerbates the mobility of the granular soil. It is also difficult to mix the
excavated cuttings into a suitable EPB plug since the harder materials may remain as lumps
within a loose matrix of granular soils and/or water. This latter problem makes it difficult to
consistently maintain the necessary pressure to control the granular soil. A further problem in
these conditions, which often vary from ring to ring, is the complexity of the tasks faced by
the shield operator. The operator has to attend to multiple controls for face openings
(“doors”), forward thrust, head RPM, screw conveyor RPM, screw conveyor gate openings,
and conditioning agent (water, bentonite slurry, foam, polymers, etc.), while also checking
the volume excavated. Where there is a delicate, and changing, balance required between the
various controls there is a risk that the operator may not achieve that balance locally.

Head access involves emptying out part of the spoil in the plenum chamber. Pressure on the
face is typically maintained by the use of compressed air or the use of “doors” that cover the
face openings. However, there is a tendency to minimise the use of compressed air, both
because of health concerns and the time taken for decompression. Because of these concerns,
it is common to see face access attempted with a significantly lower compressed air pressure
than the pressure used for the previous tunnelling, or without compressed air. The face
“doors” may not fully seal, and there may be additional gaps surrounding openings for face
tools. This may result in soil and water entering the chamber. Both of these conditions can
result in major losses of ground.

Modern EPB and Slurry machines are complex. A malfunction of the machine can affect the
ability of the operator to maintain the pressure at the face, and can therefore result in
excessive settlement. Table 4 lists the mechanical problems that resulted in excessive
settlement on the seven projects listed in Table 2, plus an additional two cases where the
settlement was in the range 100mm to 150mm. Most of these mechanical problems resulted

Cause Number
Abrasion of screw conveyor 3
Failure of main pressure bulkhead 1
Jamming of discharge gate by obstructions 2
Failure of tail-seals 1
Table 4. Mechanical failures that lead to surface settlements of over 100mm
in loss of face pressure, leading to loss of ground. In one case the tail seals failed, leading to
an inflow of ground at the tail of the shield.

4 ISSUES FOR SPECIFICATION AND FOR REVIEW

The number and frequency of major settlement incidents, including sinkholes, over EPB
shields is of concern. Measures have to be taken to control this aspect of the tunnelling. A
key issue is that of the risk assessments carried out during design and in preparation for
construction. Risk assessments were prepared before the drives in Singapore where there
were cases of major ground loss. Typically, the risk assessments included a table showing
that with an EPB machine, a good operator and suitable conditioning, the risk of a major loss
of ground was negligible. On one contract the tunnelling contractor experienced seven large
sinkholes over his two EPB machines, so a revised risk assessment was required before
tunnelling could recommence. The ‘revised’ risk assessment still came to the conclusion that
the risk of a large loss of ground was ‘negligible’, based on the same premise; surely a
triumph of optimism over experience. The sheer number of major ground losses over EPB
machines, listed in Tables 1 and 2, shows that simply having an EPB machine, a good
operator and (putatively) good spoil conditioning is not adequate to reduce the risk of a major
ground loss to an acceptable level. In the projects listed in Table 2 the tunnelling was carried
out by contractors with extensive EPB tunnelling experience, experienced staff, and a
machine built by one of the major manufacturers. In all but one case the machine was
specifically built for that particular job. Clearly, these factors alone are not sufficient to
control the risk of excessive settlement to an adequate extent.

Another ‘control’ measure frequently cited in method statements and risk assessments is the
provision and assessment of numerous settlement points over the tunnel. However, the
ground loss is initiated at the tunnel face, and progresses towards the surface. It is therefore
generally too late to do anything about a highly localised ground loss by the time it is
identified by monitoring. The monitoring merely serves to help quantify the loss of ground,
provided the settlement point survives the incident and is located at the point at which the
localised settlement occurs.

Controlling any risk requires acceptance that the risk is real and serious. Only then will the
necessary steps be taken to control that risk. A first step is to consider whether a conventional
EPB shield is actually suitable for the ground conditions anticipated. Many of the major
ground losses listed in Table 3 occurred in mixed faces of granular soil and either rock or
hard cohesive soils. Doran and Athenoux (13) argue, on the basis of the Storebaelt
experience, that a slurry shield might be a better choice in these conditions. Maidl and

Cause
Obstructions, such as an old tree trunk or an abandoned drill casing
The learning curve after the shield launch proper
An experienced operator handing over temporarily to an inexperienced
one
Old wells
Continuing tunnelling even though the tail void grouting pipes are
blocked
The head ‘plugging’ due to inadequate mixing paddles
Table 5. Other recorded causes to which large ground losses were attributed
Nellessen (14) also comment on the risks involved in the selection of EPB machines in
particular ground conditions.

Where an EPB shield is chosen, the risk assessment should consider previous experience,
including the six particular areas of risk given in Table 3. Some other risk factors that have
been mentioned in various publications are listed in Table 5. Control measures for those risks
that may be present on a particular drive should be identified. However, the issue of large,
localised settlement is potentially so important that an owner may not wish to leave the
identification and selection of control measures entirely to the contractor.

Some issues that may be considered for inclusion in specifications by the owner are covered
in Tables 6, 7 and 8. Table 6 relates to general issues, such as muck volume and pressure
control, that apply to EPB tunnelling under all circumstances. Traditionally, the excavated
volume has been estimated from the number of muck skips filled. Even this relatively crude
indicator can provide warning of major ground losses during EPB tunnelling, Shirlaw et al.
(14). However, the ability to respond to over excavation can be improved by developing
systems that measure excavated volume in real time. If this information can be linked to the
on-board computer, the operator can obtain rapid feedback on the volume excavated. Osborne
et al. (15) provide a comparative assessment of the use of belt weighing and belt scanning
devices with the counting of muck skips. In any case, the simple counting of muck skips must
still back up automated systems.

Requirements in Specifications Areas for review

Contractor to:
• Identify key personnel, including operators,
Levels of authority with respect
managers, engineers and foremen to changing method statement
• Provide Risk Assessment (i.e. changing planned face
• Provide planned face pressure for every ring of pressure) must be defined
advance, and levels of authority for varying that
Risk should not be ignored or
pressure considered negligible simply
• Provide acceptable device to monitor excavated because of use of EPB or slurry
volume automatically in real time, displayed in machine
control cabin and correlated with skip counting.
Response to over-excavation
Measurement to be made at least every 300mm must be defined.
of advance.
• Identify contingency measures
• Provide sufficient mixing paddles/arms to avoid
plugging of the head
• Provide screw conveyor of sufficient length for
expected pressures / ground conditions
Table 6. Enhanced Specification requirements: General issues

Table 7 covers the six major areas of risk identified in Table 3. Some of the ways in which
specifications can be augmented to give greater control of risks are shown, although the table
is by no means exhaustive. Some issues for particular attention during review of the
contractor’s documentation are also covered. It should be noted that the six risk categories in
Tables 7 cover only the most likely conditions for a major loss of ground, and are, to some
extent, tailored for the ground conditions that are common in Singapore.
 Condition Requirements (on contractor) in Areas for review
Specifications
Shield launch • Provide minimum length of ground
The ground treatment should
treatment (specified by owner) be checked/probed before
• Provide detailed plan for how face breaking out tunnel eye
pressures will be built up to general
First buried ring to be grouted
operating level in place as soon as practicable
• Head should be pre-filled to the
extent practicable with soil materials
prior to launch
Shield recovery • Provide minimum length of ground
The ground treatment should
treatment (specified by owner) be checked/probed before
• Provide detailed plan for how face breaking out tunnel eye
pressures will be reduced from
Face pressures to be sufficient
general operating level to zero to control ground, but not too
high for temporary works

Interfaces • Provide detailed planning of face
Adequate site investigation to

between stable pressures through interface accurately define formation
& unstable • Probe if operating at face pressures boundaries
geological units below hydrostatic

Mixed Face • Provide detailed planning of face
In tropically weathered rocks

conditions pressures through interfaces, if interpretations from vertical
identified boreholes are seldom reliable
• Probing required if operating at face
Sufficient thrust and torque to
pressures below hydrostatic. handle full range of likely
conditions
Chamber/ • Give detailed procedure for access
If access in free air or a
Head access • Provide compressed air locks. pressure below existing
Compressed air access to be available hydrostatic pressure is
within 72 hours at all times. considered, procedure to
• Provide planned face pressure for confirm stability of ground
every ring of advance, in case of need before accessing head must be
for emergency access included
• Incorporate a system to close all head
A pre-planned schedule of
openings during access to the maintenance inspections, with
chamber interventions located in areas
• Provide service penetrations through where the ground conditions
bulkhead to avoid running lines are favourable and/or the
through air locks. consequences of ground loss
are a minimum

Mechanical • Provide foam generator
Use of foam to reduce wear

problems • Provide conveyor casing with have
Check how, and how often,
replaceable abrasion resistant different parts of the machine
components; hardness to match auger will be checked for wear –
• Provide full list for all assembly items including cutting tools,
to be double-checked prior to saddles, cutting head, pressure
machine launch bulkhead, screw conveyor,
screw conveyor housing, tail
seals, discharge gate.
Table 7. Examples of enhanced specifications - particular areas of risk
To avoid a major loss of ground during machine tunnelling it is necessary to have the correct
machine for the ground conditions and operate that machine appropriately. This requires that
all of those directly involved in the tunnelling, from the machine operator to the Project
Manager, plan and carry out the work correctly at all times. This cannot be ensured simply by
improving specifications, but changes in the specifications can help to ensure that the true
level of risk is recognised, and that the planning and procedures for the work are of an
appropriate level to mitigate that risk.

A key issue is the level of specification for the machine. The machine is a major piece of
capital expenditure at the start of the work. In a competitive bid, keeping the price of the
machine to a minimum may be a significant factor in minimising the bid price. However,
minimising the cost of the machine may involve substantially increasing the risk of problems
developing during the tunnelling. As discussed by Maidl and Nellessen (14), owners and
designers have to influence the selection of the type of machine (i.e. slurry or EPB) to avoid
the use of a less expensive, but less suitable, machine. Where the choice of machine could
affect the risk to third parties then the type of machine may have to be specified by the owner
or designer. In addition, specifications can be developed that require that key aspects of the
machine meet a higher standard than the minimum. This will provide some margin against
unforeseen conditions. Some aspects of how the machine selection and design can be
influenced by the specifications are discussed by Shirlaw et al. (17).

4 CONCLUSIONS

It has been demonstrated by review of a number of EPB tunnelling projects that there have
been a significant number of instances of large, but generally localised, settlements on many
of the projects. These settlements can be separated into:

• those which are larger than would normally be anticipated by the designer (here taken as
equivalent to 3% volume loss), but are not visible to the eye
• settlements (typically over 150mm) and sinkholes which are visible to the naked eye

Such settlements have the potential to cause major damage to structures or utilities over the
tunnel and can be a risk to safety. The potential for this magnitude of settlement to occur is
therefore of concern to owners and designers, as well as contractors. Although the execution
of the work must be the responsibility of the contractor, the owner and/or designer can help to
ensure that the contractor carries out adequate planning and preparation for the tunnelling.
The owner/designer can influence the contractor in these aspects through the preparation of
specifications and in the review of the contractor’s proposals. Some issues that can be
considered in terms of enhancing specifications and review have been outlined. In particular,
the owner and designer may need to influence the selection of, or specify, the type of
machine to be used, and ensure that the specification of the machine is of a high standard.

Perhaps most importantly, the risk of large settlements or sinkholes should not be dismissed
as “negligible” just on the basis of the machine type selected for a project. This paper
demonstrates that such risks are real. It would be beneficial for future projects to quantify the
probability of such events and proactively develop mitigation measures.
 5 REFERENCES

1 SHIRLAW, J.N., ONG, J.C.W. ROSSER, H.B., TAN, C.G, OSBORNE, N.H. and
HESLOP P.J.E. “Local settlements and sinkholes due to EPB tunnelling” Proc. ICE,
Geotechnical Engineering I 56, Issue GE4, October 2003 pp 193 – 211
2 SHIRLAW, J.N. “Observed and calculated pore pressures and deformations induced by an
Earth Pressure balance shield: Discussion”, Canadian Geotechnical Journal, 32, pp181-189
3 RAJU, G.V.R., NG, C.C. and KUMAR, B.S. “Damage assessment of adjacent structures
due to bored tunnelling for Changi Airport Line”. Proc. Underground Singapore 2001, pp
65-75
4 CLOUGH, G.W. and LECA, E. “EPB Shield Tunnelling in Mixed Face Conditions”
Journal of Geot. Engineering, ASCE. Vol 119, 10. 1993, pp1640 – 1656
5 MOH, Z.C., CHANG, J.L. and HWANG, R.N. “Soft ground tunnelling in Taipei –
problems and solutions” Proc. Dr Tan Swan Beng Memorial Symposium, Singapore. 1999.
pp87-97
6 BOONE, S.J., MCGAGHRAN, S., BOUWER,G. and LEINALA, T. “Monitoring the
performance of earth pressure balance tunneling in Toronto” Proc. North American
Tunnelling, 2002, pp293-302.
7 BIGGART, A.R. and STERNATH, R. “Storebaelt eastern railway tunnel: construction”
Proc Instn Civ. Engrs Civ. Engng. 1996. pp20-39
8 KIMURA, K. 2001. Dangerous gap between theory and practice in earth pressure balance
shield tunnelling in Japan. Proceedings Modern Tunnelling Science and Technology, Swets
and Zeitlinger. 2001.
9 LEBLAIS, Y., LECA, E. and MAUROY, F. “Deplacements verticaux lies au creusement
au bouclier a pression de terre (EPBM); cas du Metro de Lille – Ligne 2: lots 1 et 2 ” Proc.
International AFTES, Underground Works: Techniques and Men, Chambery, 21 –24 October
1996, pp 483-487
10 ANON. “Oporto Metro breakthrough” Tunnels and Tunnelling International, November
2002, p7.
11 WALLIS, S. “Shenzen Metro” World Tunnelling, April 2002, pp 130-134.
12 ANON. “Spoil analysis stepped up at CTRL ground failure site” New Civil Engineer, 27
March 2003 p10
13 DORAN, S.R. and ATHENOUX, “Storebelt Eastern railway Tunnel – Denmark
Geotechnical Aspects of TBM Selection and Operation”
14 MAIDL, B. and NELLESSEN, P. “Comment on the article – Machine selection
influences” Tunnels and Tunnelling International, December 2004, pp28-29.
15 SHIRLAW, J.N., HENCHER, S.R and ZHAO, J. “Design and construction issues for
excavation and tunnelling in some tropically weathered rocks and soils” Proceedings
GeoEng2000 Technomic Publishing, PA, USA. Volume 1. November 2000, pp1286-1329
16 OSBORNE, N., AW, E.M., and LIM, W.B. “a comparison and evaluation of available
techniques for monitoring bored tunnelling excavated spoil volumes” Proc 30th ITA-AITES
World Tunnel Congress, Elsevier, May 2004, pp875-882
17 SHIRLAW, J.N., BOONE, S., SUGDEN, N.B. and PEACH, A. “Controlling the risk of
sinkholes over EPB driven tunnels – a client perspective” Proc. Geotechnical aspects of
Underground Construction, Amsterdam, 2005, in press

View publication stats