Sunteți pe pagina 1din 11

Pressurized Face Tunneling under Very

High Groundwater Heads


Steven W. Hunt
CH2M, Henderson Nevada, USA

ABSTRACT
A paper called Global Experience with Soft Ground and Weak Rock Tunneling under Very High
Groundwater Heads was presented at the North American Tunneling 2006 conference (Holzhuser et.al
2006). It summarized global experience with earth pressure balance (EPB) and slurry shield tunneling on
nine projects that encountered groundwater heads ranging from 6 to 11 bar. Since then, additional
projects have been completed with groundwater heads ranging from 7 to 15 bar. Of these, the high
pressure tunneling experience on the Lake Mead Intake No. 3 project in Nevada and on the Eurasia
Tunnel (Istanbul Strait Crossing) had the greatest lengths of tunneling at face pressure over 7 bar.
Experience with sustained pressurized face tunneling at groundwater heads over 7 bar and risk
management observations are the focus of this paper.

INTRODUCTION
High groundwater head, particularly at 7 bar or more, is a major challenge when tunnelling in high
permeability ground and on drives long enough to require cutter changes. It has a strong impact on
design and operation of Tunnel Boring Machines (TBMs) in order to protect the main bearing from
contamination, prevent excessive groundwater inflow, ensure face stability and enable access to the
cutterhead and excavation chamber for maintenance.
During the past 20 years, TBM manufacturers and suppliers have made tremendous improvements
that allow tunneling at groundwater pressures to 17 bar or more. The most significant risk with high
pressure tunneling is generally that of excavation chamber access and potentially the cutterhead for
cutter changes, cutterhead repairs and TBM maintenance.
Hyperbaric interventions are possible and safety variances have been approved for saturation dive
interventions at pressures as high as 17 bar (Lake Mead Intake No. 3). While high pressure hyperbaric
intervention equipment is available and safety approvals can be obtained at many locations, few
contractors consider the cost and risk of hyperbaric interventions over 7 bar to be acceptable as the
primary means of excavation chamber access.
Experience with high pressure tunneling prior to 2006 was summarized in a paper entitled Global
Experience with Soft Ground and Weak Rock Tunneling under Very High Groundwater Heads
(Holzhuser et.al 2006). That paper provided information on 9 projects: South Bay Ocean Outfall, San
Diego; Channel Tunnel (French Side), France; Storebaelt Tunnel, Denmark; 4th Elbe Tunnel, Germany;
Wesertunnel, Germany; Westerschelde Tunnel, Netherlands; St. Petersburg Red Line, Russia; Nara
Prefecture Water Conveyance Tunnel, Japan; and Tokyo Wan Aqua Line Tunnel, Japan see Table 1.

1
Table 1. Summary of data from selected projects with high pressure tunneling and interventions (modified from Holzhuser et.al 2006)

2
Since then, over 15 projects have had TBMs designed to allow tunneling at groundwater pressures
at 7 bar or more including: Brightwater Central BT3, Seattle (8 bar); Brightwater Central BT3C, Seattle
(7.3 bar); Arrowhead Tunnels, California (22 bar standstill, <5 bar tunneling); Shanghai, China (7 bar);
Nanjing, China (7 bar); Lee Tunnel, London (8 bar); Hallandsas, Sweden (13 bar); Katzenberg Tunnel,
Germany (9 bar); Croydon Cable Tunnel, London (8 bar); Alaskan Way, Seattle (7 bar EPB tunneling, 10
bar emergency); Melen 7 Bosporus Crossing Tunnel, Turkey (13.5 bar standstill, 4 bar EPB); LMI- 3,
Nevada (17 bar); and Eurasia Tunnel -Istanbul Strait Crossing, Turkey (12 bar). Anagnostou 2014
provides a discussion of risks and tunneling for several of these projects.
Of the more recent projects, two tunnel projects had the longest lengths bored at face pressures of
7 bar or more: Lake Mead Intake No. 3 having ~ 2310 m or 52 percent of the total tunnel; and Eurasia
Tunnel having ~2300 m or 69 percent of the total. Data and experience with pressurized face tunneling
on these projects are also summarized in Table 1 and are discussed in more detail below.

LAKE MEAD INTAKE NO. 3


Lake Mead Intake No. 3 Planning and Design Factors
Planning was a critical element of risk management for high pressure tunneling at Lake Mead. Once
the locations of the access shaft and intake structure were determined, a geologic desk study and
phased subsurface investigations were completed to allow assessment of tunneling risks along 3
alignments (Figure 1). The shorter alignments (1 and 2) would have involved high pressure tunneling up
to 14-17 bar depending on lake level, within very poor quality, high permeability metamorphic rock for
over half of the alignment lengths. The risks and estimated costs for cutter change interventions along
these alignments were considered excessive. Alignment 3 was expected to have over 80 percent of the
tunnel within Tertiary sedimentary weak rocks (Muddy Creek and Red Sandstone formations) which had
adequate strength and generally low permeability. Although significantly longer, alignment 3 was
selected because of better geology and anticipated lower risks. In hindsight, the decision to select
Alignment 3 was appropriate. The Muddy Creek and Red Sandstone formations proved to be favorable
for tunneling in mostly open mode or low pressure closed mode despite groundwater heads up to 14
bar.
A decision that was later found to have significantly increased tunneling and intervention risk was to
allow design-build tenderers to modify the selected Alignment 3 and relocate the tunnel anywhere
within a delineated corridor see green shaded zone in Figure 2. The final borings, the Geotechnical
Baseline Report, specifications were focused on risks along the selected alignment and did not fully
address the risks for an easterly alignment within the corridor.
The selected design-builder, Vegas Tunnel Constructors (VTC) planned the blue (middle) alignment
shown in Figure 2. It shortened the tunnel length, but placed it within more uncertain and potentially
much poorer ground conditions. To help mitigate this risk, the owner, Southern Nevada Water Authority
(SNWA) completed a supplementary subsurface investigation program consisting of five, deep lake
borings while VTC completed an additional land boring. The additional lake borings found that
extensions of the poorer quality Precambrian metamorphic and Tertiary volcanic rock extended into the
planned tunnel zone. The additional land boring which was located approximately 300 m from the
Access Shaft, indicated much poorer than expected Precambrian metamorphic rock with an unknown
extent. Additional land borings were proposed to reduce uncertainties and risk but were not completed.

3
Figure 1. Lake Mead Intake No. 3 alignment alternatives considered (MW-Hill TM-4)

During conventional drill-blast tunneling for


the starter tunnel, a very wide, major high angle
fault zone was encountered. The fault zone and
high groundwater head eventually led to an
inundation of the starter tunnel and launch
chamber in 2010. After over a year of effort to
recover the original starter tunnel with two
additional blow-ins, it was abandoned and a re-
aligned starter tunnel was constructed (Hunt et.al
2015).
The consequences of the fault zone and
original starter tunnel inundations were a delay
of nearly two years and over US$40 million in
cost (Nickerson et.al 2015b).
The new starter tunnel alignment resulted in
a more easterly intake tunnel alignment within
the initial third of the tunnel as shown in Figure 2
(red or right alignment). As a result, a much
higher percentage of tunnel had to be bored
within poor quality, faulted, high permeability Figure 2. Planned Alignment 3, allowed corridor,
Precambrian metamorphic rocks. The planning VTC selected and final tunnel alignments

4
phase decision to locate the tunnel along Alignment 3 to reduce tunneling and intervention risks was
compromised by the corridor option and resulting starter tunnel inundations, abandonment and
realignment.

Lake Mead Intake No. 3 High Pressure Tunneling


After the realigned Starter Tunnel was completed and the TBM was assembled, high pressure TBM
tunneling began in December 2011. The TBM was a Herrenknecht dual mode machine that operated as
a slurry mixshield in closed mode and as a single shield TBM in open mode (McDonald & Burger 2009
and McDonald et.al 2010). A key risk mitigation element of the open mode configuration was a
horizontal screw conveyor that could be retracted within minutes to seal the excavation chamber should
a high inflow condition be encountered. Conversion from open to closed mode or back occurred over 10
times during TBM tunneling.
The ground conditions anticipated and baselined in the GBR for Alignment 3 and those encountered
on the actual bored alignment are listed in Table 2. The realigned tunnel resulted in 4.2 percent more
tunnel within the very poor quality Upper Plate (Pcu) unit and 3.4 percent more tunnel in the very poor
quality Saddle Island Volcanics (Tvsi) unit.
Figure 3 shows ground
conditions along the tunnel and Table 2. Summary of geologic units: GBR Alignment 3 and actual
that the TBM was launched in
low pressure closed mode
within the Lower Plate (Pcl) unit
which was locally dewatered by
inflows into the Starter Tunnel.
The progression of TBM
tunneling after launch is
explained in Nickerson et.al
2015a and 2015b.
After an advance of about
135m, the excavation chamber
pressure was raised to near
hydrostatic of 13 bar based on
Lake Mead level to provide groundwater control and face stability for tunneling through the extremely
poor ground conditions within the Detachment Fault (DF). These fault zone conditions were similar to
those that had caused the original starter tunnel inundation in 2010 (Hunt et.al 2015). The TBM bored
through the Detachment Fault and adjacent very poor quality rock without any occurrences of over-
excavation, blocky ground face instability, squeezing ground or steering problems.
The most significant impacts from boring through the Precambrian rocks (Pcl, DF and Pcu) under
approximately 13-14 bar of hydrostatic pressure were abrasion and lack of a natural safe zone that
would allow interventions to inspect and cutters. The contractor observed that when they operated the
TBM at excavation chamber pressures less than hydrostatic, the bentonite slurry became diluted and
abrasion of the mucking system works (chamber intake lines, pumps, slurry lines, and slurry line elbows)
was more severe. Even though much of the Lower Plate and Upper Plate rock would have been stable
for tunneling at pressures less than full face pressure, it was required to help minimize abrasion.

5
Figure 3. Geologic profile along actual alignment, TBM mode and excavation chamber pressure

After advancing the TBM to Ring 234 or Station 18+65 ft, a distance of 430m without cutter changes,
TBM parameters indicated a significant cutter wear and need to inspect and change worn cutters
(Nickerson et.al 2015a and 2015b). The face pressure was reduced to assess if a free air intervention
was possible, but indicated potential inflows of over 2000 m3/h (500 l/s or 8800 gpm). VTC elected to
complete a cutter change intervention after grouting from the TBM rather than by a saturation diving
intervention. This decision was partly due to a helium shortage resulting in an inadequate supply of tri-
mix gas, but was mostly due to saturation dive risk concerns. The highest pressure saturation dive
intervention to date (2012) had been 8 bar on the Westerschelde Tunnel in Germany (Holzhuser et.al
2006). The potential health and safety impacts a full 14 bar intervention were considerable. The risks of
trapping workers in the excavation chamber if a blow-in or face instability occurred during a partial
pressure intervention were considered unacceptable (Nickerson et.al 2015a and 2015b).
An extensive grouting campaign was implemented to reduce heading inflows to a manageable level
- Nickerson et.al 2014 and in Nickerson et.al 2015a and 2015b. After 4-1/2 months of grouting and two
TBM advances into the grouted zone, the heading inflows at free air were reduced temporarily to 220
m3/h (61 l/s or 965 gpm) - a reduction of about 90 percent. TBM slurry line repairs and perimeter cutter
changes were initiated, but heading inflows progressively increased during the first two days to over 450
m3/h (125 l/s or 2,000 gpm). After 14 additional days of similar high inflows, a partial face collapse
occurred and cutter changes were stopped before the center cutters had been changed. The high
seepage pressures from 14 bar of head had caused progressive erosion of fines and grout in rock joints
resulting in a very high hydraulic gradient near the cutterhead and eventually face instability. High
seepage pressures associated with high groundwater heads makes grouting to complete a free air

6
intervention less reliable than if the same grouting effort were completed at lower hydrostatic pressure.
This risk should be recognized when planning and managing grouted ground interventions at high heads.
VTC pressurized the face back to nearly 14 bar then advanced the TBM out of the unstable grouted
zone. Soon they found metal fragments at the separation plant an indication that the center cutters
were breaking. Fortunately, a heading pump test found an isolated zone of lower permeability rock that
allowed atmospheric access at inflows of about 950 m3/h (~264 l/s or 4180 gpm). VTC proceeded with
an atmospheric intervention without grouting. Within two days the inflow rates had increased to
approximately 1000 m3/h (278 l/s or 4400 gpm) and ranged from that value to 1090 m3/h (303 l/s or
4800 gpm). After access was achieved and inspections completed, severe abrasion damage was found at
the center cutters, center cutter housings and portions of adjacent cutterhead (Nickerson et.al 2015a
and 2015b). A repair cavern had to be mined and the high inflows channeled around it. Fortunately, the
face remained stable for the 47 days required to finish repair the cutters, housings and cutterhead.
In summary, the decision to grout and complete a free air intervention took 222 days (32 weeks) to
complete during which the TBM had advanced only 35m (114 ft). In high permeability, high head
ground, grouting to complete a free air excavation chamber intervention for cutter and cutterhead
repairs is often an expensive and time-consuming operation that should be compared to the cost and
risks for both hyperbaric interventions and surface grouted safe zones that the TBM could bore into.
High pressure tunneling continued through the Upper Plate rocks until June 17, 2013 when the TBM
had finally bored into the Tertiary Muddy Creek formation. After mining about 15m into the low
permeability Muddy Creek, open mode tunneling was initiated (Figure 3). Generally stable face
conditions and low inflows of 10-35 m3/h (3-10 l/s or 44-154 gpm) were encountered during open mode
tunneling in the Muddy Creek. The contractor found that when heading inflows exceeded about 23 m3/h
(6 l/s or 100 gpm) muck spillage at the screw conveyor discharge onto the belt conveyor created an
excessive mess and unsafe work conditions. When these conditions developed, they generally switched
back to slurry shield tunneling at low face pressure (Figure 3).
High pressure tunneling was resumed around Station 80+70 (ring 1247) as a precaution for
tunneling below the Las Vegas Wash. High pressure tunneling was continued until around Station
136+88 (ring 2204) on October 21, 2014 before resuming open mode tunneling. High pressure slurry
tunneling was resumed on October 30, 2014 at Station 142+56 (ring 2298) due to decreasing cover and
proximity to the Callville basalt formation (Figure 3).
High pressure tunneling (around 7-9 bar or 1-2 bar below lake pressure) resumed until the tremie
concrete surrounding the Intake Structure was encountered. The TBM proceeded into the tremie
concrete with a cover of only 3.7 m (12 ft) at full lake pressure (8.8 bar). Breakthrough into the flooded
Intake Structure successfully occurred on December 10, 2014 (Nickerson et.al 2015a and 2015b).
None of the many, but generally small fault zones encountered from entry into the Muddy Creek
formation through breakthrough into the Intake Structure presented any problems for TBM advance.
These faults had no influence on tunneling except where the fault zone and adjacent rock had high
enough permeability to cause excessively sloppy conditions within the TBM when advancing in open
mode resulting in a need to convert to slurry mix-shield mode.
In summary, 2310 m or 52 percent of the Lake Mead Intake No. 3 intake tunnel was bored at face
pressures of 7 bar or more. Approximately 1300 m or 30% of the tunnel was bored at face pressures of
12 bar or more, which was a world record. The intake tunnel was successfully completed through
extremely faulted, very poor quality, faulted Precambrian hard rock under 14 bar of hydrostatic pressure

7
and through lower permeability, weak sedimentary rock having buried ridges of volcanics and
Precambrian rock.

EURASIA TUNNEL
Tunnel and TBM
Eurasia Tunnel consisted of a 3,340 m long bore using a 13.66 m outside diameter Herrenknecht slurry
mixshield TBM see data summary in Table 1. Precast concrete segmental rings with a 12.0 m ID were
installed for the lining (Clark et.al 2015 and Zlatanic et.al 2015). The tunnel was bored below the
Istanbul Strait in Turkey at maximum depth of 106 m below the sea Figures 4-5. The tunnel was
advanced through inter-layered siltstones/mudstones and sandstones of the Trakya Formation and a
zone of soft ground soils.

Figure 4. Eurasia Tunnel profile (Anon 2012)

Figure 5. Eurasia Tunnel face pressures (after Burger 2015)

8
The TBM was designed for maximum face (excavation chamber) pressure of 12 bar (Bappler 2012).
Actual measured faced pressures ranged from 3 to 10.8 bar Figure 5 (Burger 2015). Approximately
2300 m or 69 percent of the drive was completed at face pressure over 7 bar.
The TBM cutterhead had six man accessible spokes with and special lock systems to allow disc cutter
changes under atmospheric conditions (Bappler 2012). The spoke access allowed the 19-inch disc
cutters to be changed without hyperbaric interventions or grouting. In addition the TBM was equipped
with a disc cutter rotation monitoring system (DCRM) to determine when disc cutters needed changing.
In addition, wear detector bars were placed on the cutterhead. This combination of cutter and
cutterhead wear monitoring with atmospheric access for cutter changes helped to keep the cutterhead
in good condition and avoid the center cutter, cutter housing and cutterhead wear problem experienced
on Lake Mead Intake No. 3. While spoke access solved the disc cutter intervention challenge, it did not
preclude the need for saturation dive interventions for scraper cutter changes and other excavation
chamber maintenance work.

Hyperbaric and Saturation Dive Interventions


Hyperbaric and saturation dive interventions were completed at pressures ranging from 4.2 to 10.8 bar
during 8 deployments. The hyperbaric work was performed to: change scraper cutters; close the
submerged wall opening; repair slurry pipe leaks, repair stone crushers and suction intake screens and
other items (Carl 2015). Hyperbaric interventions occurred on a total of 39 days and involved over 5,000
man-hours Table 3. Except for deployment 1 which involved compressed air, the other interventions
were completed with mixed gas and saturation diving resulting in a new record of 10.8 bar for manned
intervention pressure.

Table 3. Summary of Eurasia Tunnel hyperbaric intervention data (Carl 2015)


Intervention Date Ring Hyperbaric Work Time living under Time working
Deployment Pressure, bar Days pressure, hrs under pressure, hrs
1 June 2014 126 4.2 12
2 Jan 2015 875 10.8 11 1065 243
3 Feb 2015 933- 10.4 8 1086 147
962
4 Mar 2015 1003 10.3 5 612 92
5 Mar 2015 1076 - 1 230 8
6 April 2015 1117 8.9 3 481 42
7 April 2015 1147 8.5 10 1016 184
8 May 2015 1147 8.5 1 269 24
Totals 39 4759 740

SUMMARY AND CONCLUSIONS


Since 2006, TBMs on over 15 projects have been designed to tunnel at face pressures ranging from 7 to
17 bar. Many of these TBMs experienced less pressure or very short advance lengths at high pressure,
e.g. Hallandsas. Two projects, Lake Mead Intake No. 3 and Eurasia Tunnel had tunnel drives completed
in 2014 and 2015 with record setting lengths (> 2300 m) of tunnel completed at pressures over 7 bar
and at maximum pressures of 14 and 10.8 bar, respectively. On both projects, high permeability, low

9
quality faulted rock conditions were handled by the TBMs without serious impacts such as face collapse,
squeezing and stuck advances.
TBM manufacturers have developed equipment that can handle very high pressures, however this
equipment requires maintenance and repairs when components wear or break. The biggest challenges
for tunneling in poor quality, high head ground do not appear to be TBM advance, but instead how to
create suitable, safe conditions for access to the excavation chamber and cutterhead for maintenance
and repairs.
Options for facilitating access to the excavation chamber and cutterhead include:
Hyperbaric interventions with compressed air up to 4 or 4.5 bar and mixed gas saturation diving
at higher pressures.
Spoke cutter access for larger TBMs (>12-13 m diameter) this only provides disc cutter change
access and does not provide access to change other cutters and complete maintenance work.
Grouting from the TBM heading to create a lower permeability zone suitable to reduce inflows
to a manageable, safe level.
Pre-excavation grouting from the surface to create a lower permeability safe zones.
Closer spaced, more thorough subsurface investigations (borings and water pressure tests) to
identify suitable natural or more readily grouted safe zones for interventions.
Saturation dive interventions can be very expensive and risky. North American contractors are
generally more reluctant than global contractors to utilize hyperbaric interventions as the primary
method of access. Even if not used as the primary means for interventions, hyperbaric intervention
capability should still be required for backup.
Grouting is an option to consider for primary access, but grouting is often unreliable and can be very
expensive, particularly from a TBM. The uncertainties of grouting effectiveness must be managed.
Based on the experience to date, more funding and effort should be given to more extensive
subsurface exploration programs that are designed to find and characterize natural low permeability
safe zones for potential interventions. More extensive subsurface investigation programs may save
considerable expense during tunneling from use of hyperbaric or grouted interventions.

REFERENCES
Anagnostou, G. 2014. Some Critical Aspects of Subaqueous Tunnelling. Muir Wood Lecture 2014. ITA-
AITES, 20p.
Anon. 2012. Eurasia Tunnel, The most advanced transportation infrastructure connecting the
continents. Republic of Turkey. 9p
Bppler, K. 2014. Constant Demand on Very Large Tunnel Boring Machine Diameters for the
Construction of Todays Infrastructure Systems. In: Proceedings of 2014 North American Tunneling
Conference. SME, Littleton, Colorado, USA. N2014.19. 160-165.
Burger, W. 2015. Chamber Interventions and High Pressure TBM Operation, In: Ozdemir (ed.)
Presentations and Proceedings of 2015 Tunneling Short Course, September 14-17, 2015, Colorado
School of Mines. 27p.
Carl, V. 2015. Istanbul Strait Road Tube Crossing Project S-762, Hyperbaric Work - Project Report,
Nordseetaucher GmbH. 7p.

10
Clark, G., Castelli, R., Richards, D., Toan, T., Arolu, B. & Gnen, O. 2015. Elements of the Istanbul
Strait Highway Tunnel. In: Proceedings of 2015 Rapid Excavation and Tunneling Conference. SME,
Littleton, Colorado, USA. R2015.19. 223-239.
Holzhuser, J., Hunt, S. W., Mayer, C. 2006. Global Experience with Soft Ground and Weak Rock
Tunneling under Very High Groundwater Heads. In: Ozdemir (ed.), North American Tunneling 2014,
Taylor & Francis Group, London. 277-289.
Hunt, S.W., Smith, E. & Moonin, E. 2015. Characterizing and Baselining Faults for Tunneling. In
Proceedings of 2015 Rapid Excavation and Tunneling Conference, SME, R2015.49. 600-615.
McDonald, J. & Burger, W. 2009. Lake Mead TBM designed for the extreme, Tunnel Talk, Nov 2009, 3p.
McDonald, J.A., Hurt, J. & Moonin, E. 2010. Lake Mead Intake No. 3. In Proceedings of 2010 World
Tunnelling Congress, ITA, W2010.54. 8p.
Nickerson, J., Ceccato, R., Bono, R., Cimiotti, C. & Fioravanti, P. 2014, Lake Mead - Intake Tunnel No. 3,
Pre-excavation grouting challenges using a high pressure slurry TBM. Geotechnical News, September
2014, 32-38.
Nickerson J., Bono, R., Cimiotti C. & Moonin E. 2015a. Lake Mead Intake No. 3 - TBM Tunneling at high
pressures. In Proceedings of 2015 World Tunnelling Congress, ITA, Session 4.36, W2015.233. 12p.
Nickerson J., Bono, R., Cimiotti C. & Moonin E. 2015b. Lake Mead Intake No. 3TBM Tunneling at High
Pressures; In Proceedings of 2015 Rapid Excavation and Tunneling Conference, SME, R2015.77. 936-
946.
Wallis, S. 2015 Bosphorus Eurasia highway TBM breaks through, TunnelTalk, 22 Aug 2015. 3p.
Zlatanic, S., Ozturk, M. & Munfah, N. 2015. Istanbul Strait Road Tube Crossing Project Uses Good
Practices to Manage Risks. In: World Tunnelling Congress 2015. ITA, No. 1.60, W2015.60. 19p.

11

S-ar putea să vă placă și