Vancouver 19 TH NC

TUNNELLING ASSOCIATION OF CANADA
19
th
National Conference
Conference Proceedings
Tunnelling Towards 2010
Vancouver, B.C. Canada
September 1720, 2006
Marriott Pinnacle Hotel
i
19
th
Canadian Tunnelling Conference Vancouver, BC, September 2006

Table of Contents

W. Chen, M. Crow, D. Young & S. Norris, USA................................................................1
Performance of an EPB-TBM beneath the Sacramento River and Levees on the Lower
Northwest Interceptor Project, Sacramento, CA, USA

R.D. Drake, D. Dugan & A.J . Pooley, USA......................................................................11
Brightwater Conveyance Project Soft Ground Tunnels and Shafts

S. Skelhorn, USA...............................................................................................................19
The Big Walnut: High Tech Tunnelling in Columbus

A. Dean & D.J . Young, USA.............................................................................................26
A Framework for Preliminary Design of Tunnel Eyes

A. Ameli, Canada...............................................................................................................37
Estimation of Water Inflow to a Tunnel during Construction

P.J . Oblozinsky, J . Kuwano & Q. Li, Canada and J apan...................................................41
Static Centrifuge Experiments on Stability of Pressure-supported Tunnel Faces in Sandy
Ground

D.J . Young, & A. Dean, USA............................................................................................46
Seismic Performance of Precast Concrete Tunnel Linings

H.T. AL-Battaineh, S. AbouRizk, S. Fernando, F. Policicchio & J . Tan, Canada...........54
Tunnelling for Success, Case Study: Glencoe Tunnel in Calgary

M. Murray, S. Redmond, R. Sage, F. Langer & D. Phelps, USA......................................59
SEM in Seattle Design and Construction of the C710 Beacon Hill Station Tunnels

D. Brox, J . Rotzien, C. Genschel, J . Messner, A. Gupta, T. Morrison & A. Saltis,
Canada................................................................................................................................70
Construction Update and TBM Excavation Planning, Seymour Capilano Twin Tunnels
Project, Vancouver

R. Humphries, M. Funkhouser & E. OConnor, Canada and USA...................................76
Design-Build Tunnels and Shafts at the San Roque Project

G. Dubois, Canada.............................................................................................................83
The Toulnustouc River Intake Tunnel

ii
A.J . Rancourt, C. Chartrand, A. Whalen & D. Bergeron, Canada.....................................87
Toulnustouc pressure tunnel leakage estimation, filling, instrumentation and control

U. Korsman, P. Nieminen & P. Salminen, Finland...........................................................95
New intelligent drilling jumbos for accurate, fast and cost-efficient tunnelling

G.A. Schmalz & S.L. De Dominicis, Switzerland...........................................................101
Gotthard Base Tunnel the worlds longest base tunnel

E. Eberhardt, C. Zangerl & S. Loew, Canada, Austria and Switzerland.........................111
The influence of geological structures in promoting asymmetrical surface subsidence
over deep tunnels in hard rock

R. Delmar, H. Charalambu, E. Gschnizter, R. Everdell, Canada....................................120
The Niagara Tunnel Project An Overview

D. Brox, S. Bean, P. Branco & T. Coulter, Canada.........................................................131
Planning for Canadas First Bored Road Tunnel in over 40 Years Kicking Horse Canyon
Project

R. Collins, P. Gonzalez & H. MacIsaac, Canada.............................................................140
Raising the Bar in Shaft Sinking at Falconbridge's Nickel RimSouth Project

P.C. Raleigh, USA ...........................................................................................................152
Tunneling the Metro do Porto - Under Pressure in Porto Granite

C. Maia & L. Babendererde, USA and Germany............................................................160
Innovative Station Design and the new Additional Face Support Systemmake Metro do
Porto a unique light rail project

B. Henry, Canada.............................................................................................................166
Design and Construction of the Canada Line Bored Tunnel Section

B. Lukajic, M. Schafer, R. Pintabona, M. Kritzer, S. J anosko & R. Switalski, USA......175
Mill Creek Tunnel Geomechanics

M. Sunnananda, P. Chanpradappha, K. Akewanlop & T. Srisirirojanakorn, Thailand...179
Responses of Pore Water Pressures during EPB Shield Tunneling in Bangkok Subsoil

I. Corbett, Canada............................................................................................................182
The Changing Face of Tunnelling in Greater Toronto

A. Coombs & D. Zoldy, Canada......................................................................................194
York DurhamSewage System19
th
Avenue/Leslie Street Interceptor Sewage Project,
Richmond Hill, Ontario, Canada

iii
Dave Long........................................................................................................................206
The Robbins Company, Kent, WA, USA

E. Hoek, Canada..............................................................................................................216
European Motorway Tunnels Keynote Address

1 INTRODUCTION
The Sacramento Regional County Sanitation
District (SRCSD) is in the process of expanding its
interceptor system from about 130 km (80 miles) to an
estimated 320 km (200 miles) of gravity and force
main pipelines to serve the growth projected for its
service region.
The River Crossings Project is a part of the 30.3
km (18.9 mile) Lower Northwest Interceptor Sewer
(LNWI), presented in Figure 1, managed by
Montgomery Watson Harza for the SRCSD. The
LNWI brings wastewater from the newly developing
areas to the north of Sacramento and the City of West
Sacramento to the regional wastewater treatment plant
south of Sacramento, near Elk Grove. The preliminary
design of the tunnels performed by URS has been
described elsewhere [1].
2 PROJ ECT DESCRIPTION
The pipeline was constructed in a soft-ground
tunnel under the Sacramento River and levees in two
locations, with precautions taken to avoid any ill
effects on the river and its flood protection levees. The
location of the Northern Sacramento River Crossing
(NSRC) and the Southern Sacramento River Crossing
(SSRC) are shown in Figure 1. The NSRC is located
about 30 m to 120 m (100 ft. to 400 ft.) east

Fig. 1. Plan of Lower Northwest Interceptor Program.
and downstream of the Interstate 80 over-crossing of
the Sacramento River. The SSRC is located upstream
Performance of an EPB-TBM beneath the Sacramento River and
Levees on the Lower Northwest Interceptor Project, Sacramento,
CA, USA
Wally Chen
Parsons Brinckerhoff Construction Services, Sacramento, CA, USA
Matthew Crow
Kellogg Brown and Root, Pasadena, CA, USA
David Young
Hatch Mott MacDonald, Pleasanton, CA, USA
Steve Norris
Sacramento Regional County Sanitation District, Sacramento, CA, USA

ABSTRACT: Two 610 m (2,000 ft.) long tunnels were driven through water bearing silts, sands and clays beneath the
Sacramento River using a 4.59 m (15.1 ft.) diameter Earth Pressure Balance Tunnel Boring Machine (EPBM). The pre-
cast concrete segmentally lined and grouted tunnel was driven at 6% grades beneath the river flood protection levees,
buildings, roads, a railroad and adjacent to a Freeway Viaduct. Continuous automatic data recording of TBM operation and
field inspectors records, accompanied by monitoring surface displacement points, inclinometers, piezometers and tunnel
convergence were used to obtain information. The paper explains the reasons for specific tunneling specifications
controlling tunnel construction in order to meet the project constraints including local environmental and regulatory
requirements. The paper will also present a summary of data to demonstrate the tunneling performance in the different
ground conditions encountered.
1
of the settlement of Freeport and about 1,200 m
(4,000 ft.) upstream of the Freeport Bridge. Aspects
of the detailed design performed by Hatch Mott
MacDonald are presented elsewhere [2].

2.1. Northern Sacramento River Crossing
This crossing, Figure 2, is about 752 m (2,466 ft.)
long overall, and consists of twin 1.524 m (60 in.)
internal diameter pipelines.

Fig. 2. Aerial photograph of NSRC.
The twin force main pipelines were installed in a
tunnel with a length of 602 m (1,975 ft.) that is lined
with pre-cast concrete segments. The remaining 150
m (492 ft.) of pipeline was constructed in temporary
tunnel construction shafts and conventional open cut.
The space between the pre-cast concrete tunnel lining
and the force main pipes is backfilled with low-
density cellular concrete (LDCC). A section view of
the pipes installed in the tunnel is shown in Figure 3.
The tunnel alignment, presented in Figure 2,
begins within a launching shaft on the north side of
the river. The vertical profile, Figure 4, shows the
tunnel invert level at the shaft about 8.5 m (28 ft.)
below grade. The tunnel continues on a tangent
downward at a 6% grade.

Fig. 3. Schematic of pipes installed in tunnel.

The alignment passes through a series of vertical
and horizontal curves beneath the low point in the
river before ascending another 6% grade to the
receiving shaft. At its deepest point, the excavated
tunnel reaches about elevation -22 m (72 ft.) and
provides about 13 m (42 ft.) of cover from tunnel
crown to river mud line.
The land on the north side of the NSRC is used
for a restaurant, single-family homes, and un-
developed city-owned and developer-owned areas.
The launching shaft is located on city-owned land in
an open field. The levees are constructed of
compacted soil with typically 2:1 (horizontal to
vertical) slopes and have crest elevations of about
11.5 m (+38 ft.) and 12 m (+39 ft.) for the north and
south sides of the river, respectively.

Fig. 4. Subsurface profile of NSRC (Launching Shaft on
right).
The levee on the north side of the river has a
cutoff wall consisting of soil, cement, and bentonite
constructed through the levee and extending 1.5 m to
2
3 m (5 to 10 ft.) below the natural ground level, which
is around elevation +6.7 m (+22 ft.). The Garden
Highway is located atop the levee north of the river
about 91 m (300 ft.) south of the launching shaft.
The south end of the NSRC is located adjacent to
the sludge drying beds for the City of West
Sacramentos Bryte Bend Water Treatment Plant.
Between the treatment plan and the river are a city
street, unused open fields and the levee.

2.2. Southern Sacramento River Crossing

This crossing, Figure 5, is about 743 m (2,437 ft.)
long overall, and consists of twin 1.676 m (66 in)
internal diameter pipelines.

Fig. 5. Aerial Photograph of SSRC.
The construction details are very similar to those
of the NSRC with a tunnel length of 635 m (2,082 ft.)
and the remaining 108 m (355 ft.) of pipeline
constructed in temporary tunnel construction shafts
and conventional open cut segments.
The tunnel construction began within a launching
shaft on the east side of the river, identified on
Figure 5. The vertical profile, Figure 6, presents the
shaft with the tunnel invert at the shaft about 9 m (29
ft.) below grade.
The tunnel continues on a tangent downward at a
6% grade. At its deepest point, the excavated tunnel
reaches about elevation -23 m (-76 ft.), 10 m (33 ft.)
of cover from tunnel crown to river mud line.
The land on both sides of the river at the SSRC is
currently being used for agricultural purposes. On the
east side, the former Southern Pacific Railroad
(SPRR) tracks are located along the levee crest and
State Highway 160 (Freeport Boulevard), a two-lane
road, runs along the landside levee toe. On the west
side, South River Road is located along the levee
crest.

Fig. 6. Subsurface profile of SSRC (Launching Shaft on
right).
The levees are constructed of earth and have crest
elevations of about +9.5 m (+31 ft.). The levee on the
east side of the river has a cutoff wall consisting of
soil, cement, and bentonite constructed through the
levee and extending 1.5 3 m (5 to 10 ft.) below the
natural ground level.
3 GROUND CONDITIONS
3.1. Description of Soil Groups
The soils encountered on both river crossings
were similar and generally consisted of silts, sands,
and clays with infrequent layers containing gravels.
Generalized geologic profiles developed by the
designers geotechnical consultant, Kleinfelder for the
NSRC and SSRC are shown on Figures 4 and 6
respectively. Descriptions of the soils encountered on
the two crossings are provided in Tables 1 and 2.
The sizes of gravel particles encountered in the
borings were generally less than 25 mm (1 in.) in
maximum dimension. However, based on reports of
larger gravels, cobbles and boulders in other borings
drilled in the area, and in other near-by projects, soil
layers containing coarse gravels, cobbles and boulders
were anticipated. The maximum clast size observed in
the tunnel spoil during construction was just over 100
mm (4 in.).
3.2. Potential for Encountering Man Made Features
of Significance
Timber piles were observed on a private parcel
directly over the NSRC tunnel alignment.
Traditionally these piles were 21 m (70 ft.) long and
installed by jetting, so it was considered that timber
3
piles could be encountered during tunneling; however,
this was not the case.

Table 1. NSRC Soil Groups
Soil Description Geologic
Unit*
Comments
Loose silty sand
and sand
Qa Sacramento River
alluvium, generally
uncemented, high
permeability.
Soft to stiff silt
and clay with
localized sand-
filled channels
Qb1 Basin deposit,
generally uncemented,
moderately to highly
plastic with low
permeability
Stiff to very stiff
clay and silt
Qb2 Basin deposit,
generally uncemented,
moderately to highly
plastic with low
permeability
Loose to
medium dense
silty sand and
sand with
localized
channels infilled
with soft to stiff
silt and clay
Qsc Active river channel
deposit, highly
variable, but generally
uncemented, coarse
grained with high
permeability
Dense to very
dense silty sand,
sand, and
localized gravel
filled channels
Qm Modesto Formation,
alluvium is relatively
dense and weakly
cemented to
uncemented, with high
permeability.
Very stiff to
hard silt and
clay and very
dense silty sand
and sand
Qrl Lower Riverbank
Formation, weakly
cemented to
uncemented, generally
highly variable in
composition and
permeability.
* Geologic unit names are based on a stratigraphic model
interpreted fromearlier mapping by Helley and Harwood,
1985 [3] and Schlemon, 1995 [4].
3.3. Clay Stickiness
High plasticity clays were observed in the borings
and were expected to cause clogging and consequent
problems and delays with moving mechanical parts.
Sticky clays were encountered, particularly in the
vicinity of Station 906+00 on the NSRC, where a
1.2 m (4 ft) thick fat clay material was encountered in
the heading.
3.4. Groundwater Levels
Groundwater levels fluctuate seasonally in
response to river levels, which can rise above the
natural bank level during the winter rainy season and
snow melt from the mountains. The maximum water
level in the river recorded during the construction of
the NSRC was at +1.5 m (5 ft.), however higher river
levels occurred after tunneling. A photograph of the
NSRC during this event is included as Figure 7.
Table 2. SSRC Soil Groups

Soil Description Geologic
Unit*
Comments
Soft to stiff silt
and clay
Qb1 Basin deposit,
generally
uncemented, with
low permeability
Soft to stiff silt
and clay with
localized
channels of
dense sand
Qb2 Basin deposit,
generally
uncemented, with
low permeability
Medium dense to
dense sand and
silty-sand
Qa Sacramento River
alluvium,
uncemented, with
high permeability.
Interbedded very
stiff silt and clay
and very dense
silty-sand and
sand
Qrl Lower Riverbank
Formation, weakly
cemented to
uncemented, and are
highly variable in
composition and
permeability.
Levee fill
materials
N/A Artificial fill, highly
variable
* Geologic unit names are based on a stratigraphic model
interpreted fromearlier mapping by Helley and Harwood,
1985 [3] and Schlemon, 1995 [4].

Fig. 7. High river levels at the NSRC in J anuary 2006.
3.5. Hazardous Gases
A photo ionization detector was used to assess the
presence of flammable gas such as methane
emanating from soil samples and accumulating in
4
monitoring well head space. The pre-construction
investigation revealed the possibility that methane can
be encountered during tunneling and Cal/OSHA
classified the tunnels as potentially gassy. During
tunneling of the SSRC methane was detected from the
spoil being discharged from the screw conveyor,
however at a level less than 10% of the lower
explosive limit (LEL).
4 TUNNELING METHODOLOGY
4.1. Specification
The specification required the use of a closed face
TBM and allowed either a Slurry TBM or an Earth
Pressure Balance (EPBM). Minimum requirements
were specified for each, with the contractor being
responsible for the detailed design and performance of
the TBM. For an EPBM the requirements included
feeding the screw conveyor from the invert of the
plenum, and providing a positive displacement device
on the screw conveyor and a compressed air lock for
face interventions.
Due to schedule constraints imposed by the State
Reclamation Board for flood protection, and the
required commissioning of the project by the end of
2006, the SRCSD initially required two TBMs so that
the two river crossings could be excavated
concurrently.
4.2. Implementation
All Bidders for the construction contract, who had
been prequalified previously, proposed EPBM rather
than Slurry TBMs. The contractor, Affholder Inc of
Chesterfield, Missouri, used a Lovat RMP 181 SE
TBM, Figure 8, refurbished for this project by Lovat
Inc.

Fig. 8. TBM Mary 1 in Lovat Workshops, Toronto,
Ontario, Canada.
The principal features of the machine are
presented in Table 3 and Figure 9.
Table 3. Features of Lovat 181 SE

Overall Geometry
Cut diameter: 4.622 m
TBM length: 8.3 m
TBM weight:145 tonnes
Minimum radius: 100 m
Open Face Area: 30%

Power
Total power available:
800kW.
Power to cutterhead: 200 kW
Variable Frequency Drive
Max cutterhead torque:
4.2 MNm (0.97 to 4.65 rpm)
Max. propulsion thrust:
1540 tonnes at 340 bar

Annular Grout
Grouting through
segments by two pairs of
7.5 HP/1 HP Peristaltic
pumps for two part
accelerated cement fly ash
grout supplied by fixed
lines.
Cutterhead Tools
Nose cone
30 Ripper teeth
interchangeable with disc
cutters.
64 Scraper teeth
Cutterhead carbide plated and
tools with carbide inserts.
Screw Conveyor
Length: 9.5 m (Primary)
Length: 13 m (Secondary)
Diameter: 610 mm
Spoil Removal
Secondary belt conveyor
discharging onto fixed
conveyor discharging to
surface. Segments
brought in by diesel
powered train.
Plenum Monitoring
6 Ea bulkhead mounted total
pressure cells
Soil Conditioning
Injection ports:
Cutterhead face - 5
Screw conveyor - 4
Segments
Segment ID: 4.013 m
Thickness: 223 mm
5 piece
parallelogram/trapezoidal ring
with trapezoidal key
Length: 1.220 m

Fig. 9. Lovat EPB-TBM (4.59 m OD) used on the LNWI
Sacramento River Crossings.
The machine generally conformed to all aspects of
the specification, including locating the screw
conveyor at the base of the plenum. However, as the
machine was a used machine the thrust and open face
area provided were less than specified.
5
In particular a two stage screw conveyor with
terminal and intermediate guillotines was provided in
place of a formal positive displacement device
operating in the following way (Figure 9).
Both conveyors are filled with material, but the
terminal guillotine is kept closed. Once full, the
intermediate guillotine is closed to isolate the second
stage conveyor from the plenum, prior to discharge.
When the terminal guillotine is closed after discharge
of excavated soil, the intermediate guillotine is opened
and the process is repeated.
After the contract award, the contractor
demonstrated that the schedule risk could be
adequately managed with one TBM in a value
engineering saving proposal that was accepted by the
owner. The TBM for the SSRC was utilized for the
NSRC, with all tunneling completed during 2005.
5 MONITORING AND CONTROL
5.1. General
The responsibility for providing information to
the Engineer on monitoring of tunnel operations was
shared between the Construction Manager, Parsons
Brinckerhoff Construction Services (PBCS) in
association with Kellogg Brown & Root (KBR) and
the Contractor. The Contractor made available data
from the TBM data logger and daily reports required
by the Contract. The Construction Manager provided
confirmatory readings of information available from
automatic data loggers and prepared computer
generated shift and ring reports. The key TBM
performance parameters were processed to produce
graphical information on surface settlement,
excavated soil weight, grout volume, face pressure,
screw speed, TBM advance rate, thrust and cutterhead
torque. This methodology was based upon the system
used for the ECIS Sewer Tunnel in Los Angeles [5].
5.2. Surface Monitoring
Tunnel centerline ground surface displacement
monitoring points (DMPs) were installed at 15 m (50
ft.) intervals. The DMPs, installed by the contractors
geotechnical consultant, Brierley Associates, as
presented on Figure 10, were generally in soil within
the temporary construction easement. However,
concrete nails and timber stakes were used on
structures and within landscaped areas respectively.
Additionally three point arrays of DMPs were
installed at 30 m (100 ft.) intervals with 5 point arrays
at 15 m (50 ft.) intervals to monitor the settlement of
critical structures and levees. Monitoring of the DMPs
was undertaken manually using a rod and Leica
NA724 level and data were processed using
spreadsheets. Inclinometers were installed to
demonstrate that there was no adverse impact on the
levees or the Interstate 80 freeway structure.

Fig. 10. Detail of displacement monitoring point in soil.
6 TUNNELING PERFORMANCE
6.1. General
Mining of the 635 m (2082 ft.) long SSRC started
on J une 3, 2005 reaching the receiving shaft on
August 23, 2006, a duration of 12 weeks. A graphical
representation of tunnel progress is included as Figure
11.
J un-05
J ul-05
Aug-05
Sep-05
1
7
9
+
0
0
1
8
0
+
0
0
1
8
1
+
0
0
1
8
2
+
0
0
1
8
3
+
0
0
1
8
4
+
0
0
1
8
5
+
0
0
1
8
6
+
0
0
1
8
7
+
0
0
1
8
8
+
0
0
1
8
9
+
0
0
1
9
0
+
0
0
1
9
1
+
0
0
1
9
2
+
0
0
1
9
3
+
0
0
1
9
4
+
0
0
1
9
5
+
0
0
1
9
6
+
0
0
1
9
7
+
0
0
1
9
8
+
0
0
1
9
9
+
0
0
2
0
0
+
0
0
Station (ft)

Fig. 11. Production rate on SSRC.
Installation of the fixed tunnel crown conveyor in
the launching shaft and demonstration of the
compressed air apparatus for potential face
interventions, prior to tunneling beneath the river
caused a two week break in production at Station
181+50. An overall average production rate of 53 m
(174 ft.) a week was achieved, although the equivalent
rate excluding the mining prior to installation of the
fixed conveyor would be 84 m (275 ft.) a week.
Mining of the 602 m (1975 ft.) long NSRC started
on October 17, 2005, with hole through at the West
Sacramento receiving shaft, Figure 12, on December
8, 2006, a duration of 7 weeks.

6

Fig. 12. Hole through at NSRC receiving shaft.
A graphical representation of tunnel progress is
included as Figure 13.
Oct-05
Nov-05
Dec-05
9
0
1
+
0
0
9
0
2
+
0
0
9
0
3
+
0
0
9
0
4
+
0
0
9
0
5
+
0
0
9
0
6
+
0
0
9
0
7
+
0
0
9
0
8
+
0
0
9
0
9
+
0
0
9
1
0
+
0
0
9
1
1
+
0
0
9
1
2
+
0
0
9
1
3
+
0
0
9
1
4
+
0
0
9
1
5
+
0
0
9
1
6
+
0
0
9
1
7
+
0
0
9
1
8
+
0
0
9
1
9
+
0
0
9
2
0
+
0
0
9
2
1
+
0
0
9
2
2
+
0
0
Station (ft)

Fig. 13. Production rate on NSRC.
Installation of the fixed tunnel crown conveyor in
the launching shaft resulted in a week long break in
production at Station 918+75. An overall average
production rate of 91.5 m (300 ft.) a week was
achieved, although the equivalent rate after
installation of the fixed conveyor was 152 m (500 ft.)
a week.
A maximum weekly production of 152 m (500 ft.)
a week was achieved, with 24 m (80 ft.) being
achieved in a single 10 hour shift. Higher overall
production rates were not realized due to the short
overall length of the tunnel and external logistic
problems. Consideration of the performance of the
tunneling in the main geological units, clay and sand
follows, using examples from the SSRC. Similar
observations were made about tunneling on the
NSRC.
6.2. Clay
The measured time to excavate for a 1.219 m
(4 ft.) long ring in clay is presented in Figure 14.
Fig. 14. Advance rate on SSRC when tunneling through
clay.
The average excavation time was 45 minutes,
with the longer times generally being associated with
managing water ingress through the screw conveyor.
The surface settlement above the tunnel is
presented in Figure 15. No data is presented between
Stations 187+50 and 192+50, as the alignment passes
beneath the river. The maximum settlement was 6 mm
(0.02 ft.).

Fig. 15. Surface settlement on SSRC when tunneling
through clay.
The actual groundwater table was measured by
reading borehole piezometers and was affected by
dewatering of the launching shaft and fluctuations in
the river level. The actual hydrostatic head on the
tunnel is presented in Figure 16 with the maximum
head being associated with the low point of the tunnel
at Station 189+00.
The average value of the six total pressure cells on
the plenum bulkhead is presented on Figure 16 and
provides an indication of actual face support pressure.
In many cases this was less than the hydrostatic head;
7
however this did not appear to affect the stability of
the ground.

Fig. 16. Face pressure on SSRC when tunneling through
clay.
6.3. Sand
The measured time to excavate for a 1.219 m (4
ft.) long ring in sand is presented in Figure 17.

Fig. 17. Advance rate on SSRC when tunneling through
sand.
The average excavation time was 23 minutes.
Significantly more consistent excavation times were
experienced in the sand.

18. Settlement on SSRC when tunneling through sand.
The surface settlement above the tunnel is
presented in Figure 18. It can be seen that settlement
varied from 9 mm (0.03 ft.) to a maximum of 107 mm
(0.35 ft.). Heave was also observed. Observations and
possible mechanisms for these ground movements are
discussed below.
The estimated support pressure required, based
upon the actual hydrostatic pressure and calculated
ground support pressures on the tunnel is presented in
Figure 19 with the reduction in head resulting from
the rising 6% tunnel grade. The average value of the
six total pressure cells on the plenum bulkhead is
presented on Figure 19 and provides an indication of
actual face support pressure. In many cases this was
equal to the hydrostatic head, with no additional
pressure to support the ground.

Fig. 19. Face pressure on SSRC when tunneling through
sand.
The history and discussion of settlement of
specific instruments against distance from the moving
tunnel face and associated 8.3 m (27 ft.) long shield is
presented in Figures 20 through 25.
Figure 20 presenting the settlement of DMP-29 at
Station 195+00 represents the largest surface
settlement of 100 mm (0.330 ft.).

Fig. 20. Settlement DMP-29 at Station 195+00; cover 45-ft.
8
It can be seen that 88 mm (0.290 ft.) of settlement
occurred either at the face or over the shield, followed
by 12 mm (0.040 ft.) behind the shield before the
ground was stabilized by grouting in the annulus
between the ground and the tunnel lining. This level
of settlement was acceptable, as it lay beneath a field
and long term monitoring indicated stability. The
settlement corresponded with some over excavation
coincident with relatively low face pressure (Figure
19) at the interface between the clay and the sand.
Station 195+50 represents a total surface settlement of
43 mm (0.140 ft.).

27 mm (0.090 ft.) of settlement occurred at the
face, followed by 15 mm (0.050 ft.) over the shield
with the ground fully stabilized by grouting. It appears
that half of the settlement is associated with face
control and half with settlement of the ground over the
shield resulting from the overcut; however no further
settlement occurred following the grouting operation
associated with installing of the segmental tunnel
lining.
Figure 22 presenting the settlement of DMP-T1 at
9 mm (0.030 ft.).

Fig. 22. Settlement DMP-T1 at Station 195+75; cover 38-ft.
It can be seen that no significant settlement
occurred at the face or over the shield, with only
minor settlement occurring after grouting of the
ground behind the shield. Although the face pressure
was relatively high (Figure 19), no heave was
observed, with an overburden depth greater than 2.5
diameters. It appears that the stand up time of the
ground was greater than the time taken to fill the
annular space between the ground and the segments
with grout.

73 mm (0.240 ft.).

occurred at the face, a further 18 mm (0.060) over the
shield, followed by a further 47 mm (0.155 ft.) after
grouting. Although face control appears good,
settlement appears to have occurred during the
annular grouting operation. It appears that the stand
up time of the ground was less than the time taken to
fill the annular space between the ground and the
segments with grout.
40 mm (0.130 ft.).
occurred at the face, 12 mm (0.040) over the shield,
followed by 18 mm (0.060 ft.) after grouting. It
appears that although face control was adequate,
settlement has resulted from the overcut and
continued through the grouting operation. It appears
that the stand up time of the ground was less than the
time taken to advance the TBM past the instrument.

9

Station 199+25 represents no net surface settlement,
however significant heave was observed.
It can be seen that 3 mm (0.010 ft.) of heave
occurred at the face, increasing to a maximum of
20 mm (0.065) over the shield, however subsequent
settlement counteracted this heave. It appears that a
relatively high face pressure in an area of just over
one diameter of cover caused the heave.
7 CONCLUDING REMARKS
The data gathered from this project displayed
classic ground response in varying soil types to
varying mining parameters in EPB tunneling. The
management of settlement and ground heave centers
around the balance of adequate face pressure and
controlled mining with adequate annulus grouting in a
timely operation. Varying soil types, depths of cover
and hydrostatic pressures are the external elements
that alter the fulcrum of such balance. Tunnels can be
driven successfully in soils beneath the Sacramento
River and its flood protection levees. Construction
risks can be managed by an owner by the engagement
of designers, construction managers and prequalified
contractors experienced in underground construction
working as a team.
8 ACKNOWLEDGEMENTS
The Authors wish to acknowledge the team led by
Neal Allen of the Sacramento Regional County
Sanitation District, and the organization led by Dan
Martz of Affholder Inc, who constructed the project,
without whom there would be nothing to report on.
The paper is based upon the data collected and
numerous presentation charts prepared by Cody
Painter of PBCS and Steve Cano of KBR.
REFERENCES
1. Nagle, G., and J . Nonnweiler. 2003. Preliminary
Design of Tunnels for the Lower Northwest
Interceptor, Sacramento, California, USA. In
Proceedings of the Rapid Excavation and Tunneling
Conference, New Orleans, 2003, R.A. Robinson and
J .M. Marquardt, 1294-1300. Society for Mining,
Metallurgy, and Exploration, Inc. Littleton, CO.
2. Nonnweiler, J ., J . Hawley, and D. Young. 2003.
Design of Lower Northwest Interceptor Tunnels
beneath the Sacramento River. In Proceedings of the
Rapid Excavation and Tunneling Conference, New
Orleans, 2003, R.A. Robinson and J .M. Marquardt,
179-190. Society for Mining, Metallurgy, and
Exploration, Inc. Littleton, CO.
3. Helley, E.J ., and D.S. Harwood. 1985. Geologic Map
of the Late Cenozoic Deposits of the Sacramento
Valley and Northern Sierran Foothills, California. U.S.
Geological Survey Miscellaneous Field Studies Map
MF-1790.
4. Shlemon, R.J ., 1995. Pleistocene Channels of the
Lower American River, Sacramento County,
California: (appended, five-page article) in Franks, A.,
and Moss G. (leaders), Geology of the Sacramento
Area, Foothills, and the Sierra Nevada Mountains:
Association of Engineering Geologists Field Trip
Guide, 1995 Annual Meeting of the Association of
Engineering Geologists and Groundwater Resources
Association, Sacramento, California.
5. Crow, M.R., and J . Holzhauser. 2003. Performance of
Four EPB-TBMs Above and Below the Groundwater
Table on the ECIS Project, Los Angeles, CA, USA. In
Conference, New Orleans, 2003, R.A. Robinson and
J .M. Marquardt, 905-926. Society for Mining,
Metallurgy, and Exploration, Inc. Littleton, CO.

10
1 INTRODUCTION
In response to ongoing and projected population
growth in the greater Seattle region of Washington
State, USA, King Countys Wastewater Treatment
Division is constructing a new regional wastewater
treatment plant with associated pipelines and other
facilities as major improvements to the Countys
regional sewage treatment system. The new $1.5
billion system has been named Brightwater.
King Countys Wastewater Treatment Division
provides sewage collection and treatment service to
1.4 million people in the region, including 18 cities
and 16 sewer districts covering 1087 km
2
(420 square
miles). The Wastewater Treatment Division operates
two regional treatment plants, numerous pump
stations and approximately 571 km (355 miles) of
pipelines. Brightwater will add a third treatment plant
to the system, reducing the load on the two existing
plants and providing increased capacity to deal with
projected future needs.
The subject of this paper is the conveyance
portion of the Brightwater project, which consists of
approximately 23 km (14 miles) of bored tunnels for
the purpose of conveying collected influent into the
treatment plant and treated effluent from the plant to a
marine outfall which will discharge into Puget Sound.
1.1. Project History and Overview
In 1999 a long range planning study to address the
issues of the rapidly growing greater Seattle region
identified the need for a third regional wastewater
treatment plant, to be operational by 2010. The siting
and environmental review process commenced in
2000. Over 100 potential sites for the treatment plant
and marine outfall and a variety of conveyance
alignments were initially considered; the number of
alternatives was progressively reduced at each stage
of the process. Two treatment plant options were
considered in the final environmental review.
Following publication of the final review the current
Project siting was approved in 2004. A more detailed
description of the site selection and procurement
processes is given by Locke and Edgerton [1].
Detailed studies and preliminary design for the
project were carried out between 2002 and 2004, and
detailed final design began in 2004. The design of
Brightwaters conveyance system is described by
Adams et al [2]. Separate construction management
contracts were awarded for the conveyance and
treatment plant in 2005. The first conveyance
construction contract was awarded in December 2005
and construction commenced in January 2006. The
remaining conveyance construction contracts are
expected to be awarded during 2006 and early 2007.
The system is scheduled to be operational by late
2010.
The Brightwater treatment plant will have an
initial average wet weather capacity in 2010 of 136
million l/day (36 million gallons/day) expanding to
204 million l/day (54 million gallons/day) by 2040. In
addition to the treatment plant, the Brightwater project
Brightwater Conveyance Project Soft Ground Tunnels and Shafts
Ronald D. Drake, P.E.
Jacobs Civil Inc., Seattle, USA
Derek Dugan
CH2M Hill Inc., Seattle, USA
Anthony J. Pooley, C.Eng., MICE
Jacobs Civil Inc., Seattle, USA
ABSTRACT: This paper describes the current status of the Brightwater Conveyance Project in King County, Washington
which consists of 21 km of soft ground tunnels with accompanying shafts and appurtenant facilities which are being
constructed as part of a new waste water treatment system for the greater Seattle area. The scope and history of the project
and the various procurement methods used are outlined. The project anticipates four pressurized face (EPB or slurry) TBMs
working simultaneously from composite slurry wall shafts and an open cut excavation during 2008 and 2009. The tunnel
drives are expected to encounter highly variable glacial geology and hydrostatic pressures of up to 7 bar. The paper
discusses TBM selection, construction methods and constraints and other construction challenges. Pipe installation in
tunnels and the challenging project schedule are described. The paper is intended to be a predecessor to future papers which
will describe how the challenges of the project were met.

11
includes approximately 21 km (13 miles) of
conveyance pipelines, a 643 million l/day (170 million
gallons/day) pump station, a marine outfall into the
Puget Sound and various appurtenant facilities for
odour control and integrated operations of the
systems.
2 CONVEYANCE SYSTEM DESCRIPTION
2.1. System Components
The Brightwater conveyance system is shown in
Figure 1 and consists of the following components:
21 km (13 miles) of bored tunnels with internal
diameters of 4.3 m to 5.1 m (14 ft to 17 ft), with a
variety of secondary lining and internal piping
configurations (as shown in Figure 2);
5 portals comprising 2 open cut excavations (one
at each end of the main tunnel) and 3 intermediate
shafts ranging from 26 m to 62 m (80 ft to 200 ft) in
depth;
2 km of microtunnels with finished internal
diameters of up to 1.98 m (78 inches);
An influent pump station (IPS) with a peak capacity
of 643.5 Ml/day (170 million gallons/day);
Several odour control facilities.
2.2. Influent System
Influent will flow into the conveyance tunnel at the
North Kenmore and North Creek portals. The influent
system comprises a 1.37m (54 inch) inside diameter
gravity pipe from the North Kenmore portal to the IPS
at the North Creek portal. The IPS will pump the flow
through 1.22 m and 1.68 m (48 and 66 inch) force
mains to the treatment plant. The microtunnels will
carry flows from the existing collection system to
each of the portals.
2.3. Effluent System
Treated effluent from the treatment plant will flow
along the tunnel under gravity to the marine outfall.
The effluent system is a 2.13 m (84 inch) diameter
gravity to full-pressure pipe from the treatment plant
to North Kenmore, increasing to 3.2 m (126 inch)
diameter to Ballinger Way portal. A partially-full
gravity flow pipe 3.96 m (13 ft) in diameter will run
from Ballinger Way to the marine outfall, decreasing
to 3.05 m (10 ft) over the last 762 m (2500 ft).
2.4. Alignment
The treatment plant is sited just north of the city of
Woodinville in Snohomish County, approximately
19.3 km (12 miles) inland from Puget Sound. The
conveyance tunnel heads south-west from the plant
for a short distance before turning and heading west
along the King County / Snohomish County border
towards the outfall location at Point Wells.
3 PROCUREMENT AND CONTRACTING
King County has adopted a variety of delivery
approaches for the design and construction of
Brightwater (Locke and Edgerton [1]). These were
chosen to suit the requirements and constraints of each
element of the project and are illustrated in Figure 3.
3.1. Treatment Plant Procurement
Design of the treatment plant was let as a single lump
sum contract. Construction will be carried out under a
General Contractor / Construction Management
(GC/CM) contract. The GC/CM contractor is limited
to self-performing 30% of the work (which it must
win by competitive bidding) and will thus sub-
contract the majority of the work. This was
considered the best approach to avoid interface risks
associated with the Owner managing multiple prime
contracts on a single project.
3.2. Marine Outfall Procurement
Several possible methods for constructing the marine
outfall were considered by King County. Experience
on past outfall projects indicated that contractors often
propose modifications to the design which are
Figure 1. Brightwater system layout
12
beneficial to the project. With this in mind, a Design-
Build contract will be used for the outfall.
3.3. Conveyance System Procurement
A conventional Design-Bid-Build approach has been
adopted for the conveyance system. The design was
let as a single cost plus fixed fee contract, providing
benefits in cost (economies of scale), coordination,
consistency and design schedule. After careful
consideration, construction was divided into three
major tunnel contracts plus an Influent Pump Station
contract and several smaller ancillary contracts.
Figure 2. Brightwater conveyance tunnel lining configurations
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.25 m (10") MIN.)
1.68 m (66")
INFLUENT
FORCE MAIN
1.22 m (48")
INFLUENT
FORCE MAIN
2.13 m (84")
EFFLUENT
PIPELINE
0.69 m (27") RECLAIMED
WATER PIPELINE
CELLULAR CONCRETE
BACKFILL
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.25 m (10") MIN.)
5.08 m (16' 8") MINIMUM
FIBER OPTIC CABLES
1.83 m (72")
EFFLUENT
PIPELINE
1.37 m (54")
INFLUENT
PIPELINE
WATER PIPELINE
CELLULAR CONCRETE
BACKFILL
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.33 m (13"))
FIBER OPTIC
CABLES
STRUCTURAL
CONCRETE
4.37 m (14' 4")
FIBER
OPTIC
CABLE
TUNNEL SECTION 1 (EAST CONTRACT)
LENGTH : 4228 m (13,872')
TUNNEL SECTION 2 (CENTRAL CONTRACT)
LENGTH : 3532 m (11,587')
TUNNEL SECTION 3a (CENTRAL CONTRACT)
LENGTH : 1378 m (4520')
TUNNEL SECTION 3b (CENTRAL CONTRACT)
LENGTH : 4721 m (15,490')
4.37 m (14' 4") 4.37 m (14' 4")
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.30 m (12"))
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.30 m (12"))
CELLULAR CONCRETE
BACKFILL
3.20 m (126")
EFFLUENT
PIPELINE
WATER PIPELINE
0.36 m (14")
RECLAIMED
WATER PIPELINE
WATER PIPELINES
EFFLUENT
TUNNEL
STRUCTURAL CONCRETE
3.96 m (13' 0") MINIMUM 3.96 m (13' 0") MINIMUM
TUNNEL SECTION 4a (WEST CONTRACT)
LENGTH : 5645 m (18,520')
TUNNEL SECTION 4b (WEST CONTRACT)
LENGTH : 764 m (2505')
BOLTED, GASKETED
CONCRETE SEGMENTAL
LINING (0.25 m (10") MIN.)
EFFLUENT
TUNNEL
3.05 m (120")
EFFLUENT
PIPELINE
BACKFILL CONCRETE
STORM WATER
STORAGE RESERVOIR
13
The division of the tunnel contracts was based on
schedule and interface considerations together with a
goal of keeping each contract below $300 million in
order to encourage competition. All three tunnel
contracts are lump sum. Construction management
services for the conveyance system have been let
under a separate contract.
3.4. Contractor Qualifications
Washington State law requires that construction
contracts are awarded to the lowest-priced,
responsive, responsible bidder. Prequalification of
bidders is not allowed in Washington State without
legislative approval. The two low bidders must
therefore submit qualifications that demonstrate an
appropriate level of experience, technical competence
and successful past performances on similar projects.
Because construction of the Brightwater conveyance
system shafts and tunnels poses some demanding
challenges, the required qualifications have been
defined in detail, separately, for each of the tunnel
contracts.
Depending on the contract, the low bidder will be
required to demonstrate specific experience in
constructing long, large diameter tunnels in a
pressurized face environment, constructing shafts
using diaphragm walls or ground freezing and
installing large diameter pipes within tunnels.
Additionally, the Central and West tunnel contracts
require experience in hyperbaric work at pressures
above 3.5 bar (50 psi).
4 CONSTRUCTION SCHEDULE
The main factor governing scheduling of the various
contracts is the October 2010 target date for flow
through the conveyance system. Since the
conveyance system is linear, all tunnels and pipes
must be ready to receive flow by this date. Figure 4
shows a simplified schedule for the major conveyance
construction contracts. The start dates for each of the
tunnel contracts are spaced at approximately six
month intervals. This timing is intended to allow time
to determine which contractor each contract will be
awarded to before bidding the next. Each six month
period includes time for bidding, review of contractor
qualifications and time to resolve possible bid
protests. The first tunnel construction contract (the
East tunnel) commenced in January 2006. The
Central and West contracts are expected to commence
in the second half of 2006 and first half of 2007
respectively. The pumping station and ancillary
contracts are scheduled around the main contracts.
The linear nature of the system means that each of
the main contracts has an interface with at least two
other main contracts (one at each end). Additional
interfaces occur with the pump station and ancillary
contracts. For this reason particular care has been
required to define and manage contract interfaces to
avoid potential interference between contractors. This
has been achieved by careful scheduling and defining
a series of progress milestones within each contract
and was one of the factors governing the division of
the tunnel contract packages.
CONTRACT 2006 2007 2008 2009 2010 2011
EAST
CENTRAL
WEST
IPS
OUTFALL
TARGET DATE FOR FLOW THROUGH SYSTEM
02/06
07/06
02/07
06/07
07/07
11/09
10/10
03/11
08/10
10/09

Figure 4. Simplified conveyance schedule
5 GEOLOGICAL CONDITIONS
5.1. Geologic Setting
In the Puget Sound region the Juan de Fuca oceanic
plate has been subducted beneath the North American
Figure 3. Brightwater final design and construction contracts
14
continental plate, creating a back-arc basin oriented
north to south, known as the Puget Trough. Seattle is
situated in the Puget Sound Lowlands, on the eastern
side of the Puget Trough and bordered by the Cascade
Mountains to the east and Puget Sound to the west.
It is generally accepted that the Seattle area has
been subject to at least 6 glaciations, most
significantly the Vashon Stade which took place
between 10,000 and 20,000 years ago, advancing in a
north-south direction.
The local terrain reflects both the structural and
glacial geological history of the area. Overall it slopes
gently westward from the Cascades to Puget Sound
(Galster & LaPrade [3]) and is characterized by north-
south aligned ridges and steep-sided valleys,
reflecting the direction of the Vashon Stade glacial
advance.
The sedimentary deposits of the area have been
formed by successive glacial and non-glacial
depositional cycles. Glacial advances occurred from
the north. Consequently the glacial deposits found in
Seattle originated in British Columbia. Non-glacial
deposits originated in the Olympic and Cascade
Mountains to the west and east respectively and were
transported by water rather than ice. The differing
sediment origins, transport mechanisms and prevailing
depositional environments have resulted in a complex
and highly variable superficial geology, with
significant differences between glacial and non-glacial
deposits. A comparison of some major characteristics
of the glacial and non-glacial deposits is presented in
Table 1.
5.2. Hydrogeology
The complex alternating sequences of coarse- and
fine-grained soils have formed a discontinuous series
of aquifers and aquitards of varying thickness, not all
of which are recharged from ground level. As a
consequence groundwater conditions vary greatly
over the conveyance system alignment and are not
consistent with the ground surface profile.
5.3. Tunnelling Conditions for Brightwater
Throughout the Brightwater project area the depth to
bedrock ranges from 120 m - 180 m (400 - 600 ft)
near the treatment plant at the east end of the project,
to over 450 m (1500 ft) near the marine outfall at the
west end [4]. Consequently all deposits encountered
by the tunnels and shafts will be late Quaternary and
Holocene glacial and non-glacial sediments.
Deposits associated with at least three glacial and
three non-glacial depositional cycles are found in the
Brightwater project area. These include glacial tills,
non-glacial soils, lacustrine clays and silts and
outwash sand and gravel. Overall the depositional
pattern is highly variable, as explained above.
However, clay deposits (both glacial and non-glacial)
tend to be more laterally extensive than coarse-
grained soils. A simplified longitudinal geological
profile of the tunnel alignment is presented in Figure
5.
Deposits which have been glacially over-ridden
(whether glacial or non-glacial in origin) are generally
over-consolidated, dense to very dense and or stiff to
hard, while those not glacially over-ridden (late
Vashon and post-glacial deposits) are generally loose
to medium dense or soft to very stiff.
In general the mineralogy and angularity of the
glacial deposits make them likely to be more abrasive
during excavation than the non-glacial deposits. The
elongated north-south depositional patterns of the
glacial deposits mean that there will probably be more
variability in the glacial deposits when tunnelling east
to west than when tunnelling through non-glacial
deposits.
Table 1. Comparison of Glacial and Non-Glacial Sediments in Puget Sound Lowland
Characteristic Glacial Sediments Non-Glacial Sediments
Deposition
pattern
Deposits elongated in north-south direction. Not elongated. Occur in broad sheets or sinuous
ribbons associated with watercourses.
Composition Relatively high quartz and feldspar content
(55% - 65%)
Relatively low quartz and feldspar content (40% -
60%)
Angularity Angular grains. Moderately well-rounded grains.
Clay deposits Generally thick, laterally extensive and uniform
deposits, with high rock flour content.
Generally inches to a few feet thick and
interbedded with silts, sands and organics. Higher
clay mineral content.
Organic content Very little. Fragments only. Some continuous organic layers. Some logs.
Boulder content High risk of boulders within glacial deposits, in
buried meltwater channels and at contacts with
non-glacial strata.
No boulders, expect at contacts with glacial
deposits.
15
5.4. Hydrostatic Head at Tunnel Depth
Hydrostatic head at tunnel invert level for the whole
alignment is shown on Figure 5 as a dashed line. For
the East contract a fairly consistent head of 2.5 to 3
bar (37 to 44 psi) is expected. Pressures for the
Central contract tunnels vary from 2.5 bar (37 psi) to a
maximum of 7 bar (103 psi) beneath the Lake Forest
Park area. For the West contract water pressure is less
than 2 bar (29 psi) over the western half of the
alignment, rising sharply to 4 to 5 bar (59 to 73 psi)
near the mid-point of the drive and remaining above 3
bar (44 psi) for the remainder.
6 TECHNICAL CHALLENGES
Construction of Brightwater conveyance system
presents a range of construction challenges, the most
significant of which are described in the following
sections.
6.1. Long Tunnel Drives
Tunnelling will be carried out in four drives with
lengths of 4428 m, 3532 m, 6099 m and 6416 m
(13,872 ft, 11,587 ft, 20,010 ft and 21,052 ft)
respectively. Sections 3 and 4 will be the longest
tunnel drives to date in the Seattle area.
6.2. TBM Selection
Selection of an appropriate TBM is a critical factor in
achieving a successful tunnel drive. For all
Brightwater contracts, pressurized face TBMs are
required for all drives, although the variability of the
ground conditions means that the choice between
Earth Pressure Balance (EPB) and slurry TBMs is not
always obvious and will require careful judgment on
the part of contractors. For the East and West
contracts selection of the TBM is left to the
contractors. However, for the Central contract slurry
TBMs have been specified for both drives. At the
time this paper was prepared only the East contract
had been awarded and the contractor elected to use an
EPB machine.
6.3. Cutterhead Inspection and Maintenance
The tunnel contract specifications require that the
contractor maintains the cutterhead in operating
condition as per the TBM manufacturers
recommendations. The intent of the inspection
requirements is to achieve an efficient preventative
maintenance regime that minimizes the risk of having
to carry out major repair work in the expected high
hydrostatic head conditions.
The specifications define in detail the requirement
for the contractors to carry out TBM cutterhead
inspections. The Owner is providing direct payments
for inspection stops in the form of a unit price
payment for each inspection undertaken. Typically,
the contractor will be required to inspect the
cutterhead after every 150 m (500 feet) of TBM
advance, using a variety of techniques including
remote camera, tool wear indicators and manned
entry. In the more challenging zones the inspection
intervals are reduced.
6.4. Compressed Air Interventions
It is envisaged that entry into the TBM cutterhead for
inspection and maintenance will be undertaken either
in a compressed air or a free air environment,
depending on the prevailing ground conditions. The
specifications require the TBMs to be fitted out with
the facility for compressed air entry into the
cutterhead chamber.
Compressed air working in Washington State is
regulated by the State of Washington Administrative
Code (WAC) 296-36. However, the contract
specifications indicate to the contractor that other
worldwide best practices may also be used in
conjunction with the WAC. The current WAC
regulations require that, except in an emergency, no
person shall be compressed to a pressure exceeding
3.4 bar (50 psi). In order to accommodate the
expected cutterhead entries on the Central and West
contracts in compressed air pressures higher than the
WAC regulations, the contractors will be required to
submit a compressed air working plan to the State of
Figure 5. Simplified geological profile along conveyance tunnel
16
Washington Department of Labor and Industries, in
order to obtain a variance to WAC 296-36.
The Central and West contractors will also be
required to engage an experienced hyperbaric
consultant with experience in managing pressures
greater than 5 bar (73 psi). This consultant will
prepare the compressed air work plan and oversee the
operations which may include the use of mixed gas
breathing under the higher pressures.
6.5. Settlement
Over most of their lengths the conveyance tunnels are
at depths exceeding 30 m (100 ft) so surface
settlement is unlikely to be a significant issue.
However there are limited reaches of tunnel on each
contract where the tunnels are shallow and will pass
beneath private residential property and major
transportation corridors. Particular challenges
include:
Tunnel Section 4 (West contract) will pass 5m (16
ft) beneath the main Burlington North Santa Fe
railroad immediately after the TBM is launched and
before the backup is fully installed.
Tunnel Section 3 (Central contract) will pass 11 m
(36 ft) beneath a residential area shortly after the
TBM is launched from the North Kenmore portal, in
an area of coarse granular soils and artesian water
pressures.
Tunnel Section 2 (Central contract) will pass 15 m
(50 ft) beneath the I-405 highway shortly before
reaching the end of the drive, in an area of granular
soils and slightly artesian water pressures.
6.6. Pipe Installation in Tunnels
Tunnel Sections 1 and 2 (East and Central contracts)
require the installation of multiple large diameter
pipes and concrete backfill within the lined tunnels
over long distances. A range of material types is
allowed, including steel, concrete and fibreglass-
reinforced polymer pipe.
The installation process is likely to be complex
and require careful planning. Issues to be addressed
include:
Transporting multiple pipes in a single operation;
Limited working space within the tunnel;
Preventing formation of voids within the backfill;
Transporting concrete over long distances.
No method of installation has been specified,
although numerous possibilities have been evaluated
during design. The actual method of installation will
be decided by contractors.
6.7. Shaft Construction
The North Creek Portal site will house two permanent
shafts approximately 24.4 m (80 ft) deep, both of
which will be formed using slurry wall techniques.
The site has limited working space and it is expected
that excavation of the Influent Shaft (IS) will be
carried out concurrently with slurry wall construction
for the complex figure-of-eight shaped Influent
Pump Station (IPS) Shaft, which will eventually house
a pump station. East contract tunnelling will be
carried out from the IS shaft.
At the Ballinger Way portal a shaft 61 m (200 ft)
deep is to be excavated. Slurry wall or ground
freezing may be used to construct the shaft. Design of
the shaft forms part of the Central contract. This site
also has a very limited working area.
6.8. Environmental Constraints
The location of the project in an urban area requires
the construction work to be managed within strict
environmental controls. Contractors will be required
to work within limits defined by the Owner,
regulatory agencies and jurisdictions to protect the
environment. Each contractor will be required to
provide a Site Environmental Manager with relevant
experience to oversee its environmental mitigation
and control plans.
The main environmental concerns deriving from
the construction works are related to site traffic, light,
noise, air pollution, working hours and site water
discharge. The specifications clearly define the limits
within which the contractors will be required to work.
Specific efforts are focused on the site water discharge
to which tight controls are imposed on activities in
and around wetlands, classification of discharge
waters and discharge levels of pollution.
Working hour limits are defined to mitigate the
impact of construction noise. However, activities
specifically relating to the tunnel excavation operation
are permitted 24 hours per day Monday to Saturday,
although removal of tunnel spoil from the sites during
the night is prohibited.
7 CONCLUSIONS
This paper has provided an overview of the
Brightwater conveyance project and the major
technical challenges which will be faced as
construction proceeds. Future papers will address
how these challenges were met.
REFERENCES
1. Locke, C., and W.W. Edgerton. 2005. King Countys
delivery approach for the Brightwater project.
Proceedings of the 2005 Rapid Excavation and
Tunneling Conference, Seattle, August 2005, eds.
Hutton, J.D., and D.W. Rogstad, 582-589. Colorado:
Society of Mining, Metallurgy, and Exploration, Inc.
2. Adams, D.N., J. Johnson, L. Maday and D.L. Pecha.
2005. Design of the Brightwater conveyance tunnels.
17
Proceedings of the 2005 Rapid Excavation and
Tunneling Conference, Seattle, August 2005, eds.
Hutton, J.D., and D.W. Rogstad, 590-601. Colorado:
Society of Mining, Metallurgy, and Exploration, Inc.
3. Galstar, R.W., and W.T. LaPrade. 1991. Geology of
Seattle, Washington, United States of America.
Bulletin of the Association of Engineering Geologists
3:235-302.
4. Camp Dresser & McKee Inc. 2005. Geology and
hydrogeology regimes Brightwater conveyance
system. Technical memorandum to King County Dept.
of Natural Resources and Parks, Wastewater
Treatment Division. April 26, 2005.
ACKNOWLEDGEMENTS
The authors are grateful to King County Wastewater
Treatment Division for their permission to publish this
paper, and for providing graphics forming the basis of
Figures 1 and 5.
18

ABSTRACT

1 INTRODUCTION
Located on the south side of Columbus Ohio,
BWOAS is the second stage of the Clean Rivers
project for the city of Columbus. Both tunnel
contracts were designed by a team lead by URS
Corporation with Lachel Felice & Associates and H.R
Gray as sub consultants. The tunnels were specified as
closed face pressurized TBM drives lined with bolted
and gasketed pre-cast cast concrete segments.

Fig. 1. General layout.
Part 1 of the scheme was awarded to the J ay
Dee/Michels/Traylor J V in 2003 and construction
commenced the same year. The BWOAS contract
(Part 2) followed and was awarded to the
McNally/ Kiewit J V in October 2004.
BWOAS (Part 2) is located approximately 10
miles south of the City of Columbus, Ohio and
extends from the intersection of Alum Creek
Drive and State Route 317 to a location just north
of Interstate 270. The

Fig. 2. Arial photograph of site.
BWOAS tunnel consists of approximately
4000m (13,200 linear feet) of 3.66m (12-foot)
internal diameter concrete sewer tunnel and
approximately 1600m (5,400 linear feet) of
1.066m (42 inch) diameter sewer, constructed in
open cut and trenchless construction.
The alignment parallels Alum Creek Drive,
the centerline of the tunnel being within 16m (50
feet) of the west edge of pavement. The tunnel
crown varies from approximately 14m to 20m (46
to 64 feet) below this major 4-lane arterial road.
The tunnel includes the crossings of six different
2-lane roads, a parking lot near an office building,
Big Walnut Creek and a 2.74m (108-inch)
diameter sanitary sewer.
The Big Walnut: High Tech Tunnelling in Columbus
Steve Skelhorn
Project Manager, McNally / Kiewit J oint Venture
The Big Walnut Outfall Augmentation Sewer (BWOAS) is the second stage of an ongoing sewer expansion for the City of
Columbus, Ohio. The main tunnel involves 4000m (13,200 feet) of 3.66m (12 foot) ID tunnel with 6 access shafts. An open-cut
connection to the nearby Rickenbacker Airport also forms part of the contract. The paper looks at the various aspects faced at
award and details the high-tech approach to overcome the various challenges. Linabond corrosion protection is to be installed after
the tunnel is complete. With planning for this activity still is in its early stages, Linabond is not covered within this paper.
19

The Big Walnut Creek is a major drainage
feature in the area and the tunnelled crossing will
be approximately 60m in length and will be
constructed with approximately 5m (17 feet) of
cover from the tunnel crown to the stream
bottom.
The project will require a drop structure to
accommodate the flow from the existing 2.74m
Big Walnut Outfall sewer. This Interconnect
Structure (ICS) will be near the crossing of the
new and existing sewer and will include a
manhole access structure to tunnel level. In
addition to the Interconnect Structure the project
will require five other manhole access structures,
which are numbered from Shaft 8 near the Alum
Creek Drive-State Route 317 intersection through
Shaft 12 at the northern end of the project.
The Construction Management Team (CMT),
within the contract document specified EPB
tunnelling. The specifications detailed a number
of measures to control the tunnelling, de-watering
is specifically restricted; shafts are to be built as
slurry walls or auger casings; the TBMs are to
incorporate the latest enhancements to control
and monitor settlement; and an extensive array of
monitoring points are to be installed.
Additionally, on completion the tunnel is to
be lined internally with the Linabond Co-Lining
corrosion protection system to ensure a required
100-year design life. Linabond was selected by
the design team specifically because it is a
separate installation operation. The designer
wished to avoid most of the problems associated
with using systems that are cast integral with the
segments. These include quality control at the
casting yard, heat welding the joints and also the
potential for damage to integral membranes as the
tunnel advances.
A more detailed description of the project
background can be found in Chapman et al. [1].
This paper focuses on the TBM tunnel and
takes a look some of the major tunnelling
methods to be used on this project to meet the
owners expectations.
2 GEOLOGY
A Geotechnical Baseline Report was issued with the
contract documents. Generally the expected strata
consist of glacial or post glacial deposits with up to 2-
bar water head. Cobbles and boulders are expected
along the entire route. The first 25% of the tunnel is
expected to be located within a competent till layer.
After this water charged sands and gravels are
expected.
A more detailed description of the
geotechnical aspects was presented by Frank and
Chapman [2].
3 TBM
The TBM selected was a Lovat full face EPB
machine. The 4.26m (168) diameter TBM was the
latest generation of Lovat machines and incorporated
many of the latest innovations for tunnelling. Many of
these were specified requirements or were added to
meet the specifications.

3.1. Cutter Head
The machine was equipped with a soft ground cutting
head incorporating ripper teeth, scraper teeth and disc
cutters. Wear protection was provided with Trimay
plating over the entire face.
The cutter head was driven by a VFD drive
consisting of four 200kW (265hp) electric
motors. A small diameter bearing was initially
offered by Lovat. Although this allowed the
screw conveyor to be mounted at the bottom of
the head, concerns were raised regarding the
potential to jam on boulders. Consequently a
larger diameter bearing was selected with the
screw mounted approximately one quarter of the
way up the face.

20

Fig. 3. Lovat RME-169 TBM.
3.2. Segment Erector
Within a 3.66m tunnel, space is limited, especially
around the ring build area. Our preference during
design was for a ring erector mounted around the
screw conveyor. Recent developments have provided
a ring type erector mounted to the rear of the
stationary shell. This style of erector, utilised
previously on a Lovat for the NEIS project in
California, Zernich et Al [3], provided greatly
improved clearance within the ring build area and
provides a much larger window for the laser beam.

3.3. Screw Conveyor
To provide the specified positive displacement screw
discharge system, a twin screw was adopted. The
rationale was that the two screws can be rotated at
different speeds, controlling face pressure with the
forward screw and belt discharge with the rear. The
double screw also provides a secondary advantage in
that the screw discharges to a horizontal section of
belt, mitigating the potential for spillage at this
location.

3.4. Grout System
A two component grout system was selected for this
project. Backfill grout to the rings was to be
controlled by pressure and not volume. With the
variability in the ground it is not possible to accurately
predict the total volume of grout required. Pumping
retarded grout to a holding tank on the TBM allows an
unlimited supply of grout not dependent upon the
train. This provides an added advantage of having a
supply of grout at the TBM allowing any leaks or
blow-out to be dealt with immediately. Additionally,
the omission of a grout transfer car allows a shorter
train and eliminates the need to split the train under
the TBM gantry during the mining phase.
To allow for pumping of the grout, a
specialised grout mix was required. The grout
selected was developed by Master Builders and
consisted of a cement / fly ash blend with a
chemical retarder and a viscosity modifier added
at the batching plant. The retarder volume could
be modified to adjust the initial set for a few
hours to several days. Bentonite was
subsequently added to the mix to mitigate the
drop-out observed in the pump lines. At the point
of injection an accelerator (sodium silicate) was
added to the grout mix.
A surface batching plant was set up. This
consisted of two silos (Cement and fly ash),
weigh batcher (supplied by Standley), and mixer
(supplied by Chemgrout). The system was
automated by a PLC. Smith and Long Ltd,
Toronto, Canada, provided the control panels and
software programming. Grout is batched in 200
litre ( cubic yard) batches and transferred to a
1.5 m
3
(2 cubic yard) holding tank for transfer to
the tunnel. A three stage Moyno pump transfers
the grout through a 50mm (2) line to the TBM.
At the TBM, Lovat supplied a grout injection
system consisting of holding tanks for the grout
and accelerator and two Moyno pumps to pump
the grout. In addition, Master Builders supplied
two peristaltic pumps to dose the accelerator. The
injection system was controlled by a PLC
integrated to the TBM PLC.
With theoretical grout volume at 1.5m
3
(2cy)
per ring, a 3m
3
(4cy) tank was supplied with the
TBM.

3.5. Ground Conditioning
A ground conditioning system was incorporated on
the TBM consisting of four independent pumps
pumping to four of five available ports ahead of the
cutter head. For the first section of tunnel, hard till
was predicted requiring foam rates of up to 80%
(foam injection ratio). After this the foam rate is
expected to be much less but will need the addition of
polymers to control the wet sands and gravels.

21

3.6. Guidance System
The TBM manufactured offered a tacs guidance
system with the TBM. Prior experience with these
units has been good and we had no hesitation in
accepting this option. The unit was fully up to date,
incorporating a Leica 1103 motorised theodolite and
the latest software which allows the TBM progress to
be plotted in real time on an imported AutoCAD
drawing of the project plan, profile or geotechnical
profile.
As the unit was supplied with the TBM and
was required to meet the requirements of Class
one Div 2 electrical specifications, tacs supplied
both the tacs control computer and the Leica
enclosure in explosion proof boxes. Although this
requires a large laser window, the box does
provide some protection to the instrument.

Fig.4. tacs guidance system.
The tunnel alignment consists of 15, 250m radius
curves. This presents potential problems with the line
of sight for the laser and requires the laser to be
moved forward every 9 rings during transition of the
curves.
4 TUNNEL RINGS
The contract specified six-piece bolted segmental
rings consisting of 3 parallel sided plates, one key and
two top plates. During bid stage it became clear that a
universal style ring would be preferable. With the
requirement to line the tunnel on completion, the
fewer bolt pockets the better as all irregularities must
be filled prior to affixing the liner. To minimise the
bolt pockets, it was elected to use dowels rather than
bolts of the circumferential joints. A universal ring, by
virtue of each plate being a trapezoid, allows for much
easier building with a dowel.
A six piece universal ring was selected
consisting of four 67-1/2 degree rhomboid and
two 45 degree trapezoid segments. The design
incorporates a 38mm taper to the ring and allows
the ring to be rotated through 360 degrees around
the tunnel axis. A ring width of 1.5m (5-feet) was
selected. Although this width creates limited
clearances for transport under the TBM gantry, it
reduces the total number of rings, again
minimising the amount of patching required.
To allow for the final key to be placed in the
top, the ring can also be rotated by 180 degrees
about the vertical, essentially providing alternate
keys in the top for adjacent rings. As this system
requires alternate rings to be delivered to the
TBM in a different order and complicates the
placement of ring packers, we decided not to
elect this system. Instead we bring every ring into
the TBM in the same order and always
commence the ring build with the same key plate.
This on occasion results in the first plate being
built above spring line and the final key plate
inserted below, however the versatility of the
erector allows this system to work without any
problems.

Fig. 5. Tunnel ring.
Ring design was carried out by Hatch Mott
MacDonald (Toronto, Canada) with assistance
from Chris Smith of CRS consultants (UK).
During the initial stages of design, the option of
using steel fibres was explored. Based on success
in European tunnels, the use of fibres simplifies
the segment casting process and also provides a
tougher ring, less susceptible to superficial
damage during handling. Unfortunately, due to
the 1.5 m width of the ring, combined with the
22

geometry of the TBM, the taper angle of adjacent
plates eliminated the possibility of using full fibre
rings. However, Hatch was able to design a
hybrid ring consisting of a reduced re-bar cage
with added fibre reinforcement. This provided
the advantage of reduced handling in the segment
casting plant, combined with the increased
durability from the fibres.
A segment casting yard was established in
Mount Vernon, Ohio, approximately 100km
north of the job site. Segment casting was carried
out by North American Segment Company,
managed by Chris Smith of CRS.

Fig. 6. Test ring erected.
CBE custom built sufficient forms to cast 16
full rings each day, working on a two shift basis.
Steam curing allowed for rapid turnaround on the
forms. Using a single type universal ring allowed
the segments to be stripped from the forms and
stacked in half rings. These stacks remained
intact until delivery to the segment unloading
station on the TBM, minimising the potential for
damage during handling.

Fig. 7. Casting yard.
5 SHAFT STRUCTURES
The specification mandated a slurry wall for the
launch shaft at Shaft 8, and gave the option to use
either slurry wall or auger casings for the intermediate
shafts. We elected to go with the auger shaft option
for the intermediates. The recovery shaft at the ICS is
located within a dewatered zone. A rectangular soldier
pile and lagging shaft will be constructed at this
location.
To safeguard the ground around each of the
shafts a jet-grout block was specified. This
consists of a block of grout around the two tunnel
eyes in the case of Shaft 8 and a block completely
encasing the base of the shafts, for the
intermediates.

5.1. Slurry Wall Shaft
Shaft 8 was a 12m (40-foot) diameter, 25m (80-foot)
deep shaft. The shaft was designed by the CMT as a
slurry wall extending 5.2m (18 feet) below the tunnel
invert with a specified 4.5m (15-foot) thick plug to be
tremmied in place. Soletanch were subcontracted to
excavate the slurry walls. The slurry wall design
incorporated soft eyes (fibre glass re-bars) and
included box-outs which were to provide keyways for
the base slab. The shaft was formed in 7 major panels
consisting of three smaller panels cast as one.
On completion of the slurry wall, excavation
commenced. With groundwater elevations only
6m below ground, excavation was to be carried
out underwater, however with a competent till
cap over the water bearing sand it was possible to
excavate the shaft to 6m above the tunnel before
flooding the shaft. The final 10m of excavation
was carried out by clam bucket in a flooded shaft.
For the box-out connections, it was necessary
to employ divers to access the base of the shaft,
remove the box out material and install the tie-in
bars for the shaft base.

Fig. 8. Slurry wall shaft.
23

On completion the shaft base was tremmied in
place before pumping out the shaft.

5.2. Auger Shafts
The Auger shafts were subcontracted to Koker
Drilling. These involved a 3m (10-foot) diameter shaft
augered within a steel casing. Shaft depths were
generally around 20m (60-feet) and all shafts were
augered to tunnel axis before backfilling with
uncompressible fill to just above the tunnel. This fill
was required to provide a seal for grouting the casing.
On completion the u-fill was removed ready for the
TBM.
The intent is to drive the TBM to the centre of
the shafts where the upper part of the head will be
visible. To safeguard the earth pressure the shafts
will be flooded prior to TBM arrival. Expanding
polyurethane grouts will be used to seal around
the annulus before pumping out the shafts. At the
time of writing the TBM was approaching the
first shaft and preparations were underway for the
arrival.
6 LAUNCH PROCEDURES
The launch shaft was 12m in diameter; however the
design requirements dictated an offset to the tunnel.
This resulted in an effective launch window of only
10m. While the TBM could be installed in this space,
there was insufficient room to enable mucking or to
allow the main gantry components to be connected.
An initial plan was to set up the full trailing
gear on the surface and use an umbilical system
to connect the TBM. However, this was
discounted due to the cost of such a system. It
was decided to install the TBM in the shaft in
sections, launch the machine with a short screw
and extend the trailing gear and conveyors as the
tunnel advanced.

Fig. 9. TBM launch.
The front two sections of the TBM were
installed within the shaft along with the relevant
gantry sections holding the TBM transformer and
switchgear. With over 70 electrical connections
running to the TBM it was originally planned to
use a split tailcan which would later be wrapped
around the cables to prevent the need to
uncouple. The major concern with this method
was to get the tailcan back together perfectly
round. Consequently the method was adjusted
and for the initial launch the full tailcan was
suspended from a shelf within the shaft. This
allowed the cables to be passed through the can
negating the need to reconnect.
7 CONTINGENCY MEASURES
Several contingency measures have been drawn up,
although to date none have been required. The major
contingencies concern access to the TBM head for
cutter teeth replacement or boulder removal.
The plans for entry in all cases follow a
sequential protocol and will be implemented in
the following order.
Attempt head entry under free air
Treat the ground around the TBM using
polyurethane grout
Install well points around the TBM and locally
de-water.
Install an in-tunnel compressed air lock and
access under compressed air (maximum expected
pressure is 55 kPa (8 psi)).
8 DATA ACQUISITION
As with most recent tunnelling contracts data
acquisition is a major component of the owners
prerequisites. The TBM is equipped with a PCL
controller which reports to a data logger continuously.
This includes most of the TBM operational data as
well as inputs from the guidance system.
The data is transmitted by modem to surface
computers and from there is downloaded to a
VPN. This enables real time data viewing over
the internet. This information is shared with the
owner as specified and with the TBM
manufacturer, which can greatly assist in
troubleshooting.
9 PROGRESS TO DATE
24

As at the time of writing, the tunnel construction was
in its early stages with 600m (2000ft) of tunnel
completed.
The initial breakout through the jet grout zone
allowed a non-EPB launch, however the cutting
through the wall took a long time. This was due
to a higher strength concrete than expected,
combined with a deliberate slow push to protect
the cutting teeth.
The initial TBM launch proved difficult. The
machine was launched in August of 2005 and in
September the first rings were built. Progress was
very slow for the initial launch and it was not
until J anuary 2006 that the entire TBM was fully
buried and the trailing gear fully hooked up.
Mining continued and at the time of writing
(May 2006), 400 rings were installed within the
tunnel.
There have been some teething problems with
the initial launch, with some of the teething
causing more problems than others. However,
these have been ironed out as the tunnel
advanced.
In terms of the major components, such as the
TBM drive, grout systems, segment erector and
concrete segments, it is too early in the project to
give a definitive report.
Average mine times at this stage are around
15 minutes with ring build times around 25
minutes.

Fig. 10. TBM within shaft.
10 CLOSING REMARKS
It is difficult to write about the progress of a tunnel in
its early days and consequently I have aimed this
paper more at the planning and set-up and not the
execution.
This project incorporates some state of the art
systems for TBM tunnelling, including a
computerised TBM, fibre reinforced concrete
segments and more data recovery than anybody
could reasonably use. It will be interesting to see
how these enhancements benefit modern
tunnelling.

REFERENCES
1. Chapman, D.R., T.L. Richardson, and G.W. Gilbert.
2005 BWARI Tunneling Underway. RETC Seattle,
2005.
2. Frank, G. and D.R. Chapman. 2005 A New Model for
Characterising the Cobble and Boulder Fraction for
Soft Ground Tunneling. RETC Seattle, 2005.
3. Zernich, Brett, B. Robinson and M. Krulc 2005.
Northeast Interceptor Sewer Case History. RETC
Seattle 2005.
25

1 INTRODUCTION
A tunnel eye is a hole in a shaft or portal shoring
bulkhead through which a Tunnel Boring Machine
(TBM) or microtunnelling machine (MTBM) passes
as it begins or completes a tunnel drive. Tunnel
eyes must be sealed against water, flowing ground,
and/or slurry pressure when closed face tunnelling
methods are used to limit loss of ground and control
settlement. The responsibility for design of the
tunnel eye often falls on the contractor, being
temporary work; however, the owners engineer
may want to limit the owners risk by requiring that
the contractors design include certain elements of
the tunnel eye such as dewatering, a soil cement
block and a mechanical seal.
Design responsibility for the performance of the
tunnel eye is often complicated by the fact that the
contractor responsible for the shaft shoring, ground
improvement (if any), and tunnelling may be three
separate entities. Consequently, there can be
several designers involved, each focusing on a
particular element of the work. The prime
contractor is responsible for ensuring the elements
work together to achieve the desired performance,
and this can be problematic when the prime
contractor is not the tunnelling contractor.
Because of the complexities involved and the
number of disputes related to tunnel eyes,
procedures are needed to assist the tunnelling
community regarding tunnel eye design in soft
ground. This paper provides an initial preliminary
framework for development of such procedures.
Regardless of the procedures discussed herein,
tunnel eyes will need to be designed on a case-by-
case basis and in accordance with the
Client/Owners risk tolerance. This paper focuses
on design of tunnel eyes in soft ground only;
excavations in rock are not considered.
Actual tunnel eyes in soft ground can be
constructed using a variety of methods from the
simple to the complex. These methods include the
following in ascending order from simplest to most
complex:
No seal or and informal non-engineered
seal,
Mechanical seal,
Ground improvement or blind panel walls,
Ground improvement or blind panel walls
with mechanical seal, and
Composite eyes.

A Framework for Preliminary Design of Tunnel Eyes
Anil Dean
David J . Young

ABSTRACT:
Design of tunnel eyes for closed face tunnelling involves a multidisciplinary approach that includes consideration of
geotechnical parameters, TBM specifications, temporary excavation and shoring design, ground improvement, tunnelling
induced settlement, scheduling, and constructability considerations. These issues are all inter-related at tunnel eyes and, as
a result, coordination of these issues within the construction contract documents becomes a significant issue. Because of
the multitude of inter-related design considerations at tunnel eyes, together with the lack of widely accepted design
procedures, construction of tunnel eyes is often problematic, which can lead to construction delays, disputes and claims.
This paper is intended to be used as a guide to assess data needs, design considerations, and methodologies for construction
of tunnel eyes. Selected tunnel eye design and construction case histories are reviewed so that tunnel contractors, designers
and construction managers can relate more easily to tunnel eye design considerations. In sum, a framework for developing
preliminary designs of tunnel eyes is presented, together with a discussion of steps required for successful tunnel eye
design and construction.

26

Figure 1. Tunnel Eyes from the Launch Chamber,
Terminal 5, Heathrow Express Rail Link Project
2 FUNCTIONS OF TUNNEL EYES
The multiple functions of tunnel eyes for a soft
ground, pressurized-face tunnel are summarized
below:
Resist earth and groundwater loads using
materials that a soft ground TBM can
excavate through while maintaining ground
stability and controlling ground
deformation.
Seal the excavation adjacent to the tunnel
against intrusion of water. Inflow of water
can lead to ground deformation around the
excavation due to piping-erosion, and due
to changes in soil volume associated with
increased effective stresses in the soil.
Inflows add to pumping and water
treatment work and can result in flooding of
the excavation.
Improve steerability and control of line and
grade at the launch. Experience has shown
that TBMs have a tendency to dive until the
body of the TBM is underground and
stabilized by the surrounding ground. This
tendency could be mitigated through
construction of a treated ground mass.
Facilitate a stepped build-up of face operating
pressures and annulus grout injection
pressures as subsequent lining rings are
grouted in. This stepped build-up can allow
for a more efficient thrust jack reaction
frame design in the launch shaft. Similarly,
it is beneficial to allow a pressure reduction
in steps at the reception shaft or portal.

3 DATA NEEDS
3.1. Ground Treatment
Geotechnical characterization, soil classification,
soil strength, and groundwater levels are needed.
Permeability should also be characterized.
Suggested calculations include, but are not limited
to, the following:
Lateral earth pressures on the shoring
system.
Hydraulic gradient and seepage velocity
along critical flow-paths. These
calculations may be omitted if ground
treatment is used and it can be shown that
permeabilities are sufficiently low to
preclude piping effects.
Strength demand and stability of the treated
soil block.
Settlement of soils beneath the block of
treated ground, considering change in
weight of the block and construction loads,
especially if soft cohesive materials
underlie the tunnel alignment.
In addition to the above calculations, annulus grout
or slurry pressure near the eye needs to be
considered to limit flow of grout or slurry back into
the shaft or portal.
4 DESIGN CONSIDERATIONS
The challenge in designing and constructing a
tunnel eye is in providing a seal capable of resisting
pressures exerted by pressurized face TBMs and
MTBMs, external groundwater and annulus
grouting pressures. There are subtle differences in
how pressures are applied to the seal during launch
and reception, but in general total pressure acting on
the seal is similar. This paper presents three
primary methods of creating the required seal in
order of sophistication and cost:
Mechanical seals, which are composed of
elastomeric compounds within a steel
frame and bolted or embedded in a
reinforced concrete collar at the
excavation headwall (Figure 2),
Ground treatment methods. A variety of
ground treatment techniques are
discussed, and
A combination of mechanical seals and
ground treatment.
27

While composite eyes, which use a combination
of ground treatment and shoring methods, are
effective in certain situations, their design is not
discussed in this paper because they have been
addressed by other authors [1,2].
Preliminary cost information suggests that for a
robust ground treatment option, a pair of tunnel eyes
needed for 6.1 m (20-foot) excavated diameter twin-
bore tunnels in California could cost in the
neighborhood of US$5
to $10 million,
depending on the
selected alignment
configurations. A
literature search revealed
that there are many
examples of mechanical
seal failures, which
suggests that mechanical
seals are not as reliable
as ground improvement,
which in turn is not as
reliable as a combination
of a mechanical seal and
treated block. There may
be justification for
relying on mechanical
seals for applications where the risk of seal failure is
acceptable, although this is not generally
recommended.
Selection of the eye system will depend on cost,
risk assessments (cost versus risk of failure),
schedule analysis, and constructability
considerations. Specific tunnel eye designs depend
on ground and groundwater conditions at the tunnel
eye location, the location of surface and subsurface
facilities, access constraints that limit ground
treatment options, and details of the tunnels and
shaft or portal excavation and shoring. These
combinations of factors need to be evaluated on a
case-by-case basis for each design. This paper can
be used as a guide to the eye system selection
process.

4.1. Temporary Excavation Shoring at Portals
The TBM must be able to excavate through the
structural shoring that is supporting the ground at
the tunnel eye. A rule of thumb is that drag-pick
equipped TBMs can excavate material with
unconfined compressive strengths of up to 3.4 to
10.3 MPa (500 to 1,500 psi).
On a recent project in Northern California, a 4.5
m (15-foot) diameter EPB machine with drag picks
excavated through jet grout blocks at the tunnel
eyes on four occasions with very little sign of pick
wear (Figure 3). The target strength of the jet grout
was 3.4 MPa (500 psi) and strengths from core
testing ranged from 0.2 to 21.4 MPa (35 to 3,100
psi) at 7 days.
Shoring systems could be either slurry wall or
sheet pile wall systems. A slurry wall system could
employ embedded reinforcing steel members, either
H-piles or rebar; however, steel should be avoided
at the eye to enable a
drag pick equipped
TBM to excavate
through. Glass fiber
reinforcement bars
are available that can
be readily excavated
with a TBM. If steel
sheet piles are used
for shoring, they
must be pulled or cut
from the tunnel eye
to permit the TBM to
excavate through.
Ground stability must
be confirmed and
ground improvement
employed if
Figure 2. Reinforced Concrete Collar at Excavation
Headwall
Figure 3. Photo Showing Little Pick Wear after a Tunnel
Drive in Northern California
28

necessary to maintain face stability in the event
sheets are cut out or pulled.
Blind panel walls can be constructed
immediately adjacent to excavation shoring systems
to resist lateral earth pressures in the tunnel eye, and
can be an economical component of tunnel eye
construction where slurry wall equipment is already
on site being used for excavation shoring. A blind
panel is constructed like a structural slurry wall, but
is typically composed of a weak cement bentonite
mix and does not include reinforcing steel or other
reinforcing materials that would damage or block
the cutterhead of the TBM.
In addition, there are construction loads applied
by the TBM and grouting equipment that need to be
considered when designing the shoring system.
Excavation shoring systems associated with tunnel
eyes on pressurized face tunnel projects also need to
resist water inflow.

4.2. TBM Launch
TBMs are typically launched from a steel frame
(Figure 4) that provides a reaction for the TBM to
push against as it propels forward. The frame is
designed to be compatible with the TBM weight,
dimensions and the anticipated thrust loading. Due
to the relatively large loads that occur at launch, the
frame is typically constructed on a reinforced
concrete slab, which can be integrated with the floor
slab of the portal structure if necessary.
To accommodate the machine and operational
requirements in the shaft, the shaft dimensions
should be larger than the width of the tunnel
opening by at least 90 centimeters (3-feet) all
around the machine for TBMs in the 6.1 meter (20
foot) diameter range.
This dimension is in
accordance with
accepted
international
tunnelling standards
[3].
After the
machine is launched,
thrust resistance is
gradually transferred
from the launching
frame to friction
between the tunnel
lining and the
surrounding ground.
After several rings
of tunnel lining have
been constructed
underground, the
launching frame may be removed and reused if
needed. Annulus grout pressures need to be
considered in the vicinity of the launch and
reception areas so as not to flow into the portal or
shaft.

4.3. TBM Reception
Reception of a TBM at a portal or shaft typically
involves several steps, which are briefly outlined
below:
Once the portal wall is reached, a survey is
conducted to check line and grade,
TBM face operating pressures are minimized
once stable or treated material is
encountered as the TBM excavation
approaches the shoring system,
After the TBM breaks through, a seal may be
needed to contain annulus grout and
groundwater as the final segment rings are
built and grouted in, and
Segment lining rings are built through the
tunnel eye for a sufficient length to
complete the permanent seal if necessary.
Appurtenant features such as seismic sleeves
are built in as required.
Line and grade are more important upon reception
because reinforced or sheet pile shoring systems are
time consuming to cut out.
5 CONSTRUCTION METHODS
5.1. Ground Treatment
The treated soil block is typically designed with two
criteria in mind:
strength and
hydraulic
conductivity.
Specific
considerations related
to these two criteria
are presented in
Table 1.
The potential
range of strengths is
dependent on both
the method of ground
treatment used and
the soil type
encountered. Ground
treatment methods
applicable for pre-
excavation treatment
Figure 4. TBM Launching Frame
29

at tunnel eyes could include jet grouting,
permeation or soil fracture grouting, deep soil
mixing, short term dewatering, and ground freezing.
Selection of the recommended type of ground
improvement is beyond the scope of this paper.
Table 1. Ground Treatment Design Considerations
Parameter Design Consideration
Strength The treated ground mass needs to be
strong enough and dimensioned to
(1) distribute lateral loads around the
soft or unreinforced eye back to
the shoring system and (2) reduce
active pressures to an acceptable
level at the soft eye.
Hydraulic
conductivity
The excavation and tunnel eye are
not expected to be watertight;
instead some small amount of
leakage should be tolerable. The
overall dimensions and hydraulic
conductivity of the treated mass
must minimize pressures within the
annulus between lining and ground
and groundwater seepage velocities
and gradients through the ground
treatment zone to less than what is
critical for piping erosion. Potential
leakage paths are illustrated on
Figure 5.

Table 2 summarizes preliminary design
recommendations for the treated soil block as well
as the basis of the recommendations and anticipated
final design considerations. Appropriate project-
specific analyses should be conducted to confirm
these preliminary design recommendations. These
recommendations can also be checked against case
history data such as those summarized by Richards
et al. [2].

Figure 5. Potential Leakage Paths at the Tunnel Eye
30

Table 2. Preliminary Design Considerations for Ground Treatment of Twin Tunnels
Parameter Preliminary Design Criteria Basis of Preliminary Design
Criteria
Final Design
Criteria
Length (measured
along tunnel)
Minimum: the length of the
TBM shield plus three complete
grouted rings.
Minimum length should also
extend until there is at least 1
diameter of cover above the
crown of the tunnel and a
minimum pillar width of
diameter between bores, if
supported by pillar interaction
analyses. Additional pillar width
must be provided in the absence
of pillar stability analyses.
Additional cover may be
required if settlement sensitive
structures are present along the
alignment.

Allow space for annulus
grouting behind TBM.
Provide adequate strength to
minimize interaction with the
second tunnel (if applicable).
Minimize surface settlement.
Final TBM and ring
dimensions and ring
constructability
considerations.
Tunnelling induced
settlement analyses.
Width (measured
perpendicular to
tunnel axis)
The tunnel diameter plus one-
third of the diameter on each
side, with a minimum of 1.5 m
(5 feet) on each side, outside of
the excavated diameter.

Strength and hydraulic
conductivity. (Strength is
required to make the annulus
self-supporting and facilitate
grouting).
Seepage and
strength analyses.
Depth (from top to
bottom of soil
cement zone)
Top: 3 m (10 feet) above top of
TBM.
Bottom: 3 m (10 feet) below
bottom of TBM.

Same as width plus additional
thickness required to limit
bending.
Same as width.
Target Strength 0.7 MPa or 100 psi (80%
achievement criteria)

Hoop stress around annulus. External loading
conditions.
Maximum
Strength
3.4 to 10.3 MPa
(500 to 1,500 psi)
Below strength where disc
cutters would be required.
Drag bit capacity.
Maximum
Hydraulic
Conductivity
1 x 10
-6
cm/s, or lower
Higher hydraulic conductivity
values can be used if it can be
demonstrated that piping/erosion
will not occur
Estimated to be equal to or less
than existing hydraulic
conductivities in granular
materials.
Exit gradients and
seepage velocity.

5.2. Mechanical Seals
The maximum pressure resisted by the mechanical
seal may be related to the injection pressure used to
displace water with grout in the annulus between the
soil and the outside of the tunnel lining. An annulus
grouting pressure (gage pressure) slightly above
hydrostatic pressure may be adopted during design for
this type of construction. Mechanical seals are
available to resist a range of expected pressures. A
seal with the appropriate pressure rating therefore
needs to be selected. An alternative would be to grout
the initial rings without pressure, and to increase the
pressure in increments until the intended injection
pressure can be developed after a few rings are
grouted. This reduces the need for a mechanical seal,
but does not necessarily eliminate the need for one.
31

This technique also
requires favorable
ground conditions or
ground treatment.
The seals need
to be durable enough
to slide along the
TBM as it
enters/exits the
opening. Lubricants
can be added to
reduce the
coefficient of
friction between the
TBM and the seal.
Seals must also have
the ability to seal
initially against the
TBM and later
against the outside
of the tunnel lining, which has a diameter that is
smaller than the TBM. An inflatable seal arrangement
may be used to seal around the outside of the segment
lining after the TBM passes. Alternatively, adjustable
steel plates like those shown in Figure 6 have been
used to support the gasket around the outside of the
segment lining after
the TBM passes.
Mechanical seals with
elastomeric lip-type
seals are less effective
for an exit seal if the
lip is pointing in an
unfavorable direction
away from the face of
the machine. Specific
design considerations
are noted in Table 3.
Depending on
anticipated ground and
groundwater
conditions, seals could
be used either alone,
or with another tunnel
eye construction
method. Additionally,
they can be provided with inflatable emergency seals
(Figure 7) that can be activated in the event of a
failure of the primary seal.

Figure 7. Phoenix TBM Seal Arrangement Showing Inflatable Seal [4]

Inflatable Seal
Figure 6. TBM Seal Arrangement Showing Adjustable Plates to
Support Entry Seal
32

A disadvantage of mechanical seals is that they
can be breached upon entry or exit of the TBM.
Failure can occur for a variety of reasons, including
debris severing the seal or failure of the connection
between the seal and the excavation shoring system.
One problem with seals encountered on past projects
is that accidental reverse movement of the TBM can
fold the gasket over and cause a leak. Seal failures are
reviewed in Tunnels and Tunnelling International
Magazine [5]. Regardless of the seal materials, failure
can still occur in the bolts that attach the seal to the
shoring system, which can have the same
consequences as a failure of the seal itself. Because
of the consequences of seal failure, and the need to
stabilize the soil within the eye, seals are frequently
used in conjunction with other ground and
groundwater control methods.
One approach to providing soil stability is to use a
blind panel method. In cases where steel sheet piles
are used for shoring, these piles can be pulled or cut in
the vicinity of the blind panel and the integrity of the
excavation can be maintained by the strength of the
blind panel itself.
Despite the possibility of seal failure, it is
important to note that seals are frequently used
successfully. Advantages of seals include ease of
construction and relatively low cost. Additionally,
they can be readily designed to accommodate virtually
any diameter of machine and lining arrangement.
Seals or parts of seals may also be reused at multiple
tunnel eye locations.
Table 3. Design Considerations for Mechanical Seals
Parameter Design Consideration
Materials The seal needs to be flexible enough to seal around the TBM and tunnel segments, which
have different diameters.
Abrasion The seal needs to be robust enough to withstand sliding as the TBM proceeds through it.
Compatible lubricants are typically applied along the seal to reduce the coefficient of
friction between the seal and the TBM.
Degradation In areas where hydrocarbon contamination is expected, the seal should be made with
materials such as neoprene that do not break down in the presence of hydrocarbons.
Excavation Size The size of the seal and frame can vary greatly, but a minimum of 90 centimeters (3 feet)
all around should be allowed from the shoring system into the tunnel heading.
Grout Pressure The seal needs to be designed to prevent inflows of annulus grout. Grout pressures can
be limited immediately adjacent to the opening to reduce the likelihood of grout flowing
into the adjacent excavation.
Hydrostatic Pressure In addition to grout pressure, seals are designed to withstand design hydrostatic
pressures, plus an appropriate factor to account for variations in hydrostatic pressure that
may occur.
Ground Conditions Hydraulic conductivity of soils surrounding the eye affects the seal design. In clays and
other soils with low hydraulic conductivity, seals would not necessarily need to resist the
full hydrostatic head, while in sands and gravels, the full hydrostatic head could be
expected within a relatively short amount of time.
Tolerance Tolerances on the diameter and position of the seal need to be established considering
that TBMs have a tendency to dive as they leave the launching frame and arrive slightly
off-target upon reception.

6 CASE HISTORIES
6.1. A River Crossing in Northern California
Recently on a tunnel project in northern California,
problems occurred at tunnel eyes at a number of the
tunnel crossings in sands beneath the water table.
At one location involving a 4.6 m (15 ft) EPBM and
a 10.7 m (35 ft) deep receiving shaft, flowing
ground entered the shaft. The source of the flowing
ground appeared to be a gap between the sheetpile
shoring system and the jet grout block outside the
shaft. The quality of the jet grout block was not in
question, and the presence of the jet grout helped
maintain circularity of the precast concrete tunnel
lining such that no significant distortion of the
lining occurred. The shaft excavation sequence
involved driving sheets, jet grouting, shaft
excavation in the wet, underwater concrete
placement of invert acting as a strut, and
dewatering. This sequence could conceivably
create a gap between the shoring and the jet grout, if
the sheetpile embedment and strutting is insufficient
to limit deflection. The critical load case for
limiting deflection is at the point when excavation is
at its maximum depth, but before invert concrete
placement. When the EPBM reached the shaft, it
hit the shoring hard enough for a gauge cutter to rip
33

through the sheet piles, creating an alternative
mechanism to create a gap between the sheetpiles
and soil cement.
A gap of less than 50 mm (2-inches) between
shoring and soil cement block was identified on a
nearby project during probing through the sheetpiles
before tunnelling [6]. Secondary grouting using
microfine cement was reportedly used to seal the
gap.
Similar gaps between soil cement blocks and
shaft shoring have been identified in the literature,
which suggests the problem of a gap between jet
grout and shaft shoring is not limited to sheet pile
shoring applications [1].
Given the difficulty experienced at the interface
between soil cement and the shaft shoring system,
there must be an emphasis in the design of an
effective seal to prevent the formation of the gap.
Once this design issue is solved, the length and
overall dimensions of the soil cement block must be
addressed. As a result, although soil cement block
dimensions are addressed above, other factors may
also come into play in sizing the soil cement block.

6.2. SR75/282 Transportation Corridor Project
The SR75/282 Transportation Corridor Project
involves construction of twin 2.3 km (1.4 mile) long
road tunnels in Coronado, California. The project is
in the preliminary design stage and both a cut-and-
cover and bored tunnel alternative are currently
under consideration.
Ground conditions along the bored tunnel
alignment consist primarily of Bay Point deposits
composed of interbedded medium dense to dense
clean to silty/clayey sands and medium stiff clays
and silts.
Preliminary design of ground improvement at
the tunnel eyes focused on the criteria presented on
Table 2 above. A pillar width study was also
performed to ensure that a separation distance of
diameter would not result in adverse impact to the
tunnel lining performance. The pillar width at the
eye itself had to be minimized as this minimized
excavation at the launch and reception portals,
which was the primary cost driver at this location.
Depth of the eye was also minimized for the same
reasons. After the tunnel eye depth and pillar width
were established, a preliminary ground
improvement layout was developed (Figure 8). As
the project is located in an urban area, and the
launch and reception excavations are adjacent to
private property and significant State highways, a
mechanical seal has been included in the design in
addition to ground improvement. The mechanical
seal will include an inflatable emergency seal
should additional protection be required. The
inflatable seal will provide an additional level of
redundancy against failure of the eye, and will also
protect against piping between the shoring system
and the portals.
Figure 8. Preliminary Draft Plan and Profile for Tunnel Eye Ground Improvement on the SR75/282
Project. The length of the ground improvement zone is approximately 92 meters (300 feet).
34

7 CONCLUSIONS
Tunnel eyes have a high incidence of unacceptable
performance during construction, that has led to
excessive settlement, sinkholes and tunnel flooding
(Figure 9).
A common failure mechanism at tunnel eyes is
piping and flowing soil entering the shaft or portal
along a narrow gap between the shoring and the soil
cement block. It is often difficult to fill this gap
with ordinary cement grout because it fills with
loose soil as it is formed and will not take grout.
Successful tunnel eye design and construction
requires a multi-disciplinary approach that involves
a number of design criteria and factors. Contractual
barriers can get in the way of producing a tunnel
eye system, where each component is optimized so
that the desired sealing performance is achieved.
The recommendation is to have a professional
engineer work for the tunnel contractor or
subcontractor to prepare a design package for the
tunnel eye that addresses each component of the
system in addition to verification testing and
remediation action plans.
This paper establishes some basic design
concepts that can be used to ensure the design is fit
for purpose, at least where external hydrostatic
pressures are in a typical range up to about 3 bar.
Seal technology is evolving rapidly, and a project is
currently under consideration in Arizona where seal
manufacturers have indicated that seals
withstanding pressures up to 12 bar can be
manufactured [7]. Procedures for preliminary
sizing and specifying ground improvement and
design considerations for mechanical seals have
been discussed. Selection of the eye concept is site
specific, and final design requires a variety of
assessments. Specific considerations include the
following:

1. Risk assessments: specific risks associated
with each of the alternative eye designs
should be analyzed in detail.
2. Construction cost estimates: these need to
be generated and balanced against the risk
analysis to develop the design. Cost
estimates will also need to be included in
the project wide estimate.
3. Schedule analysis: Schedule estimates will
need to be made and analyzed in the context
of the overall project schedule.
4. Constructability considerations.
Constructability issues will need to be
considered. Potential issues include
availability of space on the surface above
the eye, the potential for special
environmental considerations, and
disturbance of existing facilities.

Figure 9. Flooding of a Tunnel Shaft (and surrounding area) due to
Problems at the Tunnel Eye
35

8 ACKNOWLEDGEMENTS
The authors wish to thank Gary Kramer, J ohn
Hawley, Stuart Warren, and Andrew Hindmarch,
for technical review, Napoleon Purification for
figure preparation, and numerous other Hatch Mott
MacDonald personnel for their help, suggestions,
and review of this work.
REFERENCES
1. Campo, D.W., D.P. Richards, and M. Coudry. 1997.
A Review of the Grouting on Line 2 of the Cairo
Metro. In RETC Proceedings, Las Vegas, NV, pp.
22-25 J une 1997, eds J.E. Carlson and T.H. Budd.
Society of Mining, Metallurgy and Exploration.
2. Richards, D.P., A.J . Burchell, D.W. Campo, and P.
Raymond. 1996. Review of Break-in and Break-out
concepts for Tunnel Boring Shields in Saturated Soft
Ground. In Proceedings ITA Conference North
American Tunneling, Washington, D.C., 21-24 April
1996, ed. L. Ozdemir, pp. 451-460, Rotterdam: A.A.
Balkema.
3. J apan Society of Civil Engineers (J SCE), 2000.
J apanese Standard for Shield Tunneling, Third
Edition.
4. Phoenix North America, 2006. Personal
Communication.
5. Tunnels and Tunneling International. 2003. TBM
recovery after shaft seal failures, December, pp. 24-
26.
6. Doig, P.J . and A. Page. 2006. Microtunneling on the
Lower Northwest Interceptor, Sacramento,
California. In Proceedings of the North American
Tunneling 2006 Conference, Chicago, IL., 10-15
J une 2006, ed. L. Ozdemir, pp.437-443, London:
Taylor & Francis Group
7. J urich, D.M. and P.M. Kandaris. 2006. New water
intake for Navajo Generating Station at Page,
Arizona. In Proceedings of the North American
Tunneling 2006 Conference, Chicago, IL., 10-15
J une 2006, ed. L. Ozdemir, pp.229-235, London:
Taylor & Francis Group

36

Estimation of water inflow into a tunnel during construction
Ali Ameli
Geo Engineering Ltd., Vancouver, BC, Canada, Email: aameli@GeoEngineeringLtd.com

ABSTRACT: This case study discusses a reliable, quick solution to evaluate the water ingress into the excavations, while
the construction is in progress. The matter is reviewed for a 15 m diameter tunnel in sedimentary rocks in the vicinity of
Karun River through enlarging the pilot tunnel. The water inflow inside the tunnel when excavated in full was estimated
based on Darcy Formula and the amount of water discharge observed in the pilot. The approach did not require any further
drilling or field permeability testing and eliminated the need for grouting. The predicted results were in harmony with the
actual discharge level, which occurred in practice and relieved the management of facing unpredicted situations.

1 INTRODUCTION

The ingress of water into the tunnel could create
laborious conditions and unwanted consequences in
cost and project time schedule. It may increase the
construction period by 50% or even cause
abandonment of project construction operations.
Although, many contractors are accustomed to
dealing with difficult situations, this issue is often one
of the main concerns for clients and their consultants,
who are wary about project cost during the course of
construction. Different parties in a contract are
therefore concerned with potential of water inflow
into the excavations and subsequent preventive
methods, in particular when the data from site
investigation proved to be not accurate or insufficient.
The case study refers to estimation of water
inflow into a tunnel during construction by an
intuitive approach, which imposed no additional
liabilities to the client in terms of field permeability
testing, grouting and other environmental issues. The
phrase - environmental issues - was used, as drilling
and grouting would impose environmental impacts in
terms of machinery depreciations, employing
cementitious materials and associated delays to the
project. The paper explains provisions and foresight
that can be made by the engineers to minimise the
cost in terms of avoiding sophisticated analysis or
remedial actions, which would not be required if the
available data were used intuitively. This case study
highlights a problem-solving example for a decision
making process during construction.

2 THE PROJECT

The tunnel, constructed in 1990s, was part of
diversion facilities for a 2000MW hydroelectric
project in southern Iran. The 15m-diameter tunnel
with a length of about 700m was located in the close
vicinity of Karun River, having a long-term average
flow of 300 m
3
/sec. The tunnel ran almost parallel to
the river about 10 m to 30 m from the riverbank with
the crown elevation equal to the average river water
elevation.
For the construction of this tunnel, after the
excavation of deep portals, a 6x8 m
2
pilot tunnel was
excavated at the upper part of the main tunnel section.
The pilot was then widened and benched to establish
the full circular section. Access to the tunnel was
limited by geological and topographical conditions.
Water inflow to the pilot tunnel was collected
through a number of sumps and ditches and directed
to the outlet portal area where it was pumped away.
The sumps were spaced at about 150m intervals. The
location of the pumping station was constant
throughout the excavation of the diversion tunnel.

3 GEOLOGY

The project is located in Oligocene and Miocene
sedimentary rock. The diversion tunnel is situated on
the limb of an anticline in the right bank of the river.
The bedding strikes were almost perpendicular to the
river direction. The tunnel mainly passed through
thickly bedded limestone and marly limestone with a
short distance near the inlet in marly limestone and
37

marlstone of low overall rock mass strength and
modulus. The tunnel alignment was predominantly
sub-vertical to the steeply dipping bedding of
limestone and marly limestone. Generally, the
discontinuities and solution-enlarged joints were
filled with silt, clay and silty clay. The New
Australian Tunnelling principles were used for tunnel
excavation and support. The support system
comprised of wire mesh reinforced shotcrete, 4-6m
rock bolts/prestressed anchors and lattice girder
beams.

Figure 1. Tunnel cross section, dewatering and excavation stages.

4 GROUNDWATER INFLOW

Water inflow into the pilot tunnel increased after each
blasting cycle. It then decreased slightly to a steady
state level until the next round of pilot tunnel
excavation. The water inflow reached 100-120 l/sec,
when the whole length of the pilot tunnel was
completed. The estimation of water inflow into the
full section of tunnel was a matter of concern towards
the end of the drilling of the pilot tunnel and in the
threshold of excavating the full section. As well as
pumping, grouting had also been prescribed in the
specification to reduce water inflow. Unexpected
ingress of water during excavation of the bottom half
of the tunnel could compromise the construction
operations. This was due to difficulties in handling of
the groundwater at a greater depth with reference to
the access road elevations servicing the portal area.
Grouting along the tunnel to create a curtain against
water seepage into the tunnel became an attractive
option. Further, the contractor needed an estimation
of water inflow in order to provide necessary
equipment and to choose suitable methodologies for
bench excavations. Two approaches were thought of
for the prediction of water inflow as:
4.1. Trial Excavation
This was a contractor-oriented suggestion in which a
20-30m length of the tunnel would be excavated in
full section and water inflow into the excavation
estimated. Total water inflow into the tunnel would
thus be extrapolated on a linear basis. There were two
problems associated with this approach; i)
extrapolation of water inflow rate into the whole
length of the tunnel would be reliable if the
permeabilities along the tunnel axis were identical
(something difficult to assume), and ii) water seepage
from the end walls perpendicular to the tunnel axis
would introduce further complexities in the estimation
process. Therefore this approach was not considered
any further.
4.2. Analytical Approach
Darcys formula was selected to calculate the water
seeping into the excavation.
Bench II
Bench I
El. 670m
River
Outlet
Portal
Original Ground Level
Pilot
Slash
38

Q = k H (N
f
/ N
d
) A (1)

where:
Q = volume of water inflow
k = rock mass permeability
H = total head
N
f
= number of flow channels
N
d
= number of equipotentials
A = area

The following options were considered for
determination of permeability of the rock mass:

a) Execution of water tests by drilling a number of
boreholes along the tunnel axis and estimation of
permeability (Lugeon) values.
Site investigation data and field observations
during excavation showed that water mainly
seeped through the discontinuities in the rock mass
and the permeability of rock material was
insignificant. Data from the water test results in
the rock mass with sub-vertical beddings or
jointing was further questionable, as test sections
in vertical drill holes would be unlikely to intersect
steeply-dipping discontinuities.

b) The preferred approach was evaluating the rock
mass permeability through back analysis by
employing the amount of water discharged from
the pilot tunnel during excavation. The approach
required measuring the following field data:

o River water El., m
o Water discharge in pilot tunnel, m
3
/sec
o Advance in pilot tunnel, m
o (Advance in excavation of slashes, m)

River water elevations, which reflected the wet
and dry season conditions, were read from a
hydrometer station on a bridge pier close to the
tunnel outlet. Pumping rates, weirs or rate of
discharge from pipelines were used to quantify the
water discharge.

5 EVALUATION

5.1. Permeability
Permeability is the most uncertain and complicated
parameter that geotechnical engineers would face in
their analysis. In the above solution attempts were
not made to quantify the permeability values of the
rock mass around the tunnel. But the effect of this
parameter was reflected on the volume of water,
which seeped into the pilot tunnel. The pilot tunnel
with 48 m
2
cross section can be regarded as a large in-
situ test hole, which offers more reliable results than
the Lugeon tests carried out traditionally in a borehole
of ~ 5000 times smaller cross section. (Typical
borehole diameter used for Lugeon test is 100 mm).
For the accumulated amount of water discharge
into the whole length of the pilot tunnel, the variation
of rock mass permeability along the tunnel axis would
be reflected automatically. No karstic holes were
found during pilot tunnel excavation. This gave
relative confidence that no karst would be
encountered in full tunnel excavation. Ground
conditions, examination of pilot tunnel geological
mapping profiles, and aerial observations all indicated
that the pattern of rock mass discontinuities across
and around the tunnel diameter was uniform.
Therefore, applying an average permeability in
directions vertical to the tunnel axis would be a
reasonable assumption.
5.2. Water Inflow
Seepage analysis was carried out based on flow net
diagrams for different stages of tunnel excavation i.e.
after removal of the slashes and for excavation of the
benches in two stages.
The calculations, which include the tunnel section
dimensions and the area of the tunnel exposed to the
flow channels, are summarised in the form of the
following equations:

Q = Q
p
(23.76 + 1.76H)/(6 + H) = q
p
C

(2)
Q = Q
(p + s)
(12.94+0.96H)/(6 + H) = q
(p + s)
D (3)
Q
(p + s)
= Q
p
(C/D)

(4)

where:
Q = water inflow into the full section of tunnel, m
3
/sec
Q
p
= water inflow into the pilot tunnel, m
3
/sec
Q
(p + s)
= water inflow into tunnel after removing slashes
and before benching
H = head of water, m (Tunnel crown elevation was
chosen as a datum)
C = (23.76 + 1.76H) / (6 + H)
D = (12.94 + 0.96H)/ (6 + H)

The coefficients C and D for the probable ranges
of river water elevations can be obtained from Figure
2.
39

Figure 2. Values of coefficients C and D.
Eq. 2. was used to forecast expected water
inflow into the full section of the tunnel based on
inflow into the pilot.
Similarly, Eq. 3. may be used, if the amount of
water discharge into the excavation after removal of
the slashes is measured. Though, measurements were
not continued as the water inflow after removal of the
slashes was in line with the predicted values and no
critical conditions were expected at this stage.
The contract specifications had suggested to
design the pumps, collection ditches and pipe lines for
a water discharge of about 750 l/sec. This was based
on the scattered permeability tests data conducted
from the original ground surface around the tunnel
area during site investigation.
When the plot tunnel was dug, the measured water
inflow was 100-120 l/sec. Evaluations based on
Table 1. Summary of water inflow data

Discharged Water Volume, l/sec

Pilot
Tunnel
Q
p

Pilot and
Slashes
Q
(p + s)

Full
Section
Q
Contract
Specification
NA NA 750
Predicted Values
from
Eqs. 2, 3, 4.
NA 190 - 225 400 - 480
Actual Values
Measured
100 - 120 NA 400 - 500

Darcys formula and data obtained during
construction indicated that at average river water
elevation, the amount of water seeping into the whole
tunnel cross section would be ~ 4 or 2 times that of
the pilot tunnel, q-
p
, or pilot and slashes, q
(p + s)
,
respectively. The actual flows and the predicted
values, when available, are recorded in Table 1. The
values reflect the seepage for the whole length of the
tunnel after a steady state was attained.
Table 1 also shows that the permeability values
obtained from the water tests in the design stage
overestimated the actual permeability values by a
factor of 50%.
Inspectors confirmed that during the excavation
of the full section the amount of discharged water was
400 - 500 l/sec. No unexpected water ingress occurred
and only some casual grouting and sealing of springs
were reported.

6 CONCLUSIONS

The amount of water inflow when the full section of
tunnel was dug was in the same range predicted from
water inflow in the pilot tunnel. Darcys Formula
proved to be valid for estimation of water inflow in
rock mass around the tunnel.
Occurrence of insufficient or erroneous design
parameters is common in underground excavations.
Simple analysis from geotechnical data obtained
during construction and engineering intuitions could
assist the management in making cost-saving
decisions in operation planning.

664
666
668
670
672
674
676
1 2 3 4 5 6 7 8 9 10
Value
E
l
e
v
a
t
i
o
n
,

m C D
40
1 INTRODUCTION
Much of the soil deformation and surface settlement
experienced during tunnel excavation can be
associated with the inward movement of the tunnel
face. The need to control the stability of the face, and
thereby reduce the ground deformation and surface
settlement, has increased in the last decade due to
expansion of urban centers and the necessity to drive
tunnels in difficult geological conditions, often below
existing structures. The idea of controlling the face
stability by applying a pressure is not new. In 1973,
J.V. Bartlett provided a complex description and
demonstration of a slurry boring tunnel machine.
Later, several papers appeared in various journals
dealing with the tunnel face failure mechanism and
prediction of the minimum required supporting
pressure. (For example: Atkinson and Pott, 1977;
Davis et al. 1980; Mohkan and Wong, 1988; Kimura
and Mair; Leca and Dormieux, 1990; Lee et all 2003).
In 1991 and 1994, Chambon and Cort reported on
centrifuge experiments modeling tunnel face stability
supported by pressure. They have reported low
pressures at the failure occurrence.
This paper reports on centrifuge experiments on
tunnel face stability supported by air pressure and
subsequent stability analyses using the FE method,
and upper and lower bound solution. The main
objective of the performed work was to obtain the
magnitude of the supporting pressure at the moment
of the face collapse, to observe the development of
progressive failure mechanism, and to identify the
zone affected by the face collapse. This information is
of particular importance for tunneling in urban areas,
where the ground deformation may affect the existing
structures above the tunnel.
2 CENTRIFUGE MODEL DESCRIPTION
A miniature model of tunnel was manufactured for the
testing purposes in a geotechnical centrifuge, which
consisted of three main parts: aluminum lining,
junction box and a rubber bag. Due to the symmetry
of the modelled problem, only half of the tunnel was
modeled. The model configuration allowed placing
the tunnel model against the window of the centrifuge
container and video observation of the soil movement
in the vicinity of the tunnel face. An aluminum half-
tube 2 mm in thickness and 50 mm in diameter was
used to model a concrete tunnel lining. An
approximately 0.2 mm thick rubber bag formed into a
Static centrifuge experiments on stability of pressure-supported tunnel
faces in sandy ground
Pavol Oblozinsky, Ph.D.
Geotechnical Engineer, Golder Associates Ltd., Kamloops, BC, Canada
Jiro Kuwano, Ph.D
Professor, Saitama University, J apan
Qiang Li, Ph.D.
Software Engineer, Allrightsoft Co., Tokyo, J apan
ABSTRACT: The paper reports on a series of centrifuge experiments on stability of working tunnel faces in sandy ground
supported by means of pressure. A miniature model of a tunnel was manufactured for the centrifuge testing purposes,
consisting of an aluminum lining and a thin rubber bag holding the compressed air to support the tunnel face. During the
experiment, the supporting pressure was reduced gradually until that the collapse of the face occurred. The main purpose of
the experiments was to obtain the minimum supporting pressure, the soil deformation around the tunnel face and the
surface settlement. The results of the experiments were compared with the results of FEM stability analysis and the upper
and lower bound solution. The centrifuge experiments indicated that a surprisingly low pressure is sufficient to support the
soil skeleton at the tunnel face.
Sand
LDT
Air pressure
Tunnel
Face
Lining PPT

Figure 1. Centrifuge experiment set-up
41
half-cylinder shape was placed inside the aluminum
tube. At the tunnel face, the rubber bag was in direct
contact with the soil. At the other end, the rubber bag
was attached to a junction box connected to a pressure
supply located outside the centrifuge container, and
controlled from the centrifuge operating room. The
set-up of centrifuge experiments is shown
schematically in Figure 1.
A centrifuge gravity field equal to 80G was
selected for the testing program. In the 80G field, the
tunnel model corresponded to a tunnel of 4 meters
outside diameter in prototype scale. The tests were
performed for different thicknesses of the cover, with
the cover to diameter ratio, C/D equal to 2, 4 and 6,
which in prototype scale is equivalent to 8, 16 and 24
meters of cover.
The tunnel model was instrumented with a pore
pressure transducer (PPT) located in the junction box
at the tunnel invert level to measure the pressure
inside the rubber bag. A laser displacement transducer
(LDT) was used to measure the surface settlement
above the tunnel face. Small targets were placed
around and above the tunnel face into a grid to
visualize soil movements. There was no displacement
transducer inside the tunnel to measure the inward
movement at the face, as the model was originally
design for a liquid supporting medium and a suitable
transducer was not available.
3 SOIL MODEL GROUND
Dry Toyura silica sand was used to prepare the ground
model, with the following properties: The particle
density was 2.647 g/cm
3
, with grain size D
50
=0.19mm,
D
30
=0.16mm and D
10
=0.14mm.
The sand was slowly rained into the centrifuge
container to make the model ground. The relative
density of the rained sand varied from 79% to 83%
with an average density of 1.56g/cm
3
.

4 CENTRIFUGE TESTING
As the centrifuge was being accelerated to the
designed working centrifuge gravity field of 80G, the
pressure inside the tunnel was simultaneously
increased to keep at rest pressure conditions at the
tunnel face calculated based on K
0
=0.5. After
achieving the specified gravity field, the pressure in
the tunnel was reduced. The moment of the face
failure was determined based on the soil movement at
the face recorded by the video camera. At collapse,
the rubber bag was squeezed by the sand moving into
the tunnel, which generated an impulse of increased
pressure recorded by the PPT, clearly indicating the
moment of the face collapse.
The obtained values of the supporting pressure at
the face collapse are summarized in Table 1. The
pressures at the face collapse are of low magnitude,
raging from 2.6 to 5.6kPa, indicating that low
supporting pressure would be sufficient to provide
temporary stability of the tunnel face in dry conditions
during the tunnel driving process. Compressed air was
used to support the tunnel face, and it was assumed
that the supporting air pressure was uniformly
distributed over the tunnel face area. If the tunnel is
driven below ground water table, a pressure balancing
the hydrostatic pressure should be added to the
obtained minimum supporting pressure.
Table 1. Supporting pressure,
s
, at the face collapse
C/D : 2 4 6
S
[kPa] : 3.6 2.6 5.3
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 1 2 3 4 5 6 7
C/D
D
presented tests
after Chombon and Corte

Figure 2. Normalized supporting pressure at failure
against C/D ratio
D
Sand deposit
Failed wedge
C/D=2
C/D=4
C/D=6
Envelope of
progressive failure
Depression
cone
3
x
D

Figure 3. Sketch of the progressive failure envelope
42
A comparison with the centrifuge experiments
reported by Chambon and Cort (1991, 1994) is
provided in Figure 2 in dimensionless form. The
agreement between the experiments is acceptably
good considering that a different approach of failure
determination was used.
The failure mechanism is demonstrated by a
sequence of photographs, Photo 1 to 3, showing the
C/D=2 case. Photo 1 shows the targets before the face
collapse, followed by Photo 2 taken just one frame
subsequent to Photo 1, which is approximately 1/30 of
second. Photo 2 shows the loosening of the soil at the
upper portion of the face and above the tunnel crown.
Photo 3 was taken several frames later and shows the
progressively developed failure propagating upward
in a chimney like shape. In the C/D=2 case, the
progressive failure reached the ground surface and
settlement of order of 1.2 metres (in prototype scale)
was measured. In the C/D=4 case, the surface
settlement of 0.06 meters was measured, and the
propagation of the progressive failure was terminated
at a distance of approximately 3D above the tunnel. In
the C/D=6 case, negligible surface settlement was
measured and the progressive failure was terminated
again a distance of approximately 3D above the tunnel
crown. It appears that the vertical extent of the
progressive failure above the tunnel can be estimated
as 3D, where D is the tunnel diameter. The envelope
of the extent of the progressive is schematically
shown in Figure 3.
5 NUMERICAL ANALYSES
A FE code originally developed at Gunma University
(Li. Q., 2000) for slope stability and liquefaction
analysis was used to evaluate the tunnel face stability
and calculate the safety factor. The safety factor was
calculated using the shear strength reduction
technique (well described elsewhere; Zenkiewich
1975, Matsui and Sun 1992).
A 3D isoparametric 20-nodded element was used
to build up the FE mesh. The C/D=2 mesh contained
1230 elements and 6108 nodes. Additional elements
were added to this base mesh to achieve higher cover
for C/D=4 and 6 cases, keeping the mesh geometry
around and above the tunnel face unchanged in all
series of calculations. The FE mesh and the
appropriate boundary conditions are schematically
shown in Figure 4.
Elasto-perfectly-plastic constitutive law was
adopted to model the soil behavior, in which the
failure was governed by the Mohr-Coulomb equation
and the plastic potential by the Drucker-Prager
equation. The tunnel lining was modeled as an elastic
material with the material properties of the aluminum

Photo 1 Condition before collapse

Photo 2 Loosening of sand ahead of the face
triggering of the collapse

Photo3 Propagation of failure

Figure 4. FE mesh with boundary conditions, case
C/D=2
43
scaled from the model to prototype scale. The material
properties are indicated in Table 2.

The results of the FE analyses are shown in Figure
5 as a plot of the safety factor against the uniformly
distributed pressure applied on the face. It can be
noticed that the calculated safety factors are almost
identical up to the supporting pressure of 15 kPa. This
indicates that the cover above the tunnel crown has a
little affect on the required face supporting pressure
for dry soil, which is in agreement with the centrifuge
experiment results. The centrifuge experiment results
suggest that the minimum supporting pressure varies
from 2.6 to 5.3 kPa compared to 4kPa predicted by the
FE stability analysis.
Adopting the minimum pressure of 5.3 kPa from
the centrifuge experiment as the calibration pressure
for the FEM, a minimum safety factor of 1.2 should
be required by the FE analysis for a design of
supporting pressure. It is believed that the FE
approach can be applied to the effective stress analysis
to take account for the ground water.
Figure 6 shows the distribution of shear strain
around the tunnel face and can be compared with
Photo 2 showing the moment of the face collapse. The
concentration of high shear strain indicates the failed
zone at the face, which triggers the progressively
developed failure. Upper and lower bound solution
denoted as MII, suggested by Leca and Dormieux
(1990), was used to estimate the minimum supporting
pressure applied to the tunnel face for a failure
mechanism. The results are summarized in Table 3.
As noticed from the table, the minimum tunnel face
supporting pressure obtained from the centrifuge
experiments, FE analyses and upper bound solution
are in acceptably good agreement. There is an
indication that the required minimum supporting
pressure varies little with the C/D ratio. The Lower
bound solution gives much higher minimum
supporting pressure as a function of C/D ration.
6 DISCUSSION AND CONCLUSIONS
The centrifuge experiments indicate that a low
uniformly distributed pressure applied on the working
tunnel face can provide a stable support condition
during construction. The results are surprising;
however, they are in a good agreement with the FE
analyses, upper bound solution and similar centrifuge
experiments performed previously. It is suggested that
the main reason for measuring such low pressure in
the centrifuge experiments was a perfect transfer of
the supporting pressure to the surrounding soil by
means of the rubber bag, which may not be achieved
in practical engineering. Centrifuge experiments as
Table 2. Material properties for FEM analyses
Soil Lining
E [kNm
-2
] 20 000 6.9x10
7

0.3 0.34
[kNm
-3
] 15.6 27
c [kNm
-2
] 0.2 -
[deg] 42 -
Table 3. Supporting pressure,
s

C/D : 2 4 6
S
[kPa]
Centrifuge
3.6 2.6 5.3
S
[kPa]
FEM
4.0 4.0 4.0
S
[kPa]
Upper Bound
3.7 3.7 3.7
S
[kPa]
Lower Bound
37 61 86
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 5 10 15 20 25 30
Supporting pressure, m [kPa]
S
a
f
e
t
y

F
a
c
t
o
r
,

S
F
C/D=2
C/D=4
C/D=6

Figure 5. Safety factor against the uniform load
applied on the face

Figure 6. Shear strain distribution

s
=4kPa S
F
=1.03
44
well as the numerical analyses suggest that the C/D
ratio has a little influence on the required minimum
supporting pressure for moderately deep tunnels in
dry soil condition. It was observed that failure
occurred suddenly without extensive soil deformation
at the tunnel face prior to the collapse, and failure
extent progressively upward in a chimney like shape.
The extent of the caved-in soil was observed to a
distance of 3D above the tunnel crown in C/D=4 and
6 cases. The comparison of the FE analysis and the
centrifuge experiments suggests that a factor of no
less than 1.2 should be applied to the FE results to
design.

ACKNOWLEDGMENT
The centrifuge testing program was performed at
Tokyo Institute of Technology under the grant P02702
provided by the Japanese Society for Promotion of
Science (JSPS) to the first author. The financial
support is herein gratefully acknowledged.

REFERENCES
1. Atkinson, J.H, Pott, D.M., 1977, Stability of a shallow
circular tunnel in cohesionless soil, Gotechnique, 27,
No. 2, pp.: 203~215
2. Bartlett J.V., Biggart A.R., Triggs R.L.: 1973, The
bentonite tunneling machine, Proc. Civ. Engrs, 54,
605-624
3. Chambon, P., Cort, J-F., Garnier, J., Knig, D., 1991,
Face stability of shallow tunnels in granular soils,
Proc. Centrifuge 91, Balkema, Rotterdam, pp.:
99~105
4. Chambon, P., Cort, J-F., 1994, Shallow tunnels in
cohesionless soil: Stability of tunnel face, J ournal of
Geotechnical Engineering, Vol. 120, No. 7,
pp.:1148~1165
5. Davis, E.H., Gunn, M.J., Mair, R.J., Seneviratne H.N.,
1980, The stability of shallow tunnels and underground
openings in cohesive material, Gotechnique, 30, No.
4, pp.: 397~416
6. Kimura, T., Mair, R.J., 1981, Centrifugal testing of
model tunnels in soft clay, pp.: 319~322
7. Leca, E., Dormieux, L. 1990, Upper and lower bound
solutions for the face stability of shallow circular
tunnels in frictional material, Gotechnique, 40, No. 4,
pp.: 581~606
8. Li, Q., 2000, Development of a new finite element
program for liquefaction analysis of soils and its
Application to seismic behavior of embankments on
sandy ground, PhD Thesis, Gunma University
9. Matsui, T., San, K.-C., 1992, Finite element slope
stability analysis by shear strength reduction technique,
Soils and Foundation, Vol., 32, No. 1, pp.: 59~70
10. Mohkan, M., Wong, Y.W., 1988, Three dimensional
stability analysis of the tunnel face under fluid
pressure, Proc of Numerical Methods in
Geomechanics, Innsbruck, pp.:2271~2278
11. Zienkiewicz, O. C., Humpheson, C., Lweis, R. W.,
1975, Associated and non-associated viso-plasticity
and plasticity in soil mechanics, Gotechnique, 25,
No. 4, 671-689.

45
1 INTRODUCTION
Precast concrete tunnel linings (PCTL) have been
extensively used in seismically active locations
around the world, and they continue to be specified
for bored tunnel construction in seismic areas. PCTL
have been subjected to strong shaking during major
seismic events and have performed well. The
observed performance is plausible given that tunnels
are not subjected to the effects of inertia of their mass
and they are constrained by the ground around them.
This study summarizes the observed performance of
PCTL systems where such systems have been
subjected to significant seismic ground motions, and
offers some insight into seismic design procedures
based on observed performance.
2 PRECAST CONCRETE TUNNEL LINING
SYSTEMS
When utilizing a closed face Tunnel Boring Machine
(TBM) in soft ground beneath the water table, it is
necessary to construct a lining concurrently with TBM
advance to stabilize the ground and limit groundwater
inflow into the tunnel excavation. The lining is
erected in the tail shield area of the TBM and is
utilized as a reaction block by the TBM to shove
forward, as shown in Figure 1. PCTL with gaskets, as
shown in Figure 2, are frequently used as a one-pass
lining. When using a one-pass system, the lining
installed in the TBM tail shield acts as both initial and
final support. Assembled segments in a tunnel are
shown in Figure 3.

Fig. 1. Cut-away of TBM showing ring-build area in tail
shield.
Precast concrete tunnel linings were used for the
first time in Great Britain in the 1930s, in North
America in the mid 1960s, and entered widespread
Seismic Performance of Precast Concrete Tunnel Linings
David Young
Senior Tunnel Engineer, Hatch Mott MacDonald, Pleasanton, CA, USA
Anil Dean
Tunnel Engineer, Hatch Mott MacDonald, Pleasanton, CA, USA

ABSTRACT: Over the next decade and beyond, numerous tunnels will be constructed in seismic areas to facilitate the movement
of people, goods, and services. The purpose of this paper is to discuss the performance of precast concrete tunnel linings (PCTL) in
seismic events. PCTL systems are comprised of a number of segments, which are assembled in the tail-shield of the tunnel boring
machine, and are typically used in soft-ground tunnels. When PCTL are used, the final tunnel lining system can be either one-pass
or two-pass. One-pass PCTL have become the most favoured lining type for closed-face soft ground transit tunnels, due to the
overall value added to the project. However, the seismic behaviour of PCTL is not well understood, and the volume of published
work on the subject is relatively thin. As a result, tunnels that could be lined with a one-pass PCTL system are sometimes
constructed with two-pass systems at significantly higher cost. This paper, which builds on the authors previous research, is
intended to bridge the gap between the theory of PCTL design and the performance of the lining during earthquake shaking. Case
histories of PCTL performance in the Northridge (1994), Kobe (1995), Athens (1999), and Hualien (2002) earthquakes are
available. Lessons learned from these tunnels are presented. PCTL have inherent advantages over other tunnel linings when
subjected to earthquake shaking. These advantages are discussed and related to the case histories noted above. PCTL performance
in these case histories is related to previous tunnel seismic performance studies and reports by others.

46
use in the 1980s. By 1990, the use of precast concrete
segments replaced steel and cast iron segments as the
most widely used lining for tunnels in soft ground.
The main reason for use of precast concrete instead of
cast iron lining was the cost. The cost differential was
initially small, but precast concrete lining segments
are now roughly one-third to one-fourth of the cost of
comparable cast iron or steel segments. The cost
advantage of precast concrete lining segments has
increased with the high shove forces needed for
increasingly larger closed face TBMs.
Despite these early introductions, the adoption of
PCTL systems for tunnels in North America has been
slower than in Europe, J apan and elsewhere. It is only
in the past two decades that PCTL systems have
gained widespread use throughout North America
including seismically active areas such as Seattle,
Portland, San Diego and Los Angeles. One-pass
PCTL systems are currently being designed for
highway and rail transit applications in the highly
seismic San Francisco Bay, San Diego, and Los
Angeles areas.

Fig. 2. Precast concrete tunnel lining ready for transport to
the TBM.

Fig. 3. Erected precast concrete tunnel lining.
3 SEISMIC DESIGN CONCEPTS
PCTL can be subjected to transient deformation due to
seismic wave passage and permanent deformation
associated with ground failure due to liquefaction-
induced lateral spreading, fault displacement or slope
instability.
Seismic wave energy is transmitted in different
wave types and possibly in different directions
concurrently. Seismic deformation is often simplified
for design purposes to a single waveform propagating
in a single direction. Simplified wave effects can be
examined in transverse and longitudinal orientations.
Transverse wave effects produce ovalling in the
lining, while snaking and longitudinal extension and
contraction can result from various seismic
waveforms and incident angles. Various references
cover this material [1-3]. Ovalling and snaking styles
of deformation are illustrated in Figure 4. Two
primary considerations are:
i. Lining flexibility - The lining could simply
deform with the ground or it could be relatively
rigid in comparison to the ground such that it
resists transient deformation; and
ii. The size of the tunnel - If it is large enough,
wave passage could create dynamic
amplification of stresses that would need to be
considered in the lining design analyses.

Fig. 4. Primary types of tunnel deformation.
The first consideration, lining flexibility, can be
thought of in terms of the flexibility ratio [2], which is
47
a tool for assessing whether or not the structure
deforms with the ground.
The flexibility ratio is a measure of the flexural
stiffness of the tunneling medium relative to that of
the PCTL under a state of pure shear and measures the
liners resistance to ovalling. The flexibility ratio, F, is
given by the following equation [3]:

) 1 ( EI 6
)R - (1 E
3 2
m
m
F
+
= (1)

where: E
m
is the modulus of elasticity of the medium
E is the modulus of elasticity of the liner
R is the radius of the liner

m
is the Poissons ratio of the medium
is the Poissons ratio of the liner
I is the moment of inertia of the liner

Flexibility numbers above 20 indicate a flexible
structure that deforms as if it were perforated ground
according to free field ground deformation, in which
case the diametric distortion of the lining is related to
the free field shear strain and Poissons ratio of the
medium according to the following relationship [2]:

) 1 ( 2
d
max m
D
=
(2)
where:
D
d
is the diametric distortion of the liner

max
is the free field shear strain

Sensitivity analyses of the input parameters show
that a flexibility number for a PCTL application of 20
or below is quite rare, meaning PCTL systems can for
practical purposes be considered flexible structures
that simply deform with the ground.
The second consideration involving the size of the
tunnel relates to dynamic amplification of stresses
associated with a seismic wave impinging on a tunnel
has been recognized by some authors [3-5] as a
potentially important design consideration. It can be
shown that these effects are heightened for a given
earthquake as size of tunnel cross-section and
frequency of ground motion increase and as shear
wave velocity of the tunnelling medium decreases.
Typical tunnels lined with PCTL are not large enough
for these effects to become important.

3.1. Deformation Based Design
A deformation based approach is used for design
of flexible PCTL for seismic loads, either transient or
permanent deformations. In order to accommodate
the deformation imposed by the ground on the tunnel
lining, it is useful for the lining to be segmented. This
allows the seismic deformation to be absorbed at the
joints rather than in the concrete. This also means that
PCTL is ideal for seismic environments.
The flexibility of the individual segments
themselves is achieved through steel reinforcing bars
or steel fibre reinforcing and flexibility of the overall
structural system is achieved through joints between
the segments that accommodate deformations with
little or no damage. In addition, joint contact areas
contain packing materials that cushion segments and
avoid high contact stresses. Inter-segment connecting
devices such as dowels, bolts, and guide-rods (largely
used for convenience during construction) maintain
segment alignment and provide a level of redundancy
with respect to stability of the segment positions. It is
possible to design these connecting devices to
accommodate the seismic deformation.
4 PCTL PERFORMANCE
An extensive literature search was conducted to
evaluate the use of PCTL in seismic areas and PCTL
performance in seismic events. Details of the research
are provided by Dean, Young and Kramer [6], and
some of the findings are discussed below.
To illustrate the frequency of PCTL use in
earthquake prone areas, a selection of transit tunnels
was reviewed. Some significant examples of PCTL
use in seismic zones are listed on Table 1, showing
the widespread use of PCTL worldwide. PCTL
systems continue to be specified for bored tunnel
construction in seismic areas.
Approximate values for the design ground
acceleration are noted on Table 1, based on available
probabilistic seismic hazard acceleration (PSHA)
maps using a 10% probability of being exceeded in
50-years, which corresponds to a recurrence interval
of 475-years. Probabilistic seismic hazard evaluation
techniques were developed to assess hazards based on
possible earthquake magnitudes, source-site distances,
and probabilistic analyses. PSHA maps are commonly
available through the Global Seismic Hazard
Assessment Program (GSHAP), the California
Geological Survey (CGS), the United States
Geological Survey (USGS), and other agencies. The
accelerations in Table 1 from PSHA analyses are
ground surface accelerations and some attenuation of
these motions can be expected at tunnel depth.
Four case histories are summarized on Table 2
below. There is very limited documentation of PCTL
performance in seismic events, despite the fact that
many more than four tunnels with PCTL are likely to
48
have experienced ground motions comparable to those
shown on Table 2.

4.1. LA Metro
Strong shaking occurred in the LA Metro tunnels
resulting from the Northridge Earthquake of 1994,
which struck at 4:31AM local time and had a moment
magnitude of 6.7. The earthquake caused 57 fatalities
and over 5000 injuries. Property damage was
estimated at $20 billion, the costliest natural disaster
in the history of the United States at that time. The
epicenter of the earthquake was located in the San
Fernando Valley, approximately 32 km (20 miles)
northwest of Los Angeles, according to the
Earthquake Engineering Research Institute (EERI)
[7].
Earthquake design criteria for the LA Metro Red
Line was developed for both Operating Design
Earthquake (ODE) and Maximum Design Earthquake
(MDE) events. The shaking during the Northridge
Earthquake corresponded to the ODE event, and there
was no damage to the PCTL that was under
construction in the Los Angeles Basin. PCTL
tunnelling in the San Fernando Valley did not start
until after the earthquake. The Red Line was designed
to have a two-pass lining, due to the presence of
hazardous gasses along the alignment. Since the
tunnel segment of the Red Line was under
construction, the internal lining had not been installed
at the time of the earthquake. Therefore, earthquake
loading of the tunnel lining was borne solely by the
PCTL.

Table 1. Recent Use of PCTL for Transit Tunnels in Seismically Active Areas
Project Type Location Diameter Year
PSHA Ground
Acceleration
*

Reference
Taipei Metro Metro Taiwan
6 m (19.7 ft)
OD
1987-1996 >0.48g
World Tunnelling
[8]
Athens Metro Metro Greece
8.5 m
(27.9 ft)
1991-1999 0.24g
T&T International
[9]
Barcelona Metro Metro Spain
10.9 m
(35.8 ft)
2002 0.16g
T&T International
[10]
Turin Metro Metro Italy
6.9 m
(22.6 ft)
2003 0.16g
T&T International
[11]
Tehran Metro Metro Iran
6.0 m
(19.8 ft)
1997 >0.48g Lovat News Release
Shiraz Metro Metro Iran
6.0 m
(19.7 ft)
2004 0.41g T&TC [12]
Metropolitano de
Lisboa
Metro
Lisbon,
Portugal
9.8 m
(32.1 ft) OD
Under
construction
0.16g Lovat News Release
Passante
Ferroviario
High-
speed
railway
Bologna, Italy
9.4 m
(30.8 ft)
2001 0.41g Lovat News Release
Istanbul Metro
Extension
Metro
Istanbul,
Turkey
6.5 m
(21.3 ft) OD
Under
construction
>0.48g Lovat News Release
Marmaray
Bosphorus
Crossing
Rail
Bosphorus,
Turkey
8.0 m
(26.2 ft) OD
Under
construction
>0.48g Lovat News Release
Ankara Metro
Sogutozu-Kizilay
Metro
Ankara,
Turkey
5.9 m
(19.3 ft) OD
Under
construction
Metro Caracas
Linea 3
Metro
Caracas,
Venezuela

5.8m
(19 ft) OD

Under
construction
LA Metro Gold
Line
Metro
Los Angeles,
CA
5.8m
(19 ft) OD
Under
construction
Design levels:
0.41and 0.79g
Result from
mapping: 0.70g
Law/Crandall [13]
* This column indicates results of Probabilistic Seismic Hazard Assessment (PSHA) mapping from a variety of sources for
an event with a 10% probability of being exceeded in 50 years, unless otherwise noted.

49
Table 2. PCTL Earthquake Performance Case Histories
Tunnel Earthquake Earthquake
Date
Surface Horizontal
Acceleration Near
Tunnel (g)
Post Event
Tunnel
Condition
Source(s)
LA Metro Northridge 1/17/94 0.4 No Damage Monsees and Elioff
[14] and EERI [7]
Isobe Dori Shield
Tunnel
Kobe 1/17/95 0.5* Some Spalling at
Segment J oints
J SCE [15]
Athens Metro Athens 9/7/99 0.25 No Damage EERI [16]
Taipei Metro Hualien 3/31/02 0.20 No Damage Taipei Times [17]
* A range of accelerations is available - see discussion below
The maximum shaking measured within the LA
Metro Red Line tunnels was 0.27g for this event [7,
14]. The 0.4g surface acceleration estimate indicated
on Table 2 is derived from a shake map, which is
included as Figure 5. The acceleration within the
tunnels was likely lower than 0.4g due to attenuation
that occurs with depth.

Fig. 5. Northridge Earthquake shake map with Metro Red
Line Tunnel locations, base map source: EERI [7].

4.2. Isobe Dori Shield Tunnel
The Isobe Dori Shield Tunnel was under construction
at the time of the Kobe Earthquake of J anuary 17,
1995. The Kobe Earthquake had a moment magnitude
of 6.9, and struck directly beneath the densely
populated City of Kobe. The earthquake caused over
5500 fatalities and over 26,000 injuries. Property
damage was estimated at US$200 billion.
The Isobe Dori Shield Tunnel is owned by Kansai
Electric Power Company. The 931-metre long, 4.95 m
diameter tunnel had been completed and lined with
the PCTL at the time of the earthquake. The design
included construction of a concrete invert, although
construction of the invert had not started at the time of
the earthquake. Cracks 0.2mm wide were observed in
the shafts, but the PCTL remained intact, with only
some spalling observed after the event. The damage
report [15] stated the following:
there was some spalling in the grooves in the
segments between segment rings; otherwise the
structure remained undamaged
Figure 6 shows the post-earthquake condition of
the tunnel during the inspection. Accelerations in the
vicinity of the tunnel vary. Ground conditions at the
tunnel consisted of very dense gravel with
groundwater 1.8 to 3.0 metres (6 to 10 feet) below the
ground surface [15]. The tunnel depth ranged from
approximately 21 to 28 metres (69 to 92 feet) below
the ground surface with a riser section from
approximately 17 to 21 metres (56 to 69 feet) below
the ground surface. Shake maps indicate a horizontal
PGA of approximately 0.5g, so this is the value
reported on Table 2. This value is likely to be
conservative, as most nearby strong motion stations
reported readings well in excess of this value.

Fig. 6. Post earthquake inspection of the Isobe Dori Shield
Tunnel, Kobe, J apan, source: [15].

4.3. Athens Metro
No damage was reported to any of the Athens Metro
tunnels in the Athens Earthquake of September 7,
50
1999 according to EERI [16]. The earthquake, which
had a moment magnitude of 5.9, struck at 2:56PM
local time with an epicentre located in the northwest
portion of Athens. The earthquake caused 147
fatalities and hundreds of injuries [18]. Data from 14
strong motion recordings between 10 and 20 km from
the epicentre indicate peak horizontal ground
accelerations ranging from 0.04 to 0.35g (a recording
of 0.53g is regarded to be anomalous due to site
specific ground motion amplification). EERI noted
that peak ground accelerations might have exceeded
0.5g in the epicentral area. The Athens Metro tunnels
are known to be constructed using PCTL and limited
information regarding the segment design for the
tunnels can be found in World Tunnelling [19].
Several of the strong motion recording stations
were located in Metro stations. The largest horizontal
level of shaking of 0.25g, measured within the Sepolia
Station, is reported on Table 2. Accelerations in other
stations close to the epicentre may have been higher,
but were not recorded during the event.

4.4. Taipei Metro
The Hualien earthquake had a Richter magnitude of
6.8 and struck at 2:52 pm local time on March 31,
2002. The epicentre was located near the town of
Hualien in eastern Taiwan. The earthquake caused 4
fatalities and 200 injuries. While the epicentre was
180 km (110 miles) east of the capital city of Taipei,
significant ground accelerations were recorded in
Taipei, and the Taipei Metro system experienced
ground motions resulting from the earthquake.
The Taipei Times [17] reported that the Metro
was stopped for inspection at the time of the
earthquake. Metro Service was restored by 7:30 pm
on the same day. Peak horizontal ground acceleration
was measured from a seismograph, Station TAP022,
in downtown Taipei operated by the Taiwan Institute
of Earth Sciences [20]. The peak ground acceleration
from this station is reported on Table 2 as 0.2g. The
Metro tunnels are known to be constructed using
PCTL, and limited information regarding the segment
design for the Taipei Metro is included in World
Tunnelling [8].
5 LESSONS LEARNED
For purposes of evaluating tunnel performance,
the intensity of ground shaking is typically quantified
by peak ground acceleration (PGA), peak ground
velocity, peak ground displacement, and strong
motion duration. For initial assessments of potential
seismic effects, PGA at the ground surface is usually
used as an index of the shaking intensity, because
acceleration is the parameter usually recorded and
most readily estimated at the ground surface. A
summary of the performance of different types of
bored tunnels that experience shaking are plotted
against peak horizontal ground acceleration by Power
et al. [21]. While these authors did not specifically
identify performance of PCTL, PCTL performance is
grouped under reinforced concrete linings. The
seismic performances of PCTL summarized on Table
2 have been used to supplement the Power et al. data
in Figure 7. This data, although very limited, indicates
the generalized threshold between none to slight
damage in terms of PGA may be in the vicinity of
0.5g for PCTL, which is consistent with what Power
et al. presented for reinforced concrete. It is also
apparent that reinforced concrete lined tunnels
perform better than unreinforced concrete lined
tunnels. There is insufficient information from the
performance records to show whether PCTL perform
better than a reinforced, cast-in-place concrete lining.
However, from a design perspective, the PCTL
system offers more opportunity, compared to cast-in-
place, to accommodate seismic deformations by
detailing the joint surfaces and joint connectors.

Fig. 7. Empirical damage state data for bored tunnels with
PCTL damage state from this study added.

It is curious to note that the slight damage
observed in the Isobe Dori shield tunnel occurred on
51
the joint between rings, which are the circumferential
joints shown in Figure 8. This observation suggests
that snaking deformation influenced the spalling,
whereas ovalling deformation is often considered to
be a more critical seismic loading condition.
6 CONCLUSION
In general, tunnels perform well in earthquakes
because they are constrained by the ground around
them and are not subjected to inertial effects like
above-ground structures. PCTL systems perform
particularly well in earthquakes because of their
circular, largely symmetrical shape and their
flexibility relative to the ground surrounding them.
The research conducted for this study has confirmed
that PCTL perform well when subjected to seismic
ground motions, based on four case histories and a
lack of reported damage to PCTL in many more
tunnels that have been subjected to similar shaking.
Only one instance of slight damage to a PCTL was
found in an extensive search of the performance of
PCTL during seismic events. This incident was
limited to slight spalling in the Isobe Dori tunnel in
Kobe, J apan, which did not result in more than
cosmetic damage to the PCTL system.

Fig. 8. Isometric view of a PCTL.

Hundreds of tunnels have been built using PCTL
systems in seismically active areas around the world.
The widespread use of PCTL together with inherent
advantages in load carrying capacity, flexibility, cost
effectiveness and seismic performance, make PCTL
the ideal lining type for large single-pass bored tunnel
projects that are designed to withstand strong seismic
shaking.
PCTL are robust systems when subjected to
seismic loading for several reasons. PCTL are
subjected to significant loads during construction due
to thrust loads exerted by the TBM. Higher than
specified concrete strengths are often used at the
precast plant to reduce curing time. The result is that
the compressive strength and thickness of the lining
are typically more than sufficient to resist the static
and seismic loads imposed on them. Tensile strain
induced by seismic waves passing through the ground
can be distributed to the joints between the PCTL
segments, thereby minimizing tensile stresses within
the segments themselves. J oint connectors can be
designed to accommodate the deformation.
Where a seismic analysis of a PCTL is warranted,
the anticipated ground displacement at tunnel depth is
most important to design. Ground displacement can
be higher within 15 to 20 km of the epicenter where
near source ground motion effects must be considered.
Free field peak shear strain, seismic deformation
modulus and Poissons ratio of the soil are
geotechnical parameters needed for a basic analysis.
Since tunnels are frequently built using PCTL in
seismic areas, development of standard post-
earthquake tunnel reconnaissance guidelines would
greatly facilitate future tunnel earthquake design.
Additional seismological data relating to ground
motions experienced at tunnel depth could be used to
further refine the state of the practice for seismic
PCTL design in soft ground. Similar research could
be conducted for tunnels in rock as well.
ACKNOWLEDGEMENTS
The authors wish to thank Gary Kramer for his
contributions to the initial study of the performance of
precast concrete tunnel linings and to the rest of the
tunnel segment design team for the Silicon Valley
Rapid Transit project, for which the initial study was
undertaken. Special thanks also goes to Hatch Mott
MacDonald for their assistance with publishing this
paper, as well as Herrenknecht, Kawasaki, and Lovat
for information regarding projects built with their
TBMs.
REFERENCES
1. Owen, G.N., and R.E. Scholl. 1981. Earthquake
engineering of large underground structures. Report no.
FHWA / RD-80 / 195. NTIS Document number PB81-
247918. Federal Highway Administration and National
Science Foundation.
2. Hashash, Y.M.A., J .J. Hook, B. Schmidt, and J .I.-C.
Yao. 2001. Seismic design and analysis of
underground structures. Tunnelling and Underground
Space Technology, 16, 247-293.
3. Merritt, J .L., J .E. Monsees, and A.J . Hendron. 1985.
Seismic Design of Underground Structures. In
Conference, Volume 1, Society for Mining,
Metallurgy, and Exploration, Inc (SME).
52
4. Law, H.K, and I.P. Lam. 2003. Evaluation of Seismic
Performance for Tunnel Retrofit Project. J ournal of
Geotechnical and Environmental Engineering, Vol.
129, Number 7, Paper number 575, J uly.
5. Wang, Y.N., B. Hughes, H. Caspe, and M. Amini.
2006. Devils Slide Tunnels Caltrans 1
st
New
Highway tunnels in 50 years. In North American
Tunneling 2006 Conference, Chicago, Taylor and
Francis/Balkema, the Netherlands.
6. Dean, A., D.J . Young, and G.E. Kramer. 2006. The
Use and Performance of Precast Concrete Tunnel
Linings in Seismic Areas. International Association of
Engineering Geologists, 2006 Proceedings, paper
number 679.
7. Earthquake Engineering Research Institute (EERI).
1995. Northridge Earthquake Reconnaissance Report,
Volumes 1 and 2. Earthquake Spectra, EERI,
Supplement C to vol. 11.
8. World Tunnelling. 1994. Taipei Metro. P. 430,
December.
9. Tunnels and Tunnelling International. 1996.
Tunnelling problems delay Athens metro, p.16.
November.
10. Tunnels and Tunnelling International. 2002.
Barcelonas new backbone runs deep, (no page
number). March.
11. Tunnels and Tunnelling International. 2003. Multiple
TBM action on Turins metro, p.16. April.
12. Tunnelling and Trenchless Construction. 2004. Shiraz
Metro boring about to commence, p.29. August.
13. Law/Crandall. 2003. Geotechnical and Environmental
Investigation, Eastside LRT Project Underground
Segment, Volume I of III, First Street, from Clarence
Street to Lorena Street, Los Angeles, California.
October 22, 2002, Revised May 2003.
14. Monsees, J .E. and A. Elioff. 1999. Evolution of Design
LA Metro underground structures. Geo-Engineering
for Underground Facilities, Proceedings of the Third
National Conference. ASCE Geotechnical Special
Publication No. 90.
15. J apan Society of Civil Engineers (J SCE). 1995.
Preliminary Report on the Great Hanshin Earthquake,
J anuary 17, 1995.
16. Earthquake Engineering Research Institute (EERI).
1999. The Athens, Greece Earthquake of September 7,
1999. EERI Special Earthquake Report November
1999.
17. Taipei Times. 2002. Temblor rattles Taipeis MRT,
cripples cellphones.
18. Anastasiadis, A. N., M. Demosthenous, C.H.
Karakostas, N. Klimis , B. Lekidis, B. Margaris. 1999.
The Athens (Greece) Earthquake of September 7,
1999: Preliminary Report on Strong Motion Data and
Structural Response. Institute of Engineering
Seismology and Earthquake Engineering (ITSAK).
19. World Tunnelling. 1994. Athens Metro. P. 271,
September.
20. Taiwan Institute of Earth Sciences. 2002. The March
31, 2002, Taiwan Earthquake.
http://www.earth.sinica.edu.tw/~smdmc/recent/2002/2
0020331.htm
21. Power, M.S., D. Rosidi, J . Kaneshiro, S.D. Gilstrap,
and S-J Chiou. 1998. Draft Report, Summary of
Evaluation of Procedures for the Seismic Design of
Tunnels. September 9 Rev 1, Technical Report
MCEER-98-XXXX, FHWA Contract No. DTFH61-
92-C-00112, Task 112-D-5.3, prepared for the
Multidisciplinary Center for Earthquake Engineering
Research, State University of New York at Buffalo.

53
1 INTRODUCTION
Since the mid 1980s, parts of the City of Calgary,
including the community of South Calgary and
adjacent communities, have experienced street and
basement flooding from severe rainstorms. In
response to this flooding, the City of Calgary has
constructed several improvements to the stormwater
collection and conveyance system. The Glencoe
Storm Sewer Upgrade Project Phase B/C is another
step in addressing flooding problems in Calgary.
The City of Calgary, Wastewater Division engaged
the City of Edmonton, Asset Management and Public
Works, Drainage Services to undertake a storm sewer
upgrade in the South Calgary Community. The storm
sewer upgrade consists of a deep tunnel along 27
Avenue SW from 15 Street SW to west of 20 Street
SW.
The Glencoe tunnel is 935m in length, 2920 mm in
diameter and has a 0.9% slope. The depth of the
tunnel varies from 16m to 42m. This tunnel will
provide temporary storage of stormwater runoff
during major storm events to reduce street flooding in
the vicinity of the project.

Figure 1: Glencoe Tunnel Cross-Section
The Glencoe Storm Sewer Upgrade is funded
through ICAP Infrastructure Canada Alberta
Program. ICAP is a co-operative funding program
jointly shared between the Federal, Provincial and
Municipal governments. The City of Calgary has a
total of 19 stormwater improvement projects within
ICAP. Glencoe Phase B/C is the last of the
wastewater ICAP projects to be constructed. The
estimated cost for the tunnel portion is $7.8 million
within the total project cost estimate of $11.5 million.
The ICAP funding eligibility requires a completion
date of no later than March 31st, 2006.
Tunnelling for Success, Case Study: Glencoe Tunnel in Calgary
AL-Battaineh Hussien T.
Ph.D. Candidate, Department Civil and Environmental Engineering, University of Alberta, Edmonton, AB,
Canada
S. AbouRizk
Professor, Department Civil and Environmental Engineering, University of Alberta, Edmonton, AB, Canada
Siri Fernando
Director Design and Construction, Drainage Services, City of Edmonton, Alberta, Canada
Frank Policicchio
General Supervisor Tunnel Design and Construction, Drainage Services, City of Edmonton, Alberta, Canada
J ames Tan
ProgramManager Expansion, Design and Construction, Drainage Services, City of Edmonton, Alberta,
Canada
ABSTRACT: The Glencoe Tunnel project in Calgary, Alberta represents a marvelous example of collaboration between
the City of Edmonton and the City of Calgary. This is the first time the Design & Construction Section of Drainage
Services, City of Edmonton has undertaken a tunnel project outside the City of Edmonton. The proposed storm tunnel is
2920mm in diameter along 27 Ave SW starting from 15th St SW to 20th St SW with a total length of 935m. The depth of
the tunnel varies from 16m at the working shaft on 15 Street SW to 42m at the retrieval shaft on 20 Street SW. The tunnel
will reduce surface flooding by providing temporary storage of stormwater runoff during major storm events.
Given that this is the first time the City of Edmonton Tunnelling Team has ever worked outside the city, it is critical to the
success of the project that the planning phase is thoroughly developed and carefully executed. The planning phase includes
scope definition, contract set up, cost estimate, project team assembly, equipment and material procurements, risk analysis,
constructability review, geotechnical investigation, Safety and Environmental Construction Operations Plan (ECO Plan)
development, and scheduling and productivity simulations. The challenges presented in this project are the unfamiliarity
with the local conditions in terms of geology, local contractors and suppliers, the City of Calgarys business processes and
requirements, the tunnelling crews welfare and the delivery of the necessary technical and equipment supports from
Edmonton. This paper will present the results of simulation modeling to predict tunnelling productivity to meet the target
completion date.
54
The project started with the relocation of utilities
at the shaft locations in late March and early April of
2005. Construction of the tunnel working shaft got
underway in mid April of 2005.
As shown in Figure 1 the tunnel will be going
through conditions which can be categorized into
three zones: (1) 300 m of soft clay till section, (2)
100m of mixed face section at the transition zone
between clay till and bedrock, and (3) 530 m of hard
bedrock section. Tunnelling Boring Machine (TBM)
can generally be classified as either a soft ground or
hard ground machine. The available M126 TBM
owned by the City of Edmonton used in this project is
a soft ground machine. This presents a risk of not
being able to tunnel through the hard bedrock or with
a very low penetration rate.
2 CONSTRUCTABILITY ISSUES
During the planning phase of the project the following
key constructability issues were identified. Mitigation
measures were developed for each of the issues.
1. Working shaft configuration and construction
schedule sequence,
2. Unfamiliarity with the geological condition,
3. Concrete liner segment production and quality
control,
4. Remote logistic of the project in terms of
technical and equipment support and
unfamiliarity with the local conditions such as
suppliers and services,
5. Spoil removal, storage and sedimentation
control,
6. Welfare of the crews.
The following discussion will focus on the
process and results of construction schedule
simulations to provide sufficient information to the
project team to make decisions with respect to the
working shaft configuration, tunnelling sequence and
productivity in an attempt to meet the ICAP
completion deadline of March 31, 2006.
3 CONSTRUCTION SCHEDULE SIMULATION
The project completion date is the major driver for
this project. Considerable efforts were given to derive
a workable construction schedule to meet the
deadline. The parameters having direct impact on the
productivity and schedule are the size of undercut; the
productivity achievable through the three different
ground sections; the number and duration of shift; the
number of working days per week; the number of
trains and the number of TBM employed.

4 WORKING SHAFT CONFIGURATION
There are two shafts at the 15th Street location; a
working shaft and a pump station shaft. There is also
an existing 900mm water main feeding downtown
Calgary in between the two shafts. Four construction
configurations for the working shaft and the
connection between the two shafts were analyzed:
1- Option A: construct the working shaft and
pump station shaft. Connect the two shafts
with a large 30m tail tunnel by hand
tunnelling and then hand tunnel a short 6m
front undercut as shown in Figure 2.

Figure 2: Working Shaft configuration Option A
2- Option B: Construct a 30m front undercut,
connect the two shafts with a small hand
tunnel as shown in Figure 3.

Figure 3: Working Shaft configuration Option B
3- Option C: construct a short 12m front
undercut, begin tunnelling using rib and
lagging support for the first 30m (in this case
only one train can be used due to the size of
the undercut), connect the two shafts with a
small hand tunnel as shown in Figure 4.

Figure 4: Working Shaft configuration Option C
55
4- Option D: Similar to Option C but with the
use of concrete segmented liner instead of rib
and lagging support for the first 30m as
shown in Figure 5. (The installation of the
concrete liners eliminates the option of
enlarging the undercut at a later date if
necessary.)

Figure 5: Working Shaft configuration Option D
4.1. Construction Simulation
The above four construction configurations with
variable parameters were simulated under different
schemes and scenarios to determine the most viable
construction approach that would meet the target
completion date. Simulation models were developed
for each of the construction configuration using
Simphony.
Simphony is a special purpose simulation environment
developed by the construction group at the University
of Alberta. The model schematic is shown in Figure 6.

Figure 6. Glencoe Simulation Model
4.2. Penetration Rate Schemes
Two schemes of penetration rate are assumed:
Scheme #1: Variable TBM penetration rates in the
three tunnelling segments:
Section #1: 300m of clay till section assuming
good ground conditions.
Section #2: 100m of mixed face section
assuming TBM penetration rate 15% lower
than in Section #1.
Section #3: 530m of hard bedrock section
assuming a reduction in TBM penetration rate
by 25% lower than section #1.
Scheme #2: Variable TBM penetration rates in the
two tunnelling segments:
Section #1: 300m of clay till section assuming
good ground conditions.
Section #2: 630m of good soil condition in the
bedrock.

4.3. Construction Scenarios
Based on the above working shaft configurations (A,
B, C, and D) and penetration rate schemes
combination (Schema 1, 2), four scenarios (C1, C2,
D1 & D2) were analyzed. (Note: C1 is the
combination of working shaft configuration C and
scheme 1) Working shaft configuration Option A was
dropped because of the safety risk of working
underneath the 900mm water main for a long
duration. This factor was identified during risk
assessment. Option B was dropped as well because of
the required duration to construct the 30m front
undercut which would add extra time to the schedule.
Scenarios C1 and D1 assumed that the penetration
rate in the mixed face section is 1m per shift, and that
a second hard ground TBM will be required to be
employed at the 18 Street shaft location for tunnelling
in the hard bedrock section.

The results of the simulation are shown in Table 1 &
2. The simulations were carried out based on 10 hr
shift and 6 working days per week.
Table 1: Simulation Results
Option Production Rate
(m/10hr shift)
Total
Tunnelling
Duration
Total
Project
Duration
C.1 300 @ 8.1 m/shift
100 @ 0.9 m/shift
530 @ 7.7 m/shift
211 days 323 days
C.2 300 m @ 8.6 m/shift
630 m @ 11.1 m/shift
84 days 197 days
D.1 300 m @ 8.8 m/shift
100 m @ 0.9 m/shift
530 m @ 7.7 m/shift
207 days 290 days
D.2 300 m @ 8.6 m/shift
630 m @ 8.2 m/shift
115 days 198 days

Table 2: Simulation Results (cont.)
Option #of
Trains
#of TBM Duration
Days
Completion
Date
C.1 1 2 323 7/28/06
C.2 2 1 197 2/24/06
D.1 1 2 290 6/15/06
D.2 1 1 198 2/25/06

56
As shown in the analysis there is a high risk of not
being able to meet the target date due to geotechnical
uncertainty if we encountered a low productivity
through the mixed face and bedrock sections.

5 FINAL CONSTRUCTION APPROACH
Based on the results of the schedule analysis shown
above, the project team realized the need to acquire
more information regarding the bedrock condition and
to prepare a mitigation action plan if another TBM is
needed to complete the J ob, the project team decided
to proceed with the following approach:
1. Start tunnelling at 15
th
Street with working
shaft configuration C.
2. Construct shaft at 20
th
Street as soon as
possible. This will provide information on the
hardness of the bedrock and allow a better
estimation of penetration rate in the bedrock
section.
3. Based on the findings in step (2), two
scenarios of tunnelling can be taken: (1) If the
soft ground TBM can achieve reasonable
productivity then proceed with tunnelling
toward 20th street; (2) If the soft ground TBM
cannot achieve reasonable productivity (more
than 2m/day) then a second TBM (hard
ground machine) will be required. Tunnelling
will start from the 20
th
street toward 18
th

Street where an extraction shaft will be
constructed.
These approaches were further analyzed and the
results are shown in Table 3:
Table 3: Simulation Results
S
c
e
n
a
r
i
o

#

o
f

T
B
M

S
h
i
f
t

D
u
r
a
t
i
o
n

(
h
r
s
)

#

o
f

S
h
i
f
t
s
/
d
a
y

#

W
o
r
k
i
n
g

D
a
y
s

p
e
r

w
e
e
k

T
u
n
n
e
l
l
i
n
g

C
o
m
p
l
e
t
i
o
n

D
a
t
e

(
m
m
/
d
d
/
y
y
)

P
r
o
j
e
c
t

C
o
m
p
l
e
t
i
o
n

D
a
t
e

(
m
m
/
d
d
/
y
y
)

1 1 10 1 5 9/8/06 1/8/07
2 1 10 1 6 7/5/06 10/13/06
3 1 12 1 5 7/5/06 11/7/06
4 1 12 1 6 5/11/06 8/23/06
5 1 8 2 5 4/18/06 8/16/06
6 1 8 2 6 3/6/06 6/14/06
7 1 10 2 5 2/21/06 6/21/06
8 1 10 2 6 1/18/06 4/28/06
9 2 10 1 5 6/12/06 9/28/06
10 2 10 1 6 4/25/06 7/25/06
11 2 12 1 5 5/11/06 8/31/06
12 2 8 2 5 4/24/06 8/16/06
13 2 8 2 6 3/15/06 6/19/06
14 2 10 2 5 4/4/06 8/2/06
15 2 10 2 6 2/27/06 6/7/06
The target project completion date of March 31,
2006 is not achievable in any of the scenarios.
However, there are six scenarios where the
construction of the tunnel could be completed before
or close to March 31, 2006. They are Scenario 6, 7 &
8 with one TBM, 13, 14 &15 with two TBM.
These results concluded that the tunnelling
progress needs to be closely monitored in order to
provide timely advice to the City of Calgary on when
to request for an extension of the project completion
date from ICAP.
The project team also decided to modify the soft
ground TBM cutting head to handle the anticipated
hard bedrock more effectively.

6 PROJ ECT PROGRESS CONTROL
During project execution, daily productivity reporting
was undertaken. The actual productivity and
penetration rates were input into the simulation model
to project the completion date on a monthly basis.
Figure 7 shows the productivity analysis conducted on
J anuary 24, 2006, in this Figure the x-axis represent
the day and the y-axis the productivity (m/day). The
tunnel completion was projected to be March 14,
2006.

0
2
4
6
8
10
12
14
1 11 21 31 41 51 61 71 81 91 101 111 121 131
Productivity
Avg. Productivity

Figure 7 Tunnelling Productivity
7 CONCLUSION
The tunnel construction proceeded with working shaft
configuration C using one TBM and one train. The
actual production rate through the first 300m of clay
till section was 8m per shift vs. the simulated rate of
8.1m per shift, the actual rate through the 100m of
mixed face section was 4m per shift vs. the simulated
rate of 1m per shaft and the actual rate through the
530m of bedrock section was 5m per shift vs. the
simulated rate of 7.7m per shift. Overall the simulated
and the actual production rates are a good match.

57
Construction of the tunnel was completed on March
22, 2006 which correlated very well with the J an 24,
2006 projection of March 14, 2006. The City of
Calgary was advised to obtain an extension of the
March 31, 2006 deadline and was successful in
obtaining the extension. The project is presently
expected to be completed by the end of August.

This case study demonstrated the importance of the
pre-planning as a vital foundation for successful
project execution by applying risk analysis,
constructability review and productivity simulation.

Risk analyses identified risk factors and
developed mitigative measures to deal with those risk
factors. Constructability reviews brought field
experience early into the discussion in optimizing
design and construction approaches and method
resulted in direct positive impact on productivity, cost
and schedule.

The utilization of computer simulation model
such as Simphony, is a great tool to manage project
schedule by identifying and tracking key factors
which drive productivity such as the modeling of
working shaft configurations, geotechnical variation
and uncertainty, construction approach and resources
allocation.

REFERENCES
1. TBM Tunnel Simulation Template Users Guide
(2000), NSERC/Alberta Construction Industry
Research Chair.

2. AbouRizk, S., and Mohamed, Y. 2000. Simphony an
integrated environment for construction simulation. In
Proceedings of the 2000 Winter Simulation
Conference, ed. J . A. J oines, R. R. Barton, K. Kang,
and P. A. Fishwick, 1907-1914. San Diego, California:
Institute of Electrical and Electronics Engineers.

3. Hajjar, D., and AbouRizk, S., 1999. Simphony: an
environment for building special purpose construction
simulation tools. In Proceedings of the 1999 Winter
Simulation Conference, ed. P. A. Farrington, H. B.
Nembhard, D. T. Sturrock, and G. W. Evans, 998-
1006. Phoenix, Arizona: Institute of Electrical and
Electronics Engineers.

58
1 INTRODUCTION
Sound Transit (ST) is constructing the 26 km (16
mile) long light rail line from downtown Seattle
southwards to Sea-Tac Airport. By 2009 modern low-
floor light rail cars will run along the light rail route
with a maximum speed of 88 km/h (55 mph). The
light rail line will run at street level, on elevated
trackways as well as underground. The travel time
from Westlake, downtown Seattle, to the Airport will
be 36 minutes.
Contract C710 is the only mined tunnel section
and is located just south of the downtown area. In
addition to the construction of approximately 1.6 km
(one mile) long twin-bored running tunnels and a deep
mined station under Beacon Hill, the contract also
includes 800 m (one half mile) of aerial structure and
an elevated station at the eastern end. Obayashi
Corporation was awarded the construction contract in
J une 2004 at a contract price of US$280M.
The 1300 m (4,300 ft) long twin running tunnels
under Beacon Hill are being mined by an Earth
Pressure Balance Tunnel Boring Machine (EPB-
TBM) supplied by Mitsubishi Heavy Industries in
Kobe, J apan.
The deep mined Beacon Hill station is being built
from a one-square-block site located at the
intersection of Beacon Avenue South and McClellan
Street South. Future passengers will access the
Beacon Hill station by high-speed elevators that
transport them 49 m (160 ft) down to the underground
platforms.

SEM in Seattle Design and Construction of the C710 Beacon Hill
Station Tunnels
Michael Murray
Hatch Mott MacDonald
Stephen Redmond
Obayashi Corporation
Richard Sage
Sound Transit
Franz Langer
Dr. Sauer Corporation
Don Phelps
ABSTRACT: Contract C710 is currently under construction as part of Sound Transits Link Light Rail connecting
downtown Seattle with Sea-Tac Airport. The paper describes the design and construction of the deep mined station under
Beacon Hill using the Sequential Excavation Method (SEM), also known as the New Austrian Tunneling Method (NATM).

The 55 m (180 ft) deep binocular station includes platform, concourse, cross-passage and emergency ventilation tunnels
together with station egress and ventilation shafts.

The paper describes the geotechnical conditions anticipated and encountered, and the development of the design from the
preliminary design stage through the construction stage. Following the construction of a Test Shaft during the final design
stage, it was realized that the ground conditions would be difficult, so provision was made for further geotechnical
investigations and ground improvement from the surface during the construction stage.

The construction methods and design details are strongly influenced by the need to ensure safety during construction.
Excavation sequences include twin-sidewall and single-sidewall drifts. A range of pre-support measures and tool-box
items are made available and adopted as necessary.

Details are included on the rates of progress achieved in the safe and successful tunnel construction to date.

59
2 DESIGN
In October 2000, a joint venture of Hatch Mott
MacDonald and J acobs Civil Inc. (HMMJ ) was
awarded a contract for the final design of the Beacon
Hill Tunnels (D710) segment of the project. Dr. G.
Sauer Corporation (DSC) was awarded a sub-contract
by HMMJ for the design of the concourse cross adits
and platform tunnels.
The underground station layout at contract award
is shown in Fig.1 and consists of twin shafts and a
complex configuration of vehicle, pedestrian and
ventilation tunnels. The inverts of the platform
tunnels are 49 m (160 ft) below ground surface. The
platform tunnels are 116 m (380 ft) long and spaced
45 m (146 ft) apart, center to center.

Fig. 1. Station Layout at Contract Award.
The SEM was selected as a means of progressively
excavating and supporting the ground. The specified
sequences are designed to expose and stabilize the
ground in limited incremental widths and heights.
Standard support measures are used throughout and
these include specified round lengths, fiber-reinforced
flashcrete applied to newly exposed surfaces, lattice
girders and/or steel arches installed at predetermined
intervals, and reinforced shotcrete support ranging
from 20 cm to 30 cm (8 inches to 12 inches) in
thickness, depending on the final opening dimensions.
The final lining consists of fiber reinforced in-situ
concrete varying from 30 cm to 35 cm (12 to 14
inches) for the normal pedestrian access areas of the
station tunnels, platform tunnels and concourse cross
adits. For the remaining tunnels, fiber reinforced
shotcrete varying from 20 cm to 40 cm (8 to 15
inches) is specified. J unctions are reinforced with
steel reinforcement. A waterproofing system is
installed between the initial and final lining consisting
of a geotextile fleece and PVC membrane with an
injection system.

2.1. Geological Conditions
Seattle is located within the central portion of the
Puget Lowland, an elongated topographic and
structural depression bordered by the Cascade
Mountains on the east and the Olympic Mountains on
the west. The lowland is characterized by a series of
north-south trending ridges separated by deeply cut
ravines and broad valleys, the result of glacial
scouring and sub-glacial erosion. The area may have
been subjected to six or more major glaciations with
an ice thickness up to 915 m (3000 ft). The glacial
and interglacial soil units are typically of limited
lateral extent. A high degree of variation is evident
locally to the extent that some units cannot be reliably
correlated between adjacent borings.
Subsurface conditions were investigated in three
different phases associated with the Conceptual
Engineering (CE), Preliminary Engineering (PE), and
Final Engineering (FE) stages of design. A total of 73
investigation borings were drilled specifically for the
project during the period 1998 and 2003.
A simplified geologic model was developed for
the Beacon Hill Tunnel and included in a
Geotechnical Baseline Report (GBR). The model
groups the geologic units into six engineering classes
having similar physical and engineering properties as
follows:
Class 1 Loose to Dense Granular Deposits
Class 2 Soft to Very Stiff Clay and Silt
Class 3 Till and Till-Like Deposits
Class 4 Very Dense Sand and Gravel
Class 5 Very Dense Silt and Fine Sand
Class 6 Very Stiff to Hard Clay
The GBR was included as a contract document, and
was intended to assist bidders in evaluating the
requirements for excavating and supporting the
ground, and in preparing their bids. The baseline
conditions presented in the report are used by Sound
Transit to evaluate any differing site condition claims.
2.2. Test Shaft
In 2003, a 46 m (150 ft) deep Test Shaft was
constructed within the design footprint of the Beacon
Hill Station Main Shaft. The shaft was 5.5 m (18 ft)
in diameter from the ground surface to 32 m (105 ft)
below ground surface, and 1.8 m (6 ft) in diameter
from 32 m to 47 m (105 to 153 ft). The primary
objectives of the Test Shaft were to confirm the nature
of the ground and groundwater conditions during
construction, and provide an opportunity for bidders
to view the subsurface materials prior to bidding the
project. The Test Shaft indicated the extreme
60
variability of the soils and groundwater conditions
which had a profound impact on the shaft and tunnel
design. A report on the Test Shaft was included in the
contract documents in the Geotechnical Data Report.
2.3. Connector Tunnels VECP
The four 24 m (80 ft) long connector tunnels
positioned between the platform tunnels and the
running tunnels were sized to accommodate the large
transverse ventilation adits (TVA) framing into them
in addition to allowing the passage of the TBM
through them. Obayashi proposed a Value
Engineering Change Proposal (VECP) to drive three
of the four SEM connector tunnels by the TBM,
leaving behind as the final lining the pre-cast bolted
one pass lining associated with such a TBM. It was
proposed in the VECP to mine the two eastern
connector tunnels and one of the western connector
tunnels with the TBM thereby reducing a portion of
SEM excavation of the station and also reducing the
jet grout zones which would perhaps have required
inclined drilling beneath a secondary arterial street
Right-Of-Way.
This resulted in a redesign jointly developed with
HMMJ for the junctions between the TVAs and the
running tunnels by allowing the systematic removal of
the segments and providing a SEM junction chamber.
Also as part of the VECP, the west damper chamber
(DC) required reconfiguring to accommodate its
construction from the southwest TVA. Also the cast-
in-place final lining for the remaining southwest
connector tunnel would be replaced by a steel fiber
reinforced shotcrete (SFRS) final lining in the arch.
This VECP option seemed appealing initially.
However, upon further review of the schedule, the
advantages were overcome by events and the subject
was dropped. At the same time that the VECP was
being developed, additional information gradually
became available from the post-award borings and
indicated that the geological conditions at the east end
of the station were more difficult than originally
expected.
While the VECP solution would have reduced,
but not completely eliminated large portions of jet
grouting within the confines of a city street in a quiet
neighborhood, anything that could be done to reduce
the quantity of jet grout would have substantial
financial, political, and environmental benefits.
2.4. PlatformShift
Out of the connector tunnel VECP was conceived the
first Platform Shift design. The ground to the west of
the main shaft was found to be much better (stiff clays
and tills), so much so that the platforms could be
shifted 27 m (88 ft) to the west without drastic
changes in the interior details of the station itself
while at the same time utilizing the shafts (main and
ancillary) in their original locations (the slurry walls
had already been constructed). As part of the scheme,
the west TVA was relocated further west and three
additional angle drilled probes were performed to
confirm the suitability of the relocated position.
Further borehole information confirmed that the
shifted east DC and part of the east TVA were still
partly located in sands. Again to avoid the necessity
for further jet grouting from the surface or ground
improvement from within the shaft, a second scheme
to shift the east TVA to align directly with the
ancillary shaft was developed. This time the redesign
would have the benefit of deleting the east DC by
relocating the ventilation dampers into extended
platform tunnels. Further ventilation analysis
confirmed that this arrangement would be acceptable.
As a result the west TVA was reconfigured similar to
the east TVA with the dampers relocated to the
extended platform tunnels and a smaller junction
chamber replaced the west DC which was no longer
sized to house the dampers.

Fig. 2. Station Layout with Shifted Platform Tunnels
This design scheme displaced the VECP and was
developed in stages by HMMJ with input from
Obayashi and ST. This scheme (see Fig. 2) is being
implemented by ST.

3 PRETREATMENT AND SHAFT
CONSTRUCTION
3.1. Exploratory Drilling
A subsurface drilling program was specified for the
construction phase by the designers, and the
information gained enabled the geotechnical
interpretations to be refined. Obayashi carried out
more than 50 borings using mud rotary and sonic core
61
recovery methods between August 2004 and
November 2005. These were used to adjust the extent
of jet grouting and dewatering. For example, the
geological profiles at the east sections of the
northbound and southbound platform tunnels were
revised to indicate a long thick section of sand present
at the tunnel crown level. The borings were also used
for the installation of the surface instrumentation
including inclinometers, extensometers and
piezometers. The piezometers indicated that there
was approximately 15 m (50 ft) of water head in these
sands which, if not removed or the ground modified,
would have resulted in a flowing ground condition
upon excavation.
During SEM mining, systematic probing ahead of
the face is carried out. Also for each tunnel section, a
horizontal probe hole is cored to augment the
geotechnical interpretations. This information is
reviewed during the daily SEM meetings.

Fig. 3. Aerial View of Beacon Hill Construction Site
3.2. J et Grouting
J et grouting was carried out from the surface and
targeted zones of sand within the tunnel profile as
pretreatment for the SEM tunnels. J et grouting was
defined as the furnishing and installation of
overlapping jet grouted columns to allow tunnel
excavation with minimal water ingress and to provide
a stable crown and face for excavation. The volumes
and locations of ground treatment were specified on
the contract drawings and Obayashi was responsible
for the design of jet grouting to achieve the required
volumes and performances. The original contract
indicated known areas requiring jet grout pretreatment
as approximately 15 m (50 ft) along the west
Longitudinal Ventilation Adit (LVA) and a 15 m (50
ft) zone of the east Damper Chamber (DC). Condon
J ohnson/Soletanche J V was awarded the subcontract
in 2004. After some initial test columns late in 2004,
the production work continued between J anuary 2005
and December 2005. Inclined holes were originally
specified especially to alleviate restrictions of
grouting under Beacon Avenue at the west LVA and
near private property to the east of the station. With
the platform shift described earlier, this was no longer
a concern and the columns were drilled vertically
from within the staging area (see Fig. 3). The targeted
column spacing was generally on a triangular grid of
1.5 m (4.75 ft) centers. Predrilling was carried out
using a Klemm 806 rig drilling to the top of the
planned jetted zones. A Klemm KR 3012 drill rig
with a 24 m (80 ft) mast and high pressure pumps
were used for the jetting (see Fig. 4). Each hole was
surveyed with an inclinometer and results plotted to
confirm there was no divergence. Construction
quality testing included in-situ permeability testing
and core recovery with associated compressive
strength testing. Generally the strength results
achieved were between 3 MPa (400 psi), the specified
minimum, and 20 MPa (3000 psi). A total volume of
approximately 4200 m
3
(5500 yds
3
) was injected using
over 500 deep columns. The majority of this
treatment was to target the sands in the eastern
sections of the northbound and southbound platform
tunnels over a 46 m (150 ft) length and 18 m (60 ft)
length respectively, with additional jet grouting
performed in the tunnel breakout zones of the main
and ancillary shafts. In cross section, the target area
includes the sands within the tunnels and a zone at
least 1.2 m (4 ft) in thickness outside of the tunnel
initial lining in the sands.

Fig. 4. Vertical J et Grouting Rigs
3.3. Dewatering Wells
A system of vacuum-enhanced deep dewatering wells
was specified to reduce the hydrostatic pressure where
the tunnel excavations were expected to encounter
permeable soils below the water table. Sound Transit
was responsible for the well system design. Obayashi
are responsible for the proper installation, operation,
62
maintenance of the pumping wells and operation
system. Wells are generally spaced 15 m (50 ft) apart
along both sides of the tunnels, and at depths between
33 m and 55 m (110 and 180 ft). A total of 39 wells
and 10 observation wells were installed. As each
group of wells was installed, pump tested and brought
online, the drawdown effects were noticeable. The
pumps are two horsepower and have a capacity of 110
l/min (30 gpm) pumping against the heads described
above. Total pumping volume (steady state condition)
generally ranges between 110 and 190 l/min (30 and
50 gpm). The pumps are checked daily and
maintenance is performed when necessary.
3.4. Slurry Wall Shaft and Headhouse
Observations of the ground behavior from the Test
Shaft during the final design stage resulted in
redesigning both the main shaft and ancillary shaft
lining from SEM to using slurry walls.
The main shaft diaphragm is approximately 16 m
(52 ft) in diameter and is 55 m (182 ft) deep. This
work was performed by Soletanche using a hydro-
fraise machine mounted on a Liebherr crawler crane
cutting a 1 m (3 ft-4 inch) thick panel. The ancillary
shaft diaphragm is approximately 9 m (30 ft) in
diameter and has a depth of 51 m (167 ft). This work
too was done with the same hydrofraise used for the
main shaft except the cutting wheels were changed to
cut a thinner wall at 0.9 m (3 ft-2 inch).
The bentonite slurry transported the cuttings to a
separation plant complete with screens, cyclones, and
centrifuge for return to the excavation. A Cat 320 was
used to muck a pit constructed from the basement of
one of the houses demolished to clear the site. Rebar
cages were tied on site with block-outs for invert slab
niches and with pipe sleeves for instrumentation. The
cages with their attachments were lowered into the
bentonite and suspended from a structure on the guide
walls. Concrete trucks backed up to hoppers setting
on tremie pipes to deliver approximately 3500 m3
(4,600 yds3) of concrete to the main shaft and 1800
m3 (2,300 yds3) to the ancillary shaft. Since the
upper 18 m (60 ft) of circular main shaft slurry wall
and of the circular ancillary shaft slurry wall was to be
demolished while the interior excavation of the
headhouse basements was being done, a lean mix was
used in the upper reaches of the slurry wall panels.
The headhouse basement diaphragm wall was 0.8
m (2 ft-8 inch) thick and 19 m (62 ft) deep. This work
was performed by Soletanche using a conventional
cable grab mounted on a Leibherr crawler crane. The
grab deposited the material directly into trucks queued
on site. The work was orchestrated so that some of
the ancillary shaft headhouse wall panels were
constructed while some of the main shaft panels were
constructed.
3.5. Head House and Shaft Excavation
The main shaft was excavated using a Hitachi 330
Excavator with breaker and Cat 320 for excavating
and loading muck skips (see Fig. 5). Nine cubic meter
(twelve cubic yard) muck skips were lifted to the
surface and tipped into a muck bin at the collar using
a Kobelco 2000 (200 MT) lattice boom crawler crane.
The main shaft diaphragm walls which extended
through the headhouse were demolished as the
excavation advanced. This was all carefully
choreographed with the installation of several rows of
multi-strand tie-backs which extended approximately
21 m (70 ft) into the surrounding ground.
After the headhouse excavation was complete, a
cap beam approximately 1.8 m (6 ft) tall was formed
and poured, tying in dowels protruding from the top of
the slurry wall panels. Soon after, the interior
excavation of the circular shaft continued down to the
bottom using the Cat 320 excavator and the same
muck skips described above.
Upon reaching the bottom, the subgrade was
excavated to a dished shape, rebar dowels were
installed, and submarine style concrete invert was
poured as a provisional shaft bottom. Upon reaching
the design strength, the invert was backfilled with
spoils to develop a working platform for
commencement of the break-in to the SEM tunnels.

Fig. 5. West Headhouse Excavation
4 SEM CONSTRUCTION
4.1. SEM Organization
Under the oversight of the Tunnel Manager, Obayashi
employed an experienced SEM Manager to control
the day-to-day SEM activities along with a Site
Manager responsible for Beacon Hill Station.
Obayashi entered into an agreement with Beton and
Monierbau USA, Inc. (Evansville, Indiana) to provide
63
key SEM staff. Tunnel excavation and support
activities continue on a six day 24 hour working
schedule. Generally two crews are working three
shifts of 8 hours duration. Experienced SEM
Superintendents and SEM Project Engineers are on
site continuously to facilitate immediate decision
making at the face. These individuals are supported
by Walkers and Shift Engineers respectively.
During the design stage, agreement was reached
with Sound Transit that the Designer should be
represented on site during the implementation of the
SEM design. As mentioned earlier, Hatch Mott
MacDonald/J acobs J oint Venture (HMMJ ) was
responsible for the detailed design of all tunnels and
portals, shafts and mined station tunnels, including the
final lining and waterproofing system. Dr Sauer
Corporation (DSC) assisted with the SEM design and
waterproofing design for the Station as a sub-
consultant to HMMJ . During construction, HMMJ
and DSC provide a team of experienced SEM
engineers and SEM inspectors to assist the
Construction Management team (Parsons
Brinckerhoff) in providing engineering oversight of
the SEM excavation and support activities. STs
geotechnical consultant, Shannon & Wilson (S&W),
is represented on site providing oversight on
geotechnical activities.
As part of the regular communications required
for the control of the SEM work, daily on-site
meetings are held following a joint inspection of all
the SEM faces. Topics discussed include current
activities, planned activities for the next 24 hours and
instrumentation results. The meetings are always
attended by representatives of Obayashi, ST and
HMMJ /DSC and a partnering approach adopted by
the parties helps to ensure open communication.
Current progress, instrumentation results and
agreements reached during these daily meetings on
field decisions to better adjust the SEM to actual
ground conditions are entered into a J ournal Book and
signed by the Obayashi SEM Manager and the
HMMJ /DSC SEM Engineer. SEM activities are also
included in more formal Weekly Progress Meetings
used to discuss all C710 activities in a larger forum.
On a weekly basis, shotcrete strength results are
summarized and discussed at the SEM daily meetings.
This allows close control of any potential problems
and the timely agreement on any necessary mitigation
measures.
Construction Work Plans are developed by
Obayashi for each of the tunnels for review and
approval by ST. Any changes to these plans are
discussed in the SEM daily meetings. In addition,
contingency plans were developed and these include
procedures to implement additional tunnel support
measures.
A Required Excavation and Support Sheet
(RESS) is produced for all tunnel sections to assist
communications. The RESS confirms the excavation
sequence, required support, tool box items etc. and is
countersigned by the relevant parties.
Geologic mapping is performed during each
excavation cycle. The face maps are jointly agreed
between ST and Obayashi and countersigned.
Photographs are taken to complete the records. Along
with the borehole data, the geotechnical model is
constantly updated and presented to the interested
parties as interpreted geological sections.
4.2. West Longitudinal Vent Adit
The first SEM excavation was for a 3 m (10 ft) long
section of the 7 m (23 ft) wide west LVA using a top
heading, bench and invert sequence. This was turned
under as the shaft went by with the intention of
completing the remainder of this long decline later
from the bottom up. This section was completed
over a two week period in J une 2005. A Cat 320B
excavator with a milling head attachment, suitable for
mining through the jet grouted columns and the dense
clay material was used (see Fig. 6). After completion
of the west LVA, shaft excavation continued to gain
access to the Concourse Cross Adit (CCA) top
headings.

Fig. 6. West LVA Excavation
4.3. Concourse Cross Adits
The CCAs, with an excavated width of approximately
14 m (45 ft) and height of 12.5 m (41 ft) are the
largest tunnel openings on the project (see Fig. 7).
The north and south adits, each 20 m (67 ft) long,
connect the main shaft to the platform tunnels. A
grouted barrel vault pipe arch was installed in the
crown over the full length of each CCA prior to tunnel
excavation.
64

Fig. 7. Concourse Cross Adit Cross Section

4.3.1 Barrel Vault Installation
Specialty subcontractor Northwest Cascade Inc. and
Obayashi jointly installed multiple rows of perforated
steel pipes above the crown of both CCAs from the
main shaft. On the basis of more favorable conditions
in the south CCA, two rows of pipes were reduced to
one. Pipes were drilled using a Klemm KR 806-3
hydraulic rig at 45 cm (18 inch) centers with 10 cm
(4-inch) diameter used for the shortest pipes
terminating at the platform tunnel junction and a
larger 15 cm (6-inch) diameter for the longest pipes
extending beyond the headwall up to 23 m (75 ft) in
length. The pipes were drilled as lost casing using J
teeth welding into the lead casing pipe. After
cleaning, each pipe was surveyed and then weak
cement/bentonite grout dams were placed to ensure
micro-fine cement would not run along the annulus.
After the grout dams set up, stage grouting in 1.5 m (5
ft) sections using a double packer system and
microfine MC-500 portland cement grout was
performed. The refusal criteria was 85 l (3 cubic ft)
grout per 30 cm (lineal ft) pipe or holding 14 bar (200
psi) for 10 minutes in sand zones and 5 minutes in
clay/silt zones. The target strength was originally 14
MPa (2000 psi) after 48 hours but was later changed
to 3.5 MPa (500 psi) in 24 hours. The required
positional tolerance of 1 % was confirmed using a
down-the-hole Maxibore horizontal inclinometer. A
total pipe length of approximately 1980 m (6500 ft)
was drilled and grouted in 7 weeks with the crews
working two 10-hour shifts/day.
4.3.2 Excavation and Support
Due to the large size of the openings and the expected
difficult ground conditions especially in the crown,
the excavation sequence was prescribed as a twin-
sidewall drift each with top heading, bench and invert
followed by the center drift top heading, bench and
invert. Finally the temporary sidewalls were removed
in stages to form the completed ring shape. At
Obayashis request, the center drift sequence was
changed and the center drift top heading size was
increased to permit the use of the Liebherr 900
excavator (see Fig. 8). This top heading was driven
all the way to the headwall before removal of the
bench and completion of the invert closure (see Fig.
9).

Fig. 8. South CCA Center Drift Top Heading Excavation

The excavation for the south CCA commenced
with the breakout of the shaft slurry wall concrete in
August 2005. The Liebherr 900 excavator with a
purpose-made rotating boom was used to excavate the
side-wall drift top headings. Various tool box items
were used including face bolts, pocket excavation and
welded wire fabric. The overlapping 11 m (35 ft)
long probe holes drilled ahead of the face were
generally dry with the exception of the probes in the
southeast adit which provided small flows of less than
4 l/min (one gpm). A sand dyke was first encountered
in the southeast side drift with localized flowing sand
in the crown requiring the use of tool box items such
as grouted pipe spiles and well points. Pocket
excavation was necessary in this area with the
immediate application of flashcrete. The sand dyke
was again encountered in the center drift top heading,
and following the previous experience was more
effectively handled primarily with the use of pocket
excavation, grouted pipe spiles and additional
shotcrete. The south CCA was completed following
65
the construction of the main shaft base slab and
temporary backfilling in February 2006.
The north CCA commenced in October 2005 and
at the time of writing was just completed.

Fig. 9. View of North CCA Side Drifts from the South
CCA
4.4. PlatformTunnels
The southbound and northbound platform tunnels as
originally designed are approximately 103 m (338 ft)
long each, with excavated dimensions of
approximately 11 m (37 ft) wide and 10 m (32 ft) high
(see Fig. 10). The length of each platform was
increased to 132 m (434 ft) for the platform shift
redesign. The excavation and support of the
southbound platform tunnel commenced with the
breakouts from the south CCA lining for both the east
and west drives in March 2006. Two rows of grouted
pipe spiles were used as pre-support in the breakout
zones. Initially the advance length was 1.2 m (4 ft)
but this was later increased to 1.4 m (4 ft 6 inch). The
shotcrete thickness is 35 cm (14 inch) including 5 cm
(2 inch) flashcrete. Reinforcement is 2 layers of 6x6
W12xW12 mesh. Lattice girders are installed at 1.2
m (4 ft) centers close to the face of each excavated
round. Steel TH girders are installed in the temporary
sidewalls for ease of later removal. The invert of the
first side drift leads the top heading of the second by a
minimum of 7 m (24 ft). Systematic probe drilling is
carried out in each top heading generally over an 11 m
(35 ft) length with a minimum 5 m (16 ft) overlap. To
date the west drive has encountered primarily dry silty
clay conditions with some localized sand pockets.
Varying quantities of rebar spiles of length 4 m (12 ft)
and spacing 30 cm (1 ft) are driven in advance of the
crown at each round of the top heading. The
excavation sequence requires the completed invert to
be a maximum distance from the top heading face of
10 m (34 ft). Following invert construction of 2.5 m
(8 ft) in one cycle, the sequence follows with the
excavation and support of a 1.2 m (4 ft) long section
of bench then a 1.2 m (4 ft) top heading followed by
another bench and top heading again before the next
invert cycle (see Fig. 11). A larger Liebherr 932
tunnel excavator is used to excavate the platform
tunnels. The temporary side-wall is removed
generally in 2.5 m (8 ft) stages ensuring that the joints
at the crown and invert are carefully inspected and
constructed and maintaining a minimum of 2.5 m (8
ft) of intact side-wall from the completed invert. The
invert shotcrete is protected with a minimum 1 m (3
ft) layer of temporary backfill. Any local seepage
water encountered at the face is collected in pipes held
in position by shotcrete and channeled away.

Fig. 10. Platform Tunnel Cross Section
At the time of writing the east drive of the southbound
platform tunnel had been completed approximately 8
m (25 ft) in the first side drift of the top heading. A
sand lense ahead of the face, which was originally
detected from the subsurface drilling program, has
resulted in additional probing from the tunnel face.
Along with the additional probes, drainage lances
were installed to better understand this complex piece
of ground and to take advantage of whatever drainage
effects could be realized from underground.

66

Fig. 11. Platform Tunnel Excavation Sequence

Fig. 12. View of the Southbound Platform Tunnel from the
South CCA
The northbound platform tunnels are expected to start
shortly with the barrel vault installation from within
the north CCA. A double row of 18 m (60 ft) long
pipes was designed for the start of the northbound
platform tunnels in both directions. However,
because of more favourable ground conditions
identified during the supplemental geologic
exploration and as a result of the north CCA
excavation, the double row has been reduced to a
single row. Also the pipes in the west direction have
been reduced to 12 m (40 ft) lengths. Excavation of
this tunnel will commence upon completion of the
barrel vault and the excavation of the southbound
platform west drive (see Fig. 12).
4.5. Initial Shotcrete Lining
The design specifications require shotcrete
compressive strengths of 14 MPa (2000 psi) at 24
hours and 34 MPa (5000 psi) at 28 days. Dry-mix
fiber reinforced shotcrete is specified for the
minimum 5 cm (2-inch) thick flashcrete layer (see
Fig. 13). The remaining initial lining thickness of
generally 30 cm (12-inches) is sprayed using wet-mix
reinforced with two layers of welded wire fabric.
Panels for shotcrete testing are sprayed daily in the
tunnels during the initial lining shotcrete application.
Two panels for dry-mix and two panels for wet-mix
provide the necessary number of cores for off-site
testing by an independent laboratory. On-site testing
facilities are used by Obayashis QC department
primarily to check early strengths. In general the
results are consistently better than those specified.
However, following some sporadic low 24 hour test
results for the dry-mix during the early stages,
procedures were improved especially for handling the
panels during transportation. In-situ cores are taken
occasionally to verify the panel results. The sporadic
low 24 hour results, taken from panels, have been
checked and found acceptable when in-situ cores were
taken and tested. Penetration nail testing is also used
unofficially to verify the early strength gain for
shotcrete less than 12 hours old.
The shotcrete thickness is measured and
controlled in the field primarily using the lattice
girders once they have been surveyed in position.
Overbreak/overexcavation is generally filled with the
flashcrete layer.
Pre-bagged shotcrete was initially used for the
dry-mix until confidence was gained with the on-site
batcher. The on-site batcher is a 46 m
3
/hr (60 cubic
yd/hr) volumetric batcher and it is planned to be
augmented with a second 76 m
3
/hr (100 cubic yd/hr)
weigh batcher dedicated generally for wet-mix
shotcrete. Other improvements include reconfiguring
the shotcrete pump system so that the pumps are
located underground in the south CCA and supplied
by drop holes from the batcher. This will reduce the
lengths over which shotcrete is pumped thereby
reducing the risk of blockages and downtime.
Shotcrete nozzlemen are required to be
experienced and have ACI certification. In addition,
panels are sprayed by each nozzleman to allow
shotcrete cores to be visually inspected and
categorized. A shortage of skilled nozzlemen was
experienced in the early stages, and on-site training of
nozzlemen is helping to overcome this shortage. All
shotcrete is sprayed by hand generally from man-
baskets with some limited use of the Oruga shotcrete
mobile robot. When space allows, a larger
Spraymobile Robot will be introduced to the headings.
Obayashi initiated a redesign to replace the
specified welded wire fabric reinforcement with steel
fibers for the platform tunnel initial lining. Flexural
strength testing is required from beams sawed from
test panels. After some initial difficulties primarily
with the fiber dosage control and blocked shotcrete
lines, this change was temporarily suspended. Steel
67
fibers will be reintroduced once the second batch plant
is operational.

Fig. 13. Flashcrete for the Southbound Platform Tunnel
Breakout
4.6. Tool Box
The excavation sequences and ground support
measures specified on the drawings are augmented
with discretionary additional excavation and ground
support measures that are installed based on the
encountered ground conditions. These measures (tool
box items) provide pre-support, support or ground
improvement around or within the tunnel. They
include rebar spiles, grouted pipe spiles, metal sheets,
face wedge, pocket excavation, face bolts, permeation
and fracture grouting, soil nails, additional reinforced
shotcrete, and vacuum dewatering. The tool box
items are installed as approved or directed by Sound
Transit. For planning and estimating purposes a table
of quantities of tool box items was included in the
GBR and the Contract Price Schedule. The GBR
table is reproduced as Table 1.

Table 1. Tool Box Items from the GBR
4.7. Instrumentation
An extensive array of approximately 54 instruments
has been installed by Obayashi from the surface in
advance of tunnel excavation. This includes
extensometers and inclinometers for monitoring
ground movements. Readings are generally taken 2-3
times weekly depending on the proximity of the
advancing tunnel face. With the exception of the
surface settlement readings taken by CH2M Hill for
Sound Transit and readings taken in the tunnel by
Obayashi, Shannon & Wilson is responsible for taking
the readings, consolidating all data (CH2MHill,
Obayashi and S&W) and presenting the data on behalf
of Sound Transit. To date the maximum recorded
surface settlement is approximately 6 mm (0.25 inch).
Obayashi is responsible for the timely installation
of the tunnel instrumentation at each monitoring
section, generally at 15 m (50 ft) spacings in the
platform tunnels. A typical section consists of optical
targets used to monitor the deflection of the lining,
earth pressure/shotcrete stress cells and strain gauges
fixed to the lattice girders. All instruments are
installed prior to shotcrete application. Initial
readings are taken within hours to ensure valuable
data on deformation is not lost.
Obayashi employs a Professional Land Surveyor
to implement the survey program. The optical targets
are generally read to the required accuracy of 0.15
mm (0.006 ft). There was a period during the initial
excavation of the CCAs when it was necessary to use
a tape extensometer to supplement the surveyed data.
However, given the congestion at the bottom of the
shaft, improvements were made to the survey
allowing the continuation of the optical method. The
data is presented using Eupalinos software and
discussed at the Daily SEM Meetings. The data is
monitored against specified threshold and limiting
values, and also for unusual trends. Although in some
cases threshold and limiting values are reached, the
readings are generally well within expectations. For
example, the maximum recorded roof settlement of
the south CCA was 15 mm (0.60-inch).
4.8. Construction Sequence/Schedule
The SEM excavation sequences in the various mined
station adits, tunnels and other underground openings
of the station are prescribed on the drawings. The
sequencing of the mined station tunnels excavation
relative to each other is flexible. Obayashi has
freedom in selecting construction methods,
equipment, procedures and sequences, subject to the
approval of Sound Transit. The contract requires that
the platform tunnels be excavated prior to TBM
arrival.
Obayashi are employing three crews working
from the main shaft. At the time of writing, the crews
68
had just completed mining of the north CCA and are
progressing on the southwest platform tunnel, the
southeast platform tunnel, and the installation of
barrel vault pipes over the northbound platform
tunnels concurrently. In addition, there is a crew
working out of the ancillary shaft to execute the
emergency tunnel and the east transverse ventilation
adits.
The crews consist generally of 10 individuals on
each shift. After a learning curve, the cycle times
have been reduced to approximately 6 hours per round
to date (Fig. 14) which approximates to 6 m (20 ft) per
week in the platform tunnel. Where possible and
dependent on the soil conditions and the deformation
monitoring, field changes are made to assist
production and reduce the cycle times. For example,
the specified completion of all support prior to the
next excavation cycle has so far been relaxed to allow
the second layer of mesh and final shotcrete layer to
be delayed by up to a maximum of three rounds.
The total duration of all the SEM excavation is
scheduled to be 24 months out of the total 48 month
contract.

Fig. 14. Platform Tunnel Average Cycle Times
5 CONCLUSION
Soft ground SEM tunneling in such variable ground
conditions as the local water-charged glacial deposits
in Seattle presents significant technical challenges.
These challenges were recognized with the provision
of a robust design with appropriate excavation
support, pre-support, and available tool box items.
Along with having suitably experienced field staff, an
open partnering approach between the parties has
been a key to the safe and successful tunnel structures
excavated to date.
The authors would like to thank all the staff and
workforce involved in making this such an interesting
project worthy of presenting to the engineering
community, and advancing the technical boundaries
of soft ground tunneling in the USA.
REFERENCES
1. Tattersall C., M. Murray, J . Laubbichler, F. Langer.
SEM Tunneling Underway in Seattle Construction of
the Beacon Hill Station and Tunnel. NAT 2006.
2. Phelps D., J . Gildner, C. Tattersall, J . Laubbichler,
McAllister. Design and Risk Management Strategy for
the Sound Transit Beacon Hill Station and Tunnels
RETC 2005.
3. Robinson R., M. Kucker, M. Lehnen, S. Warren,
McAllister. Impacts of Geotechnical Issues on Design
of the Beacon Hill Tunnel and Station Project RETC
2005.
4. Hatch Mott MacDonald J acobs. Geotechnical Baseline
Report 2004.
5. Tattersall C., T. Gregor, M. Lehnen. Design and
impact of the Beacon Hill Station exploratory shaft
program NAT 2004.
6. Laubbichler J ., T. Schwind, G. Urschitz. Benchmark
for the future: the largest SEM soft ground tunnels in
the United States for the Beacon Hill Station in Seattle
NAT 2004.

Platformtunnel - Weekly Progress:
Hours per Sidewall Drift 4 feet Rounds of TH, B or Invert
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
1 2 3 4 5 6 7 8 9 10 11 12 13
Week
H
o
u
r
s
69
1 INTRODUCTION
The Greater Vancouver Water District
(GVWD) is currently constructing the Seymour-
Capilano Filtration Project (SCFP) to enhance the
quality of water supplied to the Greater Vancouver
area from the Seymour and Capilano reservoirs in
North Vancouver. The main element of the SCFP is
a 1,800-megalitre per day filtration plant, currently
being constructed in the Seymour Valley. This
plant, when built, will be the largest in Canada.
Considerations of cost, community impacts and
environmental issues resulted in the decision to
build a single filtration plant in the Seymour Valley,
connected to the Capilano reservoir by twin tunnels,
rather than two separate plants adjacent to each
water source. Raw water will be drawn from the
Capilano reservoir and pumped through the Raw
Water Shaft, Raw Water Tunnel and Seymour Shaft
raw water riser pipe to the head of the Seymour
Filtration Plant. Treated water will return by gravity
through the Seymour Shaft treated water riser pipe,
the Treated Water Tunnel and the Treated Water
Shaft through an Energy-Recovery Facility and
pressure-balancing tank to the Capilano distribution
mains.
The Twin Tunnels component of the project
comprises the 11m diameter, 180-m deep Seymour
Shaft as the main construction access shaft, twin,
3.8-m diameter, 7.1-km bored tunnels (the Twin
Tunnels), and two 4-m diameter, 275-m deep
raisebored shafts connecting the tunnels to surface at
Capilano. Figure 1 shows the Laydown area of the
Seymour Shaft in May 2006.
In operation the twin tunnels will be
pressurized. The Capilano Pumping Station will
pressurize the Raw Water Tunnel. Gravity will
pressurize the Treated Water Tunnel. The resulting
hydraulic grade line will be above the surface
topography at both ends of the tunnel alignment.

Figure 1. Seymour Shaft laydown area.

Construction Update and TBM Excavation Planning
Seymour Capilano Twin Tunnels Project, Vancouver
Dean Brox, J oe Rotzien
Hatch Mott MacDonald, Vancouver, B.C., Canada
Christian Genschel, J osef Messner, Arvindh Gupta
Bilfinger Berger Canada, Vancouver, B.C., Canada
Tom Morrison
Greater Vancouver Regional District, Burnaby, B.C., Canada
Andy Saltis
Pacific Liaicon & Associates/SNC Lavalin, Vancouver, B.C., Canada

ABSTRACT: The Greater Vancouver Water District (GVWD) is currently constructing the Seymour-Capilano Filtration
Project to enhance the quality of water supplied to the Vancouver area from the Seymour and Capilano reservoirs in North
Vancouver. The project, designed by Hatch Mott Macdonald, includes major underground works comprising twin 3.8-m
diameter, 7.1-km bored tunnels, the 11 m diameter, 180-m deep Seymour Shaft as the main construction access shaft, and the
twin 4-m diameter, 275-m deep Capilano raisebored shafts. This work was tendered as a single contract, which was awarded
to Bilfinger Berger (Canada) Inc. Construction started in early 2005 with the excavation of the Seymour Shaft by
conventional shaft sinking methods through 30 m of dense glacial deposits followed by 150 m of granitic bedrock. Bilfinger
Berger used lattice ring girders, mesh and dry-mix shotcrete for shaft support while sinking through the glacial deposits. Shaft
support for the bedrock portion followed the Engineers design, comprising pattern rock bolts, mesh and shotcrete. Shaft
sinking was completed in early November 2005. Ancillary shaft-base excavations and structural, electrical and mechanical
work were completed in May 2006 for TBM installation and launching planned for mid-2006.
70
A design separation of 100 m is required to
prevent cross-flow between tunnels; the tunnel
alignments converge at their east and west ends.
Steel lining will therefore be installed in the east
and west ends of both tunnels to prevent leakage.
Investigations and preliminary design work for
the SCFP date back nearly ten years. The GVWD
selected Hatch Mott Macdonald (HMM) to
participate in the preliminary design of the Twin
Tunnels and awarded the detailed design to HMM in
2002. The GVWD awarded the construction of the
Twin Tunnels as a single contract, with lump sum
and unit price components, to Bilfinger Berger
(Canada) Inc. in August 2004. Pacific Liaicon and
Associates/SNC Lavalin are responsible for
construction management and HMM are providing
resident engineering services. The project is
scheduled for completion in early 2009. The GVWD
appointed a Technical Advisory Board in late 2004.
Tenderers were provided with a Geotechnical Data
Report and Geotechnical Baseline Report. The
contract makes provision for Escrow Bid Documents
and a Dispute Review Board.
Construction began with the excavation of the
Seymour Shaft in mid-J anuary 2005. Shaft sinking
was completed in early November 2005. The
Seymour Shaft has been the first major, vertical
shaft sunk in British Columbia in more than twenty
years. It has also been one of the deepest shafts ever
sunk using a crawler crane as the primary hoisting
system. Credit is due to the Contractors staff that
shaft sinking was completed with no major accidents
or injuries.
2 PROJ ECT LOCATION
The GVWD decided early in the design process
to connect the Capilano and Seymour sites entirely
by tunnels, rather than by combinations of tunnels
and surface pipelines. Length, routing and
community impacts made this the best option.
Tunnelling, however, was affected by the
existence of deep buried valleys at both ends of the
tunnel alignment, eroded from the bedrock by
glaciers and filled with mixed, water-bearing glacial
materials. The probable cost and risk of tunnelling
through the buried glacial valleys led to the decision
to locate the tunnels entirely in bedrock, even
though this entailed substantial shaft construction.
The horizontal and vertical alignments of the
tunnels were both reached after considering many
possible options.
The tunnels will be driven westwards from the
Seymour Shaft, beneath the Lynn buried valley,
passing beneath the slopes of Grouse Mountain and
Mount Fromme and under the Capilano buried
valley to a western terminus 275 m from surface
immediately south of Cleveland Dam. There, the
tunnels will be connected to surface by twin 4-m
diameter raisebored shafts. All excavation spoil will
be hoisted through the Seymour Shaft and trucked
to a disposal site in the Seymour valley; much of
this material is being used in other parts of the
project and in the Seymour Falls Dam seismic
upgrade. Figure 2 shows the overall project layout.
The Seymour Shaft is located on the Seymour
glacial plateau formed at the confluence of the Lynn
and Seymour River valleys. The horizontal
alignment was selected during the early stages of
detailed design from northerly and southerly
alignment options, based on perceived geotechnical
risks.
The location initially selected for the Seymour
Shaft was at the west side of the Filtration Plant.
Drilling revealed water-bearing glacial materials to
100 m from surface with shattered bedrock
conditions beneath, possibly the extension of the
Lynn Creek Fault. Ground freezing was
contemplated as a means of controlling water inflow
into the shaft, but hydrological testing indicated a
risk that groundwater movement could add heat to a
freezewall faster than the freeze plant could remove
it, making this method problematical at best.
The GVWD and its consultants considered ten
alternative shaft sites, taking into account technical,
financial, social and environmental considerations.
The site finally selected was about 200 m southwest
of the original site, where seismic surveys and
drilling proved the existence of a buried bedrock
hump rising to within 30 m of surface. This hump
rose above the main water-bearing members of the
glacial stratigraphy and consisted of reasonably
competent granitic rocks. The Twin Tunnels
contractor was provided with a cleared, leveled,
drained and graded site; site operations began in the
autumn of 2004.
3 SEYMOUR SHAFT CONSTRUCTION
3.1 Shaft Excavation and Support in
Overburden

71

Figure 2. Project layout.
Bilfinger Berger Canada commenced construction of
the project with the excavation of the 11 m diameter,
180 deep, Seymour Shaft in J anuary 2005. The 30 m
deep overburden portion of the Seymour shaft was
excavated conventionally through dense glacial
materials in approximately 1.0 m lifts using a CAT
308 excavator with mucking into 6 m
3
buckets. Shaft
excavation was undertaken with two 12-hour shifts
with shift changes at 6:00 AM and 6:00 PM daily.
The base design of the support for the
overburden section of the shaft consisted of W250 x
131 ring girders at 1.2 m vertical spacing for the first
15 m of shaft, 0.9 m spacing for the next 5 m and 0.6
m for the final 10 m down to bedrock, with 6-mm
corrugated liner plate behind the ring girders, grouted
in place. Bilfinger Berger Canada proposed an
acceptable alternative support system for sinking
through the glacial deposits, consisting of lattice
girders, mesh and 250 mm thickness of dry mix
shotcrete and shown in Figure 3. The ground support
was installed immediately upon completion of
excavation of each lift. No raveling or any form of
instability occurred during excavation in the
overburden and the alternative shaft support system
performed well. Sporadic, localized groundwater
inflows were encountered during excavation and
drainage holes were completed accordingly. The dry-
mix shotcrete for the overburden portion mainly
comprised a pre-manufactured product that included
an accelerator. The strength of dry-mix shotcrete
exceeded the project specifications based on results
from test panels and from in situ cores.

Figure 3. Shaft support in overburden.

72
3.2 Shaft Excavation and Support in Bedrock

The 150-m bedrock portion of the shaft was
excavated by conventional drill and blast methods
using two Atlas Copco R3 Crawler drill rigs. A typical
blast hole production round varied from 1.5 m to 4.0
m deep. The blast design was based on 220 parallel-
drilled blast holes with a typical perimeter hole
spacing of 0.5 m. The blast design initially comprised
a wagon-wheel cut comprising four reamer holes and
a single middle blast hole. The blast design was later
changed to a nine hole cut comprising five reamer
holes with four surrounding blast holes, which
produced significantly improved results.
Shaft support for the bedrock portion of the
shaft followed the base design and comprised a
combination of a pattern of 4.0 m long, 32 mm
diameter, fully grouted and hand tensioned rock bolts
at a typical 1.5 m square spacing, welded wire mesh
and 150 mm minimum thickness of dry-mix shotcrete.
Welded wire mesh was installed around the
full shaft perimeter for the full round depth following
mucking. Shotcrete was then applied followed by the
installation of pattern rock bolts. Shotcrete was
applied to each excavation lift using a portable lifting
assembly that was designed by Bilfinger Berger
Canada and lowered to the shaft floor by a crawler
crane. The portable shotcrete lifting assembly
comprised 2-1 m
3
hoppers, 2-16 liter Aliva 252
shotcrete machines, and 2-150 liter pumps for
injection of accelerator. Water was provided by a
direct connection to a major water supply pipeline.
Water temperature from this source was about 5 C.
Shotcrete application as part of the shaft support
system was typically completed within four hours
with two nozzlemen. The dry mix shotcrete that
formed part of the shaft support system comprised
both a pre-mix product and an on-site batch mix. The
pre-mix product was Target Superstick Shotcrete,
comprising Type 10 Cement, 2.25% silica fume, fine
and course aggregate, and Target Set Accelerator at
3%. The on-site batch mix comprised Lafarge Type
10 Cement and blended aggregate conforming to ACI
Gradation 2, with varying additives. Figure 4 shows
the portable shotcrete lifting assembly used during
shaft excavation.
Shotcrete application was closely monitored
and tested during excavation as per the project
specifications with a testing frequency of three cores
(from panels and/or in situ) for every 200 m
2
of
applied shotcrete. The quality control testing was
completed by Metro Testing Services of Vancouver
on behalf of Bilfinger Berger Canada.

Figure 4. Shotcrete lifting assembly.

Both the pre-mix and on-site batch mix
shotcrete provided consistently good strength results
from the quality control and quality assurance testing.
Quality assurance testing of the shotcrete was
carried out by AMEC of Vancouver on behalf of the
GVWD. The quality assurance testing was typically
carried out at a frequency of 10% of the quality
control testing. The quality assurance testing always
comprised in situ core testing.
Figure 5 shows the drilling of rock bolts using
a single drill rig. On a few occasions a number of
hollow-core self-drilling injection anchors were
installed in closely fractured metavolcanic rock that
prevented easy insertion of rock bolts following
normal drilling. The support system for the bedrock
portion of the shaft performed well and no form of
any instability manifested along the shaft walls during
excavation. Convergence monitoring was carried out
at regular intervals behind shaft excavation and
confirmed the satisfactory performance of the shaft
support system.

Figure 5. Shaft support in bedrock.
73
4 SEYMOUR SHAFTBASE CONSTRUCTION
The shaftbase excavation adopted for construction
was not significantly dissimilar to the proposed base
design. The shaftbase was allowed to be modified by
the Contractor and Bilfinger Berger Canada elected to
adopt the same excavation width for the starter and
back chambers of approximately 14 m and 10 m
respectively but extended the length of both the starter
and back tunnels as well as the back chamber to
facilitate the set up and launching of the TBMs and
mucking logistics. The excavated lengths of the starter
tunnels, back tunnel, and back chamber were about 35
m, 50 m, and 30 m respectively. The excavated
heights of the starter and back chambers are 6 m and
about 5 m respectively. Figure 6 shows the start of
excavation of the starter tunnels in early 2006.
Prior to the first blast round in the starter
chamber a probe hole was drilled to investigate the
possible presence of a fault zone inferred from a pre-
construction borehole. No clear evidence of a major
fault zone was detected however groundwater inflows
of 50 l/min were encountered that dissipated to less
than 30 l/min after 48 hours and continues at 10 l/min.
Excavation of the starter and back chambers were
completed by drilling and blasting 1.5 m to 3.0 m full-
face rounds. The initial rock support for the starter
chamber was modified from the base design by
Bilfinger Berger Canada to include lattice girders for
the first 6 m. In addition, 6-m long rock bolts were
installed for the first 4 m. The length of the rock bolts
was subsequently reduced to 4 m for the remainder of
the chamber excavations based on inferred good
geological conditions and the initial results of the
convergence monitoring.
Welded wire mesh and dry-mix shotcrete were
applied to a thickness of 150 mm. Excavation of the
starter and back tunnels were also completed in 1.5 m
to 3.0 m full-face rounds. Spot rock bolts and full
coverage shotcrete with mesh were applied for
support.
The encountered rock conditions within the
shaftbase excavation typically comprised mixed
granitic and metavolcanic rock of varying quality.
Groundwater inflows into the shaftbase excavation
have been measured to date to be less than 20 l/min.
The final excavation for the shaftbase comprised an 8
m deep pit to receive the mucking buckets and an 8 m
deep sump.

Figure 6. Shaftbase excavation, early 2006.
5 SEYMOUR SHAFTBASE INSTALLATION
Substantial structural, mechanical and electrical work
was needed in the Seymour Shaft and shaftbase
chamber in preparation for TBM excavation. This
work lasted from March to May 2006. The sidewalls
at the heading of starter tunnels have been concreted
to facilitate the bearing of the thrust pads of the
TBMs. Rails have been laid out for the movement of
rolling stock.
A Liebherr modular gantry crane was
constructed at the shaft collar, equipped with two
hoists, each capable of hoisting 70 tons as can be seen
in Figure 1. Figure 7 shows the hoisting drums for the
gantry crane system. These are being used for
installing the TBMs and, once tunneling begins, will
be used to hoist two 28 m
3
mucking buckets built to a
Bilfinger Berger patented design. The gantry system is
unique and equipped with electronic sensors for
automatic muck removal operation. Most of the
components of the gantry are available off the shelf to
facilitate easy repairs without too much down time for
parts.
An Alimak elevator with a capacity of 2,400 kg
or 23 people was installed, running on a rail bolted to
the shaft wall. The electrical facilities were installed
to transmit power down the shaft at 12.47 kV for
transformation at the shaft base to 4,160 V for supply
to each TBM at 600 V. Two 100-kW ventilation fans
were installed on surface with ducting down the shaft.
In addition, a 400 m
3
capacity sump pit and two muck
pits each having a volume of 125m
3
have been
excavated at the shaft bottom. The entire water from
the tunnel headings and behind the TBM will be
pumped in to the sump pit and then pumped out to the
surface for further treatment and disposal. The pit for
the rock buckets was covered with two hydraulically
actuated doors. The chamber floor was concreted and
74
track, switches, car dumps and muck chutes were
installed. Concrete cradles and gripper walls were
installed for the start of TBM excavation.

Figure 7. Liebherr hoisting systems.
6 TBM SELECTION AND DESIGN
Bilfinger Berger Canada selected two new 3.8 m
diameter, 1,260-kW TBMs that were built new for the
project by Robbins at Solon, Ohio. These machines
are high-powered, open-shield, hard rock machines
with 483 mm (19-inch) cutters and support fingers
immediately behind the shield. With trailing gear,
each machine will be approximately 250 m long. The
trailing gear will include pumping and grouting
equipment and will also allow for passing of the
mucking trains. Each TBM will be equipped with two
drills for forward probing and pre-excavation
grouting. A probe-hole will be maintained
continuously ahead of the face.
Bilfinger Berger Canada has designed the TBM
mucking arrangement such that each 1.5-m push by
the TBM will fill a train of six Mhlhauser 5 m
3
side-
dumping cars. A GAI diesel locomotive will haul each
train from the TBM to the Seymour Shaftbase for
tipping. One train will fill one 28 m
3
mucking bucket.
The TBM muck will be hoisted to surface and trucked
to the GVWD designated disposal site located 4 km
away up the Seymour valley.
The Robbins TBMs were trucked from Ohio to
Vancouver in 72 tractor-trailer loads, starting in early
May 2006. TBM installation and commissioning, with
the assistance of Robbins personnel, was carried out
in the latter part of May and early J une 2006. Figure 8
shows the first TBM assembled in the shaftbase.
TBM tunneling is anticipated to begin in mid-J une
2006.

Figure 8. Installation of first TBM in shaftbase.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the
permission of the Greater Vancouver Water District
and Bilfinger Berger (Canada) Inc. to publish this
paper.

REFERENCES

BROX, D.R., PRINGLE, J ., PHELPS, D.,
PROCTER, P. MORRISON, T., and SALTIS, A. The
Seymour Capilano Twin Tunnels, Vancouver, BC,
Canada. RETC 2005 Seattle.
75
Design-Build Tunnels and Shafts at the San Roque Project

Richard Humphries, P.Eng, P.E.,
Mark Funkhouser, P.E.
Golder Associates, Vancouver, B.C., Canada and Lansing, MI, USA

Edward OConnor, P.E.
Washington Group International, Inc., New York, USA
ABSTRACT: The San Roque Multipurpose Project, in the Philippines, is one of the largest Build-Operate-Transfer
(BOT)/Design-Build hydropower projects that has been constructed. The project cost $1.1 billion and was completed in
2003. The tunnels and shafts comprise a significant portion of this cost. The project includes a 200m high embankment
dam, a 12,800 m
3
/sec spillway, a 345 MW powerhouse, eighteen tunnels and seven shafts. The tunnels vary in size from the
10.4m x 16m diversion tunnels to the 3.5m to 4.5m grouting galleries, and the shafts vary in size from the 28m wide by 80m
long by 48m deep shaft for the powerhouse and the 23m diameter surge shaft to the 1.5m diameter grouting gallery vent
shafts. The tunnels were all excavated by the drill-and-blast method and the shafts were excavated by the raise bore and
slash method. The BOT development put the design and construction on a fast track and the design-build contract offered
many opportunities for the designers and constructors to work as a team to develop the most efficient and cost effective
methods to complete construction in an accelerated schedule. One of the major changes from the initial concept was the use
of shotcrete lining for the diversion tunnels, which saved six months on the construction schedule. This paper describes the
design and construction of the tunnels and shafts.

1 INTRODUCTION
The San Roque Multipurpose Project is on the island
of Luzon in the Philippines. The project provides
hydroelectric power, irrigation water supply and flood
control. It was developed as a BOT Project by the San
Roque Power Corporation. Washington Group
International was the Design-Build contractor and
Golder Associates was the sub-consultant to
Washington for the design of the rock engineering
aspects of the project.

Figure 1 Plan of San Roque Project

As the funding for the project came from commercial
sources, there was a great need to minimize interest on
borrowed money by compressing the construction
schedule. The Design-Build approach is well suited to
this schedule compression as the full design does not
have to be complete before construction can start. In
fact, design and construction can proceed
simultaneously, with the design of structures that will
be constructed late in the schedule deferred until the
initial construction has started. At the San Roque
Project, for example, construction of the diversion
tunnels and cofferdam started as soon as the design of
these structures was complete and before the detailed
designs of the main dam and other tunnels were well
advanced.

Note: The originally planned Low level
Outlet stilling basin was changed to a flip
bucket.
76
The layout of the project is shown in Figure 1. The
central feature of the project is the 200m high San
Roque Dam, which spans the Agno River at the
downstream end of the steep gorge. This dam is a
conventional embankment dam with a central
impervious core and earth/rockfill shells [1].

Table 1 - Main Statistics of the Project

Earth/Rockfill
Dam
height.............................. 200m
side slopes......... 2:1
Cofferdam height.............................. 55m
Power Tunnel finished diameter............ 8.2m
length.............................. 1300m
Power Tunnel
Drop Shaft
finished diameter............ 8.2m
depth............................... 5.5m
Power Tunnel
Gate Shaft
finished diameter............ 12m
depth............................... 100m
Surge Shaft finished diameter............ 20m
depth............................... 100m
Powerhouse 3 x 115mw units
Shaft
Excavation
85m x 28m x 46m deep
Diversion
Tunnels
2 Tunnels - 10.4m wide x 16m high
1 Tunnel 6m wide x 6m high
length of each................. 800m
Low Level
Outlet Tunnel
length.............................. 1400m
Low Level
Tunnel Gate
Shaft
depth............................... 100m
Spillway 6 gates each 15m x 15m x 19m high
Chute 400m long x 110m wide
with flip-bucket and plunge pool

There are a total of 18 separate tunnels and seven
shafts of different sizes and shapes on the project [2,
3]. Cross sections of the major tunnels and shafts on
the project are shown in Figures 2 and 3, respectively.

2 GEOLOGY AND SEISMIC CONSIDERATIONS
The site is located in the southern Piedmont of the
Central Cordillera of Luzon. This area has undergone
uplift and associated volcanism in the relatively recent
geologic past. The bedrock geology reflects the
geologic history and consists of crystalline igneous
and metamorphic rocks, overlain in the southern part
of the site by a younger sedimentary formation. The
primary igneous and metamorphic rock types at the
site are volcanic breccia with local diorite intrusions,
and metavolcanics/metasediments. These units are
closely jointed, with uniaxial compressive strength
ranging from 50 to 100 MPa. There are shear zones
and faults throughout the site that vary from a
centimeter to 2m thick. The Klondyke Formation,
which overlies the southern quarter of the site, is the
only sedimentary unit on the project. It is comprised
of interbedded conglomerate, sandstone, and siltstone,
which is highly variable and changes over short
distances from strong sedimentary rock to
unconsolidated sediments.

The weathering profile at the site is typical of a
tropical environment. The overburden is typically
residual soil underlain by progressively less weathered
rock until fresh rock is encountered.

Seismic considerations were an important factor in the
project design as the site is located in a region of
active tectonics. The site is located approximately
eight kilometers from a splay of the Philippine Fault,
which is the most significant crustal fault in the
Philippines. A comprehensive seismic hazard
evaluation concluded that there are no active faults at
the site and the deterministic maximum credible
earthquake motion is a moment magnitude (MW) 7.2
event. The project design earthquake generated by the
fault was calculated to have a peak ground
acceleration of 0.6g.

3 GROUND BEHAVIOUR AND ROCK SUPPORT
As shown on Figure 2, the tunnels are typically
inverted U-shaped and range in size from
approximately 3m wide by 4m high for the grouting
access adits to 10.4m wide by 16m high for the two
large diversion tunnels. In general, the rock is
relatively strong and the maximum cover over the
tunnels is approximately 150m. Consequently, in-situ
stresses are relatively low and the behavior of the
tunnel and shaft excavations is controlled by the rock
structure, rather than the stresses induced in the rock
surrounding the underground openings.

The rock support for the tunnels and shafts was
designed to support blocks and wedges of rock
bounded by discontinuities, such as joints and shear
zones in the rock mass. The support consisted of rock
bolts and fiber reinforced shotcrete, with steel sets and
shotcrete for the tunnel portals, shaft collars, and the
highly jointed and shear zones. The final linings of the
tunnels and shafts vary, depending on the final use
and stability requirements, as described below.

77
Figure 2 Cross Sections of he San Roque Tunnels t

4 TUNNELS
All sizes of tunnels were constructed using the drill-
and-blast method. Rock support was installed after
each round was excavated. In general, the design was
based on three rock support categories with the
selection of the category type made by the tunnel
supervisors and the rock mechanics engineer on site.
In many locations, it was necessary to develop rock
support designs on site to address unexpected
geotechnical conditions.

Rubber tired drill jumbos were used for probe hole
drilling, blast hole drilling, and rock bolt installation.
Load-haul-dump units and low profile rock trucks
were used for mucking, and a high capacity, robotic
arm boom was used for shotcreting. Epoxy-coated
rock bolts with epoxy resin grout were used in the
permanent water tunnels while Swellex rock bolts
were in the temporary tunnels. The shotcrete is
generally steel fiber-reinforced and most tunnel and
shaft final linings are reinforced concrete.

Figure 3
Cross Sections of the San Roque Shafts
(Powerhouse Shaft not shown)
78
The most complex tunnels are described in the
following paragraphs.

4.1 Diversion Tunnels

During construction, the Agno River was diverted
around the dam footprint through three diversion
tunnels: two large tunnels, 16m high by 10.4m wide
and one smaller tunnel, 6m high by 6m wide. Each
tunnel is approximately 800m long. The climate at the
site is divided into two distinct seasons: a dry season
and a wet season. During the dry season, the river
flows are typically 100 cubic meters per second (cms),
and can be passed through the smaller diversion
tunnel. However, the selected construction design
flood was 5600cms, which would require the full
capacity of all three diversion tunnels with the
upstream portals submerged to a depth of 50m. The
55m high cofferdam to provide this head was
incorporated in the upstream shell of the main
embankment.

Advantage was taken of recent advances in tunnel
design and construction methods to speed
construction, as the diversion tunnels were on the
critical path of the project. Changes from the original
design concept include:

Tunnel Shape:
Figure 4 Comparison of 1979 and 1998 Designs
For the Diversion Tunnels
The original design cross section for
the large tunnels was an ellipse shape with a full
concrete lining, as shown in Figure 4, indicating that
the designers were concerned about the stresses in the
rock surrounding the tunnels. Based on current
understanding of rock structure control of tunnel
stability, it was possible to simplify the tunnel shape
and facilitate faster tunnel construction by changing
the shape to an inverted-U, as shown in Figure 4. The
tunnels performed well with the revised shape.

Tunnel Lining: Advances in shotcrete materials and
equipment have resulted in shotcrete with a higher
compressive strength, greater toughness and ductility,
greater erosion resistance, better consistency and
faster application. To take advantage of these
improvements, the cast-in-place concrete lining of the
tunnels was eliminated and replaced by fiber-
reinforced shotcrete. The shotcrete, in combination
with rock bolts, also provides rock support. The
shotcrete lining was applied concurrently with tunnel
excavation, thus saving the six months that would
have been required to place the concrete lining. The
shotcrete lining performed well through several wet
seasons.

Tunnel Invert Lining: On the basis of an analysis of
the erosion resistance of the rock, using the stream
power/erodability index method [4], it was possible to
eliminate the invert lining of the tunnels. The unlined
inverts and shotcrete lining of the walls and crown
passed high diversion flows during the 1999 typhoon
season. The maximum velocity was approximately 16
m/sec which included massive amounts of large bed
load material, including boulders over 2 meters in
diameter. In fact, the unlined invert and shotcrete
behaved better and was less abraded than a short
section of reinforced concrete lining immediately
downstream of the inlet portal.

Tunnel Support: Tunnel advance rates were improved
by the use of SuperSwellexexpandable rock bolts.
Their use was possible as the diversion tunnels are
temporary structures, so long term corrosion
protection was not required.

Tunnel Construction Equipment: To maximize tunnel
advance rates, the latest drill-and-blast tunneling
equipment was used, including four-boom Tamrock
Maximatic drill jumbos, Normet shotcrete machines
and high lifts, and Wagner low profile trucks and
scoop trams.

4.2 Power Tunnel

A profile of the power tunnel is shown in Figure 5.
The tunnel is approximately 1300m long, with a
finished internal diameter of 8.2/8.5m and a maximum
head of 200m. From upstream to downstream, the
tunnel consists of: an intake, which is approximately
100m below normal pool level; a low pressure sub-
horizontal tunnel with a gate shaft; a surge shaft at the
downstream end of the low pressure tunnel; a drop
shaft; and a high pressure tunnel.

79

Figure 5 Profile of Power Tunnel

The tunnel is lined with reinforced concrete except for
200 m at the downstream end, where it is steel lined.
The steel lining is required only where the internal
water pressure exceeds the minimum in-situ rock
stress. The concrete lining is required to provide
erosion resistance and to reduce hydraulic head losses
as water tightness is provided by the low permeability
rock mass and by grouting of the more permeable
zones. Extensive packer testing, hydrofrac testing and
seismic refraction testing was carried out from the
ground surface and from within the tunnel to confirm
rock mass properties. The design and type of lining
(concrete and steel) of the tunnel were based on the
results of these tests and on current pressure tunnel
design criteria [5].

5 GROUTING GALLERIES
To compress the construction schedule, a minimum
amount of grouting was done from the invert of the
core trench of the dam. This included only the blanket
grouting and the first stages of the grout curtain. The
remainder of the curtain grouting was done from
grouting galleries, which were constructed on the
centerline of the dam at varying depths below the
surface of the core trench. Access to these galleries for
construction, for grouting and for drainage was
provided by four access adits. The grades of the
galleries and adits are a maximum of 12%. The portals
for the adits are at the downstream toe of the
embankment dam, just above maximum tailwater
level, as shown on Figure 1.

6 SHAFTS
Figure 3 shows the relative sizes of the various shafts
on the project. The excavated sizes of the shafts vary
from the 28m wide x 80m long powerhouse shaft to
23m excavated diameter surge shaft to the 1.2m
diameter grouting gallery vent shafts. Like the
tunnels, the stability of the shafts is controlled by the
structure in the rock, rather than by the stresses
induced around the shafts. The collars of all shafts are
in weathered rock and transitioned to fresher rock at
depth. The larger shafts required full concrete or
structural steel support in the top few meters in the
weathered rock, while a combination of shotcrete and
rock bolts was used in the better rock at depth. The
final lining of the large, circular shafts was reinforced
concrete. The vent shafts were raise bored and then
lined with steel pipes which were concreted in place.
The powerhouse shaft was lined with shotcrete, as
described in further detail below.

Most of the shaft excavation was done by the raise-
bore-and-slash method. The two vent shafts were
excavated by raise boring alone, and the uppermost
approximately 15 m of the surge shaft was
conventionally sunk because the rock mass quality
was so poor and the raise bore hole would have been
unstable. The excavations for the shafts were
coordinated with tunnel construction to provide access
to the top and bottom of each shaft so that raise-
boring-and-slashing could be used. Usually, the raise
bore drill was set up on the shaft centerline, and a
0.3m pilot hole was drilled down to the tunnel below.
80
The 1.2m, 1.8m or 3m diameter raise bore head was
then attached in the tunnel below and the pilot hole
was reamed upward. Following this the shafts were
slashed out by blasting to the full diameter and the
excavated rock dumped down the raise bored shaft so
the muck could be removed from the tunnel below.
Rock support was installed after each round was
excavated. Final lining followed directly after
excavation.

Design of the Surge Shaft was challenging because of
the large diameter and the rock conditions in the upper
45m where the rock was poorly cemented Klondyke
conglomerate. The Surge Shaft excavation is shown in
profile on Figure 6.

The rock support in the conglomerate consisted of
400mm thick concrete rings, 1.2m high, spaced 2m,
vertically. The gaps between concrete rings were
shotcreted to prevent raveling. The lower half of the
shaft was supported by shotcrete and rock bolts. A
final 1m thick lining of concrete was placed after
excavation.

The powerhouse shaft also offered particular
challenges. From a geotechnical perspective, the
powerhouse is similar to an underground powerhouse
except that it does not have a rock arch roof. It
consists of a vertical-sided, rectangular rock cut, 85m
long by 28m wide and 46m deep. A typical section
through the powerhouse and tailrace tunnels is shown
in Figure 7 (on the next page).

The powerhouse has a concrete roof, which spans the
excavation, shotcrete covered rock walls in the upper
half of the excavation, and concrete walls in the lower
half. The excavation houses three 115 MW
turbine/generators. Three penstocks enter the
powerhouse on the north side of the excavation and
three tailrace tunnels exit the powerhouse on the
opposite side. The roof beams and the substructure
were used in the design to assist with rock support for
the high seismic loading.

Large surface excavations were required to reach the
top of the powerhouse shaft. The shaft excavation is
entirely in diorite, which is some of the best rock at
the site. The tailrace tunnels were excavated before
the powerhouse shaft so that two raise bores could be
drilled in the center of the shaft before the bulk
excavation started. The shaft was excavated from the
top down with the blasted rock being dropped down
the raise bores and removed through the tailrace
tunnels. Rock support consists of 46mm rock dowels
spaced at about 2.65m by 2m, as shown in Figure 7.
Fiber reinforced shotcrete provides rock support
between the dowels. The rock support was installed as
the excavation proceeded. As with all underground
excavations, the site geologists mapped the excavation
and the on-site rock mechanics engineer analyzed the
data and adjusted the rock support to suit the rock
conditions as they were exposed in the excavation.

7 CONCLUDING REMARKS
Re-engineering of the 1970s designed tunnel and
shaft excavations, and adopting a design-build
approach, contributed significantly to the economic
success of the project. The schedule for the
construction of the many tunnels and shafts had to be
closely coordinated with the critical path items (the
dam, the spillway and the powerhouse) so that the
entire project could be completed on schedule. Some
general observations of the geotechnical design
approach include:
Figure 6 - Section of Surge Shaft Excavation

The layout and shapes of the tunnels and shafts
were optimized to take advantage of our current
understanding of rock mechanics principles and
the capabilities of modern construction
equipment;
The use of database and analytical computer
programs for the evaluation of discontinuity data
and evaluation of tunnel and slope stability
allowed the components of the design to be
completed relatively quickly, during construction,
and modified as additional mapping data became
available from site;

81

Construction has been greatly accelerated by the
use of modern construction equipment and by the
use of modern rock support materials, particularly
fiber reinforced shotcrete, rock bolts, and rock
anchors;
The raise bore and slash method of shaft
excavation provides an efficient and effective
excavation method for a wide range of shaft sizes;
The design-build approach works well in
conjunction with modern rock engineering design
and construction techniques; however, it has
proven essential to have an experienced rock
mechanics design engineer on site to ensure that
appropriate rock support is installed to suit the
actual rock conditions that are exposed.

8

Figure 7 Typical Rock Support of Powerhouse Shaft
9 REFERENCES
1. Kessler, K. 2002, Design and Construction of San
Roque Dam Project ASDSO Southeast Regional
Conference, Lake Lanier, GA.

2. OConnor, E. 2001, Diversion Schemes for the San
Roque Multipurpose Project Waterpower XII, Salt Lake
City.

3. Humphries, R., E. OConnor, L. Gertler,, W. Warburton,
J . Daly, M. Funkhouser, 2001 Rock Engineering at the
San Roque Multipurpose Project Waterpower XII, Salt
Lake City.

4. Annandale, G., 1995, Erodability, J ournal of Hyd.
Res., Vol.33, No.4.

5. Merritt, A. 1999 Geologic and Geotechnical
Considerations for Pressure Tunnel Design, Geo-
Engineering for Underground Structures, ASCE
Geotechnical Special Publication No. 90.
82
The Toulnustouc River I ntake Tunnel

Guylaine Dubois, P.Eng., EBC inc., Quebec, QC, Canada

ABSTRACT: EBC inc., a general contractor of Quebec, Canada, was involved in the excavation of the
water intake tunnel for the Toulnustouc River hydroelectric project. The tunnel, 13 meters high by 11
meters wide, was driven full face. EBC used innovative equipment and techniques to manage a successful
project.

1. INTRODUCTION

EBC, the second largest construction company
in Quebec, is a general contractor. EBC is
involved in major civil earthwork projects as
well as in industrial and commercial building
constructions. EBC can count on a dedicated
workforce of 200 permanent employees.
In October 2001, EBC was the lowest
bidder for the excavation of an 8.3 km tunnel.
The tunnel was part of a major hydroelectric
project in Quebec, the Toulnustouc River. The
owner, Hydro-Quebec, is producing 536
MegaWatts of electricity with this project. This
great hydroelectric project was inaugurated in
August 2005 by the provincial Prime Minister,
Mr. J ean Charest. The total cost of this project is
estimated at $804 million.
Located 750 kilometers northeast of
Montreal, the project includes a 77 meter high
dam, a reservoir of 235 square kilometers and an
intake tunnel of 9.8 kilometers.

Fig. 1. Project location map.

2. TUNNEL ADVANCE

The contract bid for by EBC was to drive an 8.3
kilometers long, 13 meters high by 11 meters
wide tunnel.
EBC chose to drive this tunnel full face.
The normal way of driving a tunnel of that size
would be in two steps. First, the top of the
tunnel would be driven with a jumbo, followed
by benching the bottom section with
conventional surface drills. In construction the
tendency is to believe that full face advance is
more expensive than benching. At first, it looks
that way, but benching is not efficient and it is
time consuming, because of the need to install
the ventilation and services twice. This makes it
difficult to meet the schedule, which is probably
the most important aspect of a construction
project.
EBC chose to use the full face method,
but one problem still remained: finding a jumbo
capable of driving a tunnel that size in a full face
mode.
EBC approached the Sandvik Tamrock
team to find out if it would be possible to modify
the existing Tamrock Axera T-12 jumbo. The
reach for the T-12 is 11.9 meters and EBC
needed at least 13 meters. The Tamrock team
was successful in adapting this model of jumbo.
EBC was the lowest bidder on the contract, and
placed an order for three Tamrock Axera T12-
315 fully computerized jumbos.

83

Fig. 2. Sandvik Tamrock Axera T-12 jumbo.

EBC was the first owner of a T-12 in
North America and the first one in the world for
a T-12 capable of reaching 13 meters.
These jumbos offer other advantages
besides their full face capability. The T-12
jumbos are fully computerized. The drilling
layout is inserted in the jumbos computer; the
jumbo is stationed at the face, aligned by tunnel
laser and, finally, the operator pushes the start
button. The computer is then in charge of the
drilling. The operator can overrule the computer
and drill manually, but after seeing the final
results, anyone would prefer to leave the
computer to do the job.
The advantages of the computerized jumbo
could be summarized as follows:

Less over break;
All drilling data are kept in the jumbos
computer;
Penetration rate is maximized;
Time between holes is reduced;
The face does not have to be marked up
for drilling;
Blasting results are improved;
Less manpower for the drilling
operation.

EBC started the project with two headings,
one directly in the future reservoir and the other
one, an access tunnel, 6 kilometers downstream.
After driving 507 meters in the access tunnel,
EBC reached the main tunnel alignment, and
could then advance in three headings.

Fig. 3. Layout of tunnels.

EBC drove one heading from the
reservoir for 3.017 kilometers. The access
tunnel was driven for 507 meters. Two headings
were then advanced 2.485 kilometers east and
2.300 kilometers west, for a total of
8.3 kilometers.
EBC also proposed to the owner a new
shape for the tunnel. Two major aspects would
be improved by adopting this new shape:
Hydraulic capacity of the tunnel;
Rock stability.

Fig. 4. Base design and EBC proposal.

The new design was approved by the
owner. A typical round was 174 holes, 57 mm in
diameter and 5.4 meters deep. The cut was a
normal Canadian cut, which works well in the
hard rock of the Canadian Shield, in which the
tunnels were driven.

84

Fig. 5. Full-face blast design.

There was enough room to drill a V cut,
which EBC did at times, but during most of the
operation, EBC put two jumbos in the face to
speed up the cycle. With a V cut and two
jumbos in the face, there is more risk of boom
interference.

Fig. 6. Twin jumbos in the heading.

The drill holes were all drilled at angles,
so the holes of the next round would not be
collared in the bootlegs of the last round.
For blasting, EBC used Dyno Nobel
products and long delay detonators. Blasting
vibration was monitored at 30 meters distance
and was limited to 150 mm/s Peak Particle
Velocity.

The tunnel mucking was done with a
Caterpillar 988G wheel loader and eight
Caterpillar 773, 50 ton off-highway dump trucks.
A million cubic meters (1 000 000 m
3
) of rock
were excavated by EBC in this tunnel. A
Caterpillar 235C excavator was used for scaling
but manual scaling was done afterwards to
insure a safe environment.

3. GROUND SUPPORT

The ground was supported with 4 and 6 meter
hollow-center mechanical bolts. The drilling
was done with the Tamrock T-12 jumbo on
manual mode. Bolts were put in place using a
basket, final torque was done manually. Then
flexible wire mesh was put in place, all the way
up to the face. When far enough from blasting,
all mechanical bolts were checked again for
torque and grouted in place.
EBC also encountered poor ground
conditions; after discussion with the owner, EBC
poured in-place concrete over 100 meters of
tunnel with a semi-circular form.

4. VENTILATION

EBC used eleven fans of 2.1 meter diameter for
ventilation. This was a push-pull system. With
a fan at both ends of the vent tube, EBC was able
to push fresh air to the face and evacuate
blasting fumes efficiently.

Fig. 7. Ventilation layout.

85
In fact, EBC put two fans at the entrance
of the tunnel and two near the face.

EBC used 3 meter diameter flexible
ventilation tube, manufactured by ABC Canada.
At the time, this was the biggest diameter of
ducting that ABC had ever manufactured.
Using ventilation tube this big is an
advantage, reducing the pressure loss, and
minimizing the need for booster fans, provided
the headroom is available. In a 13 meter high
tunnel, headroom is not a problem. In fact, at
one heading, EBC was able to ventilate 3
kilometers of tunnel without a booster fan. This
was another first for EBC.

5. CONCLUSION

EBC, its management team and all its employees
involved in this project combined their efforts to
make it a big success.
Innovation and creativity were major
factors in this success. The project was
delivered on time and according to EBCs initial
budget.

86

1 INTRODUCTION
Toulnustouc project is located in the north-eastern
part of province of Qubec, Canada (Figure 1). The
526 MW hydroelectric facility was constructed in four
years beginning in November 2001.
Fig. 1. Project location.

The project includes:

- a 77 m high concrete-face dam on the
Toulnustouc River and a nearby 45 m high
dike;
- a 10 km long, 13 m x 11 m unlined headrace
tunnel operating under a maximum static head
of 183 m;
- two 138 m long, 8 m diameter, steel lined
pressure penstocks;
- a 520 MW surface powerhouse with two
Francis units;
- a 300 m long tailrace channel.

Impounding took place in February 2005 and
tunnel filling was carried out over a period of 15 days
from March 22 to April 5, 2005. The facility was
commissioned in J uly 2005.
Figure 2 presents the general layout of the project
and a longitudinal profile of the tunnel including the
principal geological features encountered. The
maximum reservoir level is 301.75 m and the
minimum water level in the tailrace channel is
127.3 m.

Toulnustouc pressure tunnel leakage estimation, filling,
instrumentation and control
A.J . Rancourt, C. Chartrand
RSW Group, Montral, QC, Canada
A. Whalen, D. Bergeron
Hydro-Qubec quipement, Montral, QC, Canada
ABSTRACT: The Hydro-Qubec Toulnustouc hydroelectric project is located approximately 150 km north of the city of
Baie-Comeau, Qubec, and was commissioned in J uly 2005. The water conveyance system to the 520 MW surface
powerhouse is assured by a 10 km long, 13 m x 11 m unlined headrace tunnel and a 175 m high vertical surge shaft,
operating under a maximum static head of 180 m.

In-situ stress measurement tests results indicated the presence of local minimum principal stresses lower than the water
pressure in the tunnel. Those minimum measured values were also lower than the stress predicted by topographic rock
cover criteria. Based on stress measurements results, the penstock steel liner behind the powerhouse was lengthened and, at
a second location, a pressure relief curtain was constructed in the low stress zone to control the propagation of potential
local hydraulic jacking and also to control pore pressures in the near surface zone that might destabilize the overburden
located above the tunnel. The paper presents the leakage estimation, the filling procedure and the monitoring program that
was carried out in order to closely follow and control progressive rock mass saturation and total tunnel leakage. The tunnel
was filled in March 2005 and has performed satisfactorily since then, with acceptable water losses within the predicted
range.
87

P
M
3000
5540 000N
266 000 E
P
M
1000
P
M
2000
264 000 E
5538 000N
268 000 E
P
M
4000
P
M
5000
262 000 E
260 000 E
258 000 E
256 000 E
P
M
9000
P M
0
P
M
6000
P
M
7000
P
M
8000
P
M
9275,745
Fig. 2. Toulnustouc hydroelectric project plan and tunnel section (vertical x 20).
88

The main purpose of this paper is to present the
assessment of expected leakage based on tunnel
geology and the existence of two low in-situ stress
areas along the tunnel, one near the powerhouse and
the other around station 7700 (Rancourt et al., 2006).
Also the paper presents the tunnel filling procedure
and the monitoring program along with the principal
results. Areas of low in-situ stress measurements are
located on Figure 2. For the low stress zone near the
powerhouse, the water pressure in the tunnel is around
1.8 MPa and the steel liner was lengthened to the
point where sufficient in-situ stresses were measured
with safety factor above 1.3. The other zone is located
2.36 km upstream of the powerhouse around station
7700, with a water pressure of approximately 1 MPa
and safety factor around 1.0. In this area, the low
stresses were observed on a 250 m long tunnel section
where the ground level is below the reservoir level. A
relief curtain was constructed to avoid high pressure
water reaching surface and destabilizing the
overburden.
Geological and geotechnical information collected
during construction was utilised to evaluate total
expected water leakage and to design the pressure
relief curtain. A detailed filling procedure was carried
out that included careful monitoring of all leakage
sources. Total leakage was calculated using falling
head tests at the intake, which measured the response
of the whole tunnel. Also the results of leakage
measurements from relief holes in station 7700 area
and from Adit No. 1 near the powerhouse are also
presented.
2 GEOLOGY AND STRUCTURE

The Toulnustouc River is a tributary of the
Manicouagan River which drains a large basin on the
north shore of the St-Laurence River. The rocks are of
Precambrian age, and are composed of mixed grey
and pink gneisses folded and cut by granitic and mafic
dykes. The rock mass is cut by at least three joint
families.
The tunnel was excavated in a good to very good
quality rock mass with GSI values around 80 % and Q
values always above 10 which will be referred to as
Class I rock type in the following text. As shown on
Figure 2, the tunnel intersected three important
geological features. These are a sub-horizontal 3 m
thick diabase dyke located between station 890 and
station 990, a 20 - 30 m thick sub-vertical shear zone
between station 4340 and station 4410 and a sub-
vertical shear zone around station 7800. Several other
narrow, widely spaced shear zones (0.1 1 m thick)
were also observed in the tunnel. The rock of the
tunnel was classified in two different rock mass types,
Class I which is the normal rock mass and Class II,
which represents the fractured rock mass. The soft
diabase dyke was completely concreted and does not
influence tunnel leakage. For the purpose of leakage
analysis, the Class II rock type is applied to two zones
of fractured rock mass, the shear zone between station
4300 and station 4400, and the shear zone between
station 7750 and station 7850. Table 1 gives the basic
properties for each rock mass class. It can be seen on
Table 1 that the Class I rock type is dominant while
the Class II rock type is only encountered along 2 %
of the tunnel.

Table 1. Basic rock mass properties
Rock
mass
type
Percentage
along the
tunnel (%)
Geological
Strength
I ndex
(GSI) (%)
Youngs
modulus
(GPa)
J oint
spacing
(m)
Class I 98 75 - 85 20 - 40 0.5 3
Class II 2 40 - 60 5 15 0.2 0.5

3 IN-SITU PERMEABILITY
MEASUREMENTS

Some seepage inflows were encountered during
tunnel construction. These were all associated with
shear zones and persistent joints. Seepage quantities
were generally small, reflecting the low porosity and
low permeability of the rock mass. However the shear
zones around station 4400 and station 7800 have both
shown a significant amount of inflow during
construction. In particular, the inflow at station 7800
was around 8 l/s during 3 days following the
excavation, after that time the inflow decreased
rapidly suggesting that water storage was not
renewed. Those zones were assumed to be more
permeable based on the assumption (Bremen and
Tognola, 2002) that there is a relationship between
zones of water inflow and zones of water leakage.
Permeability was measured using conventional
Lugeon tests and the hydraulic jacking stress
measurement curves. A total of 80 permeability
values were calculated in both critical areas. Lugeon
tests and hydraulic jacking tests carried out
respectively from the ground surface and the tunnel
level, between station 7400 and station 8400, and also
hydraulic jacking tests carried out in the tunnel for the
steel liner length, allowed an estimate to be made of
the unjacked permeability. Table 2 shows the
principal hydrogeological parameters for both rock
mass types.

89

Table 2. Rock mass hydrogeological characteristics
Permeability (m/s)
Rock mass type Class I Class II
High 7 X 10
-7
3 X 10
-6

Low 5 X 10
-8
1 X 10
-6

Average 1 X 10
-7
2 X 10
-6

No. of tests 44 6

Also jacked permeabilities for low stress zones
were estimated using the hydraulic jacking test curves
as shown on Figure 3, and Table 3 presents the
approximate values. It can be noted that the jacked
permeability is around one order of magnitude greater
than the unjacked permeability in Table 2.

Fig. 3. Typical hydraulic jacking curve and permeability
interpretation.

Table 3. J acked rock mass permeability
J acked
permeability (m/s)
High 5 X 10
-5

Low 5 X 10
-7
Average 3 X 10
-6
No. of tests 32

4 EXPECTED LEAKAGE

Based on the rock mass permeability estimates
presented above, tunnel leakage was calculated using
the equation given by Fernandez (1994) to estimate
the water outflow (or inflow) per unit length of a
circular pressure tunnel, which is given by:

( )
b
h
h h k
Q
i
0
0
2
ln
2
=

(1)
This relation is similar to the one given by Goodman
et al. (1965) for the estimation of water inflow. In
Eq. 1, k is the equivalent continuum rock mass
permeability, h
i
internal water pressure, h
0
elevation
of the original water table and b the tunnel radius. The
assumed initial water table along the tunnel is shown
on the tunnel section on Figure 2.
In order to estimate total tunnel leakage, the
tunnel was separated into several sections not
exceeding 1000 m long. For each section, initial water
table and tunnel pressure were evaluated and leakage
was calculated using Eq. 1. Two zones of Class II
rock type were considered, the shear zone between
station 4350 and 4400, and the shear zone between
station 7750 and 7850. Also, total leakage was
estimated assuming jacked conditions according to in-
situ stress measurements results presented in Rancourt
et al. (2006) which identifies two areas of low stress
with potential jacking problems. The two low stress
areas are located between station 7600 and 7800 and
near the surface powerhouse, around station 10260.
Table 4 presents estimated total tunnel leakage for
unjacked and jacked conditions.

Table 4. Total estimated tunnel leakage
Total leakage (l/s)
Low High
Unjacked conditions 15 85
Local jacked conditions 125 1970

Unjacked leakage values presented on Table 4 are
in accordance with the recommendations of Merritt
(1999) about acceptable tunnel leakage which is 10
l/s/km. With jacked conditions, most of the leakage is
associated with the low stress areas (station 7700 and
station 10260).
Figure 4 shows the expected average unjacked
leakage along tunnel axis (negative leakage values
represent water inflows). It can be seen on the figure
that in the first part of the tunnel, from the intake to
around station 4350, water inflows are expected based
on the initial assumed water table elevation that is
higher than tunnel pressure. Further downstream, the
pressure is gradually increasing and the initial water
table is lower (see Figure 2), so outflow appears
which increases toward the powerhouse.

-100
-50
0
50
100
150
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000
Stations (m)
W
a
t
e
r

l
e
a
k
a
g
e

(
l
/
s
e
c
)
Fig. 4. Average unjacked leakage along tunnel axis.
90

Based on these results and on the possibility of
having local jacked conditions and excessive leakage
that might necessitate tunnel shutdown, a filling
procedure and a monitoring program was
recommended and carried out.

5 FILLING PROCEDURE AND CONTROL

The initial filling was done slowly in a controlled
manner in order to limit deformations within rock and
liners and also to allow progressive saturation of the
surrounding rock mass and thus limiting seepage
forces.

5.1 Rate of Filling

The filling procedure was planned as shown in Figure
5, with compulsory stops and waiting times. The rate
of filling was chosen in accordance with
recommendations given by Deere (1983) as follows:

- A first stop at el. 180 (60 m rise), after a 5m/h
filling rate, with a 12 hour delay to check
behaviour near the powerhouse. The PM 7700
low stress zone is then not flooded yet.
- A second 60 m rise (to el. 240 m), at a 2.5m/h
filling rate, followed by a 48 hour delay. At
this point, the water pressure in the tunnel, at
the PM 7700 low stress zone, is 0.6 MPa,
which corresponds to the lowest local
minimum stress measured in that area.
- Five succeeding 10 m increases, at the same
filling rate, separated by 48 to 72 hour
observation delays.
-
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Time (days)
W
a
t
e
r

l
e
v
e
l

i
n

t
h
e

t
u
n
n
e
l

(
m
)
Fig. 5. Rate of filling.

The tunnel behaviour during the filling procedure
was such that waiting delays were never longer than
the minimal planned values. Water level in the tunnel
was controlled by pressure readings at the
powerhouse.

5.2 Leakage Control in the Low Stress Zone (STA
7400 STA 8400)

On the basis of the hydraulic jacking tests, this zone
showed low minimum in-situ stress between station
7700 and station 7900 and, locally between station
8100 and 8300 (Rancourt et al., 2006). To control
pressurized water from getting near the surface and
destabilizing rock or overburden, 173 boreholes of 75
mm diameter were drilled between station 7400 and
station 8400. The array of drill holes is composed of
two rows of holes, one sub-vertical and the other sub-
horizontal. Together, the two rows form an umbrella
between the tunnel and the ground surface as shown
in Figure 6 and 7. The distance between holes varies
from 5 m to 40 m depending on the location. A higher
density of holes was planned around the shear zone
region, which presented low in-situ stress (Figure 7).
A trench was excavated above the tunnel and between
station 7650 and station 7800. In that area, the sub-
vertical shear zone facilitated the erosion of the rock
surface, which resulted in a locally thicker overburden
layer, and thus in a shorter path for the water to reach
the surface.
Thirteen of the surface boreholes were
instrumented with two electronic piezometers, one in
the rock mass and one in the soil. The information
was used to follow ground saturation and evaluate
stability during each step of the filling procedure.

PRESUMED ROCK
GROUND ELEVATION
OFFSET (m)
E
L
E
V
A
T
I
O
N

(
m
)
RELIEF HOLES
TOULNUSTOUC RIVER
PRESSURE TUNNEL
184.6
-100 -80 -40 -60 -20 140 0 20 40 60 100 80 120 160 180 200 220
240
170
180
190
200
210
220
230
250
260
270
280
290
300
Fig. 6. Pressure relief holes between station 7400 and
station 8400.

91
8 6
P
M
77 10
P
M
7720
PM
77 3 0
PM
77 4 0
P
M
77 50
PM
7 76 0
PM
777 0
PM
778 0
P
M
7 7 90
P
M
781 0
PM
7 82 0
PM
7 83 0
P
M
7 840
P
M
7 85 0
P
M
7 86 0
P
M
78 70
PM
7 88 0
7471 . 5
7 4 91. 5
75 11 . 5
75 31. 5
75 71. 5
75 91 . 5
7 6 11. 5
7 6 31. 5
1
2
3
4
5
6
7 8 9
1 0
1 1
1 2
1 3
1 4 1 5
1 6 1 7 1 8
1 9
2 0
2 1
2 2
2 3
2 4
2 5
2 6 2 7
2 8
2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9
4 0
4 1
4 2
4 3 4 4
4 5
4 6
4 7
4 8 4 9 5 0 5 1
5 2
5 3
5 4
5 5 5 6
5 7
5 8
5 9
6 0 6 1 6 2 6 3 6 4
6 5
6 6
6 7
6 8
6 9
7 0 7 1
7 2
7 3
7 4 7 5 7 6 7 7
7 8
7 9
8 0 8 1
8 2
8 3
8 4
8 5
TF-06
TF-05
TF-03
TF-08
TF-07
TF-13
TF-10
TF-09
TF-11
TF-12 7.6 m
TF-14
TF-15
TF-16
TF-17
18.8 m
20.6 m
2 6 0 0 0 0 E
2 6 0 2 0 0 E
2 5 9 8 0 0 E
5 5 3 7 6 0 0 N
P . I . - 3 1
210
220
230
240
250
260
270
280
230
240
250
230
240
250
260
270
260
270
280
270
280
5 5 3 7 8 0 0 N

Fig. 7. Pressure relief holes location in the low stress zone between station 7400 and station 8400.

6 LEAKAGE MONITORING

As it is often the case with complex geotechnical
situations (Baker, 1991), the filling of the pressure
tunnel was carried out in accordance with a detailed
procedure that had been prepared in conjunction with
a Hydro-Quebec technical committee formed to
oversee all aspects of the reservoir and tunnel filling
operations. The monitoring program included:

-Flow meter in adit 1 near the powerhouse.
-Relief holes curtain and flow meters along low stress
area (station 7400 to station 8400).
-Falling head test for the whole tunnel.

6.1 Leakage at the Powerhouse

Leakage measured at the powerhouse corresponds to
the volume of water collected by the drainage system
inside the powerhouse plus the leakage measured at
the Adit No. 1. Figure 8 illustrates the total measured
leakage and the reservoir level versus time.

Fig. 8. Total measured leakage at the powerhouse and
reservoir level along time.

It can be seen on Figure 8 that water leakage
correlates with the water elevation in the reservoir.

6.2 Leakage in the Low Stress Zone

Total leakage coming out of the relief holes was
measured during and after filling of the tunnel.
Results are shown on Figure 9 where the total flow
and the water pressure are plotted with time.

290
291
292
293
294
295
296
297
298
299
300
1
4
/1
2
/
2
0
0
5
2
1
/1
2
/
2
0
0
5
2
8
/1
2
/
2
0
0
5
0
4
/0
1
/
2
0
0
6
1
1
/0
1
/
2
0
0
6
1
8
/0
1
/
2
0
0
6
2
5
/0
1
/
2
0
0
6
0
1
/0
2
/
2
0
0
6
0
8
/0
2
/
2
0
0
6
1
5
/0
2
/
2
0
0
6
2
2
/0
2
/
2
0
0
6
0
1
/0
3
/
2
0
0
6
0
8
/0
3
/
2
0
0
6
1
5
/0
3
/
2
0
0
6
2
2
/0
3
/
2
0
0
6
2
9
/0
3
/
2
0
0
6
0
5
/0
4
/
2
0
0
6
1
2
/0
4
/
2
0
0
6
Days
R
e
s
e
r
v
o
i
r

l
e
v
e
l

(
m
)
38
43
48
53
58
W
a
t
e
r

l
e
a
k
a
g
e

(
l
/
s
)
Reservoir level ( m)
Flow (l/s)
92
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
0 5 10 15 20 25 30
Time(days)
T
o
t
a
l
f
lo
w

(
l/
s
e
c
)
245
250
255
260
265
270
275
280
285
290
295
W
a
t
e
r

le
v
e
l
in

t
u
n
n
e
l
(
m
)
Total flow
Water level

Fig. 9. Total measured leakage in the relief holes at the low
stress zone versus time.

A maximum of 24 holes (14 vertical and 10
horizontal) gave water during the days following
tunnel filling. Those holes are located between station
7500 and station 7770, with a local minor flow in a
sub-vertical hole at station 8091. Ten of these holes
presented small leakage and low pressure, and the 14
others, three month after tunnel filling, have shown a
stabilized total leakage of around 6 l/sec.

6.3 Total Tunnel Leakage

Total tunnel leakage was determined during the filling
procedure by measuring the falling rate of the water
level in the tunnel with the intake gate shut. Results
are shown in Figure 10.

0
10
20
30
40
50
60
70
80
240 250 260 270 280 290 300
Water level in the tunnel (m)
A
v
e
r
a
g
e

l
e
a
k
a
g
e

(
l
/
s
)
Fig. 10. Falling head test results for the whole tunnel
according with water level in the tunnel

Figure 10 shows that total tunnel leakage at
maximum reservoir level (301.75 m) would be around
80 l/sec. The 40 l/sec flow measured at Adit No. 1 and
the 6 l/sec flow measured in the relief holes are
included in the total value. The total measured water
leakage agrees with the expected unjacked leakage
presented on Table 4, for the high permeability range.
It is also interesting to note that according to
authors such as Bouvard and Pinto (1969) and Benson
(1987), the amount of leakage should be high in the
beginning of filling due to re-establishment of the
groundwater table. However, this phenomenon is not
observed in the data from figure 10. One explanation
would be that the rock mass porosity and permeability
are very low and that the groundwater table was not
totally drained during construction.

7 CONCLUSIONS

The design of the Toulnustouc project was influenced
by the low confining stress areas discovered during
construction and the potential problems that could be
encountered during filling in these zones. Total
leakage was estimated in order to set acceptable
seepage limits and to establish a decision tree during
filling. But for a 10 km long tunnel, the initial leakage
estimates were highly dependent on parameters
chosen from the investigation results. Thus a careful
monitoring program was undertaken for tunnel initial
filling. The main results of the monitoring and control
program are as follows:

The total leakage from the tunnel is in the upper
range predicted for unjacked conditions. From an
operational view point the measured leakage rates
are insignificant and, despite the presence of in-
situ minimum stresses slightly lower than the
water pressure, there is no evidence of hydraulic
jacking.
There will always remain a possibility that with
time, local rock mass deformations and associated
permeability changes will occur due to water
pressure redistribution.
Filling and monitoring procedures have provided
a large quantity of data that were very useful for
decision making and in the evaluation of the
overall tunnel performance during first filling and
during the first year of operation.

ACKNOWLEDGEMENTS

The authors would like to acknowledge Hydro-
Quebec for the permission to use the data. Also
special thanks to Dr. R. P. Benson and M. D. K.
Murphy for their review of this paper and their very
constructive comments.

REFERENCES

1. Baker, D.G. 1991. Wahleach power tunnel monitoring,
Proc. 3rd Int. Symp. On Field Measurements in
Geotechnics, Oslo, Norway, pp. 467-479.
2. Benson, R.P. 1987. Design of Unlined and Lined
Pressure Tunnels. Canadian Tunneling, 1987/1988, pp.
37-65.
93
3. Bouvard, M., and N. Pinto. 1969. Amnagement
Capivari-Cachoeira, tude du puits en charge. La
Houille Blanche, 7, p.747-760.
4. Bremen, R. and F. Tognola. 2002 Evaluation of
leakage in a partially unlined pressure tunnel at
Casecnan. Hydropower & Dams, 1, pp. 74-78.
5. Deere, D.U. 1983. Unique Geotechnical Problems at
Some Hydroelectric Projects. 7th Panamerican Conf.
on Soil Mech. and Foundation Eng., Vancouver,
p.865-888.
6. Fernandez, G. 1994. Behavior of Pressure Tunnels and
Guidelines for Liner Design. J . of Geotech. Eng., 120,
p.1768-1789.
7. Goodman, R.E., D.G. Moye, and A. van Schalkwyk,
and I. J avandel, 1965. Ground water inflows during
tunnel driving. Eng. Geol., 2, pp. 39-56.
8. Merritt, A.H. 1999. Geologic and Geotechnical
Considerations for Pressure Tunnel Design. AmSoc
Testing & Materials, Geotechnical Special Publication,
90, p.66-81.
9. Rancourt, A.J., D.K. Murphy, A. Whalen, and R.
Benson. 2006. Extensive stress measurements program
at the Toulnustouc hydro-electric project Quebec,
Canada. Proc. of the Int. Symp. on in-situ rock stress,
Trondheim, Norway.

94
1 SUCCESS FACTORS IN TUNNELLING
The most critical success factors in tunnelling
business are cost, schedule and quality. Target is to
stay ahead of schedule and below budget. The
tendency is that the round cycle times should be
minimized and the utilization rates of the equipment
should be maximized especially at construction work
sites. Target is also to ensure superior quality in the
projects and their outcome. Achieving accurate
contours in tunnelling operations improves the total
economy of the lined tunnel projects dramatically.
In recent years more emphasis has been put to
environmental and safety issues as well. On one hand,
tunnels offer ways to protect sensitive landscapes as
well as reduce disruption, noise and vibration, which
is especially important in urban areas. Tunnels are
increasingly being demanded in sensitive areas
despite the fact that even though the cost of tunnel
construction has come down, tunnels are still more
expensive than surface roads. For example in
Australia the development of Sydneys motorway
system has relied heavily on the use of tunnels since
there is a need to reduce the environmental impacts of
motorways and to address the concerns of the local
communities. [1]
On the other hand, there is also a need to build the
tunnels in a safer and greener manner. A couple of
years ago the Scandinavian tunnelling industry was hit
by several high profile environmental disasters - like
Norways Romeriksporten Tunnel and Swedens
Hallandss Tunnel. Now more emphasis is put also on
environmental assessments of the construction
processes to avoid such disasters. [2]
In the future, we continue to see development
towards increased performance and stricter demands
on the total quality of tunnel, care of the environment,
vibration and noise levels of excavation. Also constant
measurement, control and documentation are required
from the contractor. All this puts pressure to the
features of the equipment.
2 NEW INTELLIGENT J UMBOS TO SUPPORT
THE TUNNELLING PROCESS
2.1. Equipment needs to be used for many different
tasks
The multifunctional use of drilling jumbos has been
increasing lately. The jumbos should be able to be
used e.g. in:

New intelligent drilling jumbos for accurate, fast and cost-efficient
tunnelling
Ulla Korsman
Sandvik Mining and Construction Oy, Tampere, Finland
Pekka Nieminen
Pekka Salminen

ABSTRACT: The most critical success factors in the tunnelling business are cost, schedule and quality. Round cycle times
have to be minimized and the equipment utilization rates maximized. In tunnelling operations achieving accurate contours
dramatically improves the total economy of the lined tunnel projects. In recent years greater emphasis has also been placed
on environmental and safety issues. Tunnels offer ways to protect sensitive landscapes as well as reduce disruption, noise
and vibration, especially in urban areas. At the same time tunnels also need to be built in a safer and more environmentally
responsible manner. The introduction of Sandviks new generation Tamrock i-series jumbos will bring the tunnelling to a
totally new level regarding quality of excavation, ease of operation and maintenance, as well as overall performance. The
new intelligent drilling platform provides virtually unlimited possibilities in the further development of new features and
integrated systems. State-of-the-art technology integrated with accuracy, speed and user-friendliness the latest customer
demands forms the backbone of the entire product family. The intelligent machines ability to interact with external
network systems enables efficient planning of production and rig maintenance schedules.

95
Face drilling
Probing
Bolt hole drilling
Casing / umbrella drilling
Grouting
Self-drilling anchoring

There are all kinds of other auxiliary / utility
works to be done as well. Furthermore, tunnel
profiling, hole deviation and other measurements will
be used to improve the quality of tunnelling.
Today, the machines need to be able to be
configured with many different sets of features and
components to handle the multiple tasks in varying
rock conditions. This kind of versatility can only be
enhanced by modular structures combined with
advanced intelligence.

2.2. Improved accuracy through instrumentation
and automation
The total economy of tunnelling operations can be
improved by putting more emphasis in the accuracy of
the drilling. Reductions in overbreak and underbreak
in turn lead to reductions in haulage and scaling time
as well as in volume of shotcrete.
Better accuracy is achieved through instrumented
and automated drilling equipment. The development
of instrumentation and automation in face drilling rigs
started in the 1980s. However, wide acceptance of
the new systems had to wait until early 2000.
Today there are several different instrumentation
levels available and in use. The most sophisticated
systems, i.e. fully computerized jumbos, can be
operated manually, or in semi- or fully automatic
modes. In fully automatic mode, the jumbo drills
according to the pre-programmed drilling pattern
whilst the operator concentrates on supervising the
system. Also the tunnel geometry can be controlled as
the input data includes the curvature and inclination of
the tunnel. The automatic drilling cycle consists of the
following automated main process elements:

Positioning of drilling boom and alignment of
drilling feed to correct position (look-out angle
specified in pre-programmed drill plan)
Forward movement of drilling feed feed is
supported and aligned against the rock surface
Forward movement of rock drill drilling bit is
supported against the rock surface
Collaring of the hole to specified depth using
adjustable collaring power
Power acceleration from collaring power to
adjustable full power
Drilling of the hole to specified depth using
adaptive drilling features
Hole cleaning with compressed air
Return of rock drill to back end stop and retraction
of drilling feed from the rock surface

At best the operator can simultaneously supervise
drilling with three booms. It is still recommended that
a 4-boom drilling jumbo be operated by 2 operators in
order to best utilize the machine.
Advanced instrumentation provides many
advantages. The drill pattern can be optimized with
respect to the number of drill holes in the round,
blastability, haulage and pull-out. Good repeatability
ensures that the desired quality is achieved by all
operators on the rig.
In the new Tamrock i-series tunnelling jumbos,
the most advanced systems are the full data controls
(see Figure 1). It is also possible to extend the full
data controls with additional features like optimizing
the drilling e.g. by 1) adjusting the parameters per
hole type and 2) allowing the drill cycle to be adjusted
according to rock conditions. Both of these features
make better use of the expertise of experienced
operators.

Fig. 1. Drill plan display in a new i-series Tamrock jumbo
with full data system.
The simplest systems angle-indicating
instruments show the direction of the drill feed in
reference to the tunnel laser or other reference line.

2.3. Speed means optimization of the tunnelling
cycle
Earlier, speed was linked mainly to the drilling
features of a jumbo. The efficiency of the project
relied heavily on the penetration rates. The increased
drilling power has given substantially higher
penetration rates, e.g. in Scandinavian granite the rate
has increased even up to 4 m/min, and with newer
technologies to come this rate will still be improved.
96
Today, the actual drilling speed is no longer the
bottleneck of the performance and the emphasis has
been given to a more holistic approach evaluation of
the whole tunnelling cycle (see Figure 2).

Fig. 2. Drill and blast (D&B) cycle.
Planning is one of the key functions in the D&B
process. Careful planning is essential to enable
effective operation, and it has to be flexible in order
to react to the rapid changes in the process. To be
able to get the best possible tunnel quality, the
profile of each round should be measured and the
results immediately utilized for the next round.
Drilling with highly instrumented and
automated machines is effective and fast, since the
machines can repeat the designed drill plan round
after round. In grouting and probing, additional
solutions for mechanized long-hole drilling improve
the efficiency by reducing the drilling and auxiliary
times. These solutions also improve safety, since
there is no more need for the assistant driller to stay
at the face in the front of the machine.
The time used for charging and blasting can be
optimized since they have been taken into
consideration already in the early phase of planning.
In new generation jumbos more emphasis has been
put to not only drilling the holes, but also to how the
rock should break.
Loading and hauling of the blasted rock
provides information on rock fragmentation and the
floor conditions of the round. This can be utilized in
further development of the drill plans. The data
collection on the jumbo on the other hand helps in
phasing the different tasks in the tunnelling cycle
so, that the haulage can be done in a timely manner
e.g. when blasting is restricted.
Both drilling and charging may drastically
affect the reinforcement needs. Smooth and accurate
tunnel profile reduces the amount of highly
expensive shotcrete and concrete lining.
Surveying and profiling gives exact information
on the advance of each round and on the quality of
the contour. Thus it is possible to maximize the
pull-out and revise the drill patterns for the new
rounds. With instrumented jumbos it is even
possible to make modifications on the jumbo if
needed. The changes are then recorded, so that they
can also be utilized with the actual planning
software.
97

Fig. 3. Optimized tunnel cycles for 2-front excavation with one equipment set (prepared with Tamrock Tunnel Study
software).
To get full benefits, the different stages of the
tunnelling cycle cannot be considered or optimized
separately. Each stage can affect the performance or
cost of the other stages, and a slight change in time or
money spent in one stage can result in considerably
bigger opposite change in the other (see Figure 3).

2.4. Data collection supports the optimization of the
tunnelling cycle
Due to higher safety requirements, monitoring and
data logging needs are increasing in construction
work. Contractors face also extremely strict demands
regarding environmental aspects like noise and
vibration, wastewater from site, ground water control,
and possible water seepage into the tunnel. This also
increases monitoring needs.
Thus for example in contract negotiations the way the
risk is shared between the project owner and the
contractor is determined according to data logged.
Also, geo-engineers follow closely the ground
conditions and rock mass distribution along the tunnel
(see Figure 4). The results act as a basis e.g. for
grouting.

Fig. 4. Amount of water in a tunnel (characterization example made with Rockma system).
Data is gathered on face drilling holes as well as
on bolt, grouting and probing holes. Analysis will
partly be made on the jumbo and partly in the office,
and the rest of the work will be adjusted accordingly.
Real-time data collection together with
telecommunication connections (e.g. WLAN) opens
new revolutionary opportunities for the tunnelling and
mining businesses. Information transfer service
between the rig and the office increases productivity
and saves time and money.
Sandviks SanRemo system (see Figure 5) helps
in making well-grounded production plans.
Furthermore, the system helps in preparing equipment
maintenance schedules and evaluating the service
resource needs, since it delivers data on e.g. drill rigs
condition and work phase.

98

Fig. 5. SanRemo interface in construction equipment.
The information system on the jumbo needs to be
able to be connected to external systems and the data
format needs to be compatible to other software
(IREDES interface).
CONCLUSIONS
An in-depth understanding of the entire tunnelling
process and customer needs can be achieved only
through very close cooperation between the customers
and the equipment suppliers as well as with other
parties in the project. As an example, with over 60
years of experience in rock excavation and over 20
years of experience in data drilling Sandviks
Tamrock product line is widely recognized as one of
the most visionary pioneers in the mining and
tunnelling business. Furthermore, the new Tamrock i-
series is the most comprehensive range of intelligent
tunnelling jumbos available.
Careful optimization of the whole tunnelling
process all stages together is important. The
essential parts of the process concept are:

Fig. 6. Intelligent tunnelling equipment is required for accurate, fast and cost-efficient underground rock excavation.

99
Careful planning and plans, which can be
implemented in practice
Machine features and work methods to enable high
tunnel profile quality and excavation accuracy
Specification to meet challenging underground
conditions and local requirements and legislation
Equipment to fulfill strict environmental regulations
of tunnel construction
Suitable attachments to maximize equipment
utilization (machine option)
Reliable equipment, which offers high performance
constantly and efficiently
Safe and ergonomic operational environment
Easy operation and servicing of the machines
Maintenance plans and schedules to keep the
machines in excellent running condition
Measurements, which can be used to control the
tunnelling process
Operator training and team development to
implement efficient and systematic way of working
Logistics and shipping routes to ensure the spare
parts availability

With the right equipment, careful planning &
control, the trained organization will achieve the most
desired reward high quality tunnel excavation with
minimized project time and lowest cost.
Sandviks new intelligent Tamrock drilling
platform provides virtually unlimited possibilities in
the further development of new features and
integrated systems for tunnel process development.
State-of-the-art technology integrated with accuracy,
speed and user-friendliness forms the backbone of the
entire product family.
REFERENCES
1. World Highways, October 2004
2. Tunnels and Tunnelling, J une 2004
100
1 GOTTHARD AN ENDURING CHALLENGE
The Alps have always offered a stiff test to its would-
be conquerors. The story can be traced back to Roman
times and Hannibal's celebrated feat in crossing the
mountains with elephants.
And the task now facing the Swiss economy of
building the world's longest railway tunnel, the 57 km
Gotthard Base Tunnel is no less daunting.

2 WIDENING THE SCOPE OF SBB SERVICES
Once completed, the Gotthard Base Tunnel will in-
crease SBB's transport capacity in both the passenger
and freight sectors, thereby contributing to two key
planks of the Swiss transport policy.

2.1. European high-speed rail network
The Gotthard Base Tunnel marks a further step in the
integration of the Swiss railway system in Europe's
high-speed network [Fig. 1] and will greatly increase
future passenger transport capacity.
With the completion of all planned route upgrades
by the end of 2016, the journey time between Zurich
and Milan will be slashed by a full hour to a mere 2
hours and 40 minutes. Improved connections will also
be provided at the major hubs.

Fig. 1. Integration in Europe's high-speed rail network
2.2. Shifting freight from road to rail
By enshrining the Article on the Protection of the
Alps (Alpenschutzartikel) in its constitution, the Swiss
population has unequivocally pledged itself to shifting
long-distance freight from road to rail. This aim ac-
cords with the policies and objectives of the European
Union.
Located between France, Austria and Italy, Swit-
zerland plays a pivotal role in north-south transalpine
freight traffic, particularly in connection with the
road/rail modal split. Even today, two-thirds of the
tonnage crossing the Alps passes through Switzerland
by rail.
To accommodate the growth in transalpine freight
volumes, Switzerland has set out to double its rail ca-
pacity from the current 20 million tonnes p.a. to some
40 million.
Gotthard Base Tunnel the world's longest railway tunnel
Georg A. Schmalz
Head of Projects Construction Management, Swiss Federal Railways SBB, Berne, Switzerland
Serena L. De Dominicis
MSc Civil Engineer ETH, Swiss Federal Railways SBB, Berne, Switzerland
ABSTRACT: The 57 km long Gotthard Base Tunnel currently being driven through the Swiss Alps will serve rail traffic.
The following article shows how Swiss Federal Railways (SBB) as future operator has mastered a project of this scale while
outlining the key technical challenges and solutions.

101
3 ALPTRANSIT GOTTHARD
The New Rail Link through the Alps (NRLA) or
AlpTransit for short is the name of the scheme
launched by the Swiss Confederation to boost its pas-
senger and freight transport capacity. Its aim is to pro-
vide a flat route along both the Ltschberg and
Gotthard axes.
Linking Basel and Milan via Berne and the Sim-
plon line, the Ltschberg route is scheduled to come
into operation at the end of 2007.
Work on the Gotthard axis, which connects the
same two cities via Zurich and Lucerne, is still under
construction. Its key projects are [Fig. 2]:
the 20 km long Zimmerberg Base Tunnel; the first
section is already in operation
the 57 km long Gotthard Base Tunnel
the 15 km long Ceneri Base Tunnel.

Fig. 2. Gotthard AlpTransit axis

3.1. AlpTransit Gotthard Ltd. (ATG)
A specially established 100% subsidiary of SBB
AlpTransit Gotthard Ltd. signed a contract with the
Swiss Confederation for construction of the Gotthard
axis.
With a workforce of 110, ATG appoints profes-
sionals and contractors from across Europe to provide
the necessary engineering and construction services.
All appointments fully comply with statutory
requirements on public procurement.
ATG acts as the client's agent and bears respon-
sibility for meeting all quality, cost and deadline
targets. To maximise efficiency, it employs an integral
management system, including all aspects of quality
and environmental management as well as safety at
work and information security [9]. The built-in risk
management function imposes exacting Quality
Management (QM) requirements on the appointed
engineering and contracting companies.

3.2. Costs
The total construction costs for the flat rail link along
the Gotthard axis was originally estimated at around
8 billion Swiss francs (USD 6.3 billion). Approxi-
mately 3.5 billion Swiss francs (USD 2.8 billion) has
been invested to date.
The need for additional investment on infrastruc-
ture with a design and construction timeframe of 25
years was hardly unexpected, given the inevitable ad-
vances in technology and standards over such a long
period. At around 2.3 billion Swiss francs (USD
1.8 billion), the volume of additional investment
required for the Gotthard Base Tunnel is roughly
equivalent to 30% of the base cost estimate [Fig. 3].
This has been necessitated by [9]:
Investment in safety and improvements reflecting
the current state-of-the-art, e.g. reduction of the
distance between passages, rerouting of exhaust
systems at Sedrun and Faido multifunction
stations, change from one double-track to two
single-track tunnels for the Ceneri Base Tunnel
Politically motivated delays, phasing, provisions of
the FinV (federal public transport funding
legislation), e.g. preliminary investment in branch
structures
Geological factors, e.g. unforeseen fault zones in
Bodio and Faido sections
Improvements for population and environment
Contract award and site operations.

Fig. 3. Additional investment and extra costs
Present knowledge suggests that the final costs are
likely to run to some 10 billion Swiss francs (USD
8.1 billion) [6].

102
4 GOTTHARD BASE TUNNEL
For safety reasons, the 57 km base tunnel will have
two single-track tubes, linked by cross passages every
325 m. The two so-called multifunction stations, pro-
viding crossovers, are located at the "one-third" points
in the Sedrun and Faido sections. These structures
also accommodate the emergency stations used in
case of tunnel incidents [Fig. 4].

Fig. 4. Scheme of the Gotthard Base Tunnel system
4.1. Emergency stations
The two emergency stations provide for the escape
and evacuation of passengers. Escape routes to the
parallel tube neither lead across track nor necessitate
the use of stairs or elevators. Should an incident oc-
cur, the emergency stations, service and connecting
tunnels will be supplied with fresh air [Fig. 5].
High-performance ventilation systems to extract
smoke and blow in fresh air from the outside guaran-
tee safe rescue in emergencies. While smoke is sucked
out of the tube in which the incident has occurred, a
slight overpressure of the emergency station keeps the
escape route to the other tube smoke-free. Should a
train come to a halt outside the emergency station,
travellers can use the cross passages to access the
other tube.

Fig. 5. Emergency stations

4.2. Longitudinal and cross-section
Designed as a flat route, the new Gotthard railway
reaches its highest point in crossing the Alps at an
altitude of 550 m above sea level, with a gradient not
exceeding 12 outside and 8 inside the tunnel. The
maximum overburden depth totals 2500 m.
The circular single-track tubes are 8 m in inner di-
ameter [3] and are designed with doubleshell lining
and partial ("umbrella") waterproofing. Waterproofing
also under the invert is installed wherever heavy water
infiltration or aggressive waters are encountered.
While the inner lining generally comprises 30-35 cm
unreinforced concrete, up to 120 cm wall thickness is
adopted in squeezing rock zones.

4.3. Construction programme
To cut construction time, the 57 km tunnel between
the north portal at Erstfeld and south portal at Bodio
was split into five sections.
Table 1. Gotthard Base Tunnel sections
Length Tunnelling method
Erstfeld section 7.4 km TBM
Amsteg section 11.4 km TBM
Sedrun section 6.8 km Drill & Blast
Faido section 14.6 km TBM
Bodio section 16.6 km TBM

Due to additional bypasses and intermediate points of
attack the Gotthard Base Tunnel will need a total ex-
cavation of 153.5 km for the main tunnels, access
tunnels and shafts. By the end of April 2006, 92.3 km
or 60.1% had been excavated [7]. The Gotthard Base
Tunnel will go into service in 2016 [Fig. 6].

Fig. 6. Construction programme

103
5 GEOLOGY
For a better understanding of geological conditions, a
comprehensive exploration programme was imple-
mented. This comprised ground investigations (ex-
ploratory drillings and exploration galleries), seismic
surveys and the analysis of documents from previous
construction schemes (road and rail tunnels, power
stations and military facilities).

Fig. 7. Geological formations
The Gotthard Base Tunnel crosses the following geo-
logical formations [Fig. 7]:
Aar massif: old-crystalline gneisses and younger
granites, Intschi zone
Tavetsch intermediate massif: frequent alternation
between kakirite-intercalated schists and phyllites
Urseren-Garvera zone
Gotthard massif: gneisses and old-crystalline
schists
Piora syncline
Penninic gneiss zone: comprising Leventina gneiss
in the north and Lucomagno gneiss in the south.

5.1. Temperatures
Temperatures as high as 45C were measured at tun-
nel level in the Amsteg section, where the overburden
depth under the Chrzlistock summit rises to a full
2187 m. These temperatures exceed the predicted
value of 38C and the projected 5C fluctuation
range.
To protect the health of tunnelling crews, the
SUVA (Swiss National Accident Insurance Fund) has
prescribed a maximum working temperature of 28C
with 70% humidity for the Gotthard Base Tunnel. The
underground working areas are therefore equipped
with an air cooling system.
The unexpectedly high temperatures necessitated
retrofitting to the ventilation and cooling installations
as the works proceeded. Moreover, measures have
been specified for immediate implementation when-
ever thresholds are exceeded, e.g. shorter working
times and the suspension of works for cooling provi-
sion [6].
Initial analyses single out the assumptions made
regarding the anisotropy of thermal conductivity in
the vertically stratified ground as one possible expla-
nation for the inaccuracy of predictions. Temperatures
up to 50C may well be encountered during tunnelling
in the Gotthard massif, where overburden depths will
rise to 2500 m.
Apart from the extra cost incurred for additional
measures during the construction period, the impli-
cations for the tunnel's operation phase are currently
being examined.

5.2. Geological findings
When the project started, the Piora syncline within the
Faido section appeared to pose the greatest risks in
geological terms. Yet, as revealed by detailed
explorations in 1997, the feared occurrence of sugary
dolomite failed to reach tunnel level and thus posed
no particular problems [Fig. 8].

Fig. 8. Piora exploratory system
Exploratory borings had suggested difficult tun-
nelling conditions in the Tavetsch intermediate massif
and Urseren-Garvera zone within the Sedrun section.
Fortunately, no major problems have yet been encoun-
tered during initial excavations in these zones.
In the Bodio section, the base tunnel crosses an
approximately 400 m long scree zone, called the
Ganna di Bodio. To prevent delays, an additional
1200 m bypass tunnel was driven through good rock
from which work can proceed northward and
southward along the tunnel axes.
The 950 m long Intschi zone also posed various
challenges. Here, the average daily advance achieved
by the TBM in the eastern tube slumped to 7 m per
working day, while an average 15-20 m was recorded
in the healthy parallel tube to the west.
104
The geological conditions revealed by the Got-
thard Base Tunnel have shown that even the most
thorough ground investigations cannot eliminate all
residual risks. Estimates based on the present state of
tunnelling put the residual geological risk at 10% of
the remaining tunnel construction cost [9].

5.3. Rock bursts
A rock burst experienced at the Faido site at the end
of March 2006 was perceived as a micro-earthquake
in the surrounding villages, reaching a magnitude of
2.4 on the Richter scale. However, the rock burst
caused only minor harm to the tunnel lining and no
damage was reported on the surface [7].
The Faido section had already witnessed several
violent rock bursts in 2005. These had only been pre-
dicted for the massive gneisses of the Gotthard massif
and not for the stratified gneisses with heavily split
planes near the Faido multifunction station, and
certainly not in this severity.
These rock bursts are thought to have been trig-
gered by the tunnelling works in the neighbouring
tube and stress redistribution in the region of the ma-
jor fault zone.
As a result, it was decided to make suitable provi-
sion to accommodate the deformation precipitated by
work in the western tube as long as this potentially
affected northward tunnelling in the eastern tube. At
the same time, the spacings of side tunnel and exhaust
adits at the eastern emergency station were increased
to minimise interaction [6].

6 TUNNELLING DETAILS
6.1. Tunnel boring machine (TBMs)
All sections apart from Sedrun are driven using un-
shielded hard-rock TBMs with 8.8-9.58 m cutterhead
diameters [Fig. 10].
The machines were custom-developed for the
Gotthard Base Tunnel scheme and part of their as-
sembly takes place in vast underground caverns. The
tunnelling installations (cutterhead and back-up)
stretch over a length of around 440 m [Fig. 9].

Fig. 9. Long section through tunnel boring machine
Excavation support is by systematic rock bolting
and shotcreting. Where tunnelling conditions are dif-
ficult, it is possible to incorporate steel arches directly
behind the cutterhead. The in-situ concrete invert is
placed from the rear section of the TBM.
To allow tunnel driving to proceed independently
of invert placement while minimising any mutual ob-
struction, the boring and support material is supplied
via a partially suspended back-up unit. Conveyor belts
are used for mucking out and rail trucks for transport-
ing materials to the tunnel face [1].
The maximum advance rate so far achieved in the
Gotthard Base Tunnel stands at 40.1 m per day. Given
favourable ground conditions, average rates of 20-
25 m per working day are normally feasible.

Fig. 10. Assembly of tunnel boring machine
6.2. Raise boring in Sedrun
The Sedrun section is accessed via two 800 m deep
vertical shafts. The 8.4 m diameter Shaft I was sunk
by blasting in 1998/99. Later, in 2002/03, the raise
boring method was used to construct the 7.0 m diame-
ter Shaft II [Fig. 11].
Adoption of this less violent technique enabled
substantial savings in terms of support materials and
lining concrete. The per-metre cost for Shaft II was
only half that for Shaft I as reflected by the conside-
rably shorter construction time of 12 months for Shaft
II, compared to the 17 months needed for Shaft I [4].

Fig. 11. Raise boring technique used for Sedrun Shaft II
105
6.3. Developments for excavation in squeezing rock
The Tavetsch intermediate massif features geological
formations comprising approximately 70% soft and
kakirite-intercalated schists with ductile fracture be-
haviour and approximately 30% friable, largely intact
gneisses prone to brittle fracture.
Poor strength and deformation properties, coupled
with low ground permeability (coefficients of perme-
ability, k-values between 10
-8
and 10
-10
m/s) and prob-
able initial pore water pressures of up to 8 MPa at
tunnel level, call for special measures [5]. Rock
pressure due to squeezing, pore water pressure and
tunnel face instability constitute the key hazard
scenarios.
The adopted technical solutions include [4]:
selection of circular excavation geometry
exclusive use of full-face excavator
provision for up to 70 cm overbreak to accom-
modate deformations of rock
provision for systematic tunnel face support using
long horizontal bolts
use in the initial (deformation) phase of support
materials with high deformation capacity, e.g. steel
rings with sliding connections and bolts
installation of rigid supports, e.g. sprayed concrete,
as deformation diminishes.
To maximise advance rates and avoid delays, various
excavation support types tailored to the different
hazard scenarios were specified in advance for use as
required by the encountered rock conditions [5].

Difficult tunnelling conditions always provide a
valuable opportunity for trying out new technical so-
lutions. The Gotthard Base Tunnel has been used for
testing deformable steel supports. Established in
hard coal mining, the system with sliding connections
(TH profiles) has never before been used on this scale
in tunnelling.
Excavation of the cavity is followed by installa-
tion of two concentric steel arches, each comprising
eight segments with connections of limited flexibility.
Any rock convergences of the tunnel section due to
pressure of squeezing rock causes the rings to gradu-
ally slide together [Fig. 12]. The arches achieve their
maximum load-bearing capacity when fully closed.
The test series showed the concept to be a suitable
solution for this geologically difficult zone [9].

Fig. 12. Testing of special steel supports in Sedrun section
This squeezing rock zone, in particular, posed a
major logistical challenge for the contractor. To mini-
mise the mutual obstruction of traffic on the tunnel
invert, the tunnelling installation was suspended. For
the very first time, a hanging excavation machine
familiar from mining applications is being used in a
tunnelling scheme [4]. This multi-purpose machine
allows the installation of steel arches, cutting of face-
bolts and shotcreting at the tunnel face using a spray-
ing manipulator [Fig. 13].

Fig. 13. Hanging excavation machine (Drill & Blast)
7 RELATED ASPECTS OF THE EXCAVATION
WORK
7.1. Monitoring of surface movement
The drainage of water from the ground caused by tun-
nelling schemes can sometimes induce surface settle-
ment. The Gotthard Base Tunnel passes under three
hydropower dams. Differential settlement and move-
ment at the valley sides bounding these retaining
structures may result in overloading or even structural
damage.
106
The comprehensive monitoring regime imple-
mentted in response to these risks, which featured a
ground-breaking combination of various measurement
sensors and automation systems, amounts to an
extraordinary innovation in topographical surveying
[Fig. 14].

Fig. 14. Monitoring regime for hydropower dams [8]
The monitoring regime provides for the auto-
mated, year-round surveillance of the valley profiles
at the three dams, together with a selection of repre-
senttative points. For poorly accessible locations, au-
tomatic readings are taken by GPS satellite [Fig. 15].
As the precision measurement equipment will be
in operation the whole year round over the entire ten-
year construction period, it needs to be suitably resi-
lient to mountainous conditions. The measurement
points may experience snow depths of up to 3 m.
Some are inaccessible during the winter months or
may need to withstand avalanches. The monitoring
programme will permit timely action to be taken
should the readings indicate any breach of the
tolerance limits. An extensive grouting programme
within the tunnel has been defined to be a geo-
technical auxiliary measure for the excavation of the
Gotthard Base Tunnel, should it be required.

Fig. 15. Monitoring of dam structures
7.2. Materials management
Excavation of the Gotthard Base Tunnel produces vast
quantities of material: the 24 million tonnes of muck
excavation material are equivalent to building five
Cheop's pyramids [9].
To achieve savings while conserving resources
and reducing material shipments, the client decided,
wherever possible, to use the excavated material for
construction purposes. The finely graded, chip-like
muck produced by TBMs, was without treatment not
suitable for use as concrete aggregate and could, until
recently, only be used for embankments or as landfill
for disposal sites.
The client's dissatisfaction with this situation led
to the launch, back in 1993, of a test programme in
collaboration with universities, research institutes and
the concrete industry which successfully demon-
strated the feasibility of recycling this material to pro-
duce high-grade concrete aggregate [9]. This, how-
ever, requires state-of-the-art infrastructure for aggre-
gate production plus leading-edge concrete techno-
logy.
The contractors are now provided with the proc-
essed concrete aggregate by the client, who thereby
assumes a share of responsibility in the concrete pro-
duction process.
The converting process of high-quality excavated
rock into around 5 million tonnes of concrete aggre-
gate is being carried out directly on the construction
site. The surplus material is offered to interested third
parties. It is also transported eco-efficiently by rail
and barge for use in re-naturalization schemes or as
fill for rock embankments in lakes [Fig. 16].

Fig. 16. Fill deposited in Lake Lucerne

107
7.3. Safety at work
Top priority is given at the Gotthard Base Tunnel con-
struction sites to protecting the workforce of approxi-
mately 2000 employees.
AlpTransit Gotthard Ltd. implements a three-
stage safety at work strategy for site operations [1]:
Due consideration is given to safety at work is-
sues from the very first stages of design. The fo-
cus is on finding solutions that ensure maximum
safety during work execution.
In the tendering phase, safety at work ranks
among the key criteria for contract award. All
tenderers are expected to give detailed conside-
ration to the relevant issues, with submission of a
safety at work concept as one of the require-
ments.
On-site implementation of safety procedures is
ensured through training, good-practice models,
inspections and audits. Broad-based awareness
campaigns help to establish a deeply rooted
safety culture.

Fig. 17. Project status in spring 2006

8 PROJECT STATUS IN SPRING 2006
8.1. Erstfeld section (7.4 km)
Erstfeld was the last section on which work com-
menced. Preparation measures, e.g. construction of a
water reservoir, works for a water treatment plant and
installation of a train loading facility, are currently in
progress.
The contract for this last major tunnel construc-
tion section of the Gotthard Base Tunnel was finally
awarded at the start of May 2006. The initial contract
let in August 2005 had been annulled by the Swiss
Federal Appeals Commission for Public Procurement
following an objection lodged by one of the tenderers.
This had necessitated a re-evaluation of the tender
submissions [7].
8.2. Amsteg section (11.4 km)
A further drive of approximately 750 m is required in
the eastern tube to reach the Sedrun boundary. The
length of tunnel drive completed by the beginning of
April 2006 totalled around 10.5 km.
The TBM in the western tube has resumed opera-
tion at km 9 following clearance of the blockage
caused by water infiltration and ground flow in June
2005. To clear the cutterhead, a bypass tunnel was
driven from the advancing eastern tube to allow drain-
age and grout stabilisation of the shattered rock ahead
of the cutterhead. Given that the construction pro-
gramme included reserve time for passage through
such fault zones, the section boundaries are still
anticipated to be reached on schedule [6].
Of the total of 37 cross passages, 28 have been
driven. So far, the inner concrete lining of 19 cross
passages has been cast [7].

8.3. Sedrun section (6.8 km)
In the Sedrun section, over 50% of the two single-
track tubes has now been excavated. The southward
drive has already cleared cross passage 7 (of 16),
while cross passage 4 (of 5) has been reached in the
north.
With the exception of the exhaust adits, the multi-
function station has now been fully excavated. Work
on the exhaust adit to the south-western emergency
station (length 518 m, 7 cross passages and shafts)
commenced in mid-March 2006. Excavation work for
4 of the 7 shafts for the northern exhaust adit is now
complete [7].
On the southward drive, the Urseren-Garvera
zone, classed as difficult for tunnelling, turned out to
be only 305 m long a full 205 m shorter than pre-
dicted. The more favourable geology cut the construc-
tion time by roughly one year. To speed up the break-
through between Sedrun and Faido, the options on
redefinition of the sections length awarded to the
contractors are being exercised to allow a southward
relocation of the boundary with the Faido section. The
Faido-Sedrun breakthrough will thus take place
roughly two to three months ahead of schedule [6].
The northward tunnel drive through the squeezing
rock of the Tavetsch intermediate massif is proceed-
ing slowly at a rate of approximately 1.3 m per day
(the use of hanging excavation machines drill &
blast is described under 6.3).

108
8.4. Faido section (14.6 km)
The northward drive of the western single-track tunnel
at the end of 2005 became the last of the main drives
to leave the major fault zone. Excavations are now
proceeding apace through the undisturbed Lucomagno
gneiss. The southward drive has reached the boundary
with the Bodio section.
Repair work to the northward drive of the single-
track eastern tunnel caused by a rock burst is now
completed. The challenging works required for the
eastern side tunnel overpass are making good pro-
gress. Construction of the exhaust adits and extraction
shafts is proceeding as planned [7].
The aim is to complete the majority excavation
within the multifunction station in time for the
scheduled arrival of the two TBMs from Bodio in the
summer of 2006, so the TBMs can be partially disas-
sembled and pulled through the multifunction station.
There the TBMs will be fitted with new cutterheads
with larger diameter and will be reused to advance the
tunnel towards the Sedrun section [6].

8.5. Bodio section (16.6 km)
Around 11.8 km (87% of the TBM section) of the
eastern tube and 12.2 km (approximately 87%) of the
western tube had been driven by mid-April 2006. An
approximately 1.7 km drive in the eastern and a
1.9 km drive in the western tube are left to the
breakthrough at Faido section [7].
The difficult geological conditions made the in-
stallation of steel arches in parts of both the eastern
and western tubes necessary. The unforeseeable hori-
zontal fault zone encountered in 2003 slowed down
tunnelling considerably. Nonetheless, work is still ex-
pected to finish on schedule [6].
The tunnel inner lining is being installed at a rate
of 24 m per day and tube. Approximately 7.1 km
(53%) and 8.8 km (63%) of the linings are in place in
the eastern and western tube, respectively [7].

9 SBB AS FUTURE OPERATOR
As future operator of the Gotthard Base Tunnel, SBB
is closely involved in all aspects of railway engineer-
ing. In cooperation with AlpTransit Gotthard Ltd., it
defends the interests of the rail service and is con-
cerned with all operational and maintenance issues
affecting the rail infrastructure after the tunnel will
come into operation in 2016.
The SBB will use an integral network to manage
rail operations and tunnel monitoring from a newly
built control centre sited at the southern tunnel portal.
The tunnel control centre (TCC) will be responsible
for regulation, surveillance and scheduling of rail traf-
fic under normal, incident and maintenance modes.
A run or maintain strategy will be applied to the
performance of maintenance work (including inspec-
tion). Safety matters in railways can not be delegated,
so SBB will retain full responsibility for maintenance
and fault rectification. Various works (ventilation,
cleaning etc.) will be carried out by contractors under
SBB's supervision.

Fig. 18. Maintenance concept
Maintenance is performed on the basis of a 2/2
closure cycle (productivity gain in train path availabil-
ity). One of the two tubes is completely closed for
eight hours during the nights from Saturday to Sunday
and from Sunday to Monday [Fig. 18]. Sporadic
joker closures for a maximum of 4 hours will also
permit planned deployments during the week. Work
will be co-ordinated from two maintenance units
(north and south), which also house rescue control
centres [2].
Maintenance work is complicated by environ-
mental conditions within the tunnel. Mobile gates are
used to split the standard maintenance sections into
two ventilation zones to maintain acceptable working
conditions.
Preliminary estimates put the projected mainte-
nance costs of the Gotthard Base Tunnel when in op-
eration at around 25-37 million Swiss francs p.a.
(USD 20-30 million).

109
10 OUTLOOK

Fig. 19. The top of the Gotthard massif
Next year will see the 125
th
anniversary of the present
15 km long Gotthard Tunnel, in operation since 1882
and with a summit 1150 m above sea level [Fig. 20].
While this mountain line was considered a pioneering
achieve-ment at the time of its construction, the new
57 km long Gotthard Base Tunnel will deliver a
similar testament to modern engineering.
The successful completion of such an exacting,
once-in-a-century project hinges on first-rate, inno-
vative work consistently delivered by a team of highly
skilled and committed individuals.

Fig. 20. Intercity train on the existing Gotthard line
A decision has yet to be taken on the future of the
existing Gotthard line after the Gotthard Base Tunnel
is commissioned in around 10 years' time. For SBB,
however, it is clear that the mountain line must be
retained, if only with reduced capacity. Both tourists
and the local population will thus continue to enjoy
the scenic ride along the mountain line, while fast
passenger [Fig. 21] and transit freight trains speed
through the Gotthard Base Tunnel.

Fig. 21. Future high-speed train for the Gotthard Base
Tunnel: Cisalpino ETR 610

REFERENCES
1. AlpTransit Gotthard AG, 2002. Die neue Gotthard-
bahn, Herausforderungen und Lsungen (The new Got-
thard Railway, challenges and solutions). Supplement
from baublatt_ EXTRA of 5 March 2002.
2. Bernardi, Felix P., Swiss Federal Railways SBB. 2004.
NBS AlpTransit Gotthard, Erhaltungskonzept 2004
(Maintenance concept), Berne.
3. Chabot, Jan D., Swiss Federal Railways SBB, 2004.
Presentation on the Gotthard Base Tunnel, the longest
Railway Tunnel in the world in Tokyo.
4. Ehrbar, Heinz and A. Henke. 2003. Aktuelle Erfahrun-
gen und Entwicklungen beim Bau des Gotthard Basis-
tunnel (Current findings and developments during con-
struction of Gotthard Base Tunnel). International Sym-
posium on Geotechnical Measurements and Modelling,
Karlsruhe.
5. Ehrbar, Heinz, Alp Transit Gotthard. 2002. Felssi-
cherung in druckhaftem Gebirge am Beispiel des Got-
thard-Basistunnels (Rock support in squeezing ground,
based on example of Gotthard Base Tunnel). In Tun-
nel, IUT 02, 8495.
6. Federal Office for Transport (BAV). 2005. Neue
Eisenbahn Alpentransversale (New Transalpine Rail-
way), status reports no. 20 and no. 19.
7. Official homepage on the Gotthard Base Tunnel: www.
alptransit.ch
8. Topographic maps from swisstopo (Swiss Federal Of-
fice of Topography), Wabern, Berne.
9. Zbinden, Peter and A. Sieber. Alp Transit Gotthard.
2004. Herausforderungen und Lsungen beim Bau des
lngsten Eisenbahntunnels der Welt (Challenges and
solutions for the world's longest rail tunnel scheme).
Honorary colloquium Prof. Dr. Friedhelm Heinrich
Rock Mechanics, Technical University of Freiberg,
Germany.
110
1 INTRODUCTION
A prime concern in any shallow tunnelling project is
the impact that the excavation process may have on
the surface environment. In most cases, the immediate
concern involves differential settlements caused by
ground/volume loss leading to damage to sensitive
surface structures. The experiences of the Jubilee Line
Extension in London provide a recent example of one
such case where compensation grouting was used to
minimize surface settlements resulting from ground
loss [1]. In water-bearing ground, tunnel drainage may
also lead to differential displacements due to time-
dependent consolidation and subsidence. These cases
almost exclusively involve shallow tunnels excavated
in soft, unconsolidated soils (e.g. [2-4]).
Consequently, the analytical and numerical
procedures available to predict the extent of surface
subsidence are based solely on continuum mechanics
where the subsidence profile calculated is symmetric
about the vertical tunnel axial plane. In many cases,
the profile is approximated by an inverted Gaussian
normal distribution curve (Fig. 1a). This assumption is
generally valid for soft ground conditions where the
presence of geological heterogeneity is restricted to
horizontal layering (Fig. 1b).
In fractured rock masses, geological heterogeneity
is more complicated. At scales greater than the natural
block size, deformation generally occurs by joint
opening or shear, by fault movements, or by bedding
plane slip [5]; i.e. the deformation mechanism is
discontinuous. This has been shown to be an
important factor when considering both subsidence
due to mining/ground loss (e.g. [6, 7]) or
consolidation during fluid extraction (e.g. [5]).
In hard rock tunnelling, subsidence is rarely
considered and in the past, tunnelling engineers would
not expect substantial subsidence to occur in
association with a deep tunnelling project.
Consolidation phenomenon is even less of a concern
even though large reductions in pore pressure can
occur when driving a deep tunnel through fractured
rock. In contrast to such widely held views, recent
high-precision levelling measurements above the
Gotthard highway tunnel in central Switzerland have
revealed up to 12 cm of subsidence over a 10 km
section. Zangerl et al. [8] have shown that the
temporal and spatial relationships between the
measured settlements and excavation of the Gotthard
highway tunnel point to causality between water
drainage into the tunnel and surface deformation.
This paper reports these findings and discusses the
role of geological structures in controlling the shape
of surface subsidence profiles. Emphasis is placed on
understanding the underlying mechanisms involved in
fracture consolidation given their importance with
regards to the construction and monitoring of the new
57-km long Gotthard base tunnel [9], whose
alignment passes close to several concrete dams and
other strain-sensitive surface structures.
The influence of geological structures in promoting asymmetrical
surface subsidence over deep tunnels in hard rock
Erik Eberhardt
Geological Engineering/EOS, University of British Columbia, Vancouver, BC, Canada (formally ETH Zurich)
Christian Zangerl
alpS GmbH, Centre for Natural Hazard Management, Innsbruck, Austria (formally ETH Zurich)
Simon Loew
Engineering Geology, Swiss Federal Institute of Technology (ETH Zurich), Zurich, Switzerland
ABSTRACT: Subsidence in fractured crystalline rock is rarely observed and in the past, engineers would not expect
substantial differential settlements to occur in association with a deep tunnelling project in hard rock. However, recent high
precision levelling measurements have revealed up to 12 cm of surface subsidence several hundred metres above the
Gotthard highway tunnel in central Switzerland. Large-scale consolidation resulting from tunnel inflows and pore pressure
changes in the rock mass is believed to be the controlling mechanism. This paper presents results from an extensive field
and numerical modelling investigation focussing on the processes responsible for this subsidence. Results show that the
symmetrical subsidence profiles calculated using traditional continuum techniques, whether closed-form solutions or finite
element modelling, inadequately reproduce the shape of the subsidence profile. Instead, 2-D discontinuum modelling using
the distinct-element method, whereby mapped geological structures were explicitly included, enabled important insights to
be gained with respect to the asymmetry and small-scale inflections in the shape of the subsidence profile measured.

111

Fig. 1. a) Typical surface subsidence profile based on an
inverted Gaussian normal distribution curve; b) Measured
long-term subsidence profile above a tunnel in soft soils
(after [10]).
2 CONSOLIDATION SUBSIDENCE AND HARD
ROCK TUNNELLING
Although surface subsidence related to hard rock
tunnelling is rarely considered, some precedence does
exist. Karlsrud and Sander [11] report cases in Oslo,
Norway where tunnels driven in fractured bedrock at
depths of 20 to 40 m drained the overlying sediments
resulting in up to 35 cm of subsidence. In these cases,
the soft marine clays lying above the bedrock were
considered as being the sole consolidating material,
with the fractured rock mass below only serving as a
conduit for drainage during excavation.
More pertinent to the case of rock mass
consolidation is Lombardis report [12] of the Zeuzier
double arch dam in western Switzerland, where 13 cm
of vertical settlement were measured at the dam in
relation to the driving of an investigation adit 1.5 km
away (Fig. 2a). These displacements resulted in the
development of a series of cracks up to 15 mm wide
(Fig. 2b). Initially, no clear explanation could be
provided as to the source of the settlements and the
reservoir was ordered drained. After investigating and
discounting a number of alternative causes, suspicion
fell on the construction of an adit in a confined,
fractured, marly-limestone aquifer in which
significant water inflows were recorded [12]. The adit
excavation works were then ordered stopped and the
cracks in the dam body repaired.
3 THE GOTTHARD ROAD TUNNEL
In 1998, the Swiss Federal Office of Topography
completed a routine high-precision levelling survey
over the Gotthard pass road in central Switzerland
(Fig. 3). Comparison of results with those from a
survey over the same route in 1970 revealed that up to
12 cm of downward displacement had occurred over a
10 km section (Fig. 4a). This sharply conflicted with
earlier surveys made between 1918 and 1970, which
showed upward displacements of 1 mm/year in
agreement with measured alpine uplift rates for the
region [13]. The presence of the subsidence trough
was later confirmed by geodetic triangulation
measurements supplemented with GPS data [14].
Field investigations performed to determine the
origin of the subsidence quickly ruled out localised
surface phenomena (e.g. a deep-seated landslide)
given the absence of local indicators and the 10 km
extent over which the settlements were measured [15].
Instead, spatial and temporal relationships between
the measured settlements and the nearby Gotthard
highway tunnel pointed to causality between tunnel
drainage and surface deformation.

Fig. 2. a) Geological profile through the Zeuzier double
arch dam site and investigation adit showing direction of
subsidence related ground tilt; b) Photos of the dam and
subsequent cracking of the dam. After [12].
112
The 16.9-km long Gotthard highway tunnel was
constructed between 1970 and 1977, thus fitting
temporally with the period between the 1970 and 1998
geodetic surveys. Spatially, the North-South
alignment of the tunnel closely follows that of the
levelling transect along Gotthard Pass Road several
hundred metres above (Fig. 5). The geology
encountered consists primarily of paragneisses and
granitic gneisses of the central Gotthard massif.
Labhart [16] describes these massifs as representing a
crystalline basement made up of paragneisses,
amphibolites, late-Ordovician granites and middle-
Palaeozoic metasediments, intruded by late-Variscian
plutons (e.g. the Aar, Gamsboden and Fibbia granites;
Fig. 5). These rock bodies were overprinted by alpine
metamorphism, mostly in greenschist facies.
Construction of the highway tunnel included that
of a smaller safety tunnel, which was excavated
several hundred meters ahead of the main tunnel with
a 12 to 18 month time lag. This allowed the safety
tunnel to serve both as an investigation and drainage
adit for the main tunnel. Water inflows into the safety
tunnel were measured periodically, for which a 3-km
zone was encountered in the Gamsboden granitic
gneiss that produced especially high inflows (Fig. 4b).
Steeply inclined brittle fault zones acted as the
primary drainage conduits into the tunnel, two of
which, situated 23 m apart, produced initial inflows of
300 l/s (today rates of 8 l/s are recorded along this
interval [17]). The location of this interval and that of
the 3-km zone in general, closely corresponds to the
centre of the broad subsidence trough seen in the
settlement profile, with the point of peak water inflow
coinciding with that of maximum subsidence (Fig. 4).

Fig. 3. Photo of the Gotthard Pass road and area of
investigation (looking North from Mtteli). Note the
ventilation shaft for the Gotthard highway tunnel to the
right of the photo.

Fig. 4. a) Levelling profile along the Gotthard pass road
showing measured surface subsidence relative to earlier
measured periods of alpine uplift; b) Measured initial
inflow rates into the Gotthard safety tunnel per 100 m
interval during excavation. After [18].

Based on this correlation, a working hypothesis
was developed that pointed to tunnel-induced surface
subsidence associated with deep drainage and
consolidation of the fractured crystalline rock mass.
An extensive field, laboratory and numerical
modelling campaign was then initiated to explore and
explain the processes and mechanisms underlying the
measured subsidence (see [15]). Most of the focus
was placed on the major fault zones and meso-scale
fractures mapped in the region [19] and understanding
the phenomenon of fracture consolidation, although
part of the focus was also extended towards the
consolidation behaviour, i.e. poroelasticity, of the
low-porosity intact rock (<1% intact matrix porosity).
Thus, the analysis performed approached the problem
from both the perspective of an equivalent continuum,
the assumption required by most conventional
methods of analysis, and that of a discontinuum.

113

Fig. 5. Geological map of the study area and locations of
the Gotthard highway tunnel and surface levelling traverse
over the Gotthard Pass road.
4 CONVENTIONAL ANALYSIS OF TUNNEL-
INDUCED CONSOLIDATION SUBSIDENCE
4.1. Analytical Solutions
Analytical solutions for consolidation settlement are
primarily based on Terzaghis [20] one-dimensional
consolidation theory, where the vertical strain is
calculated for a given change in pore pressure acting
across a compressible stratum:

o
o
o
o
c
p
p p
e
H
C s
+
+
= log
1
(1)

Here, s is the vertical settlement, C
c
the compression
index, H
o
the initial thickness of the compressible
stratum, e
o
the initial void ratio, and p
o
and p the
initial and change in pore pressure, respectively. An
example using this formulation is provided by
Attewell et al. [4] for a tunnel in silty alluvial clay.
The key assumption required by such analytical
solutions is that the zone of influence around the
tunnel can be delineated and that the change in pore
pressure is uniform across it. However, the pore
pressure distribution will naturally vary between the
free drainage boundary condition represented by the
tunnel opening and the far field boundary conditions.
Factors relating to plastic yielding around the tunnel,
anisotropic permeability conditions and geological
heterogeneity will all significantly influence the
evolution of pore pressures during tunnel drainage and
groundwater drawdown.
4.2. Empirical Relationships
Empirical databases compiled for subsidence
prediction are almost exclusively focussed on shallow
urban tunnels in granular and cohesive soils (e.g.
[21]). From these, design charts are developed that
depict trends for maximum subsidence and shape of
the subsidence profile based on that of a symmetrical
inverted normal distribution curve (e.g. Fig. 1a). In
one such study, Rankin [21] found that the overall
trough width of detectable surface settlements can be
estimated as three times the tunnel depth. However, it
should be noted that the case histories used in
Rankins assessment are heavily weighted towards the
influence of ground loss during excavation.
This points to a general deficiency of empirical
databases in that they are holistic and disregard
details of the underlying mechanisms. In the case of
the above cited example [21], the influence of both
ground loss and long-term consolidation are combined
together in the observations used to form the empirical
relationships. Thus, if ground loss is not an issue, as is
the case for the Gotthard highway tunnel, then such
empirical relationships are likely not applicable. As an
aside, the width of the measured subsidence trough
over the Gotthard highway tunnel is more than ten
times its depth. Care must be that the case histories on
which empirical relationships are based are applicable
to the case in question.
4.3. Numerical Analysis
To better assess the ground response to tunnel
drainage, the finite-element method (FEM) has
established itself as the conventional means by which
to solve consolidation settlement problems. Examples
include [22, 23], with variations being extended for
fractured rock masses by [12, 24, 25]. In most cases,
the rock mass is treated as an equivalent continuum,
where fractures are treated implicitly and the problem
domain is assumed to be symmetric about the vertical
tunnel axis. Solutions often invoke Biots [26] 3-D
consolidation theory, which describes the transient
coupled hydro-mechanical response of a linear elastic,
isotropic, homogeneous porous medium.
Eberhardt et al. [25] describe the incorporation of
Biots 3-D consolidation and related effective stress
laws [27], into a finite-element formulation and its
application to the Gotthard highway tunnel problem.
For this, a 2-D analysis was performed using the
finite-element code VISAGE [28], which incorporates
flow and elastic field solutions coupled through
poroelasticity. Each simulation required as input
several independent parameters including drained
114
Youngs modulus and Poissons ratio, Biots and
Skemptons coefficients, and the permeability tensor.
These were established through laboratory tests and
field-based estimates scaled to field values through
mapping and rock mass characterization [15, 25]. The
2-D mesh was designed to replicate local
topographical, geological and hydrological conditions
in the study area (Fig. 6a, b). Of key importance was
the representation of the fracture drainage network
intersecting the tunnel. This consisted of a primary
sub-vertical drainage conduit (Fig. 6a), representing
the steeply inclined brittle fault zone that produced the
peak tunnel inflows (Fig. 4b), fed by a less permeable
equivalent-continuum domain representing the
smaller-scale fracture permeability network.
Results from the continuum analysis (Fig. 6c)
showed that a good fit could be achieved with the
observed maximum settlement magnitude when
constrained by field observations. However, the fit to
the shape of the subsidence trough was poor, and for
the most part, the magnitudes of vertical
displacements were under predicted. A better fit to the
width and asymmetry could be obtained by including
additional tunnel drainage points or increasing the
horizontal to vertical hydraulic conductivity ratio [25],
but for the most part, these variations could not be
supported by field observations.

Fig. 6. a) Schematic model geometry and boundary
conditions for continuum-based Gotthard highway tunnel
consolidation subsidence analysis; b) Associated finite-
element mesh; c) Comparison of finite-element results and
measured Gotthard subsidence profile.
5 DISCONTINUUM ANALYSIS OF TUNNEL
INDUCED-CONSOLIDATION SUBSIDENCE
The nature of the rock mass affected in the Gotthard
example (fractured, crystalline rock), points to a
scenario where the shape and asymmetry of the
measured subsidence profile was controlled directly
by the consolidation of the fracture network. The
change in mechanical aperture of a compliant fracture
due to a drop in pore pressure can be shown through
the effective stress law for fracture closure to be:

n
n
n
k
u
'

= (2)

where u
n
is the change in normal deformation of the
fracture,
n
is the change in effective normal stress
and k
n
is the normal stiffness of the fracture.
Intuitively, the frequency and normal stiffness of
sub-horizontal fractures would have the most impact
on surface subsidence as closure of these fractures
would directly contribute to vertical displacements.
However, mapping of the major conductive structures
in the region (i.e. brittle faults, Fig. 7a) showed that
the majority of these structures were steeply inclined
(Fig. 7b), forming a fan-like structure along the tunnel
alignment (Fig. 7c). This required alternative fracture
consolidation mechanisms to be considered.
5.1. Conceptual Models
Zangerl et al. [8] developed several deformation
models to explain consolidation subsidence in
fractured crystalline rock. These included the closure
of sub-horizontal joints, the deformation of sub-
vertical joints and brittle fault zones through changes
in effective normal stress, and the poroelastic
consolidation of the intact rock matrix (Fig. 8).
The model for horizontal joints simply relates
vertical displacements to normal closure of the
fracture due to changes in the normal effective stress
(Fig. 8a). In the case of vertical joints, both the total
and effective stresses acting normal to the fracture
plane (i.e. horizontal) are assumed to change during
drainage, but the vertical stress remains constant (Fig.
8b). The resulting drop in the effective normal stress
would enable vertical slip to occur along the fracture.
Decreases to the horizontal total normal stress relative
to the constant vertical stress would generate strains
within the intact rock blocks, resulting in a Poissons
ratio effect where the intact rock would experience
expansion in the horizontal direction and shortening in
the vertical direction. For this case, vertical joints and
faults are differentiated where the shear and normal
stiffness values for the brittle fault zones were much
lower than those for the unfilled joints (Fig. 8c).
115

Fig. 7. a) Schematic illustration of a typical brittle fault zone mapped in the Gotthard region; b) Stereonet pole plot (lower
hemisphere) of brittle fault planes mapped within and above the Gotthard highway tunnel; c) Cross section along the
Gotthard highway tunnel showing the fan-like pattern formed by steeply inclined brittle fault zones. See Figure 5 for
corresponding geological map of area and legend. After [19].

Fig. 8. a) Conceptual deformation models showing mechanical response to fluid drainage along: a) horizontal joints, b)
vertical joints, c) vertical brittle fault zones, and d) within intact rock matrix. After [8].

116
5.2. Distinct-Element Analysis
Based on the above deformation models, a series of
numerical simulations were performed to better
understand the role of fracture consolidation in
influencing the asymmetry seen in the subsidence
profile above the Gotthard highway tunnel. To do so,
the distinct-element code UDEC [29] was used. The
distinct-element method treats the problem domain as
an assemblage of impermeable, deformable blocks for
which the dynamic equations of equilibrium are
solved until the boundary conditions and laws of
contact and motion are satisfied. The method accounts
for complex non-linear interaction between blocks
(i.e. slip and/or opening/closing along discontinuities),
and through the effective stress law and an aperture-
flow coupling relationship, is capable of modelling the
hydro-mechanical response of a fracture network to
tunnel drainage.
For the Gotthard analysis, discrete fractures were
added to the model based primarily on the steeply
inclined brittle fault zones mapped from within the
safety tunnel and at surface (Fig. 9a). Thus, the
fracture spacing and geometry for these faults were a
direct (i.e. explicit) representation of those mapped in
situ. Normal and shear stiffness values for the faults
were assumed to be constant and equal to 0.5 and 0.1
MPa/mm, respectively. The transition from elastic
shear displacement along the fault to plastic slip was
dictated using a Coulomb-slip law where cohesion
was set to zero and the joint friction angle,
j
, equalled
30. Next, horizontal joints were added to the model
to provide connectivity for fluid flow. The normal set
spacing for these joints were likewise based on
mapping data, with spacing values decreasing with
depth. Normal stiffness values for the horizontal joints
were based on a semi-logarithmic closure law, where
values were varied with depth (i.e. as a function of
increasing normal stress).
Integration of the conceptual hydrogeological
model into UDEC was achieved by calibrating the
sub-vertical hydraulic conductivities of the
representative fault zones based on their
transmissivities (as estimated by Ltzenkirchen [17]).
Hydraulic boundary conditions along the side
boundaries were set as impermeable (i.e. no flow
boundaries). A mean water table was set 500 m above
the tunnel elevation, with surface recharge being
accounted for through a constant pore pressure
condition applied to the upper boundary (Fig. 9b).
Results from the discontinuum models showed
that vertical displacements are generated through
shear deformation and slip along the steeply inclined
faults upon tunnel drainage (Fig. 9c). Furthermore, the
models also confirmed the presence of the Poisson
ratio effect as previously described. Of key
importance though, was that the distinct-element
models were able to reproduce most of the asymmetry
and small-scale inflections with respect to the shape of
the subsidence profile (Fig. 9d). It should be noted
though, that these models could only reproduce 75%
of the total surface settlement magnitudes (assuming
low normal stiffness values), as limitations in the
UDEC formulation, where the blocks are treated as
being impermeable, do not enable the contributing
poroelastic effects of the intact rock matrix to be
considered. Laboratory tests by Zangerl [15] suggest
that these are sizeable and should not be discounted.

Fig. 9. a) Distinct-element model with explicit
representation of brittle fault zones mapped within Gotthard
highway safety tunnel; b) Model boundary conditions; c)
Discontinuum results showing shear displacements along
discontinuities; d) Comparison of distinct-element results
and measured Gotthard subsidence profile.
117
6 DISCUSSION & CONCLUSIONS
Conventional methods used to calculate tunnel-
induced consolidation subsidence, both analytical and
numerical, are largely based on continuum mechanics
where the subsidence profile calculated is symmetric
about the tunnels vertical axial plane. These methods
have largely been developed based on experiences in
soft ground tunnelling where the assumption of a
homogeneous geological continuum is generally valid.
In hard rock tunnelling, the ground conditions
encountered are instead often heterogeneous and
discontinuous owing to the presence of jointing and
faults. These features cause the rock mass to behave
as a discontinuum where fractures open or slip/shear
when the effective stress conditions change. The result
is that the development of any tunnel-induced surface
displacements will likely be asymmetrical.
This was the case for a subsidence trough that
developed over the Gotthard highway tunnel, a tunnel
excavated at several hundred metres depth through
fractured crystalline rock. Completely unforeseen,
more than 12 cm of subsidence was measured above
the tunnel at a point coinciding with large tunnel
inflows along sub-vertical faults. The implications are
thus significant for the 57-km long Gotthard Base
Tunnel currently under construction, as its trajectory
passes through similar rock mass conditions as those
of the Gotthard highway tunnel as well as close to
several important concrete dams.
Comparison of numerical results based on
continuum and discontinuum techniques demonstrated
that finite-element (i.e. continuum) models were able
to reproduce the maximum settlement magnitude
measured when constrained by field observations, but
could not reproduce the asymmetric shape of the
subsidence profile. This resulted in the under-
prediction of vertical displacements away from the
centre of the subsidence trough. A better fit of the
width and asymmetry of the subsidence profile was
instead achieved by distinct-element (i.e.
discontinuum) models where mapped geological
structures within and above the tunnel could be
explicitly included.
Through this study, it has been established that
detrimental consolidation settlements in relation to a
deep hard rock tunnelling project are possible by
means of fracture drainage and consolidation. Results
further demonstrate the importance of detailed field
investigation, monitoring and selection of the correct
analysis method for the given ground conditions
encountered. One of the major limitations to the
present case study, was that pore pressure drawdown
due to tunnel drainage could not be well constrained;
prior to construction, the prospect of generating
surface displacements several hundred metres above
the tunnel was not considered and therefore data
relating to pore pressure evolution was not recorded.
In future, continuous spatial and temporal deformation
and pore pressure measurements are recommended in
cases where a deep hard rock tunnel excavation will
pass under strain sensitive surface structures.
ACKNOWLEDGEMENTS
The authors would like to thank the maintenance team
of the Gotthard highway tunnel for their kind support
of this work and the AlpTransit Gotthard AG for the
permission to publish the settlement data.
REFERENCES
1. Harris, D.I., R.J. Mair, J.B. Burland, and J.R. Standing.
2000. Compensation grouting to control tilt of Big Ben
Clock Tower. In Proceedings of the International
Symposiumon Geotechnical Aspects of Underground
Construction in Soft Ground, Tokyo, eds. O. Kusakabe
et al., 225-232. Rotterdam: A.A. Balkema.
2. Glossop, N.H. and M.P. O'Reilly. 1982. Settlement
caused by tunnelling through soft marine silty clay.
Tunnels & Tunnelling 14: 13-16.
3. Schmidt, B. 1989. Consolidation settlement due to soft
ground tunneling. In Proceedings of the Twelfth
International Conference on Soil Mechanics and
Foundation Engineering, Rio de J aneiro, 797-800.
Rotterdam: A.A. Balkema.
4. Attewell, P.B., I.W. Farmer, and N.H. Glossop. 1978.
Ground deformation caused by tunnelling in a silty
alluvial clay. Ground Engineering 11: 32-41.
5. Barton, N., L. Hrvik, M. Christianson, S.C. Bandis, A.
Makurat, P. Chryssanthakis, and G. Vik. 1986. Rock
mechanics modelling of the Ekofisk reservoir
subsidence. In Proceedings of the 27th U.S. Symposium
on Rock Mechanics, University of Alabama, ed. H.L.
Hartman, 267-274. Littleton, CO: Society of Mining
Engineers.
6. Xie, H., G. Yu, L. Yang, and H. Zhou. 1998. The
influence of proximate fault morphology on ground
subsidence due to extraction. International J ournal of
Rock Mechanics and Mining Sciences 35: 1107-1111.
7. Wu, J.-H., Y. Ohnishi, and S. Nishiyama. 2004.
Simulation of the mechanical behaviour of inclined
jointed rock masses during tunnel construction using
Discontinuous Deformation Analysis (DDA).
International J ournal of Rock Mechanics and Mining
Sciences 41: 731-743.
8. Zangerl, C., E. Eberhardt, and S. Loew. 2003. Ground
settlements above tunnels in fractured crystalline rock:
numerical analysis of coupled hydromechanical
mechanisms. Hydrogeology J ournal 11: 162-173.
9. Loew, S., H.-J. Ziegler, and F. Keller. 2000. Alptransit:
Engineering geology of the world's longest tunnel
system. In GeoEng 2000: Proceedings, International
Conference on Geotechnical & Geological
118
Engineering, Melbourne, 927-937. Lancaster:
Technomic Publishing.
10. Attewell, P.B. 1998. An overview of site investigation
and long-term tunnelling-induced settlement in soil. In
Engineering Geology of Underground Movements, eds.
F.G. Bell et al., 55-61. London: The Geological
Society.
11. Karlsrud, K. and L. Sander. 1979. Subsidence
problems caused by rock-tunnelling in Oslo. In
International Conference on Evaluation and Prediction
of Subsidence, Pensacola Beach, FL, ed. S.K. Saxena,
197-213. New York: American Society of Engineers.
12. Lombardi, G. 1992. The FES rock mass model - Part 2:
Some examples. DamEngineering III: 201-221.
13. Kohl, T., S. Signorelli, and L. Rybach. 2001. Three-
dimensional (3-D) thermal investigation below high
Alpine topography. Physics of the Earth and Planetary
Interiors 126: 195-210.
14. Salvini, D. 2002. Deformationsanalyse im
Gotthardgebiet. Report #297, Institute of Geodesy and
Photogrammetry, ETH Zurich.
15. Zangerl, C. 2003. Analysis of surface subsidence in
crystalline rocks above the Gotthard highway tunnel,
Switzerland. D.Sc. thesis, Engineering Geology, Swiss
Federal Institute of Technology (ETH Zurich), Zurich,
Switzerland. 190 pp.
16. Labhart, T.P. 1999. Aarmassiv, Gotthardmassiv und
Tavetscher Zwischenmassiv: Aufbau und
Entstehungsgeschichte. In Vorerkundung und
Prognose der Basistunnels am Gotthard und am
Ltschberg, Zurich, eds. S. Lw and R. Wyss, 31-43.
Rotterdam: A.A. Balkema.
17. Luetzenkirchen, V.H. 2003. Structural geology and
hydrogeology of brittle fault zones in the central and
eastern Gotthard massif, Switzerland. D.Sc. thesis,
Engineering Geology, Swiss Federal Institute of
Technology (ETH Zurich), Zurich, Switzerland. 264
pp.
18. Zangerl, C., E. Eberhardt, and S. Lw. 2001. Analysis
of ground settlements above tunnels in fractured
crystalline rocks. In Rock Mechanics - A Challenge for
Society, Proceedings of the Eurorock Symposium,
Espoo, Finland, eds. P. Srkk and P. Eloranta, 717-
722. Rotterdam: A.A. Balkema.
19. Zangerl, C., S. Loew, and E. Eberhardt. 2006.
Structure, geometry and formation of brittle
discontinuities in anisotropic crystalline rocks of the
Central Gotthard Massif, Switzerland. Eclogae
Geologicae Helvetiae: In Review.
20. Terzaghi, K. 1943. Theoretical Soil Mechanics. New
York: John Wiley & Sons.
21. Rankin, W.J. 1988. Ground movements resulting from
urban tunnelling: predictions and effects. Engineering
Geology of Underground Movements, eds. F.G. Bell et
al., 79-92. London: The Geological Society.
22. Gudgeon, D.L., M.F. Warner, and J. Stowell. 1988.
Prediction of settlement due to dewatering for deep
excavations. Engineering Geology of Underground
Movements, eds. F.G. Bell et al., 377-386. London:
The Geological Society.
23. Anagnostou, G. 2002. Urban tunnelling in water
bearing ground - Common problems and soil-
mechanical analysis methods. In Proceedings of the
2nd International Conference on Soil Structure
Interaction in Urban Civil Engineering, Zurich, 233-
240. Zurich: Swiss Federal Institute of Technology.
24. Abousleiman, Y., M. Bai, and J.-C. Roegiers. 1996.
Analysis of tunnel stability in fractured carbonate
rocks. In Proceedings, Third International Conference
on Computer Methods and Water Resources, Beirut,
eds. Y. Abousleiman et al., 347-362.
25. Eberhardt, E., K. Evans, C. Zangerl, and S. Loew.
2004. Consolidation settlements above deep tunnels in
fractured crystalline rock: Numerical analysis of
coupled hydromechanical mechanisms. In Coupled
Thermo-Hydro-Mechanical-Chemical Processes in
Geo-Systems, eds. O. Stephansson et al., v2, 759-764.
Amsterdam: Elsevier.
26. Biot, M.A. 1941. General theory of three-dimensional
consolidation. J ournal of Applied Physics 12: 155-164.
27. Nur, A. and J.D. Byerlee. 1971. An exact effective
stress law for elastic deformation of rock with fluids.
J ournal of Geophysical Research 76: 6414-6419.
28. V.I.P.S. 2003. VISAGE - Vectorial Implementation of
Structural Analysis and Geotechnical Engineering,
v.8.7. Bracknell, UK: Vector International Processing
Systems Limited.
29. Itasca. 2000. UDEC - Universal Distinct Element
Code, v.3.1. Minneapolis: Itasca Consulting Group,
Inc.

119
Page 1
1 THE NIAGARA RIVER WATER DIVERSION
AND GENERATION CAPACITY
1.1. Introduction
The Niagara River is an international waterway
Iorming part oI the boundary between Canada and the
United States oI America. The river is about
53 kilometers (km) in length with an average Ilow oI
approximately 6000 cubic metres per second (m
3
/s).
The hydroelectric resource at Niagara is shared
with the United States oI America in accordance with
the terms oI the 1950 Niagara Diversion Treaty. This
treaty established priority Ior scenic, domestic and
navigation purposes and allows the remaining Ilow to
be used Ior power generation. The scenic Ilow
requirement is 2832 m
3
/s during the daytime Irom
April through October and 1416 m
3
/s at all other
times. About two-thirds oI the average Niagara River
Ilow is available Ior power generation and is shared
equally by Canada and the United States.
1.2. Developing More Power
In the 1980`s, Ontario Hydro, the predecessor to
Ontario Power Generation (OPG), began exploring
the possibility oI developing more power at the Sir
Adam Beck (SAB) Niagara generating complex.
Preliminary engineering and an environmental
assessment (EA) were undertaken on the proposed
new Niagara River Hydroelectric Development
(NRHD). The NRHD EA was approved by Ontario`s
Ministry oI Environment in October 1998. The
approved project included construction oI two
additional diversion tunnels and an underground
generating station north oI the existing SAB
generating stations.
Table 1. Diversion and Generation Capacities
In-
Service
Year
Diversion
Capacity
(m
3
/s)
Station
Capacity
(MW)
Annual
Energy
(GWh)
SAB 1 1922 625 487 2,700
SAB 2 1954 1,200 1,472 9,200
SAB PGS 1958 - 122 100
Current Totals 1,825 2,081 11,800
Niagara Tunnel 2009 500 - 1,600
Future Totals 2,325 2,081 13,400
In July 2004, OPG decided to proceed with the
Niagara Tunnel Project, a design/build project Ior one
oI the diversion tunnels. This tunnel will divert a
nominal 500 m
3
/s oI water to OPG`s SAB Generating
The Niagara Tunnel Project An Overview
Russel Delmar, P.Eng.
Construction Manager, Hatch Mott MacDonald, Niagara Falls, ON, Canada
Harry Charalambu, P.Eng.
Profect Manager, Hatch Mott MacDonald, Niagara Falls, ON, Canada
Ernst Gschnitzer, Ph.D.
Profect Manager, Strabag, Niagara Falls, ON, Canada
Rick Everdell, P.Eng.
Profect Director, Ontario Power Generation, Toronto, ON, Canada
: The award oI the design/build contract Ior construction oI Ontario Power Generation`s (OPG) Niagara
Tunnel Project was made on September 1, 2005, aIter an eight month international procurement process. The project
requires delivery oI a nominal 500 m
3
/s oI water Irom the Niagara River Irom an intake located upstream Irom Niagara
Falls via a 10.4-km tunnel running underneath two existing water delivery tunnels beneath the City oI Niagara Falls, to an
outlet at the existing Sir Adam Beck generating station complex. The tunnel will be excavated by means oI the world`s
largest hard rock TBM, a 14.44-m diameter open gripper machine. Vertical alignment oI the tunnel is constrained by an
ancient buried gorge at the north end and existing abandoned power generating Iacilities at the south end. Approximately
80 oI the tunnel will be in Queenston Shale, which exhibits both squeezing and swelling characteristics. The tunnel lining
will be a two-pass system with rock bolts, mesh, steel ribs and shotcrete as the primary lining and cast-in-place concrete,
with a double layer waterprooIing membrane, as the Iinal lining. Under the contract, subsurIace geotechnical risk is
assigned by means oI a negotiated Geotechnical Baseline Report (GBR).
120
Page 2
Station complex allowing more eIIicient use oI the
Niagara River Ilow available to Canada Ior power
generation, Iacilitating an increase in average annual
energy output oI about 1600 GWh (14), which is
enough to supply the annual needs oI a city oI
approximately 160,000 residents. The second
diversion tunnel and underground powerhouse could
be constructed in the Iuture, depending on energy
requirements and project economics.
The Ilow in the Niagara River that is available to
Canada Ior power generation varies Irom about
1000 m
3
/s to 3000 m
3
/s and as indicated in Figure 1,
exceeds the existing SAB diversion capacity (canal
and two tunnels) about 65 oI the time. With the
addition oI a nominal 500 m
3
/s diversion capacity
Irom the new tunnel, the available Ilow will exceed
SABs diversion capability only about 15 oI the
time.
Fig. 1. Water Availability in the Niagara River
2 THE NIAGARA TUNNEL PROJECT
DESIGN/BUILD PROPOSAL PROCESS
In July 2004, OPG decided to proceed with Phase 1 oI
the Niagara Tunnel Project which was to procure a
contractor based on the Iollowing objectives:
to minimize the project duration
to capture contractor experience and innovations
to appropriately allocate project risks
to provide as much price certainty as possible.
It was, thereIore, decided to proceed with a
design/build process as this was considered the best
approach to achieving these objectives. The proposal
process included a number oI distinct stages:
an international invitation Ior expressions oI interest
a prequaliIication process
proposal preparation and submission
proposal evaluation and negotiation culminating in
the award oI a design/build contract.
OI the seven contractors that submitted
expressions oI interest, Iour were pre-qualiIied to
submit proposals on the basis oI a number oI
evaluation criteria that included relevant design and
build experience and saIety perIormance. Three
proposals were received aIter a Iive month proposal
period and this was Iollowed by proposal evaluation
and negotiations with all three proponents. Evaluation
criteria included the design and construction
approach, cost, risk proIile, tunnel Ilow capacity,
schedule, project team, health & saIety management,
environmental management and quality management.
A recommendation oI a design/build contractor was
made to OPG Board oI Directors in July 2005 and the
project was approved by OPG and the Government oI
Ontario by August 17, 2005.
Phase 2, the detailed design and construction oI
the project, commenced on August 18, 2005, with the
award oI the design/build contract to Strabag AG oI
Austria with local sub-contractor, DuIIerin
Construction, assisted by designers ILF Irom Austria
and Morrison HershIield oI Toronto.
2.1. Stakeholders to the Niagara Tunnel Profect
A number oI stakeholders, as summarized in Table 2,
will inIluence the successIul completion oI the
project.
Table 2. Key Stakeholders
Stakeholder Responsibility
OPG Shareholder
Government oI Ontario
Provide Direction to OPG
and Financing
Owner/Operator
OPG/Niagara Plant
Group
Provide Project Direction
and Oversight
Operate and Maintain New
Tunnel and Gates
Owner`s Representative
with Hatch Acres
Administer Design/Build
Contract
Review Design and Monitor
Construction
Design/Build Contractor
Strabag AG
Execute Design/Build
Contract
Regulators
MOE, MNR, DFO,
NPCA
Monitor Compliance with
EA Approval
Issue Permits/CertiIicates
oI Approval
Host Municipalities
Niagara Region
Niagara Falls
Niagara-on-the-Lake
Manage Forecast Tourism
Impacts
Provide Agreed Municipal
Services
Issue Municipal Permits
The Ontario Government, as OPG`s sole
shareholder and project Iinancier, endorsed and
approved the project as being consistent with its
objective oI promoting the development oI cost
competitive, environmentally Iriendly sources oI
Niagara River - OPG EntitIement - MonthIy FIow Duration Curve
Period: Jan 1926 - Dec 2003
1,000
1,200
1,400
1,600
1,800
2,000
2,200
2,400
2,600
2,800
3,000
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Exceedence [%]
F
I
o
w

[
m
3
/
s
]
35,315
42,378
49,440
56,503
63,566
70,629
77,692
84,755
91,817
98,880
105,943
F
I
o
w

[
c
f
s
]
Existing Canal and Two Tunnels
One Additional Tunnel
Two Additional Tunnels
121
Page 3
electricity generation. OPG and its Niagara Plant
Group (NPG) are the Owner and Iacility Operator
respectively. OPG selected Hatch Mott MacDonald in
association with Hatch Acres as its Owner`s
Representative to administer the design/build contract
and provide engineering review and construction
monitoring services. Strabag AG Irom Austria was
awarded the design/build contract. A number oI
regulatory agencies are responsible Ior monitoring the
terms oI the Environmental Assessment Approval and
Ior issuing permits. In addition, the project is hosted
by a number oI local municipalities that include the
Regional Municipality oI Niagara, City oI Niagara
Falls and Town oI Niagara-on-the-Lake.
A Community Impact Agreement (CIA) was
signed in December 1993, between Ontario Hydro
(now OPG), the Regional Municipality oI Niagara, the
City oI Niagara Falls, and the Town oI Niagara-on-
the-Lake. The CIA outlines compensation Ior
anticipated project impacts, and addresses:
Municipal Approvals
Liaison Committee
Monitoring and Remedial Programs
Tourism Impact Management
Transportation Impact Management
Municipal Sewage Collection and Treatment
Municipal Water Supply
Municipal Garbage Disposal
Emergency Services (where permitted and cost
eIIective).
3 DESIGN/BUILD CONTRACT
3.1. Mandatorv Requirements and Liquidated
Damages/Bonuses
Proposals Ior the Niagara Tunnel Project were based
on an invitation Ior proposal document that mandated
a number oI key technical requirements including:
construction oI a 10.4-km long TBM driven tunnel
Irom an outlet on OPG property near the existing
SAB generating stations, along a mandated
horizontal alignment, to an intake in the Niagara
River
design liIe oI the Iacility to be 90 years including
measures to deal with swelling rock conditions in
the local shale units
tunnel to deliver a nominal 500 m
3
/s water Ilow.
For construction scheduling, the Invitation
required contractors to propose substantial completion
dates which were then taken into account during
proposal evaluation (earliest substantial completion
was evaluated most beneIicially). This date was then
used as the contractual substantial completion date Ior
the successIul contractor. Similarly, proponents were
required to propose a guaranteed Ilow amount, based
on the prevailing hydraulic heads and their selected
tunnel diameter and Iinal lining characteristics. These
Ilow amounts were also taken into account during
proposal evaluation (largest guaranteed Ilow amount
was evaluated most beneIicially) and this Ilow amount
was then adopted as the contractual guaranteed Ilow
amount. The contract includes bonuses Ior exceeding
the guaranteed Ilow amount and Ior early substantial
completion, and liquidated damages Ior Iailure to
achieve the same.
3.2. Construction Schedule
A summary level project schedule is provided in
Figure 2.
Fig. 2. Summary Schedule Ior the Niagara Tunnel Project
122
Page 4
3.3. Geotechnical Baseline Report
The geotechnical baseline report (GBR) was
produced by means oI a negotiated process
whereby:
an initial GBR (GBR-A), prepared by OPG, was
included in the Invitation that allowed proponents
to revise the GBR as required to suit their
proposed means and methods.
a proposal, GBR (GBR-B) submitted with the
proponents proposal and reviewed and evaluated
as part oI the proposal evaluation with changes
made during the negotiation process.
the Iinal negotiated GBR(C) (the GBR) was then
included in the design/build contract.
3.4. Scope of Work
The project comprises three major elements oI
work, namely, Intake Iacilities, Outlet Iacilities and
Diversion Tunnel.
Intake
The tunnel Intake Iacilities consists oI a submerged
reinIorced concrete bell-mouth intake structure in
the Niagara River, beneath Bay 1 oI the existing
International Niagara Control Works (INCW)
structure, and a 170 m long underwater approach
channel in the river bed. The intake structure
accommodates sectional gates Ior closure oI the
tunnel when required Ior dewatering. To Iacilitate
construction oI the Intake in the dry, a temporary
coIIerdam will be installed upstream Irom Bay 1.
Extensive curtain grouting oI the highly permeable
bedrock Iormations will be required prior to drill
and blast excavation down to tunnel invert
elevation, a depth oI approximately 40 m. A 5 m
minimum diameter x 250 m minimum length
grouting gallery will be drill and blast excavated
along the diversion tunnel axis to enable pre-
grouting prior to arrival oI the TBM. Depending on
Intake works productivity, schedule, design,
environmental and logistical considerations, a Iull
diameter grouting gallery may be excavated thereby
reducing critical path TBM tunnelling. Other work
at the Intake includes removal oI an existing 520-m
long ice accelerating wall that extends upstream
Irom Pier 4 oI the INCW structure and installation
oI a parallel new precast concrete accelerating wall
upstream oI Pier 5. A new 360-m long precast
concrete approach wall will also be constructed
along the shoreline. Work on the approach channel,
accelerating and approach walls will be carried out
as a marine-based operation.
Fig. 3. Niagara Tunnel Project Elements and Layout oI the Work
123
Page 5
Outlet
The tunnel Outlet consists oI a reinIorced concrete
outlet structure discharging into a 390 m long canal,
excavated in rock, connecting to the existing Pump
Generating Station canal near the SAB Generating
Stations. A closure gate and hoist will be provided
at the Outlet to permit closure oI the tunnel Ior
emergencies and Ior dewatering. During
construction, the outlet canal also acts as the
assembly and launch site Ior the TBM and as the
staging area Ior logistics to and Irom the bored
tunnelling operation. A 40 m long rock plug,
between the new and existing PGS canal, will be
leIt in place Ior the duration oI construction. One oI
the Iinal construction activities will be removal oI
this plug at the time oI watering up oI the tunnel.
Diversion Tunnel
The approximately 10.4-km long Diversion tunnel
will be constructed as a two-pass tunnelling system
with boring taking place Irom the Outlet canal to the
Intake excavation. A 14.44 m diameter Robbins
open gripper rock machine will be used. This will
be the largest diameter rock machine in the world to
date. An initial rock support lining will be installed
Irom the TBM and trailing gear Iollowed by an in-
situ placed concrete lining aIter completion oI TBM
tunnelling. Approximately 1.7 million cubic metres
oI excavated spoil material will be transported Irom
the TBM by conveyor belt and disposed on OPG
property between the two existing power canals.
Five dewatering shaIts oI 0.75 m diameter will be
constructed at the lowest point oI the tunnel to allow
tunnel dewatering iI required.
3.5. Geologv
The Niagara Region is underlain by Cambrian,
Ordovician and Silurian sedimentary rocks having a
total thickness oI approximately 800 to 900 m. The
Niagara River Gorge, the main physiographic
Ieature, and the Niagara Escarpment control the
topography oI the project area. The generally Ilat
lying bedrock strata consists oI dolostone, dolomitic
limestone, sandstone and shale in which the
diversion tunnel will be excavated with about 80
oI the tunnel length in the Queenston Formation, a
siltstone/mudstone with an unconIined compressive
strength ranging Irom 19 to about 45 MPa.
The present Niagara River Gorge was Iormed
by erosion during the last major ice retreat, about
12,000 years ago. The buried St. Davids Gorge
represents an earlier river course that has been in-
Iilled with glacial outwash materials. Away Irom
the gorge areas, the bedrock is covered almost
entirely by glacial lake sediments.
The buried St. Davids Gorge is similar in shape
to the Niagara River Gorge and extends Irom Lake
Ontario through the village oI St. David`s to the
Whirlpool. The St. Davids Gorge is oriented in a
northwest direction and varies in width Irom 350 to
630 m in the vicinity oI the Niagara River. Depth to
bedrock is in the order oI 125 m in the vicinity oI
the proposed tunnel alignment and in excess oI
200 m where it intersects with the present Niagara
River at the Whirlpool. The gorge is completely in-
Iilled with deposits oI glaciolacustrine, glacial and
glacioIluvial origin. The bedrock (Queenston
Formation) over the width oI the St. Davids Gorge
is slightly weathered and relatively more Iractured
to a depth oI between 15 to 25 m below the bottom
oI the gorge. Below this depth, the rock is generally
Iresh and oI excellent quality.
Bedrock in the project area has generally well-
deIined bedding with a southerly dip oI about
6m/km and an east-west strike. Sheared, weak
bedding planes exist between many oI the rock
Iormations and within the Queenston Formation.
There are no known occurrences oI any major
Iaulting within the project area, some near-surIace
thrust with minor vertical displacement are known
to occur and are probably related to stress relieI
associated with the gorge Iormation and the high
horizontal residual stresses in the area. Some
shearing oI this type is expected in the area oI the
St. Davids Gorge. Three major near-vertical joint
sets, which strongly inIluenced the physiography oI
the project area, have been identiIied. These sets
strike parallel to the Niagara River; the St. Davids
Gorge and the Niagara Escarpment. Vertical joints
are generally widely spaced. The joint surIaces are
generally rough and Iresh to slightly weathered.
3.6. Bedrock Stratigraphv and Structure
In descending order Irom surIace, the sequence oI
rocks is as presented in Table 3.
The Queenston Formation extends well below
the deepest section oI the tunnel with thickness
greater than 300 m being reported in the literature.
Primary bedding planes are deIined as major
bedding planes between lithological units above the
Queenston Formation and between sub-units within
the Queenston Formation. Sheared primary bedding
planes reIer to those planes where some diIIerential
displacement has occurred. Within the Queenston
Formation, the primary bedding planes are major
discontinuities occurring at spacing oI about 5 m to
somewhat greater than 20 m and locally aIIecting
the rock mass quality.
124
Page 6
Table 3. Rock Sequence to be Encountered During
Tunnelling
Sequence Description Thickness
Guelph Dolostone 2 to 3 m
Lockport Dolostone 43 to 45 m
DeCew Dolostone 2 to 3 m
Rochester Shale 17 to 19 m
Irondequoit Limestone 2 to 4 m
Reynales Dolostone 3.5 to 4.5 m
Neahga Shale 1.5 to 2 m
Thorold Sandstone 2 to 3.5 m
Grimsby Sandstone 12.5 to 15 m
Power Glen Shale 10 to 12 m
Whirlpool Sandstone 4.9 to 8.5 m
Queenston Shale/Mudstone ~300 m
These planes oIten exhibit Ieatures such as
gouge or breccia (a Iew millimetres to 2 to 3 cm)
and slickensides that are consistent with lateral
structural dislocation.
Groundwater conditions in the project area are
inIluenced by depth and lithology, and vary
between the rock Iormations above the Queenston
Formation, but are relatively consistent in the
Queenston Formation. The only known aquiIers are
the Lockport and DeCew (dolostone) Formations,
whereas the remaining strata below the DeCew are
generally considered to be aquitards. Hydraulic
conductivity ranges Irom 10
-7
to 10
-3
cm/s in the
upper dolostone and limestone Iormations and Irom
4 x 10
-3
cm/s to practically impermeable
(10
-7
cm/s) in the shale Iormations. In general, the
groundwater below the DeCew Formation is highly
corrosive.
Natural gas has been encountered in some oI the
Iormations, particularly in the Rochester and
Grimsby Formations, with some minor amounts oI
gas being encountered in other Iormations,
including the Queenston.
High in-situ stresses exist in the project area
bedrock. Measurements along the tunnel horizon
show that maximum horizontal stress in the
Queenston Formation range Irom 10 to 24 MPa,
with a maximum horizontal/vertical stress ratio
varying Irom 3 to 5. In general, the orientations oI
the maximum horizontal stresses along the
alignment oI the diversion tunnel lie within the NE-
SW quadrant. The orientations oI the local stresses
are inIluenced by the presence oI major
physiographic Ieatures, namely the buried
St. Davids Gorge and the Niagara River Gorge.
The Iormations in the project area are subject to
time-dependent deIormations, initiated by the relieI
oI the relatively high in-situ stresses and swelling
on the uptake oI Iresh water. There is a well-
documented history oI rock 'squeeze aIIecting
surIace excavations. The swelling potential oI shale
units in the Niagara area is also well documented.
Swelling involves the volume increase in shale units
and is initiated by the relieI oI the high in-situ stress
in the presence oI Ireshwater. The process is
associated with chloride ion diIIusion Irom the
connate pore water in the rock.
4 TBM, TRAILING GEAR AND BACKUP
EQUIPMENT
Strabag selected a Robbins Open Gripper TBM Ior
the project. The machine, with speciIications as
reIlected in Table 4, was Iabricated in a number oI
countries including the USA (predominantly in
Robbin`s Iactories in Ohio), England (Markham
Iacilities), Canada and Europe.
Table 4. TBM SpeciIications
Description Details
HARD ROCK TBM DESCRIPTION
Machine Diameter
New cutters
(worn cutters)
14.44
14.41
Main Bearing three roller (3 axis)
Bearing liIe ~13,000 L10 hrs 224
kN cutter load
Cutters
Number oI cutters 85 x 19/20
(483/508 mm)
Loading
Individual cutter load
Average cutter spacing
311 kN
89 mm
Cutterhead Recommended Operating
Cutterhead thrust
Maximum thrust
Maximum gripper Iorce
18,426 kN
27,900 kN
71,500 kN
Cutterhead Drive Variable Frequency
Description Cutterhead
Drive/Electric
Motors/Gear
Reducers/VF Drive
Cutterhead power 15x315 kW 4,725 kW
Cutterhead Torque
at 2.4 rpm
at 5.0
18,670 kNm
8,960 kNm
Breakout Torque 28,000 kNm
Thrust Cylinder Stroke 1.82 m
TransIormer (TBM
Drives)
2 x 2500 kVA
TransIormer Backup 1 x 1000 kVA
125
Page 7
Description Details
Conveyor
Capacity 1600 tph
Belt width 1400 mm
TBM weight (approx.) 1900 t
The trailing gear was manuIactured by ROWA
oI Switzerland with Iabrication carried out in
Slovenia and Hungary. Additional backup equipment
such as the muck conveyor was procured Irom HE
Logistics oI Germany, the rubber tired transporters
Irom Plan and Teco oI Germany and ventilation
equipment Irom Cogemacoustic oI France.
5 TUNNEL DESIGN AND CONSTRUCTION
5.1. Alignment Constraints
Vertical alignment oI the tunnel, as reIlected in
Figure 5, is constrained by a number oI existing
structures and geological Ieatures. In proIile, at the
north end the tunnel passes beneath the buried
St. Davids Gorge with bedrock at an approximate
elevation oI 100 m. This requires an initial tunnel
downgrade oI 7.82 Irom the Outlet portal to pass
beneath the buried gorge with minimum rock cover
oI approximately 10 m. At this lowest point the
tunnel axis is approximately 135 m below ground
surIace.
Progressing southward, the tunnel rises at an
upgrade oI 0.1 to a point beneath the existing
diversion tunnels 1 and 2 where the cover between
new and existing tunnels is approximately 24 m. In
this vicinity the tunnel also crosses under the
decommissioned Toronto Power Station which
Ieatures a 50 m deep turbine pit and tailrace tunnels.
From these crossings the tunnel rises at an upgrade
oI 7.15 to the Intake portal beneath Bay 1 oI the
INCW structure. These alignment constraints result
in the tunnel being located predominantly in
Queenston Shale.
Fig. 4. Schematic Robbins Open Gripper TBM
126
Page 8
Horizontal alignment is predominantly
constrained by a combination oI the existing tunnel
right-oI-way (Ior existing diversion Tunnels 1 and
2), the Niagara River, orientation oI the tunnel
intake to the existing INCW structure and minimum
radii curves to suit TBM tunnelling. In plan, at the
north end Irom Sta 0000 to Sta 1426 the
alignment parallels the existing OPG power canal
on OPG property. At Sta 1426 the alignment
curves southward and converges beneath the
existing diversion tunnels Iollowing an existing
easement along Stanley Avenue beneath the
downtown core oI the City oI Niagara Falls. From
Sta 7950 the alignment curves eastward on an
azimuth perpendicular to the existing INCW
structure within an easement beneath Niagara Parks
Commission property. The Iinal design alignment
uses minimum horizontal curve radii oI 1000 m to
Iacilitate muck transportation by conveyor belt.
5.2. Tunnel Construction Sequence
The Diversion Tunnel will be excavated by an open
gripper TBM oI 14.44 m diameter with bored
tunnelling progressing Irom the Outlet end to the
Intake in the Niagara River. Concurrent with bored
tunnelling, a 'grout tunnel will be excavated by
drill and blast technique Irom the Intake. This will
enable grouting oI the highly permeable rock
Iormations adjacent to the river near the Intake, to
be carried out beIore arrival oI the TBM.
Initial lining rock support, to suit encountered
rock conditions, will be installed Irom the TBM and
trailing gear. Excavated muck will be transported
Irom the TBM to the surIace disposal area by muck
conveyor.
Fig. 5. Tunnel ProIile
In order to meet the construction schedule the
invert portion oI the cast-in-place Iinal lining will
be constructed concurrent with bored tunneling.
This will be carried out some distance back Irom the
TBM trailing gear and will include cleaning oI the
invert, diversion oI water seepage, installation and
testing oI the waterprooIing membrane system,
installation oI invert Iormwork and placing oI
concrete. For this purpose, a moving bridge will be
installed allowing transportation to and Irom the
TBM to cross this work area.
Once bored tunnelling is completed,
construction oI the top section oI Iinal lining will be
carried out using 4 x 12 m long Iorms. The
sequence oI membrane installation and concrete
placement essentially Iollows the same pattern as
the invert concrete production. AIter the Iinal lining
is installed, contact and interIace grouting will be
done as Iinal activities prior to Ilooding oI the
tunnel and Ilow testing. Further details oI the initial
and Iinal lining operations are provided below.
5.3. Tunnel Lining
The tunnel lining will be constructed by means oI a
two pass tunnel lining system that comprises the
Iollowing components:
An initial lining that includes a combination oI
steel wire mesh, steel ribs, rock bolts and shotcrete.
A Iinal cast-in-place unreinIorced concrete
lining with a waterprooIing membrane system to
ensure that both water seepage Irom the tunnel and
diIIusion oI chloride ions Irom the rock to prevent
time dependent rock deIormation (swelling) does
not occur.
Contact grouting carried out along the top oI the
Iinal concrete lining aIter completion oI concreting
127
Page 9
to close all voids in the crown and to provide a tight
interIace between rock mass, initial lining and Iinal
lining.
Finally, a high pressure interIace grouting
system is installed to pre-stress the Iinal concrete
lining and surrounding rock such the entire lining
system remains in compression over the design liIe
oI the tunnel thereby eliminating the need Ior
reinIorcing steel.
5.4. Initial Lining
The TBM and trailing gear arrangement allows Ior
installation oI rock support at two locations namely
Location1 and Location2. L1 is between 4 to 7 m
back Irom the excavation Iace and L2 between
20 to 40 m back. Due to proximity to sensitive TBM
electronic and mechanical components and
limitations to TBM advance, installation oI rock
support in general and shotcrete, in particular, at L1
is limited to the minimum required Ior personnel
saIety. In the interests oI tunnel advancement, the
preIerence is Ior installation oI the majority oI rock
support at L2. The assessment oI installation oI
initial lining rock support at locations L1 and L2
takes into consideration the requirements Ior
personnel saIety and the rate oI tunnel
advancement. Depending on encountered rock
conditions, this will result in progressive installation
oI rock support at L1 and L2 with Iull support being
completed 40 m behind the excavation Iace.
Six rock conditions have been baselined in the
GBR Ior which rock support types have been
designed as Iollows:
i Support for Rock Condition 1 (0.2 of
Tunnel)
Applied in stable rock conditions with a uniaxial
compressive strength comparable to lean concrete.
A 50 mm thick sealing layer oI shotcrete reinIorced
with mesh applied Irom L2 iI rock is sensitive to
water or will degrade when exposed to air.
ii Support for Rock Condition 2 (3 of Tunnel)
Applied in L1 to provide saIety Ior personnel
working at the Iront oI the TBM where blocks oI
ground are diIIerentiable in otherwise stable rock
conditions. It consists oI steel ribs (C 100 x 11 steel
channels) bolted with a limited number oI 2.4 m
long rock bolts to the tunnel crown and steel wire
mesh Iixed with the bolts. At L2, 70 mm oI
shotcrete is applied and additional rock bolts are
installed. II necessary in ground sensitive to water,
sealing shotcrete is applied to the invert section
similar to Support Type 1.
iii Support for Rock Condition 3 (approximatelv
11 of Tunnel)
Applied in L1 in Iriable ground, where small blocks
oI ground tend to Iall Irom the tunnel crown iI leIt
unsupported. It consists oI steel C150X16 channels
installed at 1.2 m intervals, 4.0 m long rock bolts
and steel wire mesh. Every second steel channel is
extended to springline. At L2, 100 mm oI shotcrete
and more rock bolts are installed to support the Iull
circumIerence oI the tunnel.
iv Support for Rock Tvpe 4 (approximatelv 29 of
Tunnel)
Applied at L1 where the tunnel crown and primary
bedding planes are close to intersecting and an
increasing number and size oI unstable blocks are
expected. Consisting oI steel C150X16 channels at
0.9 m intervals in the crown and 1.8 m intervals to
the sidewalls (installed Iull round in the Queenston
Formation as Type 4Q), 4.0-m long rock bolts and
steel wire mesh in the tunnel crown. At L2, 130mm
oI shotcrete and additional rock bolts are installed to
complete Iull support oI the tunnel.
v Support for Rock Tvpe 5 (approximatelv 47 of
Tunnel)
Applied in L1 in squeezing ground, where slabbing
and spalling is experienced soon aIter excavation. It
consists oI steel ribs in Iorm oI mid weight
W150X37 steel beams around the Iull
circumIerence oI the cross section at 1.8 m
intervals. As Ior support, Types 3 and 4 steel wire
mesh and 6 m long rock bolts are installed in the
tunnel crown to provide saIety oI the personal
working at the Iront oI the TBM. At L2, 160 mm oI
shotcrete is applied.
vi Support for Rock Tvpe 6 (approximatelv 10 of
Tunnel)
Applied in L1 in exceptional ground conditions,
where spalling and slabbing is experienced even in
Iront oI the excavation Iace. Heavy W200X59 steel
beams, 6-m long rock bolts and 100 mm oI
shotcrete reinIorced with steel wire mesh. An
additional 160 mm oI shotcrete and one additional
layer oI mesh reinIorcement is installed at L2.
5.5. Final Lining
The Iinal lining consists oI cast in place
unreinIorced concrete with a waterprooIing
membrane between the initial and Iinal lining and
both contact and interIace grouting aIter concreting.
128
Page 10
(i) Waterproofing Membrane Svstem
The waterprooIing membrane is designed to act as
an impermeable layer between the initial and Iinal
lining preventing seepage oI Iresh water and
diIIusion oI chloride ions to and Irom the
surrounding rock that is susceptible to swelling
thereby ensuring that time dependent deIormation
oI the rock does not occur. The waterprooIing
membrane system consists oI regulating shotcrete,
where necessary, to smooth corners and edges and
prevent damage oI the waterprooIing membrane
during placement, membrane-backed geotextile
Ileece to protect the waterprooIing membrane
against the shotcrete, a double layer waterprooIing
membrane oI 2 mm and 1.5 mm thickness
respectively consisting oI a Poly-OleIine (or
polyethylene) product, produced in 2 m wide strips
and heat welded together by double seams.
Rigorous and comprehensive quality assurance
and quality control testing is required to ensure that
the selected waterprooIing membrane material and
installation provides a Iinal lining system that is
100 waterprooI and diIIusion resistant and
includes:
laboratory testing to demonstrate chloride
diIIusion barrier characteristics
in-situ pressure testing oI each double weld seem
100 in-situ vacuum testing oI each panel oI the
double layer membrane system, where the inner
layer is manuIactured with dimples to allow
vacuum testing between the two membrane
layers.
(ii) Final Lining Construction
The sequence oI Iinal lining installation includes
initial placement oI invert concrete (and invert
waterprooIing membrane), concurrent with TBM
tunneling, Iollowed by placement oI the remainder
oI the concrete lining system aIter completion oI
tunnelling. The invert will be cast in 12 m or 24 m
long bays. A bridge across the invert concreting
operation will Iacilitate transportation and material
supply to the TBM while placement oI the invert
concrete system is in progress. Final lining oI the
tunnel section above the invert will commence once
the tunnel is excavated. BeIore the Iinal lining
system is installed above the invert, the preset rings
Ior interIace grouting and the waterprooIing
membrane system will be Iixed to the tunnel walls
and crown. Twelve metre long, adjustable diameter,
steel Iorms, placed on the previously installed invert
concrete, will be used Ior concreting the top section
oI the tunnel. Steel Iorms will be used to provide the
smooth concrete Iinish required to limit Iriction
losses to the Ilow oI water in the tunnel. The
adjustable diameter Iormwork allows adjustment oI
the diameter up to 260 mm to accommodate the
variable thickness initial lining and provide a 600 to
700 mm thick Iinal lining.
5.6. Contact Grouting
AIter placement oI Iinal lining concrete, cement
grout contact grouting oI voids in the crown
between the initial and Iinal lining will be carried
out via grouting pipes attached at regular intervals
to the intrados oI the waterprooIing membrane and
extending through the Iinal lining.
5.7. Interface Grouting
InterIace grouting oI the Iinal lining is required to
create a continuous pre-stressed compression
concrete support ring able to sustain internal water
pressure without requiring steel reinIorcement
thereby eliminating the risk oI corrosion oI
structural reinIorcement within the 90 years design
liIe. Grout, at speciIic pre-determined pressure (up
to 30 bar), will be injected through a system oI
grout-hose rings installed between the initial lining
and the waterprooIing membrane system at 3 m
centres. The grout-hose rings have pressure valves
at 3 m circumIerential centres that open under
pressures releasing grout into the joint between
initial lining and the 'geotextile Ileece backing oI
the waterprooIing system. The ends oI the grout
hoses penetrate through the waterprooIing
membrane system and the cast-in-place Iinal lining
into the tunnel. Grout blocking rings will be
installed every 12 m to control the longitudinal Ilow
oI grout.
InterIace grouting pressure Ior each individual
section oI tunnel are calculated taking the Iollowing
considerations into account:
required long term pre-stressing pressure and
associated tunnel convergence
anticipated short term pre-stressing pressure and
associated tunnel convergence including
deIormation allowance Ior shrinkage oI concrete
and temperature contraction aIter watering up
acceptable diIIerential deIormations including the
shrinkage oI concrete beIore watering up.
InterIace grouting will be controlled by precise
in-situ measurement oI lining deIormation during
and aIter interIace grouting. Pumping pressures as
deIined by structural analysis, will thereby be
controlled within allowable limits. Regrouting will
be required iI the tunnel deIormations (convergence
oI tunnel lining) are less than the values anticipated
beIore watering up (i.e., diIIerential deIormation
associated with the conservation oI the minimum
129
Page 11
long term pre-stressing pressure including
temperature contraction aIter watering up).
InterIace grouting will be carried out in two
phases with initial grouting through every second
grout hose and re-grouting, iI required, through
intermediate unused hoses.
6 PROJECT UPDATE
As oI June 2006, approximately nine months into
the project, Phase 2 design and construction work is
on schedule. At the Outlet, excavation oI the Outlet
canal is complete (except Ior the rock plug);
delivery oI the TBM major components is
approximately 50 complete and TBM assembly
has commenced. Most oI the tunneling support
equipment has been delivered, including trailing
gear, ventilation equipment, gantry crane, conveyor
equipment and transIormers. A new 13.8-kV
overhead power line and a 200-mm diameter water
supply line have also been constructed.
At the Intake, marine-based work has
commenced on the installation oI the new ice
acceleration wall, demolition oI the existing
acceleration wall, drill and blast excavation oI the
Intake approach channel and preparation Ior
installation oI sheet pile coIIerdam cells.
7 ACKNOWLEDGMENTS
The authors would like to thank Ontario Power
Generation (OPG) Ior their permission to publish
this paper.
Fig. 6. Outlet Site June 2006
Fig. 7. Intake Site June 2006
130

1 INTRODUCTION
The Kicking Horse Canyon Project (KHCP) involves
upgrading approximately 26 kilometers of the Trans-
Canada Highway in B.C. from the town of Golden to
the western boundary of Yoho National Park to a
modern four-lane standard. Due to the rugged nature
of the canyon section and the requirements of the BC
Ministry of Transportation this upgrading has required
the consideration of tunnels as part of the preliminary
stage of the project.
The KHCP is being undertaken in three phases.
Phase 1 comprised the replacement of the Yoho
Bridge which was completed in 2004. Phase 2
comprises the replacement of the Park Bridge and a
major through cut in rock to a depth of 80 m that is
currently under construction and planned for
completion in late 2007, and Phase 3 comprises
upgrading of the remaining stretch westwards from
the Yoho Bridge to Golden and east of Phase 2.
The preliminary design for the Phase 3 section has
evaluated several new highway alignments through
the canyon that include two to three short (<1 km)
tunnels as well as one alignment that includes a nearly
2.9 km tunnel. The construction for the Phase 3
section may commence in 2008 and both traditional
Design-Bid-Build and Design-Build-Finance-Operate
(DBFO) contract approaches are being considered.

Figure 1. Kicking Horse Canyon circa 1920s.
Both contract approaches have been implemented to
date on past and current works. If constructed, the
Planning for Canadas First Bored Road Tunnel in over 40 Years
Kicking Horse Canyon Project
Dean Brox
Hatch Mott MacDonald, Vancouver, B.C. Canada
Steve Bean, Paulo Branco
Thurber Engineering Limited, Victoria, B.C./Mississauga, ON, Canada
Terry Coulter
Coulter Consulting Limited, Victoria, B.C. Canada
ABSTRACT: The Kicking Horse Canyon Project (KHCP) involves upgrading approximately 26 kilometers of the Trans-
Canada Highway in BC from the town of Golden to the western boundary of Yoho National Park to a modern four-lane
standard. The KHCP is being undertaken in three Phases. Phase 1 comprised the replacement of the Yoho Bridge which
was completed in 2004. Phase 2 comprises the replacement of the Park Bridge and grading work that is currently under
construction, and Phase 3 comprises upgrading of the remaining stretches from the Yoho Bridge to Golden and the Brake
Check to the Yoho Park Boundary. The preliminary design for the Phase 3 West section between Yoho Bridge and Golden
is currently in progress and is evaluating new highway alignments through the steepest terrain section of the canyon that
include two to three short (<1 km) tunnels as well as one alignment that includes a nearly 2.9 km tunnel that would be the
longest bored road tunnel in North America. If constructed, the tunnel requirements for the highway improvement will
represent the first major road tunnels in Canada in over 40 years. The preliminary design for the road tunnels comprises
twin tube, 2-lane road tunnels. The road tunnels would be constructed within mixed sedimentary bedrock comprising
limestones, dolostones, and shales with rock strengths varying from 30 MPa to 240 MPa. State-of-the-art road tunnel fire,
life and safety requirements following NFPA requirements have been adopted given the increasing awareness for road
tunnel safety measures. Comprehensive geotechnical site investigations have been completed as part of preliminary design
studies. Portal location and tunnel design and constructability issues have been evaluated as part of the preliminary design
work and are presented.
131
tunnel requirements for the highway upgrading
represent the first major road tunnels in Canada in
over 40 years since the construction of tunnels along
the Trans-Canada Highway in the Fraser River
Canyon. Most notably, the nearly 2.9 km tunnel
option would be the longest road tunnel in North
America. Figure 1 shows a historical photograph of
the highway through the canyon section.
2 PROJ ECT LOCATION
The Kicking Horse Canyon Project is located
immediately east of the town of Golden, B.C. which is
over 700 km northeast of Vancouver, B.C. and over
250 km west of Calgary, Alberta within the East
Kootenay region of B.C. The section of the highway
improvement project comprises 26 km of existing 2-
lane highway from the town of Golden to the west
gate of Yoho National Park. The Kicking Horse
Canyon comprises a fairly narrow and winding
canyon with maximum relief of nearly 500 m
exhibiting very steep rock slopes as shown in Figure
2.

Figure 2. Kicking Horse Canyon.
The elevation of canyon section of the alignment
is approximately 1000 m and the main canyon section
of the project is generally trending east-west with the
Kicking Horse River flowing westwards through the
canyon. The main CP Rail line is present along the
lower slopes of the canyon and mostly along the
northern side of the canyon below the existing
highway. Figure 3 illustrates the project location and
phases of the project.

Figure 3. Project location and works phases.
3 ROAD TUNNEL ALIGNMENT OPTIONS
Several studies have been completed since the early
1990s to identify highway improvement solutions for
the Kicking Horse Canyon Project. The work from
some of these studies identified several highway
alignments along both the north and south sides
through the canyon section. Recent preliminary work
including terrain and natural hazard studies have
indicated that a major ancient landslide is present
along the south side of the canyon thereby precluding
a highway alignment on the south side.
The remaining alignments along the north side of
the canyon were evaluated further in terms of natural
hazards and constructability in terms of maintaining
the existing highway during construction. Following
further work to date there exist two preferred
alignment options referred to as the NB-2 and NC-2
alignments as shown in Figure 4.
The NB-2 alignment comprises a single,
approximately 2.9 km tunnel that deviates northwards
below a sharp ridge known as Frenchmans Ridge,
passes under Dart Creek valley, a 400 m wide U-
shaped hanging glacial valley, and continues eastward
below the Black Wall Bluffs. The west portal is sited
near the grade of the existing highway and the east
portal is sited about 20 m below grade and at a sharp
bend of the existing highway. Figure 3 shows the NB-
2 alignment. The maximum cover along the NB-2
alignment is about 300 m below the Black Wall
Bluffs. The minimum cover along the NB-2 alignment
occurs within the Dart Creek valley and is under
investigation at the time of writing.
The NC-2 alignment comprises three, short
tunnels with lengths of about 555 m, 565 m and 565 m
respectively for a total tunnel length of about 1685 m.
The NC-2 alignment deviates not as much northwards
as the NB-2 alignment but rather cuts through the
132
series of bluffs/ridges along the canyon. The six
portals required for the NC-2 alignment are sited both
above and below grade of the existing highway.
With the development of possible new
alignments through the canyon section it has been
recognized that it is necessary to address the
requirements of the external stakeholders. These
requirements include maintaining access for CP Rail
and river rafting businesses operating from upstream
of the canyon, and to allow safe passage through the
canyon for bicyclists. These requirements indicate that
at least one lane of the existing highway will have to
be maintained.
4 ROAD TUNNEL CROSS SECTION
Tunnel cross-section geometry was evaluated during
the early stages of the preliminary design in

Figure 4. Road tunnel alignments.
recognition of the potential cost sensitivity and traffic
safety requirements. As it was recognized that no
tunnel roadway geometry criteria exist for road
tunnels, neither in Canada nor elsewhere in the world,
a review of tunnel roadway geometry was undertaken
and three options were developed for safety and
costing considerations. The original tunnel cross
sections varied from 115 m
2
to 127 m
2
and were based
on lane widths of 3.7 m, walkways on both sides of
1.0 m, and varying shoulder widths of 1.5 m, 1.75 m,
and 2.5 m, the largest allowing for emergency vehicle
access (Figure 5). The initial tunnel cross section
options were based on adopting large sized tunnels
rather than minimized size tunnels similar to those in
Europe. Cost comparison of these initial cross sections
did not indicate significant cost increases for the
largest size tunnel of 127 m
2
. Smaller tunnel cross
section options (90 m
2
) have also been developed
based on minimum shoulder dimensions for
consideration based on acceptable international
practice (Figure 6).

5 GEOTECHNICAL INVESTIGATIONS
Comprehensive geotechnical site investigations have
been completed as part of the preliminary design work
for both tunnel alignments. Historical geotechnical
investigations have been completed throughout the
canyon dating as far back as 1985. The recent
fieldwork has comprised seismic refraction surveys at
all portal locations as well as across the Dart Creek
valley, over 1900 m of rotary core drilling, point load
strength testing, in situ packer permeability testing,
and laboratory rock strength and abrasivity testing.
Piezometers have been installed in many of the
boreholes completed along the tunnel alignments.

133

Figure 5. Large Tunnel Cross Section.

Figure 6. Small Tunnel Cross Section.

Access to some of the boreholes at portal
locations and along the tunnel alignments was
difficult and was only accomplished by helicopter. All
boreholes were drilled vertically and the deepest
borehole completed was about 250 m along the Black
Wall Bluffs. Figure 7 illustrates drilling of one of the
deepest boreholes from the upper reaches of the Black
Wall Bluffs by helicopter assistance. A limited
number of deep boreholes were required to investigate
the presence of the Black Wall Bluffs Fault zone
along the eastern section of the tunnel alignments.

Figure 7. Geotechnical drilling by helicopter.
6 TUNNELLING CONDITIONS
6.1. Site Geology
The geology along the proposed tunnel alignments
comprises sedimentary bedrock of the McKay Group
and the lower part of the Glenogle Formation that has
been subjected to eastward thrust faulting and folding
with overturning. The McKay Group of rocks
comprises five main sub-units (COMk2 to COMk6)
that can be characterized as shaley limestones (odd
numbered sub-units) and dolomitized limestones
(even numbered sub-units). These five sub-units are
present along the western and central sections of the
alignment. The shaley limestones are finely laminated
whereas the dolomitized limestones appear to be more
massive in nature. The lower part of the Glenogle
Formation can be described as a mixed
slate/shale/dolomitic siltstone with finely laminated
dolomitic beds.
Extensive bedrock outcrops are present in massive
sub-vertical rock cuts formed by the original highway
construction and appear along almost the entire tunnel
alignments. Figure 8 shows a typical large rock cut of
bedded limestone along the central section of the
tunnel alignments.

134

Figure 8. Existing rock cut.
6.2. Rock Strength and Abrasivity
Numerous point load strength index tests and uniaxial
compressive strength (UCS) test were completed on
the drill core. The rock strength of the shaley and
dolomitized limestones generally varies from 15 MPa
to 100 MPa with an average strength of about 60
MPa. The rock strength of the dolomitic shales and
siltstones generally varies from 20 MPa to 240 MPa
with an average strength of about 100 MPa. Figures 9
and 10 illustrate the variation of rock strength based
on UCS lab testing results of the limestones and
shales respectively. Rock abrasivity was evaluated
based on CERCHAR Testing. CERCHAR Abrasivity
Index (CAI) values for the limestones indicated values
generally ranging from 0.8 to 2.4 characterizing the
rock as slightly to non-abrasive. CAI values for the
shales indicated values generally ranging from 2.0 to
4.0 characterizing the rock as very abrasive.

6.3. Rock Fracturing
Rock fractures are pervasive within the limestone and
siltstone/shale bedrock in the form of bedding and
sub-vertical fractures that can be identified from both
drill core and surface outcrops. The dip of the bedding
within both the limestones and shales typically ranges
from 15 to 20 degrees. The dip direction of the
bedding is northeasterly for the limestone and
northerly for the shales. The main sub-vertical
fractures sets are prominent as orthogonal fractures
and are generally oriented both perpendicular and sub-
parallel to the strike of the bedding. Bedding fractures
are typically very smooth and planar. Figures 11 and
12 illustrate stereonets of the rock fractures within the
limestones and shales, respectively.

Figure 9. Rock strength limestones.

Figure 10. Rock strength shales.

6.4. Major Fault/Shear Zones
The geology along the tunnel alignments has been
intruded by both sub-vertical and sub-horizontal thrust
type faults. Sub-vertical fault/shears have been
mapped from outcrops below the tunnel alignment
along the CP Rail right-of-way and are expected to be
present across the tunnel alignments. The most
distinct sub-vertical faults are referred to as the West,
Middle and Dart Creek Faults. The thicknesses of the
West and Middle Faults are inferred to be no more
than 10 m while the Dart Creek Fault is inferred to be
as much as 20 m in thickness.

0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
1
0
0
1
1
0
1
2
0
UCS, LAB, MPa
F
r
e
q
n
e
y
c
,

%
0%
5%
10%
15%
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
2
0
0
2
2
0
2
4
0
UCS Lab, MPa
F
r
e
n
q
u
e
n
c
y
,

%
135

Figure 11. Rock fractures limestones.

Figure 12. Rock fractures shales.
The most prominent major fault/shear feature
along the tunnel alignments is referred to as the Black
Wall Bluffs Fault. This fault zone is a thrust fault that
is clearly visible in a sub-vertical rock cut at the
western edge of the Black Wall Bluffs. This zone is
comprised of highly distorted siltstone and shale
bedrock as shown in Figure 13 and is currently
inferred to be undulating in nature and may also be
present near the east portal. The overall orientation of
this fault is inferred to be dipping less than 20 degrees
towards the northeast. The fault is present in the
exposed rock cut over a vertical height of about 15 m
and may have an overall thickness as much as 25 m
based on a limited number of borehole intersections
from targeted drilling. It is noted that complete
recovery of this fault was obtained in one of the
boreholes that were targeted.
Interpretation of the possible length of
intersection of this major fault with the proposed
tunnels has been made based on a 3D geological
model using surface outcrop and borehole data.
Results from the model suggest that the length of the
intersection of this fault with the tunnel could be as
much as 250 m extending over much of the eastern
portion of the tunnel alignment. The geotechnical
relevance of this fault zone in terms of construction
cost is not considered to be significant enough to
justify modification of the tunnel alignment to reduce
this intersection length versus the possible increased
costs associated with higher capacity tunnel support
that may or may not be necessary.

Figure 13. Black Wall Bluffs Fault exposure.
6.5. Groundwater Conditions
The groundwater table is inferred to be at a shallow
depth below surface along the tunnel alignments. A
limited number of packer permeability tests were
completed in the dolomitized limestones along the
Dart Creek valley section of the tunnel alignments.
These tests indicated rock mass permeability of this
generally massive bedded bedrock is in the order of 5
x 10
-7
m/s to 5 x 10
-10
m/s. Given the relatively low
cover along the alignments, the magnitude of
groundwater inflows from most of the bedrock is not
expected to be significant. Significant groundwater
inflows can however be expected from the identified
sub-vertical faults at Dart Creek valley and possibly
from the Black Wall Bluffs Fault.
Artesian groundwater conditions are indicated
from two boreholes completed into bedrock in Dart
Creek. These conditions are inferred to represent
136
groundwater flow under pressure either along the top
of bedrock in the Dart Creek valley or within the
extensive thickness of glacial overburden in which the
Dart Creek stream course has disappeared and then re-
appears at a location on the north side of the existing
highway. These artesian groundwater conditions may
pose problems for surface excavation in the
overburden at Dart Creek but are not expected to have
any influence on tunnel excavation in bedrock below
Dart Creek valley.
7 FIRE, LIFE AND SAFETY DESIGN ISSUES
The fire, life and safety design requirements for road
tunnels are currently undergoing significant changes
and improvements as a result of the recent series of
significant fires within major road tunnels in Europe.
No such requirements are established for road tunnels
in Canada as a code of practice and therefore the
current approach for the proposed tunnel options is to
adopt those standards set out by the National Fire
Protection Agency (NFPA) of the United States that
are considered to represent state-of-the-art industry
practice. These standards set out requirements for the
maximum spacing of pedestrian cross passages, fire
suppression, communications, lighting, ventilation,
and operations monitoring. In Europe, there currently
exist a number of research groups dedicated to
establishing a suggested code of practice for fire, life
and safety requirements for road tunnels. The cost of
the fire, life, and safety requirements for road tunnels
is a significant portion of the overall construction
costs. Significant costs are also associated with
providing a tunnel operations center and operations
monitoring and maintenance.
8 CONSTRUCTABILITY ISSUES
8.1. Tunnel Excavation and Support
The stability of the tunnels will be predominantly
influenced by the formation of potentially unstable
rock wedges formed along the crown and haunches
due to the presence of the pervasive bedding in
conjunction with orthogonally oriented sub-vertical
fractures. Large bedding slabs can be expected to
form and be unstable, requiring regular support for
stability and safety.
Owing to the generally strong and bedded nature
of most of the bedrock it is expected that the tunnels
can be excavated by a standard 2-stage top heading
and bench approach or even possibly full face method
depending on the final size of the tunnels.
Conventional tunnel support measures comprising
pattern rock bolts with mesh and shotcrete are
expected to provide adequate support for the majority
of the tunnels. Enhanced tunnel support measures
comprising lattice girders in conjunction with mesh
and shotcrete and possibly supplemented with
forepoling or self-boring anchors may be necessary at
the intersection of major fault zones such as the Black
Wall Bluffs Fault.
Kinematic and numerical analyses have been
competed to assess excavation stability in terms of the
maximum size of potentially unstable wedges formed
around the tunnels and any potential for overstressing
to confirm the adequacy of conventional tunnel
support requirements. Figure 14 shows an example of
a large unstable wedge that may form along the
sidewall of the proposed tunnels that would be
required to be supported using pattern rock bolts
during excavation.

Figure 14. Typical large unstable wedge.

8.2. Portal Excavation
The proposed portal locations are sited in very close
proximity to the existing highway. Bedrock is present
at very shallow depth for most of the portals for the
NC-2 alignment and the west portal for the NB-2
alignment. The main challenge for excavation of the
portals in rock will be controlled blasting and
management of traffic due to the close proximity of
the existing highway.
In comparison, mixed overburden materials are
present to a depth greater than 10 m at the east portal
of both alignments. The nature of these materials is
not well defined but is believed to comprise loose
side-cast material from the construction of the original
highway.
The east portal is sited below the existing
highway and presents a challenge for excavation and
support in order to prevent any impact to the existing
highway. An initial portal layout was developed based
on commencing tunnel excavation in bedrock for
minimum tunnel costs that requires significant
137
excavation into the existing highway as shown in
Figure 15.

Figure 15. East portal excavation in rock.

A preferred alternative to this layout has been
developed based on commencing tunnel excavation as
soon as possible with minimum cover through
overburden material as shown in Figure 16.

This alternative layout provides an appropriate
buffer from the existing highway however would
require specialized tunnel support measures including
some form of pre-support such as forepoling or self-
boring anchors at a higher cost for the initial section
of tunnel.
A portal excavation layout for the east portal
that does not impact the existing highway also
provides the benefit that the highway serves as an
avalanche and rockfall catchment bench during
operations. The east portal is located immediately
adjacent to hazardous avalanche chutes that would
otherwise require some form of protection-shed
structure at the east portal.
8.3. Dart Creek Intermediate Access Tunnel
Owing to the expected relatively large portal
excavations required to be formed prior to tunnel
excavation, it has been recognized that the low cover
section of the Dart Creek valley offers an opportunity
for the construction of an intermediate access tunnel
and laydown area that provides great benefits to the
project. In addition to independent access for tunnel
excavation, the intermediate access tunnel prevents
any need for muck haulage along the existing highway
to the designated spoil disposal site in the Dart Creek
valley. Also, the intermediate access tunnel may be
used for permanent emergency access during
operations from a tunnel control center that is
currently proposed to be located near the access tunnel
at Dart Creek.
The proposed intermediate access tunnel would be
constructed entirely in bedrock after surface
excavation and the establishment of an appropriate
laydown area located immediately north of the
highway in Dart Creek as shown in Figure 17.

Figure 16. East portal excavation in overburden.

Spoil area
138

Figure 17. Dart Creek intermediate access tunnel/laydown/spoil site.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the permission of
the BC Ministry of Transportation to publish this
paper. The Owners Engineer Team for the
preliminary design work comprises Focus
Corporation, UMA-AECOM, Hatch Mott
MacDonald, and Thurber Engineering Ltd., along
with several specialist consultants. Geophysical
surveys were completed by Frontier Geoscience of
Vancouver. Geotechnical drilling was completed by
Sea to Sky Drilling of Burnaby, BC. Laboratory
testing was completed by Thurber Engineering Ltd.
and packer permeability testing was completed by
Golder Associates of Burnaby, BC. Abrasivity testing
was completed by SubTerra in Seattle, USA.
139

1 I NTRODUCTI ON
Falconbridges Nickel Rim South project is situated 35
km NE of the Sudbury city centre on the East Range of
the Sudbury Basin. The Sudbury basin was formed by a
catastrophic meteorite impact some 1.8 billion years
ago and is today one of the worlds richest mining
districts. The project objective is to access an inferred
geological resource at a depth of between 1200m and
1800m and to establish economic viability of a
potential mine through drilling-off the resource and
conducting a feasibility study based on results from
drill core analysis. The scope of this $600M project
includes surface infrastructure, two deep shafts, 11 km
of lateral development, 86 km of diamond drill holes
and supporting infrastructure on 3 underground levels.
Of the two shafts, the Main Shaft is designed for
production purposes with a finished diameter of 7.6m
(excavated diameter of 8.7m) and the Ventilation Shaft
with a finished diameter of 6.1m (excavated diameter
of 7.2m), is required for second egress and ventilation
purposes.
The project team for shaft sinking comprises
Falconbridge and two alliance style partners:
Cementation Canada; who are responsible for shaft
design and sinking; and Hatch-McIntosh Alliance
(HMA), who are responsible for overall project EPCM.
Three Falconbridge personnel have been seconded into
the execution team in the areas of Construction, Safety
and Procurement management and one acts as Chief
Raising the Bar in Shaft Sinking at Falconbridge's Nickel Rim South
Project
Rick Collins
Project Manager - Nickel RimSouth Project and Hatch Ltd.
Pedro Gonzalez
Area Manager Nickel RimSouth Project and Cementation Canada Ltd.
Hugh MacIsaac
Construction Manager - Nickel RimSouth Project and Falconbridge Ltd.
ABSTRACT: Falconbridges Nickel Rim South Project is situated on the NE rim of the Sudbury basin in Northern Ontario. The
objective of the project is to define an inferred mineral resource, at between 1200m and 1800m depth, by accessing and drilling-off
the ore body from depth to determine whether a production mine would have economic viability. Capital release for the deposit
definition phase was granted by the Falconbridge Board on March 8, 2004 on the basis of a 58 month project schedule and a cost
estimate of C$600M. The scope includes three distinct stages:
Site development and shaft-sinking set-up.
Shaft sinking two shafts.
Off-shaft development, construction and core drilling.
Two shafts are required; one for production purposes, the other for ventilation and second egress. Shaft sinking commenced in
the 6.1m finished diameter Ventilation Shaft in February 2005 with the 7.6m finished diameter, fully equipped Production Shaft
following four months later. Final depths are designed to be 1675m and 1735m respectively.

By the end of May 2006 some 2200m of shaft had been completed without a single lost-time injury and at production rates
consistently ahead of plan.
To achieve this success the Nickel Rim Project team employed Cementation Canada, one of the worlds leading shaft-sinking
groups, to design an optimum set-up for sinking. Together with an EPCM approach to project execution, provided by Hatch in
alliance with McIntosh Engineering, the team has developed a program of performance management and adopted a number of
innovative ideas and tools. The overall result has been sustained average safe-sinking rates of 4.0m and 3.5m per day in the Vent
and Production Shafts respectively.
This paper describes the Nickel Rim project approach to shaft sinking and the controls adopted in raising the bar for the shaft
sinking industry.
140
Engineer for the sinking group. During the steady-state
sinking phase a management and supervision team of
around 50 manages all project activity including
engineering, procurement, planning and construction
contracts, operation and maintenance of shafts, surface
and underground plant and equipment. Overall the
project team employs some 220 people during the shaft
sinking phase. This will increase to 500 during the later
off-shaft development phase.
Surface infrastructure and sinking set-up were
completed and sinking commenced in the Vent Shaft in
February 2005 after 11 months of the 58 month
schedule. At the time of writing the Vent Shaft had
been excavated to approximately 1300m, with the Main
Shaft following at a depth of 900m, and some 65% of
shaft sinking had been completed. The project has
experienced considerable success with its shaft sinking,
measured in terms of safety and the actual production
rates achieved by the sinking crews. The causes of the
success include a well designed equipment set-up,
careful selection and training of skilled operators, tight
project controls, a collaborative project delivery model
and a project team focused on common understanding
of the drivers for success.
This paper describes the sinking set-up and cycle,
the project delivery model and expertise. The paper
also examines techniques adopted to generate and
maintain a safe workplace resulting in sustained rates
of shaft sinking which are in excess of the project plan
and of the shaft sinking industry norms.
The shafts are being excavated in the hanging wall
of the Deposit in Felsic Norite rock a very uniform,
competent unit with only minor geological structures.
The rock has a UCS ranging from 150MPa to 250MPa.
Horizontal stress in the E-W direction is 1.8x the
Vertical Stress; in the N-S direction, the Horizontal
Stress is 1.4x the Vertical.
Shaft Statistics:
Following is some of the principal shaft and sinking
equipment data for the two shafts:
Vent Shaft
Excavated shaft diameter 7.2m.
Finished shaft diameter 6.1m.
Shaft depth at completion 1675m.
Concrete lining volume 20,000 m
3
.
Galvanized steel sets 252 Tonnes.
Sinking stage - 5 deck Galloway, weight 47
Tonnes.
Shaft Muckers 2 x Pneumatic Brutus Grabs.
Shaft J umbo 1 x Sling-down 3 boom
electric/hydraulic.
Hydraulic drills - 3 Reedrill model HPR 4519.
Muck buckets 3 x 12 Tonnes.
Headframe structural steel frame - 45 m high.
Headframe steel weight - 550 Tonnes.
Headframe auxiliary sheave - 3.05 m diameter.
Headframe sinking sheaves - 2 x 4.57 m
diameter.
Shaft Galloway winches - 4 x single drum.
Winch rope-pull capacity - 45,360 kg each.
Sinking Hoist - Refurbished Nordberg double-
drum, rope capacity 48,500 kg.
Auxiliary Hoist - New AKHE single-drum, rope-
pull capacity 16,500 kg.
Shaft stations at: 530, 1050, 1280, 1480 and
1660 metre levels.
Main Shaft
Excavated shaft diameter 8.7m.
Finished shaft diameter 7.6m.
Shaft depth at completion - 1735 m.
Concrete lining volume 24,500 m
3

Galvanized steel sets 1080 Tonnes.
Sinking Stage - 6 deck shaft Galloway, weight
78 Tonnes.
Equipping deck independent from the Galloway.
Shaft Muckers 2 x Pneumatic Brutus Grabs.
Shaft J umbos 4 x single boom,
electric/hydraulic nested in Galloway.
Hydraulic drills - 4 Reedrill model HPR 4519.
Muck buckets 4 x 15 Tonnes.
Headframe structural steel frame - 62m high
(NavCan airport proximity restriction).
Headframe steel weight - 1200 Tonnes.
Headframe auxiliary sheave - 3.05 m diameter.
Headframe sinking sheaves - 2 x 4.57m
diameter.
Headframe production sheaves - 2 x 5.49m.
Shaft Galloway winches - 4 x single drum.
Equipping stage winches - 2 x single drum.
Winch rope-pull capacity 45,360 kg each.
Sinking Hoist - New AKHE double drum, rope-
pull capacity 48,700 kg
Production Hoist - New AKHE double drum,
rope capacity 53,600 kg
Auxiliary Hoist - New AKHE single-drum, rope-
pull capacity 16,500 kg.
Shaft stations at 530, 1050, 1280, 1480, 1500,
1660 and 1700 metre levels.
141

Figure 1 -- Schematic of Nickel Rim South Shaft and Ore
Bodies.

2 SAFETY
2.1 Safety Culture
The management of the Nickel Rim South project
committed to preserving the health and safety of its
workforce from the start of execution planning.
Notwithstanding the ultimate goal of zero injuries on
the project site, statistical targets (expressed as number
per 200,000 person-hours) were developed for Lost
Time, Reportable Injury and Medical Aid frequencies
of less than 1.0, 3.0 and 5.0 respectively. Each year
these targets are investigated and tightened (current
targets 0.8, 2.4 and 4.0). At time of writing the project
is running LTI of 0.2, RIF of 0.9 and MAF of 2.5. See
Section 2.2 for shaft only statistics.
The project has fostered a culture of safety,
conducive to carefully planned, responsibly executed
work methods and elimination of unsafe conditions,
primarily by empowering all employees to act in
accordance with the Internal Responsibility System.
See Section 2.5. The whole workforce, including safety
professionals, supervision, contractors and project
management are expected to demonstrate sincerity over
safety and to lead and act by example.
All employees are required to be committed to
safety and to demonstrate their proactive commitment
on a daily basis through a continuous focus on accident
prevention within their respective work areas.
The project safety commitment and principles
dictate that every worker leaves work each day
unharmed, the underlying ethic being that it is no
longer socially acceptable to work in an unsafe manner.
Health and Safety is not treated as an extra to the
project or analyzed in cost terms; safety has been
integrated into the routine of daily business. Some of
the tools and techniques used in creating the safety is
never good enough culture are described later in this
section.
2.2 Statistics
As of May 2006, the project has completed some 1.7
million work-hours Lost-Time Accident free, stretching
back to August 5, 2004. The following table describes
current accident statistics based on total project hours
of some 1.9M hours.
Table 1. Accident Statistics May 2006.
Description Actual Target
LTI Frequency 0.22 0.8
MA Frequency 2.60 4.0
LTI (Shafts Only 0 0.8
MA (Shafts Only) 1.9 4.0

Frequency calculations are based on injuries per
200,000 work hours.
Although these interim results are encouraging the
project team remains focused on keeping the message
alive. In a project of 58 months duration there is a need
to constantly refresh process, procedures and
expectations. The fundamental concerns are that both
repetitive and one-off activities are vulnerable to the
effects of complacency, and that, even in a controlled
environment, accidents are difficult to predict. There
have been 95 first aid treatments; each of these were
investigated and considered an opportunity for a lesson
learned, to ensure that no accident ever repeats itself.
2.3 Orientation
The Nickel Rim South project requires anyone entering
the site to complete a site orientation which involves a
four hour presentation by senior members of the project
team to welcome all new employees, review the scope
of work and a provide a detailed explanation of the
project philosophies, safety programs, requirements and
expectations. Contractor management is also
encouraged to attend the orientation to demonstrate
commitment to the project and to the workforce they
manage. At the time of writing 2,690 employees have
successfully completed this orientation, representing
some 306 organizations.

Table 2. Orientation Statistics May 2006

Man Hours to Date 1,896,984
Orientation
Number of Orientations- 136
Number of Organizations- 306
Total Workers in Attendance- 2690
Man Hours to Date 1,896,984
Orientation
Number of Orientations- 136
Number of Organizations- 306
Total Workers in Attendance- 2690
142
A further four-hour induction is provided by the
Cementation training coordinator for all shaft sinking
employees; to review PPE, fall arrest systems, 5-point
safety system and other prerequisite shaft-work topics,
prior to commencing work in the shafts.
These orientations provide a platform for building a
strong relationship of mutual respect between all
workers and the project management. The safety
culture is introduced at the orientation and backed up
by a strong management team presence on all phases of
the project.
2.4 Safety Recognition
The Nickel Rim South project has rejected hourly
monetary award systems for employees working safely,
the concern being that such systems can prove to be
counter-productive by promoting the wrong
expectations and stifling good reporting. Instead a
positive recognition safety program has been
developed, whereby crews are encouraged and
recognized for attaining milestones and working injury-
free. Various safety gifts as well as lump sum cash
awards have been extended to all workers without
following any specific schedule or timetable. In this
way safety rewards are not expectations, however
regular recognition has created a positive and safe work
environment with primary attention on safe sinking
rather than reporting accident-free hours for
maximizing monetary gain.
By way of example, safety recognition was given at
the 1 million accident-free work hours milestone by a
leather coat award for every project employee with 6
months tenure.
Many site barbeques have been held to celebrate
medical aid free months, sinking milestones and house-
keeping achievements. In addition to the recognition of
safe work practices, these events provide an
opportunity for project management to interact with the
workforce in a social environment and discuss safety
and morale in the workplace.
Monthly cash draws for $1000 promote various
safety programs such as Stop and Correct and Accident
Imaging. These programs contribute to a safer work
environment by forcing what-if planning for hazard
identification and accident prevention.
Additionally, long-term safe performance awards
are made to recognize excellence by individuals. These
awards increase in value based on the length of injury
free time worked by individuals.
2.5 Safety Tools
The project safety program includes a number of
elements derived from loss control theory and
practices. Key elements of the program are: the Internal
Responsibility System (IRS), the five-point safety
check, daily safety huddles, near-miss reporting,
accident imaging, Think it Through task planning,
development of detailed procedures and extensive
training.
The IRS includes several fundamental
philosophies: everyone is responsible for their own
safety and the safety of those around them; safety
problems are to be corrected by those closest to the
hazard or, if not possible, referred to a supervisor.
Every worker has the right to refuse or stop work for an
unsafe condition and has unfettered access to the J oint
Health and Safety Committee when they feel the
problem remains or has not been addressed
satisfactorily.
The five-point safety check is a workplace audit
form including a checklist of key aspects for
maintaining an acceptable underground environment
(ventilation, dust-control, house keeping,
signage/barricades, rock support, working alone/leaky-
feeder and locking/tagging). The safety card prompts a
worker with the following 5 key checks/actions:
1. Are the entrances to your workplace in good order?
2. Is your workplace and equipment in good
condition?
3. Are the employees working safely?
4. Do an act of safety.
5. Does your crew have the ability, tools and attitude
to continue to act safely?
Workers, first and second-line supervisors all sign-
off the cards on every shift and where the answer to the
questions is no a stop and correct action must be
conducted and recorded on the card.
The Stop and Correct Program focuses on
establishing a culture conducive to discussing and
dealing with issues which may create risk or
compromise safety of workers. Employees are
encouraged to act on as many stop and corrects as
required promoting the theme that any stop and
correct action completed will help to lower and
potentially eliminate accidents. Once a worker
completes a stop and correct, they can fill out a ballot
detailing the act of safety and deposit in ballot boxes
strategically located at various locations across the
project site. Once a month the boxes are emptied,
reviewed, and a draw is completed with the winner
recognized by cash award and a custom designed
project gift.
The Think It Through task planning sheet is a tool
that was created to enhance communication and
awareness of the work at hand. Leaders and supervisors
are required to explain in writing a critical task for the
next shift, any hazard relating to the task, any special
skills or input required (such as maintenance,
mechanical, electrical or engineering), procedures to be
143
followed and any added planning which may assist in
safe execution of the work.
Most importantly, the team continues to raise the bar
by creating fresh ideas to recognize safe workers and
safe work places. These ideas help keep alive and
prolong the positive safety culture and recognize the
risks brought by complacency and routine. The goal
remains that all employees return home safely after
every day working on the project.
3 PROJ ECT DELI VERY MODEL
Falconbridge utilizes a gated approach to project
evaluation such that a potential project is required to
undergo scoping, pre-feasibility and feasibility level
studies passing review gates with specific
deliverables at each level. Capital funding for project
execution is granted on completion of a satisfactory
feasibility study projection including a solid business
case and detailed project execution plan. The Nickel
Rim South project delivery model was built on a
philosophy of engaging key skilled partners early in the
project process.
During the pre-feasibility study Falconbridge
engaged both Cementation and HMA for assistance in
developing estimates, schedules and planning in
support of the studies. Early during the feasibility study
Davy-Markham was employed to develop engineering,
manufacture and commissioning of the 5 major hoists
required for the two shafts. Each of these companies
was contracted under competitive proposals for alliance
style contracts, including pain share/gain share type
incentives, to encourage successful delivery and a
collaborative approach to project execution.
Establishing good project-partner relationships early in
the process has the following advantages:
Increased breadth of experience brought to the
project studies.
Early identification of opportunities in design and
execution.
Early risk identification and mitigation strategies.
Improved accuracy of capital estimates and work
schedules.
Improved planning of work methodology.
Creation of ownership in cost estimates,
schedules and execution plans.
An Engineering, Procurement and Construction
Management (EPCM) model was adopted for project
execution, with the small e indicating engineering
management rather than complete engineering services.
The approach taken was to have engineering completed
by specialized groups in the following principal areas:
surface mine infrastructure, shafts/head-frames,
hoisting plant, electrical reticulation and underground
mine development/systems. In this way the project was
able to engage the best available design groups for
principal specialized areas of work.
Procurement is generally conducted by competitive
tendering with sole sourcing only where specialised-
equipment supply arrangements exist with
Falconbridge or the shaft sinker. For larger construction
and supply contracts, interested bidders are pre-
qualified based on a number of key factors including:
approach to safety; pedigree in similar work; quality;
financial stability; and proposed methodology/people.
A goal of procurement by 80% local suppliers was
established in the project plan; this has been achieved
without major difficulty as Sudbury has a strong base
of competent contractors, as well as many of Canadas
mining equipment and consumable suppliers. All
project procurement and expediting is raised and
tracked through the Hatch project control system iPAS
(integrated Project Administration System), with
resulting work orders set up on the Falconbridge
accounting base for ease of payments as the work is
conducted. Budget, cost (earned value) and schedule
control along with regular forecasting is also
undertaken in iPAS.
Construction management takes over from
procurement at the time of equipment delivery to site or
kick-off meeting for construction contracts. Major
construction packages have been placed with some 25
contractors, all of whom have been selected based on
their specialised skills along with commercial
proposals. The team believes that with the alternative of
adopting a single general contractor with multiple
sub-contractors, they would have less influence on
management of the work and eventually less success.
The approach to construction management is to include
daily attendance at all work sites by team supervisory
personnel who are carefully selected for their
experience and ability to supervise work of contractors
without directing the workforce. Instructions are made
direct to contractor leadership so as not to interfere with
the correct flow of supervision. Specific focus is paid to
safe working practices and to interfaces between
contractors. The project work continues around the
clock and effective communication is believed to be the
key to success; a daily planning meeting is attended by
supervisors of all contractors on site. This forum
proved to be of great value when up to 20 contractors
were working side by side during the initial 18 months
of establishing the surface site infrastructure.
Commissioning and Maintenance are treated as
separate activities. Leaders with responsibility for work
planning and execution are allocated to each of these
important areas, dependent on the specialist skills
required.
144
Project objectives and targets are established for
each calendar year and reviewed on a weekly basis.
The following categories of work are monitored:
Safety/Health/Environment
Financial
Schedule
Quality
Workforce Effectiveness
Within each of these categories metrics have been
developed to measure progress of the work on a weekly
basis. Physical progress is measured by earned value
which establishes value of work performed against a
detailed work plan and consumption of budget.
Performance is measured against the plan via cost and
schedule indices which compare earned value with
actual costs and planned value respectively. Better than
plan performance is indicated by index values greater
than 1.0 and performance below plan by values less
than 1.0. At the time of writing the project is running
schedule and cost performance indices of 1.01 and 1.02
respectively.
4 SHAFT SI NKI NG SET-UP
The project has been fast-tracked from the start of
heavy construction in March of 2004. The fast-track
approach including just in time engineering, caused
construction challenges and a field engineering team
was employed to ensure rapid resolution of field issues
and to control scope and change management.
The project critical path runs through the Vent
Shaft; hence this shaft was collared first while the 6.0m
x 6.5m ventilation adit was excavated. The adit is 300m
long and connects to the two shafts at 66 m below the
shaft collar surface elevation.
The adit portal was collared using jack legs to
maintain perimeter accuracy and control in the
weathered rock structure. Once the portal was
established the balance of the tunnel was completed
using a two boom electric hydraulic jumbo and 4.3m
blast rounds. The intake adit connects to both shafts
and is used for fresh air ventilation into the shafts
during sinking. The adit connection to the Vent Shaft
has been closed off with a bulkhead to prevent short
circuiting of blast fumes; exhaust air flows to surface
through the shaft collars.
Both shafts were collared 15 metres down from
surface. An Alimak raise was piloted up through the
shaft center from the adit level to surface. The pilots
were then ring-drilled and slashed from the adit horizon
to surface, with waste rock removed via the adit.

Fig. 2. Alimak Pilot Raises for Shaft Collars
Shaft lining concrete was completed from surface to
the adit brow followed by assembly and lowering of the
Galloway sinking stages from the surface into the
shafts using a 650 Tonne mobile crane. The Galloway
sinking stages in both shafts are suspended on four
single-line stage ropes that are carried via sheaves
located in the upper deck of the head frames and
controlled by four independent winches.
Once the Galloways were suspended in the shafts,
surface head frame construction continued
simultaneously while equipping the Galloways with the
required sinking equipment in the collars, access for
men and equipment being provided via the adit.
The Vent Shaft is being sunk bald, meaning
concrete lined with construction services only. The
Main Shaft is being sunk and fully equipped
simultaneously; shaft steel and services installation
takes place from an independent equipping deck as part
of the sinking cycle
Both shafts are being sunk using full-face blasting
techniques. Blast holes are drilled using a sling down
three-boom electric-hydraulic jumbo in the Vent Shaft
and four single-boom electric-hydraulic jumbos nested
in the Galloway in the Main Shaft. Waste rock is
removed using Brutus Muckers nested in the bottom of
the Galloway staging.
145

Fig. 3 shows a plan schematic of the Nickel Rim property
surface layout.
5 SHAFT SI NKI NG EXPERTI SE
5.1 Team Philosophy
After 100 years of mining in the Sudbury basin but a
period of some 15 years since the last new green-field
mine development, construction of the Nickel Rim
South project is a welcome addition to the local mining
industry and to the region. Sudbury has long been
considered a world class mining hub with an abundance
of mining resources and support.
It was necessary for a safety-based shaft sinking
culture to be established within the first thirty days of
the project start. With this in mind, it was very
important to ensure that the best available management
and workforce were selected. Shaft sinking leadership
was hand-picked to provide the correct blend of
experience, expertise and specialist knowledge.
The alliance style contract with the shaft sinker
permitted some of the leadership to have continuous
involvement from early stages of feasibility, through
detailed engineering, planning, heavy surface
construction and ultimately shaft sinking execution.
This approach provided the project team with the
continuity and knowledge required to successfully
expedite a fast-track project.
Significant effort has been made to develop a one-
team project culture. This has included integration of
personnel from the various participating companies in
both function and location. Off-site team alliance
sessions were held at the start of project execution and
shortly after commencing sinking. These sessions
helped to understand and communicate objectives, roles
and responsibilities of the various team parties. In
accordance with the Internal Responsibility System
specific responsibilities for each position, including
well developed roles and responsibilities, have been
established and communicated throughout the team.
Accountability has been built into the team approach
recognizing that success for individuals is tied to
success for participating companies, the whole team
and the project. The team is empowered to participate
in all aspects of the project and possess a good
understanding of the scope of work. Nurturing respect
throughout the workforce has resulted in strong site
morale and a feeling of pride that continues to
contribute to project success with the project rather
than the various employers, as team members first
affiliation.
5.2 Recruiting
The shaft sinking Superintendents were engaged early
and tasked with recruiting the sinking crews, as they
had good knowledge of the available shaft sinking
labour pool. Involvement of Superintendents in crew
selections provides the benefit of ownership; the
leaders being committed to and familiar with their
teams. Once the crews were identified, a screening
process was initiated which included review by the
project management team and the Cementation human
resource department. On approval to hire, potential
employees were requested to complete drug and
alcohol testing, base line hearing tests and physical
fitness tests.
Due to the difficulty of finding experienced shaft
sinkers a dedicated recruiter was employed and a cross-
Canada recruitment campaign was conducted. In the
final make-up of the shaft team some 40% of
employees are long-distance and 60% local to Sudbury.
The cyclical nature of the mining industry often
provides challenges for recruiting individuals for
specialty work within mining. The shaft sinking
business is relatively small and suffers from cyclical
effects of feast or famine. Following a period of very
little shaft sinking work in Canada, there has been
resurgence in the level of shaft sinking work recently
with few experienced shaft sinkers available to safely
expedite projects.
5.3 Manpower Experience
At the beginning of the project, it was recognized that
completely staffing the project with experienced shaft
sinking manpower was not going to be possible. A
plan was adopted to use the following breakdown of
experience in a 7 man crew:
Experienced shaft men 4
Experienced development men 2
Raw recruit -- 1
The project ramped up through the shaft sinking
setup and learning curve phase with an actual of 56%
experienced miners, the balance in the crews having
little to no experience in shaft sinking. Miners were
attracted on the basis of a well prepared shaft sinking
and quality methodology, strong experienced
management and a shaft sinking incentive structure
X

Vent Adit Portal
146
based on achievable performance. In consultation with
the Owner, the shaft sinking incentive rate was
established based directly on a bench-marking study of
incentives currently being paid in the industry. The
resulting expectations and incentive charts were posted
in the wicket and dry house areas.
5.4 Training
With the cyclic nature of shaft sinking in Canada, the
previous decline in the number of shafts being sunk
resulted in many shaft men retiring or moving into
other careers. Recognizing this shortage in skilled
manpower, Cementation initiated a new-miner training
program. The new-miner training program introduced
workers with no mining experience into the industry.
After 10 weeks of training they were placed on various
projects for a nine month period to further expand their
knowledge and experience.
In addition a Brutus Mucker training program was
instigated as few competent workers were available to
operate this specific shaft-mucking equipment. A
Brutus Mucking training facility was developed on the
project site with a full scale Brutus Mucker set-up for
training. The facility provided an investment in the
success of the project through the specific training of
key skills for safe operation and advance of the shafts.
Some 30 operators were trained on the safe operation of
the Brutus Muckers in an underground set-up which
closely resembled the ultimate workplace. The result
was a pool of shaft miners with the necessary skills to
efficiently and safely operate the Brutus Muckers.
Training has taken a very large role on the project.
A total of 37 training programs have been run including
Supervisory Common Core, First Aid, WHMIS, and
Fall Protection. Some 22,000 work hours of training
have been carried out on the project to date.
5.5 Workforce Effectiveness
Workforce effectiveness is measured by staff turnover
rate. A target of one voluntary departure per $5M of
project expenditure was established at the start, and the
project is running a current rate of 0.85 on shaft sinking
personnel.
6 SHAFT SI NKI NG CYCLE
The following series of figures describes the shaft
sinking cycle:

Fig. 4. Main Shaft Bench Drilling
Drilling the bench is done by lowering the jumbo to
shaft bottom and drilling 51mm diameter holes to a
depth of 4.3m in accordance with the blast pattern
shown in Fig. 3 for the Vent Shaft. The Main Shaft
blast pattern includes one additional ring of holes,
compared to the Vent Shaft pattern.

Fig. 5. Vent Shaft Blasting Pattern
The Main Shaft blast pattern comprises 115 holes and
the Vent Shaft 94 holes to achieve the required
fragmentation (<400mm). The firing sequence is
indicated by the cap number (7 to 14 shown in Fig. 3)
with a total delay of 350 milliseconds over the blast
from the centre out to the outside ring. Blasting caps in
the bottom of each hole are connected with B-line
detonating chord. The blast is detonated via a
magnadet starter cap fed by a frequency converter
switch located on the Galloway which is connected to
the surface blasting switch via a blasting cable.
147

Fig. 6. Blast Bench
The project selected an emulsion explosive which is
pumped into the blast holes. Trace-Its are used for
perimeter control. Proper loading procedure is followed
with emulsion loading limited to approximately 0.6m
of the hole collars to prevent build up of ammonia
when waste concrete mixes with unspent emulsion.
Fresh air ventilation is supplied via a 1.37m
diameter steel ventilation duct, designed to supply
1,500m
3
per minute at the shaft bottom. Ventilation is
reversed (to a pull system) for blast fume removal
which takes approximately 45 minutes to clear the shaft
bottom and resume work on the following cycle.

Fig. 7 Muck Bench
In both shafts the removal of blasted rock,
mucking, is performed with two Brutus shaft
Muckers nested in the Galloway sinking stages. These
mucking units pick up the blasted rock from the shaft
bottom and place it into shaft buckets. The Vent Shaft
rock is hoisted from the shaft bottom to surface cycling
three 12 tonne buckets dumped into a temporary
surface dump. The 4.6 m (15 ft) diameter double drum
hoist has a speed of 900 m (3000 ft) per minute on rope
guides. The Main Shaft rock is hoisted from shaft
bottom to surface in a similar manner, while cycling
four 15 tonne buckets. The sizes of the buckets are a
function of the capacity of the hoisting plant as well as
the available clearance room through the shaft
Galloways.

Fig. 8. I nstall Ground Support
Ground support for both shafts is installed in
sequence. Crews muck approximately 2 vertical metres
of the shaft at a time and install the required ground
support. This is followed with final mucking and the
remaining ground support installations.
Screen installation commenced at around the 1050
metre depth in both shafts due to the risk of ground
bursting from stress relief. The Main Shaft typically
completes the last half of the ground support cycle
during the drilling of shaft bottom for the next round.

Fig. 9. Pour Curb Ring Forms
The shaft concrete lining is poured in 6 metre lifts as
sinking progresses. A re-usable 6 metre high set of steel
forms is used. The curb ring and A panel comprise
the first metre of the shaft forms and are initially
lowered and lined up; the fluid concrete is then poured
with an accelerator to quicken the set time. The main
forms are then lowered, positioned and the balance of
the concrete is poured.

148

Fig. 10. Pour Main Forms
Concrete is delivered by concrete buckets in the
Vent Shaft and via a slick line in the Main Shaft. The
concrete liner is a minimum of 300 mm thickness in
accordance with CSA Standard A23.3 attaining a 28
day minimum compressive strength of 25 MPa.
Target time for concreting in the Main Shaft is 7.2
hours and target concrete volume 82m
3
per 6m pour in
the Main Shaft and 67m
3
per 6m pour in the Vent
Shaft.

Fig. 11. I nstall Steel Set
The Main Shaft is equipped simultaneously with the
concrete pour. Equipping in the Main Shaft is
performed from a separate stage which permits the steel
crew to close all doors on the work stage allowing them
to work independently of the concrete crew below;
hence, reducing cycle time. The equipping stage is
suspended on ropes from two independent winches
located in the winch building along with the four
Galloway winches.
Both shafts utilize a 3m diameter single drum
auxiliary hoist and cage as a second means of egress
and for the installation of the auxiliary cage guides and
guide backers.

Fig. 12. Cleaning Bench
The last step in the cycle is to blow the shaft bottom
bench clean using a blow pipe and compressed air. The
muck is loaded into the buckets and hoisted to surface
for disposal. This provides a clear working platform for
marking up and recommencing the cycle with drilling
for the next round.
A complete Main Shaft sinking-cycle is targeted
every 30.5 hours (for 3.25m/day equivalent advance).
Steady state average cycle time achieved to- date is
28.55 hours.
A complete Vent Shaft sinking-cycle is targeted
every 24.3 hours (for 4.0m/day equivalent advance).
Steady-state average cycle time achieved to-date is
23.32 hours.
7 SHAFT SI NKI NG PERFORMANCE
Shaft sinking is subject to a continual performance
management process. Each element of the cycle has
been planned and target durations set prior to the start
of sinking. The Performance Management Engineer
ensures accuracy of detailed records and manages the
process of introducing improvement ideas. Various
initiatives are sorted into priorities depending on
variation analysis and requirements for making the
change. The Performance Management Group meets
fortnightly to assess performance results, study trends
and determine the improvement path forward.
7.1 Vent Shaft
The following statistics highlight the average cycle
element times during the learning curve and steady-
state shaft sinking periods:
Table 3. Vent Shaft Cycle Element Times

149
Cycle
Element
Learning
Curve
Average
Durations
(Hours)
Steady
State
Average
Durations
(Hours)
Steady
State
Target
Durations
(Hours)
Drilling 6.83 4.29 4.2
Loading
and
Blasting
2.19 1.81 1.7
Mucking 9.66 8.23 7.6
Bolting
(Ground
Support)
3.02 3.16 5.2
Concrete 6.54 5.83 5.6
Total
Cycle
Time
28.24 23.32 24.3
Daily
Shaft
Advance
3.0 m 4.15 m 4.0 m
Concrete
Volumes
79.1 m
3
69.8 m
3
67 m
3
As part of shaft sinking pre-planning a 4-stage learning
curve target was developed for the first 300m of both
shafts.
Figure 13 demonstrates the Vent Shaft advance rates
during steady-state shaft sinking.

Vent Shaft Advance
Steady State Shaft Sinking
0
200
400
600
800
1000
1200
1400
Feb Mar Apr May J un J ul Aug Sep Oct Nov Dec J an Feb
Month
S
h
a
f
t

D
e
p
t
h

(

M
e
t
r
e
s

)
4.2 m/d
4.1 m/d
4.0 m/d
4.3 m/d
4.8 m/d
4.1 m/d
3.9 m/d
3.9 m/d
4.2 m/d
Variance
530 Station
Development
Variance
1050 Stati on
Devel opment
Variance
1280 Stati on
Devel opment

Fig. 13 Vent Shaft Advance Steady-State Sinking
7.2 Main Shaft
The following statistics highlight the average cycle
element times during the learning curve and steady
state shaft sinking periods:
Table 4. Main Shaft Cycle Element Times
Cycle
Element
Learning
Curve
Average
Durations
(Hours)
Steady
State
Average
Durations
(Hours)
Steady
State
Target
Durations
(Hours)
Drilling 6.71 4.41 4.9
Loading and
Blasting
2.45 2.42 2.2
Mucking 15.34 12.08 12.4
Bolting
(Ground
Support)
3.74 2.08 2.0
Concrete 8.47 7.56 7.2
Total Cycle
Time
36.71 28.55 30.5
Daily Shaft
Advance
2.81 m 3.34 m 3.25 m
Concrete
Volumes
109.5 m
3
88.5 m
3
81.9 m
3

Figure 14 demonstrates the Main Shaft advance
rates during steady state shaft sinking.

Fig. 14. Main Shaft Advance-Steady State Sinking
8 CONCLUSI ONS
The Nickel Rim South project has employed a carefully
selected team and a well planned approach to shaft
sinking. The key aspects to this approach are as
follows:
1. Creation of a project safety culture based on pride
of ownership.
2. Early engagement of key organizations and
personnel in alliance style contracts.
3. Carefully planned and engineered shaft-sinking set-
up and process.
4. Establishment of and measurement against realistic
project metrics.
150
5. Utilizing a well thought out performance
management program to drive-up efficiencies of
each element of the shaft sinking process.
6. A unified one-team approach with positive
reinforcement generating project pride.
The project modus operandi and techniques
described represents somewhat of a new approach for
the shaft sinking industry and for Falconbridge
projects. However, after the first half of the project, and
with 65% of shaft-sinking complete, the benefits of this
approach have been demonstrated by safe work and
close adherence to plan in terms of cost and schedule
performance.
There is a significant amount of underground work
remaining to undertake over the second half of the
project with approximately a further 3 million effort-
hours to be expended. Safety performance has been
better than target but can never be good enough. The
bar has been raised for the shaft-sinking industry and
this high performance will drive targets for the
remainder of the shafts and underground work on the
Nickel Rim South Project.

Fig. 15. Main Shaft Hoist Room

Acknowledgements
The Authors wish to thank Falconbridge Ltd. for
permission to publish this paper.
151
1 INTRODUCTION
The Metro do Porto project comprises 70 km of light
railway; 20 km of new construction and 50 km
utilised existing railway alignments. Approximately
6.5 km have been tunnelled by two tunnel boring
machines (TBMs) using single-pass pre-cast concrete
segments with either a 7.8 or 8.0m internal diameter.
These single twin-track running tunnels are large
enough to accommodate circulation of the trains
running in either direction. The system comprises 64
stations of which 12 have been constructed
underground. Of the remaining 52 stations, 41 are
new construction and 11 have been renovated. The
construction of the underground stations was carried
out using various techniques Including Diaphragm
walls, tangent piles and NATM methods.
The bored tunnels were separated into 3 separate
drives; Line C or the Blue Line was 2.4 km in length,
Line S or the Yellow line was 2.7 km and the Line S1
the continuation of the Yellow line a further1 km of
tunnelling to the south of the Trindade station. The S1
tunnel broke into a cut and cover section adjacent to
the historic Don Luis Bridge. A separate NATM
tunnel of 300 m was constructed in order to permit
trains to be shunted between the Yellow and Blue
lines at the Trindade station, hub of the new light rail
system.

Figure 1: Bored Tunnels and stations on the Blue and
Yellow lines of the Metro do Porto.
2 GEOLOGICAL SETTING
The entire underground system was excavated
through Porto Granite. The Porto granite is a two mica
igneous rock characterised by its heterogeneity caused
by the weathering processes. This part of the Iberian
Peninsula was once located within the tropics and
suffered tectonic movements leading to the
penetration of the weathering fronts deep into the
Tunneling the Metro do Porto - Under Pressure in Porto Granite
Peter C. Raleigh
J acobs Associates, Seattle, WA, USA

ABSTRACT: The tunnels of the Porto Metro Light Rail project form the heart of the new 70km transport system for the
Metropolitan area of Oporto, Portugal. Linking the historic central Trindade district with Vila Nova de Gaia to the South and
Pvoa do Varzim and Maia to the North and to the east Gondomar, this ambitious project got off to a rocky start. Following a
tragic incident which resulted in the death of a member of the public and linked to the TBM excavation, the works were halted.
The client, construction manager and contractor were forced to take a fresh look at their respective roles and then to take the
necessary actions imposed by a government appointed Commission of Inquiry. The Commission outlined the steps required by all
concerned so that tunneling could recommence. This paper will provide a general overview of the project, discuss the events
leading up to the fatal collapse, treat the steps that were required to recover from the tragic events and conclude with the various
solutions adopted, including the controversial use of Earth Pressure Balance techniques in rock, finally resulting in the successful
completion of tunneling in late 2003.
152
system of fractures. This has left very deep
weathering profiles similar to other tropical granites
such that completely decomposed granite can be
encountered next to fresh material in any possible
depth and location.
The hydro-geological regime encountered
was also extremely complex and defined by sharp
changes in permeability. In weathered granites and
residual soils water circulation took place through the
pores whereas in fresh granites the water flow was
through the fractures. Water levels were generally
close to the ground surface and therefore represented
the biggest challenge during tunnelling. Man-made
minas or water mines and deep wells were also
found frequently both intersecting and above the
tunnel alignment.
3 PROJ ECT ORGANIZATION
The metro system has been constructed on a Design,
Build and Operate Transfer (DBOT) basis by the
consortium Normetro. The consortium has
responsibilities for the operation of the Metro system
during a 5 year period after which the concession will
return to the owner of the project; Metro Do Porto.
Figure. 2: General project organization chart during
construction of the Metro do Porto light rail tunnels
The Normetro consortium supervised the
construction of the various facilities including the
tunnels through their civil construction group
collectively know as Transmetro. The principal
contractors making up the group were Impregilio,
Soares da Costa and Somague, the latter two
companies of Portuguese origin. Following the
restructuring that took place following the accident
both Geodata of Italy and Mott-MacDonald became
responsible for design with the Mott team taking on
the resident engineering duties for the Consortium
during construction.
The Fiscalizao or Construction
management team (CM) was responsible for the
management of Safety, Quality, Schedule and Costs
and made up of a consortium of Cinclus, J acobs
(Gibb) and Earth Tech (Kaiser Engineers) known as
CGK. Ensitrans, a group comprised of engineers with
experience from the Metro do Lisboa, carried out the
project review on behalf of the Metro do Porto.
4 PROGRAMME
4.1. Line C
The 8.7 m diameter S-160 Herrenknecht earth
pressure balance (EPB) TBM began excavating line C
on J une 12
th
2000. However due to a major set-back
caused by the sudden collapse of a home and the
death of one of the occupants, the TBM was stopped
after less than 25 % of the drive had been completed.
A Commission of Inquiry set up by the government to
investigate the accident mandated change at all levels
of the project including an overhaul of the equipment,
personnel, geological model and working methods. A
considerable delay was inevitable since all of these
items had to be completed including the hiring of new
personnel prior to restarting the TBM tunneling. The
153
Metro do Porto - Line C and S Progress
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
Project Week
P
.
K
.

(
m
)
Estimated Excavation Line C
Actual Excavation Line C
Estimated Excavation Line S
Actual Excavation Line S
Estimated Excavation Line S1
Actual Excavation Line S1
P.K. 630
Heroi smo
Campo 24 do Agosto
Bol hao
Inquiry and Modifications to TBM
Tri ndad e
TBM S-160
Cuttehead Maintenance
Start: Campanha
June 12th 2000
End: Tri ndade
October 22nd, 2002
Sal gui eros
TBM S-203
Cutterhead Maintenance
Li ma
Fari a Gui maraes
End: Tri ndade
October 16th, 2003
Start: Salguieros Trench
June 3rd 2002
Marques
End: Potal to Don Luis Bridge
November 3rd, 2003
Start: Tri ndade
November 4th 2002
Refurbish TBM S-160
Al i ados
Sao Bento
Metro do Porto took a further step by engaging an
international Panel of Experts, POE, which also added
further recommendations to be implemented.
Following the restarting of the TBM drives no
major problems were encountered and the TBM
passed through the three underground stations along
the alignment finally holing through into Trindade
station on October 21, 2002.
The planned progress rate was 25 rings (35
m) per week. As can be noted in figure 3 the planned
rate was regularly achieved through all soil types
encountered along the alignment despite the heavy
maintenance required on the cutterhead carried out
almost exclusively under hyperbaric conditions.

Figure 3: Sloping line diagram showing the progress of
major TBM activities for line C and S.
4.2. Line S
Delay to the works caused by the accident required
the purchase of a second machine. The S-203 TBM
for the northern extension of the S line was designed
to be 200 mm larger in diameter to better
accommodate the dynamic envelope of the single twin
track running tunnels. It was launched from a cut and
cover portal adjacent to the Salgueiros site in J une of
2002. The TBM traversed the alignment without
mishap passing below a major highway, through the
Salgueiros, Lima, Marques and Faria Guimaraes
stations to its final breakthrough into the Trindade
station hub at the end of October 2003 several weeks
ahead of schedule.
The southern extension of the S line was
excavated using the refurbished S-160 TBM
following the completion of the C-line. After a
successful re-launch in February 2003 it excavated
through the Aliados and Sao Bento stations finally
breaking through on November 3, 2003 into the cut
and cover portal adjacent the historic Don Luis
Bridge. This bridge has now been refurbished to be
used as a dedicated link to the southern side of the
Douro River and completes the link with the city of
Vila Nova de Gaia.
The programme for payments on the contract was
set out according to agreed milestones being reached.
In the case of the tunnel a schedule was created with
specific chainages identified as milestones for
payment.
The original budget for the tunneling works was
55 million euros not including the tunnel lining, an
estimate of the final costs range to 110 million euros.

154

Figure 4: Lowering the cutterhead of the Herrenknecht S-
203 8,9m EPB-TBM into the Salgueiros cut and cover
portal.
5 COMMISSION OF INQUIRY
An inquiry into the causes of the fatality which
occurred following the collapse was commissioned by
the various government ministries involved in the
works. This report had essentially the force of law and
was binding for all the involved parties.
A list of 16 recommendations was generated
which gave rise to a complete re-organization of the
contractor and construction manager, CGK, including
the presence of TBM and Geotechnical expertise.
The list is summarized as follows:
Improve the knowledge of the geology along
the proposed tunnel route.
Improve the soil where indicated to avoid
collapses.
Revise the system of surface and deep
monitoring on a permanent basis particularly in
suspect areas.
Revise the Normetro management team and
ensure that they have the correct experience for
the job. Revise the required procedures for safe
execution of the works.
Safety first. Production second.
Operation of the TBM in closed mode only
except where the situations warrants.
Reinforce the teams with additional competent
personnel in the geotechnical area.
Provide real-time access to the information
from the TBM and settlement monitoring
Revise the project geological model.
Improve the prediction of material densities in
order to improve the excavated volume
measurements.
Improve the primary grouting and prove it by
taking cores and by making secondary proof
injections.
Probe ahead of the TBM to verify predicted
conditions.
Uncover all possible pre-existing underground
structures along the alignment which could
affect the safety of TBM operations.
Implement the PAT (Tunnel Advance Plan)
system bringing together all the information
existing along the route of the tunnel to assist in
the proper execution of the works.
The first PAT should redefine the investigation,
consolidation and monitoring that must be
carried out or had been carried out in the
accident areas.
It is fundamental that the supervision follows
the works both inside the tunnel and on the
surface. The safety of the project can not
depend upon mere contractual penalties which
could be imposed.
These recommendations formed the skeleton of a
specification under which the construction
management could operate and compel the contractor
to perform the work in a controlled systematic
manner. Daily meetings were held during tunneling
with the construction manager, designer and
contractor so that information could be exchanged
easily and any issues could be dealt with promptly by
the team. Analyses of the tunneling and TBM activity
were also made on a daily basis so that any anomalies
in the operation of the TBM and its impact
particularly on the deep instruments could be
reviewed. Although a formal Partnering process
was not implemented the value of these daily
meetings in creating a collegial atmosphere open to
dialogue among all the parties cannot be overstated.
6 A BRIEF HISTORY
Since the start of the line C tunneling in J une 2000
there had been three ground collapses. The most
serious incident occurred on 12 J anuary 2001 when a
crater formed at the surface leading to the sudden
collapse of a home resulting in one fatality. At the
time of the collapse the TBM was approximately 60m
155
ahead of the collapse zone and was halted after
driving 470 m of tunnel.
Following this unfortunate incident and the
removal of the wreckage of the house, the cavity was
measured and the resulting volume was estimated to
be 250 m
3
. The TBM had passed below the property
during the previous 16 to 18 of December, i.e. 25 to
28 days before the accident. The collapse and
resulting cavity did not damage the tunnel lining
which had remained remarkably stable.
It is uncertain in what mode the TBM was driven
over the initial stretch due to the fact that these
records were incomplete. The construction
management team did not have access to the records
from the TBM data logger system nor was the
conveyor belt scale thought to be working correctly.
At that time, the construction management team of
engineers and inspectors were performing a more
passive role of record keeping and measurement in
accordance with their contract, rather than the
proactive role which they would later embrace.
At the outset, the tunneling activities were carried
out by Transmetro on site with the design support
from Geodata who were located off-site. Prior to the
accident there had been no real construction
supervision as it is normally understood. The fatality
resulted in the reorganization of the whole
construction team following the mandate of the
Commission of Inquiry. The new structure led to the
employment of Mott-Macdonald who was
subcontracted by Transmetro to assist the original
designers and undertake the design and construction
supervision.
The new team worked on the recovery works and
prepared all the documents, specifications and
required steps to recommence the tunneling. Finally
after months of preparation including the introduction
of many new specialists and full analysis of the means
and methods proposed, excavation began anew on 18
September 2001.
7 PANEL OF EXPERTS
A panel of experts or POE was commissioned by the
Metro do Porto. The panel was composed of a group
of internationally known experts in tunneling and
underground engineering including: Dr. Ing.
Siegmund Babendererde, Prof. Antonio Silva
Cardoso, Dr. Evert Hoek, and Prof. Paul Marinos.
Their first report proved to be the most
significant. Two major modifications to the TBM
were proposed.
Automatic bentonite injection system in order to
maintain a minimum EPB pressure in any
situation.
Addition of a double piston pump permanently
attached to the screw conveyor in order to handle
material not manageable with the screw conveyor.
The contractor added his own modifications
including:
10 independent foam generators.
A new rotary fluid joint which permitted the
passage of ground conditioning, high pressure
water and hydraulic fluids.
A new twin belt weighing system
A laser scanner to detect the volume of material
passing over the belt.
Various improvements in the PLC system were
developed so that alerts and alarms could be given
to the operator immediately in the event that the
any of the levels were reached. For example, a
warning for minimum apparent density within the
chamber measured by the pressure acting on EPB
cells located on the bulkhead of the TBM. This
indicated to the driver in the form of a message
and light.
The POE met every four or five months in order
to deal with any developments or technical issues
including proposed modifications to the tunneling
approach agreed upon.
For example in their second report they did not
permit the use of foam for soil conditioning due to the
unreliable foam generating plant and the systems for
control. They instead encouraged the use of bentonite
for ground conditioning. The CM team encouraged
the use of polymer as the most useful compromise.
In their third report they gave the approval for
foam trials satisfied that the problems with the plant
had been corrected. However they raised the question
of compressed air interventions and the need for
having a proper bentonite membrane on the face in
order to avoid sudden collapses of the face due to the
use of compressed air.
Subsequent reports began to deal more with the
pressing problem of station construction as it was
clear by this time that the major problems with the
TBM drives had been overcome and that the tunnel
drives were being managed very well by the
contractor and construction management team.
8 INNOVATIVE DEVELOPMENTS
As a result of the accident many new developments
were recommended or introduced by the Commission
of Inquiry, the Panel of Experts and the Contractor.

8.1. Granite under pressure
The first major development to be universally
implemented was the exclusive use of closed mode
TBM operation. This strategy meant the operation of
156
Figure 5: Example of the TBM drive parameter table from
P.A.T. document.
the TBM in EPB mode through areas where good
quality granite was expected and proven by the probe
holes drilled ahead of the TBM. In many cases a
minimum face support pressure was adopted and used
to excavate where competent rock was encountered.
The consequences of this approach were the rapid
wearing of the cutting tools and the structure of the
cutterhead. By the time the TBM had reached the 24
do Agosto station, much of the front and periphery of
the cutterhead had to be rebuilt underground, stopping
progress on the tunnel drive for an additional 6 weeks
(note in Figure 3 above).

Figure 6: A worker changes a disk cutter tool.
Both the C line and S line TBM cutter heads
required heavy maintenance throughout the respective
drives as a result of strict adherence to closed-mode
operation despite treatment with polymer ground
conditioners added at the face. Both the client and
their experts felt that no other alternative was
considered to be 100% safe. The risks of another
major collapse as a result of moving away from this
strategy would have meant the permanent cessation of
tunneling and serious blow to the Metro scheme.

Figure 7: View of the face of showing the Porto Granite
8.2. The PAT document
The contribution made by those who developed the
PAT document should not be overlooked. In Figure 5
an extract of the summary table is shown. This
document along with its supporting calculations, plans
and tunnel sections permitted easy reference to any
position along the tunnel alignment showing buildings
and condition surveys, expected settlement and other
information invaluable to all parties during the
prosecution of the works. The fundamental parameters
included by ring and chainage were; the expected
geology, the required face support pressure, the
157
weight of material excavated, the grouting pressures
and total volume to be injected. Other information
included a table of the required starting air pressure
for the daily hyperbaric interventions which were
needed to maintain the cutting tools. The PAT became
an indispensable document for all of those involved in
the daily management of the tunneling works

8.3. Apparent Density
Another development was the concept of apparent
density which is an interpolated value of the pressure
registered on the pressures sensors located within the
plenum. The sensors were fitted to the main bulkhead
which separates the pressurized soil from atmosphere.
The uppermost three sensors were used to determine
the apparent density. The uppermost sensor pressure
was compared to that of the lower two sensors which
were separated by a fixed height such that a
differential in pressure could be noted. This
differential in pressure was used to evaluate if the
chamber was really filled with material and if the
material had sufficient apparent density to provide
effective pressure to balance any instability which
may have been present at the tunnel face. A simple
system was incorporated into the operators display to
warn the driver to slow the screw or modify the rate of
ground conditioners added if the chamber density
became too low. This concept worked very well and
was the subject of continuous observation by the
construction management team.
Further measures were implemented at the request
of the Panel of Experts. These included the fitting of a
double piston pump at the screw outlet and an
Automatic Face Support system which permitted the
injection of bentonite when the pressure in the
chamber dropped below a predetermined level for a
period of time. This system did not require input from
the operator and proved to be useful when using foam
as soil treatment. Figure 8 shows a graphic illustration
of the AFS system in automatic mode injecting
bentonite as needed to maintain the target face support
pressure. The purpose of the double piston pump was
to permit the handling of running sands in the event
that these could not be controlled easily by the screw
conveyor.

8.4. Belts Scales
Due to the fact that the contractor elected to use a
continuous belt conveyor for removing muck from the
tunnel, a positive method of volume measurement
such as the counting of muck cars was not easily
available. In light of the lost ground noted in the first
470 m of line C it was of special interest to all
involved that careful and continuous monitoring of the
muck weight and volume be carried out in order to
determine if over-excavation was occurring. The
contractor installed two belt scales on a level portion
of the fixed conveyor fitted to the TBM trailing gear
in order to measure the excavated weight of the muck.
Additionally, flow meters were installed in the
ground conditioning lines capable of providing the
total volume of liquids injected. Thus, a concise
accounting of materials injected and extracted could
be made readily indicating if over-excavation was
occurring.
This system was reliable and worked very well in
controlling the excavated weights and showing a very
controlled excavation was possible using EPB
methods. The scale readings were added to the
operators control screen and provided warnings of
over-excavation based on instantaneous measurement
of propulsion strokes relative to the total excavated
weight.

8.5. Real Time Monitoring
Modern data loggers and computer networks made
possible the rapid and reliable exchange of data such
that real-time observation of all important TBM
parameters, and instrumentation such as settlement
points, piezometers and inclinometers were available
on a continuous basis. This access to the data and the
time to interpret and act appropriately was a powerful
tool for demonstrating the contractors control of his
works and giving confidence back to a client and to
the public at large.
9 THE FOUR FUNDAMENTALS OF EPB
TUNNELING
During the course of the works some fundamental
parameters which together contribute to the success of
the EPB tunneling technique were identified:
Maintenance of appropriate face support
pressure
maintenance of material density in an
appropriate range
Control of excavated weight and volume
Control of primary grouting
10 FINAL OBSERVATIONS
The successful breakthrough into the Trindade station
following the challenging low-cover under-passing of
adjacent buildings was really the highlight of an
almost two-year struggle for the Metro do Porto to
once again feel confident that EPB-TBM tunnels
could be safely driven through the unpredictable
geology that lies below the city of Porto.
This was due in no small part to the dedicated
tunneling team which Transmetro was able to
158
assemble following the tragic accident and the
willingness of the Metro do Porto together with their
Construction Manager Cinclus J acobs Gibb and Earth
Tech to create a team which could work with
contractor to tackle the issues placed before them.
The successful completion of the tunneling on the
Metro do Porto is a reason for all parties to be proud
and for which the fans of the recent 2004 European
Cup Football held at the Portos new Dragon Stadium
had another reason to cheer as they stepped off the
metro and into to a new era of transportation for
Portugals working city.

Figure 8: Variability of EPB pressures while using foam
soil conditioning and the pressure compensation of the
Automatic Face Support system at ring 905 of the Metro
do Porto, Line C.

Metro do Porto Line C Ring 905 - Support Pressure
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1
5
:
1
1
:
0
0
1
5
:
2
6
:
0
0
1
5
:
4
1
:
0
0
1
5
:
5
6
:
0
0
1
6
:
1
1
:
0
0
1
6
:
2
6
:
0
0
Time (hh:mm:ss)
B
u
l
k
h
e
a
d

P
r
e
s
s
u
r
e

(
b
a
r
)
P1 P2 P3 P4 P5 P6 P7 Pr essur e Cells
Pr essur e compensat ion
by Act ive Face Suppor t
Syst em
P ressure release
byair diffusion
fl uctuati ng pressure
TBM on advance TBM on hold
159
1 INTRODUCTION
Traditionally, underground metro or light rail stations
have been conceived either as a rectangular cut and
cover box, or as large diameter tunnels connected to
smaller cut and cover, near surface structures for
passenger access and ventilation. In these types of rail
projects, circular or elliptical shafts are used mostly to
house ventilation and pumping equipment, as well as
to provide an emergency escape way to the surface.
The main reason why designers avoid using
circular (or elliptical) shafts to house underground
metro stations is their limitation in size.
In spite of the fact that a closed ring is known to
have structural advantages over other shapes, its use
has been limited to diameters usually less than 15 m
(49 ft).
In Oporto, Portugal, a recently commissioned
light rail system has made use of innovative design to
accommodate underground stations inside large
diameter elliptical shafts. The technical challenges
and advantages of this solution are discussed in this
publication.
Oporto was also the place where the Additional
Face Support System (AFS System) was first applied
in EPB machines. The features of this system are also
discussed in this paper.
2 WHY NOT LARGE DIAMETER SHAFTS?
It is commonly accepted that small diameter, circular
shafts behave like a closed ring. As such, nearly all
hydrostatic and lateral earth pressures acting on the
shaft wall turn into compression loading of the shaft
lining. Because commonly used shaft lining materials,
such as concrete and shotcrete, handle compression
well and primarily because a circular shaft does not
require anchors or struts, this combination is ideal for
the support of excavation works.
Designers have made extensive use of circular
geometries for shafts up to 15m in diameter. Beyond
this, whenever a large excavation footprint is required,
they tend to choose traditional box-like geometries,
which, for stability reasons, require anchors, struts, or
both. The reasons for this seem to be founded on the
notion that, as the diameter increases the loading
regime on the shaft wall shifts from mostly
compressional to mostly flexural, and becomes
similar to that of a slurry wall, typically requiring
support by anchors or struts. Also, openings in the
shaft wall create design difficulties only solved by 3D
modeling.
In Oporto, it was possible to show that these
problems can be overcome and that considerable
savings in time and money can be obtained by using
large diameter shafts. This light rail system was, at
time of construction, the largest transportation project
funded by the European Union. It included 70 km (43
mi) of surface rail lines and 66 surface stations. In
downtown Oporto, it included 11 underground
stations and 7 km (4.3 mi) of tunnels.
It was built under a design-build contract
involving contractors and design offices from Portugal
and several other countries. Two underground stations
Innovative station design and the new Additional Face Support
System make Metro do Porto a unique light rail project
Maia, Claudio
Babendererde Engineers LLC, Kent, Washington, USA.
Babendererde, Lars
Babendererde Ingenieure GmbH, Bad Schwartau, Germany.
ABSTRACT: The challenges of tunnelling in the highly heterogeneous Oporto granite led design engineers to review
and modify traditionally accepted station geometries. Conventional cut and cover box stations were changed into large
diameter NATM shafts, never before executed in Europe. This paper describes the most significant station design and
construction issues. The TBM tunnels also required innovative technology in order to deal with highly variable ground
conditions. The Additional Face Support (AFS) System was used for the first time on two large diameter EPB machines.
This paper also describes how this system works and why this unique feature is an important addition to EPB machines
facing variable ground.

160
named Marqus and Salgueiros were commissioned to
Brazilian design offices CJC Engenharia and
Figueiredo Ferraz. The innovative design, involving
the use of large shafts, was followed by technical
construction management carried out by the principal
designer (CJC), and supervised by the main author,
acting at the time as the Project Director for the civil
group.
3 MARQUS STATION
The Marqus station is made up of a central elliptical
shaft and two NATM tunnels. The shaft is 27 m (88
ft) deep and its elliptical footprint is 48 m (157 ft)
along the major axis, and 40 m (131 ft) along the
minor axis. The tunnels are 18 m (59 ft) long and have
a section of 180 m
2
(1937 ft
2
).
The heterogeneous nature of the Oporto granite
was very evident at the station site. A sharp sub-
vertical fault, running obliquely to the shaft main axis,
separated weathered soil-like granite from moderately
to slightly weathered good granite rock. This strong
heterogeneous character was one of the main reasons
for avoiding slurry walls at Marqus.
A shotcrete lined, large diameter shaft allowed
the excavation of most of the station volume in an
open cut, requiring no anchors or struts. This
contributed a great deal to a fast paced excavation,
advancing 4.5m (14.7 ft) each month. Shaft
excavation was completed in six months, between
June and November 2002.

Figure 1 View of Marqus Plaza and station footprint.
Note tunnel alignment oblique to Plaza.

An added difficulty was the requirement to
preserve the century old maple trees that embellish the
Plaza. Only six trees, in the center of the Plaza, were
allowed to be relocated. This was taken as a design
input and contributed to the final shape of the shaft.

Figure 2 Aerial view of the shaft and surrounding maple
trees.
According to the sequential excavation design
guidelines, the shaft was excavated in panels. Panels
were 1.8 m (6 ft) high, with horizontal lengths varying
between 4 m and 12 m (13 ft and 39 ft). The exposed
granite on each panel surface was immediately
protected by three layers of shotcrete and welded wire
mesh. The shotcrete wall thickness varied with depth
between 0.3m and 0.6m (1 ft to 2ft). A final liner was
provided by a cast-in-place concrete wall.
During excavation, the rock mass was dewatered
by horizontal drainage holes installed systematically
on the shaft wall (4m long, spaced at 1.8m). This
allowed the temporary shotcrete lining to remain in a
drained and depressurized condition. Inside the shaft,
deep vertical relief wells were also employed, to
reduce water uplift forces and minimize the risk of
hydraulic failure of the bottom.
The challenge with large diameter shafts is
maintaining the stability of the shaft walls around
large openings. At Marqus Station, in order to
accommodate the light rail platforms with a length of
70 m (230 ft), two NATM tunnels 18 m (59 ft) wide
and 18m long were constructed from the shaft. This
produced large openings on the shaft walls and
therefore required that a reinforced concrete (RC)
frame be built prior to excavating the tunnel.
The loading imposed on the shaft by creating
these two large openings is best modelled with three
dimensional numerical models. 3D numerical analyses
were carried out with STRAP (finite elements) to
determine bending moments, axial and shear forces
acting on the shaft wall and frame, as well as their
deformation, before and after the excavation of the
openings. The cast-in-place reinforced concrete frame,
built prior to the openings, was incorporated in the
model. It is essential for the stability of the shaft that
the RC frame be stiff enough and capable of
161
maintaining, as much as possible, the vicinity of the
opening in compression. The basic dimensions of the
RC concrete frame are shown in Figure 3. The
NATM tunnels attached to the openings were
analyzed with a two dimensional finite difference
model named FLAC.
Nvel dos carris

Figure 3 Schematic view of the RC frame.
From a construction point of view, some
important advantages can be associated with a
combination of large diameter shafts plus short
tunnels. During excavation, a large, strut-free volume
is available. Because no anchors are required, no time
is lost in waiting for the anchor subcontractor to finish
a level before excavation is allowed to proceed to the
next level.

Figure 4 View of the RC frame and adjacent NATM
platform tunnel at Marqus Station.

Also, as the shaft is excavated in sequential
panels, if ground treatment is required due to weaker
than expected geology, ground improvement
techniques may be applied at the desired location,
while the rest of the shaft perimeter may continue
with routine panel excavation or shotcreting. At
Marqus, some weak granite exposed on the shaft
wall required treatment by means of jet grouting. This
was applied in about 20% of the shaft perimeter and
the jet grouting columns had about half the shaft
depth.
During construction of the stations internal
structures, another major advantage of the large shaft
plus tunnel arrangement becomes evident. The
volume in the shaft is enough to house all station
technical rooms, commercial spaces and public areas.
The tunnels are used for platforms only.
Again, as the shaft is free of struts, the
unobstructed bottom-up construction of internal walls,
pillars, slabs and beams, becomes similar to any RC
building on the surface. Construction of all station
internal structures consumed eight months only.

Figure 5 Internal structures at Marqus Station.
From the architectural point of view, Marqus
Station, having such a unique geometry, has become
an icon station in Oporto. As metro users enter the
station and proceed downwards to platform level, the
tall standing curved walls provide a notion of space
and bold engineering design. Marqus Station has
been in operation since September 2005.

Figure 6 Finished station. View from the shaft looking
towards one platform tunnel.
162
4 SALGUEIROS STATION
Much has been written about the advantages and
disadvantages of design-build contracts. Although it
is clearly beyond the scope of this paper to approach
the topic, it is important to highlight that Marqus and
Salgueiros stations are examples of innovative
engineering design that could only be accommodated
in a design-build scheme. Such bold engineering
solutions prosper when, in additional to technical
competence, there is a strong confidence link between
the designer and the builder. Together, they have the
strength to sell the idea to the owner and all other
stakeholders. This is more likely to happen in design-
build situations, than otherwise.

The double shaft design of Salgueiros Station is
unique in Europe. The geometry is made up of two
incomplete ellipses, which combined produce an open
cut more than 80 m (262 ft) long and about 40m (131
ft) wide. The excavation depth is 24m (78 ft).

Figure 7 Footprint of Salgueiros Station.

The geology at the station site is predominantly
made up soft, highly weathered granite. Medium to
hard granite is found only at excavation bottom. The
engineering solution for the excavation works is
bolder than at Marqus.
As the two large ellipses are not excavated as
closed rings, a very stiff reinforced concrete frame,
located where the ellipses touch each other, was
required in order to provide stability.
The RC frame was constructed before the station
excavation. Two small diameter circular shafts (3.3m;
10.8ft) were excavated and filled with a rebar cage
and concrete, to form the pillars of the frame. At the
surface, a large RC beam, cast on surface ground,
provided the connection between the two pillars. The
beam is 30 m (98 ft) long, 2m (6.5 ft) high and 1.6 m
(5.2 ft) wide.
Once the pillars and the beam acquired enough
strength, the large ellipses were excavated and
supported with a temporary shotcrete lining. In order
to maintain both ellipses with uniform ground loading
on the shotcrete shell, excavation occurred on both
sides simultaneously.
The entire station volume was excavated without
struts or anchors. The ground was dewatered by
means of vertical wells prior to excavation.

Figure 8 Introducing the reinforcement cage in one of the
frame pillars at Salgueiros Station.
The elliptical shafts were excavated in panels.
Panels were 1.8 m (6 ft) high, with horizontal lengths
varying between 4 m and 12 m (13 ft and 39 ft). Each
panel was immediately supported with shotcrete and
welded wire mesh.

Figure 9 Salgueiros Station excavation. Note open panel
at the top.
The temporary shotcrete lining was applied in
three layers, so as to avoid shadowing by the wire
mesh. The thickness of the lining increased with
depth, going from 0.35 m (1.1ft) near the surface, to
0.60 m (2 ft) near the bottom. Similarly to the
Marqus geometry, the temporary shotcrete lining at
Salgueiros was designed with the use of three
163
dimensional numerical models, such as STRAP (finite
elements). Because of its round shape, the
predominant forces on the lining are compressive.

Figure 10 Salgueiros Station fully excavated and ready for
internal structures, in a space free of obstructions.
The loading scenarios were quite different in the
construction stage (Figure 10) from the final
operational stage. During construction, as the water
table had been drawn down, the loads acting on the
temporary structure were due to lateral earth pressure
and surface surcharge only. In the final stage, with the
internal reinforced concrete structure in place,
hydrostatic pressure and seismic loads were included.

Figure 11 Finished concourse level.
In the final station configuration, intermediate
slabs at concourse and mezzanine levels provide a
strutting action for the final RC lining, which has a
design life of one hundred years. For the long term,
the temporary shotcrete lining was disregarded.
5 ADDITIONAL FACE SUPPORT SYSTEM
The TBM driven tunnels in Oporto were excavated by
two large diameter (>8.5m; 28ft) Earth Pressure
Balanced TBMs (EPB-TBMs). Very early in the
excavation process, during the first TBM drive, a
large sinkhole produced a fatal accident. It became
evident that the Oporto granite could not be safely
excavated in open mode or semi-closed mode (half
muck-filled excavation chamber, with compressed air
in the upper part of the chamber, Figure 12).

Figure 12 Different modes in EPB-TBM operation.
In highly heterogeneous weathered granite, displaying
complex hydrogeological behavior and erratic
permeability, compressed air support can be
ineffective. As such, following the incident, TBM
operation was changed to strictly closed mode.
In Oporto, as the first machine excavated granites
in different stages of weathering, ground properties
varied significantly over short distances. These
complex ground conditions often produced
oscillations in muck density and also in face support
pressure.
Babendererde Engineers, contributing as a member
of the Metro do Porto Panel of Experts, proposed that
both EPB-TBMs be equipped with the Additional
Face Support System (AFS) for the first time in
tunnelling.
This is a fully automatic system that adds to the
EPB face support control the ease of operation and
reliability of the Slurry-TBMs. Independent from the
advance of the TBM, the AFS System ensures the
application of the required support pressure.
Open Mode
Semi-open Mode
Closed Mode
Compressed air
P
e
Support pressure P
- not required
- no additives
for conditioning
- moderate support
pressure by
compressed air
- no additives
for conditioning
P
a
P
e
P
S
> P
W
+ P
E
+ P
OT
- required
- additives for
conditioning
Requirements on
ground properties
- enough cohesion
- low ground
water table
- enough
cohesion
- moderate
groundwater
table
- variable
cohesion
- high
groundwater
table
P
W
P
E
P
W
P
E
P
W
P
E
P
S
Additives
164
A tank on the back-up, filled with a support
medium such as bentonite slurry, is kept under the
necessary face support pressure by compressed air.
An automatic valve opens the link between the tank
and the working chamber as soon as a critical support
pressure drop is registered by the pressure sensors.
By the compressed air, the slurry is pressed into the
TBM working chamber until the tunnel face is
supported again by the necessary face support
pressure. Upon reaching this pressure level, the valve
closes again. The system works completely
automatically. Intervention by the TBM operator is
not required and, in reality, not recommended.

Figure 14 AFS System installed on both EPB-TBMs.

The use of the AFS System on both TBMs proved
to be a successful tool. In combination with other
measures, it was possible for the contractor to
overcome the challenging Porto ground conditions
and complete both drives successfully. The AFS
System is now a proven equipment component on
other TBM drives in difficult situations.

6 CONCLUSIONS

The experience in Oporto is unique and filled with
innovation. It is important in a sense that it shows that
a lot more use can be made of circular shafts than has
been done so far.
It demonstrates that circular and elliptical shafts can
be used for more than just housing pumps, ventilation
units or emergency escape ways.
Large underground light rail stations, with cutting-
edge architectural design were accommodated in such
shafts.
Construction benefited from a strut and anchor free
space, both during excavation and also during the
erection of the internal structure. This contributed to
significant savings in time and in resources.
The AFS System, in its first application ever,
contributed to maintaining the adequate face support
pressure in the excavation chamber, which was highly
prone to oscillation due to the nature of the local
granite.
7 ACKNOWLEDGEMENTS
The authors would like to acknowledge and thank the
following entities: Metro do Porto, Normetro A.C.E.,
Transmetro A.C.E, CJC Engenharia Ltda and
Figueiredo Ferraz Cons. Eng. Proj. Ltda.
8 REFERENCES
1. Andrade, J. C. et al., 2004. Estaes suberrneas em
poos e tneis no Metro do Porto: Aspectos gerais de
projecto e acompanhamento tcnico da obra- ATO.
Conference on Geotechnics, Aveiro, Portugal.
2. Franco, S. G., et. al., 2004. Estao do Marqus em
poo no Metro do Porto. Modelao e Segurana.
3. Frana, P.T. et. al., 2004. Estao de Salgueiros em
poo no Metro do Porto. Modelao e Segurana.
4. Babendererde, L.H., 1999 TBM drives in Soft Ground-
Weak points in process engineering and their
consequences. Proc. of the World Tunnel Congress 99,
Oslo, Norway:811-815.Rotterdam: Balkema.
5. Babendererde,S. & Holzhauser, J. 1999 Der
Betriebszustand Druckluftstutzung beim Hydroschild.
Taschenbuch fur den Tunnelbau 2000: 231-252.
Essen,Gluckauf.
6. Babendererde, S., Babendererde, J. & Holzhauser, J.
2000. Difficulties with operation of slurry tunnel
boring machines.North American Tunnelling 2000,
Proc. Boston, June 2000: 317-326, Rotterdam;
Balkema.
165
1 INTRODUCTION
The Canada Line is a 19km long rapid transit system
with 16 stations with a capital cost of C$1.9B. The
cost of the infrastructure is approximately C$1.6B.
The project was awarded to InTransitBC to design,
build, partially finance, operate and maintain the line
for 35 years. The project must be complete by
November 2009 to be ready for the 2010 Winter
Olympics to be held in Vancouver and Whistler, BC.
The Bored Tunnel section consists of 2.4km of twin
bored tunnel and 3 stations. The scope of work for the
stations is limited to the works necessary to facilitate
the bored tunnel works. A joint venture between SNC
Lavalin Constructors Pacific and SELI is executing
the design and construction of this section. Work for
the bored tunnel commenced in November 2005 with
shoring and excavation works for the tunnelling
operations pit that will become the Olympic Village
Station. This paper describes the specifics of the bored
tunnel section and the construction progress up to J uly
2006.
2 STATIONS
There are four underground stations connected by the
bored tunnel. The joint ventures scope of work
includes all station works required to facilitate the
tunnel boring operations. In most instances this
requires the joint venture to carry out utility
diversions, shoring and excavation works and the
stations reinforced concrete base slabs. The
excavation for the Olympic Village Station doubles as
the tunnel boring operations site. A shotcrete and
anchor shoring system was adopted through
contaminated fill material, glacial till and sandstone.
Being a wide open area, this simplistic shoring system
was suitable. The Yaletown station has significant
challenges in that the excavation is from property line
to property line on Davie Street between two high rise
residential towers. In some locations, the 4 level
underground parking structures border the property
line and in others it is set back. Due to the depth of the
excavation (20m), the ground conditions (glacial till
with boulders and lenses of coarse, water-bearing
material), the clearance to underground structures and
the loading from those structures, a shotcrete and
anchor shoring technique was deemed unsuitable and
risky. The clearance between the tunnel and the
property line (approximately 600mm) required some
form of slender element to form the structure of the
shoring. A drilled micropile method was chosen. This
consists of a drilled hole filled with a steel pipe of
176mm diameter at 550mm or 650mm spacing and
filled with grout. Two layers of internal bracing are
required, 7 and 11m below ground and a tie back
anchor at -16m below the buildings (Figure 1). The
first tunnel drive will pass through the station prior to
excavation. A vehicular bridge is also required across
the excavation spanning 22m.
Design and Construction of the Canada Line Bored Tunnel
Section
Brendan Henry, P.Eng, C.Eng (MICE)
Vancouver, BC, Canada
The bored tunnel section of the Canada Line holds many technical and schedule challenges. Construction began in
November 2005 and will be completed in October 2008. The alignment will cross through two geological interfaces and
may encounter many boulders. It will negotiate tight 200m radius curves and pass close to the foundations of several
buildings. The tunnel boring machine will also encounter some man made obstructions such as tie back anchors from the
shoring of deep excavations of adjacent buildings. This paper describes the features of the project, the tunnel boring
machine, some of the risks and management of such and provides a construction update.

166

Figure 1. Yaletown Station excavation shoring system.

The shoring for Vancouver City Center Station also
utilizes a vertical shoring element. However, as the
ground conditions of till over sandstone are better than
those at Yaletown Station, a 100mm minipile pipe is
required and is tied back by rock anchors under the
adjacent building. As with Yaletown Station, the
TBM will pass through the station before the
excavation is complete. The TBM cutting head will be
exposed and serviced at this location, some 2km into
the drive. Waterfront Station, where the TBM is
extracted, will be built under a separate contract.
3 GEOLOGY AND ALIGNMENT
From the launch shaft, the tunnel immediately drops
into a vertical curve and down a 5.5% gradient to
clear building foundations and the body of water
known as False Creek. The first 600m are located in
weak sandstone and siltstones of the Kitsilano
member with UCS of 5 to 10 MPa that is interspersed
with cemented sandstone floaters with UCS of
~50MPa. The rock is generally massive, with very
high R.Q.D. J ointing is poorly developed or absent,
except along the bedding, which is horizontal to
shallow dipping (from north to south). Under False
Creek, the alignment starts to rise at 5.5% and passes
through a shallow interface between the sandstone and
a buried glacial valley, which is filled with dense to
very dense glacial till. The till consists of a clay-silt-
sand matrix with gravel. Fines content is typically 45
to 50 percent, although sandy horizons are known to
be present. At this interface, the high tide level in
False Creek is 30m above the tunnel invert.
The till is interspersed with granite and
granodiorite boulders that vary in size up to 2 or 3m
across and over 250MPa in strength. The tunnel
comes ashore under the raft foundation of a 33 storey
building and continues along to Yaletown Station and
Davie Street, at depths ranging from 15 to 35m before
curving through a 200m radius onto Granville Street.
Here the TBM passes under two deep foundation
multi-storey buildings with approximately 10m of
cover.

167

Figure 2. Aerial view of bored tunnel alignment.

Shortly after, the tunnels re-enter the sandstone and
pass through Vancouver City Center Station before
dipping to narrowly avoid a pedestrian underpass and
rising again to pass over an existing LRT tunnel with
only 1m clearance. The drive ends shortly after at an
extraction shaft at Waterfront Station.
Occasional medium strong basalt dykes up to two
meters thick, and with unpredictable orientations, are
expected to be encountered in the Kitsilano member,
as is minor faulting and shearing, particularly
associated with the stronger siltstones.
Considering the geology, the proximity of the
tunnel to foundations and the urban setting, an Earth
Pressure Balance (EPB) tunnel boring machine was
chosen.
4 TUNNEL BORING MACHINE
The design and fabrication of the tunnel boring
machine (TBM) was competitively tendered to several
manufacturers. Lovat Inc. of Toronto, Canada was
selected as the manufacturer. The machine will
operate in full EPB mode where necessary and foam
agents will be added to condition the rock and soils to
maintain earth pressure transfer.

4.1. TBM General Dimensions

Shield Diameter: 6026 mm
Cut Diameter: 6064 mm
Overcut Diameter: 6089 mm
Length: 9.0 m
Overall Length: 82 m
Overall Weight: approximately 400 tonne

4.2. Cuttinghead

The cuttinghead of the TBM has a mixed face design
which incorporates 17 Twin Tip Disc Cutters which
are interchangeable with Rippers. The head can be
configured with either a full face of disc cutters
(Figure 3) for the anticipated rock, a full face of ripper
teeth for the soft ground sections, or a combination to
deal with boulders in soft ground.

Figure 3. Disc cutter cuttinghead configuration.
168
Basic design features of the head include:

Incorporates either Ripper Teeth or Disc Cutters
Scraper Teeth
Eight Spoke Design
7 No. Independent Injection Ports
Chromium Carbide Plate for Abrasion
Protection
Grizzly Bars Across Cuttinghead Openings

4.3. Main Drive

The Main Drive consists of a variable frequency
electric drive (VFD) incorporating a high capacity
triple roller bearing, gear reducers with integral
pinions, water cooled electric motors and VFD units
controlling electric frequency and voltage to the
system. Specifications of the drive are as follows:

1200 kW Installed Power
1104 kW Available at the Cuttinghead
6,180 kNm of torque at 0.0 to 1.7 rpm
Maximum Speed of 3.2 rpm
Peak Starting Torque of 7,750 kNm

The above values take into consideration system
inefficiencies. Of note is the fact that the installed
torque on this machine is 18% higher than current
machines of similar diameter and 100% higher than
machines of similar diameter built only 8 years ago.

4.4. Muck removal

The TBM is equipped with a screw conveyor located
at the invert of the cuttinghead chamber (Figure 4) for
controlled removal of excavated material.
The screw conveyor is equipped with a peripheral
drive that allows the material to flow through the
drive unimpeded. This allows for more control of the
discharge and helps in aligning the location of the
trailing belt conveyor to accept the material from the
screw conveyor.
The trailing belt conveyor discharges onto a
shuttling conveyor that loads the muck cars. The use
of a shuttle conveyor increases the safety of the
operation as it negates the need for the train to move
during the mining operation, which is especially
important on the 5.5% grades.

Figure 4. Cuttinghead chamber and screw conveyor.
4.5. Segment Erector

The segment erector on the TBM uses a vacuum
system to pick up the segments. This is the first use
of a vacuum pick-up mechanism for segments on a
TBM in Canada. The segments must be kept cleaner
during handling as debris will clog the vacuum
system. The vacuum lifting device must have enough
buffer capacity to hold the segment at 90 degrees for
over 15mins in case of power failure.
5 LINING AND GROUTING
5.1. Pre-cast segmental lining

The pre-cast concrete segmental lining consists of a
250mm thick reinforced concrete section 1.4m wide
with a finished internal diameter of 5.3m and an
external diameter of 5.8m. Each ring consists of 5
segments plus a key. The ring is tapered to allow for
negotiating the tight corners. The pre-cast plant is
situated in Nanaimo on Vancouver Island and was
expanded to accommodate the casting of the tunnel
lining. The expansion required a concrete batch plant,
two 10T overhead cranes, a steam generator
connected to a fully computerized temperature control
system and the segment moulds.
Three sets of moulds were supplied by SLCP-
SELI and were manufactured by S.A.M.E in Italy.
To achieve the required production schedule, nine
complete rings per day are required to be cast. Three
cycles are required within one 24 hour period.
Strength of 15MPa is required in 5 hours in order to
achieve this schedule. A high early strength concrete
169
mix, combined with steam curing is utilized to meet
the demanding cycle time.
The segments are demoulded using a hydraulic
lifting/turning device (Figure 5a) and the segments are
then stored in the yard until they reach the design
strength of 40MPa.

Figure 5a. Segment demoulding in operation.
The segments are shipped to site by truck using BC
Ferries service from Nanaimo to Tsawwassen on a
daily basis.
Once at site, bituminous packers and rubber
gaskets are placed on the segments. Under False
Creek, a hydrophilic gasket will be used in place of a
standard rubber gasket.

5.2. Grouting system

A two component (A/B) grout system is utilized
where the retarded grout (A) is mixed by an
automated batching plant outside the tunnel and the
grout is pumped along the tunnel to a holding tank on
the TBM through a 1 line. The grout consists of
cement, fly ash and bentonite. With the retarder, the
grout can stay in the lines and tanks for a number of
days. A one inch line was chosen to keep the velocity
high to prevent possible segregation of the grout
components. The accelerator (B) component is
pumped from surface also into a separate holding tank
on the TBM. The system is designed for grouting
through the trailing shield thereby allowing for
immediate backfill of the annulus as the machine
advances. The specifications of the grout system are
as follows:

6 Injection points
6 A Component Pumps (peristaltic)
6 B Component Pumps (peristaltic)
Pig launchers/receivers for line maintenance
Computer Controlled Injection linked to TBM
advance (Figure 5b).

Figure 5b. Computer controlled grout injection system.
The benefits of the two component system are
highlighted by quick setting times providing
immediate support to the lining as it moves out of the
TBM. Grouting operations are linked to TBM advance
to ensure that as the void is being created by the
advance of the TBM it is being filled instantaneously
with grout. As the pumps shut down the A component
pumps continue to run for 30 seconds more that the B
component to prevent clogging of the injection ports.
This is the first use of this type of grout system on
a TBM in Canada. The method was developed by the
J apanese and has since been adopted in Europe and
the United States on numerous projects.
6 TBM OPERATIONAL CONTROLS AND
GUIDANCE SYSTEM
The TBM is operated from a control cabin that houses
all the controls and computer systems. The Human
Machine Interface (HMI) consists of three screens
showing the operating systems and guidance system,
six CCTV screens and all the necessary buttons and
levers. Here the operator can see all the functions of
the machine on several screens. The main screen
shows the TBM position, the thrust cylinders pressure
and extension, the cutting head speed and torque, the
screw conveyor speed and torque, the earth pressures
and many other features (Figure 6a).
The machine position and ring build is controlled
with the assistance a tacs acs guidance system. The
system utilises a laser theodolite, target and software
that is fully integrated with the TBMs computer
systems. acs shows the position and orientation of the
cutting head relative to the design alignment and

170

Figure 6a. Main screen of HMI.
suggests the orientation of the ring build. When off
alignment, the system computes a corrective curve to
avoid over corrections and kinks in the tunnel. The
system also shows the position of the lining relative to
design.

Figure 6b. tacs acs guidance system showing TBM location
along alignment.
The TBM position can be shown in real time on the
project drawing plan and profile. This is a particularly
useful tool as the operator can visualise when the
TBM passes under a building or other obstruction or
nears a different geological zone (Figure 6b).
7 FACE STABILITY & SETTLEMENT
ANALYSIS
The face support pressure is calculated for each
section of the alignment by a specialist designer
(Professor Kovri). The face support pressure, or earth
balance pressure, is calculated based on a number of
factors including overburden pressure, superimposed
pressures such as structures, hydrostatic pressures and
permeability. The face support pressure is then chosen
based on the consequences of losing face support, i.e.
the equilibrium pressure (support pressure equal to
exerted) under a building may be 100kPa but if there
is a loss of support pressure during mucking, ground
movement may occur leading to possible settlement.
Simplistically, the operational face pressure will be set
at equilibrium pressure multiplied by a factor of
safety. This factor will be chosen based upon factors
such as type of ground being excavated, potential for
anomalies in the ground (e.g. pockets of running sand
in a glacial till), proximity to structures/ground
surface, and even the type of structure and what kind
of movement it could tolerate.
The magnitude of ground movement is calculated
through empirical methods and finite difference
analysis using FLAC. Critical sections (primarily
where the tunnel is close to structures) along the
alignment were analysed for ground movement from
tunnel boring operations and station excavation
(Figure 7).

Figure 7. Settlement analysis using FLAC.
The final choice of face support pressure is made
with consideration of these computations and a risk
analysis.
171
8 RISK ANALYSIS AND SETTLEMENT
MONITORING
The risk analysis was performed by the project team
and specialist consultants including Professor Kovri.
The aim of the risk analysis was to identify all
possible risks associated with the boring operations
and to assess the probability of that risk being
realised. The approach is intentionally conservative
and all what if? scenarios are investigated. The
result of the risk analysis is to set out face support
pressures and any special measures to be taken before,
during or after boring. For example, 75m after
launching the TBM, before all back up gantries are
installed, it advances below the foundations of a
multi-storey building with only 2m of rock cover. The
building is founded on expanded base piles with
unknown penetration into the till. The alignment goes
under the parking basement of the building and just
clips the edge of the main tower. The following
information was unknown:
Full details of the structural system of the
building.
Penetration of the piles.
Loads on piles.
Rock quality at the pile locations.
Rock cover to the tunnel at pile locations.

The following information was known:
General cover to tunnel close to the building.
Alignment relative to individual foundations.
Magnitude of expected settlement from analysis.
Rock quality from two boreholes adjacent to the
building including viewing of cores.
Behaviour of the rock during mining of the 7m
span tail track tunnels (150m away).
Behaviour of the rock during in other mined
tunnels in the Kitsilano member.

The overall risk of ground movement and potential for
damage to the structure is low. However, should
anomalies exist in the ground, the potential for ground
movement is real. If ground movement were to occur
under individual footings or piles resulting in loss of
bearing the structure above may be affected. As noted
earlier, the alignment passes briefly under the main
tower of the building.
The risks for this structure could be summarized
as follows:
Risk of poor quality rock around footings.
Risk of loss of support during tunnelling resulting
in ground movement around foundations.
Risk of loss of bearing of foundations resulting in
movement of the structure.
Risk of structural deformation of main tower.
The probability of realizing any of these risks was
deemed to be very low. Options for
avoiding/mitigating these risks include:
a. Move horizontal alignment to avoid main tower.
b. Move vertical alignment to provide more cover.
c. Investigate the rock quality at each footing.

The following actions were taken to mitigate or avoid
these (very low probability) risks:
The horizontal alignment was moved 2m eastward
to avoid the footings of the main tower. This
decreased the tunnel to tunnel spacing to 4m over
a short distance which, in the Sandstone, was not
thought to create any problems during boring.
A small drill rig was taken into the basement of
the structure to investigate the rock quality at each
footing. A small drill rig was used due to the
limited headroom. Very strong glacial till was
encountered and the small rig did not have
sufficient power to penetrate this till.
The operating face pressure was set at 100kPa;
under the building this pressure is slightly higher
than the pressure of geostatic and imposed loads
providing a factor of safety against loss of
support.

The vertical alignment could not be altered at this
location so providing increased cover to foundations
was not possible.
The following settlement monitoring points were
deemed necessary to detect potential settlement:
Level monitoring points on the columns in the
foundations of buildings.
Street level monitoring points (set in the soils
below road surface).

In addition, three borehole multipoint extensometers
were set up in the overlying till and in the bedrock
between the two tunnel drives. Although the
extensometers were not thought to be necessary, the
results would provide good intelligence for ongoing
risk analysis further along the alignment.
To date, only 1mm settlement has been detected
in any of the monitoring devices. No movement has
been detected in the two extensometers reached so far.
Effectively, there has been no settlement. The
settlement analysis for the first structure showed
potential for up to 6mm of ground movement.
9 OPERATIONS & LOGISITCS
The 82m long TBM and back-up starts from a 65m
long by 18m wide by 14m deep excavation that will
become the Olympic Village Station. The last 4 gantry
cars are added as the machine is buried. The TBM is
172
launched on a 2% down gradient and immediately
starts into a vertical curve followed swiftly by a
horizontal curve of 315m radius. A train consisting of
a 35T locomotive, 4 muck cars, a personnel carrier, a
flat deck car and two segment cars services the TBM.
Two of the flat deck cars that carry the muck cars are
powered to assist in hauling up 5.5% grades. To avoid
having to move the muck cars during mining, the back
up system is equipped with a shuttling conveyor. This
not only adds to the safety of the system but means
the loco does not have to mobilise the train to move
only a few metres during mucking on steep grades.
The pit bottom is configured for two trains side by
side. With this track work configuration, the shaft has
insufficient length to accommodate the two trains and
the switches to enter the tunnel. Therefore, 15m long
tail track tunnels were excavated in the Sandstone
with a roadheader attachment to a backhoe. These
tunnels were lined with steel ribs and shotcrete. The
expected average mucking cycle for the 1.4m forward
shove is approximately 40 minutes and time taken to
build the ring is 20 minutes. During the mucking
period the lining segments are offloaded by two
unloading gantry cars and arranged for installation.
Upon completion of the mucking cycle, the train exits
the tunnel and another train enters. A gantry crane
with a SWL of 52T spans the excavation and unloads
the muck cars into a muck pool where it awaits
transportation by truck to the ocean disposal barging
site. Lining segments are loaded and the train re-
enters the tunnel after the other arrives. Due to the
relatively short length of the bore, no locomotive
passing places are required within the tunnel.
Grouting is continuous during the forward shove of
the TBM. Advance rates will average between 10 to
14m per day depending on ground conditions. The
tunnel will be driven using a combination of 3x 8 hour
shifts and 2 x 10 hour shifts with 3 shift configuration
being utilised in difficult areas of the drive.
Approximately 50 workers are required to directly
cover the tunnel boring activities.
10 CONSTRUCTION STATUS
In November 2005, the shoring and excavations works
for the tunnel boring operations pit began. The
excavation was completed in March 2006 and
construction of the tail track tunnels and reinforced
concrete base slab began immediately after (Figure
10a).

Figure 10a. Tail tunnels and R.C works at TBM pit.
In May 2006, the TBM was delivered from Lovat.
There was a significant delay in delivery of the TBM,
mainly due to the bearing delivery being delayed from
the U.S. manufacturer due to the Defense Priorities
and Allocations System (DPAS) program. Under this
program, manufacturers are required to prioritise
orders supporting national defence programs. In this
case, the manufacturer was required to complete an
order of 400, then 900, bearings for the gun turrets of
US Humvees in Iraq. The bearing arrived directly
from the factory and had to be installed vertically in
the machine. The TBM was lowered in sections into
the operations pit with a 440T derrick crane. The
shotcrete and anchor shoring system of the pit was
designed to allow the outriggers of the crane to sit on
the edge of the excavation (Figure 10b).
The TBM, unloader sleds and 5 of the 9 gantries
were installed prior to starting boring (Figure 10c).
The machine broke ground toward the end of J une and
has since advanced over 100m. Over 80m were bored
before the TBM had all gantry cars in place and the
four muck-car train available. At 170m into the bore,
the boring operations will stop for one week while the
pit bottom is reconfigured.
173

Figure 10b. Installation of stationary shell of TBM.
During this time the 8 thrust rings and thrust frame
will be removed and all track work laid down to allow
two trains side by side in the pit.

Figure 10c. Start up TBM and back up configuration.
Micropiling for Yaletown station will be completed in
September (Figure 10d). Shortly after, a temporary
bridge will be placed to span the station and
excavation work will begin. The first drive will pass
through the station before the excavation work is
complete.

Figure 10d. Commachio MC-1200 drill rig installing
micropiles in Yaletown.
Construction of the Vancouver City Center Station
begins in August. The excavation will be
approximately 80% complete when the TBM arrives,
after which the excavation works will be delayed until
the end of the first drive.
The first drive will be complete around April
2007 and the second drive 10 months later.
Construction of tunnel track bed and walkways, cross
passages and pumping station are scheduled to be
complete by October 2008.
ACKNOWLEDGMENTS
Tunnel boring machine information for this paper was
summarized by Simon McNally of Lovat Inc.
Other information was provided by SELI employees
Andrea Ciamei, Leonardo Pia, and Peter
Andrikopolous.
174
1 INTRODUCTION
The Mill Creek project is located in the Greater
Cleveland area and serves 134,000 people in 11 of the
Northeast Ohio Sewer Districts sixty-member
communities. The total contract cost for the three-
phase development will be about $150,000,000. The
first phase (MCT-1), a 3-m diameter conveyance
tunnel, was completed in 1999. The second phase
(MCT-2), a 7.3-m excavated diameter storage tunnel,
was completed in 2005. The third phase (MCT-3),
also a 7.3-m excavated diameter storage tunnel is
currently under construction with planned completion
in 2008. This paper will focus on the large diameter
tunnels excavated under Phases 2 and 3 (MCT-2&3).
Further details can be found in References [1] to [3].
2 GEOLOGICAL CONSIDERATIONS
2.1. Rock Structure
The MCT-2 and 3 tunnels are located within the
Chagrin shale rock formation. This shale is known to
contain closely bedded zones of siltstone, limestone
and sandstone layers. Shale exhibits only local
dampness in tunnel excavations. Gas, primarily
methane, is commonly encountered in this formation.
There are no supportive arguments on swelling
characteristics of Chagrin Shale; it is therefore
believed to be negligible.

2.2. Rock Strength and Stresses
The unconfined compressive strength generally ranges
from 14 MPa to 55 MPa, which can be classified as
weak to medium strong rock. Average rock strength
for Mill Creek tunnels is illustrated in Figure 1.
Higher strength values were noticed in rock samples
where siltstone and sandstone interbeds were present.
To date, in-situ stress measurements have not
been carried out in the project area. However, it is
common knowledge that the major horizontal
compressive stress in the Cleveland area rock
formations trends approximately N80E and has a
magnitude about two times the vertical stress. Based
on the previous studies performed in the region, the
GBR for Mill Creek tunnels suggests the horizontal
stresses in the rock range from 6.9 to 28 MPa. The
horizontal stress varies in relation to the tunnel
alignment.

Figure 1. Chagrin Shale Average Strength.

Mill Creek Tunnel Geomechanics
B. Lukajic, M. Schafer, & R. Pintabona
Montgomery Watson Harza, Cleveland, Ohio, USA
M. Kritzer, S. J anosko & R. Switalski
Northeast Ohio Regional Sewer District, Cleveland, Ohio, USA
ABSTRACT: The purpose of this paper is to review technical considerations and summarize the methods used in constructing the
large diameter Mill Creek tunnels in shale. The three-phase tunneling construction program encompasses nineteen (19) shafts and
three (3) tunnels, totaling 12,727 m of tunnel length. The paper describes the experience gained during design and construction,
relative to specialized techniques used for ground improvements and exploration.

M
C
T
-
1
M
C
T
-
2

M
C
T
-
3

0
5
10
15
20
25
30
35
40
U
C

S
t
r
e
n
g
t
h

(
M
P
a
)
Average =23 MPa
175

2.3. Rock Behavior while Mining
Observations during construction indicate that stress is
not the dominant mechanism in tunnel stability. This
is evidenced by the fact that the tunnel roof remained
intact, with little tendency for overbreak in massive
shale beds. Occasionally, local rock blocks loosened
due to separation along vertical joints and horizontal
bedding.
When the tunnel intersected closely bedded
shale/siltstone zones, more frequent slabbing and
loosening occurred. Slabbing normally occurred in the
crown at 11:00 and 1:00 oclock positions. An
example of thin shale slabbing is illustrated in Figure
2.

Figure 2. Chagrin Shale Thin Slab.
3 ISSUES RELATED TO GEOLOGY
Variations in geological conditions are often
encountered in tunnel projects. Such variations
sometimes require additional field explorations and
adjustments to design and construction methods.
Several geology related adjustments to the design and
construction occurred at the Mill Creek project.
First, the presence of a deep soil valley within the
alignment of the tunnel had a major bearing on design
and construction methodology of two large diameter
shafts.
Second, the presence of the soil valley required an
exploratory tunnel to be constructed to investigate the
depth of the valley and the bedrock beneath it, in
advance of the main tunnel, as depicted in Figure 3.
Third, an eight-month shutdown of the MCT-3
main tunnel TBM drive was required to mitigate
methane gas conditions.
A brief discussion of each issue is presented
below.

Figure 3. Location of the Exploratory Tunnel Under Buried
Valley.

3.1. Shaft Construction Within Buried Valley
Constructing 10-m diameter, 55-m deep shafts in a
buried valley presented several difficulties and risks to
Owner (NEORSD) and the Engineer (MWH). Failure
to successfully excavate the shafts could result in
protracted construction delays and costly claims.
Considering the fully saturated silt-sand soil condition
and the risks involved, four alternative methods were
evaluated for construction of two shafts. They
included slurry wall, jet grouting, deep soil mixing,
and ground freezing. Artificial ground freezing was
chosen as an initial support method during
construction of the shafts. It was determined that
neither slurry wall nor jet grouting could be relied on
to overcome boulder obstructions at the depth of 55
metres.
The ground freezing was performed by use of a
brine coolant circulating through a series of vertical
freeze-pipes installed at 1.2-m centers around the shaft
perimeter. The coolant circuit included a brine chiller,
down freeze pipes and two manifolds. The portions of
the shafts located within soil above the groundwater
table or in the weathered rock below the soil, were not
frozen, but were supported with steel liner plate and
steel ribs. Shaft excavations below the top of sound
rock were supported by a combination of rock dowels
and welded wire fabric. Excavation of the soft core
was completed using a conventional backhoe.
Mucking was completed utilizing a crane to hoist a
skip box. Ground freezing proved effective in
providing temporary support while excavating deep
shafts in wet sandy soils. Figure 4 illustrates the soil
excavated from the shaft, after freezing.

176

Figure 4. Excavated Soil From Frozen Shaft.

3.2. Exploratory Tunnel Below Buried Valley
Exploratory boreholes, including tomography survey,
suggested that unfavorable geological features along
the main tunnel could be encountered while tunneling
beneath the buried valley. Tunneling below the valley
presented risks that could not be mitigated effectively
from the ground surface. Furthermore, encountering
an incised part of the buried valley in the tunnel
horizon could result in mining difficulties, claims and
an expensive remediation program. This prompted the
project team to launch the exploratory tunnel program,
to define and evaluate the potential risks associated
with tunneling underneath the valley. A layout of the
3-m diameter exploratory tunnel is shown in Figure 5.

Figure 5. Exploratory Tunnel Arrangement and
Configuration.

One of the questions related to the layout of the
exploratory tunnel was its location in relation to the
main tunnel. Initially, construction of a side-drift
gallery was considered beneficial because it would
provide the means to explore rock, provide a platform
for grouting, facilitate drainage and reduce
groundwater pressure on the crown of the main
tunnel. In the final analysis, it was determined that a
centrally located exploratory tunnel would be the most
beneficial alternative. This arrangement provides
direct evidence of ground conditions in the domain of
the main tunnel. A concentrically located exploratory
tunnel was selected to minimize complications when
overboring the main tunnel. The most significant
findings of the exploratory tunnel investigations were
that the rock cover above the main tunnel crown
consisted of good quality rock and that no evidence of
a buried valley protrusion existed within the main
tunnel domain. Although no adverse geological
condition was encountered, the decision to construct
the exploratory tunnel was correct considering the
potential risks identified initially.

3.3. Gas in Tunnel
Encounters with natural gas are not uncommon in
Chagrin Shale and have occurred on previous
NEORSD projects. The GBR stipulated that natural
combustible gases (primarily methane) and poisonous
gases (such as hydrogen sulfide), under pressure, are
to be anticipated in the shafts and tunnels of the Mill
Creek project, classifying the tunnel as potentially
gassy. Methane gas was encountered while the
Contractor was drilling a down hole at the Mill Creek,
Phase 3 Tunnel. The gas was permitted to completely
dissipate. After 757 m of the tunnel drive had been
completed, several large quantities of gas entered the
tunnel from behind the TBM in the general vicinity of
the aforementioned down hole. This situation was
especially dangerous as the TBM gas monitoring
equipment was designed to detect gas near the tunnel
heading. As the frequency and volume of the gas
incursions increased, the decision was made to
suspend mining operations to safely address the gas
issue. Steps taken to mitigate the gas conditions
consisted of drilling de-gassing wells, installing a gas
monitoring system and constructing an additional 5.5-
m diameter vent shaft.
4 MAIN TUNNEL DRIVE SUMMARY
4.1. General
It is common knowledge that the ground conditions
have direct bearing on the methods of tunnel design
and construction. More specifically, they govern the
selection of the excavation method and the type of
temporary and permanent tunnel lining. In other
words, all major aspects of the tunnel work.
Based on ground conditions (closely bedded
shale), a two-pass tunneling method was designed by
the Engineer as best suited for this project.

Fiberglass rock dowels
Buried Valley
Chagrin Shale
7.3-m. dia. main tunnel
3-mdia. exploratory tunnel
177
4.2. Primary Support
A primary support system (first pass) was installed
concurrently with tunnel excavation. It consisted of
Grade 50, W6 x 20 expanded circular steel ribs at 1.2-
m centers and 150-mm thick timber lagging spaced at
a maximum of 600-mm along the tunnel perimeter. Of
all alternatives evaluated this support system was
determined to have the greatest probability of
successfully achieving desired performance
requirements. Furthermore, this lining system takes
advantage of existing experience held by the local
labor force. View of primary support is illustrated in
Figure 6.

Figure 6. Close-up of Primary Support Behind TBM.

4.3. Mining
An open face (7.3 m diameter), Robbins type machine
was used to excavate the tunnels. The primary
requirement for the TBM was its suitability to
negotiate through thinly bedded, closely jointed rock
of variable strength. Primary support was installed
within the fingershield, located immediately behind
the primary TBM shield. Ribs were initially
expanded by rib-erector system and then jacked into
the final position from the TBM platform. Typically
the excavation sequence consisted of tunnel boring
(advancing in 1.2-m. increments), rib-lagging
installation and continuous mucking via conveyor
system.

4.4. Final Lining
Design of final lining for the Mill Creek tunnel
(second pass) was selected in accordance with the
requirement for permanent tunnel support,
groundwater control and hydraulics. The final lining
consisted of cast-in-place reinforced concrete, 300-
mm thick. Because the tunnel will experience internal
hydrostatic pressures during storage, the use of steel
reinforcement in the tunnel liner was considered
beneficial. A view of tunnel liner is shown in Figure
7.

Figure 7. View of Cast-in-Place Concrete Liner MCT-2
Tunnel.

CONCLUSION
As in many underground projects, geology played a
significant role in the design and construction
methodology employed for the Mill Creek tunnels.
Firstly, because of the existence of a buried valley at
the site, two deep shafts were designed offering a
variety of construction options. Ultimately, ground
freezing was considered as the best option and proved
to be successful in this case. Secondly, the geology at
the site made it prudent to construct an exploratory
tunnel, which provided the design team with valuable
data well in advance of the main tunnel drive. Thirdly,
the gas related issue provided a valuable experience
that could be of some benefit to future tunnel
designers and constructors in the Cleveland area.
ACKNOWLEGEMENTS
The authors would like to thank Northeast Ohio
Regional Sewer District, specifically Charles Vasulka,
Director of Engineering, for his review and approval
to publish this paper. Special thanks go to Carol
Chavis for managing the paper design and production.
REFERENCES
1. Lukajic, B., R. Pintabona, M. Kritzer, and R. Switalski.
2003. Ground Freezing for Deep Shafts at the Mill
Creek Tunnel Project. Rapid Excavation and Tunneling
Conference, New Orleans, LA.
2. Schafer, M., R. Pintabona, B. Lukajic, M. Kritzer, T.
Shively, and R. Switalski. 2004. Rock Tunneling at
the Mill Creek Project. North American Tunneling
Conference. Atlanta, GA.
3. Pintabona, R., M. Schafer, B. Lukajic, M. Kritzer, T.
Shively, and R. Switalski. Exploratory Tunnel at the
Mill Creek Project. 2006. North American Tunneling
Conference, Chicago, IL.
178
1 INTRODUCTION
Heavy rainfall has caused flooding in the central part
of Bangkok and has been a persistent problem to
resolve for the Bangkok Metropolitan Administration
(BMA), Department of Drainage and Sewerage. BMA
established a $56 million solution for the existing
flood problem with construction of a water diversion
tunnel to improve capacity of the current drainage
system. The tunnel collects water from four inlet
structures at Makkasan, Sansab, Pai Singto, and Chua
Plueng, and pumps it into the Chao Phraya River at
Chong Lom pump station. The tunnel will improve
drainage efficiency in ten districts of Bangkok,
namely Bangrak, Dindang, Huay Kwang, Khlong
Toey, Pathumwan, Phayathai, Rajthevee, Sathorn,
Wattana, and Yannawa.
The main objective of this paper is to investigate
response of pore pressure around the BMA flood
protection tunnel during tunneling, using field
observation data. Two instrumented sections were
installed along the tunnel alignment in order to
capture responses of both ground movements and pore
water pressure during tunneling and after passage of
the tunneling shield. Surface settlement points, deep
settlement points, and combined inclinometer/magnet
extensometers were put in to monitor ground
movements, whereas pneumatic piezometers allowed
observations of pore pressure changes. In actuality,
most of the installed geotechnical instruments
malfunctioned as a result of rather poor quality of
instrument installation. Installation of the pneumatic
piezometers followed the technique suggested by [1].
The piezometers performed reliably and therefore
only piezometric data are reported and discussed in
this paper.
2 PROJ ECT DESCRIPTION
Ch. Karnchang Public Co. Ltd. is the contractor
building the flood protection tunnel and the four inlet
stations at Makkasan, Sansab, Pai Singto, and Chua
Plueng. Thai Engineering Consultants Co. Ltd.
represents BMA to supervise the project.
The tunnel is mined 5.31 m in diameter with an
earth pressure balance machine (EPBM) for a
horizontal distance of 6.2 km at an average depth of
30 m. The EPBM was launched from a 15-m-diameter
Responses of Pore Water Pressures during EPB Shield Tunneling in
Bangkok Subsoil
Mongkol Sunnananda
Thai Engineering Consultants Company Limited, Bangkok, Thailand
Prapon Chanpradappha
Ch. Karnchang Public Company Limited, Bangkok, Thailand
Kitti Akewanlop
Thai Engineering Consultants Company Limited, Bangkok, Thailand
Tanate Srisirirojanakorn
Department of Civil Engineering, Chulalongkorn University, Bangkok, Thailand
ABSTRACT: A flood protection tunnel of the Bangkok Metropolitan Administration is under construction to eliminate a
flood problem in many districts of Bangkok. An earth pressure balance machine of 5.31-m diameter is employed in
excavation of the tunnel through ground conditions varying from dense sand to stiff clay. The tunnel is approximately 6.2
km long with a 4.6-m inside diameter and a springline depth ranging from 25 to 30 m below ground surface. Tunnel lining
support features five main pieces and one key piece of precast segmented reinforced concrete that is 1.2 m long and 0.275
m thick. The tail void behind the lining is grouted with a mix of water, cement, bentonite, and sodium silicate. A field
instrumentation program was initiated to evaluate tunneling process and performance of the constructed tunnel. This paper
presents and discusses field monitoring results during tunneling at two test sections of pneumatic piezometers installed at
various distances from the tunnel.

179
pump shaft at Chong Lom. It travels northward
beneath the railway connecting with a 12.9-m-
diameter inlet shaft at Pai Songto and an 8-m-diameter
inlet shaft at Sansab, and finishes at a 15-m-diameter
inlet shaft at Makkasan, as displayed in Figure 1.
The tunnel slopes uphill from the Chong Lom
shaft to the Makkasan shaft, at 0.011 percent for a
horizontal distance of about 2.5 km, 0.342 percent for
a horizontal distance of about 2.4 km, and 0.010
percent for a horizontal distance of about 1.3 km.
2.1. Tunneling Procedure
The flood protection tunnel is excavated with an
EPBM and lined circumferentially with precast
concrete segments. The segmented tunnel lining
serves as the primary and secondary lining. It is made
of reinforced concrete and 1.2 m long with an outer
diameter of 5.15 m and an inner diameter of 4.6 m.
Each completed lining ring comprises five main
segments and one key segment, and weighs
approximately 130 kN. All the segments and the key
are assembled piece by piece within the tail of the
machine with a segment erector and secured together
with steel curved bolts. All the rings are fastened
together with the curved bolts as well. All bolt pockets
on the lining are later filled with a padding mortar,
flush with the inner surface of the tunnel. A hydro-
swelling seal is used to ensure the watertightness at all
circumferential and longitudinal joints of the lining.
The machine moves forward by a propulsion
system pushing against a previously installed ring. As
the machine advances, the annular space behind the
lining is grouted with a mix of water, cement, and
bentonite, with addition of sodium silicate for faster
hardening process.
The construction of the flood protection tunnel
began on October 31, 2005. It employs two working
shifts per day, twelve hours per shift from 7am to 7pm
(day shift) and from 7pm to 7am (night shift),
Mondays to Saturdays. Sundays are normally spent on
performing preventive maintenance and repair of
construction equipment. The total length of the tunnel
drive requires installation of approximately 5090
tunnel rings. As of May 16, 2006, 1264 rings have
been installed. Each 1.2-m shove was finished on the
average of 71 and 39 minutes at test sections 1 and 2,
respectively. On average, assembly of each tunnel
ring was completed in 39 and 32 minutes at test
sections 1 and 2, respectively.

Fig. 1. Project location and layout of tunnel alignment.
Bangkok
180
2.2. Earth Pressure Balance Machine
The EPBM is 7.74 m long. It is equipped with
propulsion and ground control systems and a grouting
system manufactured by Mitsubishi Heavy Industries,
Ltd., J apan, as shown in Figures 2 and 3. The
propulsion system contains eighteen hydraulic
cylinders with a total thrust capacity of 27000 kN.
Steering of the machine relies primarily on an
articulation system that features sixteen jacks, each
with a thrust capacity of 1500 kN.
The ground control system involves regulating
earth pressure inside the cutter chamber by adjusting
the discharging rate of excavated muck from the
screw conveyor. For the grouting system, injection
can be carried out at a maximum pressure of 1.2 MPa
and at a maximum rate of 0.2 m
3
/min.

Fig. 2. 5.31-m-diameter Earth Pressure Balance Machine.

Fig. 3. General assembly of EPBM.
3 GROUND CONDITIONS
In general, soil at the tunnel level varies from dense
sand to hard clay. Subsurface conditions at test
sections 1 and 2 are strikingly similar. Illustrated in
Figure 4 is a representative soil profile and Table 1
summarizes average properties of soil layers
encountered at the two test sections.

D
e
p
t
h

b
e
l
o
w

g
r
o
u
n
d

s
u
r
f
a
c
e
,

m
0
10
20
30
40
Fill
Very soft to Soft clay
dark gray, trace fine sand and shell
Mediumclay, dark gray, trace fine sand and shell
Stiff to Very stiff silty clay
light grayish brown, trace fine sand
Very stiff to Hard silty clay
grayish brown, trace fine sand
Dense to Very dense silty sand
light grayish brown
5.31-m
BMA Tunnel

Fig. 4. Representative soil profile at test sections 1 and 2.

Table 1. Average geotechnical properties of soil layers at
test sections 1 and 2
Soil
layer
w
o
,
%
w
l
,
%
I
p
,
%
,
t/m
3

N,
blows
/ft
s
uo
(UC)
t/m
2

Very
soft to
soft clay
65.3
11.0
69.3
17.0
41.8
12.2
1.64
0.07
- 1.7
0.6
Medium
clay
52.7
3.9
- - 1.73
0.02
- 3.7
0.9
Stiff to
very
stiff
clay
26.9
3.8
65.6
16.2
40.0
10.5
1.98
0.09
16.6
4.3
9.4
2.0
Very
stiff to
hard
clay
22.3
3.1
60.8
16.2
37.3
10.7
2.04
0.07
28.5
9.1
15.8
3.8
Dense
to very
dense
sand
- - - - 61.4
13.4
-
181
Note: w
o
=natural water content, w
l
=liquid limit, I
p
=
plasticity index, =total unit weight, s
uo
(UC) =undrained
shear strength from unconfined compression test. Values in
italic represent standard deviations.
4 INSTRUMENTATION AND MONITORING
PROGRAM
Two test sections of pneumatic piezometers, as shown
in Figure 1, were installed along the tunnel alignment
to monitor changes of pore water pressures at various
stages of tunneling with higher frequency as the shield
passed the piezometer arrays and with lower
frequency after the passage of the shield. One
pneumatic piezometer (P1) was installed in test
section 1. Test section 1 was located at Sta. 0+853, as
early as the tunneling shield completed the first curve
and entered the straight drive. Eight pneumatic
piezometers (P2-1 to P2-8) were utilized at test
section 2, which was at Sta. 1+446. Figure 5 is a
layout of the piezometers installed at test sections 1
and 2.
Plotted in Figure 6 are responses of pore water
pressures as the EPBM advanced at test sections 1 and
2. Changes in pore pressure above the tunnel crown
(i.e. P2-1 to P2-4) were minimal, probably because of
minimal yielding of the ground. The tunneling shield
influenced piezometric levels along the tunnel
springline (i.e. P1 and P2-5 to P2-8) as it approached
within a distance of 3 m before the test sections. The
pore pressures continued to increase as the shield
advanced, and maximized at about 10 m behind the
shield face where grout was injected behind the lining.
Pore pressures responded less dramatically with the
construction activity as the distance from the tunnel
increased. Drop of pore pressures due to ground
relaxation was expected immediately after the
tailpiece cleared the piezometers and before grouting
took place. However, field data showed no
piezometric drop and this is likely because of the lack
of piezometric measurements between 8 to 10 m
behind the shield face. In general, pore water
pressures stabilized after the shield had passed the test
sections for a distance of approximately 45 m, 1.5
times the springline depth.
Pore pressures along the tunnel springline at the
two test sections are combined to show their
distribution at different stages of construction in
Figure 7. About 10 days after the stabilization of pore
pressures, pore pressures decreased indicating the
ongoing process of ground consolidation.

-10 0 10
-40
-30
-20
-10
0
D
5.31 m
3.4 m
fromtunnel centerline
29.36 m
-10 0 10
D
e
p
t
h

b
e
l
o
w

g
r
o
u
n
d

s
u
r
f
a
c
e
,

m
-40
-30
-20
-10
0
18.5 m
below ground
surface
D
5.31 m
4.4 m
fromtunnel centerline
5.9 m9.6 m
10.6 m
21.8 m
23.8 m
25.8 m
29.11 m
Fill
Very soft to Soft
clay
Mediumclay
Stiff to Very stiff
silty clay
Very stiff to Hard
silty clay
Dense to Very dense
silty sand
Test Section 1
(Sta. 0+853)
Test Section 2
(Sta. 1+446)
Distance from
tunnel centerline, m
-5
0
5
Direction of
tunneling
Plan View
H
o
r
i
z
o
n
t
a
l
d
i
s
t
a
n
c
e
,

m
-5
0
5
Direction of
tunneling
Plan View
P1
P2-1
P2-2
P2-3
P2-4
P2-5
P2-6
P2-7 P2-8

Fig. 5. Layout of piezometers at test sections 1 and 2.

Distance to shield face, m
-50 -40 -30 -20 -10 0 10 20 30 40 50
P
o
r
e

w
a
t
e
r

p
r
e
s
s
u
r
e
,

k
P
a
0
50
100
150
200
250
300
350
400
450
P2-1
P2-2
P2-5
P2-4
P2-6
P2-7
P2-3
P2-1
P2-2
P2-5
P2-4
P2-6
P2-7
P2-3
7.74 m
5.31 m
T
A
I
L
F
A
C
E
P2-8
P2-8
P1
P1
Test Section 1 Test Section 2

Fig. 6. Piezometric response with face advance at test sections 1 and 2.

Distance fromtunnel centerline, m
3 6 9 12
E
x
c
e
s
s

p
o
r
e

w
a
t
e
r

p
r
e
s
s
u
r
e
,

k
P
a
0
50
100
150
200
250
300
1 - at shield face
2 - after grouting (10 mbehind shield face)
3 - 50 mbehind shield face
4 - 12 days after shield passage
Springline piezometers
1
2
4
3
P1
D =5.31 m
P2-5 P2-6 P2-7 P2-8
?
?
?
?
?
?
?

Fig. 7. Distribution of pore water pressures along tunnel srpingline at various stages of construction at test sections 1 and 2.

5 CONCLUSIONS
The field instrumentation and monitoring program of
the BMA flood protection tunnel project exemplified
the importance of quality of instrument installation. In
addition, it has proved that, with proper installation
technique, pneumatic piezometers can be utilized
successfully in Bangkok subsoil. Piezometric data
suggested that the piezometers in the two test sections
were functioning well. Undrained response of pore
water pressures was observed during tunneling,
whereas pore pressures dropped over the long term
suggesting ground consolidation in progress.
REFERENCE
1. Srisirirojanakorn, T. 2004. Pore pressure response and
ground displacements in Chicago clay during tunneling
and over long term. Ph.D. Thesis in Civil Engineering,
University of Illinois, Urbana, Illinois, 487 pp.
1 INTRODUCTION
The expanding population of the Greater Toronto
Area (GTA) and in particular, anticipated population
growth within the Regional Municipalities of Halton,
Peel, York and Durham as indicated on Figure 1,
require a series of significant expansions of the trunk
sanitary sewer network over that initially constructed
between about the late 60s and the 80s.

Fig. 1. General location plan.

A major part of these undertakings is a proposed 25
year plan that was initiated by the Regional
Municipality of York (York Region) in 1997 to
essentially twin and expand the existing York Durham
Sanitary System (YDSS) located within its boundaries
as indicated on Figure 2. In this respect, it should be
noted that the original YDSS system within York
Region was constructed by the province and assumed
by York Region in the 1990s.

Fig. 2. York Region 1997 YDSS system.

However, during the three year period that was
required to complete the construction of the first of
the York Region expansion projects, i.e., the Ninth
Line and 16
th
Avenue Phase 1 trunk sewers in
Markham as presented on Figure 3, a profound change
in the approach to the design, permitting and
construction methodology for similar projects within
York Region and indeed, the entire GTA has
occurred. Specifically, the combination of soil and
groundwater
The Changing Face of Tunnelling in Greater Toronto
Ivan Corbett, M.Sc., P.Eng.
GeoTerre Limited, 215 Advance Blvd., Unit 5/6, Brampton, Ontario, L6T 4V9, Canada
ABSTRACT: The expanding population of the Greater Toronto Area, and in particular, anticipated population growth within the
Regional Municipality of York (York Region) located directly north of Toronto has resulted in York Region embarking on a
significant expansion of their sanitary sewer network since 1997 beyond that initially constructed in the late 60s and 70s. Based
on experience gained with some 25 km of deep sewer tunnels located within York Region where the author has acted as
geotechnical project manager, this paper presents a summary of the geologic and topographic settings of the latest series of
projects and associated experiences with these major sewer tunnel projects that, taken in combination, have resulted in a profound
change to tunnelling within York Region and the GTA. Specifically, this paper details the experiences of York Region during the
completion of the first of its proposed sewer expansion works and how those experiences, that can be directly related to the
geologic setting of the project, have resulted in a transformation from the traditional GTA tunnelling approach of using open face
Tunnel Boring Machines (TBM) with rib and lagging primary support to a more contemporary EPB TBM in conjunction with pre-
cast segmental liners for the more recent sewer tunnel projects.

182

Fig. 3. Plan of Ninth Line and 16
th
Ave trunk sewers.
conditions along the proposed new sewer alignments,
in combination with increased environmental
awareness of both the public and the various
regulatory agencies, has resulted in a change from the
traditional GTA approach to tunnelling of using an
open face Tunnel Boring Machine (TBM) in
conjunction with a two pass tunnel liner system (rib
and lagging plus cast in place concrete) and associated
dewatering of any cohesionless deposits, to an
essentially dewaterless approach to tunnelling as
can be obtained using an Earth Pressure Balance
(EPB) and associated single pass, pre-cast segmental
liner.

Based on experience gained with some 25 km of deep
sewer tunnels located within York Region where the
author has acted as geotechnical project manager, this
paper presents a summary of the geologic and
topographic settings of the latest series of projects and
associated experiences with these major sewer tunnel
projects that, taken in combination, have resulted in
the aforementioned profound change to tunnelling
within York Region and the GTA.

2 GEOLOGIC/HYDROGEOLOGIC SETTING
First and foremost in the change in tunnelling
approach over that adopted for most of the initial
York Region trunk sewer development, is a
fundamental change in the geologic setting of the
latest series of trunk sewer expansion works.
Specifically, overburden soils of the GTA, were
deposited during the Quaternary period and consist of
a variety of glacial till, glacio-lacustrine and glacio-
fluvial sand, silt and clay. Bedrock within most of the
southern portions of York Region consists of shale of
Upper Ordovician age located at least 30 m below the
existing ground surface within the south limits of
York Region and increasing in depth toward the north.

The Quaternary deposits were laid down by two
successive glacial periods (Illinoian and Wisconsinon)
and an interglacial warmer period (Sangamonian).
The major Illinoian deposits starting with the oldest
and extending upward are the Scarborough Formation,
the Sunnybrook Till, which is in turn overlain by the
Thorncliffe Formation of gravel, sand and silt that is
associated with the Sangamonian interglacial period.

The Wisconsinon glacial period was initiated by ice
advance out of the north during which time a fairly
continuous layer of basal till known as the Newmarket
Till was deposited burying the older deposits
(Thorncliffe Formation and below that, deposits of the
Sunnybrook Till and the Scarborough Formations).
At one time, this initial ice advance included
uninterrupted ice over most of southern Ontario.
However, with time an east-west trending split
occurred between the northern ice sheet and the ice
sheet of the Lake Ontario basin to the south. The
resulting deposition within this split created the Oak
Ridges Moraine, which is a high ridge of land that
extends from the Niagara escarpment eastwards for
approximately 160 km as indicated on Figure 4.
Overburden depths as great as 200 m have been
recorded along the Oak Ridges Moraine [1].

As the Ontario Ice Lobe melted back, a combination
of deep water ice marginal glacio-lacustrine and
glacial outwash sediments were deposited on top of
the Newmarket Till. However, the final stage of the
Wisconsinon glacial period was characterized by a re-
advance of the Lake Ontario ice sheet that overrode,
and in some instances totally removed, the interstadial
deposits and deposited a capping layer of glacial till
referred to as the Halton Till over most of the GTA,
although importantly, the final advance did not reach
the crest of the previously deposited Oak Ridges
Moraine.
183

Fig. 4. Outline of Oak Ridges Moraine.

The interstadial deposits between the Halton Till and
the Newmarket Till are a known, sometimes
discontinuous source of groundwater and are referred
to as part of the Oak Ridges Moraine Aquifer
Complex (ORMAC).

Above the upper Halton Till, generally thin deposits
of lacustrine sand, silt and clay that were deposited in
lakes formed along the face of the receding Lake
Ontario ice front occur at the surface.

The net result of this deposition sequence is
substantial inter-fingering and inter-layering of
deposits along the fringes to the Oak Ridges Moraine
complex, especially in a perpendicular, north/south
direction as indicated on Figure 5. In addition,
downcut river valleys that were ultimately infilled
with predominantly cohesionless granular materials
are also known to have developed within the
Newmarket Till and ORMAC.

Fig. 5. General soil profile perpendicular to the Oak Ridges
Moraine (after [2]).

The Oak Ridges Moraine is a significant
hydrogeologic feature in Southern Ontario and its
high ground forms a regional groundwater divide
between Lake Ontario to the south and Lakes Scugog
and Simcoe to the north. The Oak Ridges Moraine is
an important groundwater recharge area.

The discontinuous granular deposits of the ORMAC
that are sandwiched between the upper Halton Till and
the lower Newmarket Till and the underlying granular
deposits of the Thorncliffe Formation located beneath
the Newmarket Till are both connected to the Oak
Ridges Moraine and act as aquifers that sub-crop
within the lower terrain lands within the more
southerly limits of York Region. The ORMAC varies
in thickness and grades from sand to silts and clays.
The Thorncliffe Formation is generally thicker, more
uniform and coarser grained.

Groundwater flow in both aquifer units is generally
toward the south and Lake Ontario. Recharge to both
the ORMAC and the Thorncliffe Formation is
stronger in the upgradient areas to the north within the
Oak Ridges Moraine. However, considerable
recharge is also realized by infiltration through the
overlying glacial till south of the moraine area.

Significant yields of good quality drinking water are
usually available from the Thorncliffe Formation,
with discontinuous and sporadic water yields available
from the ORMAC.

The initial series of York Region trunk sewer
development was located within the surficial till
deposits to the south of the Oak Ridges Moraine
whereas the latest series of York Region sewer
expansion work generally extend into and/or are
located within the more southerly limits of the
elevated terrain associated with the Oak Ridges
Moraine as indicated on Figure 5. Importantly, this
resulted in sewer profiles during the initial phase of the
York Region trunk sewer development that were
located almost entirely in the near surface cohesive
Halton Till materials where stable tunnel face
conditions generally prevailed, whereas the later series
of York Region sewer expansions are located within a
geologic setting more prone to encountering
waterbearing cohesionless deposits under high water
pressures and associated unstable tunnel face
conditions.

South
Slope
Peel
Plain
184
3 PHYSIOGRAPHIC SETTING
In keeping with the fundamental change in the
geologic setting of the latest series of sewer expansion
works within York Region, a similar change has also
occurred in the physiographic and topographic setting
of this latest series of York Region sewer expansion
works. Specifically, based on information presented in
[1], the physiography of the central portions of York
Region (and Peel and Durham Regions) is
characterized by two basic Physiographic Regions as
indicated on Figure 5, i.e., the Peel Plain and the South
Slope of the Oak Ridges Moraine. The Peel Plain
refers to a relatively flat tract of land with an overall
gentle slope to the south that extends from the toe of
the south slope of the Oak Ridges Moraine complex to
just north of the Lake Iroquois shoreline. The South
Slope refers to an area of inclined land that rises from
the relatively flat lying Peel Plain to the Oak Ridges
Moraine in the north. A key feature of the South Slope
physiographic region is a number of deeply incised,
south flowing river valleys.

Most of the original York Region trunk sewer
development works were located within the Peel Plain
physiographic region whereas most of the latest sewer
expansion works are located in the South Slope
physiographic region. Importantly, the flat lying terrain
within the Peel Plain physiographic region resulted in
relatively shallow tunnel depths, whereas the more
elevated terrain within the South Slope physiographic
region and in particular, the frequent presence of deeply
incised south flowing river valleys, have resulted in
significantly deeper tunnel alignments than previously
undertaken. The impact of the south flowing river
valleys on required tunnel profiles is most dramatically
indicated by the alignment of the Ninth Line and 16
th

Avenue trunk sewer projects as presented on Figure 6
where tunnel depths of up to 50 m below existing
grades were required to maintain a gravity based sewer
below the Bruce Creek river bed, located some 10 km
upstream from the tunnel downstream terminal outlet.

4 ENVIRONMENTAL AWARENESS
In keeping with an overall trend of increased
environmental awareness over the last 25 years, the
initial Ninth Line and 16
th
Avenue Phase I sewer
projects of the proposed 1997 York Region expansion
works felt the brunt of this increasing awareness and
in particular, the increasing willingness of the general
public to voice concerns against perceived
environmental impacts. Specifically, and as indicated
by the following case histories, soil dewatering to
allow the completion of sewer tunnelling works using
a conventional open face TBM has dramatically
changed from being a routine undertaking during
completion of the initial sewer expansion works to a
major design, permitting and public relations exercise
for the latest phase of trunk sewer expansion work.

185
Fig.6. Ground surface profile along Ninth Line and 16
th
Avenue trunk sewers.
186
5 CASE HISTORIES
5.1. Ninth Line/16
th
Avenue Phase 1 Trunk Sewers
The locations of these projects that have a combined
total length 6 km relative to the Oak Ridges Moraine
are indicated on Figure 7, with detailed soil conditions
along these alignments presented on Figure 8.
Relative to the aforementioned geologic and
physiographic project settings, the following key
elements are worthy of note:
Site location just south of the Oak Ridges Moraine.
Increasing elevation along Ninth Line as the sewer
alignment climbs up the South Slope, whereas the
elevation within 16
th
Avenue Phase 1 is quite flat.
Surface capping layer of glacial till deposits.
Zone of waterbearing cohesionless materials within
the south limits of the Ninth Line sewer that are
believed to be largely inter-glacial that give way to a
second zone of ORMAC water bearing materials
with significantly higher water heads (up to 45 m
above tunnel invert) within the north limits of the
Ninth Line sewer and along the entire 16
th
Avenue
Phase 1 limits.

Fig. 7. Location of case histories relative to the Oak Ridges
Moraine.

In terms of construction, the following key elements
are worthy of note:
Projects completed by joint venture of
McNally/Aecon through a Design/Build process.
Sewers mined between December 2000 and May
2003 using an open face, 136 inch diameter Lovat
TBM machine with partial EPB capabilities
equipped with rock cutting discs in conjunction with
a two pass lining system and routine dewatering of
waterbearing deposits as required. Partial EPB
achieved by the use of a pressure relieving gate near
the top of the TBM front excavating chamber
(Project initially specified the use of a full EPB
TBM in conjunction with a single pass, pre-cast
segmental liner to avoid the need for dewatering
related to tunnelling).
Experience during tunnelling suggests that the
foregoing approach had progress and alignment
difficulties when external water pressures were
much above the springline of the tunnel.
Shafts constructed using a combination of soldier
pile and lagging and steel liner plate temporary
support systems in conjunction with dewatering.
Dewatering permit took a period of 7 months to
obtain, significantly longer than the 2 to 3 months
originally anticipated by the design-builder. This
and other delays resulted in the need for the terminal
shaft on 16
th
Avenue Phase 1 to be relocated about
100 m west of its originally proposed location and
into an area where significant shaft dewatering was
required.
Quite a number of residential well inference claims
were alleged as a result of the dewatering, some as
far as 10 km from the alignment.
Local conservation authority (Toronto Region
Conservation Authority TRCA) very concerned
about impacts to receiving surface water courses as
a result of the concentrated discharge of high
volumes of dewatering water that was too cold in
summer and too warm in winter.
Combination of residential well and surface river
impacts resulted in negative public opinion.

5.2. 16
th
Avenue Phase 2 Trunk Sewer
The location of this project with a total length of 7.5
km relative to the Oak Ridges Moraine is indicated on
Figure 7, with detailed soil conditions along this
alignment presented on Figure 9. Relative to the
aforementioned geologic and physiographic project
settings, the following key elements are worthy of
note:
Site location just south of and parallel to the Oak
Ridges Moraine complex.
Terrain along 16
th
Avenue that decreases from the
east and west toward the low point created by the
combined floodplains of the Bruce and Berczy
Creeks.
Surface capping layer of glacial till deposits with
two major interglacial cohesionless deposits
(McCowan Road Sand and Gravel and Warden
Avenue Sand deposits of Figure 8).
Almost continuous waterbearing cohesionless
ORMAC materials over the easterly 4 km of the 16
th

Avenue Phase 2 limits (Robinson Creek Buried
Sand) with initial water heads of up to 45 m above
the tunnel invert. This deposit also includes what is
believed to be an in-filled former eroded channel
directly below the West Robinson Creek that
extends through the ORMAC and into the
underlying Thorncliffe formation.
Ninth Line/16
th
Avenue Phase 1
16
th
Avenue Phase 2
Bathurst/Langstaff
Credit Valley
187

Fig.8. Soil and groundwater conditions along Ninth Line and 16
th
Avenue Phase 1 trunk sewers.
188

Fig.9. Soil and groundwater conditions along 16
th
Avenue Phase 2 trunk sewer.

Fig.10. Soil and groundwater conditions along Bathurst Collector sewer.
189

Fig.11. Soil and groundwater conditions along Langstaff Trunk sewer.

Fig.12. Soil and groundwater conditions along Credit Valley trunk sewer extension.
190
In terms of construction, the following key elements
are worthy of note:
Projects completed by a joint venture of
McNally/Aecon through a Design/Build process.
Delay in the start of tunnel mining from J uly 2003
until April 2005 (22 months) until a construction
dewatering permit could be obtained from the
Ontario Ministry of Environment (MOE). Part of
this delay resulted from a critical re-evaluation of
the jurisdictional domains of the various approving
agencies in light of the Ninth Line/16
th
Avenue
Phase 1 experience(s).
Based on consultation between York Region and the
various approving agencies, the MOE dewatering
permit included a mutually acceptable
environmental management plan totaling
approximately $20 million to address potential
impacts to the environment. Notable portions of the
environmental management plan was a commitment
by York Region to establish a peer review board for
all similar tunnel projects, extensive network of
distribution pipes to help avoid the impacts of
concentrated discharge of dewatering water into the
surface receiving streams that was too cold in
summer and too warm in winter and improved
municipal servicing and/or advance water well
replacement prior to initiating the dewatering
system to lessen impacts within the estimated cone
of drawdown.
Easterly 4 km of sewer mined between May 2005
and J une 2006 using an open face, 136 inch
diameter Lovat TBM machine with partial EPB
capabilities equipped with rock cutting discs in
conjunction with a two pass lining system and
dewatering of the Robinson Creek Buried Sand.
Tunnel construction delayed for 7 months while
mining shaft was relocated from original terminal
shaft of 16
th
Avenue Phase 1 to lessen dewatering.
Westerly 3.5 km of sewer mined between May 2005
and J une 2006 using an open face, 112 inch
diameter Lovat TBM machine with partial EPB
capabilities as previously described in conjunction
with a two pass lining system and no dewatering.
Shafts constructed using a combination of soldier
pile and lagging and steel liner plate temporary
support systems in conjunction with dewatering.

5.3. Bathurst Collector and Langstaff Trunk Sewers
The location of these projects with a combined total
length of just under 9 km relative to the prevalent
geology of the area is indicated on Figure 7, with
detailed soil conditions along these alignments
presented on Figure 10 and 11. Relative to the
aforementioned geologic and physiographic setting of
these projects, the following key elements are worthy
of note:

Site location just south of a southerly projection
(Maple Spur) of the Oak Ridges Moraine.
Increasing elevation along Bathurst Street as the
sewer alignment climbs up the South Slope, whereas
the elevation along Langstaff Road is quite flat.
Surface capping layer of glacial till deposits that
encompass the proposed south tunnel zone but
generally located well above the more northerly
tunnel zones.
Series of waterbearing cohesionless materials within
the south limits of the Bathurst Collector sewer that
are largely below the proposed tunnel invert that
give way to a series of waterbearing cohesionless
ORMAC materials that encompass the proposed
tunnel zones within the north limits of the Bathurst
Collector sewer and most of the proposed Langstaff
Trunk sewer.

In summary, conditions within the south limits of the
Bathurst Collector sewer are very favorable and very
similar to those of the original York Region sewer
system development, whereas conditions within the
north limits of the Bathurst Collector sewer and entire
Langstaff Trunk sewer are less favorable and very
similar to conditions being more routinely
encountered in the latest series of York Region sewer
expansion works.

In terms of design and construction, the following key
Contract for delivery of sewers through a
Design/Build process signed in May 2006 with the
joint venture of McNally/Aecon.
Design-build tender documents delayed from
release in spring 2003 to summer 2005 while
dewatering permitting issues were evaluated and
approving agency jurisdictional issues that arose as
part of the 16
th
Avenue Phase 2 dewatering works
were resolved.
Key component of the project specifications is the
requirement to undertake all works without
installing any active dewatering system, i.e.,
essentially dewaterless construction for shafts and
tunnels although dewatering on a contingency basis
will be permitted under special circumstances.
191
Environmental and ecological impact assessment
more extensive and time consuming than the
geotechnical investigations.
Sewers anticipated to be mined commencing in
J anuary 2007 using a 129 inch diameter, Lovat EPB
TBM in conjunction with single pass pre-cast
segmental liner.

5.4. Credit Valley Trunk Sewer Extension
The location of this 3.2 km long tunnel project relative
to the prevalent geology of the area is indicated on
Figure 7, with detailed soil conditions along the
alignment presented on Figure 12. Relative to the
aforementioned geologic and physiographic setting of
this project, the following key elements are worthy of
note:
Site location well south of the Oak Ridges Moraine
complex and within the Peel Plain physiographic
region.
Existing terrain along the tunnel alignment with a
very gentle rise to the north in keeping with the
characteristics of the Peel Plain physiographic
region, with associated moderate tunnel depths
relative to those of York Region.
Relatively thin series of overburden deposits with
an associated simple stratigraphy, i.e., some thin
surface ice marginal glacio-lacustrine deposits
overlying predominantly cohesive clayey silt
(Halton Till) with occasional water bearing sand
pockets overlying shale bedrock.

In terms of design and construction, the following key
Hydro-geological assessment of route more
extensive and time consuming than geotechnical
investigations, even though very limited water
bearing units along the alignment.
Contract for construction of projects awarded to
Technicore in J anuary 2006.
Sewers to be mined between J uly 2006 and J une
2007 using a Technicore manufactured TBM with
full EPB front chamber capabilities in conjunction
with a rib and primary temporary support system.
Shafts to be constructed using a combination of
soldier pile and lagging and steel liner plate
temporary support.

6 SUMMARY AND CONCLUSIONS
The most recent series of major trunk sewer
expansions works within York Region have
experienced a significant change in face conditions
over those encountered during the development of its
initial trunk sewer system, i.e., predominantly stable
cohesive face conditions with occasional unstable
waterbearing cohesionless zones under moderate
water pressures to more frequent and extensive
unstable waterbearing cohesionless zones under high
water pressure.

The transformation in tunnel face conditions within
the latest series of York Region trunk sewer
expansions is a direct reflection of differing geologic,
physiographic and topographic settings versus those
of the original development of the trunk sewer system.

The soil and groundwater conditions along the
alignment of the various recent York Region sewer
expansions are consistent with the geological setting
of these projects. In fact, the extensive sub-surface
investigations completed for these projects have
assisted greatly with the confirmation and refinement
of the prevalent geology within the south slope of the
Oak Ridges Moraine.

The combination of differing geology, project
duration, magnitude of dewatering and regulatory and
public environmental awareness coalesced in the first
of the most recent York Region sewer expansion
projects, i.e., Ninth Line and 16
th
Avenue Phase 1 to
create a vastly different, more challenging and more
time consuming design and permitting regime for
similar trunk sewer projects both in York Region and
within the GTA in general. Simply stated, conditions
on the Ninth Line and 16
th
Avenue Phase 1 sewer
project have coalesced into the Perfect GTA
Tunnelling Storm.

Project scheduling concerns and a need to reduce
dewatering impacts during the completion of proposed
sewer projects that are similar to those of Ninth Line
and 16
th
Avenue Phase 1, have led to the adoption of
shaft and tunnelling construction methods that can be
completed in the absence of planned dewatering
where major deposits of waterbearing cohesionless
materials are present. Hence, York Region and the
GTA in general are poised to witness their first major
trunk sewers to be constructed using an EPB Tunnel
Boring Machine in conjunction with a single pass pre-
cast segmental liner.

192
With respect to design, the recent experience with the
York Region projects suggests that future similar
major trunk sewer projects in the GTA will be
required to spend much more effort to evaluate the
sub-surface conditions along proposed alternative
alignments, including proposed shaft locations, during
the environmental assessment and route selection
phase of the project.

With respect to construction, items like shaft
construction, production of pre-cast segments,
retraining of mining personnel, tunnel progress,
management of boulders in waterbearing cohesionless
ground and overall project costs remain a work in
progress within the current dewaterless era of
tunnelling within the GTA.

With respect to cost, the current dewaterless era of
tunnelling in the GTA must focus more on total
project costs and not just tunnel construction cost,
with particular emphasis being given to required
environmental assessment, mitigation and monitoring
should major dewatering be contemplated.

REFERENCES

1. Chapman, L.J ., and K.F. Putnam 1984. The
Physiography of Southern Ontario. Ontario Geological
Survey, Special Volume 2, 170 pp. Accompanied by
Map P.2715 (coloured), scale 1:600,000.
2. Sharp, D. R., P. J . Barnett, P.A. Brennand, B. Finley,
G. Gorell, H.A.J . Russell, and P. Stacey 1997.
Surficial Geology of the Greater Toronto and Oak
Ridges Moraine Area, Southern Ontario. Geological
Survey of Canada, Open File 3062 Scale 1:200,000.

193

During the 1990s the Regional Municipality of York
experienced the highest growth rate in the Greater
Toronto Area (GTA). By 2026, the population of the
Region is expected to reach 1,280,000. Currently,
94% of the Region's existing population relies on
sanitary services provided through the York Durham
Sewage System (YDSS). Planned population growth
and industry are expected to generate sewage flows
and water servicing beyond the existing capacity of
the YDSS.

1 YORK REGIONS VISION 2026
Creating Strong, Caring and Safe Communities.

Vision 2026 is an overall blueprint for York Region.
We will remain leaders in customer service. We will
support our businesses. We will create new
partnerships. York Region will also protect the legacy
of our natural environment for future generations."[1]

Highlights from four of the eight goals specific to
building water and wastewater infrastructure in
support of smart growth are listed below.

i. Enhanced Environment, Heritage and Culture
Securing a Green York Region
Ensuring Clean Water and Air
Promoting Conservation

ii. Managed and Balanced Growth
Promote a sustainable natural environment
Balancing growth with the environment by taking
a leadership role in environmental strategies and
conservation
Working and partnering with federal and
provincial governments, and the private sector to
develop innovative funding methods

York Durham Sewage System 19
th
Avenue/Leslie Street Interceptor
Sewer Project Richmond Hill, Ontario, Canada
Adrian Coombs, P.Eng., Senior Project Manager
Transportation & Works Department, Water & Wastewater Branch, The Regional Municipality of York,
Newmarket, ON, Canada
Derek Zoldy, P.Eng., Senior Project Manager, Tunnelling Specialist
Urban Infrastructure Group, Earth Tech Canada Inc. Toronto, ON, Canada
ABSTRACT: This paper will discuss the unique aspects of the regulatory framework for this sewer project included
within the Class EA, additional conditions imposed by the Minister of the Environment, Comparison Requirements for
Alternative Routes, Compliance with Regulatory Agencies, Project Approvals Processes, Conditions for Permit To Take
Water and Construction Mitigation and Monitoring Programs.
The York Durham Sewer System (YDSS) experiences heavy extraneous flows during wet weather events. Specific to the
19
th
Avenue Interceptor Sewer Project, if capacity in the Richmond Hill portion of the YDSS is not increased, then sewer
surcharging, under a severe event, is likely to occur in parts of the Richmond Hill system. In 1997, York Region
completed a Master Plan Study for the YDSS. This Master Plan reviewed existing conditions and future alternatives
necessary to service planned population growth in the Region and further identified several priority and strategic projects to
be implemented. In March of 2002, the YDSS Master Plan was updated to reflect current planning forecasts and to
confirm the timing of infrastructure projects to be completed in York Region. The Minister of the Environment, in
response to a request to have this and other YDSS Class EA projects subject to a Part II Order set out specific conditions
for this project. The intention of the Ministers conditions was to ensure that the environment is protected and the
environmental concerns which have been raised are addressed. Some conditions will be applicable over the duration of
the project and the Regions activities and progress on meeting these conditions will require annual reporting prior to,
during and following construction of the new sewer.
194
iii. Infrastructure for a Growing Region

iv. Engaged Communities and a Responsive Region
Meaningful consultation to develop effective
solutions together
Fiscal responsibility
2 HISTORY OF THE YDSS
2.1. Beginnings of Sustainable Sewage Treatment

The YDSS is a wastewater collection system within
the Great Lakes basin. It was constructed by the
Province of Ontario in the late 1970s and early 1980s
in response to a 1965 decision that no additional
sewage treatment plants could be built on the Humber,
Don or Rouge Rivers. The concern at the time was
that the assimilative capacity of receiving streams
could be exceeded by continued local service. The
project need was hastened through the execution of
the Great Lakes Water Quality Agreement between
the Governments of Canada and the United States of
America in 1972, and subsequent amendments. It
fulfills some of the obligations imposed upon the
Province of Ontario through the Canada-Ontario
Agreement Respecting the Great Lakes Basin
Ecosystem. What evolved from these initiatives was
the most environmentally respectful system in the
entire Great Lakes basin. [2]

Because of the regulatory need to implement the
concept for a sustainable YDSS, the initiation of the
system was exempted from the full regulatory process
of the day.

2.2. Existing YDSS

Figure 1 shows the existing YDSS trunk system and
highlights the alignment for this Interceptor. Routed
to relieve a capacity bottleneck to the south on Yonge
Street, virtually all sewage will be passed to a more
easterly north/south sewer capable of conveying the
flow.

2.3. Project Setting

The natural environmental, hydrogeological, and
geomorphological settings as well as the current
regulatory climate are collectively key to
understanding of the project baseline and the potential
impacts of the project. For example:

Within the regulation framework infrastructure, siting
is regulated by the Provincial Places to Grow Planning
Initiative; Greenbelt Plan for the Greater Golden
Horseshoe; Oak Ridges Moraine Conservation Act;
York Region 1997 Master Plan Study for the YDSS;
and the March 2002 Update to the Master Plan.

The ecological setting is comprised of:

Oak Ridges Moraine sand deposits occur beneath
portions of the route. The Oak Ridges Aquifer
Complex (ORAC) occurs within these deposits.
Domestic wells in this area are typically within the
ORAC.
Headwaters of the Rouge River include some
tributaries fed by groundwater discharge and
tributaries containing the sensitive indicator
species, brook trout and Redside Dace.
Several wetland units of the Rouge River
Headwater Wetland complex, itself is a
provincially significant wetland. Some of these
units are partially supported by groundwater.
The project area is capped by Halton Till. Along
portions of the route, the till sheet is thin (less than
3 m).

Baseline assessments include: geotechnical,
agricultural, archaeological, air pollution and
acoustical impact and built heritage and cultural
landscape were performed.

The project route extends along 19th Avenue for a
length of approximately 3.8 km from Yonge Street,
east to Leslie Street, and south on Leslie Street for
approximately 550 m. The YDSS Interceptor Sewer
is to be constructed by tunnelling except for a section
from Yonge Street to approximately 180 m east of
Yonge Street, which is to be constructed by open cut.
The tunnel section of the sewer will have a minimum
internal diameter of 2.1 m. The open cut section of
the sewer will have a minimum internal diameter of
1.65 m.
The 19th Avenue and Leslie Street route was selected
and construction methods evaluated to minimize
overall construction impacts. Significant efforts have
been made to identify and mandate construction
methods to limit the need to dewater; the number of
dewatering locations; the volume of dewatering; and
the duration of dewatering. Another feature of the
project is the existence of a significant length of
artesian conditions in conjunction with a limited
thickness of confining till cap.
For the purpose of the Environmental Approvals for
the 19
th
Avenue/Leslie Street Project three general
construction methods will be employed for this
project:
195
Figure 1
196
Open cut where the water table lies below the
trench invert;
Tunnelling with a pressurized-face tunnel boring
machine (Earth Pressure Balance (EPB) or Slurry
Shield technology) machine for tunnel sections
on 19th Avenue and Leslie Street; and
Sealed shaft construction.

Growth across all municipalities of the GTA is being
mandated by both the provincial and the federal
governments in order to accommodate increased
immigration and a rapidly growing population.
However, 69% of York Regions area is subject to the
Green Belt plan, which includes the Oak Ridges
Moraine. Physical growth is limited by both
provincial legislation and the Regions Official Plan.
Both of these ensure that growth will be managed and
sustainable, and will be focused through
intensification within our Regional Centres. This
growth requires in-time water and wastewater
infrastructure incorporating the most advanced level
of proven technology.

Figure 1 also defines the Oak Ridges Moraine (ORM)
and the recent Greenbelt Plan 2005 Area. Any new
infrastructure use to this trunk sewer system must and
will be subjected to extensive and exacting scrutiny
from the public, stakeholders and particularly
agencies and ministries with the aim of protecting
ground and source water resources and maintaining
the integrity of protected habitat. The challenge is to
meet the expectation of residents that a safe,
responsible and technically proven sewer system will
be available and on time to support growth.

All the while this same system must pass the rigorous
review and evaluation from Conservation Authorities
and the Ministry of the Environment (MOE). The
construction technology and methods must also pass
muster from other agencies when weighted against the
terrestrial, ecological, aquatic and source water
environments.
3 SMART GROWTH
The 5 Ss that Underpin Strong Infrastructure:
Smart- Sustainable- Sewage- Science-Solution
Smart
Smart growth considers planning initiatives that
recognize existing communities, urban centres,
linkage and traffic corridors, environmentally
sensitive features, and groundwater resource
protection.

Sustainable (Growth)
Growth which meets the needs of the present without
compromising the ability of future generations to meet
their own needs.

Some key principles York Region residents identified
that should shape future decisions and actions
included: Fiscal Responsibility through sustainable
and accountable government; leading-edge, effective,
timely services delivered with Quality; Safety;
Stewardship recognizing that it is everyone's duty to
protect the legacy of York Region's natural
environment and heritage.

Upgrading capacity through twinning (or expanding)
the YDSS is critical to ensuring adequate wastewater
treatment that keeps pace with smart growth - with
system development on a demand/supply basis.
Sewage
The YDSS converges at the jointly operated (York
Region and Durham Region) Duffin Creek Water
Pollution Control Plant (WPCP) located in Pickering,
and discharges treated wastewater into Lake Ontario.
A distinctive feature of the WPCP is the absence of
overflow and by-pass capabilities.

In 1997, the Region completed the Master Plan Study
for the YDSS. This Master Plan identified and
reviewed current conditions and future alternatives
necessary to meet population projections in the
Region.

In March 2002, the YDSS Master Plan was updated to
reflect current planning forecasts and to confirm the
timing of required infrastructure.

Science
Science supports the traditional gravity system the
question evolves from can we build it? to how can
we build it to protect the environment?

There are five pumping stations throughout the
YDSS; however, the system uses gravity sewers
wherever possible. Gravity-based systems are used in
more than 99% of municipally serviced systems.
These sewers are routinely installed within the
groundwater table. In comparison to pumped systems,
gravity sewers do not require any mechanical parts to
operate, and do not use electricity or burn fossil fuels.
Gravity sewers have a life span of more than a
century.

197
Gravity sewers do not pose any threat to groundwater
quality. The deep, larger diameter sewers that make
up the trunk system are constructed using tunnelling
techniques and have 150mm to 300mm thick concrete
walls. Sewage pipes do not leak sewage out into the
environment when designed and maintained properly.
If leakage does occur, it is from groundwater outside
of the pipe leaking into the sewer in relatively minute
quantities. This occurs because the groundwater
outside of the pipe is under greater pressure than the
free-flowing sewage inside the pipe; any water simply
follows the path of least resistance. Pipes are routinely
inspected by closed circuit TV cameras in addition to
regular monitoring of flow rates.
Solution
The solution is to adopt proven science and develop
appropriate design and construction methods that are
respectful of the environment. In this case
groundwater isolation with the use of sealed shaft
construction methods and Earth Pressure Balance
(EPB) tunnelling techniques were determined to be
the best fit for the project.
4 REGULATORY FRAMEWORK
4.1. The Ministers Conditions

In March 2003, the Schedule B Municipal Class
Environmental Assessment (EA) study for the
proposed Lower Leslie Street Trunk and the 19th
Avenue Interceptor Sewer was completed. On March
27
,
2003, a Notice of Study Completion was issued for
the Class EA. The study identified the preferred route
alignment for the trunk sewer as the west side of
Leslie Street and the south side of 19th Avenue,
primarily in traditional open cut construction.[3]

The Region was notified on October 1, 2004 that the
Minister of the Environment received a Part II Order
request on all unfinished YDSS projects. Although the
bump-up request was denied on this project, a number
of conditions were imposed by the Minister. As a
result, the scope of work for this project has
significantly increased and it was back to the design
board in late 2004 to reconsider and re-evaluate the
route, technology, construction methods and the
usual approach to groundwater pumping, all in
response to the 11 conditions imposed on the YDSS
Interceptor as set out below.

Below is a condensed summary of the 11 conditions
imposed by the Minister of Environment. The
Regional Municipality of York shall

1. develop monitoring and mitigation measures
for the dewatering activitiesaddress potential
impacts to well users and the natural
environmentassess the impacts of this project
on long-term sustainability of groundwater and
surface water resources.
2. hold a public information meeting to present
these mitigation and monitoring measures.
3. assess and address any cumulative impacts on
the environment of additional dewatering
activities occurring as a result of any other related
sewer projects.
4. maintain its well complaint review committee
process.
5. ensure that all technical studies, reports, and
other documents prepared for this proposed
project become part of the public record.
6. consult with the public on the application for
the Pemit To Take Water.
7. ensure that all technical studies be
independently peer reviewed.
8. conduct a new comparison of alternative route
alignmentsconsideration for the potential
dewatering impacts of each alternative.
9. conduct a detailed assessment of the impacts of
any dewatering activities for the preferred
alternative route alignment.
10. evaluate all reasonable design and construction
techniques for this proposed project .
11. submit an annual report to the Director,
Environmental Assessment and Approvals
Branch, Ministry of the Environment.
Figure 2 demonstrates where these 11 conditions fit
into the project milestones of EA, design and
construction. As of August 2006 only a final Public
Consultation Centre (PCC) to fulfill condition 2, and
the 2006 annual report remained outstanding.

198

Ministers Conditions 5, 8, 10
Ministers Conditions 1, 2, 3, 6, 7, 9
Ministers Conditions 4, 11
Figure 2. The Ministers Conditions
(Imbedded in the Project Phases) [3]
Ultimately the additional study and redesign of the
entire project has resulted in the completion date for
this crucial project being deferred by some 3 years.
4.2. Process Chronology

October 2004
- MOE denies Part II Order request and sets 11
conditions.

November 2004
- Public Information Forum (PIF) to notify
interested parties of additional study.

December 2004
- Recruitment of stakeholders for Interceptor Sewer
Advisory Committee (ISAC).

January 2005
- ISAC to discuss the Terms of Reference (ToR) for
the group.

February 2005
- Second meeting of the ISAC members were
provided data on environmental conditions &
preliminary alternatives.
- PIF to present Alternatives.
- York Region launched the YDSS Interceptor
Sewer Project Web site at http://ydss.cenet.ca as
part of its Constructive Engagement Program.

March 2005
- Third meeting of the ISAC evaluation methods.
- York Region held two Charrette sessions for
discussions about potential route alignments,
construction techniques and environmental
protection processes.

April 2005
- Evaluation of the alternatives.
- Fourth session of the ISAC to preview evaluation
of the alternatives in advance of the April PIF.

May 2005
- Report on YDSS Interceptor for Regional
Council.
- Fifth session of the ISAC on an additional
alternative.
- Meetings with Save The Oak Ridges Moraine
Coalition (STORM) to discuss conformance with
the Oak Ridges Moraine Conservation Plan
(ORMCP).

June 2005
- Second special meeting with STORM.
- Preferred alternative route alignment presented to
Regional Council for endorsement.
- Independent Public Facilitator released the
"Compendium of Comments and Responses"
report outlining the public consultation
undertaken.

July 2005
- Draft "YDSS Interceptor Sewer Study - New
Comparison of Alternative Route Alignments"
Class EA
Class EA
Minister s Conditions
Ministers Conditions
New Comparison of
Alternative Route Alignments
New Comparison of
Detailed Design
Detailed Design
Compliance
Compliance
ORWCA
Greenbelt
Places to Grow
TRCA
MNR (Lakes and Rivers)
OWRA
Approvals
Approvals
MOE Certificates of Approval
MOE Permit To Take Water
TRCA
MNR
Construction
Construction
Monitoring
Monitoring
Baseline Report
Baseline Report
Sewer Design and
Construction Techniques
Report
Sewer Design and
Report
Class EA
Class EA
Minister s Conditions
Ministers Conditions
New Comparison of
New Comparison of
Detailed Design
Detailed Design
Compliance
Compliance
ORWCA
Greenbelt
Places to Grow
TRCA
MNR (Lakes and Rivers)
OWRA
Approvals
Approvals
MOE Certificates of Approval
MOE Permit To Take Water
TRCA
MNR
Construction
Construction
Monitoring
Monitoring
Baseline Report
Baseline Report
Sewer Design and
Report
Sewer Design and
Report
199
posted to the web site and placed in viewing
rooms.

August 2005
- Draft "YDSS Interceptor Sewer Study - New
Comparison of Alternative Route Alignments"
submitted for Peer Review.

October 2005
- Comments from the Peer Review Team.
- Final New Comparison of Alternative Route
Alignments report was submitted to MOE,
incorporating the peer review comments.

November 2005
- Permit To Take Water (PTTW) submitted for Peer
Review.

December 2005
- Independently facilitated PCC held on the PTTW
application.

January 2006
- The YDSS First Annual Compliance Report 2005
(York Durham Sewage System Projects) prepared
by York Region addressing the annual reporting
requirements from the Minister of the
Environment.

March 2006
- PTTW Application and Environmental
Management Plan (EMP) submitted to MOE.

April 2006
- York Region awards Open-Cut portion of the
YDSS Interceptor Sewer project.

May 2006
- York Region receives acceptance of route and
methodology from MOE.
- York Region receives both Certificates of
Approval for the project.

July 2006
- York Region receives the PTTW for the YDSS
Interceptor, 19
th
Avenue/Leslie Street project.
- York Region awards Tunnelled portion of YDSS
Interceptor.

4.3. Transparency of Process

Excerpts from:
MEDIA BACKGROUNDER of October 2005

The Regional Municipality of York continues to make
every effort to be accessible, accountable and
transparent with respect to the twinning (expansion)
of the York Durham Sewage System (YDSS). This
has entailed extensive public involvement,
engagement and consultation.

Despite this environment of openness, York Region is
the subject of numerous attacks by a handful of anti-
growth activists. Many of the allegations being made
against York Region were fuelled by incomplete
information or understanding both of the concepts and
the technology and have no basis in fact.

Myths perpetuated about the YDSS

Local sewage treatment plants are preferable to one
centralized sewage system.
Sewage servicing can be provided through both local
treatment options and centralized treatment, such as
the YDSS. York Region has both types of systems.
Centralized systems such as the YDSS promote
intensification of development. When the YDSS was
constructed, effluent from over 30 treatment plants
was removed from local streams, demonstrably
improving water quality.

York Region is flaunting environmental regulations
and breaking the law.
All YDSS projects have undergone extensive
Environmental Assessments and either meet or exceed
stringent Ontario Ministry of the Environment
guidelines and regulations. York Region takes these
guidelines and regulations seriously. York Region
never constructs without the required approvals and
permits. During construction, projects are carefully
monitored and monthly reports are submitted to the
regulatory agencies.

The next three myths refer specifically to the 16
th
Avenue
Trunk Sewer project in York Region. The 16
th
Avenue Sewer
is a two pass tunnel constructed with a traditional TBM
which suffered severe controversy and intense regulatory
reappraisal and re-approval scrutiny when many private
wells dried up as a consequence of the dewatering
program. Key lessons learned were:
Projects must have extensive baseline monitoring of
sensitive environmental areas.
Listen to the public and provide a vehicle to voice
complaints, register issues, receive an appropriate
considered response-action. The mitigation and
monitoring program for the YDSS Interceptor adopted
the lessons learned from 16
th
Avenue.

Dewatering associated with YDSS construction is
causing permanent harm to the Oak Ridges Moraine
watershed.
200
Although the 16th Avenue project has required an
extensive amount of dewatering, it represents only 2
to 4% of the aquifer storage. The aquifer is
continuously replenished through natural recharge.
Once dewatering is finished, water levels will return
to normal and there will be no lasting impacts on the
watershed. Monitoring to date proves that water levels
have returned to normal where construction is
complete.

Dewatering into local streams is causing permanent
damage to fragile, local ecosystems and fish habitats.
The environmental monitoring program for the 16th
Avenue project includes extensive monitoring of
aquifer water levels, stream flows, creek levels, fish
species surveys, vegetation surveys, wetland surveys
and more. Dewatering discharge is directed into local
streams, up to the safe carrying capacity of the
streams. The balance is discharged into the YDSS.
Creeks near the 16th Avenue project are monitored for
water chemistry, fish and water levels. Discharge
water temperatures into local streams are controlled to
eliminate all adverse impacts to fish. All indications
are that fish are thriving.

Dewatering is causing wells to run dry.
On the 16th Avenue project, York Region has
committed $30 million to a comprehensive
environmental monitoring and mitigation program to
protect the environment and provide alternate water
supplies to residences that are temporarily affected.
As anticipated, dewatering has temporarily impacted
some private wells. The Region has a Proactive
Mitigation Program to ensure appropriate water
supplies are provided to affected residences.

Note: The well mitigation protocol will continue through
the Interceptor project as a proactive measure despite the
mandate to minimize dewatering with the construction
methodologies employed.

4.4. York Regions Commitments

A strong working relationship with all regulatory
agencies lynchpin a successful, effective and SMART
project.

Unique aspects of this sewer project fall under three
phases:
Class EA Additional conditions imposed by the
Minister of the Environment; Comparison
Requirements for Alternative Routes; public and
expert ranked selection of a preferred route;
unprecedented public consultation.
Design The design elements had to incorporate
minimizing groundwater pumping; a peer review
process involving subject matter experts in
geology, hydrogeology, construction techniques,
surface and groundwater movement and
protection, ecologists and biologists, aquatic
habitats and fisheries; owner application for
PTTW; compliance with regulatory agencies;
project approvals processes. [4]
Construction Redesign of the project from
mainly conventional open cut construction with
dewatering to sealed shafts and EPB tunnelling
methods without dewatering; detailed construction
mitigation, monitoring and reporting programs.

Figure 3 is a high-level graphic illustrating the
intricacies and interdependencies connecting all
stakeholders in this project through direct and indirect
feedback communication loops.

York Region has committed to the public and
commenting authorities that the project will be
designed and implemented in a manner that minimizes
the potential impacts to groundwater resources,
private wells, and natural features. These
commitments include the following:
Sewer construction methods to minimize
dewatering
Well inventory and notification of pumping to well
owners
Minimize groundwater pumping rates to ensure
compliance with the PTTW
Control of discharge to streams
A detailed monitoring program to include:
- Construction compliance monitoring
- Groundwater monitoring
- Stream monitoring
- Wetland monitoring
- Timely data analysis and response to issues
- Monitoring reports, summarizing all of the
monitoring results and discussing any issues
arising during the monitoring period, will be
submitted to the MOE and other interested
agencies every second month. Reports will
also be posted on the Regions Web site. The
Region will be available to meet with the
MOE, as required.
Mitigation Plans requiring the implementation of
additional measures, should monitoring identify
impacts on private wells, streams or wetlands.
201
Figure 3 - Project Communication
Permits & Approvals
Certificate of Approval
Agencies
Responsibility
Municipal & Private Sewage
Works CofA
PTTW
Work Permit
Lake and Rivers
Improvements
Fill Construction & Alteration
Waterways Permit
Letter of Advise
Independent
Peer
Review
Public
Ministers
11
Conditions
Final
Permits &
Approvals
Draft
Permits &
Approvals
Ministry of the Environment
MOE
Toronto and Region
Conservation Authority TRCA
Department of Fisheries and
Oceans DFO
Ministry of Natural Resources
MNR
Process
Planning
Capital Budget
Project Need
EA
Public Engagement
Meeting Ministers Conditions
Enhanced Study
MOE Approval of Route and
Design
Extensive Public Consultation,
Permits & Approvals
Peer Review
Tender
Award
Construct
Environmental Monitoring &
Reporting
Commission
Post Commission
Evaluation/Closeout
York
Region
2 CofAs
PTTW
TRCA Permit
MNR Work Permit
DFO Letter of Advice
TCPL Crossing Permit
CN Crossing Permit
Feedback Loop
Direct
Indirect
202
5 ABOVE AND BEYOND
We (York Region) take our stewardship of the
environment seriously, especially the delivery of high
quality drinking water and the effective treatment of
wastewater for all of our residents and businesses. The
YDSS was designed with the best practices in mind to
minimize any effects on the environment, now and
into the future. [1]

York Region has consulted and continues to consult
with the MOE, the Ministry of Natural Resources
(MNR), the Federal Department of Fisheries and
Oceans (DFO) and the Toronto Region Conservation
Authority (TRCA) to ensure that the expansion of the
YDSS not only meets, but exceeds the requirements
of environmental regulations and standards, and that
all necessary permits are obtained. YDSS projects also
have independent audits and/or peer reviews by
industry professionals. Moreover, York Region has
committed to a comprehensive Environmental
Monitoring and Mitigation Program to assist residents
who may be impacted due to Regional infrastructure
projects.

5.1. Meeting and Exceeding Regulatory
Requirements

First there was the imposition of 11 conditions, the
scrutiny of a redone EA route selection, design and
anticipated 21 months of construction. Next unfolds
an unprecedented monitoring and mitigation program,
during and after construction, which will set the bar
for future infrastructure projects in York Region.
York Region is seeking a positive legacy to the YDSS
Interceptor of an engaged public and a protected and
respected environment. The goal is a smart and
sustainable solution for sewer infrastructure that
harmonizes science and the environment.

5.1.1. Adaptive Management The tool that
implements the unique requirements

Adaptive management simply refers to the process of
anticipating the potential for various types of impacts,
establishing a monitoring program which allows these
to be detected quickly, and having a predetermined
response plan which sets out a protocol and a schedule
for a phased response, ensuring action in advance of
potentially negative impactsthrough a
Trigger/Response/Action protocol.
Although the project has been designed to minimize
potential impacts on domestic wells and the natural
environment by restricting groundwater taking; it is
prudent to implement a monitoring plan to ensure the
project design and associated mitigation measures
achieve this purpose.
The monitoring program requires an understanding
and analysis of what could go wrong. The emphasis
is on identification of early warning parameters
when they vary from their normal values or ranges of
values. These provide information on whether there
are potential concerns. The monitoring is divided into
three separate components: construction compliance,
groundwater and wells; streams and wetlands. The
following summarizes each component:
1. construction compliance monitoring is intended
to ensure the operational criteria established to
prevent adverse impacts are being achieved by the
contractor at all times;
2. groundwater and well monitoring is intended to
provide supplementary information to ensure the
zone of influence (ZOI) remains as predicted and
that well owners are not experiencing any
difficulties with their water supply;
3. stream and wetland monitoring is intended to
provide information to ensure there are no
developing issues which could result in impacts
on the health and integrity of area streams and
wetlands.

Specific triggers and response actions have been
developed for each component of the monitoring
program. These provide a high level of protection to
area wells and the natural environment.

Comprehensive monitoring reports are to be prepared
on a bi-monthly basis and provided to the MOE and
other agency contacts. Additionally, a brief, twice-
monthly construction progress and monitoring status
report is to be prepared and distributed.

The key to this program is its ability to feed
information back to the construction process, and
dictate any requirements for modifications in
procedures (i.e., adaptive management).

Control monitoring sites allow for additional analysis
of baseline trends and are intended to serve a due
diligence function. Local monitoring control sites
typically extend to more than twice the minimum
area of influence of 120 m established in the Oak
Ridges Moraine Conservation Plan to ensure the
protection of wetlands, fish habitat, significant
woodlands, significant wildlife habitats, streams,
seepage areas, and springs.

203
A remote monitoring limit of 250 m beyond the area
of construction has been established. This 250 m
distance:
is sufficiently large to incorporate any worst case
scenario.
robust to include any other potential impact;
incorporates most of the adjacent wells, streams
and wetlands.

The monitoring program is multi-disciplinary and
requires ongoing communication between the project
engineers, hydrogeologists, biologists, and fluvial
geomorphologists to be effective.
A specific monitoring program has been established
for tunnelling shafts and the open cut segment where
groundwater pumping may be required. [4]

5.1.2. Construction Contingency

For each construction event, response/contingency
plans have been prepared to describe the steps to
document the issue, identify alternative solutions and
evaluate the impact of these solutions on the natural
environment, stakeholder interests, project schedule
and cost.

For these plans to be effective the contractor, the
supervisory staff and crew must be fully engaged in
the Environmental Management Plan and buy into the
necessity of environmentally responsible construction
operations.

York Region has initiated a mandatory four hour
training workshop for all construction site staff,
including consultants, contractors and sub-trades. The
training topics include: history of the project,
environmental concerns and issues, conditions of all
approvals and permits, work area safety and
environmental compliance procedures. It is intended
that the contractors supervisory staff will carry these
messages to every individual on site through weekly
tailgate meetings. Quick reference handouts
appropriate to the level of training have been prepared
by the Region and its consultants for training
purposes.

The overall safety and environmental compliance goal
is to adequately prepare individuals at all levels of the
project to be responsible for leading and engaging
project staff, contractors, suppliers and partners in
meeting our safety goals and objectives. The safety
and environmental core values include:
Occupational injuries, environmental releases, and
property damage incidents can and must be
prevented;
Safety, Health and Environmental compliance is
everyones responsibility;
All activities will be monitored to ensure that all
hazards and controls are properly identified,
evaluated and communicated to the Region in a
timely manner; [4]
All construction site staff and crew must take
ownership of environmental responsibility find
a problem and supply a remedy with the
sanction of supervisory staff.

5.2. Enhanced Public Engagement and
Consultation

A multifaceted program was developed by an
Independent Consultant to engage Regional residents,
agencies and special interest groups in all aspects of
the YDSS Interceptor project.

The goal was not consensus or even necessarily
agreement, but rather that the public voice was heard
and acknowledged. Some of the most effective
techniques are listed here.
Media release and fact sheets
Provision of a 1-800 number for the public
Public notices of key project dates and milestones
published in local newspapers and posted to the
project Web site.
Telephone Access/Response to/from key project
staff.
Face-to-face consultation undertaken when
appropriate throughout the project through PCCs
and PIFs.
An Advisory Committee comprised of interested
residents, members of agencies, local community
representatives and special interest groups
provided their position on the project from the
environmental assessment to project permitting.
Plain English summaries, guides and newsletters
were distributed to explain the large and complex
reports. These large documents were available in
viewing rooms and libraries. Copies of
summaries were available upon request.

The unique features of the PCC used to explain the
PTTW application and approval processes were
Trade-Show Style Open Houses and Breakout Groups.

204
Trade-Show Style Open Houses included display
panels, videos and models to help the stakeholders
better understand the technical details of monitoring
and construction.

Breakout Groups of two 20 minute presentations and
discussion sessions provided a smaller group format
to receive and provide feedback to the project team.
All comments and responses were compiled into the
independent facilitators report and made available to
the public, stakeholders and agencies.

A project specific Web site provides access to a
library of all documents including milestones,
schedules and updates of interest to the public. [5]

5.3. Facts and Figures

The following list demonstrates the extraordinary
level of effort required to accomplish this project.
- Up to 70 consultant staff working at a given time,
over a 14 month period.
- 27 key staff from the Region, Agencies and
Consultants visited an EPB TBM sewer project
under construction, also in an end moraine and
using sealed shaft construction, to observe first
hand the construction methods, challenges and the
application of different (US EPA) regulations.
- The PTTW application is 29 pages with a
supporting EMP and appendices in two 75mm
binders.
- The cost of the construction environmental
monitoring and mitigation by York Region is 3%
of the overall construction cost. Additionally the
contractors have an independent cost item for
environmental management and compliance.
- On average 1.5 weeks per month were spent by
the team meeting and talking with various
regulators; and five days per month were spent
responding to public/political concerns and
meeting with members of the public.
- The project has required more than 10 project
managers through the course of the YDSS
Interceptor project life.
- Originally scheduled for completion by the end of
2005, the current projection is for completion of
the two contracts by the end of the first quarter of
2008. This delay is primarily attributed to filing
and receiving the extensive permitting and
approvals.

REFERENCES

1. Vision 2026, York Region: Ontarios Rising
Star, Towards a Sustainable Region, Fourth
Annual Report on Indicators of Progress,
Regional Municipality of York, Spring, 2006.
2. Chapman, L. J., Putman, D.F. The
Physiography of Southern Ontario Third
Edition, Ontario Geological Survey, Special
Volume 2, Ministry of Natural Resources,
Ontario, 1984.
3. YDSS Interceptor Sewer Study Report &
Appendices, New Comparison of Alternative
Route Alignments, 2005, Earth Tech in
association with Alston Associates Inc.,
Beatty & Associates, Parish Geomorphic, and
Michalski Nielsen Associates Limited.
4. YDSS Interceptor Sewer Study,
Environmental Management Plan, In Support
of a Permit to Take Water Application and
other Environmental Approvals for the 19
th

Avenue/Leslie Street Project, March 6, 2006,
Earth Tech in association with Alston
Associates Inc., Beatty & Associates, Parish
Geomorphic, and Michalski Nielsen
Associates Limited.
5. York Durham Sanitary System (YDSS)
Interceptor Sewer on 19
th
Avenue & Leslie
Street: Geotechnical Baseline Report,
Richmond Hill, Ontario, 2006, Hatch Mott
MacDonald in association with Earth Tech
Canada Inc., Alston Associates Inc., W.B.
Beatty and Associates Limited and Thurber
Engineering Ltd.

205
1 INTRODUCTION
Auger boring machines (ABMs) have been in use for
over a century, but their performance has improved
significantly with evolution of modern pipe jacking
methods. More powerful boring machines combined
with modern pipe-jacking methods, better steel casing,
improved materials for augers and tungsten-carbide
cutting structure are permitting ABMs to install longer
and larger diameter casings. With the incorporation
of rock cutting heads, ABMs have become a much
more versatile tool for utility installations while
working in limited water inflow.
ABMs typically use various shapes and styles of
rock cutting tools, usually incorporating tungsten-
carbide teeth for wear resistance. Most of these tools
are inefficient and not cost effective while cutting
medium to hard and/or abrasive rock formations in
excess of 28 MPa (4000 psi) unconfined compressive
strength.

1.1. The History of Auger Boring Machine
Capabilities

Augers have excavated earth for centuries. Various
power sources have been used in the past and have
evolved into the modern auger boring machine. ABMs
produce more than ample thrust and torque for the
average geology encountered for most utility
installations less than 120 150 m (400 500 ft) in
length. In soft ground applications, the casing is
jacked forward, cutting the periphery, while the auger
removes material from the face and pulls the spoils
back through the casing to the bore pit for removal to
the surface.

Over the past few decades manufacturers have
increased both thrust and torque of the ABM. With
improvements in materials and casing strength, thrust
capacity has increased steadily to keep pace with
those improvements, allowing for longer and longer
installations. As drives increased in length and cutting
tools improved in quality, increased torque was
required as well to maximize equipment capabilities.

The ABM method was still mostly limited to
loose soils, sands, gravels, or similar ground with little
or no water present. Before the last 10 years,
contractors cut rock with ABMs through utilizing
cutting heads with tungsten carbide teeth, such as
Christmas tree heads. These cutting attachments were
and are available in a variety of shapes. These
attachments excavate rock through cutting or ripping
away at the rock. The result is considerable torque
spikes that transfer to the ABM creating significant
wear on the entire ABM drive train. Cutting rock
with attachments such as these is effective, but not
efficient for rock strength below 28 MPa (4000 psi).
When such tools are used on rock that exceeds 28
MPa (4000 psi) in strength, then the bullet bits break
off and wear rapidly, causing the contractor to pull the
cutting head back and replace parts every few feet of
boring. In an attempt to cut rock, ABM owners in the
past have tried disc cutter heads on their machines.
Many of these cutter heads were not very well
engineered devices; instead, they were crude copies of
TBM cutterheads and were not cost effective tools for
ABM applications.

Efficient Excavation of Small Diameter Utility Installations in Hard
Rock

Dave Long
The Robbins Company, Kent, WA, USA

ABSTRACT: This paper discusses the evolution of boring equipment used with Auger Boring Machines (ABMs) for
underground utility installations in medium to hard rock. The past 10 years have seen the advent and subsequent
development of several methods of cutting hard rock with ABMs. This paper briefly summarizes the evolution of ABMs
and the subsequent development of rock cutting heads for ABMs. It then goes on to outline the most efficient rock cutting
tool used with ABMs, Small Boring Units (SBUs). The utilization of this technology is then explored through some recent
projects employing SBUs. Limitations and strengths of the ABM method are also discussed for pipe-jacked projects in
hard/abrasive rock. In closing, the paper touches lightly on the potential economic advantages of ABMs and outlines the
future potential of the ABM and tools to allow for improved economics and equipment versatility.

206

Figure 1. Typical Christmas Tree Head using tungsten-
carbide bullet bits to cut consolidated materials or weak
rock.
2 DEVELOPMENT OF ROCK CUTTING HEADS
FOR AUGER BORING MACHINES
The larger man entry tunnel boring machines (TBMs)
have been cutting harder and harder rock since the
first successful use of rolling disc cutters in the early
1950s. Disc cutter technology has evolved since then
with incremental improvements in disc ring
metallurgy and material processing, increased bearing
capacity and improved lubrication. All of these
advances have contributed to increased cutter load
capacity. In addition, the application of disc cutters
has improved as TBM manufacturers gained
knowledge of the effects of cutter spacing,
penetration-to-spacing ratios, gage-cutter
development, cutter housing retention design, muck
flow, etc.For many years, the only companies making
underground excavation equipment for a full range of

Figure 2. Typical ABM and rock cutting SBU.
diameters (e.g. microtunnelling machines, earth
pressure balance machines (EPBMs) and hard rock
TBMs) were J apanese. However, in J apan the
primary focus was on soft earth EPBM type
technologies to address domestic demands, which did
not leave room for development of a range of other
applications. Since ABMs, hard rock TBMs and
EPBMs were made by different manufacturers around
the globe, there was little technology transfer between
the different machine types.
In recent years there have been a growing number
of underground equipment manufacturers that have
recognized the need for integration of technologies
into equipment for all disciplines. Today, there are
several companies making all of these products and,
as a result, the rate of technology transfer has
increased considerably, resulting in rapid
improvements in underground construction
equipment.
Equipment manufacturers recognized a need by
ABM owners to be able to cut rock efficiently with
attention to cost and production. In response they
began to develop and test disc cutter cutting heads
specifically for auger boring machines. Commonly
referred to as small boring units (SBUs), these rock
cutting heads incorporated a half century of TBM hard
rock cutting technology into their design. Experienced
hard rock TBM manufacturers well understand the
requirements for efficient rock cutting with disc
cutters on large diameter tunneling machines.
Integration of the disc cutter method to small boring
units (SBUs) for use with ABMs presented unique
design challenges for the manufacturer. SBUs range
in size from 600 mm to 1.8 m (24 in. to 72 in.) in
diameter. Today, contractors employing SBUs with
their ABMs are successfully and economically cutting
hard rock.

207

Figure 3. Detail of Small Boring Unit. Front and profile
view.

2.1. SBU Cutters and Cutter Spacing

Unlike ABM operation in soil, when boring through
rock the casing cannot cut the periphery, or gage
diameter. Disc cutters had to be utilized to excavate a
diameter larger than the casing called over-cut in
rock to allow for casing installation. A cutter much
smaller than the tunneling industry standard, 430 mm
and 490 mm (430 and 480 mm / 17 and 19 in.)
diameter, had to be employed in order to create an
effective gage-cutting pattern (see Figure 4).

Figure 4. Typical 430 mm and 165 mm (17 in. and 6.5 in.)
cutter gage area. Using the smaller cutters provides a flat-
face, minimum periphery exposure and leaves space for
spoils paddles.

Using smaller cutters allows the gage to be cut with
approximately the same number of cutter locations
that are used on a larger TBM cutterhead, but with a
smaller gage-radius. This leaves sufficient space on
the cutterhead for spoils openings to remove the cut
rock from the face. As utility tunnel diameters
increase, it is possible to use larger cutters, which is
highly beneficial in terms of increased load capacity
and production. Today, ABM cutter diameters range
from 165 mm (6.5 in.) for smaller casing installations
to 290 mm (11.5 in.) for larger bores.

The objective for efficient rock cutting is to break the
rock into chips, rather than crushing it into fines or
ripping it from the face. Crushing the rock
unnecessarily requires more energy and results in
excessive wear on cutter rings, cutterhead and augers.
Similarly, ripping action at the rock face created by
conventional ABM cutter heads creates very high
torque spikes, creating unusually high wear on the
entire ABM drive train. In short, inefficient rock
cutting tools increase the cost of equipment operation
and maintenance. For a given distance of penetration
of the cutter into the rock per revolution of the
cutterhead, there is a preferred distance of spacing
between adjacent cutter positions. The spacing-to-
penetration ratio must be optimized for efficient rock
cutting to occur.

Cutter penetration is governed by several factors
including: rock strength, rock mass properties, cutter
load, cutter diameter and cutter ring tip-width. There
is literature widely available from the worldwide
academic community regarding theoretical methods
for estimating the penetration rate for rock
cutterheads. Most penetration estimating methods
require as minimum inputs unconfined compressive
strength (UCS), Brazilian tensile strength (Bt),
fracture spacing, fracture dip and strike, cutter
diameter, cutter tip-width, cutter load and cutterhead
speed (revolutions per minute). With additional
information, some manufacturers can also estimate the
cutter cost-per-volume excavated for the larger
tunneling projects where numerous cutter changes are
made.

Most equipment manufacturers have developed their
own proprietary algorithms, which are benchmarked
against their internal performance records from
previous projects. Having access to a great amount of
field data is imperative for the development of an
accurate penetration-estimating algorithm, and an
accurate penetration-estimating algorithm is an
essential tool for designing an efficient, rock-cutting
cutterhead.

After the rock has been cut it must be drawn from the
face and pushed to the aft side of the rotating
cutterhead to the auger for removal from the casing.
The paddles, or bucket lips, used to move the muck to
the back of the head are subjected to heavy abrasive
wear and is important to understand the flow of the
material through the machine in order to minimize the
wear on these components. They are generally made
of abrasion-resistant steels.
208
Figure 5. Diagram showing rock chip formation with
properly spaced cutter discs.

In the case of long drives, or extremely abrasive rock,
it may be necessary to change cutters before the
casing has been fully installed. This task may be
accomplished by making the gage cutters retractable
and altering the design of the shield. With this design
it is possible to withdraw the SBU back through the
casing and service it on the surface. Retractable SBUs
also include wheeled transport units, which make it
easier to pull the SBU out of the casing.

Figure 6. View from back of large SBU, through the
cutterhead, to the rock face. Note concentric cutter paths
and kerf cutting.

Figure 7. Two SBUs on factory floor showing muck
buckets or paddles, which move the cut rock aft of the
cutterhead, to be removed by the auger.
2.2. SystemThrust Requirements and Limitations

The SBU shield is welded to the lead casing during
installation. Thrust force is transmitted from the
ABM through the casing, and then into the SBU
shield. Thrust force is then transmitted into the
bearing housing, through the bearing, and into the
cutterhead and disc cutters to the rock face.

The ABM provides thrust to overcome casing to rock
(skin) friction as well as thrust to the cutters as
required to fracture the rock. The thrust required for
any specific cutterhead to efficiently fracture the rock
is a function of the properties of the rock to be cut, the
cutter ring diameter, cutter tip-width, and the number
of cutters on the head profile.

During ABM operation, the skin friction on the casing
can exceed the thrust required for fracturing the rock.
As a result, thrust requirements when using a rock
boring SBU is seldom a problem for most ABMs. In
fact, as tunnel length increases, the thrust required to
overcome skin friction can become higher than the
thrust required by the cutterhead to fracture the rock.
209
Depending on the ground conditions there are three
methods of operation to prevent exceeding maximum
thrust capacity of the SBU and cutter bearings:

In homogenous rock properties over the bore
length, it is sufficient to limit the advance rate
of the cutterhead, thereby limiting the cutter
penetration per cutterhead revolution.
In any ground condition, actual skin friction
may be monitored by retracting the SBU a
meter or so and then thrust back to the face
recording actual thrust requirements to
overcome skin friction while not cutting rock.
By adding that to the allowable cutter loading
for the particular SBU in use, the operator can
ensure there is enough cutter loading, while
also ensuring the SBU and cutters are not
being overloaded.
In geology that is highly variable and where
very hard rock is present, it is advisable to
have thrust jacks installed between the boring
unit and the liner pipe, so that boring unit
thrust can be monitored separately from the
casing thrust.

Figure 8. SBU welded to steel liner pipe, in launch chamber
ready to excavate limestone.
2.3. SystemTorque Requirements

The ABM power unit provides the torque for the
auger spoils removal as well as the torque to power
the SBU cutterhead. The torque is transmitted from
the ABM drive output into the auger string and from
the auger string into the SBU cutterhead.

The ABM system provides the torque to overcome
auger-to-casing friction as well as torque to the SBU
cutterhead. The torque required for any specific
cutterhead to work efficiently is a function of the
properties of the rock to be cut, the cutter ring
diameter, cutter tip-width, the number of cutters, the
radial location of cutters on the head, and the cutter
penetration into the rock. On an ABM the torque
required to turn the auger is a function of the auger
diameter, pitch, volume of material in the auger,
material size, presence of water, etc. On typical
drives, the torque required to turn the auger is
generally 2 to 3 times the torque required to turn the
SBU cutterhead.

Figure 9. Standard hexagonal input shaft for SBU
cutterhead drive. Note the mechanically adjustable
stabilizer pads, also used for initial steering.

The SBU cutterhead is fitted with a standard male hex
adapter onto which the female hex in the lead auger
freely pilots onto and off of, to provide torsional
power to the cutterhead. As the tunnel and auger
become very long, auger friction can become extreme.
For very long tunnels, the SBU can be fitted with a
separate cutterhead power unit, either electric or
hydraulic drive, so that all ABM torque can be
employed to turn only the auger.

In all cases, the auger friction to casing results in
torsional forces being introduced into the casing and
for this reason the casing is generally torsionally fixed
to the ABM, where the torque is reacted by machine
weight and the earth. It is important to prevent the
casing from rotating, since a rotating pipe will reduce
the spoils removal efficiency of the auger. When
using an SBU with a separate cutterhead drive motor,
the cutterhead torque is also reacted by the casing,
providing further need for good torsional fixing of the
casing at the ABM.

2.4. Steering the SBU

The SBU is typically fitted with small, manually
adjustable stabilizer pads. With the SBU shield
welded to the casing, there is limited opportunity to
steer the machine mid drive. It is imperative to
adjust the stabilizer pads as required while the initial 6
210
to 9 m (20 to 30 feet) of casing is installed in order to
get the casing started on the correct path. With a good
start, the SBU will typically maintain direction with a
high level of accuracy. During the initial 6 to 9 m (20
to 30 feet) of the drive, it is usual to remove the auger
from the casing several times to check the line and
grade of the advancing SBU.

Generally, following the initial 6 to 9 m (20 to 30
feet), the stabilizers are adjusted to prevent the natural
tendency of the rock boring head to climb and move
to the right. This tendency is due to the clockwise
rotation of the head when viewed from the rear.
Typically the stabilizer pad in the upper right quadrant
of the SBU is extended considerably more than the
other stabilizers, but all are in contact with the rock
wall to ensure stability of the disc cutters with the
rock face, which is primary for efficient excavation of
rock with disc cutters.

Figure 11. Three SBUs in factory. Note mechanically
adjustable stabilizer pads used for initial steering.
3 RECENT PROJ ECTS
To date, there have been over 300 SBUs that have
been delivered for utility installations worldwide. The
SBU+ABM has been mostly used in the Northeastern
U.S.A., due to domestic population centers and the
acceptance of the ABM method. However, the
technology is becoming more highly accepted
throughout the Americas, and into Europe as well as
other parts of the globe. Following are examples of
some of the utility projects where the SBU has been
deployed.
3.1. Louisville, Kentucky, USA -- SBU

Louisville Water Company awarded a contract for the
Westport Road Transmission Main, an improvement
of the Louisville water distribution network. Part of
the project required trenchless methods to install
parallel 760 mm and 1.2 m (30 and 48 inch) water
lines through limestone with UCS in the range of 55
to 90 MPa (8000 to 13000 psi). The rock crossing
was 100 m (328 feet) long. Work on the project
started in April 2003.

Contractor Midwest Mole of Indianapolis, Indiana
bored the tunnels with an SBU, powered by a 225 HP
ABM with five-speed drive and a 127 mm (5 inch)
hex auger output. An average advance rate of 6 m
(20 feet) per day was achieved. A duct type final liner
was installed within the steel casing.
3.2. Glasgow, Scotland, UK -- SBU

F&B Tunneling of Tickhell, Doncaster, UK elected to
use an SBU+ABM, rather than their MTBM, to drive
five 30 m (100 ft) long, 1.2 m (48 in.) diameter
casings. The geology was reported as mixed ground
to 100 MPa (14500 psi) rock. The contractor
achieved advance rates as high as 9 m (30 ft) per day.

3.3. St. Louis, Illinois, USA SBU & MSBU

Missouri American Water Co. serves more than
445000 customers in nine areas in Missouri, including
St. Charles, St. Louis and Warren counties. It is a
division of American Water Works Co. For the
Missouri American Water Imperial project located
south of St. Louis, Illinois at Mastodon State Park,
222 m (730 ft) of 1.2 m (48 inch) diameter water main
needed to be installed through limestone.

Contractor Aarrow Boring of Bridgeton, Missouri
drove the tunnels using an ABM and two different
SBUs. The ABM was rated at 106 kW (142 hp) with
five speed gearbox, a 50 mm (3 in.) hex auger output
and 3340 kN (750000 lb) of thrust.

A 1.2 m (48 in.) SBU+ABM, was used for an 80 m
(260 ft) drive in one direction. Then, a 1.2 m (48 in.)
motorized SBU was used to drive 150 m (490 ft) in
the other direction. On the second and longer drive,
the contractor employed a 200 mm (8 in.) diameter
aluminum vacuum tube for spoil removal. Average
advance rate on the project was 6 m (20 ft) per day.
3.4. Manassas, Virginia, USA -- SBU

Manassas, Virginia is a suburb of the growing
Washington, D.C. metropolitan area, and lies within
Prince William County. The Prince William County
Service Authority has a 20 year plan to update the
local water supply network. New Construction, Inc.
received contracts for construction of Phases I and II
211
of the project that included approximately 10 km
(35000 ft) of 1.06 m (42 in.) diameter water line.
Several sections of the project required trenchless
methods in order to minimize surface disruptions.
One such section was a crossing under State Route
234, a six-lane highway. The crossing was 5.5 m (18
ft) under the road surface through red shale. An open-
face tunneling machine was first used to attempt the
crossing but failed after cutting only 1.5 m (5 ft) and
the project stalled for approximately 6 weeks.

New Construction then subcontracted the crossing to
Fithian Contracting of Youngstown, Ohio. Fithian
owns three SBUs, and as one of the initial owners of
the SBU product, is very experienced in their use.
Fithian used a 1.5 m (60 in.) SBU and steel casing to
drive the 55 m (180 ft) in only three and a half days.
Twice the contractor advanced the SBU over 18 m (60
ft) in a single day, which is a record for this size of
machine. The SBU emerged within 6 mm ( in.) of
line and grade target. The 1.06 m (42 in.)
transmission line was installed within the 1.5 m (60
in.) steel casing and the annulus backfilled with sand.

Figure 12. Manassas, Virginia. 1.5 m (60 inch) diameter
SBU breaks through red shale rock face after completing a
180 feet drive.
3.5. Big Sky, Montana, USA -- SBU

Big Sky, Montana is a rapidly growing resort area
located just 30 minutes from West Yellowstone. The
Yellowstone Club, a private resort, is in the final stage
of construction which includes a gated community
with scores of high end homes and a private ski and
golf resort. A 300 million liter (80 million gallon)
reservoir was excavated to hold treated waste water
for purposes of golf course irrigation. The design
called for the installation of a 762 mm (30 in.) casing
96 m (318 ft) long. The material was thought to be fill
to soft rock with UCS values less than 27 MPa (4000
psi).

Tunnel Systems Inc. of Woodinville, Washington was
awarded the work and commenced boring in
September, 2004 utilizing a conventional Christmas
Tree style head with carbide bullet bits. After working
seven shifts averaging about 3 m (10 ft) per day, while
pulling the head several times to replace broken and
missing bits, Tunnel Systems decided to change cutter
head tooling.

It was decided to use an SBU to complete the
remaining 74 m (242 ft) in mudstone and shale,
estimated to be in excess of UCS 35 MPa (5000 psi).
The decision was made based on economics as well as
time constraints. The project was located at a 2600 m
(8500 ft) elevation, and snow was already beginning
to fall, so the need to finish the drive and mobilize
their ABM and support equipment to a lower
elevation was immediate.

Once boring resumed with the SBU, Tunnel Systems
averaged about 15 m (50 ft) per day with the 762 mm
(30 in.) SBU, finishing the installation in just five
shifts. This was their first experience with the
technology, and the SBU came out within a few
inches of planed line and grade.

Figure 13. Big Sky, Montana, 2004. A 762 mm (30 inch)
diameter SBU is used to complete the 96 m 318 ft) drive in
mudstone/shale.
3.6. Redmond, Oregon, USA SBU & MSBU

Stadeli Boring & Tunneling, Inc. of Silverton, Oregon
began using Robbins SBUs in J anuary of 2006. Their
first bore utilized a rented Robbins 762 mm (30 in.)
SBU for use on the highway 97 bypass project,
diverting traffic from the highly congested downtown
Redmond route. The Oregon Department of
Transportation estimated total cost of the project to be
about $70 million USD, of which about $6 million
USD was awarded to Hap Taylor & Sons. The project
called for a 37 m (120 ft) of 762 mm (30 in.) casing
beneath the railroad track through Deschutes
Formation, which is volcanic basalt. The bores were
necessary as a result of the relocation of existing
212
water mains that fell directly in line with required 4 m
(12 ft) deep footings in bedrock, for a yet to be
constructed overpass. The bore was monitored closely
by the ODOT and General Contractor, Hap Taylor,
due to the critical path of the bore on the overpass and
project schedule.

Stadeli had expected an average of 3 m (10 ft) a day,
based on experience, and factoring in discussions with
the manufacturer. Instead, they bored up to 3 m (10
ft) an hour, completing the bore well in advance of
schedule, allowing overpass construction to proceed
ahead of schedule.

Since the initial bore, Stadeli has completed two
subsequent 915 mm (36 in.) bores on the same
project, which were 35 m (120 ft) and 18 m (60 ft)
each with excellent results. As of August 2006,
Stadeli has a motorized version of a 1.2 m (48 in.)
SBU currently on order that will be used on two bores
in excess of 150 m (500 ft) each in the Hood River,
Oregon area, where line and grade are critical.

3.7. South Wales, UK SBU & MSBU

National Grid is the main supplier of natural gas for
the UK. In 2005, National Grid announced its plans
for a multi-million pound natural gas pipeline that
would transport gas over 120 km (75 mi) through the
Welsh and English countryside. The pipeline, when
finished, will transport up to 20 percent of the UKs
natural gas.

The general contractors, a NACAP/Land and Marine
J V, subcontracted B&W Tunnelling for 60 crossings
through hard rock in Phase I of the project. B & W
Tunnelling chose two Robbins SBUs and one Robbins
Motorized SBU, all 1219 mm (48 in.) in diameter.

The machines are excavating crossings ranging from
20 90 m (66 295 ft) in length. As of J uly 2006,
the Motorized SBU had completed its first crossing
and was averaging 1.5 2.0 m (5.0 6.6 ft) per day.
The machines are boring through mudstone and
sandstone. The two SBUs are performing at similar
rates.
4 LIMITATIONS OF THE METHOD

The reader is reminded that the discussion herein is
with respect only to cutting medium to hard rock on
small bore, pipe-jacked tunnels.

ABMs have the required torque and thrust to cut hard
rock if fitted with a proper rock cutting head and disc
cutters. The primary limitations of the SBU+ABM
method with respect to cutting rock are:

Torque requirements for the cutter head
torque and spoils removal on larger diameter
bores in excess of 122 m 152 m (400 500
ft) in length, while using full diameter augers.
Ability to deal with excessive water.
The ability to continuously steer very small
diameter SBUs, which can be done, but with
great difficulty.
The ability to only change cutters in mid-bore
on medium and large diameter SBUs, which
cannot be done on smaller SBUs, without
retracting the casing.
Thrust requirements on bores in excess of 122
m 152 m (400 500 ft).

5 FUTURE POTENTIAL DEVELOPMENT OF
THE ABM+SBU
Development of the todays ABM and tooling over
the past 30 years has been driven in part by the need
for increased power requirements, to enable larger and
longer installations as previously discussed in this
paper. Safety issues and environmental concerns have
been mandated and addressed, through contributing to
the technologic improvements of the machines, as
well as safer working environments.

It is evident that the SBU will continue to evolve to
become more versatile and include features to allow
the equipment to be utilized on more difficult jobs. In
the SBUs brief history, there have been considerable
product improvements necessitated as manufacturers
gained a better understanding of the applications.

5.1. Remotely Controlled Motorized SBUs

Remotely controlled, motorized SBUs (MSBU) utilize
the ABM to supply thrust conventionally through the
steel casing, while supplying torque through an
independent cutterhead drive motor. This torque is
transmitted through the center hex shaft drive as seen
on the standard ABM. To provide room for the drive
motor, the MSBU is supplied with small diameter
casing and auger, which is powered by the ABM for
spoil removal. In addition, a standard pipe laying laser
is used in the pit and directed to a target on the SBU
for guidance. Articulation cylinders mounted on the
SBU shield are then actuated by the operator on the
surface for line and grade corrections. The MSBU is
213
used on longer and/or larger diameter installations,
and/or where line and grade may be critical such as
gravity sewer installations, and may be fitted with
cutterhead tooling to excavate any expected geology.

5.2. Belt Conveyors for ABMs

On all ABM projects there are provisions made to
evacuate the spoils from the bore pit for removal from
the work site. The logistics to allow for spoils
removal usually include a laborer to manually clear
the spoil pile from the chute opening on the ABM. In
addition, there is more than one method to remove the
spoils from the pit including backhoes, bucket
excavators, muck boxes, etc. All this equipment is
generally owned or rented by the contractor at a cost,
and takes up considerable space, and at times is not
available.

The technology currently exists to supply small
horizontal and vertical, or near vertical belt conveyors
to remove spoils to the surface and into a truck for
removal from site. This is done commonly on the
larger TBM projects around the world simplifying the
handling of excavated material.

The use of conveyors for spoils removal with the
ABM is a concept that should be given a chance to
evolve. The cost is easily offset by the reduction of
on-site personnel and heavy equipment.

5.3. Retractable Gage Disc Cutters

One of the major fears of most contractors using an
SBU with disc cutters is cutter failure prior to
completion of the drive. As discussed previously in
this paper, some manufacturers supply retractable
gage cutters to allow for retraction of the unit back
through the casing for servicing on the surface. This
retractable cutter technology has been used on several
SBUs and TBMs, but advancements and practical use
are necessary to gain acceptance into the trenchless
utility construction industry.

This design challenge is multiplied on the SBU due to
the smaller disc cutters used, and limited room on the
gage area of the cutterhead. The mechanism allowing
for cutter retraction must be very robust, while at the
same time packaged in a very small area.

This is a challenge for all underground construction
equipment manufacturers, and is being given a higher
priority to make the ABM+SBU method a more
viable solution for longer drives in extremely
hard/abrasive rock conditions.

6 CONCLUSIONS

To date, SBUs have cut well in excess of 32.2 km (20
mi) of rock up to UCS 172 MPa (25,000 psi), on over
500 projects, primarily across North America.
Clearly, the method offers the opportunity to
economically excavate hard rock in suitable ground
conditions.

While SBUs were generally intended for drives up to
152 m (500 ft), they have been employed successfully
on longer drives. C.B. Services, in Texas, completed
a 195 m (640 ft) drive using a 914 mm (36 in.)
diameter SBU. J & J Boring, in Virginia, completed a
233 m (763 ft) long, 1372 mm (54 in.) diameter, drive
through mica schist with UCS to over 55 MPa (8000
psi). It would appear that the ABM +SBU method
could successfully complete drives of perhaps 244
305 m (800 1000 ft). Auger torque limits would
likely preclude longer drives.

If there is water present in significant quantities or at
pressure, some machine type other than the
SBU+ABM is the only choice. In the presence of
water and hard rock, the drive must be of a short
enough length, which assures arrival of the selected
machine at the reception pit, prior to failure of the
cutters. Other machines cannot be removed back
through the casing for cutter changes, and there is no
chance for man entry on these smaller machines.
Geology must be known or project costs can explode
for retrieval and/or repair on such projects.

If there is not water present in significant quantities or
at pressure, an ABM + SBU can be clearly be
employed successfully. Since many contractors
already have suitable ABMs and augers in their fleets,
it is frequently only the incremental investment in an
SBU that is the added cost for excavating rock. When
the geology is appropriate for the method, an ABM +
SBU system can be a very efficient method of
excavation and casing installation, and provides a very
cost effective solution for small bore, rock excavation
for casing installation.

To advance the growth of equipment used in
trenchless utility installations machine manufacturers,
contractors, project designers and owners alike have a
unique opportunity to reap the rewards of
improvements by becoming more involved as a
community in product evolution. As improvements
are made all concerned parties have an obligation to
214
ensure advancements are recognized and properly
applied in the utility installations.

REFERENCES
1. Roby, J . and T. Fuerst. 2004. Economic Tunneling in
Hard Rock. In International Society for Trenchless
Technology, Conference Proceedings, No-Dig.
2. Calio, V.J . and F. Nee. 2004. Multiple Tunneling
Methodologies in Atlanta Georgia. In North American
Society for Trenchless Technology, Conference
Proceedings, No-Dig.
3. Iowa State Department of Urban Design. 2004. Design
Manual / Chapter 14 Trenchless Construction. In
Iowa Statewide Urban Design and Specifications.
4. Bennet, Dr. D. 2004. Beaver Water District
Microtunneled Intake Pipelines and Intake Shafts
Through Karstic Limestone. In North American Society
for Trenchless Technology, Conference Proceedings,
No-Dig.
5. Atalah, A. 2003. Pipe J acking & Microtunneling
Operations. In Presentation to the Detroit Rehab Road
Show, May 2003.
6. Sedjo, A. 2002. Microtunneling: Poised for Growth in
the U.S. Market. In Trenchless Technology, May 2002.
7. Brown and Caldwell & Herrara Environmental. 2001.
King County Conveyance System Improvement
Project / Conveyance System Cost System / Trenchless
Technology Cost Parameters. Final Report. King
Count: Washington USA.
8. Rush, J .W. 2000. SBUs Aid in Water Line
Installation. In Trenchless Technology, December
2000.
9. Rostami, J ., L. Ozdemir, and B. Nilson. 1996.
Comparison Between CSM and NTH Hard Rock TBM
Performance Prediction Methods. Excavation
Engineering and Earth Mechanics Institute, Colorado
School of Mines.
10. University of Trondheim. 1994. Hard Rock Tunnel
Boring. In Project Report 1-94, University of
Trondheim, The Norwegian Institute of Technology.
Trondheim.
11. Deering, K., G. Dolinger, D. Krauter, and J . Roby.
1991. Development and Performance of Large
Diameter Cutters for Use on High Performance TBMs.
In Proceedings, Rapid Excavation and Tunneling
Conference.
12. Ozdemir, L., R. Miller, and F. Wang. Rock-Cutter
Boreability Parameters. In Excavation Engineering and
Earth Mechanics Institute, Colorado School of Mines.

215

European Motorway Tunnels
Evert Hoek,
TAC Conference, Vancouver, September 2006
ABSTRACT: Increasing traffic demands, particularly for commercial goods vehicles, have placed a strain on the existing
highway network in Europe. This has resulted in an expansion of the highway network with hundreds of kilometres of new
roads, tunnels and bridges at various stages of design, construction and operation. All of the motorways which form part of
the Trans European Highway Network have to meet rigorous design standards which are aimed at minimizing the number
and consequences of accidents. Tunnel fires are a special concern and minimizing the risk of such fires is a significant
component of the tunnel design standards which will be discussed in this presentation.

The Egnatia highway is a 680 km long 4 lane motorway running across northern Greece and forming part of the Trans
European Highway Network. It has 75 tunnels, with a total length of about 100 km, 1650 bridges and 50 interchanges.
About 60% of the motorway has been completed and is in operation and the project is scheduled for completion in 2008.
The crossing of the Pindos Mountains, the southernmost extension of the Alps, has confronted the designers and
contractors with many challenges because of the wide range of tectonically disturbed rock types encountered. The site
investigations, design and construction methods used in overcoming these challenges will be described. The project
management and contracting arrangements have also played an important role in the success of this project and these will
be reviewed briefly.

With financial and logistical support from the Egnatia Odos organization, a very large database on the Egnatia tunnels has
been compiled by the Department of Civil Engineering of the National Technical University of Athens. This database
includes information on site investigations, engineering geology models, rock mass characteristics, groundwater
conditions, excavation and support design, final lining design, construction methods, advance rates and detailed costs. All
of the information is cross referenced so that correlations between variables can easily be examined. It is anticipated that
this database will provide an important resource for tunnel designers in the future.
216

Vancouver 19 TH NC

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Vancouver 19 TH NC

Încărcat de

Drepturi de autor:

Formate disponibile

TUNNELLING ASSOCIATION OF CANADA

S-ar putea să vă placă și