vancouver 19 th nc

TUNNELLING ASSOCIATION OF CANADA19th National Conference

Conference Proceedings

ˆTunnelling Towards 2010˜Vancouver, B.C. Canada

September 17−20, 2006Marriott Pinnacle Hotel

i

19th 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 Japan...................................................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. O’Connor, 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 world’s 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 Canada’s 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 Rim South 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 System make 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. Janosko & 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 Durham Sewage System 19th 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 PROJECT 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 inspector’s 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 Sacramento’s 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 designer’s 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 from earlier 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 from earlier mapping by Helley and Harwood, 1985 [3] and Schlemon, 1995 [4].

Fig. 7. High river levels at the NSRC in January 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 contractor’s 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 June 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.

Jun-05

Jul-05

Aug-05

Sep-05

179+

0018

0+00

181+

0018

2+00

183+

0018

4+00

185+

0018

6+00

187+

0018

8+00

189+

0019

0+00

191+

0019

2+00

193+

0019

4+00

195+

0019

6+00

197+

0019

8+00

199+

0020

0+00

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

901+

0090

2+00

903+

0090

4+00

905+

0090

6+00

907+

0090

8+00

909+

0091

0+00

911+

0091

2+00

913+

0091

4+00

915+

0091

6+00

917+

0091

8+00

919+

0092

0+00

921+

0092

2+00

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.

Figure 21 presenting the settlement of DMP-34 at 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 Station 195+75 represents a total surface settlement of 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.

Figure 23 presenting the settlement of DMP-T6 at

Station 198+00 represents a total surface settlement of 73 mm (0.240 ft.).


It can be seen that 8 mm (0.025 ft.) of settlement 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.

Figure 24 presenting the settlement of DMP-T7 at Station 199+00 represents a total surface settlement of 40 mm (0.130 ft.).

It can be seen that 9 mm (0.030 ft.) of settlement 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


Figure 25 presenting the settlement of DMP-43 at 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 Proceedings of the Rapid Excavation and Tunneling 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 County’s Wastewater Treatment Division is constructing a new regional wastewater treatment plant with associated pipelines and other facilities as major improvements to the County’s regional sewage treatment system. The new $1.5 billion system has been named “Brightwater”.

King County’s 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 km2 (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 Brightwater’s 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, GASKETEDCONCRETE SEGMENTALLINING (0.25 m (10") MIN.)

1.68 m (66")INFLUENT

FORCE MAIN1.22 m (48")INFLUENT

FORCE MAIN

2.13 m (84")EFFLUENTPIPELINE

0.69 m (27") RECLAIMEDWATER PIPELINE

CELLULAR CONCRETEBACKFILL


5.08 m (16' 8") MINIMUM

FIBER OPTIC CABLES 1.83 m (72")EFFLUENTPIPELINE

1.37 m (54")INFLUENTPIPELINE



BOLTED, GASKETEDCONCRETE SEGMENTALLINING (0.33 m (13"))

FIBER OPTICCABLES

STRUCTURALCONCRETE

4.37 m (14' 4")

FIBEROPTICCABLE

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")






0.36 m (14")RECLAIMED

WATER PIPELINE

0.36 m (14") RECLAIMEDWATER PIPELINES

EFFLUENTTUNNEL

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')


EFFLUENTTUNNEL


BACKFILL CONCRETE

STORM WATERSTORAGE 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 manufacturer’s 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 County’s 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 Jay Dee/Michels/Traylor JV in 2003 and construction commenced the same year. The BWOAS contract (Part 2) followed and was awarded to the McNally/ Kiewit JV 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 Joint 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 TBM’s 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 owner’s 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 (¼ cubic yard) batches and transferred to a 1.5 m3 (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 1.5m3 (2cy) per ring, a 3m3 (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 owner’s 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 January 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 owner’s engineer may want to limit the owner’s risk by requiring that the contractor’s 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/Owner’s 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 Hatch Mott MacDonald, Pleasanton, CA, USA

David J. Young Hatch Mott MacDonald, Pleasanton, CA, USA

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 ¾ 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 ¾ 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, John 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 June 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. Japan Society of Civil Engineers (JSCE), 2000. Japanese 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 June 2006, ed. L. Ozdemir, pp.437-443, London: Taylor & Francis Group

7. Jurich, 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 June 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: [email protected]

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 m3/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 m2 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

Darcy’s 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 (Nf / Nd) A (1)

where:

Q = volume of water inflow

k = rock mass permeability

H = total head

Nf = number of flow channels

Nd = 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, m3 /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 m2 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, m3/sec

Q p = water inflow into the pilot tunnel, m3/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

Darcy’s 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. Darcy’s 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

Ele

vat

ion,

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 Corté 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, Japan

Qiang Li, Ph.D. Software Engineer, Allrightsoft Co., Tokyo, Japan

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.

SandLDT

Air pressure

TunnelFace

LiningPPT

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/cm3, with grain size D50=0.19mm, D30=0.16mm and D10=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/cm3.

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 K0=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, σs, at the face collapse

C/D : 2 4 6 σS [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 7C/D

σ/γD

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

Depressioncone

3xD

Figure 3. Sketch of the progressive failure envelope

42

A comparison with the centrifuge experiments reported by Chambon and Corté (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 6.9x107

ν 0.3 0.34 γ [kNm-3] 15.6 27 c [kNm-2] 0.2 - φ [deg] 42 -

Table 3. Supporting pressure, σs

C/D : 2 4 6 σS [kPa]

Centrifuge 3.6 2.6 5.3

σS [kPa] FEM

4.0 4.0 4.0

σS [kPa] Upper Bound

3.7 3.7 3.7

σS [kPa] Lower Bound

37 61 86

00.5

11.5

22.5

33.5

44.5

0 5 10 15 20 25 30

Supporting pressure, σm [kPa]

Saf

ety

Fact

or, S

F

C/D=2C/D=4C/D=6

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

Figure 6. Shear strain distribution

σs=4kPa SF=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, Géotechnique, 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., Corté, J-F., Garnier, J., König, D., 1991, Face stability of shallow tunnels in granular soils, Proc. Centrifuge ’91, Balkema, Rotterdam, pp.: 99~105

4. Chambon, P., Corté, J-F., 1994, Shallow tunnels in cohesionless soil: Stability of tunnel face, Journal 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, Géotechnique, 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, Géotechnique, 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”, Géotechnique, 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, Japan 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(EI6)R-(1 E 32

m

mF

νν+

= (1)

where: Em is the modulus of elasticity of the medium

E is the modulus of elasticity of the liner R is the radius of the liner νm is the Poisson’s ratio of the medium ν is the Poisson’s 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 Poisson’s ratio of the medium according to the following relationship [2]:

)1(2dmax mD

νγ −±=∆

(2)

where: Dd∆

is the diametric distortion of the liner

maxγ 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


Metro Caracas Linea 3 Metro Caracas,

Venezuela

5.8m (19 ft) OD


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 Joints

JSCE [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 January 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, Japan, 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, Japan, 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. Joint 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 Poisson’s 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 Proceedings of the Rapid Excavation and Tunneling 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. Journal of Geotechnical and Environmental Engineering, Vol. 129, Number 7, Paper number 575, July.

5. Wang, Y.N., B. Hughes, H. Caspe, and M. Amini. 2006. Devil’s Slide Tunnels – Caltrans 1st 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. Barcelona’s new backbone runs deep, (no page number). March.

11. Tunnels and Tunnelling International. 2003. Multiple TBM action on Turin’s 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. Japan Society of Civil Engineers (JSCE). 1995. Preliminary Report on the Great Hanshin Earthquake, January 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 Taipei’s 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/2002/20020331.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 1980’s, 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

James Tan Program Manager – 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 Calgary’s 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 Job, the project team decided to proceed with the following approach:

1. Start tunnelling at 15th Street with working shaft configuration C.

2. Construct shaft at 20th 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 20th street toward 18th 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

Scen

ario

# of

TB

M

Shift

Dur

atio

n (h

rs)

# of

Shi

fts/d

ay

# W

orki

ng

Day

s per

wee

k

Tunn

ellin

g C

ompl

etio

n D

ate

(mm

/dd/

yy)

Proj

ect

Com

plet

ion

Dat

e (m

m/d

d/yy

)

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 PROJECT 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 January 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 Jan 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 User’s 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. Joines, 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 June 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, Japan.

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 Hatch Mott MacDonald

ABSTRACT: Contract C710 is currently under construction as part of Sound Transit’s 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 Jacobs 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. Junctions 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. Platform Shift 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. Jet Grouting Jet grouting was carried out from the surface and targeted zones of sand within the tunnel profile as pretreatment for the SEM tunnels. Jet 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 Johnson/Soletanche JV was awarded the subcontract in 2004. After some initial test columns late in 2004, the production work continued between January 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 m3 (5500 yds3) 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 Jet 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/Jacobs Joint 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. ST’s 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 Journal 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 June 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 Obayashi’s 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. Platform Tunnels 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 Obayashi’s 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 m3/hr (60 cubic yd/hr) volumetric batcher and it is planned to be augmented with a second 76 m3/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 Jacobs. 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

0.05.0

10.015.020.025.030.035.040.045.050.055.0

1 2 3 4 5 6 7 8 9 10 11 12 13

Week

Hou

rs

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, Joe Rotzien Hatch Mott MacDonald, Vancouver, B.C., Canada

Christian Genschel, Josef 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 Engineer’s 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-January 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 Contractor’s staff that shaft sinking was completed with no major accidents or injuries.

2 PROJECT 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 January 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 m3 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 m3 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 m2 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 m3 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 m3 capacity sump pit and two muck pits each having a volume of 125m3 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 Mühlhauser 5 m3 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 m3 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 June 2006. Figure 8 shows the first TBM assembled in the shaftbase. TBM tunneling is anticipated to begin in mid-June 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 O’Connor, 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 m3/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.

t

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.

76

Note: The originally planned Low level Outlet stilling basin was changed to a flipbucket.

Figure 1 – Plan of San Roque Projec

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 unitsShaft 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 ................. 800mLow Level Outlet Tunnel


length.............................. 1400mLow Level Tunnel Gate Shaft


depth............................... 100mSpillway 6 gates each 15m x 15m x 19m highChute 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 SuperSwellex® expandable 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 1970’s 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. O’Connor, E. 2001, “Diversion Schemes for the San Roque Multipurpose Project” Waterpower XII, Salt Lake City. 3. Humphries, R., E. O’Connor, 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”, Journal 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 Intake 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. Jean 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 jumbo’s 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 jumbo’s

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 m3) 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 EBC’s initial budget.

86

1 INTRODUCTION

Toulnustouc project is located in the north-eastern part of province of Québec, 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 July 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, Montréal, QC, Canada

A. Whalen, D. Bergeron Hydro-Québec Équipement, Montréal, QC, Canada

ABSTRACT: The Hydro-Québec Toulnustouc hydroelectric project is located approximately 150 km north of the city of Baie-Comeau, Québec, and was commissioned in July 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

PM 3000

5 540 000 N

266

000

E

PM 1000

PM 2000

264

000

E

5 538 000 N

268 000 E

PM 4000

PM 5000

262

000

E

260

000

E

258

000

E

256

000

E

PM 9000

PM 0

PM 6000

PM 7000

PM 8000

PM 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 Index (GSI) (%)

Young’s modulus (GPa)

Joint 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. Jacked rock mass permeability Jacked

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:

( )bh

hhkQ i

0

0

2ln2 −

=π (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, hi internal water pressure, h0 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)

Wat

er le

akag

e (l/

sec)

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.

-

110120130140150160170180190200210220230240250260270280290300

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

Wat

er le

vel i

n th

e tu

nnel

(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)

ELE

VA

TIO

N (m

)

RELIEF HOLES

TOULNUSTOUC RIVER

PRESSURE TUNNEL

184.6

-100 -80 -40-60 -20 1400 20 40 60 10080 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

86

PM

771

0

PM

772

0

PM 7

7 30

PM 7

7 40

PM

77 5

0

PM 7

760

PM 7

770

PM 7

780

PM

77 9

0

PM

781

0

PM 7

820

PM 7

830

PM

784

0

PM

785

0

PM

786

0

PM

787

0

PM 7

880

7471

.5

7 491

.5

7511

. 5

7531

.5

7571

. 5

7591

.57 611

.5

7 631

.5

1

2

3

4

5

6

789

10

11

12

13

1415

1617

18

19

20

21

22

23

24

25

2627

28

293031323334

3536373839

40

41

42

4344

45

46

47

48495051

52

53

54

5556

57

58

59

6061626364

65

66

67

68

69

7071

72

73

7475

7677

78

79

8081

82

83

84

85

TF-06

TF-05

TF-03

TF-08

TF-07

TF-13

TF-10

TF-09

TF-11TF-127.6 m

TF-14

TF-15

TF-16

TF-17

18.8 m

20.6 m260 000 E

260 200 E

259 800 E

5 537 600 N

P.I.-31

210

220

230

240

250

260

270

280

230

240

250

230

240

250

260

270

260270

280

270

280

5 537 800 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

14/12/2005

21/12/2005

28/12/2005

04/01/2006

11/01/2006

18/01/2006

25/01/2006

01/02/2006

08/02/2006

15/02/2006

22/02/2006

01/03/2006

08/03/2006

15/03/2006

22/03/2006

29/03/2006

05/04/2006

12/04/2006Days

Res

ervo

ir le

vel (

m)

38

43

48

53

58

Wat

er le

akag

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)

Tota

l flo

w (l

/sec

)

245

250

255

260

265

270

275

280

285

290

295

Wat

er le

vel i

n tu

nnel

(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.

01020304050607080

240 250 260 270 280 290 300Water level in the tunnel (m)

Ave

rage

leak

age

(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. Aménagement Capivari-Cachoeira, Étude 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. Javandel, 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. Am Soc 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 Sydney’s 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 Norway’s Romeriksporten Tunnel and Sweden’s Hallandsås 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 JUMBOS 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 Sandvik Mining and Construction Oy, Tampere, Finland

Pekka Salminen Sandvik Mining and Construction Oy, Tampere, Finland

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 Sandvik’s 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 1980’s. 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.

Sandvik’s 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 rig’s 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 Sandvik’s 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.

Sandvik’s 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, June 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 Lötschberg and Gotthard axes.

Linking Basel and Milan via Berne and the Sim-plon line, the Lötschberg 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 FinÖV (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 45°C were measured at tun-nel level in the Amsteg section, where the overburden depth under the Chrüzlistock summit rises to a full 2187 m. These temperatures exceed the predicted value of 38°C and the projected ±5°C fluctuation range.

To protect the health of tunnelling crews, the SUVA (Swiss National Accident Insurance Fund) has prescribed a maximum working temperature of 28°C 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 50°C 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 TBM’s 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 TBM’s, 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 section’s 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 125th 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 Lösungen (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, 84–95.

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 Lösungen beim Bau des längsten 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 Lombardi’s 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 Mätteli). 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 Terzaghi’s [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

oc p

ppe

HCs ∆++

= log1

(1)

Here, s is the vertical settlement, Cc the compression index, Ho the initial thickness of the compressible stratum, eo the initial void ratio, and po 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 Rankin’s 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. Solution’s often invoke Biot’s [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 Biot’s 3-D consolidation and related effective stress laws [27], into a finite-element formulation and it’s 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

Young’s modulus and Poisson’s ratio, Biot’s and Skempton’s 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

nn k

u'σ∆

=∆ (2)

where ∆un is the change in normal deformation of the fracture, ∆σn′ is the change in effective normal stress and kn 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 “Poisson’s 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, φj, 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 Lützenkirchen [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 tunnel’s 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 Symposium on 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 Janeiro, 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. Hårvik, 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 Journal 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 Journal 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 Journal 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. Dam Engineering 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 Lötschberg, Zurich, eds. S. Löw 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. Löw. 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. Särkkä 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. Journal 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. Journal 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

1 THE NIAGARA RIVER – WATER DIVERSION AND GENERATION CAPACITY

1.1. IntroductionThe Niagara River is an international waterway forming part of the boundary between Canada and the United States of America. The river is about 53 kilometers (km) in length with an average flow of approximately 6000 cubic metres per second (m3/s).

The hydroelectric resource at Niagara is shared with the United States of America in accordance with the terms of the 1950 Niagara Diversion Treaty. This treaty established priority for scenic, domestic and navigation purposes and allows the remaining flow to be used for power generation. The scenic flow requirement is 2832 m3/s during the daytime from April through October and 1416 m3/s at all other times. About two-thirds of the average Niagara River flow is available for power generation and is shared equally by Canada and the United States.

1.2. Developing More PowerIn the 1980’s, Ontario Hydro, the predecessor to Ontario Power Generation (OPG), began exploring the possibility of 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 of Environment in October 1998. The approved project included construction of two additional diversion tunnels and an underground generating station north of the existing SAB generating stations.

Table 1. Diversion and Generation Capacities

In-Service Year

Diversion Capacity(m3/s)

Station Capacity(MW)

Annual Energy(GWh)

SAB 1 1922 625 487 2,700SAB 2 1954 1,200 1,472 9,200SAB PGS 1958 - 122 100Current Totals 1,825 2,081 11,800Niagara Tunnel 2009 500 - 1,600Future Totals 2,325 2,081 13,400

In July 2004, OPG decided to proceed with the Niagara Tunnel Project, a design/build project for one of the diversion tunnels. This tunnel will divert a nominal 500 m3/s of 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.Project Manager, Hatch Mott MacDonald, Niagara Falls, ON, Canada

Ernst Gschnitzer, Ph.D. Project Manager, Strabag, Niagara Falls, ON, Canada

Rick Everdell, P.Eng.Project Director, Ontario Power Generation, Toronto, ON, Canada

ABSTRACT: The award of the design/build contract for construction of Ontario Power Generation’s (OPG) Niagara Tunnel Project was made on September 1, 2005, after an eight month international procurement process. The project requires delivery of a nominal 500 m3/s of water from the Niagara River from an intake located upstream from Niagara Falls via a 10.4-km tunnel running underneath two existing water delivery tunnels beneath the City of Niagara Falls, to an outlet at the existing Sir Adam Beck generating station complex. The tunnel will be excavated by means of the world’s largest hard rock TBM, a 14.44-m diameter open gripper machine. Vertical alignment of the tunnel is constrained by an ancient buried gorge at the north end and existing abandoned power generating facilities at the south end. Approximately 80% of 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 waterproofing membrane, as the final lining. Under the contract, subsurface geotechnical risk is assigned by means of a negotiated Geotechnical Baseline Report (GBR).

120

Station complex allowing more efficient use of the Niagara River flow available to Canada for power generation, facilitating an increase in average annual energy output of about 1600 GWh (14%), which is enough to supply the annual needs of a city of approximately 160,000 residents. The second diversion tunnel and underground powerhouse could be constructed in the future, depending on energy requirements and project economics.

The flow in the Niagara River that is available to Canada for power generation varies from about 1000 m3/s to 3000 m3/s and as indicated in Figure 1, exceeds the existing SAB diversion capacity (canal and two tunnels) about 65% of the time. With the addition of a nominal 500 m3/s diversion capacity from the new tunnel, the available flow will exceed SABs diversion capability only about 15% of 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 of the Niagara Tunnel Project which was to procure a contractor based on the following 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, therefore, 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 of distinct stages:• an international invitation for expressions of interest• a prequalification process• proposal preparation and submission• proposal evaluation and negotiation culminating in the award of a design/build contract.

Of the seven contractors that submitted expressions of interest, four were pre-qualified to submit proposals on the basis of a number of evaluation criteria that included relevant design and build experience and safety performance. Three proposals were received after a five month proposal period and this was followed by proposal evaluation and negotiations with all three proponents. Evaluation criteria included the design and construction approach, cost, risk profile, tunnel flow capacity, schedule, project team, health & safety management, environmental management and quality management.A recommendation of a design/build contractor was made to OPG Board of Directors in July 2005 and the project was approved by OPG and the Government of Ontario by August 17, 2005.

Phase 2, the detailed design and construction of the project, commenced on August 18, 2005, with the award of the design/build contract to Strabag AG of Austria with local sub-contractor, DufferinConstruction, assisted by designers ILF from Austria and Morrison Hershfield of Toronto.

2.1. Stakeholders to the Niagara Tunnel ProjectA number of stakeholders, as summarized in Table 2, will influence the successful completion of the project.

Table 2. Key Stakeholders

Stakeholder ResponsibilityOPG ShareholderGovernment of Ontario

Provide Direction to OPG and Financing

Owner/OperatorOPG/Niagara Plant Group

Provide Project Direction and Oversight

Operate and Maintain New Tunnel and Gates

Owner’s RepresentativeHatch Mott MacDonaldwith Hatch Acres

Administer Design/Build Contract

Review Design and Monitor Construction

Design/Build ContractorStrabag AG

Execute Design/Build Contract

RegulatorsMOE, MNR, DFO,NPCA

Monitor Compliance with EA Approval

Issue Permits/Certificates of Approval

Host MunicipalitiesNiagara RegionNiagara FallsNiagara-on-the-Lake

Manage Forecast Tourism Impacts

Provide Agreed Municipal Services

Issue Municipal Permits

The Ontario Government, as OPG’s sole shareholder and project financier, endorsed and approved the project as being consistent with its objective of promoting the development of cost competitive, environmentally friendly sources of

Niagara River - OPG Entitlement - Monthly Flow Duration CurvePeriod: 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 [%]

Flow [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

Flow [cfs]

Existing Canal and Two Tunnels

One Additional Tunnel

Two Additional Tunnels

121

electricity generation. OPG and its Niagara Plant Group (NPG) are the Owner and facility 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 from Austria was awarded the design/build contract. A number of regulatory agencies are responsible for monitoring the terms of the Environmental Assessment Approval and for issuing permits. In addition, the project is hosted by a number of local municipalities that include the Regional Municipality of Niagara, City of Niagara Falls and Town of Niagara-on-the-Lake.

A Community Impact Agreement (CIA) was signed in December 1993, between Ontario Hydro (now OPG), the Regional Municipality of Niagara, the City of Niagara Falls, and the Town of Niagara-on-the-Lake. The CIA outlines compensation for 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 effective).

3 DESIGN/BUILD CONTRACT

3.1. Mandatory Requirements and Liquidated Damages/Bonuses

Proposals for the Niagara Tunnel Project were based on an invitation for proposal document that mandated a number of key technical requirements including: • construction of a 10.4-km long TBM driven tunnel from an outlet on OPG property near the existing SAB generating stations, along a mandated horizontal alignment, to an intake in the Niagara River

• design life of the facility to be 90 years including measures to deal with swelling rock conditions in the local shale units

• tunnel to deliver a nominal 500 m3/s water flow.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 beneficially). This date was then used as the contractual substantial completion date for the successful contractor. Similarly, proponents were required to propose a guaranteed flow amount, based on the prevailing hydraulic heads and their selected tunnel diameter and final lining characteristics. These flow amounts were also taken into account during proposal evaluation (largest guaranteed flow amount was evaluated most beneficially) and this flow amount was then adopted as the contractual guaranteed flow amount. The contract includes bonuses for exceeding the guaranteed flow amount and for early substantial completion, and liquidated damages for failure to achieve the same.

3.2. Construction ScheduleA summary level project schedule is provided in Figure 2.

Fig. 2. Summary Schedule for the Niagara Tunnel Project

122

3.3. Geotechnical Baseline ReportThe geotechnical baseline report (GBR) was produced by means of 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 of the proposal evaluation with changes made during the negotiation process.

• the final negotiated GBR(C) (the GBR) was then included in the design/build contract.

3.4. Scope of WorkThe project comprises three major elements of work, namely, Intake facilities, Outlet facilities and Diversion Tunnel.

IntakeThe tunnel Intake facilities consists of a submerged reinforced concrete bell-mouth intake structure in the Niagara River, beneath Bay 1 of 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 for closure of the tunnel when required for dewatering. To facilitate construction of the Intake in the dry, a temporary cofferdam will be installed upstream from Bay 1. Extensive curtain grouting of the highly permeable bedrock formations will be required prior to drill and blast excavation down to tunnel invert elevation, a depth of approximately 40 m. A 5 mminimum 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 of the TBM. Depending on Intake works productivity, schedule, design, environmental and logistical considerations, a full diameter grouting gallery may be excavated thereby reducing critical path TBM tunnelling. Other work at the Intake includes removal of an existing 520-m long ice accelerating wall that extends upstream from Pier 4 of the INCW structure and installation of a parallel new precast concrete accelerating wall upstream of 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 of the Work

123

OutletThe tunnel Outlet consists of a reinforced 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 of the tunnel for emergencies and for dewatering. During construction, the outlet canal also acts as the assembly and launch site for the TBM and as the staging area for logistics to and from the bored tunnelling operation. A 40 m long rock plug, between the new and existing PGS canal, will be left in place for the duration of construction. One of the final construction activities will be removal of this plug at the time of watering up of the tunnel.

Diversion TunnelThe approximately 10.4-km long Diversion tunnel will be constructed as a two-pass tunnelling system with boring taking place from 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 from the TBM and trailing gear followed by an in-situ placed concrete lining after completion of TBM tunnelling. Approximately 1.7 million cubic metres of excavated spoil material will be transported from the TBM by conveyor belt and disposed on OPG property between the two existing power canals. Five dewatering shafts of 0.75 m diameter will be constructed at the lowest point of the tunnel to allowtunnel dewatering if required.

3.5. GeologyThe Niagara Region is underlain by Cambrian, Ordovician and Silurian sedimentary rocks having a total thickness of approximately 800 to 900 m. The Niagara River Gorge, the main physiographic feature, and the Niagara Escarpment control the topography of the project area. The generally flat lying bedrock strata consists of dolostone, dolomitic limestone, sandstone and shale in which the diversion tunnel will be excavated with about 80% of the tunnel length in the Queenston Formation, a siltstone/mudstone with an unconfined compressive strength ranging from 19 to about 45 MPa.

The present Niagara River Gorge was formed 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-filled with glacial outwash materials. Away from 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 from Lake Ontario through the village of St. David’s to the Whirlpool. The St. Davids Gorge is oriented in a northwest direction and varies in width from 350 to 630 m in the vicinity of the Niagara River. Depth to bedrock is in the order of 125 m in the vicinity of the proposed tunnel alignment and in excess of 200 m where it intersects with the present Niagara River at the Whirlpool. The gorge is completely in-filled with deposits of glaciolacustrine, glacial and glaciofluvial origin. The bedrock (Queenston Formation) over the width of the St. Davids Gorge is slightly weathered and relatively more fractured to a depth of between 15 to 25 m below the bottom of the gorge. Below this depth, the rock is generally fresh and of excellent quality.

Bedrock in the project area has generally well-defined bedding with a southerly dip of about 6m/km and an east-west strike. Sheared, weak bedding planes exist between many of the rock formations and within the Queenston Formation. There are no known occurrences of any major faulting within the project area, some near-surface thrust with minor vertical displacement are known to occur and are probably related to stress relief associated with the gorge formation and the high horizontal residual stresses in the area. Some shearing of this type is expected in the area of the St. Davids Gorge. Three major near-vertical joint sets, which strongly influenced the physiography of the project area, have been identified. These sets strike parallel to the Niagara River; the St. Davids Gorge and the Niagara Escarpment. Vertical joints are generally widely spaced. The joint surfaces are generally rough and fresh to slightly weathered.

3.6. Bedrock Stratigraphy and StructureIn descending order from surface, the sequence of rocks is as presented in Table 3.

The Queenston Formation extends well below the deepest section of the tunnel with thickness greater than 300 m being reported in the literature.

Primary bedding planes are defined as major bedding planes between lithological units above the Queenston Formation and between sub-units within the Queenston Formation. Sheared primary bedding planes refer to those planes where some differential displacement has occurred. Within the Queenston Formation, the primary bedding planes are major discontinuities occurring at spacing of about 5 m to somewhat greater than 20 m and locally affecting the rock mass quality.

124

Table 3. Rock Sequence to be Encountered During Tunnelling

Sequence Description ThicknessGuelph Dolostone 2 to 3 mLockport Dolostone 43 to 45 mDeCew Dolostone 2 to 3 mRochester Shale 17 to 19 mIrondequoit Limestone 2 to 4 mReynales Dolostone 3.5 to 4.5 mNeahga Shale 1.5 to 2 mThorold Sandstone 2 to 3.5 mGrimsby Sandstone 12.5 to 15 mPower Glen Shale 10 to 12 mWhirlpool Sandstone 4.9 to 8.5 mQueenston Shale/Mudstone >300 m

These planes often exhibit features such as gouge or breccia (a few millimetres to 2 to 3 cm) and slickensides that are consistent with lateral structural dislocation.

Groundwater conditions in the project area are influenced by depth and lithology, and vary between the rock formations above the Queenston Formation, but are relatively consistent in the Queenston Formation. The only known aquifers are the Lockport and DeCew (dolostone) Formations, whereas the remaining strata below the DeCew are generally considered to be aquitards. Hydraulic conductivity ranges from <10-7 to 10-3 cm/s in the upper dolostone and limestone formations and from 4 x 10-3 cm/s to practically impermeable (<10-7 cm/s) in the shale formations. In general, the groundwater below the DeCew Formation is highly corrosive.

Natural gas has been encountered in some of the formations, particularly in the Rochester and Grimsby Formations, with some minor amounts of gas being encountered in other formations, 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 from 10 to 24 MPa, with a maximum horizontal/vertical stress ratio varying from 3 to 5. In general, the orientations of the maximum horizontal stresses along the alignment of the diversion tunnel lie within the NE-SW quadrant. The orientations of the local stresses are influenced by the presence of major physiographic features, namely the buried St. Davids Gorge and the Niagara River Gorge.

The formations in the project area are subject to time-dependent deformations, initiated by the relief of the relatively high in-situ stresses and swelling on the uptake of fresh water. There is a well-

documented history of rock “squeeze” affecting surface excavations. The swelling potential of shale units in the Niagara area is also well documented. Swelling involves the volume increase in shale units and is initiated by the relief of the high in-situ stress in the presence of freshwater. The process is associated with chloride ion diffusion from the connate pore water in the rock.

4 TBM, TRAILING GEAR AND BACKUP EQUIPMENT

Strabag selected a Robbins Open Gripper TBM for the project. The machine, with specifications as reflected in Table 4, was fabricated in a number of countries including the USA (predominantly in Robbin’s factories in Ohio), England (Markham facilities), Canada and Europe.

Table 4. TBM Specifications

Description DetailsHARD ROCK TBM – DESCRIPTION

Machine DiameterNew cutters(worn cutters)

14.4414.41

Main Bearing – three roller (3 axis)Bearing life >13,000 L10 hrs @ 224

kN cutter loadCutters

Number of cutters 85 x 19”/20”(483/508 mm)

LoadingIndividual cutter loadAverage cutter spacing

311 kN89 mm

Cutterhead – Recommended OperatingCutterhead thrustMaximum thrust

Maximum gripper force

18,426 kN27,900 kN71,500 kN

Cutterhead Drive – Variable FrequencyDescription Cutterhead

Drive/Electric Motors/Gear

Reducers/VF DriveCutterhead power 15x315 kW = 4,725 kWCutterhead Torque

at 2.4 rpmat 5.0

18,670 kNm8,960 kNm

Breakout Torque 28,000 kNmThrust Cylinder Stroke 1.82 mTransformer (TBM

Drives)2 x 2500 kVA

Transformer Backup 1 x 1000 kVA

125

Description DetailsConveyor

Capacity 1600 tphBelt width 1400 mm

TBM weight (approx.) 1900 t

The trailing gear was manufactured by ROWA of Switzerland with fabrication carried out in Slovenia and Hungary. Additional backup equipment such as the muck conveyor was procured from H+E Logistics of Germany, the rubber tired transporters from Plan and Teco of Germany and ventilation equipment from Cogemacoustic of France.

5 TUNNEL DESIGN AND CONSTRUCTION

5.1. Alignment ConstraintsVertical alignment of the tunnel, as reflected in Figure 5, is constrained by a number of existing structures and geological features. In profile, at the north end the tunnel passes beneath the buried St. Davids Gorge with bedrock at an approximate elevation of 100 m. This requires an initial tunnel downgrade of 7.82% from the Outlet portal to pass beneath the buried gorge with minimum rock cover of approximately 10 m. At this lowest point the tunnel axis is approximately 135 m below ground surface.

Progressing southward, the tunnel rises at an upgrade of 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 features a 50 m deep turbine pit and tailrace tunnels. From these crossings the tunnel rises at an upgrade of 7.15% to the Intake portal beneath Bay 1 of the INCW structure. These alignment constraints result in the tunnel being located predominantly in Queenston Shale.

Fig. 4. Schematic – Robbins Open Gripper TBM

126

Horizontal alignment is predominantly constrained by a combination of the existing tunnel right-of-way (for existing diversion Tunnels 1 and2), the Niagara River, orientation of the tunnel intake to the existing INCW structure and minimum radii curves to suit TBM tunnelling. In plan, at the north end from Sta 0+000 to Sta 1+426 the alignment parallels the existing OPG power canal on OPG property. At Sta 1+426 the alignment curves southward and converges beneath the existing diversion tunnels following an existing easement along Stanley Avenue beneath the downtown core of the City of Niagara Falls. From Sta 7+950 the alignment curves eastward on an azimuth perpendicular to the existing INCW structure within an easement beneath Niagara Parks Commission property. The final design alignment uses minimum horizontal curve radii of 1000 m to facilitate muck transportation by conveyor belt.

5.2. Tunnel Construction SequenceThe Diversion Tunnel will be excavated by an open gripper TBM of 14.44 m diameter with bored tunnelling progressing from 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 from the Intake. This will enable grouting of the highly permeable rock formations adjacent to the river near the Intake, to be carried out before arrival of the TBM.

Initial lining rock support, to suit encountered rock conditions, will be installed from the TBM and trailing gear. Excavated muck will be transported from the TBM to the surface disposal area by muck conveyor.

Fig. 5. Tunnel Profile

In order to meet the construction schedule the invert portion of the cast-in-place final lining will be constructed concurrent with bored tunneling. This will be carried out some distance back from the TBM trailing gear and will include cleaning of the invert, diversion of water seepage, installation and testing of the waterproofing membrane system, installation of invert formwork and placing of concrete. For this purpose, a moving bridge will be installed allowing transportation to and from the TBM to cross this work area.

Once bored tunnelling is completed, construction of the top section of final lining will be carried out using 4 x 12 m long forms. The sequence of membrane installation and concrete placement essentially follows the same pattern as the invert concrete production. After the final lining is installed, contact and interface grouting will be

done as final activities prior to flooding of the tunnel and flow testing. Further details of the initial and final lining operations are provided below.

5.3. Tunnel LiningThe tunnel lining will be constructed by means of a two pass tunnel lining system that comprises the following components:

An initial lining that includes a combination of steel wire mesh, steel ribs, rock bolts and shotcrete.

A final cast-in-place unreinforced concrete lining with a waterproofing membrane system to ensure that both water seepage from the tunnel and diffusion of chloride ions from the rock to preventtime dependent rock deformation (swelling) does not occur.

Contact grouting carried out along the top of the final concrete lining after completion of concreting

127

to close all voids in the crown and to provide a tight interface between rock mass, initial lining and final lining.

Finally, a high pressure interface grouting system is installed to pre-stress the final concrete lining and surrounding rock such the entire lining system remains in compression over the design life of the tunnel thereby eliminating the need for reinforcing steel.

5.4. Initial LiningThe TBM and trailing gear arrangement allows for installation of rock support at two locations namely Location1 and Location2. L1 is between 4 to 7 m back from the excavation face and L2 between 20 to 40 m back. Due to proximity to sensitive TBM electronic and mechanical components and limitations to TBM advance, installation of rock support in general and shotcrete, in particular, at L1 is limited to the minimum required for personnel safety. In the interests of tunnel advancement, the preference is for installation of the majority of rock support at L2. The assessment of installation of initial lining rock support at locations L1 and L2 takes into consideration the requirements for personnel safety and the rate of tunnel advancement. Depending on encountered rock conditions, this will result in progressive installation of rock support at L1 and L2 with full support being completed 40 m behind the excavation face.

Six rock conditions have been baselined in the GBR for which rock support types have been designed as follows:

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 of shotcrete reinforced with mesh applied from L2 if 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 safety for personnel working at the front of the TBM where blocks of ground are differentiable in otherwise stable rock conditions. It consists of steel ribs (C 100 x 11 steel channels) bolted with a limited number of 2.4 m long rock bolts to the tunnel crown and steel wire mesh fixed with the bolts. At L2, 70 mm of shotcrete is applied and additional rock bolts are installed. If 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 (approximately 11% of Tunnel)Applied in L1 in friable ground, where small blocks of ground tend to fall from the tunnel crown if left unsupported. It consists of steel C150X16 channels installed at 1.2 m intervals, 4.0 m long rock boltsand steel wire mesh. Every second steel channel is extended to springline. At L2, 100 mm of shotcrete and more rock bolts are installed to support the full circumference of the tunnel.

iv Support for Rock Type 4 (approximately 29% of Tunnel)Applied at L1 where the tunnel crown and primary bedding planes are close to intersecting and an increasing number and size of unstable blocks are expected. Consisting of steel C150X16 channels at 0.9 m intervals in the crown and 1.8 m intervals to the sidewalls (installed full 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 of shotcrete and additional rock bolts are installed to complete full support of the tunnel.

v Support for Rock Type 5 (approximately 47% of Tunnel)Applied in L1 in squeezing ground, where slabbing and spalling is experienced soon after excavation. It consists of steel ribs in form of mid weight W150X37 steel beams around the full circumference of the cross section at 1.8 m intervals. As for support, Types 3 and 4 steel wire mesh and 6 m long rock bolts are installed in the tunnel crown to provide safety of the personal working at the front of the TBM. At L2, 160 mm of shotcrete is applied.

vi Support for Rock Type 6 (approximately 10% of Tunnel)Applied in L1 in exceptional ground conditions, where spalling and slabbing is experienced even in front of the excavation face. Heavy W200X59 steel beams, 6-m long rock bolts and 100 mm of shotcrete reinforced with steel wire mesh. An additional 160 mm of shotcrete and one additional layer of mesh reinforcement is installed at L2.

5.5. Final LiningThe final lining consists of cast in place unreinforced concrete with a waterproofing membrane between the initial and final lining and both contact and interface grouting after concreting.

128

(i) Waterproofing Membrane System The waterproofing membrane is designed to act as an impermeable layer between the initial and final lining preventing seepage of fresh water and diffusion of chloride ions to and from the surrounding rock that is susceptible to swelling thereby ensuring that time dependent deformation of the rock does not occur. The waterproofing membrane system consists of regulating shotcrete,where necessary, to smooth corners and edges and prevent damage of the waterproofing membrane during placement, membrane-backed geotextile fleece to protect the waterproofing membrane against the shotcrete, a double layer waterproofing membrane of 2 mm and 1.5 mm thickness respectively consisting of a Poly-Olefine (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 waterproofing membrane material and installation provides a final lining system that is 100% waterproof and diffusion resistant and includes:• laboratory testing to demonstrate chloride diffusion barrier characteristics

• in-situ pressure testing of each double weld seem• 100% in-situ vacuum testing of each panel of the double layer membrane system, where the inner layer is manufactured with dimples to allow vacuum testing between the two membrane layers.

(ii) Final Lining Construction The sequence of final lining installation includes initial placement of invert concrete (and invert waterproofing membrane), concurrent with TBM tunneling, followed by placement of the remainder of the concrete lining system after completion of tunnelling. The invert will be cast in 12 m or 24 m long bays. A bridge across the invert concreting operation will facilitate transportation and material supply to the TBM while placement of the invert concrete system is in progress. Final lining of the tunnel section above the invert will commence once the tunnel is excavated. Before the final lining system is installed above the invert, the preset rings for interface grouting and the waterproofing membrane system will be fixed to the tunnel walls and crown. Twelve metre long, adjustable diameter, steel forms, placed on the previously installed invert concrete, will be used for concreting the top section of the tunnel. Steel forms will be used to provide the smooth concrete finish required to limit friction losses to the flow of water in the tunnel. The

adjustable diameter formwork allows adjustment of the diameter up to 260 mm to accommodate the variable thickness initial lining and provide a 600 to 700 mm thick final lining.

5.6. Contact GroutingAfter placement of final lining concrete, cement grout contact grouting of voids in the crown between the initial and final lining will be carried out via grouting pipes attached at regular intervals to the intrados of the waterproofing membrane and extending through the final lining.

5.7. Interface GroutingInterface grouting of the final lining is required to create a continuous pre-stressed compression concrete support ring able to sustain internal water pressure without requiring steel reinforcement thereby eliminating the risk of corrosion of structural reinforcement within the 90 years design life. Grout, at specific pre-determined pressure (up to 30 bar), will be injected through a system of grout-hose rings installed between the initial lining and the waterproofing membrane system at 3 m centres. The grout-hose rings have pressure valves at 3 m circumferential centres that open underpressures releasing grout into the joint between initial lining and the “geotextile fleece” backing of the waterproofing system. The ends of the grout hoses penetrate through the waterproofing membrane system and the cast-in-place final lining into the tunnel. Grout blocking rings will beinstalled every 12 m to control the longitudinal flow of grout.

Interface grouting pressure for each individual section of tunnel are calculated taking the following considerations into account:• required long term pre-stressing pressure and associated tunnel convergence

• anticipated short term pre-stressing pressure and associated tunnel convergence including deformation allowance for shrinkage of concrete and temperature contraction after watering up

• acceptable differential deformations including the shrinkage of concrete before watering up.Interface grouting will be controlled by precise

in-situ measurement of lining deformation during and after interface grouting. Pumping pressures as defined by structural analysis, will thereby be controlled within allowable limits. Regrouting will be required if the tunnel deformations (convergence of tunnel lining) are less than the values anticipated before watering up (i.e., differential deformation associated with the conservation of the minimum

129

long term pre-stressing pressure including temperature contraction after watering up).

Interface grouting will be carried out in two phases with initial grouting through every second grout hose and re-grouting, if required, through intermediate unused hoses.

6 PROJECT UPDATE

As of June 2006, approximately nine months into the project, Phase 2 design and construction work is on schedule. At the Outlet, excavation of the Outlet canal is complete (except for the rock plug); delivery of the TBM major components is approximately 50% complete and TBM assembly has commenced. Most of the tunneling support equipment has been delivered, including trailing gear, ventilation equipment, gantry crane, conveyor

equipment and transformers. 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 of the new ice acceleration wall, demolition of the existing acceleration wall, drill and blast excavation of the Intake approach channel and preparation for installation of sheet pile cofferdam cells.

7 ACKNOWLEDGMENTS

The authors would like to thank Ontario Power Generation (OPG) for 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 1920’s.

Both contract approaches have been implemented to date on past and current works. If constructed, the

Planning for Canada’s 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 PROJECT 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 1990’s 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 Frenchman’s 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 m2 to 127 m2 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 m2. Smaller tunnel cross section options (90 m2) 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 10 20 30 40 50 60 70 80 90 100

110

120

UCS, LAB, MPa

Freq

neyc

, %

0%

5%

10%

15%

0 20 40 60 80 100

120

140

160

180

200

220

240

UCS Lab, MPa

Fren

quen

cy, %

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 Owner’s 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 INTRODUCTION

Falconbridge’s 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 world’s 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 Rim South Project and Hatch Ltd.

Pedro Gonzalez Area Manager – Nickel Rim South Project and Cementation Canada Ltd.

Hugh MacIsaac Construction Manager - Nickel Rim South Project and Falconbridge Ltd.

ABSTRACT: Falconbridge’s 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 world’s 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 m3. • Galvanized steel sets – 252 Tonnes. • Sinking stage - 5 deck Galloway, weight 47

Tonnes. • Shaft Muckers – 2 x Pneumatic Brutus Grabs. • Shaft Jumbo – 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 m3 • 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 Jumbos – 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 1,896,984Orientation

Number of Orientations- 136Number of Organizations- 306Total Workers in Attendance- 2690

Man Hours to Date 1,896,984Orientation

Number of Orientations- 136Number of Organizations- 306Total 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 Joint 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 PROJECT DELIVERY 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 Canada’s 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 SINKING 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 SINKING EXPERTISE

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 member’s 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 SINKING 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 1,500m3 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. Install 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 82m3 per 6m pour in the Main Shaft and 67m3 per 6m pour in the Vent Shaft.

Fig. 11. Install 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 SINKING 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


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 m3 69.8 m3 67 m3

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 AdvanceSteady State Shaft Sinking

0

200

400

600

800

1000

1200

1400

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan FebMonth

Shaf

t Dep

th (

Met

res

)

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

Variance530 Station

Development

Variance1050 StationDevelopment

Variance1280 StationDevelopment

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


Steady State


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 m3 88.5 m3 81.9 m3

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

Fig. 14. Main Shaft Advance-Steady State Sinking

8 CONCLUSIONS

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 Jacobs 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 Póvoa 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 PROJECT 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 “Fiscalização” or Construction management team (CM) was responsible for the management of Safety, Quality, Schedule and Costs and made up of a consortium of Cinclus, Jacobs (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 June 12th 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 CActual Excavation Line CEstimated Excavation Line SActual Excavation Line SEstimated Excavation Line S1Actual Excavation Line S1

P.K. 630

Heroismo

Campo 24 do Agosto

Bolhao

Inquiry and Modifications to TBM

Trindad e

TBM S-160 Cuttehead Maintenance

Start: CampanhaJune 12th 2000

End: TrindadeOctober 22nd, 2002

Salguieros

TBM S-203 Cutterhead Maintenance

Lima

Faria Guimaraes

End: TrindadeOctober 16th, 2003

Start: Salguieros TrenchJune 3rd 2002

Marques

End: Potal to Don Luis BridgeNovember 3rd, 2003

Start: TrindadeNovember 4th 2002

Refurbish TBM S-160

Aliados

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 June 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 Ø 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 June 2000 there had been three ground collapses. The most serious incident occurred on 12 January 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 m3. 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 contractor’s 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 Jacobs 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 Porto’s new Dragon Stadium had another reason to cheer as they stepped off the metro and into to a new era of transportation for Portugal’s “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

15:1

1:00

15:2

6:00

15:4

1:00

15:5

6:00

16:1

1:00

16:2

6:00

Time (hh:mm:ss)

Bul

khea

d Pr

essu

re (b

ar)

P1 P2 P3 P4 P5 P6 P7Pressure Cells

Pressure compensationby Active Face Support System

Pressure releaseby air diffusion

fluctuating 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 Marquês 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 MARQUÊS STATION

The Marquês 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 m2 (1937 ft2).

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 Marquês.

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 Marquês 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 Marquês 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.

Nível 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 Marquês 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 Marquês, 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 station�s 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 Marquês Station.

From the architectural point of view, Marquês 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. Marquês 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 Marquês 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 Marquês.

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 Marquês 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

Pe

Support pressure P

- not required- no additives for conditioning

- moderate support pressure by compressed air

- no additives for conditioning Pa Pe

PS > PW + PE + POT

- required

- additives for conditioning

Requirements onground properties- enough cohesion

- low ground water table

- enough cohesion

- moderate groundwater table

- variable cohesion- high groundwater table PW PE

PW PE

PW PE

PS

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. Estações suberrâneas em

poços e túneis no Metro do Porto: Aspectos gerais de projecto e acompanhamento técnico da obra- ATO. Conference on Geotechnics, Aveiro, Portugal.

2. Franco, S. G., et. al., 2004. Estação do Marquês em poço no Metro do Porto. Modelação e Segurança. Conference on Geotechnics, Aveiro, Portugal.

3. França, P.T. et. al., 2004. Estação de Salgueiros em poço no Metro do Porto. Modelação e Segurança. Conference on Geotechnics, Aveiro, Portugal.

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 July 2006.

2 STATIONS

There are four underground stations connected by the bored tunnel. The joint venture’s 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

station’s 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. Jointing 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 kN•m of torque at 0.0 to 1.7 rpm • Maximum Speed of 3.2 rpm • Peak Starting Torque of 7,750 kN•m 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 Japanese 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 TBM’s 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 Kovári). 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 Kovári. 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 June 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 District’s 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. Janosko & 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.

MC

T-1

MC

T-2

MC

T-3

0

5

10

15

20

25

30

35

40

UC

Stre

ngth

(MP

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 o’clock 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-m dia. 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 finger–shield, 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 PROJECT 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., Japan, 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 m3/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.

Dep

th b

elow

gro

und

surf

ace,

m

0

10

20

30

40

Fill

Very soft to Soft claydark gray, trace fine sand and shell

Medium clay, dark gray, trace fine sand and shell

Stiff to Very stiff silty claylight grayish brown, trace fine sand

Very stiff to Hard silty claygrayish brown, trace fine sand

Dense to Very dense silty sandlight grayish brown

5.31-mBMA 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

wo,

%

wl,

%

Ip,

%

γ,

t/m3

N,

blows/ft

suo(UC)

t/m2

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: wo = natural water content, wl = liquid limit, Ip = plasticity index, γ = total unit weight, suo (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

D5.31 m

3.4 mfrom tunnel centerline

29.36 m

-10 0 10

Dep

th b

elow

gro

und

surf

ace,

m

-40

-30

-20

-10

0

18.5 mbelow groundsurface

D5.31 m

4.4 mfrom tunnel centerline

5.9 m 9.6 m10.6 m

21.8 m

23.8 m25.8 m

29.11 m

Fill

Very soft to Softclay

Medium clay

Stiff to Very stiffsilty clay

Very stiff to Hardsilty clay

Dense to Very densesilty sand

Test Section 1(Sta. 0+853)

Test Section 2(Sta. 1+446)

Distance fromtunnel centerline, m

-5

0

5Direction oftunneling

Plan View

Hor

izon

tal

dist

ance

, m

-5

0

5Direction oftunneling

Plan View

P1

P2-1

P2-2P2-3

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

Pore

wat

er p

ress

ure,

kPa

0

50

100

150

200

250

300

350

400

450

P2-1P2-2

P2-5

P2-4

P2-6

P2-7

P2-3

P2-1P2-2

P2-5

P2-4

P2-6P2-7

P2-3

7.74 m

5.31 mTAIL

FACE

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 from tunnel centerline, m

3 6 9 12

Exce

ss p

ore

wat

er p

ress

ure,

kPa

0

50

100

150

200

250

300

1 - at shield face2 - after grouting (10 m behind shield face)3 - 50 m behind 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 60’s and the 80’s.

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 1990’s.

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 16th 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 60’s and 70’s. 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 16th 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 16th 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 16th 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 16th Avenue trunk sewers.

186

5 CASE HISTORIES

5.1. Ninth Line/16th 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 16th 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 16th 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 16th 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. 16th 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 16th 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 16th 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/16th Avenue Phase 1

16th Avenue Phase 2

Bathurst/Langstaff

Credit Valley

187

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

188

Fig.9. Soil and groundwater conditions along 16th 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 July 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/16th 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 June 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 16th Avenue Phase 1 to lessen dewatering.

• Westerly 3.5 km of sewer mined between May 2005 and June 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 elements are worthy of note: • 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 16th 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 January 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 elements are worthy of note: • 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 January 2006.

• Sewers to be mined between July 2006 and June 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 16th 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 16th 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 16th 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 REGION’S 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 19th 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 19th 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 Minister’s 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 Region’s 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 19th Avenue/Leslie Street Project three general construction methods will be employed for this project:

195

Figure 1196

• 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 Region’s area is subject to the Green Belt plan, which includes the Oak Ridges Moraine. Physical growth is limited by both provincial legislation and the Region’s 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 Minister’s 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 activities…address potential impacts to well users and the natural environment…assess 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 alignments…consideration 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

Minister’s Conditions 5, 8, 10

Minister’s Conditions 1, 2, 3, 6, 7, 9 Minister’s Conditions 4, 11

Figure 2. The Minister’s 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 EAClass EA

Minister ’s ConditionsMinister’s Conditions

New Comparison of Alternative Route Alignments


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 ReportBaseline Report

Sewer Design and Construction Techniques

Report


Report

Class EAClass EA

Minister ’s ConditionsMinister’s Conditions



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 ReportBaseline Report


Report


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, 19th 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 16th Avenue Trunk Sewer project in York Region. The 16th 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 16th 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 Region’s 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 Region’s 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 & ApprovalsCertificate of Approval

Agencies

ResponsibilityMunicipal & Private Sewage Works CofAPTTWWork PermitLake and Rivers ImprovementsFill Construction & Alteration Waterways PermitLetter of Advise

IndependentPeer

Review

Public

Minister’s11

Conditions

FinalPermits & Approvals

DraftPermits & Approvals

Ministry of the Environment MOEToronto and Region Conservation Authority TRCA Department of Fisheries and Oceans DFOMinistry of Natural Resources MNR

Process

PlanningCapital Budget Project Need EAPublic EngagementMeeting Minister’s ConditionsEnhanced StudyMOE Approval of Route and Construction TechniquesDesignExtensive Public Consultation, Permits & ApprovalsPeer ReviewTenderAwardConstructEnvironmental Monitoring & ReportingCommissionPost Commission Evaluation/Closeout

YorkRegion

2 CofA’sPTTW TRCA Permit MNR Work PermitDFO Letter of Advice TCPL Crossing PermitCN Crossing Permit

Feedback LoopDirect

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 impacts–through 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 contractor’s 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 everyone’s 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 facilitator’s 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: Ontario’s 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 19th 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 19th 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 Japanese. However, in Japan 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. System Thrust 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. System Torque 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 PROJECTS

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 (¼ 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 January 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 UK’s natural gas. The general contractors, a NACAP/Land and Marine JV, 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 July 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 today’s 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 Jacking & 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